Brief Communication: Epithelial Cell Cultures From Normal and Cancerous Human Tissues 1, Z R. B. Owens, H. S. Smith, W. A. Nelson·Rees, and E. L. Springer SUMMARY-Thirty epithelial cell strains were isolated from human carcinomas and normal epithelial tissues by collagenase digestion and selective removal of fibroblasts with trYpsinVersene. Most strains were obtained from metastatic car· cinomas or epithelia of the urinary and intestinal tracts. The success rate for growth of both neoplastic and normal tissues (excluding skin) was 38%. Six of these strains showed gross morphologic and chromosome changes typical of malignant cells. Nine resembled normal epithelium. The other 15 exhibited some degree of morphologic change from normal.-J Nat' Cancer Inst 56: 843-849, 1976.
MATERIALS AND METHODS
The growth medium was Dulbecco's modified Eagle's medium (#H-21HG, Gibco, Grand Island, N.Y.), which contains 4.5 g/liter glucose. The medium was supplemented with 10% fetal calf serum (Gibco), IOp.g insulin/ml (Calbiochem, San Diego, Calif.), and 0.1 mM nonessential amino acids (Gibco, # 114). Tissue samples were collected in cold growth medium, minced immediately to about l-cm3 fragments, and kept at 4 0 C until processed for tissue culture, usually within 2 hours after surgery. Antibiotics (penicillin, streptomycin, and Fungizone) were included in the collection fluid and growth medium until the second subculture. Cell cultures were initiated by mincing the tissue fragments with apposed scalpels to about 2·mm cubes. Portions of this mince (50-100 fragments) were transferred to several 75-cm z culture flasks (Falcon Plastics, Los Angeles, Calif.) and dispersed in 15 ml of complete growth medium containing I mg collagenase/ml (Worthington Biochemical Corp., Freehold, N.J.). A gas phase of 10% CO2 in air was added to each culture flask before it was sealed.
After incubation (37 0 C) for 18-24 hours followed by gentle pi petting, many clusters of epithelial cells remained undispersed among the singly dispersed connective tissue cells. The flasks were placed upright for several minutes to allow the cell clusters to settle out, then about 90% of the supernatant collagenase solution was pipetted off and discarded or centrifuged to recover the singly dispersed cells. These were sometimes cultured separately to produce a fibroblast strain in parallel with the epithelial cells. The visible cell clusters left by this settling process were diluted in growth medium without collagenase and distributed to additional culture flasks in amounts pro· viding 2 or 3 clusters/cm 2 • Cultures were fed twice a week (20 ml/flask) and the pH was adjusted carefully at each feeding. After several days' incubation, most epithelial clusters had attached and produced monolayer islands of epithelial cell outgrowth. Although the few fibroblasts remaining in these cultures grew more rapidly and surrounded the epithelial islands, we were able to remove them by a series of 1- to 2-minute exposures to a solution of 0.05% trypsin (1 :250; Difco Laboratories, Inc., Detroit, Mich.) and 0.02% Ver· sene (Fisher Scientific Co., Fairlawn, N.J.) in calciummagnesium-free saline (STV). If properly timed, this pro· cedure removed most of the fibroblasts but left the epithelial islands relatively undisturbed. These islands were rinsed twice and again incubated with growth medium. This treatment sometimes had to be repeated several times at weekly intervals before a uniform confluent monolayer of epithelial cells was formed. The monolayer could be dispersed (by 5-10 minutes' exposure to STV at 37 0 C) and divided among 2 or 3 flasks to begin serial subcultures. Many lots of commercial fetal serum are unsuitable for the growth of human epithelial cells. We tested each new lot against a reference serum for its ability to promote maximum growth rate and saturation density in a very serum·sensitive cell strain from human fetal intestine (FHs74 Int). Saturation density was determined by hemacytometer counts of viable cells (trypan blue exclusion) from replicate cultures seeded with 3x 1(}5 cells/flask (25 cm 2). The medium was changed twice a week and cells were counted from 2 flasks every 2 days for 14 days. Saturation density was taken as the value where three successive harvests showed no increase in cell number. For electron microscopy, monolayer cultures were fixed in situ with 2.5% glutaraldehyde in 0.1 N sodium cacodylate buffer (pH 7.3) and postfixed in Dalton's chrome osmium followed by 2% ethanolic uranyl acetate. The specimens were dehydrated in a graded ethanol series, Received October 24, 1975; accepted November 3, 1975. Supported by Public Health Service contract E73-200I-NOI CP33237 within the \'irus-
MATERIALS AND METHODS
The growth medium was Dulbecco's modified Eagle's medium (#H-21HG, Gibco, Grand Island, N.Y.), which contains 4.5 g/liter glucose. The medium was supplemented with 10% fetal calf serum (Gibco), IOp.g insulin/ml (Calbiochem, San Diego, Calif.), and 0.1 mM nonessential amino acids (Gibco, # 114). Tissue samples were collected in cold growth medium, minced immediately to about l-cm3 fragments, and kept at 4 0 C until processed for tissue culture, usually within 2 hours after surgery. Antibiotics (penicillin, streptomycin, and Fungizone) were included in the collection fluid and growth medium until the second subculture. Cell cultures were initiated by mincing the tissue fragments with apposed scalpels to about 2·mm cubes. Portions of this mince (50-100 fragments) were transferred to several 75-cm z culture flasks (Falcon Plastics, Los Angeles, Calif.) and dispersed in 15 ml of complete growth medium containing I mg collagenase/ml (Worthington Biochemical Corp., Freehold, N.J.). A gas phase of 10% CO2 in air was added to each culture flask before it was sealed.
After incubation (37 0 C) for 18-24 hours followed by gentle pi petting, many clusters of epithelial cells remained undispersed among the singly dispersed connective tissue cells. The flasks were placed upright for several minutes to allow the cell clusters to settle out, then about 90% of the supernatant collagenase solution was pipetted off and discarded or centrifuged to recover the singly dispersed cells. These were sometimes cultured separately to produce a fibroblast strain in parallel with the epithelial cells. The visible cell clusters left by this settling process were diluted in growth medium without collagenase and distributed to additional culture flasks in amounts pro· viding 2 or 3 clusters/cm 2 • Cultures were fed twice a week (20 ml/flask) and the pH was adjusted carefully at each feeding. After several days' incubation, most epithelial clusters had attached and produced monolayer islands of epithelial cell outgrowth. Although the few fibroblasts remaining in these cultures grew more rapidly and surrounded the epithelial islands, we were able to remove them by a series of 1- to 2-minute exposures to a solution of 0.05% trypsin (1 :250; Difco Laboratories, Inc., Detroit, Mich.) and 0.02% Ver· sene (Fisher Scientific Co., Fairlawn, N.J.) in calciummagnesium-free saline (STV). If properly timed, this pro· cedure removed most of the fibroblasts but left the epithelial islands relatively undisturbed. These islands were rinsed twice and again incubated with growth medium. This treatment sometimes had to be repeated several times at weekly intervals before a uniform confluent monolayer of epithelial cells was formed. The monolayer could be dispersed (by 5-10 minutes' exposure to STV at 37 0 C) and divided among 2 or 3 flasks to begin serial subcultures. Many lots of commercial fetal serum are unsuitable for the growth of human epithelial cells. We tested each new lot against a reference serum for its ability to promote maximum growth rate and saturation density in a very serum·sensitive cell strain from human fetal intestine (FHs74 Int). Saturation density was determined by hemacytometer counts of viable cells (trypan blue exclusion) from replicate cultures seeded with 3x 1(}5 cells/flask (25 cm 2). The medium was changed twice a week and cells were counted from 2 flasks every 2 days for 14 days. Saturation density was taken as the value where three successive harvests showed no increase in cell number. For electron microscopy, monolayer cultures were fixed in situ with 2.5% glutaraldehyde in 0.1 N sodium cacodylate buffer (pH 7.3) and postfixed in Dalton's chrome osmium followed by 2% ethanolic uranyl acetate. The specimens were dehydrated in a graded ethanol series, Received October 24, 1975; accepted November 3, 1975. Supported by Public Health Service contract E73-200I-NOI CP33237 within the \'irus-
