Model membrane interactions and biological activity of a naphthalimide-containing BP 100

In a large variety of organisms, antimicrobial peptides (AMPs) are primary defences against pathogens. BP100 (KKLFKKILKYL-NH2), a short, synthetic, and cationic AMP, is active against bacteria and displays low toxicity towards eukaryotic cells. BP100 acquires an α-helical conformation upon interaction with membranes and increases membrane permeability. Despite the volume of information available, the mechanism of action of BP100, the selectivity of its biological effects, and its applications are far from consensual. In this work, we synthesized a fluorescent BP100 analog containing naphthalimide linked to its N-terminal end, Napht-BP100 (Napht-AAKKLFKKILKYL-NH2). The fluorescence properties of naphthalimides, especially their spectral sensitivity to microenvironment changes, are well established, and their biological activities against different types of cells are known. A wide variety of techniques were used to demonstrate that a-helical Napht-BP100 was bound and permeabilized POPC and POPG LUV. Napht-BP100, different from that observed for BP100, was bound to, and permeabilized POPC LUV. With zwitterionic (POPC) and negatively charged (POPG) containing LUVs, membrane surface high peptide/lipid ratios triggered complete disruption of the liposomes in a detergent-like manner. This disruption was driven by charge neutralization, lipid aggregation, and membrane destabilization. Napht-BP100 also interacted with double-stranded DNA, indicating that this peptide could also affect other cellular processes in addition to membrane destabilization. Napht-BP100 showed superior antibacterial activity, increased hemolytic activity compared to BP100, and may constitute an efficient antimicrobial agent for dermatological use. By conjugating BP100 and naphthalimide antimicrobial properties, Napht-BP100 was bound more efficiently to the bacterial membrane and could destabilize the membrane and enter the cell by interacting with its cytoplasmexposed DNA. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 2 February 2021 doi:10.20944/preprints202102.0098.v1


Abstract
In a large variety of organisms, antimicrobial peptides (AMPs) are primary defences against pathogens. BP100 (KKLFKKILKYL-NH2), a short, synthetic, and cationic AMP, is active against bacteria and displays low toxicity towards eukaryotic cells. BP100 acquires an α-helical conformation upon interaction with membranes and increases membrane permeability. Despite the volume of information available, the mechanism of action of BP100, the selectivity of its biological effects, and its applications are far from consensual. In this work, we synthesized a fluorescent BP100 analog containing naphthalimide linked to its N-terminal end, Napht-BP100 (Napht-AAKKLFKKILKYL-NH2). The fluorescence properties of naphthalimides, especially their spectral sensitivity to microenvironment changes, are well established, and their biological activities against different types of cells are known. A wide variety of techniques were used to demonstrate that a-helical Napht-BP100 was bound and permeabilized POPC and POPG LUV. Napht-BP100, different from that observed for BP100, was bound to, and permeabilized POPC LUV. With zwitterionic (POPC) and negatively charged (POPG) containing LUVs, membrane surface high peptide/lipid ratios triggered complete disruption of the liposomes in a detergent-like manner. This disruption was driven by charge neutralization, lipid aggregation, and membrane destabilization. Napht-BP100 also interacted with double-stranded DNA, indicating that this peptide could also affect other cellular processes in addition to membrane destabilization. Napht-BP100 showed superior antibacterial activity, increased hemolytic activity compared to BP100, and may constitute an efficient antimicrobial agent for dermatological use. By conjugating BP100 and naphthalimide antimicrobial properties, Napht-BP100 was bound more efficiently to the bacterial membrane and could destabilize the membrane and enter the cell by interacting with its cytoplasm-exposed DNA.

Introduction
Antimicrobial peptides (AMPs) can destroy or inhibit the growth of bacteria, fungi, and viruses [1]. AMPs are ubiquitous components of the innate immune system and act as endogenous antibiotics [2,3,4,5]. AMPs are positively charged and display a hydrophobicity index and hydrophobic moment, compatible with interactions with the bacterial membrane [6,7]. The antibacterial activity of AMPs arises from electrostatic interactions with bacterial membranes, rich in negatively charged components such as phosphates, lipopolysaccharides from gram-negative bacteria, or lipoteichoic acids present in gram-positive bacteria. As the negative charge density of mammalian cell membranes is lower than that of bacteria, the electrostatic component is the main element of selectivity towards bacteria in cationic peptides' action [3,4,5,6].
Peptide flip and hydrophobic residue exposure to the membrane interior [8] may follow electrostatic binding [8]. AMPs, in addition to the high density of positively charged side chains, contain tryptophan, tyrosine, and phenylalanine, with a high affinity to the membrane interface [2,4,5,9]. After interaction with the membrane, the AMPs hydrophobic/hydrophilic topological distribution acquires a secondary structure that provides the peptide with a spatial amphipathic character, favoring the interaction with the membrane interface [4,7]. Dehydration of the hydrophobic moieties, and not just the electrostatic components, can determine the bonding selectivity of AMPs to bacterial membranes [10].
Badosa and coworkers designed a series of AMP's to identify antimicrobial structure/potency relationships [11]. BP100 (KKLFKKILKYL-NH2) ( Figure 1A) combines the properties of melittin and cecropin A, acts by inhibiting the growth bacteria, exhibits low toxicity, high therapeutic index, and low sensitivity to degradation [12]. Atomic force microscopy showed that BP100 destroys the bacterial outer envelope at a minimum inhibitory concentration (MIC) of 3 μM [13]. The extent of damage is related to peptide binding and neutralization of the cell membrane's surface charge. Circular dichroism (CD) and in silico analysis showed that the membrane-bound form of BP100 had an α-helix content of 61% [13,14,15,16,17]. The acquisition of a helical secondary structure results in an amphipathic structure, ideal for a peptide/negatively charged lipid bilayer interaction. Circular-oriented dichroism (OCD) and solid-state nuclear magnetic resonance (SS-NMR) of BP100 labeled with 19 F showed that the highly mobile helix is positioned on the membrane with its long axis parallel to the membrane surface [16]. After an initial electrostatic driven approach, helical structured BP100 accommodates at the interface by flipping along its helix longer axis and inserting the hydrophobic side into the membrane hydrophobic acyl chains. Two other relevant phenomena regarding binding and flipping processes were also accessed: negative lipid clustering and peptide dehydration [8]. Microorganisms, or transformed cell disruption by BP100, and several BP100 analogs, and the effects resulting from groups linked to the parental peptide [17,18], were analyzed [18].
Advances in the synthesis of naphthalimide analogs have made it possible to explore derivatives such as mono-naphthalimides, di-naphthalimides, and naphthalimides conjugated with other compounds that exhibited different degrees of antibacterial activity depending on the attributed modifications [24,25,26]. However, the conjugating antimicrobial potential of an AMP and a DNA binding motif such as naphthalimide [27,28,29] has been overlooked. In this work, by taking advantaged biophysical and biological properties of naphthalimide, we aim to expand the understanding of NAPHT-BP100 and BP100 mechanism of action, and to obtain an improved antibacterial agent combining both membrane disruption and DNA binding capabilities.
Our investigation focused on the NAPHT-BP100 secondary structure, membrane positioning, and bilayer lipid composition impact on binding, thermodynamics of NAPHT-BP100/membrane association, lipid organization, membrane surface charge, and peptide-induced liposome size changes.
Binding data were correlated with NAPHT-BP100-induced vesicle permeabilization, allowing NAPHT-BP100 activity rationalization. The peptide-DNA interaction was also investigated. Biological activity was examined by determining the minimum inhibitory peptide concentration against gramnegative and gram-positive bacterial species. Hemolytic activity against human red blood cells was measured as a mean of evaluating peptide toxicity.

Reagents
Louis, MO), was purified, and the sodium salt was prepared and quantified as described previously [15]. 1-   The DNA concentration (in base pair) was determined spectrophotometrically using an N-1000 Nanodrop spectrophotometer at 260 nm (ε260 = 0.020 (μg/mL) −1 .cm−1), and the ratio between the absorbance at 260 and 280 nm was registered to attest the sample purity.

Model membrane preparation
Lipid stock solutions were prepared in chloroform and quantified by measuring the phosphate concentration [35]. Lipid films with the desired amount and molar ratios were prepared by mixing an adequate volume of each lipid

Fluorescence
Steady-state fluorescence spectra of NAPHT-BP100 in Tris-HCl 10 mM, pH 7.4 buffer, containing or not NaCl 300 mM, were obtained using a Hitachi Peptide-DNA interaction was studied by varying DNA concentration from 0 to 15 ng/µL. Spectra of Tris-HCl 10 mM, pH 7.4, buffer, either containing or without NaCl 300 mM, and of the vesicles in buffer without peptide, taken under the same conditions, were subtracted from the peptide spectra for correction. In addition, spectra were also corrected by peptide dilution as the lipid addition proceeded.

Circular dichroism
Circular dichroism spectra of NAPHT-BP100 in Tris-HCl 10 mM, pH 7.4 buffer were obtained using a Jasco J-720 spectropolarimeter (Jasco, Easton, MD) at room temperature. Samples were placed in a 0.1 cm optical length quartz cells and spectra were scanned from 190 to 260 nm, at a rate of 50 nm/min, with bandwidth of 2 nm, step resolution of 0.5 nm, response time of 2 seconds, and the final spectrum was the average of six scans. The Initial peptide concentration was 20 µM, and lipid concentration was varied from 0 to 1.6 mM.
The Spectra of the buffer and the vesicles in buffer were obtained under the same conditions and subtracted from the CD spectra of the peptide. Finally, the ellipticity intensities (θ, mdeg) were normalized to molar ellipticity ([θ], deg.cm 2 .dmol -1 ) using Equation 1 to eliminate the spectral dependence on optical length, peptide concentration and number of residues.
[θ] = θ / (10*C*l*N) Equation 1 where C is peptide concentration in mol/L, N is the number of residues, and l is the cell optical length in cm.

Dynamic light scattering (DLS)
LUV hydrodynamic diameter and size distribution, and electrophoretic mobility were measured in a Zetasizer Nano apparatus equipped with a 633 nm laser (Malvern, Worcestershire, UK). The LUV surface zeta potential was calculated from the electrophoretic mobility using Henry's equations (Equations 2 and 3).
where ζ is the zeta potential, UE is the electrophoretic mobility, Ɛ the dielectric constant of water, f(ka) is Henry's function, and ƞ is the viscosity of the medium.

Differential scanning calorimetry
The phase transition temperature (Tm), cooperativity (ΔT1/2) and enthalpy where FP is the fluorescence emission intensity of the well containing the peptide after 30 min, and F0 and FT are the fluorescence emission intensities of the negative (before peptide addition) and positive (after polidocanol addition and 100 % permeabilization) control, respectively.

Minimum inhibitory concentration assay
The assays for determining the minimum peptide concentration necessary to inhibit bacterial growth were carried out according to Wiegand et al. [36]. Escherichia coli (ATCC 25922), Staphylococcus aureus (ATCC 25923) and Bacillus subtilis (PY79) bacterial species were tested for the assays, which were carried out in triplicate. Initially

Hemolytic activity
The hemolytic activity of BP100 and NAPHT-BP100 was evaluated following Mojsoska et al. [37].  where APEP, ANEG and APOS refer to the sample's absorbance with the peptide, and the positive and negative controls, respectively.

UV-absorption and fluorescence -NAPHT-BP100 membrane and ds-DNA binding
The NAPHT-BP100 UV absorption spectrum shows three characteristic peaks at 343, 275 and 230 nm corresponding to the absorption from the naphthalimide moiety, the tyrosine side chain, and the peptide bond, respectively (Figure 2 A). From this result, the maximum absorption wavelength of the naphthalimide moiety was determined and used as the excitation wavelength to study peptide fluorescence properties.
The emission spectrum of NAPHT-BP100 presented peaks at 385 and 400 nm, and the fluorescence emission intensity was lower at high salt concentrations (Figure 2 B). The fluorophore was likely less hydrated in 0.3 M NaCl, causing a decrease in fluorescence emission intensity but no peak displacement [38]. As the temperature increased from 25 to 65 °C, the fluorescence emission of NAPHT-BP100 in solution decreased slightly with no changes in peak positioning or spectral shape ( Figure SM-5 A).
Upon POPC: POPG (50:50, molar ratio) LUV addition, NAPHT-BP100 fluorescence emission intensity decreased, and both emission peaks were blueshifted by 10 nm (Figure 2 B, C). Blue shifts and decrease in emission intensity decrease are associated with fluorophore transfer to a more hydrophobic environment [38]. The emission spectra of membrane-bound NAPHT-BP100 in high or low salt were similar (Figure 2 B), suggesting that the fluorophore was inside the membrane.
The free/bound ratios of membrane association of NAPHT-BP100 were calculated from the lipid concentration-dependence of the fluorescence emission, assuming a two-state model. (Figure 2 C, D, and E). The addition of POPC-LUV to NAPHT-BP100 in low or high salt, reaching lipid/peptide ratios up to 240, also triggered a decrease, demonstrating peptide-membrane interactions (Figure SM-6). However, only at high salt full binding was observed ( Figure 2D).
Interaction between NAPHT-BP100 and LUV containing 30 to 70 mol% of POPG occurs primarily due to electrostatic interactions between the negatively charged bilayer and the positively charged peptide [15,17]. The addition of LUV to NAPHT-BP100 decreased fluorescence emission to a minimum where further lipid additions ceased to produce spectral change, indicating that the peptide was fully liposome-bound.
We calculated the lipid/peptide ratio from the binding isotherms, where 50 % of NAPHT-BP100 was LUV-bound (L/P50). (Figure 2 D and E). The addition of POPG to the LUV lipids increased, as expected [15], NAPHT-BP100 binding, and the binding extent increased with salt ( Table 1). L/P50 was less dependent on the POPG contents at high salt. Note the differences between peptide affinity to LUV with 30 and 50 mol% of POPG (Table 1). Increasing the temperature from 25 to 65 °C did not change the binding degree ( Figure SM-5   B).  Naphthalimide derivative fluorescence spectra are also sensitive to the interaction of these molecules with mono-and oligonucleotides [39]. Taking advantage of this property, NAPHT-BP100 binding to ds-DNA was investigated by fluorescence spectroscopy (Figure 3). The presence of ds-DNA in solution triggered a decrease in naphthalimide fluorescence emission attesting peptide-DNA interaction (Figure 3 A), and the degree could be further explored by varying the ds-DNA concentration to obtain a binding isotherm in the same way as accomplished in the studies with LUV (Figure 3 B).
NAPHT-BP100 bound to a large extent to ds-DNA, considering that complete binding was achieved at a ds-DNA base pair/NAPHT-BP100 ratio of 2.5. At this ratio, NAPHT-BP100 five positively charged groups were stoichiometrically neutralized by five negatively charged phosphate groups of the 2.5 ds-DNA base pairs. In this case, both charge neutralization and naphthalimide intercalation in the DNA [27] would contribute to a large degree of interaction.

Circular Dichroism
NAPHT-BP100 in aqueous solution displayed a far-UV CD spectrum with a negative peak at 198 nm and a negative and low-intensity peak centered around 230 nm, indicating that the peptide was in a random/flexible structure with a low degree of helical secondary structure (Figure 4 A). This conformation was different from that of BP100 that, under the same conditions, displays a completely flexible structure [15,17]. Naphthalimide addition to the N-terminus of BP100 in NAPHT-BP100 induced the attainment of some degree of secondary structure. The CD spectra of NAPHT-BP100, with LUV (30, 50, or 70 mol% POPG), are typical of an α-helical secondary structure, displaying a positive peak at 195 nm and two negative bands at 208 and 222 nm (Figure 4 A). The interaction of NAPHT-BP100 and other related peptides [14,15,17]with negatively charged vesicles triggered a coil-to-helix transition. The 208/222 nm intensity peak ratio of the spectra of NAPHT-BP100 with 70 mol% of POPG LUV differed from the ratios calculated from the spectra of the peptide bound to LUV containing 30 or 50 mol% of POPG. These differences suggested that NAPHT-BP100, as discussed previously for BP100 [15], aggregated on the membrane surface and induced lipid clustering.
With LUV's composed of POPC, the NAPHT-BP100 CD spectrum changes indicated a coil-to-helix transition (Figure 4 A), but complete binding was not obtained even at a lipid/peptide ratio of 190. Previous SS-NMR results with BP100 showed that the membrane-bound form of BP100 and other analogs adopt a helical secondary structure regardless of the surface net charge [40].
The isodichroic point at 204 nm indicated that the NAPHT-BP100: POPG:POPC (1:1) LUV interaction was compatible with a two-state model for binding (Figure

B).
Peptide-membrane interactions between the different lipid systems studied and the techniques employed were analyzed from binding isotherms (Figure 4 C, Table 1). CD data indicated that peptide-membrane affinity increased with POPG content ( Table 1).
CD spectral deconvolution allowed quantitative analysis of the secondary structure of NAPHT-BP100 ( Table 2) [41]. In solution, NAPHT-BP100 displayed 81% of the flexible random coil structure (10.5 residues) and 19 % of α-helical secondary structure (2.5 residues). Binding to lipid membranes containing POPG-triggered a structural transition in which NAPHT-BP100 acquires approximately 75-83% of helical conformation (10 residues) and 17-25 % of its length remaining in a random coil structure. Interestingly, the bound form of NAPHT-BP100 in the presence of LUV containing 70% of POPG indicates the presence of 75% α-helix and 17% β-sheet secondary structures, supporting the notion that the peptide aggregated on the membrane surface, indicating that aggregation could occur through the formation of β-sheet structured segments.
The NAPHT-BP100 spectrum in the presence of POPC LUV was not analyzed since complete peptide binding was not achieved, and thus no representative spectrum of the peptide bound form was obtained.

Isothermal titration calorimetry
NAPHT-BP100 binding to LUV was exothermic (Figure 5). The Measured heat changes allowed for an analysis of the binding thermodynamics ( Table 3).
The calorimetric data indicated that NAPHT-BP100 binding was driven by a negative enthalpy component and a positive entropy contribution ( Table 3).

Enthalpy variation can be ascribed to electrostatic interactions between
NAPHT-BP100 and the membrane surface, while entropy variations are most likely related to peptide dehydration and hydrophobic interactions between peptide hydrophobic helix face and lipid acyl chains. Binding of some hydrotropic ions to zwitterionic interfaces is controlled by the dehydration of their hydrophobic moieties [42].

Carboxyfluorescein leakage
BP100 and its analogs induce membrane permeabilization [17], and NAPHT-BP100 also causes CF leakage from lipid vesicles. Results are expressed in terms of lipid/peptide ratios because, as previously demonstrated [43], the extent of vesicle permeabilization depends on the peptide/lipid ratio.
NAPHT-BP100-induced membrane permeability increased with peptide concentration (Figure 6 A and B). At a lipid/peptide ratio of 10, condition which, according to binding studies (Section 3.1, Figure 2 D  Lipid/peptide ratio required to release 50% of CF from LUV as a function of POPG content. Measurements performed at 37 °C. BP100 data from Carretero et al. [17].
NAPHT-BP100 permeabilized POPC LUV, but the incorporation of POPG in LUV composition determined a greater leakage extent of leakage.
Permeabilization extent and efficiency were analyzed by calculating the lipid/peptide ratio that caused 50% CF leakage (L/P50). The Calculated L/P50 ratios increased with POPG (Figure 6 D). With a lipid/peptide ratio of 20, where NAPHT-BP100 is bound to LUV, the leakage extent increased from 24.9 to 43.6 and 66.4% by increasing the POPG content from 0 to 30 and 50 mol%. Lower permeabilization of POPC LUV confirmed both CD and fluorescence binding results, which showed that NAPHT-BP100 interacts with a zwitterionic bilayer composed of POPC, although to a much lower extent when compared to POPG-containing membranes.
Comparing BP100 with NAPHT-BP100, it was clear that the additional alanine residues and naphthalimide increased the permeation effectiveness of NAPHT-BP100 and decreased the POPG-dependence for leaking.

DLS
NAPHT-BP100 triggered vesicle aggregation at peptide/lipid ratios above 0.12, and the apparent hydrodynamic diameter (Dh) increased to 1200 nm. The Dh increase was accompanied by an increase in the size distribution, as shown by the values of PdI and the measurement error (Figure 7).
Vesicle aggregation was correlated to the membrane net surface charge by analyzing Zeta potential (ZP) variations increasing peptide/lipid ratios (Figure 7 B). As previously reported for BP100 and other AMPs [17], vesicle aggregation is directly related to membrane charge neutralization caused by positively charged peptide binding to the negatively charged membrane. At a 0.12 peptide/lipid ratio, the membrane was neutral and, without vesicle-to-vesicle charge repulsion, LUV aggregated. Membrane surface charge varied from -35 mV to +20 mV when the peptide/lipid ratios changed between 0.10 and 0.15; beyond this point, added NAPHT-BP100 did not interact with vesicle aggregates, probably due to electrostatic repulsion, and no further Dh or ZP changes were observed (Figure 7 B).  The pre-transition was up-shifted with 25 µM NAPHT-BP100 (lipid/peptide ratio 80) and superimposed with the main transition (Figure 8 A, B, and C). The addition of 6, 12, or 25 µM NAPHT-BP100 transition up-shifted the main transition by 0.7 °C (Figure 8 A and C), decreased the cooperativity and increased the value of ΔT1/2 (Figure 8 A and D). The phase transition enthalpy of DPPC:DPPG mixed vesicles at pH 7.4, was ~ 9 kcal/mol and peptide addition did not significantly modify this parameter significantly (Figure 8 E). The limited destabilization of the DPPC:DPPG gel phase by NAPHT-BP100 may be related to its interfacial position in the bilayer [8].

Minimum inhibitory concentration
The additional two alanine residues and the naphthalimide group linked to the BP100 sequence improved the peptide's ability for inhibit bacterial growth against E. coli and S. aureus and did not alter the activity against B. subtilis [17]. No exact correlation regarding improving activity and specie Gram stain could be drawn. The higher inhibitory activity of NAPHT-BP100, in comparison with BP100, correlates with NAPHT-BP100 greater membrane affinity and is related to NAPHT-BP100 ability to bind bacterial cytoplasmic DNA.

Discussion
The Synthesis, properties, and biological activities of amphiphilic peptides covalently linked to groups of diverse chemical structures are of fundamental and applied interest [3,17,46,47].
Membrane-NAPHT-BP100 binding was extensively studied using different techniques (Figures 2, 4 and 5, Tables 1 and 3). The extra two alanine residues used as a spacer and the aromatic rings of the naphthalimide moiety at the BP100 N-terminal increased molecular hydrophobicity and contributed to peptide-membrane interaction by decreasing the energy to accommodate the peptide on the bilayer hydrophilic/hydrophobic interface regardless of the membrane surface charge -NAPHT-BP100, differently from BP100, was able to bind to POPC LUV. Helix stabilization given by the addition of the alanine residues at the BP100 N-terminal sequence would also play an important role in increasing peptide-membrane interactions.
ITC data analysis allowed us confirmed that peptide-membrane interaction is essentially driven by electrostatic interactions (Figure 5, Table 3 Table 3).
The main phenomena associated with entropy variation are peptide dehydration and secondary structure acquisition as well as the hydrophobic effect driving the interaction between the α-helix non-polar face and lipid acyl chains. Although helix acquisition represents a configurational entropy loss, peptide dehydration and hydrophobic effects increase the overall entropy, resulting in a significant energetic contribution in the total free energy variation.
The secondary structure in water bound to the LUV was investigated by CD ( Figure 4, Table 2). CD spectra showed that the peptide undergoes a coil to an α-helix upon binding to LUV containing POPG.
In solution, NAPHT-BP100 presents a small content of α-helix (19 %, or 2.5 residues) and is essentially in a flexible random conformation (81 %, or 10.5 residues) (Figure 4, Table 2). The small but significant content of the α-helix secondary structure is most likely an effect of the Napht-Ala-Ala segment attached to the BP100 sequence since, in solution, BP100 is completely in random conformation [8,14,15,16]. As reported, the addition of chemical groups to the BP100 N-terminal affected its structure in water by enhancing the content of α-helix [17] and alanine residues are known to play a role as α-helix stabilizers [49].
Binding to lipid bilayers containing POPG triggered the acquisition of an αhelical structure (80 %, or 10.4 residues) (Figure 4, Table 2). Conformational changes are directly related to peptide dehydration occurring as peptides interacts more closely with the bilayer surface, and to the lower availability of water molecules in the membrane surface that trigger the establishment of intermolecular hydrogen bonds in the peptide that adopts an α-helical secondary structure [8,50]. Helical wheel projection and theoretical calculations of NAPHT-BP100 indicate that the peptide, BP100, forms an amphipathic αhelix with two extra alanine residues attached to the N-terminal end lying in the hydrophobic face of the structure [51] (Figure 10). According to the Eisenberg plot [52], NAPHT-BP100 the overall hydrophobicity (<H> = 0.409) and significant hydrophobic moment (<µH> = 0.737) classify the peptide-formed helix as a membrane surface seeking. In addition, it has been reported that slight variations in peptide sequence, thus in its hydrophobicity and hydrophobic moment, can trigger considerable changes in peptide-membrane interaction and permeabilizing capabilities of the peptide [53,54]. Changes in liposome size and size distribution, and zeta potential triggered by NAPHT-BP100 were measured using dynamic light scattering (Figure 7), revealing that membrane charge neutralization plays an important role in membrane destabilization, ultimately causing liposome aggregation. The relationship between liposome charge neutralization and interaction stoichiometry could be detailed by zeta potential measurements, indicating that most of the peptide effect occurs in the lipid to peptide ratio range in which the Zeta potential is close to 0 mV and the sample has approximately one molecule of POPG to each lysine side chain of the peptide. This observation corroborates the stoichiometry calculated from ITC data ( Table 3) and brings relevant information regarding NAPHT-BP100 mechanism of action, suggesting a detergent-like mechanism in which charge neutralization triggers LUV aggregation and complete disruption.
The analysis of the secondary structure amphipathic profile indicated that the peptide tends to occupy an interfacial position in the bilayer, and, in agreement with the predicted, DSC experiments showed that the peptide does not cause a major effect on gel phase lipid organization, supporting the proposed shallow penetration (Figure 8). Thermodynamic lipid phase transition parameters of DPPC:DPPG (70:30) MLV and the effect of the peptide on these parameters were examined by differential scanning calorimetry (Figure 8). essentially, no significant influence was observed, especially regarding the overall process energy and the main transition temperature.
CF leakage assays to test peptide efficiency for permeabilizing LUV were evaluated considering the binding extent and structure of the peptide in the bilayer, the bilayer surface charge neutralization and the low effect of the peptide on lipid acyl chains (Figure 6). Interestingly, peptide activity efficiency followed the previously measured binding extent in terms of both lipid/peptide ratio and LUV POPG content. Although low binding of NAPHT-BP100 to POPC LUV was demonstrated, the peptide was able to bind and cause CF leakage from these liposomes, confirming the presence and relevance of hydrophobic interactions and other phenomena not associated with electrostatic interactions.
The Measured biological activity of NAPHT-BP100 against bacteria and human RBC corroborates the series of biophysical studies discussed so far.
Experiments with POPG containing LUV, especially at 30 mol%, mimicking PG molar concentrations found in bacteria, were correlated with the MIC results.
Observations taken in experiments with zwitterionic POPC LUV were correlated to the biological hemolytic activity assay considering the human RBC neutral membrane surface. In comparison with BP100, NAPHT-BP100 showed higher affinity to all studied LUV compositions, resulting in a higher effect on the membrane regardless of its composition. These observations can be translated to the observed greater biological activity of NAPHT-BP100 against E. coli and S. aureus, confirmed by the measured lower MIC values ( Table 4), and to greater toxicity against human RBC as demonstrated by the hemolysis test ( Figure 9). Increased activity of NAPHT-BP100 against E. coli and S. aureus could also relate to the internalization of NAPHT-BP100 into the bacteria and NAPHT-BP100-ds-DNA interaction, resulting in DNA replication and RNA transcription blockage. NAPHT-BP100 bound to a large extent to ds-DNA ( Figure 3) and naphthalimide derivatives are known to be able to intercalate between ds-DNA base pairs that affect a series of related cellular processes [27].
The increase in hydrophobicity and hemolytic activity of NAPHT-BP100 was balanced by its improved antibacterial action and did not change to a large extent peptide therapeutic index. The indication that NAPHT-BP100 could act upon the bacteria not only by destabilizing its membrane but also by binding the cell ds-DNA, meaning the possibility to act by more than a single mechanism, would also consist in a considerable improvement in terms of avoiding bacterial resistance. Although more hemolytic, NAPHT-BP100 can be applied to the design of antibacterial molecules targeting various types of cutaneous or other mucosal infections. In this context, the results found in the present work can provide subsidies for studies aimed at drug development.