Human melanoma cell invasion is inhibited in vitro by swainsonine and deoxymannojirimycin with a concomitant decrease in collagenase IV expression

A 48 h pretreatment of two malignant and invasive human melanoma cell lines with either swainsonine (an inhibitor of Golgi alpha-mannosidase II) or deoxymannojirimycin (a Golgi alpha-mannosidase I inhibitor) resulted in a dose-dependent decrease in the cells' ability to invade a reconstituted basement membrane in vitro. This effect was reversible within 48 h of removing the drugs. Treatment with either drug resulted in both cell lines being more resistant to the cytotoxic effects of the lectin leukoagglutinin (PHA-L) and more sensitive to the lectin concanavalin A which indirectly indicated a change in the cell surface oligosaccharide composition and structure consistent with the known effects of these drugs on N-linked oligosaccharide processing. A 25-33% decrease was noted in the adhesion of treated cells to either a reconstituted basement membrane or human umbilical vein endothelial cell monolayer while no change was measured in the cells' proliferative rates. A correlative decrease was observed, however, in the expression of human type IV collagenase mRNA which was recovered within 48 h of removing the drugs. These results suggest that a correlation exists between the drug-induced changes in the cell surface oligosaccharide composition and structure with a concomitant decrease in the mRNA and secreted levels of type IV collagenase and the ability of these cells to invade.

Characterization of a highly invasive and spontaneously metastatic human malignant melanoma cell line

Although the incidence of, and deaths due to, malignant melanoma are rising at a rapid rate, few experimental models mimic the highly metastatic properties associated with the pathogenesis of the human disease, making study of the disease difficult. Thus, new human models are required to understand melanoma biology, especially its metastatic properties. Here we describe C8161, a highly invasive and spontaneously metastatic human melanoma cell line, which grows progressively in the subcutis of athymic nude mice with an average doubling time of approximately 6 days. By the time the tumor reaches a diameter of 1 cm, amelanotic metastases in lymph nodes, skin, peritoneal wall, spleen and lungs have formed. By comparing C8161 to variants from other well-characterized human malignant melanomas (A375 and MeWo) with differing metastatic traits, properties presumed to be involved in metastatic propensity were examined. C8161 showed a 2- to 14-fold higher ability to invade reconstituted basement membrane barriers in the MICS and correspondingly high type-IV collagenase mRNA levels and collagenolytic activity, as compared with other melanoma cell lines. Likewise, differential adhesion to immobilized RBM or HUVEC monolayers was observed, but did not correlate to rank orders of malignant properties. Recently, a correlation between surface expression of ICAM-1 and secondary tumor formation by human melanomas has been described in several laboratories. Basal levels of ICAM-1 on C8161, A375 and MeWo human melanomas were compared, but no correlation with metastatic potential was noted. Proto-oncogene expression in C8161 cells was compared with A375P and A375M variants using Northern blot analysis. c-myc expression was 6-fold greater than both A375 variants; c-fos expression was 3.4-fold less than A375P and 1.7-fold less than A375M; c-jun in C8161 cells was 2.5-fold and 2.1-fold greater than expression in A375P and A375M, respectively. Because C8161 is so highly malignant, amenable to experimental manipulation, and its behavior in nude mice mimics the clinical course of malignant melanoma, this cell line will prove valuable for studying properties associated with human melanoma tumor progression.

Tumor progression- and metastasis-associated proteins identified using a model of locally recurrent rat mammary adenocarcinomas

A recently established model for local breast cancer recurrence using the 13762NF rat mammary adenocarcinoma was used to evaluate biologic and biochemical properties related to clinical outcome for this class of tumors. Sublines isolated from local tumor regrowths following surgical resection differed from each other and from the 'parental' cell lines for multiple phenotypes, including metastatic propensity. Local recurrence- and primary tumor-derived sublines were examined by two-dimensional polyacrylamide gel electrophoresis (2D-PAGE), lectin binding to electrophoretically separated proteins, and lactoperoxidase-catalyzed cell surface iodination; and differential protein patterns were compared to tumor progression and metastatic potential. 2D-PAGE revealed several quantitatively different spots which correlated with lung colonization potential. In particular, quantities of an apparently unique, non-cell-surface protein, P50.9 (Mr approximately 50,900, pI approximately 7.3) correlated inversely with metastatic propensity, suggesting that it may be associated with, among other possibilities, the negative regulation of the metastatic phenotype. P50.9 was unrelated to four similarly sized metastasis-associated proteins--tumor autocrine motility factor; the rat analog of tumor suppressor, p53; rat cytokeratin 14 or procathepsin D--as determined by amino acid analysis. A major wheat germ agglutinin binding sialoglycoprotein, gp93 (Mr approximately 93,000), was present in smaller amounts as cells were passaged in vivo and re-established as in vitro cultures [MTF7 greater than 'primary' tumor-derived lines (sc1, sc3) much greater than local recurrence-derived lines (LR1, LR1a, LR3, LR4, LR5, LR6)]. Besides cell surface glycoprotein losses, two of six local recurrence-derived sublines expressed a wheat germ agglutinin-binding sialoglycoprotein, gp110 (Mr approximately 110,000), previously undetected on any of the other cell lines including the parental populations. gp110 was found in LR3 and LR6 which were relatively highly metastatic; however, correlation with metastatic potential failed because gp110 was not present on the metastatic parental cell line, MTF7. These results demonstrate specific quantitative and qualitative protein differences associated with the selection of locally recurrent mammary tumors.

Transforming growth factor beta stimulates mammary adenocarcinoma cell invasion and metastatic potential

The experimental metastatic potential of 13762NF mammary adenocarcinoma clone MTLn3 was tested after pretreatment in serum-free medium containing transforming growth factor (TGF) beta 1 at 0-5000 pg/ml. Lung colonies were measured 2 weeks after inoculation in syngeneic F344 rats, and a bell-shaped dose-response curve with 2- to 3-fold increase in number of surface lung metastases was seen. Maximal enhancement occurred at the 50 pg/ml dose level. The effect was specific because addition of neutralizing anti-TGF-beta antibody blocked the stimulatory activity at all levels of TGF-beta 1 pretreatment, but when antibody was given alone, neutralizing anti-TGF-beta antibody had no effect on untreated cells. Increased metastatic potential appears to be from an increased propensity of cells to extravasate as tested in the membrane invasion culture system. MTLn3 cells penetrated reconstituted basement-membrane barriers 2- to 3.5-fold more than did untreated control cells, depending upon length of TGF-beta 1 exposure. Increased invasive potential is apparently due, in part, to a 2- to 6-fold increase in type IV collagenolytic (gelatinolytic) and a 2.4-fold increase in heparanase activity. TGF-beta 1 treatment of MTLn3 cells did not alter their growth rate or morphology in the presence of serum; however, growth was inhibited in serum-free medium. Likewise, adhesion to human umbilical vein endothelial cell monolayers or to immobilized reconstituted basement membrane or fibronectin matrices was unchanged. These results suggest that TGF-beta 1 may modulate metastatic potential of mammary tumor cells by controlling their ability to break down and penetrate basement-membrane barriers.

Quantitation of human melanoma, carcinoma and sarcoma tumor cell adhesion to lymphatic endothelium

We have used an in vitro adhesion assay to study the interaction of tumor cells with lymphatic endothelium, a dynamic event that leads to tumor metastasis in vivo. 3H-thymidine-labeled human tumor cells from: one primary Ewing sarcoma, two established melanoma cell lines, two colon and two breast carcinomas (one established line and one primary culture of each) were added to 24-well culture dishes containing confluent monolayers of bovine lymphatic endothelium. Radioactivity associated with either the cells in suspension or the attached cells was assessed and compared at frequent intervals up to 360 minutes. Generally, tumor cell attachment increased as a function of time reaching a plateau between 180 and 360 minutes. the modular media system described here facilitates the primary and secondary culture (or co-culture) of a variety of normal and transformed cells. Primary cultures with a rounded morphology (one breast and one colon carcinoma) showed the lowest preferential attachment for lymphatic endothelium. All established cell lines and the primary Ewing sarcoma cell line displayed a more fibroblastic morphology and achieved the highest adhesion profiles. There was a correlation between the malignancy and attachment potential for the melanoma and breast carcinoma cell lines. Collectively, these data show that established tumor cell lines with fibroblastic-like morphology exhibit more rapid adhesion than primary tumor cell cultures with more rounded morphologies. While this property may reflect in vitro selection and/or adaptation, it does correlate with the metastatic propensity for some human tumor cells.

Tumor-elicited polymorphonuclear cells, in contrast to "normal" circulating polymorphonuclear cells, stimulate invasive and metastatic potentials of rat mammary adenocarcinoma cells

Circulating polymorphonuclear cell (PMN) levels rise in proportion to the metastatic potential of the tumor in 13762NF mammary adenocarcinoma tumor-bearing rats. These tumor-elicited PMNs (tcPMNs) secrete high levels of the basement-membrane-degrading enzymes, type IV collagenase and heparanase, suggesting that metastatic tumor cells stimulate neutrophilia so that the tcPMNs might assist tumor cell extravasation during metastasis. To test this hypothesis, purified proteose peptone-elicited PMNs from peritoneal exudate, circulating normal PMNs, and tcPMNs were evaluated for their effects on in vitro invasive and in vivo metastatic potentials of syngeneic 13762NF mammary adenocarcinoma tumor cells. tcPMNs caused a dose-dependent increase in invasion through a reconstituted basement membrane barrier in an in vitro invasion assay. At PMN:tumor cell ratios of 30:1, invasion potential significantly (P less than 0.05) rose to 26-fold, 40-fold, and 37-fold for poorly metastatic MTLn2 cells, highly metastatic MTLn3 cells, and moderately metastatic MTF7 cells, respectively. In contrast, purified proteose peptone-elicited PMNs and circulating normal PMNs did not significantly alter invasive potential. Intravenous coinjections of purified proteose peptone-elicited PMNs did not change the number of experimental lung metastases, but tcPMNs at ratios to 50:1 significantly raised the mean number of metastases 23-fold for MTLn2, 3- to 4-fold for MTLn3, and 1.6- to 1.8-fold for MTF7. These results demonstrate that tcPMNs contribute to the metastatic propensity of mammary adenocarcinoma clones by increasing efficiency of invasion through basement membrane.

Experimental model for locally recurring mammary tumors. Development, morphology, karyotype, growth kinetics, and experimental metastatic potential

A rat model was established for evaluating the biology of locally recurring mammary tumors after surgical resection of the primary tumor. Eight distinct cell lines were independently derived from primary tumors and local recurrences after surgical removal of 13762NF rat mammary adenocarcinoma clone MTF7(T20). In vivo tumor doubling times between the "parental" MTF7(T20) cell line, primary tumor-derived cell lines sc1 and sc3, and the local recurrence (LR) sublines varied after the inoculation of 10(6) tumor cells into the mammary fat pad of female Fischer 344 rats. Doubling times were shorter for LR3, and LR4, LR5, and LR6 than their primaries sc3 and MTF7(T20), respectively, and longer for LR1 and LR1a than their primary tumor sc1. The LR sublines varied considerably for their experimental metastatic potentials. Both increases and decreases in metastatic potential were seen compared to MTF7(T20), sc1, and sc3. Karyotype analysis by G-banding revealed the presence in the LR sublines of several marker chromosomes, previously identified in MTF7 at tissue cultures 11 and 35. Two new chromosome markers were identified: M54, shared by MTF7(T20), sc1, LR4, LR5 and LR6, and M55, shared by MTF7(T20), sc1, LR1, sc3, LR3, LR4, and LR6. These data indicate that local tumor regrowth after surgical excision of the primary tumor in this model most likely selects the growth of tumor cell subpopulations already present within the primary tumor. Differences in growth kinetics, karyotype, and metastatic potential between the parental MTF7(T20), primary tumors sc1 and sc3, and their LR sublines may reflect in vivo influences on the phenotypic diversity generated during the development of local mammary tumor recurrences after surgical treatment of the primary tumor.

Use of the Membrane Invasion Culture System (MICS) as a screen for anti-invasive agents

The Membrane Invasion Culture System (MICS) assay was adapted for relatively rapid screening of compounds and used to identify anti-invasive drugs that inhibit human and murine tumor cell migration through a reconstituted basement membrane in vitro. Cell lines demonstrating low and high invasive and metastatic potentials were tested with all compounds for tumoricidal effects prior to evaluation in MICS at non-cytotoxic doses. The effect on invasive potential in the MICS assay was determined in 3 categories: (1) 48 hr drug pre-treatment prior to seeding in the MICS (exceptions: 90 min pre-treatment with pertussis toxin and, for some studies, continuous exposure for 2-7 days); (2) peptide or prostaglandins 2 hr after seeding and attachment to the membranes in MICS followed by continuous exposure; and (3) cells receiving neither drug nor peptide treatment and serving as controls in each MICS chamber. Since invasion involves cellular motility and deformability, some cytoskeleton disrupting agents were selected. Of these, vincristine, colcemid and colchicine inhibited invasion but taxol did not. Pre-treatment with cAMP agonists produced conflicting results: dibutyryl cAMP and 8-(4-chloro-phenylthio) cAMP resulted in 50% and 38% reduction in invasion, respectively, whereas 8-bromo cAMP stimulated invasive potential by 30%. Forskolin and cholera toxin both significantly reduced invasiveness. Pre-treatment with 5-azacytidine and araC, to consider the role of methylation and proliferations decreased invasive ability. Anti-metastatic drugs such as gamma-interferon and razoxane inhibited invasive potential but to varying degrees. Treatment of cells with prostaglandins E2, F2 alpha, A2, and D2 were ineffectual; however, indomethacin mildly inhibits invasion (less than 30%).(ABSTRACT TRUNCATED AT 250 WORDS)

Sensitivity of locally recurrent rat mammary tumour cell lines to syngeneic polymorphonuclear cell, macrophage and natural killer cell cytolysis

Using a recently developed model for studying the biology of locally recurrent (LR) mammary tumours in the 13762NF rat mammary adenocarcinoma system, we examined the sensitivity to polymorphonuclear cell, macrophage and natural killer cell cytolysis. The parental MTF7(T20) cell line; the 'primary' tumours which arose following subcutaneous inoculation into the mammary fat pad, sc1 and sc3; and the local recurrences (following surgical excision) LR1 and LR1a from sc1, and LR3 from sc3 were all cells generally resistant to specific PMN cytolysis. LPS-activated macrophages caused 25.1%, 38.7% and 58.8% specific cytolysis in MTF7, sc1 and LR1 cells, respectively at E:T of 20:1 and 72 h co-incubation. LR1a, sc3 and LR3 lysis ranged from 0-4.4% under the same conditions. Non-activated macrophages did not lyse any of the cell lines. Locally recurrent and 'primary' tumour cell lines were also not lysed by naive NK cells (range 0.5-4.0% cytolysis). NK cells activated with bropirimine, a potent immunomodulator currently being studied in clinical trials, and/or interleukin-2 were mildly more effective at killing LR cells. Our results show that locally recurrent tumours exhibit heterogeneous sensitivities and are different from 'primary' tumour cells in sensitivities to immune cell killing, but they are not necessarily more or less sensitive. Results with bropirimine-activated or IL-2-activated NK cells emphasize that nonspecific activation is insufficient to eliminate all tumour subpopulations.

The role of polymorphonuclear leukocytes (PMN) on the growth and metastatic potential of 13762NF mammary adenocarcinoma cells

Circulating neutrophil (PMN) levels can increase in rats bearing subcutaneously growing clones of the 13762NF mammary adenocarcinoma and the level of increase correlates with the metastatic potential of the clone. In rats with poorly metastatic MTC tumors, numbers of circulating PMN did not rise, whereas PMN levels rose 50-fold in rats bearing highly metastatic MTLn3, 12-fold in rats with weakly metastatic MTLn2, and 14-fold in those with moderately metastatic MTF7 tumors. Neutrophilia was caused partly by tumor size, but metastatic potential was a stronger determinant, suggesting that PMNs may play a role in the metastatic process. To determine whether circulating PMNs indeed contribute to cellular metastatic potential, we examined effects of PMN on various aspects of the metastatic process. Experimental metastasis assays involving i.v. co-injections of PMNs yielded a dose-dependent increase in extrapulmonary metastases for MTLn3, but no change in lung colonization potential for any of the clones examined. The change in the metastatic profile was not due to any modification in in vivo distribution of i.v. injected tumor cells or in adhesion to endothelial monolayers in vitro. PMNs also had no effect on in vitro DNA, RNA or protein synthesis and were not cytolytic (E:T 100:1). However, PMNs collected from high-passage MTLn3 tumor-bearing rats had a 50% increase in heparanase and type-IV collagenolytic activity as compared to unstimulated PMNs isolated from normal rats. These results indicate that polymorphonuclear cells may contribute to the metastatic potential of highly metastatic clones from the 13762NF mammary adenocarcinoma cells by assisting in the degradation of basement membrane during extravasation.

Growth of rat mammary adenocarcinoma cells in semisolid clonogenic medium not correlated with spontaneous metastatic behavior: heterogeneity in the metastatic, antigenic, enzymatic, and drug sensitivity properties of cells from different sized colonies

Using rat 13762NF mammary tumor cell clones of varying spontaneous metastatic potentials and biochemical properties and known phenotypic stabilities we studied the relationship between cell colony growth in a clonogenic assay and the biological and biochemical properties of cells derived from different cell colonies. The spontaneous metastatic potential of in vivo or in vitro grown 13762NF tumor cells was not related to their in vitro cloning efficiencies; cells of both low and high metastatic potential formed colonies of various sizes and shapes during 14 days of growth in 0.3% or 0.6% semisolid agarose. A highly metastatic cell clone of relatively low growth potential in agarose was examined further. Individual tumor cell colonies derived from this cell clone were removed from agarose and their properties determined. Cells from small (less than 100-microns-diameter) or large (greater than 500-microns-diameter) agarose colonies had similar self-renewal capacities in agarose and formed variously sized cell colonies when replated in agarose medium. Metastatic potential, drug sensitivity parameters, and expression of a high Mr mucin-like glycoprotein antigen and type IV collagenolytic activity known to be associated with spontaneous metastasis of 13762NF tumor cells were dissimilar in cells from different colonies, and these characteristics were independent of original tumor cell colony size in agarose. In contrast, the expression of cell surface proteins of Mr less than 300,000 were similar among cells derived from different agarose colonies. The data indicate that heterogeneity exists in the ability of 13762NF adenocarcinoma cells of different biochemical and metastatic potentials and drug sensitivities to grow in semisolid agarose. In addition, the cells that grow in agarose to form detectable colonies (greater than 50 cells) are not necessarily those with a high potential of metastasizing spontaneously to distant sites.

Biologic considerations for drug targeting in cancer patients

There is a great need for more specific targeting of chemotherapeutic agents, but development of specific therapy will be difficult in light of the barriers which separate a tumor from the vasculature, tumor cell heterogeneity and instability, technological advances necessary for drug delivery design and introduction into the clinic and mechanisms for assessment of efficacy (Table 5). While imposing, these problems are not insurmountable. A clearer understanding of the specific goal of a drug delivery/drug targeting approach will make the expectations more realistic and the chances for success greater. Primary to improving drug targeting is a better understanding of the biology of tumors. There are a number of limitations to drug targeting technology mentioned above, but, at present, the more difficult limitations are imposed by tumors themselves and the host's response to a tumor. Currently available technology does not offer Erlich's magic bullet and it does not appear that a single entity will suffice for cancer. Perhaps a mixture of drug delivery systems, each reducing the toxicity of a particular component of therapy will provide the first realistic goal in drug targeting. Each form of drug targeting is limited by biochemical and biophysical properties of the host, tumor cells, drug delivery system and interactions between them. Successes in vitro are moot without corresponding data in the (unfortunately) more complex organismic level. Development of drug targeting approaches will also require a critical, thorough evaluation of the ability to specifically deliver drug in vivo.

Degradation of basement membrane type IV collagen and lung subendothelial matrix by rat mammary adenocarcinoma cell clones of differing metastatic potentials

A series of rat 13762NF mammary adenocarcinoma cell clones and subclones of various lung metastatic potentials were examined for their abilities to degrade rat lung subendothelial matrix and purified basement membrane type IV collagen. Metastatic potentials were simultaneously determined based on the ability to form "spontaneous" lung metastases after s.c. injection or "experimental" lung metastases after i.v. injection of cells. Microvessel endothelial cells isolated from rat lung were grown in vitro, and the subendothelial matrix containing type IV collagen was metabolically labeled with [3H]proline. When mammary adenocarcinoma cells were placed on the isolated subendothelial matrix, fragmentation and solubilization of [3H]proline-labeled components were observed; highly metastatic 13762NF cells solubilized the matrix at higher rates than did poorly metastatic cells. The 13762NF cells were assayed for type IV collagenolytic activity using [3H]proline-labeled type IV collagen purified from Engelbreth-Holm-Swarm tumor as a substrate. We found excellent correlation between the type IV collagenolytic activities of living cells and their "spontaneous" lung metastatic potentials (r = 0.993). The levels of type IV collagenolytic activity in the conditioned medium depended on the cell culture conditions. In the presence or absence of acid-treated fetal bovine serum, highly metastatic cells secreted higher amounts of type IV collagenolytic enzymes in active and latent forms than did poorly metastatic cells. Incubation of procollagen type IV with medium conditioned by highly metastatic 13762NF cells and treated with trypsin resulted in the production of several large fragments characteristic of type IV collagen. The results suggest that enzymatic degradation of basement membrane type IV collagen is important in lung metastasis of 13762NF mammary adenocarcinoma cells.

Rat mammary adenocarcinoma heat-stress proteins in vivo

Rat 13762NF mammary adenocarcinoma cells (clone MTC) were heated to 42 degrees C either in vivo as a subcutaneous tumour in the rat mammary fat pad or in vitro as attached cells. Labelling in vivo or in vitro detected very similar heat-stress proteins (hsp) at 160, 112, 90, 70 and 56 kDa. Syngeneic rat endothelial and macrophage cells synthesized several cellular proteins in vitro differently than did the tumour cells in vitro, but both types of normal cells were similar to tumour cells in the hsp synthesized. Although the quantitative aspects of induction and repression of hsp may depend on cell type and microenvironment, the major tumour hsp being studied for function in vitro were qualitatively similar to those produced and labelled in vivo in response to a similar heat dose. Hsp were similar in both normal cells and tumour cells from the same host. These observations support the concept that hsp function in fundamental processes in the different microenvironmental and metabolic conditions found in vivo and in vitro. In addition, these observations suggest that prediction of tumour thermal response by measuring hsp levels may be influenced by host cell components.

Heat stress proteins and experimental cancer metastasis

The influence of stress protein synthesis and thermotolerance on blood-borne (experimental) metastatic potential was examined in the 13762NF rat mammary adenocarcinoma model. Cloned cell populations with highly reproducible and well-defined metastatic potential were treated by hyperthermia and sodium arsenite to induce a complete set of stress proteins and thermotolerance with minimal cell killing. The influence of these in vitro treatments on subsequent experimental metastasis in vivo was determined for location, frequency, size distribution and volume. Metastatic tumour burden generally decreased following induction of increased heat or arsenite stress proteins and thermotolerance; however, there was no evidence for altered size distribution or location of metastatic lesions.

Multiple phenotypic divergence of mammary adenocarcinoma cell clones. II. Sensitivity to radiation, hyperthermia and FUdR

We have shown that, with in vitro passage, subclones derived from clonal cell populations of 13762NF mammary adenocarcinoma undergo phenotypic drift and diversification in their cellular properties. Here we examine whether phenotypic divergence of 13762NF cell clones extends to therapeutic treatments used in eliminating mammary tumors and whether the apparent rates of phenotypic divergence vary for different treatments. Six subclones isolated from low passage clone MTF7 (T11; tissue culture passage 11) cells were compared to a similar number of subclones isolated from high passage clone MTF7 (T35; tissue culture passage 35) cells. Subclones derived from clone MTF7 (T11) were relatively homogeneous (not significantly different) in their inherent sensitivities to ionizing radiation, extrapolation coefficients and quasithreshold dose values (Do = 1.61-1.99 Gy; n = 0.89-3.42; Dq = 0-2.34). When the MTF7 (T11) subclones were examined for their sensitivities to 45 degrees C hyperthermic treatment, the inherent sensitivities and dose-response curve parameters (Do = 5.24-10.05 min; n = 1.08-10.47; Dq = 0.78-12.31) were heterogeneous (significantly different). In addition, the MTF7 (T11) subclones were heterogeneous (significantly different) in their sensitivities and dose-response curve parameters to 5-fluoro-2'-deoxyuridine (FUdR) treatment (slope = -0.70 to -1.59; y-intercept = 1.31 X 10(2) to 47.80 X 10(2]. The LD50 values for FUdR ranged from 14-150 nM for the MTF7 (T11) subclones. At high passage MTF7 (T35) subclones were heterogeneous in their dose-response parameters to ionizing radiation (Do = 1.17-2.05 Gy; n = 0.80-41.18; Dq = 1.79-4.94), hyperthermia (Do = 3.57-6.32 min; n = 2.08-13.54; Dq = 3.68-9.30) and FUdR (slope = -0.77 to -0.93; y-intercept = 4.64 X 10(2) to 8.83 X 10(2); LD50 = 50-160 nM). The results indicate that clonal cells diverge for distinct phenotypic properties at differing rates to form heterogeneous cell populations with unique sensitivities to various therapeutic treatments.

Multiple phenotypic divergence of mammary adenocarcinoma cell clones

The properties of cell clones derived from locally growing and spontaneous metastases of 13762NF mammary adenocarcinoma change during in vitro growth. This has been termed phenotypic drift and is reproducible in independent experiments using different cryoprotected cell stocks. To determine whether phenotypic drift in 13762NF cell clones is the result of an en bloc shift in the properties of all tumor cells, or independent phenotypic divergence of tumor cells to produce a mixed cell population, local tumor-derived clone MTF7 was subcloned at low and high culture passage numbers in vitro. Each subclone was analyzed in vitro for cell morphology, growth rate, saturation density, karyotype and ploidy, and in vivo for experimental metastatic behavior. Subclones derived from low passage clone MTF7 (T11; tissue culture passage number 11) were relatively homogeneous in their growth rates (doubling times of 16.8-17.4 h) and saturation densities (approximately 2 X 10(5) cells/cm2); yet, these same subclones were heterogeneous in their in vitro cell morphologies, experimental metastatic potentials (means range from 0 to greater than 100 tumor nodules per lung), size distributions of lung tumor nodules, marker chromosomes and modal chromosome numbers. High passage MTF7 (T35; tissue culture passage number 35) subclones had similar growth rates and saturation densities, except for subclone 2, which had a doubling time of approximately 26 h. Cell morphologies, experimental metastatic potentials (means range from 3 to greater than 600 tumor nodules per lung), size distribution of lung tumor nodules, marker chromosomes and modal chromosome numbers varied between MTF7 (T35) subclones. The results suggest that simultaneous, independent divergence of several phenotypes from a single cloned cell occurred to form a mixed cell population containing cells with independently segregated, unrelated phenotypes. Thus, the reproducibility of phenotypic drift in clonal cell populations was probably the result of tumor cell divergence and was not an en bloc shift in phenotypic properties of all cells.

Chromosome and DNA analyses of rat 13762NF mammary adenocarcinoma cell lines and clones of different metastatic potentials

Chromosome morphologies revealed by Giemsa-banded karyotypes and chromosome numbers were compared between parental tumor-, lymph node- and lung metastasis-derived rat 13762NF mammary adenocarcinoma cell lines and clones having different spontaneous metastatic potentials. Although chromosome numbers in the cell lines and clones generally correlated with DNA content by flow cytometry, ploidy did not correlate with spontaneous metastatic potentials. Chromosome number and DNA content drifted during prolonged in vitro growth in each of the cell lines and clones. Common chromosome rearrangements were found, confirming a common origin for all the cell lines and clones, and the frequency and appearance of the individual marker chromosomes fluctuated during in vitro growth. Karyotypic analyses revealed that the markers coinciding with phenotypic drift in spontaneous metastatic potential and other biological properties of parental tumor-derived clones MTC and MTF7 and lung metastasis-derived clone MTLn3 involved chromosomes 3, 4, 5, 6, and 8. Clone MTC exhibited a shift in several markers and an increase in metastatic potential at passage T20, while clone MTF7 displayed a lesser spontaneous metastatic potential at high passage (T34) concomitant with an increase in the frequency of certain marker chromosomes. Lung metastasis-derived clone MTLn3 also exhibited a shift in some marker chromosomes, colonization preference and metastatic potential to lung and lymph nodes at high tissue culture passages. The changes in marker chromosomes during in vitro passage of clones MTC and MTLn3 suggested the presence of at least two cell subpopulations which could be responsible for the observed shift in spontaneous metastatic properties. Karyotypic features of the 13762NF cell lines and clones indicate that subtle cytogenetic changes, in contrast to gross chromosomal abnormalities, may be more important in determining metastatic phenotype.

Phenotypic drift and heterogeneity in response of metastatic mammary adenocarcinoma cell clones to adriamycin, 5-fluoro-2'-deoxyuridine and methotrexate treatment in vitro

Rat 13762 mammary adenocarcinoma cell clones of differing spontaneous metastatic potentials were tested for their sensitivities to Adriamycin, 5-fluoro-2'-deoxyuridine (FUdR) and methotrexate in vitro. Cells were treated 4 hours with a single dose of drug, and colony formation was used to assay cell survival. Dose-response curves and survival parameters were calculated for local tumor-derived clones MTC and MTF7 and lung metastasis-derived clone MTLn3. The logarithmic curves were analyzed (slope and y-intercepts) for statistical comparisons. Heterogeneous sensitivities to Adriamycin and FUdR were observed, but there was no difference in the sensitivities of the clones to methotrexate cell killing. We could not find correlations between drug response and clonal origin, passage number or metastatic properties. The sensitivities of the clones to Adriamycin and FUdR changed upon in vitro passage, although drift in Adriamycin sensitivity during growth in culture was not statistically significant. The results demonstrate that clonal heterogeneity exists within the 13762 tumor and its metastases in their inherent responses to chemotherapy agents, and in the absence of host selective pressures tumor cells can spontaneously and reproducibly drift in their sensitivities to certain chemotherapeutic drugs.

Comparison of 'spontaneous' and 'experimental' metastasis using rat 13762 mammary adenocarcinoma metastatic cell clones

Rat 13762NF mammary adenocarcinoma cloned cell lines were assayed at different in vitro passage numbers and compared for their abilities to form 'spontaneous' metastases by subcutaneous injection of cells and 'experimental' metastases by intravenous injection of cells. Tumor cell clones were established from locally growing tumor and spontaneous lung metastases, and these clones were found to possess heterogeneous metastatic potentials in both metastasis assays. The rank order of clonal metastatic potentials based on either the average number of lung tumor colonies or the average total lung tumor volume was generally equivalent for 'spontaneous' and 'experimental' metastases, but some differences were noted. Ranking of 'spontaneous' metastasis by average total lung tumor volumes more closely resembled the rank order of 'experimental' metastasis than by the average number of spontaneous metastases. The results demonstrated that in the 13762NF mammary adenocarcinoma system (i) there is heterogeneity in tumor cell clonal metastatic potential using either 'spontaneous' or 'experimental' assays; (ii) these two assay methods yield generally the same rank order of metastatic potential; (iii) the metastatic potential of each of the tumor cell clones drifts with time (passage number) in cell culture, and (iv) ranking by average tumor burden calculated from total lung tumor volumes may yield a better estimate of metastatic potential than ranking by the average number of lung tumor colonies.