Modulation of Mucin mRNA (MUC5AC and MUC5B) Expression and Protein Production and Secretion in Caco-2/HT29-MTX Co-cultures Following Exposure to Individual and Combined Fusarium Mycotoxins

Abstract

Intestinal epithelial cells (IECs) are a critical component of the innate local immune response. In order to reduce the risk of pathogen infection or xenobiotic intoxication, different host defense mechanisms have been evolved. Evidence has shown that upon ingestion of food or feed contaminated with toxins (e.g., mycotoxins), IECs respond by regulating mucin secretions, which act as a physical barrier inhibiting bacterial attachment and subsequent infection-related processes. However, the effect of Fusarium mycotoxins on mucin production remains unclear. Consequently, the aim of this study was to evaluate individual and interactive effects of four common Fusarium mycotoxins, deoxynivalenol, nivalenol, zearalenone, and fumonisins B1 on mRNA expression and secretion of mucins, MUC5AC, and MUC5B, as well as total mucin-like glycoprotein secretion, using Caco-2 (absorptive-type) and HT29-MTX (secretive-type) cells and their co-cultures (initial seeding ratios Caco-2/HT29-MTX: 90/10 and 70/30). Our results showed that individual and mixtures of mycotoxins significantly modulated MUC5AC and MUC5B mRNA and protein, and total mucin-like glycoprotein secretion as measured by quantitative polymerase chain reaction, enzyme-linked immunosorbent assay, and enzyme-linked lectin assay, respectively. Additive effects were not always observed for mixtures. Also, the present study showed that in co-cultures, lower MUC5AC and MUC5B mRNA, protein and total mucin production occurred following exposure, which might suggest higher intestinal permeability and susceptibility to toxin exposure. This study demonstrates the importance of selecting an appropriate cell model for the in vitro investigation of Fusarium mycotoxin effects either alone or in combinations on the immunological defense mechanisms of IECs, and will contribute to improved toxin risk assessments.

deoxynivalenol, nivalenol, zearalenone, fumonisin B1, mucins, intestinal epithelia

The most important Fusarium mycotoxins from a human and animal health perspective are fumonisin B1 (FB1), zearalenone (ZEA), and trichothecenes such as nivalenol (NIV) and deoxynivalenol (DON). Fusarium fungi are commonly found on cereals grown in the temperate areas of America, Europe, and Asia. Fusarium mycotoxins elicit a wide spectrum of toxic effects in animals (Creppy, 2002), including the capacity to modify normal immune functions, though this role in human disease is less clear.

Although the intestine plays an important role in digestion and absorption of nutrients, the mucus lining the epithelium (also called the gastrointestinal glycocalyx) represents a pivotal barrier against ingested pathogens and toxins. Goblet cells lie scattered in both the crypts and among enterocytes, and respond to various stimuli, including food toxins (e.g., mycotoxins), pathogens, irritant gases, reactive oxidative species, and/or other changes in the microbiophysical environment (Smirnova et al., 2003). These stimuli are known to modulate mucin production and secretion. In general, mucins tightly adhere to the apical surfaces of the gastrointestinal epithelium through the formation of a complex between mucin oligosaccharides and a mucin binding protein, creating a continuous gel layer (Bouhet and Oswald, 2005). Currently, there are limited data regarding the effect of mycotoxins on mucus production. For example, following a single exposure to the mycotoxin, ochratoxin A (OTA), in Swiss mice, OTA was observed in the squamous layer of the oesophagus, in the surface mucus and within the cytoplasm of duodenal and jejuna epithelial cells (Lee et al., 1984). Additionally, broiler chicks receiving feed containing FB1 at 300 mg/kg ad libitum for 2 weeks developed intestinal goblet cell hyperplasia (Brown et al., 1992).

It is common to detect several types of Fusarium mycotoxins in the same crop, although it is rare to find a crop contaminated with only single mycotoxin (Tajima et al., 2002). Thus, it is important to evaluate the impact of such interactions on potential toxicological targets, including mucin synthesis and secretion. Previous reports have shown that exposure to mycotoxins could result in increased susceptibility to experimental or natural mucosal infections by inducing bacterial translocation and colonization across the intestinal epithelium (Oswald et al., 2003). However, no data are available concerning the mechanisms associated with mycotoxin-related effects on intestinal infections. Evidence has shown that the mucus layer lining the intestinal epithelium serves as a physical barrier, preventing microorganisms, including pathogenic agents from translocating and colonizing in the intestinal environment, and thus controlling/limiting detrimental mucosal infections (Katayama et al., 1997). Accordingly, we hypothesized that a mixture of Fusarium toxins that are naturally observed in food/feed may exert significant effects on the synthesis and secretion of mucins when compared with their effects as individual toxins. Thus, this combination of toxins may influence bacterial translocation and susceptibility to mucosal infections in intestinal epithelial cells (IECs). This study aimed to evaluate the individual and interactive effect of four common Fusarium mycotoxins, DON, NIV, ZEA, and FB1, on the mRNA expression and secretion of secretory mucins, MUC5AC and MUC5B, as well as total mucin-like glycoprotein secretion using cell culture models. The Caco-2 cell line, derived from human colon adenocarcinoma, represents the most commonly used model, which, when grown to confluence, differentiates spontaneously to form and acquire structural and biochemical properties of small intestinal enterocytes (Hilgendorf et al., 2000). HT29-MTX cells are a homogeneous subpopulation of HT29 human colon carcinoma cells selected by adaptation to 10⁻⁵ M methotrexate, which additionally produce mucins, in particular MUC2, MUC5AC, and MUC5B, with a predominant expression of MUC5AC (Laparra and Sanz, 2009). Thus, HT29-HTX can be regarded as providing a similar function to goblet cells. In the human small intestine, the percentage of goblet cells varies in the gastrointestinal tract from approximately 10% in the small intestine to 24% in the distal colon (Mahler et al., 2009). In this study, both individual cell culture models, Caco-2 and HT29-MTX, and co-cultures of Caco-2 and HT29-MTX with ratios of 90/10 and 70/30 were used in attempts to compare the cell line models and to provide a culture model that may better represent some of the cell diversity of the intestinal tract, by preservation of a significant barrier with a complete mucus layer (Hilgendorf et al., 2000). Because the stage of confluence and extent of differentiation of HT29-MTX cells might affect the secretion of mucus, cells in highly proliferative stages were used, which mimic the proliferative cells of the invaginated crypts (Smirnova et al., 2003).

MATERIALS AND METHODS

Chemicals

All mycotoxins (DON, NIV, ZEA, FB1), dimethyl sulfoxide (DMSO), phosphate buffered saline (PBS), RIPA buffer, and avidin-peroxidase conjugate were obtained from Sigma Chemical Company (St Louis, MO). DON, ZEA, and FB1 were dissolved in DMSO and NIV was dissolved in ethanol. Dulbecco's modified eagle medium (DMEM), fetal bovine serium, and thiazolyl blue tetrazolium bromide (MTT) were purchased from Gibco-Life Technology (Eggenstein, Germany). RNAiso Plus, PrimeScript RT Reagent Kit (Perfect Real Time), Premix Ex Taq (Perfect Real Time), primers, and fluorogenic probes were purchased from Takara (Dalin, China). Absolute ethanol, sodium bicarbonate, sodium carbonate, and concentrated sulphuric acid were from Merck (Darmstadt, Germany). Tween 20 and horseradish peroxidase-conjugated goat anti-mouse IgG were from Bio-Rad Laboratories (Richmond, CA). Mouse monoclonal anti-MUC5AC (ab24070) antibody was obtained from Abcam (Cambridge, UK). Goat polyclonal anti-MUC5B (sc-23024) and horseradish peroxidase-conjugated donkey anti-goat IgG antibodies were from Santa Cruz (Santa Cruz, CA). 3,3′,5,5′-tetramethylbenzidine peroxidase (TMB) substrate reagent set was purchased from BioLegend (San Diego, CA). Biotinylated wheat germ agglutinin was purchased from Vector Labs (Burlingame, CA). PBS containing 0.05% Tween 20 (PBS-Tw) was used in enzyme immunoassays.

Inscribed central composite design for complete mixtures of Fusarium mycotoxins

Screening for interactive effects between chemicals in a defined chemical mixture requires statistical designs to optimize clarity and manageability of the toxicological experiments. Consequently, an inscribed central composite design (Heussner et al., 2008), including a fractional factorial part, was applied with four factors (i.e., DON, NIV, ZEA, and FB1) in order to minimize the number of possible toxin combinations from 44 (all possible combinations of every concentration of each toxin) to 16. The design matrix is described in Table 1.

Design Matrix from Four Fusarium Mycotoxins

TABLE 1

Design Matrix from Four Fusarium Mycotoxins

	DON (2μM)	NIV (2μM)	ZEA (40μM)	FB1 (40μM)
1	0	0	0	0
2	1	0	0	0
3	0	1	0	0
4	0	0	1	0
5	0	0	0	1
6	1	1	0	0
7	1	0	1	0
8	1	0	0	1
9	0	1	1	0
10	0	1	0	1
11	0	0	1	1
12	1	1	1	0
13	1	1	0	1
14	1	0	1	1
15	0	1	1	1
16	1	1	1	1

Note. 1—with mycotoxin, 0—without mycotoxin.

Concentrations (i.e., 2μM of DON or NIV and 40μM of ZEA or FB1) chosen for the interactive experiments were based on our previous experiments in IPEC-J2 cells, which impact on cytotoxicity (Wan et al., 2013a), proinflammtory cytokines (Wan et al., 2013b), and antimicrobial peptides (Wan et al., 2013c) effects were investigated. Selected levels of respective mycotoxins reflect concentrations expected to be encountered in the gastrointestinal tract of animals or human tissues following consumption of food or feed contaminated with 2–4 mg/kg of body weight of these mycotoxins (Kouadio et al., 2007).

Cell line and culture conditions

Caco-2 cells (passages 33–45) obtained from the ATCC (HTB-37) were maintained at 37°C, 5% CO₂, 90% relative humidity in complete medium consisting of DMEM with 10% foetal bovine serum. Routinely, cells were sub-cultured once a week using trypsin-EDTA (0.25%, 0.53mM) and seeded at a density of 4 × 10⁵ cells per 75 cm² flask. Medium was changed every other day. HT29-MTX cells were donated by Dr. Thecla Lesuffleur (INSERM U178, Villejuif, France). The cells (passage 37–50) were used. For Caco-2/ HT29-MTX co-cultures, the initial seeding cell density was adopted from previous experiments (Calatayud et al., 2012; Hilgendorf et al., 2000). Caco-2 and HT29-MTX cells were grown separately and pre-determined cell numbers of each type were mixed prior to seeding to yield cell ratios of 100/0, 90/10, 70/30, and 0/100 for Caco-2 to HT29-MTX cells, respectively. Although the final proportion of Caco-2 and HT29-MTX cells was not determined by microscopy or by staining, it has previously been reported by other investigators that co-culturing Caco-2 and HT29-MTX cells under standard conditions result in a viable and relative proportionate mixture of cells from both cell lines (Chen et al., 2010), though neither Chen's nor our group formally measured the ratios of cell types at confluence. Moreover, it has also been reported recently that the co-culture ratios could be slightly changed after cell differentiation due to slightly different but comparable doubling times of the two cell lines, and that the ratio of 90/10 was found to be the most physiologically relevant (Araújo and Sarmento, 2013). Yet, this study adds new data to existing knowledge about the effect of mycotoxins on mucus production and shows that the co-culture ratios are decisive in the evaluation of this effect. All cells were screened for mycoplasma contamination with a MycoAlert mycoplasma detection kit (Lonza, Basel, Switzerland) prior to use. Each set of experiments used all four co-culture conditions (100/0, 90/10, 70/30, 0/100) with cells maintained under identical conditions as the stem cultures; medium was changed every other day. Cells were confluent within 5–6 days and were treated with toxins at day 7 or 14 using conditions described below for each assay.

MTT assay

The MTT assay was performed to assess cell viability following exposure to single and combinatorial toxin treatments. In brief, Caco-2/HT29-MTX cells (100/0, 90/10, 70/30, and 0/100) were seeded (4.8 × 10⁴ cells/well) in 96-well plates (Costar, Corning, NY) and allowed to adhere for 24 h. Cells were allowed to differentiate that took 7 days (Caco-2/HT29-MTX (0/100)) or 14 days (Caco-2/HT29-MTX (100/0, 90/10, and 70/30)) post seeding. Cells were rinsed with PBS and treated with Fusarium mycotoxins at the concentrations reported previously (Wan et al., 2013a) in serum-free media for 48 h. Following this, 10 μl/well of MTT solution (5 mg/ml in PBS) was added and 2 h later the media discarded. One hundred microliter per well of DMSO was added with shaking for five minutes to solubilize the formazan formed by viable cells. Absorbance was measured at 595 nm using an iMark Microplate Reader (Bio-Rad Laboratories). The viability of cells after treatment with Fusarium mycotoxins was expressed as a percentage compared with the control (nonmycotoxin treated cells). Experiments were repeated four times with six replicates for each treatment.

Quantitative polymerase chain reaction analysis of MUC5AC, MUC5B and 18S gene expression

Caco-2/HT29-MTX cells were seeded at 5 × 10⁴ cells/cm² in 6-well culture plates (Costar) and allowed to adhere for 24 h. Confluent cells at days 7 or 14 (as described for MTT assay) were washed with PBS and treated with Fusarium mycotoxins in serum-free media for 48 h. Total RNA was extracted using RNAiso Plus according to the manufacturer's instructions. Purified RNA was resuspended in 100 μl of nuclease-free water and stored at −80°C. RNA concentrations were measured using a NanoDrop ND-1000 Spectrophotometer (Nano-Drop Technologies, Wilmington, DE). Prior to use in quantitative polymerase chain reaction (qPCR), RNA quality was determined by ensuring a value of > 1.8 but < 2.2 for the A₂₆₀/A₂₈₀ ratio. cDNA was prepared from 1 μg of total RNA using the PrimeScript RT Reagent Kit (Perfect Real Time) according to the manufacturer's instructions. qPCR was performed to quantify mRNA transcript levels for MUC5AC and MUC5B relative to the 18S ribosomal RNA expression levels. Because the expression of 18S ribosomal RNA levels has previously been shown to be stable in Caco-2, HT29-MTX, and Caco-2/HT29-MTX co-cultures (Laparra et al., 2009), we selected this marker for quality analysis by gel electrophoresis and used it as an internal standard for mRNA quantification, as previously reported (Thellin et al., 1999). For all experiments, no differences in 18S ribosomal RNA amount between groups were observed for all genes investigated (data not shown). All samples were run on a StepOnePlus Real-Time PCR system (Applied Biosystems, Foster City, CA) using 1 μl of cDNA and Premix Ex Taq (Perfect Real Time), with final primer concentrations of 0.2μM per primer in a total volume of 20 μl. Human-specific mucin primers and fluorogenic probes (Table 2) were designed by and purchased from Takara. Samples were centrifuged briefly and thermocycled using the default fast program (1 cycle of 95°C for 30 s and 40 cycles of 95°C for 5 s, 60°C for 10 s). To ensure the reliability of qPCR data, amplicons were designed to be < 250 bp (Nolan et al., 2006). All PCR reactions were performed in duplicate. Relative changes in gene expression levels of MUC5AC and MUC5B in cultured enterocytes resulting from mycotoxin treatments were normalized against 18S using the 2^−ΔΔCT method as described previously (Livak and Schmittgen, 2001). Experiments were repeated two times independently with each treatment performed in triplicate.

Primer Sequences for Quantification of MUC5AC, MUC5B, and 18S by qPCR

TABLE 2

Primer Sequences for Quantification of MUC5AC, MUC5B, and 18S by qPCR

Primer set	Product length (bp)	Forward primer sequence (5′-3′)	Reverse primer sequence (5′-3′)	Probe sequence (5′FAM-3′)	Accession number
MUC5AC	131	CTCCTACCAATGCTCTGTA	GTTGCAGAAGCAGGTTTG	CCACAGAGCTGGATGCCACAGA	AJ001402.1
MUC5B	154	GACAGAGACGACAATGAG	CCTGATGTTTTCAAAAGTTTC	ACTAACACCACCACCAGCCA	NM_002458.2
18S	88	TGGCTCATTAAATCAGTTATG	CGGCATGTATTAGCTCTA	CGCTCGCTCCTCTCCTACTTG	M10098

Indirect enzyme-linked immunosorbent assay of MUC5AC and MUC5B

Cell cultures were prepared and treated with Fusarium mycotoxins as described above. Following 48 h of toxin incubation, cell lysates were prepared by incubating the cells with RIPA buffer on ice for 5 minutes, followed by centrifugation at 8000 g for 10 minutes at 4°C. Cell culture supernatants were collected and stored at −80°C until required for subsequent analyses. Total mucin production consists of both mucins in the cell lysate and mucins secreted into cell culture supernatant. Therefore, total mucin protein production was calculated as the sum of mucin proteins in the cell lysate and cell culture supernatant. Levels of both mucin proteins in cell lysate and supernatant were measured by indirect enzyme-linked immunosorbent assay (ELISA) as described previously (Kohri et al., 2002). Briefly, 50 μl of cell lysates or cell supernatants were diluted in RIPA buffer and incubated with bicarbonate-carbonate buffer (50 μl) at 40°C in a 96-well ELISA plate (Nunc MaxiSorb; Nalge Nunc International, Rochester, NY) until dry. Plates were blocked with 2% bovine serum albumn (BSA) for 2 h at room temperature and then incubated with 50 μl of mouse anti-MUC5AC (1:250) and goat anti-MUC5B (1:500) antibodies in PBS-Tw. After 2 h, 50 μl of either HP-anti mouse-IgG or HP-anti goat-IgG, respectively, at 1:500 in PBS-Tw were added to each well. Following a 2 h incubation, color reactions were developed with TMB substrate solution and stopped with 2N sulphuric acid. Absorbance was read at 450 nm using an iMark Microplate Reader. Experiments were repeated three times independently with each treatment performed in triplicate.

Enzyme-linked lectin assay for total mucin glycoprotein secretion

The total production of mucins consists of both mucins in the cell lysate and mucins secreted into the cell culture supernatant. Therefore, total production of mucin-like glycoprotein was calculated as the sum of mucin proteins present in cell lysate and in cell culture supernatant. Cell lysates and supernatants from co-cultures were collected as described above and assayed for total mucin secretion using the enzyme-linked lectin assay (ELLA) as previously described with modifications (Hewson et al., 2004). Briefly, 50 μl of cell lysates or cell supernatant were diluted in RIPA buffer and incubated with bicarbonate-carbonate buffer (50 μl) at 40°C in a 96-well plate until dry. Plates were blocked with 100 μl 2% BSA in PBS-Tw for 2 h at room temperature. Biotinylated wheat germ agglutinin (1:500) in 100 μl PB-Tw was added and samples were incubated for 2 h at room temperature. Plates were then incubated with 100 μl avidin-peroxidase conjugate (1:100,000) for 1 h. Colorimetric determination was performed as described above for the ELISA. Experiments were repeated three times independently with each treatment performed in triplicates. To ensure the stability of mucin protein in the supernatant, the cell supernatant, after collected, was frozen and stored at −80°C until analyses. The ELLA has been extensively used previously by other investigators to measure the total mucin protein secreted into the supernatant (Hewson et al., 2004). Under similar experimental conditions, it is confident that the gross amounts of mucus accumulation in the supernatants from control and treatment groups were accurately measured and compared as per the current state of our knowledge and art.

Statistical analysis

The results for cytotoxicity, qPCR, ELISA, and ELLA were expressed as mean ± standard error of mean (SEM) of two or three individual experiments. All data analyses were performed using the SPSS statistical package (SPSS Version 20.0 for Windows; SPSS Inc, Chicago, IL). Data were first evaluated for normality with the Shapiro–Wilk and Levene's variance homogeneity test. One-way analysis of variance (ANOVA) with the Kruskal-Wallis test, followed by the Mann-Whitney U test, was used to identify significant differences for nonparametric data. One-way ANOVA with Dunnett's multiple comparison test was used for analyzing parametric data. Differences were considered to be statistically significant when p-values were less than 0.05. Univariate analyses of variance were performed to determine if there were any associations between different toxin treatments and total mucin-like glycoprotein secretion. Effects with p-values less than 0.05 at the 95% confidence interval were regarded as significant, suggesting potential interactive effects (either synergistic or less than additive) of different Fusarium mycotoxins detected in the total mucin-like glycoprotein secretion, whereas effects with p > 0.05 at the 95% confidence interval were considered as not significant. A lack of interaction indicates that the effects are additive (i.e., combined effects would be the sum of their individual effects) (Boone, 2008). Correlations between MUC5AC and MUC5B gene expression, intracellular (cell lysate) and extracellular (supernatant) protein levels, and total mucin-like glycoprotein secretion of Caco-2/HT29-MTX (100/0, 90/10, 70/30, and 0/100) treated with individual and mixtures of DON, NIV, ZEA, and FB1 were assessed by Spearman's correlations (nonparametric).

RESULTS

Effects of Individual and Mixtures of Fusarium Mycotoxins on Caco-2/HT29-MTX (100/0, 90/10, 70/30, and 0/100) Cell Viability

In the Caco-2 monoculture, significant increases in cell viability were observed for treatment with DON, ZEA, and FB1 alone (Fig. 1A), whereas in the HT29-MTX monoculture significant increases in cell viability were observed in NIV, ZEA, FB1 treatments with decreases occurring in DON alone and the mixture of DON-NIV-FB1 (Fig. 1D). In the Caco-2/HT29-MTX 90/10 co-cultures, significant increases in cell viability were only observed for NIV and ZEA alone (Fig. 1B). In the 70/30 co-cultures, NIV increased the cell viability whereas mixtures of DON-FB1, NIV-ZEA, NIV-FB1, ZEA-FB1, DON-NIV-ZEA, DON-NIV-FB1, DON-ZEA-FB1, NIV-ZEA-FB1, and DON-NIV-ZEA-FB1 significantly decreased viability (Fig. 1C).

FIG. 1.

Effects of individual and mixtures of DON, NIV, ZEA, and FB1 on viability of Caco-2/HT29-MTX (A) 100/0, (B) 90/10, (C) 70/30, and (D) 0/100 co-cultures. Differentiated Caco-2/HT29-MTX cells were exposed to DON (2μM), NIV (2μM), ZEA (40μM), or FB1 (40μM) for 48 h. Cell viability was determined by MTT test. Results are expressed as percentage of control and are means ± SEM (n = 4). *, **, *** p < 0.05, 0.01, and 0.001, respectively, compared with control.

A heat map depicting the percentage of viability of Caco-2/HT29-MTX co-cultures (100/0, 90/10, 70/30, and 0/100) following exposure to individual or combinations of DON, NIV, ZEA, and FB1 is shown in Supplementary 1. Comparing the viability for the Caco-2/HT29-MTX co-cultures (100/0, 90/10, 70/30, and 0/100), it was shown that the viability of cells was generally higher when cells were not treated with any mycotoxins, and the addition of increasing number of mycotoxins (i.e., mixtures) results in a reduction in viability but significant reduction was observed in the 70/30 co-cultures.

Effects of Individual and Mixture of Fusarium Mycotoxins on MUC5AC and MUC5B mRNA Expression

qPCR was performed to determine the effects of individual and mixtures of Fusarium mycotoxins, DON, NIV, ZEA, and FB1 on the expression of mucin genes (MUC5AC and MUC5B) on Caco-2/HT29-MTX co-cultures. The results showed that individual and mixtures of DON, NIV, ZEA, and FB1 significantly affected mRNA expression (Fig. 2). MUC5AC mRNA levels were mostly upregulated in the Caco-2 monocultures upon treatment with DON, NIV, ZEA, and FB1 alone and in most combinations with the exception of the mixture of DON-NIV and DON-ZEA. MUC5B mRNA levels, on the other hand, were only significantly upregulated for treatments with FB1 alone and in mixtures of DON-NIV-ZEA, DON-NIV-FB1, DON-ZEA-FB1, NIV-ZEA-FB1, and DON-NIV-ZEA-FB1 (Fig. 2A).

FIG. 2.

Relative abundance of MUC5AC and MUC5B mRNA from Caco-2/HT29-MTX (A) 100/0, (B) 90/10, (C) 70/30, and (D) 0/100 co-cultures isolated at 48 h after individual or combinations of DON (2μM), NIV (2μM), ZEA (40μM), and FB1 (40μM) exposure. Results are expressed as a percentage of the control and are means ±SEM (n = 2). *, **, *** p < 0.05, 0.01, and 0.001, respectively, compared with control.

In the HT29-MTX monoculture, significant downregulation of MUC5AC mRNA levels was observed in DON, NIV, and ZEA alone, and in combinations of DON-NIV, DON-ZEA, DON-FB1, NIV-ZEA, NIV-FB1, and DON-NIV-ZEA but significant upregulation of MUC5AC was found in DON-ZEA-FB1 and DON-NIV-ZEA-FB1. Relative levels of MUC5B mRNA were significantly downregulated in NIV, DON-NIV, DON-ZEA, DON-FB1, NIV-ZEA, and DON-NIV-ZEA but were upregulated in DON-NIV-FB1, DON-ZEA-FB1, NIV-ZEA-FB1, and DON-NIV-ZEA-FB1 (Fig. 2D).

In the 90/10 co-cultures, ZEA and FB1 alone and the mixtures of ZEA-FB1 and DON-NIV-ZEA-FB1 significantly upregulated MUC5AC mRNA levels, whereas NIV alone and the mixtures of DON-ZEA, DON-FB1, NIV-ZEA, DON-NIV-ZEA, and DON-ZEA-FB1 led to significant downregulation. The mRNA expression of MUC5B was significantly upregulated by ZEA and FB1 alone and the mixtures of DON-NIV, DON-FB1, ZEA-FB1, DON-NIV-FB1, DON-ZEA-FB1, and DON-NIV-ZEA-FB1 but was downregulated by NIV, DON-ZEA, and NIV-ZEA (Fig. 2B). In 70/30 co-cultures, significant downregulation of MUC5AC mRNA expression was observed in DON, NIV, and FB1 alone and in any combination of DON, NIV, ZEA, and FB1, with the exception of the mixture of ZEA-FB1. MUC5B mRNA levels were downregulated for treatments with NIV and FB1 alone and the mixtures of DON-NIV, DON-ZEA, DON-FB1, NIV-ZEA, DON-NIV-FB1, and DON-ZEA-FB1, but were upregulated by DON and ZEA alone and the mixtures of ZEA-FB1, DON-NIV-ZEA, NIV-ZEA-FB1, and DON-NIV-ZEA-FB1 (Fig. 2C).

A heat map depicting the changes in mucin gene expression for Caco-2/HT29-MTX co-cultures (100/0, 90/10, 70/30, and 0/100) following exposure to individual or combinations of DON, NIV, ZEA, and FB1 is shown in Supplementary 2. Caco-2/HT29-MTX co-cultures (100/0, 90/10, 70/30, and 0/100) demonstrated a unique mucin expression profile following exposure to different mycotoxins. Comparison of the both MUC5AC and MUC5B profiles for the Caco-2/HT29-MTX co-cultures (100/0, 90/10, 70/30, and 0/100) shows similarity in their expression following exposure to DON, NIV, ZEA, and FB1. The mRNA expression of MUC5AC and MUC5B were relatively low when the cells were not treated with any mycotoxins. However, the addition of increasing number of mycotoxins induced MUC5AC and MUC5B mRNA expression. Caco-2 and HT29-MTX monocultures treated with mycotoxins showed higher MUC5AC and MUC5B mRNA expression compared with their co-cultures (90/10 and 70/30).

Effects of Individual and Mixture of Fusarium Mycotoxins on MUC5AC and MUC5B Protein Expression of Caco-2/HT29-MTX (100/0, 90/10, 70/30, and 0/100)

To evaluate whether the modulation of mucin mRNA abundance exhibited translated to altered mucin protein abundance, the relative levels of MUC5AC and MUC5B in cell lysates and those secreted into culture supernatants were measured by indirect ELISA (Fig. 3).

FIG. 3.

Relative levels of MUC5AC and MUC5B protein in cell lysates and supernatants from Caco-2/HT29-MTX (A) 100/0, (B) 90/10, (C) 70/30, and (D) 0/100 co-cultures isolated at 48 h after individual or combinations of DON (2μM), NIV (2μM), ZEA (40μM), and FB1 (40μM) exposure. Results are expressed as a percentage of the control and are means ±SEM (n = 3). *, **, *** p < 0.05, 0.01, and 0.001, respectively, compared with control.

In Caco-2 monocultures, treatment with DON, DON-ZEA, and DON-FB1 significantly increased MUC5AC levels in cell lysates, whereas treatment with NIV, FB1, DON-NIV, NIV-FB1, ZEA-FB1, DON-ZEA-FB1, NIV-ZEA-FB1, and DON-NIV-ZEA-FB1 downregulated MUC5AC levels. Significant decreases in the MUC5AC levels in cell supernatants were observed in Caco-2 cells following treatment with ZEA, FB1, DON-NIV, DON-ZEA, and DON-NIV-ZEA, though significant increases resulted following treatment with NIV-ZEA and ZEA-FB1. As for the MUC5B protein levels, significant increases were found in cell lysates following exposure to DON, NIV, DON-ZEA, DON-FB1, NIV-ZEA, DON-NIV-ZEA, and DON-NIV-ZEA-FB1. Similarly, significant increases in MUC5B protein levels were also observed in cell supernatants following exposure to most of the toxin mixtures except DON, ZEA, and FB1 alone (Fig. 3A). When the HT29-MTX monoculture was used, significant increases in MUC5AC protein levels were observed in cell lysates following the incubation of cells with DON, NIV, FB1, DON-ZEA, DON-FB1, NIV-FB1, DON-NIV-ZEA, DON-NIV-FB1, NIV-ZEA-FB1, and DON-NIV-ZEA-FB1, but decreases were found in DON-NIV treated culture. For the MUC5AC protein levels in cell supernatants, significant decreases were noted in most of the toxin mixtures except ZEA, FB1, and ZEA-FB1. MUC5B levels in cell lysates, on the other hand, were only significantly increased in cells following treatment with DON, but were significantly decreased following treatments with ZEA and DON-NIV-ZEA-FB1. However, for the MUC5B levels in cell supernatants, ZEA, DON-NIV, DON-FB1, NIV-FB1, ZEA-FB1, DON-NIV-ZEA, DON-NIV-FB1, DON-ZEA-FB1, and DON-NIV-ZEA-FB1 significantly increased protein levels, whereas FB1 and NIV-ZEA-FB1 significantly decreased them (Fig. 3D).

In the 90/10 co-cultures, FB1 and ZEA-FB1 significantly increased the MUC5AC protein levels in cell lysates and NIV, DON-NIV, NIV-ZEA, NIV-FB1, and NIV-ZEA-FB1 significantly decreased it. For MUC5AC levels in cell supernatants, NIV, DON-NIV-ZEA, DON-NIV-FB1, DON-ZEA-FB1, NIV-ZEA-FB1, and DON-NIV-ZEA-FB1 significantly increased the levels, but ZEA, FB1, and NIV-ZEA led to decreased levels. Conversely, no changes in the MUC5B protein levels in the cell lysates were observed following exposure to individual or mixtures of DON, NIV, ZEA, and FB1 but significant increases in the MUC5B protein supernatant levels were observed in cells exposed to DON, DON-FB1, NIV-ZEA, DON-NIV-ZEA, and DON-NIV-FB1 (Fig. 3B). In 70/30 co-cultures, significant decreases in MUC5AC protein levels were observed in cell lysates following the treatment with NIV, DON-NIV, DON-ZEA, NIV-FB1, DON-NIV-ZEA, DON-NIV-FB1, DON-ZEA-FB1, NIV-ZEA-FB1, and DON-NIV-ZEA-FB1. Similar decreases in MUC5AC protein levels were also found in cell supernatants following exposure to DON, NIV, ZEA, DON-ZEA, DON-FB1, NIV-ZEA, NIV-FB1, and DON-NIV-ZEA with no difference noted in cell supernatants following DON-ZEA-FB1 exposure. For the modulation of MUC5B protein levels, significant decreases in MUC5B protein levels were observed in cell lysates following exposure to NIV and DON-NIV but significant increases were only found in ZEA-FB1. For the cell supernatants, significant increases in the MUC5B level were noted in DON-ZEA-FB1 and NIV-ZEA-FB1 (Fig. 3C).

A heat map depicting changes in total MUC5AC and MUC5B protein profiles (sum of their respective levels in cell lysates and cell supernatants) in Caco-2/HT29-MTX co-cultures (100/0, 90/10, 70/30, and 0/100) following exposure to individual or combinations of DON, NIV, ZEA, and FB1 is shown in Supplementary 3. Caco-2/HT29-MTX co-cultures (100/0, 90/10, 70/30, and 0/100) demonstrated a unique mucin protein profile following exposure to different mycotoxins, which is in accordance with obtained respective mRNA expression profiles. Comparing both MUC5AC and MUC5B profiles for the Caco-2/HT29-MTX co-cultures (100/0, 90/10, 70/30, and 0/100), it was shown that the MUC5AC protein levels were generally higher when cells were not treated with any mycotoxins, and that addition of these agents alone or in increasing number (i.e., mixtures) leads to reductions in MUC5AC protein level. This was in contrast to the MUC5B protein levels, which increases with exposure to increasing numbers of mycotoxins.

Effects of Individual and Mixture of Fusarium Mycotoxins Mucin-Like Glycoprotein Secretion

To investigate the effects of individual and mixtures of Fusarium mycotoxins on the mucus secretion, mucin-like glycoprotein secretion in cell lysates and those secreted into culture supernatants were measured and analyzed (Fig. 4).

FIG. 4.

Relative levels of total mucin-like glycoprotein in cell lysates and supernatants from Caco-2/HT29-MTX (A) 100/0, (B) 90/10, (C) 70/30, and (D) 0/100 co-cultures isolated at 48 h after exposure to individual or combinations of DON (2μM), NIV (2μM), ZEA (40μM), and FB1 (40μM). Results are expressed as a percentage of the control and are means ±SEM (n = 3). *, **, *** p < 0.05, 0.01, and 0.001, respectively, compared with control.

In Caco-2 monocultures, all treatment groups except FB1 alone significantly decreased mucin levels in cell lysates. Only treatment with DON-FB1 significantly decreased the total mucin in the cell supernatants (Fig. 4A). However, when the HT29-MTX monoculture was examined, DON and NIV significantly increased the total mucin in cell lysates whereas ZEA, DON-NIV, DON-ZEA, DON-FB1, ZEA-FB1, DON-NIV-ZEA, and DON-ZEA-FB1 significantly decreased it. For total mucin in cell supernatants, only treatments with DON-NIV-FB1 and DON-NIV-ZEA-FB1 showed significant reductions (Fig. 4D).

As for Caco-2/HT29-MTX 90/10 co-cultures, only treatments with DON-NIV-ZEA, NIV-ZEA-FB1, and DON-NIV-ZEA-FB1 significantly decreased the total mucin levels in cell lysates whereas most of the treatment groups, except DON, NIV, ZEA, DON-ZEA, and NIV-ZEA, significantly decreased total mucin in cell supernatants (Fig. 4B). In the 70/30 co-cultures, DON, ZEA, FB1, DON-ZEA, DON-FB1, NIV-ZEA, ZEA-FB1, DON-NIV-ZEA, and DON-NIV-FB1 significantly decreased total mucin levels in cell lysates. Similarly, DON, NIV, DON-NIV, DON-ZEA, DON-FB1, NIV-FB1, ZEA-FB1, DON-NIV-ZEA, DON-NIV-FB1, DON-ZEA-FB1, NIV-ZEA-FB1, and DON-NIV-ZEA-FB1 significantly decreased total mucin in cell supernatants, whereas NIV-ZEA significantly increased it (Fig. 4C).

A heat map depicting changes in total mucin-like glycoprotein secretion profiles (sum of their respective levels in cell lysates and cell supernatants) in Caco-2/HT29-MTX co-cultures (100/0, 90/10, 70/30, and 0/100) following individual or combinations of DON, NIV, ZEA, and FB1 exposure is shown in Supplementary 4. Caco-2/HT29-MTX co-cultures (90/10, 70/30) demonstrated a unique total mucin secretion profile following exposure to different mycotoxins. Comparing total mucin secretion profiles for the Caco-2/HT29-MTX co-cultures (100/0, 90/10, 70/30, and 0/100), total mucin levels were higher when cells were not treated with any mycotoxins, although addition of increasing numbers of mycotoxin in mixtures led to reduced mucin levels, and is similar to what was observed for total MUC5AC protein analysis.

In order to determine if there were any interactions among DON, NIV, ZEA, and FB1 on total mucin-like glycoprotein secretion, univariate ANOVA was conducted (Table 3). The results reveal nonadditive interactions in mixtures of DON-NIV-ZEA (F 1, 126) = 9.67, p = 0.002, DON-NIV-FB1 (F 1, 126) = 5.03, p = 0.027, DON-ZEA-FB1 (F 1, 126) = 4.67, p = 0.033, and NIV-ZEA-FB1 (F 1, 126) = 8.15, p = 0.005 in Caco-2 monocultures. Nonadditive interactions were also observed in mixtures of DON-FB1 (F 1, 128) = 5.90, p = 0.017, NIV-ZEA-FB1 (F 1, 128) = 7.25, p = 0.008, and DON-NIV-ZEA-FB1 (F 1, 128) = 5.07, p = 0.026 in HT29-MTX monocultures. No interactions were observed in 90/10 co-cultures but nonadditive interactions were found in mixtures of DON-NIV (F 1, 128) = 6.65, p = 0.011, ZEA-FB1 (F 1, 128) = 14.93, p < 0.001, and DON-NIV-FB1 (F 1, 128) = 4.31, p = 0.04 in 70/30 co-cultures.

Results of Univariate Analyses of Multiple Fusarium Mycotoxin Exposure on Caco-2/HT29-MTX (A) 100/0, (B) 90/10, (C) 70/30, and (D) 0/100 as Analyzed by Total Mucin-Like Glycoprotein Secretion

TABLE 3

Results of Univariate Analyses of Multiple Fusarium Mycotoxin Exposure on Caco-2/HT29-MTX (A) 100/0, (B) 90/10, (C) 70/30, and (D) 0/100 as Analyzed by Total Mucin-Like Glycoprotein Secretion

	(A) Caco-2/HT29-MTX 100/0		(B) Caco-2/HT29-MTX 90/10
Combinations	F-value	p-value
DON-NIV	0.813	0.369	0.862	0.355
DON-ZEA	0.040	0.842	0.463	0.498
DON-FB1	0.084	0.772	0.032	0.858
NIV-ZEA	0.305	0.582	2.401	0.124
NIV-FB1	0.173	0.678	0.127	0.722
ZEA-FB1	0.175	0.677	0.437	0.510
DON-NIV-ZEA	9.674	0.002	0.479	0.490
DON-NIV-FB1	5.026	0.027	2.173	0.143
DON-ZEA-FB1	4.674	0.033	0.015	0.904
NIV-ZEA-FB1	8.148	0.005	1.404	0.238
DON-NIV-ZEA-FB1	1.475	0.227	0.210	0.647
	(C) Caco-2/HT29-MTX 70/30		(D) Caco-2/HT29-MTX 0/100
Combinations	F-value	p-value
DON-NIV	6.651	0.011	2.631	0.107
DON-ZEA	0.413	0.521	0.467	0.495
DON-FB1	0.016	0.898	5.900	0.017
NIV-ZEA	0.565	0.454	1.345	0.248
NIV-FB1	0.320	0.572	0.015	0.903
ZEA-FB1	14.925	< 0.001	3.038	0.084
DON-NIV-ZEA	3.100	0.081	2.618	0.108
DON-NIV-FB1	4.308	0.040	1.267	0.262
DON-ZEA-FB1	1.659	0.200	3.372	0.069
NIV-ZEA-FB1	0.070	0.792	7.248	0.008
DON-NIV-ZEA-FB1	1.902	0.170	5.070	0.026

*p < 0.05 are bolded.

Correlations Between MUC5AC and MUC5B Gene Expression, Intracellular (Cell Lysate) and Extracellular (Supernatant) Protein Levels, and Total Mucin-Like Glycoprotein Secretion

For Caco-2 monocultures, exposure to DON, NIV, ZEA, and FB1 resulted in significant positive correlations between MUC5AC and MUC5B mRNA expression (p < 0.001) as determined by qPCR. This was consistent with correlations found between intracellular or extracellular MUC5AC and MUC5B (p = 0.001 and p = 0.032, respectively) protein levels as determined by ELISA, although it is interesting to note significant negative correlations between intracellular MUC5B and extracellular MUC5AC (p = 0.015). Significant positive correlations were revealed between total MUC5AC and intracellular or extracellular MUC5AC protein levels (p < 0.001 and p = 0.015, respectively), which was also in parallel with correlations observed between total MUC5B and intracellular or extracellular MUC5B protein (p < 0.001 and p < 0.001, respectively), as well as correlations between total mucin and intracellular or extracelluar total mucin (p < 0.001 and p < 0.001, respectively). Significant positive correlations were observed between intracellular and extracellular MUC5B (p = 0.024), although no correlations were found between intracellular and extracellular MUC5AC. Intracellular total mucin levels were positively correlated with the extracellular total mucin (p = 0.044). Interestingly, significant negative correlations were observed between MUC5AC mRNA and intracellular or extracellular total mucin (p = 0.044 and p < 0.001, respectively), and significant negative correlations were only found between MUC5B mRNA and extracellular total mucin (p < 0.001). Similarly, significant negative correlations were demonstrated between total mucin and MUC5AC mRNA, MUC5B mRNA, extracellular MUC5B, and total MUC5B (p < 0.001, p = 0.003, p < 0.001, and p = 0.014, respectively). Significant negative correlations were also observed between intracellular total mucin and extracellular MUC5B and total MUC5B (p < 0.001 and p = 0.002). Moreover, significant positive correlations were observed between MUC5AC mRNA and extracellular MUC5B or total MUC5B (p = 0.034 and p = 0.031, respectively) between MUC5B mRNA and extracellular MUC5AC (p = 0.024) and extracellular MUC5B and total MUC5AC (Fig. 5A).

FIG. 5.

Heat map showing the correlation among MUC5AC and MUC5B mRNA expression, protein levels, and total mucin-like glycoprotein secretion for Caco-2/HT29-MTX (A) 100/0, (B) 90/10, (C) 70/30, and (D) 0/100 co-cultures taking into account all data from individual and multiple Fusarium toxin exposure experiments. The heat map is a visual representation of the correlation values between each pair of parameters denoted by the corresponding row and columns of the matrix. The matrix is symmetrical from the bottom-left to upper-right diagonal axis. Red represents a positive correlation, yellow represents a low correlation, and green represents a negative correlation, as shown in the color key. Statistical significances were analyzed by Spearman's correlations. *, **, *** p < 0.05, 0.01, and 0.001. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

When HT29-MTX monocultures were used, significant positive correlations were found between MUC5AC and MUC5B mRNA (p = 0.011). However, total MUC5AC was inversely correlated with total MUC5B (p = 0.003). Positive correlations were also demonstrated between total MUC5AC and intracellular or extracellular MUC5AC (p < 0.001 and p < 0.001, respectively), between MUC5B mRNA and intracellular MUC5B (p = 0.023), between extracellular MUC5B and total MUC5B (p < 0.001), between MUC5AC and intracellular or extracellular total mucin (p = 0.004 and p = 0.002, respectively), and between MUC5AC mRNA and total mucin (p = 0.001), but surprisingly intracellular MUC5B was negatively correlated with extracellular MUC5B (p = 0.045). Intracellular total mucin was positively correlated with extracellular total mucin as well as total mucin (p < 0.001 and p < 0.001, respectively). Extracellular total mucin also had a positive correlation with total mucin (p < 0.001). Conversely, significant negative correlations were found between extracellular MUC5B and intracellular or extracellular MUC5AC or total MUC5AC (p = 0.019, p = 0.02, and p = 0.003). Significant negative correlations were also observed between total MUC5B and intracellular or extracellular MUC5AC (p = 0.023 and p = 0.022, respectively) (Fig. 5D).

As in Caco-2 monocultures, in the 90/10 co-cultures, MUC5AC mRNA was positively correlated with MUC5B mRNA (p < 0.001). Positive correlations were also found between MUC5AC mRNA and intracellular MUC5AC (p = 0.001), between MUC5B mRNA and intracellular MUC5AC (p = 0.002), between extracellular MUC5AC and total MUC5AC (p < 0.001), between intracellular MUC5B and extracellular MUC5AC or total MUC5AC (p = 0.002 and p = 0.002, respectively), and between intracellular MUC5AC and extracellular MUC5B (p = 0.016). There were positive correlations between total MUC5B and intracellular or extracellular MUC5B (p = 0.007 and p < 0.001, respectively), between intracellular mucin and intracellular MUC5AC or extracellular MUC5B (p = 0.004 and p = 0.043, respectively), between total mucin and intracellular MUC5AC or extracellular MUC5B (p = 0.002 and p = 0.025, respectively), and between total mucin and intracellular or extracellular total mucin (p < 0.001 and p < 0.001, respectively). On the other hand, significant negative correlations were observed between intracellular MUC5AC and extracellular MUC5AC (p < 0.001), between MUC5AC mRNA and extracellular MUC5B or total MUC5B (p = 0.015 and p = 0.006, respectively), extracellular MUC5AC and extracellular MUC5B or total MUC5B (p < 0.001 and p = 0.002, respectively), total MUC5AC and extracellular MUC5B or total MUC5B (p = 0.001 and p = 0.009, respectively). Intracellular total mucin was also found to be negatively correlated with MUC5AC mRNA (p < 0.001), MUC5B mRNA (p = 0.016), and total MUC5AC (p = 0.003), whereas extracellular total mucin was only negatively correlated with extracellular MUC5AC (p = 0.003), total MUC5AC (p = 0.007), and intracellular total mucin (p = 0.013). Moreover, negative correlations were also observed between total mucin and MUC5AC mRNA, extracellular MUC5AC or total MUC5AC (p < 0.001, p < 0.001, and p = 0.001, respectively) (Fig. 5B).

In the 70/30 co-cultures, significant positive correlations were found between MUC5AC and MUC5B mRNA expression (p = 0.001). Significant positive correlations were also observed between MUC5AC mRNA and the extracellular or total MUC5AC levels (p = 0.001 and p = 0.037, respectively). Total MUC5AC were positively correlated with intracellular or extracellular MUC5AC (p < 0.001 and p < 0.001, respectively), which were similar to the positive correlations found between total MUC5B and intracellular or extracellular MUC5B (p < 0.001 and p < 0.001, respectively). Positive correlations were also revealed between total mucin and MUC5AC mRNA, intracellular or extracellular MUC5AC, total MUC5AC as well as intracellular or extracellular total mucin (p = 0.049, p = 0.005, p < 0.001, p < 0.001, p < 0.001, and p < 0.001, respectively), whereas negative correlations were found between total mucin and extracellular MUC5B or total MUC5B (p = 0.023 and p = 0.034). Interestingly, correlations were noted between total MUC5B and intracellular or extracellular MUC5AC (p = 0.049 and p = 0.01, respectively), between intracellular total mucin and extracellular MUC5AC or total MUC5AC (p < 0.001 and p < 0.001, respectively) and between extracellular total mucin and MUC5AC mRNA and intracellular MUC5AC (p = 0.019 and p = 0.001, respectively) (Fig. 5C).

DISCUSSION

This is the first study to investigate individual and combined effects of four common Fusarium mycotoxins, DON, NIV, ZEA, and FB1, on mucin mRNA expression, protein production, and total mucin-like glycoprotein secretion in Caco-2/ HT29-MTX co-cultures with different initial seeding ratios, 100/0, 90/10, 70/30, and 0/100.

In this study, DON, NIV, ZEA, and FB1 alone or in combinations in general differentially modulated mRNA and protein secretion of MUC5AC and MUC5B and total mucin-like glycoprotein secretion in Caco-2/ HT29-MTX co-cultures. No cytotoxic effects for DON, NIV, ZEA, and FB1 alone or in combinations in Caco-2/ HT29-MTX co-cultures (100/0, 90/10, and 0/100 but not 70/30) were observed in the MTT assay (Fig. 1). This suggests that regulations on mucin gene expression and protein production were not simply a consequence of gross IEC toxicities.

Elevated levels of MUC5AC and MUC5B transcription were observed following individual or mixtures of DON, NIV, ZEA, and FB1 exposure. Based on similar transcriptional responses for both genes, it appears that there exists a possible common regulatory mechanism between MUC5AC and MUC5B. Considering that MUC5AC and MUC5B are encoded within the same cluster at chromosome location 11p15.5, and that they have the same transcriptional orientation, similarity in size and distribution of exons, and splice site types (Moniaux et al., 2001), a common regulatory mechanism may not be surprising. The extent of changes on mRNA levels, however, varied between genes. Expression of MUC5B in mycotoxin-treated cells was less than that of MUC5AC, and these findings may suggest differential induction of both mucin genes. This is in agreement with previous investigations by others in which the regulation of MUC5AC and MUC5B gene expression and secretion after exposure to lipopolysaccharide or proinflammatory cytokines were different in HT29-MTX cells (Smirnova et al., 2003). Nevertheless, at this point, considering the similarity in transcriptional response occurring post-toxin exposure, the exact mechanism behind transcriptional regulation of both genes remains unknown.

Concerning the characteristics and function of secretory mucins, both intracellular mucin synthesis and extracellular mucin secretion were investigated to provide a more comprehensive view of how synthesis and secretion may be impacted by Fusarium mycotoxin exposure. Upon assessing protein levels of MUC5AC and MUC5B by ELISA, an expression pattern similar to that of mRNA transcript was observed; an observation in agreement with numerous nonmycotoxin-related studies comparing MUC5AC and MUC5B protein by ELISA and mRNA data (Homsi et al., 2007; Smirnova et al., 2003). Secretion of mucin from intestinal goblet cells is known to be regulated by two processes, an unregulated constitutive pathway dependent on the continuous movement of mucin granules from the golgi to the apex of the cell, and a regulated process dependent on the sudden release of mucin from central storage granules (McCool et al., 1995). The former pathway is directly associated with the rate of synthesis of mucin and, thus, mucin gene expression. The latter is independent of synthesis in the short term and is measured as the increase in mucin output over baseline levels that follows an appropriate stimulus (Hong et al., 1999). Interestingly, we observed the extent of changes in protein levels observed was less than those for mRNA transcript. These differences may be explained, at least partly, by post-transcriptional or -translational regulatory mechanisms linked to mucin molecule(s) synthesis and secretion. It is possible that increases in mRNA stability following toxin exposure may occur and has been reported with other stimuli (Chorley et al., 2006), though this remains to be demonstrated. Moreover, differences between mucin mRNA and protein secretion may reveal issues associated with detection thresholds for either mRNA or protein, of which the methods for quantifying mRNA transcript levels are more sensitive than those for protein identification and quantification (Greenbaum et al., 2003). Furthermore, it was noted that there were generally higher extracellular MUC5AC and MUC5B levels than respective intracellular levels, though this may be associated with our sampling time (i.e., accumulation of mucin in the cell supernatants with time). Several studies have also demonstrated that cellular signaling pathways, such as protein kinase C, epidermal growth factor receptor, Ras/Raf, mitogenactivated protein kinase, and mitogen-activated protein kinases, may play roles in modulation of mucins in different cells (Hewson et al., 2004). However, the study regarding the exact underlying mechanisms is currently in progress in our laboratory and thus will not be discussed further here.

Whereas the Caco-2/HT29-MTX co-culture model has been characterized as a standard model for absorption, transport, and permeability studies (Walter et al., 1996; Wikman-Larhed and Artursson, 1995), its use to study mucin production is less well established. For example, less is known about how different absorptive and goblet cell ratios modulate mucin mRNA expression, protein production, and secretion, that may impact permeability of the epithelial cell monolayer, and thus affect bacterial translocation and colonization across the intestinal epithelium (Katayama et al., 1997). In this study, when Caco-2 and HT29-MTX monocultures were used, the cultures responded to individual and mixtures of DON, NIV, ZEA, and FB1 by elevating the mucin mRNA, protein production, and total mucin glycoprotein secretion. However, results obtained from co-culture models show that mucin mRNA expression, protein production, and total mucin-like glycoprotein secretion could be modulated stepwise by varying respective cell ratios. As the initial seeding ratio of HT29-MTX cells increased, mucin mRNA expression and protein secretion decreased following the exposure to individual and mixtures of DON, NIV, ZEA, and FB1. It is possible that pure cultures may form a tighter, more intact barrier possessing smaller mean pore sizes for tight junctions that serve to decrease the paracellular permeability of toxins through the cell layer. In this heterogeneous cell population, differences in the nature and formation of junctional complexes between Caco-2 and HT29-MTX cells are reported to result in relatively larger pore sizes, which increase paracellular permeability and susceptibility of the cell monolayer to toxin exposure (Calatayud et al., 2012). Such increases in co-culture permeability have been previously reported by other investigators (Hilgendorf et al., 2000; Walter et al., 1996).

Previous studies have demonstrated that overexpression of mucins has been associated with colonic bacterial infection by Shigella dysenteriae (Raja et al., 2012) and rotavirus infection (Liu et al., 2010). The rapid elevation of mucin secretion in response to bacterial pathogen infection is thought to provide an important mechanism of protection that aids in eliminating the pathogenic agent from the intestinal tract (Moncada et al., 2003). Surprisingly, in spite of upregulation of mRNA expression and protein secretion of MUC5AC and MUC5B, our study showed decreases in total mucin-like glycoprotein secretion, which suggests mucus depletion in response to mycotoxin exposure. The disruption of the mucus layer is likely to be deleterious to the host as it affords increased opportunity for pathogen adherence and subsequent translocation (Albanese et al., 1994).

Moreover, there is no information in the literature regarding the combined effect of Fusarium mycotoxins on total mucin-like glycoprotein secretion in vitro. In this study, additive effects were noted in most of the toxin mixtures, leading to decreases in total mucin-like glycoprotein secretion when compared with single treatments of Fusarium mycotoxin. Similarly, Tajima et al. (2002) have observed additive effects on the inhibition of DNA synthesis when L929 fibroblasts were exposed to multiple Fusarium mycotoxins (DON, NIV, T-2, ZEA, and FB1). Interactive effects (i.e., synergism or antagonism), which have not been reported previously, were also observed in a number of mixtures in different co-culture models, though the reason for such interaction remains unknown. This may indicate that these commonly occurring Fusarium mycotoxins, when combined, may interact with each other so that the magnitude of resulting toxic effects generated may be potentiated or reduced by actions of other toxins (Tajima et al., 2002). Therefore, the present results with multiple Fusarium mycotoxins suggest that Fusarium mycotoxins alone may not predict their effects in natural environment where combinations of toxins are frequently observed, meaning investigations of commonly observed toxin mixtures are necessary and will enable more accurate and effective risk assessments (Speijers and Speijers, 2004). To understand the nature of these interactive effects, a molecular-level understanding of mycotoxin-mycotoxin interaction is required in the future in order to develop improved detoxification and remediation strategies aiming at understanding mycotoxin impact on animal and human health (Fink‐Grernmels, 1999).

In the light of the results obtained, it can be concluded that IECs responded differently to individual and mixtures of DON, NIV, ZEA, and FB1 through differential modulation of secretory mucins, MUC5AC and MUC5B, as well as total mucin-like glycoprotein secretion, all of which are essential components of host mucosal immunity. These findings will help identify and characterize underlying molecular mechanisms on impacts of mycotoxins on mucin synthesis and secretion. Likewise, in terms of risk assessment, interactive effects of Fusarium mycotoxins are of great concerns as this study demonstrates the complexity of mycotoxin interactions, by which the combination of toxins may exhibit more profound defect in immune responses in animals/humans. As our understanding of interactions of mycotoxins on animal and human diseases improves, such data could provide an increased level of complexity in defining and optimizing maximum permissible limits of Fusarium mycotoxins in foods and feeds.

FUNDING

Hong Kong Research Grant Council Grant (Number 765810).

Acknowledgments

We thank Dr Thecla Lesuffleur (INSERM U178, Villejuif, France) for kindly providing the HT29-MTX cell line. We also thank Professor John Bacon-Shone of the Social Sciences Research Centre of the University of Hong Kong for his statistical advice.