Biocontrol of Fusarium Species Utilizing Indigenous Rooibos and Honeybush Extracts

ABSTRACT Mycotoxins produced by several Fusarium species have a significant effect on reducing maize yield and grain quality and have led to food safety concerns. The antifungal activities of rooibos (Aspalathus linearis) and honeybush (Cyclopia species) tea extracts reduced the growth of plant pathogen Botrytis cinerea, but their efficacy against Fusarium spp. is unknown. In this study, we examined the effects of fermented and unfermented rooibos (A. linearis) and honeybush (Cyclopia subternata) aqueous extracts as well as green tea (Camellia sinensis) against 10 Fusarium species. Conidial viability was assessed by fluorescence microscopy dyes, ATP production was determined using the BacTiter-Glo assay, the mode of action was analyzed by scanning electron microscopy (SEM), and quantification of polyphenols was done using high-performance liquid chromatography with diode array detection (HPLC-DAD). Fermented rooibos extract demonstrated the highest antifungal activity (P < 0.0001) against Fusarium verticillioides MRC 826-E, Fusarium subglutinans MRC 8553, Fusarium proliferatum MRC 8549, and Fusarium globosum MRC 6647, with only 9.53%, 9.26%, 11.0%, and 12.7% ATP production, respectively, followed by antifungal activity of the fermented C. subternata extract against F. subglutinans MRC 8553, F. subglutinans MRC 8554, F. proliferatum MRC 8550, and F. verticillioides MRC 826-E with 3.79%, 6.04%, 6.04%, and 8.40% ATP production, respectively. Extract-treated conidia examined by SEM exhibited disruption of conidial hyphae and collapsed spores. Overall, the fermented rooibos and C. subternata extracts showed higher antifungal activity against the Fusarium species than the unfermented extracts. IMPORTANCE In maize subsistence farming areas in South Africa, daily consumption of maize contaminated by high level of mycotoxins contributes to long-term health effects such as immune deficiency and cancer. Biocontrol methods that are safe and cost-effective are critical to addressing this public health problem. Plant extracts known as biocides or green pesticides are alternatives to chemical pesticides due to their safety and eco-friendly properties. In South Africa, rooibos (Aspalathus linearis) and honeybush (Cyclopia species) contain polyphenols with significant antioxidant and antimicrobial properties. These indigenous herbal teas are widely available and consumed in South Africa and have potential as an innovative approach to reduce mycotoxin levels and, subsequently, human and animal exposure to these toxins. This study evaluates the efficacy of the antifungal activities of several aqueous extracts prepared from fermented and unfermented rooibos (A. linearis), honeybush (Cyclopia subternata), and green tea (Camellia sinensis) on 10 Fusarium strains.

proliferatum, and Fusarium globosum (3,4). The prevalence of F. graminearum species (FGS) is associated with mycotoxin type B trichothecenes (TCT-B), zearalenone (ZEA), nivalenol (NIV), and acetylated derivatives (15-ADON and 3-ADON) in wheat, rice, barley, rye, and oats (5). F. verticillioides, F. proliferatum, F. subglutinans, and F. globosum produce fumonisins (FB 1 , FB 2 , and FB 3 ), especially in maize and sorghum. Higher concentrations of these mycotoxins in animal feed exert toxic effects in farm animals, causing distress and reduced productivity (6). Acute intoxication with fumonisins in domestic animals has resulted in pulmonary edema in pigs and dogs, leukoencephalomalacia in horses, and brain hemorrhage and kidney and liver cancer in rabbits (7). To prevent the negative effects on animals and consumers, many countries regulate mycotoxin concentrations in feed. Common symptoms associated with consuming mycotoxin-contaminated products include frequent vomiting, diarrhea, drowsiness, anemia, eye infection, skin infection, nail infection, and esophageal cancer, as well as stunting in children. High levels of fumonisin intake in pregnant women could also result in neural tube defects in the fetus (8). In the European Union (EU), for example, maximum levels are enforced for FB 1 (9), and guidance values have been stipulated for fumonisins, ZEA, and deoxynivalenol (DON) (10). Therefore, mycotoxin concentrations in feed and food should be continuously monitored to support risk assessment.
Current fungal pesticides such as ammonia detoxification, sodium hydroxide, and sulfur dioxide have led to serious environmental (soil contamination), food security (decreased nutritional value of foods and sensory quality) and human health (toxic derivatives) concerns (11). Due to the harmful effects of chemicals, a safe, cost-efficient alternative biocotrol approach is critical to addressing the public health problem. Plant extracts known as biocides or green pesticides are deemed sustainable alternatives due to their safety and eco-friendly properties. Plant-derived compounds for disease management can provide food safety and security to reduce Fusarium infections and their associated mycotoxin production in different grains produced in South Africa and beyond (12).
In South Africa, the herbal teas rooibos (Aspalathus linearis) and honeybush (Cyclopia species) contain polyphenols with significant antioxidant and antimicrobial properties. It has been shown that aspalathin, dihydrochalcones, flavonoids, proanthocyanidin-type polyphenols, xanthone, mangiferin, and hesperidin inhibited the growth of Escherichia coli, Streptococcus pyogenes, Staphylococcus aureus, and Candida albicans (13). The antifungal susceptibility of Botrytis cinerea spore germination by rooibos and honeybush could be linked directly to polyphenols that synergistically inhibited conidial growth (14). These indigenous herbal teas are widely available and consumed daily by the people of South Africa. They have the potential as an innovative approach to reduce mycotoxin levels and human and animal exposure. Therefore, development of a cost-effective food-grade biocontrol method from indigenous South African rooibos and honeybush tea extracts for application in subsistence farming and commercial systems against Fusarium infection and mycotoxin production will contribute to sustainable mycotoxin risk management practices in maize production.
The efficacy of extracts from A. linearis and Cyclopia spp. as antifungal agents against Fusarium species has not been previously reported. Hence, this study evaluates the effect of aqueous extracts prepared from fermented and unfermented rooibos, honeybush, and green tea on the growth inhibition of 10 Fusarium strains and thereby indirectly contributes to the reduction in mycotoxin production.
Caffeine content was very high in the green tea extract, followed by epigallocatechins. Determination of Fusarium conidial viability. Conidial viability was determined in order to confirm the functionality of the spores to be utilized for the antifungal activities of the tea extracts on growth inhibition. Propidium iodide (PI) penetrates the nonviable conidia and stains red fluorescence, while Hoechst stain passively enters through the cell membrane, where intracellular esterases subsequently hydrolyze it to blue fluorescence. This double staining enabled differentiation between nonviable and viable fungal conidia. Viable conidia retained the Hoechst 33342 blue fluorescence (data not shown), and very few nonviable conidia were visible with the red fluorescent staining patterns of the PI dye.
Determination of the antifungal activity of herbal tea extracts. The efficacy of the tea extracts on Fusarium conidia growth inhibition was determined by measuring the percentage of ATP using different tea extract concentrations. The fermented rooibos tea extract at 20 mg/mL reduced conidial viability of F. verticillioides MRC 826-E, F. subglutinans MRC 8553, and F. proliferatum MRC 8549 to 9.53% (50% inhibitory concentration [IC 50 ], 1.19 6 0.05), 9.26% (IC 50 , 1.79 6 0.05), and 11.0% (IC 50 , 0.77 6 0.04) of their respective controls after 96 h of incubation (Table 2 Table 3). The  extracts prepared from the fermented honeybush tea demonstrated higher antifungal potency than the unfermented honeybush extract at the highest concentration (20 mg/mL) ( Table 4 and Table 5). Among the 10 Fusarium strains, five strains, including F. subglutinans MRC 8553, F. subglutinans MRC 8554, F. proliferatum MRC 8550, F. verticillioides MRC 826-E, and F. proliferatum MRC 8549, were highly susceptible to the fermented honeybush (Cyclopia subternata) tea, at conidial viability rates of only 3.79% (IC 50 , 6.63 6 0.06), 6.04% (IC 50 , 9.47 6 0.13), 6.04% (IC 50 , 1.75 6 0.21), 8.40% (IC 50 , 4.28 6 0.31), and 9.73% (IC 50 , 1.20 6 0.05) ATP production, respectively, to the untreated conidia ( Table 3). The   Table 6). The comparison plot of the levels of rooibos and C. subternata growth inhibition against the various Fusarium strains at various concentrations was highly significant (P , 0.0001) ( Fig. 2A), indicating a higher growth inhibition rate by rooibos extract. Furthermore, a comparison plot of the fermented and unfermented tea extracts against the Fusarium strains confirmed the fermented tea extracts to be more active (P , 0.0001) than the unfermented tea extracts (Fig. 2B). The single optimized dose (2.5 mg/mL) of amphotericin B    (AmpB) used as an inhibitor control significantly (P , 0.0001) inhibited all spore growth of all 10 Fusarium strains, in a range of 0.21% to 50.1% ( Table 2). Germination of Fusarium spores inhibited by tea extracts. Spore germination of the Fusarium conidia exposed to each respective tea at 20 mg/mL showed inhibition of conidial germ tube formation of most Fusarium strains ( Fig. 3 and Fig. 4). However, higher inhibition of conidia germination was observed with the fermented rooibos tea extract. With fermented C. subternata treatment, significant inhibition of conidial germination in F. subglutinans MRC 8553, F. proliferatum MRC 8549, F. globosum MRC 6647, and F. globosum MRC 6122 was observed. The unfermented rooibos extract showed less activity in reducing the germination rates of the Fusarium conidia. By inhibiting Fusarium conidial germination with tea extracts, the potential for infection progression can be reduced.
Effect of herbal tea extracts on Fusarium spore morphology. The effect of tea extracts on conidial spore morphology was determined using the two most resistant strains, including F. verticillioides MRC 826-E and F. graminearum MRC 4927. Treated conidial samples were examined using scanning electron microscopy (SEM) (Carl Zeiss Microscopy, Germany). Untreated (without tea) conidia on potato dextrose broth (PDB) medium incubated for 96 h at 26°C showed normal spore growth, with smooth linear hyphae that appeared regular in shape (Fig. 5). The conidial membrane of F. verticillioides MRC 826-E became highly affected after treatment with 20 mg/mL tea extract. High fungicidal disruption of the conidial membrane was observed in the samples treated with fermented rooibos and honeybush extracts. The fungicidal activity of fermented rooibos and honeybush extracts was also shown against F. graminearum MRC 4927. Hyphae of both Fusarium strains were collapsed, shrunk, and deformed, with a loss of linearity. Furthermore, craters were evident on both conidial cell membranes.

DISCUSSION
Fusarium species are one of the most destructive fungal pathogens and cause severe postharvest contamination, especially in maize, wheat, soybean and sorghum. The present study investigated the antifungal effect of indigenous South African rooibos (A. linearis) and honeybush (C. subternata) tea and observed that the tea extracts significantly inhibited the conidial growth of several Fusarium species. The fermented rooibos extract was 2-fold more active in growth inhibition than the unfermented rooibos extract. Comparing the activity of honeybush extracts against the Fusarium conidial growth inhibition, the fermented C. subternata extract proved more active than the unfermented C. subternata. The antifungal activity of unfermented C. sinensis was 2-fold less than those of rooibos and honeybush tea extracts.  Many studies have evaluated the antimicrobial effects of rooibos and honeybush teas (13,14,16,17). Exposure of B. cinerea spores to rooibos and honeybush (Cyclopia genistoides) tea extracts at 100 mg/mL inhibited growth by 33% and 16%, respectively (14). The rooibos and honeybush tea extracts, including C. subternata tested on E. coli used as Biocontrol Utilizing Rooibos and Honeybush Microbiology Spectrum a benchmark, were observed to inhibit cell growth by 60%, 80%, and 85%, respectively, after 6 h. Similar results were noted for both Staphylococcus aureus and Staphylococcus epidermis, with fermented rooibos being more active than unfermented rooibos by 85% versus 35% and 78% versus 23%, respectively (17). The minimum inhibitory activity of the fermented honeybush C. intermedia aqueous extract was shown to reduce the cell growth of C. albicans at 150 mg/mL (13). Thembo et al. (18) also reported the antifungal inhibitory effects of four South African weedy aqueous extracts against F. verticillioides, F. proliferatum, Aspergillus flavus, and Aspergillus parasiticus at a concentration of 2.5 mg/mL. However, the results in the study by Thembo et al. (18) showed little or no inhibitory effect due to the low concentration they used compared to the high concentration (20 mg/mL) used in this study. At the lower concentrations (5 mg/mL and lower), the rooibos and C. subternata extracts in this study increased the ATP percentage above 100% for most treatments after 96 h. This phenomenon could indicate fungal growth stimulation at lower tea concentrations. Other possible reasons could be that the antioxidant components present in the teas may no longer be effective enough to inhibit the conidial growth per se but rather stimulate the cell biomass (19). Simpson et al. (17) also recorded a similar effect for E. coli, where the maximum optical density (OD max ) values for all concentrations below 10 mg GAE (gallic acid equivalent) for both rooibos infusions and unfermented rooibos were contrary to expectation, as bacterial growth was enhanced rather than suppressed after 6 h. Their most reasonable explanation for the outcome was probably the inherent nutritional value, combined with reduced toxicity at lower concentrations. Coetzee et al. (14) observed B. cinerea cells in a low concentration (10 mg/mL) of rooibos and C. subternata extracts and noticed induced biomass production by more than 2-to 3-fold after 48 h. Another hypothesis could be that the compounds act synergistically in inhibiting growth, and at lower concentrations, one of the compounds may be diluted out, affecting the synergistic effect. It is also possible that glycosides are present at a higher concentration and when a diluted extract is used, FIG 3 Fermented and unfermented tea extracts inhibit conidial germination of several Fusarium strains. Fusarium strain conidia were grown in a liquid medium without (control) or with 20 mg/mL tea extract and were incubated at 26°C for 24 h before microscopic observation at a magnification of Â400. The conidial germination rate was calculated by randomly counting 50 conidia from each of three independent experiments and then plotted using GraphPad Prism 5.01 software (GraphPad Software, Inc., La Jolla, CA). *, P # 0.05, and **, P # 0.01, for comparison between control and tea extracts for each species; #, P # 0.05, and ##, P # 0.01, for comparison between fermented and unfermented tea extracts for each species (e.g., fermented C. subternata versus unfermented C. subternata).
the glycosides are still present, acting as an additional carbon source for the fungus. Third, a lower tea extract concentration could also cause growth stimulation due to some of the vital mineral nutrients of potassium (K), calcium (Ca), magnesium (Mg), and manganese (Mn) present in the PDB growth medium combined with the tea's nutrients (Ca, K, Mg, and Na) (14). Therefore, the concentration of tea extracts is of utmost importance when considering antimicrobial inhibition studies.
A 20-mg/mL treatment with fermented rooibos extract significantly reduced the germination of spores in F. verticillioides MRC 826-E, MRC 826-J, and MRC 8559, F. graminearum MRC 4927, and F. globosum MRC 6647 compared to the other teas. From the results obtained in the study, it would be difficult to point out exactly which group of compounds were responsible for the significant effect. In an antimicrobial study by Simpson et al. (17), the polyphenols from both naturally fermented rooibos and artificial infusions were individually tested at the highest concentrations (2.0 and 4.5 mg) on E. coli (strain CFT073), S. epidermis (strain Se19), and S. aureus (strain ATCC 25923). However, no changes in bacterial growth inhibition were observed compared to the control medium only (17). They concluded that the glycosylated polyphenols alone might not account for the inhibition observed when testing the natural extracts. Therefore, bacterial uptake of several polyphenols from crude extracts could have reduced the cell viability or can even represent the presence of unidentified polyphenols with antibacterial activity. We speculate that the polyphenols-most likely the glycosylated compounds present in rooibos extracts, including S-eriodictyol-6-glucoside (E6CG), S-eriodictyol-8-glucoside (E8CG), phenylpyruvic acid-2-O-glucoside (PPAG), aspalathin, isoorientin, orientin, hyperoside, rutin, isoquercitrin, luteoloside, nothofagin, vitexin, and isovitexin-as well as the nonglycosylated bioquercetin may have reduced the conidial cell viability. The majority of these polyphenols have been shown to be active against one or more inflammatory infections. For example, the flavanone glycosides and eriodictyols in fermented rooibos have been reported to have natural antioxidant and anti-inflammatory, antidiabetes, antiobesity, and skin protection activities (20).
The polyphenolic contents of the C. subternata extracts were distinct from those in C. sinensis, being caffeine free with low levels of tannin. The main polyphenols found include xanthones, flavones, isoflavones, flavanones, flavanols, and coumestans. Contrary to previous studies (21,22), the fermented C. subternata polyphenols showed improved antifungal activity compared to those in unfermented extracts, even though the total amount of polyphenols was larger (9.105 g/100 g extract) in unfermented FIG 5 Scanning electron microscopy (SEM) observations of fungal hyphal morphological changes when exposed to teas. F. verticillioides MRC 826-E and F. graminearum MRC-4927 cells were incubated in a liquid medium without (control) or with 20 mg/mL tea extract and then submitted to SEM. After 96 h, untreated hyphae and conidia with regular cell walls are visible, as indicated by the arrows in the control samples. After treatment, both hyphae and conidia suffered from malformed structure swelling; arrows in the treated samples indicate ruptured hyphae and deformed conidia with flattening of cell membrane. than in fermented (4.333 g/100 g extract) extract. Reasons for this effect are not widely reported; however, available evidence has reported a wide range of pharmacological activities of C-glycosyl mangiferin, hesperidin, xanthone, and scolymoside (23). Hesperidin was recently shown to bind to COVID-19-spike protein and interfere with the refolding of the spike, inhibiting viral infection (24). The antimicrobial activities of tea dihydrochalcones (phloretin and its glycosylated derivatives) tested against several clinical Gram-positive and Gram-negative bacterial strains were observed to inhibit the growth of S. aureus ATCC 6538, Salmonella enterica serovar Typhimurium ATCC 13311, Listeria monocytogenes ATCC 13932, and methicillin-resistant S. aureus clinical strains from 7.81 mg/mL to .1,000 mg/mL (25). The polyphenols analyzed in C. sinensis were mainly caffeine, followed by epigallocatechins. According to Coetzee et al. (14), the antimicrobial effect of C. sinensis was directly linked to the presence of epicatechin gallate and epigallocatechin gallate. These compounds were found to have an inhibitory effect on methicillin-resistant S. aureus, multidrug-resistant Pseudomonas aeruginosa, E. coli, Enterococcus faecalis, and C. albicans. This suggests that the antifungal property of the tested tea extracts may be due to these polyphenols.
The mode of action of many antimicrobial agents analyzed by SEM targets the biosynthesis of ergosterol by disrupting either cell envelope integrity or biofilm formation (26). In this study, F. verticillioides MRC 826-E and F. graminearum MRC 4927 damaged conidial cell envelope was visible before and after tea treatment at 20 mg/mL with both the fermented rooibos and C. subternata extracts. The disruption of the cell membranes may have resulted in the osmotic imbalance of the mycelial enzymatic reactions, thus preventing further morphogenesis and growth development of the fungi (27).
Conclusion. This is the first study to investigate the use and efficacy of extracts from A. linearis and C. subternata as innovative antimicrobial agents against Fusarium species. The fermented rooibos and honeybush tea extracts demonstrated higher antifungal susceptibility than the unfermented tea extracts. The effect could be the combination of several polyphenols present in the tea extracts, which may have acted on the cell envelope. The antifungal activity of the tea extracts was shown against several Fusarium strains that are known to produce mycotoxins. If the extracts could suppress conidial viability or prevent conidial germination, it could prevent the production of these mycotoxins. However, more research is required to understand the underlying mode of action and develop suitable, sustainable, effective, inexpensive botanical products that can be used to help overcome the scourge of plant fungal diseases, which could lead to increased food safety and security.

MATERIALS AND METHODS
Chemicals. The chemicals used in this study were of analytical grade, purchased from Sigma-Aldrich Preparation and characterization of herbal tea extracts. Fermented and unfermented rooibos (A. linearis) and honeybush (C. subternata) tea were obtained from the ARC-Infruitec (Agriculture Research Council), Stellenbosch, South Africa. The green tea (Camellia sinensis), imported from China, was a gift from Vital Health Foods, Kuilsriver, South Africa. Aqueous extracts were prepared by adding 1 L of freshly boiled distilled water to 100 g plant material in a 2-L conical flask and then incubating the mixtures for 30 min at 22 6 3°C, stirring every 5 min. The extracts were strained through a double layer of cheesecloth to remove the majority of the plant material. An additional filtration was performed using Whatman no. 1 filter paper (Merck, Darmstadt, Germany). The filtrates were lyophilized, and the final percentages of each tea were calculated as follows, where X is the mass of each tea material (grams), A is the mass of the empty round-bottom flask (grams), and B is the mass of the round-bottom flask plus tea extracted (grams): mass of tea (g) = (B 2 A) and % yield = [(B 2 A)/X] Â 100. All dried extracts were stored in amber bottles and preserved in a desiccator at 22 6 3°C prior to use. HPLC-DAD quantification of the individual polyphenols in the extracts was analyzed by the ARC-Infruitec South Africa (15).
Fungal strains and culture conditions. The Fusarium strains used in this study are shown in Table 7. The strains were obtained from the Applied Microbial and Health Biotechnology Institute (AMHBI) culture collection and stored at 4°C as lyophilized spore suspensions.
Rehydrated freeze-dried spore suspensions (1 mL distilled water [dH 2 O]) of the different MRC strains Na-cacodylate containing 1% (vol/vol) aqueous OsO 4 (1 h at 4°C), washed 3 times in dH 2 O, and then dehydrated in a series of graded (50% to 70% to 90% [vol/vol]) ethanol and subjected to drying with a head-mounted display (HMDS) infiltration unit. Samples were mounted onto a 12-mm aluminum SEM stub using double-sided carbon conductive tape and coated with 50 to 100-Å Au/Pd using a sputter coater (Leica EM ACE200) for 3 min and visualized and identified at different magnifications under a scanning electron microscope (Carl Zeiss Microscopy, Germany). Images were captured using SmartSEM (Zeiss Electron Imaging) and Oxford AZtec (Oxford Instruments, Abingdon, Oxfordshire, United Kingdom).
Statistical analysis of data. The data were analyzed using NCSS 2021 v.21.0.2. Data were subjected to natural log (ln) transformation of all variables and analyzed within a generalized linear model analysis of variance (ANOVA). Multiple comparisons were analyzed using the Tukey-Kramer multiple-comparison procedure. This method provides joint simultaneous confidence intervals for all pairwise differences between the means as well as provides the multiple comparison P value. Overall, a P value of ,0.05 was used to represent statistical significance. In addition, the size of the F ratios was used to measure relative sizes of differences.