Amphiphilic Tobramycin Analogues as Antibacterial and Antifungal Agents

Abstract

In this study, we investigated the in vitro antifungal activities, cytotoxicities, and membrane-disruptive actions of amphiphilic tobramycin (TOB) analogues. The antifungal activities were established by determination of MIC values and in time-kill studies. Cytotoxicity was evaluated in mammalian cell lines. The fungal membrane-disruptive action of these analogues was studied by using the membrane-impermeable dye propidium iodide. TOB analogues bearing a linear alkyl chain at their 6″-position in a thioether linkage exhibited chain length-dependent antifungal activities. Analogues with C₁₂ and C₁₄ chains showed promising antifungal activities against tested fungal strains, with MIC values ranging from 1.95 to 62.5 mg/liter and 1.95 to 7.8 mg/liter, respectively. However, C₄, C₆, and C₈ TOB analogues and TOB itself exhibited little to no antifungal activity. Fifty percent inhibitory concentrations (IC₅₀s) for the most potent TOB analogues (C₁₂ and C₁₄) against A549 and Beas 2B cells were 4- to 64-fold and 32- to 64-fold higher, respectively, than their antifungal MIC values against various fungi. Unlike conventional aminoglycoside antibiotics, TOB analogues with alkyl chain lengths of C₁₂ and C₁₄ appear to inhibit fungi by inducing apoptosis and disrupting the fungal membrane as a novel mechanism of action. Amphiphilic TOB analogues showed broad-spectrum antifungal activities with minimal mammalian cell cytotoxicity. This study provides novel lead compounds for the development of antifungal drugs.

INTRODUCTION

The frequency of invasive fungal infections, such as candidiasis, has dramatically increased due to the rising populations of immunocompromised patients as a result of AIDS and cancer therapy, as well as bone marrow and organ transplantations (1). Candida spp. are known to cause the majority of fungal infections and are the fourth most common cause of nosocomial blood infection in the United States (2). However, other fungal pathogens, such as Candida glabrata, Candida parapsilosis, Candida tropicalis, Candida lusitaniae, Crytococcus neoformans, and Aspergillus spp., are also on the rise and are causing threats to human and animal health (3). Currently, a number of antifungal drugs, such as azoles, echinocandins, and amphotericin B (AmB), are used to treat invasive fungal infections, including candidiasis, in humans. Growing fungal resistance and host side effects, however, limit their therapeutic efficacies. There is therefore a growing need to develop novel antifungal drugs.

Aminoglycosides (AGs) are compounds that consist of two or more amino sugars that are connected to a 2-deoxystreptamine scaffold via glycosidic bonds. AGs, such as tobramycin (TOB), are well-known antibiotics that are used to treat bacterial infections in humans, but they do not inhibit the growth of fungi. They bind to the prokaryotic 16S rRNA in the decoding A-site region, which induces codon misreading and inhibits translocation (4, 5). Although AGs are predominantly known for their antibacterial activities, some conventional AGs have been reported to have antifungal activities against fungal oomycetes and other true fungi (6). Despite being potent antibiotics, emerging bacterial resistance against AGs has compromised their therapeutic use. However, there has been a growing interest in structural modifications of AGs that could revive the efficacy of these drugs against resistant bacterial strains. We recently demonstrated that modifying TOB at the 6″-position by incorporating linear alkyl chains (C₆ to C₂₂) in a thioether linkage resulted in chain length-dependent antibacterial activities against vancomycin-resistant enterococci (VRE) and TOB-resistant Escherichia coli, with resistance to the parent drug TOB itself (7). However, the antifungal activities of these analogues are yet to be determined. It was recently reported that the incorporation of a C₈ linear alkyl chain at the O-4″-position and of an octanesulfonyl chain at the O-6′-position of kanamycins A and B, respectively, led to the discovery of new applications of these AGs as antifungal agents that lack bacterial activity (8–11). This prompted us to investigate the antifungal properties of our 6″-thioether TOB analogues with linear alkyl chains with chain lengths between C₄ and C₁₄ (here, we are referring to these as C₄, C₆, C₈, C₁₀, C₁₂, and C₁₄) (Fig. 1) and to further investigate the extent of the antibacterial spectra of these analogues against resistant clinical bacterial isolates that we had not previously tested. We report the antibacterial and antifungal activities as well as the cytotoxicities of our C₄ to C₁₄ TOB analogues.

FIG 1.

Scheme for the synthesis of the 6″-thioether TOB analogues C₄ to C₁₄ used in this study.

MATERIALS AND METHODS

Materials.

TOB was purchased from AK Scientific (Union City, CA). All other chemicals were purchased from Sigma-Aldrich (St. Louis, MO) and used without further purification. Compound 1 (Fig. 1) was prepared as previously described (7). Chemical reactions were monitored via thin-layer chromatography (TLC; with silica gel 60, F₂₅₀; Merck). Visualization was achieved by using one of the following methods: H₂SO₄ stain (5% in methanol) or KMnO₄ stain (1.5 g KMnO₄, 10 g K₂CO₃, 1.25 ml 10% NaOH, 200 ml H₂O). Compounds were purified by SiO₂ flash chromatography (Flash silica gel, 32 to 63 μm; Dynamic Adsorbents Inc.). ¹H and ¹³C nuclear magnetic resonance (NMR) spectra were recorded on a Varian 400-MHz spectrometer. TOB analogues (C₄ to C₁₄) were dissolved in double-distilled water (ddH₂O) at a final concentration of 10 g/liter.

The Eis (12), AAC(6′)-Ib′ (13), AAC(6′)-Ie/APH(2″)-Ia [used only for its AAC(6′) activity] (14), AAC(3)-IV (14), AAC(2′)-Ic (12), and ANT(4′) (15) resistance enzymes were purified as previously described. 5,5′-Dithiobis(2-nitrobenzoic acid) (DTNB), ATP, acetyl coenzyme A, and inorganic pyrophosphatase were bought from Sigma-Aldrich and used without further purification. The determinations of the activities of these resistance enzymes against C₄ were performed as previously reported for a group of C₆ to C₁₄ antimicrobial agents (Fig. 2) (7). 3-(4,5-Dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide was purchased from TCI America (Portland, OR, USA). Spectrophotometric and colorimetric assays were performed on a multimode SpectraMax M5 plate reader (Molecular Devices, Sunnyvale, CA) using 96-well plates (Fisher Scientific).

FIG 2.

Bar graph showing the relative activities of the listed AMEs against C₄, normalized to that of TOB (100%).

The antifungal agents posaconazole (POS), itraconazole (ITC), and fluconazole (FLC) were obtained from AK Scientific, Inc. (Union City, CA). POS, ITC, and FLC were dissolved in dimethyl sulfoxide (DMSO) at a final concentration of 5 g/liter.

The yeast strains Candida albicans ATCC 10231 (designated strain A here), C. albicans ATCC 64124 (B), and C. albicans ATCC MYA-2876 (C) were kindly provided by Jon Y. Takemoto (Utah State University, Logan, UT, USA). The yeast strains C. albicans ATCC MYA-90819 (D), C. albicans ATCC MYA-2310 (E), C. albicans ATCC MYA-1237 (F), C. albicans ATCC MYA-1003 (G), and Cryptococcus neoformans MYA-895 (H) were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). The filamentous fungal strains Aspergillus nidulans ATCC 38163 (I) and Fusarium graminearum 053 (J) were kindly provided by Jon S. Thorson and Lisa J. Vaillancourt (University of Kentucky, Lexington, KY, USA), respectively. Filamentous fungi and yeasts were cultivated at 35°C (except for F. graminearum 053 [J], which was grown at room temperature [RT]) in RPMI 1640 (catalog number R6504; Sigma-Aldrich Chemical Co., St. Louis, MO) buffered to pH 7.0 with 0.165 M morpholinepropanesulfonic acid (MOPS) buffer (Sigma-Aldrich Chemical Co.).

The bacterial strains used in this study were obtained from various sources. Mycobacterium parascrofulaceum ATCC BAA-614 (D+) and Haemophilus influenzae ATCC 51907 (F−) were purchased from ATCC. Methicillin-resistant Staphylococcus aureus (MRSA; C+) and S. aureus NorA (F+) were a gift from David H. Sherman (University of Michigan, Ann Arbor, MI, USA). Mycobacterium smegmatis MC2 155 (E+) was a gift from Sabine Ehrt (Weill Cornell Medical College, New York, NY, USA). Enterococcus faecium BM4105-RF (A+), Listeria monocytogenes ATCC 19115 (B+), S. aureus NRS22 (G+), E. coli MC1061 (E−), Klebsiella pneumoniae ATCC 27736 (G−), and Pseudomonas aeruginosa PAO1 (I−) were gifts from Paul J. Hergenrother (University of Illinois at Urbana-Champaign, Champaign, IL, USA). Staphylococcus epidermidis ATCC 12228 (H+), Streptococcus pyogenes ATCC 12384 (I+), Escherichia coli ATCC 25922 (A−), and P. aeruginosa ATCC 27853 (H−) were generously provided by Dev P. Arya (Clemson University, Clemson, SC, USA). E. coli Δ7 wild type (wt; B−), E. coli Δ7 A1408G (C−), and E. coli Δ7 G1491U (D−) were gifts from Kurt Frederick (Ohio State University, Columbus, OH, USA). Shigella flexneri 2475 pgm-24 (J−) was obtained from Anthony T. Maurelli (Uniformed Services University of the Health Sciences, Bethesda, MD, USA).

The human lung carcinoma epithelial cell line A549 (ATCC CCL-185) and the normal human bronchial epithelial cell line Beas 2B (ATCC CRL-9609) were kindly provided by David K. Orren (University of Kentucky, Lexington, KY, USA).

Protocol for formation of the Boc-protected thioether TOB analogues.

The amphiphilic tert-butyloxycarbonyl (Boc)-protected TOB analogues C₆ to C₁₄ (compounds 2b to 2f) were synthesized from TOB as previously described (Fig. 1) (7). For the preparation of the novel analogue 2a (Boc-protected C₄), 1-butanethiol (0.13 ml, 1.22 mmol) was added to a solution of analogue 1 (0.30 g, 0.24 mmol) and Cs₂CO₃ (0.12 g, 0.36 mmol) in freshly distilled dimethylformamide (3 ml). The mixture was stirred at RT overnight. Progress of the reaction was monitored by TLC (hexanes:ethyl acetate [EtOAc] ratio, 2:3; R_f, 0.60). Upon completion, the reaction mixture was diluted with EtOAc (10 ml) and washed with H₂O (5 ml). The aqueous layer was extracted with EtOAc (twice, 10 ml). The combined organic layers were washed again with H₂O (twice, 10 ml) and brine (10 ml), dried over anhydrous MgSO₄, and filtered. The solvents were removed under reduced pressure, and the crude product obtained was purified by flash column chromatography (SiO₂; pure hexanes to hexanes:EtOAc in a 2:3 ratio) to afford compound 2a (0.19 g; 76%) as a white solid: ¹H NMR (400 MHz, CD₃OD) δ 5.07 (br s, 1H, H-1′), 5.02 (br s, 1H, H-1″), 4.05 (m, 1H, H-5″), 3.70 to 3.28 (m, 13H, H-1, H-3, H-4, H-5, H-6, H-2′, H-4′, H-5′, H-6′ (2H), H-2″, H-3″, H-4″), 2.97 (br dd, J₁ = 14.4 Hz, J₂ = 2.2 Hz, 1H, H-6″), 2.62 (m, 1H, H-6″), 2.58 [t, J = 7.6 Hz, 2H, SCH₂(CH₂)₂CH₃], 2.12 (m, 1H, H-2eq), 1.99 (m, 1H, H-3′eq), 1.66 to 1.24 [m, 51H, H-2ax, H-3′ax, 5 × CO₂(CH₃)₃, SCH₂(CH₂)₂CH₃], 0.90 [t, J = 7.6 Hz, 3H, SCH₂(CH₂)₂CH₃]; ¹³C NMR (100 MHz, CD₃OD) δ 158.0, 157.9, 156.5, 156.4 (2 carbons), 98.6 (anomeric carbon), 98.1 (anomeric carbon), 82.9, 81.1, 79.3, 79.0 (2 carbons), 78.9, 78.7, 75.8, 72.5 (2 carbons), 72.1, 70.7, 65.0, 55.7, 50.0, 49.8, 49.6, 40.5, 34.4, 33.3, 32.9, 32.4, 31.6, 27.4 to 27.3 (15 carbons), 21.6, 12.7.

Protocol for Boc deprotection [e.g., synthesis of compound 3a (C₄)].

The amphiphilic TOB analogues C₆ to C₁₄ (compounds 3b to f) were synthesized from 2b to f as previously described (Fig. 1) (7). For the preparation of the novel analogue 3a (C₄), compound 2a (46 mg, 0.044 mmol) was treated at RT with neat trifluoroacetic acid (TFA; 1 ml) for 3 min. The TFA was removed under reduced pressure, and the residue was dissolved in a minimal volume of H₂O and freeze-dried to afford the novel 6″-thioether TOB derivative 3a (C₄; 49 mg, 98%) as a white foam: ¹H NMR (400 MHz, D₂O) δ 5.74 (d, J = 3.6 Hz, 1H, H-1′), 5.10 (d, J = 4.0 Hz, 1H, H-1″), 4.05 to 3.93 (m, 4H, H-4, H-5′, H-2″, H-5″), 3.89 (app. t, J₁ = J₂ = 9.2 Hz, 1H, H-5), 3.82 to 3.54 (m, 6H, H-1, H-3, H-6, H-2′, H-4′, H-4″), 3.46 (app. t, J₁ = J₂ = 10.4 Hz, 1H, H-3″), 3.43 (dd, J₁ = 13.6 Hz, J₂ = 3.6 Hz, 1H, H-6′), 3.28 (dd, J₁ = 13.6 Hz, J₂ = 6.4 Hz, 1H, H-6′), 3.05 (dd, J₁ = 14.4 Hz, J₂ = 2.4 Hz, 1H, H-6″), 2.78 (dd, J₁ = 14.4 Hz, J₂ = 8.0 Hz, 1H, H-6″), 2.63 [t, J = 7.6 Hz, 2H, SCH₂(CH₂)₂CH₃], 2.57 (app. dt, J₁ = 12.4 Hz, J₂ = J₃ = 4.4 Hz, 1H, H-2eq), 2.57 (app. dt, J₁ = 12.0 Hz, J₂ = J₃ = 4.4 Hz, 1H, H-3′eq), 2.03 (app. q, J₁ = J₂ = J₃ = 12.4 Hz, 1H, H-2ax), 1.95 (app. q, J₁ = J₂ = J₃ = 12.8 Hz, 1H, H-3′ax), 1.58 (p, J = 7.6 Hz, 2H, SCH₂CH₂CH₂CH₃), 1.39 (sextet, J = 7.6 Hz, 2H, SCH₂CH₂CH₂CH₃), 0.89 [t, J = 7.6 Hz, 3H, SCH₂(CH₂)₂CH₃]; ¹³C NMR (100 MHz, D₂O) δ 100.5 (anomeric carbon), 94.5 (anomeric carbon), 83.8, 77.6, 74.1, 72.5, 70.2, 68.0, 67.8, 64.2, 54.5, 49.3, 48.1, 47.8, 39.7, 32.2, 32.0, 30.9, 29.2, 27.7, 21.1, 12.7. Note that compounds 3a to f were stored at −20°C as a 10-g/liter stock solution in ddH₂O.

Antibacterial susceptibility testing.

MIC values of TOB analogues (C₄ to C₁₄) were determined by using the microdilution broth method for aerobic and anaerobic organisms according to CLSI standardized methodology (16, 17) with minor modifications. A variety of Gram-positive and Gram-negative bacteria were diluted 1:1,000 from overnight cultures into fresh medium and grown to an optical density of ∼0.4 at 600 nm, or on average, for 4 h. To prepare bacteria for MIC determinations, an additional 1:1,000 dilution of turbid culture was performed. Briefly, solutions of the studied compounds were added to the optimal medium for each bacterial strain (i.e., LB for MRSA [B+], S. aureus NorA [F+], S. epidermidis ATCC 12228 [H+], S. pyogenes ATCC 12384 [I+], E. coli Δ7 wt [B−], E. coli Δ7 A1408G [C−], E. coli Δ7 G1491U [D−], E. coli MC1061 [E−], K. pneumoniae ATCC 27736 [G−], and S. flexniri 2475T pgm-24 [J−]; brain heart infusion (BHI) for E. faecium BM4105-RF [A+], L. monocytogenes ATCC 19115 [B+], and S. aureus NRE22 USA600 [G+]; supplemented BHI for H. influenzae ATCC 51907 [F−]; tryptic soy medium for E. coli ATCC 25922 [A−], P. aeruginosa ATCC 27583 [H−], and P. aeruginosa PAO1 [I−]; 7H9 for M. smegmatis MC2 155 [E+]; 7H9 with ADC for M. parascrofulaceum ATCC BAA-614 [D+]), and a double dilution was performed (100 μl) in a microtiter plate. The diluted bacterial cultures were then added to these medium-drug mixtures (100 μl). The microtiter plate cultures were grown for 16 to 20 h and visually inspected for growth. MIC values were defined as the concentration of the drug in the last well showing no bacterial growth.

Antifungal susceptibility testing.

In vitro MIC values for the C₄ to C₁₄ TOB analogues against yeast cells were evaluated in 96-well plates as described in CLSI document M27-A3 (18) with minor modifications. Yeast cells were grown in RPMI 1640 for 48 h, counted using a hemocytometer, and diluted to a concentration of 5 × 10⁴ cells/ml in fresh RPMI 1640. Cell suspensions (200 μl) containing 0.97 to 125 mg/liter of TOB analogue, 0.97 to 62.5 mg/liter of POS, 0.97 to 62.5 mg/liter of ITC, or 0.97 to 125 mg/liter of FLC were added to the wells of a 96-well microtiter plate and incubated for 48 h at 35°C. The final concentration of DMSO was ensured to be <2% in all experiments. Growth of C. albicans 10231 (strain A) was not affected by this concentration of DMSO. The MIC values of TOB analogues were defined as the minimum drug concentration that yielded complete inhibition, or the MIC-0. MIC values of azoles for yeasts were determined as the lowest drug concentration that produced at least 50% growth inhibition (i.e., MIC-2) compared with the growth control well. It is important to note that at the lowest concentration tested, the growth of C. neoformans MYA-895 (H) was completely inhibited. For testing azoles against molds, on the other hand, the MIC-0 endpoint was used.

The minimal fungicidal concentration (MFC) values for the C₁₂ and C₁₄ TOB analogues were determined as previously described (19, 20) with minor modifications. To determine MFCs, yeast cells of 1 × 10⁴ CFU/ml were used to perform broth microdilution assays as described above. After 48 h of incubation, all of the MIC wells with no visible growth were homogenized with a micropipette, and aliquots of 20 μl cell content were spread on Sabouraud's dextrose agar (SDA; Difco, BD, Franklin Lakes, NJ, USA). The plates were incubated for 24 to 48 h at 35°C for colony counts. The MFC was defined as the lowest drug concentration that killed 99% of the final inoculum (with ≤3 colonies on SDA plates, in agreement with a previous report [21]). Each test was performed in triplicate.

In vitro MIC values for these analogues against A. nidulans ATCC 38163 (I) and F. graminearum 053 (J) were conducted as previously described in CLSI document M38-A2 (22). Spores were collected from sporulating cultures growing in potato dextrose agar (PDA) by filtration through sterile glass wool and enumerated using a hemocytometer to obtain the desired inoculum size. Serial dilutions of TOB analogues as well as POS, ITC, and FLC were made in sterile 96-well plates in the ranges of 0.97 to 125 mg/liter (for TOB analogues), 0.097 to 12.5 mg/liter (for POS and ITC), and 0.97 to 62.5 mg/liter (for FLC) using RPMI, and spore suspensions were added to make a final concentration of 5 × 10⁵ CFU/ml. The plates were incubated at 35°C for 48 h (except for F. graminearum 053 [J], which was incubated at room temperature). The MIC values of azoles and TOB analogues against molds were based on the complete inhibition of growth compared to the growth control, or the MIC-0 (18, 22). Each test was performed in triplicate.

Antifungal carryover and time-kill studies.

Prior to performing time-kill studies, antifungal carryover effects were evaluated as previously described (19). C. albicans ATCC 64124 (B) cell suspensions were prepared to achieve an inoculum of approximately 1 × 10⁵ to 4 × 10⁵ CFU/ml. An aliquot of 100 μl of each suspension was added to 900 μl of sterile ddH₂O (control) or to sterile ddH₂O plus TOB analogue (C₁₂ or C₁₄) at concentrations of 8, 16, or 32 mg/liter and 2, 4, or 8 mg/liter, respectively. Immediately after addition of the fungal suspension, tubes were vortexed and 100 μl of suspension was removed and spread on PDA for colony count determinations. Antifungal carryover was defined as a reduction in colony counts of a sample by >25% compared to controls. Time-kill curve studies were performed against the azole-resistant C. albicans strain ATCC 64124 (B) as previously described (23) with modifications. The initial inoculum used for this assay was 1 × 10⁵ CFU/ml in liquid RPMI 1640 medium at 35°C in 50 ml Falcon tubes (10-ml volume) with continuous agitation (200 rpm). TOB analogues (C₁₂ and C₁₄) at concentrations of 8, 16, or 32 mg/liter and 2, 4, or 8 mg/liter, respectively, were tested against C. albicans ATCC 64124 (B). Test solutions were incubated at 200 rpm at 35°C. At 0, 3, 9, 12, and 24 h, 100-μl aliquots were removed from each solution and serially diluted in sterile ddH₂O. Fifty microliters of each dilution was spread onto a potato dextrose agar plate and incubated at 35°C. FLC (16 mg/liter) and AmB (1 mg/liter) were used as controls in this assay. Colony counts were determined after 48 h of incubation. The experiment was performed in triplicate. The lower limit for accurate and reproducible quantification was 50 CFU/ml (23).

In vitro cytotoxicity assay.

The in vitro cytotoxicity assay was performed as previously described (24) with minor modifications. Human lung carcinoma epithelial A549 cells and normal human bronchial epithelial cells Beas 2B were grown in F-12K and Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS) and 1% antibiotics, respectively. The confluent cells were then trypsinized with 0.05% trypsin–0.53 mM EDTA and resuspended in fresh medium (F-12K or DMEM). The cells were transferred into 96-well microtiter plates at a density of 3,000 cells/well and were grown overnight. The following day, the media were replaced with 100 μl of fresh culture medium containing serially diluted TOB analogues at final concentrations of 0.97 to 125 mg/liter or sterile ddH₂O (negative control). The cells were incubated for an additional 24 h at 37°C with 5% CO₂ in a humidified incubator. To evaluate cell survival, each well was treated with 10 μl (25 mg/liter) of resazurin sodium salt (Sigma-Aldrich) for 3 h. Metabolically active cells can convert the blue nonfluorescent dye resazurin to the pink and highly fluorescent dye resorufin, which can be detected at wavelengths of 560 nm (excitation) and 590 nm (emission) by using a SpectraMax M5 plate reader. Triton X-100 (1%, vol/vol) resulted in complete loss of cell viability and was used as the positive control. The percent cell survival was calculated as follows: [(control value – test value) × 100]/(control value), where the control value represents results for cells plus resazurin but without drug and the test value represents cells plus resazurin plus the drug.

Propidium iodide staining.

A single colony of C. albicans ATCC 64124 (B) was used to inoculate 5 ml of potato dextrose broth (PDB) in a Falcon tube and was grown overnight at 35°C at 200 rpm. The overnight culture was further diluted by transferring 200 μl of the cell suspension to 800 μl of RPMI 1640 medium. Fifty microliters of cell suspension was then added to the RPMI medium containing no drug (negative control) or C₁₂ or C₁₄ TOB analogues at concentrations of 15.6 (0.5× MIC), 31.2 (1× MIC), and 62.5 mg/liter (2× MIC), or 3.9 (0.5× MIC), 7.8 (1× MIC), and 15.6 (2× MIC), respectively. TOB (125 mg/liter) and POS (62.5 mg/liter) were used as controls. The cell suspensions were then treated for 1 h at 35°C with continuous agitation (200 rpm). After incubation, treated cells were centrifuged and resuspended in 500 μl of phosphate-buffered saline (PBS; pH 7.2). Subsequently, cells were treated with propidium iodide (PI; 9 μM, final concentration) and incubated for 5 min at room temperature in the dark (25). Glass slides prepared with 15 μl of each mixture were observed in bright-field and fluorescence modes (using a Texas Red filter set, with excitation and emission wavelengths of 535 and 617 nm, respectively) using a Zeiss Axiovert 200M fluorescence microscope. Data were obtained from at least two independent experiments.

Annexin V and PI staining.

We chose one of the potent TOB analogues, C₁₄, to perform the assay with Annexin V and PI staining. C. albicans ATCC 64124 (B) was treated with C₁₄ at a subinhibitory concentration (4 mg/liter) for 1 h. Cells were washed in PBS and digested at 30°C for 30 min in 0.1 M potassium phosphate buffer that contained 150 mg/liter Zymolyase 20T, 40 mM 2-mercaptoethanol, and 2.4 M sorbitol at pH 7.2 (26). Protoplasts were washed in modified Annexin binding buffer (10 mM HEPES-NaOH [pH 7.4], 40 mM NaCl, 50 mM CaCl₂, and 1.2 M sorbitol), resuspended in the same buffer along with 5 μl of fluorescein isothiocyante (FITC)-Annexin V, 2 μl of PI (20 mg/liter), and RNase (250 mg/liter), and incubated at 37°C for 30 min. Cells were analyzed in bright-field and fluorescence modes (using Texas Red and FITC filter sets) using a Zeiss Axiovert 200M fluorescence microscope.

RESULTS

In vitro antibacterial assay.

The new C₄ compound along with the C₆ to C₁₄ compounds previously evaluated (7) were tested for antibacterial activity (MIC value determinations) against several strains of Gram-positive and Gram-negative bacteria (Table 1). The strains tested ranged from completely resistant to TOB (>150 mg/liter, strains E. faecium BM4105-RF [A+], MRSA [C+], S. aureus NRS22 USA600 [G+], E. coli Δ7 A1408G [C−], E. coli Δ7 G1491U [D−], and P. aeruginosa ATCC 27853 [H−]) to very susceptible to TOB (<0.3 mg/liter; strain M. smegmatis MC2 155 [E+]). The C₄ analogue showed excellent antibacterial activity (2.3 to 4.7 mg/liter) against four Gram-positive strains (L. monocytogenes ATCC 19115 [B+], M. smegmatis MC2 155 [E+], S. aureus NorA [F+], and S. epidermidis ATCC 12228 [H+]) and one Gram-negative strain (H. influenzae ATCC 51907 [F−]), and there were only 10 strains (E. faecium BM4105-RF [A+], MRSA [C+], M. parascrofulaceum ATCC BAA-614 [D+], S. aureus NRS22 USA600 [G+], E. coli ATCC 25922 [A−], E. coli Δ7 A1408G [C−], E. coli Δ7 G1491U [D−], K. pneumoniae ATCC 27736 [G−], P. aeruginosa ATCC 27853 [H−], and P. aeruginosa PAO1 [I−]) against which it was ineffective (MIC, 75 to >150 mg/liter). For the C₆ analogue, MIC values ranged from >150 mg/liter to 18.8 mg/liter, and for the C₈ compound the values ranged from >150 mg/liter to 37.5 mg/liter against all Gram-positive and Gram-negative strains tested. Compounds with the C₁₀, C₁₂, or C₁₄ modification had none to very good activity, with MIC values ranging from >150 mg/liter to 2.3 mg/liter (C₁₀), >150 mg/liter to 0.6 mg/liter (C₁₂), and >150 mg/liter to <0.3 mg/liter (C₁₄). On a broader scale, all compounds displayed better MIC values against Gram-positive strains than against Gram-negative strains.

TABLE 1.

MICs for TOB and its analogues

Bacterial straina	MIC (mg/liter) for compoundb
	C4	C6	C8	C10	C12	C14	TOB
Gram-positive strains
A+	>150	>150	>150	75	9.4	4.7	>150
B+	2.3	37.5 (75)	75 (>150)	18.8 (37.5)	4.7 (18.8)	2.3 (9.4)	1.2 (4.7)a
C+	>150	>150	>150	>150	18.8	2.3	>150
D+	75	150	150	37.5	75	9.4	1.2
E+	2.3	18.8	>150	>150	>150	>150	<0.3
F+	4.7	150 (>150)	>150 (150)	18.8 (37.5)	4.7 (9.4)	0.6 (9.4)	1.2 (9.4)
G+	>150	>150	150	9.4	2.3	1.2	>150
H+	2.3	75 (75)	150 (75)	9.4 (9.4)	1.2 (2.3)	0.6 (1.2)	<0.3 (0.3)
I+	37.5	150	37.5	2.3	0.6	<0.3	9.4
Gram-negative strains
A−	75	>150	>150	150	18.8	37.5	9.4
B−	18.8	>150	150	75	>150	4.7	2.3
C−	>150	>150	>150	75	>150	4.7	>150
D−	>150	>150	>150	75	9.4	4.7	>150
E−	9.4	150 (150)	75 (75)	18.8 (9.4)	9.4 (18.8)	9.4 (4.7)	2.3 (9.4)
F−	2.3	18.8	>150	75	>150	>150	1.2
G−	150	>150	>150	>150	75	18.8	2.3
H−	>150	>150	>150	150	18.8	18.8	>150
I−	>150	>150	150	37.5	18.8	37.5	1.2
J−	18.8	150	150	37.5	2.3	2.3	4.7

^aGram-positive strains: A+, E. faecium BM4105-RF; B+, L. monocytogenes ATCC 19115; C+, MRSA; D+, M. parascrofulaceum ATCC BAA-614; E+, M. smegmatis MC2 155; F+, S. aureus NorA; G+, S. aureus NRS22 USA600; H+, S. epidermidis ATCC 12228; I+, S. pyogenes ATCC 12384. Gram-negative strains: A−, E. coli ATCC 25922; B−, E. coli Δ7 wt; C−, E. coli Δ7 A1408G; D−, E. coli Δ7 G1491U, E−, E. coli MC1061; F−, H. influenzae ATCC 51907; G−, K. pneumoniae ATCC 27736; H−, P. aeruginosa ATCC 27853; I−, P. aeruginosa PA01; J−, S. flexneri 2475T pgm-24.

^bMICs were determined for TOB and its analogues with various alkyl chain lengths (C₄ to C₁₀) against a variety of Gram-positive and Gram-negative bacterial strains. The values in parentheses were previously reported in reference 7 for a different batch of these molecules.

The main mechanism of resistance to aminoglycoside antibiotics is their modification by resistance enzymes termed aminoglycoside-modifying enzymes (AMEs) that can be acquired by bacteria (27). There are three types of AMEs: the aminoglycoside N-acetyltransferases (AACs), the aminoglycoside O-phosphotransferases (APHs), and the aminoglycoside O-nucleotidyltransferases (ANTs). Recently, a unique AAC, the enhanced intracellular survival (Eis) protein from a variety of bacterial species, was found capable of multiacetylating aminoglycosides (12, 28–36). Knowing that our compounds had good antibacterial activity, we tested the novel C₄ compound to see if it became modified by AMEs (Fig. 2). All other compounds in this study had previously been tested against these AMEs (7). The rates at which AAC(6′)-Ib′, AAC(6′)-Ie, and Eis modified C₄ were only 50% or less of the rate at which TOB was modified by these enzymes. The converse was true for the AAC(3)-IV, AAC(2′)-Ic, and ANT(4′) enzymes, as C₄ was modified at a rate 1.5 to 2 times that of TOB with the same enzymes.

In vitro antifungal assay.

The MIC values for TOB and its C₄ to C₁₄ analogues as well as those of three known antifungal azoles (POS, ITC, and FLC) were determined against various fungal strains (Table 2). TOB analogues C₁₂ and C₁₄ showed broad-spectrum antifungal activities against yeast and filamentous fungi. The MIC values for C₁₂ and C₁₄ against tested fungal strains ranged from 1.95 to 31.2 mg/liter and 1.95 to 7.8 mg/liter, respectively. The C₁₀ TOB analogue exhibited good to no activity against various fungal strains, with MICs ranging from 3.9 to 125 mg/liter. The parent TOB and its analogues C₄, C₆, and C₈ exhibited no antifungal activities except for C₁₀ against F. graminearum 053 (strain J; 3.9 mg/liter) and only low activity against C. neoformans MYA-895 (H; 15.6 mg/liter). It should be noted that the fungal strains used in this study showed better sensitivity to C₁₂ and C₁₄ than all three azoles, with the exception of two yeast strains, C. neoformans MYA-895 (H) and C. albicans 10231 (A), and two filamentous strains, A. nidulans 38163 (I) and F. graminearum 053 (J) (Table 2). However, only C. neoformans MYA-895 (H) showed higher sensitivity to C₈, C₁₂, and C₁₄, with low MIC values of 1.95 mg/liter.

TABLE 2.

MICs determined for TOB and its analogues with various alkyl chain lengths (C₄ to C₁₄) and for three control antifungal agents (POS, ITC, and FLC) against various yeast strains and filamentous fungi

Yeast or fungal straina	MIC (mg/liter) for compoundb
	C4	C6	C8	C10	C12	C14	TOB	POS	ITC	FLC
Yeast strains
A	>125	>125	>125	31.2	31.2	7.8	>125	0.5	0.5	62.5
B	>125	>125	>125	62.5	31.2	7.8	>125	>62.5	>62.5	>125
C	>125	>125	>125	62.5	15.6	7.8	>125	7.8	7.8	15.6
D	>125	>125	>125	125	31.2	7.8	>125	31.2	31.2	>125
E	>125	>125	>125	125	31.2	7.8	>125	31.2	31.2	>125
F	>125	>125	>125	125	15.6	7.8	>125	15.6	31.2	62.5
G	>125	>125	>125	125	31.2	7.8	>125	15.6	31.2	62.5
H	31.25	7.8	1.95	15.6	1.95	1.95	62.5	<0.975	<0.975	<0.975
Filamentous fungi
I	>125	125	125	62.5	7.8	7.8	>125	0.195	0.195	62.5
J	62.5	31.2	31.2	3.9	3.9	3.9	62.5	≤1.56	<1.95	>125

^aYeast strains: A, C. albicans ATCC 10231; B, C. albicans ATCC 64124; C, C. albicans ATCC MYA-2876(S); D, C. albicans ATCC 90819(R); E, C. albicans ATCC MYA-2310(S); F, C. albicans ATCC MYA-1237(R); G, C. albicans ATCC MYA-1003(R); H, C. neoformans MYA-895. Note that here the S or R following a strain name indicates that ATCC reports the strain susceptible (S) or resistant (R) to ITC and FLC. Filamentous fungi: I, Aspergillus nidulans ATCC 38163; J, Fusarium graminearum 053.

^bFor yeast strains, MIC-0 values are reported for TOB and its analogues, whereas MIC-2 values are reported for azoles. For filamentous fungi, MIC-0 values are reported for all compounds.

The MFC values were also determined for the most active C₁₂ and C₁₄ compounds and were found to be 62.5 mg/liter and 15.6 mg/liter, respectively, in almost all strains tested, with the exception of C₁₄ against C. albicans ATCC 64124, for which the MFC was 7.8 mg/liter (Table 3). A trailing growth effect was observed for all three azoles against all yeast strains except against C. albicans 10231 (A) and C. neoformans (H). Since the majority of azoles did not completely inhibit the growth of yeast at the tested concentrations, MFC values were not determined for them.

TABLE 3.

MFCs determined for TOB analogues with an alkyl chain length of C₁₂ or C₁₄ against various yeast strains

Yeast straina	MFC (mg/liter) for analogue
	C12	C14
A	62.5	15.6
B	62.5	15.6
C	62.5	7.8
D	62.5	15.6
E	62.5	15.6
F	62.5	15.6
G	62.5	15.6
H	3.9	3.9

^aYeast strains: A, C. albicans ATCC 10231; B, C. albicans ATCC 64124; C, C. albicans ATCC MYA-2876 (S); D, C. albicans ATCC 90819 (R); E, C. albicans ATCC MYA-2310 (S); F, C. albicans ATCC MYA-1237 (R); G, C. albicans ATCC MYA-1003 (R); H, C. neoformans MYA-895. The S or R following a strain name indicates that ATCC reports the strain susceptible (S) or resistant (R) to ITC and FLC.

Antifungal carryover and time-kill studies.

An antifungal carryover effect was not observed for TOB analogues with linear alkyl chain lengths of C₁₂ and C₁₄ at 1, 2, or 4× MICs against C. albicans ATCC 64124 (B). The time-kill course of the most potent TOB analogues, C₁₂ and C₁₄, and of FLC against the representative strain C. albicans ATCC 64124 (B) over a 24-h period is shown in Fig. 3. C₁₂ at 32 mg/liter (MIC) and C₁₄ at 8 mg/liter (MIC) rapidly reduced the CFU of C. albicans ATCC 64124 (B) by ≥2 log₁₀ after 6 and 9 h of treatment, but complete killing was observed only after 24 h for C₁₂ and after 12 h for C₁₄ (Fig. 3). However, FLC at 16 mg/ml showed a fungistatic effect against C. albicans ATCC 64124 (B) for the first 6 h of growth, and after that the trend of cell growth was similar to that of control cells. Likewise, AmB at a subinhibitory concentration (1 mg/liter) maintained its fungistatic activity against C. albicans ATCC 64124 (B) for 24 h.

FIG 3.

Representative time-kill studies of 6″-thioether TOB analogues with C₁₂ and C₁₄ linear alkyl chains against azole-resistant C. albicans ATCC 64124 (strain B). (a) Cultures were exposed to C₁₂ at 8, 16, or 32 mg/liter. (b) Cultures were exposed to C₁₄ at 2, 4, or 8 mg/liter. In both panels, cultures were exposed to FLC at 16 mg/liter and AmB at 1 mg/liter or to a no-drug control.

In vitro cytotoxicity assay.

The cytotoxicity assay results with TOB analogues with linear alkyl chain lengths of C₄ to C₁₄ against A549 and Beas 2B cells are shown in Fig. 4. The IC₅₀s of C₁₄ against the A549 and Beas 2B cell lines were >62.5 mg/liter and 125 mg/liter, respectively (Fig. 4). This was a 32- to 64-fold-higher concentration than the antifungal MICs against fungi (Table 2). Similarly, the IC₅₀s of C₁₂ against both cell lines were ≥125 mg/liter, 4- to 64-fold higher than its antifungal MICs. However, C₄, C₆, C₈, C₁₀, and TOB itself show little to no toxicity against these cell lines.

FIG 4.

Cytotoxicity of 6″-thioether C₄ to C₁₄ TOB analogues against mammalian cells. A549 cells (a) and Beas 2B cells (b) were incubated at 37°C for 24 h in a CO₂ incubator with various concentrations of TOB or its 6″-thioether analogues with C₄, C₆, C₈, C₁₀, C₁₂, and C₁₄ linear alkyl chains.

Effects of C₁₂ and C₁₄ on fungal membrane integrity.

PI was used to investigate the effects of TOB analogues (C₁₂ and C₁₄) on the fungal membrane integrity of C. albicans ATCC 64124 (strain B). PI, a membrane-impermeable dye, cannot enter an intact cell unless the cell membrane is compromised, and it fluoresces upon binding to nucleic acids (37). When yeast cells were exposed to TOB analogues C₁₂ and C₁₄ at 15.6, 31.2, or 62.5 mg/liter and 3.9, 7.8, or 15.6 mg/liter, respectively, 6, 46, or 93% and 10, 36, or 92% staining of the yeast cells was observed, respectively. However, cells treated with TOB (2% staining) or POS (5%) or untreated (1%) showed negligible staining (Fig. 5).

FIG 5.

(a) Representative dose-dependent membrane permeabilization effects of 6″-thioether TOB analogues with C₁₂ or C₁₄ linear alkyl chains on azole-resistant C. albicans ATCC 64124 (strain B). From top to bottom: PI dye uptake by yeast cells without drug, with POS (62.5 mg/liter), with TOB (125 mg/liter), with C₁₂ (0.5×, 1×, or 2× MIC) or with C₁₄ (0.5×, 1×, or 2× MIC). (b) Quantitative representation of the images shown in panel a.

Annexin V and PI staining analysis.

Annexin V and PI staining assays were performed to determine whether the killing effect of TOB analogues involves apoptosis or necrosis in yeast cells. Phosphatidylserine (PS) is distributed asymmetrically in the inner leaflet of the lipid bilayer of plasma membranes of yeast cells (38). Externalization of PS on the outer surface of the plasma membrane is an early symptom of apoptosis or programmed cell death that can be detected by FITC-conjugated Annexin V staining (green) in yeast cells. Similarly, PI dye only stains necrotic cells, as intact fungal membranes are impermeable. As shown in Fig. 6, C₁₄, at subinhibitory doses, induced both early apoptosis (∼13%) and late apoptosis/necrosis (∼40%) of C. albicans ATCC 64124 (strain B) after 1 h of exposure. However, untreated yeast cells did not show any Annexin V or PI staining.

FIG 6.

Annexin V and propidium iodide stains of C. albicans ATCC 64124 (strain B) cells treated with 6″-thioether C₁₄ TOB analogue (4 mg/liter) for 1 h (right panel) and the bright-field image (left panel). Note: in the right panel, the letter a indicates early apoptotic cells that stained only green, the letter b indicates healthy cells that did not stain, and the letter c indicates late apoptotic/necrotic cells that stained green and orange.

DISCUSSION

Previously, we showed chain length-dependent antibacterial activities of TOB analogues derivatized at the 6″-position with linear alkyl chains (C₆ to C₂₂) against VRE and E. coli with resistance to the parent compound TOB itself (7). In the current study, we further expanded the antibacterial spectra of these analogues along with that of a newly synthesized C₄ TOB analogue against various clinically important bacterial isolates. Unlike TOB analogues with longer alkyl chains, C₄ demonstrated strong antibacterial activities against L. monocytogenes ATCC 19115 (strain B+; 2.3 mg/liter) and H. influenzae ATCC 51907 (F−; 2.3 mg/liter) that were also comparable to the activity of TOB itself (1.2 mg/liter) against these strains (Table 1). Among all of these analogues, C₁₂ and C₁₄ exhibited moderate to strong antibacterial activities against most of the bacterial strains, except against M. smegmatis MC2 155 (E+; >150 mg/liter) and H. influenzae ATCC 51907 (F−; >150 mg/liter) (Table 1). The MIC values of C₆ to C₁₄ determined in this study were mostly consistent with previously reported MIC values of these analogues against the specific strains of bacteria (7).

Recently, the amphiphilic kanamycin A and B derivatives K20 and FG08, respectively, with C₈ linear alkyl chains were shown to have promising antifungal properties with no or mild antibacterial properties, respectively (8–10). This report has motivated us to explore the potential antifungal activities of our C₄ to C₁₄ TOB analogues that were yet to be determined. The C₄ to C₁₄ analogues demonstrated a parabolic pattern of chain length-dependent antifungal activity, and C₁₂ and C₁₄ were found to be the most potent among them all. A similar pattern was also observed for the antibacterial properties of these analogues (7). Both C₁₂ and C₁₄ showed moderate growth-inhibitory effects against all tested fungal strains, with a relatively high degree of inhibition against C. neoformans MYA-895 (strain H) and against two filamentous fungi, A. nidulans 38163 (I) and F. graminearum 053 (J). Intriguingly, the C₈ TOB analogue showed no activity against fungi tested in this study, except against C. neoformans MYA-895 (H). In contrast to kanamycin derivatives FG08 and K20, our TOB analogues did not require an optimal chain length of C₈ to show antifungal properties (9, 19). Not only were the MIC values of C₁₂ and C₁₄ at least 1- to 2-fold lower than amphiphilic kanamycin derivatives FG08 and K20, but also C₁₂ and C₁₄ were active against A. nidulans 38163 (I) while FG08 and K20 were completely inactive against Aspergillus strains (Table 2) (9, 19). It is also worth noting that the MIC values of C₁₂ and C₁₄ against resistant fungal strains, except for C. albicans 10231 (strain A), C. neoformans MYA-895 (H), and A. nidulans 38163 (I), were lower than those of POS, ITC, and FLC (Table 2). These results not only indicate that C₁₂ and C₁₄ may have different modes of action than the conventional azoles against fungi, but also it appears that they may circumvent the azole resistance mechanisms of resistant fungi such as C. albicans ATCC 64124 (B), which has mutations in its ERG11 sequences (39, 40).

Another noteworthy feature of the C₁₂ and C₁₄ analogues is that they showed a fungicidal effect on the azole-resistant strain C. albicans ATCC 64124 (strain B), inhibiting it completely after 24 h (Fig. 3). However, as expected, FLC was fungistatic against C. albicans ATCC 64124 (B) and completely lost its activity after 6 h, which was consistent with previous reports (41). Also, C₁₂ and C₁₄ lysed 50% of mouse erythrocytes at concentrations that were 4- to 32-fold and 8- to 32-fold higher than their antifungal MICs, respectively (data not shown), and these results were consistent with the results published previously for these analogues (7). However, in this study, we expanded the reports of effects of C₁₂ and C₁₄ on nucleated mammalian cells. C₁₂ and C₁₄ did not affect A549 (human lung carcinoma epithelial cells) or Beas 2B (normal human bronchial epithelial cells) proliferation even at concentrations higher than their antifungal MIC values, thus suggesting a selective antifungal activity (Fig. 4).

It has been shown that C₁₄ targets bacterial membranes over the ribosomes, as does the parent drug (7), but the effect of C₁₄ on fungal membranes was yet to be determined. We chose the most potent TOB analogues, C₁₂ and C₁₄, to explore their abilities to affect fungal membrane integrity by measuring PI dye influx into C₁₂- or C₁₄-treated C. albicans ATCC 64124 (strain B). Both C₁₂ and C₁₄, at their MIC and at 2× MIC, triggered rapid influx of PI by yeast cells after 1 h of exposure (Fig. 5). It is noteworthy that C₁₂ and C₁₄, even at 0.5× MIC, caused more PI staining of yeast cells than POS (62.5 mg/liter) after 1 h of exposure, which further supports our proposed statement of different modes of action of TOB analogues than that of azoles against fungi. The rapidity of PI dye influx into C₁₂- or C₁₄-treated yeast cells, however, suggests that the possible cause of killing may be associated with a direct effect on fungal membranes. The membrane primary or secondary effects of TOB analogues against fungi are not known. PS externalization on the outer surface of the lipid bilayer of plasma membranes is a sign of early apoptosis in yeast cells. The yeast cells that were treated with C₁₄ at a subinhibitory concentration did show both early apoptosis (Fig. 6, indicated by the letter a), late apoptosis/necrosis (indicated by the letter c), and no apoptosis (as indicated by the letter b), as demonstrated in the Annexin V and PI assay. We speculate that the induction of early apoptosis by C₁₄ causes externalization of phosphatidylserine on the plasma membrane by making the surface highly negatively charged, which could then react with the polycationic amphiphilic aminoglycosides generated. However, thorough experimental analysis, outside the scope of this study, is desirable for the future to elucidate the exact target of these analogues on fungal membranes.

In conclusion, C₁₂ and C₁₄ showed broad-spectrum antifungal activities against yeasts and filamentous fungi as they targeted fungal plasma membranes. They were less hemolytic and showed low mammalian cell toxicities. This study provides novel lead compounds for the development of antifungal drugs, which we are currently following by synthesizing and investigating the properties of additional AGs derivatized in 6″-thioether form with long linear alkyl chains.