Effect of host-cell interactions on clonogenic carcinoma cells in human malignant effusions

Imaging and analysis of 3D tumor spheroids enriched for a cancer stem cell phenotype

Tumors that display a highly metastatic phenotype contain subpopulations of cells that display characteristics similar to embryonic stem cells. These cells exhibit the ability to undergo self-renewal; slowly replicate to retain a nucleoside analog label, leading to their definition as "label-retaining cells"; express specific surface markers such as CD44(+)/CD24(-/low) and CD133; and can give rise to cells of different lineages (i.e., they exhibit multipotency). Based on these characteristics, as well as their demonstrated ability to give rise to tumors in vivo, these cells have been defined as tumor-initiating cells (TICs), tumor-propagating cells, or cancer stem cells (CSCs). These cells are highly resistant to chemotherapeutic agents and radiation and are believed to be responsible for the development of both primary tumors and metastatic lesions at sites distant from the primary tumor. Established cancer cell lines contain CSCs, which can be propagated in vitro using defined conditions, to form 3D tumor spheroids. Because the vast majority of studies to identify cancer-associated genes and therapeutic targets use adherent cells grown in 2 dimensions on a plastic substrate, the multicellular composition of these 3D tumor spheroids presents both challenges and opportunities for their imaging and characterization. The authors describe approaches to image and analyze the properties of CSCs within 3D tumor spheroids, which can serve as the basis for defining the gene and protein signatures of CSCs and to develop therapeutic strategies that will effectively target this critically important population of cells that may be responsible for tumor progression.

Lack of association between in vitro clonogenic growth of human cervical carcinoma and tumour stage, differentiation, patient age, host cell infiltration or patient survival

Biopsies from 117 patients with cervical carcinoma were studied using a clonogenic assay to assess in vitro growth. Successful colony growth was achieved in 84 tumours (72%) with a mean colony-forming efficiency (CFE), based on total viable nucleated cell counts, of 0.18 +/- 0.49% (+/- 1 standard deviation). There was a wide range of values, from 0.003-4.28%, with a coefficient of variation of 272%. The relationship between clinical features of cervical carcinoma and in vitro colony formation was investigated. No significant association was demonstrated between in vitro growth and either clinical stage (r = 0.02), tumour differentiation (r = -0.08) or patient age (r = -0.12). There was no significant difference in tumour grade between the group of tumours which failed to grow in culture and those which grew well (p = 0.70). In addition, there was no correlation between CFE and the degree of macrophage (r = 0.001), lymphocyte (r = 0.12), or granulocyte (r = 0.08) infiltration. There was no significant difference between CFEs of tumours from patients who had died and from those who were alive and well after a minimum of 2 years' follow-up after radiotherapy (p = 0.51). Ability to form colonies in agar was not associated with a worse prognosis (p = 0.49). Although CFE is an independent biological parameter, our results suggest that, for cervical carcinoma, it is not useful as a univariate prognostic factor.

Analysis of factors influencing clonogenic growth in vitro of cells from ovarian carcinoma patients

Clonogenic growth (defined as the formation of greater than or equal to 5 colonies per 5 x 10(5) viable nucleated cells per plate) of ovarian cancer specimens assessed in our clonogenic assay system was significantly associated with the proportion of tumor cells in the suspensions plated (N = 87; P = 0.0006), although there was no quantitative relationship with the corresponding plating efficiencies. An inverse correlation was observed between monocytes/macrophages/mesothelial cells (M) proportion and clonogenic growth (P = 0.013). These associations were most evident when only effusions were considered. Univariate analyses identified tumor cell content, M proportion and, to a lesser degree, granulocyte content as the only factors out of 12 examined to be correlated with colony formation. Multivariate analysis using a logistic regression model identified the proportion of tumor cells as the only significant factor predicting clonogenic growth in vitro (P = 0.0006). The overall accuracy of prediction for growth or non-growth was 63.2%.

Mesothelial cells stimulate the anchorage-independent growth of human ovarian tumour cells

Results are presented which show that mesothelial cells (MC) from ovarian cancer patients can both stimulate and inhibit the clonogenic growth of ovarian tumour cells (TC) in a dose-dependent fashion. TC lines from both non-ovarian and ovarian tumours were variable in their response to MC. Colony formation was rarely induced when the TC population was non-clonogenic and a bladder cell line showed inhibition of colony formation in the presence of MC. Primary tumour cultures from ovarian cancer patients also showed a variable response to MC. Fibroblasts from malignant, benign and non-neoplastic sources were significantly less effective in stimulating the clonogenic growth of responsive cell lines. Conditioned medium was a poor substitute for the presence of intact viable cells, and distance between feeder cell and TC was an important factor in determining the magnitude of response. A significant relationship between the feeder effect of MC and their proliferation in soft agar was observed when epidermal growth factor was used in the medium. The relevance of the findings in the context of the pattern of spread of ovarian cancer is discussed.

Constitutive production of macrophage colony-stimulating factor by human ovarian and breast cancer cell lines

Many nonhematologic tumors produce growth factors that may influence cellular proliferation either by autocrine or by paracrine mechanisms. In the current study, human tumor cell lines were investigated for the constitutive production of macrophage colony-stimulating factor (M-CSF). Culture supernatants obtained from cell lines were analyzed using a radioimmunoassay and a radioreceptor assay specific for M-CSF. Among the various cell types analyzed, all the ovarian cell lines and a majority of the breast cancer cell lines secreted significant amount of an M-CSF-like factor. Treatment of mouse bone marrow cultures with culture supernatants from ovarian cancer cells stimulated the production of macrophage colonies. Analysis of total cellular RNA obtained from the ovarian cell lines by Northern blot showed multiple sizes of M-CSF transcripts with an abundance of a 4.2-kb message. The relative amount of M-CSF transcripts correlated with the level of immunoreactive material seen in the culture supernatants.

Retrovirus-associated ovine pulmonary carcinoma (sheep pulmonary adenomatosis) and lymphoid interstitial pneumonia. I. Lesion development and age susceptibility

To determine the lesion development of retrovirus-induced ovine pulmonary carcinoma (OPC), ten neonatal lambs were inoculated intratracheally with either 1) lung fluid preparations derived from a sheep with Type D retrovirus-associated OPC and concurrent ovine lentivirus (OvLV)-associated lymphoid interstitial pneumonia (LIP) (n = 8); or 2) lung fluid from a sheep with only OvLV-LIP (n = 2). Seven of eight neonates that received Type D retrovirus-associated OPC/OvLV-LIP lung fluid developed both OPC and LIP lesions between 9 and 32 weeks after inoculation. Mild OPC lesions consisted of foci of type II alveolar epithelial cells lining alveoli surrounded by minimal alveolar macrophage infiltrates. More severe OPC lesions consisted of multifocal aggregates of cuboidal to columnar neoplastic cells forming acini or masses associated with abundant alveolar macrophage infiltrates. Lesions of LIP consisted of peribronchiolar and perivascular lymphoid hyperplasia and heterogeneous interstitial leukocytic infiltrates. The two neonates that received OvLV-LIP lung fluid developed rapid and severe LIP, but not OPC lesions. Two lambs (inoculated as neonates with virus-free lung fluid) and three lambs (uninoculated contacts) served as controls and did not develop OPC. To investigate age susceptibility for development of OPC, 20 additional lambs within defined age groups (neonates, 2 weeks old, 5 weeks old, and 10 weeks old) received ultracentrifuged tumor homogenate. Neonatal to 5-week-old lambs inoculated with Type D retrovirus-associated OPC/OvLV-LIP tumor homogenate were equally likely to develop OPC, but lambs inoculated at 10 weeks of age were more refractory to tumor development.(ABSTRACT TRUNCATED AT 250 WORDS)

Lesions and retroviruses associated with naturally occurring ovine pulmonary carcinoma (sheep pulmonary adenomatosis)

Five sheep with ovine pulmonary carcinoma were markedly dyspneic and had sporadic coughing; two had copious watery nasal exudate. In four, lesions consisted of multifocal nodules of neoplastic cuboidal epithelial cells in acinar or papillary patterns. Electron microscopically, cells had microvilli, tight junctions, and cytoplasmic lamellar bodies typical of alveolar type II cells. One sheep had a single lung tumor of nonciliated bronchiolar epithelial cells. Vacuolated alveolar macrophages surrounded adenomatous foci. One sheep had a metastatic lesion in the caudal mediastinal lymph node. All sheep had histologic lesions of lymphoid interstitial pneumonia (LIP, ovine progressive pneumonia) consisting of peribronchiolar and interstitial lymphoid hyperplasia, and fibromuscular proliferation; all had serum precipitating antibodies to ovine lentivirus. Lung fluids or tumor homogenates contained a 26-kd peptide that crossreacted with a primate-derived type D retrovirus as detected by immunoblotting or interspecies competition radioimmunoassay. Ovine lentivirus was isolated from concentrated lung fluids or tumor tissues of four sheep tested and from tumor cell DNA of one animal transfected into ovine muscle cells. These studies document the presence of type D-related retrovirus antigen in ovine pulmonary carcinoma (OPC) in the United States and indicate that lentivirus-induced LIP is a lesion frequently associated with this disease.

Enhancement of survival and proliferation of clonogenic cells by monocyte-derived growth factors

Human peripheral blood monocytes release a factor that enhances the clonal growth of human epithelial tumor cells in soft agar. The monocyte-derived growth factor was needed for both cellular proliferation and survival. Survival of SW-13 colony-forming cells decreased linearly in the absence of monocyte-conditioned media (MO-CM). Cells failed to respond to MO-CM after four days in culture. Although MO-CM enhanced growth when cells were plated at low density, growth was also enhanced when cell density was not a limiting factor. MO-CM increased DNA synthesis of SW-13 cells growing in monolayer culture as measured by tritiated-thymidine incorporation. These findings support evidence indicating inflammatory products may play a role in maintenance of the transformed phenotype.

The human tumor clonogenic assay as a model system in cell biology

The human tumor stem cell (clonogenic) assay (HTCA) is a soft agar system designed for growing fresh human tumor specimens in vitro. The assay has been extensively used in studies both of individual patients' response to chemotherapy and for screening new agents. The technical limitations of this assay have been extensively discussed. The use of this test as a model system to study fundamentals of tumor cell growth has not been stressed. The potentials and limitations of this assay for the study of the regulation of tumor growth are presented.

Simultaneous in vitro drug sensitivity testing on tumors from different sites in the same patient

A human tumor cloning assay was used to analyze drug sensitivity profiles for 99 pairs of tumor samples taken simultaneously from the same patient. These pairs included two areas of the same primary (12 pairs), primary tumor versus metastasis (29 pairs), and metastasis versus metastasis (45 pairs). In addition, to assess variability in the culture and sensitivity techniques, 13 specimens were divided equally and processed twice. One hundred fourteen evaluable drug sensitivity tests were performed on the specimen pairs. Drug sensitivity profiles for the same specimen processed twice demonstrated good agreement (P = 0.03). Sensitivity profiles from two different areas of the same primary were also quite similar (P = 0.002). However, when one sampled a primary tumor and its metastasis for drug sensitivity testing, the agreement (in terms of percent survival) was poor (P = 0.84). Agreement was also poor for drug sensitivity profiles of one metastasis versus another metastasis (P = 0.39). This heterogeneity in drug sensitivity profiles of primary versus metastasis and of metastasis versus metastasis could account for some of the false positives and false negatives noted in clinical correlation trials with the cloning assay. It may also have implications for adjuvant treatment programs where chemosensitivity of the primary tumor can be determined, but will probably not be predictive for chemosensitivity of the micrometastases.

Enhanced chemosensitivity of cells from malignant effusions under condition of exposure to high temperature

Utilizing a clonogenic assay, the effects of hyperthermia and selected chemotherapeutic drugs on growth of cells from malignant effusions were studied. Fourteen of 25 samples obtained from 25 patients with various carcinomas formed at least 30 colonies per plate. Exposure of the cells to heat at 42 degrees C for 1 hr before the plating slightly inhibited the colony growth. The drugs, adriamycin (AM) and mitomycin C (MMC), were tested at 3 different concentrations. When the cells were treated with these two drugs for 1 hr at 42 degrees C, the percent of surviving colonies was significantly decreased, as compared to findings at 37 degrees C, in both groups, at 3 different concentrations. The combination of drugs and hyperthermia appeared to function synergistically in one-third of such cases. These results suggest that cells from malignant effusions in patients with various carcinomas were more sensitive to AM or MMC, under condition of a higher temperature (42 degrees C).

Human tumor clonogenic assays: growth conditions and applications

Increasing investigative efforts have been focused on the application of in vitro clonogenic assays for human tumors. Several semisolid matrices have been used for purposes of immobilizing the clonogenic cells, and a variety of growth factors and environmental conditions have been studied to enhance tumor colony growth. Although there remain some technical problems with these assays, the methodology has proven to have broad applicability to preclinical and clinical studies of human cancer. This report reviews some features and areas of application of this assay methodology. The ability to combine studies of drug effect with cytogenetic studies on clonogenic cells suggests that such assays may prove useful for evaluation of the emergence of drug resistance by primary and metastatic cancers.

Evaluation of the interaction of mononuclear phagocytes with ovarian carcinoma cells in a colony assay

The effect of human peripheral blood monocytes on the SW626 ovarian carcinoma line was investigated in a colony assay in agar, Percoll-enriched monocytes inhibited colony formation by SW626 carcinoma cells at effector-to-target cell (E:T) ratios as low as 0.3:1. In contrast the same effectors had little cytolytic effect in a 48 h thymidine-release assay at E:T ratios as high as 40:1. Monocyte-depleted nonadherent cells had little inhibitory capacity on SW626 colony formation, whereas unseparated mononuclear cells were intermediate between Percoll-enriched monocytes and lymphoid cells. Sorting of cells positive for the monoclonal antibody marker MO2 confirmed the monocytic nature of cells which inhibited colony formation. Ovarian carcinoma cells freshly isolated from 9 patients were heterogenous in their susceptibility to colony inhibition by mononuclear phagocytes. Cells from 4 patients were not inhibited by effector cells and in one subject promotion of colony formation by mononuclear phagocytes was observed. With 4 cell preparations inhibition of colony formation was found as with the SW626 line. Colony assays may provide a useful methodological approach, particularly when effector cells mediate low levels of killing, of doubtful biological significance, in conventional isotope release assays, or when growth promotion is to be evaluated.

Growth factors for human tumor clonogenic cells elaborated by macrophages isolated from human malignant effusions

We previously demonstrated that macrophages isolated from human malignant effusions support colony formation of autologous tumor cells in soft agar. We now demonstrate that macrophages (derived from effusions of patients with ovarian, breast, colon, or lung adenocarcinomas) secrete a soluble factor(s) that enhances the ability of a human epithelial tumor cell line (SW-13) to clone in soft agar. Macrophages increased colony growth 5 to 10-fold in a concentration dependent manner, although inhibition of cell growth was observed in the presence of high concentrations of macrophages. We attempted to increase production of tumor colony stimulating factor by exposing macrophages to lipopolysaccharide, concanavalin A, or phytohemagglutinin. Exposure of macrophages to these agents failed to increase their ability to secrete stimulatory factors. Macrophages were cultured for 1 day to 6 weeks in the presence of GCT-CM, a source of granulocyte-macrophage colony stimulating factor and the ability of these cultured macrophages to support colony growth assessed. The ability of macrophages to support colony growth declined gradually with time in culture reaching 50% of control values at 14 days, but remained at this level until 5 weeks of culture. The results of this study indicate the SW-13 cells may provide a quantitative assay for studying monocyte-derived tumor colony stimulating factors.

Non-linearity of colony formation by human tumour cells from biopsy samples

The relationship between colony numbers and concentration of cells plated is an important parameter of clonogenic assay systems. The cloning efficiency for an ideal sample should be independent of cell concentration, thus giving a straight line through the origin when colony numbers are plotted against cell concentration. A simple statistical method has been developed to test if this is the case for individual tumour samples. Colony data from 51 freshly obtained tumour samples, which had sufficient cells to plate 3 or more dilutions and gave at least 20 colonies per plate at one or more of the dilutions, were tested. The results indicated that colony formation was linear for 27 (53%) of the samples. The remaining 24 samples could be classified into 2 groups: type I, in which cloning efficiencies increased with increasing cell concentration and type II, which had reduced cloning efficiencies at high cell concentrations. Fifteen (29%) of the samples had type I non-linearity and 9 (18%) exhibited non-linearity of type II. These findings indicate that the relationship between colonies and cells plated should be examined for each biopsy sample particularly in each experiment where the effects of cytotoxic drugs are tested.

Tumor-derived chemotactic factor(s) from human ovarian carcinoma: evidence for a role in the regulation of macrophage content of neoplastic tissues

Supernatants from freshly disaggregated human ovarian carcinomas maintained in vitro for 24 hr, from primary ovarian carcinoma cultures (4-6 days in culture) and from established ovarian cancer cell lines were examined for chemotactic activity on blood monocytes in blind-well chemotaxis chambers. Tumor-cell culture supernatants induced migration of peripheral blood monocytes across polycarbonate filters with considerable heterogeneity among different tumors. Induction of migration occurred only in the presence of a gradient between the lower and upper compartments of the chamber. Chemotactic activity was characterized by means of supernatants from primary ovarian carcinoma cultures. Chemotactic factor(s) was (were) produced in serum-free conditions and the production was inhibited by emetine but not by mitomycin C. The activity was destroyed by exposure to proteolytic enzymes and by heating at 100 degrees C but was unaffected by RNase, DNase, lipase and exposure to extreme pH values or heating at 56 degrees C. Upon fractionation on Sephadex G 75, the activity eluted as a single peak in the cytochrome C region, corresponding to an apparent molecular weight of about 12 kd. The percentage of macrophages was assessed in 25 freshly disaggregated tumor specimens. Ovarian carcinomas were heterogeneous in their macrophage content with values ranging from 4 to 36%. A significant (r = 0.62; p = 0.00097), though far from absolute, correlation was found between chemotactic activity of culture supernatants and percentage of tumor-associated macrophages. Tumor-derived chemotactic factor(s) could be one of the mechanisms involved in the regulation of the macrophage content of human ovarian carcinomas.

Combined recombinant human interferon alpha 2 and cytotoxic agents studied in a clonogenic assay

A human tumor clonogenic assay has been used to test the antiproliferative effect of recombinant human leukocyte interferon alpha 2 alone and in combination with each of 8 cytotoxic agents. Cell lines derived from 6 human tumors and primary tumor cells from 13 patients have been used in these clonogenic assay studies. Results show that interferon as a single agent causes insignificant reduction in tumor cell colony survival if the short-term 1-hr cell exposure method is used; only high concentrations of interferon used in continuous cell exposure in the clonogenic assay can demonstrate a reduction in colony survival to below 50% of control values. Combinations of interferon with either doxorubicin or cisplatin frequently show additive and occasionally synergistic antiproliferative effects on tumor cell colony formation. Variations in drug concentrations and sequencing of drugs have been tested, showing that optimal antiproliferative effects of combined interferon and doxorubicin are realized when maximal concentrations of interferon and prolonged cell exposure time of both interferon and doxorubicin are employed. Combinations of interferon and doxorubicin tested in the clonogenic assay demonstrate cytotoxicity superior to that of either agent tested alone.

Infiltrating mononuclear cells in human breast carcinoma: predominance of T4+ monocytic cells in the tumor stroma

Frozen sections of 52 breast carcinomas were examined for the presence and nature of a leukocytic infiltrate. The number of "common leukocyte (T200-) antigen"-bearing cells was remarkably high in the stromal compartment of the carcinomas, whereas the tumor foci themselves were usually infiltrated only sparsely. Approximately 80% of these T200+ stromal cells carried the myelomonocyte lineage-associated antigens M-M522 and 63D3 and exhibited "non-specific esterase" activity. More than 2/3 of the stromal monocytic cells expressed the T4 antigen as verified by monoclonal antibodies (MAbs) directed against different T4-associated epitopes. A T-cell analysis showed that T4+ cells clearly outweighed the total number of T cells as defined by anti-T3 and anti-TII antibodies. In addition, a correlation became apparent between the HLA-DR phenotype and the T lymphocyte content of tumor cell areas in the sense that densely infiltrated tumor areas were invariably HLA-DR+.

Percoll density gradient separation of cells from human malignant effusions

A simple method is described for the separation of cells derived from effusions of patients with adenocarcinomas in discontinuous density gradients of Percoll. After separation, cells from different fractions were analyzed by morphologic, histochemical and immunologic criteria. Total cell recovery from 27 experiments was 67 +/- 4%. Macrophages (82%) were recovered in the intermediate density fraction (1.056-1.067 g ml-1) with a purity of 90%. Recovered lymphocytes (98%) were found in the high density fraction (1.067-1.077 g ml-1) with a purity of 92%. The majority of the lymphocytes recovered were T cells. Malignant adenocarcinoma cells (90%) were recovered in the lowest density fractions (up to 1.056 g ml-1) with a purity of 79%. Use of effective cell separation procedures should facilitate the analysis of the functional capacities of both normal and neoplastic cells derived from human malignant effusions.

Human tumor stem cell assay. A prospective clinical trial

Reported is a prospective clinical trial involving 53 patients evaluating the usefulness of a human tumor stem cell assay for selecting chemotherapeutic agents for patients with advanced malignancies. Three patients could not be directed by the cloning system results because of inadequate tumor growth or other difficulties. Cloning efficiency exceeded 90%. Fewer than 3% of soft agar cultures were contaminated by fungal elements. True-positive rates of 65% and true-negative rates of 90% for the cloning assay in predicting response (or lack of response) to chemotherapeutic agents were seen. Preliminary data on the predictive ability of the assay in determining response to adjunctive hyperthermia are presented. With the derivation of mean doubling times from the assay, individualized treatments may be designed for patients with advanced malignancies.

Enhancement of rat ACT-1 tumor clonogenicity by xenogeneic mouse macrophages

In vitro growth of rat atriocaval epithelial tumor cells (ACT-1) was enhanced by the inclusion of xenogeneic mouse adherent peritoneal exudate cells (PECs) in a two-layer soft agar system. A linear relationship was found between the number of cells plated and the number of colonies when ACT-1 tumor cells were plated at plating densities of between 1 and 5 X 10(5) cell/60 mm plate (r = 0.9, P less than 0.001). Inclusion of irradiated PECs in the bioassay for tumor stem cells resulted in a two and a half-fold increase in colony formation in three separate experiments (P less than 0.001).

In vitro colony formation by cryopreserved primary human tumor cells

The effect of cryopreservation on the ability of primary human cancer cells to form colonies in a two-layer agar system was examined. Although considerable variation occurred, concentrations of 5 or 10% dimethylsulfoxide employed with slow freezing rates allowed survival of colonies in the range 20-40% or greater of nonfrozen controls. The methods used in this study do not require elaborate freezing equipment, and can be used for the cryopreservation of a wide variety of types of cancers.

Effect of hydrocortisone on the macrophage content, growth and metastasis of transplanted murine tumors

Hydrocortisone (HC) reduced the macrophage content of four murine tumors to less than half of control values. Since HC causes monocytopenia and inhibits the recruitment of mononuclear phagocytes at sites of inflammation but does not affect the viability and proliferative capacity of macrophages, this finding suggests that the maintenance of macrophage levels in growing tumors is in part dependent upon the entry into the neoplasm of circulating monocytes. HC inhibited the growth of these tumors. The effect of HC was most marked on the mFS6 and MN/MCA1 sarcomas and 3LL carcinoma, with only marginal inhibition in the M109 carcinoma. HC augmented spontaneous and/or artificial (i.v.) metastasis in these tumor models. Cells from metastatic foci of the mFS6 sarcoma and M109 carcinoma inoculated i.m. were as susceptible to HC as cells from primary tumors. The effects of HC on macrophage content, growth and metastasis of the M109 carcinoma and mFS6 sarcoma were similar in mice with defective T-cell function (nude or thymectomized) or defective NK activity (beige or antiasialo GM1-treated) and in controls. The in vitro growth of the mFS6 and MN/MCA1 sarcomas was not modified by HC. Some inhibition of tumor-cell proliferative capacity was observed with the 3LL and M109 tumors. Tumor cell had high-affinity binding sites for glucocorticoid hormones, but these were not correlated to susceptibility in vivo. Thus it is unlikely that a direct interaction of HC with tumor cells accounts for the inhibition of tumor growth observed in vivo. To evaluate whether HC affected tumor growth by reducing the macrophage content of tumors, mFS6 sarcoma cells were transplanted, mixed with peritoneal macrophages into normal or HC-treated mice. Macrophages did not affect the growth of tumor cells in normal mice, whereas in HC-treated animals lesions from mixtures of macrophages and sarcoma cells appeared earlier and weighed more than those from tumor cells alone or tumor cells and thymocytes. However, macrophages did not reconstitute growth of the other sarcoma (MN/MCA1) in HC-treated mice. These results are consistent with the hypothesis that HC inhibited growth of some poorly immunogenic transplanted murine tumors, at least in part, by interfering with the macrophage levels in neoplastic tissues, tumor-associated macrophages providing the conditions for optimal tumor-cell proliferation at least in some neoplasms.

Improving techniques for clonogenic assays

A serum-free medium has been developed which supports colony formation by cells from several human tumor cell lines, one colon adenocarcinoma (WiDr) and four melanoma (Me43, Me85, MP6, MeIuso). This medium consists of a 1:1 mixture of an enriched Dulbecco's modified Eagle's medium (EMED) and a modified Ham's F-12 nutrient mixture (FMED) supplemented with 0.9% methylcellulose, 1% bovine serum albumin, 80 micrograms/ml human transferrin, 3 micrograms/ml insulin, 2.8 micrograms/ml linoleic acid, 2.6 micrograms/ml cholesterol, 20 microM ethanolamine, and trace elements. Colony formation by WiDr cells is linear with the numbers of cells plated, having a plating efficiency (PE) of 34%, as compared to 26% in serum-containing medium. Two of the melanoma cell lines. MP6 and MeIuso, exhibit linear relationships between colony numbers and cell concentration with PEs of 21% and 70% respectively. Colony formation by the other two melanoma cell lines appears to be nonlinear. This work represents a step toward standardizing culture conditions for human tumor clonogenic cell assays.

Tumor growth stimulation in vitro by interferons

Interferons (IFNs) are a family of polypeptides originally identified as antiviral substances. Subsequently, other properties of interferons were recognized, including inhibition of cell proliferation, and effects on the immune response and on expression of surface antigens. In this paper we present evidence that interferons, even the highly purified cloned IFNs, can stimulate clonogenic tumor growth in vitro. Of 225 human tumor (HT) samples tested with IFN in a clonogenic assay (HTCA), 30 (13.3%) showed growth stimulation (greater than 2 S.E. above control). The phenomenon was observed most frequently with acute myeloid leukemia (6/22 samples, 27.3%), and renal (2/10, 20%) and breast cancer (4/21, 19%), but significantly less frequent in melanomas (2/34, 5.9%). As an independent assessment of proliferation, tritiated thymidine uptake by tumor cells was measured autoradiographically in 21 patients with multiple myeloma. A significant increase of the thymidine labeling index was seen in 4 (19%) of the samples. Since this growth stimulatory effect was also observed with cell lines which lack any contaminating immunoreactive cells, there is strong evidence that interferons can directly stimulate the proliferation of clonogenic tumor cells in vitro. Growth stimulation by interferons occurred preferentially with lower dosages. It is important to be cognizant of potential clinical implications of tumor growth stimulation by interferons.

Modulation of tumour colony growth by irradiated accessory cells

The ability of human tumour cells to form colonies in soft agar is enhanced by the presence of autologous phagocytic/adherent cells. We investigated the effect of irradiation on the ability of the adherent cells to support human tumour colony formation. Relatively low doses of irradiation significantly increased the growth enhancing ability of adherent cells in 17/19 cases. The possibility that the enhancement was mediated by inactivation of radiosensitive contaminating lymphocytes was explored. Depletion of T lymphocytes from unirradiated adherent cells by a monoclonal antibody and complement resulted in little overall change in tumour colony growth. However, elimination of only the suppressor subset (OKT8+) of T lymphocytes resulted in increased colony growth relative to control values obtained with unirradiated adherent cells. In contrast, depletion of T lymphocytes from irradiated adherent cells by a pan T monoclonal antibody and complement decreased colony formation. Thus, the ability of irradiated macrophages to enhance tumour colony growth appeared to be mediated by a T lymphocyte. The effect of irradiation on isolated populations of macrophages and T lymphocytes was also examined. The enhanced ability of irradiated adherent cells to support tumor colony growth appeared to have been due to treatment of T lymphocytes alone. The results indicate that both adherent macrophages and lymphocytes may influence the growth of clonogenic human tumour cells.

Regulation of the immune response by prostaglandins

Of all the arachidonic acid metabolites, only prostaglandin E (PGE) has been shown to have a clear role in the regulation of cellular and humoral immune responses. In cellular immune responses such as T cell proliferation, lymphokine production, and cytotoxicity, PGE usually acts as a feedback inhibitor of the response. This is also true of macrophage and natural killer cytotoxicity. In some instances PGE is responsible for cellular activation rather than inhibition. This is clearest in the control of humoral immunity, where PGE production is a necessary component in the generation of some type of T suppressor cells. Disturbances in immune function found in several human conditions and diseases have been linked to changes in PGE mediated immunoregulation. Either increased production of PGE or increased sensitivity to PGE results in depressed cellular immunity. Conversely drugs which inhibit PGE production act as stimulants of cellular immune function in vitro and in vivo.

Effect of epidermal growth factor on clonogenic growth of primary human tumor cells

The effect of EGF on the soft agar colony-forming ability of fresh human tumor cells was assessed in 40 specimens obtained from various types of carcinoma including those of the breast, endometrium ovary, and other sites. Cells from four established human tumor cell lines (three breast and one endometrial) were also included in this study. The results showed that addition of EGF at a concentration of 50 ng/ml resulted in a 50% higher cloning efficiency in soft agar in 40% of the samples of fresh human tumors. When cells from tumor cell lines were plated in semi-solid medium containing EGF, the number of colonies formed was at least twice as high as controls. Cells from fresh tumor biopsies were assayed for EGF receptors to determine whether the correlation between the proliferative response in EGF-supplemented semi-solid medium as compared to control could be related to the number of EGF receptors present on the cells. Specific receptors for EGF were detected by using radioiodinated EGF in early-passage cell cultures from some of the tumors tested for clonogenicity. The number of receptors ranged from 0.3 to 3.27 X 10(5) per cell. Cells from two melanoma specimens possessed less than 0.3 X 10(3) EGF receptors per cell. We found no correlation between the number of EGF receptor on a cell surface and the mitogenic effect of EGF on the same tumor cells grown in semi-solid medium.

Limitations of the clonal agar assay for the assessment of primary human ovarian tumour biopsies

114 biopsy specimens from 70 patients with ovarian carcinoma at all stages of disease were submitted for assessment of clonogenic capacity in agar. A highly significant correlation was found between agar clonogenicity and patient survival after biopsy. However, problems related to inherent tumour heterogeneity, quality of sample and tissue disaggregation indicate that this technique may have limited applicability in the routine assessment of patients. Only 41 biopsy specimens (36%) from 31 patients (44.3%) complied with the prerequisite criteria for agar clonogenic assessment, namely: (a) the confirmed presence of malignant cells in the biopsy, (b) the ability to prepare a single-cell suspension, and (c) adequate viable cell numbers for assay. Furthermore, although the dominant patterns of agar clonogenic growth could be identified and correlated with stage of disease, the heterogeneity in both initial clonogenic capacity and "self-renewal" capacity assessed by the ability of primary clones to propagate in liquid culture and reclone in agar was too inconsistent for the assay to be used as a prognostic index for the individual patient.

Measurement of self-renewal in culture of clonogenic cells from human ovarian carcinoma

To test the identity of human tumour clonogenic cells and stem cells, a procedure was developed to allow quantitation of self-renewal capacity of human ovarian carcinoma clonogenic cells. Primary colonies grown from malignant effusions of 10 patients were disaggregated and replated; secondary colonies were observed to be similar to primary colonies in size, morphology and culture requirements. Density-gradient separation of tumour-cell populations demonstrated that not all primary clonogenic cells are capable of self-renewal during clonal expansion. Patient-to-patient variation in self-renewal capacity was shown to be significantly correlated with the concentration of the tumour-cell population in the effusion fluid, and preliminary evidence of a progressive increase in self-renewal was found in one patient. It was concluded that some, but not all, ovarian-tumour clonogenic cells have the stem-cell property of self-renewal, and that quantitation of such a property may identify an important prognostic variable.

Tumoricidal activity of macrophages isolated from human ascitic and solid ovarian carcinomas: augmentation by interferon, lymphokines and endotoxin

Peripheral blood monocytes and macrophages from ascitic or solid tumors were obtained from 50 patients with advanced (Stage III or IV) ovarian carcinoma. Blood monocytes and peritoneal macrophages from 76 patients undergoing surgery for benign gynecological diseases were used as controls. Macrophages were 30% (range 0.2-6.2) and 2% (range 0.2-4) of mononuclear cell suspensions from ascitic and solid tumors, respectively. Cytolytic activity was measured as release of [3H]-thymidine from prelabelled mKSARU5 (TU5) target cells at an effector to target cell ratio of 20:1. Peripheral blood monocytes and tumor-associated macrophages had baseline cytotoxicity lower than or similar to that of control peripheral blood and peritoneal macrophages respectively. In vitro exposure to partially purified human fibroblast interferon, lymphokine supernatants or endotoxin augmented the cytotoxicity of mononuclear phagocytes, and no significant difference was observed between control and ovarian cancer effector cells from peritoneal effusions. In four patients, macrophages were isolated from the solid tumor or from the malignant effusion and tested in parallel: in two out of four subjects, macrophages from the solid neoplasm showed defective cytotoxicity compared to ascites effector cells from the same subject. When primary ovarian carcinoma cultures were used as targets, tumor cells from two out of five patients tested were relatively resistant to activated macrophage cytotoxicity. Thus, tumor-associated macrophages from ascitic or solid ovarian neoplasms showed no evidence of activation, in terms of spontaneous cytotoxicity or responsiveness to stimuli: it appears unlikely that small numbers of non-activated macrophages present within primary ovarian carcinomas act as a significant mechanism of restraint of neoplastic growth, at least at this anatomical site.

I. Tumor growth in mice with depressed capacity to mount inflammatory responses: possible role of macrophages

Studies were made of the effects of various treatments on the growth in mouse feet of isografts of two methylcholanthrene-induced fibrosarcomas: C-4, of CBA/J mice, and A-2, of A/J mice. The isografts were prepared by pronase digestion of subcutaneous tumors and were injected as unseparated cell suspensions or as tumor-cell-enriched suspensions after depletion of infiltrating host inflammatory cells. The recipient mice were untreated or treated with reserpine, sublethal whole body irradiation, cyclophosphamide, or corticosteroids. Depletion of host cells from the inoculum resulted in increased growth from the same number of tumor cells. Reserpine treatment decreased the growth of both tumors, whether unseparated or tumor-cell-enriched, and whether injected into the foot or the flank. Irradiation, cyclophosphamide pretreatment, and corticosteroid pretreatment decreased the growth of normal inocula or enriched inocula or both. The effects of cyclophosphamide and corticosteroids were apparently not due to cytotoxicity to tumor cells. Normal resident peritoneal cells increased tritiated thymidine uptake by tumor cells in vitro. Sedimentation velocity separation showed the largest cells to be the most potent. It is suggested that some hot inflammatory reaction is necessary for optimal tumor growth and that murine hosts produce not only cells with antitumor effects but also cells, possibly a subpopulation of macrophages, that potentiate tumor growth.

In vitro effects on tumor cells of macrophages isolated from an early-passage chemically-induced murine sarcoma and from its spontaneous metastases

Macrophages were isolated by adherence from an early-passage chemically-induced tumor in C57BL/6 mice and from its spontaneous lung metastases. Cytolytic activity was measured as release of [3H]-thymidine from prelabelled tumor cells and cytostasis in a postlabelling assay. Normal, unstimulated peritoneal macrophages exhibited low but significant non-specific cytotoxic activity at attacker to target (A:T) ratios greater than or equal to 20:1, whereas stimulation of tumor-cell proliferative capacity was consistently observed at A:T ratios less than or equal to 2.1, Macrophages from the peritoneal cavity of tumor-bearing mice, from the primary tumor or from polled secondaries had baseline cytotoxic or growth-enhancing (depending on the A:T ratio) potential similar to that of control peritoneal macrophages. In vitro exposure to endotoxin, lymphokine supernatants or partially purified fibroblast interferon augmented the tumoricidal activity of normal and tumor-associated macrophages; tumor-associated macrophages showed no evidence of enhanced responsiveness to these activating stimuli. Tumor cells from the primary neoplasm and its spontaneous metastases were equally susceptible to macrophage tumoricidal activity. It is inferred from in vitro data that the relatively low numbers of non-activated macrophages present within poorly immunogenic metastasizing tumors may have no direct effect or actually stimulate tumor-cell proliferative capacity at the primary tumor site; conversely, disseminating tumor cells might encounter sufficient numbers of mononuclear phagocytes in the circulation and in lungs to permit the expression of macrophage cytotoxicity, at least during the early steps of metastasis formation.