Skip to main content
NCBI home page
Protein Science : A Publication of the Protein Society logoLink to Protein Science : A Publication of the Protein Society
. 2017 Jul 21;26(9):1852–1863. doi: 10.1002/pro.3224

Effect of solvent and protein dynamics in ligand recognition and inhibition of aminoglycoside adenyltransferase 2″‐Ia

Valjean R Bacot‐Davis 1,✉, Angelia V Bassenden 1, Tara Sprules 1,2, Albert M Berghuis 1
PMCID: PMC5563142  PMID: 28734024

Abstract

The aminoglycoside modifying enzyme (AME) ANT(2″)‐Ia is a significant target for next generation antibiotic development. Structural studies of a related aminoglycoside‐modifying enzyme, ANT(3″)(9), revealed this enzyme contains dynamic, disordered, and well‐defined segments that modulate thermodynamically before and after antibiotic binding. Characterizing these structural dynamics is critical for in situ screening, design, and development of contemporary antibiotics that can be implemented in a clinical setting to treat potentially lethal, antibiotic resistant, human infections. Here, the first NMR structural ensembles of ANT(2″)‐Ia are presented, and suggest that ATP‐aminoglycoside binding repositions the nucleotidyltransferase (NT) and C‐terminal domains for catalysis to efficiently occur. Residues involved in ligand recognition were assessed by site‐directed mutagenesis. In vitro activity assays indicate a critical role for I129 toward aminoglycoside modification in addition to known catalytic D44, D46, and D48 residues. These observations support previous claims that ANT aminoglycoside sub‐class promiscuity is not solely due to binding cleft size, or inherent partial disorder, but can be controlled by ligand modulation on distinct dynamic and thermodynamic properties of ANTs under cellular conditions. Hydrophobic interactions in the substrate binding cleft, as well as solution dynamics in the C‐terminal tail of ANT(2″)‐Ia, advocate toward design of kanamycin‐derived cationic lipid aminoglycoside analogs, some of which have already shown antimicrobial activity in vivo against kanamycin and gentamicin‐resistant P. aeruginosa. This data will drive additional in silico, next generation antibiotic development for future human use to combat increasingly prevalent antimicrobial resistance.

Keywords: NMR, antibiotic resistance, solution structure, ligand‐protein interactions, antibiotics, aminoglycosides

Short abstract

PDB Code(s): 5KQJ

Abbreviations

AAC

Aminoglycoside N‐acetyltransferase

AME

Aminoglycoside modifying enzyme

ANT

Aminoglycoside O‐nucleotidyltransferase

APH

Aminoglycoside O‐phosphotransferase

KCC

Kanamycin A‐derived cationic compound

NT

Nucleotidyltransferase.

Introduction

A half‐century of mass antibiotic use has seen an emergence in resistant clinical isolates. The aminoglycoside class of antibiotics remaining in clinical use today include gentamicin, tobramycin, dibekacin, and sisomicin, with gentamicin being the most frequently administered in the emergency department.1 Gentamicin is used to treat ventilator‐associated pneumonia common with cystic fibrosis or traumatic burns (Pseudomonas aeruginosa), wound infections (Pseudomonas aeruginosa, Proteus mirabilis), bacterial broncopneumonia (Klebsiella pneumoniae), urinary tract infections (E. coli, Enterobacter spp.), and catheter‐associated bacteremia (Serratia spp.). Bacteria can alter active transport, accumulate changes to the ribosome, express modifying enzymes, or any combination thereof, to deter to the lethal effects of aminoglycoside antibiotics.2 This process is known as developing aminoglycoside antibiotic resistance. Aminoglycoside resistance outside of nosocomial (hospital acquired) infections remains limited, but has been reported for strains of Pseudomonas aeruginosa, Enterobacteriaceae, E. coli, Serratia spp., and Staphylococcus aureus.1 Resistance decreases treatment success, lengthens hospital stay, and less frequently, ends in mortality (21 vs 12%).3 Targeting bacterial expression of aminoglycoside modifying enzymes (AMEs) with next generation antibiotics resistant to modification as well as AME inhibitors, is pharmacologically attractive because the majority of resistance is caused by AME activity,4 and this high‐level AME resistance leads to difficulties treating clinical illnesses.

AMEs carry out enzymatic inactivation of aminoglycosides by three mechanisms: acetylation (AACs), nucleotidylation (ANTs), and phosphorylation (APHs).5 Aminoglycoside O‐nucleotidyltransferases (ANTs) bind ATP coordinated with magnesium followed by aminoglycoside, adenylate a specific hydroxyl group of the aminoglycoside substrate, then release modified‐aminoglycoside and inorganic phosphate to complete the inactivation reaction6. Aminoglycoside resistance mediated by nucleotidyltransferase ANT(2″)‐Ia is world‐wide, with reported clinical prevalences of 4% (Japan),7 12.9% (Nigeria),8 27% (USA),7 27.6% (Spain),9 and 78.87% (Iran).10

The X‐ray crystallographic structure of ANT(2″)‐Ia has been previously determined.11, 12 However, X‐ray, SAXS, and ITC structural studies of ANT(3″)(9) (56.4% similarity with ANT(2″)‐Ia13), reveal ANTs are dynamic, contain disordered segments, and aminoglycoside binding modulates the active site, shifting the thermodynamic properties of these enzymes.14 Our NMR studies of ANT(2″)‐Ia corroborate such findings of solution protein dynamics, with ATP‐aminoglycoside binding stabilizing nucleotidyltransferase (NT) domain and C‐terminal domain interactions for aminoglycoside nucleotidylation to efficiently proceed. Together, these new structural findings will be invaluable for competent in silico next generation antibiotic design, and development of novel drugs to resensitize multi‐drug resistant microorganisms.

Efforts now center around developing semisynthetic aminoglycosides averse to enzymatic inactivation, as well as direct, AME inhibitors. Protein structural information has already proven invaluable to AME inhibitor screening and identification.15 These contemporary NMR structural studies of adenyltransferase ANT(2″)‐Ia indicate two domain conformational dynamics, with greatest modulation upon ligand binding, corroborating similar observations for aminoglycoside adenyltransferase ANT(3″)(9). Alongside previously solved crystal structures of ANT(2″)‐Ia (PDB ID: 5CFS; 5CFU),12 these ANT(2″)‐Ia NMR conformers advocate kanamycin A‐derived cationic compounds with lipid, benzene, or naphthalene R groups as attractive next generation antibiotic candidates. Kanamycin A‐derived cationic lipids have already demonstrated antibacterial activity in vivo against kanamycin and gentamicin‐resistant P. aeruginosa.16, 17 These novel structural results will direct additional in silico next generation aminoglycoside development, further antibacterial drug identification for future in vivo human use, and combat the growing worldwide epidemic of antimicrobial resistance.

Results

ANT(2″)‐ia chemical shift assignments

The structure of ANT(2″)‐Ia (residues 1–185) was solved by multidimensional heteronuclear solution NMR spectroscopy. Backbone assignments are plotted onto the 1H‐15N‐HSQC of apo ANT(2″)‐Ia (Fig. 1), from HNCA, HNCB, CA(CO)NH, CB(CACO)NH experiments (Fig. 2), and were used in subsequent side‐chain, NOE (Supporting Information Fig. S1), structural, and titration analyses.

Figure 1.

Figure 1

Open in a new tab

apo ANT(2″)‐Ia 15N‐HSQC. 1H‐15N heteronuclear single quantum coherence spectroscopy spectra of apo, wild‐type ANT(2″)‐Ia. Peak intensities for 181 backbone residues were identified and labeled accordingly

Figure 2.

Figure 2

Open in a new tab

ANT(2″)‐Ia Strip Plot. Overlays of CA(CO)NH‐HNCA and CB(CACO)NH‐HNCB were cross‐referenced for sequential Cα, Cβ chemical shift assignments for 181 residues of ANT(2″)‐Ia (97.84%). A representative, step‐wise strip plot of G160‐A169 are show as examples

15 N/ 13 C‐NOESY solvent dynamics

15N‐NOESY and 13C‐NOESY ANT(2″)‐Ia experiments collected in 10% and 100% D2O revealed rearrangements in apo ANT(2″)‐Ia α4‐α6 helices short, medium, and long range NOE hydrogen interactions. It is worth mentioning NOESY experiments of ANT(2″)‐Ia in water (90% H2O/10% D2O) provided discernible, but weak, cross peaks for hydrogen interactions between α4 (Y135), α5 (V139), α6 (H148), and α7 (L159) residues (Table 1). This confirmed α4‐α6, closed conformation synergy, as well as α6‐long‐range, open conformation‐solvent interactions for this semi‐flexible region.

Table 1.

Solution 10% D2O 15N/13C‐NOESY C‐Terminal Conformational Dynamics

15N/13C 1H 1H Residue: α4‐α6 NOE Distance (Å)
57.69 4.68 0.74 W125CA‐HA‐L159HD22 3
29.99 3.16 0.74 W125CB‐HB3‐L159HD22 3.7
126.21 7.15 0.74 W125CD1‐HD1‐L159HD22 2.7
57.69 4.68 0.74 W125CA‐HA‐L159HD23 4.1
29.99 3.16 0.74 W125CB‐HB3‐L159HD23 4.6
126.21 7.15 0.74 W125CD1‐HD1‐L159HD23 2.5
29.99 3.16 4.31 W125CB‐HB3‐L159HA 4.6
126.21 7.15 4.31 W125CD1‐HD1‐L159HA 2.3
126.21 7.15 1.57 W125CD1‐HD1‐L159HB2 4.6
57.69 4.68 1.57 W125CA‐HA‐L159HB3 4.8
29.99 3.16 1.57 W125CB‐HB3‐L159HB3 4.1
126.21 7.15 1.57 W125CD1‐HD1‐L159HB3 3.1
57.69 4.68 0.74 W125CA‐HA‐L159HD21 4.1
126.21 7.15 1.51 W125CD1‐HD1‐L159HG 4.5
126.21 7.15 0.76 W125CD1‐HD1‐L159HD13 4.8
57.69 4.68 0.76 W125CA‐HA‐L159HD13 4.8
126.21 7.15 0.74 W125CD1‐HD1‐L159HD21 3.9
17.54 0.78 4.31 I128CG2‐HG21‐L159HA 4.9
17.54 0.78 1.57 I128CG2‐HG21‐L159HB2 4.8
17.54 0.78 1.57 I128CG2‐HG23‐L159HB2 4.9
13.47 0.68 1.57 I128CD1‐HD12‐L159HB2 5
17.54 0.78 1.57 I128CG2‐HG21‐L159HB3 3.6
17.54 0.78 1.57 I128CG2‐HG22‐L159HB3 4.5
17.54 0.78 1.57 I128CG2‐HG23‐L159HB3 3.9
13.47 0.68 1.57 I128CD1‐HD12‐L159HB3 4.6
27.74 1.24 1.51 I128CG1‐HG13‐L159HG 4.6
17.54 0.78 1.51 I128CG2‐HG21‐L159HG 4.2
13.47 0.68 1.51 I128CD1‐HD12‐L159HG 4.9
38.61 1.79 0.76 I128CB‐HB‐L159HD11 4.6
27.74 1.24 0.76 I128CG1‐HG12‐L159HD11 4
27.74 1.24 0.76 I128CG1‐HG13‐L159HD11 2.5
17.54 0.78 0.76 I128CG2‐HG21‐L159HD11 2.9
17.54 0.78 0.76 I128CG2‐HG22‐L159HD11 4.6
17.54 0.78 0.76 I128CG2‐HG23‐L159HD11 3.4
13.47 0.68 0.76 I128CD1‐HD11‐L159HD11 3.1
13.47 0.68 0.76 I128CD1‐HD12‐L159HD11 2.6
13.47 0.68 0.76 I128CD1‐HD13‐L159HD11 4.2
27.74 1.24 0.76 I128CG1‐HG13‐L159HD12 4.1
17.54 0.78 0.76 I128CG2‐HG21‐L159HD12 4
17.54 0.78 0.76 I128CG2‐HG23‐L159HD12 3.9
13.47 0.68 0.76 I128CD1‐HD11‐L159HD12 3.9
13.47 0.68 0.76 I128CD1‐HD12‐L159HD12 3.1
13.47 0.68 0.76 I128CD1‐HD13‐L159HD12 4.8
38.61 1.79 0.76 I128CB‐HB‐L159HD13 4.4
27.74 1.24 0.76 I128CG1‐HG12‐L159HD13 4.5
27.74 1.24 0.76 I128CG1‐HG13‐L159HD13 3.1
17.54 0.78 0.76 I128CG2‐HG21‐L159HD13 2.3
17.54 0.78 0.76 I128CG2‐HG22‐L159HD13 3.5
17.54 0.78 0.76 I128CG2‐HG23‐L159HD13 2.2
13.47 0.68 0.76 I128CD1‐HD11‐L159HD13 3.6
13.47 0.68 0.76 I128CD1‐HD12‐L159HD13 2.2
13.47 0.68 0.76 I128CD1‐HD13‐L159HD13 3.9
38.61 1.79 0.74 I128CB‐HB‐L159HD21 4.3
27.74 1.24 0.74 I128CG1‐HG13‐L159HD21 3.1
27.74 1.24 0.74 I128CG1‐HG12‐L159HD21 4.8
17.54 0.78 0.74 I128CG2‐HG21‐L159HD21 2.6
17.54 0.78 0.74 I128CG2‐HG22‐L159HD21 4.3
17.54 0.78 0.74 I128CG2‐HG23‐L159HD21 4.1
13.47 0.68 0.74 I128CD1‐HD12‐L159HD21 4.4
38.61 1.79 0.74 I128CB‐HB‐L159HD22 4.1
27.74 1.24 0.74 I128CG1‐HG13‐L159HD22 3.7
17.54 0.78 0.74 I128CG2‐HG21‐L159HD22 1.8
17.54 0.78 0.74 I128CG2‐HG22‐L159HD22 3.1
17.54 0.78 0.74 I128CG2‐HG23‐L159HD22 3.2
13.47 0.68 0.74 I128CD1‐HD12‐L159HD22 4.1
27.74 1.24 0.74 I128CG1‐HG13‐L159HD23 4.8
17.54 0.78 0.74 I128CG2‐HG21‐L159HD23 3.5
17.54 0.78 0.74 I128CG2‐HG22‐L159HD23 4.9
17.54 0.78 0.74 I128CG2‐HG23‐L159HD23 4.9
17.54 0.78 0.76 I129CG2‐HG22‐L159HD12 4.7
61.64 4.18 4.18 I129CA‐HA‐L159HD13 4.1
17.54 0.78 0.76 I129CG2‐HG22‐L159HD13 3.6
17.54 0.78 0.76 I129CG2‐HG23‐L159HD13 4.1
17.54 0.78 0.74 I129CG2‐HG22‐L159HD22 4.6
17.54 0.78 0.74 I129CG2‐HG23‐L159HD22 4.3
58.14 4.63 6.94 Y132CA‐HA‐Y152HD1 5
58.14 4.63 6.71 Y132CA‐HA‐Y152HE1 3.5
39.3 2.88 6.71 Y132CB‐HB2‐Y152HE1 4.2
39.3 2.88 6.71 Y132CB‐HB3‐Y152HE1 2.7
132.39 6.94 6.71 Y132CD1‐HD1‐Y152HE1 4
132.35 6.93 6.71 Y132CD2‐HD2‐Y152HE1 3.1
117.76 6.71 6.94 Y132CE2‐HE2‐Y152HD1 3.5
117.76 6.71 4.63 Y132CE2‐HE2‐Y152HA 4.7
117.66 6.71 1.36 Y132CE1‐HE1‐A155HB1 3.5
117.76 6.71 1.36 Y132CE2‐HE2‐A155HB1 4.6
117.66 6.71 1.36 Y132CE1‐HE1‐A155HB3 4.1
117.76 6.71 1.36 Y132CE2‐HE2‐A155HB3 4.7
132.39 6.94 4.67 Y132CD1‐HD1‐C156HA 4.4
117.66 6.71 4.67 Y132CE1‐HE1‐C156HA 3.2
117.66 6.71 8.41 Y132CE1‐HE1‐C156H 4.6
117.66 6.71 1.51 Y132CE1‐HE1‐L159HG 3.8
132.39 6.94 0.76 Y132CD1‐HD1‐L159HD11 4.6
117.66 6.71 0.76 Y132CE1‐HE1‐L159HD11 2.8
132.39 6.94 0.76 Y132CD1‐HD1‐L159HD12 3.5
117.66 6.71 0.76 Y132CE1‐HE1‐L159HD12 1.4
132.39 6.94 0.76 Y132CD1‐HD1‐L159HD13 3.6
117.66 6.71 0.76 Y132CE1‐HE1‐L159HD13 2.6
117.66 6.71 0.74 Y132CE1‐HE1‐L159HD21 5
117.66 6.71 0.74 Y132CE1‐HE1‐L159HD22 4.9
39.3 2.88 6.71 Y135CB‐HB2‐Y152HE1 4.3
132.39 6.94 6.71 Y135CD1‐HD1‐Y152HE1 4.4
58.14 4.63 9.76 Y135CA‐HA‐H148HE2 3.5
39.3 2.88 9.76 Y135CB‐HB2‐H148HE2 4.5
39.3 2.88 9.76 Y135CB‐HB3‐H148HE2 2.9
132.39 6.94 9.76 Y135CD1‐HD1‐H148HE2 4.7
132.35 6.93 9.76 Y135CD2‐HD2‐H148HE2 3.4
117.76 6.71 9.76 Y135CE2‐HE2‐H148HE2 4.7
132.39 6.94 7.97 Y135CD1‐HD1‐H148HE1 4.8
117.66 6.71 7.97 Y135CE1‐HE1‐H148HE1 4.9
18.99 1.36 9.32 A136CB‐HB1‐Y152HH 4.9
18.99 1.36 9.32 A136CB‐HB2‐Y152HH 3.4
18.99 1.36 9.32 A136CB‐HB3‐Y152HH 5
30 2.01 9.76 E138CB‐HB3‐H148HE2 4.1
32.73 1.99 9.76 V139CB‐HB‐H148HE2 3.6
21.29 0.8 9.76 V139CG2‐HG21‐H148HE2 4.4
21.29 0.8 9.76 V139CG2‐HG23‐H148HE2 4.3
Open in a new tab

Solution structure and conformational dynamics of ANT(2″)‐ia

Structural statistics and conformer assembly of the 15 deposited ANT(2″)‐Ia NMR structures are listed and shown [Supporting Information Table S1, Fig. 3(A)] (BMRB ID: 30133; PDB ID: 5KQJ). In solution, ANT(2″)‐Ia maintains the typical α1‐β1‐α2‐β2‐X‐β4‐β3 NT fold previously characterized by X‐ray crystallography.11, 12 However, proline‐rich α‐turn 5 can adopt a 30% population, open conformation [‐6313.155 KJ/mol energy,18 model 11, Fig. 3(B)], or thermodynamically lower‐energy, 70% population, closed conformation [‐6784.851 KJ/mol, model 1, Fig. 3(C)] observed in protein crystals. Cα imposition of the open and closed ANT(2″)‐Ia conformers show the full range of C‐terminal dynamics [Fig. 3(D)]. The electronegative, aminoglycoside‐binding patch near α4 is solvent accessible in open conformation, and hidden due to α4‐α6 interactions while closed [Fig. 3(E,F)].

Figure 3.

Figure 3

Open in a new tab

Solution Structure of ANT(2″)‐Ia. (A) Ensemble of fifteen lowest energy solution structures of apo P. aeruginosa ANT(2″)‐Ia. (B) Ribbon presentation of −6313.115 KJ/mol energy, open conformation of ANT(2″)‐Ia (model 11). (C) Ribbon representation of −6784.851 KJ/mol energy, closed conformation of ANT(2″)‐Ia (model 1). (D) Surface representation of the fifteen lowest energy solution structures of apo ANT(2″)‐Ia. (E) Molecular electronic potential surface of ANT(2″)‐Ia in solution, generated using CHARMM, in the same orientation as A‐D. (F) 180° rotation along the x‐axis

In vitro functional studies of protein mutants

Alanine mutagenesis was performed to monitor in vitro ANT(2″)‐Ia aminoglycoside modification over time, as well as corroborate NMR observations. D44A, I129A, Y135A protein mutations were purified [Supporting Information Fig. S2(A)] and analyzed by 1D 1H‐NMR [Supporting Information Fig. S2(B)]. Sharp, well separated peaks in the amide/aromatic and aliphatic regions for all verified each protein mutation was well‐folded, and any observed changes in activity were not due to mutation‐induced misfolding or aggregation. Unexpectedly, highly conserved Y135A only caused a minor 1.6‐fold decrease in aminoglycoside adenylation [Fig. 4(A)]. However, aminoglycoside modification diminished 3‐fold for nearby I129A, suggesting this amino acid substitution hinders C‐terminal conformation dynamics, substrate recognition, or both. D44A caused a 4‐fold decrease in aminoglycoside modification as expected [Fig. 4(A)]. Based on NMR and mutagenesis experiments, we propose a model where residues Y74, Y134, and Y135 initially interact with aminoglycoside in open conformation to correctly position the ligand for closed conformation catalysis. I129 and α‐turn 5 (P140‐P145) dynamics aid this conformational switch [Fig. 4(C)], with b‐factors of the ANT(2″)‐Ia crystal structures supporting these observations [Fig. 4(D)].11, 12 Residues of α4 (I129, Y134, Y135) facilitate aminoglycoside positioning for nucleophilic attack and nucleotidylation to occur [Fig. 4(C)], proficiently conferring resistance to the microorganism.

Figure 4.

Figure 4

Open in a new tab

ANT(2″)‐Ia aminoglycoside modification in solution. (A) Logarithmic plot of aminoglycoside modification rates for several ANT(2″)‐Ia protein mutations. Catalytic mutation D44A (yellow) and α4 residue I129A (green) display severely reduced adenylation profiles. (B) Cartoon representation and molecular surface of ANT(2″)‐Ia NMR conformers (models 1–15) and spherical tobramycin (magenta) [PDB ID: 5CFS] with residues colored according to the protein mutagenesis activity assay graph D44 (yellow), I129 (green) Y135 (purple). (C) Overlaid cartoon representations of ANT(2″)‐Ia NMR conformers (models 1–15) illustrate α5 P140‐P145 (cyan), Y74/Y134 (pink), and K147/H148 (blue) dynamics. (D) X‐ray structure b‐factors support α5 P140‐P145 conformational dynamics in solution [PDB ID: 4WQL]

Solution structure catalytic residues

Previously solved X‐ray crystal structures of ANT(2″)‐Ia bound with AMP‐tobramycin (PDB ID: 5CFS) and adenylated tobramycin (PDB ID: 5CFU) indicated key residues involved in ATP binding and aminoglycoside modification catalysis [Fig. 5(A)].12 The NMR solution conformers of ANT(2″)‐Ia (PDB ID: 5KQJ) propose Y74, Y134, and Y135 hydrophobic stacking interactions cooperate to coordinate aminoglycoside binding recognition [Fig. 5(B,C)]. Additionally, open‐closed conformation dynamics modulate K147, H148 ionic interactions with the β‐γ phosphate moiety of ATP, involved in release of inorganic phosphate after aminoglycoside adenylation.

Figure 5.

Figure 5

Open in a new tab

ANT(2″)‐Ia reaction mechanism. X‐ray crystallography resolved ANT(2″)‐Ia residues involved in ATP‐aminoglycoside binding and aminoglycoside adenylation (PDB ID: 5CFS and 4CFU) as shown (Figure 6A). The NMR solution structures of apo ANT(2″)‐Ia were aligned with tobramycin (PDB ID: 5CFS), and key residues determined by crystallography (Y74, Y134, K147, H148) or protein mutagenesis (I129) were displayed [Fig. 5(B)]. A surface mesh depiction of these ANT(2″)‐Ia solution conformers and residues noted a prominent hydrophobic region contributed from Y74 and Y134 (Figure 5C) benefitting further consideration

1H‐15N‐HSQC chemical shift perturbation ligand binding

1H‐15N‐HSQC titration experiments were carried out on ANT(2″)‐Ia unbound (red) or bound with ‐ATP (blue), ‐Tobramycin (green), or ‐ATP and Tobramycin (purple). Global perturbation effects were resolved (Supporting Information Fig. S3) and quantified for individual amino acids (Supporting Information Fig. S4), with significant change calculated as Δδ > 0.1 ppm.19 1H‐15N‐HSQC perturbation results for catalytic residues resolved by X‐ray crystallography11, 12 or protein mutagenesis, were further examined. D44, D86, E88, I129, K147, and H148 showed significant titration effects as a results of ATP binding, while D44, Y74, E88, I129, Y134, and K147 considerably shifted from aminoglycoside interactions; D43, D44, Y74, D86, E88, and K147 varied extensively concordant with adenylation catalysis (Fig. 6).

Figure 6.

Figure 6

Open in a new tab

Expanded 15N‐HSQC ANT(2″)‐Ia titration. Overlays of the 2D 15N‐HSQC spectra of ANT(2″)‐Ia in apo form (red), or titrated with ATP (blue), tobramycin (green), or ATP and tobramycin (purple). Residues involved in adenylation as determined by X‐ray crystallography or protein mutagenesis are labeled with the residue number. Resonances show significant broadening, and/or changes in chemical shift, associated with ligand binding, and aminoglycoside adenylation

Backbone dynamics of ANT(2″)‐ia ATP‐aminoglycoside binding

Analysis of the heteronuclear {1H}‐15N NOE ratio between ANT(2″)‐Ia bound and unbound to ATP plus tobramycin is a direct method to assess conformational changes promoted due to ligand interactions. Observational changes suggest that while the majority of ANT(2″)‐Ia is a rigid structure, semi‐flexible dynamics are observed for ATP‐site residues D43 and D46, aminoglycoside modification residues R54, R67, G75, aminoglycoside‐recognition residues W125; I129‐D137, and C‐terminal residues Y152‐H185 [Fig. 7(A)]. Noticeably increasing semi‐flexibility in the C‐terminus accounted for most backbone dynamics, with strong cross‐peaks attributable to tobramycin [Fig. 7(B)] and ATP [Fig. 7(C)] coordination in respective regions.

Figure 7.

Figure 7

Open in a new tab

ANT(2″)‐Ia {1H}‐15N NOE Titration Flexibility. (A) Heteronuclear {1H}‐15N NOE values demonstrate ligand site and C‐terminus semi‐flexibility between ANT(2″)‐Ia [ATP + tobramycin] bound [1:1] and unbound [1:0] states. (B) One‐dimensional proton spectrum of tobramycin and (C) ATP were also used in ANT(2″)‐Ia NOE cross‐peak referencing

Discussion

Overall, our NMR results strongly suggest the existence of closed and open conformational states for ligand‐free ANT(2″)‐Ia in a 7:3 ratio, with ligand‐binding shifting this equilibrium to an entirely closed conformation. Analysis of NOE measurements determined and assigned by PONDEROSA‐C/S allowed NMR data interpretation augmented by existing crystal structures (PDB ID: 4WQK, 4WQL, 5CFU, 5CFS, 4XJE).11, 12 PONDEROSA‐C/S avoids the underestimated problems associated with maximum‐entropy and large‐weight approaches of ensemble construction.20, 21 The 30% population of the open conformation in apo ANT(2″)‐Ia suggests that solution ligand binding involves what is known as a conformation‐selection mechanism, which facilitates ligand binding, and modulation of binding affinity.22 Rapidly progressing spectroscopic and computational techniques have the possibility to further integrate the study of protein dynamics in solution at an atomic level. Here, we use NMR experimentation and the computational power of PONDEROSA‐C/S to reliably interpret structural NOE data, and resolve the conformational dynamics crucial to solution protein interactions, ligand recognition, and binding affinity.23

1H‐15N‐HSQC titration, {1H}‐15N NOE experiments, and solution NMR structural studies verified residues involved in ANT(2″)‐Ia aminoglycoside recognition, ATP coordination, and aminoglycoside adenylation (D43, D44, Y74, D86, E88, I129, Y134, Y135, K147, H148, R174, Y175), with major aminoglycoside recognition residues located or involved in interactions with α4 of ANT(2″)‐Ia. As protein structures have been used successfully to identify AME inhibitors15 and design next generation antibiotics, future endeavors will require the most accurate and encompassing structural information for the implementation of both to combat human disease. Here, the NMR solution structure of ANT(2″)‐Ia offers insight into the repositioning of its NT and C‐terminal domains, just as observed for ANT(3″)(9)'s NT and C‐terminal α‐bundle domain, for aminoglycoside binding in the interdomain cleft and catalysis to occur.14 Both findings are consistent with the regulatory function reported for the C‐terminus in proteobacterial members of the polβ nucleotidyltransferase superfamily (i.e., CyaA adenylyl cyclase, DNA polymerase λ), of which nucleotidyltransferases are a member.24, 25

ANT(2″)‐Ia confers resistance by adenylating the 2″‐OH group of 4,6‐disubstituted 2‐deoxystreptamines, such as tobramycin, gentamicin, and kanamycin26 [Fig. 8(A–C)]. Kanamycin A‐derived cationic lipid compound 15 [Fig. 8(D)] was previously shown to restore antibacterial activity against kanamycin and gentamicin‐resistant P. aeruginosa.16 ANT(2″)‐Ia NMR solution structure conformers advocate development of 4,6‐disubstituted 2‐deoxystreptamines cationic compounds with lipid or hydrophobic R2 groups to combat clinical aminoglycoside resistance. Although the exact mechanism of resistance to modification for compound 15 warrants further study, next generation aminoglycosides effective against AMEs are only attractive if these analogs retain ability to inhibit their target (ie. the bacterial ribosome), as well as avoid AME modification. For this purpose, we propose that Kanamycin A‐derived cationic compounds (KCCs) that have similar hydrophobic R2 groups, could target ANT hydrophobic stacking interactions (Y74, Y134), and be effective next generation antibiotics in vivo against gentamicin and kanamycin‐resistant microorganisms.16

Figure 8.

Figure 8

Open in a new tab

ANT(2″)‐Ia substrate adjacent potential next generation aminoglycoside. 4,6‐disubstituted 2‐deoxystreptamine aminoglycosides such as (A) tobramycin, (B) gentamycin, and (C) kanamycin B, can be adenylated by ANT(2″)‐Ia. (D) Kanamycin A‐derived cationic lipid compound 15/KCC1 retains antimicrobial activity in kanamycin‐resistance microorganisms

Highly static, superimposable X‐ray crystal structures of ANT(2″)‐Ia, while extremely valuable, fail to fully explain protein dynamics according to differential substrate profiles, NMR spectra, H/D exchange data, point mutations (ANT(2″)‐Ia I129A), and k cat values.11, 12, 27, 28, 29, 30 Incorporating NMR structural data suggest that ANT aminoglycoside promiscuity is driven by an inherent ANT flexibility largely unpredictable by primary sequence alone. In conclusion, thermodynamic parameters of enzyme–ligand complexes, NMR conformational dynamics, and protein behavior in solution can complement X‐ray crystallographic protein structure determination to improve in silico docking algorithms used in the initial steps of identifying next generation antibiotics, as well as compounds to inhibit AMEs. These advancements will lead to the design of new, novel antibiotics, as well as more effective inhibitors to combat clinical antimicrobial resistance.

Materials and Methods

Isotopically‐labeled protein purification and NMR buffer

BL‐21(DE3) E. coli transformed with pET22b(+)‐ANT(2″)‐Ia were grown overnight in 2xYT broth, pelleted, washed, and diluted into 15NH4Cl, 13C‐glucose, M9 minimal media for a final OD600 of 0.6. 15N‐13C‐ANT(2″)‐Ia expression was induced by 10 mM IPTG at 15°C overnight. Cells were collected, lysed, and hexa‐histidine tagged 15N‐ or 13C/15N ‐ANT(2″)‐Ia was purified as previously described.12 15N‐ or 13C/15N ‐ANT(2″)‐Ia [0.5 mM] was exchanged into 10 mM Na2HPO4, pH 7.0, 0.04% NaN3, 10 mM DTT for finalized NMR experiments.

NMR spectroscopy

Spectra were collected on a Varian INOVA 500 MHz or Varian INOVA 800 MHz spectrometer at 298 K in the Quebec/Eastern Canada High Field NMR Facility. The following spectra were recorded for backbone, side‐chain, and NOE assignments: 2D (1H‐15N)‐HSQC, 1H‐13C‐HSQC, and 3D HNCO, HNCA, HNCB, CA(CO)NH, CB(CACO)NH, C(CO)NH, HC(CO)NH, HCCH‐COSY, 15N‐NOESY, and 13C‐NOESY. Heteronuclear {1H)‐15N NOE values for ANT(2″)‐Ia 1:1 and 1:0 [ATP + tobramycin] titration were acquired at 298 K using a Bruker DRX 600 MHz spectrometer, processed with NMRPipe,31 SPARKY,32 and analyzed in Excel. Error, rigid, semi‐flexible, and flexible values were estimated as previously described.33

Data processing and structure determination

NMR data were processed with NMRPipe.31 Backbone and side‐chain chemical shifts were assigned manually using SPARKY.32 15N‐NOESY/13C‐NOESY NOE assignments and XPLOR‐NIH structural calculations were conducted automatically through the PONDEROSA‐C/S client.21, 22, 34, 35 KobaMIN was used for final refinement of XPLOR‐NIH generated structures.36 The best models determined by MolProbity were used as the final structures of ANT(2″)‐Ia.37, 38 The fifteen best NMR solution structures of ANT(2″)‐Ia were deposited into the biological magnetic resonance and protein data bank as BMRB ID: 30133 and PDB ID: 5KQJ.

ANT(2″)‐ia solvent dynamics

Ten percent and 100% D2O aliphatic and aromatic 13C‐NOESY spectra were collected on 13C/15N ‐ANT(2″)‐Ia to assess solvent‐protein and intermolecular protein‐protein short‐, medium‐, and long‐range NOE interactions.

Protein mutagenesis

pET‐22b(+) ANT(2″)‐Ia D44A, I129A, and Y135A were cloned by overlap extension PCR using appropriate primers (Supporting Information Table S2) with restriction enzymes NdeI and NotI, and verified by restriction digest.

ANT(2″)‐Ia activity assay

Twenty‐five microliter [0.1 mg/mL] ANT(2″)‐Ia or a ddH2O control were incubated at 30°C with assay buffer consisting of 25 mM HEPES, pH 7.5, 1.0 M KCl, 100 mM MgCl2, 20 mM ATP, 2.5 mM tobramycin, and 2.0 units inorganic pyrophosphatase (NEB) for a total volume of 50 μL. The reaction was allowed to proceed 2, 4, 6, 8, or 10 minutes, and stopped by the addition of 1:1 mixture of 1% ammonium molybdate and 10% ascorbic acid. 820 nm absorbance (Abs) was used as an indicator of aminoglycoside modification.39 Reactions were repeated in triplicate and normalized to the ddH2O control. One‐way ANOVA statistical analysis of the ANT(2″)‐Ia activity assay data sets calculated a P‐level value less than 0.001.

1H‐15N chemical‐shit perturbation

0.3 mM 13C/15N‐labeled ANT(2″)‐Ia were titrated with no substrate, 0.6 mM tobramycin, 0.6 mM ATP, or 0.6 mM tobramycin and 0.6 mM ATP at 25°C for one hour with a sample dilution < 10%. 15N‐HSQC spectra were collected using a Bruker DRX 600 MHz NMR spectrometer at 298 K over 3 hours. Data were processed using NMRPipe and analyzed by SPARKY. Chemical‐shift perturbation effects were calculated as previously described.19

Supporting information

Supporting Information.

Click here for additional data file. (429.6KB, tif)

Supporting Information.

Click here for additional data file. (449.2KB, pdf)

Acknowledgments

We are grateful to Dr. Albert M. Berghuis, McGill University, for supervision, Angelia V. Bassenden, McGill University, for providing the ANT(2″)‐Ia construct, and Dr. Tara Sprules of the Quebec/Eastern Canada High Field NMR Facility, for aiding in NMR data acquisition and processing.

References

  • 1. Cameron P, Jelinek G, Kelly A‐E, Brown AFT, Little M (2014) Textbook of adult emergency medicine, 4th ed. Edinburgh: Elsevier Health Sciences. [Google Scholar]
  • 2. Mingeot‐Leclercq M‐P, Glupczynski Y, Tulkens PM (1999) Aminoglycosides: activity and resistance. Antimicrob Agents Chemother 43:727–737. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Hirsch EB, Tam VH (2010) Impact of multidrug‐resistance Pseudomonas aeruginosa infection on patient outcomes. Expert Rev Pharmacoecon Outcomes Res 10:441–451. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Labby KJ, Garneau‐Tsodikova S (2013) Strategies to overcome the action of aminoglycoside‐modifying enzymes for treating resistant bacterial infections. Futur Med Chem 5:1285–1309. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Shaw KJ, Rather PN, Hare RS, Miller GH (1993) Molecular genetic of aminoglycoside resistance genes and familial relationships of the aminoglycoside‐modifying enzymes. Microbiol Rev 57:138–163. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Bongaerts GP, Molendijk L (1984) Relation between aminoglycoside 2"‐O‐nucleotidyltransferase activity and aminoglycoside resistance. Antimicrob Agents Chemother 25:234–237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Shimizu K, Kumada T, Hsieh WC, Chung HY, Chong Y, Hare RS, Miller GH, Sabatelli FJ, Howard J (1985) Comparison of aminoglycoside resistance patterns in Japan, Formosa, and Korea, Chile, and the United States. Antimicrob Agents Chemother 28:282–288. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Odumosu BT, Adeniyi BA, Chandra R (2015) Occurrence of aminoglycoside‐modifying enzymes genes (aac(6′)‐I and ant(2″)‐I) in clinical isolates of Pseudomonas aeruginosa from Southwest Nigeria. Afr Heal Sci 15:1277–1281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Fernández‐Martinez M, Miró E, Ortega A, Bou G, Gonzálex‐López JJ, Oliver A, Pascual A, Cercenado E, Oteo J, Marínex‐Martínex L, Navarro F (2015) Molecular identification of aminoglycoside‐modifying enzymes in clinical isolates of Escherichia coli resistant to amoxicillin/clavulanic acid isolated in Spain. Int J Antimicrob Agents 46:157–163. [DOI] [PubMed] [Google Scholar]
  • 10. Soleimani N, Aganj M, Ali L, Shokoohizadeh L, Sakinc T (2014) Frequency distribution of genes encoding aminoglycoside modifying enzymes in uropathogenic E. coli isolated from Iranian hospital. BMC Res Notes 7:842. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Cox G, Stogios PJ, Savchenko A, Wright GD (2015) Structural and molecular basis for resistance to aminoglycoside antibiotics by the adenylyltransferase ANT(2’’)‐Ia. MBio 6:e02180–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Bassenden AV, Rodionov D, Shi K, Berghuis AM (2016) Structural analysis of the tobramycin and gentamicin clinical resistome reveals limitations for next‐generation aminoglycoside design. ACS Chem Biol 11:1339–1346. [DOI] [PubMed] [Google Scholar]
  • 13. Li W, Cowley A, Uludag M, Gur T, McWilliam H, Squizzato S, Park YM, Buso N, Lopez R (2015) The EMBL‐EBI bioinformatics web and programmatic tools framework. Nucl Acids Res 43:W580–W584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Chen Y, Näsvall J, Wu S, Andersson DI, Selmer M (2015) Structure of AadA from Salmonella enterica: a monomeric aminoglycoside (3’’)(9) adenyltransferase. Acta Cryst D71:2267–2277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Lin DL, Tran T, Adams C, Alam JY, Herron S, Tolmasky ME (2013) Inhibitors of the aminoglycoside 6’‐N‐acetyltransferase type Ib [AAC(6’)‐Ib] identified by in‐silico molecular docking. Bioorg Med Chem Lett 23:5694–5698. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Bera S, Zhanel GG, Schweizer F (2010) Antibacterial activity of guanidinylated neomycin B‐ and kanamycin A‐derived amphiphilic lipid conjugates. J Antimicrob Chemother 65:1224–1227. [DOI] [PubMed] [Google Scholar]
  • 17. Bera S, Zhanel GG, Schweizer F (2010) Antibacterial activities of aminoglycoside antibiotics‐derived cationic amphiphiles. Polyol‐modified neomycin B‐, Kanamycin A‐, Amikacin‐, and Neamine‐based amphiphiles with potent broad spectrum antibacteria activity. J Med Chem 53:3626–3631. [DOI] [PubMed] [Google Scholar]
  • 18. Guex N, Peitsch MC (1997) SWISS‐MODEL and the Swiss‐PdbViewer: an environment for comparative protein modeling. Electrophoresis 18:2714–2723. [DOI] [PubMed] [Google Scholar]
  • 19. Qiu Y, Zhang W, Zhao C, Wang Y, Wang W, Zhang J, Zhang Z, Li G, Shi Y, Tu X, Wu J (2012) Solution structure of the Pdp1 PWWP domain reveals its unique binding sites for methylated H4K20 and DNA. Biochem J 442:527–538. [DOI] [PubMed] [Google Scholar]
  • 20. Lee W, Stark JL, Markley JL (2014) PONDEROSA‐C/S: client–server based software package for automated protein 3D structure determination. J Biomol NMR 60:73–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Lee W, Kim JH, Westler WM, Markley JL (2011) PONDEROSA, an automated 3D‐NOESY peak picking program, enables automated protein structure determination. Bioinformatics 27:1727–1728. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Marvin JS, Hellinga HW (2001) Manipulation of ligand binding affinity by exploitation of conformational coupling. Nat Struct Biol 8:795–798. [DOI] [PubMed] [Google Scholar]
  • 23. Delaforge E, Milles S, Bouvignies G, Bouvier D, S B, Salvi N, Maurin D, Martel A, Round A, Lemke EA, Jensen MR, Hart DJ, Blackledge M (2015) Large‐scale conformational dynamics control H5N1 influenza polymerase PB2 binding to importin α. J Am Chem Soc 137:15122–15134. [DOI] [PubMed] [Google Scholar]
  • 24. Barakat KH, Gajewski MM, Tuszynski JA (2012) DNA polymerase beta (pol β) inhibitors: a comprehensive overview. Drug Discov Today 17:913–920. [DOI] [PubMed] [Google Scholar]
  • 25. Fiala KA, Duym WW, Zhang J, Suo Z (2006) Up‐regulation of the fidelity of human DNA polymerase lambda by its non‐enzymatic proline‐rich domain. J Biol Chem 281:19038–19044. [DOI] [PubMed] [Google Scholar]
  • 26. Akers KS, Chaney C, Barsoumian A, Beckius M, Zera W, Yu X, Guymon C, Keen EF, 3rd , Robinson BJ, Mende K, Murray CK (2010) Aminoglycoside resistance and susceptibility testing errors in Acinectobacter baumannii‐calcoaceticus complex. J Clin Microbiol 48:1132–1138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Serpersu EH, Norris AL Effect of protein dynamics and solvent in ligand recognition by promiscuous aminoglycoside‐modifying enzymes In: Derek H, Ed. (2012) Advances in carbohydrate chemistry and biochemistry (Vol. 67). Waltham, MA: Academic Press, pp. 221–248. [DOI] [PubMed] [Google Scholar]
  • 28. Norris AL, Serpersu EH (2009) NMR detected hydrogen‐deuterium exchange reveals differential dynamics of antibiotic and nucleotide‐bound aminoglycoside phosphotransferase 3’‐IIIa. J Am Chem Soc 131:8587–8594. [DOI] [PubMed] [Google Scholar]
  • 29. Vetting MW, Hedge SS, Javid‐Majd F, Blanchard JS, Roderick SL (2002) Aminoglycoside 2’‐N‐acetyltransferase from Mycobacterium tuberculosis in complex with coenzyme A and aminoglycoside substrates. Nat Struct Biol 9:653–658. [DOI] [PubMed] [Google Scholar]
  • 30. Norris AL, Serpersu EH (2013) Ligand promiscuity through the eyes of the aminoglycoside N3 acetyltransferase IIa. Protein Sci 22:916–928. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Delaglio F, Grzesiek S, Vuister GW, Zhu G, Pfeifer J, Bax A (1995) NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J Biomol NMR 6:277–293. [DOI] [PubMed] [Google Scholar]
  • 32. Goddard TD, Kneller DG. SPARKY 3. San Francisco: University of California. [Google Scholar]
  • 33. Estrada DF, Conner M, Jeor SC, Guzman RN (2011) The structure of the hantavirus zinc finger domain is conserved and represents the only natively folded region of the Gn cytoplasmic tail. Front Microbiol 2:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Schwieters CD, Kuszewski JJ, Tjandra N, Clore GM (2003) The Xplor‐NIH NMR molecular structure determination package. J Magn Res 160:66–74. [DOI] [PubMed] [Google Scholar]
  • 35. Schwieters CD, Kuszewski JJ, Clore GM (2006) Using Xplor‐NIH for NMR molecular structure determination. Progr NMR Spectrosc 48:47–62. [Google Scholar]
  • 36. Rodrigues JP, Levitt M, Chopra G (2012) KoBaMIN: a knowledge‐based minimization web server for protein structure refinement. Nucleic Acids Res 40:W323–W328. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Davis IW, Leaver‐Fay A, Chen VB, Block JN, Kapral GH, Wang X, Murray LW, Arendall WB, Snoeyink J, Richardson JS, Richardson DC (2007) MolProbity: all‐atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res 35:W375–W383. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Chen VB, Arendall WB, Headd JJ, Keedy DA, Immormino RM, Kapral GJ, Murray LW, Richardson JS, Richardson DC (2010) MolProbity: all‐atom structure validation for macromolecular crystallography. Acta Cryst D66:12–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Rahmati‐Bahram A, Magee JT, Jackson SK (1997) Effect of temperature on aminoglycoside binding sites in Stenotrophomonas maltophilia. J Antimicrob Chemother 39:19–24. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information.

Click here for additional data file. (429.6KB, tif)

Supporting Information.

Click here for additional data file. (449.2KB, pdf)

Articles from Protein Science : A Publication of the Protein Society are provided here courtesy of The Protein Society

RESOURCES