Discovery of potential antifungal triazoles: design, synthesis, biological evaluation, and preliminary antifungal mechanism exploration

Synthesis of novel triazole-type miconazole analogues as potential antifungal agents and molecular modeling with CYP51 and experimental investigation with DNA suggested the possible antimicrobial mechanism.

Abstract

A series of triazoles as miconazole analogues was designed, synthesized and characterized by IR, NMR, MS and HRMS. All the newly prepared compounds were screened for their antifungal activities against five kinds of fungi. The bioactive assay showed that most of the synthesized compounds exhibited good or even stronger antifungal activities in comparison with the reference drugs miconazole and fluconazole. In particular, the 3,4-dichlorobenzyl derivative 5b showed a comparable or superior activity against all the tested fungal strains to standard drugs, and formed a supramolecular complex with CYP51 via the hydrogen bond between the 4-nitrogen of the triazole nucleus and the histidine residue. Preliminary experiments revealed that both of the active molecules 5b and 9c could intercalate into calf thymus DNAs, which might block DNA replication to exhibit their powerful antifungal abilities. Further studies indicated that compound 5b might be stored and transported by human serum albumin through hydrophobic interactions, specific electrostatic interactions and hydrogen bonds. These results strongly suggested that compound 5b could serve as a promising antifungal candidate.

1. Introduction

The rapidly emerging multidrug resistant microorganisms have exerted a great threat to human health. The development of novel antimicrobial drugs is an important approach to combat the increasingly frequent pathogenic drug resistance,1 which has been a major task worldwide for medicinal chemists. Currently, modification of available clinic drugs has been acknowledged as a prominent method to exploit new drugs with a view of reducing cross resistance.2

Miconazole is the first well-known commercially significant imidazole drug with a structure of α-aryl azolyl ethanol; this drug is used as a well-established treatment for many mycotic infections with low toxicity and an excellent safety profile.3 It has been generally considered that miconazole acts by competitive inhibition via directly inhibiting a cytochrome P450 (CYP450)-dependent enzyme, 14α-lanosterol demethylase (CYP51), which results in the lethal disruption of the normal sterol biosynthesis chain in fungi, but it is of minimal consequence to mammals.4 Nevertheless, the shortcomings of miconazole, including its poor aqueous solubility, limited active spectrum, lack of oral absorption and occasional undesirable side effects when administered at therapeutic doses intravenously, have imposed restriction on its efficacy and administrable mode.5 It is conceivable that conformationally constrained miconazole analogues mimicking bioactive conformations might exhibit a higher level of intrinsic antifungal potency.6 Furthermore, the analogues of miconazole or other further structural modification of the analogues could produce candidates from this series for preclinical progression.7 Therefore, much effort has been devoted to its structural modification and the development of its analogues and many excellent achievements have been obtained.8 Lots of miconazole analogues such as econazole, fenticonazole, sulconazole and so on have been successfully developed, marketed and extensively used in clinics for the treatment of fungal infection with high safety, low toxicity and few adverse effects.9

Triazole nuclei have been playing an important role in anti-infective therapy, and in particular, triazole-based fluconazole as a first-choice antifungal drug displays advantages like excellent safety profile, favorable pharmacokinetic characteristics and wide biological activities. A large number of triazole antifungal drugs have been widely used in clinics such as fluconazole, voriconazole, itraconazole and posaconazole.10 Additionally, the toxicity caused by imidazole antifungal drugs is generally considered to result from their strong coordination potency with Fe²⁺ ions, while triazole with a lower electronic density as an isostere of imidazole could helpfully reduce the toxicity due to its inferior coordination potency to imidazole nuclei.11 The structure activity relationship (SAR) also suggests that imidazole rings are susceptible to metabolic degradation in vivo, and the introduction of triazole rings to the third generation azole antifungal drugs is much less susceptible to metabolic degradation in vivo and successfully used in clinics.12 Furthermore, so far few studies have reported the insertion of triazole rings to miconazole scaffolds.

In view of the above observations and on the basis of our previous work on triazoles,13 herein we introduced a triazole nucleus as an isostere of an imidazole ring on miconazole, substituted chlorine on a benzene ring by fluorine, and replaced a dichlorophenyl group of miconazole by alkyl chains with different lengths, various substituted benzenes and heterocyclic substituents to generate a triazole class of miconazole analogues (Fig. 1). Bis-triazoles have been acknowledged to exhibit better water solubility, strengthened targeting ability and improved physicochemical properties than single triazolyl compounds by multi-site binding through the incorporation of another triazole fragment, thus enhancing their bioactivities and therapeutic effect.14 Some bis-triazole antifungal drugs like fluconazole and fosfluconazole have been widely used in clinics with high stability, low cytotoxicity and an efficient administration mode.15 This encourages the insertion of an additional triazole ring to a target framework. Moreover, it is well known that the coumarin nucleus has a large conjugated system, possessing desirable electronic and charge-transport properties, and these characteristics result in the extensive potential applications in antimicrobial aspects.16 In particular, along with the rise of resistant Candida strains, research studies on coumarins as antifungal alternatives have become more and more active.17 Thereby, coumarin fragments were also incorporated into target molecules to study their contribution to antifungal activities. Moreover, triazoliums were reported to helpfully improve antimicrobial activity,18 so some compounds were further transformed into triazoliums. This triazole type of miconazole analogues might be expected to exhibit large potentiality against fungal strains. Therefore, all the newly synthesized compounds were evaluated in vitro for their antifungal activities and their SAR was also discussed. The supramolecular interaction of the most active molecule with miconazole-targeting CYP51 was investigated, and deoxyribonucleic acid (DNA) was selected as a target to further explore the possible antifungal mechanism by studying the interaction with the active compounds. The binding behavior of the active compound with human serum albumin (HSA) by experimental and computational methods was determined in order to understand its absorption, distribution, and metabolism.

Fig. 1. Design of triazole-type miconazole analogues.

2. Chemistry

Commercially available m-difluorobenzene 1 was acylated to produce intermediate 2, which then was reacted with triazole in the weak base potassium carbonate to afford triazole derivative 3 in a high yield of 85%. Compound 3 was reduced with sodium borohydride to yield the triazolyl ethanol 4 in an almost quantitative yield of 97%, which was further treated with a series of substituted benzyl halides in dry tetrahydrofuran (THF) using sodium hydride as base under a nitrogen atmosphere to afford the corresponding ether triazoles 5a–h with yields ranging from 33.4% to 92.9% and alkyl triazole derivatives 7a–e with yields of 21.9–75.6%. Bis-triazoles 8a–c in yields of 62.1–67.2% could be obtained by nucleophilic substitution of triazolyl ethanol 4 with dibromoalkanes. Coumarin alkyl halides19 were coupled with triazolyl alcohol 4 to produce coumarin derivatives 9a–c with yields of 42.6–62.6%. Further quaternization of triazoles 5a–h using halobenzyl halides produced the corresponding triazoliums 6a–e in moderate to good yields after purification (Scheme 1). All the new synthesized conjugates were characterized using IR, ¹H NMR, ¹³C NMR, MS and HRMS (ESI‡).

Scheme 1. Reagents and conditions: (i) ClCH₂COCl, AlCl₃, CH₂Cl₂, 20 °C; (ii) triazole, K₂CO₃, CH₃CN, 25–0 °C; (iii) NaBH₄, MeOH, 0–25 °C, 4 h; (iv) substituted benzyl halides, NaH, dry THF, N₂, 0–74 °C, 5–7 h; (v) substituted benzyl halides, K₂CO₃, CH₃CN, 80 °C, 8–12 h; (vi) alkyl halides, NaH, dry THF, N₂, 0–45 °C, 4–6 h; (vii) NaH, dry THF, dibromoalkanes, N₂, 0–45 °C, 6–8 h; (viii) NaH, dry THF, N₂, coumarin alkyl halides, 0–75 °C, 8–12 h.

3. Results and discussion

3.1. Antifungal activities

The obtained results are depicted in Table 1. Minimal inhibitory concentration (MIC, μg mL^–1) is defined as the lowest concentration of target compounds that completely inhibited the growth of fungi (provided by the School of Pharmaceutical Sciences, Southwest University and the College of Pharmacy, Third Military Medical University), and determined by means of a standard two-fold serial dilution method in 96-well microtest plates according to the National Committee for Clinical Laboratory Standards (NCCLS).20Table 1 revealed that most of the newly synthesized triazoles showed comparable or even superior inhibitory activities against all the tested fungal strains to standard drugs fluconazole and miconazole, except for compounds 4, 5d, 6c, 8c and 9a without obvious antifungal efficacy.

Table 1. In vitro antifungal data for MIC (μg mL^–1) of triazoles 4–9 ^a .

Compds	Fungi					Compds	Fungi
	C. utilis	A. flavus	B. yeast	C. albicans	C. mycoderma		C. utilis	A. flavus	B. yeast	C. albicans	C. mycoderma
4	256	>256	>256	128	256	6e	32	64	32	32	8
5a	4	16	8	8	16	7a	32	128	64	4	1
5b	1	8	0.5	1	4	7b	8	128	128	64	64
5c	4	16	4	4	8	7c	16	32	32	64	32
5d	>256	>256	>256	>256	>256	7d	16	128	16	4	16
5e	4	32	16	0.5	2	7e	>256	128	8	>256	>256
5f	8	16	16	8	8	8a	8	256	8	64	256
5g	8	32	16	16	16	8b	128	64	64	128	128
5h	4	8	16	8	16	8c	>256	256	128	256	256
6a	8	32	64	16	32	9a	256	256	128	128	256
6b	32	64	16	32	64	9b	16	32	4	4	2
6c	>256	>256	128	>256	>256	9c	16	16	16	2	1
6d	32	16	64	8	16	Fluconazole	8	256	16	1	4
						Miconazole	16	256	32	4	8

^a C. utilis, Candida utilis; A. flavus, Aspergillus flavus; B. yeast, Beer yeast; C. albicans, Candida albicans; C. mycoderma, Candida mycoderma.

The substitution of the hydroxyl group in compound 4 by halobenzyl halides yielded the triazole type of miconazole analogues 5a–h (MIC = 0.5–32 μg mL^–1) with comparable or even superior inhibitory activities against all the tested fungal strains to standard drugs fluconazole (MIC = 1–16 μg mL^–1) and miconazole (MIC = 4–32 μg mL^–1) except for 4-nitrobenzyl compound 5d with MIC values greater than 256 μg mL^–1. In particular, compounds 5a, 5b, 5c and 5e could efficiently inhibit the growth of A. flavus strain in vitro with MIC values ranging from 8 to 16 μg mL^–1, which were 16-fold and 32-fold more active than fluconazole and miconazole (MIC = 256 μg mL^–1) respectively. B. yeast strains were quite sensitive to target compounds 5a (MIC = 8 μg mL^–1), 5b (MIC = 0.5 μg mL^–1) and 5c (MIC = 4 μg mL^–1) in comparison with miconazole (MIC = 32 μg mL^–1) and fluconazole (MIC = 16 μg mL^–1), and it was apparent that 3,4-dichlorobenzyl compound 5b exhibited the highest antifungal activity among these conjugates, which was 64-fold more potent than miconazole. Prominently, the 3,4-dichlorobenzyl derivative 5b displayed not only stronger antifungal efficacies, but also a broader bioactive spectrum against all the tested strains at low inhibitory concentrations (MIC = 0.5–8 μg mL^–1) in comparison with reference drugs fluconazole and miconazole. This revealed that 3,4-dichlorobenzyl triazole 5b could serve as a lead compound in the development of more effective broad-spectrum antifungal agents. 2,4-Diflurobenzyl triazole also exhibited superior or comparable efficacies (MIC = 4–16 μg mL^–1) than miconazole (MIC = 4–256 μg mL^–1) and prominently its anti-A. flavus activity was 16-fold more potent than fluconazole and miconazole. However, 4-nitrobenzyl compound 5d had nearly no inhibitory potentiality against the tested fungi with MIC values greater than 256 μg mL^–1. Additionally, 2-fluorobenzyl miconazole analogue 5e (MIC = 0.5 μg mL^–1) gave strong antifungal efficacies against C. albicans strains, which was superior to the standard drugs, fluconazole (MIC = 1 μg mL^–1) and miconazole (MIC = 4 μg mL^–1). Thus, 2-fluorobenzyl analogue 5e can hopefully serve as a new potential anti-C. albicans drug that deserves further investigation. Against fungi C. utilis and B. yeast, 2-chlorobenzyl compound 5f and 3-chlorobenzyl compound 5g showed comparable inhibitory activities with fluconazole (MIC = 8 and 16 μg mL^–1, respectively), while the 4-chlorobenzyl analogue 5h presented enhanced potencies in the inhibition of C. utilis and A. flavus with a MIC of 4 and 8 μg mL^–1 respectively as compared to compounds 5f and 5g, indicating that the 4-chloro substitution on the benzene ring was more beneficial than the replacement at other position. It could also be deduced that chloro and fluoro substituents on the benzene ring exerted more important influence on antifungal activities than the nitro group, since 3,4-dichlorobenzyl compound 5b, 2,4-difluorobenzyl derivative 5c and 2-fluorobenzyl miconazole analogue 5e could effectively inhibit the growth of most tested fungal strains at low concentrations, which were much more effective than nitrobenzyl compound 5d. It was also noticed that disubstituted benzyl triazoles (MIC = 0.5–16 μg mL^–1) generally exhibited a higher potency than these monosubstituted benzyl compounds (MIC = 0.5–32 μg mL^–1).

The quaternization of halobenzyl compounds 5a–e could afford the corresponding triazoliums 6a–e, which exhibit totally decreased antifungal activities in comparison with their precursors 5a–e. In particular, the ionized product 6c with MIC values greater than 256 μg mL^–1 showed much lower efficiencies than 2,4-diflurobenzyl compound 5c and the reference drugs. Therefore, it was suggested that the N-3 unsubstituted triazole ring was a necessary functional group to exhibit antifungal activity against fungal strains. Nonetheless, most of them with MICs in the range of 16–64 μg mL^–1 showed a better anti-A. flavus activity than fluconazole and miconazole (MIC = 256 μg mL^–1). On the other hand, alkyl derivatives 7a–e exhibited moderate to weak antifungal efficiencies towards the tested fungi in comparison with aryl compound 5a–h and the reference drugs. However, target compounds 7a–e showed better antifungal activities against A. flavus than fluconazole and miconazole. Furthermore, the propyl miconazole analogue 7a displayed comparable anti-C. albicans potencies and an 8-fold more potent inhibitory activity against C. mycoderma (MIC = 4 and 1 μg mL^–1, respectively) in comparison with miconazole (MIC = 4 and 8 μg mL^–1, respectively). Butyl substituted compound 7b gave a comparable or 2-fold more potent potency against C. utilis in comparison with fluconazole (MIC = 8 μg mL^–1) and miconazole (MIC= 16 μg mL^–1). When the alkyl group was extended to the octyl group, the produced octyl derivative 7d could restrain the growth of C. albicans at 4 μg mL^–1, which was identical to miconazole. The replacement of alkyl substituent with the dodecyl group produced the dodecyloxyl derivative 7e, which had a 2-fold increased potentiality (MIC = 8 μg mL^–1) against B. yeast in comparison with the standard drug miconazole (MIC = 16 μg mL^–1). These results indicated that the length of the alkyl chain exhibited no obvious effects on biological activities. Nevertheless, these alkyl substituted analogues behaved less active than aryl containing compounds, which manifested that aryl groups were more favorable than alkyl chains to some extent.

The incorporation of another triazole fragment at the alkyl end of compounds 7a–c afforded bis-triazole compounds 8a–c with lower antifungal potencies than 7a–c. In particular, butyl- and hexyl-bridged bis-triazoles 8b and 8c showed a middle to poor antifungal activity against all the tested strains. However, compounds 8a–c generally gave better activities against almost all of the tested fungi than triazole alcohol 4, and it was noteworthy that propyl-linked bis-triazole 8a with a MIC of 8 μg mL^–1 displayed comparable or better fungistatic abilities against C. utilis and B. yeast than the corresponding reference drugs fluconazole and miconazole, which might be the consequence of the shortening of the alkyl linker length between triazole ethers. In addition, the substitution of one triazole fragment on compounds 8a–c by a coumarin scaffold could generate coumarins 9a–c with increased inhibitory efficacies in comparison with 8a–c and precursor 4, and in particular, coumarins 9b and 9c containing (CH₂)₄ and (CH₂)₆ moieties exhibited extremely high antifungal activities against C. albicans and C. mycoderma with MIC values ranging from 1 to 4 μg mL^–1. Particularly, compound 9c gave the most prominent anti-C. mycoderma activity with a MIC of 1 μg mL^–1 in comparison with the standard drugs. It was indicated that in the coumarin series, the suitable increase of carbon linker length was favourable to the fungi inhibitory efficacies.

3.2. Molecular modeling

To further rationalize the observed antifungal activity, a flexible ligand–receptor docking investigation was undertaken. The crystal structure data (CYP51, PDB code: 5V5Z) was obtained from the RCSB protein data bank. Miconazole analogue 5b was selected to dock with CYP51. According to the docking results (Fig. 2), the binding energy and constant of CYP51 with the target compound 5b is –7.97 kcal mol^–1 and 7 × 10⁵ M^–1, respectively. The triazolyl moiety could bind with the active sites of CYP51 through a non-covalent binding mode. Noticeably, the nitrogen atom at 4-position in the triazole ring formed a hydrogen bond with a histidine residue of CYP51.

Fig. 2. Three-dimensional conformation of compound 5b docked in the active site of CYP51.

3.3. Interactions of compounds 5b and 9c with calf thymus DNA

DNA as a significant information molecule encoding genetic instructions has been regarded as an important target for studies of bioactive molecules like antimicrobial drugs to explore the possible antimicrobial action mechanism. The interaction studies of compound 5b (exhibiting good inhibition against all fungal strains) and 9c with a coumarin scaffold (displaying high efficiency in inhibiting the growth of C. albicans and C. mycoderma) with calf thymus DNA (Sigma-Aldrich, St. Louis, MO, USA) at a molecular level were carried out in vitro by a UV–vis method.

3.3.1. Absorption spectra of DNA in the presence of compounds 5b and 9c

It is well-known that hypochromism and hyperchromism are very important spectral features to distinguish the change of the DNA double-helical structure in the absorption spectroscopy. The UV–vis spectra (Fig. 3 and 4) showed that the maximum absorption peak of DNA at 260 nm exhibited a proportional increase with the increasing concentration of compounds 5b and 9c. Furthermore, it was indicated that the measured value of the 5b–DNA complex was a little smaller than the absorption value of the simply sum of free DNA and free compound 5b (inset of Fig. 3). This meant a weak hypochromic effect existed between the DNA and compound 5b, which was a consequence of the interaction between the electronic states of an intercalating chromophore and that of the DNA base and also indicated a close proximity of aromatic chromophores to DNA bases, while the measured value of the 9c–DNA complex was higher than the sum of free DNA and free compound 9c (inset of Fig. 4), which suggested a hyperchromism between 9c and DNA and further proved the conformational change in the DNA duplex. This might result from the non-covalent interactions between the complexes and DNA, which led to the part uncoiling of the DNA helix and the exposure of some previously embedded DNA bases.

Fig. 3. UV absorption spectra of DNA with different concentrations of compound 5b at pH 7.4 and room temperature. c(DNA) = 5 × 10^–5 mol L^–1 and c(compound 5b) = 0–2.3 × 10^–5 mol L^–1 for curves a–i respectively at an increment of 0.29 × 10^–5 mol L^–1. Inset: Comparison of absorption at 260 nm between the 5b–DNA complex and the sum values of free DNA and free compound 5b.

Fig. 4. UV absorption spectra of DNA with different concentrations of compound 9c at pH 7.4 and room temperature. c(DNA) = 5 × 10^–5 mol L^–1 and c(compound 9c) = 0–2.1 × 10^–5 mol L^–1 for curves a–i respectively at an increment of 0.26 × 10^–5 mol L^–1. Inset: Comparison of absorption at 260 nm between the 9c–DNA complex and the sum values of free DNA and free compound 9c.

On the basis of the variations in the absorption spectra of DNA upon binding to 5b and 9c, eqn (1) can be utilized to calculate the binding constant (K).

1

A ⁰ and A represent the absorbance of DNA in the absence and presence of compounds 5b and 9c at 260 nm, ξ_C and ξ_D–C are the absorption coefficients of compounds 5b or 9c and 5b–DNA or 9b–DNA complexes, respectively. The plot of A⁰/(A – A⁰) versus 1/[compound 5b or 9c] is constructed using the absorption titration data and linear fitting (ESI,‡ Fig. S10 and S11), yielding the binding constant, K = 1.04 × 10⁴ or 5.06 × 10⁴ L mol^–1, R = 0.9936 or 0.9964, SD = 0.3181 or 0.0729, respectively (R is the correlation coefficient. SD is the standard deviation). Coumarin 9c had a higher binding constant than triazole 5b, demonstrating that compound 9c could combine more tightly with DNA, which might be resulted from the existence of a large conjugated coumarin backbone.

3.3.2. Absorption spectra of NR interaction with DNA

Neutral red (NR) as a planar phenazine dye with low toxicity, high stability and convenient application is structurally similar to other planar dyes, acridine, thiazine and xanthene, and it has been evidenced that the binding of NR with DNA belongs to the intercalative binding type.21 Therefore, NR was used as a spectral probe to investigate the binding mode of 5b or 9c with DNA in this work. The absorption spectrum of NR upon the addition of DNA (ESI,‡ Fig. S12) indicated that a gradual decrease with the increasing concentration of DNA in the intensity of the absorption peak of NR around 460 nm was observed, and a new band around 530 nm was developed, which suggested the formation of a new DNA–NR complex. This was further proved by the isosbestic point at 504 nm.

3.3.3. Absorption spectra of competitive interaction of compounds 5b or 9c and NR with DNA

The competitive binding between NR and 5b or 9c with DNA was observed in the absorption spectra (ESI,‡ Fig. S13 and S14). With the increasing concentration of compound 5b or 9c, an obvious intensity increase was observed around 275 nm. In comparison with the absorption around 460 or 530 nm of the NR–DNA complex, the absorbance at the same wavelength presented a reverse process (ESI,‡ inset of Fig. S13 and S14). These various spectral changes suggested that compound 5b or 9c could intercalate into DNA by substituting NR in the DNA–NR complex, which might further block the DNA replication, thus exhibiting its powerful antifungal activities.

3.4. Iodide quenching experiments

Steady-state quenching could provide further information about the binding mode of molecules with DNA. It is generally accepted that if a molecule is protected from being quenched in the presence of an anionic quencher due to the repulsion between the anionic quencher and the negatively charged phosphate backbone of DNA, the binding mode belongs to intercalation.22 Accordingly, the intercalative binding of the fluorophore should reduce the extent of the fluorescence quenching in the DNA environment in comparison to the reference solution, and hence the value of the quenching constant (K_SV) of the intercalatively bound molecule should be lower than that of the non-intercalative molecule. In this work, KI quenching was introduced to investigate the possible binding mechanism. Quenching curves of compound 5b in the absence and presence of DNA are shown in Fig. 5 and 6.

Fig. 5. The fluorescence spectra of compound 5b with increasing concentration of KI. c(compound 5b) = 1.0 × 10^–5 mol L^–1; c(KI) = 0–12 × 10^–3 mol L^–1 for curves a–g respectively at an increment of 2 × 10^–3 mol L^–1; the red line shows the fluorescence spectrum of compound 5b only; T = 298 K and λ_ex = 290 nm. Inset: The Stern–Volmer plot of the fluorescence titration data of compound 5b.

Fig. 6. The fluorescence spectra of compound 5b and DNA system with increasing concentration of KI. c(compound 5b) = 1.0 × 10^–5 mol L^–1; c(DNA) = 2.0 × 10^–5 mol L^–1; c(KI) = 0–12 × 10^–3 mol L^–1 for curves a–g respectively at an increment of 2 × 10^–3 mol L^–1; the red line shows the fluorescence spectrum of compound 5b and DNA system only; T = 298 K and λ_ex = 290 nm. Inset: The Stern–Volmer plot of the fluorescence titration data of compound 5b and the DNA system.

Stern–Volmer eqn (2) was used to deduce the quenching constant:

2

where F₀ and F represent the fluorescence intensity of the compound 5b/5b–DNA system in the absence and presence of the quencher KI, respectively. K_SV (L mol^–1) is the Stern–Volmer quenching constant, and [Q] is the concentration of KI. Hence, K_SV was calculated by the linear regression of the F₀/F versus [Q] plot. The K_SV value of compound 5b (10.7 L mol^–1, R = 0.999, SD = 0.019) in the presence of DNA was lower than that (11.6 L mol^–1, R = 0.992, SD = 0.067) in the absence of DNA. It was apparent that the iodide quenching effect of compound 5b by an anionic quencher in presence of DNA was decreased, which further evidenced that compound 5b intercalated into the base pairs of DNA.

3.5. Binding behavior of compound 5b with HSA

The investigations of the interactions between drugs or bioactive small molecules and HSA are not only beneficial to providing a proper understanding of the absorption, transportation, distribution, metabolism and excretion properties of drugs, but also significant to designing, modifying and screening drug molecules.23 Therefore, the binding behavior between compound 5b and HSA (Sigma-Aldrich, St. Louis, MO, USA) was investigated (Fig. 7). It was obvious that HSA had a strong fluorescence emission with a peak at 348 nm owing to the single Try-214 residue. The intensity of this characteristic broad emission band regularly decreased with the increasing concentrations of compound 5b, but the maximum emission wavelength of HSA remained unchanged. This suggested that Trp-214 did not undergo any change in polarity, and hence compound 5b was likely to interact with HSA via the hydrophobic region located in HSA.

Fig. 7. Emission spectra of HSA in the presence of various concentrations of compound 5b: c(HSA) = 1.0 × 10^–5 mol L^–1; c(compound 5b) = 0–2.0 × 10^–5 mol L^–1 for curves a–k at an increment of 0.20 × 10^–5 mol L^–1; the blue dash line shows the emission spectrum of compound 5b only; T = 286 K and λ_ex = 295 nm.

The fluorescence quenching data can also be analyzed by the well-known Stern–Volmer eqn (3):

3

where F₀ and F represent fluorescence intensity in the absence and presence of compound 5b, respectively. K_q is the bimolecular quenching rate constant (L mol^–1 s^–1), τ₀ is the fluorescence lifetime of the fluorophore in the absence of a quencher, assumed to be 6.4 × 10^–9 s for HSA, and [Q] is the concentration of compound 5b. The results also manifested that the binding constants were moderate. Therefore, compound 5b might be stored and carried by HSA.

Molecular docking evaluation was also performed to understand the binding mode between compound 5b and HSA. The docking mode is shown in Fig. 8 with the lowest binding free energy and the binding constant of –17.32 kJ mol^–1 and 6 × 10⁴ M^–1, respectively. As clearly evidenced, compound 5b was surrounded by Lys-413, Lys-541, Arg-410, Eu-418, Ala-406, Trp-214 and Thr-540. Moreover, compound 5b partly occupied the subdomain IIA, resulting in the fluorescence quenching of Trp-214. Hydrophobic interactions existed between the aromatic ring of compound 5b and HSA. Except for the hydrophobic contacts, hydrogen bonds of 2-F, 4-F on the benzene ring as well as the 4-N atom on the triazole ring with Lys-413, Lys-541 and Ala-406 in HSA and specific electrostatic interactions were also involved in the binding process. All of these indicated that the simulation results coincided well with the above experimental analysis.

Fig. 8. Hydrophobic pocket of the HSA-target and binding model of compound 5b to HSA.

4. Conclusion

A triazole class of miconazole analogues was designed and synthesized in good yields via an easy, convenient and efficient synthetic route. All the newly synthesized compounds were characterized by IR, ¹H NMR, ¹³C NMR, MS and HRMS. The in vitro antifungal activities of these miconazole analogues were evaluated against five kinds of fungal strains. The biological results revealed that most of the new target compounds exhibited moderate to good antifungal activities against most of the tested strains in comparison with the reference drug miconazole. In particular, the 3,4-dichlorobenzyl triazole compound 5b displayed not only stronger antifungal efficacies, but also a broader bioactive spectrum against all the tested strains at low inhibitory concentrations (MIC = 0.5–8 μg mL^–1) in comparison with miconazole. SAR suggested that both the N-3 unsubstituted triazole ring and the substituents on the ether linkage were important influence factors on antifungal potency and the introduction of chlorine to the benzene ring would improve the fungi inhibitory effect. The docking study showed that the highly bioactive miconazole analogue 5b could bind with the active sites of CYP51 through the formation of a hydrogen bond with a histidine residue. The interactive investigations with DNA validated that both of the two active molecules 5b and 9c could effectively intercalate into calf thymus DNA to form compound–DNA complexes, which might block DNA replication to exhibit their powerful antifungal abilities. The binding behaviors and molecular modeling of compound 5b with HSA suggested that compound 5b might be stored and transported by HSA, where the hydrophobic interactions, specific electrostatic interactions and hydrogen bonds played important roles. All of these indicated that compound 5b was a promising antifungal candidate with good curative effect. Further research studies, including the in vivo bioactive evaluation, time-kill kinetic assay, toxicity, drug resistance and some effect factors on antifungal activities such as other heterocyclic azole rings (benzotriazole, thiazole, oxazole and their derivatives) linked at the methylene end are now in progress in our group.

Conflict of interest

The authors declare no competing interests.