Cinemicrographic study of cell proliferation pattern and interdivision times of human keratinocytes in primary culture.

The stratified squamous epithelium of the skin, the epidermis, is a renewing cell population. In order for epidermis to remain the same size, each dividing basal cell must produce on an average, one daughter cell that will remain as a germinative cell and another that will perform terminal differentiation (differential mitosis). In order to investigate cell kinetics in the epidermis, the information obtained from in vivo study is limited, and offers only indirect evidence for the determination of cell cycle time and cell proliferation pattern. Keratinocytes in cell culture are unique in formation of a multilayered cellular sheet in which the keratinocytes form a structure resembling the epidermis in vivo, and keratinize at the top. In the early days of the primary culture of human keratinocytes, when the cells proliferate to form a monolayer sheet, direct access to cell proliferation pattern and measurement of interdivision time can be done using techniques of time-lapse cinemicrography. The primary cultures at 8-20 days of incubation were employed for cinemicrographic observation when small polygonal cells appeared in groups, when numerous mitotic figures were observed, and when stratification of the cells had not yet occurred. The appropriate field was marked and followed for up to 6 days. Photographic prints were made from the 16 mm cine film, and dendrograms were made and analyzed for pattern of cell proliferation and interdivision time. Most cells in the field divided two or three times during the period of observation. Sister-sister pairs of the second and third generations divided after approximately the same interdivision times. However, some cells have never divided. Some of the sister-sister pairs differed considerably in their interdivision times. In some cultures synchronous division was quite evident. The average interdivision time was about 26 hr in the majority of cultures, and it is suggested that the estimated long cell cycle time in vivo might be overestimated due to the existence of non-cycling cells in the germinative population.