On the quest for the elusive mechanism of action of daptomycin: Binding, fusion, and oligomerization

On the quest for the elusive mechanism of action of daptomycin: Binding, fusion, and oligomerization

BBA - Proteins and Proteomics 1865 (2017) 1490–1499 Contents lists available at ScienceDirect BBA - Proteins and Proteomics journal homepage: www.el...

1MB Sizes 49 Downloads 196 Views

BBA - Proteins and Proteomics 1865 (2017) 1490–1499

Contents lists available at ScienceDirect

BBA - Proteins and Proteomics journal homepage: www.elsevier.com/locate/bbapap

On the quest for the elusive mechanism of action of daptomycin: Binding, fusion, and oligomerization

MARK

Jin Zhanga, Walter R.P. Scotta, Frank Gabelb,c, Miao Wua, Ruqaiba Desmonda, JungHwan Baea, Giuseppe Zaccaic, W. Russ Algara,⁎, Suzana K. Strausa,⁎ a b c

Department of Chemistry, University of British Columbia, 2036 Main Mall, Vancouver, BC V6T 1Z1, Canada Institut de Biologie Structurale, 41 rue Jules Horowitz, 38027 Grenoble Cedex 1, France Institut Laue Langevin, 6 rue Jules Horowitz, 38042 Grenoble Cedex 9, France

A R T I C L E I N F O

A B S T R A C T

Keywords: Daptomycin Lipopeptide SANS PCS Membrane fusion FRET MD simulations

Daptomycin, sold under the trade name CUBICIN, is the first lipopeptide antibiotic to be approved for use against Gram-positive organisms, including a number of highly resistant species. Over the last few decades, a number of studies have tried to pinpoint the mechanism of action of daptomycin. These proposed modes of action often have points in common (e.g. the requirement for Ca2 + and lipid membranes containing a high proportion of phosphatidylglycerol (PG) headgroups), but also points of divergence (e.g. oligomerization in solution and in membranes, membrane perturbation vs. inhibition of cell envelope synthesis). In this study, we investigate how concentration effects may have an impact on the interpretation of the biophysical data used to support a given mechanism of action. Results obtained from small angle neutron scattering (SANS) experiments and molecular dynamics (MD) simulations show that daptomycin oligomerizes at high concentrations (both with and without Ca2 +) in solution, but that this oligomer readily falls apart. Photon correlation spectroscopy (PCS) experiments demonstrate that daptomycin causes fusion more readily in DMPC/PG membranes than in POPC/PG, suggesting that the latter may be a better model system. Finally, fluorescence and Förster resonance energy transfer (FRET) experiments reveal that daptomycin binds strongly to the lipid membrane and that oligomerization occurs in a concentration-dependent manner. The combined experiments provide an improved framework for more general and rigorous biophysical studies toward understanding the elusive mechanism of action of daptomycin. This article is part of a Special Issue entitled: Biophysics in Canada, edited by Lewis Kay, John Baenziger, Albert Berghuis and Peter Tieleman.

1. Introduction Daptomycin is a lipopeptide antibiotic composed of 13 amino acid residues (Fig. 1), approved in 2003 for use against Gram-positive organisms, including a number of highly resistant species [1–3]. Among the amino acid residues are three D-amino acids (D-asparagine, D-alanine, and D-serine) and three uncommon amino acids, the latter including ornithine, (2S,3R)-3-methyl-glutamic acid, and kynurenine. The N-terminus of daptomycin is acylated with a n-decanoyl fatty acid chain [4]. Daptomycin is one of the few available antibiotics that are effective against many resistant bacterial strains, including methicillinresistant Staphylococcus aureus (MRSA) [5], vancomycin-intermediate Staphylococcus aureus (VISA), vancomycin-resistant Staphylococcus aureus (VRSA), and vancomycin-resistant Enterococci (VRE) [6]. Its antimicrobial activity is entirely dependent on the presence of calcium ions (Ca2 +) and lipids with negatively charged headgroups (e.g., ⁎

Corresponding authors. E-mail addresses: [email protected] (W.R. Algar), [email protected] (S.K. Straus).

http://dx.doi.org/10.1016/j.bbapap.2017.07.020 Received 28 March 2017; Received in revised form 5 July 2017; Accepted 31 July 2017 Available online 24 August 2017 1570-9639/ © 2017 Elsevier B.V. All rights reserved.

phosphatidylglycerol (PG), as found in bacteria) [1–3,7–9]. Despite the widespread use of daptomycin in hospital and clinical settings, the complete mechanism of its action is still not fully understood and is hence the subject of recent investigations [3,10–13]. Over the last few decades, several biophysical and bacterial cell studies have helped to piece together key steps in the mode of action of daptomycin. Early clues pointed to daptomycin's ability to affect peptidoglycan biosynthesis [14,15], but this was then refuted [16], only for blockage of cell wall synthesis to gain in importance again recently [3]. Likewise, it was suggested that daptomycin functions by membrane depolarization [7] or pore formation [17], but evidence to the contrary has also been published [3,18–21]. Finally, other studies have shown that daptomycin causes membrane deformation, leading to aberrant recruitment of cell-wall proteins such as DivIVA [22]. But again, other evidence [3] suggests otherwise. As recently discussed [3], this diversity of explanations regarding the mode of action of daptomycin may

BBA - Proteins and Proteomics 1865 (2017) 1490–1499

J. Zhang et al.

of biophysical data used to elucidate the mode of action of daptomycin (Fig. 2). For instance, if one wants to determine how daptomycin binds (steps 1–2) and oligomerizes (steps 3–4) in the membrane (Fig. 2A), then fusion (Fig. 2B, shown for the hemifusion step here only) could introduce anomalies or lead to erroneous interpretation of data. Previous work by Jung et al. [23] showed that fusion occurs only when a combination of daptomycin, Ca2 + and a phosphatidylglycerol-containing model membrane is used. Interestingly, this phenomenon occurs for low concentrations of daptomycin that are close to its MIC (i.e. 8 μg/ mL or 5 μM). Recently, a number of fluorescence studies [10–12] proposed that daptomycin forms well-defined membrane-bound oligomers consisting of 6 or 7 daptomycin molecules, with Ca2 + binding occurring in two steps [10]. These biophysical studies used conditions similar to those in Jung et al. where fusion occurred [23], leading us to question whether the choice of conditions has an impact on the resulting model. In this work, we have examined daptomycin's concentration-dependent properties in solution and in model membranes. To assess the potential importance of solution-phase micellar structures of daptomycin, we have used a unique combined approach of small angle neutron scattering (SANS) and molecular dynamics (MD) simulation in order to probe the size and shape of the oligomer and the packing of the daptomycin molecules within it. Unrestrained MD simulations of the oligomer provide further insight into the evolution of the micellar structure under conditions where it is diluted. Aggregation was also tested in solution in the presence of 150 mM NaCl using 1H NMR spectroscopy. We have also systematically determined conditions under which fusion occurs using photon correlation spectroscopy (PCS), and whether or not the fusion state persisted after dilution, for a number of model membrane systems: 1,2-dimyristoyl-snglycero-3-phosphocholine (DMPC)/1,2-dimyristoyl-sn-glycero-3-phospho(1′-rac-glycerol) (DMPG) (50%/50%), 1:1 1-palmitoyl-2-oleoyl-sn-glycero3-phosphocholine (POPC)/1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-(1′rac-glycerol) (POPG) (50%/50%), POPC/POPG/cardiolipin (CL) (40%/ 50%/10%), and finally POPC alone (100%). Fluorescence experiments were conducted to quantify the binding of daptomycin to DMPC/PG and POPC/POPG membranes under conditions where fusion does not occur. Finally, FRET experiments were conducted using daptomycin and NBDdaptomycin, again under conditions where there is no fusion, to investigate if daptomycin forms oligomers in a concentration-dependent manner. Taken together, these experiments provide an improved framework for more general and rigorous biophysical studies toward understanding the elusive mechanism of action of daptomycin, and are discussed in this context.

Fig. 1. Chemical structure of daptomycin, drawn at acidic pH. At neutral pH, the 3 Asp and 1 mGlu have negatively charged side-chains, resulting in a net − 3 charge for daptomycin.

be linked to the variety of methods used to elucidate the mechanism. As with all scientific endeavours, every experiment has led to new models being proposed and a deeper understanding of how this lipopeptide antibiotic might function. However, one key element, which also plays a part, particularly in biophysical studies, is daptomycin's concentration-dependent properties—namely its ability to form aggregates in solution and in membranes, and under particular conditions, to cause membrane fusion [23]. The propensity for daptomycin to form aggregates in solution was first discovered when groups tried to solve its structure by NMR [24–26]. Indeed, Ball et al. [24] and Rotondi and Gierasch [26] found sample conditions where line-broadening in the 1H NMR spectra was evident. Hence, Rotondi and Gierasch [26] carefully optimized sample preparation conditions to produce an aggregate-free sample through the slow addition of daptomycin to a well-degassed solution of 10 mM sodium phosphate buffer, pH 5.3 [26]. As detailed in the recent review by Taylor and Palmer [2], it is the ionization state of daptomycin that plays a role, with samples prepared at neutral pH [25,27] showing no or minimal aggregation, if the daptomycin concentration is at millimolar concentrations or lower. Upon addition of Ca2 +, Ball et al. [24] and Jung et al. [25] both demonstrated that line-broadening occurs, suggesting aggregation of the sample. Nevertheless, Jung et al. (and not Straus et al., as stated by Taylor and Palmer [2]) determined the structure of a Ca2 +-bound form of daptomycin, assuming that all longrange Nuclear Overhauser Effect distance restraints (NOEs) found were intramolecular. It was Buncókzi et al. [28] who then first suggested that some of these NOEs might be intermolecular. Ho et al. [29] then reanalyzed the Ca2 +-bound structure of daptomycin and also investigated the aggregate formed in detail. Using analytical ultracentrifugation [29], they1 found that daptomycin forms aggregates consisting of 14–16 daptomycin molecules in the presence of one equivalent of Ca2 +. This finding led to the suggestion that daptomycin forms a micellar structure and that this aggregated form may be important for the mode of action of daptomycin [1,24]. This view was further supported by fluorescence measurements of daptomycin (at concentration of 60 μM and above, in the presence of 1 mM Ca2 +) [27]. However, other fluorescence measurements by Muraih et al. [12] with daptomycin concentrations of 5–10 μM (i.e. close to the minimal inhibitory concentration (MIC) of 0.02–5 μM) and calcium suggest that solution-phase oligomerization is not a key step in the mechanism of action of daptomycin. As mentioned above, daptomycin has been found to cause membrane fusion under certain conditions [23]. Since a peptidoglycan layer surrounds Gram-positive bacteria, the primary targets of daptomycin, membrane fusion is unlikely to be important in the mechanism of action of daptomycin. It could, however, have an impact on the interpretation

2. Materials and methods 2.1. Materials Daptomycin was a generous gift from Cubist Pharmaceuticals (Lexington, MA). 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-(1′-rac-glycerol) (POPG), 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC), 1,2-dimyristoyl-sn-glycero-3-phospho-(1′-rac-glycerol) (DMPG), 1′,3′-bis[1,2dimyristoyl-sn-glycero-3-phospho]-sn-glycerol (CL) were from Avanti Polar Lipids, Inc. (Alabaster, AL) and used without further purification. 4-Chloro-7-nitrobenzofurazan (NBD-Cl), 4-(2-hydroxyethyl)piperazine1-ethanesulfonic acid (HEPES), and acetonitrile were from SigmaAldrich (St. Louis, MO). Sodium borate buffer (0.5 M, pH 8.0) was obtained from Alfa Aesar (Ward Hill, MA). Acetic acid, ammonium acetate, and ethylenediaminetetraacetic acid (EDTA) were from Fisher Scientific (Fair Lawn, NJ). 2.2. Preparation of NDB-daptomycin Daptomycin was labeled with NBD at the ornithine side-chain (6th residue, Fig. 1), as reported by Muraih et al. [12]. Daptomycin

1 It should be noted that in Taylor and Palmer, the ultracentrifugation data is referenced as being reported in Jung et al. [23], but it was reported in Ho et al. [29].

1491

BBA - Proteins and Proteomics 1865 (2017) 1490–1499

J. Zhang et al.

A

B

1

1

2

2

3

3

4

4

Fig. 2. Comparison of oligomerization in biophysical studies, where (A) there is no membrane fusion; and (B) where membrane fusion occurs (shown here only up to the hemifusion step). Steps 1–2: binding and insertion of individual daptomycin molecules (red) into the lipid bilayer (yellow/green). Steps 3–4: (A) oligomerization and pore formation; or (B) oligomerization and fusion.

using molecular dynamics simulations can be found in the SM. Further tests of the oligomer formation in the presence of NaCl were done using 1 H NMR (Fig. S1) and are also described in the SM.

(0.6 mM) was dissolved in 50 mM sodium borate buffer containing 20 mM EDTA, pH 8.0, and NBD-Cl (25 mM) was dissolved in acetonitrile. NBD-Cl solution (5 mL) was added to the daptomycin solution (15 mL) in a round-bottom flask, the reaction was stirred at 60 °C for 5 h, and then cooled in an ice water bath for 2 min. Acetic acid (50 mM, 20 mL) was added to terminate the labeling reaction. The crude product was purified using semi-preparative HPLC on a Waters 600 system (Waters Ltd., Mississauga, ON, Canada) equipped with Waters 2996 Photodiode Array Detector. The separation was performed on a Phenomenex C18 reversed-phase column (250 mm × 21.2 mm) using a mobile phase of acetonitrile with 20 mM ammonium acetate buffer (pH 5.5) at a flow rate of 10 mL/min. The gradient started at 30% acetonitrile, increased to 40% in 5 min, and then was kept constant up until 35 min. Fractions were collected and identified by spectrophotometry, fluorescence spectroscopy, and mass spectrometry. The molecular weight of the purified NBD-daptomycin determined by MALDI-TOF was 1783.9 g/mol, in agreement with the calculated molecular weight of 1783.77 g/mol.

2.4. Preparation of vesicles Vesicles composed of the desired lipids were prepared by evaporating the organic solvent under a dry nitrogen stream. The resulting lipid film was thoroughly dried under vacuum overnight, then hydrated with 20 mM HEPES buffer (pH 7.5) to a final total lipid concentration of 260 μM. After five freeze-thaw cycles, the lipid suspension was extruded 10 times through two stacked 50 nm polycarbonate filters using a nitrogen-pressurized liposome extruder. 2.5. Photon correlation spectroscopy The sizes of LUVs were determined using a Beckman Coulter N4 plus particle size analyzer (Mississauga, ON, Canada). Daptomycin or NBDdaptomycin and CaCl2 were added to the extruded liposomes to reach the final concentrations indicated in the results section. The sizes of vesicles were measured from 600 nm laser light scattered at a 90° angle after incubation of the samples for 15 min at 23 °C.

2.3. SANS experiments and molecular dynamics simulations SANS experiments were performed on the small angle diffractometer D22 at the Institut Laue-Langevin (ILL) (Grenoble, France), using samples prepared as described in the Supplementary Material (SM). The wavelength of the neutrons was set at λ = 6 Å. Sample volumes were adjusted to 200 μL and samples were transferred into quartz cells with a 1 mm path-length (Hellma, Müllheim, Germany). All datasets (samples and buffers, boron standard, empty quartz cell) were collected at a detector distance of 2.0 m. Acquisition times varied between 5 min (boron) to 60 min (samples). Transmissions were measured for 2 min. Further details on SANS data analysis can be found in SM. An overall shape for the daptomycin oligomer in solution was determined from SANS data. The result was an approximately ellipsoid shape (vide infra), from which an additional refinement potential energy function was derived and implemented in GROMOS [30]. Details of the complete refinement of the daptomycin oligomeric structure

2.6. Binding isotherms Binding of peptides to LUVs was measured fluorimetrically with the successive addition of different concentrations of lipid vesicles. Aliquots of a concentrated DMPC/DMPG or POPC/POPG (1:1) stock solution (6.5 mM) were added to a daptomycin solution (2 mL, 4 μM) containing 420 μM Ca2 + in HEPES buffer (20 mM, pH 7.5). The solution was stirred for 5 min to reach equilibrium after each addition of liposomes. These concentrations of reagents ensured that no fusion occurred between lipid vesicles. Although the concentration of Ca2 + was below physiological concentration (the concentration of Ca2 + in serum is 1.0–1.2 mM [31]), it was a factor of 10 larger than required for saturation of daptomycin binding. Saturation binding was measured as a function of Ca2 + concentration with 4 μM daptomycin and 260 μM 1492

BBA - Proteins and Proteomics 1865 (2017) 1490–1499

J. Zhang et al.

2.9. FRET experiments

POPC/PG (see Fig. S2). Fluorescence emission spectra were measured on a Perkin Elmer LS50B luminescence spectrometer between 400–600 nm, with an excitation wavelength of 365 nm, and emission and excitation slit widths set to correspond to a 5 nm bandwidth. Inner filter effects were negligible.

Fluorescence emission spectra were obtained with an Infinite M1000 fluorescence plate reader (Tecan US Inc., Morrisville, NC). Samples were prepared in HEPES buffer (20 mM, pH 7.5) with extruded liposome (260 μM total lipid), CaCl2, and daptomycin or NBD-daptomycin alone or together in different proportions and concentrations. For the premixed samples, daptomycin and NBD-daptomycin were mixed before adding the liposomes and calcium. In the sequential samples, a solution containing daptomycin, calcium, and liposomes was incubated for 30 min prior to the addition of NBD-daptomycin. The excitation wavelength was 365 nm, and emission was recorded between 400–600 nm. Bandwidths for excitation and emission were 5 nm.

2.7. Binding kinetics To generate progress curves for binding of peptide to liposomes, a time-series of fluorescence images were obtained using an Olympus IX83 inverted epifluorescence microscope equipped with a 4×/0.16 (magnification/NA) objective lens, a 120XL X-Cite metal-halide light source, and an sCMOS camera (ORCA-Flash4.0, Hamamatsu Photonics, Hamamatsu, Japan). Daptomycin and Ca2 + were premixed in a well in a microtiter plate and a time-series of 17 images was acquired to determine the baseline fluorescence intensity. Liposome was then added and a series of 100 images was acquired to measure the increase in kynurenine or NBD fluorescence upon binding to liposomes. The time interval for imaging was 0.58 s for daptomycin samples and 0.66 s for NBD-daptomycin samples. A 350/20 excitation filter and a 460/50 emission filter (centre wavelength/bandwidth; Chroma, Bellows Falls, Vermont, USA) were used for measurements of fluorescence from the kynurenine residue of daptomycin. For measurement of NBD fluorescence, a 450/20 (excitation) and 540/50 (emission) filter combination was used. Inner filter effects were negligible. Images were exported using MetaMorph software (Molecular Devices, Sunnyvale, CA, USA) and analyzed in ImageJ software (National Institutes of Health, Bethesda, MD, USA). Two daptomycin concentrations were tested: 1.4 μM and 3.4 μM (17.5 μL volume), with the corresponding final concentrations of Ca2 + of 180 μM and 420 μM, respectively (35 μL volume). These concentrations avoided vesicle fusion. The final concentration of lipids was 260 μM, and the volume was 47.5 μL. Each concentration was repeated three times.

3. Results 3.1. Daptomycin micellar structure and its fate in solution Previous work has shown that daptomycin forms well-defined aggregates in solution when one equivalent of Ca2 + is added and that this oligomer consists of 14–16 daptomycin molecules [29]. In order to determine the size and shape of this oligomer, samples were prepared in 10 mM phosphate buffer, pH 7, and SANS experiments run. The resulting SANS curves, shown in Fig. 3A, were fitted to extract the radii of gyration, Rg, and I(0) intensities. The fits revealed that daptomycin underwent an aggregation process in the absence of Ca2 + (aq) between 0.6–3 mM concentration (Fig. 3A, left panel). At 0.6 mM, Rg was 11 ± 2 Å, which was in good agreement with the theoretical Rg value of 12.5 Å for monomeric daptomycin. The SANS curve at this concentration was also compatible with the theoretical one calculated from a single daptomycin conformation using CRYSON (Fig. 3B, see SM). Between 0.6 mM and 3.0 mM, the SANS curves underwent a drastic change in shape and intensity, and an Rg value could not be determined for 3.0 mM and 6.0 mM samples because of inter-particle effects, probably due to unscreened electrostatic interactions. In the presence of 5× calcium (per weight, i.e. for 0.6 mM daptomycin, 45 mM CaCl2; see SM for full details), the Rg values did not vary significantly between concentrations of 0.6, 3.0, and 6.0 mM daptomycin, with an average Rg of 23 ± 2 Å, and I(0) increased approximately linearly with concentration (0.037, 0.186, 0.330). Assuming that only daptomycin monomers were present in solution at 0.6 mM in the absence of calcium, the aggregation number of the particles in the presence of calcium can be estimated (see SM for details). With the additional assumption that no free monomers are present in solution in these conditions, an aggregation number Nagg of > 8–9 was obtained. This finding is in general agreement with the aggregate sizes of 14–16 found from AUC data [29]. As there is no specific evidence of a more complex equilibrium than the simple monomer ↔ aggregate one used here, a more precise aggregation number cannot be obtained from the SANS data. The scattering curves for all measured concentrations of daptomycin can be superposed (Fig. 3A, right panel), showing that the same particle shape was preserved in the range from 0.60–6.0 mM daptomycin. Maximum particle diameters were estimated to be ~60 Å using the program GNOM (see Fig. S3; see SM for details of analysis). Based on the above data, the low-resolution envelopes of the daptomycin aggregates in the presence of Ca2 + were calculated using the programs DAMMIN [36] (Fig. 3C) and DAMMIF [37] (Fig. S4), and the average structures extracted with DAMAVER [38]. Starting from the envelopes shown in Fig. 3C, a model consisting of 14 daptomycin monomers was created using CHIMERA [39–43] and allowed to evolve using molecular dynamics simulations. Fourteen daptomycin monomers were chosen instead of the 8–9 estimated by SANS because this represents a lower boundary, and 14 represents a situation with a higher number of intermolecular interactions that could help hold a complex together. Interestingly, when the system was constrained by the 21 Å × 16 Å ellipsoid estimated from SANS, the 14

2.8. Analysis of binding data The fraction of lipid-bound peptide, fb, was determined from Eq. (1), where F is the measured fluorescence intensity, F0 is the initial fluorescence intensity in the absence of lipid, and Fmax is the saturation fluorescence intensity.

fb =

F − F0 Fmax − F0

(1)

With assumption of a simple two-state model for bound and unbound peptide (a useful first approximation with a large excess of calcium), the lipid-bound peptide was related to the apparent dissociation constant, Kd, through Eq. (2) [32]. Kd was calculated from plots of fb versus the lipid concentration.

fb =

[lipid] K d + [lipid]

(2)

Binding data was also used to determine the partition coefficient, Kp, using Eq. (3) [33], where Xb represents the molar ratio of bound peptide to total lipid and Cf denotes the equilibrium concentration of the unbound peptide.

Xb = Kp Cf

(3)

To determine Xb and Cf, the fraction of the membrane bound peptide fb was calculated using Eq. (1). Knowing the fraction of bound peptide, fb, the Cf and Xb values could be calculated. In practice, Xb was divided by a correction factor of 0.6 since it was assumed that peptides were initially bound to the outer leaflet of the LUV (60% the total lipids) [34]. The partition coefficients were determined from the initial slopes of binding isotherm curves [34,35]. 1493

BBA - Proteins and Proteomics 1865 (2017) 1490–1499

J. Zhang et al.

A

B

C

40 Å

D

E

Fig. 3. SANS data on daptomycin: (A) in the absence (left panel) and presence (right panel) of a 5-fold excess of calcium (per weight) (or 1:75 molar ratio, see Supplemental material for details), as a function of daptomycin concentration. (B) Calculated SANS curve obtained from a single daptomycin conformation using CRYSON, using the two daptomycin PDB structures 1T5N and 1T5M. (C) Low-resolution envelopes of the aggregates were calculated using the programs DAMMIN (for those obtained with DAMMIF, see Fig. S4). (D) MD simulation results of 14 daptomycin molecules restrained within an ellipsoid derived from the shapes in (C). (E) MD simulation results of the model in (D), with the restraints removed (run in methanol for 10 ns, for water simulation results see Fig. S5 and details in Supplemental Material).

daptomycin in the presence of Ca2 + is driven by electrostatic interactions and that screening of the negative daptomycin charges by Na+ minimizes this effect (Fig. S1d).

daptomycin monomers did not pack to form an organized micelle, but rather formed a loose arrangement (Fig. 3D). When the constraints were removed, the molecules drifted apart (Figs. 3E and S5). Similar behaviour may be expected for a dilute solution of daptomycin. Finally, to test whether aggregation is important under physiological conditions (i.e. in the presence of NaCl), 1H NMR experiments were run. As can be seen in Fig. S1, addition of ca. 5 fold CaCl2 (molar ratio) causes extensive line broadening (Fig. S1b), as observed previously [24,25]. Interestingly, addition of 150 mM NaCl has no effect on the linewidths (Fig. S1c). Finally, addition of both CaCl2 (5 mM) and NaCl (150 mM) results in lines that are broadened, but not to the same extent as CaCl2 alone. This suggests that oligomer formation of

3.2. Vesicle fusion As a secondary effect of binding a lipid membrane, Jung et al. have shown that daptomycin can fuse lipid vesicles when the vesicles are composed of a proportion of PG headgroups and when Ca2 + (aq) is present [23]. To determine under what conditions fusion occurs when the daptomycin concentration is close to MIC (0.02–5 μM), we measured the size of lipid vesicles using photon correlation spectroscopy at 1494

BBA - Proteins and Proteomics 1865 (2017) 1490–1499

J. Zhang et al.

Fig. 6. Binding kinetics of daptomycin to lipid membranes: (A) native daptomycin; (B) NDB-labeled daptomycin. Data for concentrations of daptomycin are shown. The lipid concentration was 260 μM. The arrow indicates the time at which daptomycin was added to the lipid.

Fig. 4. Fusion of lipid vesicles and their persistence. (A) Lipid vesicle size as a function of Ca2 + (aq) concentration for DMPC/DMPG, POPC/POPG/CL, POPC/POPG, POPC. The daptomycin and total lipid concentrations were 6.3 μM and 260 μM, respectively. Data for NBD-daptomycin with DMPC/DMPG lipids is also shown. (B) Size of daptomycin-fused POPC/POPG, POPC/POPG/CL, and DMPC/DMPG vesicles with successive 2-fold and 10fold dilutions relative to the initial concentration. The arrows in panel A indicate the samples that were used for experiments in panel B.

4

lipids were present. The latter result was consistent with the findings by Jung et al. [23], and further suggested that PG is required for daptomycin to bind to lipid membranes. It should also be noted that higher lipid concentrations delayed the onset of fusion (i.e., higher concentrations of Ca2 + (aq) are required, data not shown), and that fusion was independent of whether the solvent was water or buffer with 150 mM NaCl. In a second set of experiments, POPC/POPG, POPC/POPG/CL, and DMPC/DMPG vesicles were fused in presence of Ca2 + (aq), then serially diluted with water and the size re-measured by PCS. Fig. 4B shows that fusion persisted even after dilution. Whereas vesicles of POPC/ POPG and POPC/POPG/CL were consistent in size with dilution, there was a decrease in the size of DMPC/DMPG vesicles with dilution; however, the absolute sizes remained consistent with fused vesicles. Studies in the literature which investigated the oligomerization of daptomycin in membranes were typically carried out using e.g. 4.8 μM daptomycin + 0.91 μM NBD-daptomycin, in 250 μM DMPC/PG (1:1) and 5 mM calcium [11,12]. According to Fig. 4A, under these conditions, it is expected that membrane fusion would occur. Moreover, concentration effects were not considered. We therefore used similar fluorescence measurements to investigate daptomycin binding and oligomerization, but under conditions where fusion did not occur and as a function of concentration.

-1

(10 M )

3.3. Membrane binding and oligomerization of daptomycin under nonfusion conditions Fig. 5. (A) Representative isotherms for binding of Dap and NBD-Dap to DMPC/PG and POPC/PG vesicles. The inset shows a closer view of the data at low concentrations of lipid. (B) Summary of apparent dissociation constants, Kd, and partition coefficients, Kp, determined from the data. For Dap and DMPC/PG, the uncertainty is the standard deviation of the average for three replicate measurements.

3.3.1. Binding of daptomycin to the membrane The binding of native and NBD-labeled daptomycin to liposomes was measured from enhancement of kynurenine and NBD fluorescence in the lipid microenvironment [44]. The resultant binding isotherms are shown in Fig. 5 for DMPC/PG and POPC/PG vesicles. The data shows that binding of lipids by Dap is strongly favoured and, for a system consisting of 4.8 μM daptomycin in ca. 250 μM total lipid [12], virtually all of the daptomycin was expected to be bound. Apparent dissociation constants, Kd, and partition coefficients, Kp, were estimated from the data, and were on the order of Kd = 101 μM, and Kp = 104–105 M− 1. Interestingly, the binding of NBD-Dap to DMPC/PG vesicles appeared to be approximately 5–6-fold weaker than for native Dap, but had similar affinity for POPC/PG vesicles. Binding kinetics were also characterized using fluorescence measurements. Fig. 6 shows binding progress curves for daptomycin and

daptomycin concentrations of 6.3 μM, total lipid concentrations of 260 μM, and Ca2 + (aq) concentrations in the range of 0.2–7 mM. The lipid vesicle compositions were DMPC/DMPG (50%/50%), POPC (100%), POPC/POPG (50%/50%), and POPC/POPG/CL (40%/50%/ 10%). As shown in Fig. 4A, fusion occurred most rapidly for DMPC/ DMPG, where native daptomycin was a more effective fusion peptide than NBD-labeled daptomycin (NBD-Dap). Higher Ca2 + (aq) concentrations were required to initiate vesicle fusion with NBD-Dap. Lipids with palmitoyl/oleoyl acyl chains also required higher Ca2 + (aq) concentrations for the onset of fusion, which did not occur when no PG 1495

BBA - Proteins and Proteomics 1865 (2017) 1490–1499

J. Zhang et al.

Fig. 7. Kynurenine (Kyn) and NBD fluorescence at different total daptomycin concentrations for daptomycin alone and either sequentially-mixed or pre-mixed solutions with a 5:1 ratio of native daptomycin (Dap) and NBD-labeled daptomycin (NBD-Dap). (A) Change in the NBD/Kyn fluorescence intensity ratio as a function of total daptomycin concentration. The inset shows the Kyn fluorescence relative to a control sample without NBD-Dap. The data is an average of two replicate experiments. (B) Representative fluorescence spectra for some of the data points in panel A. The total daptomycin concentration is indicated. The inset in the bottom-right graph shows the individual contributions of Kyn and NBD to the observed fluorescence spectrum.

nor sensitization of NBD fluorescence was significant at the lowest concentration of total daptomycin. With increases in the total daptomycin concentration, progressive quenching of kynurenine fluorescence and sensitization of NBD fluorescence was observed. Notably, the samples with sequential addition of daptomycin and NBD-daptomycin yielded consistently lower NBD/kynurenine emission ratios than the analogous pre-mixed samples. When a pre-mixed sample was left for an extended period of time (~ 24 h), it was found that the relative quenching of the kynurenine fluorescence for sequential samples approached that of the pre-mixed samples (Fig. S6), suggesting slow reversibility. The foregoing FRET results suggest that daptomycin can bind to liposomes without oligomerizing at concentrations in the low MIC range, i.e. 0.02–0.75 μM. The extent of oligomerization appears to increase as the concentration of total daptomycin, and thus the amount of daptomycin bound to liposomes, increases. When this experiment is repeated with two-stage addition of daptomycin (first) and NBD-daptomycin (second) (i.e., sequential addition versus pre-mixed addition), less quenching of native kynurenine fluorescence and less sensitization of NBD fluorescence was observed (i.e., less FRET). This result suggests that there is no significant re-equilibration of the oligomers over the timescale of the experiment (ca. 30 min), and that the fraction of unoligomerized daptomycin forms new oligomers with some of the added NDB-daptomycin (if oligomers have fixed stoichiometry), or that there is an increase in aggregate size and/or number with the added NBDdaptomycin (if oligomers do not have fixed stoichiometry). Ensemble fluorescence measurements cannot separate these two possibilities, which could also occur concurrently.

NBD-Dap with DMPC/PG vesicles. Initial rates were estimated from mathematical fits to the full progress curves (see SM for details), and were converted into binding rate constants (assuming constant [Ca2 +] (aq)). For daptomycin, the forward rate constant was estimated to be 3.1 ± 0.8 mM− 1 s− 1, versus 2.1 ± 0.2 mM− 1 s− 1 for NBD-Dap. Additional analysis of the progress curves in terms of a simple two-state equilibrium model returned estimates of forward rates constants that were 1.4 mM− 1 s− 1 and 0.50 mM− 1 s− 1, and corresponding Kd values of 15 ± 3 μM and 80 ± 53 μM, which were in good agreement with the results from binding experiments (Fig. 5). 3.3.2. Oligomerization of daptomycin in vesicular membranes To study the oligomerization of daptomycin, we adopted an NBDdaptomycin FRET system reported by Muraih and coworkers [11,12]. The kynurenine residue of the daptomycin functions as a donor for an NBD acceptor label attached to the ornithine residue. As discussed later, this FRET system is not ideal; however, the nature of daptomycin imposes several restrictions and this FRET system functions adequately. The general concept is that the kynurenine residue of a native daptomycin molecule can act as a FRET donor and transfer its energy to an NBD label on an NBD-daptomycin molecule when these two molecules are part of an oligomer. In the absence of oligomerization, this intermolecular FRET pathway is not available. Daptomycin oligomerization was studied by mixing a fixed amount of lipid (260 μM) with increasing amounts of total daptomycin (0.72–4.80 μM), where the total daptomycin was a 5:1 mixture of native daptomycin and NBD-daptomycin. The Ca2 + concentration (90–600 μM) scaled with the total concentration of daptomycin (125 ×, molar ratio). Analogous to the original study by Muraih et al. [12], we completed two variations of this experiment: (i) the daptomycin and NBD-daptomycin were pre-mixed and added to the lipid at the same time; and (ii) the daptomycin and NBD-daptomycin were added to the lipid sequentially, with an equilibration time (30 min) in between. Our experiments differed from the original experiments by Muraih et al. in that (i) the calcium concentration was kept sufficiently low to avoid vesicle fusion, (ii) the concentration of total daptomycin was varied, and (iii) FRET was analyzed in terms of the NBD/kynurenine fluorescence ratio rather than just quenching of kynurenine fluorescence. The importance of these differences is discussed later. Fig. 7 shows how FRET between the kynurenine and NBD increased as a function of total daptomycin concentration for the pre-mixed and sequential experiments. The extent of FRET was measured as the NBD/ kynurenine emission ratio, which increases with greater energy transfer. From the binding experiments, it was expected that ~96% of the total daptomycin was bound to vesicles at the lowest total concentration studied (0.72 μM), with larger fractions bound at higher concentrations. Neither quenching of native kynurenine fluorescence

4. Discussion and conclusions Despite being discovered in the late 1980s, daptomycin, originally designated as LY 146032, has a mode of action that has remained elusive, even as the subject of many studies [3,7,10,27,44] (most recently reviewed in Taylor and Palmer [2]), including very recent ones [3,10]. Among the many studies, the common points are that daptomycin requires calcium and phosphatidylglycerol membranes for activity. There are, however, also many points of divergence in the proposed mechanisms and many outstanding questions: e.g. whether daptomycin forms aggregates in solution so that the resulting micellar structure delivers a large concentration to the membrane; and, whether daptomycin forms well-defined oligomers in the membrane, resulting in pore formation, potassium leakage, membrane depolarization, and consequently cell death. In this contribution, we have examined how the concentration-dependent behaviour of daptomycin affects interpretation of biophysical data and the impact on the resulting mechanism of action models. 1496

BBA - Proteins and Proteomics 1865 (2017) 1490–1499

J. Zhang et al.

without oligomerizing at low MIC concentrations (0.02–0.75 μM daptomycin). At higher concentrations (i.e. 0.8–5 μM daptomycin), the extent of oligomerization appeared concentration dependent. These results are in contrast to what was proposed by Muraih et al. [11], where the oligomerization model assumed that only one type of oligomer is formed in the bilayer and that the fluorescence measured only comes from oligomers composed only of daptomycin (i.e., no NBD-labeled daptomycin). Our data suggests rather that there is a dynamic and two-stage equilibrium for binding and oligomerization at the membrane:

Specifically, we have determined the size and shape of the daptomycin solution-phase oligomer (i.e. daptomycin + calcium ions only). This enabled us to create a model with 14 daptomycin molecules, within the envelope derived from ab initio fits of the SANS data [36,37,45]. The starting structure revealed that the molecules are not closely packed and that the interactions between monomers are weak. The initial MD simulation (Fig. 3D), where the molecules were restrained to stay within the ellipsoid (Fig. S5B), showed that the daptomycin monomers do not even form a well-defined micellar structure. Given that there is so much space between the daptomycin monomers, it is not surprising that once the restraints are removed, the oligomer falls apart (Fig. 3E). This result suggests that the relevant form of daptomycin in terms of its mode of action is the monomeric form. This is further supported by the results obtained in the presence of 150 mM NaCl, where oligomer formation is weakened due to screening effects (Fig. S1). Not only does the oligomer have a propensity to fall apart, the daptomycin is also diluted during delivery to the patient, where the concentration of daptomycin decreases from 30 mM to peak serum concentrations of 35–86 μM (in the presence of mM Ca2 +), with 75–93% of this daptomycin bound to albumin (J. Silverman, personal communication). Our results also suggest that daptomycin samples many different conformations in solution, in accordance with the four NMR structures reported in literature (3 apo [24–26], 1 with calcium present [25,29]). Next, we characterized the fusion properties of daptomycin as a function of calcium concentration and lipid composition. It was found that DMPC/PG vesicles fused most readily, with only a 100-fold excess of calcium (molar ratio relative to daptomycin) needed for fusion to occur. For other liposomes, much higher calcium concentrations were needed, with POPC/PG vesicles being the most stable. These findings correlate with the stability of these vesicles, as discussed in a recent study which examined ion leakage caused by daptomycin [21]. Indeed, Zhang et al. found that DMPC/PG vesicles were less stable than POPG/ CL vesicles, which in turn were less stable than POPC/PG vesicles. Although bacteria do not contain such a high amount of PC lipids, POPC/ PG model membranes have been found to be useful models for biophysical studies of antimicrobial peptides [46], yielding data supported by cell studies [47]. Interestingly, we also found that fusion is preserved even upon dilution of the daptomycin/liposome/Ca2 + samples by a factor of 2 or 10. This result has important consequences in the preparation of samples, since stock solutions are often used for convenience. Nevertheless, fusion can be prevented by ensuring that calcium concentrations are kept low (vide infra). As mentioned previously, fusion is not directly relevant to the mechanism of action, as it is not expected that bacterial membranes fuse as a result of daptomycin exposure. Nonetheless, fusion can have an impact on the interpretation of biophysical data, in particular in the characterization of the oligomerization state of daptomycin in model membranes (Fig. 2). Prior to determining whether using conditions where fusion is not present has an impact on the FRET data, we first determined the extent of binding and the kinetics of binding of daptomycin to DMPC/PG and POPC/PG vesicles. We found that daptomycin binds strongly (Fig. 5) and rapidly (Fig. 6) to both DMPC/PG and POPC/PG membranes, with Kd values in the micromolar range. Partition coefficient Kp values were in the range of 104–105 M− 1, suggesting that daptomycin behaves like the conjugated peptides studied by Ruzza et al. [33], i.e. peptides with a distinct hydrophobic and charged segment. Given the nature of daptomycin (Fig. 1), this suggests that both the acyl chain and the peptide segment mediate binding, with Ca2 + playing the important role of neutralizing the negative charges in daptomycin and in the PG membrane. It should be noted that the Ca2 + concentration was below physiological concentrations (to avoid fusion), but a factor 10 larger than required for saturation of daptomycin binding (Fig. S2). FRET experiments were then conducted using sample conditions were no fusion occurred and where > 95% binding of daptomycin was assured. Our results suggested that daptomycin can bind to liposomes

Dap + Lipid ⇌ Dap⋅Lipid

nDap⋅Lipid ⇌ Dapn⋅Lipid. The first stage is binding of daptomycin to the vesicle membrane, which our data suggests occurs in a concentration-dependent manner, albeit with micromolar dissociation constants. The second step is oligomerization of the membrane-bound daptomycin, with an oligomerization number of n (where n may or may not take on multiple values), where this equilibrium appears to be driven to the right by increasing concentrations of membrane-bound daptomycin. This data and corresponding hypothesis are consistent with recent findings that daptomycin does not form distinct pores in B. subtilis [3], nor in model systems consisting of POPC/PG liposomes and the potassium fluorescent indicator PBFI [21]. Further interpretation of the data presented in Fig. 7 (e.g. in terms of oligomerization number, n) was not done because the use of the FRET donor/acceptor pair of the kynurenine residue of daptomycin and the NBD label attached to another molecule of daptomycin (i.e. as used in Muraih et al. [11,12] and in this work) has inherent limitations and assumptions that warrant discussion. The kynurenine-NBD FRET system is not ideal because there is equal or greater potential for intra-molecular energy transfer than for the inter-molecular energy transfer useful for probing oligomerization (Muriah found that intra-molecular FRET occurred with ~100% efficiency [11,12]), To minimize the influence of intra-molecular FRET, it is necessary to have a low excitation flux such that only one kynurenine residue per oligomer is excited at any one time. Under these conditions, intra-molecular FRET is estimated to be relatively low probability (~ 20% with a 1:4 ratio of NBD-daptomycin and native daptomycin), provided that the oligomer packs tightly enough that the inter-molecular kynurenine-NBD separation is very close to the intra-molecular kynurenine-NBD separation. Moreover, and in light of the potential oligomerization equilibrium noted above, the FRET system cannot readily distinguish between a growing mole fraction of oligomers or a greater oligomerization number as daptomycin concentration increases. Previous quantitative analysis had to assume complete oligomerization with a fixed oligomerization number, and had to make a correction for putative inter-oligomer FRET, which was questionable given the short Förster distance (~ 2.7 nm) for the kynurenine-NBD FRET pair and modeled size of an oligomer. A more ideal experiment would utilize two daptomycin derivatives: one labeled at its ornithine with a donor dye, and second labeled at its ornithine residue with an acceptor dye, with a greater number of the latter. Such a system would avoid intra-molecular energy transfer and, in principle, changes in the oligomer mole fraction versus oligomerization number could be distinguished by correlating fluorescence intensity and lifetime data. However, as we have shown here, even a label as small as NBD (MW ~ 164 g/mol) affects the binding of daptomycin to vesicles. Larger fluorescent dye labels would be expected to have a greater effect. Common labels such as fluorescein, Cy3, Texas Red, and Cy5 have molecular weights that are 2–4-fold larger, or as much as a third of the weight of daptomycin itself, and their lipophilicity may have a significant impact on the interactions between daptomycin and vesicles. Utilization of the native fluorescent kynurenine residue and a low molecular weight acceptor such as NBD is thus far less likely to perturb the properties and behaviour of the daptomycin. Indeed, the MIC of NBD-labeled daptomycin was found to 1497

BBA - Proteins and Proteomics 1865 (2017) 1490–1499

J. Zhang et al.

be 1.85 μM, whereas as that of Alexa Fluor 350 labeled daptomycin was 8–16 fold less active than native daptomycin (0.02–5 μM) [12]. Despite the non-ideal FRET system, the significant benefit of NBD is minimization of perturbation of the system through its small size and zerolength linker. These points are important to understand for future studies. In conclusion, our results show that the concentration-dependent behaviour of daptomycin can have a significant impact on mechanism of action models derived from biophysical data. The solution-state oligomerization has been found to be unimportant, as the aggregate is loosely packed and likely to readily fall apart upon dilution, as indicated by MD simulation. Concentration-dependent fluorescence measurements suggest that membrane-binding and oligomerization may occur in multiple equilibrium steps, suggesting intricacies beyond what has typically been assumed in many previous studies. These experimental results provide an improved framework for more general and rigorous biophysical studies, which will hopefully make the quest for the mode of action of daptomycin a less elusive one.

[10] R. Taylor, K. Butt, B. Scott, T. Zhang, J.K. Muraih, E. Mintzer, S. Taylor, M. Palmer, Two successive calcium-dependent transitions mediate membrane binding and oligomerization of daptomycin and the related antibiotic A54145, Biochim. Biophys. Acta Biomembr. 1858 (9) (2016) 1999–2005. [11] J.K. Muraih, M. Palmer, Estimation of the subunit stoichiometry of the membraneassociated daptomycin oligomer by FRET, Biochim. Biophys. Acta Biomembr. 1818 (7) (2012) 1642–1647. [12] J.K. Muraih, A. Pearson, J. Silverman, M. Palmer, Oligomerization of daptomycin on membranes, Biochim. Biophys. Acta Biomembr. 1808 (4) (2011) 1154–1160. [13] D. Lin, H.Y. Lam, W. Han, N. Cotroneo, B.A. Pandya, X. Li, Structure-activity relationship of daptomycin analogues with substitution at (2S, 3R) 3-methyl glutamic acid position, Bioorg. Med. Chem. Lett. 27 (3) (2017) 456–459. [14] N.E. Allen, J.N. Hobbs, W.E. Alborn, Inhibition of peptidoglycan biosynthesis in gram-positive bacteria by LY146032, Antimicrob. Agents Chemother. 31 (7) (Jul. 1987) 1093–1099. [15] D. Mengin-Lecreulx, N.E. Allen, J.N. Hobbs, J. van Heijenoort, Inhibition of peptidoglycan biosynthesis in Bacillus megaterium by daptomycin, FEMS Microbiol. Lett. 57 (3) (Jun. 1990) 245–248. [16] T. Schneider, K. Gries, M. Josten, I. Wiedemann, S. Pelzer, H. Labischinski, H.G. Sahl, The lipopeptide antibiotic friulimicin B inhibits cell wall biosynthesis through complex formation with bactoprenol phosphate, Antimicrob. Agents Chemother. 53 (4) (Apr. 2009) 1610–1618. [17] T. Zhang, J.K. Muraih, B. MacCormick, J. Silverman, M. Palmer, Daptomycin forms cation- and size-selective pores in model membranes, Biochim. Biophys. Acta Biomembr. 1838 (10) (Oct. 2014) 2425–2430. [18] N. Cotroneo, R. Harris, N. Perlmutter, T. Beveridge, J.A. Silverman, Daptomycin exerts bactericidal activity without lysis of Staphylococcus aureus, Antimicrob. Agents Chemother. 52 (6) (Jun. 2008) 2223–2225. [19] Y. Xing, W. Wang, S. Dai, T. Liu, J. Tan, G. Qu, Y. Li, Y. Ling, G. Liu, X. Fu, H. Chen, Daptomycin exerts rapid bactericidal activity against Bacillus anthracis without disrupting membrane integrity, Acta Pharmacol. Sin. 35 (2) (Feb. 2014) 211–218. [20] J.K. Hobbs, K. Miller, A.J. O'Neill, I. Chopra, Consequences of daptomycin-mediated membrane damage in Staphylococcus aureus, J. Antimicrob. Chemother. 62 (5) (Jul. 2008) 1003–1008. [21] J. Zhang, K. Scoten, S.K. Straus, Daptomycin leakage is selective, ACS Infect. Dis. 2 (10) (2016) 682–687. [22] J. Pogliano, N. Pogliano, J.A. Silverman, Daptomycin-mediated reorganization of membrane architecture causes mislocalization of essential cell division proteins, J. Bacteriol. 194 (17) (2012) 4494–4504. [23] D. Jung, J.P. Powers, S.K. Straus, R.E.W. Hancock, Lipid-specific binding of the calcium-dependent antibiotic daptomycin leads to changes in lipid polymorphism of model membranes, Chem. Phys. Lipids 154 (2) (2008) 120–128. [24] L.-J. Ball, C.M. Goult, J.A. Donarski, J. Micklefield, V. Ramesh, NMR structure determination and calcium binding effects of lipopeptide antibiotic daptomycin, Org. Biomol. Chem. 2 (13) (2004) 1872–1878. [25] D. Jung, A. Rozek, M. Okon, R.E. Hancock, Structural transitions as determinants of the action of the calcium-dependent antibiotic daptomycin, Chem. Biol. 11 (7) (Jul. 2004) 949–957. [26] K.S. Rotondi, L.M. Gierasch, A well-defined amphipathic conformation for the calcium-free cyclic lipopeptide antibiotic, daptomycin, in aqueous solution, Biopolym. Pept. Sci. Sect. 80 (2–3) (2005) 374–385. [27] J. Qiu, L.E. Kirsch, Evaluation of lipopeptide (daptomycin) aggregation using fluorescence, light scattering, and nuclear magnetic resonance spectroscopy, J. Pharm. Sci. 103 (3) (2014) 853–861. [28] G. Bunkóczi, L. Vértesy, G.M. Sheldrick, Structure of the lipopeptide antibiotic tsushimycin, Acta Crystallogr. Sect. D Biol. Crystallogr. 61 (8) (2005) 1160–1164. [29] S.W. Ho, D. Jung, J.R. Calhoun, J.D. Lear, M. Okon, W.R.P. Scott, R.E.W. Hancock, S.K. Straus, Effect of divalent cations on the structure of the antibiotic daptomycin, Eur. Biophys. J. 37 (4) (2008) 421–433. [30] W.R.P. Scott, P.H. Hünenberger, I.G. Tironi, A.E. Mark, S.R. Billeter, J. Fennen, A.E. Torda, T. Huber, P. Krüger, W.F. van Gunsteren, The GROMOS biomolecular simulation program package, J. Phys. Chem. A 103 (19) (1999) 3596–3607. [31] L. Larsson, S. Ohman, Serum ionized calcium and corrected total calcium in borderline hyperparathyroidism, Clin. Chem. 24 (11) (1978) 1962–1965. [32] A. Hawrani, R.A. Howe, T.R. Walsh, C.E. Dempsey, Thermodynamics of RTA3 peptide binding to membranes and consequences for antimicrobial activity, Biochim. Biophys. Acta Biomembr. 1798 (6) (2010) 1254–1262. [33] P. Ruzza, B. Biondi, A. Marchiani, N. Antolini, A. Calderan, Cell-penetrating peptides: a comparative study on lipid affinity and cargo delivery properties, Pharmaceuticals 3 (4) (2010) 1045–1062. [34] G. Drin, M. Mazel, P. Clair, D. Mathieu, M. Kaczorek, J. Temsamani, Physico-chemical requirements for cellular uptake of pAntp peptide: role of lipid-binding affinity, Eur. J. Biochem. 268 (5) (2001) 1304–1314. [35] G. Schwarz, S. Stankowski, V. Rizzo, Thermodynamic analysis of incorporation and aggregation in a membrane: application to the pore-forming peptide alamethicin, Biochim. Biophys. Acta 861 (1986) 141–151. [36] D.I. Svergun, Restoring low resolution structure of biological macromolecules from solution scattering using simulated annealing, Biophys. J. 76 (6) (1999) 2879–2886. [37] D. Franke, D.I. Svergun, DAMMIF, a program for rapid ab-initio shape determination in small-angle scattering, J. Appl. Crystallogr. 42 (Pt 2) (2009) 342–346. [38] V.V. Volkov, D.I. Svergun, Small-angle scattering, J. Appl. Crystallogr. 36 (2003) 860–864. [39] E.C. Meng, E.F. Pettersen, G.S. Couch, C.C. Huang, T.E. Ferrin, Tools for integrated sequence-structure analysis with UCSF Chimera, BMC Bioinformatics 7 (1) (Jul. 2006) 339.

Conflict of interest We declare no conflict of interest. Transparency document The http://dx.doi.org/10.1016/j.bbapap.2017.07.020 associated with this article can be found in the online version. Acknowledgements SKS gratefully acknowledges the ILL for beam time on D22. Funding from the Natural Sciences and Engineering Research Council of Canada (RGPIN 240795-12) (NSERC) is also acknowledged. SKS was a recipient of a Marie Curie International Incoming Fellowship. WRA and SKS also thank the Michael Smith Foundation for Health Research for Scholar Awards. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.bbapap.2017.07.020. References [1] S.K. Straus, R.E.W. Hancock, Mode of action of the new antibiotic for Gram-positive pathogens daptomycin: comparison with cationic antimicrobial peptides and lipopeptides, Biochim. Biophys. Acta Biomembr. 1758 (9) (2006) 1215–1223. [2] S.D. Taylor, M. Palmer, The action mechanism of daptomycin, Bioorganic Med. Chem. 24 (24) (2016) 6253–6268. [3] A. Müller, M. Wenzel, H. Strahl, F. Grein, T.N.V. Saaki, B. Kohl, T. Siersma, J.E. Bandow, H.-G. Sahl, T. Schneider, L.W. Hamoen, Daptomycin inhibits cell envelope synthesis by interfering with fluid membrane microdomains, Proc. Natl. Acad. Sci. U. S. A. 113 (45) (2016) E7077–E7086. [4] M. Debono, B.J. Abbott, R.M. Molloy, D.S. Fukuda, A.H. Hunt, V.M. Daupert, F.T. Counter, J.L. Ott, C.B. Carrell, L.C. Howard, Enzymatic and chemical modifications of lipopeptide antibiotic A21978C: the synthesis and evaluation of daptomycin (LY146032), J. Antibiot. (Tokyo). 41 (8) (Aug. 1988) 1093–1105. [5] J. Bradley, C. Glasser, H. Patino, S.R. Arnold, A. Arrieta, B. Congeni, R.S. Daum, T. Kojaoghlanian, M. Yoon, D. Anastasiou, D.J. Wolf, P. Bokesch, Daptomycin for complicated skin infections: a randomized trial, Pediatrics 139 (3) (Mar. 2017) e20162477. [6] R.H. Baltz, V. Miao, S.K. Wrigley, Natural products to drugs: daptomycin and related lipopeptide antibiotics, Nat. Prod. Rep. 22 (6) (Dec. 2005) 717. [7] J.A. Silverman, N.G. Perlmutter, H.M. Shapiro, Correlation of daptomycin bactericidal activity and membrane depolarization in Staphylococcus aureus, Antimicrob. Agents Chemother. 47 (8) (Aug. 2003) 2538–2544. [8] W.E. Alborn, N.E. Allen, D.A. Preston, Daptomycin disrupts membrane potential in growing Staphylococcus aureus, Antimicrob. Agents Chemother. 35 (11) (Nov. 1991) 2282–2287. [9] J.H. Jorgensen, L.A. Maher, J.S. Redding, In vitro activity of LY146032 (daptomycin) against selected aerobic bacteria, Eur. J. Clin. Microbiol. 6 (1) (Feb. 1987) 91–96.

1498

BBA - Proteins and Proteomics 1865 (2017) 1490–1499

J. Zhang et al.

[44] J.H. Lakey, M. Ptak, Fluorescence indicates a calcium-dependent interaction between the lipopeptide antibiotic LY146032 and phospholipid membranes, Biochemistry 27 (13) (1988) 4639–4645. [45] D.I. Svergun, Determination of the regularization parameter in indirect-transform methods using perceptual criteria, J. Appl. Crystallogr. 25 (4) (Aug. 1992) 495–503. [46] J.T.J. Cheng, J.D. Hale, M. Elliott, R.E.W. Hancock, S.K. Straus, The importance of bacterial membrane composition in the structure and function of aurein 2.2 and selected variants, Biochim. Biophys. Acta Biomembr. 1808 (3) (2011) 622–633. [47] M. Wenzel, C.H.R. Senges, J. Zhang, S. Suleman, M. Nguyen, P. Kumar, A.I. Chiriac, J.J. Stepanek, N. Raatschen, C. May, U. Krämer, H.-G. Sahl, S.K. Straus, J.E. Bandow, Antimicrobial peptides from the Aurein family form ion-selective pores in Bacillus subtilis, Chembiochem 16 (7) (2015) 1101–1108.

[40] Z. Yang, K. Lasker, D. Schneidman-Duhovny, B. Webb, C.C. Huang, E.F. Pettersen, T.D. Goddard, E.C. Meng, A. Sali, T.E. Ferrin, UCSF Chimera, MODELLER, and IMP: an integrated modeling system, J. Struct. Biol. 179 (3) (Sep. 2012) 269–278. [41] E.F. Pettersen, T.D. Goddard, C.C. Huang, G.S. Couch, D.M. Greenblatt, E.C. Meng, T.E. Ferrin, UCSF Chimera? A visualization system for exploratory research and analysis, J. Comput. Chem. 25 (13) (Oct. 2004) 1605–1612. [42] C.C. Huang, E.C. Meng, J.H. Morris, E.F. Pettersen, T.E. Ferrin, Enhancing UCSF Chimera through web services, Nucleic Acids Res. 42 (W1) (Jul. 2014) W478–W484. [43] T.D. Goddard, C.C. Huang, E.C. Meng, E.F. Pettersen, G.S. Couch, J.H. Morris, T.E. Ferrin, UCSF ChimeraX: Meeting Modern Challenges in Visualization and Analysis, Protein Sci. (Jul 14 2017), http://dx.doi.org/10.1002/pro.3235.

1499