Antileishmanial Activity of a Linalool-Rich Essential Oil from Croton cajucara

The in vitro leishmanicidal effects of a linalool-rich essentialoil from the leaves of Croton cajucara against Leishmania amazonensiswere investigated. Morphological changes in L. amazonensis promastigotestreated with 15 ng of essential oil per ml were observed bytransmission electron microscopy; leishmanial nuclear and kinetoplastchromatin destruction, followed by cell lysis, was observedwithin 1 h. Pretreatment of mouse peritoneal macrophages with15 ng of essential oil per ml reduced by 50% the interactionbetween these macrophages and L. amazonensis, with a concomitantincrease by 220% in the level of nitric oxide production bythe infected macrophages. Treatment of preinfected macrophageswith 15 ng of essential oil per ml reduced by 50% the interactionbetween these cells and the parasites, which led to a 60% increasein the amount of nitric oxide produced by the preinfected macrophages.These results provide new perspectives on the development ofdrugs with activities against Leishmania, as linalool-rich essentialoil is a strikingly potent leishmanicidal plant extract (50%lethal doses, 8.3 ng/ml for promastigotes and 8.7 ng/ml foramastigotes) which inhibited the growth of L. amazonensis promastigotesat very low concentrations (MIC, 85.0 pg/ml) and which presentedno cytotoxic effects against mammalian cells.

INTRODUCTION

Top

Introduction

Parasites of the genus Leishmania are transmitted by the biteof sand flies and infect cells of the mononuclear phagocytelineage of their vertebrate hosts (2, 21). Depending both onthe virulence factors of the parasite itself and on the immuneresponse established by the host, a spectrum of diseases knownas leishmaniasis can appear, and these can be cutaneous and/orvisceral (30). Approximately 350 million people live in areasof active transmission of Leishmania, with 12 million peoplethroughout Africa, Asia, Europe, and the Americas directly affectedby leishmaniasis. More than 90% of the cutaneous cases appearin Afghanistan, Saudi Arabia, Algeria, Brazil, Iran, Iraq, Syria,and Sudan (13). Cutaneous leishmaniasis either can resolve spontaneouslyafter a few months or, depending on the causative Leishmaniaspecies, can evolve into diffuse cutaneous, relapsing cutaneous,or mucocutaneous leishmaniasis, while untreated visceral leishmaniasisleads to death in the majority of patients (2). Leishmania amazonensisis one of the principal agents of diffuse cutaneous leishmaniasis,which is usually unresponsive to all treatments known to date(31). Also, visceralization of Leishmania strains that are classicallyrestricted to cutaneous leishmaniasis has often been observedin patients with Leishmania-human immunodeficiency virus coinfection(3).

The control of leishmaniasis remains a problem because no vaccinesexist and the available chemotherapy still relies on the potentiallytoxic pentavalent antimonials, which cause serious side effectsand require long-term treatment (38). The rise in the ratesof in vitro antimonial resistance due to intermittent drug exposure(15, 16), the isolation of antimonial-resistant Leishmania strainsfrom patients with unresponsive cutaneous leishmaniasis (2,9), and recently, the numerous cases of visceral leishmaniasisamong patients infected with the human immunodeficiency virus(3) make the search for new agents for the treatment of leishmaniasisurgent. Extensive studies of new drugs with antileishmanialactivities, including both natural products and synthetic compounds,have been undertaken worldwide (9), although problems with theside effects of the chemotherapies used at present have notyet been solved.

In recent years, there has been growing interest in alternativetherapies and the use of natural products, especially thosederived from plants (39). The bark of Croton cajucara is usedin Brazilian folk medicine as an infusion to treat gastrointestinaldisorders (22). Experimental studies with laboratory animalshave identified potential applications for some purified substancesfrom C. cajucara bark extracts, which present anti-inflammatory,antiulcerogenic, antitumorigenic, antimutagenic, antiestrogenic,hypoglycemic, and triglyceride-lowering effects (1, 6-8, 14,20, 22-24, 33, 34, 40).

Linalool, a terpenic alcohol (Fig. 1), is the principal componentof rosewood (Aniba rosaeodora var. amazonica Ducke syn Anibaduckei Kostermans) and Ho tree (Cinamomon camphora) oils (37).It is also obtained as a by-product in the industrial synthesisof vitamin E (37). The antimicrobial and anesthetic activitiesof linalool-containing essential oil have been reported (4,17, 35). Recently, the linalool-rich essential oil from theleaves of C. cajucara has been purified, and its contents havebeen analyzed by gas chromatography-mass spectrometry (GC-MS)(32). Based upon the facts that the essential oil extractedfrom the bark of C. cajucara presents anti-inflammatory activityin rodents (22, 23) and that linalool-rich essential oil extractedfrom other plants presents antimicrobial properties (35), inthis work we investigated the effects of essential oil extractedfrom the leaves of C. cajucara on L. amazonensis parasites,on the interaction of these flagellates with mouse peritonealmacrophages, and on nitric oxide production by the infectedmacrophages.

View larger version (8K):
[in this window]
[in a new window]
FIG. 1. Chemical structure of linalool.

MATERIALS AND METHODS

Top

Materials and Methods

Plant material. Plant material from C. cajucara Benth was obtained from EmbrapaExperimental Farm, Amazonas, Brazil. A voucher specimen wasdeposited at the Embrapa Occidental Amazon Herbarium (registryno. IAN 165013).

Essential oil extraction. Leaves of C. cajucara were dried at room temperature and coarselypowdered. The oil was obtained by hydrodistillation (5 h) ofC. cajucara leaves with a modified Clevenger apparatus (18);the yield was 0.40% (on a dry weight basis).

Linalool purification. Linalool was purified and identified by GC and GC coupled withMS (GC-MS). GC was performed with a Hewlett-Packard (HP) 6890gas chromatograph fitted with a BP-5 fused-silica capillarycolumn (25 m by 0.33 mm; film thickness, 0.5 µm), withhelium used as the carrier gas (1 ml/min). Retention indiceswere obtained by injection of a series of n-alkanes and werecompared with those in the literature. The GC-MS system usedwas an HP 5973 MSD system coupled with an HP 6890 gas chromatograph;helium was used as the carrier gas, and the same column andconditions described above were used (32). Results were comparedwith the Wiley library of spectra.

Parasite culture. Promastigote forms of L. amazonensis (Raimundo strain, MHOM/BR/76/Ma-5)were maintained by weekly transfers in brain heart infusionmedium supplemented with 10% fetal bovine serum (FBS) at 26°C.The infectivities of the parasites were maintained by periodicinoculation into hamster footpads.

MIC evaluation. L. amazonensis promastigotes (10⁶ parasites/ml) were incubatedat 26°C for 120 h in fresh medium (brain heart infusionmedium) supplemented with 10% FBS in the absence or presenceof several concentrations (1 pg/ml to 1 mg/ml) of essentialoil or purified linalool (cell growth was determined daily byassessment of visible turbidity) in order to evaluate the MIC,as described previously (36). The MIC was considered the lowestconcentration of each substance used that prevented the growthof L. amazonensis in vitro.

Mouse peritoneal macrophages. Thioglycolate-elicited peritoneal macrophages from female Swissmice (age, 6 to 8 weeks) were collected in 0.85% saline andallowed to adhere onto coverslips placed in 24-well cultureplates for 30 min at 37°C in a 4% CO₂ atmosphere. Nonadherentcells were then removed, and the adhered macrophages were washedtwice with 0.9% saline and cultured for 24 h in culture medium(RPMI; Gibco-BRL, Gaithersburg, Md.) supplemented with 10% FBS.

Purification of amastigotes. Intact living L. amazonensis promastigotes in the stationarygrowth phase were added to plate wells containing cultured peritonealmacrophages. Free promastigotes were removed after 2 h of incubation,and the infected macrophages were kept for 48 h under the conditionsdescribed above. Free amastigotes were then collected from thesupernatants of the culture plate wells. The amastigotes werewashed twice in 0.1 M phosphate buffer (pH 7.2) and then resuspendedin RPMI culture medium for the antileishmanial activity assays.

Antileishmanial activity. L. amazonensis promastigotes or amastigotes (10⁶ parasites/ml)were incubated in RPMI culture medium in the absence or presenceof several concentrations (1 pg/ml to 1 mg/ml) of essentialoil or purified linalool at 37°C in order to evaluate parasitesurvival and cell morphology by optical microscopy at 10-minintervals. Parasite viability was assessed before and afterthe incubations by evaluation of the motility and by the trypanblue method (5) with a hemocytometer. Cell viability was determinedby using the following formula: [100 - (L2/L1)] x 100, whereL1 is the percentage of viable control cells and L2 is the percentageof viable treated cells, as described previously (11). The 50%lethal dose (LD₅₀) was determined by logarithmic regressionanalysis of the data obtained by the formula described above,as described previously (36). Cell morphology evaluation wasperformed with fresh as well as Giemsa-stained preparations,as described previously (31). The essential oil and the purifiedlinalool were diluted in dimethyl sulfoxide (DMSO; Sigma ChemicalCo., St. Louis, Mo.) at 100 mg/ml and then in RPMI. In all tests,1% DMSO (the same concentration present in the highest doseof the compounds) and medium alone were used as controls.

Electron microscopy. L. amazonensis promastigotes were incubated in the absence orpresence of 1 ng of linalool-rich essential oil per ml for 5,10, 15, 20, 25, or 30 min. The parasites were washed twice inRinger's solution (0.9% NaCl, 5.0% KCl, 5.0% CaCl₂) and fixedin a solution containing 2.5% glutaraldehyde, 4% formaldehyde,and 3.7% sucrose in 0.1 M phosphate buffer (pH 7.2) for 1 hat room temperature. The parasites were then washed in 0.1 Mcacodylate buffer (pH 7.2), and the parasites were gently scrapedoff with a rubber policeman and postfixed in a solution containing1% OsO₄, 0.8% potassium ferricyanide, and 5 mM calcium chloridein 0.1 M cacodylate buffer (pH 7.2) for 1 h at room temperaturein the dark. The parasites were then rinsed in cacodylate buffer,dehydrated in acetone, and embedded in Epon. Ultrathin sectionswere stained with uranyl acetate and were examined in a transmissionelectron microscope (900; Carl Zeiss, Oberkochen, Germany) operatedat 80 kV.

Infection of macrophages and nitric oxide production. Mouse peritoneal macrophages were obtained as described above.The parasites and/or the macrophages were either not treatedor treated with 15, 1.5, or 0.2 ng of essential oil per ml 20min prior to the macrophage-parasite interactions. The adherentcultured macrophages and the free parasites were washed onceand resuspended in fresh culture medium. Dead parasites wereremoved from the medium by centrifugation (1,000 x g, 5 min),and intact living L. amazonensis promastigotes in the stationarygrowth phase were then added to the macrophage culture platewells. The parasite-macrophage interaction studies were performedat 37°C for 90 min by using parasites and/or macrophagespretreated with the essential oil or macrophages that had alreadybeen infected with the parasites for 24 h and then treated withthe essential oil. In the last system, all promastigotes hadalready differentiated into amastigotes before the treatmentwith the essential oil. A ratio of 10 promastigotes to 1 macrophagewas used for both infection assays. After the interaction assayswere done, the coverslips were fixed and Giemsa stained, andthe percentage of infected macrophages was determined by counting600 cells in triplicate coverslips. The association indiceswere determined by multiplying the percentage of infected macrophagesby the mean number of parasites per infected cell. Associationindices were the number of parasites that actually infectedthe macrophages. The supernatants from control and L. amazonensis-infectedmacrophages were analyzed for their nitrite contents by theGriess reaction, as described previously (19). The absorbanceat 550 nm was measured, and the concentration of nitrite wascalculated by using a linear regression of a standard curve,as described previously (25). The mice were first killed toobtain mouse peritoneal macrophages for both infection withLeishmania and nitric oxide measurements, and all federal guidelinesand institutional policies for the treatment of mice were adheredto.

Cytotoxicity assay. The cytotoxicity of the drug for mammalian cells was measuredby the neutral red uptake assay (26) with mouse peritoneal macrophagesand a transformed cell line (Vero cells) (29). The cells werecultivated in 96-well microtiter plates (150 µl containing10⁵ cells/ml in Eagle minimal essential medium [Eagle MEM]/well)at 37°C in a humidified 5% CO₂ atmosphere. The medium wascompleted by the addition of L-glutamine (0.10 g/liter), HEPES(2.38 g/liter), penicillin G (10⁵ IU/liter), streptomycin sulfate(0.10 g/liter), and 4% FBS. After a 24-h incubation, 50 µlof ethanolic or aqueous crude extracts, essential oil, or linaloolwas added to the cell cultures at the respective MICs and LD₅₀sfor L. amazonensis. A total of 50 µl of Eagle MEM wasadded to the control cells. After further incubation for 48h, control and treated cells were washed three times with phosphate-bufferedsaline (PBS) solution (pH 7.2). A total of 100 µl of neutralred solution (0.3% in Eagle MEM) was added to each well. Aftera 3-h incubation at 37°C in this solution, the cells werethen washed three times with PBS. A total of 100 µl ofa solution containing acetic acid (1%; vol/vol) and ethanol(50%; vol/vol) was added to the wells, and the optical densitiesof the supernatants were measured at 540 nm.

Statistical analysis. All experiments were performed in triplicate. The mean and standarddeviation of at least three experiments were determined. Statisticalanalysis of the differences between mean values obtained forthe experimental groups was done by Student's t test. P valuesof 0.05 or less were considered significant.

RESULTS

Top

Results

Inhibition of parasite growth. The MICs of C. cajucara essential oil and purified linaloolfor the growth of L. amazonensis promastigotes were 85.0 and22.0 pg/ml, respectively.

Antileishmanial activity. The effects of C. cajucara essential oil and purified linaloolon the viability of L. amazonensis were tested. LD₅₀s for promastigoteswere 8.3 ng/ml for essential oil and 4.3 ng/ml for purifiedlinalool, and the LD₅₀s for amastigotes were 22.0 ng/ml foressential oil and 15.5 ng/ml for purified linalool. Figure 2shows the time course of the viabilities of L. amazonensis promastigotes,amastigotes, and mouse peritoneal macrophages in the absenceor presence of linalool-rich essential oil. Essential oil at15.0 ng/ml was able to kill 100% of the parasites in 60 min.On the other hand, mouse macrophages were unaffected by essentialoil at 15.0 ng/ml (Fig. 2).

View larger version (17K):
[in this window]
[in a new window]
FIG. 2. Time courses of the viabilities of L. amazonensis promastigotes and mouse peritoneal macrophages in the absence or presence of linalool-rich essential oil (15.0 ng/ml) extracted from C. cajucara. The viabilities of both parasites and macrophages were calculated as described in the text. •, L. amazonensis promastigotes (control); ${circ}$ , L. amazonensis promastigotes and essential oil; ${blacksquare}$ , L. amazonensis amastigotes (control); ${square}$ , L. amazonensis amastigotes and essential oil; ${blacktriangledown}$ , mouse peritoneal macrophages (control); ${triangledown}$ , mouse peritoneal macrophages and esssential oil

Transmission electron microscopy. Untreated and treated (15.0 ng of essential oil per ml) promastigoteswere observed by transmission electron microscopy, and photomicrographsof the promastigotes are shown in Fig. 3A to E, which show promastigoteswith different degrees of damage. Disruption of flagellar membranes,mitochondrial swelling, and gross alterations in the organizationof the nuclear and kinetoplast chromatins were detected. After30 min in the presence of essential oil the parasites were completelydestroyed.

View larger version (138K):
[in this window]
[in a new window]
FIG. 3. Effects of linalool-rich essential oil (15.0 ng/ml) extracted from C. cajucara on promastigote forms of L. amazonensis observed by transmission electron microscopy. (A) Control parasites; (B to E) parasites treated for 5 (B), 10 (C), 15 (D), and 30 (E) min, showing promastigotes with different degrees of damage. Note the disruption of flagellar membranes (arrowheads in panels B and C), the mitochondrial swelling (C and D), and the gross alterations in the organization of the nuclear and kinetoplast chromatins (C and D). After 30 min in the presence of essential oil the parasites were completely destroyed (E). N, nucleus; K, kinetoplast, F, flagellum. Bars, 1 µm.

Cytotoxicity for mammalian cells. The effects of the essential oil and linalool (at the MICs andLD₅₀s for L. amazonensis) on mouse peritoneal macrophages anda transformed cell line (Vero cells) were tested as describedpreviously (12). No cytotoxic effects on mammalian cells wereobserved at the concentrations used (data not shown).

Infection of macrophages. Figure 4 shows the effects of the essential oil on the L. amazonensis-macrophageinteraction. The parasites and/or the mouse peritoneal macrophageswere either not treated or treated with different concentrations(15.0, 1.5, and 0.2 ng/ml) of essential oil 20 min prior tothe macrophage-parasite interactions. When macrophages werepretreated with essential oil, the association indices were50% lower than those in the control system (in which controlmacrophages and control parasites were used), regardless ofthe concentration of essential oil used in those assays. Whenparasites were pretreated with 15, 1.5, and 0.2 ng of essentialoil per ml, the association indices were 50, 30, and 25% lower,respectively, than those in the control system; these resultswere approximately the same as those obtained for pretreatedmacrophages and parasites. When the macrophages were preinfectedwith L. amazonensis for 24 h and then treated with 15 ng ofessential oil per ml, the association index was 50% lower thanthat for the control system.

View larger version (34K):
[in this window]
[in a new window]
FIG. 4. Effects of linalool-rich essential oil extracted from C. cajucara on the L. amazonensis-macrophage interaction. The parasites and/or the mouse peritoneal macrophages were either not treated or treated with 15, 1.5, and 0.2 ng of essential oil per ml 20 min prior to the macrophage-parasite interactions. The adherent cultured macrophages and the free parasites were washed once and resuspended in fresh culture medium. Dead parasites were removed from the medium by centrifugation (1,000 x g, 5 min), and intact living L. amazonensis promastigotes were then added to the macrophage culture plate wells. Association indices were determined by light microscopy, with 600 cells in triplicate coverslips counted after 90 min of interaction. Each bar represents the mean ± standard error of at least three independent experiments, which were performed in triplicate. The association indices from assays performed with macrophages and/or parasites pretreated with essential oil are significantly different from the association indices for control (nontreated) macrophages.

Nitric oxide production. Mouse peritoneal macrophages were either noninfected or infectedwith L. amazonensis, and then the culture supernatants wereevaluated for nitrite contents. The parasites and/or the macrophageswere either not treated or treated with 15, 1.5, and 0.2 ngof essential oil per ml 20 min prior to the macrophage-parasiteinteractions. Figure 5 shows that noninfected macrophages thatwere treated with essential oil produced 40 to 50% more nitricoxide than the control (noninfected and nontreated) macrophages.Infected macrophages produced 30% more nitric oxide than noninfectedmacrophages. When infected macrophages were pretreated with15, 1.5, and 0.2 ng of essential oil per ml, the levels of nitricoxide production were 220, 150, and 130% higher, respectively,than those for the control infected macrophages. When parasiteswere pretreated with 15 ng of essential oil per ml, the levelof nitric oxide production was 170% higher than that for thecontrol infected macrophages, although no significant differencewas obtained when the parasites were pretreated with 1.5 or0.2 ng of essential oil per ml. When both macrophages and parasiteswere pretreated with 15 ng of essential oil per ml, the levelof nitric oxide production was 150% higher than that for thecontrol infected macrophages, although no significant differencewas obtained when the parasites were pretreated with 1.5 and0.2 ng of essential oil per ml. When the macrophages were preinfectedwith L. amazonensis for 24 h and then treated with 15 ng ofessential oil per ml, the level of nitric oxide production was60% higher than that for the control infected macrophages.

View larger version (25K):
[in this window]
[in a new window]
FIG. 5. Effects of linalool-rich essential oil extracted from C. cajucara on nitric oxide production by mouse peritoneal macrophages (m ${phi}$ ). The parasites and/or the macrophages were either not treated or treated with 15, 1.5, and 0.2 ng of essential oil per ml 20 min prior to the macrophage-parasite interactions. The adherent cultured macrophages and the free parasites were washed once and resuspended in fresh culture medium. Dead parasites were removed from the medium by centrifugation (1,000 x g, 5 min), and intact living L. amazonensis promastigotes were then added to the macrophage culture plate wells. The supernatants from control and L. amazonensis-infected macrophages were recovered, and the nitrite concentration of each system was determined by the Griess reaction, as described in the Materials and Methods section. Each bar represents the mean ± standard error of at least three independent experiments, which were performed in triplicate. The asterisks indicate that the result is significantly different from that for the infected macrophages when neither the macrophages nor the parasites were treated with essential oil.

DISCUSSION

Top

Discussion

The ability to survive and multiply within macrophages is afeature of several infectious agents including Trypanosoma cruziand Leishmania. In order to sustain a chronic infection, parasitesmust subvert macrophage-accessory cell activities and ablatethe development of protective immunity (2). Nevertheless, themost important mechanism for the killing of Leishmania and thecontrol of leishmaniasis is the production of nitric oxide bymacrophages of draining lymph nodes (11).

New drug therapy regimens have taken advantage of the knowledgeobtained from studies on Leishmania-macrophage interactions(2). Concerning mouse peritoneal macrophage infection with L.amazonensis, when the macrophages were pretreated with C. cajucaraessential oil, as well as when the macrophages were preinfectedwith the parasites and then treated with the essential oil,the association indices were 50% lower than those for the controlsystem (in which control macrophages and control parasites wereused), regardless of the concentration of essential oil usedin those assays (Fig. 4). Therefore, the finding that macrophagespretreated with C. cajucara essential oil produced twice theamount of nitric oxide as the nontreated macrophages is hardlysurprising (Fig. 5). Interestingly, recent studies have demonstratedthat sand fly saliva suppresses macrophage leishmanicidal activity,inhibiting nitric oxide production (10, 22). This activity hasbeen attributed to the sand fly peptide maxadilan, which diminishesthe ability of macrophages to produce nitric oxide and killLeishmania in vitro (11). Cooperation between Leishmania speciesand their vectors is presumably a result of coevolution of thevector and the parasite. The epidemiological consequences ofthis restriction are that the spread of leishmaniases is restrictedby the distribution of suitable vectors (27, 28).

Linalool-rich essential oils extracted from other plants haveantimicrobial properties (35), so we decided to test the effectsof C. cajucara extracts on the growth and viability of L. amazonensis.The LD₅₀s of the essential oil and purified linalool from C.cajucara for both L. amazonensis promastigotes and amastigoteswere very low. It is remarkable that 15.0 ng of essential oilper ml was able to kill 100% of both promastigotes and amastigotesin 60 min (Fig. 2). On the other hand, mouse macrophages wereunaffected by 15.0 ng of essential oil per ml (Fig. 2). Also,the MICs of essential oil purified from Helichrysum italicumfor fungal growth (4) are much higher than the MIC of essentialoil from C. cajucara for L. amazonensis presented here (85.0pg/ml).

Mitochondrial swelling and important alterations in the organizationof the nuclear and kinetoplast chromatins were observed by electronmicroscopy when L. amazonensis parasites were treated for 20to 30 min with 15.0 ng of essential oil from C. cajucara perml (Fig. 3). Although linalool was extremely potent when useddirectly on L. amazonensis parasites, it had little effect whenused in the assays for measurement of the association betweenmacrophages and parasites, as well as assays for measurementof the levels of nitric oxide production by the infected macrophages(data not shown). These data could be explained by synergisticeffects of the different compounds of the plant extracts usedin this work, which is a phenomenon widely known for severalother systems (41). On the other hand, the present work showsthat the decrease in the association between macrophages andparasites, concomitant with the increase in the level of nitricoxide production by the infected macrophages when both celltypes were pretreated with essential oil, was less than thatwhen macrophages alone were pretreated. In this case one couldinfer that linalool and another substance of the essential oilcould present distinct and yet opposite effects on the macrophages.

The results presented in this paper further support the antimicrobialactivity of linalool-rich essential oil. The extreme toxicityof C. cajucara leaf extracts for L. amazonensis, with no effectupon mammalian cells, enables linalool-rich essential oil tobe a source of a new lead compound for novel antileishmanialdrugs.

ACKNOWLEDGMENTS

We thank Rafael Linden and Marcio L. Rodrigues for criticalcomments on the manuscript and Paulo C. Miguel, Jaqueline F.de Amorim, and Barbara C. G. Da Silva for technical assistance.

This work was supported by the Brazilian agencies PRONEX (grant0885), FAPERJ, CNPq, CNPq/UFRJ (PIBIC), FUJB, Embrapa, and FINEP.

FOOTNOTES

* Corresponding author. Mailing address: Universidade Federal do Rio de Janeiro, Instituto de Microbiologia Prof. Paulo de Góes, CCS, Bloco I, Cidade Universitária, Ilha do Fundão, 219491-590, Rio de Janeiro, R.J., Brazil. Phone: 55-21-2562-6711. Fax: 55-21-2560-8344. E-mail: angela.lopes{at}micro.ufrj.br.

REFERENCES

Top