The Journal of Nutrition Vol. 128 No. 10 October 1998, pp. 1794-1801

Physiologic Concentrations of Zinc Affect the Kinetics of Copper Uptake and Transport in the Human Intestinal Cell Model, Caco-2^1,2,3

Philip G. Reeves⁴, Mary Briske-Anderson, and LuAnn Johnson

U.S. Department of Agriculture, Agricultural Research Service, Grand Forks Human Nutrition Research Center, Grand Forks, ND 58202-9034

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Copper uptake was enhanced and copper transport was reduced in Caco-2 cells cultured in media containing high concentrations of zinc. Here we show that physiologic zinc concentrations also affect copper movement into and out of Caco-2 cells. Cells were seeded onto Falcon membranes with high pore density and maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, nonessential amino acids, glucose and glutamine. In one experiment, the cells were exposed to media containing either 5 or 25 µmol zinc/L from d 14 to 21 after seeding. Then, copper uptake and transport, in both apical and basolateral directions, were measured by using ⁶⁴Cu. Cells exposed to 25 µmol zinc/L had a 25% higher (P < 0.05) uptake of ⁶⁴Cu from the apical side than those exposed to 5 µmol zinc/L. There was no effect of zinc on ⁶⁴Cu uptake from the basolateral side, even though the amount of label taken up was as much as threefold higher (P > 0.05) than from the apical side. Transport of ⁶⁴Cu across the cell layer in both directions was 50% less (P < 0.05) in cells exposed to 25 µmol zinc/L vs. 5 µmol zinc/L. In another experiment, zinc-exposed cells were labeled with ⁶⁴Cu and efflux of the label to the apical and basolateral sides was measured over time. The rate of efflux to the apical side was linear and not affected by zinc. However, there was a 37% reduction (P < 0.05) in ⁶⁴Cu efflux to the basolateral side by the higher zinc concentration. Curve-fit analysis showed that the basolateral efflux was made up of an exponential and a linear component. Cellular zinc concentrations were proportional to the zinc concentrations in the media. Although the data suggest that high media zinc inhibited the copper efflux transporter and enhanced the influx transporter, copper did not accumulate in the cell.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Higher than normal amounts of zinc in the diet can lower the copper status of both humans and animals. The mechanism has not been completely defined, but apparently it involves an interaction between zinc and copper at the intestinal level (Oestreicher and Cousins 1985, Van Campen and Scaife 1967). An earlier hypothesis for the mechanism states that zinc-induced intestinal metallothionein (MT)⁵ sequesters copper and inhibits its absorption (Fischer et al. 1981 and 1983; Ogiso et al. 1974 and 1979). Copper then accumulates in the enterocyte and is lost from the body by sloughing of the cells. We have evidence that copper does not accumulate in the enterocyte under these conditions and that copper status can be reduced with high dietary zinc in mice that do not possess the MT gene; therefore, they cannot make MT (Masters et al. 1994, Reeves 1998). This suggests that MT induction by zinc is not entirely responsible for reduced copper absorption.

Recent studies have brought attention to the mechanism of copper transport in cells because of the discovery that the etiology of Menkes disease resides in a genetic mutation of a membrane-bound copper transporting P-type ATPase (Mercer et al. 1993, Vulpe et al. 1993). The malfunction of this protein may retard the movement of copper across the enterocyte basolateral membrane into the blood. Whether this copper-transporter is affected directly by high zinc is not known.

Reeves et al. (1996) showed recently that high concentrations of zinc in the medium of Caco-2 cells caused a significant reduction in the rate of copper transport across the monolayer. This investigation is continued in this report and shows that the rate and kinetics of copper transport in this cell type also are affected by physiologic concentrations of zinc. The data also suggest that the mechanism may involve an inhibition of copper efflux from the cell, but that copper does not accumulate in the cell.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Reagents.  Caco-2 cells were obtained from the American Type Culture Collection (Rockville, MD) at passage 17. Falcon polyethylene terephthalate high density (PET-HD; 1.0 × 10⁸ pores/cm², 0.45-µm pore size) membrane inserts were purchased from Becton-Dickinson Labware (Lincoln Park, NJ). Cell culture media and chemicals were obtained from Sigma Chemical (St. Louis, MO) and/or Gibco Laboratories (Grand Island, NY). Radioactive copper (⁶⁴Cu; specific activity, 18.9 GBq/µg Cu) was prepared at the Missouri University Research Reactor (Columbia, MO). Nonradioactive copper and zinc were added to cell culture media as solutions of the minerals CuSO₄·5H₂O and ZnSO₄·7H₂O, respectively.

Cell culture.  After the cells were subcultured to passages between 26 and 37, they were seeded (58,000 cells/cm²) onto PET-HD membrane inserts that hung inside the chambers of six-well plates. Growth medium [1.5 mL; Dulbecco's modified Eagle's medium (DME) containing an additional 100 mL of fetal bovine serum (FBS), 4.5 mg glucose, 4 mmol glutamine, and 0.1 mmol of nonessential amino acids/L] was placed inside the insert (apical side) and 2.5 mL was placed in the chamber (basolateral side). The culture medium was changed at 3-d intervals until the cells had grown to confluence, differentiated and had begun to display characteristics of intestinal epithelium (6-9 d); then the medium was changed every 2 d until the experiment was terminated (21 d).

Time course of ⁶⁴Cu uptake and transport.  On the day of the experiment, 1.5 mL of fresh media was placed in each insert and 2.5 mL in each chamber. After 4 h of equilibration, the experiment began. When studying uptake and transport in an apical to basolateral direction, 0.1 mL of medium containing 62 kBq of ⁶⁴Cu/nmol copper was placed into the insert. When studying uptake and transport in a basolateral to apical direction, 0.1 mL of labeled medium was placed in the chamber. Both apical and basolateral media contained 0.5 µmol copper and 5 µmol zinc/L. All labeling procedures were conducted at 37°C with the plates on a plate warmer. For long intervals, the labeled plates were returned to the incubators; for short intervals, the plates remained on the warmer.

At prescribed intervals after the addition of radioactive copper, inserts were removed and put through a series of ice-cold washes to remove residual copper. These washes consisted of 10 s in HEPES buffer (10 mmol HEPES, 140 mmol NaCl, 7 mmol KCl and 5.6 mmol glucose/L, pH 7.4), 10 s in EDTA-HEPES (10 mmol EDTA/L of HEPES buffer), and again for 10 s in HEPES buffer. Excess buffer was aspirated from inside the inserts, and the outside of the inserts was blotted to remove residual buffer. The membranes were cut out, placed individually into small tubes and immediately assayed for radioactivity in a gamma counter (Packard Instruments, Cobra Auto-Gamma, Meriden, CT). Because the half-life for ⁶⁴Cu is only 12.7 h, the counter was programmed to correct the counts to the beginning of the experiment. These values were used to compute the amount of ⁶⁴Cu taken up into the cell and were expressed as a fraction of moles of copper per milligram of cellular protein.

When the counting was done, the cells were dissolved away from the membranes with a solution (2 mL/filter) of 200 mmol NaOH and 7 mmol sodium dodecylsulfate (NaOH-SDS)/L. Protein content of this mixture was determined by the bicinchoninic acid method (Sigma Chemical). Medium remaining in each chamber was quantitatively pipetted into small tubes and counted. These values were used to determine the amount of ⁶⁴Cu traversing the cell monolayer and were expressed as a fraction of moles of copper per milligram of cellular protein.

Time course of ⁶⁴Cu efflux from cells.  On d 20 after the cells were seeded, the growth medium was replaced with fresh medium containing 47 kBq of ⁶⁴Cu/nmol of copper. Twenty-four hours later, the inserts were removed, washed for 10 s in warm saline and placed in new chambers, each containing 2 mL of DME-FBS without ⁶⁴Cu. One milliliter of this medium was placed inside the insert and the plates were incubated at 37°C. At prescribed intervals, the inserts were removed, the apical media were quantitatively transferred to counting tubes and the inserts were washed and treated as described above. The basolateral media were quantitatively transferred to counting tubes. Both apical and basolateral media were assayed for ⁶⁴Cu, and the values were used to calculate the amount of copper released from the cells with time. The values were expressed as a fraction of moles of copper per milligram of cellular protein.

Concentration-dependent copper uptake and transport.  The preparation for this portion of the study was similar to that for the time course of ⁶⁴Cu uptake and transport, except that the apical medium contained a series of added copper concentrations ranging from 0.32 to 126 µmol/L. ⁶⁴Cu (69 kBq) was added to each insert at each copper concentration. This meant, of course, that the specific activity of ⁶⁴Cu decreased as the amount of unlabeled copper increased. However, because the amount of copper taken up and/or transported by the cell increased as the copper concentrations increased, there was enough ⁶⁴Cu to be efficiently detected at all points. Fifteen minutes after addition of the radioactive copper, the cells were harvested as described above and the amounts of ⁶⁴Cu entering the cells and crossing the cell monolayer were determined. The values were expressed as a fraction of moles of copper per milligram of cellular protein.

Effect of changing the concentrations of zinc in the culture medium on ⁶⁴Cu uptake and transport.  When the effect of zinc concentration in the culture medium on copper uptake and transport was to be determined, the cells were prepared in the same manner as described above, except that from d 14 to 21 after seeding, cells were incubated in media containing various concentrations of zinc, including 5, 25, 35 or 140 µmol/L.

Determination of zinc and copper concentrations in cells and culture media.  Zinc concentrations in the media were determined by flame atomic absorption spectrometry (FAAS) after the samples had been diluted appropriately with deionized water. Copper was determined on undiluted media by graphite furnace atomic absorption spectrometry (GFAAS). When cellular minerals were determined, aliquots of cells were placed in HNO₃ (3.2 mol/L) in mineral-free polypropylene tubes, capped, and digested overnight in a waterbath at 37°C. Samples were diluted appropriately and analyzed for copper and zinc by GFAAS and FAAS, respectively. Commercially available internal and external standards of zinc and copper were analyzed simultaneously with the samples to ensure adequate quality controls (Sigma Chemical).

Statistical analysis.  Six replicates were used for each time interval and each copper concentration. Because the variability of replicate experiments can change between cells of different ages, we attempted to maintain the cell passage number for all experiments in a narrow range of 26 to 37.

Empirical models were used to analyze the progression of copper uptake and efflux over time (Sokal and Rohlf 1969). To analyze uptake over time in the apical to basolateral direction, a power function was used of the form

(1)

where  is a scaling coefficient, k_U is a rate constant and t is time (min). The power function is not considered to be a physiological model of the uptake process in this case. However, it closely approximated the behavior of the observed uptake in the region of interest, thus allowing the comparison of the rate constants between the two concentrations of zinc.

The equation used to model the amount of uptake over time in the basolateral to apical direction was of the form

(2)

where U_max is the maximal amount of uptake (pmol/mg protein), k_U is the first-order rate constant (min¹), t is time (min) and e is the base of the natural system of logarithms. The initial rate of uptake was estimated by taking the first derivative of Equation 2 at t = 0. An estimate of the corresponding standard error was obtained by using the standard errors of the regression coefficients and propagation of error formulas. Student's t test was used to compare these initial rates for the two concentrations of zinc.

The transport of copper from the basolateral to the apical side, as well as from the apical to the basolateral side, was modeled using the power function

(3)

where  is a scaling coefficient and k_T is a rate constant.

To analyze the efflux of copper to the basolateral side of the cell layer, the following exponential model was used:

(4)

where E_max is the maximal amount of efflux (pmol/mg protein), k_E1 is the first-order rate constant (min¹) associated with an exponential or a rapid component of efflux, k_E2 is the first-order rate constant [pmol/(min·mg protein)] associated with a linear term that was used to approximate a slower compartment of efflux, t is time (min) and e is the base of the natural system of logarithms. Because the observed efflux of ⁶⁴Cu to the apical side was somewhat erratic and highly variable, simple linear regressions were used to model its behavior and to compare effects of media zinc.

A three-parameter rectangular hyperbola model was used to analyze the dependence of the rate of copper uptake on copper concentration. The form of this model is as follows:

(5)

where [Cu] represents the concentration of ⁶⁴Cu (µmol/L) in the media, J_max,U is the maximum rate of Cu uptake [pmol/(min·mg protein)], K_U is the media copper concentration at the half-maximal rate of uptake (µmol/L) and k_d is a first-order rate constant for what was considered to be a nonsaturable or diffusional component of uptake [nL/(min·mg protein)].

The nonlinear regression procedure in SAS/STAT (SAS Institute, Cary, NC) was used to estimate the coefficients for all nonlinear models; the linear regression procedure in SAS/STAT was used to model efflux to the apical side. Multiple regression techniques were used to test whether the estimated values of the coefficients (e.g., J_max,U, K_U or k_d) in the models were significantly different (P < 0.05) between the different concentrations of zinc in the media or if a single set of coefficients could be used to model the data from all experimental concentrations of zinc. Because the models used were nonlinear, the standard errors of the regression coefficients are asymptotic approximations.

The SigmaPlot computer program (SPSS, Chicago, IL) was used to generate the graphs. Individual points in the graphs are means ± SEM for six replicate samples. In some cases, the error bars are hidden by the plot symbols.

The amounts of label taken up, released or transported at 160 min were statistically analyzed by using two-way ANOVA. The concentrations of zinc and copper in Caco-2 cells were statistically analyzed by using one-way ANOVA followed by Tukey's contrasts (Tukey 1949).

Definition of the experimental system.  We used the term "uptake" to define the movement of ⁶⁴Cu into the cell, the terms "efflux" or "release" to define the movement of the label out of the cell, and the term "transport" to define the movement of ⁶⁴Cu across the cell monolayer. At the beginning of each labeling experiment, we assumed that the system was in equilibrium with respect to unlabeled copper. In some experiments, only a small amount of labeled copper was added to the system; therefore, we assumed that the system was not perturbed, that no net movement of copper occurred and that the tracer followed linear system kinetics (Jacquez 1987). In other experiments, enough copper was added to the system to upset the equilibrium. In these cases, we assumed that net movement of copper did occur in an attempt to re-establish equilibrium.

	RESULTS

Abstract Introduction Methods Results Discussion References

Time course of ⁶⁴Cu uptake and transport.  The data for ⁶⁴Cu uptake into Caco-2 cells from the apical side of the membrane were fitted to Equation 1. The results showed that uptake was significantly (P < 0.05) higher in cells incubated in the presence of 25 µmol zinc/L than in 5 µmol zinc/L (Fig. 1, upper panel). Uptake of ⁶⁴Cu from the basolateral side of the membrane was curvilinear (Fig. 1, lower panel), and the data were fitted to the model shown in Equation 2. We obtained initial rates (rate ± SEM) of 1.84 ± 0.30 and 1.75 ± 0.40 pmol/(min·mg protein) for 5 and 25 µmol zinc/L, respectively. There was not a significant (P > 0.05) difference in initial rates between zinc concentrations. At 160 min, the amount of copper taken up from the basolateral side was three to four times greater than that from the apical side (P < 0.001).

View larger version (21K): [in this window] [in a new window]   Fig 1. Media-zinc concentration affects 64Cu uptake into Caco-2 cells. Cells were cultured for 7 d in media containing either 5 or 25 µmol zinc/L. Apical or basolateral media were then labeled with 64Cu without changing the total copper concentration, and the amount of label taken up into the cell monolayer was determined with time. Individual points are the mean ± SEM for six replicate membranes. The rate of movement of 64Cu into the cells from the apical media was significantly (P < 0.05) greater when cells were cultured with 25 µmol zinc/L than with 5 µmol zinc/L (upper panel). Zinc had no effect on the rate of 64Cu movement from the basolateral side into the cell layer (lower panel).

High media zinc affected ⁶⁴Cu transport across the monolayer differently than it affected ⁶⁴Cu uptake into the cell (Fig. 2). We fitted these data to the model in Equation 3 and found that the rate of transport across the monolayer incubated with 25 µmol zinc/L for 7 d was significantly (P < 0.05) lower than when the monolayer was incubated with 5 µmol zinc/L. This difference occurred in both directions, apical to basolateral or basolateral to apical. This amount (25 µmol/L) decreased the amount of ⁶⁴Cu transported from apical to basolateral or from basolateral to apical side by ~50% (P < 0.05) in 160 min.

View larger version (22K): [in this window] [in a new window]   Fig 2. Media-zinc concentration affects 64Cu transport across a monolayer of Caco-2 cells. Cells were cultured for 7 d in media containing either 5 or 25 µmol zinc/L. Apical or basolateral media were then labeled with 64Cu without changing the total copper concentration, and the amount of label transported across the cell monolayer was determined with time. Individual points are the mean ± SEM for six replicate membranes. At 160 min, the amount of 64Cu transport across cells cultured with 25 µmol zinc/L of media was ~50% less (P < 0.05) than that in cells cultured with 5 µmol zinc/L. Zinc affected copper movement in both directions in a similar manner.

Time course of ⁶⁴Cu efflux from cells.  Caco-2 cell monolayers were labeled with ⁶⁴Cu and the amounts released to both the apical and basolateral cell sides were measured. The efflux of ⁶⁴Cu to the apical side of the membrane was linear but somewhat erratic and with a large standard error. Elevating media zinc had no apparent effect on the rate of copper efflux (P > 0.05) (Fig. 3, upper panel).

View larger version (22K): [in this window] [in a new window]   Fig 3. Media-zinc concentration affects the efflux of copper from Caco-2 cells labeled with 64Cu. Cells were cultured for 7 d in media containing either 5 or 25 µmol zinc/L. Cells were then incubated for 24 h with 64Cu in complete media without changing the total copper concentration. Cells were washed and the amount of label released from the cell monolayer was determined over time. Individual points are the mean ± SEM for six replicate membranes. Media-zinc concentration had no effect on 64Cu efflux to the apical side of the cell layer, but efflux to the basolateral side was significantly (P < 0.05) less for cells incubated with 25 µmol zinc/L than those incubated with 5 µmol zinc/L (see Fig. 4 and Table 1).

A nonlinear regression analysis of the data was performed for basolateral efflux, by using the model illustrated in Equation 4 (Fig. 3, lower panel). With a longer time frame, it might have been more appropriate to use a double exponential model; however, with the short time frame used here, the model in Equation 4 gave a much better fit. The results showed that copper efflux was composed of two discernible components at both concentrations of zinc (Fig. 4). maximal efflux, E_max, for the exponential component was ~50% less in cells incubated in 25 µmol zinc than those incubated in 5 µmol zinc (P < 0.05) (Table 1). Although high media-zinc concentration tended to enhance the exponential rate constant, k_E1, compared with that in cells with the lower zinc concentration, the variability was too high to obtain a significant difference (P > 0.05). The rate constant, k_E2, of the linear component was reduced by ~30% in cells incubated in 25 µmol zinc/L (P < 0.05). At 160 min, the amount of label released to the basolateral side was two to three times greater than the amount released to the apical side (P < 0.005) (Fig. 3).

View larger version (18K): [in this window] [in a new window]   Fig 4. An expansion of the data in the lower panel of Figure 3 illustrating the two components of copper efflux from Caco-2 cells. Individual points are the mean ± SEM for six replicate membranes. The data for each curve were fitted to the model in Equation 4. The solid line without symbols represents the exponential component of the curves and the dashed line represents the linear component. Constant values for each component are given in Table 1.

Zinc effect on concentration-dependent copper uptake and transport.  Data for the experiment to determine the effect of media-zinc concentrations on concentration-dependent copper uptake in Caco-2 cells were fitted to the model in Equation 5. Increasing the concentrations of zinc in the medium from 5 to 35 and from 5 to 140 µmol/L significantly (P < 0.05) elevated the J_max,U of ⁶⁴Cu uptake by 80 and 180%, respectively (Fig. 5 and Table 2). However, K_U was not significantly affected by 35 µmol zinc/L compared with 5 µmol zinc/L, but was significantly increased (P < 0.05) when cells were incubated with 140 µmol zinc/L. The nonsaturable component, k_d, was 30% less (P < 0.05) for cells receiving 35 µmol zinc/L than for those receiving 5 µmol zinc/L; however, there was no difference (P > 0.05) in the rate between cell incubated with 5 and 140 µmol zinc/L, respectively. The higher amount of zinc in the media did not significantly (P > 0.05) affect copper transport across the cell monolayer (Fig. 6).

View larger version (18K): [in this window] [in a new window]   Fig 5. Media-zinc concentration affects 64Cu uptake into Caco-2 cells incubated with variable amount of media copper. Cells were cultured for 7 d in media containing 5, 35 or 140 µmol zinc/L. Cells were then incubated for 15 min in variable amounts of media copper containing 64Cu. Cells were washed and the amount of 64Cu taken into the cell layer was determined. The data were fitted to a three-parameter hyperbolic model shown in Equation 5 in the text. The solid line without symbols represents the regulated component of the uptake process and the dashed line in each panel represents the unregulated or nonsaturable component of uptake. Individual points are the mean ± SEM for six replicate membranes. As the concentration of media zinc increased, the values for Jmax,U and KU increased (P < 0.05) (Table 2). However, the value for kd was higher when cell were incubated in 35 µmol zinc/L than when incubated in 5 µmol zinc/L. There was no significant difference in kd between cells incubated with 5 and 140 µmol zinc/L.

View larger version (26K): [in this window] [in a new window]   Fig 6. Media-zinc concentration does not affect 64Cu transport across Caco-2 cells incubated with variable amounts of media copper. Individual points are the mean ± SEM for six replicate membranes. Treatment of cell was similar to that described in Figure 5. The inset expands the data for the lower concentrations of media copper. Media-zinc concentration had no significant (P > 0.05) effect on 64Cu transport across the cell monolayer.

The concentration of zinc and copper in Caco-2 cells was affected by the concentration of zinc in the media (Fig. 7). The cells were incubated from d 14 to 21 in media containing 5, 35 and 140 µmol zinc/L. Cells were then harvested and extracted with nitric acid and assayed for zinc and copper content. The results showed that as the media-zinc concentration increased, the amount of cellular zinc increased proportionately and significantly (P < 0.001) from the basal concentration. However, 35 µmol zinc/L did not affect cellular copper concentration, but 140 µmol zinc/L lowered the cellular copper concentration by ~20% (P < 0.01) compared with either of the other concentrations of zinc.

View larger version (19K): [in this window] [in a new window]   Fig 7. Culturing Caco-2 cells in media with high concentrations of zinc increases the amount of zinc, but decreases the amount of copper in the cells. The experimental design was similar to that depicted in Figure 5. After the cells had been counted for their content of 64Cu, copper was extracted from the cells with HNO3 and the amount of zinc and copper was determined by atomic absorption. Bars represent the mean ± SD of 60 replicates for each concentration of media zinc. Bars with different superscript letters are significantly different at P < 0.001 (one-way ANOVA with Tukey's contrasts).

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Effect of media zinc on copper uptake, transport and efflux in Caco-2 cells.  The uptake and transport of copper in most cell types are probably controlled by specific copper transport proteins. Recent evidence suggests that the protein responsible for copper uptake into the cell is a product of the hCTR1 gene (Zhou and Gitschier 1997). The protein responsible for copper efflux from the cell is a P-type ATPase, a product of the Menkes gene (Vulpe et al. 1993). This gene is expressed in most tissues except liver. Previous experiments have shown that high dietary zinc can reduce the copper status of both animals and humans, and the mechanism is thought to involve the interference of zinc in the absorption of copper at the level of the intestinal cell. Studies also have shown that higher than normal zinc concentrations in the growth media of Caco-2 cells enhance copper uptake but reduce copper transport (Reeves et al. 1996). Whether the expression of these two copper transport proteins is affected by high concentrations of zinc is not known.

The main purpose of this study was to examine further the effect of media-zinc concentrations on the uptake, transport and efflux of copper in cells by using the Caco-2 cell as a model. Previous studies showed that these cells required a period of preconditioning with zinc in the medium before an effect on copper movement across the cell could be seen. Adding zinc immediately before transport studies were performed did not affect copper transport. Therefore, all of our uptake and transport studies were done after the cells had come to confluence and then treated for 7 d with zinc. The results showed that media zinc as low as 25 µmol/L significantly enhanced the uptake rate of copper into the cell from the apical side of the cell layer, but had no effect on the rate of uptake from the basolateral side. If the hCTR1 gene product is involved in regulating copper uptake in this cell, then zinc may be inducing an overexpression of the gene product in the apical membrane.

On the other hand, zinc might be stimulating the transport mechanism itself. The kinetics of copper uptake showed that high zinc concentrations in the medium enhanced both the J_max,U and K_U. This suggests that zinc was acting as an uncompetitive enhancer of uptake. It could be argued that high media zinc is stimulating MT production and that this metal-binding protein is sequestering copper and pulling it into the cell. One could speculate that the rate of shuttling of copper across the apical membrane by the transporter is proportional to the rate of copper release inside the cell, i.e., the transporter has to give up bound copper before it is recycled back to the luminal surface to pick up more copper. MT could be acting as an acceptor of intracellular copper. Because high media zinc induces intracellular MT, the intracellular capacity to accept the copper would be increased, thus allowing for a higher rate of copper movement. Although this might be a plausible explanation for a zinc-enhanced rate of movement of copper into the cell from the apical side, we did not see a zinc-enhanced movement of copper into the cell from the basolateral side. In addition, instead of an accumulation of cellular copper with an increase in media zinc and a likely increase in MT (Reeves et al. 1996), we actually observed a significantly lower cellular copper concentration.

The uptake rate of copper from the basolateral side of the Caco-2 cell was much faster than uptake from the apical side. This is consistent with results for uptake of other trace elements, such as with manganese (Finley and Monroe 1997) and zinc (Finley et al. 1995) in this cell type.

Although copper uptake into Caco-2 cells was enhanced by high media-zinc concentrations, copper transport across the cell monolayer was depressed. This is consistent with the proposed mechanism for the effects of high dietary zinc on copper status in animals and humans, i.e., that zinc is inhibiting copper absorption from the gut. The rate of transport in cells cultured with 25 µmol zinc/L of medium was ~50% lower than in those cultured with 5 µmol/L. In addition, the rate of copper efflux from the cells to the basolateral side was reduced by the higher media-zinc concentration. In this case, if the P-type ATPase copper transporter is regulating copper efflux in this cell, then zinc may be causing an underexpression of the protein in the basolateral membrane. However, recent evidence suggests that the transporter is located in the Golgi apparatus and is translocated to the plasma membrane when cellular copper is elevated (Petris et al. 1996). Therefore, high zinc might be affecting this translocation process. Although the higher amount of zinc is five times the normal concentration of zinc in complete culture media, it is not out of range of physiologic concentrations; the normal range of plasma zinc concentration is 15-25 µmol/L, and that in the gut lumen could be much higher.

Some of the observations in this study seem to be inconsistent with the expected results. One such observation was that zinc had no effect on ⁶⁴Cu efflux to the apical side of the cell but reduced ⁶⁴Cu transport across the cell when measured in the basolateral to apical direction. Because each condition is measuring copper movement in the same direction, one might expect that both should be either inhibited or not inhibited. However, because transport measures a more complex, slower process with different rate-limiting steps than uptake or efflux by themselves, it is possible that the two processes would show different results.

The second inconsistent observation was that the initial values (0-5 min) for ⁶⁴Cu efflux to the apical side of the cell were higher than expected; then there was a gradual rise with time (Fig. 3). Part of the problem here might be in the technique used. When cells were labeled with ⁶⁴Cu, the specific activity of the labeled copper in the cell and apical membranes was very high. Before efflux measurements were begun, the cells were washed in warm saline to remove the adhering label. The observation that efflux for the first 5 min was higher than expected, suggests that nonspecifically bound ⁶⁴Cu at the apical surface was immediately released, and the results represent a procedural artifact that overshadows the more slowly regulated efflux of copper by the transporter. This is a reasonable scenario because the apical side of the cell contains numerous microvilli with a very large surface area to which the label could adhere. The short period of washing the cells may have been insufficient to release all of the lightly bound copper.

A third observation was that ⁶⁴Cu uptake from the apical side of the cell and ⁶⁴Cu transport to either side were curvilinear instead of linear as might be expected. Yet, others also have observed this response in Caco-2 cells with copper (Reeves et al. 1996), zinc (Finley and Monroe 1997) and manganese (Finley et al. 1995). Ferruzza et al. (1995) observed a short lag period in the transport of lysine across Caco-2 cells with time and stated that the event was caused by the time required for transport across the membrane on the donor side, through the cell and out of the membrane on the basolateral side. However, because we are measuring the movement of labeled copper under equilibrium conditions, and the samples were stirred for only a few seconds after the label was added, we may be observing the effects of the unstirred water layer. This phenomenon was described generally by Dietschy and Westegaard (1975) and more specifically for Caco-2 cells by Hidalgo et al. (1991). However, we did not see the distortion in transport kinetics as predicted and described mathematically by Winne (1973). Nonetheless, the curvilinear phenomenon was present in our cells treated with both concentrations of zinc and does not weaken the major finding that treating Caco-2 cells with as little as 25 µmol zinc/L of medium reduces the rate of copper movement across the cell layer.

In summary, physiologic concentrations of media zinc enhanced the uptake of copper into Caco-2 cells from the apical side of the membrane and, at the same time, inhibited the movement of copper out of the cell to the basolateral side. The mechanisms to explain these outcomes may lie in the effect of zinc on the expressions or activities of distinctive copper transport proteins in either membrane.

	FOOTNOTES

Manuscript received 18 February 1998. Initial reviews completed 30 March 1998. Revision accepted 12 June 1998.

	ACKNOWLEDGMENTS

The authors thank Ashley Eraas, Jared Spletstoser and Carrie Sagsveen for technical support in the Cell Culture Laboratory, and Brenda Skinner and Lana Stallard for assistance with experimental procedures.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Dietschy, J. M. & Westergaard, H. (1975) The effect of unstirred water layers on various transport processes in the intestine. In: Intestinal Absorption and Malabsorption (Csaky, J. Z., ed.). Raven Press, New York, NY.
• Ferruzza S., Ranaldi G., Di Girolamo M., Sambuy Y. The transport of lysine across monolayers of human cultured intestinal cells (Caco-2) depends on Na⁺-dependent and Na⁺- independent mechanisms on different plasma membrane domains. J. Nutr. 1995; 125:2577-2585
• Finley J. W., Briske-Anderson M., Reeves P. G., Johnson L. K. Zinc uptake and transcellular movement by CACO-2 cells: studies with media containing fetal bovine serum. J. Nutr. Biochem. 1995; 6:137-144
• Finley J. W., Monroe P. Mn absorption: the use of CACO-2 cells as a model of the intestinal epithelia. J. Nutr. Biochem. 1997; 8:92-101
• Fischer P.W.F., Giroux A., L'Abbé M. R. The effect of dietary zinc on intestinal copper absorption. Am. J. Clin. Nutr. 1981; 34:1670-1675[Abstract/Free Full Text]
• Fischer P.W.F., Giroux A., L'Abbé M. R. Effects of zinc on mucosal copper binding and on the kinetics of copper absorption. J. Nutr. 1983; 113:462-469
• Hidalgo I. J., Hillgren K. M., Grass G. M., Bochardt R. T. Characterization of the unstirred water layer in Caco-2 cell monolayers using a novel diffusion apparatus. Pharm. Res. 1991; 8:222-227[Medline]
• Jacquez, J. A. (1987) Application of tracers to the study of membrane transport processes. In: Membrane Physiology (Andreoli, T. E., Hoffman, J. F., Fanestil, D. D., & Schultz, S. G., eds.). Plenum Publishing, New York, NY.
• Kleinbaum, D. G. & Kupper, L. L., (1978) Applied Regression Analysis and Other Multivariable Methods, pp. 95-105. Duxbury Press, North Scituate, MA.
• Masters B. A., Kelly E. J., Quaife C. J., Brinster R. L., Palmiter R. D. Targeted disruption of metallothionein I and II genes increases sensitivity to cadmium. Proc. Natl. Acad. Sci. U.S.A. 1994; 91:584-588[Abstract/Free Full Text]
• Mercer J.F.B., Livingston J., Hall B., Paynter J. A., Begy C., Chandrasekharappa S., Lockhart P., Grimes A., Bhave M., Siemieniak D., Grover T. W. Isolation of a partial candidate gene for Menkes disease by positional cloning. Nat. Genet. 1993; 3:20-25[Medline]
• Oestreicher P., Cousins R. J. Copper and zinc absorption in the rat: mechanism of mutual antagonism. J. Nutr. 1985; 115:159-166
• Ogiso T., Mariyama K., Sasaki S., Ishimura Y., Minato A. Inhibitory effect of high dietary zinc on copper absorption in rats. Chem. Pharm. Bull. 1974; 22:55-60
• Ogiso T., Ogawa N., Miura T. Inhibitory effect of high dietary zinc on copper absorption in rats. II. Binding of copper and zinc to cytosol proteins in the intestinal mucosa. Chem. Pharm. Bull 1979; 27:515-521
• Petris M. J., Mercer J.F.B., Culvenor J. G., Lockhart P., Gleeson P. A., Camakaris J. Ligand-regulated transport of the Menkes copper P-type ATPase efflux pump from the Golgi apparatus to the plasma membrane: a novel mechanism of regulated trafficking. EMBO J. 1996; 15:6084-6095[Medline]
• Reeves P. G. Copper metabolism in metallothionein-null mice fed a high-zinc diet. J. Nutr. Biochem. 1998; 9:598-601
• Reeves P. G., Briske-Anderson M., Newman S. M. Jr. High zinc concentrations in culture media affect copper uptake and transport in differentiated human colon adenocarcinoma cells. J. Nutr. 1996; 126:1701-1712
• Sokal, R. R. & Rohlf, F. J. (1969) Biometry, pp. 441-442. W. H. Freeman, New York, NY.
• Tukey J. W. Comparing individual means in the analysis of variance. Biometrics 1949; 5:99-114
• Van Campen D. R., Scaife P. U. Zinc interference with copper absorption in rats. J. Nutr. 1967; 91:473-476
• Vulpe C., Levinson B., Whitney S., Packman S., Gitschier J. Isolation of a candidate gene for Menkes disease and evidence that it encodes a copper-transporting ATPase. Nat. Genet. 1993; 3:7-13[Medline]
• Winne D. Unstirred layer, source of biased Michaelis constant in membrane transport. Biochim. Biophys. Acta 1973; 298:27-31[Medline]
• Zhou B., Gitschier J. hCTR1: a human gene for copper uptake identified by complementation in yeast. Proc. Natl. Acad. Sci. U.S.A. 1997; 94:7481-7486[Abstract/Free Full Text]

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences