ABSTRACT
Serially propagated cells derived from the steer thyroid gland preserve several specialized characteristics, some demonstrable for as long as 8 months (over 15 passes). For about 5 passes (3 months), follicular-like arrays of cells develop. Thyroid-stimulating hormone (TSH) or thyrotropin (1 and 10 mu./ml) induces the formation of supernumerary nucleoli, and promotes at the ultrastructural level the prompt (within 10 min) appearance of microvilli, pseudopodia, and intracytoplasmic droplets. These cells have a mean plating efficiency (PE) of 16·5 ±4·5 % (standard deviation) and with 3 × 104 cells as the standard inoculum, a mean doubling time (Td) of 43·7± 1·2 h. The linear variation of Td with the number of cells seeded probably signifies strong cell-cell interactions. TSH (1–50 mu./ml) influences both PE and Td, with 1 mu./ml producing the maximum stimulation. TSH (0·001–100 mu./ml) also induces the discharge of incorporated radio-iodide from the cultured thyroid cells, achieving a peak by 2–4 h, titres of 0·1–1 mu./ml being most potent. Long-acting thyroid stimulator (LATS) (0·2 u./ml) likewise effects release of 131I into growth medium (1 experiment), but over a more protracted period. After labelling with [3H]leucine, radioautographs demonstrate that the synthesis of proteins is stimulated by exposure to TSH (1 mu./ml). Precipitation with trichloro acetic acid indicates that these cells persist in synthesizing 131I-tagged iodoproteins, an activity optimally stimulated by 1 mu./ml TSH and marked by 3 h. Hence, such thyroidal cell lines afford a useful model for studying differentiation and hormonal effects.
INTRODUCTION
Cell lines derived from the thyroid gland of the steer have been serially propagated as monolayers for several months with the retention of several specialized morpho logical and functional properties. While descriptions of cultures of the thyroid gland have been given by Pulvertaft, Davies, Weiss & Wilkinson (1959); Hagmüller & Leslie (1962); Hung, Winship, Bowen & Houck (1966, 1967, 1968); Flanagan, Barron, Beutner & Witebsky (1966) and of thyroid cells by Hilfer (1962); Tong (1964); Hung & Winship (1964); Kerkof, Long & Chaikoff (1964); Kerkof, Raghupathy & Chaikoff (1964); Mallette & Anthony (1966); Kalderon & Wittner (1967); Spooner (1970), this study is concerned with the manifestations of differentiation by dispersed primary cells and their descendants. In addition to characterizing the cell lines, some of the responses elicited by thyrotropin (TSH) and the long-acting thyroid stimulator (LATS) will be given.
METHODS AND PROCEDURES
Establishing thyroid cell lines
To initiate the culture, steer thyroid glands were obtained from a slaughterhouse and tran sported promptly in the cold to the laboratory. After trimming and mincing, the glandular frag ments were rinsed several times with Hanks’s balanced salt solution (BSS) containing mycostatin (200 u./ml), and digested with collagenase (1 mg/ml in Hanks’s BSS) at 37 °C in trypsinization flasks (Rappaport, 1956) for approximately 2 h. The dispersed cells were grown as monolayers in Eagle’s minimum essential medium supplemented with 10 % foetal bovine serum and penicillin streptomycin (50 u./ml) in water-saturated incubators kept at 37 °C, gassed with a 5 % CO2/95 % air mixture. Thereafter, subcultures were ‘passed’ about every 14 days, utilizing trypsin (0·05 %) digestion and gentle agitation for the release of the cells. For the stock cultures as well as the experimental populations, growth medium and hormone (if present) were renewed almost weekly. Our observations were made on lines derived from 2 explants surviving 4 months (10 passes) and 8 months (15 passes), respectively (Siegel & Lautenschlager, 1969; Siegel, Siegel & Lautenschlager, 1969).
Characterizing dimensions and morphology
To estimate cellular dimensions with an ocular micrometer, cells were overgrown on to microscope slides, fixed with Hanks’s BSS-95 % ethanol (1:1) and absolute alcohol, and stained with haematoxylin.
For the preparation of electron micrographs, approximately 3 × 106 isolated log-phase cells from the fourth pass were transferred to a 100-mm diameter Petri dish and incubated for several days. Adapting the method given by Brinkley, Murphy & Richardson (1967), the result ing monolayer cultures were fixed in situ with 3 % glutaraldehyde in Hanks’s BSS followed by 1 % osmium tetroxide, pre-stained in uranyl acetate, dehydrated, and embedded in Epon.
Determining doubling time and plating efficiency
To express cellular proliferation, the mean doubling time (Td) was estimated graphically from the temporal variation of multiplicity (average number of cells per colony), a parameter introduced by Puck, Marcus & Cieciura (1956). As another index of growth and aggregation, determinations were made of plating efficiency (PE), the number of colonies that develop relative to the number of single log-phase cells plated. Procedures given earlier (Siegel & Tobias, 1966a) were followed for obtaining Td and PE.
Assaying discharge of 131/into medium
Determinations were made with the cultured thyroid cells analogous to the bioassay procedures that depend upon the release of localized 131I induced by TSH (Nagataki, Shizume, Matsuda & Ishii, 1959; Rosenberg, Athans, Ahn & Behar, 1961) and by LATS (McKenzie, 1958). For these studies about 3–5 × 106 cells were exposed to Na 131I (10 μCi/ml) for 1–2 h. Following 3 rinses with Hanks’s BSS, 0·001–100 mu./ml of bovine TSH (Armour’s Thytropar) was introduced into the medium bathing the cells. The radioactivity of aliquots at intervals thereafter was assessed with a well-type scintillation counter. Duplicate or triplicate samples were prepared from replicate cultures; at least 4000 counts were collected for each sample.
Assessing protein synthesis
To study polypeptide formation, radioautographs were prepared (Siegel & Tobias, 1966a) after the cells were labelled for 1 h with [3H]leucine (specific activity 2 Ci/ml), diluted to 1 μCi/ml of medium, and then immersed in TSH (1 mu./ml) for varying times.
The release of tagged iodoproteins into the medium was followed by assaying the fraction of 131I precipitated by trichloroacetic acid (TCA). For these determinations, the cells were incubated overnight with Na 131I (1 μCi/ml), rinsed thrice with Hanks’s BSS, and treated with TSH (1–50 mu./ml) for intervals up to 7 days. Cold 10% TCA was added to aliquots of the medium and the precipitate was separated by centrifugation and decanting. The resulting pellet was twice resuspended in cold 1 % TCA and centrifuged, dissolved in 0·1 N NaOH, and counted.
RESULTS
Cytology and ultrastructure
Soon after seeding, the isolated cells proliferate and aggregate to form concentric rings. About the third week, centres of densely packed overgrown cells, or ‘knots’, comprise a portion of many of these follicular-like arrays. Highly structured sheets of cells develop over a month or two, permitting the identification of networks of knots, lacunae, and connecting membranes (Figs. 4–6). Following the fifth pass (approximately 3 months after explanting) such highly organized patterns are less frequently encountered. Thyrotropin in concentrations of 0·1–10 mu./ml produces no discernible modification in gross morphology of the explant and subsequent passes.
As the cells replicate, they acquire many pseudopodia possessing dendritic processes which interdigitate with those of adjoining cells (Fig. 7). The cytoplasm of these cells appears stringy and foamy, containing many vacuoles that are especially prominent in the vicinity of the nucleus (Fig. 8). As multicellular colonies develop, the cells become smaller and rounder with fewer and shorter pseudopodia.
Measurements were made of the cultured thyroid cells with a calibrated ocular reticle. Assuming an ovoid shape, and examining 250 cells from the first 3 passes, the cells’ major and minor axes are 83·3 ± 27·6 /tm (standard deviation) and 28·0 ± 10·7 μm, respectively. For the same preparations, the corresponding nuclear axes are 19·4 + 4·0 μm and 12·2 ± 2·0 μm.
The number of nucleoli per nucleus was counted to ascertain the influence of successive subculturing and of treatment with TSH. For this cell line, 3 nucleoli per nucleus is the modal number. The frequency distribution of nucleoli does not vary significantly with the number of passes, at least for the first 7 subcultures. Exposing the cells from the fourth pass to TSH (1 and 10 mu./ml) for 8 days, however, produces a marked increase in nuclei containing supernumerary nucleoli (Fig. 1).
Effect of TSH on the frequency distribution of nucleoli in the thyroid cell line. Cells of the fourth pass were exposed to TSH for 8 days. Each point represents the mean frequency (per cent) of 1000 cells possessing the indicated number of nucleoli. ○, Controls; ◐, 1 mu./ml TSH; •, 10 mu./ml TSH.
Effect of TSH on the frequency distribution of nucleoli in the thyroid cell line. Cells of the fourth pass were exposed to TSH for 8 days. Each point represents the mean frequency (per cent) of 1000 cells possessing the indicated number of nucleoli. ○, Controls; ◐, 1 mu./ml TSH; •, 10 mu./ml TSH.
At the ultrastructural level, the cultured cells appear and behave like their in vivo counterparts. Among the prominent intracytoplasmic features appearing on electron micrographs of these cultured thyroid cells are a highly developed, granular endo plasmic reticulum, many sinuous mitochondria, large vacuoles or droplets, and smaller electron-dense bodies (Fig. 9) – organelles very similar to those found on examining the ultrastructural anatomy of the intact gland (Wissig, 1964; Klinck, Oertel & Winship, 1970). Following a 10-min exposure to TSH (1 and 10 mu./ml), there is a marked increase in the number and size of the intracytoplasmic droplets (Fig. 10), a response long known to occur in vivo (De Robertis, 1942; Nadler, Sarkar & Leblond, 1962; Wollman & Spicer, 1963). Another reaction to thyrotropin is the occurrence of more and longer projections at the cell margins. In some fields of stimulated cells, pseudopodia are nearly or in contact with one another, consistent with the supposition that in situ these droplets, presumably engulfed thyroglobulin (Wetzel, Spicer & Wollman, 1965), form by endocytosis (de Duve & Wattiaux, 1966).
Proliferation and colony formation
For the cultured thyroid cells, Td varies linearly with the number seeded (Fig. 2), over the range 1·5–6·0 × 104 cells. Using 3 × 104 cells as the standard inoculum, the mean Td of 3 independent determinations is 43·7±1·2 h. Several experiments in dicate that the proliferation of these cells (from the fourth to sixth passes) is enhanced by the presence of thyrotropin in the growth plates (Fig. 3). Thus, the addition of 10 mu./ml TSH produces a relative reductionof 27 % in Td (mean of 4 determinations); the corresponding diminution effected by 1 mu./ml is 32% (2 determinations).
Variation of mean doubling time (Td) with the number of log-phase single cells seeded. Points plotted are derived from the slopes of linear portions of growth curves (similar to Fig. 3) from cells of the fifth pass. At least 100 colony-forming units were counted daily for 5 or more days to establish the growth curves.
Variation of mean doubling time (Td) with the number of log-phase single cells seeded. Points plotted are derived from the slopes of linear portions of growth curves (similar to Fig. 3) from cells of the fifth pass. At least 100 colony-forming units were counted daily for 5 or more days to establish the growth curves.
Effects of TSH (10 mu./ml) on the proliferation of the thyroid cell line (fourth pass). At zero time, 3 × 104 single cells in the log phase with a multiplicity of 1·00 were deposited into 100-mm Petri dishes. Points plotted represent the mean multiplicity of at least 100 colony-forming units, counted under phase-contrast microscopy. ○, Control (Td = 42 h); •, TSH (Td = 33 h).
Effects of TSH (10 mu./ml) on the proliferation of the thyroid cell line (fourth pass). At zero time, 3 × 104 single cells in the log phase with a multiplicity of 1·00 were deposited into 100-mm Petri dishes. Points plotted represent the mean multiplicity of at least 100 colony-forming units, counted under phase-contrast microscopy. ○, Control (Td = 42 h); •, TSH (Td = 33 h).
Photomicrograph of cultured thyroid cells overgrown on to microscope slides, 19 days subsequent to first pass. The reorganizing and proliferating cells assume an ordered pattern, including the formation of a knot, or focus of dense packing. Hae-matoxylin; × 500.
Photomicrograph taken through the bottom of the culture bottle with the inver ted tissue-culture microscope. Cultured thyroid cells, 10 days subsequent to second pass. Note the follicular-like arrays, × 120.
Similar to Fig. 5; 19 days after the third pass. Another example of a follicular pattern assumed by the cultured thyroid cells; 3 opaque knots are near the ‘follicles’. × 30.
Photomicrograph of cultured thyroid cells overgrown on to microscope slides, 10 days after the seventh pass. Field chosen to illustrate the many, long pseudopodia with dendritic processes. Haematoxylin; × 300.
Similar to Fig. 7. Note the foamy, stringy cytoplasm containing many vacuoles, particularly in the vicinity of the nucleus. × 500.
Note the granular character of the endoplasmic reticulum, the presence of sinuous mitochondria, vacuoles or droplets, and smaller electron-dense bodies.
Treated with TSH (10 mu./ml) for 10 min. Numerous large droplets are present in the cytoplasm.
The mean PE for this cell line is 16·5 ±4·5%, based on 8 determinations (3–5 dishes each). The addition of 1 mu./ml TSH elevates PE from 19·0 to 24·3 % (Student’s ‘t’ test, significant at the level of P < 0·005); however, 50 mu./ml TSH reduces PE to 8·0% (P < 0·001).
Release of 131I by TSH and LATS
The discharge of radio-iodine from these cultured cells evoked by TSH is similar to that found in other systems (Nagataki et al. 1959; Rosenberg et al. 1961). Thus, the peak 131I concentration of the culture medium occurs about 2–4 h following the introduction of thyrotropin; the data from a representative experiment are given in Table 1. The causes for the irregularities in these counting data are being sought and may relate to the adsorption of 131I by glass and plastic surfaces. After 4 h exposure to TSH (1 mu./ml), the 131I titre of the medium bathing treated cells is 1·3–4·2 times greater than the control levels; this response appears unaffected by the number of passes, at least through the ninth (Table 2). The concentrations found most effective in inducing the 131I discharge range between 0·01–1 mu./ml TSH. On the other hand, higher titres of TSH (10–100 mu./ml) inhibit the release of 131I.
Release of131/to the medium by cultured thyroid cells (fourth pass) following exposure to TSH (1 mu./ml)
Relative release of131/by steer thyroid cell line after 4 k of treatment with TSH (7 mu./ml) for the first 9 subcultures
LATS remains a substance shrouded in mystery, distinguished functionally from TSH by its protracted action. Utilizing thyroid cells of the fifth pass, the release of 131I caused by LATS was studied. For this experiment (Table 3), the effect of 0·2 u./ml LATS was compared to that produced by 0·1 mu./ml TSH at 4, 8, and 24 h. Like the action of LATS in vivo, 131I concentration in the medium remains elevated at 24 h, when the 131I level for TSH-treated cells is below that for the controls.
Release of131l to the medium by cultured thyroid cells (fifth pass) following exposure to 0·1 mu./ml TSH and 0·2 u./ml LATS
The specificity of the interaction between target cell and its tropic hormone was further affirmed by experiments which indicate that TSH (0·1 mu./ml) does not influence the release of 131I from T-1 human kidney cells (first cultured by van der Veen, Bots & Mes, 1958), which are highly sensitive to thyroxine and triiodothyronine (Siegel & Tobias, 1966a, b).
Synthetic activities and TSH
The enhancement of protein synthesis with TSH is indicated by radioautographs prepared after [3H]leucine labelling of the cultured thyroid cells (seventh pass) for 1 h. These reveal an increase of 27·8% in tritium uptake following treatment with TSH (1 mu./ml) for 1 h, relative to the controls, by scoring the developed grains over 200 cells of each population.
That this cell line through the tenth pass synthesizes iodoproteins is demonstrated by counting the TCA-precipitable fraction of the culture medium following Na131I uptake; the protein-bound 1311 levels are always significantly greater than background. Furthermore, TSH (1–50 mu./ml) stimulates the output of iodoproteins, an effect marked by 3 h; 1 mu./ml TSH proves the most active titre.
DISCUSSION
Our morphological observations support the conclusion that the thyroid cell retains the potential for directing glandular development (Pulvertaft et al. 1959; Hilfer, 1962; Tong, 1964; Kerkof, Long & Chaikoff, 1964; Kerkof, Raghupathy & Chaikoff, 1964; Spooner, 1970). Indeed, the characterization of our cultured cells allows extending this inference: the formation of the functional follicular arrays originates from information that is evidently transmissible from the explanted differentiated cell to its descendants over many generations. Expression of glandular development remains latent, being triggered perhaps by the presence of a critical number of cells or by the synthesis of a requisite titre of some intercellular principle.
The evidence relating TSH to the rate of formation or to the number of thyroid follicles remains confusing. As pointed out by Boyd (1964), the administration of TSH to chick embryos promotes the appearance of follicles, but absence or profound reduction of the hormone does not impede the development of normal thyroid gland anatomy in full-term anencephalic monsters. No effects of TSH were discerned on the gross morphology of both the dissociated chick embryo thyroid cells (Hilfer, 1962) and our steer thyroid cells; however, Kerhof, Long & ChaikofT (1964) found that only after 2-3 days of hormonal treatment would isolated sheep thyroid cells rearrange themselves into follicular arrays. If our cell line requires TSHto develop follicular-like arrays, then apparently adequate amounts are supplied by the foetal bovine serum, a constituent of the growth medium.
Cellular interactions and adhesiveness exhibited by the cultured thyroid cells are evidently promoted by thyrotropin. Thus, the presence of highly developed pseudo podia possessing complex arboreal processes that interdigitate with their neighbouring counterparts, and the strong dependence of Td on the number of cells seeded are manifestations of appreciable intercellular involvement. The influence exerted by TSH on Td and PE of these cultured cells also reflects enhancement of surface activity essential for the evolution of stable, specialized structures. Apparently, the enzymic treatment given the thyroid explant and its progeny does not impair the adhesiveness of the cell surface requisite for attachment to the substratum and cell–cell association (Mallette & Anthony, 1966).
Little being known about the role of the nucleolus in the secretion of thyroid hormones, it is difficult to interpret the appearance of supernumerary nucleoli following treatment of the cultured thyroid cells with TSH. This response may be linked to RNA metabolism in view of the involvement of the nucleolus with ribosomal RNA (Perry, 1964). It is noteworthy that thyroid hormones likewise effect an increase in nucleoli of rat liver (Stenram, 1957) and T-i human kidney cells (Siegel & Tobias, 1966 a).
The preliminary data presented here disclose that thyroid cells persist in producing proteins, whose formation can be influenced by TSH over many subcultures. Cyto chemical determinations (E. Siegel & J. Lammert, in preparation) reveal that the elaboration of lipids, glycoproteins, mucosubstances and acid phosphatases by the thyroid cell line also continues to be affected by TSH. Recently, Spooner (1970) has reported that for at least 32 days, approximately 17 generations, primary clonal cultures of embryonic chick thyroid synthesize the thyroidal hormones.
Thyroid cell lines afford a host of research opportunities, especially for pursuing relatively long-term (2 months or 30 generations, at least) studies. As examples, this in vitro approach should prove fruitful in investigating the molecular determinants of differentiation (Davidson, 1965), the sub-cellular fate of TSH (Mato vino vie & Vickery, 1959; Begg & Munro, 1965), and the relationship of TSH to cyclic-AMP (Pastan & Katzen, 1967; Maayan & Ingbar, 1968; Kaneko, Zor & Field, 1969). Refinements of the 131I-release measurements should yield sensitive bioassays for TSH and LATS. Enhanced understanding of the elaboration of the iodinated thyronines by cultured thyroid cells should emerge from the application at the ultrastructural level of radio autographic (Nadler, 1965) and cytochemical (Flanagan et al. 1966; Hilfer, 1962; Mallette & Anthony, 1966; Kalderon & Wittner, 1967; Wetzel et al. 1965) methods.
ACKNOWLEDGEMENTS
This investigation was supported by U.S. Public Health Grant No. AM-11881 from the National Institute of Arthritis and Metabolic Diseases. I thank Dr J. H. Frenster for his generosity in making available the facilities of his electron-microscope laboratory and Dr J. P. Kriss for the gift of the LATS sample. I am indebted to Mrs Elsie P. Siegel, M.S. for the preparation of the electron micrographs and to Miss Holde H. Lautenschlager and Miss Jeanne Lammert, M.S. for assistance in the laboratory.
