OUP user menu

FK520 interacts with the discrete intrahelical amino acids of multidrug transporter Cdr1 protein and acts as antagonist to selectively chemosensitize azole-resistant clinical isolates of Candida albicans

Shweta Nim, Manpreet K. Rawal, Rajendra Prasad
DOI: http://dx.doi.org/10.1111/1567-1364.12149 624-632 First published online: 1 June 2014

Abstract

FK520, a homolog of antifungal FK506, displays fungicidal synergism with azoles in Candida albicans and inhibits drug efflux mediated by ABC multidrug transporter. This study establishes the molecular basis of interaction of FK520 with Cdr1 protein, which is one of the major ABC multidrug transporters of C. albicans. For this, we have exploited an in-house library of Cdr1 protein consisting of 252 mutant variants where the entire primary structure of the two transmembrane domains comprising of 12 transmembrane helices was subjected to alanine scanning. With these mutant variants of Cdr1 protein, we could identify the critical amino acids of the transporter protein, which if replaced with alanine, not only abrogated FK520-dependent competitive inhibition of drug efflux but simultaneously decreased susceptibility to azoles. Notably, the replacement of most of the residues with alanine was inconsequential; however, there were close to 13% mutant variants, which showed abrogation of drug efflux and reversal of fungicidal synergy with azoles. Of note, all the intrahelical residues of Cdr1 protein, which abrogated inhibitor's ability to block the efflux and reversed fungicidal synergy, were common. Taken together, our results provide evidence of cross-talk of FK520 with Cdr1 by interacting with the select intrahelical residues of the protein to chemosensitize isolates of Candida.

Keywords
  • Fungal infections
  • multidrug resistance
  • drug transporters
  • modulators

Introduction

Widespread fungal infections caused by opportunistic pathogen Candida albicans in immunocompromised patients are very common (Flevari et al., 2013). The prolonged usage of available antifungals also results in the development of tolerance to drugs, a phenomenon termed as multidrug resistance (MDR; White et al., 1998; Perea & Patterson, 2002). Various mechanisms that contribute to the development of MDR include point mutations in ERG11 or over-expression of its gene product, the target enzyme of azoles (i.e. lanosterol 14α-demethylase; Martel et al., 2010), and in the over-expression of the drug-efflux-pump-encoding genes CDR1, CDR2 and MDR1, which belong to the ATP-binding cassette (ABC) and major facilitator superfamily (MFS) transporters, respectively (Prasad et al., 1995; Sanglard & Odds, 2002; Hiller et al., 2006). Among various strategies employed to combat MDR, blocking or modulating the drug efflux pump proteins represents an attractive approach. In this context, studies from our as well as from other groups have identified modulators and inhibitors of efflux pump proteins, which can potentiate the efficacy of antifungals in use. For example, a natural polyphenol curcumin, quorum-sensing molecule farnesol, and an antabuse drug clinically used for treating alcoholism, disulfiram, show fungicidal synergism with the azoles and other drugs (Shukla et al., 2004; Sharma et al., 2009; Sharma & Prasad, 2011). Various drugs such as FK506, enniatins, milbemycin, synthetic-d-octapeptides, cyclosporine, ibuprofen, and unnarmicins are identified as fungal ABC transporter inhibitors that predominantly chemosensitize the fungal cells by competitively or noncompetitively inhibiting the drug efflux (Maesakia et al., 1998; Pina-Vaz et al., 2005; Tanabe et al., 2007; Prasad et al., 2011; Hayama et al., 2012).

The core stress pathway of Candida is activated by two related pathways; calcium–calmodulin-mediated calcineurin and cAMP pathway, which modulate fluconazole tolerance (D'Souza & Heitman, 2001). The two inhibitors of well-conserved calcineurin, FK506, and cyclosporine A are lethal to C. albicans and other fungal cells. Notably, if either of cyclosporine A or FK506 is given in combination with the fungistatic antifungals such as fluconazole, it results in fungicidal synergy (Cruz et al., 2002; Blankenship et al., 2003; Uppuluri et al., 2008). There are also studies to suggest that FK506-dependent fungicidal synergy could also be independent of calcineurin pathway; however, the molecular basis of FK506 and multidrug transporter protein interactions are not established (Schuetzer-Muehlbauera et al., 2003). Our earlier studies have shown that similar to FK506, its homolog FK520 could also inhibit drug efflux mediated by Cdr1 protein (Saini et al., 2005). In this study, the molecular basis of FK520 with Cdr1 protein is re-examined, and the specificity of interaction was compared with other ABC transporter Pdr5 of Saccharomyces cerevisiae and a MFS transporter Mdr1 of C. albicans.

The fungal ABC transporter Cdr1 protein is composed of two transmembrane domains (TMDs) and two nucleotide-binding domains (NBDs). Recently, we have probed the nature of the drug-binding pocket by performing systematic mutagenesis of the primary sequences of both TMDs. The entire primary sequence of the two TMDs of Cdr1 protein, comprising 12 transmembrane segments (TMS), was subjected to alanine-scanning mutagenesis with each of the 252 residues replaced with an alanine. It was shown that while the replacement of the majority of the amino acid residues with alanine yielded neutral mutations, close to 30% of the variants lost resistance to drug efflux substrates completely or selectively. The study led to the identification of critical amino acids within a drug-binding cavity, which, upon mutation, abolished resistance to all drugs tested (Rawal et al., 2013). In this study, we have exploited the in-house mutant library of Cdr1 protein to show the specific interaction of FK520, with the multidrug transporter. This study shows that the majority of the replacements of intrahelical residues yielded neutral mutations with no impact on the activity of FK520; however, there were several mutant variants that showed specific interaction of calcineurin inhibitor with the select amino acids of transporter protein and also demonstrated the molecular basis of FK520 with Cdr1 protein.

Materials and methods

Materials

Media chemicals were obtained from Hi-Media (Mumbai, India). Miconazole, ketoconazole, rhodamine 6G (R6G), nile red (NR), FK520, and deoxyglucose were obtained from Sigma Chemical Co (St. Louis, MO). Fluconazole was a kind gift from Ranbaxy Laboratory, Delhi, India.

Yeast strains and culture media

All the yeast strains were grown in the yeast extract peptone dextrose medium at 30 °C as described in our previous publications (Rawal et al., 2013). The S. cerevisiae strain used as a heterologous host for the expression of Cdr1 protein was AD1–8u, provided by Richard D. Cannon, University of Otago, Dunedin, New Zealand. The host AD1–8u having seven major ABC transporters deleted was suitably modified to clone GFP-tagged Cdr1 protein and its mutant variants (Nakamura et al., 2001).

Efflux of R6G and NR

The yeast cells of exponential phase were used for all the experiments. R6G and NR efflux was performed with freshly harvested cells as described previously (Maesaki et al., 1999; Ivnitski-Steele et al., 2009). The competition assays were performed as described in earlier publications (Sharma et al., 2009) where 100 μM of FK520 was added 5 min prior to the addition of 10 μM of R6G/NR. Cells were pelleted, washed twice with phosphate-buffered saline (PBS; without glucose), and resuspended as a 2% cell suspension, which corresponds to 108 cells (w/v) in PBS without glucose. The cells were then de-energized for 45 min in deoxyglucose (5 mM) and dinitrophenol (5 mM) in PBS (without glucose). The de-energized cells were pelleted, washed, and then resuspended as a 2% cell suspension (w/v) in PBS without glucose, to which R6G was added at a final concentration of 10 μM and incubated for 40 min at 30 °C. The equilibrated cells with R6G/NR were then washed and resuspended as a 2% cell suspension (w/v) in PBS without glucose. Samples with a volume of 1 mL were withdrawn at the indicated time and centrifuged at 9000 g for 2 min. The supernatant was collected, and absorption was measured at 527 nm. Energy-dependent efflux (at the indicated time) was measured after the addition of glucose (2%) to the cells resuspended in PBS (without glucose). Glucose-free controls were included in all the experiments. For competition assays, FK520 (100 μM) was added to the de-energized cells 5 min before the addition of R6G/NR and allowed to equilibrate.

Growth inhibition assay

The cytotoxic effect of FK520 on S. cerevisiae control cells (AD1–8u) and on cells over-expressing Cdr1 protein (AD-CDR1), Pdr5 protein (AD-PDR5), and Mdr1 protein (AD-MDR1) was determined by growth inhibition assay. Yeast cells (104) were seeded into 96-well plates in the absence and the presence of varying concentrations of FK520 (1–200 μg mL−1) and were grown for 48 h at 30 °C. Absorbance was measured using a microplate spectrophotometer at 600 nm. The percentage of growth inhibition was calculated considering the growth of cells without the inhibitor as 100% as (mean absorbance in test wells)/(mean absorbance in control wells) × 100.

Drug susceptibility assay

Inhibitory effect of FK520 in presence of azoles to Candida cells was evaluated by checkerboard assays and spot assays as previously described (Sharma et al., 2009). The interaction of FK520 with ketoconazole, miconazole, and fluconazole was evaluated by the checkerboard method recommended by the CLSI (formerly NCCLS). A range of concentrations were tried: 0.03 to 32 μg mL−1 for fluconazole, 0.125 to 64 μg mL−1 for ketoconazole, 0.125 to 80 μg mL−1 for miconazole and 0.156 to 200 μg mL−1 of FK520.

Results

FK520 competitively modulates R6G efflux mediated by ABC transporter

We had earlier cloned CDR1 gene as GFP-tagged proteins (AD-CDR1) stably over-expressed from a genomic PDR5 locus in a S. cerevisiae (AD1–8u) host that was derived from a Pdr1–3 mutant strain with a gain-of-function mutation in the transcription factor PDR1, resulting in the constitutive hyperinduction of the PDR5 promoter (Nakamura et al., 2001). For all the analysis, we used AD-CDR1 strains and compared it with the host strain (AD1–8u). The inhibitory effect of FK520 was reinvestigated and compared with other ABC and MFS transporters. For comparison, we studied the efflux of a well-known fluorescent substrate R6G. There was no significant extracellular efflux of R6G in control AD1–8u cells as compared with AD-CDR1 cells, which showed time-dependent elevated extracellular R6G levels (due to efflux) upon energization of starved cells (Fig. 1a). The addition of FK520 totally prevented the R6G efflux from the AD-CDR1 cells, which was evident from the very low extracellular R6G levels and was also comparable with the host AD1–8u cells (Fig. 1a). The inhibitor FK520 could also block the efflux of R6G in cells over-expressing Pdr5 protein (AD-PDR5); however, such was not the case with the Mdr1 protein over-expressing cells (AD-MDR1) where the addition of FK520 was ineffective in preventing the efflux of NR (Fig. 1b and c, respectively). Another fluorescent substrate NR was used to evaluate the impact of FK520 in case of Mdr1 protein as R6G is not a substrate of Mdr1 transporter (Ivnitski-Steele et al., 2009). The efflux of radiolabeled fluconazole (H3-FLC) and fluorescent NR, which are also substrates of Cdr1 protein, could also be similarly blocked by FK520 (data not shown).

1

Analyses of functional properties of Cdr1 protein in response to the treatment of FK520. Extracellular R6G concentrations in Saccharomyces cerevisiae control cells (AD1–8u) and in cells over-expressing (a) Cdr1 protein (AD-CDR1), (b) Pdr5 protein (AD-PDR5), (c) Mdr1 protein (AD-MDR1). In case of AD-MDR1 cells, the efflux of NR was performed as a test substrate as R6G is not a substrate of Mdr1 protein. ‘Y’-axis represents the extracellular concentration of NR/R6G at wavelength of 527 nm. To see the modulatory effect on the efflux of R6G (10 μM) or NR (10 μM), a 10× concentration (100 μM) of FK520 was used. (d) Lineweaver–Burk plot of Cdr1-protein-mediated R6G efflux in the presence of FK520. The X-axis (1/S) represents the indicated concentrations (μM) of R6G, and Y-axis (1/V) shows the rate of release of R6G in the absence and in the presence of (2×), (6×) and (8×) of FK520. The rate of each reaction was calculated as nmoles of R6G released min−1 mg−1 dry weight of cells. The energy-dependent R6G or NR efflux was initiated by adding glucose (2%, indicated by an arrow) and quantified by measuring the absorbance of the supernatant at 527 nm. The values indicated by the bars represent the means ± standard deviations (indicated by error bars) of three independent experiments. (e) Percentage growth inhibition of control cells (AD1–8u) and of cells over-expressing ABC/MFS transporters in presence of indicated concentrations of FK520. The experiments in triplicates were conducted as described in Materials and methods. The values represent mean ± SD of three independent experiments.

The Lineweaver–Burk plot of R6G efflux revealed that FK520 competitively inhibited the efflux of R6G associated with an increase in km values (7.87–25 μM), with no change in Vmax for R6G (Fig. 1d).

FK520 is not a substrate of MDR pumps

The fact that FK520 inhibits drug transport suggested that it could be acting as a competing substrate of the multidrug transporters. The percentage growth inhibition of yeast cells control (AD1–8u) and ABC/MFS transporter-expressing cells (AD-CDR1, AD-PDR5, AD-MDR1) by FK520 was similar in all the cells over-expressing either of the transporter and was comparable with host cells (Fig 1e). In other words, if FK520 was a substrate of Cdr1 protein, one would have expected less inhibition of growth due to the efflux of the inhibitor. This implied that the over-expression of different pump proteins could not affect the percentage inhibition of growth, which was comparable with host cells. It would mean that FK520 could not be transported by the efflux pump proteins over-expressing cells.

FK520 interacts with the select residues of Cdr1 protein

We observed that the Cdr1 protein over-expressing cells (AD-CDR1) showed synergy with azoles, which was also reflected in the reduced efflux of R6G and H3-FLC in presence of nonlethal concentration of FK520 (data not shown). For evaluating the molecular interaction of FK520 with Cdr1 protein, we screened the entire library of mutant variant of TMDs of Cdr1 protein (Rawal et al., 2013) for their ability to reverse the fungicidal synergy between drugs and the inhibitor and if that could also be correlated with the efflux of drug. Notably, in our screen, we left out 55 mutant variants, which upon replacement with alanines became highly susceptible to azoles. The fungicidal synergy was evaluated in remaining 197 mutant variants, which were inconsequential with respect to their susceptibility toward the drugs such as miconazole, ketoconazole, and fluconazole. Our screen revealed that while most of the Cdr1 protein mutant variants expressing cells behaved similar to the native protein-expressing cells (data not shown), there were 32 (13%) variants, which showed the reversal of the fungicidal synergy with the drugs (Fig. 2a). The reversal of synergy was also reflected in R6G efflux, which also could not be blocked by FK520 in these mutant variants expressing cells (Fig. 2b). Thus, a single substitution of an amino acid of Cdr1 protein could reverse the ability of FK520 to abrogate the efflux of drugs.

2

Screening of mutant library of Cdr1 protein to establish synergistic and modulatory role of FK520. (a) To establish the FK520 synergy, cells expressing mutant variants of Cdr1 protein were screened for the reversal of fungicidal synergy between FK520 and indicated drugs by employing spot assays. The panel depicts spot assays of those mutant variants (32 amino acid variants) expressing cells that showed reversal of synergy. (b) The panel depicts the extracellular concentrations of R6G recovered after efflux assays of all the mutant variants expressing cells included in upper panel (except L563A, which is susceptible to R6G). Y-axis depicts extracellular concentration of R6G with and without FK520.

FK520 interactive residues are distributed selectively among transmembrane helices

The interactive residues whose individual replacement to alanine abolished FK520-dependent synergy and failed to show any impact of FK520 on the efflux of R6G were found to be scattered to almost all the transmembrane helices (TMHs) of both the TMDs. However, TMH5, TMH8, and TMH9 appeared to harbor the most of the interactive residues implying specific and preferred interaction of FK520 with the helices of the transporter protein (Fig. 3a).

3

Critical residues of Cdr1p involved in FK520 interaction. (a) The cartoon depicts helix wise distribution of the critical residues of Cdr1 protein whose replacement leads to the reversal of fungicidal synergism and abrogation of efflux. (b) Venn diagram depicts critical residues, which reversed the fungicidal synergy of different azoles.

With the data set, we could also compare important residues for displaying reversal of the synergy with different azoles. The Venn diagram of Fig. 3b depicts all the residues that were critical to FK520 interaction with different drugs. It is apparent from the comparison that several interactive residues for FK520-dependent reversal of synergy were common among all the three tested azoles. However, there were quite a few residues, which were exclusive for a specific drug. For example, the single replacement of M667, V674, M1229, I1290, or of W1486 residue was specific to the reversal of fungicidal synergy between fluconazole and FK520 while the replacement of C632 or F1239 exclusively affected the synergy between miconazole and the inhibitor. These data also coincide with our previous finding where while mapping the entire drug-binding pocket, it was revealed that while each drug has a set of exclusive interactive residues for the drug binding and transport, there were several residues, which were critical for all the drugs (Rawal et al., 2013).

FK520 shows Cdr1-protein-dependent chemosensitization of azole-resistant isolates

In the following experiment, we could show that FK520 acts as antagonist to selectively chemosensitize azole-resistant (AR) clinical isolates of C. albicans. The checker board assays revealed that AD1–8u cells that lacked seven ABC transporters including Cdr1 protein could elicit synergy between FK520 and azoles (Fig. 4). However, there was a dramatic effect when we compared it with Cdr1 protein over-expressing AD-CDR1 cells, which showed the reversal of azole resistance in the presence of the inhibitor resulting in up to 100-fold change in minimum inhibitory concentrations (MICs; Fig. 4a). Notably, because of highly susceptible AD1–8u cells for fluconazole (MIC 0.5 μg mL−1), one could not see synergy between fluconazole and FK520; however, the inhibitor could show fungicidal synergy in AD-CDR1 cells with 16-fold lower MIC than the drug alone (Fig. 4a). The Cdr1-protein-dependent reversal of resistance was best illustrated, when we used matched pair of clinical isolates. Gu4 and Gu5 are pair of azole-susceptible (AS) and AR clinical isolates, which were recovered from a single patient over the period of time (Sanglard et al., 1995; Franz et al., 1998, 1999). It is shown that high resistance to azoles in Gu5 cells is majorly due to an over-expression of Cdr1 protein. Interestingly, the resistance to azoles could be totally reversed by FK520 in Gu5 cells, which showed up to 32-fold change in MICs as compared to the susceptible Gu4 cells, where only twofold change in MIC was observed (Fig. 4a). The Cdr1-protein-dependent interaction with FK520 was further evident from another pair of AS (F2) and AR (F5) isolates where resistance is attributed to an over-expression of Mdr1 protein. FK520 could not show any demonstrable difference in fold change in MICs between F2 and F5 cells implying that the over-expression of MFS transporter MDR1 could not impact the synergy between the inhibitor and azoles (Fig. 4a). This was also well supported by the NR efflux assays performed with AS/AR clinical isolates. The NR efflux in Gu5 cells was abolished more than 50% by the presence of FK520 while it had no such impact in F5 cells (Fig. 4b).

4

FK520 shows Cdr1-protein-dependent synergy with the drugs. (a) The table depicts fold changes in MIC values to demonstrate restricted interaction of FK520 with multidrug transporters by checker board assays with AD1–8u and AD-CDR1 and clinical isolates Gu4/Gu5 and F2/F5. The isolates GU5 and F5 are highly resistant to fluconazole due to an over-expression of Cdr1 protein and Mdr1 protein, respectively. (b) The transport of NR in the presence of FK520 was observed in Gu4/Gu5 and F2/F5 in clinical isolates. As R6G is not transported by Mdr1 protein, NR efflux was used for comparison with both the pair of isolates Gu4/Gu5 and F2/F5 strains.

Discussion

The fungicidal synergism of cyclosporine A and FK506 in combination with fluconazole is well established (Cruz et al., 2002; Uppuluri et al., 2008). This synergy is exerted via calcineurin-conserved core stress pathway in C. albicans. The ability of these inhibitors to inhibit drug extrusion mediated by efflux pump proteins led to the suggestions that multidrug transporters could also contribute to the observed synergy between azoles and FK506. In this study, we used another homolog inhibitor FK520 to demonstrate that indeed it could inhibit the drug efflux mediated by ABC efflux proteins of C. albicans. We also show that the modulatory effect of FK520 in preventing drug efflux is the result of its specific interaction with select amino acids of Cdr1 protein. Notably, the residues, which are critical in inhibiting the efflux of R6G or NR, upon mutation also abolished the fungicidal synergy. This further reinforces an efflux-pump-dependent synergy between FK520 and azoles.

It was further demonstrated with the use of genetically matched AS and AR clinical isolates of C. albicans. The data show that the reversal of fungicidal synergy with azoles could be achieved by ABC-transporter-dependent (as in Gu5) and ABC-transporter-independent mechanisms (as in Gu4 and F2; Fig. 4a). Thus, we provide evidence for a direct interaction of FK520 with the select intrahelical amino acids of Cdr1 protein that plays an important role in displaying fungicidal synergism independent of calcineurin pathway. Considering the other modulators of Cdr1 protein such as curcumin, disulfiram and farnesol, which potentiate azole sensitivity in Candida cells (Shukla et al., 2004; Sharma et al., 2009; Sharma & Prasad, 2011), it would be worthwhile to evaluate whether their potency is also dependent upon their specific interactions with the ABC or MFS multidrug transporter proteins. The screening of entire Cdr1 protein mutant library with other modulators is expected to reveal the putative binding sites necessary to block the azole transport, which would help in the rational drug designing and new therapeutic strategies. The molecular interaction of FK520 was restricted to ABC proteins only as efflux pump proteins belonging to MFS super family did not interact with the inhibitor. Thus, FK520 could not inhibit NR efflux from AR strain (F5), which showed over-expression of MFS transporter Mdr1 protein. This was not the case with another pair of AS/AR clinical isolates (Gu4/Gu5), which showed enhanced expression of Cdr1 protein (Sanglard et al., 1995). The enhanced modulatory effect of FK520 upon Cdr1 protein in AR isolate (Gu5) was also accompanied with the reversal of synergy. The efflux-pump-dependent synergy was further confirmed when we used AD1–8u and AD-CDR1 cells. Although FK520 in combination with azoles could show synergy in AD1–8u cells but it was several folds pronounced in AD-CDR1 cells where Cdr1 protein was over-expressed. The drastic change in MIC from 16 to 100-fold between AD1–8u and AD-CDR1 cells well illustrates the interaction of FK520 with the transporter protein. The library of mutant variants of Cdr1 protein helped to elucidate the molecular basis of FK520-dependent fungicidal synergy and its ability to modulate efflux of drugs. A set of several mutant variants of Cdr1 protein showed that a single substitution of an amino acid with alanine could result in the reversal of fungicidal synergy and in inhibitor's ability to abrogate efflux of drugs. Notably, such critical residues of Cdr1 protein, which interact with FK520, are distributed across various TMHs; however, few TMHs harbor more responsive critical residues than others. Together, our results demonstrate specific molecular interaction between select intrahelical residues of Cdr1 protein and the inhibitor FK520 that not only impact the efflux activity but also affect synergy between the inhibitor and the drugs.

Acknowledgements

This work was supported in parts by research grants to RP from Department of Biotechnology, Government of India (BT/01/CEIB/10/III/02) and (BT/PR 14879/BRB 10/885/2010). None of our co-authors have conflict of interest to declare.

Footnotes

  • Editor: Carol Munro

References

View Abstract