Isolation, Molecular Identification, and Mycotoxin Production of Aspergillus Species Isolated from the Rhizosphere of Sugarcane in the South of Iran

Knowledge of the genetic diversity detected among fungal species belonging to the genus Aspergillus is of key importance for explaining their important ecological role in the environment and agriculture. The current study aimed to identify Aspergillus species occurring in the rhizosphere of sugarcane in the South of Iran, and to investigate their mycotoxin profiles. One-hundred and twenty-five Aspergillus strains were isolated from the soil of eight major sugarcane-producing sites, and were molecularly identified using sequences of partial β-tubulin (benA) and partial calmodulin (CaM) genes. Our molecular and phylogenetic results showed that around 70% of strains belonged to the Aspergillus section Nigri, and around 25% of species belonged to the Aspergillus section Terrei. Species belonging to both sections are able to produce different mycotoxins. The production of mycotoxins was measured for each species, according to their known mycotoxin profile: patulin (PAT) and sterigmatocystin (STG) for Aspergillus terreus; ochratoxin A (OTA) and fumonisins for Aspergillus welwitschiae; and OTA alone for Aspergillus tubingensis. The data showed that the production of OTA was detected in only 4 out of 10 strains of A. welwitschiae, while none of the A. tubingensis strains analyzed produced the mycotoxin. Fumonisins were produced by 8 out of 10 strains of A. welwitschiae. Finally, none of the 23 strains of A. terreus produced STG, while 13 of them produced PAT. The occurrence of such mycotoxigenic plant pathogens among the fungal community occurring in soil of sugarcane fields may represent a significant source of inoculum for the possible colonization of sugarcane plants, since the early stages of plant growth, due to the mycotoxin production capability, could have worrisome implications in terms of both the safety and loss of products at harvest.


Species Identification
To identify Aspergillus species occurring in the sugarcane rhizosphere in the South of Iran, and to investigate their mycotoxin profiles, 64 samples were collected from eight major sugarcane-producing sites (Table 1). In total, 125 strains of Aspergillus were isolated and, based on morphological characteristics, were primarily divided into five sections, including Circumdati, Flavipedes, Nigri, Terrei, and Usti. However, the majority of strains were included in sections Nigri and Terrei. For precise species identification, all isolates were subjected to a DNA-based analysis of partial β-tubulin (benA) and partial calmodulin (CaM) genes. The concatenated sequences of benA and CaM, generated in this study, were aligned against the sequences of 13 reference Aspergillus species available at GenBank. Based on our phylogenetic analysis (Figure 1), nine species could be identified: Aspergillus calidoustus (1 strain), Aspergillus japonicus (1), Aspergillus luchuensis (1), A. niger (2), Aspergillus ochraceus (1), Aspergillus templicola (5), A. terreus (29), Aspergillus tubingensis (60), and Aspergillus welwitschiae (25).  Tappeh  8  32  21  24  30  24  Total  64  125  23 48 20 9 Among the 125 Aspergillus strains, A. tubingensis occurred at a rate of 48%, A. welwitschiae at 20%, and A. terreus at 23%, together representing 89% of the Aspergillus population ( Figure 2). In most of the Iranian regions, the percentage of Aspergillus section Nigri strains was higher than that of Aspergillus section Terrei. The highest occurrence of A. tubingensis was observed in Amir Kabir, Imam Khomeini, Debal Khazaei, Mirza K. Khan, and Salman Farsi, with a 75%, 50%, 73%, 82%, and 81% occurrence, respectively. In only one case (from Dehkhoda), A. terreus displayed the highest occurrence (80%). On the other hand, A. welwitschiae was the most frequent species in Karun and Haft Tappeh regions (42% and 30%, respectively).

Mycotoxin Production
Mycotoxin production of a subset of strains, representative of the different sites of isolation, was tested for potential toxigenic species: A. welwitschiae (10 strains), A. tubingensis (15 strains), and A. terreus (23 strains) (see Table 2).
No production of ochratoxin A (OTA) was observed among the 10 tested strains of A. welwitschiae, when grown on CY20S; however, on YES medium, four out of the 10 strains produced OTA, in the range of 32 and 142 µg/g. None of the 15 A. tubingensis strains, grown on both CY20S and YES media, were able to produce OTA. Fumonisin B2 (FB2) was produced by eight out of 10 A. welwitschiae strains tested (MTN108, MTN46, MTN11, MTN17, MTN105, MTN78, MTN103, and MTN36) in the range of 0.1 and 4.4 µg/g, when grown on CY20S. Furthermore, 13 out of the 23 tested strains of A. terreus, on YES, were able to produce patulin (PAT) in the range 0.2 and 523.56 µg/g. On the other hand, no production of PAT was recorded by the same strains when grown on CYA. Finally, sterigmatocystin (STG) was not produced by any strain of A. terreus, on both YES and CY20S (Table 2). A. = Aspergillus; n.t. = not tested; * = on YES medium; LOD FB2 = 0.03 µg/g; LOD OTA = 0.015 µg/g; LOD PAT = 0.005 µg/g based on a signal to noise ratio of 3:1.

Discussion
Agricultural crops grow in soils, which usually host enormous types and numbers of micro-organisms, that, from one side, are responsible for soil health by nutrient cycling, while from the other side, can opportunistically infect the plants [24][25][26]. Among micro-organisms, fungal species of the genus Aspergillus are soilborne micro-organisms, and their predominance in rhizospheric and non-rhizospheric soils has been extensively studied [27]. In this study, we focused our analyses on the isolation and identification of Aspergillus species, since they are well-known mycotoxin-producers and their occurrence in sugarcane soils can be the cause of a subsequent contamination of sugarcane plants that grow in such contaminated soils [28].
The occurrence of Aspergillus species was evaluated in the rhizosphere of eight sugarcane sites in the South of Iran (see Figure 3 and Table 3), where the sugarcane is the main agro-food product. The number of Aspergillus strains isolated in this study varied from region to region: Imam Khomeini and Haft Tappeh sites showed the highest and lowest number of isolates, respectively. These differences in the distribution of strains among the regions might be related to significant specific environmental and biological features of each site, such as the microbiome profile [29,30], fungal antagonist occurrence [31], and soil moisture and temperature [32]. However, inadequate knowledge on the sites sampled prevents us from drawing further conclusions.  The presence of toxigenic Aspergillus species in the soil associated with sugarcane indicates that a contamination of sugarcane plants in the field and its by-products post-harvest can occur. The heat treatment with which sugarcane juice is processed to obtain molasses and non-refined sugar, leads to the removal of fungal contaminants at this stage. However, since we proved a high presence of toxigenic Aspergillus species in the rhizosphere of sugarcane plants, the contamination of the early stages of plant development could be high. Therefore, in the field, environmental conditions suitable for mycotoxin production in planta by the Aspergillus species can easily occur and also lead to the contamination of final products. As a consequence, this contamination can cause a high toxigenic risk for consumers, but can also interfere with sugarcane industrial processing.
Our molecular and phylogenetic results showed that around 70% of strains belong to the Aspergillus section Nigri, and around 25% of species belong to the Aspergillus section Terrei. Although no morphological differences were observed among strains assigned to A. terreus, a phylogenetic tree divided them into two clades. Interestingly, MTN64 and MTN71, which were morphologically identified as A. terreus, were closely related to A. hortai (CBS124230). A correct identification of Aspergillus species is a key aspect, since many species produce mycotoxins and each species has its own mycotoxin profile. Therefore, a precise risk assessment is strongly linked to the use of advanced diagnostic tools.
As reported above, the species identified in this survey belonged to two main Aspergillus sections-Terrei and Nigri-characterized by distinct mycotoxin profiles: PAT and STG for section Terrei and OTA and FB2 for section Nigri. The in vitro mycotoxin production by different species tested was analyzed according to their known mycotoxin profile: PAT and STG for A. terreus and OTA and FB2 for A. welwitschiae. However, the production of OTA was also tested for A. tubingensis because its ability to produce mycotoxin has been inconsistently reported. Susca et al. [33] demonstrated the lack of some genes in the OTA cluster, most likely related to their inability to produce OTA, while Medina et al. [34] reported OTA production for few strains of A. tubingensis. However, our data confirmed the inability of A. tubingensis to produce OTA, since none of the strains tested could produce any trace of the mycotoxin. The fact that this species is the most frequently detected in our study is comforting since it reduces the risk related to the occurrence of mycotoxigenic Aspergillus species.
On the contrary, the presence of A. welwitschiae strains that produce both OTA and FB2 is worrisome, since both of these mycotoxins are considered by the International Agency for Cancer Research (IARC) group 2b, meaning potentially carcinogenic. Only a minority of strains analyzed were able to produce only OTA, while the majority of A. welwitschiae strains (80%) also produced FB2, in agreement with previous reports [33,35]. Some authors have demonstrated that OTA production in A. niger is affected by temperature [36], while in A. welwitschiae, it is affected by water activity and the medium [37]. Our data confirm the high influence of media in the in vitro production of OTA by A. welwitschiae. No strain produced the mycotoxin on CY20S, while 40% of strains produced it if grown on YES. In addition, since the temperature was proved to affect OTA production, environmental and geographical variations clearly influence the potential of OTA production by A. welwitschiae in the field of a given area.
To the best of our knowledge, in Iran, there is currently a lack of information on the mycotoxin contamination of sugarcane and related by-products. Here, we reported the potential contamination by FB2, OTA, and PAT of sugarcane plants. Additionally, our data suggest the need for further investigations for developing a better understanding of fungal contamination and related mycotoxins in planta, in order for proper management protocols to be adopted and to minimize the risk of contamination of this prestigious food. Finally, the high variability of the Aspergillus species profile in the soils of different regions, the high influence of temperature on their ability to colonize soils, and the rapidity with which the population structure of the genus Aspergillus can change in the soil, together suggest that, by the time a study like this one is published, the data presented in it are historical rather than a profile of an existing Aspergillus population. However, genetic diversity and flux in populations of the genus Aspergillus may allow us, in Blake's words, 'to see the world in a grain of sand, and eternity in an hour'.

Conclusions
Our study showed the presence of toxigenic Aspergillus species in the rhizosphere of sugarcane. Such occurrence could represent a toxigenic risk for the crop culture in the following year, due to the usual practices of harvesting in Khuzestan province. By testing the in vitro toxin production of fungal species isolated, this study showed their capability of producing OTA, PAT, and FB2, and therefore, this might be the cause of possible contamination of sugarcane plants. These data demonstrate the need for further investigations aimed at assessing the possible risk for human and animal health, due to the consumption of sugarcane products contaminated by mycotoxins.

Samples
Sugarcane fields were sampled from different regions in the South of Iran, depending on the history of cultivation. The sampling was conducted in different stages of sugarcane development, preferably in cold seasons (autumn and winter), with at least 20 well-distributed samples per field, including sugarcane root with the surrounding rhizosphere [38]. Samples were placed in suitable paper envelopes and transferred to the laboratory. After transferring samples to the laboratory, the rhizosphere soil around the root was carefully separated from the root. Samples were divided into soil and root subunits. Soil samples were kept in a ventilated condition and preferably at a temperature of 15-20 • C for 12 to 24 h (depending on the moisture content of the soil) and then transferred to the refrigerator (2-5 • C) for further examinations [39].

Fungal Isolation and Morphological Identification
The soil dilution plating technique modified by Johnson et al. [40] and Warcup [41] was used to isolate Aspergillus species from the soil. Soil samples were crushed by a 2 mm sieve. About 10 gram of the crushed soil was poured into a graduated cylinder. One hundred milliliters of sterile water was added to the cylinder. The suspension was transferred to the 250 mL Erlenmeyer flask and mixed for 30 minutes. One milliliter of suspension was transferred to test tubes containing 9 mL of 0.12% water agar medium. The 1:100, 1:1000, and 1:5000 dilutions were prepared. One milliliter of 1:1000 and 1:5000 dilutions was transferred to Petri dishes and then 10-12 mL of PCA (Potato Carrot Agar) medium [42] containing melted and cooled 1.5% agar and 1% oxgall was added. In total, 200 ppm of penicillin and streptomycin antibiotics were also added to avoid bacterial contamination. The plates were moved gently so that the suspension became uniform. For each dilution, three plates were considered. The plates were incubated in the dark at 17 • C for 10 days. Pure cultures were obtained by single spore isolations. Some representative strains-48 out of the 125 included in the study-were deposited in ITEM (Agri-Food Toxigenic Fungi Culture Collection, ISPA-CNR Bari, Italy, http://server.ispa.cnr.it/ITEM/Collection/) and IRAN (Iranian Fungal Culture Collection, Iranian Research Institute of Plant Protection, Tehran, Iran, http://gcm.wfcc.info/cc/iran).
For morphological characterization, pure cultures were grown on Malt Extract Agar (MEA), and were incubated at 25 • C for 7 days. Macroscopic traits, such as the colony appearance, color, pigmentation, and growth rate, were recorded according to standard protocols [27,[43][44][45].

Molecular Identification
For molecular identification, isolates were grown on Potato Dextrose Agar (PDA). Direct PCR from fungal mycelia was done using Phire Plant Direct PCR Master Mix (F-160L, Thermo Fisher Scientific, Waltham, USA). For direct PCR, a small piece of growing mycelium from a 3-4-day-old colony was swiped with a sterile pipette tip and re-suspended in 20 µL of 1× Phire Plant Direct PCR Master Mix prior to PCR. Mycelia were crushed with a 100 µL pipette tip and vortexed briefly, and then collected down at the tube bottom with a spin in centrifuge. The supernatant (1 µL) was used as a template for a 20 µL PCR reaction. All amplifications were performed according to standard Direct PCR protocols (Tables 1 and 2) by using a GeneAmp 9700 thermalcycler (Applied Biosystems, Foster City, CA, USA). Partial β-tubulin (benA) and partial calmodulin (CaM) genes were amplified using primers described in the literature: primers BT2a and BT2b, and CL1 and CL2A [46,47]. After amplification, amplicons were purified with the enzymatic mixture EXO/SAP (Exonuclease I, E. coli/Shrimp Alkaline Phosphatase) and used as a template for bidirectional DNA sequencing. Sequencing was performed with the BigDye v3.1 terminator kit (Applied Biosystems, Foster City, CA, USA), following the manufacturer's instruction, and analyzed on an ABI 3730 XL Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). Alignment of the two strands was performed using the software package BioNumerics 5.1 (Applied Maths, Sint-Martens-Latem, Belgium)), with manual adjustments, where necessary. Sequences of benA and CaM genes, generated in this study, were deposited in GenBank with the following accession numbers: from LR693748 to LR693997.

Sequence Data Analysis
To obtain a previous species identification in order to select reference sequences for Aspergillus species to be used in the phylogenetic analysis, sequences of benA and CaM were searched for on the GenBank database using the Basic Local Alignment Tool (BLASTN, NCBI BLAST website). The sequences of 125 strains were used to perform phylogenetic analysis. A set of 16 reference sequences for Aspergillus species, identified by BAST analysis, were downloaded from GenBank and used for phylogenetic analysis. All sequences were aligned using the MUSCLE algorithm [48] with MEGA7 software ver. 7.0.14 [49]. The evolutionary history was inferred by using the Maximum Likelihood method based on the Tamura-Nei model in MEGA7 software [50]. To evaluate the support for inferred topologies, the percentage of trees in which the associated taxa clustered together was calculated by bootstrapping with 1000 replicates [51]. The percentage of trees in which the associated taxa clustered together is shown next to the branches. Initial trees for the heuristic search were obtained automatically by applying Neighbor-Join and Bio NJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach, and then selecting the topology with a superior log likelihood value. The tree was drawn to scale, with branch lengths measured in the number of substitutions per site. The analysis involved 138 nucleotide sequences. Codon positions included were 1st + 2nd + 3rd + Noncoding. All positions containing gaps and missing data were eliminated.
OTA stock solution was prepared by dissolving the solid commercial toxin in methanol (1 mg/mL). The exact concentration of OTA was determined according to the Association of Official Analytical Chemists (AOAC) Official Method 2001.01 [52]. Appropriate aliquots of the stock solution were brought to dryness and reconstituted with acetonitrile/water/acetic acid (99:99:2, v/v/v) to obtain standard solutions of OTA in the range 0.05-0.10 µg/mL. Fumonisin calibration solutions were prepared by diluting the samples with acetonitrile/water (1:1, v/v), to obtain solutions with a concentration in a range of 0.01-1.00 µg/mL for FB1 and FB2. Standard solutions were stored at −20 • C and warmed to room temperature prior to use.
PAT stock solution was prepared by dissolving the solid commercial toxin in ethanol (10 µg/mL). The exact concentration of standard solution was determined as reported by the European Committee for Standardization [53]. Aliquots of the stock solution were transferred to 4 mL amber glass vials and evaporated to dryness under a stream of nitrogen at 50 • C. The residue was dissolved in water/acetonitrile (85:15, v/v) to obtain a desired final concentration of 80 to 800 ng/mL.
Mycotoxin stock solution of STG (1 mg/mL) in acetonitrile was transferred to 4 mL amber silanized glass vials and evaporated to dryness under a stream of nitrogen at 50 • C. The residue was dissolved in water/methanol (60:40, v/v) to obtain calibrant standard solutions from 0.5 to 5.0 µg/mL. Standard solutions were stored at -20 • C and warmed to room temperature before use. Strains were grown for 14 days on YES (OTA, PAT, and STG), CYA (PAT and STG), and CY20S (FB2) media. One gram of agar was extracted with 5 mL of extraction solution on an orbital shaker for 60 minutes: OTA and PAT acetonitrile/methanol/water (90: 90: 80, v/v/v), FB2 methanol/water (70:30, v/v), and STG methanol/water (80: 20, v/v). Two milliliters of extracts were evaporated to dryness under a stream of nitrogen at 50 • C.
The quantification of mycotoxins was determined differently. OTA residue was dissolved with 1 mL of acetonitrile/water/glacial acetic acid (99:99:2, v/v/v) and filtered using RC through a 0.20 µm regenerated cellulose filter and determined by HPLC/FLD. OTA quantification was performed according to Susca et al. [35]. The quantification limit (LOQ) was 0.05 µg/g based on a signal to noise ratio of 10:1. FB2 residue was dissolved with 1 mL of the extract acetonitrile/water (30:70, v/v), filtered using RC 0.20 µm filters (Phenomenex, Torrance, CA, USA), and determined by HPLC/FLD previously derivatized with o-phtaldialdehyde (OPA). FB2 quantification was performed according to Susca et al. [35]. The LOQ of the method was 0.1 µg/g, based on a signal to noise ratio of 10:1.
PAT residue was dissolved with 1 mL of water/acetonitrile (85:15, v/v) and filtered using RC through a 0.20 µm regenerated cellulose filter and determined by HPLC/DAD. A total of 50 µL of extract was injected into HPLC apparatus Agilent 1260 Series. The analytical column was a Luna-C18 (4.6 × 150 mm, 5 µm) (Phenomenex, Torrance, CA, USA) preceded by C18 guard column (4 × 3 mm, Phenomenex, Torrance, CA, USA). The temperature of the column was maintained at 30 • C. The mobile phase was a mixture of water as solvent A and acetonitrile as solvent B, eluted at a flow rate of 1 mL/min. A gradient elution was performed as follows: 5% B solvent that was linearly increased by 100% in 15 min. In these analytical conditions, the retention time of PAT was about 5 min. The diode array detector (DAD) was set at a wavelength of 276 nm. The LOQ was 0.015 µg/g, based on a signal to noise ratio of 10:1 [54].
STG residue was dissolved with 1 mL of water/acetonitrile (75:25, v/v) and filtered through RC 0.20 µm. A total of 50 µL of extract was injected into HPLC apparatus Agilent 1260 Series. The analysis of toxins was performed using the following analytical method [55]. The analytical column was a Luna-C18 (4.6 × 150 mm, 5 µm) (Phenomenex, Torrance, CA, USA) preceded by C18 guard column (4 × 3 mm, Phenomenex, Torrance, CA, USA). The mobile phase consisted of a mixture of acetonitrile/water (40:60, v/v) at a flow rate of 0.8 mL/min. The array detector (DAD) was set at wavelengths of 248 and 328 nm. The temperature of the column was maintained at 30 • C. In this analytical condition, the retention time of ST was about 6.5 min. ST was quantified by measuring peak areas at the retention time of ST standard solutions and comparing these areas with the relevant calibration curve. The LOQ was 0.012 µg/g, based on a signal to noise ratio of 10:1.6.