Fungicides Cuprozin Progress and SWITCH Modulate Primary and Specialized Metabolites of Strawberry Fruits

Abstract

Numerous pesticides, including fungicides, are applied every year to crop plants. However, such application may affect the plant metabolism and thus crop quality. Strawberry is an economically important crop, but the fruits are highly susceptible, especially to fungal diseases. In the present study, the effects of two fungicides, Cuprozin progress and SWITCH, on the metabolism of two cultivars and the wild strawberry were tested, focusing on primary (amino acids, (in)organic acids, sugars, total phenolics) and specialized metabolites (aroma volatiles), which determine the fruit flavor. The fungicide treatment significantly affected 11 out of 57 metabolites, while 20 of those differed between strawberry types and 27 were affected by the interaction of both factors. Given these modifications in metabolites in response to the treatments, the taste and quality of the strawberries may pronouncedly change when plants are treated with fungicides.

Introduction

More than a third of the world’s habitable land is used for agriculture.¹ To avoid the loss of crops due to an infestation with different antagonists, plants are treated with numerous pesticides as a precautionary measure. Against pathogenic fungi, such as the gray mold (Botrytis cinerea), fungicides are commonly used pesticides in fruit cultivation to protect the crop and to prevent economic losses.² For example, fungicides based on copper have been used against molds since the end of the 19th century and are approved in both conventional and organic farming.³ In addition to the broad antifungal spectrum of such fungicides, trace amounts of copper serve as a nutrient for plants, thus enhancing crop quality.⁴ However, heavy metals such as copper cannot be transformed or metabolized and can accumulate in the food chain.⁵ The application of pesticides is also associated with other ecological harms, such as contamination of soils and groundwater, as well as toxic effects on nontarget organisms, such as insects acting as pollinators of many plant species, mollusks, and birds.^2,5 Because of these side effects, application restrictions have been imposed for various pesticides, including copper-based fungicides (Regulation 2002/473/EC).⁶ Despite the risks, pesticides are still used or have to be used. Therefore, it is essential to examine the consequences of their use.

Apart from harmful effects on various organisms and the environment,⁵ pesticides can also cause measurable changes to the metabolism of treated plants.⁴ Copper acts as a contact fungicide that kills fungi by not allowing the mycelia to grow upon spore germination.⁷ However, copper exposure can significantly affect plant metabolites involved in central pathways, such as the galactose metabolism and the tricarboxylic acid (TCA) cycle, leading to suppressed growth, reduced yield, and induced senescence.^8,9 Systemic fungicides, such as anilinopyrimidines or phenylpyrroles, can also stop the fungal dispersal and infection after penetration of the plant tissue.⁷ For example, anilinopyrimidines act on the methionine biosynthesis, inhibiting fungal protein synthesis, while phenylpyrroles interfere with the osmotic signal transduction in fungi.¹⁰ In plants, anilinopyrimidines and phenylpyrroles can inhibit photosynthesis, alternate pigment concentrations, and decrease the chlorophyll content.⁴ Pesticides can also activate the defense responses of plants, increasing their resistance.¹¹ Overall, there is a wide variety of modes of action, in which plant diseases are controlled, depending on the active ingredient. These modes of action may also affect the metabolism of pesticide-treated plants in different ways.

The cultivated strawberry (Fragaria × ananassa Duchesne, Rosaceae) has a worldwide production of 8.9 million tons.¹² However, it is also among the most susceptible crops, and thus a wide variety of fungicides are applied before or during flowering to inhibit the germination of fungal spores.¹³ Due to centuries of crossbreeding and cultivation of different strawberry species, a high species- and cultivar-specific responsiveness to changes and stresses during the development and ripening is known.^14,15 In strawberries, sugars (e.g., glucose, fructose, sucrose), some free amino acids (e.g., valine, leucine, and isoleucine), as well as aroma volatile metabolites determine the flavor, i.e., sweetness, sourness, and aroma of the fruit, respectively.¹⁶⁻¹⁹ Organic acids such as malic acid, fumaric acid, shikimic acid, and ascorbic acid are involved in the sourness, the regulation of the pH value, and the stabilization of anthocyanins.²⁰ Anthocyanins belong to the class of phenolic compounds, which can serve as antioxidants.¹⁵ The unique flavor of strawberries is primarily due to the aroma volatile metabolite mesifuran and the sugar alcohol furaneol, as well as the acid methyl anthranilate.²¹ A high variation between different genotypes has been found in metabolite composition.²⁰

This study aimed to investigate the effects of fungicide application on the flavor of the fruits of two varieties of the horticultural strawberry Fragaria × ananassa and the wild strawberry Fragaria vesca. Flowers of these three strawberry “types” (two horticultural cultivars and wild strawberry) were either kept untreated (control) or were treated with a copper-based broad spectrum fungicide, which is approved in organic farming, or a commonly used fungicide that acts on the methionine biosynthesis and on the histidine-kinase involved in osmotic signal transduction of fungi.¹⁰ Extracts of ripe fruits were analyzed to investigate treatment-specific effects on total soluble phenolics contents, amino acids, organic acids, sugars, and aroma volatile metabolites. Considering that the two fungicides have different modes of action and might influence not only the metabolism of the fungi but also that of the plants, we expected that the targeted plant metabolites would be differentially modulated by the fungicide treatments, affecting the flavor profiles of the fruits. Moreover, we expected the three strawberry types to show specific responses to the fungicide treatments due to the known metabolic differences between strawberry species and varieties.²⁰ Our findings highlight the significant impact of strawberry type and fungicide treatment on different parameters of fruit quality, which may influence the taste of the fruits for humans.

Materials and Methods

Plant Cultivation and Fungicide Treatments

All plants were grown on a 1:1 mixture of steamed potting soil (type: P, Fruhstorfer Pikiererde, Hawita Group, Vechta, Germany) and river sand. Wild strawberry plants (WS; F. vesca) were grown from seeds (Rühlemann, Kräuter und Duftpflanzen, Horstedt, Germany) and kept in a greenhouse (16:8 h light:dark, fluctuating temperature and humidity), where they were grown in 1 L pots. For overwintering, these plants were kept in a climate chamber at 4 °C and 8:16 h light:dark for 5.5 months to initiate flowering in the following year. Afterward, they were transferred back to the greenhouse and re-potted in 4 L pots (15 × 15 × 23 cm). Two cultivars of Fragaria × ananassa, Darselect (DS; early cultivar, fruit ripening from the end of May to mid-June) and Malwina (MW; late cultivar, ripening time from mid-June to mid-July), were purchased as precultivated “frigo” plants (Kraege Beerenpflanzen GmbH & Co.KG, Telgte, Germany) and planted in 4 L pots. DS and MW were each potted at two time points (in total four time points), approximately two weeks apart, to increase the number of simultaneously flowering plants and ripe fruits. Plants were placed in trays, positioned randomly in the greenhouse, water was added daily to the trays, and plants were fertilized once a week (Wuxal Professional Fertilizing, MANNA, Ammerbuch, Germany; NPK fertilizer solution 8-8-6 with trace nutrients: 8% nitrogen, 8% diphosphorus pentoxide, 6% potassium oxide).

Plants of each strawberry type were randomly divided into one of three fungicide treatments, either kept as control (CTR) or treated with the copper hydroxide-based fungicide Cuprozin progress (CU; Spiess-Urania Chemicals GmbH, Hamburg, Germany) or SWITCH, containing cyprodinil and fludioxonil as active substances [FR (fruit rot); Syngenta Agro GmbH, Basel, Switzerland]. The fungicide products were used according to the mean values recommended by the suppliers and were applied with separate compression sprayers. CU has a copper hydroxide concentration of 383 g/L, and the spraying resulted in an application of approx. 9.67 mg per plant per application time point. CU was applied four times: once before, twice during, and once after flowering. The interval between the first three CU applications was 7–8 days each, and the last application followed 17–23 days after the third. FR has a concentration of 2 g/L, and the spraying resulted in an application of approx. 4.69 mg cyprodinil and 3.13 mg fludioxonil per plant per time point. FR was applied twice, once before and once during flowering, with an interval of 16 days. For fungicide treatments, plants were transferred to a separate room and returned after drying the fungicides to avoid cross-contamination between plants of different treatments. In total, the experiment consisted of three strawberry types (DS, MW, WS) × three fungicide treatments (CTR, CU, FR) × 10 replicates per type and treatment (total of 90 pots).

Harvest of Fruits and Leaves

Once all plants in the corresponding strawberry type had at least two ripe fruits, two randomly selected ripe fruits (completely red, not too firm or too soft) and one medium-aged leaf were harvested per plant (DS: May 19 and 26, 2020; MW: June 16 and 30, 2020; WS: evenly distributed across all four time points). For the fruit samples, the tip, as well as two outer slices from the sides of each of both fruits were cut and pooled (DS, MW: 0.7–2.5 g fresh weight, fw, and WS: 0.1–1.5 g fw). Six disks (⌀ 2.8 cm) were cut with a cork borer from one medium-aged leaf per plant and pooled for the leaf samples. All samples were immediately weighed, frozen in liquid nitrogen, and then lyophilized. The dry material was homogenized and stored in a desiccator until analysis. In addition, all remaining fruits and flowers per plant were harvested. The potential yield was determined as the sum of all fruits, including the two taken for analyses and the flowers. Moreover, all fruits were weighed, and the mean fruit weight per plant was calculated.

Photometric Analysis of Total Soluble Phenolics in Fruits and Leaves

Samples (3 mg of fruit, 2 mg of leaf material) were extracted 3-fold in methanol (Fisher Scientific, Geel, Belgium). Supernatants were combined after centrifugation, following a method as described in Mikulic-Petkovsek et al.¹⁷ Three technical replicates were measured per sample against a calibration curve of gallic acid (Carl Roth GmbH & Co.KG, Karlsruhe, Germany). Plant extracts (7.5 μL) and pure extraction solvent (as controls) were diluted with 100 μL Millipore water, and the background absorption was measured at 750 nm in a photometer (Thermo Scientific Multiskan FC, Waltham, Massachusetts). Then, 12.5 μL of Folin-Ciocalteu reagent (Sigma-Aldrich, Chemie GmbH, Munich, Germany) diluted 1:10 with Millipore water, and 100 μL of 1 M sodium carbonate solution (Carl Roth) were added to each sample. The plate was incubated at 45 °C for 20 min and measured again at 750 nm. The phenolic contents of samples were determined in relation to a gallic acid calibration curve after subtraction of the background absorption.

Amino Acid Analysis of Fruits

The amino acid composition in the fruits was analyzed according to Barber and Müller.²² Fruit material (4 mg) was extracted 3-fold in 80% methanol (LC-MS quality; VWR International GmbH, Leuven, Belgium), including the internal standards norvaline for primary and sarcosine for secondary amino acids (Agilent Technologies, Waldbronn, Germany). The supernatants were combined, filtered (0.2 μm poly(tetrafluoroethylene) filter, Phenomenex, Aschaffenburg, Germany), and frozen at −80 °C until further analysis. Samples were analyzed using a high-performance liquid chromatograph equipped with a ZORBAX Eclipse Plus C18 column (250 mm × 4.6 mm, 5 μm) with guard column (Agilent Technologies, Santa Clara, California) coupled with fluorescence detection (HPLC-FLD; 1290 Infinity HPLC and 1260 Infinity FLD, Agilent Technologies). Each sample was derivatized in the autosampler at 6 °C with borate buffer (0.4 M, pH = 10.2; Agilent Technologies), ortho-phthaldialdehyde reagent (OPA; 10 mg mL/L in borate buffer and 3-mercaptoproprionic acid; Agilent Technologies), 9-fluorenyl-methyl chloroformate reagent (FMOC; 2.5 mg/mL in acetonitrile; Agilent Technologies), and injection diluent [mobile phase A: 1.4 g Na₂HPO₄ (>99.5%; AppliChem GmbH, Darmstadt, Germany), 3.8 g Na₂B₄O₇ × 10 H₂O (≥99.5%; Sigma-Aldrich), and 32 mg NaN₃ (≥98%; Carl Roth) solved in 1 L Millipore water, pH = 8.2 and mixed with 85% phosphoric acid (AppliChem) in a ratio of 1:0.004 (v:v)]. Samples were injected and analyzed using a gradient of eluent A and eluent B [4.5:4.5:1 (v:v:v) methanol, acetonitrile (both LC-MS quality, VWR International), and Millipore water], starting with 2% B for 0.84 min, increasing to 57% B within 53.4 min, reaching 100% B at 53.5 min, followed by a cleaning and column equilibration step. Primary amino acids (OPA-derivatized) and later eluting secondary amino acids (FMOC derivatized) were detected at an excitation wavelength of 340 or 260 nm and emission wavelength of 450 or 325 nm, respectively, with the switching point at 36.6 min. In addition to the samples, blanks and reference standards were measured. Amino acid data were analyzed with OpenLab ChemStation revision C.01.07 (Agilent Technologies). Only amino acids that could be identified by matching retention times to those of reference standards were included in the dataset. Each amino acid was quantified using its peak area and an amino acid-specific calibration response factor relative to the respective internal standard and the sample dry weight. The amino acid-specific calibration response factors were calculated based on additional measurements of standards in different concentrations within the linear detection range, with the internal standards being applied in the concentration that was also used for the plant samples. Means of the factors along the concentration range were then used for quantification. Only amino acids were included in the analysis that did not occur in the blanks or only occurred in much lower concentrations (max. 10% of the corresponding peak) in the blanks than in the samples and that were found in at least half of the replicates of at least one treatment group.

Sugar and (In)organic Acid Analyses of Fruits

Sugars and (in)organic acids were analyzed from strawberry fruits as in Schweiger et al.²³ Therefore, 4 mg of fruit material was extracted in a 2.5:1:1 mixture (v:v:v) of methanol, water, and chloroform (both AppliChem), containing ribitol (Sigma-Aldrich) as internal standard. After phase separation, the aqueous phase was dried, and aliquots methoxymated with O-methylhydroxylamine hydrochloride (MeOX) (Sigma-Aldrich) dissolved in pyridine (Sigma-Aldrich), and subsequently silylated with N,O-bis(tri-methylsily)trifluoroacetamide (BSTFA) supplemented with 1% tri-methylsilyl chloride (Sigma-Aldrich); both incubated for 30 min at 50 °C. The derivatized samples were diluted in pyridine and analyzed using gas chromatography coupled with mass spectrometry (GC-MS; GC 2010plus—MS QP2020, Shimadzu, Kyoto, Japan) on a VF-5 MS column (30 m × 0.2 mm ID, 10 m guard column, Varian, Palo Alto, CA) with helium as carrier gas. The GC-injection temperature was 250 °C and operated with a split ratio of 1:10 and a flow rate of 1.14 mL/min. The initial oven temperature of 100 °C was held for 1 min, increased by 2 °C/min to 200 °C, then increased by 10 °C/min to 280 °C, held for 5 min, then increased to 300 °C with 20 °C/min and held for 5 min. The MS source temperature was 230 °C, the interface temperature was 250 °C, and the solvent delay was 10 min. The line spectra (range: 40–600 m/z) of separated metabolites were acquired in quadrupole MS mode at a scan rate of 0.3 s/full scan.

The Shimadzu GC-MS Postrun analysis program (version 4.45) was used to analyze the chromatograms. In addition to the fruit extracts, blank samples and alkane standard mixtures (C8–C20 and C21–C40, Sigma-Aldrich) were measured to calculate retention indices (RI) according to Van den Dool and Kratz for targeted metabolites.²⁴ For identification, RI and mass spectra were compared with the NIST database (NIST14, National Institute of Standards and Technology, Gaithersburg, Maryland) and an in-house database. After subtraction of contaminations using the blank samples, the peak areas of identified metabolites within each sample were normalized to the peak area of the internal standard and sample dry weight.

Collection of Fruit Aroma Volatile Metabolites and Analysis

Fruit aroma volatile metabolites were collected from a fruit puree according to a modified protocol of Ulrich and Olbricht.²⁵ Dry fruit material (about 20 mg) was combined with 20% NaCl solution, rehydrated, centrifuged, and the supernatant transferred to a 2 mL Eppendorf reaction tube. As an external standard, 1-bromodecane (Sigma-Aldrich) diluted in n-heptane (CHEMSOLUTE, Th. Geyer GmbH & Co. KG, Renningen, Germany) was added to the extracts together with two polydimethylsiloxane (PDMS, 1 mm internal diameter, 1.8 mm external diameter; Carl Roth) tubes and stirred for 40 min. Subsequently, the PDMS tubes were removed, cleaned with Millipore water, dried with a lint-free paper towel, and stored at −80 °C until further use. Collected samples on the PDMS tubes and blank samples were desorbed using a thermal desorption unit (TD, TD-20, Shimadzu) and measured by GC-MS on a VF5-MS column with helium as carrier gas. The column oven temperature started at 50 °C and was held for 5 min, then increased by 10 °C/min to 250 °C, further increased by 30 °C/min to 280 °C and was finally held for 2 min. The MS ion source temperature was 230 °C with an interface temperature of 280 °C. The scan was performed from the first to 28th min, and the line spectra (range: 30–400 m/z) of separated metabolites were acquired in quadrupole MS mode at a scan rate of 0.2 s/full scan.

The KI, the FFNSC3-database (University of Messina, Messina, Italy) specifically for flavor and fragrance analysis, the Good Scents Company Information System database (http://www.thegoodscentscompany.com/), the PubChem database,²⁶ and Adams²⁷ were used for putative identification of metabolites. Only annotated aroma volatile metabolites were used for further analyses. After subtraction of contaminations using the blank samples, the peak areas of metabolites within each sample were normalized to the peak area of the internal standard and the sample dry weight. Furthermore, all metabolites, whose mean intensity in at least one combination of treatment × strawberry type was higher than twice their mean intensity in the blank samples, were kept in the dataset. Likewise, metabolites that were present in at least 50% of the samples in at least one of the treatment × strawberry type combinations were kept (similar to Schrieber et al.²⁸). The final dataset included 21 metabolites.

Statistical Analyses

All statistical analyses and data visualization were performed with RStudio²⁹ in R 4.1.2³⁰ with a significance threshold of α = 0.05. All figures were generated using the package plyr.³¹ Linear mixed-effects models (LMM) were performed for response factors with a normal error distribution and a generalized linear (mixed-effects) model [GL(M)M] for responses with a Gamma or binominal distribution using the R package lme4.³² Model variance homogeneity and normal distribution of residuals were checked by visual inspection.³³ If necessary, the data were shifted by 1 × 10^–07 along the x-axis to positive x-values to fit a Gamma distribution. By the marginality rule model simplifications via the dropterm function (package MASS(34)) were carried out based on Chi-square likelihood ratio tests for (G)LMM to generate p-values for factor interactions and individual factors (R package car(35)). To test the effect of fungicide treatments on the potential yield and mean fruit weight, GLMs (Poisson, link: log and Gamma, link: inverse, respectively) with the factorial predictors fungicide treatment (CTR, CU, and FR), strawberry type (WS, DS, and MW) and their interaction were performed. The effects of the factorial predictors fungicide treatment, strawberry type, and their interaction on chemical traits were tested using separate GLMMs for the total content of soluble phenolics (link: identity), the total amino acid concentration (sum of 19 amino acids) and the total content of aroma volatile metabolites (sum of 21 metabolites) (Gamma, link: log and binominal, link: logit). The total sugar content (sum of 10 metabolites) was analyzed with an LMM (Gaussian, link: identity). The total (in)organic acid content (n = 3) could not be analyzed due to singularity in the model. The fruit fresh weight per plant was included as random factor for all four models. To test effects on individual metabolites, (G)LMMs were calculated, including the factorial predictors fungicide treatment, strawberry type, and their interaction, as well as fruit fresh weight per plant as random effect. To visualize the metabolic composition of (a) amino acids, (b) sugars, and (c) aroma volatile metabolites of each fruit, separate nonmetric multidimensional scaling (NMDS) with Wisconsin double standardization of square-root transformed data were conducted for these three datasets, using the Kulczynski distance (R package vegan(36)). Additionally, permutational multivariate analyses of variance using distance matrices (ADONIS) based on Kulczynski distance were used to test the effects of the factors fungicide treatment, strawberry type, and their interaction on the three sets.

Results and Discussion

Effects of Fungicide Treatment and Strawberry Type on the Potential Yield and Fruit Weight

Overall, the potential yield, i.e., the sum of flowers and fruits, was relatively low in the plants growing in our greenhouse. The yield was not influenced by the fungicide treatment. Still, it was significantly influenced by the strawberry type, with DS plants having with around eight flowers and fruits the highest yield and MW plants having on average three fruits less (Figure S1). Likewise, the fungicide treatment did not affect the mean fruit fw but differed significantly between plants of the three strawberry types (Figure 1). Horticultural strawberry fruits were on average around 19 times heavier than the WS fruits, while within the cultivars, the fruits of MW were 1.3 times heavier than those of DS (Figure 1). Differences in both the potential yield and average fruit fw between the strawberry species and cultivars were in line with our expectations and are well known from different strawberry cultivars and varieties.³⁷ However, our results contrast with the plant supplier’s data, as our average fresh weight value per fruit was approx. 60% (DS) and 40.8% (MW) lower than that postulated by the supplier. In the field, DS plants should reach a fruit yield of around 650–900 g with 23–24 g average fw per fruit and MW plants a yield of around 750 g with 21–22 g average fw per fruit.³⁸ Reasons for the much lower yields in our experimental plants could be the light conditions in the greenhouse, the composition of the used substrate, or limitations by the pot size since the pots were filled entirely with roots at the end of the experiment.

Figure 1.

Mean fruit fresh weight per plant of different strawberry types [Fragaria × ananassa: Darselect (DS), Malwina (MW); F. vesca: wild strawberry (WS)], which were treated with fungicides [control (CTR), Cuprozin progress (CU), SWITCH (FR)]. Data are presented as box-whisker plots with interquartile ranges (IQR; boxes) including medians (horizontal lines) and whiskers (extending to the most extreme data points with a maximum of 1.5 times the IQR), while black dots indicate the means; individual values are given as circles. Significant difference (p < 0.05) of the linear model is shown in the graph; n = 8–10 replicates per fungicide treatment and strawberry type (=groups).

Effects of Fungicide Treatment and Strawberry Type on Total Phenolic Contents

The total phenolic content in the fruits, given as gallic acid equivalents, was significantly affected by the interaction of the fungicide treatment and strawberry type (Figure 2). Differences were found in CTR fruits between strawberry types, with fruits of the WS plants having highest and DS plants having lowest phenolic contents. With regard to fungicide treatments, the CU treatment resulted on average in a higher phenolic content in fruits than the CTR in DS plants. In contrast, no difference was observed between CU-treated and CTR fruits in WS and MW plants. FR-treated fruits showed on average a lower phenolic content than CTR fruits in WS plants (Figure 2). Ascorbic acid, which is highly abundant in strawberry fruits,³⁹ may have dominantly contributed to the total phenolics. In strawberry fruits of the cultivars Elsanta and Honeoye, an increase in phenolic content after treatment with the fungicide Signum (active substances: boscalid and pyraclostrobin; mode of action: inhibition of mitochondrial respiratory chain; BASF, Ludwigshafen, Germany) was found.^10,17 In contrast, the content of total phenols decreased in the peel and pulp of apples (Malus domestica) due to treatments with the inorganic fungicide sodium bicarbonate and a commonly used fungicide for apple cultivation.⁴⁰ Differences in fruit phenolic content have been reported for different strawberry species and cultivars.^14,15 Phenols enable plants to withstand oxidative stress³⁹ and ultraviolet radiation,⁴¹ and have antifungal potential.⁴² Due to the fungicide treatment and thus likely lower stress levels, the phenolic compounds may have been downregulated in our experimental plants, allowing for a relocation of resources to other metabolic pathways. It should be noted that the FR fungicide left a whitish residue on the leaves of some plants in our experiment (personal observation), which may offer a protection layer against ultraviolet radiation and may also contribute to a decrease in the phenolic content. While fruit phenolic contents were responsive to the fungicide treatment in the present study, the fungicide did not affect the phenolic content of leaves (Figure S2). However, leaf phenolic contents differed between strawberry types (Figure S2). These findings are in contrast to other studies, based on the same quantification method, which found elevated phenolic contents in the leaves of cucumber (Cucumis sativus) or tobacco (Nicotiana tabacum) due to fungicide treatments.^43,44 Apart from species-specific effects, the effect of the fungicide treatment on phenolic contents may depend on the used fungicide class and the regulation of the metabolic pathway of individual phenolic compounds.¹¹

Figure 2.

Phenolic contents (measures as gallic acid equivalents) in ripe fruits of different strawberry types [Fragaria × ananassa: Darselect (DS), Malwina (MW); F. vesca: wild strawberry (WS)], which were treated with fungicides [control (CTR), Cuprozin progress (CU), SWITCH (FR)]. Data are presented as box-whisker plots with interquartile ranges (IQR; boxes) including medians (horizontal lines) and whiskers (extending to the most extreme data points with a maximum of 1.5 times the IQR), while black dots indicate the means; individual values are given as circles. Significant difference (p < 0.05) of the linear model is shown in the graph; n = 8–10 replicates per fungicide treatment and strawberry type (=groups); dw—dry weight.

Effects of Fungicide Treatment and Strawberry Type on Fruit Amino Acids

A total of 22 amino acids were detected in the fruits, of which 19 could be well quantified. Three amino acid peaks (asparagine, glutamine, and a combined peak of arginine and alanine, which co-eluted at the same retention time) were overloaded in the chromatograms, so only their presence or absence could be considered (Table S1). The three overloaded amino acid peaks were neither included in the total amino acid concentration (sum of 19 amino acids) nor the NMDS. The presence of asparagine and the combined peak of arginine and alanine were significantly higher in WS plants than in DS and MW plants, while the presence of glutamine was not affected by the tested factors (Table S1). The concentration of 11 amino acids was significantly influenced by the fungicide treatment × strawberry type interaction (e.g., phenylalanine, lysine, proline, isoleucine), four were affected by the factors fungicide treatment and strawberry type individually (glycine, valine, tryptophan, leucine), one was affected by the fungicide treatment only (methionine), and three amino acids (aspartic acid, histidine, hydroxyproline) were influenced by the strawberry type only (Table S1).

The total amino acid concentration (sum of 19 amino acids) differed significantly between the fruits of the different fungicide treatments, with fruits of the CU treatment having on average a 9% higher and fruits of FR treatment a 26% higher amino acid concentration than fruits of control plants (Figure 3A, Table S1). In contrast, the strawberry type had no effect on the total amino acid concentration. There were also some outliers in the different groups, especially in the WS-CU group and MW-FR group, which had multiple times higher values than the mean value of the corresponding groups (Figure 3A). Fruits of these outlier WS plants had a higher concentration especially of hydroxyproline, compared to the horticultural strawberry. Outliers of DS and MW fruits had higher concentrations of aspartic acid, glycine, serine, and glutamic acid than those of WS plants. These distinct increases in concentration could be related to different responses to environmental stress.²⁰ In general, many of the individual amino acids were influenced by the interaction of fungicide treatment × strawberry type, which means that it depends on the specific strawberry type and the respective fungicide treatment in which direction the fruits were modified (Table S1).

Figure 3.

Amino acid concentrations and composition in fruits of different strawberry types [Fragaria × ananassa: Darselect (DS), Malwina (MW); F. vesca: wild strawberry (WS)], which were treated with fungicides [control (CTR), Cuprozin progress (CU), SWITCH (FR)]. (A) Total amino acid concentration (sum of 19 amino acids), presented as box-whisker plots with interquartile ranges (IQR; boxes) including medians (horizontal lines) and whiskers (extending to the most extreme data points with a maximum of 1.5 times the IQR), while black dots indicate the means; individual values are given as circles; (B) mean amino acid composition (averaged over replicates within groups) and (C) nonmetric multidimensional scaling (NMDS; with Kulczynski distance matrix) of the absolute amino acid concentrations with scores (colored symbols; samples within each group are surrounded by convex hulls and the corresponding medians of the groups shown as gray crosses) and loadings (blue metabolite names); full names of the metabolites are given in Table S5. Significant differences (p < 0.05) of the linear model in (A) and the ADONIS in (C) are shown in the graphs; n = 7–10 replicates per fungicide treatment and strawberry type (=groups); dw—dry weight.

Especially the concentrations of three amino acids, namely valine, leucine, and isoleucine, were on average 12–48% higher in fruits of the CU treatment, while they were 4–11% lower in fruits of the FR treatment compared to fruits of control plants (Figure 3B, Table S1). In fruits of the CU treatment, the concentrations of two further amino acids, phenylalanine and methionine, were around 3.4 and 2.2 times higher, respectively (Figure 3B). In fruits of the FR treatment, the concentrations of phenylalanine and methionine were about 1.7 and 1.1 times higher, respectively, compared to fruits of control plants (Figure 3B). The fungicide-induced changes in these five amino acids may lead to modifications in the fruit aroma volatile composition because these amino acids are essential precursors of volatile compounds such as aldehydes, alcohols, acids, and esters in ripening fruits.¹⁸ The effect of the fungicide SWITCH on the amino acid methionine is particularly intriguing because one of the active compounds of this fungicide is cyprodinil, which inhibits the methionine biosynthesis of the fungi.¹⁰ In plants, methionine feedback regulates its own de novo biosynthesis with cystathionine synthesis as a regulator, but methionine can also be regained through the methionine salvage pathway.⁴⁵ Via the common branching point of S-adenosylmethionine (AdoMet), methionine biosynthesis is linked with the biosynthetic pathway of ethylene,⁴⁵ which is upregulated in response to different abiotic challenges.⁴⁶ The significant fungicide treatment effect on methionine concentrations in the fruits of FR-treated plants found in the present experiment highlights that not only the metabolism of the targets (fungi) but also that of the host plants can be affected. Fungicide-induced changes in plant metabolism may result in the disruption of important functions, such as growth and signaling processes, as well as plant stress responses.¹¹

The amino acid composition of the strawberry fruits depicted in an NMDS showed no separation in relation to the fungicide treatment. In contrast, the composition differed significantly among strawberry types, with fruits of WS plants showing a distinct amino acid composition compared to the horticultural strawberry fruits, while fruits of MW CTR plants overlapped largely with all samples of horticultural strawberries and treatment groups (Figure 3C). Differences in the amino acid profiles between strawberry types had been expected because distinct compositions in primary metabolites, including amino acids, sugars, and organic acids, have been reported for different varieties of strawberries.²⁰ The composition of these primary metabolites can also be influenced by different abiotic factors, such as rainfall or temperature, as well as farming conditions in strawberries.^20,47

Effects of Fungicide Treatment and Strawberry Type on Fruit Sugars and (In)organic Acids

In total, 10 sugar metabolites, including six identified sugars and two putative sugars (showing at least the characteristic masses of sugars, such as 103, 147, 173, 205, 217, and 307 m/z), and two sugar alcohols (sorbitol and cyclic polyol myo-inositol) were detected in the fruits, next to two organic acids and the inorganic phosphoric acid (Table S2). The content of five metabolites (ribose, rhamnose, myo-inositol, malic acid, citric acid) was influenced by the fungicide treatment × strawberry type interaction, two sugars (glucose, fructose) were affected by the fungicide treatment, and three other sugars (xylose, sucrose, putative sugar 1) were influenced by the strawberry type (Table S2). The content of sorbitol, phosphoric acid, putative sugar 2, and the sum of (in)organic acids (three metabolites) could not be statistically analyzed due to zero inflation in the dataset (Table S2). Due to the low acid content and zero inflation in the dataset, we focus mainly on the sugar metabolites.

The sum of all 10 sugar metabolites differed significantly between fruits of the different strawberry types, with fruits of DS plants having on average a 37.7% higher content and those of MW a 23.2% higher content than fruits of WS plants (Table S2). The main sugars found in the fruits were glucose and fructose, which is in line with many other studies on strawberry fruits (e.g., Akhatou et al.²⁰ or Fait et al.¹⁶). These two sugars were significantly affected by the treatment. In WS and MW plants, the glucose content was on average 2.8 and 5.5 times higher in fruits of the FR treatment than in those of the CU and CTR treatment, respectively (Figure 4A, Table S2). Likewise, the fructose content in fruits of the FR treatment was 1.6 and 1.3 times higher than in those of the CU and CTR treatment, respectively (Figure 4B, Table S2). In contrast to glucose and fructose, the sucrose content was significantly different between strawberry types, being 3.1 and 5 times higher in fruits of MW plants than in those of DS and WS plants, respectively (Figure 4C, Table S2). Notably, the ratio of glucose, fructose, and sucrose shifted in the fruits due to the fungicide treatment of the plants (Figure 4E). This ratio is considered important because it affects the taste of fruits and is an important factor for nutritive and nutritional fruit quality.⁴⁸ Especially in fruits of WS and MW plants, the proportion of glucose increased compared to fruits of control plants (Figure 4E). The fungicide-induced shift in sugar ratios in fruits of DS plants was different from those in the other strawberry types. Fruits of DS control plants had a higher proportion of glucose, those of CU-treated plants had a higher proportion of fructose, and fruits of the FR treatment had a higher proportion of sucrose (Figure 4E). Glucose is a product of photosynthesis but also, together with fructose, an enzymatic breakdown product of sucrose, whose breakdown provides energy (ATP) for the cell.²⁰ In plant metabolism, glucose is an essential sugar for structural, nuclear, and biochemical processes as well as for fruit ripening.⁴⁸ Thus, treatment with fungicides may upregulate AdoMet, which is converted from methionine at the expense of ATP utilization.⁴⁶ This higher need of ATP for ethylene biosynthesis may explain the shifts in sugar ratios in the fruits of the fungicide-treated plants. Increased levels of sugars have been previously reported from strawberry fruits treated with a fungicide (Signum, BASF) that targets the succinate-dehydrogenase in the TCA cycle.¹⁷

Figure 4.

Sugar contents and composition of different strawberry types [Fragaria × ananassa: Darselect (DS), Malwina (MW); F. vesca: wild strawberry (WS)], which were treated with fungicides [control (CTR), Cuprozin progress (CU), SWITCH (FR)]. (A) Glucose, (B) fructose, and (C) sucrose content are presented as box-whisker plots with interquartile ranges (IQR; boxes) including medians (horizontal lines) and whiskers (extending to the most extreme data points with a maximum of 1.5 times the IQR), while black dots indicate the means; individual values are given as circles; (D) nonmetric multidimensional scaling (NMDS; with Kulczynski distance matrix) of the sugar composition with scores (colored symbols; samples within each group are surrounded by convex hulls and the corresponding medians of the groups shown as crosses) and loadings (blue metabolite names), full names of the metabolites are given in Table S5; and (E) glucose:fructose:sucrose ratio. Significant differences (p < 0.05) of the linear models in (A–C) and the ADONIS in (D) are shown in the graphs; n = 8–10 replicates per fungicide treatment and strawberry type (=group).

The profiles of sugar metabolites of the nine different fungicide treatments and strawberry type combinations overlapped largely in an NMDS. Nevertheless, significant differences were found due to fungicide treatment and strawberry type (Figure 4D). As discussed above, these shifts in the primary metabolite composition of the fruits of fungicide-treated plants may reflect the impacts of the used fungicides on plant metabolism and may contribute to changes in the overall fruit quality, shelf life, and ripening, as found in other strawberry fruits (cultivar: Chandler).⁴⁹

Effects of Fungicide Treatment and Strawberry Type on the Fruit Aroma Volatile Metabolites

In total, 21 aroma volatile metabolites were detected in the fruits, including two metabolites each from the metabolite classes of acetate esters, acids, alcohols, and furanones, five aldehydes, four lactones, three aromatic metabolites, and one terpenoid (Table S3). Of these 21 metabolites in the fruits, 11 were influenced by the fungicide treatment × strawberry type interaction (e.g., mesifurane and nonanal), while (E)-nerolidol was affected by fungicide treatment and strawberry type individually (Table S3, Figure 5B). Further, two metabolites were affected by the treatment [(E)-hex-2-enal and vanillin], and two were influenced by the strawberry type (δ-decalactone and hexyl acetate) (Table S3, Figure 5B). Five metabolites [e.g., benzaldehyde and (E)-hept-2-enal] showed no significant differences. Due to zero inflation, the metabolite 5-methyl-hexanoic acid could not be statistically analyzed (Table S3). The detected aroma volatile metabolites are all known for strawberry fruits^27,50,51 and arise from different pathways of the plant metabolism (for details see Table S4), which include amino acids (valine, leucine, isoleucine, phenylalanine, and methionine), sugars (fructose, glucose), and organic acids (malic acid, citric acid).^20,52 The changes in amino acids and sugars found in fruits after the different fungicide treatments compared to the fruits of the control plants may have affected the aroma volatile metabolites. These aroma volatile metabolites arise from different metabolic pathways involving different nonvolatile metabolites.⁵³ Thus, correlations can hardly be drawn between individual compounds.

Figure 5.

Aroma volatile metabolite contents and composition of different strawberry types [Fragaria × ananassa: Darselect (DS), Malwina (MW); F. vesca: wild strawberry (WS)], which were treated with fungicides [control (CTR), Cuprozin progress (CU), SWITCH (FR)]. (A) Total aroma volatile metabolite (sum of 21 metabolites), (B) mesifurane, (C) (E)-nerolidol, (D) γ-decalactone, and (E) 2,5-dimethyl-3(2H)-furanone were presented as box-whisker plots with interquartile ranges (IQR; boxes) including medians (horizontal lines) and whiskers (extending to the most extreme data points with a maximum of 1.5 times the IQR), while black dots indicate the means; individual values are given as circles and (F) nonmetric multidimensional scaling (NMDS; with Kulczynski distance matrix) of the aroma volatile metabolite composition with scores (colored symbols; samples within each group are surrounded by convex hulls and the corresponding medians of the groups shown as gray crosses) and directional pulling loadings (blue metabolite name), full names of the metabolites are given in Table S5. Stress values are given in the right corner. Significant differences (p < 0.05) of the linear models in (A–E) and the ADONIS in (F) are shown in the graphs; n = 8–10 replicates per fungicide treatment and strawberry type (=group).

The total content (i.e., normalized peak area) of aroma volatile metabolites (sum of 21) was significantly affected by the fungicide treatment × strawberry type interaction. Fruits of MW and DS plants had 1.4 times and 4.8 times higher content, respectively, than fruits of WS plants which showed the lowest average content (Figure 5A). Cultivated strawberries have been crossbred and selected for higher concentrations of specific aroma volatile metabolites than wild strawberries.⁵⁴ Such differences in profiles could also be observed in the present study. The effect of the fungicide treatment depended on the strawberry type, whereby the contents of aroma volatile metabolites in fruits of the CU and FR treatment were lower in WS and DS plants compared to fruits of control plants and vice versa in fruits of MW plants (Figure 5A). Interestingly, although the content of the precursor metabolites (amino acids, sugars) in the fruits of the DS plants was not affected by the two fungicides, the fungicide treatments caused a stronger decrease in the total content of aroma volatile metabolites (Figures 3A, 4A–C, and 5A). In particular, the contents of the aroma volatile metabolites mesifurane, (E)-nerolidol, and γ-decalactone differed between the fruits of the different fungicide treatments and strawberry types (Figure 5B,C, Table S3). The terpenoid (E)-nerolidol was only present in the two horticultural cultivars (Figure 5B) which indicates that these fruits may have a greener flavor with fruity nuances compared to the WS fruits (Table S4⁵³). The content of the lactone γ-decalactone was particularly higher in fruits of DS plants than those of WS and MW plants, and higher in fruits of the CTR treatment than in those of fungicide-treated plants (Figure 5D). This higher content in γ-decalactone may give the fruits a fruity flavor with a syrupy, fatty nuance (Table S4⁵³). 2,5-Dimethyl-3(2H)-furanone was significantly affected by the fungicide treatment × strawberry type interaction (Figure 5E, Table S3). This metabolite is a breakdown product of the key strawberry aroma metabolite furanone [4-hydroxy-2,5-dimethyl-3(2)-furanone], which is unstable in air and aqueous solutions and sensitive to higher temperatures,⁵¹ which may be the reason why we only found the breakdown product.

In line with the changes in the total contents, the aroma volatile metabolite profiles were significantly influenced by the fungicide treatment, strawberry type, and their interaction, as depicted in the NMDS (Figure 5F). A particularly clear separation was found between the profiles of fruits of DS plants compared to both fruits of WS and MW plants, which were largely overlapping. Given the various changes in aroma volatile metabolites in response to the treatments, the taste of the strawberries and their quality may be strongly affected by different fungicides.

The flavor profiles of strawberry fruits depend on numerous factors, including genetic variability,¹⁹ domestication,⁵⁴ and cultivation techniques.⁴⁷ Our results indicate that fungicide treatment is a further factor contributing to differences in flavor profiles among strawberry fruits. Overall, we found fungicide-related alterations in 11 out of 57 metabolites and strawberry type-related differences in 20 out of these metabolites. In addition, 27 metabolites were significantly affected by the interaction of fungicide treatment and strawberry type. With these results, our study is in line with many studies, which present evidence that fungicides influence crops and their fruits at different metabolic levels, resulting in disturbed development, changes, or losses of quality^4,8 but also in activation of defense responses.¹¹ In addition, fungicide residues can occur in fruits. After the second application of cyprodinil with the recommended dosage, fungicide residues in strawberry fruits can increase the health risk, and it is recommended not to use this type of active agent.⁵⁵ Instead, alternatives may be used that are also permitted in organic farming, such as, for example, copper-containing products, products based on potassium bicarbonate, sodium bicarbonate, or sulfur/lime sulfur.^40,42

During the development of strawberry fruits, the metabolic composition of phenols, amino acids, sugars, and organic acids changes over time.^15,16 In future experiments, fungicide effects on different fruit stages could be analyzed to understand at which developmental stage fruit metabolism may be most impacted. Furthermore, since fungicide applications start even before the flowering stage of the strawberry plant,¹³ their effects on flower odor, nectar, and pollen composition, as well as on pollinators should be investigated in detail.

Glossary

Abbreviations Used

DS

Darselect

MW

Malwina

WS

wild strawberry

CTR

control

CU

Cuprozin progress

FR

SWITCH

HPLC

high-performance liquid chromatograph coupled with fluorescence detection

OPA

ortho-phthaldialdehyde reagent

FMOC

9-fluorenyl-methyl chloroformate reagent

MeOX

O-methylhydroxylamine hydrochloride

BSTFA

N,O-bis(tri-methylsily)trifluoroacetamide

GC-MS

gas chromatography coupled with mass spectrometry

PDMS

polydimethylsiloxane

TD

thermal desorption unit

LMM

linear mixed-effects models

GL(M)M

generalized linear (mixed-effects) model

NMDS

nonmetric multidimensional scaling

TCA cycle

tricarboxylic acid cycle

AdoMet

S-adenosylmethionine

Data Availability Statement

The Dataset and R-script are available on Dryad at DOI: https://datadryad.org/stash/share/XxNLAM1dvdYajf1r5V340S79k-tbyIVYlGBFSLgDl6M.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jafc.2c06584.

• Sum of numbers of flowers and fruits (Figure S1); gallic acid equivalents of leaves of different strawberry types (Figure S2); output of (generalized) linear mixed-effects models of all metabolites (Tables S1–S3); pathways and flavor description of aroma volatile metabolites (Table S4); and metabolites with their details (Table S5) (PDF)

Author Contributions

A.-C.V., E.J.E., and C.M. conceived and designed the experiments; A.-C.V. performed the experiments and analyzed the data with the help of E.J.E. and C.M.; A.-C.V. conducted all statistical analyses, prepared all figures and tables, and wrote the original draft of the manuscript; C.M. edited the writing. All authors have read the manuscript and given final approval for publication.

This work was funded by the Scholarship program of the German Federal Environmental Foundation (Deutsche Bundesstiftung Umwelt; DBU).

The authors declare no competing financial interest.