Absorption and metabolism of the food contaminant 3-chloro-1,2-propanediol (3-MCPD) and its fatty acid esters by human intestinal Caco-2 cells
- First Online:
- Cite this article as:
- Buhrke, T., Weißhaar, R. & Lampen, A. Arch Toxicol (2011) 85: 1201. doi:10.1007/s00204-011-0657-6
- 535 Views
3-Chloro-1,2-propanediol (3-MCPD) fatty acid esters are formed upon thermal processing of fat-containing foods in the presence of chloride ions. Upon hydrolytic cleavage, these substances could release free 3-MCPD. This compound is toxicologically well characterised and displayed cancerogenic potential in rodent models. Recently, serious contaminations of different food products with 3-MCPD fatty acid esters have been reported. In regard to a risk assessment, the key question is to which degree these 3-MCPD fatty acid esters are hydrolysed in the human gut. Therefore, the aim of the present project was to examine the hydrolysis of 3-MCPD fatty acid esters and the resulting release of free 3-MCPD by using differentiated Caco-2 cells, a cellular in vitro model for the human intestinal barrier. Here, we show that 3-MCPD fatty acid esters at a concentration of 100 μM were neither absorbed by the cells nor the esters were transported via a Caco-2 monolayer. 3-MCPD-1-monoesters were hydrolysed in the presence of Caco-2 cells. In contrast, a 3-MCPD-1,2-diester used in this study was obviously absorbed and metabolised by the cells. Free 3-MCPD was not absorbed by the cells, but the substance migrated through a Caco-2 monolayer by paracellular diffusion. From these in vitro studies, we conclude that 3-MCPD-1-monoesters are likely to be hydrolysed in the human intestine, thereby increasing the burden with free 3-MCPD. In contrast, intestinal cells seem to have the capacity to metabolise 3-MCPD diesters, thereby detoxifying the 3-MCPD moiety.
Keywords3-MCPD3-MCPD fatty acid esterFood contaminantRisk assessmentCaco-2
3-Chloro-1,2-propanediol (3-monochloropropane-1,2-diol; 3-MCPD) is a well-known food contaminant that was first detected in acid-hydrolysed vegetable proteins (Velísek et al. 1978). Later, it was shown to be present in various heat-processed foods such as bread, cereals and meat (Hamlet et al. 2002). Studies on the occurrence and formation of 3-MCPD led to the assumption that this substance is formed from triacylglycerides in the presence of chloride ions under acidic conditions when a chloride anion may replace an acyl group by a nucleophilic attack on a carbon atom that is activated by neighbouring ester groups (Collier et al. 1991; Hamlet et al. 2002). Further acid-mediated ester hydrolysis then may lead to the release of free 3-MCPD. The occurrence of 3-MCPD fatty acid esters in acid-hydrolysed vegetable proteins has been reported first in 1980 by Velísek et al., and in the following years, 3-MCPD esters have been detected in numerous heat-processed foods and vegetable oils (Hamlet and Sadd 2004; Sveikovská et al. 2004; Zelinková et al. 2006).
No toxicological data on 3-MCPD fatty acid esters are available so far; however, free 3-MCPD is toxicologically well characterised. A maximum tolerable daily intake (TDI) of 2 μg kg−1 body weight has been recommended for 3-MCPD by the Scientific Committee on Food of the European Union (European Commission 2001). In the following years, it was shown that the substance is not genotoxic in vivo (Robjohns et al. 2003), but mutagenic effects were observed at high concentrations in certain in vitro experiments (El Ramy et al. 2007). In a two-year cancerogenic study, there was increased incidence for the development of tumours in kidney and testis in male rats (Cho et al. 2008a). There was no incidence for cancerogenicity in a 90-day study with mice; however, adverse effects were observed again for kidney and testis (Cho et al. 2008b). The observation that kidney was the main target organ in the toxicity studies might be due to the fact that the well-known nephrotoxic compound oxalic acid is one of the main metabolites of 3-MCPD (Jones et al. 1981).
In contrast to free 3-MCPD, the toxicological characterisation of 3-MCPD fatty acid esters is somehow puzzling due to the large number of different existing compounds. 3-MCPD can be esterified with all those saturated and unsaturated fatty acids that exist in living systems. Acyl groups might be bound only to one of the two hydroxyl groups (3-MCPD monoesters) or to both positions (3-MCPD diesters). Therefore, food samples burdened with 3-MCPD esters do in fact contain a mixture of numerous different 3-MCPD mono- and diesters (Zelinková et al. 2006). However, it has been assumed that 3-MCPD fatty acid esters ingested with food would probably be completely hydrolysed because (i) the ester bonds could be cleaved under the acidic conditions of the stomach and (ii) due to their structural similarity to mono- and diacylglycerols, the esters might be accepted as substrates for intestinal lipases. Thus, in the context of toxicological characterisation of 3-MCPD fatty acid esters, attention has primarily been drawn on the putative ester hydrolysis so far. It has been reported that 3-MCPD esters were hydrolysed by lipases from Aspergillus oryzea (Hamlet and Sadd 2004). Seefelder et al. showed that by using an in vitro system based on the activity of pancreatic lipase in the presence of porcine bile extract, various 3-MCPD fatty acid esters were hydrolysed within minutes (Seefelder et al. 2008). In this study, we examined the metabolism of free 3-MCPD and of 3-MCPD fatty acid esters by human intestinal cells. The well-established Caco-2 system is a widely used in vitro model for the intestinal barrier, because differentiated Caco-2 cells are known to form a tight cellular monolayer with morphological and biochemical properties very similar to those of enterocytes of the small intestine including the formation and excretion of “HDL-like” particles from ingested lipids (Pinto et al. 1983; Artursson et al. 2001; Liao and Chan 2000; Chateau et al. 2005). Here, we show for the first time that (i) free 3-MCPD is able to migrate through a Caco-2 monolayer, that (ii) 3-MCPD monoesters are hydrolysed by cellular lipases, thereby releasing free 3-MCPD, and that (iii) the diester 3-MCPD-1,2-dipalmitate is obviously metabolised by the cells.
Materials and methods
All chemicals were purchased from Merck (Darmstadt, Germany) in the highest available purity. 3-Chloro-1,2-propanediol (3-MCPD) was from Sigma Chemicals (Deisenhofen, Germany). 1-Lauroyl-3-chloropropanediol (3-MCPD mono-laureate), 1-oleoyl-3-chloropropandediol (3-MCPD monooleate) and 1,2-bis-palmitol-3-chloropropanediol (3-MCPD dipalmitate) were purchased from Toronto Research Chemicals (Ontario, Canada). For easy reading, the trivial names given in brackets are used in the following.
Stock solutions of 100 mM in DMSO were prepared for 3-MCPD, 3-MCPD monolaureate and 3-MCPD monooleate. Due to its low solubility in DMSO, 3-MCPD dipalmitate was dissolved in ethyl acetate (100 mM stock solution). For the incubation experiments, the respective stock solution was diluted 1:1,000 in complete medium (Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal calf serum, 100 U/ml penicillin and 100 μg/ml streptomycin) to give a final concentration of 100 μM of the respective substance and a final concentration of 0.1% DMSO or 0.1% ethyl acetate in case of the diester, respectively. In the control experiments, DMSO or ethyl acetate was also added to the complete medium to a final concentration of 0.1%.
The human colon adenocarcinoma cell line Caco-2 was obtained from the European Collection of Cell Culture (ECACC, Porton Down, UK), and the culture media and supplements were obtained from PAA Laboratories GmbH (Pasching, Austria). Caco-2 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% foetal calf serum, 100 U/ml penicillin and 100 μg/ml streptomycin in a humidified atmosphere of 5% CO2 at 37°C. Cells were passaged every 3–4 days by treatment with 0.1% trypsin and 0.04% EDTA and then plated at a density of 1.3–2 × 104 cells/cm2.
For transport analysis, Caco-2 cells were grown on Transwell™ inserts (4.71-cm² growth area, 0.4-μm pore size, polycarbonate membranes; Corning Costar Co., Cambridge, MA) and cultured for 20 days to allow complete cellular differentiation and formation of a Caco-2 monolayer (Cogburn et al. 1991; Buesen et al. 2002). The incubation experiment was started by replacing the medium of the apical chamber with medium containing the respective test substance. The integrity of the cell monolayer was routinely checked before and after the transport experiments by measuring the transepithelial electrical resistance (TEER) (Ebert et al. 2005).
For the determination of 3-MCPD, cell culture supernatants were diluted tenfold in 20 mM sodium phosphate buffer pH 6.0 supplemented with 20% NaCl (w/v). This solution was directly applied to GC/MS analysis. To determine the intracellular content of 3-MCPD, Caco-2 cells were scraped out of the cell culture flask, washed 3 times with phosphate-buffered saline (PBS) and re-suspended in 1 ml of sterile distilled water. Cell lysates were prepared by sonication and subsequent removal of cellular debris by centrifugation. The supernatant was diluted ten-fold in 20 mM sodium phosphate buffer pH 6.0 supplemented with 20% NaCl (w/v). This solution was directly applied to GC/MS analysis.
For the determination of 3-MCPD fatty acid esters, cell culture supernatants were lyophilised. Caco-2 cells were scraped out of the cell culture flask, washed 3 times with phosphate-buffered saline (PBS), one time with sterile distilled water and finally lyophilised. Lyophilised samples were suspended in 1 ml of a mixture of t-butylmethylether and ethyl acetate (80 + 20/vv). Unsolved matter was allowed to settle down, and then, 0.5 ml of supernatant was used for ester cleavage and GC/MS analysis.
For the determination of free 3-MCPD, GC/MS analysis was performed according to Divinova et al. (2004). After addition of deuterium-labelled 3-MCPD as internal standard, sample solution was derivatised with phenylboronic acid. The resulting cyclic phenylboronic derivative was extracted with hexane, and the hexane layer was analysed by GC/MS.
Ester-bound 3-MCPD was determined by GC/MS analysis according to Option A of DGF Standard Method C III 18b (Deutsche Gesellschaft für Fettwissenschaft: DGF Standard Method C III 2009). After addition of deuterium-labelled 3-MCPD as internal standard, 3-MCPD esters were cleaved with NaOCH3 (0.5 M in methanol). Derivatisation was performed with phenylboronic acid in NaCl solution. The resulting cyclic phenylboronic derivative was extracted with hexane, and the hexane layer was analysed by GC/MS.
Absorption and metabolism of 3-MCPD
To examine whether 3-MCPD was absorbed and possibly metabolised by human intestinal cells, differentiated Caco-2 cells were incubated with various concentrations and for various incubation times with 3-MCPD. Subsequent to the incubation experiment, the amount of 3-MCPD in the cell culture supernatants as well as in the lysates prepared from the incubated cells was determined by GC/MS analysis. Up to 1 mM 3-MCPD was applied to the cells; however, the 3-MCPD concentration in the cell culture medium was constant and never decreased within a period of at least 24 h (data not shown). Moreover, 3-MCPD was never detected in cellular lysates prepared from Caco-2 cells from these incubation experiments. From these results, we conclude that 3-MCPD was neither absorbed nor metabolised by Caco-2 cells. The substance was stably present in the cell culture medium and was not affected by the cells.
Diffusion of 3-MCPD across a Caco-2 monolayer
Hydrolysis of 3-MCPD fatty acid esters by Caco-2 cells
Similar results were obtained by using 3-MCPD monooleate for the incubation experiment (Fig. 2b). In the control experiment, this substance was completely stable and no spontaneous hydrolysis of the ester was observed within 24 h. In the presence of Caco-2 cells, 3-MCPD monooleate was hydrolysed; however, hydrolysis was slower compared to the Caco-2-mediated cleavage of 3-MCPD monolaureate. About 70% of the applied dose of 3-MCPD monooleate was hydrolysed by the cells after 24 h (Fig. 2b).
When 3-MCPD dipalmitate was used for the incubation experiment, no release of free 3-MCPD was observed within a period of 72 h (data not shown). Thus, in contrast to the monoesters, the diester was not cleaved by the cells.
Cellular lysates prepared from the cells after the incubation experiment were also analysed for the presence of free 3-MCPD; however, free 3-MCPD was not detected in these cellular lysates, neither for the monoesters nor for the diester. In case of the monoesters, this indicates that hydrolysis seems to take place outside the cells. Thus, membrane-bound enzymes or secreted enzymes released from Caco-2 cells might facilitate monoester hydrolysis, but these enzymes do obviously not accept the diester as a substrate.
Recovery of 3-MCPD as free substance and bound to the respective ester in the transport study
Total amount of 3-MCPD as free substance and bound to the respective ester in both Transwell™ chambers (μM)
96.3 ± 4.8
104.3 ± 8.0
102.2 ± 8.7
96.0 ± 29.1
90.3 ± 38.0
99.1 ± 13.7
96.2 ± 22.7
81.8 ± 30.4
102.6 ± 24.0
91.2 ± 25.3
83.2 ± 33.3
83.4 ± 25.5
91.7 ± 22.9
87.6 ± 30.6
73.8 ± 23.1
101.4 ± 23.8
88.0 ± 33.1
28.0 ± 12.9
98.4 ± 22.2
85.1 ± 34.8
12.4 ± 5.9
Similar results were obtained by using 3-MCPD monooleate for the Transwell™ experiment (Fig. 3b). Caco-2-mediated cleavage of 3-MCPD monooleate was not as fast as the cleavage of the monolaureate ester; however, the overall picture of the disappearance of the ester in the apical chamber and the concomitant appearance of free 3-MCPD in both chambers was comparable to the results obtained for the monolaureate ester. Again, adding up the amount of both substances in both chambers yielded about 80–90% of the initial concentration of the ester (Table 1), indicating the lack of absorption and/or metabolism of the ester and of free 3-MCPD by the cells.
By using 3-MCPD dipalmitate for the Transwell™ experiment, a time-dependent decrease in the substance in the apical chamber was observed and only marginal amounts of the diester were detected in the basolateral chamber over a period of 24 h (Fig. 3c). In contrast to the results obtained for the monoesters, there was no free 3-MCPD detected either in the apical or in the basolateral chamber (data not shown). This finding was in line with the result described earlier, that no 3-MCPD was released from 3-MCPD dipalmitate upon incubation with Caco-2 cells. Thus, in the Transwell™ experiment, there was a clear decrease in the total amount of the diester over the time (Table 1) that could not be explained by extracellular cleavage of the ester. Therefore, in contrast to the monoesters, the diester must have been absorbed and/or metabolised by the cells.
To examine whether the 3-MCPD diester was absorbed and stored inside the cells, Caco-2 cells were incubated with 100 μM 3-MCPD dipalmitate. After various incubation times, cell culture supernatants were collected and cellular lysates were prepared. From these samples, the amount of free 3-MCPD as well as the amount of ester-bound 3-MCPD was determined. Again, free 3-MCPD was not detected in any sample, indicating that free 3-MCPD was not released from the diester by hydrolytic cleavage of the ester bonds. Time-dependent decreasing amounts of ester-bound 3-MCPD were detected in the cell culture supernatant as already shown in Fig. 3c. In case of the cellular lysates, only marginal amounts of ester-bound 3-MCPD were detected; these values were close to the detection limit of the analytical method. Less than 0.1% of the applied dose of 3-MCPD dipalmitate was detected in these lysates at the chosen times of incubation (data not shown), indicating that the diester was not simply absorbed and stored inside the cells. From these data, we conclude that the diester was metabolised by the cells.
According to the development of improved analytical methods in the past decade, 3-MCPD fatty acid esters have been detected in numerous food samples, in some cases in unexpected high amounts. Special attention has been drawn on the finding of significant amounts of 3-MCPD fatty acid esters in infant formula and in follow-up formula by German food control authorities in 2007. The contamination came from palm oil contaminated with 3-MCPD fatty acid esters that was used for the production of the instant formula. In the subsequent risk assessment from the Federal Institute for Risk Assessment, it was stated that infant formula and follow-up formula may contain harmful amounts of 3-MCPD esters, because assuming complete hydrolysis of the esters, the resulting amount of free 3-MCPD would then exceed the recommended TDI about tenfold (Federal Institute for Risk Assessment 2007). Thus, the central question that arose in this context was whether 3-MCPD fatty acid esters are hydrolysed in the human intestinal tract or not.
Seefelder and colleagues showed that 3-MCPD esters are quickly hydrolysed by pancreatic lipase in the presence of porcine bile extract. Under these in vitro conditions mimicking the conditions of the intestinal lumen, 3-MCPD-1-monoesters were completely hydrolysed within 1 min and complete hydrolysis of the diester 3-MCPD-1-palmitate-2-oleate was achieved after 90 min (Seefelder et al. 2008). The authors, however, argued that in particular 3-MCPD fatty acid diesters may not necessarily contribute to the overall burden of free 3-MCPD under in vivo conditions, because due to their structural similarity, it was proposed that 3-MCPD diesters may be metabolised by intestinal cells in the same way as they metabolise diacylglycerols. The intestinal metabolism of diacylglycerols includes the hydrolytic cleavage of the ester bond at sn-1 position, followed by the uptake of the resulting 2-monoacylglycerols by the cell, re-esterification with cellular fatty acids to form triglycerides, packaging of those compounds in the form of triglyceride-rich lipoproteins (chylomicrons) and excretion of those particles into the blood stream. There is, however, so far no experimental evidence that 3-MCPD diesters are metabolised in the same way.
By using the Caco-2 cell line as an in vitro model for the human intestinal barrier, our studies revealed that free 3-MCPD was neither absorbed nor metabolised by the cells. 3-MCPD is a polar but uncharged compound. In general, specific membrane-integral transport proteins are required to transport this type of compounds into the cell. As an example, aquaglyceroporins mediate the cellular uptake of glycerol and related small polar solutes (Asai et al. 2006). In spite of being structurally closely related to glycerol, our data indicate that aquaglyceroporins do obviously not transport 3-MCPD and that Caco-2 cells do not possess any other transporter that accepts 3-MCPD as a substrate.
Although not being absorbed or metabolised by Caco-2 cells, we could show by transport studies that free 3-MCPD is able to migrate through a Caco-2 monolayer. Differentiated Caco-2 cells are known to form a tight monolayer comprising intact tight junctions that function as a barrier for paracellular diffusion processes; however, it was shown for several small polar substances such as mannitol or glucose that these substances can cross that barrier (Ballard et al. 1995; Cogburn et al. 1991). In view of being a small and polar molecule, it was likely that also 3-MCPD was able to migrate through that barrier. As shown in Fig. 1, the concentration of 3-MCPD equilibrated on both sides of the Transwell™ system within 4 h, indicating a simple diffusion kinetics. There was no incidence for an active transport process that might have resulted in the time-dependent enrichment of the substance in one of the two Transwell™ chambers.
In case of 3-MCPD fatty acid esters, it was shown that 3-MCPD-1-monoesters were hydrolysed on the apical side of a polarised Caco-2 monolayer. Therefore, these compounds are accepted as substrates by cellular lipases other than pancreatic lipase that was not present in the experimental set-up. It is likely that the hydrolytic cleavage took place outside the cell, because free 3-MCPD accumulated in the cell culture supernatant and was never detected in cellular lysates. There are numerous possible secreted or membrane-bound esterases that might accept 3-MCPD monoesters as a substrate. As an example, the family of extracellular phospholipases A1 belongs to the pancreatic lipase gene family. These enzymes are expressed and secreted by various mammalian cells and specifically cleave the ester bond at the sn-1 position of phospholipids (Aoki et al. 2007). In case of Caco-2 cells, it was shown that the cells contain lipase activity not only in the cytosolic fraction but also in the apical membrane. This membrane-bound lipolytic activity was capable of converting triacylglycerols to 2-monoacylglycerols (Spalinger et al. 1998).
In contrast to 3-MCPD-1-monoesters, the diester 3-MCPD-1,2-dipalmitate did not release free 3-MCPD upon hydrolytic cleavage. In the presence of Caco-2 cells, the diester disappeared from the cell culture supernatant over the time, but it did not show up in cellular lysates. In these experiments, free 3-MCPD was detected neither in the cell culture supernatant nor in cellular lysates, indicating that the cells had metabolised the diester in a way that the 3-MCPD moiety had been converted to a product that was no longer accessible to our 3-MCPD-specific analytical method. The observed “disappearance” of the diester in the presence of Caco-2 cells might be a hint supporting the assumption by Seefelder et al. (2008). They proposed that 3-MCPD diesters might be metabolised by intestinal cells in the same way as they metabolise diacylglycerols such as phospholipids. Provided that 3-MCPD diesters are accepted as phospholipid-related compounds by the lipid-metabolising enzyme machinery of the cell, 3-MCPD diesters would first be hydrolysed at the sn-1 position outside the cell. The resulting 3-MCPD-2-monoester would then be incorporated by the cell and further re-esterified by the action of the enzyme monoacylglycerol acyltransferase. This enzyme facilitates triacylglycerol formation in intestinal cells very efficiently in vivo; however, the monoacylglycerol pathway was shown to be almost inactive in Caco-2 cells (Levin et al. 1992). In Caco-2 cells, triacylglycerol formation is predominantly mediated via the phosphatidic acid pathway that involves glycerol-3-phosphate acyltransferase in order to synthesise phospholipids (Trotter and Storch 1993). In both scenarios, however, one would expect the substitution of the chloride of the 3-MCPD moiety by an acyl residue in the course of the intracellular re-esterification process. This might simply explain the “disappearance” of the 3-MCPD diester in our experiments. In the future, further studies have to be conducted in order to determine the fate of the 3-MCPD moiety of the diester e.g. by using radioactive labels.
In sum, from a toxicological point of view, 3-MCPD-1-monoesters may simply be considered as additional free 3-MCPD, because we and others (Seefelder et al. 2008) showed that these monoesters are quickly hydrolysed, thereby releasing free 3-MCPD. In case of 3-MCPD diesters, there is some incidence that these compounds may be metabolised by intestinal cells in a way that the 3-MCPD moiety is lost. This observation could be interpreted in a way that intestinal cells may have the capacity to detoxify 3-MCPD diesters. On the other hand, it was shown that 3-MCPD diesters are hydrolysed by pancreatic lipase, thereby releasing free 3-MCPD (Seefelder et al. 2008). Therefore, there are two possible mechanisms for the metabolism of 3-MCPD diesters in vivo, one of them—luminal hydrolysis by pancreatic lipase—leading to an increase in the burden of free 3-MCPD, whereas the second possible mechanism—intracellular metabolism by intestinal cells—would not lead to increased amounts of the free substance. To address this question, toxicokinetic studies have to be conducted in the future to determine which of the two possible mechanisms is favoured under in vivo conditions.
We thank Linda Brandenburger, Albrecht Maier and Dieter Köhl for technical assistance.
Conflict of interest
The authors declare that they have no conflict of interest.