As shown in Scheme 1, firstly, 2-nitrophenols were reduced by SnCl2·2H2O to afford 2-aminophenol derivatives (1a–c) , which were further cyclized with cyanogen bromide (BrCN) in methanol to give the key intermediates 5-substituted 2-aminobenzoxazoles (2a–d). Then, C-5-alkyl-substituted 2-aminobenzoxazole derivatives (3a–y) were smoothly obtained by Suzuki cross-coupling reaction. Meanwhile, amides (4a–j) and sulfonamides (5a–i) derivatives were conveniently prepared by acylation reaction. The structures of all the target compounds were characterized by M. p., 1H NMR, 13C NMR and HRMS.
According to the mycelium linear growth rate method , all the target compounds (2a–d, 3a–y, 4a–j and 5a–i) were screened in vitro for their antifungal activities at 50 μg/mL against eight phytopathogenic fungi [e.g., Fusarium sulphureum (FS), Thanatephorus cucumeris (TC), Fusarium oxysporum (FO), Fusarium graminearum (FG), Botrytis cinerea (BC), Valsa mali (VM), Sclerotiua sclerotiorum (SS) and Alternaria solani (AS)]. Hymexazol, a commercial agricultural fungicide, was used as the positive control.
As depicted in Table 1, most of the target compounds displayed good-to-excellent inhibitory effects on the growth of the tested phytopathogenic fungi at 50 μg/mL. Generally, these compounds can be divided into three groups according to the average inhibition rates. Compounds 3a–e, 3h, 3m, 3u and 3v showed the highest activity with average inhibition rates of 80.7–97.5%, compounds 2b, 2c, 3f, 3i, 3k, 3p, 3q, 3t, 3x, 4h and 5e exhibited moderate activity with average inhibition rates of 46.1–70.2%, and the other compounds displayed weak activity against the tested fungi. For FS strain, it was worth mentioning that thirty compounds exhibited more pronounced antifungal activity (> 52.2%) than the positive control hymexazol (50.4%); particularly, the inhibition rates of compounds 3a, 3b, 3d, 3e and 3v reached 100% at the concentration of 50 μg/mL. For TC strain, seven compounds 3a, 3b, 3c, 3e, 3h, 3m and 3v showed considerable antifungal activity with inhibition rates over 97.8%, which were much better than that of hymexazol (89.7%). Furthermore, compounds 3a–3f, 3m and 3v were found to exhibit higher antifungal activity against FO and FG strains than hymexazol, and compounds 2b, 2c, 3a–3e, 3h, 3m and 3u possessed significant antifungal activity against BC, SS and AS (53.4–100%) in comparison with hymexazol (42.9–88.6%). To our delight, almost all the target compounds displayed better antifungal activity against VM than hymexazol (10.1%). In general, compounds 3a, 3b, 3c, 3e and 3m presented more promising antifungal activity against a broad spectrum of phytopathogenic fungi than the commercial agricultural fungicide hymexazol.
Moreover, some interesting results of structure–activity relationships (SARs) were found as follows (Fig. 2): (1) Introduction of chlorine and bromine atoms is very beneficial to enhance the antifungal activity (2b and 2c vs. 2a and 2d). (2) Amino group is very essential for the antifungal activity, acylation and sulfonylation which can obviously decrease the activity against some tested fungi. (3) Compared with the intermediate compound 2c, selection of suitable aryl groups replacing bromine atom can significantly improve the antifungal effects, such as the inhibition rates of compounds 3a (Ph), 3b (2-FC6H4), 3c (4-FC6H4), 3d (2-ClC6H4), 3e (3-ClC6H4), 3m (3,5-diCH3C6H3), 3u (3-furyl) and 3v (3-thienyl) against most of strains were over 90%. (4) It is noteworthy that introduction of electron-withdrawing groups (such as F, Cl, CF3, NO2 and CN) at the 5-position on phenyl ring of compound 3a could result in more potent compounds than electron-donating groups (such as OH, CH3, Et, NH2) excepted for compounds 3g and 3m (3b–3f, 3h, 3p and 3q vs. 3j, 3l, 3n and 3o). (5) The number and position of chlorine atom also have some influence on the antifungal activity. For instance, mono-chloro compounds (3d, 3e and 3f) exhibited much better antifungal activity than the corresponding bis-chloro compound (3g), and the effect of chlorine atom position on antifungal activity was meta-(3e) > ortho-(3d) > para-(3f). (6) The target compounds bearing furan (3u) and thiophene (3v) ring at the 5-position of compound 2a displayed better inhibition effects than those bearing a pyridine (3w), 2-chloropyridine (3x) and benzothiophene ring (3y). The aforementioned result demonstrates that the antifungal effect can be dramatically influenced by the substituents group on the NH2 and 5-position of 2-aminobenzoxazole.
To further evaluate the inhibitory of the most promising synthesized compounds, the median effective concentrations (EC50) values of compounds 3a–e, 3h, 3m, 3u and 3v against eight phytopathogenic fungi were tested. As shown in Table 2, it was noticed that the nine compounds exhibited impressive antifungal effects against FS, FO, VM, SS and AS, which were better than that of hymexazol. For example, compounds 3a, 3b and 3m exhibited the best anti-FS effects in vitro, with the EC50 values as low as 3.96 μg/mL, 4.47 μg/mL and 4.10 μg/mL, respectively; compounds 3a–e, 3m and 3v exhibited 5.1–12.6 folds more potent activities than hymexazol against FO strain; five compounds 3a, 3c, 3e, 3h and 3m exhibited remarkable antifungal activity against SS strains in vitro, with the corresponding EC50 values of 6.50 μg/mL, 2.82 μg/mL, 1.48 μg/mL, 5.00 μg/mL and 3.82 μg/mL, much superior to hymexazol (25.12 μg/mL). Furthermore, all the target compounds (except 3d and 3u) possessed higher antifungal effects than hymexazol against TC and BC strains. Regarding FG strains, compounds 3a (EC50 = 9.68 μg/mL) and 3b (EC50 = 6.91 μg/mL) are identified with excellent antifungal competence, compounds 3m, 3u, 3v, 3c and 3e displayed moderate activity (EC50 = 12.16–16.60 μg/mL), and compound 3d with the EC50 value of 27.8 μg/mL, which was comparable with that of hymexazol (32.36 μg/mL). This result suggested that the tested fungal displayed high susceptibility to the nine compounds. Meanwhile, the effects of compound 3d on the growth of FS and FO strains at different concentrations are shown in Fig. 3. It’s obviously that the antifungal efficiency significantly depended on the drug concentrations.
The morphological changing of Fusarium solani (FS) and Alternaria solani (AS) was then viewed under the light microscope. From Fig. 4, it can be seen that the FS control group mycelium had an eel shape, smooth surface, uniform size and much-branched, but the compound 3a treatment group mycelium appeared invagination, shriveling and few-branched (FS-CK vs. FS-3a). Meanwhile, the AS mycelium of treatment with compound 3a could produce more less oval-shaped spores than that of control group (AS-CK vs. AS-3a). This phenomenon indicated that these compounds might exert antifungal effect by significantly inhibiting the growth and differentiation of fungal.
Finally, the in vivo antifungal effects of the promising compounds were conducted against Botrytis cinerea on tomato at 100 μg/mL. As shown in Fig. 5, compounds 3a, 3c, 3e and 3m exhibited much better preventative effect than hymexazol (23.1%), with the corresponding preventative rates of 46.7%, 48.3%, 55.1% and 53.3%. The results verified that 2-aminobenzoxazole derivatives could be used as a lead compounds for the discovery of novel agrichemicals.
Molecular docking studies
Studies in the literature have reported that lipid transfer protein Sec14p (Saccharomyces cerevisiae) might be a potential target of inducing fungicidal activity by benzoxazole derivatives . In an effort to elucidate the hypothesis that our compounds acted on sec14p, molecular docking of compounds 3a, 3b, 3d, 3h, 3m and 3u into the homology model according to the binding site of benzoxazoles on reported lipid binding pocket of Sec14p was performed, respectively. As shown in Fig. 6, the six-test compounds shared very similar binding modes with the literature reported. Benzoxazole ring was oriented pi-pi stacked interaction with Tyr 151 and pi-alkyl interaction with Arg208. The substituted phenyl ring made pi-pi stacked, pi-donor hydrogen bond and pi-alkyl interactions with Phe212 and Thr175 and Met209, respectively. This result suggested that the tested compounds were compatible with the active site of sec14p.