Inhibition of growth and ochratoxin A production in Aspergillus species by fungi isolated from coffee beans

Abstract

Ochratoxin A (OTA) is a mycotoxin found in several agricultural commodities. Produced by Aspergillus spp., it is nephrotoxic and hepatotoxic and can be carcinogenic. Preventive measures are preventing fungal growth and OTA production. In this study, fungal strains (Rhizopus oryzae, Lichtheimia ramosa, Aspergillus westerdijkiae, Aspergillus niger, Aspergillus tamarii, Aspergillus sp., and Aspergillus fumigatus) isolated from coffee beans were identified for their abilities to inhibit the growth of Aspergillus ochraceus, Aspergillus westerdijkiae, Aspergillus carbonarius, and Aspergillus niger, and OTA production. All fungi strains tested were able to inhibit growth of the four Aspergillus species and OTA production, where A. niger showed the best results in both tests. L. ramosa showed the lowest growth-reducing potential, while the other fungal strains had a growth-reducing potential higher than 70% against all Aspergillus species tested. Regarding OTA production, L. ramosa and Aspergillus sp. completely inhibited the mycotoxin production by A. ochraceus and non-toxigenic strain A. niger completely inhibited OTA production by A. niger. Our findings indicate that the strains tested can be used as an alternative means to control growth of OTA-producing fungi and production of the mycotoxin in coffee beans.

Introduction

Ochratoxin A (OTA) is a mycotoxin produced by several species of Aspergillus and Penicillium that can contaminate many food commodities prior to harvest or during storage [1]. OTA is nephrotoxic, immunotoxic, hepatotoxic [2–5], and a potential carcinogen (group 2B) by the International Agency for Research on Cancer (IARC) [6, 7]. Several toxic effects of OTA have been described, such as inhibition of protein synthesis, oxidative stress, and DNA damage [8–11]. As a preventive measure, the European Union has set the maximum permitted levels of OTA at 5 μg/kg for roasted coffee (beans or ground) and 10 μg/kg for instant coffee [12]. In Brazil, the maximum permitted level of OTA in both roasted coffee (beans or ground) and instant coffee is 10 μg/kg [13].

OTA can be found in a variety of agricultural products including cereals, coffee, raisins, grapes, and spices [1, 14]. Other food items are also contaminated such as milk, animal tissues, muscle, eggs, wine, and beer [15–18]. In Brazil, the presence of OTA in coffee beans and its contamination levels vary significantly. Batista et al. evaluated the incidences of ochratoxigenic fungi and OTA in coffee beans at different stages of maturation and processing, from southern Minas Gerais (Brazil) [19]. The authors found that, among 289 samples analyzed, OTA was detected in 56%; in the other 217 samples, OTA levels were below 5.0 μg/kg of coffee. However, they described that in 82 samples of coffee swept from ground OTA was detected at levels above 100 μg/kg, which is higher than the maximum permitted levels of OTA in Brazil and the European Union.

In temperate regions, OTA is mainly produced by Penicillium species, whereas in tropical and subtropical areas it is produced mostly by Aspergillus species [20, 21]. Some species of Aspergillus, such as Aspergillus westerdijkiae, Aspergillus steynii, Aspergillus ochraceus and related species (section Circumdati), Aspergillus carbonarius, and a small number of isolates of Aspergillus niger (section Nigri), were described as OTA producers in coffee [22–25].

Several strategies have been employed to prevent and control the growth of OTA-producing fungi in beans, such as the use of chemical and antifungal products post-harvest. However, these products lead to an increased risk of toxic residues in food [26–28]. In addition, indiscriminate use of antifungals may result in the development of fungal resistance [29]. Thus, biological methods to control the production of OTA in food have recently raised expectations. Researchers have demonstrated the potential of bacteria, yeasts, and filamentous fungi in inhibiting the growth of OTA-producers fungi, decreasing the production or leading to mycotoxin degradation.

Microorganisms have the potential to inhibit, degrade, or absorb mycotoxins, as Fiori et al. evaluated four yeast strains against A. carbonarius OTA producer and tested their ability to remove OTA from grape juice. According to the authors, the yeasts Candida friedrichii, Candida intermedia, and Lachancea thermotolerans reduced the concentration of OTA present in contaminated grape juice by up to 75%, showing their enormous capacity to absorb toxins [7]. Yeast can also inhibit OTA, as Masoud and Kaltolft have shown when testing yeasts involved in coffee fermentation for their abilities to inhibit fungal growth and ochratoxin A production by A. ochraceus. The yeasts Pichia anomala, Pichia kluyveri, and Hanseniaspora uvarum reduced the growth of A. ochraceus (over 60%) and prevented the biosynthesis of OTA on MEA medium [21]. Piotrowska and Zakowska assayed lactic acid bacteria for the elimination of OTA and found that the largest decrease in OTA was caused by Lactobacillus acidophilus (70.5%) and Lactobacillus rhamnosus (87.5%) [30]. Studies involving fungi for mycotoxin elimination are scarce in the literature. Barberis et al. studied the effects of fungi isolated from soil on OTA production by A. carbonarius and identified several genera capable of inhibiting OTA production, such as Aspergillus, Trichoderma, Cladosporium, Acremonium, and Geotrichum [31].

This paper aimed to evaluate the growth inhibition potential of eight fungal strains isolated from green coffee beans against the OTA producers A. carbonarius, A. niger, A. westerdijkiae, and A. ochraceus, and also assay their potential to decrease the OTA production by Aspergillus species.

Material and methods

Biological material

Microorganisms from stored green coffee bean samples were isolated by direct plating of four beans in sabouraud (SDA) medium and incubated at 28 °C for 7 days. Eight fungal colonies were collected, purified, and identified by the genus using classical methodology (micro and macromorphology) as Rhizopus spp. (2 isolates), Lichtheimia sp. (1 isolate), and Aspergillus spp. (5 isolates) [32, 33]. All isolates were deposited in the microbial culture collection of LabMicro (Laboratory of Microbiology and Molecular Biology, Federal University of Paraná, Brazil).

The reference strain of OTA-producing Aspergillus ochraceus, INCQS 40013 (ATCC 22947), was obtained from the National Institute of Quality Control in Health (INCQS) of the Oswaldo Cruz Foundation, Rio de Janeiro, Brazil, and the strains of OTA-producing Aspergillus carbonarius 187, Aspergillus niger 1, and Aspergillus westerdijkiae 91 were provided by the Laboratory of Molecular Biology of the State University of Londrina, Londrina, Brazil. The OTA production of standard Aspergillus strains was evaluated by high-performance liquid chromatography (HPLC) (Table 1).

Table 1.

Production of OTA by standard strains of Aspergillus used in inhibition tests

Strain	OTA production, μg/L	Source
A. westerdijkiae 91 A. niger 1 A. carbonarius 187	1.45 0.40 12.30	Molecular Biology Laboratory of UEL, State University of Londrina, Londrina, Paraná, Brazil
A. ochraceus 40013 (ATCC 22947)	1.84	National Institute of Quality Control in Health (INCQS) of the Oswaldo Cruz Foundation

Identification of fungi isolated from coffee beans

DNA extraction was performed according to Badali et al., and fragments of the ITS region (rDNA) were amplified using specific primers and sequenced [34]. In some instances, partial β-tubulin gene was sequenced as well. Primers ITS1 (5′-TCCGTAGGTGAACCTGCGG-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′) were used to amplify the ITS region, and Bt2a (5′-GGTAACCAAATCGGTGCTGCTTTC-3′) and Bt2b (5′-ACCCTCAGTGTAGTGACCCTTGGC-3′) were used to amplify part of the β-tubulin gene [35, 36].

ITS region and partial β-tubulin gene DNA amplification reactions were both performed in a volume of 25 μL and contained 1× buffer, 2.5 mM MgCl₂, 0.30 mM dNTPs, 0.03 U/μL of Taq DNA polymerase (Applied Biosystems), 0.5 μM each primer for ITS reaction, and 0.2 μM each primer for β-tubulin reaction. The PCR reaction from ITS region was carried out in an automatic thermocycler using the following conditions: initial denaturation at 94 °C for 2 min; 30 cycles at 94 °C for 35 s, 52 °C for 30 s, and 72 °C for 1 min; and final extension at 72 °C for 7 min. The PCR reaction conditions from β-tubulin were initial denaturation at 94 °C for 5 min; 35 cycles at 94 °C for 50 s, 60 °C for 50 s, and 72 °C for 1 min; and final extension at 72 °C for 7 min. The fragments were sequenced on an ABI 3500 automated sequencer (Applied Biosystems), and the sequences were aligned and edited with the Staden program version 1.6, aligned with the MEGA 7 program and compared with other sequences in the NCBI database by the BLAST program. The parameters used to analyze the sequence in the BLAST program were identify and cover = 100%.

Growth inhibition

All fungi were grown on sabouraud (SDA) medium for 7 days at 28 ± 0.5 °C. Spore concentrations of fungi isolated from coffee beans were estimated by microscopy using a counting chamber (Neubauer), and the suspensions were diluted to final concentrations of 10⁶ spores/mL. The spore suspensions (0.1 mL) were spread on the surface of SDA medium, in Petri dishes, with water activity (a_w) adjusted to 0.980 with glycerol, for better fungi growth [37]. A 6-mm plug of culture medium was removed from the plate center, and another 6-mm plug containing mycelium of an OTA-producing fungus replaced it. Plates were incubated at 28 ± 0.5 °C for 7 days, and the experiment was performed in 4 replicates. Each OTA-producing fungi was grown on SDA medium at the same conditions for negative controls.

Colony diameters were measured after 7 days of growth, and growth inhibition was assessed using the following formula: % Inhibition = ((Dc − Ds)/Dc) × 100, where Dc is the diameter of the control and Ds is the diameter of the sample [38].

Inhibition of OTA production

After 7 days of incubation, OTA was determined as proposed by Barberis et al. with some modifications [31]. From the plates used for growth inhibition tests, 3 agar plugs (6 mm in diameter) were removed at a distance of 1 cm each from the center of the colony (inoculation point). OTA was extracted by adding 1.5 mL of methanol; the mixture was allowed to stand for 1 h in the dark and centrifuged at 1200×g for 5 min. The solutions were filtered through glass wool, evaporated to dryness, re-dissolved in a mobile phase (acetonitrile/water/acetic acid, 50:48:2), filtered through a 0.22 μm syringe-driven filter unit (Millex, Millipore, Bedford, MA), and injected into the HPLC system. To determine the reduction of OTA production, we used the above formula.

OTA detection

HPLC analysis was performed according to Bragulat et al., with adaptations [39]. Detection of OTA was performed by using an HPLC Varian system equipped with a model 410 (50492) autosampler and a model 363 (00956) fluorescence detector. Separation was performed on a Varian C18 column, 250 × 4.6 mm, 5 μm, under isocratic conditions. The mobile phase (acetonitrile/water/acetic acid, 50:48:2) was pumped at a speed of 1.0 mL/min; the excitation and emission wavelengths were set to 330 nm and 470 nm, respectively. The injection volume was 50 μL, and the retention time was around 8 min. Standard curves were constructed with 0.1 to 15 μg/L of OTA (Sigma-Aldrich, St. Louis, MO), and the mean correlation coefficient was 0.999774. The detection limit of this method was 0.5 ng/L, and the limit of quantification was 0.1 μg/L. The toxin was quantified by correlating peak areas of sample extracts and those of the standard curves.

Statistical analysis

Analysis of variance (ANOVA) for repeated measures [40] was employed in order to access the differences in growth and OTA inhibition, followed by the Tukey’s method for multiple comparisons procedures (MCP) [41] when necessary. Residual analysis was applied to verify model assumptions. All computations were performed in the R statistical computing environment [42].

Results and discussion

Fungi identification

Microorganisms capable of reducing the growth of OTA producers arise as an alternative for the control of OTA in food products, as well as the mycotoxin production. With this purpose, the fungal isolates from coffee beans were evaluated according to their capacity to reduce the growth of and production of OTA by Aspergillus species.

The DNAs from the 8 isolates selected were sequenced and deposited in GenBank (Table 2).

Table 2.

Identification of the strains used and their GenBank accession numbers

Code	Identification	GenBank accession No.	Sequenced gene
C113	Rhizopus oryzae	KP784371	ITS
C183	Rhizopus oryzae	KP784372	ITS
C118	Lichtheimia ramosa	KP784373	ITS
C107	Aspergillus westerdijkiae	KJ599602	ITS
C187	Aspergillus niger	KJ599619	ß-tubulin
C122	Aspergillus tamarii	KP784375	ITS
C143	Aspergillus fumigatus	KP784370	ß-tubulin
C176	Aspergillus sp.	KP784374	ITS

These species have been previously described as regular coffee beans contaminants [43–46]. Alvindia et al. isolated fungi epiphytes and endophytes from coffee beans before and after surface disinfection and they found the genus Cladosporium, Rhizopus, Fusarium, Acremonium, Nigrospora, Aspergillus, and Penicillium after surface disinfection, while the genera Aspergillus and Penicillium epiphytes were found [47].

Growth inhibition

All fungi tested inhibited growth of A. ochraceus, A. westerdijkiae, A. carbonarius, and A. niger (Fig. 1), showing significant growth inhibition, when compared with the control treatment (critical p value < 0.025).

Fig. 1.

Percentages of growth inhibition of the four Aspergillus species by the fungi isolated from coffee beans

Strain L. ramosa C118 showed the lowest growth inhibition potential (18.3% A. ochraceus, 46.32% A. westerdijkiae, 67.9% A. carbonarius, 63.95% A. niger). The other fungal strains had growth inhibition potentials exceeding 70% against all tested species of Aspergillus. Strain A. niger C187 completely inhibited the growth of A. westerdijkiae.

The use of fungi to control the growth of mycotoxin-producing fungi is barely described in the literature, while yeasts and bacteria have been assayed by most authors [7, 21]. Zhu et al. assessed growth inhibition of A. ochraceus and A. carbonarius by yeasts, and found that strains Metschnikowia aff. Fructicola M179, Pichia kluyveri M117, Candida zemplinina M3, and Saccharomyces cerevisiae M114 and C297 exhibited significant inhibitory effects on both fungi tested [48]. The inhibition rates ranged from 26.45 to 100% for A. carbonarius and 79.56 to 100% for A. ochraceus. In this study, the inhibition percentages were different from those reported by Zhu et al., 67.9 to 92.57% for A. carbonarius and 18.3 to 96.59% for A. ochraceus [48]. Cubaiu et al. reported that strains of S. cerevisiae were evaluated for inhibition of A. carbonarius and A. ochraceus, and all yeast strains displayed an ability to inhibit fungal growth at levels of inhibition of up to 65% [49].

Piotrowska et al. assessed growth inhibition of A. westerdijkiae by lactic acid bacteria (Lactobacillus) and S. cerevisiae [50]. Growth of A. westerdijkiae was completely inhibited by S. cerevisiae, but it was more resistant to lactic acid bacteria. In this study, the percentage of inhibition of A. westerdijkiae ranged from 46.36% for L. ramosa to 100% for A. niger. Kogkaki et al. evaluated the interaction of A. carbonarius OTA producers with grape-associated fungal strains, reporting a slight inhibitory effect shown by Aspergillus ibericus on growth rates of two strains of A. carbonarius [51]. However, in general, fungi rarely have considerable effects on the growth of A. carbonarius.

Unlike the data reported by Kogkaki et al., in the present study, all fungi tested inhibited growth of the four Aspergillus species producing OTA, with the lowest percentage of inhibition for A. carbonarius (67.9%) shown by L. ramosa C118 [51].

The possible inhibition mechanism responsible for the Aspergillus species inhibition is the nutrient and space competition. It is a biocontrol mechanism, where the control agent uses the nutritional resources before the pathogen, being a great advantage when the fungi have a fast growth covering a large area [52]. Those findings enlighten the importance of searching new fungi strains, isolated from different plants, which can be potential controllers of mycotoxin producers.

Inhibition of OTA production

All fungi tested were able to inhibit OTA production by Aspergillus species. Strains L. ramosa C118, A. niger C187, A. tamari C122, and A. fumigatus C143 had the highest effects on fungi belonging to the section Circumdati (A. ochraceus and A. westerdijkiae). While strains A. niger C187 and A. fumigatus C143 showed higher potentials for reducing OTA production in the fungi belonging to the section Nigri (A. niger and A. carbonarius) (Fig. 2).

Fig. 2.

Percentages of OTA inhibition in the four Aspergillus species by the fungi isolated from coffee beans

Barberis et al. evaluated the effects of soil fungi on OTA production by A. carbonarius and reported that all A. niger strains tested completely inhibited the production of OTA [31]. In our study, the non-toxigenic strain A. niger C187 completely inhibited the production of OTA by the toxigenic strain of A. niger and provided significant inhibition of OTA production by all other Aspergillus species compared with the control. However, 100% inhibition of OTA production by A. carbonarius was not achieved. Abrunhosa et al. evaluated the OTA biodegradation potential of fungi isolated from grapes and showed that A. niger 27 aggregate strains degraded more than 80% of the OTA [53]. The authors concluded that the OTA degradation activity in these strains could be due to carboxypeptidase. Bejaoui et al. tested black Aspergillus species (A. carbonarius, A. niger aggregate, and A. japonicus) and concluded that A. niger could be interesting for OTA detoxification processes [54]. Abrunhosa and Venâncio isolated and purified the A. niger enzyme responsible for the hydrolysis of OTA [55]. According to the authors, the results suggested that the OTA hydrolytic enzyme is a metalloenzyme similar to carboxypeptidase A, corroborating some studies that describe the OTA hydrolytic activity of carboxypeptidase A55 [56, 57].

Two strains of Rhizopus oryzae were evaluated in our study and showed OTA inhibition rates above 78%. R. oryzae C113 displayed the following percentages of inhibition: 78.41% for A. westerdijkiae and A. ochraceus, 90.35% for A. carbonarius, and 94.84% for A. niger. C183 showed the following percentages of inhibition: 97.47% for A. westerdijkiae, 97.56% for A. ochraceus, 82.67% for A. carbonarius, and 89.61% for A. niger. For both strains, strong statistical evidence was obtained for the inhibition of OTA production by A. carbonarius, compared with the control treatment. Varga et al. assessed the potential degradation of mycotoxins by isolates of Rhizopus species [58]. Ochratoxin A was successfully degraded by Rhizopus stolonifer, R. microsporus, R. homothallicus, and R. oryzae isolates, and the Rhizopus isolates were able to degrade more than 95% of ochratoxin A within 16 days. The kinetics of OTA degradation by Rhizopus was compared with that of A. niger. The authors highlighted the similarities between the kinetics of OTA degradation by A. niger and Rhizopus and based on the product OTA degradation produced by Rhizopus suggested that carboxypeptidase A activity might be responsible for OTA decomposition by these Rhizopus isolates, similar to A. niger. According to Valero et al., other fungi have OTA degradation capacities due to carboxypeptidase A activity, such as Alternaria, Cladosporium, and Trichoderma [59].

Barberis et al. tested Absidia sp., which inhibited about 90% of OTA production by A. carbonarius [31]. In this work, strain L. ramosa C118 (formerly Absidia idahoensis var. thermophila) [60] inhibited by 100% the OTA production by A. ochraceus. Inhibition of OTA production by A. carbonarius (60.03%) was lower than that in the study by Barberis et al. [31]. Abrunhosa et al. tested fungi from the Aspergillus sections Flavi and Circumdati [53]. One strain of A. ochraceus (section Circumdati) inhibited OTA production by 95 to 100%, while the strain representing the section Circumdati in our study, A. westerdijkiae C107, reduced the production of OTA by 43.45% in A. carbonarius, 76.89% in A. niger, 94.96% in A. westerdijkiae, and 97.97% in A. ochraceus. Regarding strain A. flavus C176, representative of the section Flavi, there was a 100% reduction in the production of OTA by A. ochraceus. In a previous study, Abrunhosa et al. demonstrated that two isolates of A. flavus (section Flavi) degraded 50 to 80% of OTA [53].

Strain R. oryzae C113 had the lowest reduction potential for the fungi of the section Circumdati (78.41% for A. westerdijkiae and A. ochraceus). Regarding the fungi of the section Nigri, strain A. westerdijkiae C107 showed a lower OTA-reducing potential (76.89% for A. niger and 43.45% for A. carbonarius). Valero et al. reported three factors involved in OTA reduction by fungi: (1) limitation of fungal growth, which generally leads to a reduction in the production of OTA; (2) consumption by antagonistic fungi-specific nutrients required to synthesize OTA; (3) OTA degradation by other fungi [59].

From all these findings reported in the literature, we observed a reduction of growth in OTA-producing species, which may have been caused by space competition. This inhibition mechanism may also even excrete substances diffusing towards other fungi and blocking their growth and synthesis of OTA.

Conclusions

OTA is a mycotoxin found in several agricultural products, and it raises food safety problems. The use of biological methods (yeasts, bacteria, and fungi) is considered by many investigators as one of the most promising ways to control mycotoxin production. Non-toxigenic strain A. niger C187 showed the best results in terms of inhibition of growth and OTA production. All tested fungal strains caused inhibition of growth and OTA production by Aspergillus species.

Our results indicate that selected fungal strains may be efficiently developed as agents to control OTA in food products. These data can also encourage new research on fungi, aiming at their use as potential agents reducing mycotoxin contamination.