Cytotechnology 13: 175-184, 1993. 01993 Kluwer Academic Publishers. Printed in the Netherlands.

Epithelial properties of human intestinal Caco-2 cells cultured in a serum- free medium

Kei Hashimoto I and Makoto Shimizu 2

SSchool of Food and Nutritional Sciences, University of Shizuoka, 52-1 Yada, Shizuoka 422, Japan; eDepartment of Agricultural Chemistry, The University of Tokyo, Bunkyo-ku, Tokyo 113, Japan

Received 27 July 1993; accepted in revised form 17 November 1993

Key words: Caco-2, epithelial monolayer, human intestine, peptide transport, serum-free culture

Abstrac t

Human intestinal Caco-2 cells were cultured under serum-free conditions on an insoluble collagen and FCS matrix (Caco-2-SF), and a comparison was made between several characteristics of Caco-2 and Caco-2-SF cells. Their morphological appearance was identical. Slight differences were found in cell growth and expression of brush border enzymes between Caco-2 and Caco-2-SF cells. Similar levels of activity of Gly- Gly transport were expressed in both types of cell. Caco-2 cells cultured on permeable filters showed high transepithelial electrical resistance (TEER), indicating the high monolayer integrity. The transepithelial transport activity for glucose, alanine and Gly-Gly was detected by measuring the change in short-circuit current (Msc) after adding each of these nutrients to the apical chamber. In Caco-2-SF cells, such parameters as TEER and AIsc were reduced drastically, suggesting that the monolayer integrity and cell polarity that are important for transepithelial transport were not attained. These parameters, however, could be restored by adding FCS or by milk whey. The result suggested that FCS and milk whey contain factors which regulate the formation of the tight junctions and, consequently, the development of cell polarity. Thus the Caco-2-SF cell-culture system will provide a useful model for studying factors which regulate the intestinal transepithelial transport functions.

Abbreviations: BCECF -- 2',7'-bis(carboxyethyl)-5(6)-carboxyfluorescein; TEER -- transepithetial electrical resistance; LY -- lucifer yellow CH lithium salt

Introduct ion

The small intestine has such functions as digestive ability and nutrient absorption carried out by highly differentiated mature and columnar cells. Turnover of these cells is very rapid, with a mean cell duration time of 2--3 days in most mammals, and

3--5 days in humans (Cairnie et al., 1965a, 1965b). The differentiated columnar cells on the villus develop a luminally-orientated brush-border mem- brane possessing such enzymes as disaccharidases, di- and tri-peptidases, phosphatases (Simon et al., 1979; Semenza, 1986), and several transporters for hexoses (Hediger et al., 1987, 1989), amino acids

176

(Satoh et aL, 1989) and dipeptides (Matthews, 1975). These characteristics make intestinal epithe- lial cells an excellent model for studying the regulation of cell growth and functional differen- tiation.

The human colon adenocarcinoma cell line Caco-2 (Fogh et al., 1977) provides a useful in vitro cell culture model for studying the differen- tiation function of intestinal enterocytes, because Caco-2 cells have been observed to have sponta- neous differentiation ability and to exhibit various enterocytic characteristics (Pinto et al., 1983). The formation of brush-border microvilli (Hughson and Hopkins, 1990) and tight junctions (Hughson and Hopkins, 1990; Rousset, 1986), as well as the expression of brush-border membrane enzymes (Matsumoto et al., 1990), have all been reported. Dome formation on an impermeable support (Pinto et al., 1983; Grasset et al., 1984) is also one of the characteristics of Caco-2, and this phenomenon suggests that Caco-2 expresses a variety of trans- port activities. The transport activities of Caco-2 cells have been recently reported for such sub- stances as hexoses (Riley et al., 1991; Blais et al.; 1987), amino acids (Hidalgo and Borchardt, 1988), dipeptides (Thwaites et al., 1993) and taurocholic acid (Hidalgo and Borchardt, 1990). It is, therefore, considered that Caco-2 would be particularly useful for studying the intestinal epithelial transport characteristics and their regulatory mechanisms (Rousset, 1986).

In order to reveal the regulatory factors that affect cell functions, a serum-free culture of the cells is beneficial, because the effect of serum factors which might affect the development of cell functions can be ruled out in such a system. Re- cently, Jumarie and Malo (1991) have found that Caco-2 cells could be cultured in a serum-free medium (ITS medium). They reported that Caco-2 cells grown in an ITS medium showed normal structural and functional differentiation, expressing many brush-border membrane enzyme markers. The transport characteristics of the cells in the ITS medium, however, were not described in their study. In our present study, we cultured Caco-2 cells on an insoluble collagen and FCS matrix in a serum-free medium, and investigated the properties

of the cells grown in the serum-free system (Caco- 2-SF) with particular emphasis on the transport functions in comparison with those of Caco-2 cells.

Materials and methods

Materials

The Caco-2 cell line was obtained from American Type Culture Collection (Rockville, MD). Dulbec- co's modified Eagle's medium (DMEM) was obtained from Nissui (Tokyo, Japan), and fetal calf serum (FCS) was from Bioserum (Victoria, Austra- lia). Cosmedium (a serum-free medium, which contains insulin (5 mg/1) and transferrin (5 mg/1), but not albumin) was obtained from Cosmobio (Tokyo, Japan), and Millicell-HA with 0.45 ~tm cellulose membranes of 12 mm in diameter was purchased from Millipore (Molsheim, France). Snapwell with 0.4 ~tm polycarbonate membranes of 12 mm in diameter, diffusion chambers, a gas manifold and a block heater were from Costar (Bedford, MA). A type I collagen solution was obtained from Nitta Gelatin (Osaka, Japan), while penicillin and streptomycin were obtained from Wako (Osaka, Japan). Lucifer yellow CH lithium salt (LY) was from Molecular Probes (Eugene, OR), all other chemicals being of reagent grade.

Cell culture

Caco-2 cells were maintained in DMEM sup- plemented with 10% FCS, 4 mM L-glutamine, 50 IU/ml of penicillin and 50 pg/ml of streptomycin. They were incubated at 37~ in a humidified atmosphere of 5% CO2 in air. The monolayers became confluent 3 to 4 days after seeding at a number of 7 x 105 to 1.2 x 10 6 cells per d~100 mm- dish and the cells were passaged at a split ratio of 4 to 8 by trypsinization with 0.25% trypsin and 0.8 mM disodium ethylenediamine tetraacetate in phosphate-buffered saline (PBS).

A confluent monolayer of Caco-2 cells cultured in the presence of FCS was trypsinized, and the obtained cells were washed with DMEM containing 0.1% bovine serum albumin. The cells were resus-

pended in a serum-free medium (Cosmedium) and then seeded in a dish precoated with collagen and FCS. The coating was carried out by incubating culture dishes with type I collagen (100 ~tg/ml) for 30 min, rinsing with DMEM twice, incubating with FCS (10% in DMEM) for 5 min, and then rinsing with DMEM twice. By this procedure, fibronectin, which was contained in FCS, was thought to be specifically attached to collagen molecules (Kohira and Nakamura, 1988). The Caco-2 cells were observed to adhere to the dish surface and start growing within one day. The cells cultured in the serum-free medium (Caco-2-SF) could be passaged at a split ratio of 2 to 5 every 2 to 4 days.

For the transport studies, cells were grown in Millicell-HA or Snapwell coated with type I col- lagen and/or FCS. The cells were seeded at a density of 105 cells in 500 ~tl of the medium, which was changed every 1 or 2 days.

Protein and enzyme assays

Protein content was determined with a Bio-Rad protein assay reagent (Bio-Rad, Richmond) (Brad- ford, 1976). Sucrase activity was measured by determining the glucose content, which was pro- duced by the hydrolysis of sucrose, with a glucose test kit (Wako) (Trinder, 1969). Alkaline phospha- tase activity was measured by using p-nitrophenyl phosphate as a substrate (Bergmeyer et al., 1983). Protein content is expressed as ~tg per well, and enzyme activity as milliunits per mg of protein: one unit is defined as the activity which hydrolyzes 1 jamol of substrate per min.

lntracellular variation of 1t* concentration meas- urement

Variation of intracellular H + concentration was determined with the fluorescent dye, 2',7'-bis(car- boxyethyl)-5(6)-carboxyfluorescein (BCECF). Cells cultured in a 96-well plate were incubated with BCECF/AM (2 t~M) for 20 min at 37~ the non- diffusible BCECF being generated in situ by the action of cytoplasmic esterases (Rink et al., 1982). The excess dye was washed twice with a sodium- free Ringer solution, and 200 jal of the sodium-free

177

Ringer solution was then added. The fluorescence signal at 530 nm (excitation at 490 nm) was re- corded by an MTP-32F microplate reader (Corona). Percent of the decrease in fluorescence intensity by the addition of 50 mM Gly-Gly (at 15 min) to that in the absence of Gly-Gly (at zero time) was calculated (AF), which represents an increase in the intracellular H + concentration.

Transepithelial electrical measurements

Cells were seeded on a collagen-coated Millicell- HA cellulose membrane at a density of 1.7 x 105 cells/era 2 on day 0. The cells were rinsed with a Ringer solution, and then mounted as a flat sheet in Ussing chambers with an exposed area of 0.3 cm 2. To measure the transport activity of glucose and alanine, a Ringer solution at pH 7.4 was added to both sides. For Gly-Gly, the apical side was bathed with 1.9 ml of a sodium-free Ringer solution at pH 5.5, in which sodium was substituted by choline chloride, and the basal side similarly at pH 7.4. Each solution was thermostatically maintained at 37~ and gassed with 5% CO2/O 2. To measure the transepithelial potential difference (pd) and passing current, calomel electrodes in series with 1 M KC1 agar bridges were employed. The transepithelial electrical resistance (TEER) was derived from Ohm's law by measuring the pd change induced by an external current pulse. The change in short- circuit current (AIsc) following the addition of 50 mM glucose, alanine or Gly-Gly was measured by passing an external current to nullify the change in pd. All electrical measurements were made after a 15-min period of stabilization.

Transport studies

Snapwell-grown cell monolayers were rinsed with a Ringer solution, and then placed in a diffusion cell maintained at a constant temperature of 37~ by a block heater. At zero time, 4 ml of an LY solution (1 mg/ml in a Ringer solution at pH 7.4) was added to the apical compartment, and an LY- free Ringer solution likewise to the basal side. Mixing in the diffusion cell was achieved by a 5% CO2/O 2 airlift after 1 hour, after which the medium

Preparation of human or bovine whey

was collected and LY was determined by measur- ing its fluorescence intensity in a F-4000 spectro- photofluorometer (Hitachi) at excitation and emis- sion wavelengths of 430 nm and 540 nm, respec- tively. The percentage of LY transported per 1.2 cm 2 was calculated.

Human or bovine skim milk was centrifuged at 100,000 xg for 2 hr to precipitate caseins, and the neutral whey was obtained. The protein content of human and bovine whey samples was 4.0 mg/ml and 3.7 mg/ml, respectively.

Statistical analyses

500

Student's t-test was used to compare means and ranges.

Results

Growth and enzymatic characterization of the cultured cellS

Changes in the protein content per well during the culture of Caco-2 and Caco-2-SF cells are shown in Fig. 1. As reported elsewhere, protein content closely matches the number of cells per plate (Blais et aL, 1987), thus giving a good estimate of cell growth. No initial lag phase was observed, and the values increased until day 11 for Caco-2-SF and until day 14 for Caco-2, when a steady state was reached. Confluence was reached at day 3 and day 4 in Caco-2 and Caco-2-SF cells, respectively. Both

4 0 0 -

" • 300"

:::L

r -

200" n

1 0 0 "

0 0 25

I I I I

5 10 15 20 Days

178

Fig. 1. Protein content of Caco-2 (Q) and Caco-2-SF (O) cells against time after seeding. Cells were seeded in 24-well plates at a density of 1.5 x 1041cm 2 on day 0. The values presented are the mean (n = 3), and the vertical bars represent S.E, *Significantly different (p < 0.01) from Caco-2 cells. **Significantly different (p < 0.001) from Caco-2 cells,

Fig. 2. Phase contrast micrographs of confluent cultures of (A) Caco-2 cells (day 3) and (B) Caco-2-SF cells (day 4). The cells were seeded in petri dishes at a density of 1.5 x 104/cm 2 on day 0. Magnification is x25.

179

25 400t B A

t J 30

g 20 .

0 5 10 15 20 25 0 5 10 15 20 25 Days Days

Fig. 3. Growth-related differentiation of brush-border membrane enzymes in Caco-2 (0 ) and Caco-2-SF (O) cells. Sucrase (A) and alkaline phosphatase (B) activity was measured against time after seeding. The values presented are the mean (n = 3), and the S.E. bars are shown, unless the error bar is smaller than the symbol. *Significantly different (p < 0.01) from Caco-2 cells. **Significantly different (p < 0.001) from Caco-2 cells.

cells appeared cuboidal in shape, and their mor- phological appearance was similar (Fig. 2). In the stationary phase, the amount of protein per well in the Caco-2 cells was slightly greater than that in the Caco-2-SF cells.

The activity of both sucrase and alkaline phos- phatase was each measured during the growth of Caco-2 and Caco-2-SF cells (Fig. 3). The enzyme activities were low at the beginning and then increased regularly, however, sucrase activity of Caco-2-SF was reached to steady state at day 14. The maximal alkaline phosphatase activity of Caco- 2 was nearly half that observed in Caco-2-SF. On the other hand, little difference was found in sucrase activity between the two cell lines.

H+-coupled Gly-Gly transport of Caco-2 and Caco- 2-SF cells cultured on an impermeable substrate

The activity of H*-eoupled Gly-Gly transport in Caco-2 and Caco-2-SF cells was examined by determining the intracellular variation of H + con-

centration by the trapped fluorescent indicator, 2',7'-bis(carboxyethyl)-5(6)-carboxyfluorescein. In both Caco-2 and Caco-2-SF cells, the fluores- cence signal decreased similarly (Table 1), in- dicating that the intracellular H + concentration was increased by adding Gly-Gly to the culture me- dium. It is, therefore, considered that H+-coupled Gly-Gly transport was exhibited in both cells.

Transepithelial electrical properties of Caco-2 and Caco-2-SF cells

The monolayer integrity was evaluated by measur- ing the transepithelial electrical resistance (TEER). As shown in Fig. 4, TEER of the Caco-2 mono- layer remained high during the culture. But the Caco-2-SF monolayer grown on a permeable membrane had a low TEER value during the first 4 days of culture, and this value rapidly decreased after that. From the 9th day onwards, only a slight- ly measurable TEER value remained, suggesting that the monolayer integrity had been lost.

180

Table 1. Effects of Gly-Gly on the fluorescence of BCECF loaded in Caco-2 and Caeo-2-SF cells

Fluorescence unit AF (%)"

0 min 15 min

Caco-2 1383 -+ 13 1012 _+ 13" 26.8 _+ 0.5 Caco-2-SF 603 _+ 7 443 _+ 27** 26.4 _+ 4.8

Cells cultured in 96-well plates for 8 days were loaded with BCECF, and then preincubated in a sodium-free Ringer solution at pH 5.5. 50 mM Gly-Gly was added at zero time. Autofluorescence, which was determined from a well seeded with cells but not loaded with BCECF, was subtracted from the fluorescence intensity of the BCECF-loaded cells. Time dependent decrease of fluorescence intensity during the experiments was corrected by multiplying the data by the ratio of the fluorescence intensity of Gly-Gly-absent wells at zero time to at 15 min. Values represent the mean _+ S.E. of 3 independent determinations. *Significantly different (p < 0.01) from the value at zero time. **Significantly different (p < 0.001) from the value at zero time. aDecrease of fluorescence intensity (AF) was not significantly different between Caco-2 and Caco-2-SF cells.

The active transport activity was determined by measuring short-circuit current (Isc). As Isc repre- sents the net ion transport across the cellular layer, the change in Isc (AIsc) following the addition of a substrate to the apical chamber can be used as a simple index of cation-coupled nutrient transport. Glucose, alanine and Gly-Gly transport were distinctly observed in Caco-2 cells at day 9, al-

300

250-

E o �9 200-

5

o150- o c t~ 0~

"~100- O rr

0 I v i v I i

0 5 10 15 20 25 Days

Fig. 4. Transepithelial electrical properties of Caco-2 and Caco- 2-SF monolayers. Transepithelial electrical resistance of Caco-2 and Caco-2-SF monolayers against time after seeding. The val- ues presented are the mean of different determinations as indi- cated over the data points, and the vertical bars represent S.E. *Significantly different (p < 0.01) from Caco-2 cells. **Significantly different (p < 0.001) from Caco-2 cells.

though the activity (AIsc) for glucose and alanine was much lower than the Gly-Gly transport activity (Table 2). Unlike Caco-2 cells, the transepithelial transport of Gly-Gly was very low in Caco-2-SF cells when the change of Isc was measured after adding substrate (Table 2), and no glucose or alanine transport was detected (Table 2).

Induction of functional differentiation in Caco-2-SF cells

The loss of monolayer integrity in Caco-2-SF cells may indicate that the enterocytic differentiation potential of the cells was, at least partly, eliminated by culturing in the serum-free medium. In order to evaluate the differentiation potential of Caco-2-SF, additional experiments were undertaken.

FCS was added to the culture medium of Caco- 2-SF, and the change in TEER value of the Caco-2- SF monolayer was studied. By adding 10% FCS to the culture medium 3 days before measurement, the TEER value, which had been almost negligible before the FCS treatment, was increased to half of the TEER value of the Caco-2 monolayer (Table 3, Fig. 4), suggesting that the monolayer integrity had been much improved.

The flux of the leakage marker, Lucifer yellow (LY), was also markedly changed by adding FCS. After a 15-day culture of Caco-2-SF, the flux of LY was 5.35 _ 0.69 %/h (mean +_ S.E. of 3 deter- minations). Three days of treatment by FCS before measurement, however, decreased this value to 0.63

Table 2. Transepithelial transport of glucose, alanine and Gly-Gly in Caco-2 and Caco-2-SF cells

181

Substrate Alsc (gA)

Caco-2-SF Caco-2

Control (+) FCS

Glucose N.D.a 0.49 +- 0.10 0.37 -+ 0.05 Alanine N.D. a 0.57 _+ 0.03 0.41 _+ 0.07 Gly-Gly 0.63 -+ 0.27 2.18 -+ 0.26* 2.22 _+ 0.08

Caco-2 cells were cultured for 9 days. Caco-2-SF cells were cultured for 6 days, and then cultured with FCS for 3 more days. At day 9, cell monolayers were rinsed with a Ringer solution and then mounted as a flat sheet in an Ussing chamber. Cells for the controls were cultured in a serum-free medium for 9 days. Values represent the mean _+ S.E. of 4 independent determinations. *Significantly different (p < 0.01) from the control. anot detected.

_+ 0 . 1 1 % / h (mean _+ S.E. of 3 determinations),

which is in good agreement with the result obtained

by the T E E R measurement (Table 3). These results

suggest that the formation of tight junct ions be-

tween the cells and organizat ion of the cell mono-

layer was induced by some soluble factors in FCS.

The active transport activity of Caco-2-SF cells

was also recovered by adding 10% FCS to the

culture med ium accompanying the TEER increase.

The transepithelial glucose, alanine and Gly-Gly

transport activity was increased to almost the same

level of activity as that of Caco-2 cells (Table 2).

These data indicate that Caco-2-SF cells did not

lose their functional differentiation abili ty in the

serum-free medium even after several passages.

Induct ion of the enterocytic differentiat ion of

Caco-2-SF cells by food-derived substances was

attempted. The addition of 10% bovine or human

Table 3. Effect of FCS on TEER of Caco-2-SF monolayers

Resistance (f2.cm 2)

Control (+) FCS

9 days 1.9 _-. 2.3 88.4 +_ 5.4 14 days 0.6 + 0 56.3 _+ 10.2

Transepithelial electrical resistance of Caco-2-SF monolayers against time after seeding. FCS was added to the medium at day 6 or day 11 and cultured for three more days, then TEER was measured at day 9 and day 14 as described in Materials and Methods. Values represent the mean __. S.E. of 4 independent determinations. *Significantly different (p < 0.05) from the control.

Table 4. Effect of whey on TEER and transepithelial Gly-Gly transport activity (Alsc) of Caco-2-SF monolayers

Resistance (f2.cm z) AIsc (~tA)

Control 28.2 -2_ 2.1 0.47 _+ 0.10 Human whey 99.0 _+ 12.9"** 0.95 _ 0.10" Bovine whey 67.8 _+ 6.0*** 0.83 _+ 0.05**

Ceils were seeded on a collagen-coated Millicell-HA cellulose membrane, and cultured for 6 days. Bovine or human whey was then added to the medium for 3 days, after which TEER and AIsc were measured as described in Materials and Methods. Values represent the mean _+ S.E. of 4 independent determinations. *Significantly different (p < 0.05) from the control. **Significantly different (p < 0.02) from the control. ***Significantly different (p < 0.01) from the control.

182

whey to the culture medium increased the TEER value and transepithelial Gly-Gly transport activity (Table 4).

Discussion

Intestinal epithelial cells are known to be able to degrade and absorb nutrients and solutes from the lumen of the intestine. The existence of many carriers involved in the active transport of such nutrients as amino acids (Satoh et al., 1989) and hexoses (Hediger et al., 1987,1989) in these cells has been reported.

Caco-2 cells have also been reported to express Na+-coupled glucose and alanine transport (Grasset et al., 1984; Riley et al., 1991). Recently, the presence of an H§ peptide transport system in Caco-2 cells has been proved by Thwaites et al.

(1993). The peptide transport system has a broad specificity and narrow affinity range (Abe et al.,

1987), being different from the amino acid trans- port system. It compensates the Na§ amino acid-transport system by which hydrophobic amino acids are predominantly transported (Preston et al.,

1974). The existence of the peptide transport systems is, therefore, thought to be of nutritional importance. However, the properties of the peptide transporter and the regulatory mechanism for this transport system are still obscure. Caco-2 cells, particularly those cultured in the absence of FCS, would provide a good model for studying regula- tory factors for epithelial peptide transport.

In Caco-2-SF cells the brush-border enzyme activity was not reduced, as has been reported by Jumarie and Malo (1991). Moreover, the Caco-2-SF cells were observed to express Gly-Gly transport, when the activity was determined by measuring with a fluorescent dye (Table 1). Caco-2-SF cells cultured on a permeable filter, however, showed little transepithelial Gly-Gly transport activity (Table 2). This difference may partly be due to the characteristics of the electrophysiological method, which requires an adequate TEER value for meas- uring the Isc. However, a more likely explanation would be the lack of epithelial cell membrane polarity in Caco-2-SF cells. Maintenance of the

epithelial cell membrane polarity strongly depends on a tight junctional barrier (Gumbiner and Simons, 1986). In contrast to Caco-2 cells cultured in an FCS-containing medium which formed tight junc- tions, Caco-2-SF cells in a serum-free medium were unable to adequately form tight junctions as observed by TEER measurement (Fig. 4). Gly-Gly transport is considered not to have been properly organized, although Caco-2-SF cells themselves expressed it (Table 1). However, it would be possible that the decrease in fluorescence intensity of BCECF observed in Table 1 was not only due to variation of the intracellular H + concentration but due to either variations in BCECF/AM cell uptake or variations in esterase activity. Further inves- tigations are therefore necessary to confirm the Gly-Gly transport activity. When FCS was added to the culture medium, the TEER value for Caco-2-SF was increased (Table 3), suggesting that a tight enough junction was formed to act as a barrier to paracellular leakage. As a result, nutrient transport was organized with proper polarity, and alanine, glucose and Gly-Gly transport was accordingly observed (Table 2). These results suggest that soluble FCS in the medium played an important role in maintaining the epithelial polarity, which was accompanied by the expression of transepithe- lial transport activity. To investigate the factors contained in FCS, several factors, like transforming growth factor-[31 (TGF-~1), epidermal growth factor (EGF), hydrocortisone, glucagon and triiodothyro- nine were tested, but they did not have the TEER- increasing activity (data not shown). The factors which affected the tight junction formation and epithelial polarity remain to be investigated.

In the present study, some samples derived from food materials were assayed by using our Caco-2- SF system. Among the samples, both human and bovine whey induced enterocytic differentiation of Caco-2-SF cells; that is, their TEER value and the transepithelial Gly-Gly transport activity was increased (Table 4). Damerdji et al. (1988) have reported that bovine whey contains growth ac- tivators and inhibitors. It will be interesting to discover the factors which induce the epithelial functions of Caco-2-SF cell monolayers from whey, because they may play an important role in the

development of intestinal functions in neonates by being administered orally. The isolation and iden- tification of these factors are now in progress.

In conclusion, Caco-2-SF resembles Caco-2 in cell growth, morphological appearance, expression of brush-border enzymes and dipeptide transport. The Caco-2-SF cell monolayer, however, lost TEER and transepithelial transport activity, which could be recovered by culturing the cells in the presence of appropriate factors contained in FCS or in such food as milk whey. Caco-2-SF cells, therefore, represent a very useful model with which to assay factors regulating the formation of a tight junction and the maintenance of membrane polar- ity, which are very important for the development of epithelial functions.

Acknowledgement

We are grateful to Dr. T. Hoshi for helpful discus- sions.

References

Abe M, Hoshi T and Tajima A (1987) Characteristics of transmural potential changes associated with the proton- peptide co-transport in toad small intestine. J. Physiol. 394: 481--499.

Bergmeyer HU, Gral31 M and Walter H-E (1983) Biochemical reagents for general use. In: Bergmeyer HU (ed.) Methods of enzymatic analysis. 3rd ed. (pp. 126--328) Verlag Chemic GmbH, Weinheim.

Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248-- 254.

Blais A, Bissonnette P and Berteloot A (1987) Common characteristics for Na§ sugar transport in Caco-2 cells and human fetal colon. J. Membr. Biol. 99: 113-125.

Cairnie AB, Lamerton LF and Steel GG (1965a) Cell prolifera- tion studies in the intestinal epithelium of the rat. I. Deter- mination of the kinetic parameters. Exp. Cell Res. 39: 528--538.

Cairnie AB, Lamerton LF and Steel GG (1965b) Cell prolifera- tion studies in the intestinal epithelium of the rat. II. Theoret- ical aspects. Exp. Cell Res. 39: 539-553.

Damerdji O, Derouiche F, Legand C, Capiaumont J, Bour JM, Maugras M, Belleville F, Nabet P, Paquet D and Linden G (1988) Utilization of whey fractions as a substrate for fetal

183

calf serum in culture media. Biotech. Techn. 2: 253--258. Fogh J, Fogh JM and Offeo T (1977) One hundred and twenty-

seven cultured human tumor cell lines producing tumors in nude mice. J. Natl. Cancer Inst. 59: 221--226.

Grasset E, Pinto M, Dussaulx E, Zweibaum A and Desjeux JF (1984) Epithelial properties of human colonic carcinoma cell line Caco-2: electrical parameters. Am. J. Physiol. 247: C260--C267.

Gumbiner B and Simons K (1986) A functional assay for proteins involved in establishing and epithelial occluding barrier: identification of an uvomorulin-like polypeptide. J. Cell Biol, 102: 457--468.

Hediger MA, Coady MJ, Ikeda TS and Wright EM (1987) Expression cloning and eDNA sequencing of the Na§ co-transporter. Nature 330: 379--381.

Hediger MA, Turk E and Wright EM (1989) Homology of the human intestinal Na+/glucose and Escherichia coli Na+/pro - line cotransporters. Proc. Natl. Acad. Sci. 86: 5748--5752.

Hidalgo IJ and Borchardt RT (1988) Amino-acid transport in a novel model system of the intestinal epithelium (Caco-2 cell). Pharm. Res. 5: PD950.

Hidalgo IJ and Borchardt RT (1990) Transport of bile acids in a human intestinal epithelial cell line, Caco-2. Biochim. Biophys. Acta 1035: 97--103.

Hughson EJ and Hopkins CR (1990) Endoeytic pathways in polarized Caco-2 cells: identification of an endosomal compartment accessible from both apical and basolateral surfaces. J. Cell Biol. 1 I0: 337--348.

Jumarie C and Malo C (1991) Caco-2 cells cultured in serum- free medium as a model for the study of enteroeytic differen- tiation in vitro. J. Cell. Physiol. 149: 24--33.

Kohira T and Nakamura T (1988) Primary culture of hepato- cytes. Exp. Medicine 6:89--97 (in Japanese).

Matsumoto H, Erickson RH, Gum JR, Yoshioka M, Gum E and Kim YS (1990) Biosynthesis of alkaline phosphatase during differentiation of the human colon cancer cell line Caco-2. Gastroenterology 98:1199-1207.

Matthews DM (1975) Intestinal absorption of peptides. Physiol. Rev. 55: 537--608.

Pinto M, Robine-Leon S, Appay M-D, Kedinger M, Triadou N, Dussaulx E, Lacroix B, Simon-Assman P, Haffen K, Fogh J and Zweibaum A (1983) Enterocyte-like differentiation and polarization of the human colon carcinoma cell line Caco-2 in culture. Biol. Cell 47: 323--330.

Preston RL, Schaeffer JF and Curran PF (1974) Structure- affinity relationships of subsu'ates for the neutral amino acid transport system in rabbit ileum. J. Gen. Physiol. 64: 443-- 467.

Riley SA, Warhurst G, Crowe PT and Turnberg LA (1991) Active hexose transport across cultured human Caco-2 cells: characterization and influence of culture conditions. Biochim. Biophys. Acta 1066: 175--182.

Rink TJ, Tsien RY and Pozzan T (1982) Cytoplasmic pH and free Mg 2+ in lymphocytes. J. Cell Biol, 95: 189-196.

Rousset M (1986) Human colon carcinoma lines HT-29 and Caco-2: two in vitro models for the study of intestinal differentiation. Biochemie 68: 1035--1040.

184

Satoh O, Kudo Y, Shikata H, Yamada K and Kawasaki T (1989) Characterization of amino acid transport systems in guinea-pig intestinal brush-border membrane. Biochim. Biophys. Acta 985: 120-126.

Semenza G (1986) Anchoring and biosynthesis of stalked brush border membrane proteins: glycosidases and peptidases of enterocytes and of renal tubuli. Annu. Rev. Cell Biol. 2: 255--313.

Simon PM, Kedinger M, Raul F, Grenier JF and Haffen K (1979) Developmental pattern of rat intestinal brush-border enzymic proteins along the villus-crypt axis. Biochem. J. 178: 407--413.

Thwaites DT, Brown CDA, Hirst BH and Simmons NL (1993) Transepithelial glycylsarcosine transport in intestinal Caco-2 cells mediated by expression of H+-coupled carriers at both apical and basal membranes. J. Biol. Chem. 268: 7640-7642.

Trinder P (1969) Determination of blood glucose using an oxidase-peroxidase system with a non-carcinogenic chromo- gen. J. Clin. Path. 22: 158--161.

Address for correspondence: K. Hashimoto, School of Food and Nutritional Sciences, University of Shizuoka, 52-1 Yada, Shizuoka 422, Japan.