Pharmaceutical Biology 1388-0209/99/3703-0208$15.00 1999, Vol. 37, NO. 3, pp. 208-215 O Swets & Zeitlinger ANTIMICROBIALACTIVITY OF NEOTROPICALWOOD AND BARKEXTRACTS Ivelisse Roviral, Amy Berkovl, Ainsley Parkinson1, Scott Mori3 and Barbara Meurer-Grimes1 lDepartment of Biological Sciences, Lehman College and The Graduate School of the City University of New York, 250 Bedford Park Boulevard West, Bronx, New York 10468, U.S.A. 21nstitut Français de Recherche Scientifique pour le Développement en Coopération (ORSTOM), Laboratoire d’Entomologie forestière, Centre de Cayenne, B. P. 165,97323 Cayenne Cedex, France 31nstitute of Systematic Botany, The New York Botanical Garden, Bronx, New York 10458, U.S.A. ABSTRACT INTRODUCTION A total of 203 wood and bark extracts obtaiiiedfiom a The tree species of the lowland Neotropical rainforests Neotropical lowland rainforest in French Guiana were in northem South America constitute a vast natural and tested for antimicrobial activity against a panel offour economic resource of our planet. Waterman and huinaii pathogens. Iïihibitoiy activity against Staphy- lococcus aureus was regularly obsewed. Significant McKey (1983) propose that the bark might actually be growth inhibition against Enterococcus faecilis was the ‘most defended part of a tropical tree’. Medicinal found in species belonging to the plant genus Sloanea uses of various woods and, especially, of barks have (Elaeocaipaceae) and various Sapotaceae. ílie plant been reported from many cultures that live in close con- genera Eschweilera, Gustavia and Couratai (Lecythi- daceae) showed notable growth iiihibitory activity tact with these forests. Grenand et al. (1987) mention against both Escherichia coli aiid Enterococcus fae- hundreds of ethnomedicinal applications of bark and cilis. Fractionation of the active extracts showed that wood decoctions used by the local Creole, Palikur, the activities against Grani-positive and Graïn-nega- Boni, and Wayãpi of French Guiana. Examples include tive bacteria appear to be associated with different the wood of an Oimosia species (Fabaceae) prepared conipounds. In some cases several replicates were as a fever remedy by the Palikur, the bark of the large sampled, showing that inti-aspeciJicactivity levels ïnay falcata vary. The results of the screening experiments are coïn- tree Eperua (Caesalpiniaceae) used in a decoc- pared to ethnoinedicinal infoimation available froni tion as a dental analgesic by the Boni and Creoles, the riortheriz South Anierica and adjacent areas. macerated bark of Gustavia augusta (Lecythidaceae) is used by Creoles to treat small children for vomiting and by the Palikur as a remedy against leishmaniasis, and bark df a species of Micl-opholis (Sapotaceae) is used by the Palikur for its antidiabetic properties (Grenand et al., 1987). Keywords: Elaeocarpaceae, Tiliaceae, Sterculiaceae, Bomba- Recently, wood and bark samples were collected caceae, Lecythidaceae, Sapotaceae, Caesalpiniaceae, from several hundred tree species growing in the low- Fabaceae, Mimosaceae, Staphylococcus aumus, Candida land rainforests of the Sinnamary River Basin in north- albicans, Enterococcus faecilis, Escheï-ichia coli, ethno- em French Guiana and Les Eaux Claires (near Saül) in medicine, French Guiana. Central French Guiana. Botanical, ecological, and phy- tochemical information on these specimens has already been published, because they have been subject to a Address correspondence to: Barbara Meurer-Grimes, AMRAD Discovery Technologies Pty Ltd., 576 Swan Street, massive study on the host plant associations of ceram- Richmond, Victoria 3 121, Australia E-mail: bmeurer- bycid beetles (Tavakilian et al., 1997; Meurer-Grimes -__ __ c_ -_ grimes @ amrad.com.äu & Tavakilian, 1997; Berkov, 1999). We selected over 3 ANTIMICROBIAL ACTlVITY OF WOOD AND BARK EXTRACTS 209 200 of these specimens representing the most abun- Samples with activities against more than one pathogen dantly sampled plant families, both in number of indi- were fractionated into an organic and a water phase. viduals and in number of species, to be tested for their Half of each extract was partitioned between 50 mlof antimicrobial properties against a panel of four human dichloromethaneand 25 mlof water. Both phases were pathogens. Our samples are mostly derived from the evaporated and reconstituted in a small volume of sol- Fabaceae, Caesalpiniaceae, and Mimosaceae (often vent for storage. For agar disk diffusion assays, these collectively referred to as Leguminosae or Fabales), the extracts were adjusted to a final concentration of 500 Lecythidaceae (Brazil nut family), and the Sapotaceae, mg dry weighdm1 solvent or 50 mg/ml (for as well as a few other smaller families. Replicates of dichloromethane-phases). Twenty microliters of each many species were collected at one or both localities extracthtaction were applied to a 6 mm paper disk. allowing for a preliminary assessment of the variability found within one species. The results of the screening Bioassays experiments are discussed in reference to ethnomedic- Growth inhibitory activity was tested against four inal information. microorganisms: Staphylococcus aureus (ATCC 25923), Candida albicans (ATCC 60193), Escherichia coli (ATCC 25922), and Enterococcus MATERIALS AND METHODS faecilis (ATCC 29212) by using the Kirby Bauer agar disk diffusion technique. Staphylococcus aureus and Wood and Bark Samples E. coli were plated onto Mueller-Hinton agar plates, The 203 wood samples investigated in this study were E. faecilis onto Mueller-Hinton agar plates with 5% derived from approximately 102 different plant taxa sheep blood, and C. albicans onto Sabouraud dextrose (the exact number of species could not be determined agar plates. For bioassays, plates were inoculated because of the yet unidentified specimens or unde- from freshly prepared suspensions (Tryptic soy broth) scribed species). The majority of the samples were at 5 CPU that were obtained from freshly isolated, obtained from a Neotropical lowland rainforest in uncontaminated colonies. Extract impregnated 6 mm northern French Guiana in the Sinnamary River Basin paper disks were applied to the inoculated plates, during the field seasons of 1992 and 1993. The wood along with either penicillin or sulconazole (10 samples were cut as 1 cm3 cubes (total of 10 to 15) yg/disk) as a positive control and 80% aq. methanol from felled trees that were simultaneously used for (20 yl/disk) as a negative control. All bioassay plates studies of their faunas of woodboring longicorn beetles were incubated for 24 h at 37°C. Growth inhibition (Cerambycidae) (Tavakilian et al., 1997; Meurer- zones were measured in mm, and compared to inhibi- Grimes & Tavakilian, 1997). Additional collections tion zones of the control disks. from twigs and branches of 25 trees belonging to five species of Lecythidaceae were made at Les Eaux Claires (near Saül, Central French Guiana), during RESULTS AND DISCUSSION 1995 and 1996. These trees are subject of an in-depth study of the longicorns associated with the Brazil nut The screening results of methanolic wood and bark family (Berkov & Tavakilian, in press; Berkov, 1999; extracts (Table 1) are arranged by plant taxon in taxo- Berkov et al., submitted). The samples were stored in nomic order according to Cronquist (1981). In addition, methanol for transportation and then kept frozen until Table 1 contains ethnomedicinal information on the further processing. Voucher specimens of all samples plant species investigated in our study. The information were collected, identified by specialists, and deposited is limited to the use of wood and bark preparations, but in major herbaria (NY,P, CAY). Unidentified taxa (see includes reports from regions other than French Table 1) are currently under revision (Poncy, pers. Guiana, mostly from adjacent countries in northern comm.) or being processed. South America or Central America where the same or closely related species are known to occur. Extraction and Fractionation Fractionation of the active extracts indicated The wood cubes, including the remaining methanol, whether the multiple activities observed in the crude were homogenized in a blender and extracted twice in extracts could be due to the presence of more than one methanol. The combined filtered extracts were evapo- bioactive compound. These results are summarized in rated to dryness and reconstituted in 10 mlof methanol. Table 2, 210 I.ROVIRA ET AL. Table 1. Antimicrobial activity of neotropical wood and bark extracts. Plant taxon1 Collection S. aurezrs E. faecilis E. coli C. albicans Ethnomedicinal numbes information Elaeocqaceae Sloanea a# latifoliiim L 1692 +6 - febrifuge (Grenand et al., 1987: 224; Schultes & Rauffauf, 1994) Sloanea sp. L1700 + - L1708 ++ - L1883 + - L1886 - - Sloanea sp. 1 M23601 + - Sloanea sp. 2 M23564 ++ - Sloanea sp. 3 M23527 + - Sloanea sp. 5 M23687 + - M23690 i- - Tiliaceae Lueheopsis rugosa L1878 (+) Sterculiaceae Sterculiafrondosa L1762 +f S. pnrriens M23692 + bronchial infections (Grenand et al., 1987: 414) Theobroma subincaniini LI821 + Bombacaeae Catostemmafragrans L1680 + L1794 + M23417 ++ Pachira insignis L1824 + Lecythidaceae Corythophora amapaensis M241452,3 +++ M241 162,3 +++ M24147223 ++ M24148223 +++ M241742,4 + M24174233 i-++ C. riìnosa L1704 (wet) + L1704 (dry) + Couratari stellata M240922,3 + M24093233 +f M240942*3 + M240952,3 + M2411l2l3 ++ Eschweilera alata L1741 + M23537 + E. apiciilata M2345 1 + E. collina M23483 + M23597 + E. congestijlora M23423 (+) M23478 + M23631