REGULAR PAPERS

Structure-activity relationship of berberine and derivatives acting as antifungal compounds

¹Enriz, R. D. ; ²Freile, M. L.

¹ Facultad de Química, Bioquímica y Farmacia. Universidad Nacional de San Luis, Chacabuco 917. (5700) San Luis, Argentina.
² Facultad de Ciencias Naturales. Universidad Nacional de la Patagonia San Juan Bosco Km 4, Comodoro Rivadavia. (9000) Chubut, Argentina
FAX: +54-2652-431301, E-Mail: mfreile@unpata.edu.ar

Received April 24^th, 2006. In final form May 10^th, 2006
Dedicated to Prof. Imre G. Csizmadia on the occasion of his 75^th birthday

Abstract
A structure-activity relationship study of berberine and structurally related compounds is reported. Berberine, was chemically modified in order to obtain canadine and oxyberberine. The in vitro evaluation of antifungal activity of berberine, canadine, oxyberberine, liriodenine, o-methyl-mostachatoline and other four derivatives is also reported here. Our results allow us to determine the minimal structural requirements to produce the antifungal response and can provide a guide for the design of new compounds with these properties.

Resumen
Se presenta un estudio de correlación estructura-actividad de berberina y compuestos estructuralmente relacionados. Se modifico químicamente a berberina con el objeto de obtener canadina y oxiberberina. Se informa también la actividad antifúngica in vitro de bereberina, canadina, oxyberberina, liriodenina, o-metil-mostachatolina y otros cuatro derivados. Nuestros resultados nos han permitido determinar los mínimos requerimientos estructurales para producir la respuesta antifúngica y pueden dar una guía para el diseño de nuevos compuestos con estas propiedades.

Introduction
In the southern regions of Argentina (Patagonia), there are about seexteen different species of Berberis (Berberidaceae) commonl y know as "calafate" or " michay". These plants are representative of this geographical zone and can vary in size from small shrubs to small trees.
Although some reports [1-4] indicate that other species of Berberis have been used as a remedy for different diseases, it appears that, at least in the south of Argentina, these plants had not really been used with therapeutical purposes. First-hand interviews with local Indian communities in the state of Chubut (Argentina) indicate that the Indian people do not use "Calafate" as remedy, they use the roots only as a dye [5].
Recently, we have reported that the aqueous extracts of B.heterophylla do not possess significant antimicrobial activity [6]. In addition the agar dilution method showed that none of the aqueous extracts tested displayed significant antifungal activity against dermatophytes fungi. These results account for, at least in part, the reason the Indian people of Patagonia do not use infusions of Berberis heterophylla for therapeutically. In contrast to those results, berberine (compound I in Figure 1) isolated from B. heterophylla displayed a moderate but significant antibacterial and antifungal activity against Staphylococcus aureus and different Candida spp. Also we evaluated berberine against a panel of Candida spp obtained from the clinical isolated. It is interesting to note that berberine displayed antifungal activity not only against standardized strains, but against clinical isolated of Candidas. We tested it against 7 clinical strains from different body humors from different immunocompromised patients. MIC (minimum inhibitory concentration) values ranged from >128 to 16 mg/ml were obtained for berberine indicating a moderate but significant activity against C. glabrata, C. albicans, C. lusitaniae, C. krusei and C. Parapsilosis [7].

Figure 1 –structures of alkaloids reported here and of structurally related compounds.

A large number of studies have been performed in order to shed some light on the structural aspects and bioactivities of berberine and its congeners. However, compared with these aspects, the action mechanism of these alkaloids, at least at molecular level, has received relatively little attention. In this regard, previously we performed a computer-assisted study reporting the importance of aromatization within the putative bio-medical action mechanism of berberine and related cationic alkaloids with double iso-quinolinoid skeleton [8]. In this supposed mechanism of action of berberine, to prevent DNA replication, the first step is aromatization. In contrast to the covalent dehydrogenation, which is endothermic, the aromatization under ionic conditions was found to be exothermic [9]. Our results indicated that in the aromatization process the ease of hydride ion removal parallels the stabilization energy of the aromatic compounds to be formed. Comparing the nucleophylic additions to the pi-systems, LUMO (Lowest-Unoccupied-Molecular-Orbital) energy values suggested a greater accessibility of the N(+) heterocycles in comparison to the polycicle aromatic hydrocarbons. Thus, it appears that the quaternary nitrogen and the aromatic polycyclic and planar structure of berberine is the pharmacophoric pattern to produce the antifungal effect in this alkaloid.
With the aim to finding an experimental support for the above hypothesis, we report here the obtention of some analogues of berberine, their antifungal effects as well as the structure-activity relationship study performed on these alkaloids acting as antifugal compounds.

Experimental section
Chemistry
Canadine (compound II in Table 1)
Berberine chloride (105 mg) was dissolved in aqueous acetic acid 66.6 % (9 ml), Zn powder was added (10.54 g) with an aqueous HCl solution 35% (23 mL). The solution was refluxed for 2 h at 100 ºC, when TLC indicated a complete reaction. The reaction mixture was neutralized with an aqueous HONH₄ 30% (20 mL). It was extracted with CH₂Cl₂ (3 x 15 mL). The combined CH₂Cl₂ extracts were dried, filtered, and evaporated to yield Canadine (65.4 mg). Canadine was purified by column chromatography and then identified by ¹H and ¹³C NMR and compared with spectral data from literature [10].

Oxyberberine (XIII)

Table 1. MICs (µg/ml) obtained for the different compounds. Active values are denoted in bold.

A) C. albicans (00-604); B) C. glabrata; C) C. lusitaniae; D) C. krusei; E) C. parapsilosis (951706); F) C. albicans (00-622); G) C. parapsilosis (00-629).
* Taken from reference 7

Berberine chloride (50 mg, 0.134 mmol) was dissolved in 20 % aqueous KOH (20 mL). The solution was refluxed for 6 h at 80 ºC, when TLC indicated complete reaction. The reaction mixture was extracted with CH₂Cl₂ (3 x 15 ml). The combined CH₂Cl₂ extracts were dried, filtered, and evaporated to yield oxyberberine (46.5 mg, 98.5%). The oxyberberine was purified by column chromatography and then identified by ¹H and ¹³C NMR and compared with the spectral data from literature[11].
Compounds III (see reference [12]) IV and V [13]; and the oxoaporphines XIV, XV and XVI (reference [14]) have been previously reported and were kindly provided by Professor Dr. D. Cortes-Martinez (Universidad de Valencia, España).

Biological assays
Microorganism and media
The microorganisms used for the fungistatic evaluation were obtained from the Department of Mycology INEI ANLIS "Dr. Carlos G. Malbrán" Bs.As. Argentina: Candida albicans 00-604 (soft bone), Candida glabrata 00-547 (hisopado), Candida haemulonii 982822,1 (hemocultive), Candida lusitaniae 00-623 (BAL) and Candida parapsilosis 00-629 (cirugy injury) were used. All of them from clinical isolated.

Antifungal assays
The antifungal activity was evaluated using the microdilution technique in media RPMI 1640. These bioassays were performed following references [15 –17].

Results and Discussion
As was previously mentioned, in the putative mechanism of action for berberine to prevent DNA replication, the first step is aromatization. With the aim to finding an experimental support for the above hypothesis, as a first step of our study we obtain canadine (compound [II] in Figure 1) from berberine. Canadine was inactive against all the fungi tested here (Table 1). It should be noted that there are few structural differences between berberine and canadine. The nitrogen atom is not quaternary in canadine, besides one of the rings is not aromatic in this alkaloid (compare structures (I) and (II), Figure 1). At this stage of our study we assume that the electronic effects appear to be more important than conformational factors in the production of the antifungal response. Next we evaluate three compounds structurally related to berberine and canadine. They are compounds III, IV and V. These molecules were kindly provided by Professor Diego Cortes-Martinez (University of Valencia, Spain). It should be noted that these compounds were chosen considering their different molecular flexibility. Compounds IV and V are conformationally restricted due to the presence of a double bond in the connecting chain. In addition these molecules possess a NCOCH₃ moiety. None of these compounds displayed significant antifungal activity (Table 1). These results provide additional support for the hypothesis suggesting that the quaternary nitrogen, the aromatic polycyclic and the planar structure of berberine is the pharmacophoric pattern to produce the antifungal effect of these alkaloids.

Alkaloids possessing a carbonyl group
Evaluation of the antifungal activity of onychine [VI], an azafluorene alkaloid, showed it to have anticandidal (MIC of 3.12 mg/ml) and anticryptococcal (MIC of 3.12 mg/ml) activities [18-19]. Subsequent structure-activity relationship studies revealed that the carbonyl moiety plays a key role in the activity of these compounds. Thus, partial reduction of the carbonyl moiety led to inactive dihydroonycine (VII). Evaluation of the chlorinated isomeric agents (VIII -X), revealed that modification of the 4-methyl group to a chloromethyl moiety enhanced as well as broadened the antifungal profile of azafluorenones [20-22]. In the last two decades some structurally related and biologically active polycyclic aromatic alkaloids were isolated or re-synthesized [23,24]. Several of these were evaluated for antifungal activity. Among these, sampagine [XI] has an EC₅₀ (median effective concentration) of 1.6 m/ml against C. albicans [23] and zones of inhibition of 25-31 nm against C. albicans, C. glabatra and C tropicalis. which are comparable to ketoconazole. Sampagine also showed potent activity against C. Neoformans and A. Fumigatus. Other polyaromatic structures were evaluated for their antifungal activity. Meridine (XII) has been reported to inhibit the growth of C. Albicans, C neoformans by interacting with nulceic acid biosynthesis [25].
Berberine was modified in order to obtain a polyaromatic derivative possessing a carbonyl group, oxyberberine (XIII). Unfortunately oxyberberine was inactive against all the fungi tested (see Table 1).
In parallel we evaluate the oxoaporphines liriodenine (XIV), o-methyl-mostachatoline (XV) and compound XVI, [14,26], all alkaloids possessing a carbonyl group in their structures and structurally related to oxiberberine. These compounds showed marginal or null antifungal activity. Thus, compound XVI only displayed a moderate antifungal effect (MIC=128 µg/ml); whereas compounds XIV and XV were inactive (Table 1).
Berberine isolated from B. Heterophylla displayed an interesting in vitro/in vivo antifungal activity against a panel of pathogenic fungi [6,7] In addition berberine displayed a lower toxic effect in comparison with ketoconazole. Unfortunately all the compounds tested here were inactive or at least displayed lower antifungal activity than those reported for berberine. On the bases of our results it appears that berberine is the best starting structure to look for new cationic alkaloids with double isoquinolinoid skeleton acting as antifungal agents.

Conclusions
Previously we have reported the importance of aromatization within the putative bio-medical action mechanism of berberine and related cationic alkaloids with double iso-quinolinoid skeleton. It appears that the quaternary nitrogen and the aromatic polyciclic and planar structure of berberine could be the pharmacophoric pattern for these alkaloids acting as antifungal agents. All the compounds tested and reported here give an additional support for this hypothesis.

Acknowledgements

This work was supported by grants from National University of San Luis, Argentina, ANPCyT (PICTR -00260). The authors thank Professor Dr. Diego Cortes for samples of some alkaloids. R.D Enriz is a member of CONICET, Argentina.

References

[1] Bottini, C.M. ; Bran, D., Arbustos de la Patagonia. Calafates y Michay In:Presencia, Casamiquela , C. (Ed.), INTA (30), Argentina 1993, pp. 5-9.         [ Links ]

[2] Donoso Zegers, C.; Ramírez Garcia, C. Arbustos Nativos de Chile In: Cúneo V.M. (Ed.), Colección Naturaleza de Chile, Valdivia-Chile, 1994, pp. 27-32.         [ Links ]

[3] Ivanoska, N.; Philipov, S. Int. J. Immunopharmacol. 1996, 18 (10), 553.         [ Links ]

[4] Pamplona, R.J. Berberis vulgaris In Safeliz, S.L. (Ed.), Enciclopedia de las Plantas Medicinales, Madrid (España), 1996, pp. 22-54.         [ Links ]

[5] Freile, M.L. Personal interview on June 30th, 2001 at the National University of Patagonia SJB with Nahuel Pan Chief's wife (Tehuelche community in the Nahuel Pan zone).         [ Links ]

[6] Freile, M.L.; Giannini, F.; Sortino, M.; Zamora, M.; Juarez, A.; Zacchino, S.; Enriz, R. Acta Farm Bonaerense. 2006, 25(1), 83.         [ Links ]

[7] Freile, M.L.; Giannini, F.; Pucci, G.; Sturniolo, A.; Rodero, L.; Pucci, O.; Balzareti, V.; Enriz, R. Fitoterapia. 2003, 74, 702.         [ Links ]

[8] Freile, M.L.; Masman, M.; Suvire, F.; Zacchino, S.; Balzaretti, V.; Enriz, R. J. Mol. Struct.(TEOCHEM) . 2001, 546: 243.         [ Links ]

[9] Freile, M.L. Busqueda de nuevos Compuestos Antifungicos y citotoxicos. Estudio de Correlacion estructura-actividad y Mecanismo de Acción. Tesis Doctoral UNPSJB 2002, pp. 128.         [ Links ]

[10] Pinho, P.M.; Pinto, M.M.M.; Kijjoa, A.; Pharadi, K.; Diaz, J.G.; Herz, W. Phytochemistry. 2002, 31, 1403.         [ Links ]

[11] Gonzalez, M.C.; Zafra-Polo, M.C.; Blázquez, M.A.; Serrano, A.; Cortes, D. J. Nat Prod. 1997, 60 (2) 108.         [ Links ]

[12] Bermejo, A.; Andreu, I.; Suvire, F.; Leonce, S.; Caignard, D.H.; Renard, P.; Pierre, A.; Enriz, R.; Cortes, D. J. Med. Chem. 2002, 45, 5058.         [ Links ]

[13] Andreu, I.; Cortes, D.; Protais, P.; Casesels, B.K.; Chagraoui, A.; Cabedo, N. Bioorg. Med. Chem. 2000, 8, 889.         [ Links ]

[14] Leboeuf, M.; Cortes, D.; Hocquemiller, R.; Cavé, A. Planta Medica. 1983, 48, 185.         [ Links ]

[15] Fromtling, R.F.; Galgiani, J.N.; Faller, M.A.P. Antimicrob Agents Chemoter. 1993; 37, 39.         [ Links ]

[16] Faller, M.A.P.; Rinaldi, M.G.; Galgiani, J.N. Antimicrob Agents Chemoter. 1990, 34, 1648.         [ Links ]

[17] Faller, M.A.P.; Grant, C.; Morthland, V.; Rhine Chalberg, J.J. J. Clin. Microb. 1994; 32(2), 506.         [ Links ]

[18] Koyama, J.; Sugita, T.; Suzuta, Y. Heterocycles 1979, 12, 1017.         [ Links ]

[19] Hufford, C.D.; Liu, S.; Clark, A.M.; Oguntimein, B.O. J. Nat Prod. 1987, 50, 961.         [ Links ]

[20] Bracher, F. Prarmazie 1993, 48, 521.         [ Links ]

[21] Bracher, F. Arch Pharm /Wheinheim, 1994, 327-371        [ Links ]

[22] Koyama, J.; Okatani, T.; Tagahara, K. Heterocycles 1989, 29, 1649.         [ Links ]

[23] Kitahara, Y.; Onikura, H.; Shibano, Y.; Watanabe, S.; Mikami, Y.; Kubo, Y. Tetrahedrom, 1997, 53, 6001.         [ Links ]

[24] Feliu, L.; Ajana, W.; Alvarez, M.; Joule, J.A. Tetrahedron, . 1997, 53, 4511.         [ Links ]

[25] McCarthy, P.; Peterson, T.; Gunawardana, G. 31 st ICAAC, 1991, Abstract Nº 212.         [ Links ]

[26] Cortes, D.; Hocquemiller, H.; Leboeuf, M.; Cavé, A. J. Nat. Prod. 1986 , 49(5), 878.         [ Links ]