REDOR constraints on the peptidoglycan lattice architecture of Staphylococcus aureus and its FemA mutant

Abstract

The peptidoglycan of Gram-positive bacteria consists of glycan chains with attached short peptide stems cross-linked to one another by glycyl bridges. The bridge of Staphylococcus aureus has five glycyl units and that of its FemA mutant has one. These long- and short-bridge cross-links create totally different cell-wall architectures. S. aureus and its FemA mutant grown in the presence of an alanine-racemase inhibitor were labeled with D-[1-¹³C]alanine, L-[3-¹³C]alanine, [2-¹³C]glycine, and L-[5-¹⁹F]lysine to characterize some details of the peptidoglycan tertiary structure. Rotational-echo double-resonance (REDOR) NMR of isolated cell walls was used to measure internuclear distances between ¹³C-labeled alanines and ¹⁹F-labeled lysine incorporated in the peptidoglycan. The alanyl ¹³C labels in the parent strain were preselected for C{F} and C{P} REDOR measurement by their proximity to the glycine label using ¹³C-¹³C spin diffusion. The observed ¹³C-¹³C and ¹³C-³¹P distances are consistent with a tightly packed architecture containing only parallel stems in a repeating structural motif within the peptidoglycan. Dante selection of D-alanine and L-alanine frequencies followed by ¹³C-¹³C spin diffusion rules out scrambling of carbon labels. Cell walls of FemA were also labeled by a combination of D-[1-¹³C]alanine and L-[¹⁵N]alanine. Proximity of chains was measured by C{N} and N{C} REDOR distances and asymptotic plateaus, and both were consistent with a mixed-geometry model. Binding of an ¹⁹F-labeled eremomycin analogue in the FemA cell wall matches that of binding to the parent-strain cell wall and reveals the proximity of parallel stems in the alternating parallel-perpendicular mixed-geometry model for the FemA peptidoglycan lattice.

Introduction

We recently described the rotational-echo double-resonance analysis of the peptidoglycan of the cell walls of the FemA mutant of Staphylococcus aureus into which four specific labels were incorporated (Figure 1). Distances from the ¹⁹F label to the ¹³C labels (preselected by carbon-carbon spin diffusion) were interpreted by a model in which a peptidoglycan structural motif was a hybrid of tightly packed alternating parallel and perpendicular stems (1). The preselection was restricted to proximity to the glycyl-carbon label (see Figure 1). In this report we describe measurements in which the preselection site is moved to the D-alanyl and L-alanyl carbon-label sites. In addition, we contrast C{F} and C{P} REDOR results for short-bridge FemA with those for its long-bridge parent strain (which has an all-parallel-stem lattice) with respect to inter-stem proximities, location of wall teichoic acid, and preferred binding sites for a fluorine-labeled eremomycin glycopeptide antibiotic (2). These comparisons add detail to the lattice architecture of S. aureus which should be useful for future modeling.

Figure 1.

Location of the labels (red, ¹³C; green, ¹⁹F) of FemA peptidoglycan, and (red, ¹³C; yellow, ³¹P) of the repeating unit of wall teichoic acid.

Materials and Methods

Growth and labeling of FemA whole cells

Growth conditions were described before (3). The same method was used for both the parent strain of S. aureus (BB255) and its FemA mutant (UK17). To prevent the scrambling of L-[3-¹³C]alanine and D-[1-¹³C]alanine through alanine racemase, alaphosphin (L-alanyl-L-1-aminoethylphosphonic acid), an alanine-racemase inhibitor, was added to a final concentration of 10 μg/ml in two steps (3). The FemA and parent-strain whole cells were harvested after 6 hours of growth by centrifugation at 8,000 g for 10 min at 4 °C in a Sorvall GS-3 rotor. The cells were washed twice in 50 mL of ice-cold 40 mM triethanolamine buffer, pH 7.0, and then resuspended in 10 mL of 40 mM triethanolamine buffer and lyophilized. Cell walls were isolated as described in detail previously (3).

Dipolar recoupling

REDOR is a solid-state NMR method that recouples heteronuclear dipolar interactions under magic-angle spinning (4) and so can be used to determine inter-nuclear distances. REDOR is a difference experiment in which two spectra are collected, one in the absence of heteronuclear dipolar coupling (full echo, S₀ spectrum), and the other in the presence of the coupling (dephased echo, S spectrum). In the S₀ spectrum, dipolar dephasing is refocused due to spatial averaging resulting from motion of the rotor in magic-angle spinning. In the S spectrum, the spin part of the dipolar interaction is manipulated by the application of rotor-synchronized dephasing π-pulses to prevent full refocusing. The extent of the dephasing is related to the spin-pair dipolar coupling and hence the internuclear separation (4,5).

Solid-state NMR spectrometer and REDOR Pulse Sequence

Experiments were performed at 12 Tesla with a six-frequency transmission-line probe having a 12-mm long, 6-mm inner-diameter analytical coil, and a Chemagnetics/Varian ceramic spinning module. Samples were spun using a thin-wall Chemagnetics/Varian (Fort Collins, CO/Palo Alto, CA) 5-mm outer diameter-zirconia rotor at 7143 Hz, with the speed under active control and maintained to within ±2 Hz. A Tecmag Libra pulse programmer (Houston, TX) controlled the spectrometer. A 2-kW American Microwave Technology (AMT) power amplifier was used to produce radio-frequency pulses for ¹³C (125 MHz), and a 1-kW AMT amplifier for ³¹P (202 MHz). The ¹H (500 MHz) and ¹⁹F (470 MHz) radio-frequency pulses were generated by a 2-kW Creative Electronics tube amplifiers driven by 50-W AMT amplifiers. All final-stage amplifiers were under active control (6). The π-pulse lengths were 9 μs for ¹³C and ¹H, 6 μs for ³¹P, and 5 μs for ¹⁹F. Proton-carbon-matched cross-polarization transfers were made in 2 ms at 56 kHz. Proton dipolar decoupling was 100 kHz during data acquisition. The S and S₀ alternate-scan strategy compensated for short-term drifts in REDOR experiments. Standard XY-8 phase cycling (7) was used for all refocusing observe-channel π pulses (inserted at the end of each rotor period during dipolar evolution) and dephasing π pulses (inserted in the middle of each rotor period) to compensate for pulse imperfections. Frequency-specific ¹³C chemical shifts were selected prior to ¹³C{¹⁹F} REDOR experiments using rotor-asynchronous Dante irradiation, z-axis storage, and mixing times between 200 and 400 ms with no ¹H decoupling (8, 9). Typically, spectra from 100-mg cell-wall samples were the result of the accumulation of 16,384 scans at room temperature.

Calculated REDOR dephasing

REDOR dephasing was calculated using the modified Bessel function expressions given by Mueller et al. (10) and de la Caillerie and Fretigny (11) for a spin-½ pair. A plot of ΔS/S₀ with respect to time (t=NT_r), yields the dipolar coupling constant and hence the internuclear distance (r_IS). The distance and spin-pair concentration (asymptotic dephasing maximum) were allowed to vary to minimize the root-mean-square deviation between the experimental and calculated dephasing (12).

Results and Discussion

Glycyl Dante frequency selection for the parent strain of S. aureus

The long-bridge parent S. aureus strain has an intense Dante-inverted glycyl peak at 42 ppm (Figure 2). Spin diffusion in 200 ms from the glycyl label to the D-[1-¹³C]alanyl label arises only from the nearest-neighbor glycyl unit in the bridge. Two alanyl peaks are partially resolved: a cross-linked D-Ala-4 peptide peak at 175 ppm, and a weaker uncross-linked D-Ala-5 carboxyl-carbon shoulder at 178 ppm (1,2). The deconvoluted relative intensities are 4:1, which indictes that cross-linking is 80% for this cell-wall sample. Based on the lattice structure of the parent strain (2), the only cross-linked D-Ala ¹³C label within 5 Å of a lysyl fluorine is part of the same stem and this accounts for the minor ΔΔS/Δ S of 5% (Figure 2, left), a factor of 4 less than that observed for the FemA cell wall where inter-stem proximities are possible (1). The C{P} ΔΔS/ΔS is also small (Figure 2, right) and is associated only with the cross-linked D-Ala ¹³C label. This result is consistent with the fact that the surface of the peptidoglycan structural motif of the S. aureus parent strain displays mostly bridges (see Figure 9 of ref 1) with the uncross-linked sites buried within more hydrophobic regions (Figure 8 of ref 2).

Figure 2.

Dante-selected C{F} (left) and C{P} (right) REDOR of intact cell walls of the parent strain of S. aureus grown in media containing D-[1-¹³C]alanine, L-[3-¹³C]alanine, [2-¹³C]glycine, and L-[5-¹⁹F]lysine with the alanine racemase inhibitor, alaphosphin. Dante differencing with inversion of the glycyl-carbon peak (42 ppm) preceded REDOR dephasing. Four data blocks were collected resulting in spectra with and without Dante irradiation, each with and without ¹⁹F (or ³¹P) dephasing. The Dante differences (ΔS) are shown at the bottom of the figure and are the reference spectra for REDOR dephasing (ΔΔS) shown above. Spinning sidebands are designated by “ssb.”

D-alanyl and L-alanyl FemA Dante inversions

Dante inversions at 175 (Figure 3, bottom) and 178 ppm (Figure 3, top) both result in contacts of similar strengths with Gly and L-Ala ¹³C labels (Figure 4, left and right, respectively). However, the glycyl-label chemical shifts are different by 3.3 ppm. The unusual glycyl down-field shift (46.1 ppm) shows (i) the conformational strain associated with an uncross-linked stem (D-alanyl shift of 178 ppm) in the mixed-geometry lattice of FemA (see Figure 10 of ref 1); and (ii) the ability of Dante selection to make frequency-specific connectivities despite the broad lines of a heterogeneous cell wall.

Figure 3.

Dante frequency selection for the FemA cell-wall sample. A train of 1-μs ¹³C radio-frequency pulses separated by 5 μs, with the carrier frequency centered at 178 ppm (top) or 175 ppm (bottom), was followed by z-axis storage for 200 ms. The resulting spectra are designated as “S.”

Figure 4.

Dante difference spectra from the FemA cell-wall inversions of Figure 2 showing ¹³C-¹³C spin diffusion (for 200 ms) from the carbonyl-carbon label (175 ppm, left; 178 ppm, right) to the glycyl label (near 42 ppm) and the L-alanyl label (15 ppm). ΔS = S₀ – S. Spinning sidebands are designated by “ssb.”

The L-alanyl ¹³C methyl contacts are also of similar strengths with the two Dante D-Ala inversions, and this rules out the possibility of substantial D-Ala and L-Ala scrambling. Such scrambling would have resulted in D-[1-¹³C]Ala-4-D-[3-¹³C]Ala-5 pairs with a separation of only three bonds between ¹³C labels. These pairs would have resulted in a much stronger spin-diffusion contact for Dante inversion at 175 ppm than at 178 ppm.

The L-Ala methyl-carbon ¹³C peak shows an unusual low-field shoulder at 21.2 ppm (Figure 3, top). When this shoulder is Dante inverted (Figure 5, right), a stronger contact with the glycyl label is observed than for inversion of the main peak at 17.5 ppm (Figure 5, dotted lines). We attribute this result to strained conformations resulting from stems that are separated by only a few Ångstroms (Figure 10, ref 1) where the bridge of one stem is apparently forced near the L-alanyl head of a nearest-neighbor stem.

Figure 5.

Dante difference spectra from L-alanyl FemA cell-wall inversions showing ¹³C-¹³C spin diffusion (for 400 ms) from the methyl-carbon label (17.5 ppm, left; 21.2 ppm, right) to the glycyl label (near 42 ppm) and the D-alanyl label (near 175 ppm). Spinning sidebands are designated by “ssb.”

Contact of D-alanyl ¹³C-label with L-alanyl ¹⁵N-label in FemA cell walls

We attribute the minor contribution to the C{N} REDOR dephasing of Figure 6 (top left) that is associated with a one-bond distance (2) to natural-abundance carbonyl carbons near L-Ala ¹⁵N labels (see Figure 1, left). Some minor scrambling of ¹³C and ¹⁵N labels may also be present resulting in D-Ala-D-Ala ¹³C-¹⁵N bonds. The major component (3.4-Å ¹³C-¹⁵N separation) arises from the proximities of the heads and tails of adjacent parallel stems (Figure 7, top left and expanded inset). Both parent and FemA peptidoglycan lattices have high concentrations of locally parallel adjacent stems (1,2). N{C} REDOR dephasing reveals the same inter-stem contacts (Figure 6, bottom left). The asymptotic dephasing limit for the C{N} dephasing is 12% (½ × ½ × ½) and depends on the ¹⁵N isotopic enrichment (50%, 3), the fraction of D-Ala label in wall teichoic acid (50%, 13), and the fraction of D-Ala-D-Ala termini in conformationally strained locations pushed away from the stem head and towards the glycyl bridge (50%, 1). The N{C} asymptotic limit is twice as large (25%) because only 50% of the stem-head L-Ala ¹⁵Ns that are in stems not conformationally strained (50% of the total) are also near a D-Ala (Figure 9, ref 1).

Figure 6.

(Left) C{N} and N{C} REDOR dephasing (ΔS/S₀) of ¹⁹F-labeled LCTA-1110 complexed to cell walls of the FemA mutant of S. aureus grown on media containing D-[1-¹³C]alanine and L-[¹⁵N]alanine, with the alanine racemase inhibitor, alaphosphin (10 μg/ml), as a function of the dipolar evolution (open circles). The error in the integrated REDOR difference is estimated as the diameter of the open-circle symbols. (Right) C{F} and N{F} REDOR dephasing of the same sample (closed circles). The calculated dephasings for two single-distance components are shown in red and blue, and the combined dephasing curve (25% shorter-distance component and 75% longer-distance component), as a solid line.

Figure 7.

(Top left) A cross-section of the proposed peptidoglycan-tertiary structure for the FemA mutant of S. aureus (ref 1). The cross-section consists of sixteen glycan chains in a 4×4 matrix where the glycan backbones (represented by gray circles) are propagating perpendicular to the plane of the paper. The stems and bridges are represented by green and red rectangles, respectively. A cartoon of the glycopeptide LCTA-1110 (bottom right) is shown bound to a D-Ala-D-Ala uncross-linked peptide stem based on the lattice model of ref 1. Details of the highlighted region of the top-left panel are shown in the light-orange insert (bottom left). REDOR distance measurements (blue numbers) connect labels (red arrows). (Top right) Chemical structure of LCTA-1110.

Drug Binding

Adjacent parallel stems offer the same binding site for eremomycin analogues (14) in FemA (Figure 7) as in the parent strain of S. aureus (2). FemA has a mixture of parallel and perpendicular stems but LCTA-1110 prefers the parallel-stem sites (Figure 7, left). The REDOR-measured C-F distances of 4.8 and 8.8 Å from the ¹³C label of D-Ala to the ¹⁹F of LCTA-1110 (Figure 6, top right), combined with the 3.4-Å ¹⁵N-¹³C D-Ala-L-Ala interstem distance (Figure 6, left), combine for a calculation of 5.8 Å for the distance between the ¹⁵N label of L-Ala and the ¹⁹F of LCTA-1110 (Figure 7, bottom left inset). This value is in reasonable agreement with the observed REDOR distance of 4.8 Å (Figure 6, bottom right). We therefore conclude that the glycopeptide drug is bound in a hydrophobic region of the peptidoglycan lattice (1,2).

Abbreviations

¹³C{¹⁹F} or C{F}

carbon-channel observation with fluorine dephasing

¹³C{³¹P} or C{P}

carbon-channel observation with phosphorus dephasing

¹³C{¹⁵N} or C{N}

carbon-channel observation with nitrogen dephasing

¹⁵N{¹³C} or N{C}

nitrogen-channel observation with carbon dephasing

NAG

N-acetylglucosamine

NAM

N-acetylmuramic acid

REDOR

rotational-echo double resonance