Antifungal efficacy of ten selected South African plant species against Cryptococcus neoformans

Abstract

Fungal infections are a major threat to public health care. Cryptococcosis in humans and animals, caused by Cryptococcus neoformans, is a life-threatening disease. In a random antifungal screening of acetone leaf extracts of 400 tree species against Cryptococcus neoformans, the following plant species had good activity: Zanthoxylum capense (Thunb.) Harv. (Rutaceae), Morusmesozygia Stapf (Moraceae), Calodendrum capense (L.f.) Thunb. (Rutaceae), Catha transvaalensis Codd (Celastraceae), Cussonia zuluensis Strey (Araliaceae), Ochna natalitia (Meisn.) Walp. (Ochnaceae), Croton sylvaticus Hochst. ex C. Krauss (Euphorbiaceae), Maytenus undata (Thunb.) Blakelock (Celastraceae), Celtis africana Burm.f. (Ulmaceae), and Cassine aethiopica Thunb. (Celastraceae). Hexane, dichloromethane, acetone, and methanol extracts of these 10 plants were tested against Cryptococcus neoformans using bioautography and microdilution assays. Acetone extracted the highest quantity of plant material. Dichloromethane and hexane extracts of Maytenus undata showed clear bands in bioautography while the other species did not produce good results in bioautography. Maytenus undata extracts had promising antifungal activity against C. neoformans, with average minimum inhibitory concentration (MIC) of 0.09 mg/mL after 24 h and 0.18 mg/mL after 48 h incubation. Croton sylvaticus and Catha transvaalensis extracts also had good activity, with average MIC values of 0.07 mg/mL and 0.09 mg/mL, respectively. Because of the clear bands on bioautograms and low MIC values compared to the other plant species investigated, M. undata was identified as a good candidate for further studies.

Keywords::

• Antifungal
• bioautography
• Catha transvaalensis
• Cryptococcus neoformans
• fungal infections
• Maytenus undata
• microdilution
• Morus mesozygia
• Zanthoxylum capense

Introduction

Fungal infections cause a major threat to public health care systems (Cos et al., 2006), particularly with the number of immunocompromised and immunosuppressed patients on the rise over the past decade as a result of the human immunodeficiency virus/acquired immunodeficiency syndrome (HIV/AIDS) pandemic. This has resulted in increased numbers of patients contracting opportunistic fungal infections (Diamond, 1991), and a concomitant rise in morbidity and mortality. The non-availability of effective, cheap antifungal drugs for the treatment of systemic fungal infections and toxicity of available drugs such as amphotericin B (Saral, 1991) contribute to a rise in mortality caused by fungal infections.

There are several synthetic and natural product-based drugs available for the treatment of fungal infections, but they are not consistently effective against pathogenic fungal infections (Lazar & Wilner, 1990; Gearhart, 1994; Goa & Barradell, 1995). Excessive use of antibiotics has resulted in the emergence of antimicrobial resistance, with profound effects on human and animal health. These developments and the associated increase in fungal infections (Beck-Sagué & Jarvis, 1993) have intensified the search for new, safer, and more efficacious agents to combat serious fungal infections (Ghannoum & Rice, 1999).

Cryptococcus neoformans causes cryptococcosis in humans and animals. This is a life-threatening disease that develops following inhalation and dissemination of the organism from the lungs to the central nervous system (Uicker et al., 2005). Infection with the pathogenic fungus C. neoformans often results in pneumonia and meningitis in HIV-infected patients (Brol et al., 2002). C. neoformans is the second most important fungus causing disease in AIDS patients, after Candida albicans (Powderly, 1993; Brol et al., 2002).

Medicinal plants have been used for centuries as remedies for human and animal diseases because they contain components of therapeutic value (Nostro et al., 2000), and this healing knowledge is often verbally passed from generation to generation (Van Wyk et al., 1997). Many higher plants produce economically important constituents or compounds such as flavors, fragrances, pharmaceuticals, and pesticides. However, most species of higher plants have never been studied for biologically active constituents, and hence new sources of commercially valuable substances remain to be discovered (Balandrin et al., 1985).

Plants are highly exposed to unfavorable conditions such as extreme weather conditions and fungal, bacterial, parasitic, and pest attacks. Due to these conditions plants have developed mechanisms to overcome stress by producing secondary metabolites which act as a protective system for their survival in harsh environments. Metabolites produced from these plants may present a significant contribution toward the development of new drug leads from plants against fungal, bacterial, parasitic, and HIV infections; hence, plants are a good source of anti-infective agents.

The Phytomedicine Programme at the University of Pretoria has for several years undertaken the systematic screening of acetone leaf extracts of trees against bacteria and fungi of medical importance. A database of close to 500 plant species and their activities against fungi and bacteria has been developed (Phytomedicine Programme at www.up.ac.za/phyto). Plants used in this project were selected from the Phytomedicine Laboratory database based on several aspects such as plants with a low minimum inhibitory concentration (MIC) value and a high total activity (Eloff, 2004) against C. neoformans. The purpose of this article is to evaluate the activity of the extracts from different plant species in order to select the most promising plant species for in depth research.

Materials and methods

Plant material

The plants used for screening were collected from the Lowveld National Botanical Gardens (NBG) during the summer of 2005. These plants were dried, ground to a powder, and stored at room temperature in glass storage containers in the dark in the Phytomedicine Labo-ratory. The plant species chosen were: Zanthoxylum capense (Thunb.) Harv. (Rutaceae), Morus mesozygia Stapf (Moraceae), Calodendrum capense (L.f.) Thunb. (Rutaceae), Catha transvaalensis Codd (Celastraceae), Cussonia zuluensis Strey (Araliaceae), Ochna natalitia (Meisn.) Walp. (Ochnaceae), Croton sylvaticus Hochst. ex C. Krauss (Euphorbiaceae), Maytenus undata (Thunb.) Blakelock (Celastraceae), Celtis africana Burm.f. (Ulmaceae), and Cassine aethiopica Thunb. (Celastraceae). The plants were identified by labels on the trees and by Mr. Rudi Kotze, Curator of the NBG; voucher specimens of trees labeled in the NBG are deposited in the NBG Herbarium.

Extraction

Powdered leaves (4 g) of the 10 plant species were extracted separately with 40 mL of four different solvents of increasing polarity (hexane, dichloromethane, acetone, and methanol) in 50 mL centrifuge tubes. The tubes were shaken on a Labotec shaking machine for 1 h. The extracts were centrifuged at 300 × g for 10 min and the supernatant was filtered through Whatman No. 1 filter paper into pre-weighed glass vials and placed under a stream of cold air to dry.

Thin layer chromatography (TLC) fingerprinting of the extracts

Plant extracts were dissolved in acetone to a concentration of 10 mg/mL. Aliquots of 10 µL (100 µg) were loaded on each of three aluminum-backed TLC plates (silica gel 60 F₂₅₄; Merck) and eluted in three mobile systems of differing polarity developed in the Phytomedicine Laboratory, University of Pretoria (Kotzé & Eloff, 2002). The TLC systems used were as follows:

• Benzene:ethanol:ammonia (18:2:0.2 ) (BEA, non polar);

• Chloroform:ethyl acetate:formic acid (10:9:2) (CEF, intermediate polarity);

• Ethyl acetate:methanol:water (10:1.35:1) (EMW, polar).

The developed TLC plates were evaluated under ultraviolet (UV) light at 254 and 365 nm to detect UV active absorbing spots/plant constituents. The plates were then sprayed with vanillin spray reagent (0.1 g vanillin dissolved in 28 mL methanol, with the addition of 1 mL sulfuric acid) and heated at 110°C to optimal color development.

Fungal cultures

Fungal cultures of Cryptococcus neoformans isolated from an untreated cheetah were obtained from the Microbiology Laboratory collection (Department of Veterinary Tropical Diseases, Faculty of Veterinary Science, University of Pretoria). The cultures were maintained on Sabouraud dextrose (SD) agar at 4°C and were inoculated in SD broth at 37°C and incubated prior to conducting bioautography and microdilution assays. The fungal inoculum (adjusted with fresh medium to approximately 2 × 10⁶ cells/mL) used during microdilution and bioautography assays was quantified using a hemocytometer.

Bioautographic assay of the extracts

Replicate chromatograms were prepared as described above without spraying with the vanillin spay reagent. The plates were left uncovered in a dark place for several days in a stream of air to allow the eluting solvent to evaporate completely from the plates before being sprayed with a day-old actively growing suspension of Cryptococcus neoformans. The TLC plates were then incubated for 24 h at 37°C under 100% relative humidity to allow the microorganism to grow on the plates. After overnight incubation the bioautograms were sprayed with an aqueous solution of 2 mg/ mL p-iodonitrotetrazolium violet (INT; Sigma) and incubated for 30 min for observation of clear zones indicating inhibition of fungal growth by bioactive compounds in the extract. The bioautography procedure for fungi is described in more detail by Masoko and Eloff (2006).

A set of chromatograms sprayed with vanillin was used as reference chromatograms for the bioautography plates. The R_f values of active zones were correlated with those of bands on the reference chromatograms.

Microdilution assay

The two-fold serial dilution microplate method of Eloff (1998a) was used to determine the MIC values of plant extracts. This method has been used to evaluate antibacterial activities of plant extracts (McGaw et al., 2001), and Masoko et al. (2005) have modified the method for evaluating antifungal activity. Residues of plant extracts were resuspended into acetone to a concentration of 10 mg/mL. The plant extracts (100 µL) in triplicate for each experiment were serially diluted two-fold with water in 96-well microtiter plates. A 100 µL aliquot of C. neoformans suspension (2 × 10⁶ cells/mL) was added to each well. Acetone was used as a solvent control and distilled water was used as a negative control. Amphotericin B (160 µg/mL initial concentration) was the positive control. As an indicator of growth, 40 µL of 0.2 mg/mL of INT dissolved in water was added to each well, and the covered microtiter plates were incubated at 35°C overnight to ensure adequate color development. The MIC was recorded as the lowest concentration of extract that inhibited fungal growth. The colorless tetrazolium salt acts as an electron acceptor and is reduced to a formazan product by biologically active organisms (Eloff, 1998a). The MIC values were read after 24 and 48 h. Where fungal growth is completely halted (lethal concentration), the solution in the well remains clear after incubation with INT, but inhibition of growth is measured as the first concentration of plant extract that causes a decrease in color intensity of the formazan salt (inhibitory concentration). The experiment was repeated three times to confirm the results.

Total activity

Eloff (2000) mentioned that not only MIC but also the quantity extracted should be taken into account to compare the activity of different plant extracts. Total activity, reported in units of mL/g, indicates the degree to which the active compounds in 1 g of plant material can be diluted and still inhibit the growth of the tested fungal microorganism. This takes into account the quantity extracted from plant material and is calculated as follows:

The higher the total activity of a plant extract, the more effective the original plant (Eloff, 2000). If the total activity is calculated at each step of a bioassay-guided fractionation procedure it is easy to determine whether there is a loss of biological activity during isolation, and also synergistic effects can be discovered. This situation is equivalent to the terms efficacy and potency used in pharmacology (Eloff, 2004).

Results and discussion

Quantity extracted

Figure 1 indicates the extracting efficiency of the different solvent systems against selected plant species. The acetone extract of Catha transvaalensis (species 4) produced the highest amount of extract of all the plant species (14.6%), and Zanthoxylum capense (species 1) the lowest (1.15%). C. transvaalensis (species 4) and Cassine aethiopica (species 10) were the most extracted plant species, with average extraction efficiencies of 5.57 and 4.62%, respectively. The least extracted plant species were Morus mesozygia (species 2) and Z. capense (species 1), with average extraction efficiencies of 1.69 and 0.97%. This clearly shows that plant metabolites are selectively extracted by different solvents. This depends on the nature or type and polarity of a metabolite and a solvent system used. It is expected that non-polar constituents are extracted by hexane and dichloromethane (DCM) and highly polar constituents are extracted by methanol (MeOH). Acetone extracts compounds of a wide range of polarity, and is miscible with aqueous and organic solvents (Eloff, 1998b). Acetone, with an average extracting efficiency of 5.44%, was the best extracting solvent for the 10 plant species amongst the four solvents used because it generally extracted the highest quantity of plant material compared to the other solvents, followed by MeOH with 3.39%, hexane with 0.68%, and DCM (0.59%) extracting the lowest quantity. It is evident from Figure 1 that the non-polar solvents (hexane and DCM) extracted the lowest quantity of plant constituents in relation to the more polar solvents (acetone and MeOH).

Figure 1.  Quantity of extracts of 10 plant species (1, Zanthoxylum capense; 2, Morus mesozygia; 3, Calodendrum capense; 4, Catha transvaalensis; 5, Cussonia zuluensis; 6, Ochna natalitia; 7, Croton sylvaticus; 8, Maytenus undata; 9, Celtis africana; 10, Cassine aethiopica) extracted using different solvents. Hex, hexane; DCM, dichloromethane; ACE, acetone; MeOH, methanol.

TLC fingerprinting

Figure 2 indicates the separation or distribution of extracted plant constituents using TLC. BEA was the best mobile phase system for separation of almost all the extracts, compared to CEF and EMW, because many compounds were distributed, or separated, efficiently using BEA (Figure 2). This indicated that these species have a variety of non-polar compounds present in a high concentration. It is expected that the more polar extractants would extract few of the very non-polar compounds. In general, the separations of the extracted compounds in both the CEF and EMW systems were not as efficient, because the majority of the extracted compounds were too non-polar for these solvent systems and moved to the top of the TLC plate, making it difficult to count the number of extracted compounds visible after spraying with vanillin sulfuric acid spray reagent (Figure 2). Kotzé and Eloff (2002) obtained similar results after extracting Combretum microphyllum with different solvents of differing polarities.

Figure 2.  TLC chromatograms of 10 plant species (left to right: Zanthoxylum capense (Zc), Morus mesozygia (Mm), Calodendrum capense (Cc), Catha transvaalensis (Ct), Cussonia zuluensis (Cz), Ochna natalitia (On), Croton sylvaticus (Cs), Maytenus undata (Mu), Celtis africana (Ca), and Cassine aethiopica (Ce) extracted with hexane, DCM, acetone, and methanol (left to right), developed in BEA, CEF, and EMW (top to bottom), sprayed with vanillin sulfuric acid in methanol.

Several compounds were visualized when chromatograms were sprayed with vanillin sulfuric acid. Using BEA as eluent and also investigating chromatograms under UV light prior to spraying, the total numbers of compounds in all the extracts were compared. The highest number of bands on the TLC plate was visible in the DCM extract, followed by the hexane and acetone extracts, with the methanol extract showing the least number of bands. This indicated that most of the compounds present were relatively non-polar and soluble in non-polar extractants. The compounds were separated well by the non-polar BEA eluent. There were fewer polar compounds present when using the more polar eluents CEF and EMW. Both CEF and EMW chromatograms of the acetone and methanol extracts of Catha transvaalensis had yellowish polar compounds with low R_f values.

Bioautography of plant extracts

Figure 3 indicates bioautograms of 10 plant extracts against C. neoformans. The activity of plant extracts against C. neoformans is indicated by the presence of clear bands/spots against a reddish background on a TLC plate. Actively growing microorganisms on a TLC plate reduce the colorless p-iodonitrotetrazolium violet (INT) to a formazan product. Maytenus undata showed a wide clear band on the bioautograms of the DCM and hexane extracts developed in BEA and EMW, respectively. Other plant extracts did not show any clear spots/bands against C. neoformans on bioautograms. These compounds were extracted by more non-polar solvents such as hexane and DCM and not by acetone or methanol, implying that the nature of the active constituents tends more toward non-polarity. Acetone extracted a high quantity of plant constituents from M. undata compared to both hexane and DCM; however, this extract did not show any zones of fungal growth inhibition. The TLC bioautogram method is a useful method in identify the number of biologically active plant metabolites in an extract for microorganisms that are able to grow on a TLC plate. We have found from time to time that extracts with good antimicrobial activity do not yield visible zones of inhibition. This may be explained if some plant constituents act synergistically to inhibit microbial growth in the crude extract but are not active against the target organisms when separated by TLC. Another explanation for this observation may be that the active compounds are volatile and may evaporate from the plate during the process of removing the TLC solvents. A third possibility is that the separated compounds may be chemically altered by e.g., photo-oxidation on the TLC plate, and therefore lose activity.

Figure 3.  TLC bioautogram of 10 plant species (from left to right: Zanthoxylum capense (Zc), Morus mesozygia (Mm), Calodendrum capense (Cc), Catha transvaalensis (Ct), Cussonia zuluensis (Cz), Ochna natalitia (On), Croton sylvaticus (Cs), Maytenus undata (Mu), Celtis africana (Ca), and Cassine aethiopica (Ce) extracted with DCM or hexane, developed in BEA and EMW and sprayed with Cryptococcus neoformans. Clear zones on the bioautogram indicate fungal growth inhibition.

Microdilution assay

Table 1 presents the minimum inhibitory concentration (MIC) values and the total activity of the 10 plant species extracted with four different solvents against Cryptococcus neoformans. Tetrazolium violet (INT) was used as an indicator for microbial growth. The biologically active microorganism reduces the colorless INT to a reddish formazan product. The MIC value is recorded as the first concentration of plant extract that causes a decrease in color intensity of solution in a well (Eloff, 1998a). This table shows that the Maytenus undata extracts had very low MIC values compared to those of the other plant species. Also, what is evident from the table is the high total activity of M. undata extracts, especially the methanol extract with 2050 mL/g and DCM extract with 550 mL/g, respectively, after 24 h incubation with the test organism. This shows that 1 g of M. undata methanol and DCM extracts can be diluted with 2050 mL or 550 mL, respectively, and still inhibit C. neoformans growth. After 48 h, M. undata DCM and MeOH extracts showed a moderate total activity of 92 mL/g and 128 mL/g, respectively. This indicates that the inhibition was fungistatic and that C. neoformans was able to overcome the activity of these extracts after 24 h.

Table 1.  Quantity extracted from 4 g of plant material, average MIC values, and total activity of 10 plant species against C. neoformans

Plant species	Quantity extracted (mg)				Incubation period 24 h								Incubation period 48 h
					MIC value (mg/mL)				Total activity (mL/g)				MIC value (mg/mL)				Total activity (mL/g)
	HEX	DCM	ACE	MeOH	HEX	DCM	ACE	MeOH	HEX	DCM	ACE	MeOH	HEX	DCM	ACE	MeOH	HEX	DCM	ACE	MeOH
Z. capense	9	15	46	83	0.04	0.02	0.08	0.32	56	188	144	65	0.64	0.18	0.08	0.945	4	21	144	22
M. mesozygia	13	11	111	122	0.08	0.04	0.12	0.24	41	69	232	127	0.64	0.12	0.16	0.64	5	123	174	48
C. capense	39	38	190	104	0.08	0.02	0.03	0.32	122	475	1583	81	0.64	0.20	0.04	0.64	15	48	1188	41
C. transvaalensis	69	36	584	264	0.02	0.04	0.12	0.16	865	225	1217	413	0.64	0.24	0.06	0.48	27	38	2433	138
C. zuluensis	24	31	238	103	0.16	0.02	0.16	0.32	38	388	372	81	0.64	0.20	0.12	0.64	9	39	496	40
O. natalitia	9	5	155	152	0.16	0.02	0.08	0.32	14	63	485	119	1.25	0.24	0.08	0.32	2	5	485	119
C. sylvaticus	13	20	137	89	0.02	0.02	0.12	0.12	163	250	286	186	0.24	0.24	0.64	0.32	14	21	54	70
M. undata	50	44	305	164	0.08	0.02	0.24	0.02	156	550	318	2050	0.16	0.12	0.12	0.32	78	92	636	128
C. africana	6	7	182	41	0.32	0.02	0.12	0.40	5	90	379	26	0.32	0.36	0.06	0.08	5	5	758	128
C. aethiopica	38	28	225	233	0.04	0.02	0.12	0.40	238	350	469	146	0.64	0.48	0.12	1.25	15	15	469	47

HEX, hexane; DCM, dichloromethane; ACE, acetone; MeOH, methanol.

M. undata had an average MIC value of 0.09 mg/ mL after 24 h and 0.18 mg/mL after 48 h of incubation (Table 2). Also, some of the plant species such as C. sylvaticus and C. transvaalensis indicated lower/equal average MIC values (0.07 mg/mL and 0.09 mg/ mL, respectively) compared with those of M. undata, but their activity was not detected using bioautography (Figure 3). M. undata had the highest average total activity of 769 mL/g after 24 h. After 48 h both C. transvaalensis and C. capense showed higher average total activities of 659 mL/g and 323 mL/g, respectively, than M. undata with an average total activity of 234 mL/g (Table 2). In general, this result indicates that, after 48 h, C. transvaalensis extracts were almost three times more active than M. undata extracts against C. neoformans, and these extracts may contain fungicidal rather than fungistatic compounds.

Table 2.  Average MIC values and total activity of C. neoformans against the different plant extracts. Each value represents the average of three replicates

Plant species	24 h		48 h
	MIC value (mg/mL)	Total activity (mL/g)	MIC value (mg/mL)	Total activity (mL/g)
Z. capense	0.15	113	0.46	48
M. mesozygia	0.12	117	0.39	88
C. capense	0.11	565	0.38	323
C. transvaalensis	0.09	680	0.36	659
C. zuluensis	0.17	220	0.40	146
O. natalitia	0.15	170	0.47	153
C. sylvaticus	0.07	221	0.36	41
M. undata	0.09	769	0.18	234
C. africana	0.22	125	0.21	224
C. aethiopica	0.15	300	0.62	137

The MIC method is more trustworthy than the bioautographic method because plant constituents are dissolved, kept in solution, and not separated from each other, allowing all the possible plant constituents in an extract to be in full contact with the microorganism. This allows the detection of a possible synergistic effect in a plant extract. Like any other technique or method there are some limitations associated with this method. One of those limitations is precipitation of non-aqueous plant constituents in a well containing aqueous medium, making it impossible for the plant constituents to fully interact with microorganisms; hence, microbial growth might not be inhibited, making the scoring of the MIC values difficult. Another limiting factor is that it is not possible to identify active constituents from a plant extract indicating activity against a microorganism. The color of the plant extract may make it difficult to read the MIC values, resulting in false positive results. Even if this method has some limitations it is considered to be an effective method for the analysis of antimicrobial activity of plant extracts because it is possible to detect the loss or gain of activity.

Conclusions

Different solvents extract different types of plant constituents owing to their varying polarities. This feature enables the extraction of plant constituents to be more selective. In this study, bioautography (Figure 3) indicated heightened fungal growth inhibition by Maytenus undata extracts, especially the hexane and DCM extracts, compared to extracts of the remaining nine plant species screened. However, this does not mean that other plant species are not active against C. neoformans. Table 2 revealed that the lowest average MIC value was observed with the M. undata extracts. M. undata showed a well-defined fungal growth inhibition against Cryptococcus neoformans on TLC plates in bioautography, and good antifungal activity in the microdilution assay. Because active constituents from other plant species were not easily visible in bioautography, the plant constituents responsible for low MIC values observed during the microdilution assay against C. neoformans could not be localized on the TLC plates. A means of avoiding this problem would be to use more volatile solvents as TLC eluents that would evaporate quickly from the TLC plate. Decreased exposure to oxygen and light may reduce chemical changes or volatilization of potentially active compounds that are exposed on TLC plates.

The M. undata extracts indicated the presence of antifungal constituents using bioautography, and low MIC values in the microdilution assay against C. neoformans. This facilitates the identification and isolation of active constituents from M. undata. Current research is concentrating on identifying the active constituents in M. undata.

Acknowledgements

The Curator of the Lowveld National Botanical Gardens in Nelspruit, Mr Rudi Kotze, kindly allowed collection of plant material. One of the authors (L.J.M.) is grateful to the Claude Leon Foundation for a Postdoctoral Fellowship.

Declaration of interest

The National Research Foundation (NRF) and University of Pretoria (UP) provided funding.