Comparison between efficacy of allicin and fluconazole against Candida albicans in vitro and in a systemic candidiasis mouse model

Authors


  • Editor: Matthias Brock

Correspondence: Pei Pei Chong, Department of Biomedical Sciences, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia. Tel.: +60 894 723 02; fax: +60 894 361 78; e-mail: cpp@medic.upm.edu.my

Abstract

The efficacy of allicin compared with fluconazole in alleviating systemic Candida albicans infections was evaluated both in vitro and in vivo through a systemic candidiasis mouse model. Determination of in vitro minimum inhibitory concentrations (MICs) for different C. albicans isolates revealed that both allicin and fluconazole showed different MICs that ranged from 0.05 to 12.5 μg mL−1 and 0.25 to 16 μg mL−1, respectively. A time–kill study showed a significant effect of allicin (P<0.01) against C. albicans, comparable to that of fluconazole. Scanning electron microscopy observation revealed that, similar to fluconazole, allicin produced structural destruction of C. albicans cell surface at low MIC and lysis or puncture at high MIC concentrations. Treatment of BALB/c mice systemically infected with C. albicans showed that although the allicin treatment (at 5 mg kg−1 day−1) was slightly less efficacious than fluconazole treatment in terms of the fungal load reduction and host survival time, it was still effective against C. albicans in terms of mean survival time, which increased from 8.4 to 15.8 days. These results demonstrate the efficacy of anticandidal effects of allicin both in vitro and in an animal model of candidiasis and affirm the potential of allicin as an adjuvant therapy to fluconazole.

Introduction

Recently, the incidence of systemic candidiasis, which is caused by Candida spp., predominantly Candida albicans, has increased (Chowta et al., 2007). This increase over the last two decades has caused a rise in the use of antifungal drugs (Pereira-Cenci et al., 2008). Azoles such as fluconazole or ketoconazole are usually used for treatment of systemic fungal infections. However, one of the biggest problems faced in clinical practice is the emergence of resistance to most of these azole drugs due to mutation (Odds et al., 2003; Looi et al., 2005). Clinically adverse effects are also seen with the use of azoles (Al-Mohsen & Hughes, 1998). Therefore the most urgent challenge in pharmaceutical research is the discovery and development of new antifungals from plant and microbial sources.

Allicin (diallyl thiosulfinate), one of the sulfur compounds from garlic, has been shown to possess antifungal activity (Yamada & Azuma, 1977). It has been shown that after crushing fresh garlic cloves, allinase rapidly converts the released allin (precursor of allicin) into allicin (Ankri & Mirelman, 1999). Allitridium (diallyl trisulfide), one of the breakdown products from allicin, has also been found to show antifungal activity in vitro (Davis et al., 2003) and in vivo (Davis et al., 1990). Ajoene, one of the other products from allicin, has been shown to have antifungal and antiparasitic properties (Ledezma et al., 2008). Potent antifungal activity of allicin against major species of Candida in vitro has been reported in our previous work (Khodavandi et al., 2010) but, thus far, there have been very few studies investigating the activity of allicin as an anticandidal agent in vivo. In the present study, the anticandidal activity of pure allicin demonstrated its strong potential activity both in vitro and in vivo.

Materials and methods

Antifungal agents

Allicin was acquired from Alexis Biochemicals Co. (purity ≥98%, Batch No. ALX-350-329, San Diego, CA) and dissolved at a concentration of 10 mg mL−1 in a mixture of methanol, water and formic acid (60 : 40 : 0.1), and then stored at −20 to −80 °C until use. Fluconazole was purchased from Sigma Chemicals Co. (St. Louis, MO). The stock solution was prepared by dissolving in dimethyl sulfoxide at 5 mg mL−1. The stock solutions were stored frozen at −70 °C until use. For the in vitro study, allicin and fluconazole drug dilutions ranged from 0.05 to 25 μg mL−1 and 0.03 to 64 μg mL−1, respectively, but for the in vivo work, the two dosages of these drugs were 1 and 5 mg kg−1 day−1, respectively.

In vitro assays

The in vitro efficacy of allicin and fluconazole was measured against C. albicans ATCC 14053 and nine clinical isolates obtained from the diagnostic microbiological laboratories of the University of Malaya Medical Center, Kajang Hospital, Selangor and Seremban Hospital, Negeri Sembilan by the broth microdilution method (NCCLS M27 A2). The inhibitory effect of antifungal agents at different time intervals was then studied (C. albicans ATCC 14053). All clinical samples were isolated from patients with systemic candidiasis.

The lowest concentrations of antifungals that can inhibit 50% and 90% of Candida growth (minimum inhibitory concentrations, MIC50 and MIC90, respectively) compared with untreated growth control were determined as described previously (Khodavandi et al., 2010). In summary, 100 μL of antifungal agents in different concentrations (in standard RPMI 1640 medium with 0.2% glucose buffered to pH 7.0 with 0.165 M morpholinophosphonyl sulfate) were inoculated with 100 μL of inoculum containing between 5 × 102 and 2.5 × 103 yeast cells mL−1 in a 96-well microplate (Brand 781660, Wertheim, Germany). The microplates, including drugs and cells, were incubated at 35 °C and MICs were measured at 530 nm from two independent experiments in three separate technical replicates using an EMax® microplate reader after 24 h.

For investigation of the inhibitory effect of allicin and fluconazole on growth of Candida, two different inoculum sizes of C. albicans ATCC 14053 (1 × 104 and 1 × 106 cells mL−1 in RPMI 1640 as described before) were grown in the presence of 0.1 μg mL−1 of allicin and 2 μg mL−1 of fluconazole based on MIC concentration. After 0-, 2-, 4-, 6-, 8-, 12- and 24-h incubation at 35 °C, 100 μL of this solution was collected and 10-fold serial dilutions were made and plated on Sabouraud dextrose agar (Difco Laboratories, Detroit, MI). Colonies were counted after 24-h incubation at 37 °C and the number of CFU was calculated.

Scanning electron microscopy for morphological observation of Candida after treatment with allicin

A suspension of C. albicans ATCC 14053 containing 1 × 106 cells mL−1 in RPMI 1640 was mixed with different dilutions of allicin and fluconazole (1/2 × MIC, 1 × MIC and 10 × MIC) and incubated at 35 °C for 24 h. Cells were fixed in 2% v/v glutaraldehyde in phosphate-buffered saline (pH 7.2) and washed with sodium cacodylate buffer. For postfixation, samples were rinsed in 1% osmium tetroxide for 2 h at 4 °C, washed again with sodium cacodylate buffer and then dehydrated with ascending ethanol series. After that, samples on coverslips were put into a critical point dryer and then stuck onto the stub. The specimens were coated with gold and observed through a JEOL JSM 6400 scanning electron microscope.

In vivo assays

For the experiments on the animal model of systemic candidiasis, 4–6-week-old female BALB/c mice were infected intravenously through the tail vein with 200 μL per mouse of C. albicans ATCC 14053 (5 × 106 yeast cells mL−1). The mice were divided into five experimental groups of 12 mice each. In the first two groups of mice, allicin (200 μL per mouse) was administered intravenously once daily for 5 days beginning 1 h after Candida injection (postinfection) at 1 and 5 mg kg−1 day−1, respectively (Shadkchan et al., 2004). For the third and fourth groups of mice, fluconazole (200 μL per mouse) was administered via the intraperitoneal route once daily for 5 days starting 1-h postinfection at 1 and 5 mg kg−1 day−1, respectively (Rex et al., 1998). For the untreated control group, 200 μL of normal saline was injected into each mouse at 1-h postinfection.

The infection was followed up for 28 days and evaluated in terms of mortality and morbidity. For studies of tissue burden, two randomly chosen mice were sacrificed from each experimental group on days 2, 4, 7, 10, 14 and 28 after infection (Shadkchan et al., 2004). Kidneys, liver and spleen from each mouse were aseptically removed and homogenized in 1 mL of sterile normal saline and cultured on Sabouraud dextrose agar plates as explained in the time–kill study, and assessed by determination of fungal colonization of viscera.

Histopathologic analyses were performed for a qualitative confirmation of the result. Tissues were fixed in 10% formalin, then blocked by paraffin wax and cut with a microtome (Leica, model RM2025) in 4-μm thickness. Hematoxylin and eosin and periodic acid-Schiff staining were used to observe the tissue and presence of fungal elements.

All animal care procedures were supervised and approved by the University of Putra Malaysia Animal Ethics Committee (ACUC NO.: UPM/FPSK/PADS/BRUUH/00278).

Statistical analysis

For quantitative statistical analysis of inhibitory effects of drugs in vitro and also reduction of fungal load in tissues of mice, data were analyzed in terms of normality and one-way anova was carried out. Moreover, a log rank test was used to determine the survival time. P values of <0.05 were considered significant. Statistical analysis was performed using spss version 17 software (SPSS Inc., Chicago, IL). All experiments were carried out at least in triplicate.

Results

Table 1 shows the MICs of allicin and fluconazole against C. albicans ATCC 14053 and some clinical isolates. The results are representative of two independent experiments arranged in triplicate. The MIC50 and MIC90 of these isolates ranged from 0.05 to 0.78 μg mL−1 and 0.1 to 12.5 μg mL−1, respectively for allicin, and from 0.25 to 4 μg mL−1 and 2 to 16 μg mL−1, respectively, for fluconazole. All samples were sensitive to fluconazole and drug resistance was not seen.

Table 1.   Interaction of allicin and fluconazole against clinical and standard isolates of Candida albicans by broth microdilution assay after 24-h incubation at 35°C
 AllicinFluconazole
Isolates*MIC50/MIC90MIC rangeMIC50/MIC90MIC range
  • *

    Strains with numbers only without the ATCC prefix are clinical isolates from local hospitals.

  • MIC values are μg mL−1 from three independent experiments. The MIC values of strains ATCC 14053 and 3092 were determined previously and shown in our previous publication (Khodavandi et al., 2010).

C. albicans ATCC 140530.05/0.10.025–0.20.5/20.25–2
C. albicans 30920.39/0.780.39–1.560.5/40.5–8
C. albicans 30720.1/0.780.05–0.780.25/20.25–8
C. albicans 31091.56/3.131.56–6.252/162–16
C. albicans 30711.56/6.250.1–6.251/40.5–8
C. albicans 30640.78/3.130.78–3.132/82–16
C. albicans 30550.39/3.130.39–6.254/164–16
C. albicans 01633.13/12.53.13–254/82–16
C. albicans 31510.2/3.130.1–6.250.5/20.5–8
C. albicans 26050.39/6.250.39–12.51/80.25–16

The potency of allicin and fluconazole in decreasing the cell number of C. albicans ATCC 14053 after 0, 2, 4, 6, 8, 12 and 24 h was significant compared with the control growth (Fig. 1). Figure 1a and b indicate the inhibitory effect of allicin and fluconazole on different inoculum sizes of C. albicans. The significant reduction of Candida treated with allicin and fluconazole started after 4-h incubation (P<0.01) in comparison to untreated control for both inoculum sizes (Fig. 1).

Figure 1.

 Time–kill curves of allicin (0.1 μg mL−1) and fluconazole (2 μg mL−1) against Candida albicans ATCC 14053 at different time intervals: (a) 1 × 106 Candida mL−1; (b) 1 × 104 Candida mL−1. Antifungals: (□) allicin, (▴) fluconazole and (•) untreated control.

Candida albicans cells grown in RPMI 1640 medium at 35 °C showed typical yeast cells with a smooth surface after 24 h, but cells treated with increasing concentration of allicin or fluconazole displayed changes in surface morphology, with the cell surface becoming rough and irregular. According to Lemar et al. (2005) the main reason for this phenomenon could be a decreased cytoplasmic volume. It was also observed in the present study that higher concentrations of the antifungal agents (such as 10 × MIC) destroyed the cell surface, inducing puncture in allicin-treated samples and causing cell lysis in fluconazole-treated samples (Fig. 2).

Figure 2.

 Scanning electron micrograph of Candida albicans ATCC 14053 treated with allicin (a) or fluconazole (b) in different concentrations based on MICs: (i) 1/2 × MIC; (ii) 1 × MIC; (iii) 10 × MIC; and (iv) untreated control.

The results of fungal load determination in the liver, kidney and spleen at different time points indicated a significant reduction of CFU g−1 of the tissue (P<0.001) starting from the second day postinfection for different dosages of the antifungals. In addition, the reduction of Candida cells CFU in tissues after 28 days postinfection ranked from 5 mg kg−1 day−1 fluconazole >1 mg kg−1 day−1 fluconazole >5 mg kg−1 day−1 allicin >1 mg kg−1 day−1 allicin (Table 2).

Table 2.   Tissue fungal loads obtained from mice infected with Candida albicans ATCC 14053 and treated with allicin, fluconazole or untreated
Day after infectionAllicin*AllicinFluconazole*FluconazoleUntreated control
  • Results in a row with different superscripts differ significantly (P<0.05) using Duncan's test.

  • *

    1 mg kg−1 day−1.

  • 5 mg kg−1 day−1.

Log10 CFU g−1 of liver ± SD
 23.82 ± 0.41a3.82 ± 0.41a3.64 ± 0.18a3.64 ± 0.36a4.13 ± 0.12b
 43.77 ± 0.18b3.65 ± 0.38b3.74 ± 0.04b3.37 ± 0.27a4.18 ± 0.09c
 73.79 ± 0.03d3.16 ± 0.42a3.60 ± 0.01c3.36 ± 0.07b4.45 ± 0.07e
 103.77 ± 0.05d3.36 ± 0.07c3.26 ± 0.09b3.00 ± 0.05a5.47 ± 0.05e
 143.67 ± 0.09d3.37 ± 0.07c2.90 ± 0.10b2.53 ± 0.51a6.65 ± 0.03e
 283.62 ± 0.06d3.09 ± 0.07c2.85 ± 0.05b2.16 ± 0.20a
Log10 CFU g−1 of spleen ± SD
 23.88 ± 0.17b3.66 ± 0.40a3.76 ± 0.04ab3.73 ± 0.15ab3.91 ± 0.09b
 43.73 ± 0.24c3.23 ± 0.46a3.63 ± 0.08bc3.42 ± 0.20ab4.01 ± 0.05d
 73.83 ± 0.03c3.34 ± 0.23b3.29 ± 0.03b3.15 ± 0.19a4.30 ± 0.05d
 103.46 ± 0.08d3.27 ± 0.10c3.03 ± 0.09b2.82 ± 0.42a5.89 ± 0.08e
 143.39 ± 0.07d3.27 ± 0.08c2.88 ± 0.10b2.70 ± 0.18a6.33 ± 0.05e
 283.35 ± 0.05c3.26 ± 0.10c2.80 ± 0.04b2.49 ± 0.21a
Log10 CFU g−1 of kidney ± SD
 23.86 ± 0.28a3.74 ± 0.36a3.80 ± 0.32a3.67 ± 0.05a4.09 ± 0.23b
 43.90 ± 0.06c3.71 ± 0.24ab3.75 ± 0.03b3.60 ± 0.13a4.28 ± 0.06d
 73.46 ± 0.23b3.12 ± 0.20a3.74 ± 0.02c3.65 ± 0.05c5.10 ± 0.04d
 103.24 ± 0.15a3.15 ± 0.14a3.24 ± 0.10a3.17 ± 0.11a5.11 ± 0.04b
 143.19 ± 0.08c3.08 ± 0.06b2.74 ± 0.04a2.77 ± 0.10a6.33 ± 0.05d
 283.19 ± 0.10d3.08 ± 0.04c2.68 ± 0.10b2.03 ± 0.13a

As described before, the mortality and morbidity of the treated mice were evaluated for 28 days postinfection. Table 3 also shows the mean survival time (MST) of mice treated with different drugs. Moreover, based on statistical analysis of log rank=13.449 in this study, comparison of the mean of survival time between treated and control groups indicated significant differences (P<0.05) (Fig. 3).

Table 3.   The survival time in different groups of treated mice
Mice-treated groupsMST ± SE (days)*Mortality after 28 days (%)
  • *

    The results are representative of three independent experiments for 12 mice in each group.

  • P value for treated groups compared with untreated control.

Allicin (1 mg kg−1 day−1)13.900 ± 3.97380 (P=0.163)
Allicin (5 mg kg−1 day−1)15.800 ± 3.88350 (P=0.067)
Fluconazole (1 mg kg−1 day−1)18.300 ± 3.78740 (P<0.05)
Fluconazole (5 mg kg−1 day−1)22.900 ± 3.22620 (P<0.005)
Saline treated8.400 ± 2.077100
Figure 3.

 Cumulative mortality of mice (n=10) infected with Candida albicans ATCC 14053 and treated with allicin at 1 and 5 mg kg−1 day−1 (P=0.163, 0.067, respectively) and fluconazole at 1 and 5 mg kg−1 day−1 (P<0.01). Mice treated with different dosages of antifungal agents: (▴) fluconazole, 5 mg kg−1 day−1; (▾) fluconazole, 1 mg kg−1 day−1; (▪) allicin, 5 mg kg−1 day−1; (◆) allicin, 1 mg kg−1 day−1; (•) untreated control.

Discussion

Previous reports have demonstrated the antifungal activity of allicin in vitro against Aspergillus, Trichophyton and Candida spp. (Yamada & Azuma, 1977; Aala et al., 2010). On the other hand, the antifungal potential of allicin against Aspergillus spp. was presented by Shadkchan et al. (2004) using a mouse model of systemic aspergillosis. In this study, we attempted to investigate the potency of allicin against C. albicans, the predominant fungal species isolated from human infections.

Allicin alone could exhibit antifungal activity, and when used in synergy with antimicrobial agents, it increased the efficacy of the therapeutic agents (Aala et al., 2010; Khodavandi et al., 2010). For example, combination of allicin with amphotericin B and fluconazole has been demonstrated to have a significant synergistic effect in a mouse model of systemic candidiasis (An et al., 2009; Guo et al., 2010). Garlic and some of its derivatives destroy the Candida cell membrane integrity (Low et al., 2008), inhibit growth (Lemar et al., 2002) and produce oxidative stress (Lemar et al., 2005) in C. albicans. Most of these abilities are related to an SH-modifying potential, because the activated disulfide bond of allicin has an effect on thiol-containing compounds such as some proteins; however, the main targets of allicin on Candida are not well understood. It has been demonstrated that the antifungal activity of allicin in vivo may be related to some secondary metabolites such as ajoene, diallyl trisulfide and diallyl disulfide, because the chemical structure of allicin is too unstable and converts to these secondary products immediately (Miron et al., 2004). Nonetheless, little is known about the potential in vivo activity of allicin against Candida.

In this study, we used fluconazole as the standard anticandidal drug for comparison against allicin. The MICs of allicin and fluconazole against C. albicans fell within the ranges 0.05–12 and 0.25–16 μg mL−1, respectively (Table 1), which is similar to findings from previous reports (Ankri & Mirelman, 1999; Khodavandi et al., 2010). All of the samples were sensitive to fluconazole and drug resistance was not seen.

The time–kill study demonstrated a significant inhibition of Candida growth comparing untreated controls against those treated with allicin and fluconazole, using inoculum sizes of 1 × 106 Candida cells mL−1 (P<0.05) and 1 × 104 Candida cells mL−1 (P<0.001) after 2- and 4-h incubation, respectively. This demonstrates that allicin decreased the growth of C. albicans almost as efficiently as fluconazole (P>0.05) for both inoculum sizes of Candida, demonstrating a comparable ability to inhibit the growth of the yeast cells (Fig. 1). The presence of pits on the cell surface and cellular collapse with high concentrations of allicin indicates that the cell membrane could be one of the targets of allicin in Candida (Lemar et al., 2002), whereas fluconazole in high concentrations can destroy the Candida cell entirely (Fig. 2).

It has been reported that the use of high doses of organosulfur compounds from garlic, such as diallyl sulfide, diallyl trisulfide, allyl methyl sulfide, allyl methyl trisulfide and dipropyl sulfide, predispose to the development of preneoplastic lesions in rat livers (Lee & Park, 2003). It is also important to note that allicin could be toxic for mammalian cells in high concentrations (>60 μg mL−1), but the lethal dosage for fungus is lower (Rabinkov et al., 1998). In this study, two dosages of antifungal agents, 1 and 5 mg kg−1 day−1, were selected. The results showed that allicin could reduce the morbidity and the fungal load in tissues of mice infected with C. albicans. However, these effects cannot be directly attributed to allicin, as it is not stable and converts immediately to other products such as ajoene, which may also have antifungal potential.

The fungal load in liver of treated mice showed a significant reduction with increasing time intervals. Although after 1-week postinfection, the fungal load in mice treated with 5 mg kg−1 day−1 of allicin was lower (log10 mean CFU g−1=3.16 ± 0.42) compared with the other treated groups, mice treated with 5 mg kg−1 day−1 of fluconazole showed a more significant decrease in fungal load (log10 mean CFU g−1=2.16 ± 0.20) thereafter. The results seen in other organs were similar to those seen in the liver (Table 2). Our findings also showed that the fungal load for all concentrations of antifungals during the first week were approximately similar, but after this time the differences between treated groups were significant. This may be due to the intrinsic murine immune responses of BALB/c mice (Ashman & Papadimitriou, 1988) infected at sites surrounding the infection for as long as 5 days postinfection, whereas treated mice were able to suppress Candida infection after at least 1 week. On the other hand, our data suggest that the conditions were approximately constant after 2 weeks postinfection until the last day of the experiments.

Data analysis showed a significant reduction in mortality for the two groups treated with fluconazole when compared with untreated control (P<0.05), whereas no significant difference was observed between the allicin groups treated with 1 and 5 mg kg−1 day−1 dosages and the untreated control group at levels P=0.163 and P=0.067, respectively. However, the survival study suggests that allicin could increase the MST until 16 days, whereas the untreated control group showed an MST of 8.5 days. The percentage of mortality was reduced to 50% by treatment with allicin (Table 3, Fig. 3). The results from the MIC determination seem to suggest a more significant anticandidal potential in vitro of allicin than of fluconazole. However, the time–kill curve showed that allicin is comparable to fluconazole in terms of fungal load reduction.

The combined results from both the survival studies and fungal load reduction studies in the present work demonstrate that allicin is slightly less efficacious than fluconazole in the treatment of candidiasis. Therefore, it is necessary to discover better treatment modalities or to increase the dosage of allicin, which will require further experiments. Designing a new model of treatment with different dosages of allicin and related compounds as well as a combination of allicin with azoles should be investigated. The molecular mechanisms of the actions of allicin could be investigated further to determine its probable targets in Candida cells.

Acknowledgements

This project was funded through the Research University Grant Scheme (RUGS) sponsored by the university and a Science Fund sponsored by the Ministry of Science, Technology and Innovation.

Ancillary