Antifungal Activity of Salicylanilides and Their Esters with 4-(Trifluoromethyl)benzoic Acid

Searching for novel antimicrobial agents still represents a current topic in medicinal chemistry. In this study, the synthesis and analytical data of eighteen salicylanilide esters with 4-(trifluoromethyl)benzoic acid are presented. They were assayed in vitro as potential antimycotic agents against eight fungal strains, along with their parent salicylanilides. The antifungal activity of the presented derivatives was not uniform and moulds showed a higher susceptibility with minimum inhibitory concentrations (MIC) ≥ 0.49 µmol/L than yeasts (MIC ≥ 1.95 µmol/L). However, it was not possible to evaluate a range of 4-(trifluoromethyl)benzoates due to their low solubility. In general, the most active salicylanilide was N-(4-bromophenyl)-4-chloro-2-hydroxybenzamide and among esters, the corresponding 2-(4-bromophenylcarbamoyl)-5-chlorophenyl 4-(trifluoromethyl) benzoate exhibited the lowest MIC of 0.49 µmol/L. However, the esterification of salicylanilides by 4-(trifluoromethyl)benzoic acid did not result unequivocally in a higher antifungal potency.


Introduction
The spread of fungal infections has recently been increasing because of the eminent presence of the predisposing risk factors, e.g., malnutrition, invasive surgical procedures, treatment with OPEN ACCESS broad-spectrum antibiotics, corticosteroids or other immunosuppressive agents, immunodeficiency related to diseases like diabetes mellitus or human immunodeficiency virus infections. Mycoses may either be superficial or systemic [1,2]. Immunocompromised patients can develop opportunistic mycoses caused by an expanding spectrum of fungal pathogens, including those with problematic susceptibility to current antifungal drugs [3]. Invasive fungal infections are a major cause of morbidity and mortality in these patients. Candida albicans, Cryptococcus neoformans and Aspergillus fumigatus comprise the most common etiologic agents, but there is an increasing number of infections caused by rare pathogens, including e.g., non-albicans Candida species, opportunistic yeast-like fungi like Trichosporon spp. or non-fumigatus Aspergillus spp. [4].
Pathogenic fungi can also develop resistance to various drugs with unshared mechanisms of action [3]. That is why the development of new antifungal agents still remains challenging, despite the intensive searching in recent times.
Salicylanilides (2-hydroxy-N-phenylbenzamides) have displayed a wide range of potentially interesting biological effects including those against protozoa, fungi, bacteria and mycobacteria or viruses [5,6]. Their mechanism of the action towards microbes is believed to be multiple, with a lot of molecular targets and effects [7].
The temporary masking by an esterification of a salicylanilide free phenolic group, which is presumed to be necessary for the antimicrobial activity, may be beneficial, e.g., due to improved bioavailability and membrane permeability, a high activity and/or a lower toxicity [7]. A correlation between hydrophilic/lipophilic balance (lipophilicity) necessary to penetrate through biological barriers and the antimicrobial activity with an optimal span has been described repeatedly [15,16].
The antifungal properties of salicylanilide esters (acetates, esters with N-protected amino acids) were summarized in our earlier review [7]; later, a mild and sporadic activity was found for salicylanilide benzoates (MIC  3.9 µmol/L) [17] and more importantly for salicylanilide pyrazinoates (MIC  1.95 µmol/L) [18] with a general result that salicylanilide esters are more active and probably fungicidal against filamentous fungi than against Candida strains with only a limited fungistatic activity.
Based on these facts, we synthesised a new series of salicylanilide esters with 4-(trifluoro-methyl) benzoic acid (illustratively, esters of imidazole derivatives with 3-and 4-(trifluoro-methyl)benzoic acids expressed some activity against Candida strains [19]) and evaluated them as well as their parent salicylanilides as potential antifungal agents.
In general, the most antifungal active salicylanilide assayed was N-(4-bromophenyl)-4-chloro-2hydroxybenzamide (1j). Bromine and 4-trifluoromethyl moiety-containing salicylanilides expressed mostly a significant activity towards T. asahii, A. corymbifera and T. mentagrophytes, while dichlorosalicylanilides were only active against the last strain. It is difficult to postulate distinct structure-activity relationships due to a lot of partially inactive derivatives; for some strains the activity is only sporadic.
Similarly, the fact that for most of the esters it was not possible to determine their MIC values, complicates the systematic evaluation of how the esterification by 4-(trifluoromethyl)benzoic acid influenced the activity of the parent salicylanilides. In some cases the esterification improved the activity (e.g., 1e vs. 2e and 1r vs. 2r for Candida sp., 1j vs. 2j for filamentous fungi, 1p vs. 2p for T. mentagrophytes), for others the impact was the reverse (e. g., 1j vs. 2j for yeasts, 1c vs. 2c for  moulds, 1f vs. 2f, 1k vs. 2k). However, only the esters inhibited A. fumigatus, C. tropicalis and also C. krusei (additionally with 1j) at low MIC values. Additionally, some 4-(trifluoromethyl)benzoates (2h, 2j, 2r) share submicromolar MICs for T. mentagrophytes after 72 h of incubation.
Although lipophilicity is generally presumed to belong to the factors influencing antifungal activity, there is no sharp correlation in this series, as demonstrated above for the pairs of salicylanilide vs. its ester. The most lipophilic salicylanilide 1e was not the unambiguously most active derivative; its MIC values were even below-average. Similarly, the least lipophilic fluorinated salicylanilides 1k, 1l, 1m and 1n showed a better activity in some cases than more lipophilic molecules like 1e or 1p.
Similarly, compounds with a comparable calculated logP showed non-uniform MIC values (e. g., 1b vs. 1j, 2h vs. 2p). Moreover, if the antifungal properties depend only on the lipophilicity, the salicylanilides derived from 5-chlorosalicylic acid and 3-substituted anilines should be unanimously more active than those synthesized from 4-chlorosalicylic acid and 4-substituted anilines; but this clear relationship was not observed and N-(4-bromophenyl)-4-chloro-2-hydroxybenzamide (1j) revealed the most significant in vitro antifungal features. However, the logP values presented here are only calculated values, not experimentally found; therefore their significance should be viewed with caution.
Similarly, it was shown that the individual biological impacts of related compounds with various substitution patterns may be in part often influenced by the volume of the substituents. This bulkiness can be expressed, e.g., by a bulk parameter like MR (molar refractivity) [9,22]. The esterification of salicylanilides by 4-(trifluoromethyl)benzoic acid led to the larger molecules with a calculated increment of MR of 34.74. However, as pointed out previously, this modification did not improve the activity unequivocally. Similarly, bulkier derivatives 1e, 1f, 2e and 2f with two chlorines on the aniline ring (MR for these atoms together = 9.6) did not present uniformly superior activity, and only some MICs are excellent, when the remainder are average, weak or comparatively high. On the other hand, the replacement of chlorine (MR = 4.8) by a bulkier bromine (MR = 7.6) on the salicylic acid ring resulted in the improvement of the activity towards T. asahii and A. corymbifera (1q vs. 1s), while for other strains this brought no additional benefit. Similarly, anilines containing bromine (MR = 7.6) showed a higher antifungal activity than those substituted by chlorine (4.8), fluorine (−0.4) and it is slightly superior to a trifluoromethyl group (MR = 4.0). These data indicate, analogously to lipophilicity, that there is no simple linear correlation between biological activity and steric parameter MR.
When compared to other recently evaluated salicylanilide esters, 4-(trifluoromethyl)benzoates did not surpass the activity of salicylanilide acetates [16], although some MICs are comparable. Contrarily, benzoates [17] are much less potent in vitro than the here presented esters. The comparison with salicylanilide pyrazinoates [18] and esters with N-benzyloxycarbonyl amino acids [23] is not unambiguous, while these derivatives showed more consistent activity; however, MICs of many 4-(trifluoromethyl)benzoates exceeded them. Esters with N-acetyl-L-phenylalanine showed less and less uniform antifungal potency [5]. In some cases, the esterification of parent salicylanilides by various carboxylic acids led to a decreased activity.

General Methods
All the reagents and solvents were purchased from Sigma-Aldrich (Seelze, Germany) or Penta Chemicals (Prague, Czech Republic) and they were used as received. Reactions and purity of products were monitored by thin layer chromatography with a toluene/ethyl-acetate 4:1 mixture as eluent for salicylanilides and a toluene/methanol 9:1 mixture as eluent for esters; plates were coated with 0.2 mm Merck 60 F254 silica gel and were visualized by UV irradiation (254 nm). Melting points were determined on a Büchi Melting Point machine B-540 apparatus using open capillaries and the reported values are uncorrected.
Elemental analysis (C, H, N) were performed on an automatic microanalyser CHNS-O CE instrument (FISONS EA 1110, Milano, Italy). Infrared spectra (ATR) were recorded on FT-IR spectrometer Nicolet 6700 FT-IR in the range of 400-4,000 cm −1 . The NMR spectra were recorded on a Varian Mercury-Vxbb 300 (300 MHz for 1 H and 75.5 MHz for 13 C; Varian, Inc., Palo Alto, CA, USA) at ambient temperature using deuterated dimethylsulfoxide (DMSO-d 6 ) solutions of the samples. The chemical shifts δ are given in ppm, with respect to tetramethylsilane as an internal standard. The coupling constants (J) are reported in Hz.
The calculated logP values (ClogP), that are the logarithms of the partition coefficients for octan-1ol/water, were determined using the program ACD/ChemSketch (Freeware) version 12.01 (Advanced Chemistry Development, Inc., Toronto, ON, Canada); while the program CS ChemOffice Ultra version 12.0 (CambridgeSoft, Cambidge, MA, USA) did not allow us to distinguish lipophilicity for particular position isomers providing the same logP data. Only the mean values without deviations are reported in this work. The increment of molar refractivity due to esterification of salicylanilides by 4-(trifluoromethyl)benzoic acid was calculated using the program ACD/ChemSketch (Freeware) version 12.01.

Synthesis of Salicylanilides
A substituted salicylic acid and corresponding aniline (both 0.002 mol) were suspended in chlorobenzene (20 mL) and PCl 3 (0.001 mol) was added. The reaction was carried out with vigorously stirring in a microwave reactor (530 W, 600 rpm, MicroSYNTH Milestone) for 20 min to refluxing. The reaction mixture was filtered while hot, let stand at 20 °C and then at +4 °C for 24 h. The crude product was filtered off and recrystallized from boiling 96% ethanol to obtain the pure product, in all cases white crystals. Salicylanilides, whose physical and spectral properties are not presented here, were synthesized and reported previously [24,25].

Conclusions
In summary, we have synthesized nineteen salicylanilides and eighteen corresponding esters with 4-(trifluoromethyl)benzoic acid. New compounds were characterised and all derivatives were evaluated as potential antimycotic agents towards eight fungal strains. It was not possible to determine MICs of a range of the esters because of a low solubility and/or a precipitation in the testing medium. Salicylanilides and their esters affected the fungal growth from 0.49 µmol/L; however, their activity is not uniform and especially against Candida spp. it seems to be most likely sporadic. However, some of the here described compounds may represent derivatives with a promising in vitro antifungal properties.