|Home | About | Journals | Submit | Contact Us | Français|
Cutaneous and mucocutaneous Candida infections are considered to be important targets for antimicrobial photodynamic therapy (PDT). Clinical application of antimicrobial PDT will require strategies that enhance microbial killing while minimizing damage to host tissue. Increasing the sensitivity of infectious agents to PDT will help achieve this goal. Our previous studies demonstrated that raising the level of oxidative stress in Candida by interfering with fungal respiration increased the efficiency of PDT. Therefore, we sought to identify compounds in clinical use that would augment the oxidative stress caused by PDT by contributing to reactive oxygen species (ROS) formation themselves. Based on the ability of the antifungal miconazole to induce ROS in Candida, we tested several azole antifungals for their ability to augment PDT in vitro. Although miconazole and ketoconazole both stimulated ROS production in C. albicans, only miconazole enhanced the killing of C. albicans and induced prolonged fungistasis in organisms that survived PDT using the porphyrin TMP-1363 and the phenothiazine methylene blue as photosensitizers. The data suggest that miconazole could be used to increase the efficacy of PDT against C. albicans, and its mechanism of action is likely to be multi-factorial.
In addition to its application to cancer therapy (1), it is clear that photodynamic therapy (PDT) has considerable potential as an antimicrobial treatment, especially for cutaneous and mucocutaneous infections (2). To achieve that goal, it will be important to devise PDT strategies that enhance microbial killing while minimizing damage to host tissue. The fungus Candida colonizes the mucosal and cutaneous epithelial surfaces of the body as a commensal organism in the majority of the human population (3). Impairment of innate and/or adaptive defenses leads to conversion from commensal to pathogen, leading to clinical candidiasis (4). Hence, superficial Candida infections are considered important targets for antimicrobial PDT (2). C. albicans is the most prevalent pathogenic Candida species and is the focus of this report.
We have recently demonstrated that a high degree of photosensitizer selectivity for C. albicans can be achieved in the context of infection in a mouse model, and shown that PDT conditions that cause minimal host tissue damage result in significant organism reduction (5). We have also identified a point of vulnerability to PDT in Candida, which occurs during successful adaptation of the fungus to respiratory deficiency following antifungal drug treatment and host-induced environmental stress (6, 7). This is important clinically since Candida can use a strategy of either undergoing transient respiratory deficiency (8)or uncoupling oxidative phosphorylation (9) to evade both host defenses and antifungal treatment, as well as to enhance its virulence in vivo (10). Thus, PDT against Candida infection can potentially be applied effectively under circumstances where other treatments may fail following fungal adaptation to stress. Our observations also open up the possibility of enhancing the efficacy of PDT against primary Candida infection through a two-step strategy by interference with fungal respiration or its associated functions.
In the treatment of tumors with PDT in experimental models, there is precedent for increasing the efficacy of PDT by the use of combined therapy (reviewed in (11)). As examples, a decrease in the expression of angiogenic and proinflammatory factors by administration of cyclooxygenase-2 inhibitors resulted in significantly improved survival following PDT of tumor-bearing mice compared to individual treatments alone (12); administration of the anti-angiogenic drug Avastin was also shown to improve the efficacy of PDT in a xenograft model of Kaposi’s sarcoma in nude mice (13). Pertinent to the treatment of candidiasis, inhibition of the mitochondrial respiratory chain increased the growth-inhibitory effect of natural phenolic compounds against the opportunistic filamentous fungal pathogens Aspergillus flavus and A. fumigatus (14). Finally, the targeting of cell stress response pathways in pathogenic fungi, including Candida, is being explored as a means of both enhancing the activity of currently used antifungal agents and abrogating mechanisms of antifungal resistance (15–18). Herein, we examine antifungal agents already in clinical use for their ability to enhance the efficacy of PDT against Candida.
Our previous studies (7)demonstrated that blocking respiratory pathways in C. albicans, C. glabrata and Saccharomyces cerevisiae at sites known to lead to increased oxidative stress increased the efficacy of PDT-induced phototoxicity against these organisms. On the basis of these observations, we sought to identify compounds in clinical use that would have a similar effect of augmenting the oxidative stress caused by PDT by contributing to ROS formation themselves. The ability of the antifungal miconazole to induce ROS in Candida (19)and S. cerevisiae (20)suggested it would enhance the efficacy of PDT. Ergosterol is the primary sterol in fungal membranes. A well-characterized mode of action of miconazole is its effect on membrane synthesis through inhibition of the cytochrome P-450 lanosterol 14α-demethylase during ergosterol biosynthesis, resulting in accumulation of toxic 14α-methylated sterols. Accumulation of these toxic intermediates and downstream effects impairing membrane function account, at least in part, for the antifungal action of miconazole and related compounds (21). However, early studies on the antifungal action of miconazole showed that, in addition to effects on ergosterol synthesis, it inhibited respiratory pathways in C. albicans (22)and mitochondrial ATPase in both S. cerevisiae and C. albicans (23). Therefore, we predicted that inhibiting known metabolic pathways in C. albicans with compounds such as miconazole in order to increase the sensitivity of the fungus to the oxidative stress induced by PDT would increase the efficacy of the initial PDT dose against C. albicans.
C. albicans SC5314 (24) was used for all studies. Cells were grown overnight to early stationary phase in yeast extract-peptone-dextrose broth (Difco, Detroit, MI), washed twice in dH2O and resuspended to O.D.600nm = 1.0. PDT was performed in the presence of 10 μg/ml of the cationic photosensitizer meso-tetra (N-methyl-4-pyridyl) porphine tetra tosylate (TMP-1363). The organism suspension (2 ml) was added to a six well culture dish and irradiated at a fluence of 1.0 J/cm2 with broadband visible light from a 48 × 48 cm light box equipped with a bank of fluorescent lamps (Sylvania GRO-LUX, 15 W, part no. F15T8/GRO). The irradiance at the surface of the light box was 4.0 mW/cm2, and the spectrum of light was such that approximately 67% of the power was emitted within the range of 575–700 nm. Untreated organisms and organisms treated with TMP-1363 but shielded from light were used as controls. For PDT with methylene blue (MB), clinical grade reagent (10 mg/ml aqueous solution; American Regent, Shirley, NY) was diluted to 100 μg/ml in dH2O and incubated with organisms for 30 min with periodic mixing. Organisms were pelleted by centrifugation and resuspended to their original volume in dH2O to return them to O.D.600nm = 1.0. Cells were irradiated at a fluence of 7.2J/cm2. Untreated organisms and organisms treated with MB but shielded from light were used as controls. Cells were serially diluted 10-fold in dH2O; 2 μl of each dilution was spotted on YPD plates and incubated at 37°C for 24–48 h as indicated.
Miconazole, in the form of its nitrate salt, clotrimazole and ketoconazole (Sigma-Aldrich, St. Louis, MO) were dissolved in absolute methanol at stock concentrations of either 10 or 12.5 mg/ml depending on the experiment. Fluconazole (Sigma-Aldrich) was dissolved in dH2O at a stock concentration of 5 mg/ml. Working solutions were prepared by dilution in dH2O, and specific treatment conditions are described in the figure legends.
Early stationary phase organisms of C. albicans (O.D.600 = 1.0) were treated in a 6-well tissue culture dish with the corresponding imidazole at 25 μg/ml and 2′,7′-dichlorofluorescein diacetate (H2DCFDA, Molecular Probes, Eugene, OR; 5 μg/ml)for 2 h prior to PDT using 10 μg/ml TMP-1363 and a fluence of 1.0 J/cm2. Induction of fluorescence of H2DCFDA (19, 20) after oxidation was used as an indicator of ROS production. Following PDT, fluorescence was measured in a SpectraMax M5 (Molecular Devices, Sunnyvale, CA) plate reader (Ex = 485 nm/Em = 538 nm).
Early stationary phase organisms of C. albicans SC3514 (O.D.600 = 1.0) were either left untreated or treated with 100 μg/ml methylene blue (MB) for 30 min. Excess MB was removed following centrifugation and organisms were resuspended to the original density in dH2O. Following irradiation with a broadband light source (7.2 J/cm2), 100 μl of cell suspension at either the original density (MB irradiated) or a 10−1 dilution (untreated, MB shielded) were mixed with 4 ml molten YPD top agar, poured onto YPD plates and allowed to solidify at room temperature for 30 min. Four blank paper filter discs (0.25 inch diameter; BBL Microbiology Systems, Becton Dickinson and Co., Cockeysville, MD) were placed on each plate and miconazole was added to individual discs in amounts ranging from 0 to 62.5 μg, respectively. Images of plates were taken following 24 and 48 h of incubation at 37°C.
Each experimental group was assayed in duplicate and experiments performed two to three times as indicated in the figure legends. Data represent the mean from combined replicate experiments ±S.D. using the pair-wise Student’s t-test. In all cases, P values of <0.05 were considered significant.
We initiated our studies by determining whether pre-treatment of C. albicans SC5314 with the antifungal agent miconazole increased its sensitivity to killing by PDT. The structure of miconazole and the other antifungal agents used in this study are shown in Figure 1. PDT was performed on early stationary phase organisms of C. albicans SC3514 using the tetra-cationic, porphyrin-based photosensitizer meso-tetra (N-methyl-4-pyridyl) porphine tetra tosylate (TMP-1363; Frontier Scientific, Inc., Logan, UT) (6). As shown in Figure 2, miconazole treatment alone induced a modest reduction (P<0.01) in colony forming units (CFU) compared to untreated cells. While the PDT conditions employed increased C. albicans killing significantly compared to miconazole alone (P<0.001), combined treatment with miconazole plus PDT using TMP-1363 was significantly more effective than either treatment alone (P<0.001). Organisms treated with miconazole + TMP-1363 but shielded from light were not killed to a greater degree than miconazole alone (data not shown). Hence, in keeping with our previous findings using TMP-1363 as a photosensitizer (6, 7), the effects of miconazole + TMP-1363 on C. albicans were irradiation-dependent and not attributable to dark toxicity under the conditions of our assay.
We next evaluated a combination of PDT and 25 μg/ml of the triazole fluconazole, the most widely used systemic azole (21). In combination with PDT, fluconazole neither enhanced the killing of C. albicans nor induced fungistasis, a transient inhibition of growth. In contrast, miconazole exposure at the same concentration in combination with PDT (Mcz + PDT) resulted in fungistasis of surviving organisms in the first (undiluted) cell spot for at least 24 h (Figure 3).
A “threshold level” of miconazole exposure was necessary for fungistasis, since growth was observed in the second and third cell spots where the miconazole concentration was decreased by dilution (10−1 and 10−2, respectively) prior to spotting cells on the plate. After 48 h incubation, the fungistasis induced by miconazole in combination with PDT was overcome (Figure 3, bottom row).
Since the fungistatic effect seen was miconazole exposure-dependent (Figure 3), we next asked how the timing and duration of exposure to miconazole modulated fungistasis and fungicidal activity against C. albicans SC5314 (Figure 4). In the top four rows, cells were treated with miconazole but not subjected to PDT. Early stationary phase yeast pre-treated with miconazole for 2 h grew comparably to untreated cells. Subsequent exposure to miconazole in the dilution series post-irradiation was fungistatic, but had only a modest effect on organism viability. In the bottom four rows, cells were treated with miconazole and subjected to PDT. As seen previously in Figure 2, pre-incubation with miconazole increased the fungicidal activity of PDT. Exposure to miconazole following PDT had the same adverse effect on viability as pre-PDT exposure to miconazole, and induced fungistasis as seen previously in Figure 3. Exposure to miconazole before and after PDT (bottom row) resulted in the highest degree of fungicidal activity.
The ability of miconazole to induce reactive oxygen species (ROS) in Candida (19)and S. cerevisiae (20)suggested it would enhance the efficacy of PDT. Therefore, we compared two other commonly used imidazole antifungal agents ketoconazole and clotrimazole to miconazole for their ability to induce ROS in C. albicans, and to augment PDT using TMP-1363. Induction of ROS in C. albicans (Figure 5) was measured spectrofluorometrically by the conversion of 2′,7′-dichlorofluorescein diacetate (H2DCFDA, Molecular Probes, Eugene, OR) to fluorescent 2′,7′-dichlorofluorescein after removal of acetate groups by cellular esterases and oxidation within the cell (19). Preliminary experiments showed that 25 μg/ml of miconazole was the lowest concentration capable of inducing substantive ROS production in the conditions of our assay. Increasing the miconazole concentration up to 100 μg/ml did not result in significantly higher ROS levels as indicated by the relative fluorescence of 2′,7′-dichlorofluorescein (data not shown). Hence, we used a concentration of 25 μg/ml in our comparison (Figure 5). While the concentration used resulted in slight differences in the molarity of these three imidazoles, it was well in excess of their established minimal inhibitory concentrations of <0.5 μg/ml (25). Ketoconazole stimulated ROS to a slightly higher level compared to miconazole (P<0.04), and clotrimazole stimulated ROS to a lower level than either ketoconazole or miconazole (P<0.001). Although the fluorescence measurements used to detect ROS were performed after PDT, incubation of C. albicans with TMP-1363 alone did not result in fluorescence above untreated control cells in this assay. Only in combination with clotrimazole did TMP-1363 appear to augment ROS production (P<0.003) compared to each respective imidazole alone.
Surprisingly, there was no direct correlation between the ability of animidazole antifungal to stimulate ROS in C. albicans and its ability to increase the efficacy of PDT. Ketoconazole did not enhance C. albicans killing by PDT, nor did it induce fungistasis in surviving cells (Figure 6). Furthermore, clotrimazole induced less ROS formation in C. albicans than miconazole and ketoconazole (Figure 5), yet clotrimazole enhanced killing by PDT and induced fungistasis in surviving cells post-PDT (Figure 6). These data demonstrated that, combined with PDT, miconazole and clotrimazole, but not fluconazole (Figure 3) and ketoconazole, were more effective against C. albicans in vitro, in both fungicidal capacity and fungistasis of survivors, than either PDT or azole treatment alone.
Since TMP-1363 is not yet approved for clinical applications in humans, we determined whether using methylene blue (MB) as a photosensitizer would provide a similar degree of enhancement in PDT of C. albicans. MB is widely used in various clinical settings (26, 27)and has efficacy in PDT of C. albicans in vitro (28, 29). There are several reports of treatment of oral candidiasis using MB for PDT in murine models of infection (30–32), and MB analogues (33)are currently in clinical trials for antimicrobial PDT of infected leg ulcers (www.photopharmica.com).
Our earlier studies demonstrated that TMP-1363 was superior to MB in PDT of C. albicans in vitro when used at the same concentration (5). However, by increasing both the MB concentration and fluence used previously (5), PDT yielded approximately 1–2 log10 killing of early stationary phase C. albicans SC5314 (data not shown). Surviving cells were then exposed to increasing amounts of miconazole. We adapted a recently described miconazole disc diffusion assay (34) to assess both fungistasis and fungicidal activity at multiple concentrations of miconazole in a single plate, which allowed more flexibility than the CFU assay used in the PDT studies with TMP-1363 described above.
Cells treated with MB and irradiated were compared to MB-treated cells shielded from light and untreated, irradiated cells (Figure 7). Plates were photographed at 24 and 48 h post-PDT. In all cases, the methanol vehicle (lower right disc) used to dissolve miconazole showed no inhibition of C. albicans. In untreated cells (top panels), a zone of fungistasis, seen as partial clearing around the disc, remained constant over 48 h at all three miconazole levels (2.5, 12.5 and 62.5 μg; see Figure 7 legend), and a small zone of fungicidal activity, seen as a zone of complete clearing, was observed at the two highest levels of miconazole. Compared to untreated cells, MB-treated, shielded cells (middle panels) displayed more pronounced fungistasis and a higher degree of fungicidal activity. Hence, MB and miconazole in combination have a degree of dark toxicity not seen with TMP-1363 and miconazole. Strikingly, MB-treated, irradiated cells (bottom panels) showed the most pronounced degree of fungistasis at all miconazole levels tested after 24 h (bottom left panel), and a substantially increased zone of fungicidal activity compared to the other two treatment groups after 48 h.
At minimal inhibitory concentrations, miconazole, like the other members of the azole family of antifungals, interferes with ergosterol synthesis by targeting primarily Erg11p, lanosterol 14α-demethylase, generating toxic sterol intermediates that induce fungistasis (21, 35). Unlike most other members of the azole family, miconazole is also fungicidal against Candida at higher concentrations (36, 37). Furthermore, innate or acquired resistance by Candida to miconazole is reported to be fairly rare (38, 39). However, phenotypic resistance to miconazole-mediated killing of C. albicans develops in early stationary phase (36) at 16 μg/ml, and treatment of C. albicans biofilms with 250μg/ml miconazole generated phenotypically resistant “persisters” at a frequency of 1–2% of the population (40).
The enhanced fungistatic and fungicidal activity against C. albicans we describe using combined treatment with miconazole plus PDT was observed against planktonic, early stationary phase C. albicans cells using 25 μg/ml miconazole. Under the conditions of our assay, miconazole alone at this concentration had relatively modest fungicidal and fungistatic activity. The prolonged fungistasis of cells that survived treatment with miconazole plus PDT in vitro is reminiscent of the postantibiotic (41)or postantifungal (42) effect induced by the polyene (41)and echinocandin (43, 44)classes of antifungal agents. Interestingly, earlier studies showed imidazoles induced, at best, a slight postantifungal effect in vitro (41, 42). The fungistasis induced by combined treatment with miconazole plus PDT could provide an advantage for host defenses to help control infection more effectively. This provides rationale for further investigation of the anti-Candida effects of miconazole plus PDT against C. albicans in terms of both mechanism of action and therapy of infection, particularly in the context of biofilms (45). Drug efflux pump activity is increased in C. albicans biofilms compared to planktonic cells (46), and could contribute to miconazole resistance (47, 48). In addition, the ATP-binding cassette (ABC) class of multidrug efflux systems in C. albicans modulates the efficacy of one of the photosensitizers used in this study, methylene blue (49).
Current treatment guidelines (50)encourage topical application of miconazole or clotrimazole as the first line of drugs for mucosal candidiasis in AIDS and other immunocompromised patients. A drawback to miconazole or clotrimazole use, particularly for mucosal infection, is the need for multiple topical applications, often daily, for a prolonged time period (50, 51). Hence, the induction of fungistatic and fungicidal effects by miconazole plus PDT may also lead to increased spacing between doses. The miconazole concentration in most commercial topical preparations is in the range of 10 to >40 mg/ml (52). At these concentrations, unpleasant taste is given as a reason for patient non-compliance in treatment of oral infection (52).
Thus, a long-term clinical objective will be to overcome the potential limitations of miconazole in therapy of mucosal and cutaneous candidiasis by combination with PDT, maximizing the advantages of both treatments. The combination of miconazole with PDT may also yield improved efficacy at lower miconazole concentrations, with the added benefit of improved patient compliance. Furthermore, combined treatment with miconazole plus PDT not only has efficacy against C. albicans, but has the potential to be applied to other pathogenic Candida species as well. This improved approach to topical antifungal therapy could have the positive effect of reducing the incidence of disseminated infection, and help reserve the newest class of antifungal agents, the echinocandins, for the most serious and recalcitrant systemic Candida infections.
Our initial hypothesis was that antifungal agents like miconazole that increased ROS production in C. albicans would augment the sensitivity of the fungus to the oxidative stress induced by PDT. Other investigators (19, 20)reported that fluconazole was somewhat less able than miconazole to induce ROS by C. albicans, which may account, at least in part, for the inability of PDT to be augmented by fluconazole. Our own preliminary studies gave similar results (data not shown). Furthermore, clotrimazole increases the sensitivity of C. albicans to exogenous ROS (53), and superoxide dismutases (SODs) contribute to the development of miconazole-tolerant “persisters” in C. albicans biofilms (40). While SODs provide protection against PDT in the Gram-negative bacterium E. coli (54), they do not appear to influence the sensitivity of the Gram-positive Staphylococcus aureus to PDT even though the expression of SODs is up-regulated (55). The contribution of SODs to sensitivity to PDT in C. albicans has not been investigated. The results we obtained would not appear to support our initial hypothesis that azole-induced ROS enhanced the efficacy of PDT, since the imidazole ketoconazole induced a high degree of ROS induction comparable to miconazole (Figure 5), but did not appear to enhance either fungistasis or fungicidal activity against C. albicans by PDT (Figure 6). This is in accordance with our earlier studies demonstrating that the induction of catalase by mild pre-treatment with hydrogen peroxide does not protect C. albicans against PDT using Photofrin as a photosensitizer (56).
Alternatively, the fungistatic effect of miconazole and clotrimazole could be linked to their relative lipophilicity compared to the more hydrophilic nature of fluconazole and ketoconazole (57), allowing more effective uptake of miconazole and clotrimazole at the same concentration. Furthermore, miconazole perturbs a number of other fungal structures and functions, in addition to sterol metabolism and induction of ROS. Miconazole induces changes in mitochondrial ultrastructure and organization (20, 58). Cell membrane integrity (59), membrane rafts (60) and membrane protein function (61) are also perturbed by miconazole. Calcium signaling modulates miconazole sensitivity in both the non-pathogenic yeast S. cerevisiae and C. albicans (62–66), and miconazole is an antagonist of calmodulin (67). Like miconazole, calcium signaling antagonists disrupt cell integrity in these fungi (59). As mentioned earlier, miconazole inhibits respiratory pathways in C. albicans (22)and mitochondrial ATPase in both S. cerevisiae and C. albicans (23). A number of these antifungal effects, including inhibition of cytochrome P-450 lanosterol 14α-demethylase and pathways involved in respiration, would be predicted to be related to the ability of miconazole to bind to the hydrophobic amino acid pocket designed to harbor heme co-factor in heme-dependent proteins (68, 69). Thus, the mechanism by which miconazole potentiates PDT is likely to be multi-factorial.
The ability of miconazole to potentiate MB-based PDT is important for two reasons. Both compounds are approved for clinical use, and the application of miconazole post-PDT would obviate the need for a prolonged pre-treatment to increase the sensitivity of the fungus to PDT. The dark toxicity seen with the two compounds in combination (Figure 7) may be related to the activity of MB as a membrane-active compound against infectious agents (70). Since both MB (49)and miconazole (47)are potential substrates for drug efflux pumps in C. albicans, it will be important to determine whether PDT in combination with miconazole is also more effective than PDT alone in the context of an infection. Therefore, future studies will be directed at examining the role of miconazole and closely related imidazoles for the ability to potentiate PDT in an experimental model of candidiasis (5).
This work was supported by grants CA68409 (T.H.F) and Post-Baccalaureate Research Program for Minority Students (PREP) GM064133-2106 (S.B.S) from the National Institutes of Health.
†This paper is part of the Symposium-in-Print on “Antimicrobial Photodynamic Therapy and Photoinactivation”