The anti-aflatoxigenic mechanism of cinnamaldehyde in Aspergillus flavus

Aflatoxin B1 (AFB1), the predominant and most carcinogenic naturally polyketide, is mainly produced by Aspergillus flavus and Aspergillus parasiticus. Cinnamaldehyde has been reported for inhibiting the growth and aflatoxin biosynthesis in A. flavus. But its molecular mechanism of action still remains largely ambiguous. Here, the anti-aflatoxigenic mechanism of cinnamaldehyde in A. flavus was investigated via a comparative transcriptomic analysis. The results indicated that twenty five of thirty genes in aflatoxin cluster showed down-regulation by cinnamaldehyde although the cluster regulators aflR and aflS were slightly up-regulated. This may be due to the up-regulation of the oxidative stress-related genes srrA, msnA and atfB being caused by the significant down-regulation of the diffusible factor FluG. Cinnamaldehyde also inhibited aflatoxin formation by perturbing GPCRs and oxylipins normal function, cell wall biosynthesis and redox equilibrium. In addition, accumulation of NADPH due to up-regulation of pentose phosphate pathway drove acetyl-CoA to lipids synthesis rather than polyketides. Both GO and KEGG analysis suggested that pyruvate and phenylalanine metabolism, post-transcriptional modification and key enzymes biosynthesis might be involved in the suppression of AFB1 production by cinnamaldehyde. This study served to decipher the anti-aflatoxigenic properties of cinnamaldehyde in A. flavus and provided powerful evidence for its use in practice.

particular, it can depress the production of aflatoxin by A. flavus 1,8 . It also stimulates apoptosis and inhibits tumor growth 14 , and has been reported as an effective agent against several cancers effectively 15,16 .
Out of 10 essential oils previously studied by our group, cinnamaldehyde was the most effective against fungal growth and aflatoxin production by A. flavus 17 . Production of aflatoxin B 1 (AFB 1 ) was completely inhibited by cinnamaldehyde at lower concentration (0.4 mM) without influencing A. flavus growth, and at the concentration of 0.8 mM cinnamaldehyde showed complete inhibition of fungal growth and AFB 1 production 1 . This was similar to the results reported by Sun 8 , which indicated that fungal growth and aflatoxin production were significantly inhibited by cinnamaldehyde in dose-dependent manner by modulating the oxidative stress in A. flavus. It takes inhibitory action against bacteria 11,18,19 , yeast and filamentous molds [20][21][22] by depressing intracellular ATP 23 , cell wall biosynthesis 24 , and altering the membrane structure and integrity.
Using qPCR, Yin et al. 24 found that cinnamaldehyde (0.005%) significantly inhibited AFB 1 production in A. flavus and A. parasiticus. The expressions of the majority of aflatoxin gene cluster were down-regulated by more than 4-folds, especially pksA (aflC), nor-1 (aflD), and norA (aflE). Our previous studies showed that AFB 1 production was largely reduced in A. flavus treated with cinnamaldehyde at the low concentration in YES medium 1 . In the presence of cinnamaldehyde (0.4 mM), aflM was significantly down-regulated by more than 5963-fold, following by aflP, aflR, aflD and aflT. The decreased transcription levels of aflatoxin cluster genes subsequently resulted in the reduction of AFB 1 production. Although many researchers develop desire at exploring the anti-aflatoxigenic mechanism of cinnamaldehyde, the detailed molecular mechanism behind in the inhibition of aflatoxin biosynthesis by cinnamaldehyde still remains largely ambiguous.
RNA-seq, a high-throughput sequencing technology used to sequence complementary DNA, has been applied to transcriptomic studies, including anti-fungi response mechanism to essential oils. Wang et al. 12 found that cinnamaldehyde inhibited P. expansum by modulating the oxidative stress and down-regulating the ergosterol biosynthesis using transcriptional profiling analysis. In another report, the transcriptome profiling of A. flavus exposed to antioxidant gallic acid was used in exploring the response mechanism 25 . The gallic acid played a pivotal role in fungal development via over-expression of brlA while the velvet complex didn't show a significant differential expression. In addition, other regulators were also involved in the inhibitory mechanism of gallic acid. In another transcriptional profiling analysis of A. flavus exposure to 5-azacytidine (5-AC), the up-regulation of brlA was also found 25 .
The main aim of this study was to investigate the role of cinnamaldehyde in the inhibition of fungal development and secondary metabolite biosynthesis of A. flavus via RNA-seq approach. The differentially expression genes between cinnamaldehyde treated and untreated A. flavus were obtained and further analyzed. Especially, the anti-aflatoxigenic mechanism of cinnamaldehyde was revealed. This work may also contribute to better understanding on the aflatoxin biosynthesis and regulation.

Results
Overall transcriptional response profile of A. flavus to natural cinnamaldehyde. To explore the latent detailed molecular mechanism response to natural cinnamaldehyde on A. flavus, a transcriptomes analysis was implemented to evaluate the response at mRNA level. Averagely, A. flavus YC-15 untreated and treated with cinnamaldehyde generated 10.63 million and 11.11 million raw reads, respectively. From these raw reads, 8.84 million and 9.26 million clean reads were obtained after purity filtering. And, 58.84% and 68.36% of total clean reads from control and treatment groups were mapped to the reference genome sequence while only 0.02% were aligned to rRNA genes. The mRNA data revealed that 1032 genes were significantly differentially transcribed between the A. flavus treated with cinnamaldehyde and the untreated sample. Among them, 427 genes' transcripts showed up-regulation and 605 genes showed down-regulation in cinnamaldehyde-treated group compared with the untreated group.

Functional classification and pathway analysis of differential expression genes (DEGs). The
DEGs between the A. flavus treated with cinnamaldehyde (R75) and control group (CK) provided a potential anti-aflatoxigenic mechanism of cinnamaldehyde related to A. flavus. These 1032 DEGs related to a large quantity of regulatory and metabolic process were identified (with FDR ≤ 0.05, log 2 Ratio ≥1 or ≤1) between R75 and CK according to the FPKM values. In order to analyze the functions of 1032 DEGs, GO functional and KEGG metabolic pathways enrichment analyses were performed. GO analysis revealed that these significantly DEGs were mainly involved in oxidoreductase activity, RNA binding, Nuclease activity and translation ignition factor activity ( Table 1, Fig. 1). KEGG analysis revealed that these significantly DEGs were mainly involved in RNA transport, ribosome biogenesis, pyruvate metabolism, phenylalanine metabolism, sulfur relay system and sulfur metabolism ( Table 2). Genes involved in biosynthesis of conidial pigment, aflatrem, aflatoxin and cyclopiazonic acid. The expression profile referred to the biosynthesis of conidial pigment (#10), aflatrem (#15), aflatoxin (#54), and cyclopiazonic acid (#55) were evaluated and shown in Table 3. In pathway #10, O-methyltransferase family protein (AFLA_016120) and a hypothetical protein (AFLA_016130) were down-regulated, but arp1 gene was up-regulated. In pathway #15, the majority of cluster genes showed low-level expression. In pathway #55, MFS multidrug transporter (AFLA_139460) was slightly down-regulated, nevertheless the genes encoding a hybrid PKS/NRPS enzyme (AFLA_139490), FAD dependent oxidoreductase (AFLA_139470), and tryptophan dimethylallyltransferase (AFLA_139480) were up-regulated. Our previous studies confirmed that cinnamaldehyde can repress the aflatoxin production and development in dose-dependent manner 1 . Based on the transcriptome data, 25 of 34 genes in aflatoxin biosynthetic gene cluster were down-regulated to varying degrees including the key structural genes aflC, aflD, aflE, aflG, aflH, aflI, aflL, aflM, aflO, aflP and aflQ in A. flavus treated with 0.60 mmol/L of cinnamaldehyde. Surprisingly, both transcription regulator genes aflR and aflS in aflatoxin cluster showed a slight up-regulation. However, the lower level of aflS/aflR ratio was observed compared with the untreated group, subsequently resulting in the down regulation of most structural genes. The aflU gene, encoding a P450 monooxygenase and probably involving in the biosynthesis of AFG 1 and AFG 2 rather than AFB 1 , was up-regulated after treatment of cinnamaldehyde. The aflT gene, encoding a membrane-bound protein presumed to be involved in aflatoxin secretion, was not affected by cinnamaldehyde. Furthermore, genes involved in aflatoxin biosynthesis pathway were analyzed through qPCR and the results were shown in Fig. 2. Meantime, the sugar cluster genes sugR, orf, glcA and hxtA were significantly up-regulated by cinnamaldehyde.

Genes involved in fatty acids β-oxidation and pentose phosphate pathway. Fatty acids
β-oxidation in peroxisome and mitochondria promoted aflatoxin formation 26 . And there is a competition in acetyl-CoA between lipid synthesis and polyketides formation 27 . The transcriptional levels related fatty acids β-oxidation and pentose phosphate pathway were shown in Table S1. A large number of fatty acids β-oxidation-related genes were significantly down-regulated. The most strongly down-regulated gene in peroxisome was AFLA_009410, followed by AFLA_135240, AFLA_091060, and AFLA_090720. However, most of the genes in pentose phosphate pathway were up-regulated in A. flavus exposure to cinnamaldehyde, for example, Zwf1 (AFLA_086620), Sol (3AFLA_080390), and Gnd1 (AFLA_036840).

Figure 1.
The gene ontology annotation of differential expression genes.   www.nature.com/scientificreports www.nature.com/scientificreports/ Genes involved in ergosterol biosynthesis. The plasma membrane plays a crucial role in maintaining homeostasis, exchanging materials, and transduction of information. And ergosterol is one key constituent of fungal membrane 19,28 . The transcriptional levels related ergosterol pathway was shown in Table S2. Transcriptional levels of several genes were down-regulated after cinnamaldehyde treatment, for example, sterol delta 5,6-desaturase Erg3 (AFLA_018090), squalene monooxygenase Erg1 (AFLA_061500), and C-14 sterol reductase (AFLA_051080, AFLA_111350).

ID TERM (molecular functions) p-value q-value
Genes involved in fungal development. The regulation of secondary metabolism is associated with fungal growth and development. From the expression profile data, we found that the expression patterns of some gene referred to conidiophores development were down-regulated when A. flavus was treated with cinnamaldehyde (Table S3). For the velvet complex, veA did not show a significant differential expression while velB, leaA, and vosA, were slightly up-regulated exposure to cinnamaldehyde. fluG (AFLA_039530), encoding a protein comprising an N-terminal amidohydrolase domain and a C-terminal glutamine synthetase domain 29 , was down-regulated. And esdC, an early sexual development gene, was mildly down-regulated. Nevertheless, development regulator FlbA was up-regulated. BrlA mediating conidiophores, and AbaA controlling phialide differentiation were also up-regulated. In addition, RodA and RodB, conidial hydrophobic genes, both showed strong up-regulation.  Table S4. Among the 47 relevant genes, ap-1, atfB and msnA were all up-regulated. The cellular receptors gprC, gprF, gprK, gprM, gprR, gprP, gprS, the oxylipins ppoA, ppoB and ppoC, the MAP kinase genes mkk2, fus3, pbs2, mpkA, sakA, bck1, ste11, sskB and ste7, and catalase gene cat1, catA, and superoxide dismutase gene sod1, mnsod were all up-regulated to varying degrees. AfPXG, encoding calcium binding protein caleosin, and GPCRs (gprA, gprB, gprD, gprG, gprH) were all down-regulated.

Discussion
Cinnamaldehyde is gradually regarded as safer food additive in food processing and manufacturing comparing to chemical fungicides. The inhibitory effects and mechanism of cinnamaldehyde on fungal growth and mycotoxin have been reported by many researchers 1,8 . In our previous study, 0.4 mM cinnamaldehyde inhibited AFB 1 production with the rate of 68.9%, and 0.8 mM cinnamaldehyde could completely suppress A. flavus growth 1 . In this study, the mechanism of A. flavus growth and aflatoxin formation dysfunction exposure to cinnamaldehyde was investigated by RNA-seq analysis. Moreover, the anti-fungal and anti-aflatoxigenic properties of cinnamaldehyde were discussed and conclusions were drawn based on the results of the previous studies and this study.
Aflatoxin synthesis is supported by the action of enzymatic cascade and involves 21 steps 30 . In A. flavus, this process is managed by a gene cluster in which aflR and aflS serve as regulators 31,32 . In our RNA-seq data, 25 genes of the aflatoxin biosynthesis cluster were down-regulated after treatment with cinnamaldehyde although aflR and aflS were up-regulated. With the exception of an up-regulated result for aflF, all the structural genes in the cluster were down-regulated. However, none of these was completely suppressed. The most strongly down-regulation gene was aflD, followed by the key structural genes aflG, aflH, aflP, aflM, aflI, aflL and aflE. The expression levels of all genes in the cluster were confirmed by q-PCR (Fig. 2). The aflF, encoding a dehydrogenase, is involved in the conversion of NOR to AVN 33 . The expression level of aflF was up-regulated, but its two homology protein genes aflD and aflE both were down-regulated. The gene aflT, encoding a membrane-bound protein presumed to be involved in aflatoxin secretion, was not modulated after treatment with cinnamaldehyde. And similar results were also reported in A. flavus treated with piperine 34 and eugenol 34,35 . These findings indicated that cinnamaldehyde suppressed aflatoxin biosynthesis by down-regulating the transcript levels of most structural genes.
An astonishing result is that the transcriptional factor aflR and cofactor aflS showed a mild up-regulation in A. flavus treated with cinnamaldehyde. In our previous study 35 , the expression level of aflR and aflS showed slight up-regulation in A. flavus treated with eugenol although aflatoxin production was significantly inhibited by www.nature.com/scientificreports www.nature.com/scientificreports/ eugenol. Similarly, 5-Azacytidine (5-AC) suppressed fungal development and aflatoxin synthesis while aflR and aflS were mildly up-regulated 36 . The results are also much like with the findings reported by Zhao and his colleagues 25 . They found that aflR and aflS were up-regulated slightly while structure gene showed down-regulation exposure to an antioxidant gallic acid which could inhibit A. flavus development and aflatoxin production. The similar result that the aflatoxin cluster regulators aflR and aflS were up-regulated slightly although most structural genes were down-regulated in A. flavus treated with different anti-aflatoxigenic natural compounds, suggesting the stable expression of aflR and aflS.
The velvet complex was critical for conidiation and aflatoxin formation in A. flavus 37,38 . In the deletion mutant of veA, the expression of key aflatoxin genes including aflR, aflD, aflM and aflP was completely suppressed. Consequently, aflatoxin was halted 39 . However, veA did not show significant differential expression although almost all structural genes were down-regulated. The oxidative stress-related genes such as msnA, srrA, atfB and pacC, which were positively regulated by veA, were up-regulated after cinnamaldehyde treatment. The similar result was obtained in A. flavus treated with eugenol 35 . The LaeA and velB genes, encoding the other two proteins of velvet complex, were slightly up-regulated. Interestingly, a velvet-related gene FluG were significantly down-regulated in A. flavus treated with cinnamaldehyde. FluG, composed of an N-terminal amidohydrolase domain and C-terminal glutamine synthetase domain, was assumed for synthesizing a diffusible factor 40 . Chang et al. (2013) reported that VeA, VelB, and LaeA, combined with FluG, were indispensable to maintaining conidiation program, sclerotial formation, and aflatoxin biosynthesis in A. flavus 41 . These results suggested that FluG may play an important role in the anti-aflatoxigenic mechanism by cinnamaldehyde.
Acetyl-CoA, the fundamental structure element of all known fungal polyketides, is mainly produced from fatty acids β-oxidation and glycolysis of sugars. For aflatoxin biosynthesis, fatty acids β-oxidation is a major contributor to acetyl-CoA 26 . It was reported that pentose phosphate pathway activity was associated with lower content of aflatoxin. Zhao et al. 25 found that gallic acid inhibited the aflatoxin formation via up-regulation of pentose phosphate pathway. Incubated in aflatoxin inhibitory medium, A. flavus pentose phosphate pathway was accelerated leading to NADPH accumulation 42 . Ultimately, acetyl-CoA was converted into lipid biosynthesis rather than polyketide formation 27 . In our data, there were large number of fatty acids β-oxidation-related genes showing significant down-regulation after cinnamaldehyde treatment, such as AFLA_019280 (peroxiredoxin), AFLA_052400 (isocitrate lyase) and AFLA_009410 (delta (3,5)-delta (2,4)-dienoyl-CoA isomerase) (Table S3). Meantime, most of the genes involved in pentose phosphate pathway were up-regulated exposure to cinnamaldehyde including AFLA_041580 (estradiol 17 beta-dehydrogenase), AFLA_115890 (acyl-CoA oxidase) and AFLA_080390 (6-phosphogluconolactonase Sol). These results suggested that the down-regulation of fatty acids β-oxidation and the up-regulation of pentose phosphate pathway were also associated with the anti-aflatoxigenic mechanism of cinnamaldehyde.
Cinnamaldehyde was considered to make its antifungal effects on perturbing cell wall biosynthesis, ergosterol biosynthesis and ATPase 43 . The 4 genes associated with cell wall, AFLA_098380, AFLA_083360, AFLA_014260 and AFLA_100100, were down-regulated. The similar phenomenon had been reported that cinnamaldehyde caused several genes involved in cell wall biosynthesis dysfunction 12,24 . Ergosterol is one of the principal sterol ingredients in the fungal membrane and is crucial for survival due to the ability in maintaining cell membrane fluidity, permeability, and pheromone signaling [44][45][46] . In the present work, the transcriptional level of several genes related ergosterol was down-regulated, for example squalene monooxygenase Erg1 (AFLA_061500), C-14 sterol reductase (AFLA_051080 and AFLA_111350). The Erg1 gene of S. cerevisiae encodes squalene epoxidase, a key enzyme in the ergosterol pathway. Disruption of the gene resulted in a lethal phenotype when cells grew under aerobic conditions, even in the presence of ergosterol 47 . C-14 sterol reductase (AFLA_051080 and AFLA_111350) were all down-regulated. Double deletion of Erg25 genes was lethal in A. fumigatus 48 . Cinnamaldehyde weakened ergosterol biosynthesis which resulted in the disruption of the intracellular ATP, and some essential irons equilibrium 43 . In E. coli and Listeria monocytogenes, cinnamaldehyde inhibited the membrane-bound ATPase activity 49,50 . In the present study, some genes related to mitochondrial ATPase activity were repressed, for example, mitochondrial F1F0 ATP synthase subunit (AFLA_129660, AFLA_032070 and AFLA_043330).
RNA-binding was found to be the most dysregulated function after cinnamaldehyde treatment using GO enrichment analysis. Our previous study found the similar results in A. flavus treated with eugenol 35 . Therefore, similar with eugenol, the post-transcriptional regulation may play an important role in the anti-aflatoxigenic mechanism of cinnamaldehyde. KEGG metabolic pathway analysis showed that pyruvate metabolism and phenylalanine metabolism were the main dysregulated metabolic pathway after cinnamaldehyde treatment. Pyruvate locates intersection of intermediary metabolism, which refers to multiple metabolic processes covering gluconeogenesis, lipogenesis and energy production 51 . As a metabolic switch, the pyruvate dehydrogenase complex (PDH) was considerable for carbon metabolism because of turning pyruvate into acetyl-coA 52 . Acetyl-CoA and malonyl-CoA are precursor substances in aflatoxin formation 53 . Besides, PDH was crucial for morphology and pathogenicity in different fungal species 54,55 . Amino acid metabolism plays an important role in aflatoxin biosynthesis. It was reported that phenylalanine metabolism was dysregulated in A. flavus treated with 2-phenylethanol 56 . In addition, phenylalanine was lightly incorporated into aflatoxin in A. flavus 57 . These results suggested that pyruvate metabolism and phenylalanine metabolism dysfunction might result in the reduction of aflatoxin biosynthesis.
Different stress can perturb cellular redox equilibrium, resulting in enhancive reactive oxygen species (ROS) levels named oxidative stress 58 . Excessive accumulation of ROS can jeopardize DNA, proteins and lipids, leading to cellular dysfunction 59 . Several researchers have thought that oxidative stress is a pre-condition for aflatoxin biosynthesis in A. flavus and A. parasticus 60,61 . The hypothesis is associated with the tentative that aflatoxin biosynthesis protects the fungus against oxidative stress. Reverberi et al. 59 introduced a P33 gene into A. flavus resulting in enhanced ROS accompanying aflatoxin accumulation. On the contrary, antioxidants such as gallic acid and ethylene reduced the oxidative stress in A. flavus leading to the decrease of aflatoxin content 25 . GPCRs and www.nature.com/scientificreports www.nature.com/scientificreports/ oxylipins are tied in oxidative process. The expression levels in regard to oxidative-related genes were shown in Table S3. After cinnamaldehyde exposure, 7 GPCRs and 2 oxylipins genes showed significant differential expression. In this study, we found that gprC, gprF, gprK, gprM and grpS were significantly up-regulated with AFB 1 inhibition in A. flavus treated with cinnamaldehyde. Similar results were obtained in A. flavus treated with eugenol in our previous study 35 . The genes, grpC, gprF, gprK, gprM and gprS were also up-regulated after eugenol treatment. Caceres et al. 34 also reported that over-expressed gprK accompanied with lower content of AFB 1 . Oxylipins pathway includes four genes, ppoA, ppoB, ppoC, and afPXG in A. flavus 62 . Affeldt et al. 63 reported that high content of oxylipins was associated with lower levels of aflatoxins. Simultaneous silencing via RNAi of ppoA, ppoB and ppoC and AfPXG resulted in an increase of aflatoxin biosynthesis 62 . Caceres et al. 34 also found that over expression of ppoB and ppoA was correlated with AFB 1 inhibition by piperine. In present study, the expression levels of ppoA, ppoB and ppoC were all up-regulated, suggesting the decreased oxylipins genes expression was associated with AFB 1 inhibition by cinnamaldehyde. All these results suggest that the up-regulation of GPCRs and oxylipins genes was involved in AFB 1 inhibition by cinnamaldehyde.
In A. flavus and A. parasiticus, there were several bZIP transcription factors referring to aflatoxin biosynthesis and oxidative stress response. Among these, SrrA, AtfB, AP-1, and MsnA were characterized as co-regulators 60,[64][65][66][67] . In this study, we found that genes belonging to bZIP-type family were involved in the anti-aflatoxigenic mechanism of cinnamaldehyde. SrrA, an orthologue of S. cerevisiae Skn7 and Saccharomyces pombe Prr1, controlled key functions in response to osmotic and oxidative stress and was considered as a regulator in aflatoxin biosynthesis 64 . AP-1, a highly conserved protein in mammalian, yeast and fungi 60,68,69 . AP-1 may play crucial roles in sensing ROS because of high cysteine content in N-and C-terminal 70 . Over-expression of napA, an ortholog of AP-1, resulted in secondary metabolite inhibition in A. nidulans which implied napA was a negative regulators in secondary metabolite synthesis 71 . In A. parasiticus, the ApyapA disruption resulted in more aflatoxin production 60,65 . In this study, the ap-1 showed up-regulation accompanying with aflatoxin inhibition in A. flavus cinnamaldehyde exposure. Similar results were also obtained by Caceres et al. 34 . They found that the AP-1 was up-regulated with aflatoxin inhibition in A. flavus after piperine treatment. AtfA mediates several processes in vegetative hyphae, contributes to stress tolerance and changes secondary metabolism in A. nidulans 72 , A. oryzae 73 , and A. fimugatus 74 . AtfB, an orthologue of AtfA, is an important regulator referring to aflatoxin production and oxidative stress via binding to CER sites of aflatoxin biosynthesis genes promoter 64,67 . This CRE binding site was found in 7 genes promoter regions 34 . In the present study, AtfA did not show significant differential expression while AtfB was up-regulated by cinnamaldehyde. Caceres et al. 34 also found AtfB was up-regulated with decreased production of aflatoxin after piperine treatment. MsnA has an important effect on fungal growth, aflatoxin and kojic acid formation, and oxidative stress 64 . In A. flavus and A. parasiticus, MsnA disruption resulted in aflatoxin and ROS accumulation 75 . In our previous study, we also found that transcript factor MsnA played a negative role in aflatoxin biosynthesis 35 . Similar result was obtained in A. flavus treated with cinnamaldehyde. Taken together, srrA, atfB, ap-1, and msnA were all up-regulated after cinnamaldehyde exposure. These results implied that bZIP transcription factors SrrA, AtfB, AP-1, and MsnA up-regulation played a direct negative role in aflatoxin formation after cinnamaldehyde treatment.
Antioxidant enzymes SOD and CAT which were regulated by the bZIP transcription factors make crucial effect on defense against ROS 12 . Many publications have reported that some inhibitors could suppress aflatoxin formation via positive regulating the antioxidant enzymes activities. However, different aflatoxin inhibitors act on different type of antioxidant enzymes. For example, piperine and β-glucans from Lentinula edodes led to lower AFB 1 production with higher CAT activity 34 . Oppositely, eugenol and ascorbic acid sharply depressed the AFB 1 biosynthesis accompanying with high SOD activity 34 . In addition, gallic acid may equilibrium ROS by activating the glutathione-and thioredoxin-dependent antioxidant system instead of changing CAT and SOD activities 25 . In this study, we found that antioxidant enzymes catalase gene (cat, cat1, and catA), and superoxide dismutase gene (sod1, and mnSOD) were all up-regulated in A. flavus treated with cinnamaldehyde. However, Sun et al. 8 reported that exposure to cinnamaldehyde only resulted in higher SOD activity using the hydroxylamine analysis. The different results may imply that (1) reveals a dose effect; (2) exists a post-translational modification of CAT. These results made it clear that cinnamaldehyde enhanced CAT and SOD activities as part of its anti-aflatoxigenic mechanism. Figure 3 shows the hypothetical gene modulation mode of action on aflatoxin formation and fungal growth in A. flavus treated with cinnamaldehyde at transcription levels. The signal transduction disorder happens when cinnamaldehyde regulates the expression of GPCRs and oxylipins genes. Velvet complex together with FluG modulates conidiation, sclerotial production, and aflatoxin biosynthesis. However, the differential expression of LaeA, veA, and VelB was not significant. The down-regulation of FluG may trigger the expression of stress response transcriptional factor gene srrA, which results in up-regulation of bZIP transcriptional factor ap-1, zinc finger transcriptional factor msnA, and CREB/ATF family member atfB. Ultimately, the redox system is perturbed and then antioxidant enzymes are activated. In addition, AP-1, MsnA, AtfB, as negative regulatory factors, modulate aflatoxin biosynthesis gene cluster. For conidia development, early asexual development factor FlbA is modulated by velvet complex and FluG. Up-regulation of FlbA activates FadA and SfaD which play a negative role in the expression of esdC. Besides, FlbA causes the up-regulation of BrlA which triggers over-expression of AbaA and wetA. Taken together, down-regulation of esdC and over-expression of BrlA, AbaA, and wetA facilitate asexual development.
To sum up in Figs 3 and 4, cinnamaldehyde inhibits the aflatoxin biosynthesis and fungal growth of A. flavus via (1) reducing the fatty acid oxidation level by modulating several oxidation-related genes which leads to marked reduction of aflatoxin precursor acetyl-CoA; (2) increasing the NADPH accumulation by HMP which competes with aflatoxin biosynthesis; (3) weakening ergosterol synthesis which does damage to cell membrane integrity accompanied with altering the intracellular ATP and some indispensable iron equilibrium; (4) disturbing the redox system and then activating antioxidant enzymes which are deemed as key elements for www.nature.com/scientificreports www.nature.com/scientificreports/ regulating aflatoxin-related genes. These results uncovered in this study play a critical role in understanding the anti-aflatoxigenic mechanism of cinnamaldehyde in A. flavus and may accelerate its use in practice. Moreover, these results should assist further studies on the mechanism of action of inhibitor against fungal growth and mycotoxin production.

Conclusion
The results of this study put forward a mechanism to explain the transcription regulation concerning the inhibitory effect of cinnamaldehyde on aflatoxin biosynthesis via RNA-seq. On basis of early studies, we draw a conclusion that (1) the decline in aflatoxin biosynthesis is on account of the down-expression of most of structural genes of aflatoxin cluster after treatment with cinnamaldehyde; (2) accumulation NADPH drives acetyl-CoA to lipid synthesis rather than polyketide formation; (3) the down-expression of diffusible factor FluG working with the velvet complex and the concomitant up-regulation of the oxidative stress-related genes srrA, msnA, and atfB; (4)