Peptoid synthesis was performed using IRORI technology at room temperature reported by Kölmel et al. [17] using permutation of four different side chains (Fig. 2). Library 1 was synthesized using N-(2-prop-2-yn 1-yl) glycine (Nprg), N-(p-chlorobenzyl)glycine (Npcb), N-(4-aminobutyl)glycine, (Nlys), and N-(benzyl)glycine (Nphe) for side chains. Library 2 was performed with N-(benzyl)glycine (Nphe), N-(1-aminotetradecanyl)glycine (Ntetradec), N-(4-Hydroxybenyl)glycine (Nphb) and N-(4-fluorobenzyl) glycine (Npfb) side chains. Library 3 was performed with N-(2-prop-2-yn-1-yl)glycine (Nprg), N-(p-chlorobenzyl)glycine (Npcb), N-(4-Hydroxybenyl)glycine (Nphb) and N-(1-aminohexanyl)glycine (Nhe) side chains. All CPPos were labeled with Rhodamine B at the C-terminus of the peptoid as described in Kölmel et al. [17]. All products were verified to have $>$95% purity by RP-HPLC and lyophilized to dry powders prior to biological testing. Mass spectrometry was used to confirm the molecular weights of the purified products.

Fig. 2.

Peptoid monomer side chain structures.

Minimum inhibitory concentration or MIC is the concentration at which a compound is able to inhibit the growth of bacteria. All experiments were conducted in microtiter plates. 75 $\mu$L of 1 $\times$ 10${}^6$ CFU/mL bacterial inoculum in Mueller-Hinton broth (MHB) was added to 75 $\mu$L of peptoid solution in MHB (prepared in serial dilution). Negative control without peptoids was used while the positive control was performed with vancomycin treatment. The MIC${}_{90}$ was defined as the lowest concentration of peptoid that inhibited 90% of bacterial growth after incubation at 37 ${}^{\circ }$C for 18 hours at 150 rpm. Absorbance at OD${}_{600}$ was measured using a Tecan Infinite M2 to determine bacterial growth. MIC values reported were reproducible between three independent experimental replicates.

MIC for mycobacteria was determined similar to S. aureus in a microtiter plate with minor modifications. 100 $\mu$L of bacterial culture at a density of 5 $\times$ 10${}^6$ CFU/mL were incubated with an equal volume of drug containing Middlebrook 7H9 media and incubated at 37 ${}^{\circ }$C for 5 days at 150 rpm. Bacterial growth was measured by absorbance (OD${}_{600}$) similar to S. aureus. The negative control was a drug free culture which was considered as 100% growth rate. 90% growth was evaluated relative to the drug free wells. 0% growth was defined in wells containing only media without cultures.

Molar attenuation coefficient of Rhodamine B in water and octanol at I = 550 nm was determined by diluting Rhodamine B in octanol and water to a final concentration of 20 $\mu$M. The solution was measured in a 96-well-plate (Costar 3596, 96 Well Cell Culture Cluster, sterile), using a 96-well plate reader (Ultra Microplate Reader ELx808, BioTEK Instruments, INC). Peptoids were diluted to a final concentration of 160 $\mu$M in 250 $\mu$L water. Afterwards, 250 $\mu$L octanol was added and each mix was vortexed for 2 min and centrifuged for 3 min with a centrifugal force of 3000 g to separate the octanol from the aqueous phase. Phases were separated and the absorbance at I = 550 nm of 200 $\mu$L octanol phase and water phase was measured in a 96-well-plate (Cstar 3596, 96 Well Cell Culture Cluster, sterile) by using a 96-well plate reader (Ultra Microplate Reader ELx808, BioTEK Instruments, INC). Each absorbance of peptoid was measured at least three times. Data were averaged and standard deviation was calculated.

Cell culture was carried out under sterile conditions. 1 $\times$ 10${}^4$ HeLa (human cervix carcinoma) cells or 2 $\times$ 10${}^4$ RAW 264.7 cells were plated into each well of a 8-well $\mu$slide from IBIDI (Ibitreat, Martinsried, Germany) and cultured in 200 $\mu$L Dulbecco’s modified Eagle’s medium, high glucose, supplemented with 10% Fetal calf serum and 1 U/mL penicillin/streptomycin at 37 ${}^{\circ }$C, 5% CO${}_2$. The purified peptoids were dissolved in distilled water to yield a 2 mM stock solution and were further diluted with 10% DMEM. The cells were incubated with the peptoids at a final concentration of 10 $\mu$M at 37 ${}^{\circ }$C, 5% CO${}_2$. Cellular uptake of the peptoids was measured by live-cell imaging after 24 hours. Visualization of the peptoids was achieved by confocal microscopy using a Leica TCS-SP5 II, equipped with a DMI6000 microscope (Germany). The peptoids were excited at 561 nm using a DPSS laser. The Mitotracker Green (Invitrogen, Karlsruhe, Germany) for the detection of the mitochondria was excited at 488 nm using an argon laser and the Hoechst 33342 for the detection of the nuclei was excited at 364 nm using a UV laser. The objective was an HCXPL APO CS 63.0 $\times$ 1.2 Water UV. The exposure was set to minimize oversaturated pixels in the final images. Fluorescence emission was measured at 417–468 nm for the detection of the nuclei, at 499–552 nm for the detection of the mitochondria, and 593–696 nm for the detection of the rhodamine B labelled peptoids. Image acquisition was conducted at a lateral resolution of 1024 $\times$ 1024 pixels and 8-bit depth using LAS-AF 2.0.2.4647 acquisition software (Leica LAS AF Lite 3.1, Leica Microsystems CMS GmbH, https://leica-las-af-lite.software.informer.com/4.0/).

HEK 293T cells were cultured in DMEM media. A peptoid solution plate (100 $\mu$L per well) was prepared by serial dilution of peptoid stocks in media. Peptoid solution were transferred to a 96-well plate of day-old cell monolayers containing 100 $\mu$L per well media ($\approx$1 $\times$ 10${}^4$) cells per well. The plate was then incubated for 24 hours at 37 ${}^{\circ }$C and 5% CO${}_2$. MTS reagent (Promega) (40 $\mu$L per well) was added to each well, and the plate was incubated at 37 ${}^{\circ }$C for 3 hours in darkness, after which absorbance at 490 nm was read. The percentage of inhibition was calculated as below.

(1)$\text{Percentage of inhibition}=\left[1-\left(\mathrm{A}-{\mathrm{A}}_{\text{testblank}}\right)/\left({\mathrm{A}}_{\text{control}}-{\mathrm{A}}_{\text{blank}}\right)\right]$ $\times 100$

where A is the absorbance of the test well and A${}_{control}$ the average absorbance of wells with cells exposed to media and MTS (no peptoid). A${}_{testblank}$ (media, MTS, and peptoid) and A${}_{blank}$ (media and MTS) were background absorbance measured in the absence of cells.

Peptoid solution (100 $\mu$L) with concentrations corresponding to 1$\times$, 2$\times$ and 4$\times$ MIC was first prepared and added to an equal volume of bacterial solution containing bacterial counts of approximately 10${}^4$ CFU/mL in each well of a 96-well plate. The plates were incubated at 37 ${}^{\circ }$C and samples were collected at different timepoints and plated on agar plates for determination of viable counts. The results were depicted as mean log (CFU/mL) $\pm$ standard deviation.

Mid-log phase culture of S. aureus ATCC 29737 was diluted in fresh MHB to achieve an OD600 of 0.15. It was then washed twice with PBS and re-suspended in PBS. The suspension was incubated with 1$\times$ MIC of selected peptoids and 50 nM SYTOX green for 2 hours at 37 ${}^{\circ }$C. The incubation mixture was washed twice in PBS and re-suspended in same volume of PBS. Flow cytometry was performed, and data was analysed using CytoFLEX LX Flow Cytometer and CytExpert (Beckman Coulter, Brea, CA, USA). The bacteria suspension was also incubated with 35% ethanol and 50 nM SYTOX green for 1 hour at 37 ${}^{\circ }$C as the positive control. Controls of unstained cells, and peptoid treatment in the absence of SYTOX green were also analysed to account for background noises. Values reported were reproducible between three independent experimental replicates using ANOVA tests.

We synthesized libraries of AMPos to determine whether the antimicrobial activities of peptoids are affected by structural and sequence modifications in a similar fashion as AMPs [21, 22]. The design of the peptoids in these libraries was derived from changes in lipophilicity and amphiphilicity. Sequence-specific N-substituted glycine oligomers can be efficiently synthesized via primary amines as submonomers to incorporate a large number of diverse side chain functionalities [14]. This method iterates sequential steps of bromoacylation and nucleophilic displacement of bromide using any kind of amine primary amine submonomer synthons. N-acetylated linear oligomers were synthesized on the Rink amide resin to obtain C-terminal amides. Thereafter, the N- termini were acetylated with acetic anhydride prior to TFA cleavage (Fig. 1A). In total, 401 peptoids were synthesized using split mix combinatorial solid-phase synthesis. Amines used for synthesis in Library 1 were 4-aminobutylglycine (Nlys), prop-2-yn-1-amine (Nprg), benzylamine (Nphe) and 4 chlorobenzylamine (Npcb). Amines used for synthesis in Library 2 are tetradecylglycine (Ntetradec), 4 fluorobenzylglycine (Npfb), benzylamine (Nphe) and 4-hydroxybenzylglycine (Nphb). Amines used for synthesis in Library 3 are hexylamine (Nhe), prop-2-yn-1-amine (Nprg), 4-hydroxybenzylglycine (Nphb) and 4-chlorobenzylamine (Npcb). For microscopic analysis, all peptoids have been labelled with Rhodamine B. All other experiments have also been performed with labelled peptoids. The monomers used for the synthesis are shown in Fig. 2. The positive charge was introduced by incorporating lysine-like (Nlys) monomers, while the overall lipophilicity was altered using a variety of short and aromatic monomers. The chemical structures for the side chains used in creating the libraries were mimics of natural amino acid side chains, in addition to a broader selection of bulky hydrophobic side chains such as Nphb, Npfb, and Npcb. To vary the amphiphilicity of the peptoids, we used different permutations of the monomers in the peptoid sequence.

401 compounds were synthesized, of which 362 were screened for MIC${}_{90}$ against S. aureus ATCC 29737 (Fig. 3). 11 compounds were identified as primary hits with MIC${}_{90}$ of 10 $\mu$M which were then observed for cell penetration properties in HeLa cells. They were then tested for antimicrobial activity against mycobacteria and drug-resistant S. aureus clinical isolates. Their selectivity between mammalian and bacterial cells was then determined by evaluating cytotoxicity against mammalian cell lines. Six peptoids were chosen for determining time concentration kill kinetics, as well as uptake of SYTOX Green to establish whether membrane permeabilization was indeed the mechanism of action. Finally, three molecules, one as a representative from each library, were tested for uptake by macrophages as both mycobacteria and S. aureus can be localized inside mammalian macrophages. This would also enable us to determine whether cellular uptake was dependent on the cell type.

Fig. 3.

Schematic of screening of hits and workup of lead compounds.

362 peptoids were screened for their antimicrobial activity against S. aureus ATCC 29737 (Fig. 4). The highest tested concentration was 200 $\mu$M. The potent molecules were identified as those with an MIC of 10 $\mu$M, as denoted by the red cut-off line on the graph. MIC values for compounds with no activity even at 200 $\mu$M were denoted as $>$200 $\mu$M. The potent peptoids were then tested against drug-resistant clinical isolates of S. aureus, Mycobacterium bovis (M. bovis) BCG (as a BSL-2 substitute of M. tuberculosis) and Mycobacterium chelonae (M. chelonae) (as a representative of non-tuberculous mycobacteria). Of the peptoids screened, those with activity against all tested strains are summarized in Table 1.

Fig. 4.

Scatter plot of primary screen of peptoids against S. aureus ATCC 29737. Three libraries, comprising a total of 362 compounds were screened for potency. A cut-off of 10 $\mu$M was used as shown by the red line.

For the initial MIC screening, S. aureus was incubated with different concentrations of peptoids for 18 hours. MICs were determined by measuring OD${}_{600}$. The MIC was defined as the concentration that can inhibit 90% of bacterial growth (MIC${}_{90}$). The active peptoids with MIC 10 $\mu$M were then screened against the mycobacterial species and Escherichia coli (E. coli) (data not shown) to determine whether they were potential antimicrobials. Interestingly, we found no activity against E. coli which led us to conclude that these peptoids are selectively active only against Gram-positive bacteria and mycobacteria which are known to be closely related to Gram positive bacteria [23]. Recent studies have discovered that hydrophobicity is strongly suggested to enhance selective activity of antimicrobial peptides (AMPs) against Gram-positive bacteria due to the insertion of peptides into Gram-positive membranes while weakening their ability to penetrate Gram-negative outer membrane [24, 25, 26]. Membrane components such as charged lipopolysaccharides as present on Gram-negative bacteria outer membranes present additional electrostatic barriers to peptide and peptoid entry. It is likely that our active peptoids with similar membrane penetrative abilities as AMPs are unable to penetrate the Gram-negative OM as easily as they did for Gram-positive bacteria. Table 1 summarizes the sequences of the active peptoids, MICs against the different strains of bacteria and the cLogP value which denotes the lipophilicity of the molecules. The cLogP value was in the range between 0.36 and 1.52.

The lowest MICs${}_{90}$ were within the range 1.5 $\mu$M and 12.5 $\mu$M. It has been shown in literature that compounds containing fluorine and chlorine incorporated in them generally show high antimicrobial activity [27, 28]. Here, we examined whether the inclusion of fluorine and chlorine can enhance the antimicrobial activity of our peptoids. All peptoids demonstrated a decent antimicrobial effect (MIC${}_{90}$ 1.56 $\mu$M to 6.25 $\mu$M) against S. aureus ATCC 29737 with peptoid 1, 10, and 11 being effective against both strains of mycobacteria. Moreover, all shortlisted peptoids exhibited a potent activity (MIC${}_{90}$ 1.56 $\mu$M to 12.5 $\mu$M) against MDR S. aureus. The higher MICs in mycobacteria compared to S. aureus can be attributed to the differences in the cell wall structure of the two bacterial species. Mycobacteria are known to have an outer layer of mycolic acid making them relatively more impermeable to antibiotics, thus harder to kill [23].

Cell membranes are known to be selectively permeable and hence limit the penetration of large molecules or non-lipophilic molecules such as hydrophilic and large molecular weight drugs [17, 29]. As a result, antimicrobials are unable to reach intracellular targets or intracellular microbes. This in turn restricts the arsenal of drugs that can be employed to efficiently target intracellular microbes [30]. Existing studies have shown the use of viral vectors as shuttle for AMPs or membrane disruption for delivery, but this can potentially result in unintended cytotoxicity. Since peptoids have previously been shown to overcome barriers for cell penetration without membrane destruction, our libraries of peptoids were similarly tested for these abilities. Hela cells have been widely used in several studies to determine the cell penetrative ability of peptides [31, 32]. Due to their robust proliferative capability, we have chosen them as hosts for large scale screening of our peptoids in this study. It has also been established that most of the peptide entry mechanisms are endocytic in nature and do not involve specific cell receptors, hence HeLa cells though not perfect will act as good surrogates for determining the cell penetrating ability of our peptoids. Here, we showed that all our peptoids except for peptoid 6, localized in the mitochondria. We used the Pearson correlation coefficient (Rr) to estimate the co-localization in the mitochondria, using the ImageJ Plugin JACoP (ImageJ bundled with 64-bit Java, ImageJ, https://imagej.nih.gov/ij/download.html, https://imagej.net/plugins/jacop). All peptoids showed a coefficient of 0.75–0.95, except for peptoid 6 (0.43) which in this case could indicate endosomal localization rather than mitochondrial entry. The Rr for the peptoids were as follows: Rr (1) = 0.75, Rr (2) = 0.824, Rr (3) = 0.917, Rr (4) = 0.957, Rr (5) = 0.869, Rr (6) = 0.438, Rr (7) = 0.911, Rr (8) = 0.922, Rr (9) = 0.916, Rr (10) = 0.944, Rr (11) = 0.909. This is confirmed by the confocal images in Fig. 5 which showed yellow colour in the merged column except for peptoid 6. This indicates that all the peptoids except 6 localized in the mitochondria making them ideal for targeting intracellular pathogens or even as delivery agents to transport antimicrobial agents into cells.

Fig. 5.

Cellular uptake of library 1 (peptoid 1, 2), library 2 (peptoid 3, 4, 5, 6, 7, 8) and library 3 (8, 9, 10, 11) compounds in HeLa cells. 1 $\times$ 104 HeLa cells were treated with 10 $\mu$M of peptoid tagged with Rhodamine B for 24 hours at 37 ${}^{\circ }$C. For co -staining of the nuclei and the mitochondria the cells were treated with Hoechst 33342 (blue) and 100 nM MitoTracker Green (green). Eventually the cells were analyzed by fluorescence confocal imaging. The last column shows the merges of the respective emission channels of each line by using the following PMTs for the emission: 417–468 nm for the detection of the nuclei (blue), 499–552 nm for the detection of the mitochondria (green), and 593–696 nm for the detection of the rhodamine B labelled peptoids (red). Scale bar = 20 $\mu$m.

We evaluated the biocompatibility of selected oligomers in HEK 293T cells by using the MTS assay. CC${}_{50}$ was defined as the concentration that is cytotoxic to 50% of the cells. The cytotoxicity was estimated using a tetrazolium salt (MTS)-based colorimetric assay and is reported as the concentration that inhibits metabolic activity by 50% (CC${}_{50}$). We defined the selectivity ratio (SR) as the quotient of CC${}_{50}$ and the MIC${}_{90}$ S. aureus ATCC 29737, hence the SR is an estimate of a compounds’ tendency to kill bacterial rather than mammalian cells. Table 2 summarizes the MIC${}_{90}$ for S. aureus ATCC 29737, CC${}_{50}$ and SR. Peptoid 5, 6 and 7 showed lower toxicities to HEK 293T cells and the highest selectivity with an SR of 32. High selectivity ratios represent peptoids which preferentially kill bacterial rather than mammalian cells.

We then investigated the ability of the peptoids 1, 2, 5, 6, 10, and 11 to effectively kill S. aureus ATCC 29737 as shown in Fig. 6. Peptoid 1 had a killing efficiency of 99.99% ($>$3-log) at 1$\times$ MIC at 4 h. Peptoid 2 reduced the bacterial load by 2-log, however, the activity was only time-dependent and not dose-dependent as increasing the concentration to 2$\times$ and 4$\times$ did not affect the efficacy. The killing efficiency of peptoids 5, 6, 10 and 11 was found to be 99.99% ($>$3-log) at 2$\times$ MIC at 8 h. Furthermore, at 4$\times$ MIC, there was complete eradication of all bacteria after 4 h incubation respectively. On the other hand, vancomycin, which is an existing drug used to treat S. aureus infections, was ineffective in killing 99.99% of the bacteria within 24 h and did not show dose dependent activity. Thus, we were able to demonstrate efficient bactericidal activity against S. aureus to be both time- and dose-dependent for all the peptoids except for peptoid 2.

Fig. 6.

Time-concentration kill kinetics of peptoids and Vancomycin against S. aureus ATCC 29737. Vancomycin was used as a positive control. The CFU/mL values represent the means $\pm$ standard deviations from two independent experiments.

To examine whether the selected peptoids caused membrane permeabilization of S. aureus ATCC 29737, cultures were treated with 1$\times$ MIC of the peptoids for 2 hours. Flow cytometry was employed to estimate the portion of bacterial population with permeabilized membranes. Membrane activity of test compounds was determined by measuring the uptake of SYTOX Green, a fluorescent dye which can only enter cells with compromised membrane integrity and stain intracellular nucleic acids. During flow cytometry analysis, this dye allows for differentiation between the bacterial cells which have been permeabilized and those which are intact. As shown in Fig. 7, compared to the untreated control, treatment with all selected peptoids resulted in SYTOX Green uptake of a portion of the cultures. However, the percentage of damaged cells varied with the different peptoids. Whereas peptoids 1 and 11 permeabilized the membrane of more than half of the bacterial population (54.6% and 54.4% respectively), membrane activity of peptoids 2 and 10 were less prominent (21.2% and 34.7% respectively). Peptoids 5 and 6 barely affected the membrane, causing less than 10% of the culture to become permeable to SYTOX Green. Combining these results together with the time-concentration kill kinetics and the net charge of the peptoids, we can conclude that for peptoids 10 and 11 membrane disruption occurred before the bactericidal activity. This results in lysis leading to cell death. Peptoid 1 is cytotoxic as seen by the low selectivity index. It is likely that Peptoid 1 lysed cell membranes, bacterial or mammalian, extensively thus allowing uptake of SYTOX Green and rapid bactericidal activity. Peptoid 2, 5 and 6 exhibited a low uptake of SYTOX Green indicating that the membrane is intact. They remain, however, bactericidal as seen from the time kill kinetics, implying that there may be more than one mechanism of action involved in the activity of these peptoids.

Fig. 7.

Membrane permeabilization of S. aureus ATCC 29737 induced by the peptoids, as reflected by the percentage of cells in the culture positive for SYTOX green uptake.

Granuloma, which are compact aggregates of immune cells such as macrophages, are the hallmark structures of tuberculosis. It is historically regarded as a host-protective structure that shields off the infected area from healthy tissue [33]. This is problematic as most drugs cannot penetrate the granuloma, and thus cannot reach the inside-residing bacteria. As a result, the bacteria can become dormant with the potential of future reactivation, especially when patients become immune-compromised [34]. Thus, it is important that our peptoids are able to penetrate into macrophages to exert their antimicrobial function. We incubated RAW 264.7 cells with peptoid solution (1 representative peptoid from each library) (labelled with Rhodamine B) and imaged the localization of the peptoids in living cells using confocal microscopy (Fig. 8). Cellular uptake in macrophages was detected for the selected peptoids. The confocal microscopy images showed, that the peptoids accumulated in the mitochondria of the cells. This finding provides a strong justification to the design of cell penetrating peptoids which can enter and localize within macrophages. This suggests that our peptoids could be a good starting point for development of new antimycobacterial drugs or as adjunctive therapy in combination with existing regimens.

Fig. 8.

Cellular uptake of peptoid 1, 3, 11 RAW 264.7 cells. 2 $\times$ 10${}^4$ RAW 264.7 cells were treated with 10 $\mu$M of peptoid tagged with Rhodamine B for 24 hours at 37 ${}^{\circ }$C. For co-staining of the nuclei and the mitochondria the cells were treated with Hoechst 33342 (blue) and 100 nM Mitochondria GreenTM (green). Eventually, the cells were analyzed by fluorescence confocal imaging. The last column shows the merges of the respective emission channels of each line by using the following PMTs for the emission: 417–468 nm for the detection of the nuclei (blue), 499–552 nm for the detection of the mitochondria (green), and 593–696 nm for the detection of the rhodamine B labelled peptoids (red). Scale bar = 20 $\mu$m.