Elucidating the Eradication Mechanism of Perillyl Alcohol against Candida glabrata Biofilms: Insights into the Synergistic Effect with Azole Drugs

Abstract

Increased incidences of fungal infections and associated mortality have accelerated the need for effective and alternative therapeutics. Perillyl alcohol (PA) is a terpene produced by the hydroxylation of limonene via the mevalonate pathway. In pursuit of an alternative antifungal agent, we studied the effect of PA on the biofilm community of Candida glabrata and on different cellular pathways to decipher its mode of action. PA efficiently inhibited growth and eradicated biofilms by reducing carbohydrate and eDNA content in the extracellular matrix. PA reduced the activity of hydrolytic enzymes in the ECM of C. glabrata biofilm. The chemical profiling study has given insights into the overall mode of action of PA in C. glabrata and the marked involvement of the cell wall and membrane, ergosterol biosynthesis, oxidative stress, and DNA replication. The spectroscopic and RT-PCR studies suggested a strong interaction of PA with chitin, β-glucan, ergosterol, and efflux pump, thus indicating increased membrane fluidity in C. glabrata. Furthermore, the microscopic and flow cytometry analysis emphasized that PA facilitated the change in mitochondrial activity, increased Ca²⁺ influx via overexpression of voltage-gated Ca²⁺ channels, and enhanced cytochrome C release from mitochondria. In addition, PA interferes with DNA replication and thus hinders the cell cycle progression at the S-phase. All these studies together established that PA mitigates the C. glabrata biofilms by targeting multiple cellular pathways. Interestingly, PA also potentiated the efficacy of azole drugs, particularly miconazole, against C. glabrata and its clinical isolates. Conclusively, the study demonstrated the use of PA as an effective antifungal agent alone or in combination with FDA-approved conventional drugs for fungal biofilm eradication.

Introduction

Globally, considerable mortality and economic burden have been contributed by fungal pathogens that constitute the human mycobiome and reside in the gastrointestinal (G.I.) tract.¹ A compromised immune state of the host turns these friendly microbes into hostile ones that eventually invade the G.I. tract and reach other body parts through the bloodstream.² Members of the genus Candida are clinically the most prominent human fungal pathogen responsible for superficial and deep invasive fungal infections.³ Nearly 15 Candida species responsible for causing diseases are known worldwide, of which C. albicans is the leading cause of invasive candidiasis, followed by C. glabrata, C. parapsilosis, and C. tropicalis. However, significantly higher mortality was observed in patients with C. glabrata infections in the United States.⁴ Antifungal susceptibility data highlighted the multidrug-resistant trait of C. glabrata, as it is resistant to the majority of conventional drugs.⁵ The virulence traits and phenotypic features of C. glabrata are beneficial for its pathogenicity and make it different from other Candida species.⁶

The hallmark feature of clinically relevant fungal species is the biofilm formation, a complex, multicellular structure of surface adhered, metabolically least active cells encased in self-secreted polymeric substance called extracellular matrix.⁷ Biofilm formation is a multistep process comprising cell adhesion, cell growth and division, ECM secretion, maturation, and dispersion.⁸ The fungal cells depict maximum resistance to antifungals during the later stages of biofilm development, i.e., mature biofilm due to the dominance of metabolically less active persister cells.⁹ Moreover, high cell density, limited nutrients, and overexpression of drug efflux transporter and ergosterol genes in biofilm cells contribute to the drug-resistant phenotype of Candida biofilm cells.¹⁰

A promising avenue in antimicrobial research is to explore phytoactive molecules or secondary metabolites that have been in use from ancient traditional medicine. Essential oils (EOs) are rich sources of phytoactive molecules and have been gaining popularity because of their safety, wide potential applications, and significant consumer acceptance.¹¹ EOs are a mixture of volatile, lipophilic compounds that often belong to terpenes, terpenoids, phenols, and flavonoids.¹² Among these, terpenes are the most diverse and abundant class of phytoactive molecules with potent antimicrobial applications.¹³ Perillyl alcohol (PA) is a monoterpene present in various EOs like lavender, peppermint, lemongrass, etc.¹⁴ PA is known to exhibit wound-healing, antiproliferative, immune-modulatory, antiaging, and hypoglycemic activity.¹⁵ Preclinical studies of PA on a malignant brain tumor showed promising results.¹⁶ Furthermore, it has been reported that PA is also a potent antioxidant and neuroprotectant.¹⁷ The antifungal activity of PA has been reported against planktonic and sessile growth forms of Candida species with little evidence on antifungal activity, particularly in C. albicans.^18,19

The present study investigated the antifungal activity of PA in C. glabrata, and its clinical isolates against planktonic and sessile growth form to increase the likelihood of transforming PA into a therapeutic product. The impact of PA on the biochemical composition, the molecular and stress response pathways, the cell cycle, and the membrane permeability in C. glabrata, along with its clinical isolate(s)/gene-knockout strains, has been deciphered. To further establish the antifungal potency of PA as an adjuvant, we studied its interaction with azole and nystatin by the checkerboard method. The data suggested that PA effectively inhibited C. glabrata growth, both alone and in synergy with azoles. PA modulated different cellular biosynthetic pathways, cell cycle progression, and expression of several genes to eradicate C. glabrata biofilms.

Results

Growth Susceptibility of C. glabrata Isolates to PA

The growth inhibiting and killing concentrations of PA against C. glabrata strains were obtained by the broth microdilution method and are presented in Table 1. The MIC₉₀ (minimum inhibitory concentration) and MFC (minimum fungicidal concentration) values of PA are 200 and 250 μg/mL, respectively, for C. glabrata wild-type strain, which is the lowest among all the tested strains. For clinical isolates (CCG1–4), the MIC₉₀ value ranged between 250 and 350 μg/mL, whereas MFC was in the range of 300 to 450 μg/mL (Figure S1). Further, both the biofilm inhibiting and eradicating concentrations of PA were estimated in terms of BIC₈₀ (the concentration at which 80% of biofilm formation is inhibited) and BEC₈₀ (the concentration at which 80% of preformed biofilm eradicated) respectively (Table 1; Figure S2). The BIC₈₀ value of PA was observed to be lowest for CG, 200 μg/mL; and the highest BIC₈₀ value was observed to be 400 μg/mL for CCG3. A similar trend was observed for BEC₈₀ values of PA against tested CG strains. The observed variations in the susceptibility of CG strains toward PA suggests differential physiological features that are responsible for the strains’ virulence. On the basis of these susceptibility results, the wild-type strain CG and the resistant clinical isolate CCG3 have been chosen for further detailed mechanistic studies.

Table 1. MIC₉₀, MFC, BEC₈₀, and BIC₈₀ Value (μg/mL) of PA for C. glabrata and Its Clinical Isolates.

	MIC90	MFC	BIC80	BEC80
CG	200	250	200	400
CCG1	250	300	250	450
CCG2	300	350	300	500
CCG3	350	450	400	>500
CCG4	300	400	300	450

To dissect the PA’s mode of antifungal action against C. glabrata biofilm cells, we visualized the topological features in biofilm structure using field-emission scanning electron microscopy (Figure 1A). Microscopic pictures of CG and CCG3 control biofilms depicted an intact and complete biofilm structure, where the cells were healthy and oval to elongated in shape. However, upon PA exposure (150 μg/mL), fragmented patches of biofilm were observed with severe damage to the morphology of cells. These observations evidence the potency of PA in disintegrating CG and CCG3 biofilms by reducing the extracellular matrix (ECM).

Figure 1.

Effect of PA on C. glabrata biofilm. (A) FESEM micrographs of control and PA-treated biofilm cells of CG and CCG3. Scale bar represents 50 and 5 μm. Biochemical analysis of the ECM components of CG and CCG3 in the presence and absence of PA (100 μg/mL): (B) carbohydrate, (C) protein, (D) eDNA, (E) proteinase activity, and (F) phospholipase activity.

Compositional Analysis of ECM of PA-Treated Biofilm

The ECM is extremely crucial for biofilm integrity and the sustenance of a pathogen in a host. As observed in FESEM (field-emission scanning electron microscope) micrographs, PA dismantled the biofilm and reduced the ECM content. Thus, to validate the impact of PA, we studied the biochemical and enzymatic components of ECM of CG and CCG3 biofilms (Figure 1B–G). Among the biochemical components, the carbohydrate content was considerably increased, whereas the total protein was significantly decreased in the ECM of biofilms for both the strains (Figure 1B, C). The eDNA content of the ECM remains unchanged for the CG biofilm, whereas it increased significantly for the CCG3 biofilm upon PA exposure, thus suggesting differential biochemical responses (Figure 1D).

Further, the activity of hydrolytic enzymes such as proteinase and phospholipase present in the ECM were measured for PA-treated CG and CCG3 biofilms. The results suggested that the activity of both the proteinase and phospholipase were decreased considerably in both wild-type and clinical isolate as compared to untreated control upon treatment of with PA (Figure 1E, F). Further, it has been observed that the loss of activity of these hydrolytic enzymes were significantly more in the CCG3 as compared to CG. These results suggest that PA differentially regulates the ECM biochemical composition of CG and CCG3 to dismantle the biofilm.

Growth Response to Pathway Stressors and PA

The chemical profiling method has been employed to delineate the putative cellular pathways involved in PA-mediated antifungal activity against CG and CCG3. The sensitivity in CG and CCG3 growth toward stressors of different pathways in response to PA at subinhibitory concentration was determined using a plate spot assay (Figure 2). The stressors that mimic the pathways related to cell wall, cell membrane, oxidative stress, DNA damage/replication, ergosterol synthesis, endoplasmic reticulum, and mitochondrial activity that are often responsible for cell survival have been targeted. The interpretation on the involvement of a pathway in mediating the PA response is based on the growth sensitivity of the cell toward stressors. In principle, cells susceptible to a stressor in the presence of PA suggest direct/indirect involvement of that particular pathway in regulating the antifungal activity of PA.

Figure 2.

Effect of PA (100 μg/mL) on different cellular pathways of CG and CCG3 in the presence of subinhibitory concentrations as studied by plate spot assay. Ascorbic acid (AA, 4 mM); calcofluor white (CFW, 10 μg/mL); Congo red (CR, 50 μg/mL); dithiothreitol (DTT; 15 mM); ethylmethane sulfonate (EMS; 0.0125%); hydroxyurea (HU, 25 mM); miconazole (Mcz, 1 μg/mL); nitric oxide (NO, 3.75 mM); nystatin (Nyst, 2 μg/mL); tert-butyl-hydroperoxide (TBO, 1 mM); cobalt chloride (CoCl₂, 0.75 mM); calcium chloride (CaCl₂, 0.3 M); and sodium dodecyl sulfate (SDS, 0.02%).

The results suggested that the growth of both CG and CCG3 are compromised upon PA treatment in the presence of CFW (calcofluor white), CR (congo red), NO (nitric oxide), and HU (hydroxyurea) in both the strains. However, PA affected the growth of CG only in the presence of Mcz (miconazole) and Nyst (nystatin); whereas no sign of compromised growth was observed in CCG3. Likewise, upon SDS stress, the growth of CCG3 was found to be compromised, whereas that of CG remained unaffected. This discrepancy in the sensitivity to stressors between the strains points to the susceptibility variations in the two strains as the reason and highlights the participation of specific cellular pathway in mediating the antifungal activity of PA. In summary, the putative cellular pathways accountable for PA-mediated antifungal activity in CG and CCG3 include the cell wall and membrane synthesis, ergosterol biogenesis, oxidative stress, and the cell cycle. Henceforth, the impact of PA on the identified cellular pathways was substantiated through pathway-specific assays in CG and CCG3 and knockout mutants of C. glabrata.

PA Interacts with Cell Wall Components

The fungal cell wall is a two-layered structure, where the inner layer is composed of chitin and the outer layer has glucan and mannan.²⁰ The effect of PA on chitin and β-glucan of the CG and CCG3 cell wall was estimated using β-1,3-glucanase and CFW, respectively. Spectroscopic values depicted no difference in the cell lysis activity of β-1,3-glucanase in CG cells with or without PA until the first 140 min of incubation (Figure 3A). In the presence of PA, the cell lysis values for CG were significantly different from those of the untreated control from 160 min, and this difference kept on increasing with every 20 min time interval. After 260 min, the CG absorbance value was reduced to nearly 62 and 78% in the presence and absence of PA, respectively. These data suggest that PA potentiated the activity of the β-1,3-glucanase enzyme by interacting with β-glucan, thus making the cell wall fragile enough to facilitate cell lysis. In contrast to this, PA treatment did not show any sign of increased cell lysis in CCG3 in comparison to the control even after 260 min of incubation. This incongruity in CG and CCG3 sensitivity to β-1,3-glucanase in the presence of PA revealed the differences in the glucan content between the strains.

Figure 3.

Effect of PA exposure on cell wall and cell membrane components of CG and CCG3: (A) Spectroscopic estimation of β-glucan using β-glucanase; (B) estimation/visualization of chitin under fluorescence microscopy using CFW stained cells; (C) growth susceptibility of Cgkre1Δ, Cgkre2Δ, and Cgecm33Δ toward SDS in the presence of PA analyzed by plate spot assay. (D) Analysis of plasma membrane fluidity by measuring the DPH intensity in PA-treated cells. (E) Spectroscopic measurement of ergosterol content and (F) growth susceptibility of Cgerg5Δ and Cgupc2BΔ toward Mcz and NYST in the presence of PA as analyzed by plate spot assay. (G) Microscopic analysis of the activity of ABC transporters using R6G. All scale bars represent 50 μm.

Further, the influence of PA on cell wall chitin was estimated in CFW-stained C. glabrata cells using FACS and then visualized under a fluorescence microscope. CFW is a chemi-fluorescent blue dye that binds to chitin; any variation in its intensity reflects the changes in the chitin content. The fluorescence intensity of CFW-labeled CG cells was decreased from 48 ± 2 to 12 ± 2 upon PA treatment at 100 μg/mL (Table S1). Although no noticeable change was observed in fluorescence intensity of CCG3 (52 ± 3) at 100 μg/mL PA (47 ± 1), a considerable reduction in intensity was recorded at 150 μg/mL of PA (31 ± 2). Similar observations on CFW intensity change were marked from fluorescence microscopy images of CG and CCG3 cells upon PA exposure (Figure 3B). The blue fluorescence intensity of CG was decreased with increasing concentration of PA (100 to 150 μg/mL), and the significant change in CFW intensity in CCG3 cells was observed only at highest concentration of PA (150 μg/mL) tested. The results indicated that the PA targeted the chitin present in cell wall where CG is more sensitive to that of CCG3.

As the observations of glucanase and CFW studies pointed toward the action of PA on cell wall components, the effect of PA on the growth of cell wall biosynthesis knockout strains of C. glabrata was also investigated in the presence of SDS (Figure 3C). The knock out strains considered were Cgkre1Δ (cell wall glycoprotein), Cgkre2Δ (glycosylation and mannoprotein synthesis), and Cgecm33Δ (GPI-anchored protein involved in cell wall synthesis and organization). The increased sensitivity of all the chosen knockout strains in the presence of PA toward SDS further substantiated the results of interference of PA in chitin and glucan biogenesis, thus establishing the fact that PA attacks the cell wall of C. glabrata.

PA Perturbs Cell Membrane Stability

Membrane fluidity is crucial for cell survival and functioning. As PA is hydrophobic in nature, it has to perturb/interfere with the membrane stability in order to cross plasma membrane.²¹ To dissect the membrane perturbing nature of PA, the membrane fluidity was studied using DPH that fluoresce upon intercalation with lipid membrane. The fluorescence intensity of DPH in CG and CCG3 cells increased significantly in the presence of PA (100 μg/mL) as compared to control cells suggesting increased membrane fluidity (Figure 3D). Upon further increase in PA concentration (150 μg/mL), the DPH fluorescence was reduced in both strains but more significantly in CG which might be due to increased cell death in response to PA. The change in DPH fluorescence intensity reflects communication of PA with plasma membrane and its components like ergosterol and drug efflux pumps (ABC transporters).

To further confirm this, we elevated the ergosterol content and activity of ABC transporters of CG and CCG3 in the presence of PA. The ergosterol estimation was performed in CG, CCG3 and knock out mutants (Cgupc2BΔ, Cgerg5Δ) of ergosterol pathway. Erg5p is a sterol desaturase and Upc2Bp is a transcriptional regulator of ergosterol biosynthesis. The spectroscopic data revealed that with varying concentrations of PA, the ergosterol content increased and then decreased in CG cells (Figure 3E). However, in CCG3, the ergosterol amount was increased only in the presence of 150 μg/mL of PA. The ergosterol content was decreased considerably in both knockouts (Cgupc2BΔ, Cgerg5Δ) at 100 μg/mL of PA, whereas no significant amount was detected at the highest concentration of PA. Though PA differentially affected ergosterol content, it is evident that in both of the strains PA is interacting with membrane ergosterol. Further, in the presence of PA along with ergosterol biosynthesis pathway stressors (MCZ, NYST), the growth of both Cgerg5Δ and Cgupc2BΔ were significantly compromised (Figure 3F), thus evidencing the interference of PA with ergosterol biosynthesis in CG and CCG3.

Further, the activity of ABC transporters was investigated in CG and CCG3 cells in the presence of PA using R6G fluorescence. The amount of R6G retained in the cell after PA treatment is indirectly proportional to the ABC efflux pump activity. Untreated CG and CCG3 cells emitted almost no fluorescence because of complete efflux of R6G, suggesting the highest activity of efflux pumps (Figure 3G). The fluorescence of CG and CCG3 cells was increased upon PA exposure, thus highlighting the PA-mediated inhibition of ABC efflux pump activity. Therefore, all the above results of ergosterol and ABC efflux pumps in CG and CCG3 stand together with the outcome of membrane fluidity, establishing the fact that PA targets the membrane permeability for its antifungal activity.

PA Generates Oxidative Stress and Damages Mitochondria to Disrupt CG Biofilm

Oxidative stress generation is the broadly adopted antimicrobial strategy of phytoactive molecules in response to which the cell activates the antioxidant system.²² Different parameters that could be utilized as markers of oxidative stress generation include activity of antioxidant enzymes, ROS estimation, change in mitochondrial membrane potential (MMP), cytochrome C, etc. In the present study, all these oxidative stress indicators were estimated in the presence of PA in CG and CCG3 cells. Upon PA treatment, the activity of catalase enzyme was moderately increased in CG biofilm cells and significantly enhanced in CCG3 biofilm cells (Figure 4A). This enhanced catalase activity, which is part of antioxidant system indicates the potential generation of ROS (reactive oxygen species) in the presence of PA. Therefore, to substantiate the observations, we stained PA-treated CG and CCG3 biofilm cells with DCFDA fluorescent dye to measure the ROS and counter stained them with PI to estimate the ROS-mediated cell death (Figure 4B, C). The fluorescence intensities of both the dyes were increased in PA-treated CG and CCG3 cells, thus suggesting the involvement of PA in ROS generation and associated cell death.

Figure 4.

(A) Spectroscopic estimation of the catalase enzyme activity in CG and CCG3 cell cytoplasm upon PA treatment. The fluorescence intensity of (B) DCFDA and (C) PI measured in PA exposed CG and CCG3 cells at 100 μg/mL and 150 μg/mL of PA. (D) Effect of PA on mitochondrial membrane potential (MMP) as evaluated by flow cytometry. Graph represent the fluorescence intensity of CG, CCG3, and Cgatm1Δ cells labeled with Rho-B. (E) Rho-B labeled cells visualized under a fluorescence microscope. All scale bars represent 50 μm. (F) Growth susceptibility of Cgatm1Δ toward oxidative stress in the presence of PA as studied by a plate spot assay. (G) Spectrophotometric analysis depicting the effect of PA (100 μg/mL) on oxidative stress and Cyt C release (in both the mitochondria and the cytosol) in CG and CCG3.

In general, oxidative stress results in modulation of MMP, cytochrome C (cyt C) release, ATP production, and apoptosis. Therefore, the role of PA on MMP in CG, CCG3, and Cgatm1Δwas evaluated using a cationic fluorescent dye, rhodamine B (Rho B) (Figure 4D, E; Table S2). ATM1 encodes an ATPase-coupled transmembrane transporter in mitochondrial membrane. FACS analysis of CG cells evidenced for 2-fold increase of fluorescence intensities in the presence of 100 μg/mL of PA, whereas the intensity of cells reduced to similar level of control cells at 150 μg/mL of PA (Figure 4D). In CCG3, irrespective of PA concentration, the fluorescence intensity was increased 6–7 folds in comparison to untreated control. It has been observed that the Rho B fluorescence intensity in Cgatm1Δ cells was as similar to that of CG cells upon PA treatment. The observations of Rho B fluorescence microscopy presented in Figure 4E were also in agreement with the results of FACS. Further, the highly sensitive growth response of Cgatm1Δ to oxidative stressor (TBO, ascorbic acid) in the presence of PA also substantiated the alteration of MMP in the presence of PA (Figure 4F).

Further, to estimate the extent of mitochondrial damage caused by PA, we measured the amount of Cyt C in mitochondria and cytosol of CG and CCG3 cells upon PA treatment spectrophotometrically. In CG, the Cyt C content was reduced to 43 ± 1.9% in mitochondria and increased to 132 ± 3.2% in cytosol upon PA treatment (Figure 4G). Similarly, in the presence of PA, the Cyt C content of CCG3 cells was reduced in the mitochondria to 51 ± 2.0% and increased in cytosol 143 ± 3.3% (Figure 4G). This increased movement of Cyt C from the inner mitochondrial membrane to cytosol indicates the onset of apoptosis events. All these results collectively establish the regulatory role of PA toward mitochondrial damage and the associated oxidative stress generation that lead to cell death.

PA Inhibits Cell Cycle Progression

PA reduced CG and CCG3 growth in the presence of HU (DNA replication inhibitor) in a plate spot assay (Figure 2). To confirm the involvement of PA in DNA replication events, we studied cell cycle using flow cytometer. Cell cycle study gives an idea about time and frequency of DNA replication, and ultimately the cell division. The PA-treated CG and CCG3 cells were stained with PI, which reflects the amount of DNA bound (Figure 5A). Upon PA treatment, the G1-subpopulation decreased in both CG and CCG3 as compared to control. Interestingly, in the presence of PA the number of S-phase cells increased, whereas that of G2/M decreased in both the strains as compared to control (Figure 5A). This suggests S-phase cell cycle arrest in CG in response to PA, thus preventing the DNA replication.

Figure 5.

(A) Influence of PA (100 μg/mL) on CG and CCG3 cell cycle progression as analyzed by PI-stained cells using FACS. The comparative number of cells calculated in each phase of the cell cycle is shown in the bar graph. (B) Graph representing the fold change in the gene expression of CG and CCG3 cells upon treatment with 100 μg/mL PA. The expression was studied by RT-PCR using SYBR green in terms of Ct value.

PA Differentially Regulated Transcription of CG Genes

From all the above studies, the action of PA on the cell wall/membrane components and DNA replication was evident. These observations were further substantiated with the transcriptional expression study of selected genes in CG and CCG3 using RT-PCR (Figure 5B). The genes undertaken were ERG4 (ergosterol biosynthesis), SWI4 (DNA binding transcription factor), EPA6 (adhesin), CHS1, CHS3 (chitin biosynthesis), ECM7 (Ca²⁺ channel), RNR21 (dNTP synthesis), KRE1 (glucan biosynthesis), and CDR1 (ABC transporter). The expression of genes involved in the cell wall and membrane biogenesis, CHS1, CHS3, ERG4, and CDR1, was differentially modulated. The expression of CHS1 and KRE1 increased in both CG and CCG3, whereas that of CHS3 was upregulated in CCG3 and remained unchanged in CG upon PA treatment. This variation suggests that PA discriminatively interacted with chitin in CG and CCG3, substantiating the CFW staining results. The expression profiles of ergosterol biosynthesis-related genes ERG4 and efflux pump CDR1 were found to be reduced in both the strains upon PA treatment, and CDR1 expression is in line with the R6G studies described above. However, ERG4 expression and the estimated ergosterol content results are in sharp contrast to each other, which could be due to activation of sterol importers, and other feedback mechanisms. Interestingly, the expression of SWI4 and ECM7 were significantly increased in PA-treated CG and CCG3 cells, whereas RNR21 was observed to be reduced, thus supporting the growth-sensitive observations of CG to HU in the presence of PA. These observed fold change data of selective genes support the observed phenomenon of interference of PA in chitin/ergosterol/β-glucan synthesis, DNA replication events and efflux activity of ABC transporter in C. glabrata (Figure 5B).

PA Showed Selective Synergistic Response with Azole Drugs

To assess the suitability of PA as an adjuvant that would enhance efficacy of conventional drugs, we studied the synergistic interaction of PA with fluconazole, miconazole, and nystatin in CG and CCG3 using checkerboard analysis (Table 2). In CG, PA exhibited a synergistic interaction with fluconazole and miconazole, whereas it showed an additive effect with nystatin. In sharp contrast to this, PA showed synergy specifically with miconazole against CCG3; no significant interactions were observed with the other drugs, fluconazole and nystatin. The results suggest the possibility of using PA in synergy with antifungal agent miconazole for inhibiting C. glabrata growth.

Table 2. Interaction of PA with Conventional Drugs Studied by Checker Board Method.

strain	MIC PA (μg/mL)	MIC Flz (μg/mL)	MIC Mcz (μg/mL)	MIC PA + Flz (μg/mL)	MIC PA + Mcz (μg/mL)	FICI (PA + Flz)	FICI (PA + Mcz)
CG	200	64	4	32 + 16	64 + 0.25	0.41	0.38
CCG3	350	256	32		64 + 16	>0.5	0.43

Discussion

Mechanistic Insights into the Biofilm Eradication Potential of Perillyl Alcohol

The prevalence of fungal infections and effective antifungal demand is growing at a farfetched pace. Phytoactive molecule-based antifungal research is showing commendable potential in pursuit of antifungal therapeutic search.²³ The key information lagging to turn PA into a therapeutic adjuvant is mechanistic insights into its antifungal action. Henceforth, the present investigation delineated the effectiveness and mode of action of PA against biofilm growth of C. glabrata and its clinical isolates. PA has efficiently inhibited biofilm formation and eradicated preformed biofilm of C. glabrata, where it has exhibited cidal activity. Similar growth inhibitory activity of PA was evaluated against Fusobacterium nucleatum, Porphyromonas gingivalis, and C. albicans.^18,24 One of the most unique features of phytochemicals is their synergistic effects with the known therapeutic formulations. Various terpenes have been reported to reduce the MIC value of azoles and thus exhibited synergy against fungal cells.^25,26 PA has also exhibited synergy with azole drug(s) fluconazole and miconazole against CG and with miconazole against fluconazole resistant clinical isolate CCG3 (Table 2). Enhanced activity of CDR1 efflux pumps is one of the key reason for low susceptibility of CG toward azole.^27,28 PA has reduced both the activity and the expression of CDR1, thus increasing the antifungal activity of fluconazole and/or miconazole against chosen C. glabrata strains.

Plant-derived molecules often target multiple cellular pathways,²⁹ and in the present study, similar observations were marked with PA. The spotting and fluorescence data evidenced that PA initiate its journey of killing C. glabrata cells by making its way across cell wall and membrane barrier through interaction with different components, namely, chitin, β-glucan, ABC drug efflux pumps, and ergosterol. These inferences are in agreement with earlier report on PA antifungal activity against C. albicans.¹⁹ PA not only physically affected chitin and β-glucan, but also modulated expression of genes (CHS1, CHS2, KRE1) involved in their biosynthesis. In addition, terpenes are known to differentially modulate ergosterol content in a fungal cell membrane. The ergosterol content estimated in CG, CCG3, and knock out mutant strains in the presence of PA has also confirmed a deep impact of PA on C. glabrata ergosterol, which is in line with the earlier studies on geraniol, eugenol and eucalyptol.³⁰⁻³² Besides membrane component, membrane fluidity was also perturbed in response to PA treatment in C. glabrata and CCG3 as observed by DPH fluorescence (Figure 3D). DPH selectively binds to phosphatidylethanolamine (PtdEtn) present on outer leaflet of membrane and reflects membrane fluidity alterations. PtdEtn is usually present in inner leaflet of plasma membrane being transferred and accumulated on the outer leaflet because of the increased activity of Cdr1, thus marking the enhanced membrane fluidity.³³⁻³⁵ Interestingly, the CDR1 expression and R6G efflux results observed in the study were in sharp contrast to the DPH-based membrane fluidity results as both CDR1 expression and activity was found reduced in CG and CCG3 in the presence of PA. These uncorrelated observations potentially suggest the involvement of other pumps in externalization of PtdEtn.

Upon reaching inside the cell, PA targets key cellular pathways either simultaneously or by initiating a sequence of events that include oxidative stress generation, lipid peroxidation, perturbed mitochondrial activity, nucleic acid damage, cell cycle arrest, growth inhibition, promote necrosis, or apoptosis. Boncan et al.³⁶ reported that terpenes generate oxidative stress and activate associated cellular response pathways. ROS generation and enhanced activity of the catalase enzyme has been observed upon PA treatment. The increased ROS has resulted in alteration of MMP, increases Ca²⁺ influx, and Cyt C flow toward cytosol from mitochondria. The presence of an optimum level of Ca²⁺ ion is crucial for mitochondrial functioning or ATP production, and any irregularity in its balance leads to apoptosis.³⁷

Besides oxidative stress generation, PA interfered with cell cycle progression and associated pathways, which mediate the further response accordingly.¹⁷ The cell cycle arrest was observed in citral, eucalyptol at S-phase; linalool, linalyl acetate at G1; geraniol at G1/S; and benzyl benzoate at G2/M.^25,30,31,38 In the present study, cell cycle analysis data exhibit S-phase PA arrest in CG and CCG3 cells upon PA exposure and corroborated with the results of inhibition of DNA replication events and henceforth the cell division (Figure 2). CG and CCG3 showed sensitivity to HU in the presence of PA, indicative of participation of PA in the inhibition of DNA replication, as HU interfere primarily during S-phase and activates intra-S checkpoint pathways.³⁹ In yeast cells, G1/S arrest often results when damage occurs to DNA, and the cell launches repair only if the damage is repairable; otherwise the cell commences apoptosis.^40,41

Altogether, the experimental evidence of the present study provides a plausible hierarchy of the mechanism on the antifungal mode of PA in C. glabrata as depicted in Figure 6. Upon reaching the surface of biofilm community, PA breaches the extracellular matrix barrier by altering the equilibrium among various biochemical components. Following this, once PA comes in contact with biofilm cells, (I) it attacks the cell wall and membrane by interacting with ergosterol, glucan, and chitin, thus altering the membrane fluidity by changing the distribution of PtdEtn in the inner and outer leaflets of the plasma membrane. Such alterations in the cell wall/membrane create passage of entry to PA into the cytoplasm. (II) In cytoplasm, PA effectively blocked the activity of ABC efflux pumps, which altered the cellular transportation and plausible exclusion of antifungal agents, resulting in synergistic antifungal activity with miconazole. (III) PA also initiated ROS generation events and activated the cellular antioxidant system (catalase), changes MMP, enhance Ca²⁺ influx and facilitates Cyt C leakage. All these together could potentially trigger early apoptosis events. (IV) Further, upon entering into the nucleus, PA inhibited the cell cycle progression by interfering with the DNA replication events, and arrested cell cycle in S-phase. (V) PA also modulated the transcription of cell wall/membrane component biogenesis genes. Activation of all these stress pathways in CG hampered the cell signaling events sequentially/simultaneously, and thus did not allowing the cell to counter act the damage, henceforth resulting in its collapse (Figure 6).

Figure 6.

(I–V) Schematic representation of plausible antifungal mechanism of perillyl alcohol in C. glabrata. (I) PA disseminates through ECM and enters the cell by disintegrating the cell wall and membrane components (chitin, β-glucan, and ergosterol) of C. glabrata. Upon reaching the cytoplasm, PA (II) blocks the activity of ABC transporters; (III) generates ROS, increases Ca²⁺ influx and Cyt C release, and results in the onset of early apoptotic events; (IV) halts the cell cycle progression at the S phase; and (V) modulates the transcription of genes.

Conclusions

In conclusion, the current study has established the effectiveness of perillyl alcohol against wild-type and clinical isolates of C. glabrata. PA inhibited the growth by targeting cell wall/membrane biogenesis, generating oxidative stress and perturbing cell cycle events that ultimately hinder cell growth and survival. The MIC of PA alone is 200 μg/mL against C. glabrata and ranged from 250 μg/mL to 350 μg/mL for clinical isolates. Indeed, reports suggested that PA exhibits no sign of toxicity up to 350 μg/mL in a C. elegans in vivo model. The MIC value of PA in CG/CCG3 is further reduced to 64 μg/mL as it exhibited synergy with miconazole. This synergy is the merit of exploiting PA as antibiofilm therapeutic as PA improved azole antifungals effectivity against C. glabrata. The findings of the present investigation further strengthen the use of PA either alone or in combination with the conventional drugs for treating superficial fungal infections as PA is recognized as safe by U.S. FDA and U.S. Flavor and Extract Manufacturers Association (FEMA).

Experimental Section

Strains and Chemicals

The details of all the Candida glabrata strains used in this study were given in Table S3. The strains were routinely maintained on YPD (yeast peptone dextrose) agar plates. For experimental procedures, sabouraud dextrose broth, RPMI (Roswell Park Memorial Institute) and YPD broth were used. The chemicals and media components were procured from Himedia, Sigma-Aldrich, and SRL, India. PA (purity 96%) stock solution of 10 mg/mL was prepared in dimethyl sulfoxide.

Minimum Inhibitory and Fungicidal Concentration

The minimum planktonic growth inhibitory and cidal concentrations of PA against C. glabrata and its clinical isolates were measured using yeast broth microdilution CLSI guidelines.⁴² Briefly, log-phase cells were exposed to different concentrations of PA (ranging from 500 to 50 μg/mL) at a cell density of 2.5 × 10³ cells/mL in RPMI media. The plate was incubated at 37 °C for 48 h. The fungicidal concentration of PA was estimated by spotting the PA-treated cells on YPD agar plates.

Biofilm Studies

Biofilm inhibition and eradication activity of PA against C. glabrata and its isolates was studied in 96-well plates in RPMI media following the methods described elsewhere.⁴³ Briefly, log -phase cells at a cell density of 5 × 10⁷ cells/mL in PBS were seeded into a multiwell plate (MTP). After 90 min of incubation at 37 °C, the PBS was replaced with RPMI with or without PA. After 24 h of incubation, wells were washed and the biofilm was quantified in terms of metabolic activity using XTT reduction assay. For mature biofilm studies, plates were incubated for 48 h with RPMI only at 37 °C and then for 20 h with RPMI containing PA. The absorbance was measured at 492 nm using multimode microtiter plate reader (BIOTEK, Agilent).

FESEM Experiment

Morphological changes in the C. glabrata biofilm cells upon exposure to subinhibitory concentration (100 μg/mL) of PA was observed using FESEM. Both untreated and treated biofilm samples were prepared on the surface of polystyrene disc (1 cm²) as described by Gupta et al.³¹ The biofilm after PA treatment was fixed in 2.5% glutaraldehyde and dehydrated by exposing cells to the gradient of ethanol. The air-dried discs were visualized under FESEM (Carl Zeiss, Germany) at magnification ranging from 1000 to 5000× and 15 kV following gold sputtering.

Plate Spot Growth Assay

The growth response of C. glabrata cells to different stressors in the presence of PA (100 μg/mL) was studied by a plate spot assay.⁴⁴ Briefly, log-phase cells at a density of 1 × 10⁵, 1 × 10⁴, 1 × 10³, 1 × 10², and 1 × 10¹ cell/mL in PBS were spotted on the surface of YPD agar plates containing stressors either with or without PA. The plates were photographed after 18–20 h of incubation at 37 °C.

ECM Extraction and Biochemical Analysis

The biochemical composition of ECM of C. glabrata biofilm was measured in terms of carbohydrate, protein, and eDNA upon PA treatment as per the methods described elsewhere.³² The ECM extracted by sonication was used for carbohydrates estimation using phenol-sulfuric acid. The amount of protein in ECM was measured using a BCA kit, whereas Nanodrop was used for measuring the eDNA content following ethanol precipitation.

Enzymatic Assay

The different enzymes present in ECM were spectrophotometrically estimated in the form of phospholipase, catalase, and proteinase.³² The substrate required for phospholipase estimation includes 1.6 mM phosphatidylcholine, 20 mM AlCl₃, 0.124% BTB, 50 mM Tris-Cl (pH 7.5), and 0.25% triton X-100. The absorbance of the reaction mix was taken at 630 nm. For proteinase activity, ECM was mixed with 1% w/v azocasein and incubated for 1 h, and then the reaction was terminated using TCA. The absorbance of the supernatant was recorded at 440 nm following NaOH treatment. The catalase activity was measured in biofilm cells after their lysis using glass beads. For catalase activity estimation, the absorbance of the cell lysate was recorded after mixing H₂O₂ at 240 nm.

Membrane Fluidity

The effect of PA treatment on the fluidity of the plasma membrane of C. glabrata cells was studied using DPH (1, 6- diphenyl-1,3,5-hexatriene).⁴⁵ The PA-treated log-phase cells were fixed with 0.37% formaldehyde and after washing with PBS, the cells were frozen and thawed twice. Cells were then stained with 0.6 mM DPH for 45 min, and the fluorescence intensity was measured using a fluorimeter at 350 nm excitation and 425 nm emission.

Chitin Estimation

Chitin quantification was performed in mid log-phase C. glabrata cells after 4 h of incubation at 37 °C in YPD containing 100 μg/mL and 150 μg/mL of PA.³¹ The density of washed cells was adjusted to 1 × 10⁶ cells/mL, and then cells were fixed with 4% paraformaldehyde solution. Washed cells were stained with calcofluor white (2.5 μg/mL) for 15 min in the dark. The fluorescence intensity of cells was measured using a BD FACSverse flow cytometer, and the morphology of stained cells was visualized under 60× in a fluorescence microscope.

β-Glucanase Sensitivity

The solidarity of β-1,3-glucan in the C. glabrata cell wall upon PA treatment was analyzed spectrophotometrically.⁴⁶ Briefly, log-phase cells were exposed to a subinhibitory concentration of PA for 4 h and, after washing, were resuspended in reaction buffer (40 mM 2-mercaptoethanol; 50 mM Tris-Cl, pH 7.0) at an O.D. of 1.0. Cells were mixed with 2 U/mL stock of β-1,3-glucanase and the absorbance was measured at 600 nm after every 20 min of incubation at 37 °C for 250 min in a multiwell plate reader (Biotek Inc., Winooski, VT, USA).

Rhodamine 6G Efflux

The efflux pump activity of C. glabrata cells was estimated by a rhodamine 6G uptake assay in the presence of PA.⁴⁷ Overnight grown cells were added into glucose-containing media at a cell density of 1 × 10⁸ cells/mL. After 4 h of incubation at 37 °C, washed cells were resuspended in 0.1 M KCl containing R6G (10 mM). The cells were washed and suspended in 0.1 M KCl with or without PA (100 μg/mL and 150 μg/mL) and cells were visualized after 20 min of incubation at 60× under a fluorescence microscope.

Ergosterol Estimation

The ergosterol content in the cell wall of C. glabrata cells was measured spectrophotometrically by scanning the sample in the range of 230–300 nm. Briefly, log-phase cells in SDB with PA at subinhibitory concentrations (100 μg/mL and 150 μg/mL) were incubated for 24 h at 37 °C. After incubation, the cells were pelleted down and lysed and the sterol was collected and the amount of ergosterol was quantified using the following equation:.³⁰

where F is the dilution factor; 290 and 518 are the E values for crystalline ergosterol and 24(28) dehydroergosterol, respectively.

Mitochondrial Membrane Potential

To investigate the change in mitochondrial membrane potential (MMP) upon treatment with PA (100 μg/mL and 150 μg/mL), we incubated the log-phase cells in SDB broth with and without PA for 6 h.³⁰ After incubation, the cells were suspended in 100 nM rhodamine B containing 10 mM HEPES buffer (pH 7.4). Cells were washed and Rho-B intensity was analyzed using FACS (BD Biosciences FACSVerse, San Jose, California). Also, cells were visualized under fluorescence microscope with excitation at 555 nm and emission at 579 nm.

Reactive Oxygen Species Estimation

In response to PA (100 μg/mL and 150 μg/mL) treatment, the amount of reactive oxygen species (ROS) level in the fungal biofilms was estimated using DCFDA and propidium iodide (PI) mixture.³² DCFDA fluorescence was recorded at 485 nm excitation and 520 nm emission wavelength. The emission and excitation wavelengths of PI were set at 617 and 543 nm, respectively, using multimode microtiter plate reader (BIOTEK, Agilent).

Cytochrome C Measurement

Actively growing cells resuspended at a density of 1 × 10⁷ cells/mL in YPD broth containing PA (100 μg/mL) at 30 °C for 24 h. Cells were pelleted down, washed, and resuspended in the homogenization medium containing 2 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 2% glucose, and 50 mM Tris (pH 7.5). To isolate mitochondria, we centrifuged the supernatant for 45 min at 20 000 rpm. Both supernatant and pellet were used for estimating Cyt C in cytoplasm and mitochondria spectrophotometrically at 550 nm after reduction with ascorbic acid at room temperature for 5 min.

Cell Cycle Analysis

The cell cycle study was performed as previously described by Gupta et al.³⁰ Log phase C. glabrata cells treated with GR for nearly 6 h at 37 °C and then washed twice with PBS (pH 7.0). After washing cells were fixed for overnight in 70% ethanol and were incubated in 50 μg/mL of PI and 10 μg of RNaseA on the next day. After 30 min of incubation at 4 °C, the cells were analyzed using FACS (BD Biosciences FACSVerse, SanJose, California).

Transcription Analysis

The changes in expression pattern of selected C. glabrata genes in response to PA exposure at 100 μg/mL was evaluated by qRT-PCR.³² Total cellular RNA was isolated from PA-treated and untreated log-phase cells using RNA extraction kit (RNeasy kit, QIAGEN). The quality and quantity of extracted RNA was recorded using Nanodrop and used for cDNA synthesis. The sequences of primers used for the expression of transcript study including housekeeping genes are given in Table S4. SYBR green mix (Applied Biosystem), 100 ng of cDNA template, and 200 nM gene specific primers were used. Data are represented as relative fold change, which was determined by the ΔΔCT method using formula 2^–ΔΔCT.

Checkerboard Analysis

The synergistic interaction of PA with fluconazole, miconazole and nystatin was studied by checkerboard analysis in CG and CCG3.⁴⁸ The log-phase cell density in RPMI was adjusted to 1 × 10³ cells/mL. Antifungal drugs, fluconazole, miconazole, and nystatin (0.25 to 128 μg/mL), and PA (2 to 128 μg/mL) were added into a 96-well plate containing 100 μL of cell suspension. The plates were incubated for 24 h at 37 °C before the absorbance was recorded at 600 nm. The fractional inhibitory concentration index (FICI) was calculated using following formula:

A FICI value of ≤0.5 denotes synergism; FICI between 0.5 to 1.0 additive interaction; FICI ≥ 1.0 no interaction; FICI ≥ 4.0 signify antagonism.

Statistical analysis

All the experiments were performed in triplicates and values presented here as mean and standard deviation obtained from the three different observations for each assay. Further, Student’s t test was performed for statistical analysis. A value of P < 0.05 was considered statistically significant (*) and P < 0.01 as highly significant (**).

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsbiomedchemau.1c00034.

• Figure S1, plate spot assay-based MIC and MFC of PA in CG and its clinical isolates; Figure S2, biofilm inhibitory and eradication activity of PA against CG and its clinical isolates; Figure S3, cell cycle analysis in PA-treated CG and CCG3 using FACS; Table S1, geomean fluorescence intensity of CFW-labeled CG and CCG3 cells; Table S2, geomean fluorescence intensity of Rho-B-labeled CG and CCG3 cells; Table S3, strains used in present study; Table S4, sequence of RT-PCR primers (PDF)

Author Contributions

K.M.P. and P.G. conceived the research plan. P.G. performed the experiments and analyzed the data, and P.G. and K.M.P. wrote the manuscript. Both the authors have read and confirmed the final version of the manuscript.

The authors declare no competing financial interest.