Food Chemistry 135 (2012) 2086–2094 Contents lists available at SciVerse ScienceDirect Food Chemistry journal homepage: www.elsevier .com/locate / foodchem Chemical and antifungal investigations of six Lippia species (Verbenaceae) from Brazil Cristiano Soleo Funari a, Fernanda Patrícia Gullo b, Assunta Napolitano d, Renato Lajarim Carneiro c, Maria José Soares Mendes-Giannini b, Ana Marisa Fusco-Almeida b, Sonia Piacente d, Cosimo Pizza d, Dulce Helena Siqueira Silva a,⇑ a NuBBE, Institute of Chemistry, São Paulo State University, Araraquara, CP 355, CEP 14801-970, SP, Brazil b Faculty of Pharmaceutical Sciences, São Paulo State University, Rodovia Araraquara-Jaú, km 1, 14801-902 Araraquara, SP, Brazil c Federal University of São Carlos, Department of Chemistry, 13565-905 São Carlos, SP, Brazil d Dipartimento di Scienze Farmaceutiche, Università degli Studi di Salerno, via Ponte Don Melillo, 84084 Fisciano, SA, Italy a r t i c l e i n f o a b s t r a c t Article history: Received 5 January 2012 Received in revised form 23 May 2012 Accepted 22 June 2012 Available online 29 June 2012 Keywords: Verbenaceae Lippia spp Dereplication Antifungal Candida spp. Cryptococcus neoformans 0308-8146/$ - see front matter � 2012 Elsevier Ltd. A http://dx.doi.org/10.1016/j.foodchem.2012.06.077 ⇑ Corresponding author. Tel.: +55 19 33016659; fax E-mail address: dhsilva@iq.unesp.br (D.H.S. Silva). The Lippia genus is used in ethnobotany as food, beverages, seasoning and antiseptic remedies, among others. The chemical compositions of fifteen extracts of six Lippia species were investigated comparatively by HPLC–PDA. To avoid data replication of previous works on this genus,Lippia lupulina Cham. root ethanol extract was selected for isolation procedures based on Principal Component Analyses (PCA) of such data. Seven compounds previously unreported in this genus were isolated from this extract (a triterpene, two furanonaphtoquinones, a furanochromone, an isoflavone, a stilbene and an iridoid). The activities of extracts, fractions and pure compounds towards Candida albicans, Candida krusei, Candida parapsilosis and Cryptococcus neoformans were investigated. Two fractions from the extract of Lippia salviaefolia leaves showed marked inhibition of fungal growth, in addition to verbascoside and asebogenin, which showed MICs lower than 15.6 lg/ml and may be promising leads for the development of new antifungal agents, especially against C. neoformans. � 2012 Elsevier Ltd. All rights reserved. 1. Introduction The Lippia genus comprises approximately 200 species distrib- uted throughout the Central and South Americas as well as in trop- ical Africa, and it is estimated that Brazil hosts 70–75% of the known species (Arthur, Joubert, De Beer, Malherbe, & Witthuhn, 2011). It is used in ethnobotany worldwide as food, beverages, seasoning and remedies (Pascual, Slowing, Carretero, Sánchez, & Villar, 2001). Some Lippia species have antiseptic and healing properties, among other uses (Lorenzi & Matos, 2002; Pascual et al., 2001). Infusions of leaves and flowers of Lippia lupulina Cham. from the Cerrado biome have been employed by local people from Minas Gerais State (Southeastern Brazil) to threat mouth and throat infections (Rodrigues & Carvalho, 2001), but only one report of the chemistry of its essential oil is available in the literature (Zoghbi, Andrade, Silva, & Maia, 2002). Leaf and stem ethanol extracts ofLippia salviae- folia Cham. contain flavonoids and phenylpropanoids, including aromadendrin (12) and phloretin (13). The formal counteracted oxidative stress in human embryonic kidney HEK-293 cells and the latter inhibited human melanoma M14 cancer cell growth ll rights reserved. : +55 19 33016692. and induced concentration dependent apoptosis (Funari et al., 2011). Lippia sidoides Cham., also investigated in the present work, has been widely used in Northeastern Brazil as a general use antiseptic (Lemos et al., 2007). Dried and milled leaves, flowers and fruits of this plant have been used as a substitute for Thymus vulgaris in spice mixtures for pizzas and meats (Lorenzi & Matos, 2002). Recently, this species was included in the Brazilian Health Ministry priority list of 71 species for phytoceutical product development (Ministério da Saúde & DAF/SCTIE/MS, 2009) due to its reported antiseptic properties. Immunocompromised patients, such as HIV-infected individu- als, transplant recipients and cancer patients, are especially vulner- able and die mainly due to opportunistic invasive fungal infections (IFIs) (Chandrasekar, 2010). The most common causative agents of these infections are Candida spp., Aspergillus species, and Cryptococ- cus neoformans (Kriengkauykiat, Ito, & Dadwal, 2011). Amphotericin B and fluconazole are the drugs of choice for treatment of cryptococcosis. However, some recent isolates have shown resistance to fluconazole. In addition, amphotericin B has high toxicity and therefore its use should be limited (Mdod et al., 2011). Polyenes, azoles and echinocandins are now the main clas- ses of antifungal drugs available to control these infections, but http://dx.doi.org/10.1016/j.foodchem.2012.06.077 mailto:dhsilva@iq.unesp.br http://dx.doi.org/10.1016/j.foodchem.2012.06.077 http://www.sciencedirect.com/science/journal/03088146 http://www.elsevier.com/locate/foodchem C.S. Funari et al. / Food Chemistry 135 (2012) 2086–2094 2087 with the changing spectrum of pathogens and their increasing resistance to these antifungal agents, together with possible side effects produced by current therapies, the development of new antifungal scaffolds is critical (Chandrasekar, 2010). The important role of natural products in the development of new antimicrobial agents (antibacterial, antifungal, antiviral and antiparasitic), entities or therapies is well documented (Newman & Cragg, 2007). In this study, the chemical composition of fifteen ethanol extracts of six different Lippia species from Brazil were compared by high performance liquid chromatography coupled to a photodiode array detector (HPLC–PDA). To avoid replication procedures, the root extract of L. lupulina Cham. (EERLlup) was se- lected for fractionation and isolation. Isolated compounds were subsequently tested against opportunistic human pathogenic fun- gal strains (e.g., Candidaparapsilosis, Candida krusei, Candida albicans and C. neoformans). In addition, extracts, fractions and compounds previously extracted from leaves and stems of L. salviaefolia Cham., which presented chromatographic profiles and chemical composi- tions similar to L. sidoides Cham., were also investigated for their antifungal properties. 2. Experimental 2.1. General information One- and two-dimensional Nuclear Magnetic Resonance (NMR) experiments were performed on a Bruker DRX-600 spectrometer at 14.1 T or on a Varian INOVA 500 and Bruker DRX-500 spectrome- ters at 11.7 T. Isolated compounds were analyzed by electrospray ionization ion trap mass spectrometry (ESI-ITMSn) using a Thermo- Finnigan Spectra System HPLC coupled to an LCQ Deca ion trap mass spectrometer (ThermoFinnigan, San Jose, CA, USA). The mass spectra were acquired in negative and positive modes. The ESI source parameters were a capillary voltage of �4.0 V, a spray volt- age of 5 kV and a tube lens offset of 20 V in negative mode. In positive mode, the spectra were acquired with a capillary voltage of 23 V, a spray voltage of 5 kV and a tube lens offset of 50 V. In both ionization modes, the capillary temperature was 280 �C. Data were acquired in MS scanning mode. Semi-preparative HPLC–UV analyses were carried out on an Agilent 1100 series instrument equipped with a G-1312 binary pump, a G-1328A rheodyne injector and a G-1365B multiple wave- length detector. Preparative HPLC–UV analyses were carried out using a Varian Prep-Star 400 system. Column chromatography (CC) separations were performed over silica gel (0.035–0.070 mm, Acros Organics, USA) or Sephadex LH-20 (Pharmacia Biotech, Sweden). Analytical TLC analyses were performed on silica gel w/ UV plates (Sigma–Aldrich, USA). Spots on TLC plates were visual- ized under UV light and by spraying anisaldehyde–H2SO4 reagent followed by heating at 120 �C (Wagner & Bladt, 1996). Specific rotation measurements were performed at k 289 nm in a Jasco P-1020 polarimeter (Japan) with a cylindrical glass cell (10 mm I.D. � 10 mm, model CG1-10, Jasco, Japan). 2.2. Plant material Aerial parts of L. salviaefolia Cham. and Lippia velutina were col- lected in Mogi-Guaçu (São Paulo-Brazil) in 2006 (voucher specimens n� Lima 90 and n� Brumati TI73, respectively) and identified by Dr. Maria Inês Cordeiro of the Herbarium Maria Eneida P. Kaufmann – Instituto Botânico de São Paulo, São Paulo, Brazil. Aerial parts of Lip- pia balansae Briq. and Lippia lasiocalycina Cham. were collected in Santa Cruz do Rio Pardo and Pratânia (São Paulo), respectively, in 2008 (voucher specimens n� FEA 402 and n� FEA 3556, respectively) and identified by Dr. Giselda Durigan of the Herbarium Coleção Botânica da Floresta Estadual de Assis, São Paulo, Brazil. Aerial parts and roots of L. lupulina Cham. and L. sidoides Cham. were collected in Iaras state (São Paulo) in 2009 (voucher specimens n� FEA 3638 and n� FEA 3639, respectively) and also identified by Dr. Giselda Durigan. 2.3. Extraction Plant materials were dried in an oven with air circulation at 45 �C and then ground in a knife mill. Each extraction was per- formed in ethanol at a ratio of 7:2 (v/w) in three steps, each 24 h long at room temperature. The solutions were concentrated at 40 �C to give extracts of the flowers (EEFLb), leaves (EELLb) and stems (EESLb) from L. balansae Briq.; leaves and stems combined (EELSLlas) from L. lasiocalycina Cham.; leaves (EELLsid), stems (EESLsid) and roots (EERLsid) from L. sidoides Cham.; flowers (EEFL- lup), leaves (EELLlup), roots (EERLlup) and stems (EESLlup) from L. lupulina Cham.; and leaves (EELLv) and stems (EESLv) from L. velu- tina. In addition, L. salviaefolia Cham. leaves and stems were previ- ously extracted and concentrated following the same procedure described above to obtain EELLsal and EESLsal, respectively (Funari et al., 2011). 2.4. Dereplication by HPLC–PDA Each extract (10 mg) was dissolved in MeOH (1 ml) and filtered through a PTFE membrane (0.20 lm, Sartorius AG, Germany). These solutions (15 ll) were analyzed in a Shimadzu HPLC equipped with a degasser (DGU-20A3), two pumps (LC 20AT), an auto-sampler (SIL- 20A), a photodiode array detector (SPD-M20A) and an oven (CTO- 20A). Separation was achieved on two coupled C18 columns (Phenomenex Onyx Monolithic, 100 � 4.6 mm coupled to a Phe- nomenex Synergi Hydro-RP, 250 � 4.6 mm, 4 lm particle size) using H2O (solvent A) and MeOH (solvent B). The elution was carried out at 1 ml/min using the following gradient: 30–50% B (0–25 min) and 50–100% B (25–70 min). Detection was achieved at 254 nm, and the compounds were identified by comparison of their retention times and UV spectra with the following reference compounds isolated previously from Lippia salviaefolia Cham.: (2S)- and (2R)- 30,40,5,6-tetrahydroxyflavanone-7-O-b-glucopyranoside (1a/1b), (2S)- and (2R)-30,40,5,8-tetrahydroxyflavanone-7-O-b-glucopyrano- side (2a/2b), (2S)- and (2R)-eriodictyol 7-O-b-D-glucopyranoside (3a/3b), forsythoside B (4), 6-hydroxyluteolin-7-O-b-glucoside (5), verbascoside (6), aromadendrin (7), naringenin (8), phloretin (9), asebogenin (10) and sakuranetin (11) (Funari et al., 2011). 2.5. Multivariate curve resolution and principal component analyses Principal component analyses (PCA) were not performed di- rectly on chromatograms but instead using the individual contri- bution of each peak/substance found by Multivariate Curve Resolution–Alternating Least Squares (MCR–ALS) in four regions of the chromatograms. Each HPLC–PAD analysis gave rise to a Xtxw matrix with t retention times and w wavenumbers; thus, each retention time dataset was a UV–vis spectrum (or a sum of spectra) for a single substance. MCR–ALS decomposes this matrix into two matrices, so that: Xtxw ¼ C�txnSt wxn þ Etxw where C is a matrix that contains the relative concentration of n substances at the t retention times, S is a matrix containing the pure spectrum of the n substances and E is the matrix of errors or lack of fit. If the relative concentration of a substance in C is summed from all t retention times, a relative concentration or area for this sub- stance is obtained from that chromatogram. If matrix X is built using more than one chromatogram, MCR–ALS can avoid problems 2088 C.S. Funari et al. / Food Chemistry 135 (2012) 2086–2094 from variation in the retention times because it uses both retention time and UV–vis spectrum data to identify each substance in all of the chromatograms. Thus, MCR–ALS finds the contribution of every substance presenting the same UV spectrum (over a short time interval where peak shift can occur) in each sample. Thus, PCA is performed on a matrix which contains the samples (or chromato- grams) in the rows and the relative concentrations of each substance in each sample, in the columns. All chemometric proce- dures were performed in a Matlab 2011a (Mathworks, Inc., Natick, MA, USA) environment. MCR–ALS was downloaded from http:// www.mcrals.info/. PCA was accomplished using routines developed in our laboratory. The shifting of chromatographic bands was cir- cumvented using the area of the peaks, taking each peak area as one variable for each sample. 2.6. Isolation procedures The crude ethanol extract of L. lupulina roots (EERLLup, 9.5 g) was dissolved in MeOH–H2O 8:2 (v/v) (330 ml) and extracted with Hexane (4 � 150 ml) to give fraction FHex1 (1.6 g) and a precipi- tate (6.9 g). H2O was added to the hydromethanolic phase up to 45:55 (v/v) and it was extracted with EtOAc (3 � 250 ml) to give fraction FAc1 (0.6 g). The hydromethanolic phase was then concen- trated, diluted in H2O (500 ml) and extracted with n-BuOH (3 � 170 ml) to give fractions FBu1 (0.4 g) and FAq1 (1.2 g). All fractions were then concentrated at 40 �C. FHex1 (0.6 g) was sub- mitted to medium pressure liquid chromatography (MPLC) over silica gel (101.5 g; 39.5 � 2.6 cm i.d.) and eluted with Hex–CHCl3 in a linear gradient (20–100% B in 170 min., 17 ml/min) to afford compounds 12 (335.3 mg) and 13 (8.0 mg). FAc1 (0.6 g) was sub- mitted to liquid chromatography at atmospheric pressure (LC) over silica gel (25.0 g; 8.0 � 2.6 cm) and eluted with Hex–EtOAc 1:0, 1:1 and 0:1 (v/v) to give subfractions B1–B3 (75 ml each). Purification of B2 (20.0 mg, 20 injections) by HPLC–UV was performed on a C18 column (Waters Symmetry, 300 � 7.8 mm) with H2O–MeOH (1:1 v/v), at 3 ml/min and UV detection at 220 and 254 nm to yield compounds 14 (3.4 mg, tR = 13.4 min.), 15 (0.3 mg, tR = 16.9 min.) and 16 (5.0 mg, tR = 17.5 min.). Fraction FBu1 (1.0 g) was chro- matographed by size exclusion chromatography (SEC) on a Sepha- dex LH 20 column (180.0 � 3.0 cm) eluted with MeOH (3.0 l) to give compounds 17 (18.5 mg), 18 (13.5 mg) and subfraction C1. The latter (40 mg, 8 injections) was purified by HPLC–UV on a semi-preparative C18 column (Phenomenex Synergi Hydro-RP, 250 � 21.2 mm) with H2O–MeOH (7:3 v/v) at 12.0 ml/min and UV detection at 230 nm to afford 19 (5.0 mg, tR = 35.1 min.). EELL- sal and EESLsal were previously submitted to liquid–liquid extrac- tion to give FHex2, FAc2, FBu2 and FAq2, and FHex3, FAc3, FBu3 and FAq3, respectively (Funari et al., 2011). 2.7. Antimicrobial susceptibility testing This study evaluated the antifungal activity of crude extracts and fractions of Lippia against the pathogenic yeasts Candida par- apsilosis (ATCC 22019), Candida albicans (ATCC 90028), Candida krusei (ATCC 6258) and Cryptococcus neoformans (90012) from the mycology collection of the Clinical Mycology Laboratory, Department of Clinical Analyses, School of Pharmaceutical Sci- ences, UNESP, Araraquara. The minimum inhibitory concentrations (MICs) were determined according to the microdilution method described by Rodriguez-Tudela, Barchiesi and the Subcommittee on Antifungal Susceptibility Testing (2008) using 96-well plates with serial dilutions of stock solutions of natural compounds in DMSO into RPMI-1640 culture medium. The concentrations tested ranged from 250 to 0.48 lg/ml (Scorzoni et al., 2007). Inoculum was prepared in RPMI-1640 without sodium bicarbon- ate supplemented with L-glutamine and 2% glucose and buffered with 0.165 M MOPS at pH 7.0. Yeast suspensions were prepared to a final concentration of 1.0 � 104 CFU/ml in RPMI-1640 and 100 ll was added to each well. The plates were incubated in a shaker at 37 �C and 150 rpm for 24 (Candida species) or 48 h (C. neoformans). Amphotericin B and fluconazole were used as positive controls. MICs were read at 490 nm using a plate reader after visualization with Alamar Blue. The interpretation of the results was performed according to Scorzoni et al. (2007). MICs below 75 lg/ml were re- garded as strong antifungal activity, whereas MICs between 75 and 150 lg/ml indicated moderate activity and MICs from 150 to 250 lg/ml indicated low activity. No antifungal activity was associ- ated to MICs greater than 250 lg/ml. 3. Results and discussion 3.1. Dereplication by HPLC–PDA In our continuous effort to investigate species of the genus Lip- pia native to Brazil, 15 ethanol extracts of six different Lippia spe- cies were analyzed by HPLC–PDA. The method was developed for the ethanol extracts of L. salviaefolia Cham. leaves (EELLsal) and stems (EESLsal) because the compounds 1–11 (Fig. 1) were previ- ously isolated from them (Funari et al., 2011). After following the usual steps for HPLC method development employing a C18 sil- ica-based packed column (Snyder, Kirkland, & Glajch, 1997), a C18 monolithic column was coupled in series prior to it, improving separation without exceeding the backpressure allowed by our HPLC–PDA system (250 bar). Analyses of all extracts and reference compounds were performed using this method and the presence of 1–11 was determined in each extract based on retention times and UV spectra. The results are summarized in Table 1. Ethanol extracts from L. balansae (EELLb), L. velutina (EELLv) and L. sidoides (EELLsid) leaves and from L. balansae (EEFLb) flowers exhibited similar chromatographic profiles to the observed for the extract from L. salviaefolia leaves (EELLsal). The unusual inter- converting flavanone glucosides 1a/1b and 2a/2b were detected in all of them. Their partial interconversions via a common chalcone intermediate were previously proposed (Funari et al., 2011). This intermediate could be achieved by means of an acid-catalyzed keto-enolic tautomerization with C ring opening and C2–C3 double bond formation. The subsequent Michael-type nucleophilic attack of the 60-hydroxyl or the 20-hydroxyl on the a,b-unsaturated ke- tone might lead to compounds 1a/1b or 2a/2b, respectively (Supplementary data). Flavanone glucosides 3a/3b, flavone gluco- side 5, flavanones 8 and 11 and dihydrochalcones 9 and 10 were detected in these extracts (Table 1). Different chromatographic profiles were observed for the ethanol extracts of L. lasiocalycina leaves and stems combined (EELSLlas) as well as from L. lupulina leaves and flowers (EELLlup and EEFLlup, respectively), in which no flavonoids were detected. Phenylpropanoids forsythoside B (4) and verbascoside (6) were detected in the former two extracts, while only 6 was detected in the latter. Regarding the chromatographic profiles of extracts from stems, L. balansae (EESLb) and L. velutina (EESLv) were similar to L. salviae- folia (EESLsal), while L. sidoides extract (EESLsid) had only partial similarity with the latter. Dissimilar profiles were observed for the extracts from stems of L. lupulina (EESLlup) and from leaves plus stems of L. lasiocalycina (EELSLlas). No flavonoids (reference com- pounds) were detected in EESLlup and EELSLlas, but phenylpropa- noids 4 and 6 were found as major compounds in these extracts. In addition, the two extracts of roots investigated in this study, L. sidoides (EERLsid) and L. lupulina (EERLlup), showed different chro- matographic profiles, but phenetyl glucosides 4 and 6 were de- tected in both (Table 1). Three chromatograms representing very http://www.mcrals.info/ http://www.mcrals.info/ Fig. 1. Compounds employed as chemical markers during dereplication studies by HPLC–PDA (1–11) and compounds isolated from L. lupulina Cham. root extract (EERLlup) (12–19). Glucopyranosyl, rhamnopyranosyl, apiofuranosyl and caffeoyl are indicated as Glc-, Rha-, Api- and Caf-, respectively. C.S. Funari et al. / Food Chemistry 135 (2012) 2086–2094 2089 dissimilar HPLC–PDA profiles relative to the reference extract EELLsal are shown in Fig. 2. To improve our dereplication and to support the selection of a new extract for further study, a cluster analysis was performed on HPLC–PDA data acquired from the fifteen extracts under analy- sis. The shifting of chromatographic peaks is the main problem when PCA is performed on a chromatographic dataset. This prob- lem comes from the lack of bilinearity among the chromatographic profiles of each sample (Carneiro, Braga, Bottoli, & Poppi, 2007). As a result, the model identifies more than one principal component for one chromatographic profile, leading to a misinterpretation of the cluster formation. This problem can be solved by peak align- ment or, as employed in this work, by using each peak area as one variable for each sample. MCR–ALS is a deconvolution method employed to find chromatographic peaks with identical UV spectra among all the chromatograms (on small intervals of the chromato- grams where peak shifting might occur). This method then pro- vides a relative area for identical peaks and was performed on the chromatograms of the extracts separated in four small regions to avoid interpretation mistakes. These mistakes could occur with any compounds that have similar UV spectra but large differences in retention time. This procedure was helpful in determining the area of peaks for the same compound but with different retention times across the samples. PCA was then performed on the relative concentrations of the compounds found by MCR–ALS in the extracts and combined to give a general overview of the cluster separations (Fig. 3). As shown in Fig. 3, EELLb, EELLsid and EELLv composed a clear cluster with the reference extract EELLsal (C, F, G and E in Fig. 3, respectively), corroborating our visual analyses. EEFLb was very similar to the reference extract EELLsal (A and E in Fig. 3, respec- tively), with samples EESLsal and EESLb (M and K in Fig. 3, respec- tively) having less similarity. Despite the qualitative similarities among the chromatographic profiles of EESLb, EESLsal and EESLv (K, M and O in Fig. 3, respectively), the latter was far from the first two in PCA. This observation might be explained by the relative intensities among peaks containing variations from one sample to another. In this case, the PCA again indicated dissimilarities. It should be noted that EERLlup (I in Fig. 3) was out of any cluster, corroborating the visual observations described above. In addition to showing a dissimilar profile when compared with the previously studied extracts EELLsal and EESLsal (E and M in Fig. 3), EERLlup pre- sented intense peaks at retention times (tR) greater than 35 min., which could not be identified from the reference compounds Ta bl e 1 O cc ur re nc e of re fe re nc e co m po un ds in fi ft ee n ex tr ac ts of si x Li pp ia sp p. C om po u n d t R (m in ) k m ax (n m ) L. sa lv ia ef ol ia L. ba la ns ae L. ve lu ti na L si do id es L. la si oc al yc in a L. lu pu lin a Le av es St em s Le av es St em s Fl ow er s Le av es St em s Le av es St em s R oo ts Le av es + st em s Le av es St em s Fl ow er s R oo ts (2 S) - an d (2 R) -3 0 ,4 0 ,5 ,6 -T et ra h yd ro xy fl av an on e- 7- O -b - gl u co py ra n os id e (1 a/ 1b ) 13 .7 28 6; 32 9 + + + + + + + + (2 S) - an d (2 R) -3 0 ,4 0 ,5 ,8 -T et ra h yd ro xy fl av an on e- 7- O -b - gl u co py ra n os id e (2 a/ 2b ) 15 .8 28 6; 32 9 + + + + + + + (2 S) - an d (2 R) -E ri od ic ty ol 7- O -b -D -g lu co py ra n os id e (3 a/ 3b ) 19 .4 28 3; 32 8 + + + + + Fo rs yt h os id e B (4 ) 19 .6 29 1; 33 2 + + + + + + + + + 6- H yd ro xy lu te ol in -7 -O -b -g lu co si de (5 ) 21 .1 28 1; 34 6 + + + + + V er ba sc os id e (6 ) 21 .1 28 9; 33 2 + + + + + + + + + + A ro m ad en dr in (7 ) 21 .5 29 6; 33 0 + + + + + + N ar in ge n in (8 ) 38 .7 28 9; 32 6 + + + + + + + + + Ph lo re ti n (9 ) 39 28 6; 32 6 + + + + + + + + A se bo ge n in (1 0) 48 .7 28 5; 33 5 + + + + + + + Sa ku ra n et in (1 1) 50 28 7; 33 5 + + + + + + + + + 2090 C.S. Funari et al. / Food Chemistry 135 (2012) 2086–2094 available (Fig. 2). These results led to the selection of this extract for fractionation and isolation procedures to identify unreported non- volatile compounds in Lippia genus. 3.2. Isolation and characterization of the root extract of L. lupulina The ethanol extract of roots of L. lupulina (EERLlup), a species native to the Cerrado biome (São Paulo State, Brazil), was selected for phytochemical investigation (see Subsection 3.1). It exhibited a variety of secondary metabolite classes (Fig. 1). Its partition hex- ane fraction (FHex1) was chromatographed using MPLC and gave the triterpenes oleanonic (12) and oleanolic acids (13) (Mahato & Kundo, 1994). From the ethyl acetate fraction (FAc1), the fur- anonaphtoquinones stenocarpoquinone (14) (Schmeda-Hirschmann & Papastergiou, 2003) and avicequinone E (15) (Williams et al., 2006) were isolated by LC and HPLC–DAD, in addition to the furanochromone prim-O-glycosylcimifugin (16) (Sasaki, Taguchi, Endo, & Yosioka, 1982). The n-butanol fraction (FBu1) was submitted to SEC to afford the isoflavone triglycoside biochanin A (7-O-b-D-apiofuranosyl-(1 ? 5)-b-D-apiofuranosyl-(1 ? 6)-b-D- glucopyranoside (17) (da Silva, Velozo, & Parente, 2000), the stil- bene glycoside piceid (18) (Lu, Berthod, Hu, Ma, & Pan, 2009), and a subfraction which was further purified by HPLC–DAD to afford the iridoid glycoside b-dihydrohastatoside (19) (Teborg & Junior, 1991). Their identification was carried out by 1D and 2D NMR and ESI-MS experiments and compared with the literature. Compound 12 was isolated as a white solid. ½a�25 D + 72.7 (c 0.49, MeOH). ESI-ITMS m/z 455 [M+H]+ (calc. for C30H46O3 + H). 13C NMR spectral data (11.7 T, CDCl3, TMS, d ppm) d: 39.0 (C-1), 34.0 (C-2), 218.6 (C-3), 47.3 (C-4), 55.1 (C-5), 19.5 (C-6), 32.1 (C-7), 39.1 (C-8), 45.8 (C-9), 36.7 (C-10), 23.4 (C-11), 122.0 (C-12), 143.7 (C-13), 41.7 (C-14), 27.5 (C-15), 22.8 (C-16), 46.4 (C-17), 41.0 (C-18), 45.8 (C-19), 30.5 (C-20), 33.7 (C-21), 32.3 (C-22), 26.3 (C-23), 21.3 (C-24), 14.9 (C-25), 16.7 (C-26), 25.7 (C-27), 182.3 (C-28), 32.9 (C-29) and 23.4 (C-30). Compound 13 was isolated as a white solid. ESI-ITMS m/z 495 [M+K]+ (calc. for C30H48O3 + K). 13C NMR spectral data (11.7 T, CDCl3, TMS, d ppm) d: 38.4 (C-1), 27.2 (C-2), 79.0 (C-3), 38.7 (C-4), 55.2 (C-5), 18.3 (C-6), 32.6 (C-7), 39.3 (C-8), 47.6 (C-9), 37.1 (C-10), 22.9 (C-11), 122.6 (C-12), 143.6 (C-13), 41.6 (C-14), 27.7 (C-15), 23.4 (C-16), 46.5 (C-17), 41.0 (C-18), 45.9 (C-19), 30.7 (C-20), 33.8 (C-21), 32.4 (C-22), 28.1 (C-23), 15.5 (C-24), 15.3 (C-25), 17.1 (C-26), 25.9 (C-27), 183.0 (C-28), 33.0 (C-29) and 23.6 (C-30). Compound 14 was isolated as a yellow solid. ½a�25 D + 42.3 (c 0.17, MeOH). On line UV spectrum: kmax at 248, 254, 292 and 345 nm. ESI-ITMS m/z 281 [M+Na]+ (calc. for C15H14O4 + Na). 1H NMR spectral data (11.7 T, CD3OD, TMS, d ppm) d: 4.90 (dd, J = 9.2 and 10.5 Hz, H-2), 3.16 (dd, J = 10.5 and 16.9 Hz, H-3), 3.20 (dd, J = 9.2 and 16.9 Hz, H-3), 7.06 (dd, J = 1.5 and 7.5 Hz, H-5), 7.81 (ddd, J = 1.5, 7.5 and 8.0 Hz, H-6), 7.77 (ddd, J = 1.5, 7.5 and 8.0 Hz, H-7), 8.09 (dd, J = 1.5 and 7.5, H-8), 1.28 (s, H-11), 1.38 (s, H-12); 13C NMR spectral data extracted from HMBC and HSQC experiments (11.7 T, CD3OD, TMS, d ppm) d: 93.1 (C-2), 28.7 (C-3), 125.4 (C-3a), 183.3 (C-4), 134.1 (C-4a), 126.4 (C-5), 134.8 (C-6), 133.7 (C-7), 126.5 (C-8), 132.8 (C-8a), 178.6 (C-9), 161.7 (C-9a), 72.0 (C-10), 25.3 (C-11) and 25.3 (C-12). Compound 15 was isolated as a yellow solid. ESI-ITMS m/z 275 [M+H]+ (calc. for C15O5H14 + H). 1H NMR spectral data (14.1 T, CD3OD, TMS, d ppm) d: 4.91 (m, H-2), 3.17 (m, H-3), 7.60 (d, J = 7.7 Hz, H-5), 7.70 (dd, J = 7.7 and 8.0 Hz, H-6), 7.25 (d, J = 8.0 Hz, H-7), 1.29 (s, H-11), 1.40 (s, H-12); 13C NMR spectral data extracted from HMBC and HSQC experiments (14.1 T, CD3OD, TMS, d ppm) d: 93.1 (C-2), 28.6 (C-3), 126.2 (C-3a), 182.6 (C-4), 134.6 (C-4a), 118.9 (C-5), 137.6 (C-6), 124.5 (C-7), 162.7 (C-8), , ( ) 0 10 20 30 40 50 min EELSLlas 4 6 10 8 9 1a/1b 2a/2b EELLsal 5 7 11 3a/3b 0 500 250 4300 6 0 150 EELLlup 4 6 EERLlup 0 50 100 0 500 250 mAU Fig. 2. Selected representative HPLC–PDA chromatograms of the ethanol extracts investigated: leaves of L. salviaefolia Cham. (EELLsal) and L. lupulina Cham. (EELLlup), roots of L. lupulina Cham. (EERLlup) and leaves plus stems of L. lasiocalycina Cham (EELSLas). Identified peaks: (2S)- and (2R)-30 ,40 ,5,6-tetrahydroxyflavanone-7-O-b- glucopyranoside (1a/b), (2S)- and (2R)-30 ,40 ,5,8-tetrahydroxyflavanone-7-O-b-glucopyranoside (2a/b), (2S)- and (2R)-eriodictyol 7-O-b-D-glucopyranoside (3a/b), forsytho- side B (4), 6-hydroxyluteolin-7-O-b-glucoside (5), verbascoside (6), aromadendrin (7), naringenin (8), phloretin (9), asebogenin (10) and sakuranetin (11). C.S. Funari et al. / Food Chemistry 135 (2012) 2086–2094 2091 115.8 (C-8a), 162.0 (C-9a), 72.0 (C-10), 25.0 (C-11) and 25.0 (C-12). Compound 16 was isolated as a redish solid: ESI-ITMS m/z 491 [M+Na]+ (calc. for C22O11H28 + Na). 1H NMR spectral data (11.7 T, CD3OD, TMS, d ppm) d: 6.41 (s, H-3), 6.67 (s, H-8), 4.64 (d, J = 14.7 Hz, 2-CH2-), 4.79 (d, J = 14.7 Hz, 2-CH2-), 4.79 (m, H-20), 3.37 (m, H-30), 1.27 (s, 40-CH3), 1.33 (s, -40-CH3), 4.45 (d, J = 7.5 Hz, H-100), 3.27–3.40 (m, H-100, H-300, H-400 and H-500), 3.67 (m, H-600) and 3.90 (m, H-600); 13C NMR spectral data extracted from HMBC and HSQC experiments (11.7 T, CD3OD, TMS, d ppm) d: 164.6 (C-2), 110.5 (C-3), 176.5 (C-4), 112.1 (C-4a), 156.7 (C-5), 118.2 (C-6), 166.6 (C-7), 94.1 (C-8), 160.8 (C-8a), 66.9 (2-CH2-), 92.2 (C-20), 28.6 (C-30), 71.9 (C-40), 25.3 (40-CH3), 103.9 (C-100), 74.6 (C-200), 77.7 (C-300), 71.3 (C-400), 78.0 (C-500) and 62.5 (C-600). Compound 17 was isolated as a brown solid.½a�25 D � 76.0 (c 0.1, MeOH). Online UV spectrum: kmax at 259 and 324 nm. ESI-ITMS m/ z 733 [M+Na]+ (calc. for C32O18H38 + Na). 1H NMR spectral data (11.7 T, CD3OD, TMS, d ppm) d: 8.20 (s, H-2), 6.56 (d, J = 2.3 Hz, H-6), 6.73 (d, J = 2.3 Hz, H-8), 7.52 (d, J = 8.8 Hz, H-20), 7.01 (d, J = 8.8 Hz, H-30), 7.01 (d, J = 8.8 Hz, H-50), 7.52 (d, J = 8.8 Hz, H-60), 3.85 (s, 40-OCH3), 5.02 (d, J = 7.0 Hz, H-100), 3.49–3.54 (m, H-200, H-300), 3.38 (t, J = 9.2 Hz, H-400), 3.72–3.74 (m, H-500), 3.65 (dd, J = 7.0 and 10.9 Hz, H-600a), 4.09 (br d, J = 10 Hz, H-600b), 5.01 (d, J = 2.5 Hz, H-10 0 0), 4.00 (d, J = 2.5 Hz, H-20 0 0), 3.81 (d, J = 9.8 Hz, H-40 0 0a), 4.08 (d, J = 9.8 Hz, H-40 0 0b), 3.57 (d, J = 9.8 Hz, H-50 0 0a), 3.81 (d, J = 9.8 Hz, H-50 0 0b), 4.99 (d, J = 2.7 Hz, H-10 0 0 0), 3.95 (d, J = 2.7 Hz, H-20 0 0 0), 3.78 (d, J = 9.8 Hz, H-40 0 0 0a), 3.98 (d, J = 9.8 Hz, H-40 0 0 0b), 3.59 (m, H-50 0 0 0). 13C NMR spectral data (11.7 T, CD3OD, TMS, d ppm) d: 155.7 (C-2), 124.5 (C-3), 182.4 (C-4), 163.9 (C-5), 100.7 (C-6), 164.2 (C-7), 95.4 (C-8), 159.1 (C-9), 107.9 (C-10), 124.1 (C-10), 130.8 (C-20, C-60), 114.6 (C-30, C-50), 161.1 (C-40), 55.5 (40-OCH3), 101.3 (C-10 0), 74.1 (C-20 0), 77.6 (C-30 0), 71.3 (C-40 0), 76.9 (C-50 0), 68.8 (C-60 0), 110.2 (C-10 0 0), 78.3 (C-20 0 0), 79.6 (C-30 0 0), 74.9 (C-40 0 0), 71.5 (C-50 0 0), 110.2 (C-10 0 0 0), 77.4 (C-20 0 0 0), 80.5 (C-30 0 0 0), 74.8 (C-40 0 0 0) and 65.2 (C-50 0 0 0). Compound 18 was isolated as a brown solid. ½a�25 D � 26.2 (c 0.32, MeOH). Online UV spectrum: kmax at 215, 306 and 318 nm. ESI-ITMS m/z 429 [M+K]+ (calc. for C20O22H8 + K). 1H NMR spectral data (11.7 T, CD3OD, TMS, d ppm) d: 7.37 (d, J = 9.0 Hz, H-2, H-6), Fig. 3. Projection of the samples in the score space, evidencing that extract EERLlup (I) was outside of the clusters. The scores were calculated by PCA from the relative concentrations of compounds present in the samples obtained by MCR–ALS. Ethanol extracts: EEFLb (A), EEFLLup (B), EELLb (C), EELLlup (D), EELLsal (E), EELLsid (F), EELLv (G), EELSLas (H), EERLlup (I), EERLsid (J), EESLb (K), EESLlup (L), EESLsal (M), EESLsid (N) and EESLv (O). 2092 C.S. Funari et al. / Food Chemistry 135 (2012) 2086–2094 6.78 (d, J = 9.0 Hz, H-3, H-5), 7.03 (d, J = 16.5 Hz, H-a), 6.85 (d, J = 16.5 Hz, H-b), 6.80 (t, J = 2.0 Hz, H-20), 6.46 (t, J = 2.0 Hz, H-40), 6.66 (t, J = 2 Hz, H-60), 4.91 (d, J = 7.5 Hz, H-10 0), 3.43–3.50 (m, H- 20 0, H-30 0 and H-50 0), 3.39 (t, J = 9.1 Hz, H-40 0), 3.72 (dd, J = 5.9 and 12.1 Hz, H-60 0a) and 3.96 (dd, J = 2.3 and 12.1 Hz, H-60 0b). 13C NMR spectral data (11.7 T, CD3OD, TMS, d ppm) d: 130.3 (C-1), 128.9 (C-2, C-6), 116.5 (C-3, C-5), 158.5 (C-4), 130–4 (C-a), 126.7 (C-b), 141.5 (C-10), 107.1 (C-20), 160.5 (C-30), 104.2 (C-40), 159.6 (C-50), 108.4 (C-60), 102.5 (C-10 0), 75.0 (C-20 0), 78.1 (C-30 0), 71.6 (C- 40 0), 78.3 (C-50 0) and 62.6 (C-60 0). Compound 19 was isolated as a colourless solid. Online UV spectrum: kmax at 231 nm. 1H NMR spectral data (11.7 T, CD3OD, TMS, d ppm) d: 5.59 (d, J = 2.2 Hz, H-1), 7.41 (s, H-3), 4.03 (t, J = 6.2 Hz, H-6), 1.21 (ddd, J = 6.2, 9.0 and 13.2 Hz, H-7a), 2.0 (ddd, J = 6.2, 7.8 and 13.2 Hz, H-7b), 1.50 (m, H-8), 1.94 (dd, J = 2.2 and 10.3 Hz, H-9), 1.04 (d, J = 6.6 Hz, H-10), 3.62 (s, 11-OCH3), 4.49 (d, J = 7.9 Hz, H-10), 3.10 (dd, J = 7.9 and 9.2 Hz, H-20), 3.18–3.30 (m, H-30, H-50), 3.15–3.24 (m, H-40), 3.57 (dd, J = 5.7 and 12.0 Hz, H-60 a), 3.80 (dd, J = 2.0 and 12.0 Hz, H-60 b). 13C NMR spectral data (11.7 T, CD3OD, TMS, d ppm) d: 95.7 (C-1), 154.0 (C-3), 112.7 (C- 4), 72.8 (C-5), 76.8 (C-6), 40.4 (C-7), 31.5 (C-8), 55.7 (C-9), 19.7 (C-10), 167.9 (C-11), 51.5 (11-OCH3), 100.0 (C-10), 74.3 (C-20), 77.4 (C-30), 71.6 (C-40), 78.1 (C-50) and 62.5 (C-60). Compounds 12–19 (Fig. 1) are reported here for the first time in L. lupulina Cham. Oleanolic acid (13) was previously isolated from Lippia triphylla by Ono et al. (2008), while oleanonic acid (12) had not been isolated in the Lippia genus. Both 12 and 13 showed insec- ticidal activity against Sitophilus oryzae (L.) (Pungitore, García, Gianello, Sosa, & Tonn, 2005) and toxicity towards M. tuberculosis H37Rv (Caldwell, Franzblau, Suarez, & Timmermann, 2000). Com- pounds 14–19 have not been reported previously in the Verbena- ceae family. Avicequinone E (15) was active against human ovarian cancer cells line A2780 with an IC50 of 8.8 lM, while steno- carpoquinone (14) presented an IC50 of 50 lM in the same test (Williams et al., 2006). No reports of the biological activity of biochanin A-7-O-b-D-apiofuranosyl-(1 ? 5)-b-D-apiofuranosyl- (1 ? 6)-b-D-glucopyranoside (17) were found in the literature. Piceid (18) is abundant in nature and in food, such as grapes, cocoa and peanuts, and some of its biological activity has been reported, mainly its strong antioxidant activity (Counet, Callemien, & Collin, 2006). No reports on the biological activity of b-dihydrohastatoside (19) were found in the literature, but iridoid glycosides are known to have a wide range of activities (Tundis, Loizzo, Menichini, Statti, & Menichini, 2008). 3.3. Antifungal assay The clinical importance of systemic mycosis has increased rap- idly in recent years, mainly due to the increasing incidence of AIDS and immunocompromised or severely ill patients. Some Lippia spe- cies have been employed in ethnomedicine as antimicrobial agents (Lorenzi & Matos, 2002; Pascual et al., 2001), such as L. lupulina Cham. (Rodrigues & de Carvalho, 2001) and L. sidoides Cham. (Lemos et al., 2007). In this study, the influence of EERLlup, EELLsal and EESLsal, together with their partition fractions, was investi- gated against opportunistic human yeast pathogens (C. albicans, C. krusei, C. parapsilosis and C. neoformans). Table 2 summarizes the results obtained for these samples. EELLsal was the most active among the crude extracts, with MICs of 125 lg/ml for C. albicans, C. krusei and C. parapsilosis, and 62.5 lg/ml for C. neoformans (Table 2). According to Scorzoni et al. (2007), EELLsal presents moderate activities against the Can- dida strains used and strong activity against C. neoformans. EESLsal, which presented MICs of 250 lg/ml for Candida strains and 125 lg/ ml for C. neoformans, was considered moderately active against all yeasts tested. The MICs for EERLlup were outside the maximum concentration tested (250 lg/ml). EELLsal and EELLsid showed similar chromatographic profiles and chemical compositions (Table 1). Because L. sidoides Cham. is largely used in ethnophar- macology as a general antiseptic (Lorenzi & Matos, 2002), such similarities might be associated with the antifungal activities observed for EELLsal (Table 2). Among the partition fractions of the three crude extracts, FBu1 and FAq1 from EERLlup and FAc2 from EELLsal showed stronger activities than their original crude extracts for all tested strains (Table 2). FBu1 presented stronger activity against C. albicans, C. krusei and C. neoformans, with MICs of 62.5, 15.6 and 31.2 lg/ml, respectively, and moderate activity against C. parapsilosis (MIC of 125 lg/ml). FAq1 presented stronger activities against C. krusei and C. neoformans with MICs of 62.5 lg/ml for both strains, and Table 2 Antifungal activity of the ethanol extracts of leaves (EELLsal) and stems (EESLsal) of L. salviaefolia, roots of L. lupulina (EERLlup) and of their partition fractions (MICs in lg/ ml).a Extract/ fraction Candida albicans Candida krusei Candida parapsilosis Cryptococcus neoformans EERLlup >250 >250 >250 >250 FHex1 >250 >250 >250 >250 FAc1 125 125 >250 125 FBu1 62.5 15.6 125 31.2 FAq1 125 62.5 250 62.5 EELLsal 125 125 125 62.5 FHex2 >250 >250 >250 62.5 FAc2 62.5 31.2 62.5 31.2 FBu2 125 62.5 125 31.2 FAq2 250 125 250 62.5 EESLsal 250 250 250 125 FHex3 250 250 >250 250 FAc3 125 125 >250 125 FBu3 250 125 >250 250 FAq3 >250 >250 >250 >250 a Best MIC values are boldfaced. C.S. Funari et al. / Food Chemistry 135 (2012) 2086–2094 2093 moderate activities towards C. albicans and C. parapsilosis (MICs of 125 and 250 lg/ml, respectively). FAc2 showed strong activity against all tested strains, with MICs of 62.5 lg/ml towards C. albi- cans and C. parapsilosis, and 31.2 lg/ml towards C. krusei and C. neoformans. Thus, the purified compounds used for antifungal activity evaluation were selected from the most active fractions FBu1 (17 and 18) and FAc2 (7–11). Additionally, 4 and 6, the inter- converting isomers 1a/1b and 2a/2b from FBu3, and 12 from FHex1 were also assayed. Table 3 shows the antifungal activity measured for these compounds. Sakuranetin (11) and oleanonic acid (12) did not show any activity at the maximum concentration assayed (250 lg/ml). On the other hand, eight pure compounds (4, 6–10 and 17–18) and a mixture of the interconverting isomers (1a, 1b, 2a and 2b) inhib- ited at least one strain with MIC 6 250 lg/ml. Verbascoside (6), a phenylpropanoid isolated from the ethanol extract of stems of L. salviaefolia (EESLsal) (Funari et al., 2011), showed the strongest activity with an MIC of 25.0 lmol/l (or 15.6 lg/ml) against C. neo- formans, which was approximately 6 times less active than the po- sitive control amphotericin B (MIC of 4.3 lmol/l). This strain was the most susceptible because 4, 6, 10 and 17 as well as the mixture of isomers 1a, 1b, 2a and 2b showed MICs < 100 lmol/l. Biochanin Table 3 Antifungal activity of compounds isolated from ethanol extracts of roots from L. lupulina (EE (1–11) (MIC in lg/ml and lmol/l)a. Compound/strain Candida albicans Candida krusei (lg/ml) (lmol/l) (lg/ml) Isomeric mixture (1/2)b 62.5 134.1 62.5 Forsythoside B (4) 125 165.3 250 Verbascoside (6) 125 200.3 125 Aromadendrin (7) 250 868.0 250 Naringenin (8) 250 919.1 250 Phloretin (9) 250 912.4 250 Asebogenin (10) >250 >868.0 250 Sakuranetin (11) >250 >874.1 >250 Oleanonic acid (12) >250 >550.7 >250 Biochanin A triglycoside (17)c 125 176.0 62.5 Piceid (18) 125 320.5 125 Amphotericin B 2.0 2.2 64 Fluconazole 2.0 6.5 2.0 a Best MIC values are boldfaced. b (2S)- and (2R)-30 ,40 ,5,6-Tetrahydroxyflavanone-7-O-b-glucopyranoside and (2S)- and c Biochanin A 7-O-b-D-apiofuranosyl-(1 ? 5)-b-D-apiofuranosyl-(1 ? 6)-b-D-glucopyra A 7-O-b-D-apiofuranosyl-(1 ? 5)-b-D-apiofuranosyl-(1 ? 6)-b-D- glucopyranoside (17) showed MIC of 88.0 lmol/l towards C. krusei, which was close to that observed for amphotericin B (MIC of 69.3 lmol/l). This isoflavone triglycoside presented similar inhibi- tory activity against C. neoformans (MIC of 88.0 lmol/l) and only twofold lower activity towards C. albicans and C. parapsilosis (MICs of 176 lmol/l) (Table 3). The presence of isoflavone glucoside 17 in EERLlup might be associated with the well-known antifungal activ- ity of isoflavones and their wide occurrence in plant roots. It should be noted that compounds 4, 6–10 and the mixture of interconverting isomers 1a/1b/2a/2b, which were active against one or more strains, were detected in the extracts of L. sidoides Cham. leaves and stems (Table 1). Infusions or ethanol extracts and tinctures prepared with aerial parts of this species have been widely used in Brazil as a general antiseptic (Lorenzi & Matos, 2002). Recently, this species was included in the Brazilian Health Ministry’s priority list of species for phytotherapeutic product development (Ministério da Saúde, 2009). Several compounds derived from various species of Lippia have been studied for their antioxidant and antimicrobial activities and for their use as food seasonings. It has been shown that L. pseudo- thea presented an MIC of 625 mg/ml towards C. albicans, and L. sidoides showed an MIC of 625 mg/ml for both C. albicans and C. neoformans. The technique used to determine these antifungal activities was bio-autography and indicated that the antioxidant activity of the extracts was due to coumarins and flavonoids. Ter- penoids and the same flavonoids were associated to the observed antimicrobial properties (Fabri, Nogueira, Moreira, Bouzada, & Scio, 2011). Our results confirmed the antimicrobial potential of Lippia spp. in addition to revealing their effective inhibition of major hu- man fungal pathogens. This is the first report on L. salviaefolia and L. lupulina producing compounds with potential antifungal activity against C. krusei and C. neoformans. Given the poor arsenal of antifungal drugs and the problems regarding toxicity and increased fungal resistance to the usual therapies, the treatment of human mycoses is not always effective. For this reason, there is a growing interest in finding novel, effec- tive antifungal drugs. Considering the immense Brazilian biodiver- sity, exploring plants for novel antifungal compounds should be a priority. In addition to having increased the chemical knowledge on thir- teen extracts from five species of the Lippia genus by comparison with phytochemical markers, HPLC–PDA combined with statistics led us to the selection of the ethanol extract of roots of L. lupulina RLlup) (12, 17 and 18) and of leaves (EELLsal) and stems (EESLlsal) from L. salviaefolia Candida parapsilosis Cryptococcus neoformans (lmol/l) (lg/ml) (lmol/l) (lg/ml) (lmol/l) 134.1 62.5 134.1 31.2 66.9 330.7 125 165.3 62.5 82.7 200.3 125 200.3 15.6 25.0 868.0 125 434.0 125 434.0 919.1 >250 >919.1 125 459.6 912.4 125 456.2 62.5 228.1 868.0 >250 >868.0 15.6 54.2 >874.1 >250 >874.1 >250 >874.1 >550.7 >250 >550.7 >250 >550.7 88.0 125 176.0 62.5 88.0 320.5 125 320.5 125 320.5 69.3 8.0 8.6 4.0 4.3 6.5 1.0 3.3 0.1 0.2 (2R)-30 ,40 ,5,8-tetrahydroxyflavanone-7-O-b-glucopyranoside. noside. 2094 C.S. Funari et al. / Food Chemistry 135 (2012) 2086–2094 for further phytochemical studies. These studies resulted in the isolation of eight compounds, including seven previously unre- ported in the Lippia genus. The reported strategy was effective at avoiding replication of time-consuming isolation procedures, which might lead to compounds previously isolated from L. salviae- folia leaves and stems (Funari et al., 2011). Furthermore, the ethyl acetate and n-butanol fractions from the ethanol extract of L. sal- viaefolia leaves (FAc2 and FBu2, respectively) showed strong inhi- bition of fungal growth, along with verbascoside (6) and asebogenin (10), previously isolated from L. salviaefolia. Therefore, these natural products might be considered promising prototypes for the development of new antifungal agents, especially against C. neoformans. Acknowledgements The authors would like to thank the São Paulo State Research Foundation (FAPESP) for PhD fellowship (to CSF) and research funding (#04/07932-7, awarded to DHSS); Coordination of Improvement of Higher Education Personnel (CAPES) and Brazilian National Council for Scientific and Technological Development (CNPq), for students and research fellowships (to CSF, FPG, AMFA and DHSS). Thanks also go to Dr Maria Inês Cordeiro, from Instituto de Botânica do Estado de São Paulo, and Dr Giselda Durigan, from Floresta Estadual de Assis, São Paulo, for botanical identifications, and to the Laboratory of Phytochemistry, from São Paulo State Uni- versity (UNESP-Brazil), especially to José de Sousa Lima Neto, for the specific rotation measurements. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.foodchem.2012. 06.077. References Arthur, H., Joubert, E., De Beer, D., Malherbe, C. J., & Witthuhn, R. C. (2011). Phenylethanoid glycosides as major antioxidants in Lippia multiflora herbal infusion and their stability during steam pasteurization of plant material. Food Chemistry, 127, 581–588. Caldwell, C. G., Franzblau, S. G., Suarez, E., & Timmermann, B. N. (2000). Oleanane triterpenes from Junellia tridens. Journal of Natural Products, 63, 1611–1614. Carneiro, R. L., Braga, J., Bottoli, C., & Poppi, R. J. (2007). Application of genetic algorithm for selection of variables for the BLLS method applied to determination of pesticides and metabolites in wine. Analytica Chimica Acta, 595, 51–58. Chandrasekar, P. (2010). Diagnostic challenges and recent advances in the early management of invasive fungal infections. European Journal of Haematology, 84, 281–290. Counet, C., Callemien, D., & Collin, S. (2006). Chocolate and cocoa: New sources of trans-resveratrol and trans-piceid. Food Chemistry, 98, 649–657. da Silva, B. P., Velozo, L. S. M., & Parente, J. P. (2000). Biochanin A triglycoside from Andira inermis. Fitoterapia, 71, 663–667. Fabri, R. L., Nogueira, M. S., Moreira, J. R., Bouzada, M. L., & Scio, E. (2011). Identification of antioxidant and antimicrobial compounds of Lippia species by bioautography. Journal of Medicinal Food, 14, 840–846. Funari, C. S., Passalacqua, T. G., Rinaldo, D., Napolitano, A., Festa, M., Capasso, A., et al. (2011). Interconverting flavanone glucosides and other phenolic compounds in Lippia salviaefolia Cham. ethanol extracts. Phytochemistry, 72, 2052–2061. Kriengkauykiat, J., Ito, J. I., & Dadwal, S. S. (2011). Epidemiology and treatment approaches in management of invasive fungal infections. Clinical Epidemiology, 3, 175–191. Lemos, T. L. G., Monte, F. J. Q., Santos, A. K. L., Fonseca, A. M., Santos, H. S., Oliveira, M. F., et al. (2007). Quinones from plants of northeastern Brazil: Structural diversity, chemical transformations, NMR data and biological activities. Natural Products Research, 21, 529–550. Lorenzi, H., & Matos, F. J. A. (2002). Plantas medicinais no Brasil: Nativas e exóticas cultivadas. Nova Odessa: Instituto Plantarum. Lu, Y., Berthod, A., Hu, R., Ma, M., & Pan, Y. (2009). Screening of complex natural extracts by countercurrent chromatography using a parallel protocol. Analytical Chemistry, 81, 4048–4059. Mahato, S. B., & Kundo, A. P. (1994). 13C NMR spectra of pentacyclic triterpenoids – A compilation and some salient features. Phytochemistry, 37(6), 1517–1575. Mdod, R., Moser, S. A., Jaoko, W., Baddley, J., Pappas, P., Kempf, M. C., et al. (2011). Antifungal susceptibilities of Cryptococcus neoformans cerebrospinal fluid isolates from AIDS patients in Kenya. Mycoses, 54, 438–442. Ministério da Saúde (2009). RENISUS – Relação Nacional de Plantas Medicinais de Interesse ao SUS, DAF/SCTIE/MS, Brasília. Newman, D. J., & Cragg, G. M. (2007). Natural products as sources of new drugs over the last 25 years. Journal of Natural Products, 70, 461–477. Ono, M., Oda, E., Tanaka, T., Iida, Y., Yamasaki, T., Masuoka, C., et al. (2008). DPPH radical-scavenging effect on some constituents from aerial parts of Lippia triphylla. Journal of Natural Medicine, 62, 101–106. Pascual, M. E., Slowing, K., Carretero, E., Sánchez, D. M., & Villar, A. (2001). Lippia: Traditional uses, chemistry and pharmacology. A review. Journal of Ethnopharmacology, 76, 201–214. Pungitore, C. R., García, M., Gianello, J. C., Sosa, M. E., & Tonn, C. E. (2005). Insecticidal and antifeedant effects of Junellia aspera (Verbenaceae) triterpenes and derivatives on Sitophilus oryzae (Coleoptera: Curculionidae). Journal of Stored Products Research, 41, 433–443. Rodrigues, V. E. G., & de Carvalho, D. A. (2001). Plantas medicinais no domínio dos Cerrados. Lavras: UFLA. Rodriguez-Tudela, J. L., & Barchiesi, F. Subcommittee on Antifungal Susceptibility Testing (AFST). (2008). EUCAST definitive document EDef 7.1: Method for the determination of broth dilution MICs of antifungal agents for fermentative yeast. Clinical Microbiology and Infections, 14, 398–405. Sasaki, H., Taguchi, H., Endo, T., & Yosioka, I. (1982). The constituents of Ledebouriella seseloides Wolff. I. Structures of three new chromones. Chemical & Pharmaceutical Bulletin, 30, 3555–3562. Schmeda-Hirschmann, G., & Papastergiou, F. (2003). Naphthoquinone derivatives and lignans from the paraguayan crude grug ‘‘Tayï Pytá’’ (Tabebuia heptaphylla, Bignoniaceae). Zeitschrift für Naturforschung, 58, 495–501. Scorzoni, L., Benaducci, T., Fusco-Almeida, A. M., Silva, D. H. S., Bolzani, V. S., & Mendes-Giannini, M. J. S. (2007). The use of standard methodology for determination of antifungal activity of natural products against medical yeasts Candida sp and Cryptococcus sp. Brazilian Journal of Microbiology, 38, 391–397. Snyder, L. R., Kirkland, J. J., & Glajch, J. L. (1997). Practical HPLC method development. New York: John Wiley & Sons. Teborg, D., & Junior, P. (1991). Iridoid glucosides from Penstemon nitidus. Planta Medica, 57(2), 184–186. Tundis, R., Loizzo, M. R., Menichini, F., Statti, G. A., & Menichini, F. (2008). Biological and pharmacological activities of iridoids: Recent developments. Mini-Reviews in Medicinal Chemistry, 8, 399–420. Wagner, H., & Bladt, S. (1996). Plant drug analysis. Berlin: Springer-Verlag. Williams, R. B., Norris, A., Miller, J. S., Razafitsalama, L., Andriantsiferana, R., Rasamison, V. E., et al. (2006). Two new cytotoxic naphthoquinones from Mendoncia cowanii from the rainforest of Madagascar. Planta Medica, 72, 564–566. Zoghbi, M. D. B., Andrade, E. H. A., Silva, M. H. L., & Maia, J. G. S. (2002). Volatile constituents of Lippia lupulina Cham. Flavour and Fragrance Journal, 17, 29–31. http://dx.doi.org/10.1016/j.foodchem.2012.06.077 http://dx.doi.org/10.1016/j.foodchem.2012.06.077 Chemical and antifungal investigations of six Lippia species (Verbenaceae) from Brazil 1 Introduction 2 Experimental 2.1 General information 2.2 Plant material 2.3 Extraction 2.4 Dereplication by HPLC–PDA 2.5 Multivariate curve resolution and principal component analyses 2.6 Isolation procedures 2.7 Antimicrobial susceptibility testing 3 Results and discussion 3.1 Dereplication by HPLC–PDA 3.2 Isolation and characterization of the root extract of L. lupulina 3.3 Antifungal assay Acknowledgements Appendix A Supplementary data References