Flow Cytometry Is a Powerful Tool for Assessment of the Viability of Fungal Conidia in Metalworking Fluids

ABSTRACT

Fungal contamination of metalworking fluids (MWF) is a dual problem in automated processing plants because resulting fungal biofilms obstruct cutting, drilling, and polishing machines. Moreover, some fungal species of MWF comprise pathogens such as Fusarium solani. Therefore, the development of an accurate analytical tool to evaluate conidial viability in MWF is important. We developed a flow cytometric method to measure fungal viability in MWF using F. solani as the model organism. To validate this method, viable and dead conidia were mixed in several proportions and flow was cytometrically analyzed. Subsequently, we assessed the fungicidal activity of two commercial MWF using flow cytometry (FCM) and compared it with microscopic analyses and plating experiments. We evaluated the fungal growth in both MWF after 7 days using quantitative PCR (qPCR) to assess the predictive value of FCM. Our results showed that FCM distinguishes live from dead conidia as early as 5 h after exposure to MWF, whereas the microscopic germination approach detected conidial viability much later and less accurately. At 24 h, microscopic analyses of germinating conidia and live/dead analyses by FCM correlated well, although the former consistently underestimated the proportion of viable conidia. In addition, the reproducibility and sensitivity of the flow cytometric method were high and allowed assessment of the fungicidal properties of two commercial MWF. Importantly, the obtained flow cytometric results on viability of F. solani conidia at both early time points (5 h and 24 h) correlated well with fungal biomass measurements assessed via a qPCR methodology 7 days after the start of the experiment.

IMPORTANCE This result shows the predictive power of flow cytometry (FCM) in assessing the fungicidal capacity of MWF formulations. It also implies that FCM can be implemented as a rapid detection tool to estimate the viable fungal load in an industrial processing matrix (MWF).

INTRODUCTION

Metalworking fluids (MWF) is a generic term for any chemical used to reduce heat and friction and to safely remove small metal parts in automated mechanical processes such as grinding, cutting, drilling, and polishing (1–7). In general, MWF can be classified into four groups based on their chemical composition, straight oils, emulsifiable oils, semisynthetic MWF, and synthetic MWF (8). Being rich in carbohydrates, these MWF are prone to microbial contamination. The presence of microbial flora in MWF presents a dual problem. First, they hamper the MWF functionality by adhering to surfaces, forming biofilms. These biofilms are complex-structured microbial aggregates that often comprise multiple genera of microbes embedded in a polysaccharide type of matrix (1). Second, as these microbial communities often consist of pathogenic species, they can pose a serious health risk for operators who come into contact with these MWF either directly or through aerosols (8).

In contrast to the vast amount of information on bacterial contaminations in MWF, information on fungal contamination is scarce although fungi represent a large part of the microbial contaminations. Liu et al. (9) found 1.7 × 10⁵ CFU/ml of fungi in in-use contaminated MWF. Fungi in MWF belong to a diverse set of genera, such as Cladosporium, Candida, Cephalosporium, Penicillium, Aspergillus, and Fusarium (8–11). Some of these fungi are pathogenic and associated with hypersensitivity pneumonitis, asthma, and other allergies or skin and eye infections. On top of this pathogenicity, genera such as Penicillium, Aspergillus, and Fusarium contain toxigenic species, which entails extra health issues because these species have the ability to produce toxic secondary metabolites called mycotoxins (12).

The genus Fusarium consists of a diverse group of species and species complexes (13). Members of the genus were previously reported in MWF (8, 11, 14). In addition, fusariosis is the next most common mold infection in humans after aspergillosis (15). Fusariosis is mainly associated with locally invasive symptoms such as keratitis and onychomycosis. Nevertheless, the fungus can also spread, causing disseminated infections (16, 17). Fusariosis is mainly caused by members of the Fusarium solani species complex (SC) and the F. oxysporum SC. Other reports on fusariosis were related to F. incarnatum-F. equiseti SC, F. chlamydosporum SC, and F. dimerum SC (15). To control these fungal pathogens, MWF are provided with antifungal compounds. The biocides 3-iodo-2-propynyl butylcarbamate, sodium ortho-phenylphenate, and 2-N-octyl-4-isothiazolin-3-one have previously been reported in MWF (18, 19). However, their use often entails health issues for the operator, and their efficacy is not always consistent mainly due to the formulation complexity and the changing microbial composition during the life-span of the MWF as well as other factors (19).

There are currently no high-throughput state-of-the-art quantitative methods to monitor either the viability of fungal conidia in MWF or the fungicidal effects of MWF on these conidia. Classic plating techniques are routinely used, but these do not take into account nonculturable fungal conidia that are in a quiescent or dormant state. Moreover, classic plating is time-consuming and does not provide detailed real-time information on the physiological state of the fungal conidia.

In the present study, we therefore pursued a multidisciplinary approach by first comparing the novel flow cytometric method with the microscopic technique to corroborate the real-time viability data provided by flow cytometry (FCM). Second, we further confirmed the FCM data by assessing the fungal regrowth on potato dextrose agar (PDA) medium after a 24-h exposure to MWF. Finally, we used quantitative PCR (qPCR) to determine the fungal biomass after the prolonged incubation of fungal conidia in the MWF. The latter method can confirm the predictive value of the FCM data on the fungicidal capacity of the MWF. This study is timely, as it meets the current increasing needs of industrial stakeholders and the academic world to extend traditional microbiological test methods with novel high-throughput consensus methods which can later be adopted by industrial stakeholders (20).

RESULTS

Optimization of the flow cytometric assessment of F. solani conidial viability.

In the first step, we established a proper negative-control treatment to assess the viability of conidia not exposed to MWF but to each of 4 candidate control media. For this purpose, we kinetically monitored the amounts of viable and dead conidia in phosphate-buffered saline (PBS), distilled water, saline (0.9% NaCl), and potato dextrose broth (PDB) medium at 0 h, 5 h, and 24 h, as shown in Fig. S1. These experiments demonstrated that PBS was the most suitable negative-control treatment, inducing only 8.0% (2.6 to 11.2%, n = 6) dead conidia at 24 h. The three other negative-control treatments gave a higher dead rate at 24 h, and the PDB medium contained the highest percentage of dead conidia at that time point. Based on this preliminary experiment, all subsequent experiments were done with PBS as the negative-control treatment.

In the second step, the flow cytometer settings (i.e., laser voltages, threshold to remove debris, compensation due to bleeding between both fluorochromes, etc.) were optimized to maximally discriminate viable or membrane-intact versus dead or membrane-damaged conidia, which were obtained by boiling them for 30 min at 100^oC. Dead conidia were then mixed with viable conidia in different ratios (i.e., 100:1, 75:25, 50:50, 25:75, and 0:100), and different ratios of green fluorescent nucleic acid stain SYTO 9 to propidium iodide (PI) were tested on these mixed conidial populations. There were no major differences between the different dye ratios (data not shown); therefore, we opted to continue with a 1:1 SYTO 9/PI ratio in further analyses, as also suggested by the manufacturer and previously used by our group for bacteria. This concordance with the bacterial viability assessment also allows further research of the viability of mixed populations of bacteria or fungal conidia.

As MWF are complex matrices containing mineral oils, fatty acids, and biocides among other substances, appropriate sample preparation was required to allow flow cytometric assessment of the conidia present in these MWF. Optimization of the isolation process was done for two selected commercial MWF (MWFA [data not shown] and MWFB). Incubation in MWFA resulted in no residual pollution after a 1-step isolation of the conidia by centrifugation. Direct flow cytometric analysis of samples containing 20% MWFB (the concentration advised by the manufacturer) resulted in residual pollution of the scatterplot compared to the PBS control. This pollution was attributed to the oil vesicles present in this MWF matrix. Therefore, we included several wash steps with 0.9% NaCl to remove it. These optimization data demonstrated that a 3-step washing procedure with 0.9% NaCl resulted in the lowest level of residual pollution (Fig. 1).

FIG 1.

Forward scatter cytogram/sideward scatter cytogram (FSC/SSC) dot plots presenting the isolation of F. solani conidia from the MWFB matrix by performing several wash steps. The A region corresponds to the conidial population, and the B region corresponds to the residual pollution.

Flow cytometric assessment of the viability of F. solani conidia exposed to MWF.

The optimized flow cytometric method was used to monitor the viability of F. solani conidia in the two commercially available MWF (MWFA and MWFB) during 24 h with time intervals at 0 h and 5 h (Fig. 2). In order to simulate suboptimal conditions potentially occurring during practical use of MWF, a dilution series of both MWF was applied. This dilution series comprised concentrations of 20%, 10%, and 5%. For MWFA, none of the tested concentrations resulted in a reduction of the viability of F. solani conidia at 0 h and 5 h compared to the negative control (PBS). At time point 24 h, concentrations of 5%, 10%, and 20% MWFA reduced the viability of F. solani conidia to 76%, 70%, and 23%, respectively. Nevertheless, the efficacy of the highest concentration of MWFA was significantly lower than that of the positive control (prothioconazole fungicide) at all tested time points, and none of the concentrations inhibited F. solani viability completely (Fig. 2a).

FIG 2.

Flow cytometric SYTO 9/PI dot plots presenting F. solani conidial viability at different time points of exposure to MWFA (a) and MWFB (b) compared to PBS (negative control) and prothioconazole fungicide (positive control). The A region corresponds to the subpopulation of viable cells with an intact plasma membrane, and the B region corresponds to the subpopulation of dead cells with irreversibly damaged membranes.

In marked contrast, MWFB displayed a very high to complete fungicidal effect at all tested time points (Fig. 2b). At the time points 5 h and 24 h and at concentrations of 10% and 20%, it showed fungicidal efficacies comparable to (for 10%, both at 5 h and at 24 h; for 20% at 24 h) or even better than (20% at 5 h) those of the positive control.

Remarkably, at time point 0 h, which is the sample taken immediately after exposure of F. solani conidia to each of the two MWF evaluated, a significant reduction of viability was observed at all tested concentrations. Moreover, at this time point MWFB performed significantly better than the positive control (prothioconazole fungicide).

All dot plots serve as representative examples for one run. The data used for statistical purposes, however, are the mean of 3 replicates.

Microscopic assessment of the viability of F. solani conidia exposed to MWF to validate the flow cytometric analysis of conidial viability.

In order to validate the novel flow cytometric method from an industrial point of view, the FCM results were compared with those of classic microscopy to assess the viability of conidia (Fig. 3a and b), as the latter approach is a common practice in microbial studies in industry. However, at time points 0 h and 5 h, none of the treatments showed any sign of conidial germination. At time point 24 h, the negative control (PBS) as well as MWFA at concentrations of 5%, 10%, and 20% showed the beginning of conidial germination, which reflected the viability of the conidia after the PBS and MWFA exposures. Moreover, at time point 24 h the Pearson correlation coefficient between microscopic results and flow cytometric results showed P values lower than 0.001 for all three independent experiments. This significant correlation highlights the congruence between FCM data and microscopy data.

FIG 3.

Comparison of the flow cytometric (FCM) viability assessment (a), the microscopic assessment (microscopy) of conidial germination (b), and the binary scoring of regrowth (c) of F. solani at different time points after exposure to MWFA or MWFB compared to PBS (negative control) and prothioconazole fungicide (positive control). FCM measures the percentage of viable conidia, and microscopy measures the percentage of germinating conidia. The black, gray and white bars represent measurements at 0 h, 5 h, and 24 h of exposure, respectively. Different letters above the bars show statistical differences (i.e., comparison among letters with the same number of prime symbols [′]; P < 0.05). The regrowth of F. solani was assessed when conidia were removed from the wells after 24 h and regrown on PDA plates for 5 days. The symbols show growth (+) or no growth (−) on the PDA plates after 5 days of incubation. Data from all 4 repetitions are shown.

Binary scoring of fungal regrowth after pre-exposure to MWF and prothioconazole.

Using a plating approach, regrowth of F. solani after the fungus had been exposed for 24 h to MWF or prothioconazole was assessed on PDA medium (Fig. 3c). This plating experiment confirmed that none of the tested concentrations of MWFA was fungicidal, as regrowth of F. solani was observed at all MWFA concentrations. However, pre-exposure of F. solani for 24 h to MWFB or prothioconazole resulted in a complete fungicidal effect, as none of the PDA samples showed fungal outgrowth at 5 days after inoculation. This result confirmed the results obtained with FCM (Fig. 3a). Moreover, for the negative-control treatments, all repetitions showed outgrowth when put on PDA plates, whereas microscopic analysis at 48 h showed only a limited amount of germination (Fig. 3b).

qPCR assessment of fungal biomasses to confirm the loss of conidial viability predicted by flow cytometric viability analysis.

Using a qPCR approach, the fungicidal effects of both MWF were additionally assessed after a 7-day exposure of the F. solani isolate (Fig. 4) to confirm the flow cytometric viability data measured at 0 h, 5 h, and 24 h. A concentration of 400 mg/liter of prothioconazole inhibited fungal outgrowth completely, and the fungal biomass, represented by the amount of DNA/sample, equaled the amount of conidia added at time point 0 h. Both MWF efficiently reduced the amount of fungal biomasses at all tested concentrations. Nevertheless, MWFA was again significantly and consistently less effective than MWFB at all tested concentrations. In contrast, at concentrations of 10% and 20% MWFB completely inhibited fungal growth, and the amount of fungal biomass was comparable to the biomass of the spores added at time point 0 h. The lethal effects of 10% and 20% MWFB were similar to results obtained by FCM at time points 5 h and 24 h (Fig. 2b). Moreover, in this concentration range, the effect of MWFB was comparable to the effect of the fungicide prothioconazole at a dose of 400 mg/liter. A concentration of 5% MWFB retarded the fungal growth but did not completely inhibit the fungus, which again confirmed the flow cytometric results (Fig. 2b).

FIG 4.

Evaluation of the fungicidal effects of MWFA (black bars) and MWFB (white bars) via a qPCR-based technique at time point 7 days. Different letters above bars show statistical differences within one MWF (i.e., comparison among letters with the same number of prime symbols [′]; P < 0.05). Asterisks above bars show statistical differences within one treatment. Black and white bars for PBS and prothioconazole treatments represent the same treatment.

For MWFA, none of the tested concentrations was fungicidal, which also confirmed the flow cytometric results obtained at earlier time points.

DISCUSSION

In marked contrast to the wealth of information on the use of FCM to quantify and study the viability of predominantly bacteria and also yeasts in industrial settings (20), information on the use of FCM to study fungi remains scarce. Nevertheless, it is a growing field of interest as industrial stakeholders press for scientific reports introducing novel tools to assess the viability of fungi in industrial matrices. As a result of this need for fine-tuned monitoring techniques, FCM was recently reported for the first time to estimate spore inoculum quality in filamentous bioprocesses (21). Even in nonindustrial settings, there are fewer than 10 reports about the use of FCM to study fungi, including articles on phytopathogenic fungi (22, 23), fungal load in aerosols and air (24–26), disinfection strategies (27), and human-pathogenic fungi in hospital environments (28, 29). To date there are no studies available on the use of FCM for viability assessment of conidia in MWF. Therefore, in the present study we aimed to develop a novel method that allows accurate prediction of the fungicidal potential of MWF using FCM viability measurements.

FCM is a valuable technology for quantifying microbial load in diverse matrices. Moreover, in combination with appropriate fluorescent stains, microbial viability can be accurately assessed. The use of FCM in MWF has been described only by Chang et al. (30), albeit for the detection and quantification of mycobacteria. Moreover, in that study the microorganisms were not isolated from the MWF prior to viability staining, resulting in significant interferences in the viability assessment. One year later, the same group (31) attempted such isolation using immunomagnetic separation and centrifugation and again reported suboptimal results. In contrast, we were able to isolate F. solani conidia when added to MWF by using centrifugation at 4°C and incorporating several wash steps before staining to avoid matrix interference with viability. This isolation and staining protocol was also successfully validated with Trichoderma reesei (data not shown). The genus Trichoderma was also described as an important contaminant of MWF (32), and our technique allowed clear differentiation between live and dead populations of conidia of this additional fungal species.

First, FCM viability measurements and microscopic conidial germination assays were carried out in parallel at 0 h, 5 h, and 24 h. Although we obtained good correlations between the germination data and flow cytometric viability assessment for all three time points, the microscopic analyses consistently resulted in very low percentages of germination, not exceeding 45% at 24 h, pointing to low viability of the conidia. In marked contrast, the FCM measurements showed almost 100% viable populations (SYTO 9⁺/PI^-) in control treatments. This apparent discrepancy likely originates from the fact that PI positivity implies membrane permeability, but that structurally intact conidia might be in a dormant state. Hence, their outgrowth does not occur although their membranes are intact. Such differences between “not viable,” “metabolically inactive,” and “quiescent” or “dormant” conidia were recently reported by Ehgartner et al. (21). In that study, FCM was used to assess conidial viability by dual staining of the conidia, albeit with a different combination, i.e., fluorescein diacetate (FDA), a marker for metabolic activity, and PI. The results of the study also show the significant added value of FCM compared to the traditionally used techniques to evaluate conidial viability, such as assessment of germination by microscopy or determination of CFU. However, these authors used this viability as a quality parameter, independent of the industrial matrix. The existence of similar subpopulations has also been depicted in microorganisms other than fungi (33–35). In analogy with those FCM strategies, it is of critical importance to take into account the number of viable conidia, even if these do not yet germinate at the time of sample assessment. Indeed, the latter quiescent subpopulation might pose contamination issues later on when conditions become more optimal for their germination. This hypothesis was confirmed by our PDA plating assays, in which it was demonstrated that despite the low levels of conidial germination in the negative-control samples at time point 24 h, a proliferate regrowth was observed when conidial suspensions were transferred to PDA plates for 5 days.

For the assessment of fungal load in industrial matrices, the determination of CFU is the most widely accepted method. However, implementation of these culture-based techniques in a complex matrix such as MWF is challenging due to their inherent inaccuracy and time-consuming character, precluding real-time analysis of fungal viability (21, 29). In addition, the predictive power of these techniques is poor, as they provide only a snapshot of the viable fungal load at a given time point. Most important in the context of our research, the potential presence of dormant or quiescent spores is overlooked. The same pitfalls apply at least partially for the use of molecular-based qPCR techniques in this context (14).

As a proof-of-principle for our optimized FCM method, we successfully evaluated the fungicidal capacity of two selected commercial MWF (MWFA and MWFB). MWFB differed from MWFA by the presence of 3-iodo-2-propynyl butyl carbamate, which is a known biocide used to control microbial contamination in MWF (19) among others, while the MWFA formulation contained no fungicide at all. Our novel method nicely demonstrated that, as expected, MWFB had a greater antifungal effect than MWFA. Moreover, we were able to predict the fungicidal effect of both MWF after 7 days of incubation, based on the results seen at 0 h, 5 h, and 24 h of incubation. This promising result indicates that our protocol can lead to a vast array of applications for FCM technology in the screening of MWF for their fungicidal capacity. In general, there is an important added value in using flow cytometry to monitor fungal spore activity in industrial processes, both when fungal spore activity is needed (20) and when it is especially unwanted, as is the case for MWF.

Remarkably, in the microscopic approach, we showed that exposure of conidia to a 5% concentration of MWFA boosted germination compared to conidia suspended in PBS, which displayed hardly any germination. We have two possible explanations for this unexpected result. First, a low concentration of MWFA might provide extra nutrients to the conidia, such as carbohydrates, which are absent in PBS and might induce germination in the former matrix. Another possibility is that these low concentrations induce minor oxidative stress responses such as the formation of H₂O₂. It is known that during germination events, very small amounts of H₂O₂ are beneficial and necessary in the primordial germination and hyphal extension events and involved in de novo synthesis of cell wall and membrane components during germination and hyphal extension (36, 37). Since MWFB was fungicidal at all tested concentrations, these beneficial effects were not observed even at its lowest concentration tested.

Our experiments showed that the predictive power of FCM is much greater than that of microscopy. Predictive power refers to the capacity of a technique to provide reliable predictions of future effects based on short-view data with respect to fungal outgrowth. This was demonstrated in the accumulated F. solani biomass qPCR assay upon exposure to both MWF for 7 days. These qPCR data demonstrated that MWFB was fungicidal at concentrations of 10% and 20%, while MWFA was not lethal for F. solani at any of the tested concentrations. Remarkably, the same conclusion could already be drawn based on the FCM analysis as early as 5 h, while none of the microscopic analyses at any of the tested time points could provide such conclusive results on the effects of MWF on viability of F. solani.

When we compared the microscopic and FCM methods, it became obvious that the microscopic approach was much more labor-intensive and resulted in highly variable results compared to FCM. The latter technology was faster, had a lower error rate, and provided a higher sensitivity compared to the microscopic approach. For these reasons, the FCM method reported here is strongly suggested to be valuable in predicting whether a fungal presence will result in fungal outgrowth and as such hamper industrial processes. Although qPCR analysis is a valid alternative, the cost for FCM is much lower. When comparing the total cost for FCM and qPCR, we calculated that these are comparable with regard to consumables and extraction kits. However, the time needed to perform a DNA extraction and qPCR analysis is significantly longer than the time to perform a FCM analysis. Therefore, our straightforward novel protocol might be implemented in an industrial setting to evaluate the fungicidal properties of a newly developed MWF or as an at-line tool. This will not only allow more efficient development of such novel MWF formulations, but can also contribute to the reduction of fungal biofilm formation and thus better functioning of automated machines and additionally reduce the health risk for operators being exposed to MWF.

MATERIALS AND METHODS

Fungal strain, production of conidia, and experimental design.

The strain DSMZ 16235 used in the present study was obtained from DSMZ (Deutsche Sammlung von Mikroorganismen und Zellkulturen, Braunschweig, Germany) and maintained as a conidial suspension in glycerol at −80°C. Before each experiment, the F. solani isolate was regrown from the stock on solid PDA medium under near-UV light (black light) for 7 days at 25°C to obtain conidia.

After 7 days, conidia were scraped from the PDA plates with 0.01% Tween 80 with a Drigalski spatula and counted under a Bürker chamber. The conidial suspension was centrifuged, the Tween 80 solution was removed and replaced by PBS, and the suspension was counted with a Bürker chamber. A dilution of MWF in PBS or a reference fungicide (InputPro with active ingredient prothioconazole, dose rate 400 mg/liter) was prepared to a concentration of 5 × 10⁶ conidia/ml. For the MWF, serial dilutions were made in the amounts of 20%, 10%, 5%, and 2.5%. Conidia suspended in PBS were used as negative controls. Incubation with prothioconazole was used as a positive control for dead conidia. All treatments were set up in triplicate, and the experiment was repeated three times. Conidial germination and membrane integrity-based viability were assessed at 0 h, 5 h, and 24 h via microscopy and FCM. In addition, at 7 days the fungal biomass was measured using a DNA-based qPCR methodology. All experiments were done in three independent experiments. Each experiment consisted of three biological repeats per treatment.

Selection of commercial metalworking fluids.

In order to research the use of FCM to quantify the viability of F. solani conidia in MWF, two commercially available MWF were selected. The first MWF (MWFA) consisted of 1,2,3-propanetriol (glycerol) and 2′2′-oxydiethanol, and no specific antifungal compounds were present. The second MWF (MWFB) contained mineral oil, tall oil, fatty acids (reaction products with ethanol amine), fatty acids (reaction products with triethanolamine), 3,3′-methylenebis(5-methyloxazolidine), and 3-iodo-2-propynyl butylcarbamate. The latter compound has previously been reported to be an efficient fungicidal agent in MWF (19).

Flow cytometric assessment of conidial viability based on membrane integrity.

Membrane integrity was assessed using the LIVE/DEAD BacLight kit (Molecular Probes Eugene, OR, USA) as described by the manufacturer. This bacterial viability kit is widely used in FCM and consists of two nucleic acid stains. Green fluorescent SYTO 9 is cell permeant and can freely enter all cells, either viable or dead. In contrast, red fluorescent propidium iodide (PI) can enter only membrane-comprised cells. The kit was designed by Molecular Probes for different bacterial species. In our study it was optimized for use in a strain of the fungus F. solani. To validate this dual-staining technique for its application on fungal conidia, different protocols were tested. We searched for the optimal concentrations of SYTO 9 and PI to obtain a clear division between the live and dead fungal spore populations. More specifically, the protocol was validated by measurement of different populations of fungal spores obtained by mixing a known percentage of live and dead spores. In our setup, after incubation (0 h, 5 h, and 24 h) conidia were washed twice in sterile saline (4,000 × g, 10 min, 4°C) to remove MWF and finally resuspended in 997 μl of sterile saline. These samples were stained with 3 μl of a mixture of SYTO 9 (5 μM final concentration) and PI (30 μM final concentration) and incubated for 10 min in the dark at room temperature.

Flow cytometric measurements were performed immediately after staining on a FACSCanto flow cytometer (Becton, Dickinson and Company, Erembodegem, Belgium), with 488-nm excitation from a blue solid-state laser at 50 mW. Optical filters were set up such that green fluorescence was measured at 530 ± 15 nm (SYTO 9, FL1) and red fluorescence was measured above 670 nm (PI, FL3). All acquired data were processed using FacsDiva software (Becton, Dickinson and Company, Franklin Lakes, NJ, USA).

Microscopic assessment of conidial germination and binary scoring of fungal growth.

At time points 0 h, 5 h, 24 h, and 48 h, the conidial germination was assessed under a microscope (inverted Olympus IX81 microscope). From each vial, 5 μl of conidial suspension was taken and put on a coverslip. Five μl of aniline blue (1%) in lactic acid was added, and the mixture was covered with a microscope slide. For each treatment and each replicate, 30 conidia were counted and the percentage of germinated conidia was calculated.

In addition, at time point 24 h, 10 μl of conidial suspension was removed from each well and incubated on PDA medium at 25°C for 5 days to assess fungal regrowth. These experiments were scored using a binary score, i.e., presence or absence of growth. This score reflects the live/dead stain in the FCM analyses. Each plating consisted of four biological repeats.

qPCR assessment of fungal biomasses.

At 7 days, the fungal biomass was assessed using a qPCR technique. The mycelium was transferred from the incubation tubes into 1.5-ml microcentrifuge tubes and lyophilized for 24 h. The mycelium was homogenized by a pestle under liquid nitrogen. Genomic DNA extraction was carried out using the Invisorb Plant minikit (Stratec) as described by the manufacturer, and DNA was used in the qPCR analysis. In parallel, genomic DNA of a pure F. solani strain, DSMZ 16235, grown on PDB was extracted, and DNA concentration was measured using a Quantus fluorometer (Promega). Subsequently, the concentration was adjusted to 2 ng DNA/μl. From this concentration, a 10-fold dilution series was made which served as a standard curve for fungal quantification.

A qPCR approach in which the ITS1 region of the ribosomal DNA (rDNA) of F. solani was amplified was used with primers Fsol1 (5′ to 3′) CTCATCAACCCTGTGAACATACC and Fsol2 (5′ to 3′) ATGCCAGAGCCAAGAGATCC (38). PCRs were performed in a final volume of 25 μl containing 12.5 μl of 2× PCR SYBR Green buffer (Promega), 1 μl of each primer (5 μM), and 10 μl of fungal DNA. The PCR conditions consisted of an initial denaturation at 95°C for 2 min, followed by 40 cycles of denaturation at 95°C for 30 s, and annealing at 60°C for 30 s. After a final extension step of 72°C for 5 min, a protocol was included for the dissociation curve, and samples were cooled at 4°C. The amount of fungal DNA was calculated from cycle threshold (C_T) values using the standard curve. A standard curve based on pure F. solani DNA was included, and the total amount of fungal biomass was assessed based on the C_T values in the amplification plots.

Data analysis.

For statistical evaluation, the SPSS vs.22 software package was used. Since normality assumptions of parametric tests were not met, differences between groups of data were tested for significance using a nonparametric Kruskal-Wallis test followed by Dunn's test with a sequential Bonferroni correction for multiple comparisons, at α = 0.05/n with n the number of pairwise comparisons.

In experiments where two treatments were compared, the Mann-Whitney U test was applied, as the conditions for parametric analyses were not met.

For correlation studies to compare microscopic results with flow cytometric results, we applied the pairwise Pearson correlation coefficient.