2-Deoxystreptamine (2-DOS) aminoglycoside antibiotics exert their antimicrobial activities by targeting the decoding region A site of the rRNA and inhibiting protein synthesis. A prokaryotic specificity of action is critical to therapeutic utility of 2-DOS aminoglycosides as antibiotics. Here, isothermal titration calorimetry (ITC) and fluorescence studies are presented that provide insight into the molecular basis for this prokaryotic specificity of action. Specifically, the rRNA binding properties of the 2-DOS aminoglycosides paromomycin and G418 (geneticin) are compared, using both human and Escherichia coli rRNA A site model oligonucleotides as drug targets. Paromomycin and G418 differ with respect to their specificities of action, with only paromomycin exhibiting a specificity for prokaryotic versus human ribosomes. G418 binds to both the human and E.coli rRNA A sites with a markedly lower affinity than paromomycin, with the affinities of both drugs for the human rRNA A site being lower than those they exhibit for the E.coli rRNA A site. Paromomycin induces the destacking of the base at position 1492 (by E.coli numbering) upon binding to the E.coli rRNA A site, but not the human rRNA A site. By contrast, the binding of G418 induces the destacking of base 1492 when either rRNA A site serves as the drug target. In the aggregate, these results suggest that binding-induced base destacking at the rRNA A site is a critical factor in determining the prokaryotic specificity of aminoglycoside action, with binding affinity for the A site being of secondary importance.

2-Deoxystreptamine (2-DOS) aminoglycoside antibiotics bind specifically to the central region of the 16S rRNA A site and interfere with protein synthesis. Recently, we have shown that the binding of 2-DOS aminoglycosides to an A site model RNA oligonucleotide is linked to the protonation of drug amino groups. Here, we extend these studies to define the number of amino groups involved as well as their identities. Specifically, we use pH-dependent 15N NMR spectroscopy to determine the pK(a) values of the amino groups in neomycin B, paromomycin I, and lividomycin A sulfate, with the resulting pK(a) values ranging from 6.92 to 9.51. For each drug, the 3-amino group was associated with the lowest pK(a), with this value being 6.92 in neomycin B, 7.07 in paromomycin I, and 7.24 in lividomycin A. In addition, we use buffer-dependent isothermal titration calorimetry (ITC) to determine the number of protons linked to the complexation of the three drugs with the A site model RNA oligomer at pH 5.5, 8.8, or 9.0. At pH 5.5, the binding of the three drugs to the host RNA is independent of drug protonation effects. By contrast, at pH 9.0, the RNA binding of paromomycin I and neomycin B is coupled to the uptake of 3.25 and 3.80 protons, respectively, with the RNA binding of lividomycin A at pH 8.8 being coupled to the uptake of 3.25 protons. A comparison of these values with the protonation states of the drugs predicted by our NMR-derived pK(a) values allows us to identify the specific drug amino groups whose protonation is linked to complexation with the host RNA. These determinations reveal that the binding of lividomycin A to the host RNA is coupled to the protonation of all five of its amino groups, with the RNA binding of paromomycin I and neomycin B being linked to the protonation of four and at least five amino groups, respectively. For paromomycin I, the protonation reactions involve the 1-, 3-, 2'-, and 2"'-amino groups, while, for neomycin B, the binding-linked protonation reactions involve at least the 1-, 3-, 2', 6'-, and 2"'-amino groups. Our results clearly identify drug protonation reactions as important thermodynamic participants in the specific binding of 2-DOS aminoglycosides to the A site of 16S rRNA.

Spectroscopic and calorimetric techniques were employed to characterize and contrast the binding of the aminoglycoside paromomycin to three octamer nucleic acid duplexes of identical sequence but different strand composition (a DNA.RNA hybrid duplex and the corresponding DNA.DNA and RNA.RNA duplexes). In addition, the impact of paromomycin binding on both RNase H- and RNase A-mediated cleavage of the RNA strand in the DNA.RNA duplex was also determined. Our results reveal the following significant features: (i) Paromomycin binding enhances the thermal stabilities of the RNA.RNA and DNA.RNA duplexes to similar extents, with this thermal enhancement being substantially greater in magnitude than that of the DNA.DNA duplex. (ii) Paromomycin binding to the DNA.RNA hybrid duplex induces CD changes consistent with a shift from an A-like to a more canonical A-conformation. (iii) Paromomycin binding to all three octamer duplexes is linked to the uptake of a similar number of protons, with the magnitude of this number being dependent on pH. (iv) The affinity of paromomycin for the three host duplexes follows the hierarchy, RNA.RNA > DNA.RNA >> DNA.DNA. (v) The observed affinity of paromomycin for the RNA.RNA and DNA.RNA duplexes decreases with increasing pH. (vi) The binding of paromomycin to the DNA.RNA hybrid duplex inhibits both RNase H- and RNase A-mediated cleavage of the RNA strand. We discuss the implications of our combined results with regard to the specific targeting of DNA.RNA hybrid duplex domains and potential antiretroviral applications.