Oligonucleotides

For nucleic acid targets, buffer systems which maintain pH 6-8 with relatively high concentrations of ammonium acetate (50-250 mM) and organic co-solvents (10-50%) are generally effective [9-11]. The concentration of buffer/salt must be high enough to allow proper base pairing of the strand(s) and to ensure that the melting transition temperature is well above ambient. The buffer used most widely in our laboratory for screening relatively small RNA motifs (18-50mers) against compound collections or natural product fractions is comprised of 150

mM NH4OAc, 33% isopropyl alcohol, and no more than 1.5% DMSO [12-14]. A buffer system with the appropriate pH and ionic strength is equally important for native folding in nucleic acids as for proteins as binding to a denatured oligonucleotide, much like binding to a denatured protein, will not yield information representative of the same molecules in vivo. Similarly, denatured oligonucleotides generally exhibit a broad distribution of relatively high charge states, while nondenatured nucleic constructs most often produce mass spectra which are dominated by one or two distinct charge states at relatively high m/z. Denaturing solution conditions are ideal for obtaining accurate mass measurements of PCR products where it is desirable to thoroughly denature the duplex and obtain mass measurements for the forward and reverse strands of the amplicon independently, as opposed to a single mass measurement on a DNA duplex [15, 16]. Independent mass measurements of the forward and reverse strands allow one to unambiguously determine the base composition of a given amplicon [17, 18] (or mixture of amplicons) which has significant relevance in ESI-MS based microbial identification [16, 19], microbial forensics [20], and human forensic strategies [21]. Alternatively, nondenaturing solution conditions and gentle desol-vation are important for maintaining natively folded macromolecular targets with which to form noncovalent complexes. The appearance and information content of the spectra can differ drastically depending on the solution conditions employed. For example, Fig. 10.2 illustrates mass spectra obtained from solutions containing the 27mer RNA construct which represents the 16S ribosomal A-site, an important target for bacterial drug discovery [22]. The spectrum in Fig. 10.2a was acquired from a solution comprised of 33% isopropyl alcohol with 25 mM piperidine/imidazole with an approximate pH of 8.5. The spectrum in Fig. 10.2b was acquired from a solution containing the same concentration of the 27mer RNA construct and the same concentration of isopropyl alcohol, but with 150 mM NH4OAc with a pH of 7.0. The solution conditions employed in Fig. 10.2b allow the interrogation of specific interactions between the 27mer target and small molecule ligands, as the solution allows the target to maintain a native conformation amenable to the formation of specific noncovalent complexes (see below).

A number of larger RNA motifs, which have more complex high-order structures in vivo, rely on divalent metal ions (e.g. Mg2+) to adopt correct secondary and tertiary structures. One such construct that represents a potentially valuable antibiotic target is a 58-nucleotide domain of the 23S ribosomal subunit to which the L11 protein binds. This structural motif is highly conserved among prokar-yotes and participates in GTP hydrolysis reactions involving several ribosomal factors. The crystal structure of the 58-nucleotide construct bound to the L11 protein was obtained by Draper and coworkers [23] and provides valuable insight into the functional operation of this part of the ribosome. A naturally occurring antibiotic, thiostrepton (MW = 1663 Da) is known to bind to this motif and to inhibit key interactions at the GTPase center. While thiostrepton has poor drug properties, owing to low solubility, poor oral bioavailability, and synthetically daunting

Fig. 10.2 Effect of buffer composition on the ESI-FTICR-MS spectrum of a 27mer RNA construct representing the 16S A-site. The spectrum in (a) was acquired from a solution comprised of 33% isopropyl alcohol and 25 mM piperidine/imidazole with an approximate pH of 8.5. The spectrum in (b) was acquired from a solution containing the same concentration of the 27mer RNA construct and the same concentration of isopropyl alcohol, but with 150 mM NH4OAc at pH 7. These buffer conditions facilitate folding of the construct into a native structure that can serve as a drug binding substrate.

Fig. 10.2 Effect of buffer composition on the ESI-FTICR-MS spectrum of a 27mer RNA construct representing the 16S A-site. The spectrum in (a) was acquired from a solution comprised of 33% isopropyl alcohol and 25 mM piperidine/imidazole with an approximate pH of 8.5. The spectrum in (b) was acquired from a solution containing the same concentration of the 27mer RNA construct and the same concentration of isopropyl alcohol, but with 150 mM NH4OAc at pH 7. These buffer conditions facilitate folding of the construct into a native structure that can serve as a drug binding substrate.

multi-ringed structure, it does serve as a proof-of-principle for what could be a very significant strategy for new classes of antimicrobial agents [24-26]. A mutant of this motif found in thermophiles (A1061) is particularly stable and thus an ideal substrate for ESI-MS based affinity screening. The A1061-thiostrepton complex was used to determine appropriate solution and interface conditions for the system.

This construct was initially evaluated with the same buffer system used for smaller RNA motifs and it was found that the ammonium acetate/isopropyl alcohol buffer provided only partial complexation of the thiostrepton and a relatively wide charge state distribution, indicative of a partially denatured conformation in solution. As Mg2+ was previously implicated as a key to proper folding in vivo [27, 28], a study was undertaken to characterize the magnesium-dependent folding of the A1061 construct in solution as measured by thiostrepton binding. Figure 10.3 shows the (M-8H)8— charge state of the A1061 construct, and the resulting A1061-thiostrepton complex that results when Mg2+ is added. In the absence of Mg2+, only trace levels of the complex is observed and the spectrum is dominated by the 8— charge state of the unbound RNA. At increasing concentrations

Fig. 10.3 (M-8H)8" charge state of the 58mer A1061 RNA construct (see text) in the presence of an excess of thiostrepton. The buffer solution contains 20% MeOH and 25 mM NH4OAc. In the absence of adequate Mg2+ ion, the A1061 construct is denatured and does not bind thiostrepton. At increasing Mg2+ concentration, the A1061 adopts a native conformation and binds a stoichiometric amount of thiostrepton. The peaks labeled "N+1" refer to a synthetic impurity arising from a nontemplated nucleotide. The even number of Mg2+ ions bound to the RNA and complex is indicated by "#Mg2+: 0 2 4 6'' above the corresponding lines in the spectrum.

Fig. 10.3 (M-8H)8" charge state of the 58mer A1061 RNA construct (see text) in the presence of an excess of thiostrepton. The buffer solution contains 20% MeOH and 25 mM NH4OAc. In the absence of adequate Mg2+ ion, the A1061 construct is denatured and does not bind thiostrepton. At increasing Mg2+ concentration, the A1061 adopts a native conformation and binds a stoichiometric amount of thiostrepton. The peaks labeled "N+1" refer to a synthetic impurity arising from a nontemplated nucleotide. The even number of Mg2+ ions bound to the RNA and complex is indicated by "#Mg2+: 0 2 4 6'' above the corresponding lines in the spectrum.

of Mg2+, the abundance of the complex increases as does the amount and extent of Mg2+ adducts. These Mg2+ ions, while clearly assisting in the proper folding of the RNA construct have unwanted side-effects in the mass spectrum in the form of adducts. At 50 mM Mg2+, a significant portion of the A1061 is unbound while a nearly equal portion is complexed with thiostrepton. Interestingly, the signal from the unbound A1061 is dominated by the unadducted species, while the complex is dominated by the singly and doubly adducted species. Furthermore, with 200 mM Mg2+, the majority of the A1061 is in the form of the A1061-thiostrepton complex and the signal is dominated by species containing five Mg2+ ions. While such solution conditions allow the detection of the complex, the relatively complex spectra and multiply adducted nature of the complexes limit the utility for screening applications in which multiple compounds with unknown binding properties are to be screened simultaneously.

In contrast, Draper and coworkers have shown that organic solvent such as methanol can actually help larger RNA constructs fold properly under salt conditions which would otherwise yield incompletely or improperly folded RNA con-

Fig. 10.4 A1061 folding in the presence of MeOH. In a solution containing 25 mM NH4OAc and a molar excess of thiostrepton, the A1061-thiostrepton complex is not observed at significant abundance with less than 25% MeOH. The spectrum acquired from a solution containing 50% MeOH is dominated by the A1061-thiostrepton complex consistent with the properly folded conformation. The peaks labeled "N+1" refer to a synthetic impurity arising from a nontemplated nucleotide.

Fig. 10.4 A1061 folding in the presence of MeOH. In a solution containing 25 mM NH4OAc and a molar excess of thiostrepton, the A1061-thiostrepton complex is not observed at significant abundance with less than 25% MeOH. The spectrum acquired from a solution containing 50% MeOH is dominated by the A1061-thiostrepton complex consistent with the properly folded conformation. The peaks labeled "N+1" refer to a synthetic impurity arising from a nontemplated nucleotide.

structs [29]. This is an important observation as solution conditions in which nonvolatile salts are employed to induce proper RNA folding (e.g. Mg2+) yield relatively complex mass spectra with poor signal to noise, as the peaks which represent the complexes of interest are spread over multiple states of adduction. As organic solvents are directly compatible with electrospray ionization and are completely removed during desolvation, an alternative buffer formulation lacking divalent metal cations but containing higher proportions of methanol was evaluated. Figures 10.4, 10.5 illustrate the effect of increasing methanol concentration on the A1061-thiostrepton complex in the absence of Mg2+. In an aqueous solution containing 6 mM A1061 with a slight stoichiometric excess of thiostrepton and 25 mM NH4OAc, the complex is not detected above the chemical noise background. When 20% methanol is added, a relatively weak signal consistent with the complex is observed, while a solution containing 50% methanol produces a spectrum that is dominated by the A1061-thiostrepton complex. The titration profile in Fig. 10.5 suggests a relatively sharp transition in the A1061 structure between 30% and 40% methanol. Note also that, other than a synthetic impurity related to an additional nontemplated nucleotide (peak labeled N+1), the spectrum acquired with 50% methanol is relatively clean and readily interpretable -a situation much more amenable to screening of compounds with unknown binding properties.

Was this article helpful?

0 0

Post a comment