The interactions of selected antibiotics with the trans-acting antigenomic delta ribozyme were mapped. Ribozyme with two oligonucleotide substrates was used, one uncleavable with deoxycytidine at the cleavage site, mimicking the initial state of ribozyme, and the other with an all-RNA substrate mimicking, after cleavage, the product state. Mapping was performed with a set of RNA structural probing methods: Pb2+ -induced cleavage, nuclease digestion, and the selective 2′-hydroxyl acylation analyzed by primer extension (SHAPE) approach. The experimental results combined with molecular modeling revealed different binding sites for neomycin B, amikacin and actinomycin D inside the ribozyme structure. Neomycin B, an aminoglycoside antibiotic, which strongly inhibited the catalytic properties of delta ribozyme, was bound to the pocket formed by the P1 stem, the P1.1 pseudoknot, and the J4/2 junction. Amikacin showed less effective binding to the ribozyme catalytic core, resulting in weak inhibition. Complexes of these aminoglycosides with Cu2+ ions were bound to the same ribozyme regions, but more effectively, showing lower Kd values. On the other hand, the Cu2+ complex of the cyclopeptide antibiotic actinonomycin D was preferentially intercalated into the P2 and the P4 double-stranded region, and was three times more potent in ribozyme inhibition than the free antibiotic. In addition, some differences in SHAPE reactivities between the ribozyme forms containing all-RNA and deoxycytidine-modified substrates in the J4/2 region were detected, pointing to different ribozyme conformations before and after the cleavage event.
selective 2′-hydroxyl acylation analyzed by primer extension
Antibiotics constitute a powerful class of drugs that target a variety of microorganisms, such as bacteria, fungi, and other parasites. Currently, a broad spectrum of antibiotics is used to treat a wide range of infections, but some of them show undesirable side effects. It is therefore necessary to better understand the molecular mechanism of the action of antibiotics in the cell. Typically, the growth of bacteria is reduced by antibiotics interacting with several cellular targets: proteins, cell wall components, and nucleic acids [1, 2]. Many antibiotics targeting RNA molecules bind with high affinity to rRNA, mRNA, or tRNA, affecting the process of translation of the genetic message .
Recently, the crystal structure of the bacterial ribosome in complex with several antibiotics has been resolved. Analysis of the data revealed that some antibiotics interact directly with 16S rRNA in the small ribosomal subunits and with 23S rRNA in the large ribosomal subunits [4, 5]. For example, several aminoglycoside antibiotics, such as kanamycin, tobramycin, gentamicin, paromomycin, neomycin, apramycin, and streptomycin, bind to the A site of the 30S ribosomal subunit. These antibiotics interact with 16S rRNA, causing errors in decoding, and affect the mRNA path or inhibit the mRNA–tRNA translocation process . On the other hand, oxazilidone and macrolide antibiotics bind to the 50S large ribosomal subunit, affecting the activity of the peptidyltransferase center, which is crucial for the biological function of the ribosome, i.e. peptide bond formation .
Important information concerning the principles underlying the interactions of RNA with antibiotics can be obtained by analyzing their effects on RNA catalysis performed by ribozymes. The advantages of using ribozymes as model systems are their complex structure and the ease of testing ribozyme activities in relation to RNA structure and function . In addition, the structure and function of many different ribozymes, including the large introns group I and II, as well as small ribozymes, such as hammerhead, hairpin, delta, and the recently discovered glmS, have been examined and elucidated [9-11]. In particular, the advantages of delta ribozymes include the known crystal structures of several variants of the genomic variant, and a similar structure of genomic and antigenomic ribozymes, as confirmed by numerous biochemical experiments [12-15].
A set of antibiotics that interact with the ribozymes has been well characterized in a number of studies [8, 16, 17]. They can basically be divided into two groups: nonspecific antibiotics, which inhibit the catalytic activity of many ribozymes; and specific antibiotics, which inhibit individual ribozymes. A member of the first group is neomycin B, an aminoglycoside antibiotic that strongly inhibits the catalytic activity of small ribozymes, such as hammerhead, hairpin, and delta of the genomic and antigenomic type, as well as large ribozymes, such as RNase P and ribozymes of group I intron . Similar properties have been demonstrated in the case of 5-epi-sisomicin, another aminoglycoside antibiotic that effectively inhibits the catalytic activity of the hairpin, delta and group I intron ribozymes. On the other hand, kanamycin acts very specifically, affecting the cleavage reaction only in the case of RNase P, and not inhibiting group I intron and delta ribozymes .
It should be noted that many small ligands, including antibiotics, may exist in the cell as complexes with available metal ions . Mg2+ and Ca2+ are the most common divalent metals present in blood plasma, and their estimated concentrations are ~ 1 and 3 mm, respectively . The concentration of transition metals, such as Cu2+, Zn2+, and Fe2+, has been calculated to be 10–15 μm. Particular attention has been paid to Cu2+, which is third in abundance after Fe3+ and Zn2+, and which is present in most tissues and fluids. It was found that Cu2+ ions are mainly bound to specific proteins such as blue copper proteins or enzymes, including the oxidase families and superoxide dismutase . Moreover, in many pathological conditions, the level of serum copper increases by as much as four-fold to five-fold of the normal level [20-22]. On the other hand, the concentration of antibiotics in the human organism depends on many factors, such as chemical properties, solubility, stability in body fluids, and ability to penetrate membranes, and, in some instances, such as vancomycin, it may be as high as 10 μm .
Despite substantial progress in the elucidation of antibiotic–RNA interactions , their exact mechanism is still being discussed. Also, the mechanism of their impact on the catalytic activity of ribozymes is not fully understood. To design antibiotics that specifically bind to a desired structure of the ribozyme, or more generally to the RNA structure, it is important to identify factors influencing the binding process. Recently, we have suggested that the way in which antibiotics are positioned on the ribozyme is of crucial importance for effective inhibition. Moreover, protonation of their functional groups or complexation with Cu2+ ions can dramatically change the impact of antibiotics on the catalysis of the delta ribozymes . In this article, we present the results of biochemical mapping of the binding sites of selected antibiotics and their complexes with Cu2+ to the antigenomic delta ribozyme. We also discuss possible mechanisms by which the studied antibiotics can impact on the catalytic activity of ribozyme variants.
Results and Discussion
Characterization of research models and experimental approaches
The aminoglycosides amikacin and neomycin B and the polypeptide antibiotic actinomycin D are able to coordinate Cu2+ ions and form stable complexes under physiological conditions, a fact that we have previously reported [26-28]. In this study, we mapped their binding sites and their complexes with Cu2+ to the trans-acting antigenomic delta ribozyme. It has been shown that these antibiotics and their complexes impact differently on the RNA catalysis reaction . Amikacin and amikacin–Cu2+ do not affect the reaction. Similarly, actinomycin D is inactive as an inhibitor, but actinomycin D–Cu2+ effectively inhibits the ribozyme. Both neomycin B and neomycin B–Cu2+ strongly inhibit delta ribozyme cleavage  (unpublished data of our laboratory).
To localize the antibiotic-binding sites within the ribozyme structure, four mapping methods were applied: Pb2+ -induced cleavage, nuclease S1 and RNase T1 digestion, and the selective 2′-hydroxyl acylation analyzed by primer extension (SHAPE) approach. The Pb2+ method is very useful not only for detecting differences in the Watson–Crick base pairing, but also for detecting more subtle differences in the RNA conformation resulting from changes in stacking or tertiary interactions or ligand binding . The Pb2+ method has been applied previously to analyze the folding of several ribozymes, aptamers, and fragments of larger viral and cellular RNAs [32-34]. Similar structural specificity is claimed to be characteristic of the SHAPE approach, which has recently been introduced for the structural analysis of RNA . This approach utilizes high-throughput selective acylation of the 2′-OH group of ribose by N-methylisatoic anhydride (NMIA). The resulting 2′-O-adducts are usually detected as stops to primer extension with 32P-end-labeled primers and PAGE electrophoresis, and quantified by whole-trace Gaussian integration. SHAPE reactivities therefore provide direct model-free information regarding the overall level of structure, or architecture, for any RNA [36, 37]. Regions with low reactivities, of < 0.25 (see 'Experimental procedures'), indicate domains with a significantly base-paired secondary RNA structure, whereas median SHAPE reactivities, of ≥ 0.5, indicate largely unstructured regions. SHAPE was previously used to study local nucleotide-flexible regions of RNA polynucleotide chains of many RNAs, including those of HIV and p53 mRNA [38, 39].
In our studies, the trans-acting version of antigenomic delta ribozyme was applied as a research model (Fig. 1). A great advantage of this model is its favorable characterization in many biochemical experiments, including those carried out in our laboratory [40-43]. The ribozyme has a compact structure that highly discriminates between low molecular mas compounds, and even isomers of the same compounds . Two different variants of a 20-nucleotide oligonucleotide substrate were used: a cleavable all-RNA substrate that contained a cytidine at the cleavage site; and an uncleavable substrate that contained deoxycytidine (dC) in this position. When the first substrate was used with the trans-acting ribozyme, in the presence of catalytic Mg2+ ions, we obtained the 3′-product of ribozyme cleavage reaction – the cleaved form of the ribozyme. Such a ribozyme model mimics the postcleavage ribozyme conformation. However, when the oligonucleotide substrate containing dC at the cleavage site was used, the ribozyme remained in an uncleaved form, reflecting the situation before the catalytic cleavage event (Fig. 1).
Probing the structure of the cleaved and uncleaved forms of the delta ribozyme
Probing of the structure of the delta ribozymes was performed with Pb2+ -induced cleavage, nuclease digestion, and SHAPE reactivities, with the previously described reaction conditions [32-35, 39] in the presence of 5 mm MgCl2. Although the concentration of Mg2+ was higher than 1 mm, which is the level in the cell , such conditions better stabilize the spatial RNA folds, and have often been used in RNA structural and functional studies .
The application of the all-RNA oligonucleotide substrate in the presence of Mg2+ resulted in the formation of the final 3′-product of the cleavage reaction. In this cleaved ribozyme form, Pb2+ -induced cleavages were identified in the following regions: intensive in the J4/2 junction, as well as at the 3′-end of the ribozyme; medium inside the P3 region, and L3 and L4 loops; and weak in the P2 stem (Figs 1 and S1). These results are fully consistent with those described previously for the 3′-product of the genomic ribozyme . On analysis of the nuclease S1 and RNase T1 digestion patterns, two regions were accessible to these nucleases. Moderate digestion was observed in the P3/L3 region, and very weak digestion in the L4 loop (Figs 1, S2 and S3).
SHAPE analysis of the cleaved ribozyme form identified NMIA acetylation sites at the following nucleotides: A14 and A20 in the P2/P3 region, and C27, G28, C29, G30, G31 and U32 in the L3/P3 loop–stem region (Figs 1 and S4). Relatively lower modification extents were also observed at G35 and A36 in the P1 stem, as well as at G40, G41 and G42 in the P1.1 pseudoknot. In the L4 loop, a high degree of NMIA modification occurred, involving the whole loop region. In the J4/2 region, the highest reactivity was detected at G75. Moreover, G80, U77 and A78 were modified to a lesser extent. Interestingly, C76 was not accessible to the NMIA reagent. Comparison of the SHAPE patterns for the uncleaved and cleaved ribozyme forms revealed a significant difference in the J4/2 region (Figs 1, S4 and S5). The 2′-OH group of U77 in the uncleaved ribozyme was fully modified, but, in the cleaved form, this residue was modified only partially (20–30%). Also, the 2′-position of the ribose moiety of A78 showed greater accessibility to the reagent in the case of uncleaved ribozyme containing dC substrate. The modification level of C79 remained the same in both ribozyme forms.
A previous crystallographic analysis of the cleaved genomic variant of delta ribozyme (3′-product) and uncleaved ribozyme containing the C75U mutation has revealed several substantial differences [12, 13]. First, the metal ion is absent in the 3′-product structure, but, in the uncleaved ribozyme, the ion is bound to the C75 base and G1 phosphate in the catalytic center. Additionally, the release of the 5′-part of the substrate strand after cleavage moves the C75 deeper into the catalytic site, and this residue forms a hydrogen bond with the 5′-O′ of the leaving group. This suggests that the J4/2 region is less available for contact with other ligands. However, the recently determined crystal structure of the genomic delta ribozyme with a higher (1.9 Å) resolution does not confirm the previous observations . The determined structures of the uncleaved ribozyme and of the cleaved 3′-product are very similar. These discrepancies may be caused by the use of two distinct catalytically inactive ribozyme models: a C75U mutant ribozyme, and a ribozyme containing a substrate strand with dC modification at the cleavage site [12, 13]. However, our SHAPE data showed some changes just in the J4/2 region, which became more accessible to NMIA in the antigenomic delta ribozyme associated with the dC substrate strand. This observation is in line with the previously suggested conformational changes for the uncleaved ribozyme with the C75U mutation .
The impact of amikacin and amikacin–Cu2+ on the delta ribozyme structure
In the first step, we analyzed the impact of Cu2+ ions on the structure of the cleaved form of the antigenomic delta ribozyme, as a control reaction. The presence of 50 μm Cu2+ did not affect the Pb2+ -induced cleavage pattern (Fig. S1), whereas the nuclease S1 digestion pattern was slightly modified. In the case of SHAPE analysis, only minor differences in the L3 and J4/2 regions were found (Fig. S4). Therefore, we concluded that Cu2+ had almost no impact on the ribozyme structure, which was supported by our previous experiments, and that there was no influence of this metal ion on the ribozyme catalytic activity [29, 43].
We have previously demonstrated that neither amikacin nor amikacin–Cu2+ change the activity of antigenomic delta ribozyme . Not surprisingly, identical Pb2+ -induced cleavage patterns for the cleaved ribozyme with no antibiotic and with amikacin or amikacin–Cu2+ were obtained (Figs 2 and S1). Also, the presence of amikacin did not change the RNase T1 and nuclease S1 digestion patterns (Figs 2, S2 and S3). Furthermore, as compared with the cleaved ribozyme form with no antibiotic, the SHAPE pattern observed in the presence of amikacin was slightly altered in the P3 region (C27, G28, G30, and G31), and these modification sites were suppressed (Figs 2 and S4). In the J4/2 region, the modification of G75 was reduced, but that of U77 was higher. Moreover, there was almost no difference in the SHAPE patterns in the presence of amikacin and in the presence of amikacin–Cu2+ . Regarding the other ribozyme model, i.e. the uncleaved form, a higher extent of A78 and A79 modification was observed in the presence of amikacin (Figs 3 and S5). However, there was no difference in the SHAPE patterns in the presence of amikacin and in the presence of amikacin–Cu2+ . This observation suggests that both of them interact with the ribozyme in a similar manner.
Interactions of neomycin and of neomycin–Cu2+ with the delta ribozyme
It has been shown that neomycin B, both as a free antibiotic and as a complex, strongly inhibits the catalytic activity of antigenomic delta ribozyme  (our unpublished data). In the presence of both, stronger Pb2+ -induced cleavages in a cleaved ribozyme form were observed at G81 in the J4/2 junction than with no antibiotic added (Figs 2 and S1). On the other hand, C76 in this region was protected from Pb2+ -induced cleavage. The binding of neomycin–Cu2+ to the ribozyme structure did not significantly affect the nuclease S1 digestion pattern. In addition, RNase T1 also digested the cleaved form of the ribozyme in a similar way (Figs 2 and S2). The only stronger digestion site in the presence of neomycin–Cu2+ was located in the L3 loop, and the weaker ones were located in the L4 loop and J4/2 junction.
A comparison of the SHAPE patterns of the cleaved ribozyme with neomycin B or neomycin B–Cu2+ added and of the ribozyme with no antibiotic added revealed the most evident differences in the P1.1 pseudoknot and J4/2 regions (Figs 2 and S4). Strikingly, G40 was fully modified. Interestingly, such a phenomenon was observed in both the cleaved and uncleaved ribozyme forms. Also, the 2′-OH groups of A78, A79 and G80 were fully accessible to NMIA modification. In addition, nucleotides of the L3 and L4 loops were modified in the presence of neomycin B or neomycin B–Cu2+, as observed in the ribozyme without these compounds. Only a slight increase in the ribose U77 modification level was observed in the uncleaved ribozyme (Figs 3 and S5). Thus, neomycin B and neomycin B–Cu2+ interacted in a similar way, regardless of whether a cleaved or uncleaved form of the antigenomic HDV ribozyme was tested.
Actinomycin D and actinomycin D–Cu2+ impact differently on the delta ribozyme structure
Actinomycin D, a polypeptide antibiotic, has a molecular mass twice as high as that of the examined aminoglycosides. Therefore, a more prominent impact of actinomycin and actinomycin–Cu2+ on the delta ribozyme could be expected. Indeed, a specific SHAPE pattern of the uncleaved ribozyme in the J4/2 region was found (Figs 3 and S5). The 2′-OH group of the C76 ribose was fully modified, and this modification was an exception among all of the investigated antibiotics. However, the modification of A78 and A79 that was seen without the antibiotic completely disappeared when actinomycin D or actinomycin D–Cu2+ was added. Also, the SHAPE analysis of the cleaved ribozyme showed some alteration of the structure after the addition of actinomycin D and actinomycin D–Cu2+ (Figs 2 and S4). A lack of NMIA modifications at the 3′-end of the L3 loop and a changed modification pattern of the J4/2 region were observed. The major difference concerned a full modification of the 2′-OH group of the U77 ribose as compared with the ribozyme with no antibiotic added. The binding of the Cu2+ ion(s) to actinomycin D and its consequent complex formation did not significantly change the SHAPE pattern.
Similar Pb2+ -induced cleavage patterns were obtained for the cleaved form of the ribozyme in the presence of actinomycin or actinomycin D–Cu2+ (Figs 2 and S1). Only in the case of actinomycin D–Cu2+ were additional weak nuclease S1 digestion sites at G28, C29 and G30 detected. Moreover, new RNase T1 digestions were generated at G70, G71, G74 and G75 in the P4 stem (Figs 2 and S2). Additionally, weak digestion occurred at G85 in the P2 stem.
Actinomycin D preferentially binds to dsDNA, whereas the affinity of this antibiotic for single-stranded DNA and RNA molecules is significantly lower . The phenoxazone ring of the molecule intercalates into the DNA helix preferentially between the G–C base pairs, and the cyclic pentapeptide moiety is located in the major groove of DNA. Such an actinomycin D binding mode strongly affects the DNA-dependent RNA polymerase activities in the cell . Possibly, the additional positive charge of actinomycin D–Cu2+ increases its binding affinity inside the P4 and P2 stems. The P4 and P2 stems in the crystal structure of genomic delta ribozyme form an A-RNA-type regular helix [12-14]. However, the P4 double-stranded region of the antigenomic delta ribozyme possesses a nonstandard G–G base pair. This could be the reason for the greater accessibility of this region than of the base-paired P2 region to actinomycin–Cu2+.
Interestingly, in the cleaved form of the ribozyme, digestions in the P4 and P2 stems were detected only when RNase T1 was used as a probe. These regions were resistant to digestion with nuclease S1. These differences between RNase T1 and nuclease S1 could be explained by the different sizes of these enzymes. Nuclease S1 is a bulky enzyme (molecular mass of > 30 kDa) in comparision with RNase T1 (molecular mass of 11.2 kDa), and some RNA regions accessible to RNase T1 might be protected from the second enzyme, mainly stretches involved in the formation of a higher-order RNA structure.
The properties of binding of the studied antibiotics and their complexes to the delta ribozyme
In order to determine the binding affinities of the examined antibiotics for antigenomic delta ribozyme, changes in the intensity of the Pb2+ -induced cleavages in the presence of increasing concentrations of antibiotics were used. In this way, the corresponding dissociation constants (Kd values) for the antibiotics and their complexes with Cu2+ ions were estimated. The cleavage reactions were carried out with 1 mm Pb2+ and 10–300 μm antibiotic, respectively, and a 50% reduction in the efficiency of J4/2 junction cleavages was taken as the value of Kd. Among the analyzed antibiotics, the lowest Kd value was determined for neomycin B (26.0 ± 3.1 μm). For neomycin B–Cu2+, the Kd decreased over three-fold to 8.5 ± 1.2 μm (Fig. 4). The Kd values for actinomycin D and actinomycin D–Cu2+ were 75.5 ± 9.0 μm and 29.2 ± μm, respectively. The Kd value for amikacin was not determined, because amikacin did not inhibit the Pb2+-induced cleavages, whereas amikacin–Cu2+, with a Kd value of 136 ± 40 μm, did so weakly. Thus, the Kd values characterizing the binding of antibiotics to the antigenomic delta ribozyme can be arranged in the following order: neomycin B–Cu2+, actinomycin D–Cu2+, neomycin B, actinomycin D, amikacin–Cu2+, and amikacin. The same order has been obtained previously by examining the inhibition of the catalytic activity of the delta ribozyme in the presence of these compounds. We have just demonstrated that neomycin B–Cu2+ has the strongest inhibitory properties, and amikacin has the weakest ones .
Molecular modeling of the genomic delta ribozyme showed a high electron density of the ribozyme catalytic center [46, 47]. In addition, the Mg2+ present in this region, located close to the catalytic C75 residue, withdrew electrons, increasing its negative charge . Presumably, the binding of neomycin B–Cu2+ to the antigenomic delta ribozyme is more effective than that of free antibiotic, because the bound metal ion increases the affinity of the complex for the catalytic center with a high electron density. In the case of actinomycin D–Cu2+, the binding site is different, and the inhibition effect is likely to be a result of the intercalation into the P4 double-stranded region. An additional positively charged Cu2+ may enhance the electrostatic interaction between the complexes and the negatively charged J4/2 region of the delta ribozyme. This suggestion is strongly supported by finding that the replacement of the positively charged amino group in the kanamycin B molecule on the hydroxyl group supresses inhibitory properties of this antibiotic in the self-splicing of group I intron and the formation of the Rev=RRE complex .
Computer modeling of antibiotic binding to the delta ribozyme
To determine the locations of the examined antibiotics inside the structure of an antigenomic delta ribozyme, the molecular modeling technique was applied. The genomic and antigenomic types of the ribozyme have almost identical structures in solution, as confirmed by numerous probes . Therefore, it seemed reasonable to use the previously resolved crystal structure of the genomic variant in our modeling experiments [12-14].
Amikacin is located in the neighborhood of the J4/2 junction (Fig. 5A). Neomycin B occupies a similar region (Fig. 5B). However, its location is shifted, and the antibiotic interferes with the nucleotide at position 1. Neomycin B is also positioned closer to the catalytic C76. Thus, both aminoglycosides seem to be similarly located in the pocket formed by the J4/2 junction and P2 and P3 stems. A subtle change in the location of these antibiotics inside the delta ribozyme structure is important with regard to the function.
Actinomycin D is bound to the delta ribozyme in a different way from aminoglycosides (Fig. 5C). Computer modeling shows two binding sites, and this antibiotic intercalates mainly into the P2 and P4 stems. Although this binding site seems to be far away from the catalytic pocket, i.e. the nucleotide at position 1 and the catalytic cytosine, actinomycin D–Cu2+ strongly affects the catalytic activity of the delta ribozyme. It seems, therefore, that the binding of actinomycin D–Cu2+ induces conformational changes that make the ribozyme inactive.
The binding sites of selected antibiotics and their complexes with Cu2+ ions in two forms of the antigenomic delta ribozyme, before and after cleavage, were mapped. In the cleaved ribozyme, the SHAPE pattern observed in the presence of amikacin was slightly altered in the P3 and J4/2 regions. Moreover, there was almost no difference in the modification patterns in the presence of amikacin and amikacin–Cu2+ . Also, in the uncleaved ribozyme, changes in the J4/2 region were observed in the presence of amikacin.
Both neomycin B and neomycin B–Cu2+ strongly inhibit the catalytic activity of the antigenomic delta ribozyme  (our unpublished data). Upon addition of neomycin B or neomycin B–Cu2+ to the cleaved ribozyme, the most evident differences in the SHAPE patterns were detected in the P1.1 pseudoknot and J4/2 regions. Moreover, neomycin B and neomycin B–Cu2+ interacted with the cleaved and uncleaved form of the ribozyme in a similar manner.
Actinomycin D and actinomycin D–Cu2+ induced specific changes in the SHAPE pattern of the uncleaved ribozyme in the J4/2 region. Only in the case of actinomycin D–Cu2+ were additional weak nuclease S1 digestion sites detected in the L3 loop. Moreover, new RNase T1 digestions were generated in the P4 stem, and also in the P2 stem. Possibly, an additional positive charge of actinomycin D–Cu2+ increased its binding affinity inside the P4 and P2 stems.
The Kd values characterizing the binding of the studied antibiotics or their complexes with Cu2+ to the ribozyme fall within the range of 8–136 μm. The corresponding values can be arranged in the following order: neomycin B–Cu2+, actinomycin D–Cu2+, neomycin B, actinomycin D, amikacin–Cu2+, and amikacin. This order conforms to the one obtained previously by examining the inhibition of the ribozyme catalytic activity in the presence of these compounds .
The molecular modeling technique showed that the aminoglycoside antibiotics amikacin and neomycin B seem to be located similarly, in the pocket formed by the J4/2 junction and P2 and P3 stems. However, the location of neomycin B was shifted and the antibiotic interfered with the nucleotide at position 1, and was also positioned closer to the catalytic C76. On the other hand, actinomycin D was bound to the ribozyme in a different way, intercalating mainly into the P2 stem. Although its binding site was far away from the nucleotide at position 1 and the catalytic cytosine, actinomycin D–Cu2+ strongly affected the catalytic activity of the delta ribozyme.
Previously, several factors have been suggested that affect the delta ribozyme cleavage reaction by antibiotics, such as the displacement of the catalytic Mg2+, an impact on the catalytic cleavage reaction trajectory, and the protonation state of antibiotics or their complexation with Cu2+ ions . In this study, we showed that some antibiotics and their Cu2+ complexes inhibiting the delta ribozyme could bind to its different regions or bind in different ways. However, their high binding affinity for the ribozyme seemed to be of primary importance for effective inhibition.
All chemicals were from Serva (Heidelberg, Germany) or Fluka (Buchs, Switzerland). Antibiotics were purchased from Sigma-Aldrich (St Louis, MI, USA). T7 RNA polymerase, DNA Taq polymerase, T4 polynucleotide kinase, nuclease S1, NTPs and dNTPs were from MBI Fermentas (Vilnius, Lithuania). RNase T1 was from Sigma-Aldrich (St Louis, MI, USA). Finally, [32P]ATP[γP] (5000 Ci/mmol) was from Hartmann Analytic (Braunschweig, Germany). RNA oligomers R20 and R20dC were synthesized by Thermo Fisher Scientific (Waltham, USA).
DNA template construct
The dsDNA templates for the in vitro transcription of the trans-acting antigenomic delta ribozyme was prepared as follows. Two DNA oligomers were synthesized: A, 5′-GAAAAGTGGCTCTCCCTTAGCCAT-CCGAGTGCTCGGATGCCCAGGTCGGACCGCGAGGAGGTGGAGATGCCC-3′; and B, 5′-TAATA-CGACTCACTATAGGGCATCTCCACC-3′ (in both oligomers, the complementary sequences are underlined, and letters in italics mark the T7 RNA polymerase promoter). Equimolar amounts of pairs of oligomers (A–B) were annealed. The reaction mixture contained 1.5 μm A–B oligomers, 10 mm Tris/HCl (pH 7.0), 2 mm MgCl2, 150 mm KCl, 0.1% Triton X-100, 200 μm each dNTP, and 60 units/mL DNA Taq polymerase. The reaction was performed on a Biometra UNO II thermocycler for six cycles of 30 s at 93 °C, 30 s at 55 °C, and 1 min at 72 °C. The dsDNA was extracted with phenol/chloroform (1 : 1), and precipitated with ethanol at − 20 °C overnight. The dsDNA template was recovered by centrifugation (15 000 g, 20 min), dissolved in TE buffer, and used in transcription reactions.
Trans-acting ribozyme synthesis
The in vitro transcription reaction of ribozyme was performed as follows, with: 0.5 μm dsDNA template, 40 mm Tris/HCl (pH 8.0), 10 mm MgCl2, 0.001% Triton X-100, 2 mm spermidine, 5 mm dithiothreitol, 1 mm each NTP, 750 U/mL RNase inhibitor, and 2000 U/mL T7 RNA polymerase. The transcription mixture was incubated at 37 °C for 4 h. Subsequently, the transcription products were precipitated and purified by electrophoresis on 8% PAGE gels containing 1 mm EDTA and 7 m urea. The band corresponding to the ribozyme was localized by UV shadowing and cut out, and RNA was eluted from the gel with 0.3 m sodium acetate (pH 5.2) and 1 mm EDTA, ethanol-precipitated, and dissolved in sterile water containing 0.1 mm EDTA.
The trans-acting antigenomic delta ribozyme was prepared by mixing the 5′-32P-labeled R20 or R20dC substrate (~ 50 000 c.p.m., 0.1 pmol) supplemented with 100 pmol of unlabeled appropriate substrate with 100 pmol of delta ribozyme in the reaction buffer (50 mm Tris/HCl, pH 7.5, 0.1 mm EDTA, 10 mm NaCl, 5 mm MgCl2) to obtain a 1 : 1 ribozyme/substrate ratio. The mixture was subjected to a denaturation–renaturation procedure in a buffer containing 50 mm Tris/HCl (pH 7.5) and 0.1 mm EDTA, incubated for 2 min at 100 °C, chilled on ice for 10 min, and finally incubated for 10 min at 37 °C. Then, the antibiotics or their complexes with Cu2+ were added to a final concentration of 50 μm, and the mixture was incubated for an additional 10 min at 37 °C and used in probing experiments.
Pb2+ -induced cleavage
The 32P-labeled ribozyme samples were supplemented with tRNA carrier (Boehringer) to a final RNA concentration of 8 mm in a final buffer consisting of 10 mm Tris/HCl (pH 7.4), 40 mm NaCl, and 5 mm MgCl2. Subsequently, Pb(OAc)2 solution was added to 0.5 mm, and the reactions proceeded at 37 °C for 10 min. The reactions were terminated by mixing aliquots with 8 m urea/dyes/20 mm EDTA solution, and samples were loaded on 12% PAGE, 0.75% bis-acrylamide and 7 m urea gels. Electrophoresis was performed at 2000 V for 2–3 h, and this was followed by autoradiography at − 70 °C with an intensifying screen or phosphorimaging with a Typhoon 8600 analyzer (Molecular Dynamics).
For Kd determination, selected antibiotics and their complexes with Cu2+ were added to appropriate concentrations, and cleavage reactions were initiated by adding Pb(OAc)2 to 0.5 mm; mixtures were incubated for 30 min.
The 32P-labeled ribozymes were renatured and supplemented with tRNA carrier as described above. Limited digestions with nuclease S1 and RNase T1 were carried out in the following buffer: 10 mm Tris/HCl (pH 7.2), 40 mm NaCl, and 5 mm MgCl2. In reactions with nuclease S1, ZnCl2 was also present at 1 mm. Reactions were performed at 37 °C for 30 min, with 300 U/mL nuclease S1 and 50 U/mL RNase T1. The reactions were terminated by adding 8 m urea/dyes/20 mm EDTA solution and freezing samples on dry ice. The reaction products were analyzed by PAGE, and visualized by autoradiography or phosphorimaging.
The reaction mixture containing 20 pmol of RNA (0.4 μm final concentration) in renaturation buffer (10 mm Tris, pH 8.0, 100 mm KCl, 0.1 mm EDTA), and, in a final volume of 20 μL, was heated at 90 °C for 3 min and slowly cooled (0.1 °C/s) to 4 °C. The folding buffer (40 mm Tris, pH 8.0, 5 mm MgCl2, 130 mm KCl, 0.1 mm EDTA) was then added, and water was added to give a final volume of 146 μL. The sample was incubated at 37 °C for 10 min, and separated into two parts for reactions. The RNA solution was mixed with 3.4–7.3 μL of 180 mm NMIA (Invitrogen) in dimethylsulfoxide (8–16 mm final concentration). The control reaction mixture contained dimethylsulfoxide without NMIA. Both reaction mixtures were incubated for 50 min at 37 °C (50 °C, or 60 °C in temperature melting experiments), and RNA was precipitated with 0.3 m sodium acetate (pH 5.2), 1 μL of glycogen (20 mg/mL), and three volumes of ethanol. After centrifugation (15 000 g, 20 min), RNA was resuspended in 10 mm Tris (pH 8.0) and 0.1 mm EDTA. Sites of modification were analyzed by primer extension reaction in standard conditions .
Briefly, modified RNA (2 pmol) was mixed with 5′-end-32P-labeled DNA primer (2 pmol) and water in a final volume of 12 μL. The reaction mixture was incubated at 95 °C for 1 min, at 60 °C for 5 min, and at 52 °C for 2 min, and 8 μL of reverse transcription mix was then added; the final reaction mixture comprised 50 mm Tris/HCl (pH 8.3), 75 mm KCl, 3 mm MgCl2, 6 mm dithiothreitol, 500 μm each dNTP, and 50 units of reverse transcriptase (Superscript III; Invitrogen, USA). The reactions were performed at 52 °C for 10 min. cDNA samples were then treated with 1 μL of 4 m NaOH, placed on ice, heated at 95 °C for 5 min, placed on ice, and treated with 160 mm Trizma. Dideoxy sequencing markers were generated in the same way, with unmodified RNA and thymidine or adenosine dideoxy terminating nucleotides (0.2 mm). In the next step, cDNAs were precipitated and resuspended in water. Following addition of 8 m urea/dyes/20 mm EDTA solution, the samples were denatured for 2 min at 95 °C and loaded on 8% polyacrylamide/0.75% bis-acrylamide/7 m urea gels. After electrophoresis, the gel was analysed by phosphoroimaging with a Typhoon 8600 apparatus (Molecular Dynamics).
The model of the structure of the antigenomic HDV ribozyme was built with the Accelrys discovery studio 3.5 suite of programs (Accelrys Software, San Diego, CA, USA) by homology modeling based on the crystal structure of genomic substrate (Protein Data Bank ID 1CX0), according to the secondary structure. The fragment of the P2 region of the antigenomic variant covering G12–A14 and thew corresponding G86–C87 was truncated to maintain the structural homology of the model to the crystal structures of genomic-type ribozyme. 3′-Ends and 5′-ends were extended by the short sequences U(–2)C(–1) and C(90)U(91)U(92), corresponding to unpaired regions of the substrate form of antigenomic HDV ribozyme.
The modeled structure was subjected to minimization in vacuum followed by minimization in the implicit solvent model. Harmonic restraints with a 10.0 kcal·mol−1 Å−2 force constant were applied to the RNA heavy atoms. Then, the restraints were reduced to 1.0 kcal·mol−1 Å−2, and the minimization was continued in the implicit solvent model. After this, two rounds of 10-ps molecular dynamics simulations at 297 K and 310 K were applied, and, finally, the whole system was energy-minimized. Simulations were performed with charmm35 and charmm27  field for nucleic acids  in a continuum solvent described by a generalized Born model, based on GBSW methods with default parameters as implemented in charmm. For GBSW simulations, the Langevin dynamics were determined with the friction coefficient 10 ps−1.
Molecular docking of the ligands was performed with haddock in the regions defined by the experimental results of structural mapping and the analysis of the electrostatic potential of the surface of the RNA molecule [49, 50].
This work was supported by Wroclaw Research Center EIT+ under the project ‘Biotechnologies and Advanced Medical Technologies – BioMed’ (POIG 01.01.02-02-003/08-00), financed by the European Regional Development Fund (Operational Programme Innovative Economy, 1.1.2).