Accurate nucleic acid concentrations by nuclear magnetic resonance.

Determination of the concentration of biochemical samples often yields values with uncertainties of 10-20% or more. This paper details a protocol for use with 500- to 600-MHz NMR spectrometers to measure approximately 1mM concentrations within +/-1-3% accuracy. With suitable precautions, all compounds have equal NMR "absorption coefficients" for protons. About 2mg of sample are needed for proteins and nucleic acids with MW=5000, although less accurate determinations could be made with smaller amounts. The technique utilizes standardized internal reference reagent compounds, cacodylic acid or 3-(trimethylsilyl)propionic-2,2,3,3-d(4) acid sodium salt. Spectra were signal-averaged using long interpulse delays so that integrals of nonexchangeable protons could be quantified relative to the reference standard. Accurate determinations require careful optimization of the homogeneity of the magnetic field and meticulous attention to sample preparation and spectral processing. The main source of error is usually the accuracy of micropipets. If sample preparation errors could be eliminated, the lower limit of accuracy with the current generation of NMR spectrometers is probably near 0.4%. However, this would require >99.5% sample purity. Methods are described to establish the concentration of the standards, and applications are illustrated with DNA mono- and oligonucleotides. Similar procedures should apply to proteins, polysaccharides, and other biomolecules, with about the same accuracy and precision.

Structure of SL4 RNA from the HIV-1 packaging signal.

The NMR-based structure is described for an RNA model of stem-loop 4 (SL4) from the HIV-1 major packaging domain. The GAGA tetraloop adopts a conformation similar to the classic GNRA form, although there are differences in the details. The type II tandem G.U pairs have a combination of wobble and bifurcated hydrogen bonds where the uracil 2-carbonyl oxygen is hydrogen-bonded to both G,H1 and G,H2. There is the likelihood of a Na(+) ion coordinated to the four carbonyl oxygens in the major groove for these G.U pairs and perhaps to the N7 lone pairs of the G bases as well. A continuous stack of five bases extends over nearly the whole length of the stem to the base of the loop in the RNA 16mer: C15/U14/G13/G5/C6. There is no evidence for a terminal G.A pair; instead, G1 appears quite unrestrained, and A16 stacks on both C15 and G2. Residues G2 through G5 exhibit broadened resonances, especially G3 and U4, suggesting enhanced mobility for the 5'-side of the stem. The structure shows G2/G3/U4 stacking along the same strand, nearly isolated from interaction with the other bases. This is probably an important factor in the signal broadening and apparent mobility of these residues and the low stability of the 16mer hairpin against thermal denaturation.

Changes in 13C NMR chemical shifts of DNA as a tool for monitoring drug interactions.

The antibiotic drug, netropsin, was complexed with the DNA oligonucleotide duplex [d(GGTATACC)]2 to explore the effects of ligand binding on the 13C NMR chemical shifts of the DNA base and sugar carbons. The binding mode of netrospin to TA-rich tracts of DNA has been well documented and served as an attractive model system. For the base carbons, four large changes in resonance chemical shifts were observed upon complex formation: -0.64 ppm for carbon 4 of either Ado4 or Ado6, 1.36 ppm for carbon 2 of Thd5, 1.33 ppm for carbon 5 of Thd5 and 0.94 for carbon 6 of Thd5. AdoC4 is covalently bonded to a heteroatom that is hydrogen bonded to netropsin; this relatively large deshielding is consistent with the known hydrogen bond formed at AdoN3. The three large shielding increases are consistent with hydrogen bonds to water in the minor groove being disrupted upon netropsin binding. For the DNA sugar resonances, large changes in chemical shifts were observed upon netropsin complexation. The 2', 3' and 5' 13C resonances of Thd3 and Thd5 were shielded whereas those of Ado4 and Ado6 were deshielded; the 13C resonances of 1' and 4' could not be assigned. These changes are consistent with alteration of the dynamic pseudorotational states occupied by the DNA sugars. A significant alteration in the pseudorotational states of Ado4 or Ado6 must occur as suggested by the large change in chemical shift of -1.65 ppm of the C3' carbon. In conclusion, 13C NMR may serve as a practical tool for analyzing structural changes in DNA-ligand complexes.

Regioselective synthesis of 1,3,5-13C3 and 2,4-13C2-labeled 2-deoxyribonolactones.

The synthesis of 1,3,5-13C3- and 2,4-13C2-labeled 5-O-bromobenzyl-2-deoxyribonolactones 2, precursors to 13C-enriched nucleoside phosphoramidites for solid-phase synthesis of DNA oligonucletides, is described. An equimolar combination of these two multiply labeled lactones affords a "population-labeled" mixture of isotopomers which exhibits an approximately 50-fold increase in the sensitivity of 13C-NMR compared to natural abundance measurements. The 13C-13C 2-bond and 4-bond coupling constants are reported for the lactones; all are <2Hz, confirming that this labeling scheme should be especially useful for NMR-relaxation measurements.

Three-dimensional folding of an RNA hairpin required for packaging HIV-1.

An NMR-based structure is presented for a 20 mer hairpin model of the SL3 stem-loop from the HIV-1 packaging signal. The stem has an A-family structure. However, the GGAG tetraloop appears to be flexible with the second (G10) and fourth (G12) bases extruded from the normal stacking arrangement. The A-base (A11) occupies a cavity large enough for it to jump rapidly between stacking upon G9 (in the loop) and G13 (from the base-pair adjacent to the loop). The H-bonding loci of G10, A11, and G12 are unoccupied in the free RNA structure. The loop should be easily adaptable to binding by the HIV-1 nucleocapsid protein or loop receptors.

Full 1H NMR assignment of a 24-nucleotide RNA hairpin: application of the 1H 3D-NOE/2QC experiment.

The subject RNA models the binding site for the coat protein of the R17 virus, as well as the ribosome recognition sequence for the R17 replicase gene. With an RNA of this size, overlaps among the sugar protons complicate assignments of the 1H NMR spectrum. The cross peaks that overlap significantly in 2D-NOE spectra can frequently be resolved by introducing a third, in our approach the double-quantum, frequency axis. In particular the planes in a 3D-NOE/2QC spectrum perpendicular to the 2Q axis are extremely useful, showing a highly informative repeating NOE-2Q pattern. In this experiment substantial J-coupling confers special advantages. This always occurs for geminal pairs (H5'/H5" for RNA plus H2'/H2" for DNA), as well as for H5/H6, for H3'/H4' in sugars with substantial populations of the N-pucker, for H1'/H2' for S-puckered sugars, and usually for H2'/H3'. For the 24-mer RNA hairpin the additional information from the 3D-NOE/2QC spectrum allowed assignment of all of the non-exchangeable protons, eliminating the need for stable-isotope labeling.

Structure of the HIV-1 nucleocapsid protein bound to the SL3 psi-RNA recognition element.

The three-dimensional structure of the human immunodeficiency virus-type 1 (HIV-1) nucleocapsid protein (NC) bound to the SL3 stem-loop recognition element of the genomic Psi RNA packaging signal has been determined by heteronuclear magnetic resonance spectroscopy. Tight binding (dissociation constant, approximately 100 nM) is mediated by specific interactions between the amino- and carboxyl-terminal CCHC-type zinc knuckles of the NC protein and the G7 and G9 nucleotide bases, respectively, of the G6-G7-A8-G9 RNA tetraloop. A8 packs against the amino-terminal knuckle and forms a hydrogen bond with conserved Arg32, and residues Lys3 to Arg10 of NC form a 310 helix that binds to the major groove of the RNA stem and also packs against the amino-terminal zinc knuckle. The structure provides insights into the mechanism of viral genome recognition, explains extensive amino acid conservation within NC, and serves as a basis for the development of inhibitors designed to interfere with genome encapsidation.

Improved proton assignment for DNA by application of aliasing and dispersive-absorptive phasing to two-quantum COSY spectra.

A unique combination of aliasing and dispersive-absorptive (DA) phasing of two-quantum correlated spectroscopy (2 Q-COSY) NMR data is shown to enhance proton chemical-shift assignments in DNA oligonucleotides by (i) reducing the time necessary for acquiring NMR data or, alternatively, improving the spectral resolution in a given time, (ii) reducing the number of spectra necessary for NMR data processing and analysis, and (iii) increasing the complexity of oligonucleotide sequences and structures which are accessible to 2D NMR analysis. Aliasing allows a reduction in the size of the acquired data without significant risk of losing information. Phasing the 2Q-COSY dispersive in the F2 dimension reduces the primary antiphase doublet into a pseudo-singlet and increases the apparent signal-to-noise. A single 2Q-COSY spectrum can provide an amount of chemical-shift information comparable to that from a series of COSY, relayed-COSY, and/or spin-lock COSY spectra optimized for various coupling constants. The low signal-to-noise inherent in the most popular two-quantum-filtered correlated spectroscopy (2QF-COSY) of samples with naturally broad lines is largely avoided due to less cancellation. There is no diagonal in a 2Q-COSY which can obscure correlations between protons which are nearly isochronous. As an example of this efficient application, the assignment of 139 of the 143 proton resonances from a single 2Q-COSY and a 2D-NOE spectrum of the DNA hexadecamer [d(AAATATAGCTATATTT)]2 is demonstrated.

Monte Carlo estimation of errors in 13C-NMR relaxation studies of a DNA oligomer duplex.

An analysis of errors has been done with the Monte Carlo method for natural abundance 13C-NMR relaxation studies of a DNA duplex. Repeated measurements of the longitudinal relaxation time, T1, and the heteronuclear NOE were made at 90.6 MHz on the duplexed DNA pentanucleotide, [d(TCGCG)]2. The deviations averaged over all carbons were 13% for T1 and 9% for NOE. These relative deviations were applied to generate 100 values of T1 and NOE with normal distributions about the measured mean values for each carbon. A new version of MOLDYN, called McMOLDYN, has been written, which was used to generate 100 values of T1 and NOE with normal distributions corresponding to the measured errors; the same error distributions were also applied to measurements at 125.8 MHz. The order parameter, S2, and the effective internal correlation time, tau e, in the Model-Free Approach have been optimized from the distributions simulated by McMOLDYN. McMOLDYN also permits the automated entry of multiple sets of initial guesses for the output parameters S2, tau e, and tau m. In addition, McMOLDYN adds cross-relaxation terms from chemical shift anisotropy, increasingly important as spectrometer magnetic fields get higher. Between the two parameters optimized, S2 has the smallest relative error, estimated at 15% on average, which means that S2 is a well-defined parameter. However, tau e is very poorly defined with the average relative error estimated 85%; it is typically found in the range of 30-300 ps.

Proton NMR and structural features of a 24-nucleotide RNA hairpin.

The three-dimensional conformation of a 24-nucleotide variant of the RNA binding sequence for the coat protein of bacteriophage R17 has been analyzed using NMR, molecular dynamics, and energy minimization. The imino proton spectrum is consistent with base pairing requirements for coat protein binding known from biochemical studies. All 185 of the nonexchangeable protons were assigned using a variety of homonuclear 2D and 3D NMR methods. Measurements of nuclear Overhauser enhancements and two-quantum correlations were made at 500 MHz. New procedures were developed to characterize as many resonances as possible, including deconvolution and path analysis methods. An average of 21 distance constraints per residue were used in molecular dynamics calculations to obtain preliminary folded structures for residues 3-21. The unpaired A8 residue is stacked in the stem, and the entire region from G7 to C15 in the upper stem and loop appears to be flexible. Several of these residues have a large fraction of S-puckered ribose rings, rather than the N-forms characteristic of RNA duplexes. There is considerable variation in the low-energy loop conformations that satisfy the distance constraints at this preliminary level of refinement. The Shine-Dalgarno ribosome binding site is exposed, and only two apparently weak base pairs would have to break for the 16S ribosomal RNA to bind and the ribosome to initiate translation of the replicase gene. Although the loop form must be regarded as tentative, the known interaction sites with the coat protein are easily accessible from the major groove side of the loop.

Computer-assisted assignment of peptides with non-standard amino acids.

A comprehensive peptide assignment program and its application to a cyclic peptide, cyclosporin A, are presented in this paper. A group of graph theoretical algorithms using fuzzy logic are discussed with the aid of examples from cyclosporin A. The algorithms deal with heavily overlapped peaks, recover disjointed and distorted spin coupling networks, and include strategies for sequence-specific assignment. A procedure to extend the Protein Knowledge Base for automatically assigning non-standard amino acid residues is also presented. The program is capable of completely automated assignment for small peptides (approximately 20 residues). For such molecules, it is insensitive to whether the peptide chain is cyclic or acyclic, and to whether amide protons are present or absent. For larger peptides/proteins, more user interaction is required and the sequence-specific assignment step usually must proceed through fragments smaller than the full length to avoid problems due to occurrence of a combinatorial explosion. The program can be applied as a rigorous tool to check manual assignments. The fuzzy graph theoretical concepts built in the program are illustrated with 2D proton spectra of a peptide, but may be extended to higher-dimensional spectra, other biopolymers, natural products and other organic structures.

13C-NMR of the deoxyribose sugars in four DNA oligonucleotide duplexes: assignment and structural features.

Natural-abundance 13C-NMR spectra have been obtained for four self-complementary DNA oligonucleotides: [d(TAGCGCTA)]2, [d(GGTATACC)]2, [d(CG)3]2, and [d(TCGCG)]2; this paper focuses on the deoxyribose resonances. Assignments were made by a combination of the two-dimensional proton-detected heteronuclear correlation experiment and comparison of 1D spectra, accounting for 31P coupling, base composition, and similarities in chemical shift versus temperature profiles (delta vs T). Large shielding and deshielding of the sugar resonances (between 2.0 and -1.9 ppm) are observed upon thermal dissociation of the duplex. The shapes of the delta vs T profiles correlate strongly with the purine/pyrimidine nature of the base attached at C1' in these duplexes that have a substantial fraction of residues within alternating purine-pyrimidine sequences. The correlation is primarily associated with changes in the equilibrium distribution of furanose pseudorotational states that may arise in part from the relief of interstrand purine-purine steric clashes.

13C-NMR relaxation in three DNA oligonucleotide duplexes: model-free analysis of internal and overall motion.

Natural-abundance 13C-NMR spectra of [d(TCGCG)] (1), [d(CGCGCG)]2 (2), and [d(GGTATACC)]2 (3) were measured at 90.6 MHz to obtain 13C-1H NOEs and T1 relaxation times; relaxation data were also measured at 125.7 MHz for 1 and 2 and at 62.9 MHz for 1. Analysis of the relaxation data was performed in the context of the "model-free" approach of Lipari and Szabo [Lipari, G., & Szabo, A. (1982) J. Am. Chem. Soc. 104, 4546-4559], leading to the following conclusions: (i) Optimized values for the overall correlation times of 0.9 ns for 1 and 1.4 ns for 2 are close to those predicted by light-scattering results on similar molecules [Eimer et al. (1990) Biochemistry 29, 799-811]. (ii) For the nonterminal residues, the "order parameter", S2, is around 0.8 for the protonated base carbons and 0.6 for the sugar carbons, indicating less spatial restriction on the sugar carbons (in the model-free approach, the order parameter is 1 for a rigid body and 0 for a system with completely unrestricted internal motion). (iii) The order parameters for the terminal residues vary over a wide range with the smallest values around 0.2-0.3 for the HO-13C5' and the 13C3'-OH; rational trends are seen in the variation of S2 with chain position in the terminal residues. (iv) The analysis shows that the order parameters are accurate within 15%. (v) The "effective internal correlation time", tau e, is very short for the sugar carbons (30-300 ps) and less well-defined, but probably also short, for the bases. (vi) The analysis indicates that most of the relaxation in DNA is accounted for by S2 and the tau e is so short that a good approximation to any relaxation property, P (e.g., T1, T2, 13C-1H NOE, 1H-1H cross-relaxation rate), is P = S2Prigid, where Prigid is the value for the property in a system without internal motion (the analysis assumes the same isotropic overall motion for both the rigid and flexible bodies).

Cellobiose dehydrogenases of Sporotrichum (Chrysosporium) thermophile.

Both cellobiose dehydrogenases of Sporotrichum (Chrysosporium) thermophile, ATCC 42464, obtained after fractionation with DEAE-Trisacryl chromatography and named cellobiose dehydrogenase I and II have been purified to homogeneity by different chromatographic techniques. Both enzymes are slightly glycosylated flavocytochrome-b proteins with similar catalytic properties but with distinct molecular masses (91 kDa and 192 kDa for enzymes I and II, respectively) and isoelectric point (4.1 versus 3.45). Examination by SDS/PAGE clearly showed that the larger enzyme II is a homodimer, whose subunit is close to, but different from dehydrogenase I which is homogeneous by this technique. After limited digestion of both enzymes with papain, two main fractions with residual activity are formed, one carrying the heme, the other being the flavin component; each fraction is characterized by its particular chromatographic behaviour. The flavin carrying component shows an atypical (for flavoprotein) three-banded spectrum indicative of the presence of a flavin derivative. Both enzymes react very slowly with oxygen clearly forming some superoxide radicals and possibly hydrogen peroxide. Cellobiose and other cellodextrins are oxidized at their reducing glycosyl moiety to the corresponding aldonic acid. With the use of the autooxidable phenazinemethosulphate, cellulose (either in a hydrated form or crystalline) is also oxidized at free reducing ends so that appreciable amounts of cellobionic acid are released upon enzymatic hydrolysis.

Actinomycin D induced DNase I hypersensitivity and asymmetric structure transmission in a DNA hexadecamer.

DNase I cleavage rates and nmr chemical shifts are shown to change for DNA sequences distal to an intercalated actinomycin D molecule in a duplex hexadecamer upon drug binding. Both sets of observations suggest that the source of these changes is a DNA-mediated structural response. The nmr results imply the response is transmitted preferentially in a 5'-to-3' direction from the drug binding site. An inequivalent response of the two strands to a ligand-induced conformational change immediately suggests a mechanism for distinguishing the sense and antisense strands of DNA.

Enzymatic and NMR analysis of oligoribonucleotides synthesized with 2'-tert-butyldimethylsilyl protected cyanoethylphosphoramidite monomers.

The regioisomeric integrity of the internucleotide phosphate linkage in synthetic RNA using 2'-tert-butyldimethylsilyl protection was examined using enzymatic and NMR techniques. Two sets of DNA-RNA hybrid nonamers, T3XT5 and T5XT3 (where X = rA, rC, rG and U) and the tetramer AGCU were analyzed. Enzyme catalyzed hydrolysis of the nonamers with ribonuclease T2 showed that the linkage at the ribonucleotide was the desired 3'-5'. A control nonamer with a 2'-5' linkage was subjected to the enzyme, and showed no cleavage. High-resolution proton NMR of the tetramer also gave a favorable comparison with the same molecule obtained by non-chemical means.

Toward a computer assisted analysis of NOESY spectra: a multivariate data analysis of an RNA NOESY spectrum.

A multivariate data-representation of a portion of the H-NOESY spectrum of an RNA octamer duplex was used to explore the possibility of using Principal Component Analysis and Partial Least Squares Discrimination for pattern recognition. In this case, it is found that the methods can: (i) distinguish slices containing signal from those containing only noise, (ii) locate slices containing overlapping signals, and (iii) in some cases to segregate slices with unique aspects such as those from terminal nucleotides, overlapping signals, purine-H8, pyrimidine-H6 and adenine-H2 containing slices. These properties can easily be included in a scheme to automate spectral analysis. The formulation described here does not distinguish patterns needed to automate sequential assignment of resonances in NOESY spectra of RNA.

Actinomycin D induced DNase I cleavage enhancement caused by sequence specific propagation of an altered DNA structure.

Two DNA hexadecamers containing one central 5'-GC-3' base step have been examined by footprinting methodology in the presence and absence of actinomycin D. The results of these studies, coupled with imino proton NMR measurements indicate that the antitumor drug causes a change in DNA conformation at a distance from the actinomycin intercalation site in a molecule of sequence d[ATATATAGCTATATAT] that does not occur in d[AAAAAAAGCTTTTTTT]. The experiments demonstrate that DNase I rate enhancements associated with actinomycin D binding are caused by ligand alteration of equilibrium DNA structure.

13C NMR assignments of the bases in oligodeoxynucleotides: an automated procedure using Bayesian statistics.

A statistical method, Bayes Maximum Likelihood, has been applied to the classification of base 13C NMR resonances in DNA oligomers. An accuracy of 100% for carbon class discrimination was achieved for a preliminary training set of four oligomers using the following four parameters: (1) the chemical shift; (2) the temperature at which the spectrum was obtained; (3) the difference in chemical shift from the C5 resonances; and (4) a sequence factor representing the neighboring nucleotides. Classification of a fifth oligomer, previously assigned and not contained in the original training set, gave reasonable carbon class assignments.