Synergistic Activity of Chloroquine with Fluconazole against Fluconazole-Resistant Isolates of Candida Species

Abstract

The in vitro activity of chloroquine and the interactions of chloroquine combined with fluconazole against 37 Candida isolates were tested using the broth microdilution, disk diffusion, and Etest susceptibility tests. Synergistic effect was detected with 6 of 9 fluconazole-resistant Candida albicans isolates, with Candida krusei ATCC 6258, and with all 12 fluconazole-resistant Candida tropicalis isolates.

INTRODUCTION

Candida spp. represent the major fungal pathogens responsible for invasive candidiasis, including candidemia (1–4). The incidence of candidemia has increased significantly worldwide, with high morbidity and mortality in recent years (5–10). Because the percentage of infections caused by non-Candida albicans Candida species has increased, the susceptibility of Candida spp. to fluconazole (FLU) has decreased (4). In addition, the sensitivity of Candida spp. to first-line antifungals is inversely related to the mortality caused by candidemia (3, 4). Furthermore, previous studies have indicated that nonantifungal agents combined with FLU show synergistic activity against FLU-resistant Candida spp. in vitro (11–13), and attempts at combination therapy to address treatment failures have already been made (14). Therefore, developing new antifungal therapeutics for infections caused by FLU-resistant Candida spp. is necessary.

The antimalarial drug chloroquine (CQ) is thought to have a broad antifungal capacity, as well as the ability to inhibit the growth and morphogenesis of C. albicans and sensitize biofilms of C. albicans to antifungal azoles (15–21). However, whether CQ has activity against FLU-resistant Candida spp. is unclear. In the present study, we aimed to investigate the in vitro susceptibilities of FLU-resistant Candida spp. to CQ alone and to the combination of CQ and FLU.

Thirty-seven Candida isolates, including 13 C. albicans, 14 Candida tropicalis, 5 Candida parapsilosis, 3 Candida glabrata, and 2 Candida krusei isolates, were tested in this study. All of our isolates, which were stored at the Peking University Research Center for Medical Mycology, were collected from patients with candidemia. Isolates were identified by conventional methods such as the use of CHROMagar Candida (CHROMagar, Paris, France), the API 20C system (bioMérieux, Marcy l'Etoile, France), and molecular identification as described in a previous study (22).

Broth microdilution (BMD) susceptibility was determined as described in the CLSI M27-A3 protocol (23). Inoculum concentrations were adjusted to 1 × 10⁶ to 5 × 10⁶ CFU/ml by using a 0.5 McFarland standard at the 530-nm wavelength, ensuring that the final concentration in the 96-well plates was 0.5 × 10³ to 2.5 × 10³ CFU/ml. Antifungal drugs were obtained as standard powders. The final concentrations of fluconazole (Sunve Pharm, Shanghai, China) ranged from 0.063 to 64 μg/ml, and the chloroquine diphosphate salt (Sigma-Aldrich, Saint Louis, MO, USA) (molecular weight, 515.9) concentrations ranged from 8 to 512 μg/ml. Both were dissolved in distilled water. Synergy testing was evaluated using the checkerboard technique as previously described (24). After 24 h of incubation at 35°C, the MIC₅₀ values (MICs), the lowest concentration required to support ≥50% growth inhibition compared with the growth in the control wells, were determined. A fractional inhibitory concentration index (FICI) value of ≤0.5 indicates synergy, a FICI value of >4 indicates antagonism, and a FICI value of >0.5 and ≤4 indicates no interaction (25).

Disk diffusion susceptibility testing was done based on the CLSI M44-A2 protocol (26) with minor modifications, and the Etest was performed at the same time. Mueller-Hinton-plus-GMB (glucose-methylene blue) agar plates and RPMI-2G agar plates (27) with or without 1,031.8 μg/ml chloroquine diphosphate salt were used for disk diffusion testing and the Etest, respectively. Blank paper disks (8 mm in diameter) with 25 μg FLU and FLU Etest strips (bioMérieux) were placed on the plates. The results were read and photographed after the plates were incubated at 35°C for 24 h in the dark because CQ is sensitive to light. Etest MICs was read at the intersection of the scale of the strip and the first discernible growth inhibition ellipse or as 80% inhibition of visual growth according to the manufacturer's guidelines. The disk diffusion testing was determined by measuring the zone diameter to the nearest whole millimeter at the point at which there was a prominent reduction in growth. Pinpoint microcolonies at the zone edge were ignored. For the Etest, synergy was defined as a decrease of ≥3 dilutions in the MIC (28), whereas for the disk diffusion analysis, changes of the inhibition zone were used to reflect whether chloroquine has the ability to enhance the activity of fluconazole but did not apply for judging the synergy accurately. C. parapsilosis ATCC 22019 and C. krusei ATCC 6258 were also used as quality controls. Each test was performed in triplicate on three different days; the MIC values of BMD and Etest analysis had no significant differences, and the mean values were used for disk diffusion testing.

The MICs and FICI for FLU combined with CQ against the Candida spp. are presented in Table 1. The MICs of CQ against all isolates were ≥512 μg/ml. Synergy was observed in 6 of 9 FLU-resistant C. albicans isolates, C. krusei ATCC 6258, and all of the FLU-resistant C. tropicalis isolates. No interaction was observed among 3 FLU-resistant C. albicans, 4 FLU-susceptible C. albicans, and 2 FLU-susceptible C. tropicalis isolates, 1 C. krusei isolate, all of the C. parapsilosis isolates, and all of the C. glabrata isolates.

TABLE 1.

The results of broth microdilution, Etest, and disk diffusion susceptibility tests for fluconazole alone or in combination with chloroquine against Candida species^a

Species (no. of isolates) and isolate	BMD MIC (μg/ml)						Etest MIC (μg/ml)		DD diam (mm)c
	Single		Combined		FICIb	Outcome	FLU	FLU + CQ	FLU	FLU + CQ
	FLU	CQ	FLU	CQ
C. albicans (13)
BMU 02990	≥64	>512	0.5	32	0.039	SYN	>256	1.5	NZ	50
BMU 02989	>64	>512	0.5	64	0.066	SYN	>256	1	NZ	32
BMU 02971	>64	>512	0.5	32	0.035	SYN	>256	1	NZ	37
BMU 02972	≥64	>512	1	128	0.141	SYN	>256	0.5	NZ	30
BMU 03997	≥64	>512	0.5	32	0.039	SYN	>256	1.5	NZ	34.5
BMU 00531	≥64	512	4	8	0.078	SYN	>256	6	NZ	30
BMU 00416	64	>512	64	128	1.125	NI	>256	>256	NZ	NZ
BMU 00716	64	>512	32	64	0.563	NI	>256	>256	NZ	NZ
BMU 00717	64	>512	64	16	1.016	NI	>256	>256	NZ	NZ
BMU 05621	0.5	>512	0.5	16	1.016	NI	0.75	1	30	32
BMU 05627	0.5	>512	0.5	8	1.008	NI	0.5	0.5	30	30
BMU 05680	0.5	>512	0.5	8	1.008	NI	0.5	1	32	32
SC 5314	0.25	>512	0.25	32	1.031	NI	4	4	30	32
C. tropicalis (14)
BMU 06203	>64	>512	1	32	0.039	SYN	>256	1.5	NZ	40
BMU 06199	>64	>512	4	8	0.039	SYN	>256	2	NZ	28
BMU 06491	>64	>512	1	128	0.133	SYN	>256	4	NZ	33
BMU 02721	≥64	>512	1	8	0.023	SYN	>256	0.75	NZ	45
BMU 05462	≥64	>512	2	16	0.047	SYN	>256	2	NZ	32
BMU 05690	≥64	>512	4	128	0.188	SYN	>256	2	NZ	28
BMU 06027	≥64	>512	1	16	0.031	SYN	>256	2	NZ	31
BMU 06172	≥64	>512	0.5	16	0.023	SYN	>256	1.5	NZ	43.5
BMU 06194	≥64	>512	2	8	0.039	SYN	>256	1	NZ	36
BMU 06028	64	>512	0.5	8	0.016	SYN	>256	2	NZ	40
BMU 07166	64	>512	1	8	0.023	SYN	>256	6	NZ	26
BMU 06179	64	>512	1	8	0.023	SYN	>256	1.5	NZ	40
BMU 06358	2	>512	2	256	1.25	NI	0.75	1	30	36
BMU 05695	1	>512	1	8	1.008	NI	1.5	4	30	32
C. parapsilosis (5)
BMU 05030	64	>512	32	512	1	NI	>256	16	NZ	NZ
BMU 05485	8	>512	8	8	1.008	NI	>256	>256	NZ	NZ
ATCC 22019	1	>512	1	16	1.016	NI	1.5	1	32	32
BMU 05504	0.25	>512	0.25	8	1.008	NI	0.25	0.5	32	38
BMU 05752	0.25	>512	0.25	8	1.008	NI	1.5	1.5	32	35
C. glabrata (3)
BMU 05505	64	>512	64	512	1.5	NI	>256	>256	NZ	NZ
BMU 05032	8	>512	4	512	1	NI	4	0.75	16	28
BMU 05481	0.5	>512	0.25	8	0.508	NI	0.19	1	30	32
C. krusei (2)
ATCC 6258	>64	>512	16	128	0.25	Synergy	≥96	4	NZ	13
BMU 07410	32	>512	32	8	1.008	NI	32	32	NZ	10

^aFLU, fluconazole; CQ, chloroquine; BMD, broth microdilution; FICI, fractional inhibitory concentration index; SYN, synergy; NI, no interaction; DD, disk diffusion; NZ, no zone.

^bThe off-scale MICs had been converted to the next highest 2-fold dilution for calculating FICIs.

^cData represent inhibition zone diameters.

The disk diffusion testing and Etest also confirmed the synergistic interaction. The inhibition zone diameters of FLU disks combined with CQ for the FLU-resistant C. albicans BMU 02972, FLU-resistant C. tropicalis BMU 07166, and C. krusei ATCC 6258 isolates were 39 mm, 38 mm, and 13 mm, respectively (Fig. 1, column a, row 2 [a2], c2, and f2), whereas FLU alone showed no inhibition zone (Fig. 1, a1, c1, and f1). The inhibition zone diameters of FLU disks combined with CQ against FLU-susceptible C. albicans SC5314, C. tropicalis BMU 06358, and C. parapsilosis ATCC 22019 were 32 mm, 39 mm, and 32 mm, respectively (Fig. 1, b2, d2, and e2), whereas for FLU alone they were 30 mm, 36 mm, and 32 mm (Fig. 1, b1, d1, and e1). Etest MICs of FLU strips combined with CQ for BMU 02972, BMU 07166, and ATCC 6258 were 0.5 μg/ml, 4 μg/ml, and 4 μg/ml, respectively (Fig. 1, a4, c4, and f4), but the strips with FLU alone showed no obvious inhibition zone (Fig. 1, a3, c3, and f3). Whether in combination with CQ or not, the presence of FLU made no obvious difference in the Etest MICs for both SC5314 isolates (both 4 μg/ml), isolate BMU 06358 (1 μg/ml and 0.75 μg/ml), and isolate ATCC 22019 (1 μg/ml and 1.5 μg/ml) (Fig. 1, b3, b4, d3, d4, e3, and e4).

FIG 1.

Disk diffusion susceptibility testing and Etest of fluconazole alone (lines 1 and 3) or combined with 1,031.8 μg/ml chloroquine (lines 2 and 4) against Candida isolates. Column a represents a fluconazole-resistant C. albicans isolate (BMU 02972); column b represents a fluconazole-susceptible C. albicans isolate (SC5314); column c represents a fluconazole-resistant C. tropicalis isolate (BMU 07166); column d represents a fluconazole-susceptible C. tropicalis isolate (BMU 06358); column e represents C. parapsilosis isolate ATCC 22019; and column f represents C. krusei isolate ATCC 6258.

Our results showed for the first time that CQ had poor antifungal activity against Candida spp. but enhanced the activity of FLU in vitro. The increase in FLU resistance among Candida spp. highlights the importance of developing new antifungal compounds. Studies have shown that CQ and caspofungin exhibited synergistic activity against C. albicans, C. glabrata, Aspergillus fumigatus, and Saccharomyces cerevisiae in vitro (29), and CQ inhibited the growth of yeast species such as Histoplasma capsulatum and S. cerevisiae because of iron deprivation (17, 30, 31). The iron deprivation would cause a decrease in the level of membrane sterols and then cause an increase in membrane fluidity, which could enhance the passive diffusion of drugs and result in heightening drug susceptibility (32). Furthermore, iron deprivation would lead to the downregulation of ERG11 (33), while the upregulation of ERG11 has been observed to cause azole resistance in C. albicans, C. glabrata, and C. tropicalis clinical isolates (34, 35). In addition to those, fluconazole resistance caused by the formation of biofilms that related to the activation of the drug efflux pumps and low membrane ergosterol content was reversed significantly by the presence of CQ (21). However, studies have indicated that the adaptive response, which depends on calcineurin activity, is related to the sterol synthesis and leads to a higher sensitivity of C. albicans to FLU (11). Therefore, for FLU-resistant Candida spp., the change in drug susceptibilities with a combination of FLU and CQ could be explained by the reasons outlined above, but the exact mechanism remains to be elucidated.

In vivo studies related to pharmacokinetics and pharmacodynamics have demonstrated that CQ efficacy is dependent on both the total dose and the regimen design, although the data (maximum concentration of drug in serum [C_max], half/life [t_1/2], area under the concentration-time curve [AUC], and time required for the drug concentration to exceed the MIC under steady-state pharmacokinetic conditions [T_MIC]) differed between different subjects (36–39). It is known that the antimalarial CQ is rapidly and almost completely absorbed from the gastrointestinal tract, with approximately 55% bound to nondiffusible plasma constituents in the plasma. In clinical work, for treatment of acute attacks of malaria in adults, an initial dose of 1 g CQ phosphate (equal to 600 mg CQ base) followed by an additional 500 mg CQ phosphate (equal to 300 mg CQ base) after 6 to 8 h and a single dose of 500 mg CQ phosphate (equal to 300 mg CQ base) on each of two consecutive days has been used, but concentrations in the plasma of patients are reported to be relatively low (<3.04 μg/ml) (40). Therefore, whether the CQ concentration used in this in vitro study is suitable for achieving successful treatment of FLU-resistant Candida infections in vivo is doubtful because that concentration of CQ in the plasma samples was under the influence of different disease and drug interactions. Thus, this report provides evidence useful for determining the antifungal action of CQ and its derivatives or other formulations in the future.

The mechanisms of FLU resistance in our tested isolates are not entirely clear; further exploration for determining the synergistic effects of FLU combined with CQ against FLU-resistant Candida spp., in particular, the role of ERG11, is essential. Taking advantage of two characteristics, low cost and safety, CQ combined with FLU may provide a new combination therapy against FLU-resistant Candida spp., especially in C. tropicalis infections.