Fumonisin B1 Production by Fusarium Species and Mycotoxigenic Effect on Larval Zebrafish

Fumonisin B1 (FB1) is a common mycotoxin produced by Fusarium species particularly F. proliferatum and F. verticillioides. The toxin produced can cause adverse effects on humans and animals. The objectives of this study were to detect the production of FB1 based on the amplification of FUM1 gene, to quantify FB1 produced by the isolates using Ultra-fast Liquid Chromatography (UFLC) analysis, to examine the embryotoxicity effect of FB1 and to determine EC50 toward the larvae of zebrafish (Danio rerio). Fifty isolates of Fusarium species were isolated from different hosts throughout Malaysia. Successful amplification of the FUM1 gene showed the presence of this gene (800 bp) in the genome of 48 out of 50 isolates. The highest level of FB1 produced by F. proliferatum isolate B2433 was 6677.32 ppm meanwhile F. verticillioides isolate J1363 was 954.01 ppm. From the assessment of embryotoxicity test of FB1 on larvae of zebrafish, five concentrations of FB1 (0.43 ppm, 0.58 ppm, 0.72 ppm, 0.87 ppm and 1.00 ppm) were tested. Morphological changes of the FB1 exposed-larvae were observed at 24 to 168 hpf. The mortality rate and abnormality of zebrafish larvae were significantly increased at 144 hpf exposure. Meanwhile, the spontaneous tail coiling showed a significant difference. There were no significant differences in the heartbeat rate. As a conclusion, the presence of FUM1 in every isolate can be detected by FUM1 gene analysis and both of the species produced different concentrations of FB1. This is the first report of FB1 produced by Fusarium species gave a significant effect on zebrafish development.


IntroductIon
Mycotoxins contamination occurs frequently in the area with a hot and humid climate where it is the most suitable condition for the growth of the fungi. Factors that influence the contamination of mycotoxins in food and feedstuff are related to a favourable environmental condition such as unsterile storage area (Zain 2011). One of the most common mycotoxins is Fumonisin B 1 (FB 1 ) produced by Fusarium species. At least 10 different species of Fusarium can produce this toxin, however, Fusarium proliferatum and F. verticillioides are the most common fungi associated with crops (WHO/FAO 2001). Apart from crops, both species can also be isolated from varieties of plants and are spread by rain and other abiotic factors.
Production of FB 1 by Fusarium species is dependent on a biosynthetic gene cluster FUM, which consists of FUM1 to FUM19 where FUM1 is a gene that responsible for the production of FB 1 (Glenn et al. 2008). FB 1 is classified into Group 2B carcinogens by the International Agency for Research on Cancer (IARC). FB 1 is toxic and causing leukoencephalomalacia and it has been reported to be associated with human oesophageal cancer and birth defects (Marasas 2001). The occurrence of mycotoxin contamination has become a great concern because it is often associated with acute and chronic diseases (Logrieco et al. 2002). Besides, it also can compromise the safety of food and feed supplies in humans and animals.
As mycotoxins contamination in plant-based products become serious problems, it is important to conduct this study in order to detect and prevent human and animal exposure to toxic substances. To detect accurately the fumonisin production in F. proliferatum and F. verticillioides, a diagnostic method based on polymerase chain reaction (PCR) is the most suitable and rapid detection by using specific FUM1 gene pairs as it detects the genes responsible to produce fumonisin. Meanwhile in order to examine the effect of FB 1 towards human and animal health uses of zebrafish (Danio rerio) as a model organism is important as it has been reported that 70% of human genes have zebrafish orthologues (Howe et al. 2013). Since currently there are no reports on the exposure of FB 1 on zebrafish, from this study, it can provide more information on the effect of FB 1 thus can relate to human health and development. In addition, there were different types of mycotoxins that have been tested on zebrafish embryo such as ochratoxin, T-2 toxin and zeralenone. Wu et al. (2016) have reported that zebrafish embryo that were exposed with ochratoxin have abnormal heart looping and small heart chambers. The toxin also has disrupts the renal morphology of the embryonic zebrafish. Meanwhile, another mycotoxin exposure on zebrafish embryo was T-2 toxin where higher concentration of T-2 toxin can significantly increase the mortality rate and also caused the malformations on tails and cardiovascular defects (Yuan et al. 2014). Zeralenone exposure on zebrafish showing defects on eye development and also upward curvature of the body axis (Bakos et al. 2013).
The objectives of this study were: (i) to detect the production of FB 1 based on the amplification of FUM1 gene and to quantify FB 1 produced by the isolates using Ultra-fast Liquid Chromatography (UFLC) analysis, and (ii) to examine the embryotoxicity effect of FB 1 on the larvae of zebrafish (D. rerio) and (iii) to determine EC 50 of FB 1 towards zebrafish.

Fungal source and Preservation
Fifty isolates of Fusarium proliferatum and F. verticillioides were obtained from Mycology Laboratory, Department of Biology, Universiti Putra Malaysia and Plant Pathology Laboratory, School of Biological Sciences, Universiti Sains Malaysia. The isolates were previously isolated from different hosts and cultured on potato dextrose agar (PDA) and incubated at 27 ± 2°C for five days. All pure cultures were transferred onto Spezieller Nahrstoffarmer Agar (SNA) and sterile filter paper for preservation purpose. The culture stocks were kept in -20°C for further use. Species identification of the isolates was confirmed by using species-specific sequence analysis of ProF1/ProR1 and VertF1/VertR1 (Najihah et al. 2017).

dnA extraction and FUM1 Gene Amplification
The isolates were cultured on PDA and incubated at 27 ± 2°C for 7 days and genomic DNA was extracted by using Qiagen DNeasy® UltraClean® Microbial Kit according to the protocols provided by the manufacturer. FUM1 gene detection has been carried out by using primers FUMF1 (5'-GACACCTCCTTCTTCTC-3') and FUMR1 (5'-GTGCCGGTTCCGTGTGCTTC-3') (Proctor et al. 1999). Standard PCR master mix GoTaq® Flexi DNA Polymerase Promega Corporation for a total of 20 µL was consisted of green buffer, 2 mM dNTP, 25 mM MgCl 2 , 10 mM of each forward and reverse primer respectively, 1 unit Taq polymerase, nuclease-free water and 20 ng DNA. The amplifications were performed by using Biometra TProfessional Basic Thermocyclers. The PCR condition was performed as follows; initial denaturation at 95°C for 5 min, 40 cycles of denaturation at 95°C for 10 s, annealing at 60°C for 10 s, extension at 72°C for 30 s and final extension at 72°C for 5 min (Proctor et al. 1999). Gel electrophoresis was performed by using 1.5% agarose gel immersed in 1X Tris Borate-acid EDTA (TBE) buffer amended with FloroSafe DNA stain (1 st Base Asia). Approximately 5 µL for each DNA ladder 100 bp (Thermo Scientific) and PCR products were loaded and electrophoresed for 35 min at 90 V.

Inoculation and extraction of FB 1
All 50 isolates of F. proliferatum and F. verticillioides were cultured on PDA and incubated at 27°C ± 2°C for 7 days. The spores were dislodged by using 10 mL of sterile distilled water and filtered by using gauze. The final concentration of conidial suspension was adjusted to 1 × 10 5 conidia mL -1 (Nur Ain Izzati et al. 2008). Inoculums were prepared by adding 85 g of cornmeal into 100 mL of distilled water and autoclaved at 121°C for 20 min. Another 5 mL of sterile distilled water was added and re-autoclaved for 30 min at 121°C. About 1 mL of conidial suspension was added into the sterile cornmeal and shaken for three continuous days to allow even distribution meanwhile for control cornmeal; 1 mL of sterile distilled water was added. The cultures were done in triplicates and then incubated at 28°C ± 2°C for 28 days (Nur Ain Izzati & Nithiyaa 2015). After the incubation period, 25 g of colonised cornmeal was ground in 100 mL of methanol: water (3:1; v/v). The extract then was filtered through Whatman No. 1 filter paper. The extract was then dried using rotary evaporator at 60°C, re-dissolved in 1 mL methanol and kept in -20°C freezer before analysed (Shephard 1998).
The sample was quantified by derivatised the extract by adding 50 µL of the sample with 200 µL of OPA reagent. Quantification of FB 1 emitted by the Fusarium isolates was conducted using a Shimadzu (Kyoto, Japan) UFLC equipped with two LC-20AD pumps, a SIL-20A autosampler, a Shimadzu RF-10A XL fluorescence detector (at 335 nm excitation wavelength and 440 nm emission) and a Shimadzu SPD-20AV UV/VIS detector. A reversed-phase LC column, Thermo Scientific, Hypersil GOLD, C18, 250 mm × 4.6 mm i.d with a particle size of 5 µm was used. Approximately 20 µL of derivatised samples and standard were injected into the column using deionised water as a mobile phase A and 80% methanol: 0.1 M sodium dihydrogen phosphate (78:22/v:v) as a mobile phase B with flow rate of 0.8 mL min -1 (mobile phase A) and 0.2 mL min -1 (mobile phase B) for 10 min. Concentrations of FB 1 were determined by comparing the formed peak observed using a fluorescent detector with the standard curve (Shephard 1998).

Embryotoxicity Effect of FB 1 in Zebrafish Larvae
Adult wild type zebrafish (5-8 months old with size 3 cm-5 cm) were provided by Aquatic Laboratory, Department of Biology, Universiti Putra Malaysia. All the adult zebrafish were maintained in recirculating freshwater aquarium system at 25°C-27°C on a ratio of 14 h light: 10 h dark of light cycle to induce a reproductive cycle of the fishes (Matthews et al. 2002). At 4 h post-fertilisation (hpf), fertilised eggs that reached the gastrulation stage were chosen for FB 1 exposure. Standard FB 1 from Fusarium moniliforme (≥ 98%, Sigma-Aldrich, St Louis, MO, USA) was diluted to final concentration (0.43, 0.58, 0.72, 0.87 and 1.00 ppm) in embryonic media solution. Three replicates of 96 well plates were used. One embryo for each well (n = 48 embryos per exposure group) was exposed to FB 1 and control that contained embryonic media in a semi-static condition and incubated at 28 ± 2°C.
Percentage of mortality rate and morphological deformities were observed at 24, 48, 72, 97, 120, 144 and 168 hpf by using a dissecting microscope model Motic SMZ-161 (Motic Instrument, China). Tail coiling and heartbeat rate were counted within 1 min at 24 hpf and 48 hpf, respectively for every larva according to Organisation for Economic Co-operation and Development (OECD) guideline (OECD 2012), in which the heartbeat is visible and can be observed clearly at 48 hpf (Kalueff et al. 2013). Pericardial edema, yolk sac edema, non-inflated swim bladder and spinal curvature were determined using the guidelines of Scherz et al. (2008), Chakraborty et al. (2009), Hill et al. (2009) and Tsai et al. (2011, respectively. Effective concentration (EC 50 ) was calculated based on the concentrationresponse curve. The concentrations were converted to log value, and then a graph was plotted response against the log of the FB 1 concentration. The EC50 was converted by taking the anti-log. The statistical analyses were conducted using SPSS version 21.0. One-way analysis of variance (ANOVA) followed by Duncan test p < 0.05 was set as a criterion for statistical significance to determine the differences between all concentrations to control treatments. Data were represented as mean ± standard deviation.

FUM1 Gene detection
All 48 isolates showed a positive result of FUM1 gene where 800 bp of amplicons were amplified and clearly visible. Twenty-one isolates were F. verticillioides and 27 isolates were F. proliferatum. Two isolates of F. proliferatum, B68, and B92 were unable to amplify the FUM1 gene.

Quantification of FB 1 Production using uFlc Analysis
From the standard curve, concentrations of the samples can be calculated based on the gradient formula (y = 5287.2x + 31641) from standard curve and peak area of the chromatogram. Table 1 shows the concentration of FB 1 produced by F. verticillioides and F. proliferatum. Eighteen isolates of F. verticillioides produced FB 1 meanwhile three isolates, B106, B1371 and 7706 did not produce FB 1 . The highest concentration of FB 1 was produced by isolate J1363 with the concentration of 954.01 ppm meanwhile the lowest concentration was isolate 7696 with a concentration of 12.26 ppm. For F. proliferatum, 26 isolates produced FB 1 meanwhile three isolates B68, B92 and P202 gave no result. The highest FB 1 concentration was produced by isolate B2433 with concentration 6677.32 ppm meanwhile the lowest FB 1 concentration was produced by isolate N2215 with concentration 16.48 ppm. From these results, it proved that F. verticillioides and F. proliferatum isolates are capable of producing a high level of FB 1 . However, there were six isolates (three isolates from each species) that were not capable of producing FB 1 and no detection from UFLC analysis.
The effects of FB 1 exposure toward zebrafish larvae can be seen as early as 24 hpf where blood clotting was formed at 0.87 ppm and 1.00 ppm (Fig. 1). From the control larvae and concentration lower than 0.72 ppm, no significant morphological abnormality was observed. However, starting at 48 hpf, more abnormalities development can be seen from exposure at the lowest concentration of 0.43 ppm, the zebrafish showed spinal curvature and also pericardial edema was observed at 0.58 ppm and 0.72 ppm of FB 1 exposure. The severity of pericardial edema increases at 72 hpf onwards with exposure 0.72 ppm, 0.87 ppm and 1.00 ppm (Figs. 2 and 3).

Tail Coiling, Heartbeat and Mortality Rate of Larval Zebrafish
Tail coiling rate of zebrafish reduced significantly as the concentration of the FB 1 increased. Control zebrafish showed the highest rate of tail coiling which was 3.63 ± 1.58 per min and had a significant difference compared to other concentrations (Table 2). Meanwhile the tail-coiling rate of the highest FB 1 concentration, 0.87 ppm and 1.00 ppm had no significant difference.   Heartbeat rate that was measured at 48 hpf showed no significant difference at all different concentrations. A normal heartbeat rate of zebrafish is at 120-180 beats per min (bpm). As shown in Table 2, the range of heart rate of control and treated zebrafish was from 144.33 bpm-161.29 bpm, which was still under normal range. At 48 hpf, heart morphology only showed mild abnormalities and mild pericardial edema. Fig. 4 shows the mortality rate of zebrafish larvae that were exposed to FB 1 according to hpf. For the control treatment, the mortality rate at 24 hpf was 8.33% and there were no significant changes until 168 hpf. The mortality happened at control treatment were due to unfertilised eggs. For all FB 1 treatment, the mortality rate started showing significant changes at 144 hpf. Zebrafish larvae that were exposed to 0.43 ppm of FB 1 had 18.77% of the mortality rate at 120 hpf and increased significantly to 31.27% of mortality rate at 144 hpf. The highest concentration of FB 1 was 1.00 ppm and the mortality rate at 120 hpf was 25.03%, increased significantly to 54.17% at 144 hpf and rising to 70.87% at 168 hpf. To determine the mortality effect of FB 1 towards zebrafish, effective concentration (EC 50 ) was calculated based on concentration-response curve (Fig. 5). Based on the curve, EC 50 of FB 1 towards zebrafish was 0.569 ppm.

Amplification of FUM1 in F. proliferatum and F. verticillioides
Out of 48 isolates were used in FUM1 gene detection study, 21 isolates were F. verticillioides and 27 isolates were F. proliferatum. This result can be compared with previous research (Nor Azliza et al. 2017) detected 780 bp of FUM1 gene in 11 isolates of F. verticillioides, 49 isolates of F. proliferatum, 24 isolates of F. fujikuroi and 10 isolates of F. oxysporum from various range of host plants.
Another report by Dissanayake et al. (2009) found that 13 out of 20 isolates of F. proliferatum isolated from Welsh onion have successfully amplified 780 bp of FUM1 gene specific fragments.
The absence of FUM1 gene in two isolates of F. proliferatum (B68 and B92) might be caused by mutation where there were defects in FUM1 gene expression or might be due to the movement of gene cluster within genome during the evolution of the species (Proctor et al. 2004). It is important to identify the presence of FUM1 gene before quantifying the level of FB 1 produce in order to see the correlation between the presences of the gene and the level of the mycotoxin produced.

FB 1 Production using uFlc Analysis
There were six isolates (three isolates from each species) that were not capable of producing FB 1 and no detection from UFLC analysis. Even though FUM1 gene was amplified in Fusarium isolates, it does not reflect the FB 1 production. The biosynthesis of FB 1 also can be influenced and altered by the changes in abiotic factors. The average concentration of FB 1 produced by F. verticillioides and F. proliferatum was 143.99 ppm and 655.92 ppm respectively which is lower than previous studies (Choi et al. 2018) where F. verticillioides produced 0.6-4749.1 µg g -1 (ppm) FB 1 and F. proliferatum produced 0.9-7969.0 µg g -1 (ppm) FB 1 . Meanwhile another research by Lee et al. (2012) has reported that F. verticillioides produced 2686.4 µg g -1 (ppm) FB 1 and F. proliferatum produced 974.0 µg g -1 (ppm) FB 1 . High water activity will lead to higher fungal growth thus more fumonisin will produce (Samapundo et al. 2005). Besides that, the C:N ratio also can influence the production of fumonisin by activating selective genes that are outside of the FUM gene cluster. For example, when C source is increasing, ZFR1 and FST1 genes will regulate the FB 1 biosynthesis (Jimenez et al. 2003;Jurado et al. 2008). Meanwhile, when the N source is decreasing, FUM1, FUM8 and AREA genes will significantly increase and induce the production of fumonisin (Kohut et al. 2009).

Morphological Abnormalities of Zebrafish Larvae Exposed to FB 1
The severity of pericardial edema increases at 72 hpf onwards with exposure 0.72 ppm, 0.87 ppm and 1.00 ppm (Figs. 2 and 3). Since currently there is no evidence and reports of FB 1 exposure on zebrafish, all the comparisons can only be made with other studies of different type mycotoxins and chemicals exposure on zebrafish. Pericardial edema is one of the signs of kidney failure where the fluid accumulated in the body cavity and causing swollen of the heart (Hanke et al. 2013). The decrease in plasma oncotic pressure causes the excess fluid buildup indicates failure of the kidney where the pronephric tubules were unable to reabsorb the essential protein (Kemp & Conte 2012;Hanke et al. 2013). This is significant to a human disease called proteinuria where there are excess proteins released through urine.
Severe yolk sac edema can be observed at 120 hpf with exposure of 0.72 ppm and higher. The previous report described the effect of zebrafish larvae that were exposed to thifluzamide fungicide also showed yolk sac edema and pericardial edema . Yolk sac edema also was one indication of kidney failure that involved the osmotic pressure. Kidney failure can cause permeability defects and ionic imbalance thus lead to edema and creating a positive feedback loop. The edema later will disrupt the function of kidney and circulatory function and further worsen the edema. This positive feedback loop of edema is irreversible (Hill et al. 2004).
Spinal curvature can be seen at all concentrations as early as at 48 hpf. The spine can be observed either curve upward, curve downward or appear undulate. Yin et al. (2015) stated that the exposure of fusaric acid towards zebrafish have caused all the embryos to develop undulated spine when treated with 400 µM-600 µM (72 ppm-108 ppm) of fusaric acid. However, no report on when expose to FB 1 .

Tail Coiling, Heartbeat and Mortality Rate of Larval Zebrafish
Spontaneous tail coiling was reported as the first behaviour of zebrafish showing the development of the motor activity. Tail coiling is independent of sensory stimulation starts at 17 hpf and gradually decreasing at 26 hpf (Brustein et al. 2003). During coiling, the tail movement was non-myogenic and originates from spinal cord, primary motoneurons and a restricted number of interneurons (Hirata et al. 2005). Fumonisin exposure might have caused interference on the development of the nervous system as it significantly reduced the tail-coiling rate of zebrafish.
The effect of FB 1 on heart just started and did not interrupt the heart rate thus revealed the non-significance difference of heart rate. Exposure of pyridinebased pesticide (PPF) on zebrafish embryos and the effects of heart rate showed that there was no significant difference in the heart rate in 0.16 and 0.33 µg mL -1 till 96 hpf (Maharajan et al. 2018). However, embryos that were exposed to higher concentration of 1.66 µg mL -1 at 48 and 72 hpf significantly increase the heart rate.
From the result, it showed that the effect of FB 1 on mortality of zebrafish larvae was late at day 6 (144 hpf) of exposure and increasing as exposure time increased. The mortality also happened due to several abnormalities and malfunction of the system in zebrafish, for example, circulatory and excretory systems. Exposure of 50 µM of citrinin at 72 hpf showed viability of 71.6% ± 10.5% and reduced completely to 0% at 120 hpf where none of the zebrafish survived (Wu et al. 2013). The study also stated that the viability of zebrafish decreases as the concentration and time of exposure to citrinin increased. T-2 toxin exposure to zebrafish larvae have found that 0.80 µmol L -1 caused 100% mortality at 24 hpf. Meanwhile, zebrafish treated with 0.20 µmol L -1 , 0.25 µmol L -1 , 0.30 µmol L -1 and 0.40 µmol L -1 have caused 0%, 5.5%, 30.5% and 83.3% mortality rate respectively at 24 hpf (Yuan et al. 2014). These results indicated that FB 1 , citrinin and T-2 toxin have different levels of toxicity towards zebrafish depending on the concentration level and time of exposures. Since this is the first report of exposure of FB 1 on zebrafish larvae, higher concentrations should be applied on the larvae as well as increasing the time of exposures to get the 100% mortality rate. Based on this study, EC 50 of FB 1 towards zebrafish was 0.569 ppm. The previous report found that the EC 50 of T-2 toxin towards zebrafish was 0.18 µmol L -1 (0.084 ppm) (Yuan et al. 2014). The value is lower as compared to FB 1 and it indicates that FB 1 requires higher concentration for the effective response of 50% mortality.
conclusIon Detection of FUM1 gene on F. proliferatum and F. verticillioides revealed that 48 out of 50 isolates gave positive amplification of FUM1 gene and from quantification of FB 1 by using UFLC, 18 out of 21 isolates of F. verticillioides produced FB 1 ranging from 12.26 ppm-954.01 ppm. Meanwhile, 26 out of 29 isolates of F. proliferatum produced FB 1 ranging from 16.48 ppm-6677.32 ppm. Embryotoxicity test of FB 1 on zebrafish showing significant effects on the mortality rate and morphological abnormalities. Meanwhile the EC 50 of FB 1 towards zebrafish was 0.569 ppm. In addition, knowledge on the FB 1 effects on larval zebrafish will help to understand further the pathology of associated acute and chronic diseases in an aquatic animal.