Effects of retinoic acid on the growth and morphology of hamster tracheal epithelial cells in...

Virchows Arch B (1987) 54:38-51 g' Aw/dvB �9 Springer-Verlag 1987

Effects of retinoic acid on the growth and morphology of hamster tracheal epithelial cells in primary culture*

Elizabeth M. McDowell 1, Theresa Ben 2, Bill Coleman 1' 3, Seung Chang 1, Carnell Newkirk 1, and Luigi M. De Luca 2 1 Department of Pathology, University of Maryland School of Medicine, 10 South Pine Street, Baltimore, MD 21201 2 Differentiation Control Section, Laboratory of Cellular Carcinogenesis and Tumor Promotion, National Cancer Institute,

National Institutes of Health, Bethesda, MD 20892 3 Maryland Institute for Emergency Medical Services Systems, Baltimore, MD 21201, USA

Summary. Hamster tracheal epithelial cells were grown in primary culture on a collagen gel substrate in hormone-supplemented serum-free Ham's F12 medium with 1 0 - 8 M retinoic acid ( R A + ) , or without retinoic acid ( R A - ) . On days 1 and 2, the colonies were composed of large (secretory) cells and lesser numbers of small (basal) cells; ciliated cells were rare. At these times, cell number, thymidine incorporation, and total labelling indices (small and large cells, combined) were similar in R A + and R A - cultures, but the large cells became flat in R A - medium on day 2. On days 3- 5, thymidine incorporation and total labelling indices were less in R A - than R A + cultures, and on days 4-6, cell numbers were decreased in R A - cultures. On day 3, the large cells of the R A - colonies had flattened further and clusters of small basal cells had formed. On day 4, the RA + colonies were composed of densely-packed cuboidal secretory cells, small basal cells were inconspicuous; the total labelling index was about 27%. The R A - colonies were composed of large flat secretory cells and numerous small basal cells which were clustered in groups; the total labelling index was about 7%. Since large and small cells could be discrimi- nated by size in R A - colonies, a labelling index was generated based on cell size. On days 2, 3 and 4, the labelling index of the small basal cells in the R A - colonies was 44%, 43% and 24% respectively, whereas the labelling index of the large secretory cells fell rapidly over the same period (56%, 14% and 2%). On days 5 and 6, the cuboidal secretory cells in the RA + cultures had differentiated further and the cells were stratified focally. Some new ciliated cells had formed on day 6. In R A - cultures, mucous granules were not observed

* Supported in part by USPHS Grant #k 24722

Offprint requests to: E. McDowell at the above address

in the large flat cells and ciliated cells were not seen. The total labelling indices were 11% and 0.35% in R A + cultures, and 0.5% and 0.25% in R A - cultures on days 5 and 6, respectively.

The study shows that the target cell for vitamin A in the hamster tracheal epithelium is the secretory (mucous) cell. When retinoic acid was deficient, the secretory cells flattened and their capacity to divide was greatly diminished. Since the basal cells continued to replicate when the secretory cells did not, the population density of the basal cells increased disproportionally, which could be interpreted erroneously as a "basal cell hyperplasia." Real hyperplasia and epidermoid metaplasia were late secondary events which occurred in this study following focal disintegration of the epithelial cell sheet. Then the large secretory cells became keratinized and expressed an epidermoid phenotype.

Key words: Trachea - Epithelium - Retinoids - Cell culture - Hamster

Introduction

Wolbach and Howe (1925; 1928) demonstrated in their classic studies that vitamin A had profound effects on cell proliferation and differentiation in many organs including the tracheal epithelium. They concluded that normal differentiation was disturbed during vitamin A deficiency and often this was accompanied by increased rates of cell replication. Thus, for many years, it was assumed that hyperplasia and metaplasia, which occur in the tracheal epithelium during long-standing vitamin A deficiency, were primary effects brought about by the absence of vitamin A from the diet.

E.M. McDowell et al. : Effects of retinoic acid on tracheal epithelial cells in culture 39

This premise was challenged recently by the finding in rats and hamsters, that prior to the onset of hyperplasia and keratinizing epidermoid metaplasia, cell replication was markedly reduced by vitamin A deficiency (Zile et al. 1981 ; McDowell et al. 1984a; Strum et al. 1985), confirming an earlier report (Sherman 1961).

Furthermore, it has been widely accepted that the cell type most affected by vitamin A deficiency in the tracheal epithelium was the basal cell (reviewed by Shapiro 1986). This premise has also been challenged, since in vivo experiments in hamsters showed a marked decrease in the replication of columnar secretory (mucous) cells during the early stages of deficiency, whereas the basal cells were far less affected (McDowell et al. 1984a). Upon restoring vitamin A to the diet, the secretory cells proliferated, new secretory cells and ciliated cells were produced, and a normal mucociliary epithelium was reformed (McDowell et al. 1984b).

Characterization of the primary effects of any insult and separation from effects which result from superimposed secondary phenomena are not readily accomplished in vivo. Assuming that a sat- isfactory model is available, the primary effects of an insult are more easily defined in vitro. Hamster tracheal epithelial cells will reconstitute a functional mucociliary epithelium in serum-free primary culture, when grown on a collagen-gel substrate in hormone-supplemented medium in the presence of vitamin A (Wu et al. 1985). This in vitro model provides an exciting new tool for the study of tracheal epithelial cells in health and disease.

In this study, we have grown hamster tracheal epithelial cells in serum-free primary culture, with and without retinoic acid, to further define and characterize the effects of vitamin A on secretory cells and basal cells in the respiratory epithelium.

Materials and methods

Cell culture. Male Syrian golden hamsters (strain C R : R G H ) were obtained from the NCI-Frederick Cancer Research Facili- ty at 5-6 weeks of age. The tracheal epithelial cells were isolated by overnight protease digestion at 4 ~ C. The cells were cultured on a collagen gel, in a serum-free hormone supplemented Ham's F I 2 medium (Wu etal . 1985). Retinoic acid (dissolved in DMSO) was added at 10 -8 M or was omitted from the culture medium. The cells were plated into 35 mm plastic culture dishes (Costar | Cambridge, Ma, USA) or into Lab-Tek Permanox| plastic 1-chamber slides (Miles Scientific, Naperville, I11, USA) at a density of 1.2-1.5 x 103 cells/cm 2. The cells were incubated at 37 ~ C for 6-8 days in a humidified atmosphere of 95% air and 5% CO2. The culture medium was changed daily except on the first day after plating.

Cell counts, thyrnidine incorporation and labelling indices. Cell cultures, grown in 35 mm culture dishes (with 10 -8 M retinoic

acid, or without), were labelled for 2 h at 37~ with 4 ~tCi [3H]-thymidine (specific activity 5.0 Ci/mmol, Amersham Inter- national, Amersham, United Kingdom) per ml, in 2 ml of medium. The cells were washed three times with 0.1% unlabetled thymidine in PBS and released from the collagen gel by incuba- tion for 1 h at 37~ in 0.35% collagenase, followed by treatment with 0.1% trypsin and 0.02% EDTA for 15-30 min. After centrifugation (2000 rpm for 10 min), the cell pellet from each culture dish was resuspended in 1 ml of trypsin-EDTA solution. A 200 ~tl sample was taken, the cells were counted using a parti- cle counter (Coulter Electronics, Hialeah, FI, USA), and the number of cells per dish was calculated. The 800 I~1 of cell sus- pension remaining was then transferred to an Eppendorf tube and centrifuged. The cell pellets were stored frozen at - 8 0 ~ C until all the time points had been collected, at which time thymidine uptake was measured. After thawing in iced water, each cell pellet was resuspended in 1 ml 1 N NaOH (room tempera- ture for 10-20 min) to dissolve the cells. Then, 500 ~tl of bovine serum albumin (1 mg/ml), 2 ml of perchloric acid (0.2 N) and 150 ~tl of 10 N perchloric acid were added to each tube. The radioactivity in the supernatant was counted in a scintillation counter and the pellet was washed with 2-5 ml 0.2 N perchloric acid until 1 ml of the supernatant gave counts of 100 cpm or less. The pellet was then resuspended in 500 ~tl of 0.5 N perchloric acid and the DNA was hydrolysed at 90 ~ C for 20 min, followed by chilling in iced water. After centrifugation at 2000 rpm for 10 min, 100 ~tl of the supernatant (now containing the DNA) was counted in 10 ml Aquasol (NEN Research Prod- ucts, Boston, Ma, USA), using the scintillation counter. The results were calculated as cpm/cell for each culture dish. These procedures were carried out daily in two separate experiments, using duplicate or triplicate dishes for each treatment (with or without retinoic acid), on 6 consecutive days. In a third experiment, the cells were counted daily for 8 days, starting on day 2 (duplicate dishes for each treatment), but thymidine incorporation was not measured.

Cultures grown in plastic chamber slides were also labelled with [3H]-thymidine for 2 h, as described above. After washing three times with unlabelled thymidine in PBS, the cells were fixed for 30 min in 95% ethanol and air dried. The slides were dipped in Kodak NTB2 emulsion, drained and stored at 4 ~ C in the dark for 4 weeks. The autoradiograms were then developed in D-19 (Kodak, Rochester, New York, USA) and the cells were stained with hematoxylin. At least two slides were prepared on each day bearing cultures grown in the presence or absence of retinoic acid. Total labelling indices (percentage of large and small labelled cells) were derived by counting 1000 cells in each of 2 slides for each treatment, every day for 6 consecutive days. In addition, a labelling index according to cell size was derived on days 2, 3 and 4 in cultures grown without retinoic acid, because in this condition the cells were easy to distinguish based on size (see Results). On days 2 and 3, 500 large and 500 small cells were counted in each of 3 slides (i.e., 1500 large cells and 1500 small cells, each day). On day 4, 600 large and 600 small cells were counted in each of 2 slides (i.e., 1200 large cells and 1200 small cells). The percentage of labelled cells in the large and small cell populations was then determined.

Statistical analyses. Three types of analyses were used as appro- priate to the different types of data collected. The cell counts (such as in Fig. 1 A) were analyzed using the ordinary z2-test for a 2 • 6 table (combined data) and a 2 • 1 table (daily data). Labelling indices (as in Figs. 1 C, and 10) were analyzed using the linear logistic model as described in Coleman and McDow- ell (1985), except that the commercial SAS computer package (SAS Institute, Box 8000, Cary, NC USA) was used, which

40 E.M. McDowell et al. : Effects of retinoic acid on tracheal epithelial cells in culture

has a slightly different test for individual parameters. The radioactivity counts from [3H]-thymidine (as in Fig. 1 B) cannot simi- larly be regarded as being binomially distributed but rather have a Poisson distribution (De Groot 1975; Cinlar 1975) with parameter 2. This parameter equals n times ~t, where ~ is an underlying parameter, perhaps different for each combination of day and treatment, and n is the corresponding total number of cells examined. Using this model, parameters were fit using maximum likelihood as described in Coleman and McDowell (1985).

Morphological studies. The morphology of the cell colonies was monitored by phase contrast microscopy in three separate experiments. Representative specimens were prepared for high resolution light microscopy, autoradiography and transmission electron microscopy. Cells growing on the collagen gel substrate were fixed by adding 2 ml of 4% formaldehyde-1% glutaralde- hyde in a 200 mOSM phosphate buffer (McDowell and Trump 1976) to 2 ml of the culture medium. After 10-15 min, full- strength fixative replaced the half-strength mixture, and the cells were stored in the same until processing. For autoradiography, [3H]-thymidine (the same as above) was added to the culture medium for 1 h at 37 ~ C before fixation. The colonies were prepared for high resolution light microscopy, autoradiography and transmission electron microscopy, as described previously (McDowell et al. 1984a).

R e s u l t s

The growth kinetics of hamster tracheal epithelial cells cultured with (RA + ) or without retinoic acid ( R A - ) are shown in Fig. 1. Growth of cells was inhibited when RA was omitted from the culture medium. The number of cells per dish in R A + and R A - cultures was similar on days 1 and 2 (Fig. 1 A). The number of cells had decreased in the R A - cultures in one experiment on day 3 (p < 0.0001, Fig. 1 A), and in all three experiments on days 4, 5, and 6 (p-- <0.0001). Thymidine incorporation and labelling indices were also similar on days 1 and 2 in RA + and R A - cultures, but on days 3, 4, and 5, the values were lower (p <0.0001) in the R A - cultures. Thymidine incorporation and total labelling indices (large and small cells) were very low in both RA + and R A - cultures on day 6 (Fig. 1, B, C).

Days 1-3

On the first day, the morphology of RA + and R A - cultures was similar. Some ciliated cells were present but the colonies were composed for the most part of large and small non-ciliated cells. The large cells (flattened secretory cells) had an abun- dant cytoplasm and phase-lucent nucleus. The small cells (basal cells) had a scant cytoplasm and phase-dense nucleus. Large and small cells synthe- sized DNA, contributing to a total labelling index of about 8% (Fig. 1C).

A_. 100

80.

. c

60. ? o

E 4O. c

20.

/ \

//

o i 2 3 4 s /,

B. 2 0 0

160

x

~ 120

, 8 0

4O

C_. 50.

tt t ~. 40_

10.

i 2 5 4 S 6 TIME IN DAYS

i 2 a 4 S 6

Fig. 1 A-C. Graphs A, B and C are derived from the same experiment. On day 0, 1.2x 103 tracheal cells/cm 2 were seeded onto collagen gel substrate in 35 mm culture dishes (A and B) or chamber slides C. The cells were grown in primary culture with retinoic acid 1 0 - a M (solid line) or without retinoic acid (dashed line). The medium was changed daily, starting on day 2. A Growth kinetics. Cells attached to the substrate were collected and counted after treatment with collagenase and trypsin. Duplicate dishes were used to determine the mean cell number at each time point. Cell numbers were determined with a parti- cle counter (Coulter). B Incorporation of [3H]-thymidine by the cells shown in A. C Total labelling indices. Percentage of [3H]-thymidine labelled nuclei in small and large cells, growing on the collagen gel substrate in chamber slides

The colonies had increased in size on days 2 and 3. Mosaics of large and small non-ciliated cells formed a simple (one-layered) epithelium in RA + and R A - cultures. Ciliated cells were uncommon. Differences between RA + (Figs. 2, 5, 6) and R A - colonies (Figs. 3, 4, 7-9) were obvious on days 2 and 3 because the large cells became progressively flattened in R A - colonies, and their surface area was increased compared with that of large cells in the R A + colonies which maintained a more cuboidal shape. Accordingly, the R A - colonies were much larger than the RA + colonies, but the cell number was about the same or less (Fig. 1 A).


Figs. 2-9. Hamster tracheal epithelial cells on days 2 and 3 of culture. Fig. 2. RA + colony, day 2. Most of the cells are large, with broad cytoplasm and phase-lucent nucleus (arrows). A few cells are small; their cytoplasm is scant and the nuclei are phase-dense (arrow heads). Phase contrast, light micrograph x 300. Fig. 3. R A - colony, day 2. The colony is composed of large and small cells but the surface area of many of the large cells is greater than in the RA + colony (Fig. 2). Many of the large cells are multinucleated (arrows). Small cells (arrow heads). Phase contrast, light micrograph • 300. Fig. 4. R A - colony, day 2. The nuclei of some small cells (circled) and large cells (arrows) are labelled with [3H]-thymidine. Several large cells are multinucleated (lower arrow). Autoradiogram, hematoxylin stain, light micrograph x 120. Fig. 5. R A + colony, day 3. Many nuclei are labelled with [3H]-thymidine, but because the large cells are cuboidal and tightly-packed it was for the most part impossible to discriminate between large cells and small cells. Autoradiogram, hematoxylin stain, light micrograph x 120. Fig. 6. Vertical section through RA + colony, day 3. The large cells are tightly-packed and cuboidal (arrows). Cells at the edges (arrow heads) rise steeply from the collagen gel. Glycol methacrylate section, PAS-lead hematoxylin stain, light micrograph • 140. Fig. 7. R A - colony, day 3. Numerous small cells (S) are in a cluster. The large cells (L) are very flat and their edges are spread thinly over the collagen gel (arrows). Phase contrast, light micrograph x 200. Fig. 8. Vertical section through R A - colony, day 3. The large cells are flat and their edges (arrow heads) are spread thinly over the collagen gel (cg). Glycol methacrylate section, PAS-lead hematoxylin stain, light micrograph x 140. Fig. 9. R A - colony, day 3. Small cells (S) are in clusters and many of their nuclei are labelled with [3H]-thymidine. Very few nuclei of the large cells (L) are labelled. Autoradiogram, hematoxylin stain, light micrograph • 120


60. m , m m small = m ~ = l a r g e

t t

50 t t t

"~ 40

~ :

~ 30.

20. t z t

t

k lO ~

2 3 4

Days in Culture minus RA

Fig. 10. The large cells were very flat in RA- cultures on days 24 and so it was possible to discriminate between the nuclei of large and small cells. The percentage of small (basal) cells or large (secretory) cells with [3H]-thymidine-labelled nuclei are shown in the graph. Small cells, solid line; large cells, dashed line

On day 3, the cytoplasm spread as a thin veil over the collagen gel substrate, from the edges of the large cells at the periphery of the R A - colonies (Figs. 7, 8), whereas the exterior aspect of large cells at the periphery of the RA + colonies rose steeply from the collagen gel (Fig. 6).

On day 2, the total labelling indices were similar in R A + and R A - colonies, about 50%. The labelling index remained high in RA + cultures on day 3 (about 47%) but had fallen to 28% in RA-- cultures (Fig. 1 C). Because the large cells were so flat in the R A - colonies, thereby greatly increasing their surface area, it was relatively easy to discriminate between large and small cells in the autoradiograms (Figs. 4, 9) and to generate a labelling index according to cell size (Fig. 10). On days 2 and 3, the labelling index of the small cells in R A - cultures was the same, about 43%, falling (p= <0.0001) to about 24% on day 4. However, the labelling index of the large cells was about 56% on day 2, but had fallen to 14% on day 3 (p< 0.0001) and to 2.4% on day 4 (p <0.0001) (Figs. 4, 9, 10). On each day the difference in the labelling index between large and small cells was highly sig- nificant (p= <0.0001). Therefore, the fall in the total labelling index in the R A - cultures on day 3 (Fig. 1 C) was due to a marked and rapid decline in DNA synthesis in the large (secretory) ceils. Nests of small cells (Fig. 7), many of which were labelled (Fig. 9) were seen in the R A - colonies

on day 3. Beginning on day 2, many of the large cells were multinucleated (Figs. 3, 4).

The colonies were examined ultrastructurally on days 2 and 3. On day 2, cytoplasmic ribosomes were dispersed and the rough endoplasmic reticulum was scant in large cells in R A - and R A + cultures. The RER was better developed on day 3 and cytoplasmic ribosomes were aggregated into polyribosomes. The intercellular spaces were not wide and simple lateral membranes were joined by small desmosomes. The apical cell membranes which interfaced with the culture medium were joined by junctional complexes.

Day 4

The RA + colonies were composed of a monolayer of non-ciliated cells. The thickness of the epithelial sheet varied as the large cells became more cuboidal and dense-packed in focal areas (Figs. 11, 12). Consequently, it became increasingly difficult to discriminate between large and small cells, but it was obvious that most of the epithelial sheet was composed of large cuboidal cells (Figs. 12, 15), many of which were synthesizing DNA (Figs. 11- 13). The total [3H]-thymidine labelling index was about 27% (Fig. 1 C) and several mitotic cells were observed (Fig. 14). The RER was better developed than on day 3 and mucous granules were seen in some of the cuboidal secretory cells. The exterior aspect of the cells at the periphery of the colonies rose steeply (Fig. 14) and the edges of the cells were not spread on the collagen gel (Fig. 15).

The R A - colonies were also composed of a monolayer of non-ciliated cells but the morphology was very different from the R A + colonies (Figs. 16-19). Small and large cells were readily distinguished because the large cells were very flat and the small cells were numerous and clustered in groups (Figs. 16, 17). About 24% of the small cells were in DNA synthesis compared with only 2.4% of the large cells (Figs. 10, 18, 19). The total labelling index (small and large cells combined) was very low in R A - cultures, about 7% (Fig. 1 C). The RER was better developed in the large cells than on day 3, but mucous granules were not observed.

Day 5

In the RA + cultures, some of the cuboidal cells had differentiated into functional secretory cells (Fig. 20) and in a few cells the granules were abun- dant (Fig. 22). Ciliated cells were absent. At focal


Figs. 11-19. Hamster tracheal epithelial cells on day 4 of culture. Fig. 11. RA+ colony. Many of the nuclei are labelled with [3H]-thymidine. The cuboidal cells are tightly packed in focal areas. Autoradiogram, hematoxylin stain, light micrograph x 100. Fig. 12. Vertical section through RA + colony showing tightly-packed cuboidal cells (center to left) and flatter cells (center to right). The nucleus of one of the cuboidal cells is lightly labelled with [3H]-thymidine (arrow). Glycol methacrylate section, autoradiogram, PAS-lead hematoxylin stain, light micrograph • 375. Fig. 13, Vertical section through RA + colony showing large cuboidal cell. The nucleus is labelled with [aH]-thymidine. Glycol methacrylate section, autoradiogram, PAS-lead hematoxylin stain, light micrograph • 1200. Fig. 14. Tangential-vertical section through edge of RA + colony. The cuboidal cells are tightly-packed and rise steeply (at left) from the collagen gel. Two of the cuboidal cells are in mitosis (rn). Glycol methacrylate section, PAS-lead hematoxylin stain, light micrograph x 350. Fig. 15. Periphery of an RA + colony. A few cells with phase-dense nuclei are small (circles), but the bulk of the colony is composed of tightly-packed cuboidal cells, with phase-lucent nuclei. Phase contrast, light micrograph x 140. Fig. 16. R A - colony. Clusters of small cells (at right) are readily distinguished from the large flat cells (L, at left). Many of the large ceils are multinucleated, Phase contrast, light micrograph x 380. Fig. 17. Vertical section through R A - colony, similar to that shown in Fig. 16. Large flat cells (left of arrow) merge with a cluster of small cells (right of arrow). Collagen gel (cg). Glycol methacrylate section, PAS-lead hematoxylin stain, light micrograph • 380. Fig. 18. R A - colony. Nuclei of several small cells in a cluster are labelled with [3H]-thymidine, but large cells surrounding the cluster are not labelled. Autoradio- gram, hematoxylin stain x 160. Fig. 19. Vertical section through R A - colony. The nucleus of a small cell is labelled with I3H] - thymidinc; the large cell at left is r~ot labelled. Glycol methacrylate section, autoradiogram, PAS-lead hematoxylin stain, light micrograph x 1200

Figs. 20--26. Hamster tracheal epithelial cells on day 5 of culture. Fig. 20. Vertical section through RA + colony. The large cells are cuboidal. The rough endoplasmic reticulum (rer) is moderately well developed and a mucous granule (m) is present in the apical cytoplasm. Autophagic vacuole (av); collagen gel (cg). Electron micrograph • Fig. 21. RA+ colony. Nuclei of several cells are labelled with [3H]-thymidine. Cells are stratifying in circular foci. Autoradiogram, hematoxylin stain • 125. Inset: Vertical section through a stratified focus shows labelled nuclei in columnar cells which reach to the surface, and in cells deep in the epithelium (arrows). Glycol methacrylate section, autoradiogram, PAS-lead hematoxylin stain, light micrograph x 800. Fig. 22. RA+ colony. A cluster of mucous granules in the apical cytoplasm of a cuboidal secretory cell. Electron micrograph • 20000. Fig. 23. Vertical section through R A - colony. The rough endoplasmic reticulum (rer) is moderately well developed in the large flat cell, but mucous granules are not seen. The cell is multinucleated and nuclear heterochromatin is dispersed. Collagen gel (cg). Electron micrograph • 7200. Fig. 24. R A - colony. Only 2 cells in this field (circles) have nuclei labelled with [3H]-thymidine. The labelling index was very low on day 5 in R A - cultures (Fig. 1 C). Autoradiogram, hematoxylin stain, light micrograph • I10. Fig. 25. Vertical section through the edge of an R A - colony shows a thin cytoplasmic extension over the collagen gel

(cg). Note thick band of microfilaments (mf), running beneath the apical plasma membrane. Electron micrograph • 7500. Fig. 26. Periphery of an R A - colony showing extreme flattening of the large cells, comparable to that shown in Fig. 25. Phase contrast, light micrograph • 120


Figs. 27-33. Hamster tracheal epithelial cells on day 6 of culture. Fig. 27. Vertical section through RA + colony. The cells are stratified at center, but the colony is one cell thick at right. Glycol methacrylate section, PAS-lead hematoxylin stain, light micrograph x 500. Fig. 28. Vertical section through RA + colony. This part of the epithelial sheet is multilayered. Large pale-stained ciliated cells (c) have formed in the upper stratum. PAS-positive mucous granules (m) are present in a secretory cell. Glycogen (g). Glycol methacrylate section, PAS-lead hematoxylin stain, light micrograph x 800. Fig. 29. RA + colony. The cells are tightly- packed and stratified in focal areas but none of the nuclei of cells in this field are labelled with [3H]-thymidine. Autoradiogram, hematoxylin stain, light micrograph x 125. Fig. 30. R A + colony. Typical appearance of an area where ciliary beating was seen by phase contrast microscopy. Phase contrast, light micrograph x 500. Fig. 31. Horizontal section through R A - colony. The large cells (L) are very flat. One of the cells is vacuolated (v). The other shows cytoplasmic blebs (b). Broad bands of microfilaments (mJ) course through the cytoplasm. Intercellular spaces (is) are wide. Electron micrograph x 4000. Fig. 32. R A - colony. Phase contrast appearance of large cells (L), similar to those in Fig. 31. Some of the cells are vacuolated (v). The intercellular spaces (is) are wide. Phase contrast, light micrograph x 500. Fig. 33. R A - colony. None of the nuclei in large (L) or small cells (S) in this field are labelled with [3H]-thymidine. Autoradiogram, hematoxylin stain, light micrograph x 125


Figs. 34-37. Hamster tracheal epithelial cells on days 6--8 of culture. Changes are shown that occurred in RA- cultures, associated with lysis of the collagen gel and disintegration of the epithelial cell sheet. Figs. 34, 35. Day 6. Cells which lifted from the collagen gel have clumped together to form irregular branching cords (C). A vertical section through the line in Fig. 34 would appear as shown in Fig. 35. The cells in the cords are rounded but viable. Flat cells have migrated from the cords and cover the surface of the plastic dish. Figure 34 Phase contrast, light micrograph x 220. Figure 35: Glycol methacrylate section, PAS-lead hematoxylin stain, light micrograph x 220. Fig. 36. Day 8. Large (L) and small (S) cells participate in swirling patterns, characteristic of epidermoid metaplasia. Phase contrast, light micrograph x 220. Fig. 37. Day 8. Typical ultrastructure of cells shown in Fig. 36. The perinuclear bundles of keratin tonofilaments (T) are well-developed in the large cells (altered secretory cells). The cells are joined by desmosomes (d); intercellular spaces (is) are very wide. Electron micrograph x 5550

areas, the cells were clustered tightly together and piled up one upon the other (Fig. 21). About 11% of all epithelial cells were in D N A synthesis (Fig. 1 C) and labelling with [3H]-thymidine was seen in the nuclei of columnar cells that reached to the surface, and in cells deep in the epithelial sheet (Fig. 21, inset).

In the R A - cultures, the R E R was quite well- developed in the large flat cells, but mucous granules were not observed. The nuclear heterochromatin was dispersed and mult inucleated cells were quite common (Fig. 23). At the periphery o f the colonies, the large cells were extremely flat (Fig. 26). Broad bands of microfilaments were seen beneath the apical cell surface and thin cytoplasmic extensions spread over the collagen gel (Fig. 25). The total labelling index was very low (about 0.35%, Fig. 1 C) and very few small or large cells were labelled with [3H]-thymidine (Fig. 24).

Day 6

The cells were stratified in the thicker parts o f the RA + cultures (Figs. 27, 28) but very few cells were labelled with [3H]-thymidine (Fig. 29). The phase contrast appearance of the epithelial cell sheet appeared similar to day 5, except ciliated cells were now quite common (Figs. 28, 30).

Stratification was not observed in R A - cultures. The large cells were extremely flat and their surface area was extensive (Figs. 31, 32). Longitu- dinal bundles of microfilaments coursed through the cytoplasm (Fig. 31) and some cells showed cytoplasmic vacuolat ion (Figs. 31, 32). The intercellular spaces were wide and the peripheral cytoplasm of some cells showed blebbing (Fig. 31). The total labelling index was very low (about 0.25%) and very few small or large cells were labelled with [3H]-thymidine (Figs. 1 C, 33).

E.M. McDowel l et al. : Effects o f retinoic acid on tracheal epithelial cells in culture 47

Additional observations

The foregoing morphological descriptions specifically refer to growth of hamster tracheal epithelial cells on the collagen gel substrate. However, starting on day 3 in R A - cultures and day 5 in RA + cultures, focal lysis of the collagen gel occurred, accompanied by lifting of the epithelial cell sheet. The sheet then broke up into many small holes and the cells clumped together to form broad cords which attached to the bottom of each dish in irregular branching patterns (Fig. 34). The cells in these cords rounded up but remained viable (Fig. 35). Small and large cells then migrated from the cords over the surface of the plastic dish to form a thin monolayer (Figs. 34, 35). This sequence occurred to varying extents in all three experiments. How- ever, in one experiment, where growth of the cells was greatest, it appeared as if cells which had migrated from the irregular branching cords proliferated to provide swirling patterns of large and small cells in the RA-cultures on days 7 and 8, typical of epidermoid metaplasia (Figs. 36, 37). Well-developed perinuclear bundles of tonofilaments (keratin) were present in many of the large cells which expressed the epidermoid phenotype (Fig. 37).

Discussion

The use of culture techniques that nurture epithelial cells in serum-free media will foster meaningful studies of the multiplicity of mechanisms that control cell growth and differentiation. Many labora- tories have cultured tracheobronchial epithelial cells in attempts to reproduce a differentiated and functional mucociliary epithelium in vitro (Van Scott etal. 1986; Wu 1986). Success has been achieved by Wu et al. (1985) using hamster tracheal epithelial cells, grown on a collagen-gel substrate, in a serum-free hormone-supplemented medium. This model is especially well-suited to study factors which influence growth and differentiation of tracheal epithelial cells.

Although the mechanisms are as yet poorly- defined, the extracellular matrix upon which epithelial cells grow in vitro, is extremely important for establishing 3 dimensional cell-to-cell relation- ships, determining cell polarity and shape, regulat- ing growth, and modulating differentiation (Gos- podarowicz et al. 1978; Hay 1981; Kleinman et al. 1981; Bissel et al. 1982; Montesano 1986; Watt 1986). If hamster tracheal cells are seeded onto a thin collagen film instead of a thick collagen gel, growth and differentiation of the cells is altered and the mucociliary state is not reproduced, even

if the cells are grown in the hormone-supplemented medium containing retinoids (Wu etal. 1985). Therefore, the observations reported in this study refer specifically to cells growing on the thick collagen gel, unless stated otherwise.

It is well established that vitamin A is essential for normal growth, differentiation and mainte- nance of epithelial cells (Lotan 1980; De Luca and Shapiro 1981; Sporn and Roberts 1983; Zile and Cullum 1983; Schroder et al. 1983; Jetten 1984; Wolf 1984; Goodman 1984; Shapiro 1986). How- ever, the widespread assumption that hyperplasia and epidermoid metaplasia are primary effects of vitamin A deficiency in the tracheobronchial epithelium appears to be ill-founded. When tracheobronchial cells from different species were grown in hormone-supplemented media in the absence of retinoids, cell proliferation was decreased and the total number of cells was reduced (Klann and Mar- chock 1982; Lechner et al. 1982; Jetten and Smits 1985; Groelke et al. 1985; Wu et al. 1985; B6ning et al. 1986; Niles et al. 1986).

The present observations are consistent with these reports. Moreover, the changes which occur in cell culture are consistent with those morphological changes which occur in the tracheal epithelium early during vitamin A deficiency in vivo and in organ explants in vitro, long before epidermoid metaplasia is established. These early changes in- clude a marked reduction in cell proliferation, changes in cell shape from columnar to low cuboidal, an increased density of basal cells and a decreased density of ciliated cells (Sherman 1961; Wong and Buck 1971; Harris et al. 1972; Boren et al. 1974; Clark and Marchok 1979; Zile et al. 1981 ; Chopra 1982; McDowell et al. 1984a; Strum et al. 1985; Huang et al. 1986). In hamsters, prior to complete cessation of cell growth in vivo, the replication of secretory cells was reduced 14-fold, whereas basal cell replication was reduced only 3-fold. This was accompanied by dispersion of heterochromatin in the nuclei of the secretory cells (McDowell et al. 1984a) and occasional multinucleation (McDowell, unpublished work). These changes occurred at the weight plateau stage, when serum and liver levels of vitamin A had fallen be- low detectable limits (Strum et al. 1985).

In the present study it was obvious that cell flattening and reduced DNA synthesis affected the population of large cells in the R A - cultures, which were derived from columnar secretory (mucous) cells. These changes were accompanied by multinucleation and dispersion of nuclear heterochromatin. The small cells, which derived from basal cells, continued to synthesize DNA and repli-


cated when DNA synthesis was greatly reduced in the large secretory cells. Since the basal cells continued to proliferate when the secretory cells did so only minimally, clusters of small basal cells formed in the R A - colonies. The clusters of basal cells were well established by day 4, and are consistent with reports of increased basal cell density during the early stages of vitamin A deficiency in vivo (Wong and Buck 1971; Harris etal. 1972; Boren et al. 1974; McDowell et al. 1984a; Strum et al. 1985).

The disparate effects that absence of vitamin A imparts on the replication rates of secretory cells and basal cells gives the basal cells a growth advan- tage which has been erroneously interpreted as a basal cell "hyperplasia" (see McDowell et al. 1984a for discussion). Although the number of basal cells increased relative to the number of secretory cells (because the basal cells continued to replicate when the secretory cells did not) there was no evidence that the basal cells differentiated into any other cell type, or that the "hyperplastic" basal cells provided for an epidermoid metaplastic epithelium (Wolbach and Howe 1925, 1928; Wil- helm 1954; Wong and Buck 1971; Harris et al. 1972; Port et al. 1974; Chopra 1982). On the con- trary, since the basal cells clustered in groups, our results support the idea that basal cells are a self- maintaining pool that do not differentiate into other cell types (McDowell et al. 1985). The function of basal cells in the normal tracheal epithelium is still uncertain, but they may play a role in epithelial adhesion. Unlike secretory cells and ciliated cells, basal cells of the tracheobronchial epithelium are attached to the basal lamina by hemidesmo- somes (Frasca et al. 1968; Kawanami et al. 1979; Tandler et al. 1981 ; McDowell et al. 1983). During vitamin A deficiency, when secretory cell replication is minimal, the basal cells may play a life- saving role by increasing the adhesion of secretory cells and preventing their loss from the epithelium. It is obviously crucial that the mitotically quiescent secretory cells be preserved during vitamin A deficiency, because upon repletion these cells divide to provide new secretory cells and ciliated cells (McDowell et al. 1984b), thereby restoring a normal mucociliary epithelium.

Vitamin A influences the biosynthesis of glycoproteins (De Luca 1977; Wolf 1984) and is required for normal differentiation of tracheal secretory cells and mucus production (Bonanni et al. 1973; Bonanni and De Luca 1974; Goldman and Baseman 1980; Clark and Marchok 1979; Clark et al. 1980; Wu et al. 1982, 1985). In the present study, a few mucous granules were seen at the

apices of some of the large cuboidal cells in RA + cultures on day 4; the granules were larger and more numerous on day 5. Mucous granules were not observed in the large flat cells in the R A - cultures.

Retinoic acid is reported to enhance the synthesis of receptors for several different growth hormones (Jetten 1980; Heath et al. 1981 ; Taylor et al. 1985). Therefore, retinoic acid may be required for the synthesis of certain growth factor receptors (which are glycoproteins), in the plasma membranes of tracheal secretory cells. If the synthesis of receptors for a growth factor was impaired by retinoid deficiency, a situation analogous to omis- sion of that growth factor would be produced. Fur- thermore, the addition of certain growth factors to various cell types caused morphological alterations which were associated with changes in the organization of the actomyosin cytoskeletal net- work (Brunk etal. 1976; Chinkers etal. 1979; Bockus and Stiles 1984; Goshima et al. 1984; Ka- dowaki et al. 1986). Moreover, complex relation- ships exist between the microfilamentous actin cy- toskeleton, transmembrane receptors and the extracellular matrix, which play a major role in determining cell growth and adhesion and cell shape and differentiation (Bissell et al. 1982; Ben-Ze'ev 1986; Watt 1986). Agonist-receptor interactions have been shown to generate the two second mes- senger molecules inositol-tris phosphate and diacylglycerol, by phospholipase-C cleavage of phos- phatidyl inositol 4,5-bisphosphate (Berridge 1984). Defective cell surface receptor activity, such as might be caused by vitamin A deficiency, may also be responsible for insufficient signal transduction from the cell surface. An indication of this possibil- ity has been found in intestinal epithelium from vitamin A-deficient hamsters (De Luca et al. 1987).

The present study, and the previous study in vivo (McDowell et al. 1984a), show that retinoids are important for maintaining normal growth, shape and differentiation of tracheal secretory cells. Our observations emphasize that marked de- creases in the replication of secretory cells and changes in their shape, are the primary effects of vitamin A deficiency in the tracheal epithelium. The exact nature of the stimulus for the onset and production of epidermoid metaplasia, which occurs following prolonged vitamin A deficiency in vivo (Strum et al. 1985), remains unknown. How- ever, the formation of epidermoid metaplasia was accelerated in the tracheas of vitamin A deficient animals in vivo, when an injury was applied to the epithelium (Wilhelm 1954; Bang and Bang 1969; Bang et al. 1975; Lane and Gordon 1979).


The rapidly induced metaplastic lesion shared many of the morphological characteristics of the metaplastic tracheal epithelium of vitamin A-replete animals and humans, following diverse forms of injury. Hyperplasia, followed by epidermoid metaplasia, are stages that occur normally during epithelial regeneration (reviewed by McDowell et al. 1980; Keenan 1987). In vitamin A-replete animals, epidermoid metaplasia is transient and paves the way for restoration of the mucociliary state, but in vitamin A-deficient animals, the tran- sition to mucociliary differentiation is blocked and a metaplastic state persists until death.

In the present study, changes characteristic of keratinizing epidermoid metaplasia (Figs. 36, 37) only occurred following focal disruption of the epithelial cell sheet. The reason for the epithelial disruption in vitro is unknown and probably has no direct bearing on the situation in vivo. However, consistent with the in vivo observations, epidermoid metaplasia arose in the R A - cultures after an epithelial injury, further emphasizing that epidermoid metaplasia is a secondary event that occurs late during vitamin A deficiency, as an attempt (al- beit thwarted) to regenerate the tracheal epithelium.

Retinoids are known to influence cellular adhesion phenomena. They greatly enhance fibroblast to plastic-adherence (Adamo etal. 1978, 1979; Bertram et al. 1981; Kamei 1983) and cell-to-cell adhesiveness of pancreatic islets cells (Chertow et al. 1983). They reduce cell-to-cell adhesiveness of cloned BHK21/C13 fibroblasts (Kamei 1983) and reduce the adhesiveness to the culture dish of rat tracheal cells (Klann and Marchok 1982), colon adenocarcinoma cells (Shapiro and Poon 1979), mouse mammary carcinoma cells (Walker- Nasir et al. 1985), and human epidermis (Tammi and Tammi 1986). Altered adhesiveness of tracheal epithelial cells, due to deficiency of vitamin A, may influence cell-to-cell adhesion, thereby simulating a wound in the tracheal epithelium. In turn, this may trigger events which stimulate cell proliferation, leading to hyperplasia and epidermoid metaplasia.

In summary, this study demonstrates that the primary target for retinoids in the hamster tracheal epithelium is the secretory (large) cell population. In the absence of retinoic acid, the secretory cells flattened and their capacity to divide was much reduced, whereas the basal cells continued to replicate so that their number increased disproportionally. There was no evidence that the basal cells differentiated into any other cell types or that the "hyperplastic" basal cells expressed an epidermoid

(squamous) phenotype. Epidermoid metaplasia was a late secondary event which only occurred in this study following focal disruption of the epithelial cell sheet. Small (basal) and large (secretory) cells participated in the metaplastic lesions in the R A - cultures and the altered secretory cells expressed the epidermoid phenotype and were heav- ily keratinized as described in vivo (McDowell et al. 1984a).

Acknowledgement. The authors thank Janet S. Folmer, Perry Comegys and Delores Michael for their help and expertise in preparing ultrathin sections, publication photomicrographs and autoradiograms respectively. They also thank Dr. J.M. Strum for her critical review of the manuscript.

References Adamo S, Akalovsky i, De Luca LM (1978) Retinoic acid-

induced changes in saturation density and adhesion of transformed mouse fibroblasts. Proc Am Assoc Cancer Res 19:Abs107

Adamo S, De Luca LM, Akalovsky I, Bhat PV (1979) Retinoid- induced adhesion in cultured transformed mouse fibroblasts. J Natl Cancer inst 62:1473-1478

Bang, BG, Bang FB (1969) Replacement of virus-destroyed epithelium by keratinized squamous cells in vitamin A-de- prived chickens. Proc Soc Exp Biol Med 132:50-54

Bang FB, Bang BG, Foard M (1975) Acute Newcastle viral infection of the upper respiratory tract of the chicken. Am J Pathol 78:417-426

Ben-Ze'ev A (1986) The relationship between cytoplasmic organization, gene expression and morphogenesis. TIBS 11 : 478-481

Berridge MJ (1984) lnositol triphosphate and diacylglycerol as second messengers. Biochem J 220:345-360

Bertram JS, Mordan L J, Blair SJ, Hui S (1981) Effects of retinoids on neoplastic transformation, cell adhesion, and membrane topography of cultured 10TI/2 cells. Ann New York Acad Sci USA 359:218-236

Bissell M J, Hall HG, Parry G (1982) How does the extracellular matrix direct gene expression? J Theor Biol 99:31 68

Bockus BJ, Stiles CD (1984) Regulation of cytoskeletal archi- tecture by platelet-derived growth factor, insulin and epidermal growth factor. Exp Cell Res 153:186-197

Bonanni F, De Luca L (1974) Vitamin A-dependent fucose- glycopeptide from rat tracheal epithelium. Biochim Biophys Acta 343 : 632-637

Bonanni F, Levinson SS, Wolf G, De Luca L (1973) Glycopro- teins from the hamster respiratory tract and their response to vitamin A. Biochim Biophys Acta 297:441-451

B6ning W, Kakunaga T, Emura M, Mohr U (1986) Growth stimulation by retinol apparently independent of EGF receptor. Exp Cell Res 163:255-260

Boren HG, Pauley J, Wright EC, Kaufman DG, Smith JM, Harris CC (1974) Cell populations in the hamster tracheal epithelium in relation to vitamin A status. Int J Vitam Nutr Res 44:382 390

Brunk U, Schellens J, Westermark B (1976) Influence of epidermal growth factor (EGF) on ruffling activity, pinocytosis and proliferation of cultivated human glia cells. Exp Cell Res 103:295-302

Chertow BS, Baranetsky NG, Sivitz WI, Meda P, Webb MD, Shih JC (1983) Cellular mechanisms of insulin release. Ef- fects of retinoids on rat islet cell-to-cell adhesion, reaggrega- tion, and insulin release. Diabetes 32:568-574


Chinkers M, McKanna JA, Cohen S (1979) Rapid induction of morphological changes in human carcinoma cells A-431 by epidermal growth factor. J Cell Biol 83:260-265

Chopra DP (1982) Squamous metaplasia in organ cultures of vitamin A-deficient hamster trachea: cytokinetic and ultra- structural alterations. J Natl Cancer Inst 69:895-905

Cinlar E (1975) Introduction to stochastic processes. Prentice- Hall, Inc, Englewood Cliffs, N J, pp 70-105

Clark JN, Marchok AC (1979) The effect of vitamin A on cellular differentiation and mucous glycoprotein synthesis in long-term rat tracheal organ cultures. Differentiation 14:175-183

Clark JN, Klein-Szanto AJP, Pine AH, Stephenson KB, Mar- chok AC (1980) Reestablishment of a mucociliary epithelium in tracheal organ cultures exposed to retinyl acetate: a biochemical and morphometric study. Eur J Cell Biol 21:261 268

Coleman B, McDowell EM (1985) Development of hamster tracheal epithelium. III. Illustration of statistical methods for proportional data in biology. Anat Rec 213:457-463

De Groot MH (1975) Probability and statistics. Addison-Wes- ley Publishing Co, Reading, Mass, p 209

De Luca LM (1977) The direct involvement of vitamin A in glycosyl transfer reactions of mammalian membranes. Vi- tam Horm 35:1-57

De Luca LM, Shapiro SS (1981) Modulation of cellular interactions by vitamin A and derivatives (retinoids). Ann New York Acad Sci 359 : 1-428

De Luca LM, Grover A, Shankar S (1987) Retinoic acid en- hances the iocorporation of (2-3H) myoinositol into inositol bis-phosphate and inositol tris-phosphate in vita- minA deficient hamster intestinal mucosa. Fed Proc [Abstr # 5081] 46:1189

Frasca JM, Auerbach O, Parks VR, Jamieson JD (1968) Elec- tron microscopic observations of the bronchial epithelium of dogs. Exp Mol Pathol 9:363-379

Goodman DS (1984) Vitamin A and retinoids in health and disease. New Engl J Med 310:1023-1031

Goldman WE, Baseman JB (1980) Glycoprotein secretion by cultured hamster trachea epithelial cells: A model system for in vitro studies of mucus synthesis. In Vitro 16:320- 329

Goshima K, Masuda A, Owaribe K (1984) Insulin-induced formation of ruffling membranes of K B cells and its correlation with enhancement of amino acid transport. J Cell Biol 98 : 801-809

Gospodarowicz D, Greenberg G, Birdwell CR (1978) Determi- nation of cellular shape by the extraceUular matrix and its correlation with the control of cellular growth. Cancer Res 38:4155-4171

Groelke JW, Coalson J J, Baseman JB (1985) Growth require- ments of ferret tracheal epithelial cells in primary culture. Proc Soc Exp Biol Med 179:309-317

Harris CC, Sporn MB, Kaufman DG, Smith JM, Jackson FE, Safflotti U (1972) Histogenesis of squamous metaplasia in the hamster tracheal epithelium caused by vitamin A deficiency or benzo[a]pyrene-ferric oxide. J Natl Cancer Inst 48 : 743-761

Hay ED (1981) Collagen and embryonic development. In: Hay ED (ed) Cell biology of the extracellular matrix. Plenum Press, New York, pp 379-409

Heath J, Bell S, Rees AR (1981) Appearance of functional insulin receptors during the differentiation of embryonal carcinoma cells. J Cell Biol 91:293-297

Huang FL, Roop DR, De Luca LM (1986) Vitamin A deficiency and keratin biosynthesis in cultured hamster trachea. In Vitro 22 : 223-230

Jetten AM (1980) Retinoids specifically enhance the number of epidermal growth factor receptors. Nature 284:626-629

Jetten AM (1984) Modulation of cell growth by retinoids and their possible mechanisms of action. Fed Proc 43:134-139

Jetten AM, Smits H (1985) Regulation of differentiation of tracheal epithelial cells by retinoids. In: Retinoids, differentiation and disease. Ciba Fdn Symposium 113 Pitman, Lon- don, pp 61-76

Kadowaki T, Koyasu S, Nishida E, Sakai H, Takaku F, Yahara I, Kasuga M (1986) Insulin-like growth factors, insulin, and epidermal growth factor cause rapid cytoskeletal reorgani- zation in KB cells. J Biol Chem 261:16141-16147

Kamei H (1983) Effect of retinoic acid on cell-cell adhesiveness in cloned BHK 21/C 13 cells which form piling-up colonies. Exp Cell Res 148:11-20

Kawanami O, Ferrans V J, Crystal RG (1979) Anchoring fibrils of the normal canine respiratory system. Am Rev Respir Dis 120:595-611

Keenan KP (1987) Cell injury and repair of tracheobronchial epithelium. In : McDowell EM (ed) Lung carcinomas. Chur- chill Livingstone, Edinburgh, pp 74-93

Klann RC, Marchok AC (1982) Effects of retinoic acid on cell proliferation and cell differentiation in a rat tracheal epithelial cell line. Cell Tissue Kinet 15:473-482

Kleinman HK, Klebe RJ, Martin GR (1981) Role of collage- nous matrices in the adhesion and growth of cells. J Cell Biol 88:473-485

Lane BP, Gordon RE (1979) Regeneration of vitamin A deficient rat tracheal epithelium after mild mechanical injury. Differentiation 14: 87-93

Lechner JF, Haugen AA, McClendon IA, Pettis EW (1982) Clonal growth of normal adult human bronchial epithelial cells in a serum-free medium. In Vitro 18:633-642

Lotan R (1980) Effects of vitamin A and its analogs (retinoids) on normal and neoplastic cells. Biochim Biophys Acta 605 : 33-91

McDowell EM, Trump BF (1976) Histologic fixatives suitable for diagnostic light and electron microscopy. Arch Pathol 100:405-414

McDowell EM, Combs JW, Newkirk C (1983) A quantitative light and electron microscopic study of hamster tracheal epithelium with special attention to so-called intermediate cells. Exp Lung Res 4:205-226

McDowell EM, Hess FG, Trump BF (1980) Epidermoid metaplasia, carcinoma in situ, and carcinomas of the lung. In: Trump BF, Jones RT (eds) Diagnostic electron microscopy, vol 3. John Wiley and Sons, New York, pp 37-96

McDowell EM, Keenan KP, Huang M (1984a) Effects of vitamin A-deprivation on hamster tracheal epithelium. A quantitative morphologic study. Virchows Arch [Cell Pathol] 45:197-219

McDowell EM, Keenan KP, Huang M (1984b) Restoration of mucociliary tracheal epithelium following deprivation of vitamin A. A quantitative morphologic study. Virchows Arch [Cell Pathol] 45:221-240

McDowell EM, Newkirk C, Coleman B (1985) Development of hamster tracheal epithelium. II. Cell proliferation in the fetus. Anat Rec 213:448-456

Montesano R (1986) Cell-extracellular matrix interactions in morphogenesis: an in vitro approach. Experentia 42 : 977-985

Niles RM, Kim KC, Loewy BP (1986) Retinoic acid alters cell configuration, increases growth, and stimulates mucin release in cultured tracheal epithelial cells. J Cell Biol [Abstr #e 744] 103 : 201 a

Port CD, Baxter DW, Harris CC (1974) Surface morphology of tracheal epithelium in vitamin A deficiency and reversal.


In: Karbe E, Park JF (eds) Experimental lung cancer. Car- cinogenesis and bioassays. Springer Berlin, Heidelberg, New York, pp 257-264

Schroder EW, Rapaport E, Black PH (1983) Retinoids and cell proliferation. Cancer Surv 2:223-241

Shapiro SS (1986) Retinoids and epithelial differentiation. In: Sherman MI, (ed) Retinoids and cell differentiation. CRC Press, Boca Raton, pp 29 59

Shapiro SS, Poon JP (1979) Retinoic acid induced alterations of growth and morphology in an established epithelial cell line. Exp Cell Res 119:349-357

Sherman BS (1961) The effect of vitamin A on epithelial mitosis in vitro and in vivo. J Invest Dermatol 37:469-480

Sporn MB, Roberts AB (1983) Role of retinoids in differentiation and carcinogenesis. Cancer Res 43 : 3034-3040

Strum JM, Latham PS, Schmidt ML, McDowell EM (1985) Vitamin A deprivation in hamsters. Correlations between tracheal epithelial morphology and serum/tissue levels of vitamin A. Virchows Arch [Cell Pathol] 50:43-57

Tammi R, Tammi M (1986) Influence of retinoic acid on the ultrastructure and hyaluronic acid synthesis of adult human epidermis in whole skin organ culture. J Cell Physiol 126: 389-398

Tandler B, Sherman J, Boat TF (1981) EDTA-mediated separation of cat tracheal lining epithelium. Am Rev Respir Dis 124: 469-475

Taylor A, Hogan BLM, Watt FM (1985) Biosynthesis of EGF receptor, transferrin receptor and colligin by cultured human keratinocytes and the effect of retinoic acid. Exp Cell Res 159: 47-54

Van Scott MR, Yankaskas JR, Boucher RC (1986) Culture of airway epithelial cells: Research techniques. Minireview. Exp Lung Res 11 : 75-94

Walker-Nasir E, Codington JF, Lampert LA, Jeanloz RW

(1985) Effects of retinoic acid on ascites cells of the TA3 mouse mammary carcinoma. Experentia 41:402-404

Watt FM (1986) The extracellular matrix and cell shape. TIBS 11:482-485

Wolbach SB, Howe PR (1925) Tissue changes following deprivation of fat-soluble A vitamin. J Exp Med 42 : 753-777

Wolbach SB, Howe PR (1928) Vitamin A deficiency in the guinea-pig. Arch Pathol 5:239-253

Wolf G (1984) Multiple functions of vitamin A. Physiol Rev 64:873-937

Wong YC, Buck RC (1971) An electron microscopic study of metaplasia of the rat tracheal epithelium in vitamin A deficiency. Lab Invest 24: 55-66

Wilhelm DL (1954) Regeneration of the tracheal epithelium in the vitamin A-deficient rat. J Pathol Bact 67:361-365

Wu R (1986) In vitro differentiation of airway epithelial cells. In : Schiff LJ (ed) In vitro models of respiratory epithelium. CRC Press, Boca Raton, pp 1-26

Wu R, Nolan E, Turner C (1985) Expression of tracheal differentiated functions in serum-free hormone-supplemented medium. J Cell Physiol 125 : 167-181

Wu R, Groelke JW, Chang LY, Porter ME, Smith D, Nettes- heim P (1982) Effects of hormones on the multiplication and differentiation of tracheal epithelial cells in culture. In: Sirbasku D, Sato GH, Pardee A (eds) Growth of cells in hormonally defined media. Cold Spring Harbor Laborato- ry, Cold Spring Harbor, New York, pp 641-656

Zile MH, Cullum ME (1983) The function of vitamin A: Cur- rent concepts. Proc Soc Exp Biol Med 172:13%152

Zile MH, Bunge EC, De Luca HF (1981) DNA labeling of rat epithelial tissues in vitamin A deficiency. J Nutr 111 : 777-788

Received April 20 / Accepted June 15, 1987

Effects of retinoic acid on the growth and morphology of hamster tracheal epithelial cells in...

Documents

Transcript of Effects of retinoic acid on the growth and morphology of hamster tracheal epithelial cells in...