TectoRNA: modular assembly units for the construction of RNA nano-objects

Structural information on complex biological RNA molecules can be exploited to design tectoRNAs or artificial modular RNA units that can self-assemble through tertiary interactions thereby forming nanoscale RNA objects. The selective interactions of hairpin tetraloops with their receptors can be used to mediate tectoRNA assembly. Here we report on the modulation of the specificity and the strength of tectoRNA assembly (in the nanomolar to micromolar range) by variation of the length of the RNA subunits, the nature of their interacting motifs and the degree of flexibility of linker regions incorporated into the molecules. The association is also dependent on the concentration of magnesium. Monitoring of tectoRNA assembly by lead(II) cleavage protection indicates that some degree of structural flexibility is required for optimal binding. With tectoRNAs one can compare the binding affinities of different tertiary motifs and quantify the strength of individual interactions. Furthermore, in analogy to the synthons used in organic chemistry to synthesize more complex organic compounds, tectoRNAs form the basic assembly units for constructing complex RNA structures on the nanometer scale. Thus, tectoRNA provides a means for constructing molecular scaffoldings that organize functional modules in three-dimensional space for a wide range of applications.

Eutectic phases in ice facilitate nonenzymatic nucleic acid synthesis

Polymeric compounds similar to oligonucleotides are relevant to the origin of life and particularly to the concept of an RNA world. Although short oligomers of RNA can be synthesized nonenzymatically under laboratory conditions by second-order reactions in concentrated solutions, there is no consensus on how these polymers could have been synthesized de novo on the early Earth from dilute solutions of monomers. To address this question in the context of an RNA world, we have explored ice eutectic phases as a reaction medium. When an aqueous solution freezes, the solutes become concentrated in the spaces between the ice crystals. The increased concentration offsets the effect of the lower temperature and accelerates the reaction. Here we show that in the presence of metal ions in dilute solutions, frozen samples of phosphoimidazolide-activated uridine react within days at -18 degrees C to form oligouridylates up to 11 bases long. Product yields typically exceed 90%, and approximately 30% of the oligomers include one or more 3'-5' linkages. These conditions facilitate not only the notoriously difficult oligouridylate synthesis, but also the oligomerization of activated cytidylate, adenylate, and guanylate. To our knowledge, this represents the first report to indicate that ice matrices on the early Earth may have accelerated certain prebiotic polymerization reactions.

Purification and cloning of cytotoxic ribonucleases from Rana catesbeiana (bullfrog)

Ribonucleases with antitumor activity are mainly found in the oocytes and embryos of frogs, but the role of these ribonucleases in frog development is not clear. Moreover, most frog ribonuclease genes have not been cloned and characterized. In the present study, a group of ribonucleases were isolated from Rana catesbeiana (bullfrog). These ribonucleases in mature oocytes, namely RC-RNase, RC-RNase 2, RC-RNase 3, RC-RNase 4, RC-RNase 5 and RC-RNase 6, as well as liver-specific ribonuclease RC-RNase L1, were purified by column chromatographs and detected by zymogram assay and western blotting. Characterization of these purified ribonucleases revealed that they were highly conserved in amino acid sequence and had a pyroglutamate residue at their N-termini, but possessed different specific activities, base specificities and optimal pH values for their activities. These ribonucleases were cytotoxic to cervical carcinoma HeLa cells, but their cytotoxicities were not closely correlated to their enzymatic specific activities. Some other amino acid residues in addition to their catalytic residues were implicated to be involved in the cytotoxicity of the frog ribonucleases to tumor cells. Because the coding regions lack introns, the ribonuclease genes were cloned by PCR using genomic DNA as template. Their DNA sequences and amino acid sequences are homologous to those of mammalian ribonuclease superfamily, approximately 50 and approximately 25%, respectively.

Analysis of the RNA-editing reaction of ADAR2 with structural and fluorescent analogues of the GluR-B R/G editing site

ADARs are adenosine deaminases responsible for RNA editing reactions that occur in eukaryotic pre-mRNAs, including the pre-mRNAs of glutamate and serotonin receptors. Here we describe the generation and analysis of synthetic ADAR2 substrates that differ in structure around an RNA editing site. We find that five base pairs of duplex secondary structure 5' to the editing site increase the single turnover rate constant for deamination 17-39-fold when compared to substrates lacking this structure. ADAR2 deaminates an adenosine in the sequence context of a natural editing site >90-fold more rapidly and to a higher yield than an adjacent adenosine in the same RNA structure. This reactivity is minimally dependent on the base pairing partner of the edited nucleotide; adenosine at the editing site in the naturally occurring A.C mismatch is deaminated to approximately the same extent and only 4 times faster than adenosine in an A.U base pair at this site. A steady-state rate analysis at a saturating concentration of the most rapidly processed substrate indicates that product formation is linear with time through at least three turnovers with a slope of 13 +/- 1.5 nM.min(-1) at 30 nM ADAR2 for a k(ss) = 0.43 +/- 0.05 min(-1). In addition, ADAR2 induces a 3.3-fold enhancement in fluorescence intensity and a 14 nm blue shift in the emission maximum of a duplex substrate with 2-aminopurine located at the editing site, consistent with a mechanism whereby ADAR2 flips the reactive nucleotide out of the double helix prior to deamination.

Sequence- and regio-selectivity in the montmorillonite-catalyzed synthesis of RNA

The six binary montmorillonite clay-catalyzed reactions of the 5'-phosphorimidazolides of adenosine, cytidine, guanosine and uridine were performed and the eight dimers from each reaction were separated and analyzed by HPLC. A 16-51-fold higher yield of the 5'-purine-pyrimidine dimers over that of the 5'-pyrimidine-purines was observed. The total yield of the 5'-purine-pyrimidine dimers was in the 50-70% range while that of the 5'-pyrimidine-purine dimers was 1.3-7.0%. Less sequence selectivity was observed in the homodimers formed. Regioselectivity for the formation of 3', 5'-phosphodiester bonds over that found in the absence of clay was observed. The 5'-purine-pyrimidine, 5'-pyrimidine-pyrimidine and 5'-purine-purine dimers had 3', 5'-links in about half of their phosphodiester bonds. The percent phosphodiester links in the 5'-pyrimidine-pyrimidine dimers was 18%, a value close to that observed in the absence of the montmorillonite catalyst. The montmorillonite-catalyzed reaction of all four activated nucleotides was performed and the 24 products were separated and analyzed. The trends observed in the binary reactions were confirmed and the results also showed that the relative reactivity of the activated monomers was A > G > C > U in the ratio 8.2:4.8:1.3:1 respectively. No 5'-pyrimidine-purines with a 5'-U and pG3' pU, pC3' pA and pC3' pG were detected. These studies suggest that a limited population of RNAs would have formed in catalyzed prebiotic reactions.

NMR structures of r(GCAGGCGUGC)2 and determinants of stability for single guanosine-guanosine base pairs

Nucleotides in RNA that are not Watson-Crick-paired form unique structures for recognition or catalysis, but determinants of these structures and their stabilities are poorly understood. A single noncanonical pair of two guanosines (G) is more stable than other noncanonical pairs and can potentially form pairing structures with two hydrogen bonds in four different ways. Here, the energetics and structure of single GG pairs are investigated in several sequence contexts by optical melting and NMR. The data for r(5'GCAGGCGUGC3')(2), in which G4 and G7 are paired, are consistent with a model in which G4 and G7 alternate syn glycosidic conformations in a two-hydrogen-bond pair. The two distinct structures are derived from nuclear Overhauser effect spectroscopic distance restraints coupled with simulated annealing using the AMBER 95 force field. In each structure, the imino and amino protons of the anti G are hydrogen bonded to the O6 and N7 acceptors of the syn G, respectively. An additional hydrogen-bond connects the syn G amino group to the 5' nonbridging pro-R(p) phosphate oxygen. The GG pair fits well into a Watson-Crick helix. In r(5'GCAGGCGUGC3')(2), the G4(anti), G7(syn) structure is preferred over G4(syn), G7(anti). For single GG pairs in other contexts, exchange processes make interpretation of spectra more difficult but the pairs are also G(syn), G(anti). Thermodynamic data for a variety of duplexes containing pairs of G, inosine, and 7-deazaguanosine flanked by GC pairs are consistent with the structural and energetic interpretations for r(5'GCAGGCGUGC3')(2), suggesting similar GG conformations.

Extended in vitro selection for synthesis of novel molecular recognition oligonucleotide derivatives

In vitro selection of RNA aptamer containing biotin-carrying nucleotide was carried out used for development of a new type of molecular sensor. Cytidine 5'-triphosphate (CTP) carrying biotinyl group at the N6-position was used in this technique. A pool of biotin-containing RNAs, which binds specifically to adenosine 5'-triphosphate (ATP), was obtained and used for competitive binding assay of ATP. The selected nonnatural RNA possesses many biotinyl groups to render it a high sensitivity toward ATP.

Pyrimidine motif triplexes containing polypurine RNA or DNA with oligo 2'-O-methyl or DNA triplex forming oligonucleotides

Triplex forming oligonucleotides (TFOs) are potentially useful in targeting RNA for antisense therapeutic applications. To determine the feasibility of targeting polypurine RNA with nuclease-resistant oligonucleotides, TFOs containing 2'-deoxy or 2'-O-methyl (2'-OMe) backbones, designed to form pyrimidine motif triplexes with RNA, were synthesized. TFOs were made which can form trimolecular triplexes, or bimolecular, 'clamp' triplexes with polypurine RNA and DNA. It was found that the relative stabilities of the triplexes formed followed the order: M.DM(clamp)>>>D.DD approximately M.DD>M. RM>D.DM>M.RD approximately M.DM, where M is a 2'-OMe, D is a DNA and R is an RNA backbone. The third strand is listed first, separated by a dot from the purine strand of the Watson-Crick duplex, followed by the pyrimidine strand of the duplex. The results described here provide insight into the feasibility of using TFOs containing a 2'-OMe backbone as antisense agents.

A plausible mechanism for gene correction by chimeric oligonucleotides

Self-complementary chimeric oligonucleotides that consist of DNA and 2'-O-methyl RNA nucleotides arranged in a double-hairpin configuration can elicit a point mutation when targeted to a gene sequence. We have used a series of structurally diverse chimeric oligonucleotides to correct a mutant neomycin phosphotransferase gene in a human cell-free extract. Analysis of structure-activity relationships demonstrates that the DNA strand of the chimeric oligonucleotide acts as a template for high-fidelity gene correction when one of its bases is mismatched to the targeted gene. By contrast, the chimeric strand of the oligonucleotide does not function as a template for gene repair. Instead, it appears to augment the frequency of gene correction by facilitating complex formation with the target. In the presence of RecA protein, each strand of a chimeric oligonucleotide can hybridize with double-stranded DNA to form a complement-stabilized D-loop. This reaction, which may take place by reciprocal four-strand exchange, is not observed with oligonucleotides that lack 2'-O-methyl RNA segments. Preliminary sequencing data suggest that complement-stabilized D-loops may be weakly mutagenic. If so, a low level of random mutagenesis in the vicinity of the chimera binding site may accompany gene repair.

Transverse relaxation optimized triple-resonance NMR experiments for nucleic acids

Triple resonance HCN and HCNCH experiments are reliable methods of establishing sugar-to-base connectivity in the NMR spectra of isotopicaly labeled oligonucleotides. However, with larger molecules the sensitivity of the experiments is drastically reduced due to relaxation processes. Since the polarization transfer between 13C and 15N nuclei relies on rather small heteronuclear coupling constants (11-12 Hz), the long evolution periods (up to 30-40 ms) in the pulse sequences cannot be avoided. Therefore any effort to enhance sensitivity has to concentrate on manipulating the spin system in such a way that the spin-spin relaxation rates would be minimized. In the present paper we analyze the efficiency of the two known approaches of relaxation rate control, namely the use of multiple-quantum coherence (MQ) and of the relaxation interference between chemical shift anisotropy and dipolar relaxation - TROSY. Both theoretical calculations and experimental results suggest that for the sugar moiety (H1'-C1'-N1/9) the MQ approach is clearly preferable. For the base moiety (H6/8-C6/8-N1/9), however, the TROSY shows results superior to the MQ suppression of the dipole-dipole relaxation at moderate magnetic fields (500 MHz) and the sensitivity improvement becomes dramatically more pronounced at very high fields (800 MHz). The pulse schemes of the triple-resonance HCN experiments with sensitivity optimized performance for unambiguous assignments of intra-residual sugar-to-base connectivities combining both approaches are presented.

Experimental and theoretical studies of the effects of deoxyribose substitutions on the stability of the UUCG tetraloop

Experimental and theoretical thermodynamic studies of the consequences of 2'-hydroxyl substitution in the RNA UUCG tetraloop show distinct position dependence consistent with the diverse structural contexts of the four-loop ribose hydroxyls in this motif. The results suggest that even for simple substitutions, such as the replacement of the ribose hydroxyl (2'-OH) with hydrogen (2'-H), the free energy change reflects a complex interplay of hydrogen bonding and solvation effects and is influenced by the intrinsic pucker preferences of the nucleotides. Furthermore, theoretical studies suggest that the effect of these mutations in the single-strand state is sequence dependent, in contrast to what is commonly assumed. Free energy perturbation simulations of ribose-deoxyribose mutations in a single-strand dodecamer and in trinucleotide models suggest that in the denatured state, the magnitude of the free energy change for deoxyribose substitutions is determined to a larger extent by the identity of the nucleotide (A, C, G or U) rather than its structural context. Single-strand mutational effects must be considered when interpreting mutational studies in molecular terms.

Using phosphorothioate-substituted RNA to investigate the thermodynamic role of phosphates in a sequence specific RNA-protein complex

Part of the binding affinity and specificity in RNA-protein complexes is often contributed by contacts between the protein and backbone phosphates that are held in position by the RNA structure. This study focuses on the well-characterized interaction between a dimer of the MS2 coat protein and a small RNA hairpin. Using a short oligoribonucleotide which contains all the necessary sequence elements required for tight protein binding, a single phosphorothioate linkage was introduced at 13 different positions. In each case, the R(P) and S(P) stereoisomers were separated and their affinities to the MS2 coat protein were determined. Comparison of these biochemical data with the crystal structure of the protein-hairpin complex indicates that introduction of a phosphorothioate only affects binding at sites where a protein-phosphate contact is observed in the crystal structure. This means that phosphorothioate-containing oligoribonucleotides should also be useful for mapping phosphate contacts in RNA-protein complexes for which no crystal structure is available.

The structure of ribonuclease P protein from Staphylococcus aureus reveals a unique binding site for single-stranded RNA

Ribonuclease P (RNaseP) catalyses the removal of the 5'-leader sequence from pre-tRNA to produce the mature 5' terminus. The prokaryotic RNaseP holoenzyme consists of a catalytic RNA component and a protein subunit (RNaseP protein), which plays an auxiliary but essential role in vivo by binding to the 5'-leader sequence and broadening the substrate specificity of the ribozyme. We determined the three-dimensional high-resolution structure of the RNaseP protein from Staphylococcus aureus (117 amino acid residues) by nuclear magnetic resonance (NMR) spectroscopy in solution. The protein has an alphabeta-fold, similar to the ribonucleoprotein domain. We used small nucleic acid molecules as a model for the 5'-leader sequence to probe the propensity for generic single-stranded RNA binding on the protein surface. The NMR results reveal a contiguous interaction site, which is identical with the previously identified leader sequence binding site in RNaseP holoenzyme. The conserved arginine-rich motif does not bind single-stranded RNA. It is likely that this peptide segment binds selectively to double-stranded sections of P RNA, which are conformationally more rigid. Given the essentiality of RNaseP for the viability of the organism, knowledge of the S. aureus protein structure and insight into its interaction with RNA will help us to develop RNaseP and RNaseP protein as targets for novel antibiotics against this pathogen.

Applicability of urea in the thermodynamic analysis of secondary and tertiary RNA folding

The equilibrium folding of a series of self-complementary RNA duplexes and the unmodified yeast tRNA(Phe) is studied as a function of urea and Mg(2+) concentration with optical spectroscopies and chemical modification under isothermal conditions. Via application of standard methodologies from protein folding, the folding free energy and its dependence on urea concentration, the m value, are determined. The free energies of the RNA duplexes obtained from the urea titrations are in good agreement with those calculated from thermal melting studies [Freier, S. I., et al. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 9373]. The m value correlates with the length of the RNA duplex and is not sensitive to ionic conditions and temperature. The folding of the unmodified yeast tRNA(Phe) can be described by two Mg(2+)-dependent transitions, the second of which corresponds to the formation of the native tertiary structure as confirmed by hydroxyl radical protection and partial nuclease digestion. Both transitions are sensitive to urea and have m values of 0.94 and 1.70 kcal mol(-)(1) M(-)(1), respectively. Although the precise chemical basis of urea denaturation of RNA is uncertain, the m values for the duplexes and tRNA(Phe) are proportional to the amount of the surface area buried in the folding transition. This proportionality, 0.099 cal mol(-)(1) M(-)(1) A(-)(2), is very similar to that observed for proteins, 0.11 cal mol(-)(1) M(-)(1) A(-)(2) [Myers, J., Pace, N., and Scholtz, M. (1995) Protein Sci. 4, 2138]. These results indicate that urea titration can be used to measure both the free energy and the magnitude of an RNA folding transition.

Normalization of array hybridization experiments in differential gene expression analysis

For detecting and confirming differentially expressed genes it is necessary to have a trustworthy reference. So called 'housekeeping genes' are frequently used for this purpose as internal standard. However, if the influence of new experimental conditions is to be analyzed it is not safe to assume a priori that the expression of these genes is not affected. Therefore two synthetic poly(A)-RNAs were generated by PCR and in vitro transcription. They were used as external standards for normalization of northern blots and cDNA arrays where non-regulated genes as internal reference were not available.

Nonenzymatic oligomerization reactions on templates containing inosinic acid or diaminopurine nucleotide residues

The template-directed oligomerization of nucleoside-5'-phosphoro-2-methyl imidazolides on standard oligonucleotide templates has been studied extensively. Here, we describe experiments with templates in which inosinic acid (I) is substituted for guanylic acid, or 2,6-diaminopurine nucleotide (D) for adenylic acid. We find that the substitution of I for G in a template is strongly inhibitory and prevents any incorporation of C into internal positions in the oligomeric products of the reaction. The substitution of D for A, on the contrary, leads to increased incorporation of U into the products. We found no evidence for the template-directed facilitation of oligomerization of A or I through A-I base pairing. The significance of these results for prebiotic chemistry is discussed.

A Watson-Crick base-pair-disrupting methyl group (m1A9) is sufficient for cloverleaf folding of human mitochondrial tRNALys

We have previously shown by chemical and enzymatic structure probing that, opposite to the native human mitochondrial tRNA(Lys), the corresponding in vitro transcript does not fold into the expected tRNA-specific cloverleaf structure. This RNA folds into a bulged hairpin, including an extended amino acid acceptor stem, an extra large loop instead of the T-stem and loop, and an anticodon-like domain. Hence, one or several of the six modified nucleotides present in the native tRNA are required and responsible for its cloverleaf structure. Phylogenetic comparisons as well as structural analysis of variant transcripts had pointed to m(1)A9 as the most likely important modified nucleotide in the folding process. Here we describe the synthesis of a chimeric tRNA(Lys) with m(1)A9 as the sole modified base and its structural analysis by chemical and enzymatic probing. Comparison of this structure to that of the unmodified RNA, the fully modified native tRNA, and a variant designed to mimic the effect of m(1)A9 demonstrates that the chimeric RNA folds indeed into a cloverleaf structure that resembles that of the native tRNA. Thus, due to Watson-Crick base-pair disruption, a single methyl group is sufficient to induce the cloverleaf folding of this unusual tRNA. This is the first direct evidence of the role of a modified nucleotide in RNA folding.

Dimerization of HIV-1 genomic RNA of subtypes A and B: RNA loop structure and magnesium binding

Retroviruses encapsidate their genome as a dimer of homologous RNA molecules noncovalently linked close to their 5' ends. The dimerization initiation site (DIS) of human immunodeficiency virus type 1 (HIV-1) RNA is a hairpin structure that contains in the loop a 6-nt self-complementary sequence flanked by two 5' and one 3' purines. The self-complementary sequence, as well as the flanking purines, are crucial for dimerization of HIV-1 RNA, which is mediated by formation of a "kissing-loop" complex between the DIS of each monomer. Here, we used chemical modification interference, lead-induced cleavage, and three-dimensional modeling to compare dimerization of subtype A and B HIV-1 RNAs. The DIS loop sequences of these RNAs are AGGUGCACA and AAGCGCGCA, respectively. In both RNAs, ethylation of most but not all phosphate groups in the loop and methylation of the N7 position of the G residues in the self-complementary sequence inhibited dimerization. These results demonstrate that small perturbations of the loop structure are detrimental to dimerization. Conversely, methylation of the N1 position of the first and last As in the loop were neutral or enhanced dimerization, a result consistent with these residues forming a noncanonical sheared base pair. Phosphorothioate interference, lead-induced cleavage, and Brownian-dynamics simulation revealed an unexpected difference in the dimerization mechanism of these RNAs. Unlike subtype B, subtype A requires binding of a divalent cation in the loop to promote RNA dimerization. This difference should be taken into consideration in the design of antidimerization molecules aimed at inhibiting HIV-1 replication.

High-performance liquid chromatography purification of homogenous-length RNA produced by trans cleavage with a hammerhead ribozyme

An improved method is presented for the preparation of milligram quantities of homogenous-length RNAs suitable for nuclear magnetic resonance or X-ray crystallographic structural studies. Heterogeneous-length RNA transcripts are processed with a hammerhead ribozyme to yield homogenous-length products that are then readily purified by anion exchange high-performance liquid chromatography. This procedure eliminates the need for denaturing polyacrylamide gel electrophoresis, which is the most laborious step in the standard procedure for large-scale production of RNA by in vitro transcription. The hammerhead processing of the heterogeneous-length RNA transcripts also substantially improves the overall yield and purity of the desired RNA product.

A novel 5 displacement spin-labeling technique for electron paramagnetic resonance spectroscopy of RNA

An RNA spin-labeling technique was developed using the well-characterized interaction between the HIV Rev peptide and the Rev response element (RRE) RNA as a model system. Spin-labeled RNA molecules were prepared by incorporating guanosine monophosphorothioate (GMPS) at the 5' end using T7 RNA polymerase and then covalently attaching a thiol-specific nitroxide spin label. Three different constructs of the RRE RNA were made by strategically displacing the 5' end within the native three-dimensional structure. Nitroxide-to-nitroxide distance measurements were made between the specifically bound RNA and peptide using electron paramagnetic resonance (EPR) spectroscopy. The dipolar EPR method can reliably measure distances up to 25 A, the calculation of which is derived from the 1/r3 dependence of the broadening of EPR lines in motionally frozen samples. This RNA-labeling technique, dubbed 5' displacement spin labeling, extends the usefulness of the dipolar EPR method developed for analysis of protein structure. The advantage of this technique is that it is applicable to large RNA systems such as the ribosome, which are difficult to study by other structural methods.

A simple and efficient method to reduce nontemplated nucleotide addition at the 3 terminus of RNAs transcribed by T7 RNA polymerase

DNA templates modified with C2'-methoxyls at the last two nucleotides of the 5' termini dramatically reduced nontemplated nucleotide addition by the T7 RNA polymerase from both single- and double-stranded DNA templates. This strategy was used to generate several different transcripts. Two of the transcripts were demonstrated by nuclear magnetic resonance spectroscopy to be unaffected in their sequence. Transcripts produced from the modified templates can be purified with greater ease and should be useful in a number of applications.

[Nucleic acid syntheses and molecular recognition]

Syntheses of nucleic acids are of importance not only in the natural product chemistry but also in the molecular recognition study in biology. This review describes syntheses of RNA and DNA, which were found to be related each other. Deoxyinosine probes were developed for cloning genes with degenerated codons. Synthetic genes for c-Ha-ras and T4 endonuclease V provided new approaches in recognition of nucleic acids by proteins. Antibodies specific for photo-damaged pyrimidines in DNA have been studied by cloning and mutating their genes.

Derivation of RNA aptamer inhibitors of human complement C5

Specific aptamer inhibitors of the human complement C5 component were produced by the SELEX methodology of directed evolution of nucleic acid ligands. The SELEX procedure started with a pool of random-sequence, 2'F-pyrimidine-modified nuclease-stabilized RNA, and after twelve rounds of iterative C5 binding and nucleic acid amplification an evolved RNA pool was obtained which contained the highest affinity binders to the C5 protein. The evolved RNA pool was then cloned and sequenced, and individual clones were analyzed for binding and function. Twenty-eight clones (out of sixty) were identified which bound C5 (termed aptamers). Seven of these aptamers formed a closely related sequence homology family; these aptamers bound C5 with a Kd 20-40 nM and also inhibited human serum hemolytic activity. In addition, these aptamers inhibited zymosan-induced generation of C5a. Aptamer inhibition of both C5b and C5a suggests that aptamer binding inhibits cleavage of C5 by the C5 convertase of both pathways. One of the inhibitory aptamer sequences was truncated to yield a 38-mer 2'F RNA aptamer which retained C5 binding and inhibitory activity. The structure of this aptamer is predicted to be a stem-loop containing thirteen base pairs, and also containing two bulges. The affinity of this aptamer was improved by performing a second biased SELEX experiment, where the randomized starting RNA pool uses a template where the individual base compositions are biased toward a specific sequence. This second SELEX experiment produced an aptamer with a Kd of 2-5 nM which retained functional activity. Another SELEX to rat C5 produced an aptamer with binding and inhibitory properties virtually identical with the human aptamer. The human and rat aptamers are being evaluated for complement inhibition in vitro and in vivo as potential therapeutics for treatment of human disease.

Ternary conjugates of guanosine monophosphate as initiator nucleotides for the enzymatic synthesis of 5'-modified RNAs

We give a detailed account on the enzymatic synthesis of RNA conjugates by T7 RNA polymerase using modified initiator nucleotides during transcription. Following two different routes, ternary conjugates of guanosine-5'-monophosphate, poly(ethylene glycol), and anthracene were synthesized via phosphoramidite intermediates and characterized by a variety of spectroscopic techniques. Up to a degree of polymerization nPEG of about 17, these conjugates were efficiently incorporated into RNA by T7 RNA polymerase at the 5'-termini, thereby giving access to RNA conjugates required for biochemical studies as well as for the exploration of the catalytic potential of ribonucleic acids. The resulting conjugates are intact and functional.

Crystal structure of a 14 bp RNA duplex with non-symmetrical tandem GxU wobble base pairs

Adjacent GxU wobble base pairs are frequently found in rRNA. Atomic structures of small RNA motifs help to provide a better understanding of the effects of various tandem mismatches on duplex structure and stability, thereby providing better rules for RNA structure prediction and validation. The crystal structure of an RNA duplex containing the sequence r(GGUAUUGC-GGUACC)2 has been solved at 2.1 A resolution using experimental phases. Novel refinement strategies were needed for building the correct solvent model. At present, this is the only short RNA duplex structure containing 5'-U-U-3'/3'-G-G-5' non-symmetric tandem GxU wobble base pairs. In the 14mer duplex, the six central base pairs are all displaced away from the helix axis, yielding significant changes in local backbone conformation, helix parameters and charge distribution that may provide specific recognition sites for biologically relevant ligand binding. The greatest deviations from A-form helix occur where the guanine of a wobble base pair stacks over a purine from the opposite strand. In this vicinity, the intra-strand phosphate distances increase significantly, and the major groove width increases up to 3 A. Structural comparisons with other short duplexes containing symmetrical tandem GxU or GxT wobble base pairs show that nearest-neighbor sequence dependencies govern helical twist and the occurrence of cross-strand purine stacks.

Kinetic preference for the 3'-5'-linked dimer in the reaction of guanosine 5'-phosphorylmorpholinamide with deoxyguanosine 5'-phosphoryl-2-methylimidazolide as a function of poly(C) concentration

The formation of the internucleotide bond in diguanylate synthesis was studied in aqueous solution at pH 8 and 0.2 M Mg2+ in the presence and absence of polycytidylate, poly(C). The investigation was simplified by using guanosine 5'-phosphorylmorpholinamide, mor-pG, which can act only as a nucleophile, and deoxyguanosine 5'-phosphoryl-2-methylimidazolide, 2-MeImpdG, which can act only as an electrophile. The time-dependent product distribution was monitored by high-performance liquid chromatography (HPLC) and liquid chromatography mass spectrometry (LC/MS). In the absence of poly(C) the reaction between mor-pG and 2-MeImpdG yielded small amounts of the dimer mor-pGpdG with a regioselectivity of 2'-5':3'-5' = 3.5. In the presence of poly(C) dimer yields increased and a reversal in regioselectivity occurred; both effects were in proportion to the concentration of the polymer. The results can be quantitatively explained with the proposition that poly(C), acting as the template, catalyzes the reaction between template-bound monomers by about a factor of 4-5 over the reaction in solution and yields dimers with a regioselectivity of 2'-5':3'-5' approximately 0.33. These findings illustrate the intrinsic preference of guanosine monomers to correctly self-assemble on the appropriate template.

Formation of RNA oligomers on montmorillonite: site of catalysis

Certain montmorillonites catalyze the self condensation of the 5'-phosphorimidazolide of nucleosides in pH 8 aqueous electrolyte solutions at ambient temperatures leading to formation of RNA oligomers. In order to establish the nature of the sites on montmorillonite responsible for this catalytic activity, oligomerization reactions were run with montmorillonites which had been selectively modified (I) at the edges by (a) fluoride treatment, (b) silylation, (c) metaphosphate treatment of the anion exchange sites (II) in the interlayer by (a) saturation with quaternary alkylammonium ions of increasing size, (b) aluminum polyoxo cations. High pressure liquid chromatography, HPLC, analysis of condensation products for their chain lengths and yields indicated that modification at the edges did not affect the catalytic activity to a significant extent, while blocking the interlayer strongly inhibited product formation.

Creation of RNA molecules that recognize the oxidative lesion 7,8-dihydro-8-hydroxy-2'-deoxyguanosine (8-oxodG) in DNA

We used in vitro evolution to obtain RNA molecules that specifically recognize and bind with high affinity to the oxidative lesion 7, 8-dihydro-8-hydroxy-2'-deoxyguanosine (8-oxodG) in DNA. A pool of approximately 10(15) RNA molecules containing a random insert of 45 nucleotides in length was subject to 10 successive rounds of chromatographic enrichment using an 8-oxodG affinity matrix, reverse transcription, PCR amplification, and RNA synthesis. Selected RNA molecules bind to 8-oxodG located at the 3' terminus (Kd Collapse

Key Words Collapse

MESH Headings

• 8-Hydroxy-2'-Deoxyguanosine
• Base Sequence
• Binding Sites
• Binding, Competitive
• Cloning, Molecular
• DNA/chemistry
• DNA/metabolism
• DNA Damage
• DNA Primers/genetics
• Deoxyguanosine/analogs & derivatives
• Deoxyguanosine/metabolism
• Directed Molecular Evolution
• In Vitro Techniques
• Molecular Sequence Data
• Nucleic Acid Conformation
• Oxidation-Reduction
• RNA/chemical synthesis
• RNA/genetics
• RNA/metabolism

Collapse

Grants

• F32 CA073141 NCI NIH HHS
• AG-01 751 NIA NIH HHS
• F2CA73141A NCI NIH HHS
• R-35-CA3990S NCI NIH HHS

Collapse

Pre-steady-state kinetic characterization of RNA-primed initiation of transcription by HIV-1 reverse transcriptase and analysis of the transition to a processive DNA-primed polymerization mode

Single-turnover and equilibrium measurements were carried out to determine the basis of the apparently slow, nonprocessive polymerization reaction catalyzed by HIV-1 reverse transcriptase (RT) during transcription initiation, when both the primer and template are composed of RNA. Comparison of the binding and kinetic parameters of a 20-mer, all-RNA primer/35-mer template substrate to one identical in sequence but composed of a 20-mer, all-DNA primer/35-mer RNA template reveals striking differences. Equilibrium titrations yielded a dissociation constant (Kd) >200 nM for the RNA/RNA-RT complex which is at least 200-fold higher than that of the DNA/RNA-substrate (Kd approximately 1 nM). The affinity of the RT-RNA/RNA complex for dTTP was found to be at least 500 times lower (Kd approximately 3.4 mM) than that of the RT-DNA/RNA complex (Kd approximately 6.6 microM). The single-turnover dNTP incorporation time course using the RNA-primer substrate, the DNA-primer substrate, or a series of RNA-primer substrates preextended with one to eight deoxynucleotides showed that dNTP incorporation occurs with a biphasic exponential burst of +1 extension product, followed by a linear phase. At least three different RT-bound forms of the p/ts exist: a fast, kinetically competent form (single-turnover rate approximately 10-50 s-1); a slow form (rate approximately 0.3-1 s-1); and a form that is dead-end (no turnover). The studies further revealed that a switch to a fast, kinetically competent p/t occurs after six dNTPs are incorporated into the RNA primer, with the switch being defined as the transition from a minority to a majority of the p/t bound in the optimal manner.

In vitro selection of phosphorothiolated aptamers

A pool of RNA molecules that contained exclusively phosphorothioate internucleoside linkages was used as a starting point for the selection of aptamers that bind to basic fibroblast growth factor (bFGF), and appear to act as heparin mimics.

Correlating structure and stability of DNA duplexes with incorporated 2'-O-modified RNA analogues

Chemically modified nucleic acids are currently being evaluated as potential antisense compounds for therapeutic applications. 2'-O-Ethylene glycol substituted oligoribonucleotides are second-generation antisense inhibitors of gene expression with promising features for in vivo use. Relative to DNA, they display improved RNA affinity and higher nuclease resistance. Moreover, chimeric oligonucleotides with 2'-O-methoxyethyl ribonucleoside wings and a central DNA phosphorothioate window have been shown to effectively reduce the growth of tumors in animal models at low doses. Using X-ray crystallography, we have determined the structures of three A-form DNA duplexes containing the following 2'-O-modified ribothymidine building blocks: 2'-O-methoxyethyl ribo-T, 2'-O-methyl[tri(oxyethyl)] ribo-T, and 2'-O-ethoxymethylene ribo-T. In contrast to 2'-O-ethylene glycol substituents, the presence of a 2'-O-ethoxymethylene group leads to slightly reduced RNA affinity of the corresponding oligonucleotides. The three structures allow a qualitative rationalization of the differing stabilities of duplexes between oligonucleotides comprising these types of 2'-O-modified ribonucleotides and complementary RNAs. The stabilizing 2'-O-ethylene glycol substituents are conformationally preorganized for the duplex state. Thus, the presence of one or several ethylene glycol moieties may reduce the conformational space of the substituents in an oligonucleotide single strand. In addition, most of these preferred conformations appear to be compatible with the minor groove topology in an A-type duplex. Factors that contribute to the conformational rigidity of the 2'-O-substituents are anomeric and gauche effects, electrostatic interactions between backbone and substituent, and bound water molecules.

Non-enzymatic, template-directed ligation of 2'-5' oligoribonucleotides. Joining of a template and a ligator strand

Decauridylate containing exclusively a 2'-5' phospho-diester bond ([2'-5']U10) served as a template for the synthesis of oligoadenylates [oligo(A)s] from the 5'-phosphorimidazolide of 2'-5' diadenylate (ImpA-2'p5'A). Joining of [2'-5']U10and ImpA2'p5'A also took place in substantial amounts to yield long-chain oligoribonucleotides in the template-directed reaction. An unusual CD spectrum ascribed to helix formation between [2'-5']U10and [2'-5'](pA)2was observed under the same conditions as that of the template-directed reaction. The 3'-5' linked decauridylate ([3'-5']U10) also promoted the template-directed synthesis of oligo(A)s from ImpA2'p5'A, but more slowly compared with [2'-5']U10. The results indicate that short-chain RNA oligomers with a 2'-5' phosphodiester bond could lead to longer oligoribonucleotides by template-directed chain elongation.

Physicochemical and biochemical properties of 2',5'-linked RNA and 2',5'-RNA:3',5'-RNA "hybrid" duplexes

In recent publications, oligonucleotides joined by 2',5'-linkages were found to bind to complementary single-stranded RNA but to bind weakly, or not at all, to single-stranded DNA [e.g., P. A. Giannaris and M. J. Damha (1993) Nucleic Acids Res. 21, 4742-4749]. In this work, the biochemical and physicochemical properties of 2',5'-linked oligoribonucleotides containing mixed sequences of the four nucleobases (A, G, C, and U) were evaluated. CD spectra of RNA:2', 5'-RNA duplexes were compared with the spectra of DNA:DNA, RNA:RNA, and DNA:RNA duplexes of the same base sequence. The CD results indicated that the RNA:2',5'-RNA duplex structure more closely resembles the structure of the RNA:DNA hybrid, being more A-form than B-form in character. The melting temperature (Tm) values of the backbone-modified duplexes were compared with the Tm values of the unmodified duplexes. The order of thermal stability was RNA:RNA > DNA:DNA approximately RNA:DNA approximately DNA:RNA > RNA:2',5'-RNA > 2',5'-RNA:2',5'-RNA >> DNA:2',5'-RNA (undetected). RNA:2',5'-RNA duplexes are not substrates of the enzyme RNase H (Escherichia coli, or HIV-1 reverse transcriptase), but they can inhibit the RNase H-mediated cleavage of a natural DNA:RNA substrate. Structural models that are consistent with the selective association properties of 2',5'-linked oligonucleotides are discussed.

Minimal tRNA(Ser) and tRNA(Sec) substrates for human seryl-tRNA synthetase: contribution of tRNA domains to serylation and tertiary structure

The recognition process of tRNA(Ser) and tRNA(Sec) by human seryl-tRNA synthetase (SerRS) was studied using T7 transcripts representing defined regions of human tRNA(Ser) or tRNA(Sec) and the influence of the tRNA elements on serylation and tertiary structure was elucidated. The anticodon arms of both tRNAs showed no contribution to serylation in contrast to the acceptor stems and the long extra arms. D and T arms were only involved in formation of the L-shaped tRNA structure, not in the recognition process between tRNAs and SerRS. This is the first report of microhelices adapted from human tRNAs being aminoacylated by their homologous synthetase.

Interactions of the antiviral quinoxaline derivative 9-OH-B220 [2, 3-dimethyl-6-(dimethylaminoethyl)- 9-hydroxy-6H-indolo-[2, 3-b]quinoxaline] with duplex and triplex forms of synthetic DNA and RNA

The binding of an antiviral quinoxaline derivative, 2,3-dimethyl- 6 - (dimethylaminoethyl) - 9 - hydroxy - 6H - indolo - [2,3 - b]quinoxaline (9-OH-B220), to synthetic double and triple helical DNA (poly(dA).poly(dT) and poly(dA).2poly(dT)) and RNA (poly(rA). poly(rU) and poly (rA).2poly(rU)) has been characterized using flow linear dichroism (LD), circular dichroism (CD), fluorescence spectroscopy, and thermal denaturation. When either of the DNA structures or the RNA duplex serve as host polymers a strongly negative LD is displayed, consistent with intercalation of the chromophoric ring system between the base-pairs/triplets of the nucleic acid structures. Evidence for this geometry also includes weak induced CD signals and strong increments of the fluorescence emission intensities upon binding of the drug to each of these polymer structures. In agreement with intercalative binding, 9-OH-B220 is found to effectively enhance the thermal stability of both the double and triple helical states of DNA as well as the RNA duplex. In the case of poly(dA).2poly(dT), the drug provides an unusually large stabilization of its triple helical state; upon binding of 9-OH-B220 the triplex-to-duplex equilibrium is shifted towards higher temperature by 52.5 deg. C in a 10 mM sodium cacodylate buffer (pH 7.0) containing 100 mM NaCl and 1 mM EDTA. When triplex RNA serves as host structure, LD indicates that the average orientation angle between the drug chromophore plane and the helix axis of the triple helical RNA is only about 60 to 65 degrees. Moreover, the thermal stabilizing capability, as well as the fluorescence increment, CD inducing power and perturbations of the absorption envelope, of 9-OH-B220 in complex with the RNA triplex are all less pronounced than those observed for the complexes with DNA and duplex RNA. These features indicate binding of 9-OH-B220 in the wide and shallow minor groove of poly(rA).2poly(rU). Based on the present results, some implications for the applications of this low-toxic, antiviral and easily administered drug in an antigene strategy, as well as its potential use as an antiretroviral agent, are discussed.

Quantification of multiple human cytochrome P450 mRNA molecules using competitive reverse transcriptase-PCR

We developed a quantitative competitive reverse transcriptase-polymerase chain reaction (QC RT-PCR) assay to measure mRNA levels of seven human cytochrome P450 (P450, CYP) genes and microsomal epoxide hydrolase (EH) simultaneously. This assay employs an exogenous recombinant RNA (rcRNA) molecule as an internal standard that shares PCR primer and hybridization probe sequences with CYP1A1, CYP1A2, CYP2A6/7, CYP2D6, CYP2E1, CYP2F1, CYP3A4/5/7, and EH mRNA. Because each rcRNA molecule contains several primer sequences, an entire battery of genes that exhibit differential responsiveness to various classes of xenobiotics may be measured simultaneously from one population of cDNA molecules. In this study, we demonstrated the precision and power of the assay using small amounts of human liver total RNA. We also report for the first time quantitative profiles of P450 and EH mRNA abundance in eight human livers. Cytochrome P450 2E1 mRNA maintained the highest abundance (average 6.67 x 10(7) molecules/microg of total RNA) and least variation (13 fold) in all livers examined. Cytochrome P450 1A2, CYP2A6/7, CYP2D6, CYP3A4/5, and EH mRNAs were approximately one order of magnitude less abundant than CYP2E1 transcripts, with CYP2D6 levels exhibiting the greatest variation (220 fold) between individuals. This QC RT-PCR assay should prove valuable for measuring basal and induced mRNAs in different cell types in vitro, as well as in biomonitoring applications where individuals are exposed or hypersusceptible to certain xenobiotic-initiated toxicities.

Evolving reaction-diffusion ecosystems with self-assembling structures in thin films

Recently, new types of coupled isothermal polynucleotide amplification reactions for the investigation of in vitro evolution have been established that are based on the multi-enzyme 3SR reaction. Microstructured thin-film open bioreactors have been constructed in our laboratory to run these reactions spatially resolved in flow experiments. Artificial DNA/RNA chemistries close to the in vitro biochemistry of these systems have been developed, which we have studied in computer simulations in configurable hardware (NGEN). These artificial chemistries are described on the level of individual polynucleotide molecules, each with a defined sequence, and their complexes. The key feature of spatial pattern formation provides a weak stabilization of cooperative catalytic properties of the evolving molecules. Of great interest is the step to include extended self-assembly processes of flexible structures-allowing the additional stabilization of cooperation through semipermeable, flexible, self-organizing membrane boundaries. We show how programmable matter simulations of experimentally relevant molecular in vitro evolution can be extended to include the influence of self-assembling flexible membranes.

A new method to synthesize competitor RNAs for accurate analyses by competitive RT-PCR

A method to synthesize competitor RNAs as internal standards for competitive RT-PCR is improved by using the long accurate PCR (LA-PCR) technique. Competitor templates synthesized by the new method are almost the same in length, and possibly in secondary structure, as target mRNAs to be quantified except that they include the short deletion within the segments to be amplified. This allows the reverse transcription to be achieved with almost the same efficiency from both target mRNAs and competitor RNAs. Therefore, more accurate quantification can be accomplished by using such competitor RNAs.

Design of leader sequences that improve the efficiency of the enzymatic synthesis of 2'-amino-pyrimidine RNA for in vitro selection

The application of nucleic acids obtained by in vitro selection from a large pool of molecules with random sequences in medical diagnosis or therapy requires nucleic acids with enhanced stability in biological fluids. Chemical modifications introduced after selection are likely to alter the structure and the properties of the selected molecules. Therefore, the chemical modifications used must be present throughout the selection. This can be achieved for example by the incorporation of 2'-amino-pyrimidine nucleotides into RNA in the transcription step. Though modified molecules could be transcribed from some generally designed dsDNA templates, the efficiency of transcription and reverse transcription and reverse transcription was very low making this strategy too inefficient. Templates and primers with varying amounts of pyrimidines in the constant flanking region of the RNA molecule were designed and their efficiency in transcription and reverse transcription tested. The obtained 2'-amino-pyrimidine RNA molecules showed enhanced stability in serum and RNAse cocktails. Here we present optimized leader sequences flanking the random core-sequence and reaction conditions that allow the reliable utilization of this modification in in vitro selection.

Helical structure formation between complementary oligonucleotides. Minimum chain length required for the template-directed synthesis of oligonucleotides

Helix formation between various combinations of 3'-5' linked oligoribouridylates and oligoriboadenylates from dimer to dodecamer has been studied to gain information on the chain-length requirement for the template-directed condensation of oligoribonucleotides. We have measured the helix formation under high oligoribonucleotide concentration in the presence of magnesium ion at 0-50 degrees C by UV or CD, as many model processes of oligoribonucleotides replication have been carried out under such conditions. Adenylic acid, (pA), diadenylic acid, (pA)2, or triadenylic acid, (pA)3, forms a helix with poly(U) or oligo(U) with a chain length of more than eight. On the other hand, neither uridylic acid, (pU), nor diuridylic acid. (pU)2, can form a helix with oligo(A) or poly(A). Triuridylic acid, (pU)3, or the longer oligo(U) forms a helix with oligo(A) with a chain length of over six. The results suggest that a trimer is the minimum unit as an incorporating nucleotide for conducting any set of nonenzymatic template-directed synthesis, A --> U and U --> A, as the nonenzymatic template-directed condensation of oligoribonucleotides correlates well with the results of helix formation of complementary oligoribonucleotides. We have further found the partial helix formation between 2'-5' linked decauridylate, (pU)10, and pA or 2'-5' linked (pA)2 at 0 degrees C, which indicates the possibility of the template activity of long 2'-5' linked oligonucleotides for the nonenzymatic oligonucleotide synthesis.

Emergence of template-and-sequence-directed (TSD) syntheses: II. A computer simulation model

The initiation of the bio-geochemical scenario described in Part I serves in the present work as the basis for computer modeling, where the central process of the simulation algorithm. i.e., peptide-catalyzed oligomeric growth, is based on mass action equations. The computer model starts with a minimal system in which catalyzed growth processes of proto-RNA templates and small peptides take place, starting from their building blocks. The emerging populations of random oligomers also include a very small fraction of proto-tRNAs and a small fraction of catalytic peptides. Using simplifying assumptions regarding catalyzed proto-RNA template-replication, as well as selectivity of certain molecules and processes, the proportion of proto-tRNA in the proto-RNA molecular population increases rapidly; it is followed by TSD peptide synthesis, based on an ad hoc genetic code and specific peptide catalysts allocated for this synthesis. Consequently, a feedback system is initiated in which TSD peptides involved in the relevant catalytic reactions of the TSD syntheses also start to accumulate. The initial sporadic formation of TSD peptides is thus replaced gradually by cycles of positive feedback and autocatalysis characterized by accumulation of catalytic peptides and Proto-tRNAs and TSD-Reaction-Takeover. The model system which can be considered a 'toy model' can synthesize its templates and catalysts under a wide range of reaction parameters and initial concentrations, thus demonstrating a robustness which is essential for molecular evolution processes. The critical stage of the buildup of a molecular mechanism for the initiation of a minimal TSD reaction cycle has thus been described; because of the centrality of TSD reaction cycles in biology, it is assumed to be central also in the origin of life processes.

Synthesis and two-dimensional electrophoretic analysis of mixed populations of circular and linear RNAs

Spontaneous cleavage of the less abundant form of tobacco ringspot virus satellite RNA is readily reversible. Capitalizing on earlier observations by Feldstein and Bruening that small 'mini-monomer' RNAs derived from this molecule and containing little more than covalently attached ribozyme and substrate cleavage products are able to efficiently circularize, we have constructed a series of self-circularizing RNAs of precisely known size. Mixtures of linear and circular RNAs synthesized in vitro and containing 225-1132 nt could be completely resolved using a novel two-dimensional denaturing polyacrylamide gel electrophoresis system. Similar analyses of a complex mixture of coconut cadang-cadang viroid RNAs revealed the presence of relatively large amounts of a previously undescribed 'fast-slow' heterodimeric RNA species in infected palms. Only a single DNA template is required to prepare each pair of circular and linear RNA markers.

Effect of dinucleoside pyrophosphates on the oligomerization of activated mononucleotides on Na(+)-montmorillonite: reaction of 5'-phosphoro-4-(dimethylamino)pyridinium [4-(CH3)2NpypA] with A5'ppA

The oligomerization of adenosine 5'-phosphoro-4-(dimethylamino)pyridinium (4-(CH3)2-NpypA) and diadenosine 5',5'-pyrophosphate (A5'ppA) (9:1) on Na(+)-montmorillonite was studied. The oligomers were isolated and analyzed by selective enzymatic hydrolyses and the oligomeric composition and the percent of 3',5'-phosphodiester linkages present in each fraction was determined. The longest oligomers formed (11-mers) are slightly shorter than those produced in the absence of A5'ppA (12-mers). Smaller amounts of A5'ppA are incorporated into the oligomers than in the ImpA/A5'ppA reaction. The regioselectivity of 3',5'-phosphodiester bond formation is comparable to that of the oligomerization of 4-(CH3)2NpypA alone. An explanation of these data is proposed and the possible effect of dinucleoside pyrophosphate on prebiotic RNA formation is discussed.