Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2020 Dec 22;6(1):388–400. doi: 10.1021/acsomega.0c04752

Molecular Dynamics Simulation of the Interaction of Two Linear Battacin Analogs with Model Gram-Positive and Gram-Negative Bacterial Cell Membranes

Aparajita Chakraborty †,‡,§, Elisey Kobzev †,§,∥, Jonathan Chan ‡,⊥, Gayan Heruka de Zoysa #, Vijayalekshmi Sarojini #,∇, Thomas J Piggot ○,◆,*, Jane R Allison †,‡,§,¶,&,*
PMCID: PMC7807746  PMID: 33458490

Abstract

graphic file with name ao0c04752_0009.jpg

Antimicrobial peptides (AMPs) are a potential solution to the increasing threat of antibiotic resistance, but successful design of active but nontoxic AMPs requires understanding their mechanism of action. Molecular dynamics (MD) simulations can provide atomic-level information regarding how AMPs interact with the cell membrane. Here, we have used MD simulations to study two linear analogs of battacin, a naturally occurring cyclic, lipidated, nonribosomal AMP. Like battacin, these analogs are active against Gram-negative multidrug resistant and Gram-positive bacteria, but they are less toxic than battacin. Our simulations show that this activity depends upon a combination of positively charged and hydrophobic moieties. Favorable interactions with negatively charged membrane lipid head groups drive association with the membrane and insertion of hydrophobic residues, and the N-terminal lipid anchors the peptides to the membrane surface. Both effects are required for stable membrane binding.

Introduction

Modern medicine relies heavily on antibiotics for treating bacterial infections. However, increasing bacterial resistance against many antibiotics has become one of the biggest threats to global health and food security.1,2 The problem is expected to become much more serious in the not-too-distant future with the emergence of increasing numbers of multidrug and extremely drug resistant (MDR and XDR) bacterial strains.3 This is driving a continuous search for new synthetic and natural antibacterial agents,4,5 one promising source of which is antimicrobial peptides (AMPs).6−10 The promise of AMPs remains to be fulfilled, however, with only a few peptides entering clinical trials.7,11,12 This may be, in part, because their mechanism of action is difficult to study experimentally at an atomic level,13 hindering truly rational design of active but less toxic analogs.14 Indeed, as yet, no universal sequence–activity relationship has been discovered, although it is becoming apparent that this may require consideration of dynamic structural ensembles rather than static structures.15−19

AMPs are one of the largest classes of membrane lytic peptides produced by invertebrates and vertebrates, including microbes themselves.20 AMPs are part of the cell-mediated immune response and are synthesized upon induction by pathogens and also by bacteria themselves in response to competition for food and resources.20,21 Eukaryotic AMPs, which are genetically encoded and synthesized on the ribosome, typically consist of 10–50 amino acids and are classified according to their secondary structure or lack thereof.21 Bacterial AMPs may be non-ribosomally synthesized and thus can also include non-proteinogenic amino acids, decreasing their susceptibility to proteolytic enzymes and be glycosylated; acylated; halogenated; or hydroxylated, formylated, or lipidated.21 Non-ribosomally synthesized AMPs can also be cyclic, increasing their rigidity. The major features shared by all AMPs are that they are soluble in aqueous environments and can partition into lipid environments, such as the cell membrane.9

The generally cationic nature of AMPs along with the anionic nature of bacterial membranes results in preferential targeting of AMPs to bacterial rather than zwitterionic eukaryotic membranes.21−24 The interaction of AMPs with cell membranes is then facilitated by their high hydrophobic amino acid content.21,24,25 Nonribosomal AMPs can also include features, such as lipids, which are likely to further favor membrane interaction. The selectivity of AMPs depends upon differences in membrane composition between different organisms.

AMPs generally cause cell death through membrane disruption and eventual cell lysis, although there are some instances of AMPs inhibiting biofilm production or crossing the cell membrane and inhibiting cellular functions.21,26−28 Even amongst membrane-lytic AMPs, however, there are a variety of mechanisms of action. For ribosomally synthesized AMPs, in particular those that are α-helical, three major mechanisms of membrane disruption have emerged: “barrel-stave”, “toroidal-pore”, and “carpet”.20,21,27,28 Less is known about the mechanisms of action of other types of AMPs, however.24

Battacin is a cyclic, lipidated nonribosomal AMP first isolated from Paenibacillus tianmuensis.(29) It is composed of both d- and l-amino acids and the unnatural amino acid α,γ-diaminobutyric acid (Dab) in both its d- and l- forms, which give it resistance to proteases and thus make it a potential candidate for therapeutics. It has been found to be active against Gram-negative MDR bacteria and Gram-positive bacteria in both in vitro and in vivo.29 Along with high efficacy against MDR bacterial strains, however, naturally occurring battacin is nephrotoxic and neurotoxic29 and thus is not used clinically.

To overcome the toxicity of naturally occurring battacin, we previously designed and synthesized novel structural analogs.30 Remarkably and in contrast to other AMPs, such as polymyxin, two of these linear peptides, octapeptide 17 (hereafter octapeptide) and its derivative pentapeptide 30 (hereafter pentapeptide) (Figure 1), were found experimentally to be active, in terms of both lysing bacteria and dispersing preformed biofilms, against both model Gram-negative (Escherichia coli) and Gram-positive (Staphylococcus aureus) bacteria.30,31 Unlike many AMPs,26,32,33 the activity of these battacin analogs does not rely on the presence of a defined secondary structure when bound to the membrane,30,34 suggesting that the mechanism of action might also differ. It is extremely difficult to experimentally obtain a detailed atomic level insight into their mechanism of action, however.

Figure 1.

Figure 1

Open in a new tab

Chemical structures of (a) octapeptide 17 (octapeptide) and (b) pentapeptide 30 (pentapeptide). The Dab side chains are shown in their deprotonated state.

Molecular dynamics (MD) simulations can provide time-dependent information about the structure of AMPs and the nature and energetics of their interactions. MD simulation of AMP–lipid interactions has a long history.35−38 Limitations such as the use of coarse-grained (CG) models39−44 and simple membranes, particularly those containing only phosphatidylcholine lipids,16,41−45 which is not representative of bacterial membranes, can produce results different to those observed with more realistic models.15,16,25,39,40,46,47 It has been suggested that the different results sometimes observed using the Martini CG force field may be due to a higher energetic cost of pore formation.48 Almost all simulation studies, however, have been of AMPs that form stable α-helical or β-hairpin structures upon binding to and/or insertion into the membrane.15−17,25,32,44−47,49−59 In general, these AMPs are more likely to act via a pore-forming mechanism.58 Less is known, however, about the mechanism of action of unstructured AMPs such as the linear battacin analogs.

We have, therefore, carried out MD simulations with atomistic cell membrane and peptide models to investigate how each of the two linear battacin analogs, octapeptide and its truncated derivative pentapeptide, interact with model Gram-negative (E. coli) and Gram-positive (S. aureus) cell membranes. We find that for both peptides, and in keeping with both the general mechanism by which AMPs selectively target bacterial cell membranes21,24 and residue-specific experimental results for octapeptide30 and polymyxins,60,61 the positively charged Dab residues are the most important for the initial interaction with and binding to the membrane surface. The hydrocarbon tail of the N-terminal lipid plays a major role in membrane permeation and, along with the hydrophobic residues Leu and d-Phe, anchors the peptide to the membrane surface, again in keeping with experimental results for these peptides.30 Reducing the positive charge on the peptide by deprotonating the d-Dab and Dab residues generally reduces hydrogen bond formation and Coulombic interactions with the membrane lipids, which in turn reduces the structural stability of the peptides and the insertion of the hydrophobic moieties into the membrane. Stable initial binding to the membrane surface, which is promoted by positively charged residues, is therefore crucial for membrane penetration by the hydrophobic moieties and thus explains the experimentally observed importance of the central hydrophobic dipeptide unit Leu-d-Phe, its flanking Dab residues, and the N-terminal lipid of octapeptide, which are recapitulated in the pentapeptide.30

Results and Discussion

Choice of Model Cell Membranes

To investigate the mechanism by which these peptides disrupt the cell membrane of both Gram-positive and Gram-negative bacteria and to search for any differences in the mechanism between the two types of bacteria, we simulated a single copy of each peptide with each of two cell membrane models, a S. aureus membrane and an E. coli inner membrane (IM). These two species were chosen both to represent pathogenic Gram-positive and Gram-negative bacteria, respectively, and for comparison with the experimental tests of peptide efficacy, which used the same species.30

We focused on the E. coli IM because it is both more informative with regard to the mechanism of action of these peptides and because it is more computationally efficient. Naturally occurring battacin is known to permeate the outer membrane (OM) of Gram-negative bacteria; it is its interaction with the IM that leads to the death of the microorganism.29 The interaction with the IM is therefore most relevant to understanding the mechanism of action.

Mechanism of Interaction with Model Cell Membranes

For each peptide–membrane system, we first carried out five independent simulations of 50 ns to monitor the initial approach to and interaction with the membrane. Three of each set of five simulations were chosen at random and extended for a further 450 ns to monitor the long-term behavior and initiation of penetration. Only the data pertaining to these extended simulations are presented here.

Interaction with the S. aureus Membrane

Both peptides approach the S. aureus membrane within 25 ns of simulation in all five 50 ns simulations. Conformational clustering of the three extended simulations shows that unlike when they are alone in solution, interaction of the peptides with the membrane surface results in formation of stable structures that group into just one or two major clusters (Figure 2a-–f). Unlike most widely studied AMPs, however,26,32,33 these stable structures did not exhibit a defined secondary structure, in keeping with the circular dichroism analysis of octapeptide.30

Figure 2.

Figure 2

Open in a new tab

Time series of cluster formation by (a–c) octapeptide, (d–f) NH2-octapeptide, (g–i) pentapeptide, and (j–l) NH2-pentapeptide during simulation in the presence of the model S. aureus membrane.

Both peptides partially embed into the head group region of the lipid bilayer (Figure 3a,c, Supporting Information Figure S1). For octapeptide, the alkyl tail of the lipidated N-terminal residue (5-methylhexanoyl-d-Dab) and d-Phe residue (5) and, to a lesser extent, Dab residues 2 and 6 of octapeptide insert most into the membrane (Figure 3a). The insertion of d-Phe is in keeping with the insertion of hydrophobic side chains observed for the α-helical AMPs, pleurocidin,25 pardaxin,62 and GF-17,63 and the insertion of lipid moieties has been observed for polymyxin B164 and for other lipidated battacin analogs.65 For pentapeptide, only residue 5-methylhexanoyl-d-Dab 1 inserts into the membrane in all three replicate simulations, along with residue Dab 2 in replicate 1 only (Figure 3b). In general, the insertion is much shallower for pentapeptide than for octapeptide.

Figure 3.

Figure 3

Open in a new tab

Partial atom densities with respect to the membrane normal for the lipids and for each residue of (a) octapeptide, (b) NH2-octapeptide, (c) pentapeptide, and (d) NH2-pentapeptide as labeled during simulation in the presence of the model S. aureus membrane.

The peptides also form hydrogen bonds with the lipid head groups (Supporting Information Figures S3a–c and S4a–c). For octapeptide, Dab residue 2 forms hydrogen bonds with membrane lipids in all three simulations, Dab residues 1 and 6 form hydrogen bonds in two of the simulations, and Dab 3, Leu 4, Dab 7, and Leu 8 form hydrogen bonds in one simulation, with only residue d-Phe 5 failing to form hydrogen bonds in any of the three simulations. For pentapeptide, Dab residue 2 forms hydrogen bonds with the membrane in all three simulations, d-Phe 4 and Dab 5 form hydrogen bonds in two simulations, and Dab 1 and Leu 3 form hydrogen bonds in just one of the simulations.

The Dab residues form the majority of the hydrogen bonds to the membrane, especially for octapeptide, with both the backbone and side chain amides participating. Dab residues are known to be important for the binding of antimicrobial peptides to bacterial cell membranes due to their cationic nature,61,64,66−68 and MD simulations of polymyxin B1 with E. coli inner and outer membranes revealed that a large proportion of the hydrogen bonds between polymyxin B1 and the membrane lipids involved Dab.64

To test the importance of the positively charged Dab side chain on membrane binding, we performed an additional three independent 500 ns simulations for each peptide in which the Dab residues contained an NH2 group instead of an NH3+ group. While NH2 is uncharged, it is still capable of forming hydrogen bonds, albeit to a lesser extent.

The NH2–octapeptide still forms reasonably stable structures after binding to the membrane, particularly in the second replicate simulation, where it is predominantly in one cluster (Figure 2g–i), but these again did not comprise a defined secondary structure. During the second replicate simulation, hydrogen bonds are formed to just one lipid, an LPG, initially involving the side chain amides of Dab residues 1 and 2, and later between the side chain amides of Dab residues 2 and 3 (Supporting Information Figure S3d–f). In contrast, in the first and third replicate simulations, which are less structurally stable, only three hydrogen bonds are formed in total, mostly involving backbone amides. Overall, the NH2-octapeptide forms fewer hydrogen bonds with the membrane than the NH3+-octapeptide.

In all three simulations of the NH2-pentapeptide, the stability of the pentapeptide structure was reduced compared to the NH3+-pentapeptide, even after binding to the membrane (Figure 2j–l). There is no clear correspondence between structural stability and hydrogen bond formation (Supporting Information Figure S4d–f), however.

Overall, these results suggest that while uncharged Dab residues are still able to form hydrogen bonds with the membrane lipids, the stronger interactions between positively charged Dab residues and membrane lipids stabilize the peptide structure, particularly in the case of the pentapeptide.

The importance of Coulombic interactions is further highlighted by the much greater magnitude of the Coulombic interaction potential energies compared to the van der Waals interaction potential energies for all systems (Figure 4).

Figure 4.

Figure 4

Open in a new tab

(a, c, e, g) Van der Waals and (b, d, f, h) Coulombic interaction potential energies between (a, b) octapeptide, (c, d) NH2-octapeptide, (e, f) pentapeptide, and (g, h) NH2-pentapeptide and each of the three lipids in the S. aureus membrane during each of the three replicate MD simulations as labeled.

The lipid with which each peptide has the most favorable interactions differs between simulation replicates. It appears that there may be a slight preference for the interaction with PG and LPG, which would be in keeping with emerging evidence that AMP resistance tends to reduce the levels of PG, and sometimes LPG, in bacterial cell membranes.69 It is more likely, however, that this apparent preference simply reflects their much greater concentration (54 and 36%, respectively; Table 1). In fact, the amount of favorable interactions with DPG is perhaps surprising, given that it comprises only 5% of the lipids, although due to its size (four fatty acid tails), it effectively comprises 10% of the surface area. Overall, it appears that neither peptide has specific lipid preferences, and substantially, more sampling would be required to investigate this further.

Table 1. Lipid Composition of the S. aureus and E. coli IM Model Bilayers; The Two Leaflets of each Bilayer Had Identical Lipid Composition.
model system lipid headgroup lipid tail %
S. aureus phosphatidylglycerol (PG) 15C anteiso-branched 57%
  lysine-phosphatidylglycerol (LPG) 15C anteiso-branched 38%
  diphosphatidylglycerol (DPG/cardiolipin) 15C anteiso-branched 5%
E. coli IM phosphatidylethanolamine (PE) 1-palmitoyl, 2-cis-vaccenyl (PV) 75%
  phosphatidylglycerol (PG) 1-palmitoyl, 2-cis-vaccenyl (PV) 20%
  diphosphatidylglycerol (DPG/cardiolipin) 1-palmitoyl, 2-cis-vaccenyl (PV) 5%
Open in a new tab

To facilitate design of novel antimicrobial peptides with enhanced activity, the contribution of each residue to the membrane interaction was also extracted.

For the octapeptide, residues 5-methylhexanoyl-d-Dab 1, d-Phe 5, and Leu 8 have the most favorable van der Waals interactions (Figure 4a), in keeping with the hydrophobic lipid tail and side chains of these residues. For pentapeptide, residues 5-methylhexanoyl-d-Dab 1 and d-Phe 4 have the most favorable van der Waals interactions (Figure 4e). This is again in keeping with the hydrophobic lipid tail and side chains of residues 1 and 4 and the partial atom densities (Figure 3a). Insertion of residue 5-methylhexanoyl-d-Dab 1 of both peptides is visible in the partial atom densities (Figure 3), but the d-Phe and Leu side chains do not insert as far due to their shorter length.

The favorable Coulombic interactions are, in general, of higher magnitude compared to the van der Waals interactions (Figure 4b,f). For both peptides, d-Dab and Dab residues provide the majority of the favorable Coulombic interactions with the membrane lipid head groups, which agrees with the importance of these residues identified by experimental alanine scanning.30 These interaction potential energies were further decomposed to show that the main contribution comes from the NH3+ groups of the Dab residues (data not shown). The importance of the positively charged NH3+ group is confirmed by the substantially reduced peptide–membrane Coulombic interaction energy in the simulations of the NH2-peptides (Figure 4d,h). The importance of Dab residues has also been highlighted by MD simulations64 and experimental studies of polymyxin,61,66,67,70 and results for other AMPs have shown positively charged amino acids to be critical.68,71

Interestingly, we find that the NH2-pentapeptide and NH2-octapeptide also have reduced van der Waals interaction potential energies with the membrane core compared to the NH3+ cases (Figure 4c,g). This is due to reduced insertion of the hydrophobic portions of the 5-methylhexanoyl-Dab, d-Phe, and Leu residues into the membrane core (Figure 3). It seems, therefore, that stable membrane binding is required to facilitate insertion of the hydrophobic portions of the peptides into the membrane.

Together, these results indicate that the favorable Coulombic interactions between the d-Dab and Dab residues and the negatively charged lipid head groups provide membrane binding and thus structural stability to the peptides, which in turn allow for insertion of the hydrophobic residues into the membrane and further stabilization of the peptide–membrane interaction through van der Waals interactions. Such a mechanism has been suggested for other AMPs21,24,71 and would be in keeping with results of MD simulations of polymyxin B1.64

Interactions with the E. coli IM

In the case of the E. coli IM, both peptides were again found to approach and form hydrogen bonds with the membrane rapidly, within 30–50 ns of simulation (Supporting Information Figures S5a–c and S6a–c).

As when interacting with the S. aureus membrane, both peptides become more structurally stable once bound to the membrane surface, but to a lesser degree, especially for the pentapeptide (Figure 5a–g). Whereas on the S. aureus membrane, both peptides grouped predominantly into a single cluster; here, ∼60% of the simulation time was spent in the most populated 5–6 clusters. This lack of a single, stable, and well-defined secondary structure is in keeping with the experimental circular dichroism results for octapeptide30 and contrast to many AMPs.26,32,33

Figure 5.

Figure 5

Open in a new tab

Time series of cluster formation by (a–c) octapeptide, (d–f) NH2-octapeptide, (g–i) pentapeptide, and (j–l) NH2–pentapeptide during simulation in the presence of the model E. coli IM.

Interestingly, the lipidated N-terminal residue (5-methylhexanoyl-d-Dab) of the octapeptide did not insert into the membrane in any of the three replicate simulations; rather, residues Dab 7 and Leu 8 inserted slightly more than the other residues but not beyond the lipid head groups (Figure 6a, Supporting Information Figure S2). Overall, the insertion of the octapeptide was greatly reduced compared to with the S. aureus membrane.

Figure 6.

Figure 6

Open in a new tab

Partial atom densities with respect to the membrane normal for the lipids and for each residue of (a) octapeptide, (b) NH2-octapeptide, (c) pentapeptide, and (d) NH2-pentapeptide as labeled during simulation in the presence of the model E. coli IM.

In contrast, insertion of the pentapeptide was greater than with the S. aureus membrane, with the hydrophobic side chains of residues Leu 3 and d-Phe 4 inserting most in replicate 1 and the lipid tail of 5-methylhexanoyl-d-Dab 1 inserting most in replicates 2 and 3 (Figure 6c). The hydrophobic Leu-d-Phe is common amongst lipopeptides72 and found to be critical for antimicrobial activity of octapeptide30 as well as implicated in the interaction of polymyxin with bacterial cell membranes.60 Insertion of hydrophobic amino acid side chains has also been observed for the α-helical AMPs, pleurocidin,25 pardaxin,62 and GF-17,63 and of lipid moieties for polymyxin B1 and other linearized battacins.64,65

As with the S. aureus membrane, both peptides formed hydrogen bonds with the lipid head groups (Supporting Information Figures S5a–c and S6a–c). These primarily involved the Dab residues, along with the Phe and Leu backbone amide and, in one replicate simulation of octapeptide, the C-terminal amide cap.

To test the importance of hydrogen bond formation between the Dab NH3+ moieties and the phospholipid head groups, we again performed an additional three independent simulations for each peptide in which the Dab residues contained an NH2 group instead of an NH3+ group.

As with the S. aureus membrane, this change decreased the structural stability of the octapeptide (Figure 5d–f), but it had less effect on the pentapeptide (Figure 5j–l). The NH2-octapeptide inserted less deeply into the membrane than the NH3+-octapeptide (Figure 6b, Supporting Information Figure S2a,b), but two of the three simulations of NH2-pentapeptide showed insertion of similar depth to that of NH3+-pentapeptide (Figure 6d, Supporting Information Figure S2c,d), in keeping with their similar structural stability.

The NH2 peptides formed slightly fewer hydrogen bonds to the membrane lipids than the NH3+ peptides, with none at all formed in replicate 2 of the octapeptide and only one in replicate 3 of the pentapeptide (Supporting Information Figures S5d–f and S6d–f). In all other simulations, however, the Dab residues, including the side chain amides, still formed the majority of the hydrogen bonds.

To provide a more detailed understanding of the factors determining cell specificity, we again calculated the interaction potential energy between each peptide and each residue of each type of lipid, and split this into its Coulombic and van der Waals contributions (Figure 7).

Figure 7.

Figure 7

Open in a new tab

(a, c, e, g) Van der Waals and (b, d, f, h) Coulombic interaction potential energies between (a, b) octapeptide, (c, d) NH2-octapeptide, (e, f) pentapeptide, and (g, h) NH2-pentapeptide and each of the three lipids in the E. coli membrane during each of the three replicate MD simulations as labeled.

For both peptides, more residues have favorable Coulombic interaction energies with phosphatidylethanolamine (PE) lipids, followed by PG and DPG. This largely reflects the much greater concentration of PE (70%) compared to the other types of lipids (PG, 15%; DPG, 5%; see Table 1), however, suggesting again that neither pentapeptide nor octapeptide have specific lipid preferences, especially given that such preferences will be difficult to elucidate without more substantial sampling.

For both peptides, the Dab residues again make a major contribution to the favorable Coulombic interactions with the membrane lipid head groups, along with the NH2 moiety in the terminal Leu residue (Figure 7e,g). The latter was particularly important for pentapeptide, most likely due to its lower number of Dab residues. These results are again in keeping with alanine scanning of octapeptide30 and computational and experimental studies of other AMPs.61,64,66−68,70,71

As with the S. aureus membrane, these interactions stabilize the peptide on the membrane surface and thus facilitate insertion of the hydrophobic residues of the peptide into the membrane. The latter is visible as negative van der Waals energies pertaining to interactions between the 5-methylhexanoyl-d-Dab, the d-Phe side chain, and occasionally, the Leu side chain with the hydrophobic membrane core (Figure 7a,c). The insertion of these moieties into the membrane core is visible in the partial electron densities (Figure 6).

For the NH2 peptides, the Coulombic interaction energies again almost disappear (Figure 7e–h), and the van der Waals interaction energies are reduced (Figure 7a–d). For the NH2-octapeptide, residues Leu 4, d-Phe 5, and to a lesser extent, 5-methylhexanoyl-d-Dab 1 and Dab 8 show some favorable van der Waals interaction energies, although the partial atom densities reveal little insertion (Figure 6b). For the NH2-pentapeptide, residues 5-methylhexanoyl-d-Dab 1, Leu 3, and Phe 4 have favorable van der Waals interactions with PE and PG lipids, in keeping with the insertion of these residues into the membrane core (Figure 6d).

The reduction in Coulombic interaction energy and slight reduction in hydrogen bond formation by the NH2 peptides appears to reduce the insertion of the hydrophobic portions of the octapeptide into the membrane core but has less effect on the insertion of the hydrophobic moieties of the pentapeptide.

Conclusions

The two peptides studied here are unusual in that they have activities against both Gram-negative and Gram-positive model species,30 which we sought to understand by analyzing in detail the contributions that each residue and each type of lipid make to the peptide–membrane interaction. Our simulations did not seem to reveal any specific lipid preferences with either model membrane by either of the peptides, although substantially more replicates of each system would be required to provide a more definitive answer.

With both model membranes and peptides, the alkyl tail of the lipidated N-terminal residue (5-methylhexanoyl-d-Dab) and the d-Phe and Leu residues insert most into the membrane core, anchoring the peptide to the membrane. The Dab residues improve membrane binding through increased hydrogen bond formation with both membranes, which in turn improves the peptide structural stability. The lack of a well-defined secondary structure and indeed of a specific membrane-bound structure is in agreement with experimental circular dichroism data30 but is in contrast to many other AMPs,26,32,33 which form specific secondary structures upon membrane binding that appear to be crucial for membrane disruption. Hydrogen bond formation, structural stability, and insertion were generally reduced when the NH3+ side chains of the d-Dab and Dab residues were changed to NH2. An exception to this was the pentapeptide, for which there was little change.

In general, we found that favorable Coulombic interactions between Dab residues and the negatively charged lipid head groups drive membrane association and are required for stable membrane binding and insertion of the hydrophobic moieties, which anchor the peptides to the membrane. This mechanism, while previously unconfirmed for these linear battacin analogs, is in keeping with the results of studies of how a range of other, predominantly helical AMPs interact with the cell membrane.21,24,71 It is also in keeping with the importance of Dab30,61,64,66,67,70 and other positively charged amino acids,68,71 along with hydrophobic amino acids25,30,62,63 and/or lipid moieties64,65 for effective membrane disruption by a range of AMPs, including those studied here. Only individual peptide molecules were studied here, but our results suggest that pentapeptide and octapeptide are likely to first associate with the membrane surface, with selectivity coming largely through electrostatic complementarity but ultimately disrupt the membrane integrity through insertion of their hydrophobic components.

Computational Methods

Coordinates

Initial coordinates for both peptides were obtained using the Avogadro software package73 and refined by energy minimization and a 500 ps equilibration as outlined below. Pre-equilibrated coordinates for lipid bilayers representative of the S. aureus and E. coli inner membranes (Table 1) were taken from refs (64, 74). Peptide molecules were placed at least 1.4 nm (the cutoff distance for calculation of inter-atomic interactions) from the membrane (Supporting Information Figures S1 and S2), and peptide and membrane coordinates were combined by simply concatenating the coordinate files.

Parameters

The natural amino acids in both peptides were modeled using standard GROMOS 54A775 parameters. Parameters for l-Dab were obtained by removing one CH2 group from the side chain of Lys, and d-Phe and d-Dab were obtained by inverting the stereochemistry of and reordering the atoms surrounding the Cα atom of the l-Phe and l-Dab parameters. The terminal amine of d-Dab was modeled as NH3+, representative of its state at pH 7, unless otherwise specified. Partial charges for the NH2 state were obtained by analogy to the deprotonated state of lysine. GROMOS-CKP74,76−78 parameters, which are compatible with the GROMOS 54A7 force field, were used for the phospholipids, including the 5-methylhexanoyl portion of 5-methylhexanoyl-d-Dab.

Molecular Dynamics Simulations

All the simulations were performed using the GROMACS MD software package79 version 2016.3. All bond lengths were constrained using the LINCS algorithm80 allowing for a 2 fs time step, and periodic boundary conditions were applied. The energy of the complete peptide–membrane system was minimized using the steepest descent algorithm until the maximum force changed by less than 1000 kJ·mol–1·nm–1 and then solvated using the SPC water model81 and minimized again. The total number of water molecules depended on the box size (Table 2). Each system was then neutralized by the addition of Na+ ions (Table 2) and again, the potential energy was minimized. Each system was equilibrated for 500 ps under isothermal–isobaric (NpT) conditions with the temperature maintained at physiological temperature, 310 K, using the Berendsen thermostat82 with a time constant of 1 ps, and the pressure was maintained at 1 bar using the Berendsen barostat82 with semi-isotropic pressure coupling, a time constant of 1 ps and an isothermal compressibility of 4.575 × 10–4 (kJ·mol–1·nm–3)−1. For the production runs, the temperature was maintained at 310 K using the Nosé–Hoover thermostat83,84 with a time constant of 1 ps, and the pressure of 1 bar was maintained using semi-isotropic pressure coupling using the Parrinello–Rahman barostat85 with a time constant of 5 ps and an isothermal compressibility of 4.575 × 10–4 (kJ·mol–1·nm–3)−1. For both the equilibration and production runs, long-range electrostatic interactions outside a cutoff of 1.4 nm were treated using the reaction field86 algorithm, and van der Waals interactions were truncated at 1.4 nm.

Table 2. Box Dimensions and the Total Number of Water Molecules and Na+ Ions for Each Set of Three Replicate Simulations.

  S. aureus E. coli IM
  box dimensions (nm)     box dimensions (nm)    
peptide x y z water Na+ ions x y z water Na+ ions
octapeptide 8.32510 7.20975 11.78254 14,959 51 6.00364 6.30451 10.53374 7882 31
pentapeptide 8.32810 7.21234 11.78078 14,977 53 6.00447 6.30538 10.25824 7501 33
NH2-octapeptide 8.31097 7.19750 11.97127 14,635 56 6.10694 6.41299 9.65750 7914 36
NH2-pentapeptide 8.30443 7.19184 12.10536 14,973 56 5.97560 6.27507 9.44980 5867 36
Open in a new tab

Each peptide was first simulated alone in solution for 500 ns. Each peptide–membrane system, comprising one copy of a given peptide with one of the two types of membrane, was first simulated in quintuplicate for 50 ns, and three of these were extended to 500 ns.

Analysis

All analysis was carried out using GROMACS tools unless otherwise specified. Conformational clustering was carried out using the GROMOS87 clustering method. An RMSD cutoff of 0.25 nm was selected as this value results in more than 60% of the sampled structures being grouped into clusters. Only the 20 most populated clusters were analyzed. Partial electron densities along the z axis (perpendicular to the plane of the membrane) were calculated with the system vertically centered to the middle of the lipid bilayer. Intermolecular hydrogen bonds between the peptide and the membrane were calculated using the Visual Molecular Dynamics (VMD) software88 and plotted as a time series with the help of Python and Tcl scripts. Only hydrogen bonds that appeared over 10% of the simulation were plotted. The interaction potential energies are the short-range (within the nonbonded cutoff) nonbonded potential energies between the specified groups, divided into the contributions from the Coulombic and van der Waals (Lennard–Jones) force field terms. While it is the free energy that ultimately determines the behavior of the system, analyzing the potential energies in this way provides insight into which groups of atoms are interacting and the nature of these interactions.

Acknowledgments

We acknowledge the use of high-performance computing facilities provided by the Centre for Theoretical Chemistry and Physics, Massey University (Auckland) and by the New Zealand eScience Infrastructure (NeSI, https://www.nesi.org.nz) as part of this research. NeSI is funded jointly by NeSI’s collaborator institutions and through the Ministry of Business, Innovation, & Employment’s Research Infrastructure programme. J.R.A. is supported by a Rutherford Discovery Fellowship (15-MAU-001), A.C. is supported by a University of Auckland Doctoral Scholarship, and E.K. was supported by a Commonwealth PhD Scholarship.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.0c04752.

  • Snapshots showing the time evolution of the peptide–membrane systems; time-series of the formation of hydrogen bonds between peptide residues and membrane lipids.

The authors declare no competing financial interest.

Supplementary Material

ao0c04752_si_001.pdf (2.7MB, pdf)

References

  1. Lammie S. L.; Hughes J. M. Antimicrobial Resistance, Food Safety, and One Health: The Need for Convergence. Annu. Rev. Food Sci. Technol. 2016, 7, 287–312. 10.1146/annurev-food-041715-033251. [DOI] [PubMed] [Google Scholar]
  2. WHO . Antibiotic Resistance; World Health Organisation; 2020. [Google Scholar]
  3. Medina E.; Pieper D. H. Tackling Threats and Future Problems of Multidrug-Resistant Bacteria. Curr. Top. Microbiol. Immunol. 2016, 398, 3–33. 10.1007/82_2016_492. [DOI] [PubMed] [Google Scholar]
  4. Levy S. B.; Marshall B. Antibacterial Resistance Worldwide: Causes Challenges and Responses. Nat. Med. 2004, S122. 10.1038/nm1145. [DOI] [PubMed] [Google Scholar]
  5. Stanton T. B. A Call for Antibiotic Alternatives Research. Trends Microbiol. 2013, 111. 10.1016/j.tim.2012.11.002. [DOI] [PubMed] [Google Scholar]
  6. Brogden K. A. Antimicrobial Peptides: Pore Formers or Metabolic Inhibitors in Bacteria?. Nat. Rev. Microbiol. 2005, 3, 238–250. 10.1038/nrmicro1098. [DOI] [PubMed] [Google Scholar]
  7. Hancock R. E. W.; Sahl H. G. Antimicrobial and Host-Defense Peptides as New Anti-Infective Therapeutic Strategies. Nat. Biotechnol. 2006, 24, 1551–1557. 10.1038/nbt1267. [DOI] [PubMed] [Google Scholar]
  8. Hancock R. E. Cationic Peptides: Effectors in Innate Immunity and Novel Antimicrobials. Lancet Infect. Dis. 2001, 156. 10.1016/S1473-3099(01)00092-5. [DOI] [PubMed] [Google Scholar]
  9. Zasloff M. Antimicrobial Peptides of Multicellular Organisms. Nature 2002, 415, 389–395. 10.1038/415389a. [DOI] [PubMed] [Google Scholar]
  10. Uhlig T.; Kyprianou T.; Martinelli F. G.; Oppici C. A.; Heiligers D.; Hills D.; Calvo X. R.; Verhaert P. The Emergence of Peptides in the Pharmaceutical Business: From Exploration to Exploitation. EuPA Open Proteomics 2014, 58. 10.1016/j.euprot.2014.05.003. [DOI] [Google Scholar]
  11. Haney E. F.; Straus S. K.; Hancock R. E. W. Reassessing the Host Defense Peptide Landscape. Frontiers in Chemistry. 2019, 10.3389/fchem.2019.00043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Haney E. F.; Hancock R. E. W. Peptide Design for Antimicrobial and Immunomodulatory Applications. Biopolymers 2013, 572. 10.1002/bip.22250. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Sani M. A.; Separovic F. How Membrane-Active Peptides Get into Lipid Membranes. Acc. Chem. Res. 2016, 49, 1130–1138. 10.1021/acs.accounts.6b00074. [DOI] [PubMed] [Google Scholar]
  14. Li J.; Koh J. J.; Liu S.; Lakshminarayanan R.; Verma C. S.; Beuerman R. W. Membrane Active Antimicrobial Peptides: Translating Mechanistic Insights to Design. Front. Neurosci. 2017, 11, 73. 10.3389/fnins.2017.00073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Wang Y.; Schlamadinger D. E.; Kim J. E.; McCammon J. A. Comparative Molecular Dynamics Simulations of the Antimicrobial Peptide CM15 in Model Lipid Bilayers. Biochim. Biophys. Acta 2012, 1818, 1402–1409. 10.1016/j.bbamem.2012.02.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Wang Y.; Chen C. H.; Hu D.; Ulmschneider M. B.; Ulmschneider J. P. Spontaneous Formation of Structurally Diverse Membrane Channel Architectures from a Single Antimicrobial Peptide. Nat. Commun. 2016, 7, 13535. 10.1038/ncomms13535. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Jiang X.; Yang K.; Yuan B.; Gong B.; Wan L.; Patil N. A.; Swarbrick J. D.; Roberts K. D.; Schreiber F.; Wang L.; Velkov T.; Li J. Simulations of Octapeptin-Outer Membrane Interactions Reveal Conformational Flexibility Is Linked to Antimicrobial Potency. J. Biol. Chem. 2020, 15902. 10.1074/jbc.ra120.014856. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Amos S. B. T. A.; Vermeer L. S.; Ferguson P. M.; Kozlowska J.; Davy M.; Bui T. T.; Drake A. F.; Lorenz C. D.; Mason A. J. Antimicrobial Peptide Potency Is Facilitated by Greater Conformational Flexibility When Binding to Gram-Negative Bacterial Inner Membranes. Sci. Rep. 2016, 6, 37639. 10.1038/srep37639. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Chen C. H.; Starr C. G.; Troendle E.; Wiedman G.; Wimley W. C.; Ulmschneider J. P.; Ulmschneider M. B. Simulation-Guided Rational de Novo Design of a Small Pore-Forming Antimicrobial Peptide. J. Am. Chem. Soc. 2019, 141, 4839–4848. 10.1021/jacs.8b11939. [DOI] [PubMed] [Google Scholar]
  20. Shai Y. Mechanism of the Binding, Insertion and Destabilization of Phospholipid Bilayer Membranes by α-Helical Antimicrobial and Cell Non-Selective Membrane-Lytic Peptides. Biochim. Biophys. Acta 1999, 1462, 55–70. 10.1016/S0005-2736(99)00200-X.. [DOI] [PubMed] [Google Scholar]
  21. Zhang L. J.; Gallo R. L. Antimicrobial Peptides. Curr. Biol. 2016, 26, R14–R19. 10.1016/j.cub.2015.11.017. [DOI] [PubMed] [Google Scholar]
  22. Mahlapuu M.; Håkansson J.; Ringstad L.; Björn C. Antimicrobial Peptides: An Emerging Category of Therapeutic Agents. Front. Cell. Infect. Microbiol. 2016, 6, 194. 10.3389/fcimb.2016.00194. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hancock R. E. W.; Diamond G. The Role of Cationic Antimicrobial Peptides in Innate Host Defences. Trends Microbiol. 2000, 8, 402–410. 10.1016/S0966-842X(00)01823-0. [DOI] [PubMed] [Google Scholar]
  24. Lee T.-H.; Hall K. N.; Aguilar M.-I. Antimicrobial Peptide Structure and Mechanism of Action: A Focus on the Role of Membrane Structure. Curr. Top. Med. Chem. 2015, 25. 10.2174/1568026615666150703121700. [DOI] [PubMed] [Google Scholar]
  25. Talandashti R.; Mahdiuni H.; Jafari M.; Mehrnejad F. Molecular Basis for Membrane Selectivity of Antimicrobial Peptide Pleurocidin in the Presence of Different Eukaryotic and Prokaryotic Model Membranes. J. Chem. Inf. Model. 2019, 59, 3262–3276. 10.1021/acs.jcim.9b00245. [DOI] [PubMed] [Google Scholar]
  26. Nguyen L. T.; Haney E. F.; Vogel H. J. The Expanding Scope of Antimicrobial Peptide Structures and Their Modes of Action. Trends Biotechnol. 2011, 29, 464–472. 10.1016/j.tibtech.2011.05.001. [DOI] [PubMed] [Google Scholar]
  27. Kumar P.; Kizhakkedathu J.; Straus S. Antimicrobial Peptides: Diversity, Mechanism of Action and Strategies to Improve the Activity and Biocompatibility in Vivo. Biomolecules 2018, 10.3390/biom8010004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Raheem N.; Straus S. K. Mechanisms of Action for Antimicrobial Peptides With Antibacterial and Antibiofilm Functions. Frontiers in Microbiology. 2019, 10.3389/fmicb.2019.02866. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Qian C. D.; Wu X. C.; Teng Y.; Zhao W. P.; Li O.; Fang S. G.; Huang Z. H.; Gao H. C. Battacin (Octapeptin B5), a New Cyclic Lipopeptide Antibiotic from Paenibacillus Tianmuensis Active against Multidrug-Resistant Gram-Negative Bacteria. Antimicrob. Agents Chemother. 2012, 56, 1458–1465. 10.1128/AAC.05580-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. De Zoysa G. H.; Cameron A. J.; Hegde V. V.; Raghothama S.; Sarojini V. Antimicrobial Peptides with Potential for Biofilm Eradication: Synthesis and Structure Activity Relationship Studies of Battacin Peptides. J. Med. Chem. 2015, 58, 625–639. 10.1021/jm501084q. [DOI] [PubMed] [Google Scholar]
  31. Glossop H. D.; De Zoysa G. H.; Hemar Y.; Cardoso P.; Wang K.; Lu J.; Valéry C.; Sarojini V. Battacin-Inspired Ultrashort Peptides: Nanostructure Analysis and Antimicrobial Activity. Biomacromolecules 2019, 20, 2515–2529. 10.1021/acs.biomac.9b00291. [DOI] [PubMed] [Google Scholar]
  32. Perrin B. S. Jr.; Pastor R. W. Simulations of Membrane-Disrupting Peptides I: Alamethicin Pore Stability and Spontaneous Insertion. Biophys. J. 2016, 111, 1248–1257. 10.1016/j.bpj.2016.08.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Wimley W. C.; Hristova K. Antimicrobial Peptides: Successes, Challenges and Unanswered Questions. J. Membr. Biol. 2011, 239, 27–34. 10.1007/s00232-011-9343-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kihara S.; De Zoysa G. H.; Shahlori R.; Vadakkedath P. G.; Ryan T. M.; Mata J. P.; Sarojini V.; McGillivray D. J. Solution Structure of Linear Battacin Lipopeptides-the Effect of Lengthening Fatty Acid Chain. Soft Matter 2019, 15, 7501–7508. 10.1039/c9sm00932a. [DOI] [PubMed] [Google Scholar]
  35. La Rocca P.; Biggin P. C.; Tieleman D. P.; Sansom M. S. P. Simulation Studies of the Interaction of Antimicrobial Peptides and Lipid Bilayers. Biochim. Biophys. Acta 1999, 1462, 185–200. 10.1016/S0005-2736(99)00206-0. [DOI] [PubMed] [Google Scholar]
  36. Tieleman D. P.; Sansom M. S. P. Molecular Dynamics Simulations of Antimicrobial Peptides: From Membrane Binding to Trans-Membrane Channels. Int. J. Quantum Chem. 2001, 83, 166–179. 10.1002/qua.1208. [DOI] [Google Scholar]
  37. Matyus E.; Kandt C.; Tieleman D. Computer Simulation of Antimicrobial Peptides. Curr. Med. Chem. 2007, 14, 2789–2798. 10.2174/092986707782360105. [DOI] [PubMed] [Google Scholar]
  38. Mondal J. A Brief Appraisal of Computational Modeling of Antimicrobial Peptides’ Activity. Drug Dev. Res. 2019, 80, 28–32. 10.1002/ddr.21472. [DOI] [PubMed] [Google Scholar]
  39. Woo H. J.; Wallqvist A. Spontaneous Buckling of Lipid Bilayer and Vesicle Budding Induced by Antimicrobial Peptide Magainin 2: A Coarse-Grained Simulation Study. J. Phys. Chem. B 2011, 115, 8122–8129. 10.1021/jp2023023. [DOI] [PubMed] [Google Scholar]
  40. Thøgersen L.; Schiøtt B.; Vosegaard T.; Nielsen N. C.; Tajkhorshid E. Peptide Aggregation and Pore Formation in a Lipid Bilayer: A Combined Coarse-Grained and All Atom Molecular Dynamics Study. Biophys. J. 2008, 95, 4337–4347. 10.1529/biophysj.108.133330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Santo K. P.; Berkowitz M. L. Difference between Magainin-2 and Melittin Assemblies in Phosphatidylcholine Bilayers: Results from Coarse-Grained Simulations. J. Phys. Chem. B 2012, 116, 3021–3030. 10.1021/jp212018f. [DOI] [PubMed] [Google Scholar]
  42. Balatti G.; Ambroggio E.; Fidelio G.; Martini M.; Pickholz M. Differential Interaction of Antimicrobial Peptides with Lipid Structures Studied by Coarse-Grained Molecular Dynamics Simulations. Molecules 2017, 22, 1775. 10.3390/molecules22101775. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Catte A.; Wilson M. R.; Walker M.; Oganesyan V. S. Antimicrobial Action of the Cationic Peptide, Chrysophsin-3: A Coarse-Grained Molecular Dynamics Study. Soft Matter 2018, 14, 2796–2807. 10.1039/c7sm02152f. [DOI] [PubMed] [Google Scholar]
  44. Bond P. J.; Parton D. L.; Clark J. F.; Sansom M. S. P. Coarse-Grained Simulations of the Membrane-Active Antimicrobial Peptide Maculatin 1.1. Biophys. J. 2008, 95, 3802–3815. 10.1529/biophysj.108.128686. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Leontiadou H.; Mark A. E.; Marrink S. J. Antimicrobial Peptides in Action. J. Am. Chem. Soc. 2006, 128, 12156–12161. 10.1021/ja062927q. [DOI] [PubMed] [Google Scholar]
  46. Rice A.; Wereszczynski J. Probing the Disparate Effects of Arginine and Lysine Residues on Antimicrobial Peptide/Bilayer Association. Biochim. Biophys. Acta, Biomembr. 2017, 1859, 1941–1950. 10.1016/j.bbamem.2017.06.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Sani M. A.; Le Brun A. P.; Separovic F. The Antimicrobial Peptide Maculatin Self Assembles in Parallel to Form a Pore in Phospholipid Bilayers. Biochim. Biophys. Acta, Biomembr. 2020, 1862, 183204. 10.1016/j.bbamem.2020.183204. [DOI] [PubMed] [Google Scholar]
  48. Bennett W. F. D.; Tieleman D. P. Water Defect and Pore Formation in Atomistic and Coarse-Grained Lipid Membranes: Pushing the Limits of Coarse Graining. J. Chem. Theory Comput. 2011, 7, 2981–2988. 10.1021/ct200291v. [DOI] [PubMed] [Google Scholar]
  49. Romoli O.; Mukherjee S.; Mohid S. A.; Dutta A.; Montali A.; Franzolin E.; Brady D.; Zito F.; Bergantino E.; Rampazzo C.; Tettamanti G.; Bhunia A.; Sandrelli F. Enhanced Silkworm Cecropin B Antimicrobial Activity against Pseudomonas Aeruginosa from Single Amino Acid Variation. ACS Infect. Dis. 2019, 5, 1200–1213. 10.1021/acsinfecdis.9b00042. [DOI] [PubMed] [Google Scholar]
  50. Upadhyay S. K.; Wang Y.; Zhao T.; Ulmschneider J. P. Insights from Micro-Second Atomistic Simulations of Melittin in Thin Lipid Bilayers. J. Membr. Biol. 2015, 248, 497–503. 10.1007/s00232-015-9807-8. [DOI] [PubMed] [Google Scholar]
  51. Pino Angeles A.; Leveritt J. M. III; Lazaridis T. Structure of Transmembrane Pores Stabilized by Antimicrobial Peptides Magainin and PGLa. Biophys. J. 2015, 108, 549a–550a. 10.1016/j.bpj.2014.11.3015. [DOI] [Google Scholar]
  52. Leveritt J. M. III; Pino-Angeles A.; Lazaridis T. The Structure of a Melittin-Stabilized Pore. Biophys. J. 2015, 108, 2424–2426. 10.1016/j.bpj.2015.04.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Sahoo B. R.; Fujiwara T. Membrane Mediated Antimicrobial and Antitumor Activity of Cathelicidin 6: Structural Insights from Molecular Dynamics Simulation on Multi-Microsecond Scale. PLoS One 2016, 11, e0158702 10.1371/journal.pone.0158702. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Lai P. K.; Kaznessis Y. N. Insights into Membrane Translocation of Protegrin Antimicrobial Peptides by Multistep Molecular Dynamics Simulations. ACS Omega 2018, 3, 6056–6065. 10.1021/acsomega.8b00483. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Zhao L.; Cao Z.; Bian Y.; Hu G.; Wang J.; Zhou Y. Molecular Dynamics Simulations of Human Antimicrobial Peptide LL-37 in Model POPC and POPG Lipid Bilayers. Int. J. Mol. Sci. 2018, 19, 1186. 10.3390/ijms19041186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Jafari M.; Mehrnejad F.; Doustdar F. Insight into the Interactions, Residue Snorkeling, and Membrane Disordering Potency of a Single Antimicrobial Peptide into Different Lipid Bilayers. PLoS One 2017, 12, e0187216 10.1371/journal.pone.0187216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Lipkin R.; Pino-Angeles A.; Lazaridis T. Transmembrane Pore Structures of β-Hairpin Antimicrobial Peptides by All-Atom Simulations. J. Phys. Chem. B 2017, 121, 9126–9140. 10.1021/acs.jpcb.7b06591. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Perrin B. S. Jr.; Fu R.; Cotten M. L.; Pastor R. W. Simulations of Membrane-Disrupting Peptides II: AMP Piscidin 1 Favors Surface Defects over Pores. Biophys. J. 2016, 111, 1258–1266. 10.1016/j.bpj.2016.08.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Mukherjee S.; Kar R. K.; Nanga R. P. R.; Mroue K. H.; Ramamoorthy A.; Bhunia A. Accelerated Molecular Dynamics Simulation Analysis of MSI-594 in a Lipid Bilayer. Phys. Chem. Chem. Phys. 2017, 19, 19289–19299. 10.1039/c7cp01941f. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Velkov T.; Thompson P. E.; Nation R. L.; Li J. Structure-Activity Relationships of Polymyxin Antibiotics. J. Med. Chem. 2010, 53, 1898–1916. 10.1021/jm900999h. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Kanazawa K.; Sato Y.; Ohki K.; Okimura K.; Uchida Y.; Shindo M.; Sakura N. Contribution of Each Amino Acid Residue in Polymyxin B3 to Antimicrobial and Lipopolysaccharide Binding Activity. Chem. Pharm. Bull. 2009, 57, 240–244. 10.1248/cpb.57.240. [DOI] [PubMed] [Google Scholar]
  62. Jafari M.; Mehrnejad F.; Aghdami R.; Chaparzadeh N.; Razaghi Moghadam Kashani Z.; Doustdar F. Identification of the Crucial Residues in the Early Insertion of Pardaxin into Different Phospholipid Bilayers. J. Chem. Inf. Model. 2017, 57, 929–941. 10.1021/acs.jcim.6b00693. [DOI] [PubMed] [Google Scholar]
  63. Jahangiri S.; Jafari M.; Arjomand M.; Mehrnejad F. Molecular Insights into the Interactions of GF-17 with the Gram-Negative and Gram-Positive Bacterial Lipid Bilayers. J. Cell. Biochem. 2018, 119, 9205–9216. 10.1002/jcb.27187. [DOI] [PubMed] [Google Scholar]
  64. Berglund N. A.; Piggot T. J.; Jefferies D.; Sessions R. B.; Bond P. J.; Khalid S. Interaction of the Antimicrobial Peptide Polymyxin B1 with Both Membranes of E. Coli: A Molecular Dynamics Study. PLoS Comput. Biol. 2015, 11, 1004180. 10.1371/journal.pcbi.1004180. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Yim V.; Kavianinia I.; Knottenbelt M. K.; Ferguson S. A.; Cook G. M.; Swift S.; Chakraborty A.; Allison J. R.; Cameron A. J.; Harris P. W. R.; Brimble M. A. “CLipP”Ing on Lipids to Generate Antibacterial Lipopeptides. Chem. Sci. 2020, 11, 5759–5765. 10.1039/d0sc01814g. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Domingues M. M.; Inácio R. G.; Raimundo J. M.; Martins M.; Castanho M. A. R. B.; Santos N. C. Biophysical Characterization of Polymyxin B Interaction with LPS Aggregates and Membrane Model Systems. Biopolymers 2012, 338. 10.1002/bip.22095. [DOI] [PubMed] [Google Scholar]
  67. Mares J.; Kumaran S.; Gobbo M.; Zerbe O. Interactions of Lipopolysaccharide and Polymyxin Studied by NMR Spectroscopy. J. Biol. Chem. 2009, 11498. 10.1074/jbc.M806587200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Lad M. D.; Birembaut F.; Clifton L. A.; Frazier R. A.; Webster J. R. P.; Green R. J. Antimicrobial Peptide-Lipid Binding Interactions and Binding Selectivity. Biophys. J. 2007, 3575. 10.1529/biophysj.106.097774. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Lee T. H.; Hofferek V.; Separovic F.; Reid G. E.; Aguilar M. I. The Role of Bacterial Lipid Diversity and Membrane Properties in Modulating Antimicrobial Peptide Activity and Drug Resistance. Curr. Opin. Chem. Biol. 2019, 52, 85–92. 10.1016/j.cbpa.2019.05.025. [DOI] [PubMed] [Google Scholar]
  70. Morrison D. C.; Jacobs D. M. Binding of Polymyxin B to the Lipid A Portion of Bacterial Lipopolysaccharides. Immunochemistry 1976, 813. 10.1016/0019-2791(76)90181-6. [DOI] [PubMed] [Google Scholar]
  71. Chan D. I.; Prenner E. J.; Vogel H. J. Tryptophan- and Arginine-Rich Antimicrobial Peptides: Structures and Mechanisms of Action. Biochim. Biophys. Acta 2006, 1758, 1184–1202. 10.1016/j.bbamem.2006.04.006. [DOI] [PubMed] [Google Scholar]
  72. Velkov T.; Roberts K. D.; Li J. Rediscovering the Octapeptins. Nat. Prod. Rep. 2017, 34, 295–309. 10.1039/c6np00113k. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Hanwell M. D.; Curtis D. E.; Lonie D. C.; Vandermeersch T.; Zurek E.; Hutchison G. R. Avogadro: An Advanced Semantic Chemical Editor, Visualization, and Analysis Platform. Aust. J. Chem. 2012, 4, 17. 10.1186/1758-2946-4-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Piggot T. J.; Holdbrook D. A.; Khalid S. Electroporation of the E. Coli and S. Aureus Membranes: Molecular Dynamics Simulations of Complex Bacterial Membranes. J. Phys. Chem. B 2011, 115, 13381–13388. 10.1021/jp207013v. [DOI] [PubMed] [Google Scholar]
  75. Schmid N.; Eichenberger A. P.; Choutko A.; Riniker S.; Winger M.; Mark A. E.; Van Gunsteren W. F. Definition and Testing of the GROMOS Force-Field Versions 54A7 and 54B7. Eur. Biophys. J. 2011, 40, 843–856. 10.1007/s00249-011-0700-9. [DOI] [PubMed] [Google Scholar]
  76. Chandrasekhar I.; Kastenholz M.; Lins R. D.; Oostenbrink C.; Schuler L. D.; Tieleman D. P.; Van Gunsteren W. F. A Consistent Potential Energy Parameter Set for Lipids: Dipalmitoylphosphatidylcholine as a Benchmark of the GROMOS96 45A3 Force Field. Eur. Biophys. J. 2003, 32, 67–77. 10.1007/s00249-002-0269-4. [DOI] [PubMed] [Google Scholar]
  77. Kukol A. Lipid Models for United-Atom Molecular Dynamics Simulations of Proteins. J. Chem. Theory Comput. 2009, 5, 615–626. 10.1021/ct8003468. [DOI] [PubMed] [Google Scholar]
  78. Piggot T. J.; Holdbrook D. A.; Khalid S. Conformational Dynamics and Membrane Interactions of the E. Coli Outer Membrane Protein FecA: A Molecular Dynamics Simulation Study. Biochim. Biophys. Acta. 2013, 1828, 284–293. 10.1016/j.bbamem.2012.08.021. [DOI] [PubMed] [Google Scholar]
  79. Abraham M. J.; Murtola T.; Schulz R.; Páll S.; Smith J. C.; Hess B.; Lindah E. Gromacs: High Performance Molecular Simulations through Multi-Level Parallelism from Laptops to Supercomputers. SoftwareX 2015, 1-2, 19–25. 10.1016/j.softx.2015.06.001. [DOI] [Google Scholar]
  80. Hess B.; Bekker H.; Berendsen H. J. C.; Fraaije J. G. E. M. LINCS: A Linear Constraint Solver for Molecular Simulations. J. Comput. Chem. 1997, 18, 1463–1472. . [DOI] [Google Scholar]
  81. Berendsen H. J. C.; Postma J. P. M.; van Gunsteren W. F.; Hermans J. Interaction Models for Water in Relation to Protein Hydration. Intermolecular Forces 1981, 14, 331–342. 10.1007/978-94-015-7658-1_21. [DOI] [Google Scholar]
  82. Berendsen H. J. C.; Postma J. P. M.; Van Gunsteren W. F.; Dinola A.; Haak J. R. Molecular Dynamics with Coupling to an External Bath. J. Chem. Phys. 1984, 81, 3684–3690. 10.1063/1.448118. [DOI] [Google Scholar]
  83. Nosé S. A Molecular Dynamics Method for Simulations in the Canonical Ensemble. Mol. Phys. 2002, 100, 191–198. 10.1080/00268970110089108. [DOI] [Google Scholar]
  84. Hoover W. G. Canonical Dynamics: Equilibrium Phase-Space Distributions. Phys. Rev. A 1985, 31, 1695–1697. 10.1103/PhysRevA.31.1695. [DOI] [PubMed] [Google Scholar]
  85. Parrinello M.; Rahman A. Polymorphic Transitions in Single Crystals: A New Molecular Dynamics Method. J. Appl. Phys. 1981, 52, 7182–7190. 10.1063/1.328693. [DOI] [Google Scholar]
  86. Van Gunsteren W. F.; Berendsen H. J. C.; Rullmann J. A. C. Inclusion of Reaction Fields in Molecular Dynamics: Application to Liquid Water. Faraday Discuss. Chem. Soc. 1978, 66, 58–70. 10.1039/DC9786600058. [DOI] [Google Scholar]
  87. Daura X.; Gademann K.; Jaun B.; Seebach D.; van Gunsteren W. F.; Mark A. E. Peptide Folding: When Simulation Meets Experiment. Angew. Chem., Int. Ed. 1999, 38, 236–240. . [DOI] [Google Scholar]
  88. Humphrey W.; Dalke A.; Schulten K. VMD - Visual Molecular Dynamics. J. Mol. Graphics 1996, 14, 33–38. 10.1016/0263-7855(96)00018-5. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

ao0c04752_si_001.pdf (2.7MB, pdf)

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES