Growing colorectal tumors: minimizing microbial and stromal competition and assessing in vitro selection pressures.

Development of primary colorectal cancer cell lines ishampered by contamination from regional microbes, overgrowthof stromal cells, and purported genetic drift from selectionpressures in vitro. We initiated 32 primaryadenocarcinomas, 3 recurrences and 6 distant metastases incell culture. Twelve cell lines from eleven tumors weregenerated (26.8%) overall. Nine of 32 primary tumorsyielded 10 cell lines, 5 were lost to contamination, 13 wereoverwhelmed by stromal cells, and 5 demonstrated no growth.Addition of isobutyl methyl xanthine (IBMX) to culturelimited fibroblastoid growth. There was no associationbetween tumor location (p = 0.535, mid-P), degree ofdifferentiation (p = 0.850, mid-P) or clinicopathologic stage(p = 0.400, mid-P), and the ability of cells to becomeestablished in culture. The majority of cell lines hadsimilar nuclear DNA content and expression of cell-surfaceantigens compared with their parent tumors. Microbialcontamination and stromal cell overgrowth present thegreatest obstacle to capturing a representative bank ofcolon tumors in vitro.

Acquisition and enhanced expression of the metastatic phenotype following transfections of genomic mouse tumor DNA containing human SCLC gene sequences.

Previous primary and secondary co-transfections of genomic DNA from a metastatic human small cell lung cancer cell line into NIH/3T3 cells resulted in a murine fibrosarcoma cell line (Tx93B) that produced frequent spontaneous lung metastases in subcutaneously injected tumor-bearing nude mice. In order to transfer the acquired metastatic behavior to additional cell lines that could then be tested in syngeneic immunocompetent animals, DNA from Tx93B cells was transfected without additional neo gene into Balb/c embryo fibroblasts, which led to the isolation of a tertiary transfectant cell line (D3) of low spontaneous metastatic potential in normal Balb/c mice. Subsequent cell lines established serially from lung metastases in mice injected with D3, and metastatic descendants of D3 (all selected for the original neo marker in G-418), resulted in three generations of metastatically variant cell lines capable of causing pulmonary metastases in 11.1%, 54.6%, and 89.5%, respectively, of subcutaneously injected animals, and in 100% of normal mice injected intraperitoneally. There was no apparent ras-family oncogene participation in the metastatic behavior of either of the two DNA donor cell lines or in the metastatically variant tertiary transfectants. Gelatin zymography indicated that the secretion of gelatinolytic enzymes in vitro by the variant cell lines was inversely proportional to their metastatic capability. Human Alu repeat gene sequences detected in the metastatic variants suggested that co-transfected metastasis-associated genes present in the original human DNA donor cell may have contributed to acquisition of the metastatic phenotype by the tertiary transfectant cell lines. The increase in metastatic potential observed in successive generations of the D3-derived tumor cell lines, further suggested that selection for cells having increased metastatic capability had occurred during passage in vivo accounting for the phenotypic change. Because of their common origin and progressively metastatic nature these cell lines may prove useful in the identification of metastasis-associated genes accessible through the use of differential expression cloning strategies.

Abnormalities in structure and expression of the retinoblastoma gene in small cell lung cancer cell lines and xenografts in nude mice.

The putative retinoblastoma gene (Rb) is a tumor suppressor gene which is believed to cause retinoblastomas when both alleles are inactivated, leading to lack of the encoded Mr 110,000-116,000 phosphoprotein. Inactivation of the Rb gene has also been found in several other tumor types, including small cell lung cancer (SCLC). Absence of the 4.7 kilobase mRNA has been found to be frequent in SCLC, and it has been reported that the Rb Mr 110,000-116,000 protein product is always absent, even in tumors expressing Rb mRNA. Using Western blotting technique with a monoclonal antibody directed against the Rb protein, we investigated the expression of the Mr 110,000-116,000 Rb protein in SCLC tumors grown as xenografts in nude mice and/or as cell lines. Rb messenger RNA expression was determined by Northern blotting, and gross structural gene alterations were investigated by Southern blotting. Tumors established from 23 patients were studied. Seven of the tumors did not express Rb protein, whereas expression was detectable in 13. Three tumors were not investigated for protein expression. Only two tumors expressed Rb mRNA without detectable Rb protein expression. Gross DNA alterations were found in four tumors, of which only one expressed Rb mRNA. Our results demonstrated frequent absence of Rb mRNA and protein in SCLC, but apparently normal Rb mRNA and protein were both expressed in more than one-half of the tumors.

Biosynthesis of procalcitonin in small cell carcinoma of the lung.

Immunoreactive calcitonin (CT) secreted by DMS 53, a cell line derived from human small cell carcinoma of the lung, consists almost entirely of molecular species larger than the mature hormone (Mr 3,420). Messenger RNA isolated from DMS 53 cells and nude mouse tumors was translated in wheat germ systems, and the products were precipitated with CT-specific antisera. Analyses of the translation products by electrophoresis on 15% polyacrylamide-sodium dodecyl sulfate gels indicated synthesis of a Mr 16,500 preprohormone that was reduced to Mr 14,500 by cotranslation with microsomal membranes. Immunoprecipitation of CT from media from pulse-labeled cultures revealed two major products (Mr 16,500 and Mr 14,500) and up to three minor secreted polypeptides (Mr 9,400, 8,400, and 6,800). Intracellular CT from cell homogenates appeared almost entirely as a single major product (Mr 14,500) and possibly 3-4 minor components (Mr 16,500; 9,200, 8,400, and 6,800). No glycosylated forms of CT were demonstrable by lectin binding methods or labeling attempts with tritiated sugars. The presence of multiple CT species in DMS 53 cells suggests significant post-translational processing of the larger precursor molecules and the accumulation and secretion by small cell carcinoma of the lung of several intermediate immunoreactive forms via a glycosylation-independent secretory pathway.

Immunocytochemistry: its evolution and criteria for its application in the study of epon-embedded cells and tissue.

The word immunocytochemistry is currently used to describe a number of methods that can be employed to localize antigens within cells by means of antigen-specific antibodies. In this article we will review a number of these methods, including immunofluorescence, immunoperoxidase, avidin-biotin, and colloidal-gold techniques. The advantages and disadvantages of the various methods are discussed, special attention being focused upon immunocytochemical staining of plastic-embedded tissue. Studies on the light microscope level show that embedding tissue in plastic prior to immunoperoxidase staining not only improves visualization of antigen-specific staining but also provides an accurate and efficient means of prescreening tissue for antigen prior to immunocytochemical staining on the electron microscope level. Varying section thickness between 1 and 3 microns does not significantly influence staining, whereas the fixative used to preserve the tissue under study does. On the electron microscope level, the colloidal gold technique appears superior to immunoperoxidase staining. It is both esthetically more pleasing and highly sensitive. Of five different colloidal gold methods tested, the most sensitive is the two-step technique that employs an antigen-specific primary antibody followed by a gold-labeled secondary antibody. Throughout this article, special emphasis is placed on the use of proper controls, both on the light and electron microscope levels. Where possible, such controls should include substitution of specific antiserum with normal serum; the use of antigen-adsorbed antiserum; the use of antisera with specificities for antigens not present in the tissue being studied; the use of tissue previously shown to be stainable for the antigen; and if cultured cells are being studied, the use of a number of cell types that do not contain the antigen.

Evaluation of drug efficacy in vitro using human small cell carcinoma of the lung spheroids.

Five human small cell carcinoma of the lung (SCCL) cell lines selected from 25 established cultures were grown as three-dimensional spheroid tumor models in either spinner culture or in static, agar-coated multiwells. Volume doubling times for the cell lines were approximately 4.5 days. Decreases in spheroid volumes after exposure to a variety of chemotherapeutic agents were used as indicators of drug activity. To further quantify cell killing in SCCL spheroids by chemotherapeutic agents 24 hours after exposure to drugs, a technique was employed that measured maximum levels of incorporation of 125IUdR after continuous labeling for 48 hours. The results of the use of this assay report for SCCL spheroid responses to various concentrations of doxorubicin hydrochloride, cytosine arabinoside, mechlorethamine hydrochloride, cisplatin, or etoposide. Some evidence for an intertumor heterogeneous response to chemotherapy is presented for some of the drugs tested. This assay was also used to characterize a potentiated cell kill when etoposide is combined with cisplatin and to identify activity by a new compound, diazoacetylcholine iodide (DACI), which was synthesized as an agent targeted for SCCL cells.

Immunocytochemical staining of cytocentrifuge prepared cultured cells: nonspecific staining and its elimination.

In an attempt to localize hormones in cytocentrifuge-prepared cultured cells of small cell carcinoma of the lung (SCCL), various modifications of the immunoperoxidase (PAP) procedure (Sternberger, 1979) were tested. When using glutaraldehyde, formaldehyde, or p-benzoquinone fixation (Pearse & Polak, 1975) and rabbit antibodies in primary or bridging steps of the PAP procedure, nonspecific staining (false positives) could be elicited with the majority of rabbit antibodies tested, but not with antibodies from other animal sources. This problem could be eliminated by fixation of cells either with formalin-acetone (Mason et al., 1975) or, when using antibodies from a source other than rabbit, glutaraldehyde. It was not possible to localize ACTH in DMS-79, a human SCCL line known to produce this hormone. However, calcitonin was localized in the calcitonin-producing SCCL line DMS-53. Failure to localize ACTH in DMS-79 may be due to the lower levels of this hormone in DMS-79, as compared to the levels of calcitonin in DMS-53. This study emphasizes the importance of proper controls before concluding successful localization in a given immunocytochemical preparation of cultured cells.

Cytogenetics of small cell carcinoma of the lung.

Nineteen cell lines derived from various malignant tissues of 15 patients with small cell carcinoma of the lung (SCCL) have been studied. The results showed heterogeneity in all cell lines, with no one consistent abnormality among them. Cell lines from 11 of the patients had minute and double minute chromosomes, and cell lines from 2 patients had abnormally banding regions, designated as ABRs, as distinguished from homogeneously staining regions (HSRs). The latter 2 and several of the former cell lines were derived from specimens taken before the patients were placed on therapy. All but 2 of the cell lines had a constant marker load, consisting of 24%-35% of the complement. Some markers remained stable through months and years of culture life, while other markers came and went. Chromosomes #1, #6 and #11 were most frequently involved in marker formation in the cell lines, and these were compared to similar markers in direct bone marrow preparations. Chromosome #1 markers were of variable structure, whereas #6 and #11 most often took the form of 6q- and 11p+ markers, with breakpoints most frequently at 6q23-25 and 11p11-12. A 3p- marker was found in a minority of cell lines. All of these markers were also found in direct marrow preparations from some patients with SCCL. Nonmonoclonal tumors arose from inoculation of bimodal cell lines into nude mice, but population selection by undetermined mechanism was evident. Cytogenetic parameters showed no positive correlation with hormone production by these cell lines.

Unlabeled antibody methods in electron microscopy: a comparison of single and multistep procedures using colloidal gold.

A comparative study of five unlabeled antibody methods was conducted on the electron microscopic level using bridging techniques and colloidal gold. The study was based on the principles of the single-step colloidal gold (GLAD) method (Larsson L: Nature 282:743, 1979) and the multistep single- and double-bridge techniques used in postembedding immunoperoxidase procedures (PAP) (Sternberger LA: Immunocytochemistry, 2nd ed. Wiley, New York, 1979). Using medullary thyroid carcinoma and the same lot of primary antiserum (goat anti-calcitonin) for each procedure, it was shown that adequate localization of calcitonin with the single-step GLAD method was attainable only at dilutions of 1:100 or lower. The single-bridge technique using goat anti-calcitonin, sheep anti-goat immunoglobulin (Ig)G, and goat anti-calcitonin and antigen-coated gold, respectively, worked well at dilutions of up to 1:5000 but not at dilutions of 1:10,000, while single- and double-bridging techniques utilizing goat anti-calcitonin, sheep (Sh) anti-goat IgG, and sheep anti-goat IgG-coated gold produced good localization at a 1:10,000 dilution of primary antiserum. A two-step method using goat anti-calcitonin and sheep anti-goat IgG-coated gold, respectively, appeared to be the most sensitive technique, with adequate antigen localization occurring at a dilution of 1:25,000. While in our hands the two-step method appeared superior in sensitivity to the single-bridge IgG-coated gold technique, each method has its own advantages depending on the individual needs of the researcher.

Calcitonin as an indicator of the response of human small cell carcinoma of the lung cells to drugs and radiation.

A calcitonin (CT)-producing cell line (DMS53) established from human small cell carcinoma of the lung was grown as three-dimensional multicellular spheroids in spinner culture or on agar in multiwells, and as tumors in nude (athymic) mice. CT release into the media was directly proportional to spheroid volume. The response of these cells following exposures to X-irradiation, Adriamycin, or diazoacetylcholine iodide was assessed by monitoring levels of CT released into the media by individual spheroids. Levels of CT in the blood of nude mice bearing DMS53 xenografts were directly proportional to tumor volume and decreased proportionally with tumor response to X-irradiation and cisplatin treatment. These results suggest that the DMS53 spheroid and xenograft models may be useful systems to monitor responses to therapy utilizing CT as an indicator of tumor burden.

Bombesin and calcitonin secretion by pulmonary carcinoma is modulated by cholinergic receptors.

Three established cell lines derived from human small cell carcinoma of the lung, and known to produce significant amounts of peptide hormones were used to evaluate the regulation of hormone secretion by cholinergic agonists. In two of the cell lines (DMS 53, DMS 153) acetylcholine chloride, bethanechol chloride, and carbamylcholine at the concentrations of 10(-3)M to 10(-5)M stimulated secretion of bombesin and calcitonin as measured by RIA. The third cell line, DMS 406, was not significantly stimulated. Inhibition of induced stimulation by the cholinergic antagonist atropine, but not hexamethonium, indicated the presence of muscarinic rather than the nicotinic type of cholinergic receptors on the stimulatable cells. These receptors appear to mediate hormone secretion comparably to normal endocrine cells.

Cell surface glycoprotein patterns of cell lines derived from human small cell carcinoma of the lung.

The expression of major cell surface glycoproteins (sgp) of seven established human pulmonary small cell carcinoma (SCC) cell lines and two autologous non-neoplastic lymphoblastoid and fibroblastic cell lines were studied by the galactose oxidase tritiated sodiumborohydride labelling technique. The general sgp pattern of SCC cell lines was different from that of the autologous non-malignant cell lines and the various other normal and malignant hematopoietic cells, gliomas and melanomas previously studied by the same technique, thus confirming that the sgp pattern seems to represent a molecular "fingerprint" of various human cell types. The SCC cell lines could be subdivided into two major subgroups with respect to expression of characteristic sgps. One group of lines was characterized by having prominent sgp 52000 (52 K) Dalton (D), 50 KD, 40 KD, 34 KD and 10 KD. In the other group the major sgps had apparent molecular weights of 110 KD, 75 KD and 10 KD, respectively. In addition to these common basic "group specific" sgps each SCC line expressed individually distinct sgps.

Bombesin production by human small cell carcinoma of the lung.

A series of continuous cell lines of human small cell carcinoma of the lung (SCCL) have been evaluated for the production of bombesin (BN). In early established cultures BN was detected in the medium of 9 out of 11 cell lines and in 6 out of 7 cell homogenates examined. Levels in the medium were frequently higher in cultures of later passages compared to earlier passages of the same line and low levels developed in the two previously negative cell lines. Plasma concentrations were greater than 80 pmol/l in 2 out of 27 (7%) randomly selected patients with SCCL. A culture (DMS 406) established from the tumor of a patient with the highest plasma level (1240 pmol/l) was the highest producer in vitro. The results indicate that BN, which has been demonstrated immunocytochemically to be present in normal bronchial mucosal cells, is frequently produced by SCCL in vitro but elevated plasma levels are infrequently found in patients with this neoplasm.

Hormone production by cultures of small-cell carcinoma of the lung.

Continuous cell lines have been established from a variety of biopsy and postmortem species of tumor from patients with small-cell carcinoma of the lung (SCCL) and have been maintained over several years. The medium from the cultures has been assayed for peptide, glycoprotein, and steroid hormones. Significant amounts of 14 hormones including calcitonin, adrenocorticotropin (ACTH), parathormone, luteinizing hormone, chorionic gonadotropin, glucagon, growth hormone, somatostatin, prolactin, beta-endorpin, lipotropin, oxytocin-neurophysin, vasopressin-neurophysin, and estradiol have been demonstrated. Up to ten different hormones have been produced by a single cell line. Most produce ACTH and all evaluated so far produce estradiol. These studies indicate that cells from SCCL have a potential for producing a wide variety of hormones and that this characteristic can be maintained for prolonged periods of culture in vitro.

Isolation and growth characteristics of continuous cell lines from small-cell carcinoma of the lung.

Sixteen continuous tumor-cell cultures have been isolated from 91 tissue specimens from patients with small-cell carcinoma of the lung. Biopsy and autopsy specimens of primary and metastatic tumors have been utilized. The developing cell lines were recognized by proliferation of tumor cells in the culture from one to 14 weeks after explanation and have been maintained for up to four years. Primary lung tumor, bone marrow aspirations, pleural effusions and other metastases have all been productive explant material for the development of cell lines. Their human origin has been demonstrated by chromosome and/or isoenzyme analysis. Dense core vesicles, characteristically found in small-cell tumor cells were observed by electron microscopic examination of cultured cells. Growth rates in vitro have been measured and the in vitro cycle time in tumors of one cell line (DMS 79) has been compared with in vivo cycle time in tumors arising from DMS 79 cells in nude athymic mice.

Animal model for small cell carcinoma of the lung. Effect of immunosuppression and sex of mouse on tumor growth in nude athymic mice.

An animal model for the experimental study of small cell carcinoma of the lung (SCCL; human) has been developed. Tumors arising from continuous cell lines of cultured tumor cells as well as transplantable tumors of human SCCL in nude athymic mice have been characterized as to growth rate and morphology. The effect of antilymphocyte and antithymocyte sera and the sex of the host mouse were evaluated.