In vitro culture of rabbit esophageal epithelial cells

GASTROENTEROLOGY

1981;81:30-6

In Vitro Culture of Rabbit Esophageal Epithelial Cells J. J. STEF, F. F. ZBORALSKE, Department of Radiology Stanford, California

and M. A. KARASEK

and Dermatology,

An in vitro model using esophageal epithelial explant cultures has been developed. Esophageal mucosa from New Zealand white rabbits was cut into 1.5-mm* sections and fixed to a plastic petri dish using chick plasma. Appropriate outgrowth medium was added, and the explants were incubated in a humidified CO, incubator at 37°C. Cell outgrowth was measured directly, and lipid content was determined qualitatively. Esophageal epithelial cells grow as confluent sheets of multilayered cuboidal cells. Three growth phases were identified. Increased oxygen tensions inhibited growth. Optimum growth occurred at pH 7.4. Growth was significantly inhibited at pH 6.4, and no growth occurred at pH 5.8. Cells may be released by trypsin and five subcultures obtained. These results demonstrate that monolayer cultures of normal esophageal epithelium may be routinely established and maintained in cell culture for extended periods from small samples of tissue and that factors affecting growth may be quantitatively determined. Investigations of the growth and cellular dynamics of skin epithelial cells in organ, explant, and cell culture have led to a better undr rstanding of the basic biochemistry and physiology of cell differentiation and to models of human skin disease (1).Although several esophageal epithelial diseases are recognized clinically (e.g., reflux esophagitis, Barrett’s esophagus, and carcinoma), few in vitro studies of the basic biology of esophageal epithelial cells have been reported, and all have utilized organ culture techniques (2-4). Rosztdczy (2) studied the effect of influenza

virus

on

human

fetal

esophagus

and

Received September 24,198O. Accepted January 19,1981. Address requests for reprints to: F. Frank Zboralske, M.D., Department of Radiology, Stanford University Medical Center, Stanford, California 94035. This work was supported in part by Grant AM-14121 of the U.S.P.H.S. An abstract of this work appeared in Clin Res (1960; 28:73A). 0 1981 by the American Gastroenterological Association 0016-5085/81/070030-a7502.50

Stanford

University

School of Medicine,

demonstrated that human esophageal epithelium could be maintained and support influenza virus replication for at least 4 days. Stenn and Stenn (3), utilizing chemically defined medium, demonstrated that organ cultures of adult mouse esophageal mucosa incorporated tritiated thymidine for 6 days. In their study, autoradiography revealed that 93% of the basal cells took up the label, whereas no appreciable labeling was found in other cells of the preparation. Harris et al. (4) maintained large (1 cm’) fullthickness human esophageal tissue, obtained at the time of surgery or immediate autopsy, for 7 days before studying the metabolism of carcinogens. Although both murine and human esophageal epitheha1 cells survive in organ culture for several days, multiplication of cells is limited, quantitation is difficult, and little new growth of cells is observed. A need therefore exists for more quantitative longterm models of the normal esophageal epithelial cell. This report describes the growth of normal rabbit esophageal epithelial cells in monolayer culture using explant culture methodology and quantitates various parameters affecting their growth in longterm culture.

Materials and Methods Materials Materials used included: Eagle’s minimum essential medium (MEM) with Earle’s salts and without sodium bicarbonate in powder form (Grand Island Biological Co., GIBCO, Grand Island, N.Y.); McCoy’s 5A medium (modified) without sodium bicarbonate in powder form (GIBCO); CMRL medium 1066 with L-glutamine in liquid form (GIBCO); fetal bovine serum (Irvine Scientific Sales Co., Irvine, Calif.); streptomycin (Pfizer Pharmaceuticals, Chicago, Ill.); Penicillin G (Parke & Davis Co., Detroit, Mich.); plastic petri dishes, 35-mm diameter (Lux Scientific Corp., Thousand Oaks, Calif.); trypsin l-250 (hog pancreas) (ICN Pharmaceuticals, Cleveland, Ohio); heparinized chick plasma (GIBCO); bovine fibrinogen, fraction I

CULTURE OF RABBIT ESOPHAGEAL EPITHELILJM

July 1981

31

MUCOSAL LAYER EXTRACTED

MB&S&RlFICED EXTRACTED

c

CELL GROWTH QUANTlTlZED

EXPLANTS FIXED TO PETRI DISH

ANTIBlOTlC SOAK MUCOSA CUTS

Figure 1. Diagram of the various stages of the primary epithelial explant preparation.

(Calbiochem-Behring Corp., La Jolla, Calif.); and bovine thrombin (Parke & Davis Co.). Tissue Fresh rabbit esophageal epithelium was obtained from 35 adult New Zealand white rabbits killed with intravenous sodium pentothal. After the esophagus was removed from the thorax, the tissue was placed in Hanks’ balanced salt solution (HBSS) supplemented with 200 pg/ ml streptomycin and 400 U/ml penicillin. The esophagus was then cut open longitudinally, and the mucosa was separated from the muscularis externa by clamping forceps on the muscular edge and peeling the mucosal layer off with another forceps. This separated the esophagus through its submucosa. The resultant mucosal tissue was washed three times in supplemented HBSS and soaked for an additional 30-60 min in the final wash. Esophagus

Cultures

Primary explant cultures. The mucosa was cut into sections approximately 1.5 x 1.5 mm*. Three tissue sections (constituting one primary culture) were then explanted on a disposable 35-mm plastic petri dish which had 1 drop of chick plasma containing 18 mg/ml fibrinogen. To fix the explant to the petri dish, one drop of bovine thrombin (100 U/ml) was then added, and the dish was incubated at 37°C for 20 min or longer until a clot formed. After clot fixation of the explants, 2 ml of culture media were added, and the explants were incubated at 37°C. Three culture dishes were utilized for each experiment. A schematic of the various stages of preparation is shown in Figure 1. Serial planting,

Cultivation

Confluent cultures, reached lo-16 days after exwere utilized for subsequent subculture. To de-

tach these cells from the petri dish, they were incubated in Ca’+, Mgz+-free phosphate-buffered saline (PBS) containing 0.1% trypsin and 0.33% EDTA, pH 7.3, for 6 min at 37% The detached cells were dispersed into single cells by gentle pipetting with a Pasteur pipette, and the cells were collected by centrifugation at 800 g for 1 min. The packed cells were then resuspended in MEM containing 10% fetal bovine serum, The isolated cells for serial cultivation were plated at not less than 8-12 X lo4 cells per dish on either a gel of acid-soluble collagen prepared in a manner as described by Liu and Karasek (5) or on a plastic dish. A 1: 2 split was made during the first three subcultures of this cell line whereas subsequent subcultures were passaged at a 1: 1 split.

Determination

of Cell Growth

Cell outgrowth in primary explant cultures was determined by direct measurement using a microscope ocular fitted with a calibration scale. An average outgrowth value was obtained by measuring linear growth in four quadrants as shown in Figure 1 and dividing by 4. The measurements were taken at approximately Z-day intervals until the plateau stage of growth was reached. In subcultures, cell numbers were determined by direct cell count with a hemocytometer. Parametric

Studies

Media. Three outgrowth media were studied: Eagle’s minimum essential medium (MEM), McCoy’s, and CMRL. Each media was supplemented with 10% fetal bo(100kg/ml) and penicillin G vine serum and streptomycin (ZOOU/ml). All media were studied in an atmosphere of air and CO, sufficient to maintain the pH at 7.4 (approximately 6% CO,).

32

GASTROENTEROLOGY

ETAL.

Vol. 81, No. 1

Nistology Before staining, explant samples phosphate-buffered formaldehyde for subsequently stained for lipid using oil mucopolysaccharides with Alcian blue, the method of Preece (6).

were fixed in 10% 24 h. Cells were red 0 and for acid both according to

EM Prep Rabbit esophageal cells in the second and third passages were fixed in situ by the addition of 3% phosphate-buffered glutaraldehyde solution (pH 7.4) at 4°C. The sheet was rinsed in phosphate buffer, gently peeled from the plastic, and postfixed in 2% osmium tetroxide. The cells were dehydrated in a graded series of ethanol and embedded in Epon 812. Ultrathin sections were cut and stained with aqueous uranyl acetate and lead (5%) for 1 h at pH 4.5. The sections were examined in a Siemens Elmishop 102 electron microscope, and representative areas were recorded at a magnification of 22,600.

I

1 d 5

lo

15

20

DAYS POST EXPLANT Figure 2. Outgrowth curves of normal rabbit esophageal epithelium. This figure depicts the normal range of outgrowth under optimal pH (7.4) and oxygen (20%) conditions. Five percent of explants continued linear growth until confluency was reached, as represented by the dotted line (--).

pH. Eagle’s MEM was prepared at pHs of 4.2, 5.8, 8.4, and 7.4 by addition of either 0.1 N HCl or 0.1 N NaOH in proportions necessary to achieve the desired pH when equilibrated at 6% CO, at 37’C. Oxygen. Small desiccators were gassed using a manifold technique which mixed appropriate volumes of oxygen, nitrogen, and CO,. Oxygen tensions of 20% (atmospheric), 40%, 60%, 80%, and 95% were prepared.

Statistical

Analysis

The statistical significance of differences in all studied parameters was determined by Student’s grouped t-testing (7). The 5% level was used in determining the significance of the statistical tests.

Results Outgrowth

Kinetics

Effect of media. Growth curves using MEM with 10% fetal bovine serum under standard conditions of pH 7.4 and 20% oxygen concentration are shown in Figure 2. Three growth phases were usually identified: a lag phase of 24-48 h, a linear growth phase of 7-10 days, and, finally, a plateau phase for

/

Figure 3. Morphology of epithelial cells in primary culture. The dark area shows the explant tissue. Epithelial cells are shown growing in platelike fashion.

July 1981

CULTURE OF RABBIT ESOPHAGEAL

EPITHELIUM

33

Figure 4. H & E stain of esophageal epithelial

cells during the linear growth phase. Numerous mitotic figures (arrows) can be seen indicating tive cell multiplication.

additional lo-14 days at which time cell disintegration occurred. Figure 3 shows an explant in the linear growth phase. The cells have a typical epithelial morphology characterized by a polyhedral configuration and grow in close apposition. An H & E stain of epithelial cells during linear growth is shown in Figure 4. Numerous mitotic figures can be seen indicating active cell division rather than mere migration of cells from the primary explant tissue. In approximately 5% of experiments, cells did not reach a plateau phase but continued to grow in linear fashion until confluency was reached as indicated by the dotted line in Figure 2. A comparison of other outgrowth media with MEM demonstrated that McCoy’s had no statistically significant difference in outgrowth, while tissue grown in CMRL ceased growth at day 8 with subsequent cell disintegration (Figure 5). Eagle’s MEM was subsequently utilized for all additional studies. Effect of pH. Figure 6 shows the effect of four different pHs on rabbit esophageal epithelial cell growth. No growth was observed at pHs 4.2 and 5.8. Although some growth occurred at pH 6.4, the growth rate was significantly decreased when compared with the control pH of 7.4 (p < 0.05). Effect of oxygen concentration. Figure 7 illustrates the effect of various degrees of oxygen concentration on cell outgrowth. All oxygen tensions in excess of 20% showed statistically significant differences in outgrowth kinetics with decreasing outgrowth occurring with increasing oxygen tensions. Oxygen levels of 80% and 95% departed from the normal range at day 5, 60% at day 7, and 40% between days 8 and 9. Serial cultures. Cells plated on a plastic sur-

an

face

showed

poor

attachment,

deteriorated

at

a

ac-

rapid rate, and did not reach confluence, On a collagen surface the cells, in all passages, demonstrated the same morphologic characteristics (stratification, increase in cell diameter, and lipid accumulation) as the cells in primary culture. The growth curve of rabbit esophageal cells on a collagen gel in first passage is shown in Figure 8. During the growth phase of S, subculture a doubling time of 48 h was observed. Cells could be serially passaged five times and maintained for more than 100 days. Histology Numerous refractile droplets were seen in many of the outgrowth cells (Figure 9). Staining with oil red 0 demonstrated a high concentration of lipids within these droplets. When freshly excised rabbit esophagus was fixed and stained with oil red 0,

01 0

1

I

I

1

I

1

1

I

2

4

6

8

10

12

14

16

DAYS

POST-EXPLANT

I

18

Figure 5. Effect of three different culture media on esophageal epithelial outgrowth. Eagle’s MEM and McCoy’s showed similar growth patterns with MEM supporting slightly greater cell mass. CMRL failed to support prolonged epithelial growth.

34

GASTROENTEROLOGY

STEF ET AL.

similarly stained droplets were observed in the stratified layer of the epithelium (histology not shown). Alcian blue staining for acid mucopolysaccharides produced a weak blue stain adjacent to the plasma membrane of the outgrowth cells. A similar pattern of staining was seen in freshly excised rabbit esophagus.

1OOOJ

Qoo800700.

*p
600

Electron Microscopy

Vol. 81, No. 1

1

Electron microscopy of the cells in second and third subpassages demonstrated cell stratification, microvilli, and numerous desmosomal attachments between the cells (Figure 10).

Discussion In this study, we have demonstrated that rabbit esophageal cells can be (a) reliably and routinely grown in cell culture when obtained from small (1.5 mm’) esophageal mucosal tissue specimens and (b) studied quantitatively for more than 100 days. This is in contrast to organ culture methodology wherein esophageal epithelial survival time is short and the kinetics of cell growth difficult to quantitate. The growth stages of rabbit esophageal epithelial

lo

5

,

,

15

20

DAYSPOST EXPLANT Figure 7. The effect of oxygen concentration on epithelial growth. Increased cell toxicity is noted with increasing oxygen concentrations. RA = room air.

cells in explant culture are similar to those described for the skin epithelial cell (8). The cells in outgrowth show a lag phase of 24-48 h before active growth begins followed by a linear and a plateau growth phase. That the cells grow in linear rather than logarithmic fashion is presumptive evidence for both a replicating and a differentiated culture. The kinetic results are confirmed at the light and electron microscope levels where multilayers of the cells in varying stages of differentiation are observed (Figures 4 and

‘P
MI’S POST EXPLANT Figure 6. The effect of pH on rabbit esophageal epithelial cell growth in primary culture. No growth is observed at or below pH 5.8. The growth rate at pH 6.4 was significantly decreased when compared with the standard pH of 7.4.

2

4

6

6

lo

12

DAYS POST SUSCULTURE Figure 8. Sl growth curve for esophageal epithelial cells on collagen gels.

July 1981

CULTUREOFRABBITESOPHAGEALEPlTHELIUM

35

Figure 9. Phase-contrast micrograph of rabbit esophageal epithelial cells in primary culture. These cells are stratified and show numerous refractile lipid droplets (X 640).

10). After the plateau phase is reached, the cultures deteriorate. Approximately 95% of the 35 rabbit esophagi studied demonstrated this growth pattern. Attempts to subculture cells in the plateau phase of growth were not successful. Cells from approximately 5% of the animals studied did not reach a plateau phase but continued to multiply and produced confluent dishes of typical esophageal epithelial cells. These cells could be successfully subcultured on a collagen gel. In early passages (up to the 3rd) a split of 1:2 could be obtained; in later cultures, splits of only 1:1 were successful. The growth potential of the cells gradually decreased after five passages. The epithelial nature of these passaged cells was confirmed by the electron microscopic demonstration of cell stratification, microvilli and numerous desmosomal attachments between the cells (Figure 10). Both rabbit skin (9) and esophageal epithelial cells show improved attachment and growth on collagen. Although the function of collagen in enhancing the growth of epithelial cells in cultures is not known, it is likely that the collagen surface duplicates some regions of the basement membrane to which all epithelial cells are attached. Outgrowth kinetics are strongly affected by oxygen tension and pH. The effect of oxygen concentration on cell outgrowth followed a trend of increasing toxicity with increasing concentrations, a finding also reported for other cell types (10). At pH 6.4,cell growth was significantly inhibited; at pH 5.8 and lower, no growth was observed. That cell growth was significantly decreased at pH 6.4and inhibited

at pH 5.8may have significance in the clinical setting of human reflux esophagitis where, based on our experimental work in the rabbit, a chronic, albeit intermittent, exposure to a decreased pH might be expected to alter the replacement kinetics of the esophageal epithelial cell. Outgrowth studies of rabbit esophageal cells exposed to intermittent fluctuations of acid pH may serve as an in vitro model of this disease. The detection of large amounts of lipids in the cytoplasm of the esophageal epithelial cell in culture was an unexpected observation. When freshly excised rabbit esophagus was fixed and stained with oil red 0, lipid inclusions were also observed in the stratified layers of the esophagus in vivo. Similar fat deposits have been previously reported in human esophageal epithelial cells in vivo by Hopwood et al. (11). These investigators were uncertain whether these droplets represented normal metabolic products or were an artifact of abnormal metabolism due to the relative anoxic state of the superficial cells they studied. Although the chemical nature or function of these lipids is presently unknown, the fact that they were synthesized both in our in vivo and in vitro experiments suggests that they have an unexplained physiologic significance. In these studies we have directed our efforts at the growth parameters of the normal rabbit esophageal epithelial cell. Because this approach permits study of cell kinetics of mucosal tissue only 1.5 mm’, it indicates that this approach may also be applicable for studies of mucosal biopsies from human esophageal epithelial disease.

36

STEF ET AL.

GASTROENTEROLOGY

Vol. 81, No. 1

Figure 10. Electron micrograph of in vitro rabbit esophageal cells in cross section. At the lower left of the photograph is the surface of the petri dish and at the upper right is the free surface of the explant. Note the stratification as manifested by the four separate cells seen in the micrograph. Numerous microvilli (black arrows) and desmosomes (white arrows) are also seen (X 9984).

References

1. Karasek M. In vitro growth and maturation of epithelial cells from postembryonic skin. J Invest Dermatol 1975;65:60-6. 2. Roszt6czy I, Sweet C, Toms GL, et al. Replication of influenza virus in organ cultures of human and simian urogenital tissues and human foetal tissues. Br J Exp Path01 1975;56:322-8. 3. Stenn KS, Stenn JO. Organ culture of adult mouse esophageal mucosa in a defined medium. J Invest Dermatol 1976;66:302-5. 4. Harris CC, Autrup H, Stoner GD, et al. Metabolism of benzo(a)pyrene, n-nitrodimethylamine and n-nitrosopyrrolidine and identification of the major carcinogen-DNA adducts formed in cultured human esophagus. Cancer Res 1979;39:4491-6.

5. Liu SC, Karasek MA. Isolation and growth of adult human epidermal keratinocytes in cell culture. J Invest Dermatol 1978;71:157-62. 6. Preece A. A manual for histological technicians. 2nd ed. Great Britain: J.A. Churchill Pub. Co., 1972. 7. Snedecor GW, Cochran WG, Statistical methods. 6th ed. Ames, Iowa: Iowa State University Press, 1967. 8. Karasek MA. In vitro culture of human skin epithelial cells. J Invest Dermatol 1966;27:533. 9. Liu SC, Karasek MA. Isolation and serial cultivation of rabbit skin epithelial cells. J Invest Dermatol 1978;70:288-93. 10. Cells and tissues in culture. In: Willmer EN, ed. New York: Academic Press, Inc., 1965. 11. Hopwood D, Logan KR, Coghill G, et al. Histochemical studies of mucosubstances and lipids in normal human oesophageal epithelium. Histochem J 1977;9:153-61.