Characterization of the Intestinal Absorption of Arsenate

Jan 15, 2010 - Abstract Image .... One of the most commonly used models is the Caco-2 intestinal cell line, .... transport experiments were carried ou...
0 downloads 0 Views 361KB Size
Chem. Res. Toxicol. 2010, 23, 547–556

547

Characterization of the Intestinal Absorption of Arsenate, Monomethylarsonic Acid, and Dimethylarsinic Acid Using the Caco-2 Cell Line Marta Calatayud, Jose Gimeno, Dinoraz Ve´lez, Vicenta Devesa,* and Rosa Montoro Instituto de Agroquı´mica y Tecnologı´a de Alimentos (CSIC), Apdo. 73, 46100-Burjassot, Valencia, Spain ReceiVed August 12, 2009

Many toxicological studies have been conducted with arsenic species in target organ cell lines. However, although epithelial gastrointestinal cells constitute the first barrier to the absorption of contaminants, studies using intestinal cells are scarce. The present study examines absorption through the intestinal epithelium of the pentavalent arsenic species most commonly found in foods [arsenate, As(V); monomethylarsonic acid, MMA(V); and dimethylarsinic acid, DMA(V)], using the Caco-2 cell line as a model. Different concentrations (1.3-667.6 µM) and culture conditions (media, pH, addition of phosphates, and treatment with ethylenediaminetetraacetic acid) were evaluated to characterize such transport. The apparent permeabilities indicate that the methylated species show low absorption, whereas As(V) is a compound with moderate absorption. The kinetic study shows only a saturable component for MMA(V) transport in the range of concentrations assayed. The existence of paracellular transport was shown for all of the species, with greater significance in the case of the methylated forms. As(V) absorption was inhibited by 10 mM phosphate, and a phosphate transporter therefore could take part in intestinal absorption. Acidification of the medium (pH 5.5) resulted in a marked increase in As(V) and DMA(V) permeability (4-8 times, respectively) but not in MMA(V) permeability. This makes it necessary to consider the possible existence of absorption in the proximal intestine and even in the stomach, where the environment is acidic; alternatively, an H+-dependent transporter may be involved. The results obtained constitute the basis for future research on the mechanisms involved in the intestinal absorption of arsenic and its species, a decisive step in relation to their toxic action. Introduction Water and food constitute the main sources of arsenic exposure for millions of people throughout the world. Of the chemical species found to date in these matrices, the inorganic species, arsenite [As(III)] and arsenate [As(V)],1 are the most toxicsfollowed by dimethylarsinic acid [DMA(V)], monomethylarsonic acid [MMA(V)], and the tetramethylarsonium ion. Other arsenical species found in foods (trimethylarsine oxide, arsenocholine, arsenobetaine, arsenosugars, and arsenolipids) are considered to be essentially nontoxic. The arsenic present in water, fundamentally in inorganic form, is regarded as a human carcinogen (type IA) (1). Human exposure to inorganic arsenic is also associated with the development of dermatological disorders (hyperkeratosis and hyperpigmentation), cardiovascular alterations (hypertension), endocrine disease (increased incidence of type 2 diabetes mellitus), neurobehavioral disorders (lowered intelligence quo* To whom correspondence should be addressed. Tel: (+34)963 900 022. Fax: (+34)963 636 301. E-mail: [email protected]. 1 Abbreviations: As(V), arsenate; MMA(V), monomethylarsonic acid; DMA(V), dimethylarsinic acid; MMA(III), monomethylarsonous acid; DMA(III), dimethylarsinous acid; AQ7, aquaporin 7; AQ9, aquaporin 9; Glut 1, glucose transporter isoform-1; ECACC, European Collection of Cell Cultures; DMEM, Dulbecco’s modified Eagle’s medium; FBS, fetal bovine serum; NEAA, nonessential amino acids; DMEMc, complete Dulbecco’s modified Eagle’s medium; TEER, transepithelial electrical resistance; LY, Lucifer yellow; Papp, apparent permeability coefficient; HBSS, Hanks buffered solution salts; PBS, phosphate-buffered saline; MES, o-2-(Nmorpholine) ethanesulfonic acid; Pi, inorganic phosphate; OATP2B, organic anion-transporting polypeptide 2B; OATP-C, organic anion-transporting polypeptide C.

tient), and respiratory problems (2-5). The pentavalent methylated species also exhibit a degree of toxicity. Chronic DMA(V) exposure through drinking water in rats and mice has been shown to promote tumors in different organs and to act as a carcinogen in rat bladder malignancy (6). In contrast, MMA(V) does not appear to be carcinogenic, since 2 years of exposure to MMA(V) through food in rats yielded no tumors, although non-neoplastic lesions were identified in the rectum, colon, and cecum (7). Many toxicological studies have been made with As(V), As(III), and their metabolites [monomethylarsonous acid, MMA(III); MMA(V); dimethylarsinous acid, DMA(III); and DMA(V)] in target organ cell lines: hepatocytes, bladder cells, and keratinocytes. These studies have contributed information on the metabolism of arsenical species (8-10), the characterization of transporters (11-14), and the determination of mechanisms of cyto- and genotoxicity (15-17). However, although the epithelial cells of the gastrointestinal tract constitute the first barrier to contaminants penetrating the body via the oral route, few studies using intestinal cells have been published to date (18, 19). The evaluation of population exposure to arsenic is generally based on the estimation of arsenic intake, using the data obtained from food and drinking water analyses. Such an evaluation offers only an approximation that fails to consider a parameter that is crucial to toxic exposure: bioavailability. This parameter is defined as the fraction of an ingested compound that is solubilized in the course of gastrointestinal digestion and which reaches the systemic circulation following absorption from the gastrointestinal lumen. The resulting bioavailability is a con-

10.1021/tx900279e  2010 American Chemical Society Published on Web 01/15/2010

548

Chem. Res. Toxicol., Vol. 23, No. 3, 2010

sequence of different processes: ingestion of the toxin through food and/or drinking water, toxin release during the digestion process, absorption by the intestinal epithelium, and first pass metabolism. The study of the processes involved in the pathway that translocates arsenicals from the lumen of the gastrointestinal tract to the systemic circulation provides a better understanding of the determinants of bioavailability. Although intestinal absorption is clearly a decisive step in relation to bioavailability, it is also the least studied step. Different methods have been developed for analyzing intestinal absorption. One of the most commonly used models is the Caco-2 intestinal cell line, established from human colon adenocarcinoma cells. These cells are able to differentiate spontaneously, giving rise to a monolayer possessing many of the functional and morphological features of mature human enterocytes. The differentiated cell monolayer is polarized, with microvilli on the apical border, intercellular tight junctions, secretion of enzymes inherent to the brush border membrane, and the expression of transporters characteristic of the small intestine in the apical and basolateral membranes (20). For this reason, permeability through the Caco-2 cell monolayer is used to predict in vivo permeability in humans, to examine absorption mechanisms, and to study the effect of transporters in relation to drug (21-23) and micronutrient permeability (Ca, Fe, Zn, and Cu) (24-26). However, this cell line has been little used in application to the study of food contaminants and particularly toxic trace elements. With regard to arsenic, the Caco-2 cell line has only been used in application to the study of the uptake and transport of As(V), As(III), and organoarsenical standard solutions and of the arsenic present in rice, seaweed, and seafood products (18, 19). Some studies have characterized the role of certain transporters in the cellular uptake of arsenical species. Liu et al. (11, 27) have shown the transport of As(III) and MMA(III) by members of the aquaporin family [aquaporin 7 (AQ7) and aquaporin 9 (AQ9)]. It has also been shown that the glucose permease glucose transporter isoform-1 (Glut 1) facilitates the transport of As(III) and MMA(III) (12). Some studies point to the possibility of As(V) transport mediated by the phosphate transporters (14). Most of these studies have been made in cell models unrelated to the intestinal epithelium or involving trivalent species. On the other hand, some of the transporters studied have a low expression in the intestinal epithelial cells, as is the case with aquaporins AQ7 and AQ9 (28). As a result, there are few data on the mechanisms of transport of pentavalent arsenical species through the intestinal epithelium. The present study examines absorption through the intestinal epithelium of the pentavalent arsenical species most commonly found in foods [As(V), MMA(V), and DMA(V)] (29), using the Caco-2 cell line model. The study of apical-basal and basal-apical uptake and transport using different arsenic concentrations and culture conditions offers novel information involved in intestinal absorption.

Materials and Methods Arsenical Species. The As(V) standard solution (1000 mg/L) was acquired from Merck (Merck, Darmstadt, Germany). The standard solutions of MMA(V) and DMA(V) were prepared by dissolving the appropriate amount of the following salts in water: CH3AsO(ONa)2·6H2O for MMA(V) (Carlo Erba, Italy) and (CH3)2AsNaO2·3H2O for DMA(V) (Fluka Chemika Biochemika, Spain). Cell Culture. The Caco-2 cells were obtained from the European Collection of Cell Cultures (ECACC; number 86010202, Salisbury, United Kingdom). Cell maintenance was carried out in 75 cm2 flasks

Calatayud et al. to which 10 mL of Dulbecco’s modified Eagle’s medium (DMEM) with glucose (4.5 g/L) and pyruvate (Gibco, BRL Life Technologies, Paisley, Scotland) were added at pH 7.4. The DMEM was supplemented with 10% (v/v) fetal bovine serum (FBS) (Gibco), 1% (v/v) nonessential amino acids (NEAAs) (Gibco), 1% (v/v) 1 M HEPES (N-2-hydroxyethyl piperazine-N′-2-ethanesulfonic acid), 1% (v/v) antibiotics (penicillin/streptomycin) (Gibco), and 1% (v/ v) fungizone (Gibco) (complete Dulbecco’s modified Eagle’s medium, DMEMc). The cells were incubated at 37 °C in an atmosphere with 95% relative humidity and a CO2 flow of 5%. The medium was replaced every 2 days. When the cell monolayer reached 80% confluence, the cells were detached with a solution of trypsin (2.5 g/L) and ethylenediaminetetraacetic acid (EDTA) (0.2 g/L) (Gibco), followed by reseeding at a density of 5 × 104 cells/cm2. Cell differentiation and the posterior tests were carried out in double chamber wells (24 mm diameter; pore size, 0.4 µm; Transwell, Costar Corp., NY) equipped with a porous support on which the Caco-2 cells form monolayers allowing the diffusion chamber to be divided into two compartments: apical (upper) and basal (lower). The cells were seeded at a density of 7.5 × 104 cells/ cm2, adding 1.5 mL of culture medium to the apical compartment and 2 mL to the basal compartment. The cells were incubated at 37 °C in an atmosphere with 95% relative humidity and a CO2 flow of 5%, replacing the medium every 2-3 days until cell differentiation was reached (14-15 days postseeding). All of the cultures were used between passages 66 and 70. Integrity of the Cell Monolayer. The monolayer integrity was assessed by measuring the transepithelial electrical resistance (TEER) and the permeability of the paracellular transport marker lucifer yellow (LY). A Millicell-ERS (Millipore Corp.) was used for the TEER measurements. During the growth and differentiation period, the cell monolayer status was evaluated daily from the sixth postseeding day onward. The cell monolayer was considered to be fully formed when stable values of g200 Ω cm2 were obtained. During the uptake, transport, and permeability experiments, TEER was measured at different time points, including the start and end of the experiments. The permeability coefficient of LY, which is transported only through the intercellular junctions, is used to assess the integrity of the monolayer. Permeabilities of under 0.5 × 10-6 cm/s are taken to be indicative of an intact monolayer. LY permeability was measured adding the marker (100 µM) to the apical or basal compartment of the control wells and the arsenic-treated wells. LY fluorescence was measured at an excitation/emission wavelength of 485/520 nm on the acceptor side, using a microplate fluorescence reader (PolarSTAR OPTIMA, BMG-Labtech, Germany). To evaluate possible interactions between LY and arsenical species uptake and transport, parallel experiments were made with and without the paracellular marker, which revealed the absence of interferences between the LY and the studied arsenical species. Cell Viability. At the end of each experiment, the number of viable cells was determined with 0.4% trypan blue solution (Sigma) (30). Uptake, Transport, and Permeability Experiments. The uptake and transport experiments were carried out in DMEMc medium for 72 h. The apparent permeability (Papp) tests were conducted in Hanks buffered solution salts (HBSS) with NaCO3 (Sigma, Spain), supplemented with 20 mM HEPES (pH 7.2). Before the experiment was started, the cells were kept in contact with the media for 15 min to allow adjustment. The standard solutions of As(V), MMA(V), and DMA(V) were prepared in the corresponding media and were added to the acceptor compartment, apical or basal, according to the direction of transport studied, apical-basolateral or basolateral-apical, respectively. The concentrations of the tested arsenical species in DMEMc were as follows: 1.3, 6.7, 13.4, 66.8, and 667.6 µM for As(V); 1.3, 6.7, 13.4, and 66.8 µM for MMA(V); and 6.7, 13.4, and 66.8 µM for DMA(V). The concentration of the tested arsenical species in HBSS was 66.8 µM.

Intestinal Absorption of As(V), MMA(V), and DMA(V)

Chem. Res. Toxicol., Vol. 23, No. 3, 2010 549

At the established time points, samples (300 µL) were removed from the acceptor compartment and were replaced with an equal volume of fresh medium. In the tests with DMEMc medium, aliquots were obtained every 15 min up to 120 min and after 4, 6, 8, 24, 48, and 72 h. In the tests with HBSS medium, aliquots were obtained every 15 min up to 120 min and after 4, 6, and 8 h. The total arsenic determination was carried out in the aliquots obtained at each time point as well as in the cell monolayer and in the donor medium collected at the end of the experiment. The apparent permeability coefficients (Papp) were calculated from eq 1:

Papp ) (dC/dt)(Vr /ACo)

(1)

where dC/dt is the flux (µmol/s) determined by the slope of the cumulative concentration (µM) of arsenic in the receptor chamber over time (s), Vr is the acceptor compartment volume (apical, 1.5 mL; basal, 2 mL), A is the surface area occupied by the cell monolayer (4.71 cm2), and Co is the initial concentration in the donor compartment (µM). Study of Paracellular Transport. Paracellular transport was evaluated through the modulation of intercellular junctions using 5 mM EDTA in phosphate-buffered saline (PBS) without Ca2+ and Mg2+. The cell monolayers were preincubated with the PBS-EDTA solution for 5 min. Posteriorly, we independently added 1.5 mL of standard solution of arsenic species (66.8 µM) prepared in HBSS medium without Ca2+. Incubation was carried out for a maximum of 240 min, collecting aliquots after 15, 30, 60, 120, and 240 min. These aliquots were in turn used to determine the concentrations of total arsenic, with calculation of the Papp values using eq 1. The efficiency of tight junction modulation with EDTA was monitored using the paracellular marker LY and the TEER values. Effect of pH upon Permeability. The study to determine the influence of pH upon As(V) (66.8 µM), MMA(V) (33.4 µM), and DMA(V) (66.8 µM) transport was carried out in HBSS medium supplemented with o-2-(N-morpholine) ethanesulfonic acid (MES) (pH 5.5) on the apical side and HBSS-HEPES (pH 7.2) on the basal side. During the test period (240 min), aliquots were collected after 15, 30, 60, 120, and 240 min for the determination of total arsenic. The Papp values were calculated from eq 1. Effect of Phosphate upon As(V) Transport. To investigate the influence of phosphate upon the transport of As(V) (66.8 µM), permeability was assessed in phosphate-free saline medium at pH 7.2 (mM composition: NaCl, 130; KCl, 10; CaCl2, 1; MgCl-6H2O, 1; NaHCO3, 4; D-glucose, 5.5; and HEPES, 20) and in the same medium added with inorganic phosphate (Pi) in the form of NaH2PO4 (1.1 and 10 mM). During the test period (240 min), aliquots were collected after 30, 60, 90, 120, 180, and 240 min. These aliquots were used to determine total arsenic, and the Papp values were calculated from eq 1. Determination of Arsenic. Deionized water (18.2 MΩ cm) was used for the preparation of reagents. All reagents used were of analytical grade or higher: 65% nitric acid (Merck), magnesium nitrate hexahydrate (Merck), palladium (Fluka Chemie), and Triton X-100 (Merck). The total arsenic concentration in the media and in the cells was quantified using atomic absorption spectrometry with a Zeeman graphite chamber (model Analyst 600, Perkin-Elmer Hispania, S.A., Spain). Graphite tubes with an inserted L’vov platform were used. The graphite furnace program [temperature (°C)/ramp time (s)/ heating time (s)] used for the determination of arsenic was as follows: drying (120 °C/10 s/20 s; 300 °C/5 s/15 s), pyrolysis (900 °C/10 s/20 s; 1400 °C/10 s/30 s), atomization (2100 °C/0 s/5 s), cleaning (2450 °C/1 s/3 s), and cooling (20 °C/0 s/30 s). The matrix modifier used consisted of a mixture of 0.05 mg of palladium (Pd) and 0.003 mg of Mg(NO3)2 in 10 µL of 1% HNO3 (v/v). A certified reference material was used to assess the accuracy of the total arsenic determinations [EnviroMAT drinking water (EP-H-2), SCP Science, Canada]. For arsenic speciation, the samples were analyzed without prior treatment, using hydride generation coupled to a cryogenic trap

and atomic absorption spectrometry (31). With this methodology, the arsines of the trivalent arsenical species [As(III), MMA(III), and DMA(III)] are selectively generated at pH 6. At pH 1, the arsines of the trivalent and pentavalent arsenical species are generated. Statistical Analysis. All tests were performed in triplicate in independent cultures. Data are means ( SE. The results were subjected to statistical analysis using single-factor analysis of variance, carried out by SPSS version 15.0 statistical package. The differences were considered significant for p < 0.05. Nonlinear regression for arsenic species data in DMEMc was fitted to a Michaelis-Menten equation using Graph-Pad Prism version 4.

Results Apical-Basolateral Uptake and Transport in DMEMc. The control of membrane integrity based on TEER and cell viability showed both of these parameters to decrease after 72 h of treatment with the highest As(V) concentration (667.6 µM): TEER decreased 38% and cell viability 42%. In contrast, neither of these parameters was affected by lower As(V) concentrations or by exposure to any of the tested concentrations of the methylated species [MMA(V) and DMA(V)]. The evaluation of membrane integrity using LY yielded different results. Up to 24 h, the percentages of LY transport to the basolateral side and the Papp coefficient proved normal (1.2-1.5%; Papp < 0.2 × 10-6 cm/s) following exposure to any of the concentrations of As(V), MMA(V), and DMA(V), indicating that the intercellular junctions remain functional and that membrane integrity is complete. In contrast, a significant increase in LY transport was recorded after 48 and 72 h (increase, 23-34%; Papp 0.6-1 × 10-6 cm/s), showing intercellular junction damage, not evidenced by the TEER measurements. This behavior beyond 24 h was also seen in the controls not exposed to the arsenical species. Consequently, such junctional damage was not an effect of treatment with the arsenical species but of the culture maintenance time. Taking these data into account, it is considered that transport up to 24 h takes place through the intact cell monolayer, while the values recorded after 48 and 72 h may be due to destruction or disruption of the intercellular junctions and thus must be interpreted with some caution. For this reason, the present study only takes into account exposure times of 24 h or less. Figure 1 shows apical-basolateral transport (expressed as ng of As/106 cells) versus time corresponding to the concentrations of As(V) (Figure 1A), MMA(V) (Figure 1B), and DMA(V) (Figure 1C). For each of the tested concentrations of As(V) and DMA(V), transport increases linearly versus exposure time for up to 8 h. However, for MMA(V) (66.8 µM), two welldifferentiated segments are seen in the transport curve. Up until 2 h, there is a very marked increase in transport toward the basolateral side, with a significant reduction in transport rate after this 2 h time point. In the case of DMA(V), transport was detected from the start of the experiment (15 min), in contrast to the situation observed for both MMA(V) and As(V). Figure 1 also shows the amount (in ng of arsenic/106 cells) transported versus concentration for As(V), MMA(V), and DMA(V) after 24 h of exposure. For As(V) and DMA(V), the amount reaching the basolateral side increases linearly with the increase in concentration. In the case of MMA(V), after 24 h of exposure, this linear trend was only maintained up to 13.4 µM. These results reflect a difference in behavior of MMA(V) versus As(V) and DMA(V). When the data for arsenical species were fitted to a Michaelis-Menten equation, it was seen that only the transport of MMA(V) had a saturable component. The estimated Km value was 2.02 µM and the Vmax was 4.6 × 10-5 pmol 10-6 cells min-1.

550

Chem. Res. Toxicol., Vol. 23, No. 3, 2010

Calatayud et al.

Figure 1. Apical-basolateral transport in DMEMc medium. (A) Transport of As(V) to basolateral compartment, expressed as ng of total arsenic/ 106 cells, vs time, following Caco-2 cell exposure to 1.3 (b), 6.7 (2), 13.4 (9), and 66.8 µM ([). The subgraph represents the total arsenic content in basolateral compartment in ng of arsenic/106 cells at 24 h vs concentration of As(V) added to the apical compartment. (B) Transport of MMA(V) to basolateral compartment, following Caco-2 exposure to 1.3 (b), 6.7 (1), 13.4 (9), and 66.8 µM ([). The subgraph represents the total arsenic content in basolateral compartment in ng of arsenic/106 cells at 24 h vs concentration of MMA(V) added to the apical compartment. (C) Transport of DMA(V) to basolateral compartment following Caco-2 exposure to 6.7 (b), 13.4 (1), and 66.8 µM (9). The subgraph represents the total arsenic content in basolateral compartment in ng of arsenic/106 cells at 24 h vs concentration of DMA(V) added to the apical compartment. Bars represent the standard deviation from at least three replicates.

Intestinal Absorption of As(V), MMA(V), and DMA(V)

Figure 2. Apical-basolateral transport following Caco-2 cell exposure to 66.8 µM As(V) (b), 66.8 µM MMA(V) (1), and 66.8 µM DMA(V) (9) in HBBS. Bars represent the standard deviation from at least three replicates.

The cell retention of the pentavalent arsenical species was seen to be low (