Secretion of transforming growth factors by primary human tumour cells

We examined the ability of primary human tumour cells to secrete diffusible factors capable of stimulating anchorage independent growth of normal rat kidney fibroblast (NRK) cells. Conditioned media (CM) prepared from cells derived from 31/43 patients with adenocarcinoma of the breast, colon, ovary or lung were found to induce growth of NRK cells in soft agar. The ability of the CM to induce anchorage independent growth was enhanced in 25/35 cases by the presence of epidermal growth factor (EGF). The CM did not compete with EGF for binding to the EGF receptor site. CM from cells derived from nonmalignant effusions also supported the growth of NRK cells in soft agar. There was no significant difference in the ability of the CM derived from malignant or normal cells to support NRK colony growth. The ability of primary human tumour cells to clone in soft agar was compared to the ability of these cells to produce diffusible colony stimulating factors for NRK cells. No correlation was observed between the ability of the primary human tumour cells to clone in soft agar and their ability to induce anchorage independent growth of NRK cells. The secretion of substances with TGF like activity may be a property of many types of primary human cells.     

Direct cloning of human neuroblastoma cells in soft agar culture

An in vitro soft agar technique was used in an attempt to culture neuroblastoma cells from 71 bone marrow, 3 lymph node, and 2 solid tumor specimens from 18 patients with neuroblastoma. One-half of each specimen was sent for routine pathology studies and one-half was cultured in the soft agar system. Colonies appeared within 10 days in histologically positive bone marrows. Light microscopy, electron microscopy, catecholamine secretion, and karyology provided evidence that the colonies were composed of neuroblastoma cells. There were 38 instances in which histological study of the specimen demonstrated neuroblastoma cells. The soft agar system showed colony growth in 30 of these samples (79%). There were a total of 38 specimens that were histologically negative for neuroblastoma. Thirty of these 38 specimens showed no growth in the stem cell assay. Eight histologically negative specimens from 6 patients formed colonies in the soft agar system. Five of these 6 patients showed tumor histologically on prior or subsequent marrow examinations. In addition to a significant correlation between histological and soft agar culture results (p < 0.001), there exists a highly significant positive correlation between the number of colonies per plate and the histological status of the specimen (p < 0.005). Serial marrow samples were cultured on 7 patients. There appears to be an association between the number of colonies that develop in the plate and the clinical course and prognosis of the patient. Decreasing plating efficiencies (number of colonies per number of cells plated) correlated with tumor response. Increasing plating efficiencies indicated tumor relapse. A plating efficiency of greater than or equal to 0.1% portended a particularly poor prognosis. Neuroblastoma grows well in this soft agar culture system. This excellent growth provides a good model for both clinical and basic science studies of neuroblastoma.     

Expression of transforming growth factor-beta 1 and growth in soft agar differentiate prostate carcinoma-associated fibroblasts from normal prostate fibroblasts

Carcinoma-associated fibroblasts (CAF) promote tumor progression of pre-neoplastic epithelial cells. To investigate the basis of this phenomenon, we compared the properties of fibroblasts cultured from normal human prostate (NHPF) to prostate CAF. NHPF and CAF were assayed for growth potential, cell death and proliferative capacity by measuring population doubling time, cell cycle distribution and capability to form colonies in soft agar. Resistance to genotoxic (UV radiation: 0-50 J/cm2) and chemotoxic (0-200 nM Taxol) agents were compared between CAF and NHPF by measuring cell viability and cell cycle analysis. Transforming growth factor beta1 (TGF-beta1) immunoreactivity was assessed in non-malignant and malignant prostatic tissue. No detectable differences were found when comparing CAF and NHPF with respect to population doubling time, cell cycle distribution and response to genotoxic and chemotoxic agents. The mean number of colonies in soft agar was 120.5 for CAF vs. 18.2 for NHPF (p < 0.05). Because TGF-beta1 and matrix metalloproteinase (MMP)-9 have been associated with growth of fibroblasts in soft agar and tumor promotion, we measured the expression of these factors in NHPF and CAF by ELISA. There was no difference in expression of MMP-9; however, TGF-beta1 was expressed in higher concentrations in CAF than in NHPF (p < 0.0014). Furthermore, TGF-beta1 expression was higher in the carcinoma-associated stroma of prostate cancer tissue than stroma of non-malignant prostatic tissue. Increased capability of CAF as compared to NHPF to form colonies in soft agar may be due to a higher expression of TGF-beta1 and correlates with the ability of CAF to promote malignant progression of prostate epithelial cells.     

Immunological detection and quantitation of alpha transforming growth factors in human breast carcinoma cells

Summary Alpha transforming growth factors (TGFs) were immunologically detected in the concentrated conditioned medium (CM) prepared from four human breast cancer cell lines and from primary cultures of human mammary epithelial cells, and in the tissue extracts prepared from normal, benign, and malignant breast biopsies. Immunoreactive TGFs were quantitated by a competitive radioimmunoassay (RIA) using affinity-purified polyclonal sheep anti-rat TGF antibodies which react with human TGF but not with human epidermal growth factor (EGF). The relative level of RIA-detectable TGFs in the CM from the breast cancer cell lines MCF-7, ZR-75-1, T47-D, and MDA-MB-231, and from the CM of primary cultures of human mammary epithelial cells, ranged from 0.02 to 0.85 ng/ml. MCF-7 or ZR-75-1 cells grown in the presence of 17-estradiol (10^–8 M) for 48 h were found to release two- to three-fold more TGFs into their CM than the same cells grown in the absence of estrogen. In detergent extracts prepared from normal breast tissue, a benign fibrocystic lesion, fibroadenomas and primary breast carcinomas, the relative TGF concentrations were found to range from 1.5 to 6 ng/mg cell protein. No significant correlations were found between the TGF levels and the pathological state of the tissues, the estrogen receptor status of the tumors, or the relative amounts of theras gene protein p21ras in the tissues as determined by Western immunoblot analysis. The question of biological relevancy of TGF for human mammary tumors will require further studies on (a) synthesis and turnover of TGF, (b) the relationship between immunoreactivity and biological activity of TGF, and (c) differences in biological responsiveness of mammary tumor cells.     

Growth modulation of mouse keratinocytes by transforming growth factors

The effects of exogenously added transforming growth factor (TGF alpha and TGF beta on the growth of BALB/MK cells were examined. TGF alpha supplanted the epidermal growth factor (EGF) requirement in these cells. In contrast, TGF beta reversibly inhibited the growth of BALB/MK cells by abrogating the stimulatory actions of EGF or TGF alpha. The inhibitory effects of TGF beta appeared to be mediated by events distal to EGF ligand-receptor interactions. Growth inhibition of BALB/MK cells by TGF beta did not result in the induction of differentiation. This finding is different from the growth inhibition of these cells induced by elevated calcium levels (1.5 mM) which was tightly coupled to terminal differentiation. The BALB/MK cells were found to express TGF alpha mRNA, as well as TGF beta mRNA and protein. In addition, TGF alpha, as well as EGF, enhanced TGF alpha gene expression. These studies suggest a role for endogenous TGFs in regulating BALB/MK proliferation. TGF alpha provides a positive growth signal, while TGF beta is a potent inhibitor of growth even in the presence of such positive modulators as TGF alpha and EGF.     

Comparison of intra- and extracellular transforming growth factors from nontransformed and chemically transformed mouse embryo cells

Multiple transforming growth factors (TGFs) have been isolated from the serum-free conditioned media and acid-ethanol cell extracts of nontransformed AKR-2B and chemically transformed AKR-MCA mouse cells. The TGF activity was analyzed using Bio-Gel P-60 chromatography in 1 M acetic acid and tested for colony-stimulating activity in soft agar using AKR-2B, AKR-2B (clone 84A), NRK, and AKR-MCA cells as indicators. Both intracellular and extracellular TGF activity from AKR-MCA and AKR-2B cells show a major peak of AKR-2B and epidermal growth factor-potentiated NRK colony-stimulating activity that coelute in the 13,000 +/- 2,000 molecular weight region. In the 24,000 +/- 7,000 molecular weight range, AKR-MCA cells produce intracellular and extracellular NRK colony-stimulating activity that is not potentiated by epidermal growth factor, while the intracellular and extracellular NRK colony-stimulating activity produced by AKR-2B cells requires added epidermal growth factor for colony formation. Also important in determining the growth and morphological characteristics of a cell line, besides the production of a TGF, is the ability of a cell to respond to TGF activity. We have shown that the transformed AKR-MCA cells form more colonies than AKR-2B cells in response to certain TGF activities. This suggests that the increased responsiveness of AKR-MCA cells to TGFs may be important in determining its phenotype.     

Multiplicity of transforming growth factors in human malignant effusions

Human malignant effusions were found to contain transforming growth factor (TGF) activity capable of stimulating anchorage independent growth of nontransformed rodent fibroblasts. Bio-Gel P-60 chromatography of acid-ethanol extracts demonstrated the presence of three populations of TGF activities in 57% of malignant effusions. Two activities were similar to those of TGF alpha and TGF beta as judged by their size (Mr approximately equal to 6,000 and approximately equal to 25,000, respectively), biological activity (ability to stimulate anchorage independent growth of NRK fibroblasts in the absence or presence of epidermal growth factor, respectively), and capacity to competitively inhibit binding of 125I-labeled epidermal growth factor to A-431 membranes and 125I-TGF beta to baby hamster kidney fibroblasts, respectively. In addition a third factor which stimulated anchorage independent growth of nontransformed rodent fibroblast and human colonic epithelial cells was also recovered following Bio-Gel P-60 chromatography of extracts from several cytology positive human malignant effusions of patients with colonic and breast carcinoma as well as other malignancies. The latter malignant effusion related transforming growth factor was not present in benign or cytology negative effusions. Malignant effusion related TGF factor was inactivated by sulfhydryl reducing agents, heat, and trypsin treatment but was stable in 1% acetic acid and ethanol. Partial purification was accomplished by chromatography of an acid-ethanol extract on Bio-Gel P-60 followed by high performance liquid chromatography with C18-mu Bondapak to yield a nearly pure protein with apparent molecular weights of 64,000 by sodium dodecyl sulfate-polacrylamide gel electrophoresis when run in nonreducing conditions and 32,000 when run in reducing conditions. Malignant effusion related TGF was able to stimulate anchorage independent growth of nontransformed fibroblasts in the absence of other growth factors. It did not competitively inhibit binding of 125I-labeled epidermal growth factor, 125I-TGF beta, or 125I-labeled platelet derived growth factor. Therefore, this factor isolated from human malignant effusions may be distinct from previously described transforming growth factors. Collectively these observations indicate that human malignant effusions contain a multiplicity of transforming growth factors. It is possible that the malignant effusion related transforming growth factors play a role or reflect the metastatic growth properties of various tumors.     

Effects of physiological oxygen concentration on human tumor colony growth in soft agar

Cloning efficiencies of 67 fresh human tumor specimens, consisting of 29 ovarian, 10 lung, and 7 each of colon, breast, mesenchymal, and miscellaneous tumors, from 58 patients were studied using the technique of Hamburger and Salmon to evaluate the effect of incubation in a 5% rather than 20% oxygen environment. Under the low-oxygen tension, carcinomas exhibited an average of 170% increase in cloning efficiency (p less than 0.01). The number of carcinomas forming at least 30 colonies/dish increased from 22 to 27. Mesenchymal tumors, however, exhibited a 20% decrease in cloning efficiency (not significant). A mixture of 5% oxygen, 5% carbon dioxide, and 90% nitrogen gives a higher cloning efficiency than 20% oxygen, 5% carbon dioxide, and 75% nitrogen for certain human carcinomas in semisolid agar.     

Cloning of human solid tumors in soft agar

Of 1,014 human solid tumors of various histologic types, 690 (68%) showed evidence of colony formation within 2 to 4 weeks. Tumors which grew particularly well were colon carcinoma (104/175), melanoma (134/155), lung carcinoma (62/85), breast cancer (100/140), ovarian carcinoma (50/67), and sarcoma (72/122). Histologic examination indicated that the colony-forming cells retained functional and morphologic features similar to those of the original tumor. Plating efficiencies varied between 0.01% and 0.2%, and the numbers of colonies observed formed a direct linear correlation with the number of cells plated. Recovery of viable tumor cells was increased when enzymatic tumor dissociation was used rather than a mechanical method. A simplified, supplemented medium resulted in improved cloning efficiencies when compared to previously reported methods (Hamburger and Salmon, 1977 b).     

Partial purification of transforming growth factors from human milk

Crude, delipidated milk and the acid:ethanol extracts of primary human breast tumors contain several activities that biologically resemble transforming growth factors (TGFs) in that they promote the anchorage-independent growth of normal rat kidney and Mm5mt/c1 mouse mammary tumor cells in soft agar. Three major TGF species with isoelectric points (pl) of about 4.0, 6.0-6.5, and 7.0 have been detected in both tumors and milk. The pl 4.0 species from milk has been purified about 10,000-fold by isoelectric focusing and high-performance liquid chromatography. This species, designated milk-derived growth factor II (MDGFII), coelutes from gel filtration columns with an authentic human epidermal growth factor standard when using a low ionic strength eluting buffer. However, on the same column, MDGFII is completely resolved from human epidermal growth factor with high ionic strength eluting buffers. Nevertheless, MDGFII purified by the latter technique still competes with 125I-epidermal growth factor for receptor binding to A431 cell membranes. Additionally the TGF activity of MDGFII present in the pl 4.0 fraction of milk is markedly inhibited by anti-epidermal growth factor receptor antibody preparations. Consequently MDGFII appears to be an alpha-TGF. MDGFII is a pepsin-sensitive, disulfide reducing agent-sensitive, heat-stable protein that may be physiologically important for the mammary gland or the neonate.     

Colony growth in soft agar of human melanoma, sarcoma, and lung carcinoma cells disaggregated by mechanical and enzymatic methods

The effect of mechanical and enzymatic disaggregation on human malignant melanoma, soft-tissue sarcoma and lung carcinoma colony growth in soft agar was studied. The enzymatic disaggregation was advantageous in most cases of melanoma and sarcoma, giving a larger number of colonies and increasing the probability of achieving growth in soft agar. Enzymatically treated pulmonary carcinoma cell populations had lower clonogeneic potential, especially in the case of anaplastic carcinomas. Morphological studies showed that the cells growing in soft-agar colonies had the same characteristics as those of the original tumor. A linear relationship was obtained between the number of enzymatically and mechanically treated tumor cells plated and the number of colonies. Delayed plating decreased the number of colonies.     

Some factors influencing the behaviour of BHK 21 Cl 13 cells in soft agar medium

As part of an attempt to reproduce Styles' cell transformationassay, the effect of serum concentration on the growth of normaland 4-nitroquinoline-1-oxide (4-NQO)-treated BHK 21 Cl 13 cellsin soft agar medium was examined. Using medium with 10% newborncalf serum, dilution of control cultures from 5 x 10⁴ to 1.56x 10³ cells/ml caused little increase in the number of countable(>0.3 mm diameter) colonies, but 4NQO caused a marked dose-relatedincrease. In contrast, using 20% of the same batch of serum,4NQO-treated groups and controls diluted to comparable viabilitycounts showed very similar increases in countable colonies.Clones >0.3 mm diameter isolated from control cultures with20% serum did not appear to be transformed when grown in agarwith 10% serum. These data indicate that factors other thanneoplastic transformation can influence the growth of BHK 21Cl 13 cells in soft agar medium.     

Dexamethasone, medroxyprogesterone acetate, and vindesine modulation of human malignant melanoma growth in soft agar

We have previously shown biopsies from human malignant melanoma, as well as a cell line (MELUR) derived from human malignant melanoma cells, to contain dexamethasone (DEX) - and medroxyprogesterone acetate (MPA) - binding macromolecules exhibiting properties of glucocorticoid receptors from well-known target organs. In this paper, we demonstrate that the cloning efficiency of MELUR melanoma cells is inhibited by DEX and MPA in a dose-dependent manner. MELUR tumour colony-forming units (MELUR-TCFU) were inhibited more than 40% by either DEX or MPA at 1 X 10(-7) mol/l with concentrations needed to obtain a 50% reduction in TCFU counts for both hormones of 3 X 10(-9) mol/l. MELUR cell growth in soft agar was also modulated by vindesine (VI). At 50 ng/ml, VI produced a 70% reduction of colony formation. The combined drug effects of either DEX or MPA and VI on colony formation were additive. The data indicate that the proliferative activity of MELUR cells is regulated by DEX and MPA, both acting additively to VI.     

The effects of alpha and gamma interferons on human lung cancer cells grown in vitro or as xenografts in nude mice

We have compared the effects of alpha and recombinant gamma interferons (IFNs) on the growth of human lung cancer cell lines in vitro. There was a diversity of response amongst the lines studied, the most sensitive being COR-L23 (a large cell anaplastic carcinoma line) and POC (a small cell line). In these two lines, IFN-gamma was found to be more potent than IFN-alpha. During cell growth of line POC in the presence of IFN-gamma no significant shift in cell cycle distribution occurred. When lines COR-L23 and POC were grown as xenograft tumours in nude mice, daily injection of 4 X 10(5) units per mouse per day of IFN-gamma produced no discernible retardation of tumour growth.     

Interactions between MYC and transforming growth factor alpha alter the growth and tumorigenicity of liver progenitor cells

The MYC oncogene induces both cell proliferation and apoptosis. The apoptotic function of MYC is thought to inhibit carcinogenesis; thus, when disrupted, tumorigenic potential is increased. Both MYC and transforming growth factor alpha (TGFalpha) are commonly over-expressed in hepatocellular carcinomas, and transgenic mice expressing these genes rapidly develop tumors via the suppression of MYC-induced apoptosis by the growth factor. However, the nature of the interactions between MYC and TGFalpha are not well understood. Specifically, it is unclear whether TGFalpha acts only as an anti-apoptotic factor in its interactions with MYC or whether it has substantial effects on cell growth. We investigated whether TGFalpha can provide additional mitogenic signals if it is not required to act as an anti-apoptotic factor. We demonstrate that expression of MYC and TGFalpha in liver progenitor cells (known as oval cells) results in enhanced cell proliferation in culture and the generation of poorly differentiated tumors after inoculation into nude mice. We further demonstrate that while the apoptosis-deficient T58A and S71F alleles of MYC retain their ability to promote oval cell proliferation, they have opposite growth interactions with TGFalpha. The T58A allele has a stimulatory effect on both proliferation and tumorigenicity. In contrast, co-expression of the S71F allele reduces proliferation and slows tumor development. We conclude that the tumorigenic growth effects of MYC in TGFalpha-expressing liver progenitor cells are not solely dependent on its apoptotic activity.     

Urinary transforming growth factors in neoplasia: separation of 125I-labeled transforming growth factor-alpha from epidermal growth factor in human urine

Purified human epidermal growth factor (hEGF) from urine promotes anchorage-independent cell growth in soft agar medium. This growth is enhanced by transforming growth factor-beta (TGF-beta), and is specifically inhibited by hEGF antiserum. Transforming growth factors of the alpha type (TGF-alpha), potentially present in normal human urine or urine from tumor-bearing patients, also promote anchorage-independent cell growth and compete with EGF for membrane receptor binding. Consequently, TGF-alpha cannot be distinguished from urinary hEGF by these two functional assays. Therefore, a technique for separation of TGF-alpha and related peptides from urinary EGF based on biochemical characteristics would be useful. Radioiodination of characterized growth factors [mouse EGF (mEGF), hEGF, and rat TGF-alpha (rTGF-alpha)], which were then separately added to human urine, was used to evaluate a resolution scheme that separates TGF-alpha from the high level of background hEGF present in human urine. Commercially available methyl bonded microparticulate silica efficiently adsorbed the 125I-labeled mEGF, 125I-labeled hEGF, and 125I-labeled rTGF-alpha that were added to 24-h human urine samples. Fractional elution with acetonitrile (MeCN) of the adsorbed silica released approximately 70 to 80% of the 125I-labeled mEGF and 125I-labeled hEGF between 25 and 30% MeCN, and over 80% of the 125I-labeled rTGF-alpha between 15 and 25% MeCN, with retention after dialysis of less than 0.2 and 1.7% of the original urinary protein, respectively. Consequently, a single-step enrichment of about 400-fold for mEGF and hEGF, and 50-fold for rTGF-alpha were achieved rapidly. Subsequent Bio-Gel P-10 chromatography indicated that 125I-labeled mEGF and 125I-labeled hEGF eluted later than would be predicted on the basis of their reported molecular weight of approximately 6000, whereas 125I-labeled rTGF-alpha eluted from Bio-Gel P-10 at an approximate molecular weight of 8000 to 9000. 125I-labeled rTGF-alpha bound to carboxymethyl cellulose and eluted at a less acidic pH than did hEGF. On reverse phase high performance liquid chromatography using a linear gradient of 18 to 35% MeCN over 120 min, 125I-labeled rTGF-alpha was comparatively hydrophilic, eluting at 22% MeCN in contrast to the more hydrophobic 125I-labeled hEGF, which eluted at 27% MeCN. These observed biochemical differences between hEGF and TGF-alpha provide a rationale for resolution of these and perhaps other related putative transforming growth factors from bulk urine of tumor-bearing patients.     

Production of transforming growth factors by human colon cancer lines

Three human colon cancer lines (SW 480, SW 620, WIDR) were characterized as to their production of molecules with transforming growth factor (TGF)-like activity. Production of both TGF alpha-like and TGF beta-like activity was quantitated, as were cellular receptors for these molecules, and growth response in soft agar to exogenous epidermal growth factor (EGF) (as a substitute for TGF alpha) and TGF beta. Serum-free medium conditioned by these cells showed differing amounts of TGF alpha-like and TGF beta-like competing activity in EGF and TGF beta radioreceptor assays. Likewise the cells showed differing abilities to bind 125I-labeled EGF and TGF beta. SW 620 cells produced relatively large quantities of TGF alpha-like activity and had no detectable EGF receptors; specific TGF beta binding was observed. SW 480 cells produced the most TGF beta-like activity and had no measurable TGF beta membrane receptors, but EGF receptors were detectable. WIDR cells had both EGF and TGF beta membrane receptors and produced relatively low levels of EGF and TGF beta receptor-competing activity. All three of the cell lines grew spontaneously in soft agar (in medium containing 10% serum). In contrast to other carcinoma cell lines, exogenous EGF and TGF beta had no significant effect on soft agar growth of the colon carcinoma cells. The production of both TGF alpha-like and TGF beta-like polypeptides by colon carcinoma cell lines has been shown, yet involvement of these factors in autostimulatory activity could not be demonstrated. The possibility that these endogenous factors could be involved in paracrine stimulation of stromal cells remains to be explored.     

Time course of ovarian tumour growth in soft agar culture

Single time point assessment is usually employed in the Human Tumour Cloning System as the only parameter for in vitro growth. This does not seem to give a fair expression of the dynamic biological properties of tumour growth and time dependent effects, e.g. of cytotoxic drugs. We studied the time course of colony formation in temporal growth patterns (TGPs) and compared this method of growth evaluation with conventional single time point assessment in 57 samples of ovarian tumour cultures in the HTCS. A first advantage of the use of TGPs is that more cultures become evaluable, as this assessment over time can detect a rise in the number of colonies in dishes where colony-like clumps have initially been seeded. Thus only 28 of the cultures were evaluable for single time point assessment, whereas 57 were available for TGP evaluation. Growth was more often seen at TGP evaluation (14/57) than at single day assessment (8/57). Evaluation of growth over the course of time potentially allows detection of sensitivity to drugs. Furthermore TGPs reflect the dynamics of biological growth. These features cannot be studied in single time point assessment.