Amphipathic Guanidine-Embedded Glyoxamide-Based Peptidomimetics As Novel Antibacterial

Amphipathic Guanidine-embedded Glyoxamide-based Peptidomimetics as Novel Antibacterial Agents and Biofilm Disruptors

Shashidhar Nizalapura, Onder Kimyonb, Eugene Yeea, Kitty Hoa, Thomas Berryc, Mike Manefieldb, Charles G. Cranfieldc, Mark Willcoxd, David StC Blacka and Naresh Kumara*

Abstract

Antimicrobial resistance in bacteria is becoming increasingly prevalent, posing a critical challenge to global health. Bacterial biofilm formation is a common resistance mechanism that reduces the effectiveness of antibiotics. Thus, the development of compounds that can disrupt bacterial biofilms is a potential strategy to combat antimicrobial resistance. We report herein the synthesis of amphipathic guanidine-embedded glyoxamide-based peptidomimetics via ring-opening reactions of N-naphthoylisatins with amines and amino acids. These compounds were investigated for their antibacterial activity by the determination of minimum inhibitory concentration (MIC) against S. aureus and E. coli. Compounds 35, 36, and 66 exhibited MIC values of 6, 8 and 10 µg mL–1 against S. aureus, respectively, while compounds 55, 56 showed MIC values of 17 and 19 µg mL–1 against E. coli, respectively. Biofilm disruption and inhibition activities were also evaluated against various Gram-positive and Gram-negative bacteria. The most active compound 65 exhibited greatest disruption of established biofilms by 65% in S. aureus, 61% in P. aeruginosa, and 60% in S. marcescens respectively, at 250 µM concentration, while compound 52 inhibited the formation of biofilms by 72% in S. marcescens at 250 µM. We also report here the in vitro toxicity against MRC-5 human lung fibroblast cells. Finally, the pore forming capability of the three most potent compounds were tested using tethered bilayer lipid membrane (tBLM) technology.

Introduction

Antimicrobial resistance is an increasing medical problem, with various Gram-positive and Gram-negative bacteria showing resistance to many current antibiotics.1 This is due to the fact that existing antibiotics generally inhibit bacterial processes that are essential for survival, thus stimulating bacterial evolution by creating a selective pressure for drug-resistant mutations to persist.2, 3 The Centers for Disease Control and Prevention (CDC) has estimated that drug-resistant bacteria cause 23,000 deaths and 2 million illnesses annually in the United States.4 The rapid increase in the number of multidrug resistant strains, combined with the slow development of new antibiotics, presents a challenge for medicinal chemists to develop effective antibacterial therapies.5 Recent successful approaches have involved the structural modification of existing drugs such as antifungal azoles, antibacterial β-lactams and quinolones.6, 7 However, this strategy merely delays the development of bacterial resistance as the analogues usually operate via similar mechanisms to the parent compounds.

aSchool of Chemistry, UNSW Australia, Sydney, NSW 2052, Australia.

bSchool of Biotechnology and Biomolecular Sciences.

cMolecular Biosciences Team, School of Life Sciences, University of Technology Sydney.

dSchool of Optometry and Vision Science, UNSW Australia, Sydney, NSW 2052, Australia.

University of New South Wales, Sydney, Australia.

UNSW Australia, Sydney, NSW 2052, Australia.

*E‒mail: * Tel: +61 29385 4698; Fax:+61 29385 6141.

A biofilm is a large network of bacteria protected by a self-produced matrix of exopolymeric substances, including polysaccharides, proteins and extracellular DNA.8 Bacteria within these networks are 10-1000 times more resistant to antibiotics than planktonic cells.9 Biofilms are found in 60-80% of chronic infections, including infections associated with cystic fibrosis and endocarditis.10, 11 Additionally, biofilms are often formed on medical implants, such as catheters, artificial hips and contact lenses.9, 12 In the United States alone, 17 million new biofilm infections occur every year, leading to 550,000 fatalities per year.13 In this context, it is increasingly important to develop novel antimicrobial drugs with different mechanism of action in order to supplement or assist existing antimicrobial drugs, thereby reducing the rate of development of antibacterial resistance.

Natural host-defensive antimicrobial peptides (AMPs) and their mimics are attracting increasing focus as potential alternatives to classical antibiotics. Unlike conventional antibiotics, AMPs act by disrupting membranes of bacteria, which makes it difficult for the bacteria to develop resistance to AMPs.14 However, AMPs have several intrinsic limitations, namely susceptibility to degradation by proteases or peptidases, in vivo toxicity, non-selective action and high manufacturing costs.15 In order to overcome these limitations, focus has turned to various peptidomimetics such as α-peptides16,

Figure 1. Examples of reported small molecular antimicrobial peptide mimics (SMAMPs).

______

Figure 2. Overview of the design of novel guanidine-embedded N-naphthoyl-phenyl glyoxamide based peptidomimetics.______

β-peptides17, peptoides18, α-AApeptides19, and γ-AApeptides20, and more recently small molecular antimicrobial peptide mimics (SMAMPs) which show significant antibacterial and anti-biofilm activities against both Gram-negative and Gram-positive bacteria.21-30

Pyne and coworkers have synthesized binaphthyl-1,2,3-triazole peptidomimetics.28 The most active compound (1) (Figure 1) possessed excellent antibacterial activity with a minimal inhibitory concentration (MIC) of 2 μg mL−1 towards Staphylococcus aureus and S. epidermidis.28 Halder designed aryl-alkyl-lysine-based peptide mimics that mimicked the mechanism of action of natural AMPs, and the most active compound (2) (Figure 1) exhibited an MIC of 1.5 μg mL−1 against Escherichia coli as well as excellent antibiofilm activity against established E. coli and Pseudomonas aeruginosa biofilms, with EC50 values of of 20 and 19 µM respectively.30 De Grado and co-workers synthesized guanidine-based peptidomimetics, with compound 3 (Figure 1) exhibiting excellent antimicrobial activity against S. aureus and E. coli values with MIC of 0.045 and 0.4 μg mL−1 respectively.24

In our previous research, we reported a straightforward method for the synthesis of N-naphthoylisatins and their glyoxamide derivatives.29 These compounds exhibited selective antibacterial and antibiofilm activities against S. aureus, particularly compound 4 (Figure 1) which had an MIC of 39 μg mL−1 and showed 50% disruption of established biofilms at 250 µM.29

In continuation of our interest towards the development of novel N-naphthoyl-phenyl glyoxamide-based peptidomimetics, we were inspired to explore the structure-activity relationship (SAR) of this scaffold. The guanidine moiety is represented in various natural products31, antibiotics,32, 33 and synthetic peptidomimetics21, 23, 34-36 that show potent and broad-spectrum antibacterial activity against both Gram-positive and Gram-negative bacteria. Hence, we targeted compounds in which the ammonium moiety in the previously developed scaffold 5 was replaced with a guanidinium moiety (Series-I, Figure 2). Additionally, the amino acids lysine and arginine are a common cationic motifs in natural AMPs and synthetic peptidomimetics37, 38 showing antimicrobial activity. Therefore, replacement of the tail group with a guanidyl-lysine moiety was also proposed (Series-II, Figure 2). Finally, an extended scaffold was targeted, in which an arginine residue was coupled to the terminal lysine residue of the Series-II compounds (Series-III, Figure 2).

In this study, we report the synthesis of novel guanidine-embedded glyoxamide-based peptidomimetics by the ring-opening reaction of N-naphthoylisatins with amines and amino acids. These compounds were evaluated for their antimicrobial activity against S. aureus and E. coli, as well as their anti-biofilm activity against S. aureus, P. aeruginosa, Serratia marcescens and E. coli. In addition, we also investigated their in vitro toxicity against MRC-5 normal human lung fibroblasts cells. Finally, we report the membrane pore forming ability of the active compounds as measured using tethered bilayer lipid membranes in association with AC electrical impedance spectroscopy.

Results and Discussion

Chemistry

The synthetic guanylated N-naphthoyl-phenyl glyoxamide-based peptidomimetics were classified into three series (Series-I, II and III, Figure 2) based on the number of guanidine moieties and peptide bonds present in the molecule. The design of these molecules centered around the replacement of the ammonium group in 5 with the guanidinium ion, producing initially the Series-I compounds (6). The Series-II and III compounds were designed by incorporating either one or two guanylated amino acid units, respectively (7 and 8). The common starting point in the synthesis of all three series involved the N-acylation of isatins (9-13) to generate N-naphthoylisatins (14-18) (Scheme 1). This was achieved in good yields of 65-80% through the use of naphthoyl chloride and sodium hydride in DMF as previously reported.29

The synthesis of Series-I compounds was achieved via ring-opening reactions of N-naphthoylisatins (14-18) with tert-butyl (3-aminopropyl) carbamate in DCM at room temperature for 2 h, giving glyoxamide derivatives (19-23) in good to excellent yields (80%-90%) (Scheme 1). Cleavage of the Boc protecting group was performed with 4 M HCl/dioxane in DCM at room temperature for 2 h to generate the hydrochloride salts (24-28) in good yields (85%-95%). Subsequently, these hydrochloride salts were treated with N,N′-Di-Boc-1H-pyrazole-1-carboxamidine and Et3N at room temperature in DCM for 16 h, affording the di-Boc-guanidine derivatives (29-33) in moderate to good yields (50%-65%). Finally, treatment of these compounds with 4M HCl/dioxane in DCM at room temperature for 16 h yielded the guanidinium glyoxamide derivatives (34-38) in moderate to excellent yields (85%-95%).

The synthesis of the Series-II compounds commenced with the ring-opening reaction of N-naphthoylisatins (14-18) with N-Boc lysine methyl ester hydrochloride in sat. NaHCO3/DCM (1:1) at room temperature for 16 h, which afforded glyoxamides (39-42) in moderate yields (60%-75%) (Scheme 2). Interestingly, no product was formed when this reaction was performed using Et3N in DCM as described for the synthesis of the Series-I glyoxamide intermediates (19-23). Deprotection of 39-42 with 4 M HCl/dioxane in DCM gave amine hydrochlorides (43-46) in quantitative yields (95%-100%), and subsequent treatment with N,N′-Di-Boc-1H-pyrazole-1-carboxamidine in DCM furnished compounds 47-50 in 50%-65% yields. Final Boc-protection was achieved using 4 M HCl/dioxane in DCM in 4 h, however this also caused hydrolysis of the methyl ester group, thus generating carboxylic acids 51-54 in 60%-65% yield. Selective deprotection of the Boc groups could be achieved using TFA/DCM followed by salt exchange with 4 M HCl/dioxane in Et2O to obtain compounds 55-56 in 65%-85% yields.

In order to synthesis the dipeptide-based compounds of Series-III, esters 39 and 40 were saponified with LiOH in a THF/MeOH/water mixture at room temperature for 16 h, giving acids 57 and 58 in 90% and 95% yield, respectively (Scheme 3). These acids were then coupled with methyl 2-amino-5-((tert-butoxycarbonyl) amino) pentanoate using EDC•HCl/HOBt and (i-Pr)2NEt in acetonitrile at 0 °C to room temperature for 2 h, affording amides 59 and 60 in yields of 35% and 55%, respectively. The initial cooling of this reaction was essential in order to avoid decomposition of the starting materials; no product was observed if the reaction was performed entirely at room temperature. Deprotection of the Boc groups under the previously detailed conditions then gave compounds 61 and 62 in 60% and 70% yield. Subsequent reaction with N,N′-Di-Boc-1H-pyrazole-1-carboxamidine in DCM gave compounds 63 and 64 in 55%-65% yield. Finally, deprotection using TFA in DCM at 0 °C for 3 h and subsequent salt exchange with 4 M HCl/dioxane in DCM afforded compounds 65 and 66 in 70% and 85% yield, respectively.

Antibacterial activity

The antibacterial activities of the guanylated peptidomimetics (34-38, 51-56, 61, 62, 65 and 66) were investigated by the determination of their minimum inhibitory concentration (MIC) against S. aureus, according to a previously published protocol.29 The results of the MIC assay are given in Table 1. The Series-I compounds (34-38) showed moderate to excellent antibacterial activity (MIC) against S. aureus. Compound 34, representing the parent scaffold with nosubstitution on the phenyl moiety, possessed an MIC value of 23 µg mL-1. The introduction of an electron-withdrawing fluorine atom to the phenyl ring (37) of the parent scaffold decreased activity (MIC = 30 µg mL-1). However, introduction of abulkier bromine (35, MIC = 6 µg mL-1) or chlorine (36, MIC = 8 µg mL-1) substituent led to dramatically improved antibacterial activity. Conversely, addition of a comparatively more electron-donating methyl group at the phenyl ring (38) drastically diminished antimicrobial activity against S. aureus, returning an MIC value of 93 µg mL-1. However, against E. coli the Series-I compounds exhibited generally low antibacterial activities (MIC >110 µg mL-1 or >250 µM), with the exception of compound 37 (MIC = 59 µg mL-1).

Scheme 1. General scheme for the synthesis of Series-I guanylatedglyoxamide based peptidomimetics.

______

Scheme 2. General scheme for the synthesis of Series-II guanylated glyoxamide based peptidomimetics.

______

Scheme 3. General scheme for the synthesis of Series-III guanylated glyoxamide based peptidomimetics.

______

Minimum inhibitory concentration (MIC (µg mL-1))
Compound / S. aureus / E. coli
Series-I / 34 / 23 / >110
35 / 6 / >110
36 / 8 / >110
37 / 30 / 59
38 / 93 / >110
Series-II / 51 / 53 / 66
52 / 77 / 77
53 / 108 / 68
54 / 54 / 69
55 / >140 / 17
56 / >140 / 19
Series-III / 61 / 130 / 39
62 / 23 / 21
65 / 18 / 90
66 / 10 / 50
Gentamicin / <0.6 / 2.5

Table 1: Antibacterial activity (MIC) of compounds against S. aureus and E. coli.

The Series-II monopeptide compounds exhibited lower antibacterial activities against S. aureus, but higher activities against E. coli than their Series-I counterparts, with MIC values of 53 to >140 µg mL-1 and 17 to 139 µg mL-1, respectively. The methyl ester compounds (55-56) (MIC >135 µg mL-1) were less active than their corresponding acid derivatives (51-53), which showed MIC values of 53 to 108 µg mL-1. This might be due to the potential for the acid derivatives to form a zwitterionic species, therefore increasing the hydrophilicity of the tail portion of the molecule. Furthermore, the Series-II compounds did not show similar trends with respect to aromatic substitution as compared to the Series-I compounds. Here, the parent scaffold (51) and the methyl analogue (54) were the most active, with MIC values of 53 and 54 µg mL-1 against S. aureus, respectively. The presence of electron withdrawing groups in 52 (R = Br) and 53 (R = F), significantly reduced antibacterial activity (MIC values of 77 and 108 µg mL-1, respectively) against S. aureus. On the other hand, the Series-II glyoxamide acid derivatives 51-54 displayed good MIC values in the range of with 66 to 77 µg mL-1, respectively, against E. coli. Interestingly, the methyl ester compounds 55 and 56 were even more active against E. coli, with MIC values of 17 and 19 µg mL-1, respectively.

Finally, the Series-III dipeptides were generally the most active series of compounds synthesized. Similar to the Series-I compounds, the presence of bromine in the phenyl ring (62, 66) improved activity compared to the parent scaffold (61, 65). Additionally, the guanidinyl derivatives (65, 66) were also more active than the corresponding quaternary ammonium salts (61, 62). Compound 66, bearing both a bromo substituent on the phenyl ring and guanidinium groups, displayed the highest activity among the Series-III compounds against S. aureus (MIC = 10 µg mL-1), and it also showed moderate activity against E. coli (MIC = 50 µg mL-1). The related guanidinium salt 65 lacking the bromine substituent also displayed good antimicrobial activity against S. aureus (MIC = 18 µg mL-1), but had diminished activity against E. coli (MIC value = 90 µg mL-1). Interestingly, the bromo-substituted ammonium salt 62 also displayed good antibacterial activity, with MIC values of 23 and 21 µg mL-1 against S. aureus and E. coli, respectively.

Figure 3. Disruption of established biofilms after 24 h treatment with 250 µM of glyoxamide based peptidomimetics. The control represents the pre-established biofilms without any compounds. Error bars indicate the standard error of the mean (SEM) of three independent experiments.______

Overall, structure-activity relationships (SAR) could be deduced regarding the antimicrobial activity of the guanidine-embedded glyoxamide derivatives against S. aureus and E. coli. Generally, with

the exception of the Series-II compounds, the presence of a bromine substituent on the phenyl ring improved activity. Against S. aureus, a short 3-carbon linker or dipeptide linker was better than a single peptide linker, and additionally, the free carboxylic acids were more active than their corresponding methyl esters. Against E. coli, the monopeptide acids and dipeptide esters were more active than those with 3-carbon linkers and monopeptide esters. Furthermore, the incorporation of a guanidinium group improved antimicrobial activity compared to the corresponding quaternary ammonium compound.

Antibiofilm activity

The biofilm disruption activities of the compounds was investigated using crystal violet staining according to a previously published protocol.39 The ability of compounds (34-38, 51-56, 61, 62, 65 and 66) to disrupt the established biofilms of the Gram-positive bacteria S. aureus and the Gram-negative bacteria P. aeruginosa, S. marcescens and E. coli, was determined at a concentration of 250 µM. In this assay, bacterial cultures were grown on a polystyrene substratum overnight prior to treatment with the compounds. To assay inhibition of biofilm formation, bacterial cultures of P. aeruginosa and S. marcescens were grown in the presence of 250 µM of test compounds. The results for compounds that exhibited more than 40% biofilm inhibition are depicted in Figure 3 (disruption of established biofilms) and Figure 4 (inhibition of biofilm formation), with the data for all remaining compounds given in the Supporting Information (S.I. Figure 7, 8).

Overall, the three series of glyoxamide derivatives exhibited low to good levels of disruption against the established biofilms of Gram-positive and Gram-negative bacteria (Figure 3). The Series-I compounds (34-38) showed minimal effects on the biofilm integrity, despite some of them being the most potent in terms of antibacterial activity. Compound 35 (bromo derivative) disrupted the established biofilms of S. marcescens and E. coli by 49% and 46%, respectively, while compounds 36 (chloro derivative) and 37 (fluoro derivative) were able to reduce the established biofilms of E. coli by 43% and 53%, respectively.

On the other hand, the Series-II compounds showed better biofilm disruption activities despite only having moderate MIC values. The acids (51-54) possessed superior activity against biofilms to the esters (55-56). Compound 52 (bromo derivative) exhibited good biofilm disruption of 42% against S. aureus, while compound 53 (fluoro derivative) showed broad-spectrum biofilm disruption activities, with 46%, 42% and 40% disruption against S. aureus, P. aeruginosa and S. marcescens, respectively. Furthermore, compound 51 showed 48% disruption against S. marcescens, while compound 54 displayed 46% disruption against P. aeruginosa biofilms.

The series-III compounds displayed the greatest levels of biofilm disruption against both the Gram-positive and Gram-negative strains. Compound 61, with an unsubstituted phenyl ring and two