Fusaric Acid immunotoxicity and MAPK activation in normal peripheral blood mononuclear cells and Thp-1 cells

Fusaric acid (FA), a food-borne mycotoxin, is a potent divalent metal chelator. The human immune system is complex and susceptible to environmental insult however, the immunotoxity of FA remains unknown. We investigated the immunotoxicity of FA on human peripheral blood mononuclear cells (PBMCs) and Thp-1 cells. FA was cytotoxic to PBMCs (IC50-240.8 μg/ml) and Thp-1 (IC50-107.7 μg/ml) cells at 24 h. FA induced early apoptosis but significantly decreased caspase activity in PBMCs, a characteristic of paraptosis. In Thp-1 cells, FA induced apoptosis and increased caspase −9 and −3/7 activities. In PBMCs, FA maintained mitochondrial membrane potential and decreased protein expression of Bax whilst increasing expression of p-Bcl-2; FA induced oxidative stress and depleted ATP levels in both cell types. In Thp-1 cells, FA increased mitochondrial membrane depolarization and decreased p-Bcl-2 expression. In PBMCs, FA significantly up-regulated the MAPK protein expression of p-ERK and p-JNK but down-regulated p-p38 expression. In Thp-1 cells, FA up-regulated MAPK protein expression of p-ERK whilst p-JNK and p-p38 expression were down-regulated. In conclusion FA induced programmed cell death and altered MAPK signaling in healthy PBMCs and Thp-1 cells strongly suggesting a possible mechanism of FA induced immunotoxicity in vitro.

Mitochondria are important in maintaining cellular redox homeostasis and activation of the intrinsic apoptotic pathway. The evaluation of FA induced mitochondrial membrane integrity by flow cytometry in PBMCs showed that despite increased MDA levels, there was no effect on the mitochondrial membrane potential; however, in Thp-1 cells, mitochondrial membrane depolarization was significantly increased (1.95 fold; 33.58 ± 1.425% vs. 65.48 ± 0.329%; p = 0.0007) when compared to the controls (Fig. 3B).

Discussion
The promising role of divalent ion chelators in proliferative and virulent diseases has led to growing interest [35][36][37] . FA, a picolinic acid analogue and potent divalent metal chelator, has shown potential as an anti-cancer, anti-microbial and anti-viral agent [35][36][37] . Recently, the role of divalent ion chelators in proliferative and virulent diseases has been extensively studied. However, the toxicity of FA, a food-borne mycotoxin, on the immune system containing a diverse population of cells, has to date not been studied.
In this study, we showed the immunotoxic potential of FA to both healthy PBMCs (diverse population of immune cells) and the distinct immune Thp-1 cells. Interestingly, FA inhibited Thp-1 cell proliferation at an IC 50 value less than half that of PBMCs ( Fig. 1A and B). This result is in agreement with other studies that showed cytotoxic effects of FA on WI-38 cells (fibroblastic cells), LoVo cells (colorectal adenocarcinoma cells) and MDA-468 cells (human breast adenocarcinoma cells) in which FA had preferentially inhibited the proliferation of cancerous cells (LoVo and MDA-468) when compared to the normal cells (WI-38) 35 . Our data suggests that FA may exert selective toxicity to distinct immune cell types as evidenced by the Thp-1 response, albeit a leukemic cell line. Additionally, in comparison to the anti-neoplastic drug, ellipticine failed to inhibit PBMC and Thp-1 cell viability; which further exemplifies the potency of FA.
Contrary to the study by Fernandez-Pol (1998), FA significantly increased the externalization of PS in both PBMCs and Thp-1 cells by 26.16% and 18.19%, respectively (Fig. 2). Given that the externalization of PS occurs during both apoptosis and paraptosis, activities of caspases −8, −9 and −3/7 were assayed to determine type of cell death induced by FA in both cell lines; apoptosis requires caspase activation whilst paraptosis is independent of caspase activation. FA substantially decreased caspase activities in PBMCs (Table 1), strongly suggesting that paraptosis was the preferred mode of cell death. In Thp-1 cells, however, FA significantly increased caspase −9 and −3/7 activities were (Table 2), indicating the induction of intrinsic apoptosis in Thp-1 cells. Percentage of protein expressions were represented as mean ± SD. *p < 0.05 relative to respective control; **p < 0.005 relative to respective control; ***p < 0.0001 relative to respective control.
Immune cells respond to stimuli by activating MAPK signaling to amplify other signals to elicit an appropriate physiological response for programmed cell death 14,38 . Previous studies showed that prolonged activation of MAPK signaling induced cell death via ROS-activation of MAPK signaling pathways 16 . ROS are continuously generated by cellular processes, with the mitochondrion being the major source 6 . Excessive ROS generated during oxidative phosphorylation can cause oxidative damage to proteins, DNA and phospholipids 16 ; oxidative degradation of lipids results in the formation of lipid peroxides such as MDA 39 . FA significantly increased MDA levels in both PBMCs and Thp-1 cells (Fig. 3A), indicative of a prolonged oxidative stress environment. Furthermore, FA disrupted mitochondrial membrane potential by increasing mitochondrial depolarization in Thp-1 cells (Fig. 3B). This could be due to the weak acidic nature of the carboxylic acid group of FA 40 . Weak acids act as proton carriers across lipid membranes, thereby disrupting the proton gradient along the electron transport chain (ETC) 41,42 . Interestingly, FA did not disrupt the mitochondrial membrane potential in normal PBMCs (Fig. 3B) despite the significant depletion of ATP levels in both normal PBMCs and Thp-1 cells (Fig. 3C). In Thp-1 cells, this is substantiated by the decreased cell viability (decreased redox potential) and increased mitochondrial membrane depolarization. Also, the activation of ATP dependent caspases −9 and −3/7 may further deplete ATP levels 43 . In PBMCs, however, the decreased ATP levels may be due to increased activation and prolonged activation of ATP dependent protein kinases.
Intracellular ROS not only alters cellular integrity but is also important to MAPK signaling cascades 6 ; FA induced increased ROS production and up-regulated protein expression of ERK in Thp-1 cells (Fig. 5C and D). Although ERK signaling pathways are well known for their role in promoting cell survival, recent studies have demonstrated their ability to potentiate apoptosis 44 . Prolonged activation of ERK may be due to the inhibition of tyrosine phosphatases, a group of enzymes responsible for the removal of phosphate groups on phosphorylated tyrosine residues, hence inactivating the protein 3 . However, tyrosine phosphatases are sensitive to increased ROS and become oxidized, thereby inhibiting their activity and prolonging ERK activation 3,45 .
JNK and p38 MAPK signaling pathways are generally directed towards initiating cell death upon activation by stress signals. Recently, however, these signaling pathways have been associated in both cell death and survival 38 . In Thp-1 cells, FA significantly decreased p-JNK activation and p-p38 protein expressions ( Fig. 5C and D). A study by Pedram et al. (1998) documented the cross-talk between the ERK and JNK MAPKs where the activation of JNK by ERK MAPK was followed by the activation of ERK by vascular endothelial growth factor (VEGF) whilst JNK stimulated ERKs proliferative signaling. Therefore, it can be inferred that a decrease in JNK activity hinders the cross-talk between JNK and ERK MAPKs, preventing survival signaling by ERK 46 .
Additionally, JNK and ERK MAPKs regulate the expression of Bcl-2 family proteins that are central in regulating the mitochondrial apoptotic death pathway 3,47 . Bcl-2 inhibits apoptosis by forming a complex with pro-apoptotic proteins such as Bax 47,48 . Phosphorylation of Bcl-2 compromises it's protein stability and affects dimerization with Bax 49 . Thus, dissociation from the complex at the mitochondrial membrane leads to the formation of mitochondrial permeability transition pore (mPTP) and subsequent caspase activation 50 . In support of the increased caspase −9 and −3/7 activities in Thp-1 cells, FA decreased p-Bcl-2 expression resulting in apoptotic cell death (Fig. 4C and D). JNK signaling regulates the expression of Bcl-2 and is up-regulated in response to JNK activation. FA activated ERK death signaling, decreased p-Bcl-2 expression and induced apoptosis in Thp-1 cells, but had no significant effect on Bax expression ( Fig. 4C and D). This may be due to the deletion of the p53 gene in the Thp-1 cell line as p53 acts as a transcription factor for Bax expression and recruitment to the mitochondrial membrane 24, 51-55 . Additionally, increased ERK activity regulates mitochondrial membrane potential 56 and corresponds with the increased caspase −9 and −3/7 activities, and the subsequent activation of cell death in Thp-1 cells by FA. In PBMCs, FA increased p-Bcl-2 expression ( Fig. 4A and B) and decreased Bax expression ( Fig. 4A and B), with a corresponding decrease in caspase −8, −9 and −3/7 activities. Increased p-Bcl-2 expression helps maintain the mitochondrial membrane integrity and subsequent mitochondrial membrane potential by preventing the release of cytochrome c, activation of caspase −9 and the initiation of intrinsic apoptosis, further validating the induction of paraptosis in normal PBMCs by FA.
Although the molecular activation of paraptosis remains unknown, studies have suggested the involvement of MAPK signaling in the induction of cell death. In PBMCs, FA significantly increased the expression of ERK and JNK whilst decreasing p38 expression (Fig. 5A and B). Sperandio et al., (2000) reported that ERK and JNK activity mediated paraptosis stimulation by insulin-like growth factor 1 receptor, and that inhibition of these MAPKs prevented the induction of paraptosis in 293 T cells 57 . Another study by Yumnam et al., (2014) showed the involvement of ERK MAPK in hesperidin-induced paraptosis of human hepatocellular carcinoma (HepG2) cells 58 . Sugimori and colleagues (2015) recently showed that activated JNK induced paraptosis induction in HL-60 and NB4 human promyelocytic leukemic cell lines and in bone marrow blasts treated with benfotiamine 59 . Contrary to the studies by Sperandio et al., (2000) and Yumnam et al., (2014), benfotiamine inhibited the activity of ERK in bone marrow blasts and had no effect on ERK activity in HL-60 and NB4 cell lines 57,58 . This suggests that the involvement of MAPK in the induction of paraptosis may be dependent on the cell line and type of activation. Additionally, caspase −9 was reported to be a direct target of ERK MAPK, and that phosphorylation at threonine 125 on caspase −9 inhibits its pro-apoptotic activity 60 . These findings support the activation of MAPK signaling pathways in the induction of paraptosis in PBMCs treated by FA.

Conclusion
FA is immunotoxic to both healthy PBMCs and Thp-1 cells, albeit at a higher concentration in PBMCs. The cancerous Thp-1 cells are highly susceptible to FA toxicity. Collectively, the results show that the host response to FA exposure augmented MAPK signaling and induction of apoptosis in Thp-1 cells (via the mitochondrial apoptotic pathway) and paraptosis in PBMCs. This study shows that FA, a common food borne mycotoxin, is toxic to the human immune system. This data may help develop a better understanding of the immune risks associated with FA consumption. This has great importance in socio-economically challenged countries where the majority population relies on corn as a food staple.

Material and Methods
Materials. Cell culture reagents for PBMC maintenance and FA (Gibberella fujikuroi) were purchased from Sigma Aldrich (Johannesburg, SA). The Thp-1 cells and media were purchased from ATCC (University Boulevard Manassas, USA) and Scientific group (Johannesburg, SA), respectively. Luminometry reagents were obtained from Promega (Madison, USA).
Cell viability. The cytotoxicity of FA on PBMCs and Thp-1 cells was analyzed using the WST-1 reagent.
Briefly, PBMC and Thp-1 cells (20,000 cells/well) were seeded into a 96-well microtitre plate. The cells were incubated with varying FA concentrations (30-300 µg/ml) in triplicate (200 µl/well) for 24 h (37 °C, 5% CO 2 ). Ellipticine, anti-neoplastic agent, was used as a positive control to FA toxicity (Data shown in Supplementary Tables: S1 PBMC; S2 Thp-1). A positive control of cells with RPMI only and a negative control with RPMI/WST-1 reagent solution was also seeded. Following incubation, the plate was centrifuged at 24 °C, 400xg for 10 min. The supernatant was then aspirated and 110 µl/well of a RPMI/WST-1 reagent solution (1:10) was added and incubated for 3 h (37 °C, 5% CO 2 ). The optical density of the colorimetric reaction was measured at a wavelength of 450 nm and reference wavelength of 620 nm using a spectrophotometer (Bio-Tek uQuant, Winooski, VT, USA). The percentage cell viability was calculated by standardizing untreated (control) cells to 100% and then comparing FA treated cells to the control cells (Detailed calculation shown as Supplementary Information). The concentration of half maximum inhibition (IC 50 ) was determined using GraphPad Prism v5.0 software. All assays were performed in triplicate, twice independently. LDH activity. The LDH cytotoxicity detection kit (Roche, Mannheim, Germany) was used to confirm damaged/ necrotic cells. In brief, cell homogenates (100 μl) were seeded into a 96-well opaque polystyrene microtitre plate in triplicate. Subsequently, 100 μl of a substrate mixture containing a catalyst (diaphorase/NAD + ) and dye solution (INT/sodium lactate) was added to each homogenate and incubated in the dark for 25 min (RT). The optical density was measured (500 nm) using a spectrophotometer (Bio-Tek uQuant, Winooski, VT, USA). The results were reported as mean optical density.
Evaluation of oxidative stress. Lipid peroxidation. The TBARS was used to determine FA generation of ROS 61 . Briefly, cell homogenates (400 μl) were added to a 7% phosphoric acid (400 μl) and a thiobarbituric acid (1%W w/v)/butylated hydroxytoluene (0.1 mM) (TBA/BHT) solution. A positive control containing MDA (1 μl) and a negative control containing 3 mM hydrogen chloride were prepared. All samples were heated in a water bath (100 °C, 15 min) and allowed to cool (RT). Thereafter, lipids were extracted with butanol (1.5 ml) and were measured on a spectrophotometer (Bio-Tek uQuant, Winooski, VT, USA) at 532 nm with reference wavelength of 600 nm. The mean optical density for each sample was calculated and divided by the absorption coefficient (156 mM −1 ). The results were expressed in μM.
Mitochondrial membrane potential. Mitochondrial membrane potential was measured using the JC-1 Mitoscreen kit (BD Biosciences, Johannesburg, SA) and flow cytometry. Briefly, 100 µl of a JC-1 working solution was added to each sample (200,000 cells in 100 µl PBS) and incubated in the dark for 30 min (RT). Following incubation, 100 µl flow cytometry sheath fluid was added to each sample and were analyzed on the AccuriTM C6 flow cytometer. A total of 20 000 events were gated using AccuriTM C6 flow cytometer Fl-1 channel (525 nm) (BD Biosciences, Johannesburg, SA). The results were expressed as a percentage.
ATP levels. Intracellular ATP levels were measured using the ATP CellTitre Glo reagent (Promega, Madison, USA). Following treatment, 20,000 cells/well were seeded into a 96-well opaque polystyrene microtitre plate in triplicate. The reagent (20 μl/well) was added to each sample and incubated in the dark for 30 min (RT). Thereafter, the luminescence was measured on a ModulusTM microplate luminometer (Turner Biosystems, Sunnyvale, USA). The data was expressed as relative light units (RLU).
Statistical analysis. Statistical analysis was performed using GraphPad Prism v5.0 software (GraphPad Software Inc., La Jolla, USA). GraphPad Prism Software was used for the unpaired t-test with Welch's correction to assess the differences between samples. Level of significance (p) was established at a p < 0.05.