Transcriptional regulation by antitermination. Interaction of RNA with NusB protein and NusB/NusE protein complex of Escherichia coli

A recombinant heterodimeric NusB/NusE protein complex of Escherichia coli was expressed under the control of a synthetic mini operon. Surface plasmon resonance measurements showed that the heterodimer complex has substantially higher affinity for the boxA RNA sequence motif of the ribosomal RNA (rrn) operons of E.coli as compared to monomeric NusB protein. Single base exchanges in boxA RNA reduced the affinity of the protein complex up to 15-fold. The impact of base exchanges in the boxA RNA on the interaction with NusB protein was studied by (1)H,(15)N heterocorrelation NMR spectroscopy. Spectra obtained with modified RNA sequences were analysed by a novel generic algorithm. Replacement of bases in the terminal segments of the boxA RNA motif caused minor chemical shift changes as compared to base exchanges in the central part of the dodecameric boxA motif.     

Modifying the specificity of an RNA backbone contact.

The interaction between the MS2 bacteriophage coat protein homodimer and its cognate RNA hairpin is facilitated by 21 different RNA-protein contacts. In one of these contacts, the 2'-hydroxyl group at ribose -5 of the RNA acts as a hydrogen bond donor to Glu63 in one subunit of the protein. Previous experiments showed that substitution of ribose -5 with deoxyribose resulted in a 24-fold decrease in binding affinity between RNA and protein. Using a protein where the two MS2 monomers were fused to increase stability, the contribution of this contact to the overall binding affinity was investigated by site-directed mutagenesis. When Glu63 was substituted with glutamine, aspartate, or alanine, the binding affinity of the hairpin for the protein was weakened by 12 to 100-fold, similar to that observed with deoxyribose at position -5. However, the specificity of the three mutant proteins for RNAs with various modifications at the 2'-position of ribose -5 differed dramatically. While the Glu63Asp protein resembled the wild-type protein in preferring the 2'-hydroxyl group over a proton or a bulky 2'-substituent, both the Glu63Ala and Glu63Gln proteins preferred bulky 2'-substituents over the 2'-hydroxyl group by more than 100-fold. These experiments emphasize the ease with which the specificity of a protein-nucleic acid interaction can be changed at thermodynamically important sites.     

Preference for ribose over deoxyribose in loop-closing base pairs of extra stable nucleic acid hairpins

We have investigated the effect of switching ribose to deoxyribose at the closing base-pair of an extra-stable RNA hairpin. Specifically, we studied the sequence 5'-GGAC(UUCG)GUCC, a dodecanucleotide that folds into a well-characterized, "extra stable" RNA hairpin structure. Recently, we showed that hairpins containing a 2',5'-linked (UUCG) loop instead of the native 3',5'-linked loop also exhibit extra-stability (Hannoush and Damha, J. Am. Chem. Soc., 2001, 123, 12368-12374). In this article, we show that the ribose units located at the loop-closing positions (i.e., rC4 and rG9) contribute significantly to the stabilization of RNA hairpins, particularly those containing the 3',5'-UUCG loop. Interestingly, the requirement of rC4 and rG9 is more relaxed for DNA hairpins containing the 2',5'-UUCC loop and, in fact, they may be replaced altogether (ribose--> deoxyribose) without affecting stability. The results broaden our understanding of the behavior of highly stable (UUCG) hairpin loops and how they respond to structural perturbation of the loop-closing base pairs.     

RNA-binding specificity of E. coli NusA

The RNA sequences boxA, boxB and boxC constitute the nut regions of phage λ. They nucleate the formation of a termination-resistant RNA polymerase complex on the λ chromosome. The complex includes E. coli proteins NusA, NusB, NusG and NusE, and the λ N protein. A complex that includes the Nus proteins and other factors forms at the rrn leader. Whereas RNA-binding by NusB and NusE has been described in quantitative terms, the interaction of NusA with these RNA sequences is less defined. Isotropic as well as anisotropic fluorescence equilibrium titrations show that NusA binds only the nut spacer sequence between boxA and boxB. Thus, nutR boxA5-spacer, nutR boxA16-spacer and nutR boxA69-spacer retain NusA binding, whereas a spacer mutation eliminates complex formation. The affinity of NusA for nutL is 50% higher than for nutR. In contrast, rrn boxA, which includes an additional U residue, binds NusA in the absence of spacer. The K_d values obtained for rrn boxA and rrn boxA-spacer are 19-fold and 8-fold lower, respectively, than those for nutR boxA-spacer. These differences may explain why λ requires an additional protein, λ N, to suppress termination. Knowledge of the different affinities now describes the assembly of the anti-termination complex in quantitative terms.     

Preference for Ribose Over Deoxyribose in Loop-Closing Base Pairs of Extra Stable Nucleic Acid Hairpins

We have investigated the effect of switching ribose to deoxyribose at the closing base-pair of an extra-stable RNA hairpin. Specifically, we studied the sequence 5′-GGAC(UUCG)GUCC, a dodecanucleotide that folds into a well-characterized, “extra stable” RNA hairpin structure. Recently, we showed that hairpins containing a 2′,5′-linked (UUCG) loop instead of the native 3′,5′-linked loop also exhibit extra-stability (Hannoush and Damha, J. Am. Chem. Soc., 2001, 123, 12368–12374). In this article, we show that the ribose units located at the loop-closing positions (i.e., rC ₄ and rG ₉ ) contribute significantly to the stabilization of RNA hairpins, particularly those containing the 3′,5′-UUCG loop. Interestingly, the requirement of rC4 and rG9 is more relaxed for DNA hairpins containing the 2′,5′-UUCG loop and, in fact, they may be replaced altogether (ribose → deoxyribose) without affecting stability. The results broaden our understanding of the behavior of highly stable (UUCG) hairpin loops and how they respond to structural perturbation of the loop-closing base pairs.     

Multiple-quantum HCN-CCH-TOCSY experiment for 13C/15N labeled RNA oligonucleotides

A multiple-quantum 3D HCN-CCH-TOCSY experiment is presented for the assignment of RNA ribose resonances. The experiment makes use of the chemical shift dispersion of N1 of pyrimidine and N9 of purine to distinguish the ribose spin systems. It provides an alternative approach for the assignment of ribose resonances to the currently used COSY- and TOCSY-type experiments in which either ¹³C or ¹H is utilized to distinguish the different spin systems. Compared to the single-quantum version, the sensitivity of the multiple-quantum HCN-CCH-TOCSY experiment is enhanced on average by a factor of 2 for a 23-mer RNA aptamer complexed with neomycin.     

Structural characterization of the 5' untranslated RNA of hepatitis C virus by vibrational spectroscopy

Raman and FTIR spectroscopy have been used to characterize the structure of 5'untranslated region (5'UTR, 342-mer RNA) of the HCV genome. The study of the 750-850 cm(-1) Raman spectral domain of the ribose-phosphate backbone reveals that the percentage of nucleobases involved in double helix-loop junctions is 19+/-1%, which is very close to that of a theoretical secondary structure model (18.7%) proposed on the basis of comparative sequence analysis and thermodynamic modelling. In addition, about 68+/-2% of the bases are helically ordered having C(3')-endo ribofuranose pucker. FTIR-monitored H/D exchange provides the following results: (a) base-paired guanine and cytosine nucleobases show the lowest rate of isotopic exchange, and some synchronous intensity changes of marker bands of A.U pair and single stranded adenine are consistent with the presence of A(*)A.U triplets; (b) the vibrational coupling between the ribose ether C-O stretching and 2'OH bending motions reveals that helical regions of 5'UTR RNA are characterized by hydrogen bonding between the 2'OH ribose groups and the ether oxygen atoms of neighbouring ribose residues.     

Ab-inito quantum mechanical calculations of NMR chemical shifts in nucleic acids constituents. II. Conformational dependence of the 1H and 13C chemical shifts in the ribose

The magnetic shielding constant of the different 13C and 1H nuclei of a deoxyribose are calculated for the C2' endo and C3' endo puckerings of the furanose ring as a function of the conformation about the C4'C5' bond. For the carbons the calculated variations are of several ppm, the C3' endo puckering corresponding in most cases to a larger shielding than the C2' endo one. For the protons the calculated variations of chemical shifts are all smaller than 1.3 ppm, that is of the order of magnitude of the variation of the geometrical shielding produced on these protons by the other units of a DNA double helix, with a change of the overall structure of the helix. The computations carried out on the deoxyribose-3' and 5' phosphates for several conformations of the phosphate group tend to show that the changes of conformation of the charged group of atoms produce chemical shift variations smaller than the two conformational parameters of the deoxyribose itself. The calculations carried out for a ribose do give the general features of the differences between the carbon and proton spectra of deoxynucleosides and nucleosides. The comparison of the measured and calculated phosphorylation shifts tend to show that the counterion contributes significantly, for some nuclei of the deoxyribose, to the shifts measured. The calculated magnitude of this polarization effect on carbon shifts suggests a tentative qualitative interpretation of carbon spectra of the ribose part of DNA double helices.     

Statistical analysis of atomic contacts at RNA-protein interfaces

Forty-five crystals of complexes between proteins and RNA molecules from the Protein Data Bank have been statistically surveyed for the number of contacts between RNA components (phosphate, ribose and the four bases) and amino acid side chains. Three groups of complexes were defined: the tRNA synthetases; the ribosomal complexes; and a third group containing a variety of complexes. The types of atomic contacts were a priori classified into ionic, neutral H-bond, C-H...O H-bond, or van der Waals interaction. All the contacts were organized into a relational database which allows for statistical analysis. The main conclusions are the following: (i) in all three groups of complexes, the most preferred amino acids (Arg, Asn, Ser, Lys) and the less preferred ones (Ala, Ile, Leu, Val) are the same; Trp and Cys are rarely observed (respectively 15 and 5 amino acids in the ensemble of interfaces); (ii) of the total number of amino acids located at the interfaces 22% are hydrophobic, 40% charged (positive 32%, negative 8%), 30% polar and 8% are Gly; (iii) in ribosomal complexes, phosphate is preferred over ribose, which is preferred over the bases, but there is no significant preference in the other two groups; (iv) there is no significant prevalence of a base type at protein-RNA interfaces, but specifically Arg and Lys display a preference for phosphate over ribose and bases; Pro and Asn prefer bases over ribose and phosphate; Met, Phe and Tyr prefer ribose over phosphate and bases. Further, Ile, Pro, Ser prefer A over the others; Leu prefers C; Asp and Gly prefer G; and Asn prefers U. Considering the contact types, the following conclusions could be drawn: (i) 23% of the contacts are via potential H-bonds (including CH...O H-bonds and ionic interactions), 72% belong to van der Waals interactions and 5% are considered as short contacts; (ii) of all potential H-bonds, 54% are standard, 33% are of the C-H...O type and 13% are ionic; (iii) the Watson-Crick sites of G, O6(G) and principally N2(G) and the hydroxyl group O2' is more often involved in H-bonds than expected; the protein main chain is involved in 32% and the side chains in 68% of the H-bonds; considering the neutral and ionic H-bonds, the following couples are more frequent than expected-base A-Ser, base G-Asp/Glu, base U-Asn. The RNA CH groups interact preferentially with oxygen atoms (62% on the main chain and 19% on the side chains); (iv) the bases are involved in 38% of all H-bonds and more than 26% of the H-bonds have the H donor group on the RNA; (v) the atom O2' is involved in 21% of all H-bonds, a number greater than expected; (vi) amino acids less frequently in direct contact with RNA components interact frequently via their main chain atoms through water molecules with RNA atoms; in contrast, those frequently observed in direct contact, except Ser, use instead their side chain atoms for water bridging interactions.     

Assembly of an RNA-protein complex. Binding of NusB and NusE (S10) proteins to boxA RNA nucleates the formation of the antitermination complex involved in controlling rRNA transcription in Escherichia coli

Analytical ultracentrifugation and fluorescence anisotropy methods have been used to measure the equilibrium parameters that control the formation of the core subcomplex of NusB and NusE proteins and boxA RNA. This subcomplex, in turn, nucleates the assembly of the antitermination complex that is involved in controlling the synthesis of ribosomal RNA in Escherichia coli and that also participates in forming the N protein-dependent antitermination complex in lambdoid phage synthesis. In this study we determined the dissociation constants (K(d) values) for the individual binary interactions that participate in the assembly of the ternary NusB-NusE-boxA RNA subassembly, and we showed that multiple equilibria, involving both specific and nonspecific binding, are involved in the assembly pathway of this protein-RNA complex. The measured K(d) values were used to model the in vitro assembly reaction and combined with in vivo concentration data to simulate the overall control of the assembly of this complex in E. coli at two different cellular growth rates. The results showed that at both growth rates assembly proceeds via the initial formation of a weak but specific NusB-boxA complex, which is then stabilized by NusE binding. We showed that NusE also binds nonspecifically to available single-stranded RNA sequences and that such nonspecific protein binding to RNA can help to regulate crucial interactions in the assembly of the various macromolecular machines of gene expression.     

The roads to and from the RNA world

The historical existence of the RNA world, in which early life used RNA for both genetic information and catalytic ability, is widely accepted. However, there has been little discussion of whether protein synthesis arose before DNA or what preceded the RNA world (i.e. the pre-RNA world). We outline arguments of what route life may have taken out of the RNA world: whether DNA or protein followed. Metabolic arguments favor the possibility that RNA genomes preceded the use of DNA as the informational macromolecule. However, the opposite can also be argued based on the enhanced stability, reactivity, and solubility of 2-deoxyribose as compared to ribose. The possibility that DNA may have come before RNA is discussed, although it is a less parsimonious explanation than DNA following RNA.     

Two-and three-dimensional HCN experiments for correlating base and sugar resonances in 15N, 13C-labeled RNA oligonucleotides

Summary New 2D and 3D ¹H-¹³C-¹⁵N triple resonance experiments are presented which allow unambiguous assignments of intranucleotide H1'-H8(H6) connectivities in ¹³C-and ¹⁵N-labeled RNA oligonucleotides. Two slightly different experiments employing double INEPT forward and back coherence transfers are optimized to obtain the H1'-C1'-N9/N1 and H8/H6-C8/C6-N9/N1 connectivities, respectively. The correlation of H1' protons to glycosidic nitrogens N9/N1 is obtained in a nonselective fashion. To correlate H8/H6 with their respective glycosidic nitrogens, selective ¹³C-refocusing and ¹⁵N-inversion pulses are applied to optimize the magnetization transfers along the desired pathway. The approach employs the heteronuclear one-bond spin-spin interactions and allows the 2D ¹H-¹⁵N and 3D¹H-¹³C-¹⁵N chemical shift correlation of nuclei along and adjacent to the glycosidic bond. Since the intranucleotide correlations obtained are based exclusively on through-bond scalar interactions, these experiments resolve the ambiguity of intra-and internucleotide H1'-H8(H6) assignments obtained from the 2D NOESY spectra. These experiments are applied to a 30-base RNA oligonucleotide which contains the binding site for Rev protein from HIV.