Density-dependent Inhibition Of Growth Research Articles

Density-dependent stimulation and inhibition of cell growth by agents that disrupt microtubules.

We have previously reported that agents that disrupt microtubules, such as colchicine, inhibit the growth stimulation of lymphocytes and arrested fibroblasts; other workers have recently reported enhanced stimulation of fibroblasts in the presence of the same drugs. In the present studies, we resolve this conflict by demonstrating that the stimulatory and inhibitory effects of microtubule disruption depend upon the density and the cell type of the treated cultures. Our analysis included an examination of three variables: (i) cell density (sparse or confluent), (ii) cell type (resting fibroblasts from mouse or chicken embryos or from the permanent 3T3 mouse fibroblast line), and (iii) treatment with colchicine and related drugs in the presence or absence of various mitogens such as serum, insulin, and epidermal growth factor. We found that colchicine augmented mitogenesis in confluent cultures of all cell types. Colchicine by itself appeared to be mitogenic only for confluent chicken embryo fibroblasts. In sparse cultures with minimal cell-cell contacts, however, there were differences between embryonic cells and the 3T3 cell line. In confirmation of our previous reports, disruption of microtubules by colchicine inhibited the mitogenic stimulation of sparse cultures of embryonic chicken or mouse fibroblasts. In contrast, fibroblasts of the permanent 3T3 line in sparse cultures were stimulated by some mitogens despite the presence of colchicine. The augmentative effects of colchicine on the stimulation of confluent cultures were synergistic with the mitogens, and colchicine allowed response to otherwise submitogenic doses of growth factors. Kinetic studies indicated that the stimulatory and inhibitory effects of colchicine are separable and that both can operate simultaneously. The experiments suggest that the regulation of growth by nutrient deprivation and the regulation by density dependence proceed at least in part by different mechanisms. All of the results suggest that microtubular integrity can be associated with the expression of either negative or positive controls on cell growth, depending upon the confluence or lineage of the cells in culture.

Role of serum components in density-dependent inhibition of growth of cells in culture. Platelet-derived growth factor is the major serum determinant of saturation density.

The effects of platelet-derived growth factor and plasma components on saturation density in cultures of 3T3 cells were investigated. Both of these components of whole blood serum affect saturation density; however, when 3T3 cells become quiescent at high density in medium containing whole blood serum, only platelet-derived growth factor and fresh whole blood serum are capable of stimulating proliferation. Addition of fresh plasma- derived serum has little effect on cell growth. These results suggest that the platelet factor is the major determinant of saturation density in cultures of 3T3 cells maintained in medium supplemented with whole blood serum. Experiments were performed to investigate the mechanism by which platelet-derived growth factor regulates saturation density. We investigated the possibilities of inactivation of growth factors by proliferating cells, and the effects of cell density on the response of 3T3 cells to platelet-derived growth factor. The amount of platelet- derived growth factor required to initiated DNA synthesis increases with increasing cell density. Some inactivation of growth factors by growing cells was detected, but this depletion was only evident at high cell density. We propose that density-dependent inhibition in cultured 3T3 cells is the result both of an increased requirement for the platelet- derived growth factor as the cultures become more crowded and of inactivation of growth factor activity by growing cells.

Further characterization of a thermosensitive transformation variant of mouse fibroblasts.

Temperature-sensitive (ts) variants that express phenotype at low (33 degrees) but not at high (38.5 degrees) temperature were isolated from a mouse fibroblast strain C3H2K cells. Among these variants, cloned ts-12B cells showed at 38.5 degrees a density-dependent inhibition of growth typical of normal fibroblasts cultured in vitro. However, at low temperature they lost this capacity and grew to a higher saturation density. The ts variant was also temperature-sensitive as regards serum requirement: it required a higher concentration of serum for growth at 38.5 degrees than at 33 degrees. However, the cells behaved at both temperatures like the parent strain, possessing anchorage-dependence for growth and fibronectin, whereas they were like transformed cells with respect to release of high fibrinolytic activity. No type-C virus core protein p30 was detected at either temperature. Thus, various parameters of transformation in vitro were independently regulated in these variant cells.

Expression of the transformed phenotype and tumorigenicity in somatic cell hybrids.

We have previously shown by examining the anchorage dependence, density-dependent inhibition of growth, LETS protein and microfilament bundles that the transformed phenotype of the parental tumours are suppressed in hybrids between rat embryo fibroblast (REF) and mouse tumour cells (TA3B). Hybrids between TA3B and Syrian hamster sarcoma cells (BHK-B1) also show suppression. We now demonstrate that tumorigenicity in nude mice is also suppressed in TA3B X REF and B1 X TA3B hybrids. Tumours arise from the suppressed hybrids by the selective outgrowth of variants with properties different from the majority of cells inoculated. These tumour variants always had an altered cytoskeletin but 1 out of 14 cases retained anchorage-dependent growth. Selection in culture for anchorage-independent growth selected for tumorigenicity. The coordinate suppression of the transformed phenotype and tumorigenicity could be explained by postulating a pleiotropic control mechanism affecting both. However, the occasional dissociation of tumorigenicity from anchorage dependence in the variants suggests that the targets for the control mechanism must be different.

Density-Dependent Growth Inhibiting Interactions between 3T3 and SV40-3T3 Cells in Mixed Cultures

3T3 cells are shown to reduce SV40-3T3 cell population growth in a denisty-dependent manner in co-cultures of 3T3 and SV40-3T3 cells. The development of this inhibitory activity roughly parallels the development of density-dependent inhibition of growth in homogeneous 3T3 control cultures. The extent of reduction of SV40-3T3 growth can be manipulated by pretreatment of 3T3 cells with a high serum concentration. SV40-3T3 growth rates are reduced by factors between 10 and 20 under optimum inhibitory conditions as compared to SV40-3T3 growth in control cultures.

Adenovirus type 12 gene 401 function and maintenance of transformation

Rat (3Y1) and hamster embryo brain cells were transformed by wild-type adenovirus type 12 or the DNA-minus temperature-sensitive mutant ts401. The ts401-transformed 3Y1 cells, but not the wild-type transformants, displayed a temperature-sensitive response with respect to the following characteristics of the transformed phenotype: morphology, saturation density, growth rate, cloning in soft agar, colony formation on plastic at low cell densities in 1% serum medium, and the T antigen(s). Temperature shift-down experiments showed that the density-dependent inhibition of growth of the ts401-transformed cells was reversible, as was, to some extent, the low efficiency of colony formation at low cell densities in 1% serum. Examination of hamster transformants for their ability to clone in soft agar at permissive and nonpermissive temperatures showed that this property was temperature dependent, again only in the ts401 transformants and not in the wild-type transformants. Alteration in uptake of 2-deoxyglucose or in intracellular cyclic AMP content was not a characteristic of the adenovirus-transformed phenotype in the 3Y1 cells. The findings suggest that an active 401 function is required for maintenance of the adenovirus-transformed cell pheno-type.

Differential effects of hydrocortisone on both growth and collagen metabolism of human fibroblasts from normal and keloid tissue

Cultured fibroblasts isolated from normal and keloid tissue do not differ in their growth characteristics or in the rate of collagen synthesis under routine culture conditions. The addition of hydrocortisone to the culture media results in significant differences in both growth and collagen synthesis between these cell types. Collagen synthesis is inhibited 60% in normal cultures by hydrocortisone (0,5 micrograms/ml) and the population size at which density-dependent growth inhibition is achieved is increased. Keloid-derived fibroblasts grow to a lower maximum density in the presence of hydrocortisone, while their rate of collagen synthesis is not significantly reduced. The rate of non-collagen protein synthesis is increased significantly by hydrocortisone in both cell types. Comparison of normal and keloid-derived cultures obtained from a single individual suggests that the keloid phenotype with respect to both growth and collagen synthesis is restricted to the fibroblasts isolated from the keloid nodule.

Colony Formation by Simian Virus 40-Transformed Human Parapharyngeal Cells Cultured in Semisolid Agar<xref ref-type="fn" rid="FN1">2</xref>

A simian virus 40 (SV40)-transformed line of human parapharyngeal cells (SV-TGo) was cultured in semisolid agar to determine its ability to grow in the absence of an anchoring substratum and to evaluate any phenotypic changes that might have resulted during the isolation of sublines specifically selected for anchorage independence. After 2--3 weeks and 14--15 population doublings in culture, SV-TGo plated with over 1,000% higher efficiency than negative controls (F2408 cells). Sublines, 0.3--2.0 mm in diameter, were isolated and transferred to Leighton tubes in which they underwent an additional 0--7 divisions before senescence after 39--44 total population doublings. Subline phenotype was identical to the original parental phenotype, including epithelioid morphology, organized pattern of growth, extreme sensitivity to density-dependent inhibition of growth, and continuous production of infectious SV40 as detected by the combined tests of cocultivation and direct isolation. Limited division potential was within the range observed for the parental line. The ability to grow in agar without identifiable phenotypic changes was therefore confirmed for this line of SV40-transformed human epithelioid cells.

Suppression of the transformed phenotype in somatic cell hybrids.

Somatic cell hybrids between mouse mammary tumour cells (TA3B) and diploid rat embryo fibroblasts (REF) or between TA3B and Syrian hamster sarcoma cells (BI) were examined for the in vitro characteristics of transformed cells as soon as possible after cell fusion. Unlike the parental tumour cells as three of four TA3B X REF and five BI X TA3B independent hybrid lines had low colony-forming efficiencies in agar, exhibited density-dependent inhibition of growth and did not form colonies on confluent monolayers of 3T3 cells, demonstrating that the transformed phenotype was suppressed in these hybrids. In addition tests of some of the hybrid lines for tumour production in nude mice showed that this was also suppressed. Suppression was more stable in the TA3B X REF than in the BI X TA3B hybrids, variants of the BI X TA3B hybrids with the properties of transformed cells could be readily isolated by subculturing cells that had grown in agar. Tumour growth selected for hybrids with the characteristics of transformed cells, and derivatives of the hybrids selected to show the transformed phenotype readily produced tumours. These correlations suggest that the transformed phenotype and malignancy may be under the same control in these cells. The phenomenon of suppression may be explained by the hypothesis that neoplastic transformation results from recessive mutations in genes which control the normal phenotype. On this model the finding of suppression in hybrids between two different tumour lines is interpreted as complementation and indicates that the mutations are not the same in all cell lines.

Microfilament bundles, lets protein and growth control in somatic cell hybrids

Hybrid cell lines between normal rat embryo fibroblasts and TA3B mouse tumour cells, or between TA3B and BI hamster sarcoma cells, have been examined for the expression of the cell surface large external-transformation-sensitive (LETS) protein and the organization of microfilament bundles. LETS protein was detected by lactoperoxidase-catalysed radioiodination and microfilament bundles were visualized by indirect immunofluorescence with antibodies directed against actin or myosin. Hybrids which exhibited normal growth control characteristics had high levels of LETS protein and extensive microfilament bundles. Neoplastic transformation appears to be suppressed in these hybrids. Hybrids which had the growth control characteristics typical of transformed cells had reduced or zero levels of LETS protein and few microfilament bundles. These results confirm previous studies on the expression of the transformed phenotype in these hybrids and demonstrate that there is a good correlation between normal growth control and the presence of LETS protein and microfilament bundles. However, the changes in cell surface LETS protein and in the organization of microfilament bundles often appeared to be quantitative reductions rather than all-or-none effects. The magnitude of the alterations in the levels of LETS protein and in the organization of microfilaments appeared to correlate with the range of transformed characteristics exhibited by the cells. One transformed hybrid in particular, selected for growth in agar, had some surface LETS protein, some microfilament bundles and retained density-dependent inhibition of growth.

Effect of imposed serum deprivation on growth of the mouse 3T3 cell. Dissociation from changes in potassium ion transport as measured from [86Rb)rubidium ion uptake.

Decreased serum concentrations that substantially alter the growth of normal 3T3 cells alter neither the active and non-active components of unidirectional (86)Rb(+) influx nor the intracellular K(+) content when compared with cells in exponential growth. Thus the changes in K(+) transport (measured with (86)Rb(+) as an analogue for K(+) movements) that occur on density-dependent growth inhibition of the mouse 3T3 cell are not mimicked by serum deprivation of the cells before density inhibition.

Ruthenium red-positive surface layer, extracellular filamentous material and intercellular junctions in hybrids between tumour and normal cells: abundant gap junctions correlate with density-dependent inhibition of growth.

The cell surface and intercellular junctions of transformed and non-transformed hybrid cells, obtained from fusion of murine mammary adenocarcinoma cells (TA3B) and normal rat embryofibroblasts (REF), were compared with those of the parental cells, using ruthenium red (RR) staining of cells fixed in situ. An RR-positive layer of variable thickness was found on the surface of all cell types. Measurements of the thickness of this layer on the free surface of the cell cultures deomonstrated a significant difference between transformed and non-transformed cells. The thickness distribution of the RR layer was similar on the surface of REF and non-transformed hybrid cells, but there was significant variation among all the transformed cell lines. Extracellular filamentous RR-positive material, usually in direct contact with the cell surface, was present in confluent cultures of REF and non-transformed hybrid cells, but was absent in the transformed cell lines. Gap junctions were few in both parental cell lines and rare or absent in transformed hybrids; a large increase in number and size of gap junctions was found in non-transformed hybrids. Abundance of long gap junctions was correlated with density-dependent inhibition of growth, since non-transformed hybrids grew in single layers whereas both normal and tumour parental cell types, and transformed hybrids grew in multilayers. Tight junctions were frequently encountered in TA3B and transformed hybrid cells but were not seen in REF cells and only occasionally seen in non-transformed hybrids. Intermediate-type junctions occurred in all cells, but desmosomes were found only rarely in TA3B cells and never in the other cell types.

Retinol induces density-dependent growth inhibition and changes in glycolipids and LETS.

CHEMICAL carcinogenesis and viral transformation are prevented or inhibited by vitamin A or its analogues1,2. Vitamin A may also play a part in membrane glycosylation as a lipid intermediate3,4. Recently, effects of retinoic acid on proliferation and density-dependent growth of mouse L cells5 and many other untransformed and transformed cells in vitro have been described6. In view of these observations, we have studied the effect of retinol on cell growth behaviour, glycolipid synthesis, and surface-exposed profile of glycoproteins in hamster fibroblasts NIL, NILpy and mouse (BALB/c) 3T3 cells. We report here that culturing NIL or 3T3 cells in vitamin A-containing medium markedly enhances, whereas a medium with ultraviolet-irradiated serum markedly reduces, contact orientation and cell-density dependent inhibition of cell growth. Associated changes of cell surface membrane GM3 level, ganglioside contact response, and in LETS (‘Gap a’) were observed.

Ultrastructure and biological markers of neoplastic change in adult mouse epithelial cells transformed in vitro.

The ultrastructure and in vitro growth properties of 5 tumorigenic mouse submandibular-gland epithelial cell lines were studied. In all lines, in vitro acinus formation occurred, and well differentiated epithelial cells showing epithelial microvilli and desmosomes and cytoplasmic tonofilaments were present. None of the cells showed specific ultratructural features of the normal differentiated submandibular-gland ducts. All the lines formed colonies in semi-solid agar and on confluent monolayers of BALB/c 3T3 cells, and all lacked density-dependent inhibition of growth, as demonstrated by a random distribution of [3H]TdR labelling throughout growing colonies. These 3 growth properties appear to be reliable in vitro markers for epithelial neoplastic transformation in this system, but colony-forming efficiency in agar is lower than that reported for many transformed mesenchymal cells.

Cell saturation density is not determined by a diffusion-limited process.

THE mouse cell line 3T3 ceases cell division at a confluent saturation density despite regular medium replacement. Also, when a confluent quiescent monolayer of 3T3 is wounded, cells migrate into the wound free of cell contact and initiate DNA synthesis, while cells in the monolayer remain arrested1. As the cells in both areas are in the same bulk medium, this experiment proves that depletion of growth factors in the bulk medium is itself not sufficient to account for density-dependent inhibition of growth. Two hypotheses have been proposed to explain the wounding experiment. The first states that direct cell–cell contact, perhaps involving the interaction of a receptor with its ligand on an adjacent cell, provides a negative signal for growth1. The second hypothesis states that growth ceases at a confluent cell density because of the establishment of a diffusion boundary layer which maintains the supply of one or more growth factors below a critical level2. A diffusion boundary layer results because convective movement of liquids near a solid surface approaches zero. In these conditions, the rate of delivery of molecules to and from the surface is limited by the rate of diffusion through the boundary layer. Increased fluid velocity on confluent cells causes initiation of DNA synthesis2,3, a finding apparently supporting the diffusion barrier hypothesis. Theoretically, increasing the fluid velocity over the cell layer should decrease the thickness of the hypothetical boundary layer and therefore increase the rate of delivery of a diffusion-limited growth-controlling substance4. However, in these experiments it was not possible to rule out an effect of fluid velocity on cell–cell contact. Yet if growth is a diffusion-limited process, then increasing viscosity of the cell culture medium should decrease the saturation density. We report here that increased viscosity has no effect on saturation density or on the rate of cell growth, providing strong evidence against the diffusion barrier hypothesis.

The role of adenosine 3′, 5′‐cyclic monophosphate in the density‐dependent regulation of growth and tyrosinase activity of b‐16 melanoma cells

Incubation of cultured B-16 melanoma cells with 1-methyl-3-isobutyl xanthine (MIX) produced a sustained rise in intracellular adenosine 3',5'-cyclic monophosphate (cAMP) which preceded an increase in the specific activity of tyrosinase (EC 1.10.3.1). Cultures of two clones of melanoma cells, one having a mean population doubling time twice that of the other, showed density-dependent inhibition of growth. The tyrosinase activity of each line increased progressively during logarithmic growth, reaching maximal values shortly after the cultures achieved confluence. Intracellular cAMP levels fell during logarithmic growth, being minimal in confluent cultures. The stimulatory effects of MIX and confluence on tyrosinase activity were additive. Cells plated at high density had a lower tyrosinase activity than cells allowed to achieve a similar density by successive division from sparsely planted cultures although the intracellular cAMP levels of such cultures were not different. We support the observations of other investigators that agents which increase intracellular cAMP concentrations can both inhibit cell division and stimulate tyrosinase activity. There are, however, mechanisms for increasing tyrosinase activity and inhibiting cell division which are expressed as B-16 melanoma cells approach confluence and which are not mediated by an increase in intracellular cAMP concentrations.

Analysis of malignancy in human cells: malignant and transformed phenotypes are under separate genetic control.

Human cell hybrids derived from malignant HeLa and normal fibroblast parental cells expressed many of the transformed properties of the HeLa parent but their tumor-producing capability was suppressed. Hybrids derived from HeLa/HeLa fusions retained both their transformed and malignant phenotypes. Thus, an apparent separation of the control of the transformed versus malignant phenotype is indicated. Furthermore, several transformed properties--including lack of density-dependent inhibition of growth, lectin agglutination, lowered requirement for serum growth factors, and anchorage independence--are expressed coordinately in the nontumorigenic hybrids. This finding suggests that none of these properties by themselves, or in concert, endows a cell with tumorigenic potential.

Quantitation of residual bodies in cultured human glial cells during stationary and logarithmic growth phases

Recent report have stated that lipofuscin granules, similar to those found in post-mitotic cells in vivo, accumulate in phase II glial cells during density dependent inhibition of growth. No stereological studies have been carried out to date to investigate the degree or rate of this accumulation. Parallel cultures of cells were studied during log growth and subsequent density dependent inhibition of growth for a period of up to four months. During this period some cultures were subcultivated and the effects of renewed log growth on cytoplasmic organelles and especially the content of residual bodies were studied. We found that the mean volume percentage of the cytoplasm taken up by residual bodies increased steadily from the point of density inhibition of growth. The 24 hours tritiated thymidine ( 3H-TdR) incorporation fell below 2% 21 days after subcultivation. In log phase the mean volume percentage of cytoplasm taken up by residual bodies was 9.6, seven days after subcultivation. Four weeks later, and two weeks after the almost complete cessation of cell multiplication, this figure had risen to 15.3. Four and six weeks after density dependent growth inhibition, the figures were 29.1 and 32.2 respectively. Longer periods in stationary phase did not give any further significant increase in the residual body volume percentage of the cytoplasm. During this period cell volume increased by a factor of 1.9, nuclear volume decreased slightly and the cytoplasm, excluding the secondary lysosomes of the residual body variety, increased by a factor of approximately 1.4. The average cell accumulation of residual body structures increased by a factor of 7 thus constituting the greater part of the cell volume increase. Following subcultivation the volume of residual body structures decreased rapidly and after only seven days of log growth had been lowered to 22.5 vol.% of the cytoplasm and after a further 14 days to 9.8 which approaches the starting level. During prolonged density inhibition of growth there was a gradual cell loss with a parallel enlargement of the remaining cells which kept the monolayer intact and confluent. These results show that phase II glial cells accumulate residual bodies during density dependent growth inhibition. The fact that the volume percentage decreases rapidly following subcultivation shows that cells can decrease their load of residual bodies rapidly by dilution during cell division. Cell volume increases during prolonged periods of density inhibition of growth are to a considerable degree the result of this residual body accumulation.

Cell cycle regulation by growth factors and nutrients in normal and transformed cells

SV3T3 cells, originally responsive to epidermal growth factor (EGF) and displaying density-dependent inhibition of growth, lose responsiveness to the growth factor after several passages and then proliferate without restriction, but continue to display EGF receptor sites at the cell surface. Proliferation of primary fetal rat hepatocytes is not stimulated by EGF, but cells bind it to an extent comparable to that of responsive 3T3 cells. Therefore presence of EGF receptors does not imply that cells are responsive to the growth factor. The relevance of some growth-factor-induced events for DNA synthesis initiation is dicussed. In various primary and secondary cell cultures, Ca++-levels appear to be involved in controlling cell proliferation. In contrast, in 3T3-4a cells, levels of Ca++ ions are not tightly coupled to DNA synthesis initiation; effects of growth factors are not mediated by extracellular Ca++ ions, but cells have a Ca++-sensitive restriction, point in G1. In various cell types in primary or secondary culture or in 3T3-4a cells, polyamine, levels are not tightly coupled to induction of proliferation. Therefore growth-factor-induced ornithine decarboxylase is not an event essential for DNA synthesis initiation. Normal but not transformed cells have a spermidine/spermine-sensitive restriction point in G1. Although rRNA synthesis appears to be necessary for induction of proliferation, preliminary data obtained by double-beam flow microfluorometry suggest that cellular RNA levels might not affect rate of entry into S phase and, furthermore, that 3T3-4a cells can enter S without accumulating RNA above levels present in quiescent cells. It appears that none of the events induced during the prereplicative phase that have been studied in 3T3 cells are essential for DNA synthesis initiation under normal culture conditions.

Evidence for Long-Range Effects in Density-Dependent Inhibition of Proliferation (DDIP) in 3T3 Cells

Nonhomogeneously seeded monolayer cultures of 3T3 cells were labelled for 6 h with 3H-TdR and the proportion of labelled cells in different parts of each culture system examined in autoradiographs. In each it was found that there is an inverse relationship between the local cell density and the proportion of labelled cells, a finding which is consistent with our previously published results with these cells, and which suggests that short-range inhibitory interactions are involved. Further analysis of the data, which is presented in this paper, shows that a correlation exists between the local labelling index and the over-all cell density in the culture, indicating the existence of long-range interactions between cells. We show that the pattern of density-dependent growth inhibition is explicable in terms of a model which considers the culture medium to be composed of two fluid compartments: a stagnant boundary layer about 12 μ in depth (the pericellular compartment), and the bulk medium (supracellular compartment). It is shown that the properties of this model are consistent with the observations for both long- and short-range cellular interaction in the limitation of proliferation in monolayer cultures.