1. Introduction
Fungicides are critical for the control of many economically important plant diseases and have been an integral component of crop production for decades [
1,
2,
3]. All fungicides impose a selection pressure on their target pathogen populations, and some fungicides are prone to the development of resistance. This is especially true for fungicides with a single target site, as mutations in the target gene can lead to the selection of a resistant sub-population in the pathogen population with fungicide applications [
4,
5,
6,
7]. Some widely used examples of such fungicides include succinate dehydrogenase inhibitors (SDHIs), quinone outside inhibitors (QoIs), de-methylation inhibitors (DMIs), and phenylamides (PAs). Fungicides with multi-site modes of action are less likely to select for resistance [
8].
Managing fungicide resistance has been a priority since it began to appear in the 1960s and 1970s [
8]. Concerns over the loss of fungicide efficacy have resulted in the development and implementation of various resistance management strategies. Such strategies include applying mixtures of fungicides with different modes of action (often a site-specific with a multi-site), rotating fungicides with different modes of action, restricting the number of applications per season, maintaining the manufacturer’s recommended dose, and implementing cultural control practices as part of an integrated pest management (IPM) program [
8]. An important consideration regarding fungicide mixtures is the potential interaction between different fungicides, which may lead to antagonistic or synergistic activity against target pathogens [
9,
10,
11]. Identifying such interactions is crucial for the effective and efficient deployment of new fungicide chemistries prior to their incorporation into an IPM program.
An element of IPM that has gained attention over the past several years is the use of biologically-based pest control solutions, known as biocontrols or biopesticides. Mechanisms of biocontrols and biopesticides include antibiosis, resource competition, and disease resistance-induction [
12]. Regarding plant disease management, other than the use of resistant host varieties, biocontrols are considered as living organisms that suppress the activity or reduce populations of plant pathogens. Products that achieve the same outcome, but do not contain living organisms, and are fermented or extracted from natural sources, are considered biopesticides [
12]. Enzymes are integral to the antagonistic activity of certain biocontrol microorganisms. Lytic enzymes, such as chitinases, glucanases, and proteases, are secreted by a number of microbes and are known to be suppressive to plant pathogenic fungi and bacteria [
13,
14,
15,
16,
17]. However, deploying these enzymes directly for plant disease management has received little attention, probably in part due to the costs that are associated with obtaining commercially relevant quantities of enzyme, and subsequently stabilizing the enzymes for industrial use.
Advances in the field of biocatalysis, particularly in the areas of protein engineering enzyme immobilization and stabilization, are making the use of enzymes in agriculture increasingly feasible [
18,
19,
20]. As pesticide resistance and government regulation continue to decrease the number of effective compounds that are available to farmers [
8,
21], enzymes represent a promising new frontier in crop protection. Enzyme immobilization can be achieved through several approaches, including adsorption of the proteins onto a carrier, protein encapsulation inside a carrier, or cross-linking. Enzyme immobilization increases enzyme stability and has been shown to increase enzyme activity in certain systems [
18]. For example, immobilization of horseradish peroxidase by adsorption onto magnetic nanoparticles and the subsequent entrapment by self-assembly increased enzyme activity and reduced inhibition by the H
2O
2 substrate and reaction products when compared to the free enzyme [
22,
23]. Maximizing enzyme activity and stability will likely be paramount for the economic viability and, ultimately, the adoption of stabilized enzymes as alternatives to conventional synthetic pesticides by the agriculture industry.
Lactoperoxidase (LPO) is an enzyme with antimicrobial properties and occurs naturally in tears, saliva, and milk [
24,
25]. Lactoperoxidase is extracted from bovine milk and can be obtained relatively simply [
26], making it an attractive option for experimentation as well as commercialization. LPO catalyzes the oxidation of thiocyanate and iodide ions to antimicrobial hypothiocyanite (OSCN
−) and hypoiodite (OI
−), respectively, in the presence of hydrogen peroxide (H
2O
2). The H
2O
2 can be provided by the activity of glucose oxidase (GOx), which is extracted from the fungus
Aspergillus niger, on β-
d-glucose in the presence of oxygen. These unstable compounds, in turn, oxidize sulfhydryl groups in the cell membranes of microbes, leading to the inhibition of glucose transport, glycolysis, respiration, and ultimately cell death [
24]. The LPO system, which here refers to LPO, GOx, and substrates, is known to inhibit Gram positive and Gram negative bacteria, as well as fungi [
27]. Furthermore, once OSCN
− and OI
− are depleted, H
2O
2 accumulates through the activity of GOx which leads to additional oxidative stress on microbial cells [
28].
The objectives of this study were to (i) test the ability of a stabilized LPO and GOx system to inhibit the growth of P. ultimum, F. graminearum, and R. solani in vitro; and, (ii) determine if the combined effect of stabilized LPO and GOx with five commercial fungicides resulted in synergistic, additive, or antagonistic activity when compared to each active ingredient alone.
3. Discussion
The stabilized LPO system inhibited the in vitro growth of plant pathogenic
P. ultimum,
F. graminearum, and
R. solani.
Pythium ultimum was more sensitive to the LPO system than
F. graminearum and
R. solani, requiring approximately 30 times less enzyme to achieve comparable levels of inhibition. The difference in sensitivity between fungi and oomycetes was not surprising, given that different organisms are known to vary in their sensitivity to the LPO system [
27]. To date, most of the studies involving the LPO system have focused on human bacterial pathogens, such as
Staphylococcus spp.,
Pseudomonas aeruginosa,
Escherichia coli,
Streptococcus spp., and yeast like
Candida albicans [
27,
28,
29]. A few studies have demonstrated activity against spores of filamentous fungi, such as
Aspergillus fumigatus,
Aspergillus niger,
Mucor rouxii, and
Byssochlamys fulva [
30,
31]. To our knowledge, this is the first report of activity against the vegetative state of filamentous fungi and oomycetes, which may have important implications for crop disease management.
The LPO system is complex, and the concentration of each enzyme and substrate in the system ultimately affects its antimicrobial activity, making it difficult to directly compare results across studies. Bosch et al. demonstrated that LPO at a concentration of 387 nM in an LPO system similar to the one used here was sufficient to kill 1,000,000 CFU of
C. albicans,
E. coli,
Staphylococcus aureus, and
P. aeruginosa when exposed to the LPO system in solution for at least two hours [
27]. We found that a slightly lower concentration of LPO (323 nM) was sufficient to inhibit, but not kill, the fungi
R. solani and
F. graminearum, and a much lower concentration (20 nM) was sufficient to completely kill the oomycete
P. ultimum. Although the LPO concentrations that were used in the Bosch et al. study were relatively similar to the highest LPO concentration that was included in this study, the assay conditions were very different [
27]. The majority of studies on the antimicrobial activity of the LPO system conducted to date, including the Bosch et al. study, have involved bacteria and yeasts that were exposed to a non-stabilized LPO system in a liquid suspension [
27]. The effect of fungicides on filamentous fungal growth is typically assayed by exposing the fungus to the active ingredient in a petri dish containing solid growth media and measuring the relative reduction in mycelial growth. The disparity in assay conditions, along with differences in LPO system parameters and the use of free versus stabilized enzymes further contributes to the challenge of comparing the antimicrobial activity of the LPO system across studies. Utilizing a standardized set of assay conditions for further study of the LPO system on filamentous fungi would be helpful for building on existing work and establishing a body of knowledge on this system.
When the stabilized LPO formulation was combined with five commercial fungicides, the effects ranged from synergistic to antagonistic and were dependent on the pathogen X fungicide X concentration combination. This is in agreement with previous studies, which found that synergistic enhancement of fungicidal activity is dependent on the sensitivities of the target species, or strains within a species, to each individual component of the mixture [
9]. For example, Samoucha and Cohen (1986) found that a mixture of the systemic fungicide metalaxyl and any of five different contact fungicides resulted in the synergistically enhanced control of potato late blight caused by either a metalaxyl-sensitive or metalaxyl-resistant isolate of
Phytophthora infestans [
32]. However, the synergistic activity was much greater against the metalaxyl-resistant isolate when compared to that of the metalaxyl-sensitive isolate. Similar results were observed with mixtures of metalaxyl and mancozeb against metalaxyl-sensitive and -resistant isolates of the cucumber downy mildew pathogen
Pseudoperonospora cubensis [
33].
The goal of the fungicide portion of this study was to characterize the interactions of select fungicides in combination with the stabilized LPO system with regard to two species of fungi and one oomycete. Fungicides were chosen based on whether they were labeled for use against each individual species, which is why each fungicide was not tested against each species. However, mancozeb was tested on both P. ultimum and F. graminearum, and Tilt® was tested on both F. graminearum and R. solani, allowing for inter-species comparisons of LPO-fungicide interactions. Although a pattern of synergy was observed with both of the species that were tested with mancozeb, the relatively large experimental errors between F. graminearum replicates made it difficult to detect statistically significant synergy in that experiment. Significant synergy was observed with P. ultimum at the lowest mancozeb concentration tested. In the Tilt® experiments, a pattern of antagonism was observed with both of the species tested, but once again the large experimental errors between replicates allowed for the detection of only one case of statistically significant antagonism with F. graminearum at the second to lowest Tilt® dose that was tested. The variance was generally greater in experiments involving F. graminearum and R. solani when compared to P. ultimum.
Fungicides are critical for crop disease management. Some fungicides, however, have been shown to have phytotoxic effects at high concentrations that are required to effectively control disease, and the number of fungicides that are available to farmers is declining due to increasing government regulation and loss of efficacy due to resistance [
8,
21,
34,
35,
36]. The introduction of novel fungicide chemistries that meet government regulatory standards, and the protection of existing chemistries against the threat of obsolescence due to resistance will likely be crucial for the assurance of food security in the future. Our results show that the stabilized LPO system may be a viable tool for controlling crop diseases that are caused by fungi and oomycetes, and that it does not affect the germination or vigor of tomato seedlings under nutrient rich conditions. Future studies of the LPO system as a potential seed treatment should evaluate its efficacy in greenhouse and field trials, under low nutrient conditions, and other environmental stresses. Furthermore, the stabilized LPO system was shown to be compatible with several commercial fungicides, and its broad-spectrum activity makes it a good candidate for co-formulations with the goal of controlling multiple diseases in a single application or reducing the probability of selecting for resistance in treated pathogen populations. Further research is required to examine the mechanism of diffusion into the complex soil environment by the stabilized LPO system. The stabilized LPO system may also meet organic certification standards because it utilizes naturally derived enzymes, making it a potentially valuable tool for organic and conventional crop protection.
4. Materials and Methods
4.1. Isolates
Isolates of Pythium ultimum (isolate Geneva16), Fusarium graminearum (isolate GZ014NNY98), and Rhizoctonia solani (isolate AC1-A1) were all originally collected in New York State and were obtained from collaborators in the Section of Plant Pathology and Plant-Microbe Biology at Cornell University. F. graminearum and R. solani were maintained on potato dextrose agar (PDA) (Sigma-Aldrich, St. Louis, MO, USA) and P. ultimum was maintained on cornmeal agar (CMA) (Sigma-Aldrich).
4.2. Fungicides, Enzymes and Reagents
Commercial fungicide formulations that were used in the study were Daconil® (active ingredient (AI): chlorothalonil 29.6%, GardenTech, Lexington, KY, USA), Tilt® (AI: propiconazole 41.8%, Syngenta, Basel, Switzerland), Stratego YLD® (AI: prothioconazole 10.8% + trifloxystrobin 32.3%, Bayer, Leverkusen, Germany), mancozeb flowable with zinc (AI: mancozeb 37%, Bonide, Oriskany, NY, USA), and tea tree oil (AI: tea tree oil 100%, Mason Natural, Miami Lakes, FL, USA). Potassium iodide (KI), ammonium thiocyanate (NH4SCN), carboxymethyl cellulose (CMC), glucose, lactoperoxidase (LPO, ≥150 U/mg), and glucose oxidase (GOx, 177 U/mg) were obtained from Sigma-Aldrich.
4.3. Optimizing Enzyme Formulation for In-Vitro Assays, and Enzyme Stabilization
The concentration of LPO and GOx in the dry enzyme film disks was optimized prior to beginning fungicide interaction experiments to identify enzyme concentrations that resulted in a measurable reduction in growth for each species without being completely inhibitive. LPO and GOx were always included in a 1:1.3 M LPO:GOx ratio, and the final enzyme concentration that is reported is the amount of each enzyme (LPO and GOx) in each 50 µL enzyme disk.
P. ultimum was tested at four LPO and GOx concentrations, and
F. graminearum and
R. solani were tested at three LPO and GOx concentrations (
Table 1). Optimization tests were done in duplicates. Dry enzyme film disks were made by combining 3 µL KI (1M), 5 µL NH
4SCN (1M), 175 µL 4% carboxymethyl cellulose (CMC), varying volumes of stabilized LPO + GOx, and 3 µL blue food dye brought up to a final volume of 1 mL with distilled deionized water (DDI H
2O). Enzyme stabilization was performed, as previously described [
22,
37]. Briefly, LPO (125 µg/mL, pH 7.4) and GOx (330 µg/mL, pH 7.4) were mixed to achieve a 1:1.3 M LPO:GOx ratio and stored on ice. Magnetite nanoparticles (NP) (1.277 mg/mL, pH 3, approximately 5 mL stock) were ultrasonicated at 40% amplitude for 1 min, cooled to ambient temperature (approximately 21 °C) in a water bath, and pipette mixed with the LPO:GOx enzyme suspension in a 1:1 enzyme:NP ratio. Dry substrate film disks were made by combining 30 µL KI (1M), 50 µL NH
4SCN (1M), 350 µL 4% CMC, 500 µL glucose (1M), and 3 µL red food dye brought up to a final volume of 1 mL with DDI H
2O. Dyes were included to differentiate enzyme disks from the substrate disks and were not biologically active or structural components of the disks. Each solution was pipette mixed several times and vortexed briefly. Solutions were dispensed in 50 µL aliquots onto parafilm and were dried at an ambient temperature in a vacuum oven containing desiccant at −50 kPa. The diameter of the dried enzyme and substrate film disks is approximately 5 mm. After approximately 2 h, dry enzyme and substrate disks were stored in the dark at 4 °C until use.
To test the sensitivity of each species to the enzyme formulation, one substrate disk was removed from the parafilm sheet and was placed on the center of a petri dish containing CMA (
P. ultimum) or PDA (
F. graminearum and
R. solani). Next, one enzyme disk was removed from the parafilm sheet and was placed on top of the substrate disk. Finally, one culture plug was placed on the disks mycelia-side down.
P. ultimum was tested using 7 mm-diameter culture plugs, and
F. graminearum and
R. solani were tested using 4 mm-diameter culture plugs. Once non-treated control colonies had grown to nearly the edge of the plate (two days for
P. ultimum, four days for
R. solani, five days for
F. graminearum) two perpendicular colony diameter measurements were taken for
P. ultimum and
R. solani. Colonies of
F. graminearum were non-circular, therefore photos of each plate were taken and colony areas were measured using the public domain image processing program ImageJ (available at
https://imagej.nih.gov/ij/).
4.4. Evaluating Phytotoxicity of Stabilized Enzyme Formulation
Tomato seeds (Solanum lycopersicum, ‘Celebrity F1’ Lot #53815, Johnny’s Selected Seeds, Winslow, ME, USA) were coated with the stabilized enzyme formulation as follows: For each seed, a dry substrate disc was placed on parafilm. A single tomato seed was placed on top of the dry substrate disc, and then covered with a dry enzyme disc. A 10 µL drop of sterile DDI H2O was added to the stack to adhere the dry enzyme and substrate discs to the seed. The coated seeds were allowed to dry for 1 hour. Following the drying time, the coated seeds were rolled between gloved fingers to ensure good contact between the seed and the substrate and enzyme discs. Five seeds were coated with four concentrations of the stabilized enzyme formulation, plus five seeds coated with just inert components of the coating and five seeds uncoated as controls. Coated seeds were planted in pre-moistened Miracle-Gro potting mix (Scotts Miracle-Gro Company, Marysville, OH, USA) and were placed under t5 high-output grow lights (Hydroplanet, Wheat Ridge, CO, USA). Pots were watered with DDI H2O until seedlings emerged, then watered with label rate Miracle-Gro fertilizer (Scotts Miracle-Gro Company, Marysville, OH, USA). After four weeks, plant height was measured and seedlings were harvested. Soil was washed from the roots with water, and then root and shoots were separated and weighed.
4.5. Sensitivity of P. ultimum, R. solani, and F. graminearum to Fungicides Alone and in Combination with the Stabilized Enzyme Formulation In-Vitro
Commercial fungicides formulated as aqueous suspensions were thoroughly mixed to ensure homogeneity and diluted in sterile DDI H
2O. Fungicide dilutions are shown in
Table 8. Whatman
® qualitative grade 1 filter paper (FP) (Sigma-Aldrich) was cut into 7 mm disks using a 3-hole punch, and disks were autoclaved prior to use. FP disks were impregnated with 5 µL of each of four fungicide dilutions per fungicide tested resulting in a known mass of active ingredient (AI) per FP disk except for tea tree oil, which does not include mass of AI per unit volume on the label. Control FP disks were impregnated with 5 µL sterile DDI H
2O. Fungicide X enzyme disk interactions were tested by placing one substrate disk on the center of each petri dish containing PDA (
F. graminearum and
R. solani) or CMA (
P. ultimum), followed by one enzyme disk, as previously described. Next, one fungicide-impregnated FP disk was applied on top of the enzyme and substrate disks. Finally, one culture plug was place on top of the FP disk mycelia-side down. Final enzyme concentrations in enzyme disks were 4 nM for
P. ultimum and 119 nM for
F. graminearum and
R. solani based on results from preliminary enzyme formula optimization experiments. Plugs of
P. ultimum measured 7 mm in diameter and plugs of
F. graminearum and
R. solani measured 4 mm in diameter. Preliminary testing revealed that
P. ultimum is more sensitive to the enzyme treatment than
F. graminearum and
R. solani, so enzyme concentrations and plug sizes were chosen to achieve a measurable growth reduction, without being completely inhibitive, in enzyme-only treatments as compared to non-treated controls. Each experiment included the same fungicide dilution series plated without substrate and enzyme disks, as well as a substrate + enzyme disk-only treatment and a non-treated control. All of the plates, including controls, contained a FP disk, and each treatment was replicated once. Plates were left on the bench at ambient temperature for 2, 4, or 5 days for
P. ultimum,
R. solani, and
F. graminearum, respectively, by which time control colonies had nearly grown to the plate edge. Two perpendicular colony diameter measurements were recorded for
P. ultimum and
R. solani. Colonies of
F. graminearum measured using the public domain image processing program ImageJ.
P. ultimum and F. graminearum were insensitive to mancozeb in impregnated-FP tests, so an additional experiment was done using mancozeb-amended media. CMA and PDA (for P. ultimum and F. graminearum, respectively) were amended with mancozeb to achieve final concentrations of the fungicide formulation of 10 mg/L, 5 mg/L, 2 mg/L, 0.5 mg/L, and 0 mg/L (controls). The same experimental design used in the impregnated-FP tests was used for amended-media tests, except FP was not included in any of the treatments or controls.
4.6. Statistical Analysis and Determination of Synergism, Additivity, and Antagonism
Reduction in growth was calculated using the following formula (1):
Experiments involving
F. graminearum used colony area rather than colony diameter in the formula due to non-circular colony morphologies. Statistical analyses were done using R (version 3.3.0, R Foundation for Statistical Computing, Vienna, Austria). Each fungicide-by-pathogen combination was analyzed separately. Analysis of variance was done using the agricolae package [
38] and treatment means were separated using the Tukey-Kramer honest significant difference (HSD) test (
p < 0.05).
The effects of fungicide combined with stabilized enzyme formulation were characterized as synergistic, additive, or antagonistic based on the methods described by Couch et al. [
9]. Briefly, an additive effect is observed when the level of disease control imparted by the co-operative action of two or more fungicides is equal to the control, as predicted by the response of each fungicide applied by itself. Synergism is observed when the disease control imparted by the joint application of two or more fungicides is greater than the control that would be predicted by an appropriate reference, in this case, the additive effect. Similarly, antagonism is observed when the disease control imparted by the joint application of two or more fungicides is less than the control predicted by an appropriate reference, again, in this case, the additive effect. Theoretical additive values (i.e., expected combined effects) were calculated by adding the observed effect of each fungicide alone with the observed effect of the enzyme formulation alone. Tukey’s HSD value at
p < 0.05 was then applied to the theoretical additive value to determine whether actual combined effects were synergistic, additive, or antagonistic. For example, if the theoretical additive value was 50% and the Tukey’s HSD value was 10%, then the observed combined effect would be considered synergistic at >60%, additive between 40% and 60%, and antagonistic at <40%. If the theoretical additive value was >100%, then it was considered 100% (complete growth inhibition) for the purposes of this calculation.