Aryl-alkyl-lysines: Novel agents for treatment of C. difficile infection

Clostridium difficile infections (CDIs) are a growing health concern worldwide. The recalcitrance of C. difficile spores to currently available treatments and concomitant virulence of vegetative cells has made it imperative to develop newer modalities of treatment. Aryl-alkyl-lysines have been earlier reported to possess antimicrobial activity against pathogenic bacteria, fungi, and parasites. Their broad spectrum of activity is attributed to their ability to infiltrate microbial membranes. Herein, we report the activity of aryl-alkyl-lysines against C. difficile and associated pathogens. The most active compound NCK-10 displayed activity comparable to the clinically-used antibiotic vancomycin. Indeed, against certain C. difficile strains, NCK-10 was more active than vancomycin in vitro. Additionally, NCK-10 exhibited limited permeation across the intestinal tract as assessed via a Caco-2 bidirectional permeability assay. Overall, the findings suggest aryl-alkyl-lysines warrant further investigation as novel agents to treat CDI.

Activity against different strains of Clostridium difficile. We had earlier shown that the NCK compounds could infiltrate the membrane of bacteria and also retain activity in acidic pH, other enzymes and in blood [22][23][24] . We rationalized that because of these properties the NCK compounds would also be active against C. difficile. Thus we tested the antibacterial activity of the NCK series of compounds against different strains of C. difficile (Table 1). Vancomycin and metronidazole, which are clinically used to treat CDI, were used as comparator drugs. As depicted in Table 2, NCK-6 exhibited poor anticlostridial activity with a minimum inhibitory concentration (MIC) value of 32 µg mL −1 , while the MIC of NCK-8 ranged from 2 µg mL −1 to 8 µg mL −1 . Both NCK-10 and NCK-12 were quite active against all tested strains of C. difficile. The MIC of NCK-10 ranged from ≤0.5 µg mL −1 to 2 µg mL −1 while that of NCK-12 ranged from 1 µg mL −1 to 4 µg mL −1 , respectively. Against strains P8, P13 and Isolate 10, which were all susceptible to vancomycin (MIC of 0.5 µg mL −1 ) and metronidazole (MIC < 0.25 µg mL −1 ), NCK-10 was active at MICs of 1 µg mL −1 to 2 µg mL −1 . Vancomycin displayed an MIC of 1 and 2 µg mL −1 against Isolate 1 and P4, respectively, while NCK-10 was active at 2 µg mL −1 against both the strains. NCK-12 was also active against these strains with MIC of 4 µg mL −1 . NCK-10 displayed most potent activity against strains Isolate 5 and P15 with MIC values less than or equal to 0.5 µg mL −1 . Against the rest of the strains NCK-10 was active at 1 µg mL −1 while the MIC of vancomycin ranged from 0.5 µg mL −1 to 1 µg mL −1 . The broth microdilution assay revealed NCK-10 was comparable to that of vancomycin in inhibiting C. difficile growth in vitro.
Cytotoxicity of the compounds against human colonic epithelial cells. Earlier we had reported that the NCK compounds were selectively active towards bacterial cells [22][23][24] . Herein, we checked the toxicity of the compounds against human colonic epithelial (HRT-18) cells (Fig. 2). All active NCK compounds were non-toxic at the concentrations that they inhibited the growth of C. difficile. The compounds were tested at four different concentrations (128 µg mL −1 , 64 µg mL −1 , 32 µg mL −1 and 16 µg mL −1 ). It was observed that all compounds were safe to HRT-18 cells at 16 µg mL −1 . Although the cells survived 32 µg mL −1 of NCK-6, NCK-8 and NCK-12, when treated at the same concentration with NCK-10, 60% of the cells survived. Except for NCK-6, some toxicity was www.nature.com/scientificreports www.nature.com/scientificreports/ observed for all the active compounds at concentrations of 64 µg mL −1 and higher. The hemolytic concentration 50 (HC 50 , the concentration at which 50% of human erythrocytes are lysed) of NCK-10 was reported earlier to be 95 µg mL −1 22 .
Caco-2 permeability assay. It is important for compounds used to treat CDI to concentrate at the site of infection (in the colon) and not cross the gastrointestinal tract. We thus evaluated if the most potent compound, NCK-10, would permeate across the gastrointestinal tract via a standard Caco-2 bidirectional permeability assay. The assay revealed that NCK-10 (P app <0.01 × 10 −6 cm s −1 from the apical to basolateral compartments) exhibits limited ability to permeate across the Caco-2 monolayer from the apical to basolateral compartment. This result was similar to ranitidine (P app = 0.186 × 10 −6 cm s −1 from the apical to basolateral compartments (Table 4)), a drug known to exhibit limited permeability across the gastrointestinal tract. Ranitidine is typically used as a low-to-moderate permeability control in Caco-2 permeability assays, while warfarin is used as a high permeability control. Talinolol is susceptible to efflux by P-glycoprotein and is thus used as a control drug to determine if a test agent may be susceptible to efflux 28 . The mean apparent permeability values and efflux ratio values obtained for the control drugs were all similar to values presented in previously published reports [28][29][30] . The results from the Caco-2 bidirectional permeability assay suggest that NCK-10 would concentrate primarily in the gastrointestinal tract which is highly beneficial for treatment of C. difficile infections.

Discussion
C. difficile has gained notoriety as the main pathogen responsible for antibiotic-associated diarrhea and nosocomial diarrhea. Moreover, life threatening pseudomembranous colitis and toxic megacolon can also result from C. difficile 31 . Although antibiotics are being used to treat CDI, resistance has been reported against metronidazole and treating hypervirulent strains of C. difficile is a major difficulty 32 . The scientific community has been trying to use various alternatives to antibiotics as a means to treat CDI including testing antimicrobial peptides and their polymeric mimics 33 . This led us to explore the activity of the aryl-alkyl-lysines against C. difficile. Aryl-alkyl-lysines previously were reported to possess good activities against a variety of pathogens both in vitro and in vivo 23,24 . Ease of synthesis and broad-spectrum activity makes this class of compounds interesting leads as antibacterial agents 22 .
We identified three compounds with moderate to good potency against C. difficile. The compound containing a decyl chain was the most active compound (NCK-10). This has been observed against other bacteria and fungi as well. Intriguingly enough, NCK-10 was found to be almost as effective as vancomycin against most strains of C. difficile tested with a MIC value of 1 µg mL −1 or lower.
A challenge with treating CDI is it leads to overgrowth of Enterococci, including strains resistant to vancomycin 7 . Interestingly, we found that NCK-10 was active at 2 µg mL −1 and lower against all strains of Enterococcus www.nature.com/scientificreports www.nature.com/scientificreports/ tested, including vancomycin-resistant strains. The NCK compounds were safe to human colonic epithelial cells at concentrations above their MIC against C. difficile.
The final point we investigated was the ability of the most potent compound (NCK-10) to cross the gastrointestinal tract. The Caco-2 bidirectional permeability assay is a useful tool in preclinical drug discovery to predict the likelihood of a compound being able to cross the barrier posed by the gastrointestinal tract to gain entry to the bloodstream 34 . For CDI it is important to have compounds that concentrate in the gastrointestinal tract and do not cross into the bloodstream. From the bidirectional permeability assay results obtained, NCK-10 appears   Table 3. Antibacterial activity of the NCK compounds and vancomycin against vancomycin-resistant and vancomycin-sensitive strains of Enterococci. www.nature.com/scientificreports www.nature.com/scientificreports/ unable to effectively cross the gastrointestinal tract. In fact NCK-10 exhibited similar activity to ranitidine, a drug known to exhibit poor permeability across the intestinal tract.
In summary, this study succinctly reports the broad range of activity displayed by the NCK series of compounds against strains of C. difficile. Although further studies regarding the mechanism of action of these compounds against C. difficile is warranted, previous studies with these compounds suggest they exert their antibacterial effect by disrupting the bacterial membrane 22 . However, the potency displayed by the aryl-alky-lysine compounds indicate further research must be invested towards use of membrane-active small molecules to treat CDI.

Materials and method
Antibacterial activity of the compounds against Clostridium difficile. The broth microdilution assay was utilized following the Clinical and Laboratory Standards Institute (CLSI) guidelines with slight modifications 35 . Briefly, C. difficile strains were grown on anaerobic blood agar plates (Becton Dickinson, BD) and suspended in brain heart infusion supplemented broth (Brain heart infusion medium, BD, supplemented with yeast extract, L-cysteine, Vitamin K 1 and Hemin, Sigma) to achieve a bacterial concentration of ~10 5 CFU/ml. The bacterial suspension was seeded in 96-well plates containing drugs/compounds, in duplicate, at the required concentration. Plates were then incubated at 37 °C anaerobically for 48 hours. MICs reported are the lowest concentration of each agent that inhibited the visual growth of bacteria.
Antibacterial activity of the compounds against enterococcal species. The broth microdilution assay was used to determine the MIC of the NCK compounds and control antibiotics against different strains of E. faecalis and E. faecium following the guidelines of the Clinical and Laboratory Standards Institute 36 . Bacteria were incubated aerobically with the NCK compounds or the controls at 37 °C for 16-18 hours before recording the MIC 37-39 .

Cytotoxicity of the compounds against human ileocecal colorectal adenocarcinoma cell line (HRT-18).
Compounds were assayed at concentrations of 16 µg mL −1 , 32 µg mL −1 , 64 µg mL −1 , and 128 µg mL −1 against a human ileocecal colorectal adenocarcinoma cell line (HRT-18) to determine the potential toxic effect to mammalian cells, as described previously 40,41 . Briefly, ~2 × 10 4 cells suspended in 100 µL of RPMI-1640 supplemented with 10% horse serum were seeded in a 96-well plate and incubated at 37 °C with 5% CO 2 . The cells were cultured for 24 hours until reaching ~90% confluency. The cells were further incubated with the aforementioned concentrations of the compounds for 2 hours. Afterwards, culture media were discarded and the cells in each well were washed with PBS and 100 µL of cell culture media were added prior to addition of the assay reagent MTS 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H -tetrazolium) (Promega). The plates were incubated for 4 hours at 37 °C in a humidified 5% CO 2 atmosphere. The absorbance was recorded at 490 nm and the corrected absorbance readings (actual absorbance readings for each treatment subtracted from background absorbance) were taken using a kinetic ELISA microplate reader (SpectraMax i3x, Molecular Devices, Sunnyvale, CA, USA). The quantity of viable cells after treatment with each compound was expressed as a percentage of the control, DMSO. Asterisk (*) denotes a significant difference from the DMSO-treated cells using 2-way analysis of variance (ANOVA) followed by Dunnett's pairwise comparison.
Caco-2 permeability assay. The Caco-2 bidirectional permeability assay was conducted as described in previous reports 29,30,41,42 . The Caco-2 cell line is similar to human enterocytes and expresses P-glycoprotein and other relevant metabolic enzymes that affect absorption of orally-administered xenobiotics across the intestinal mucosa 34,43,44 . Caco-2 cells were grown in tissue culture flasks, trypsinized (by rinsing cells with phosphate-buffered saline and then treating with trypsin-EDTA for 5-10 minutes at 37 °C with 5% CO 2 until cells detached), suspended in Dulbecco's Modified Eagle Medium with 10% fetal bovine serum, 1% nonessential amino acids and penicillin-streptomycin, and then seeded onto wells of a Millipore 96-well Caco-2 plate. Cells were cultured at 37 °C with 5% CO 2 until confluent (approximately three weeks) and media was changed at relevant time points. For Apical to Basolateral (A → B) permeability, the test article was added to the apical (A) compartment and the amount of permeation (P app , as calculated below) to the basolateral (B) compartment was determined; for Basolateral to Apical (B → A) permeability, the test article was added to the B compartment and the amount of permeation to the A compartment was determined. To test tight junctions and monolayer integrity, the A-side buffer contained 100 µM Lucifer yellow dye in transport buffer (1.98 g/L glucose in 10 mM HEPES, 1 × Hank's Balanced Salt Solution) at pH 6.5 while the B-side buffer was transport buffer with pH 7.4. Caco-2 cells were incubated with these buffers for two hours, and the receiver side buffer was removed for analysis by LC/MS/MS. To verify the tight junctions and integrity of Caco-2 cell monolayers, aliquots of the cell buffers were analyzed by fluorescence (Lucifer yellow transport ≤ 2%). Any deviations from control values are reported.

Test Agent
Mean A → B P app (cm s −1 ) Mean B → A P app (cm s −1 ) Efflux Ratio (Re) Notes  Table 4. Caco-2 bidirectional permeability analysis for NCK-10 and control drugs. a N.D. = not determined (post-assay recovery of NCK-10 was too low from A → B).