Improvement of Human Melanoma Colony Formation in Soft Agar Using Boiled Instead of Autoclaved Agar

Transcription

1 International Journal of Cell Cloning 1: 8591 (1983) Improvement of Human Melanoma Colony Formation in Soft Agar Using Boiled Instead of Autoclaved Agar Stephen P. Thomson, Mark D. Wright, Frank L. Meyskens, Jr. Department of Internal Medicine and Cancer Center, University of Arizona, Tucson, AZ, USA Key Words. Soft agar. Human melanoma colony Melanoma Clonogenic. graphics. Abstract. The effect of agar sterilized by either boiling or autoclaving on human melanoma colony formation in soft agar was compared using cells from 17 biopsies of metastatic malignant melanoma. The frequency of colony formation was significantly increased for cells grown in boiled agar in 8 samples (47%), unchanged in 8 samples (47%), and decreased in only one sample (6%). There were increases in both cluster and colony formation for the melanomas which had augmented colony formation when grown in boiled agar. There was also qualitative morphological improvement, including rounder, smoother cells and less extracellular debris surrounding the colonies. These data suggest that melanoma colony formation is enhanced when cells are grown in agar which has been sterilized by boiling rather than autoclaving. Introduction Many types of human tumors are grown in clonogenic assays which use agar as a semisolid support [l41. Sterilization of the agar is typically done by autoclaving. Dixon et al. [5] have recently shown that autoclaved agar contains an inhibitor of granulocytemacrophage colony growth. We therefore have compared agar sterilized by autoclaving to agar sterilized by boiling for the support of formation of human melanoma colonies grown from cells obtained from 17 biopsies of melanoma tissue. Quantitative and qualitative improvement in melanoma colony growth was demonstrated. '@ 1983 AlphaMed Press, Inc /83/$2./

2 Thomson/Wright/ Meyskens 86 Materials and Methods Culture Techniques Tumor tissue was obtained by excisional biopsy of subcutaneous nodules from patients with metastatic malignant melanoma in accord with a protocol approved by the University of Arizona Committee on Human Subjects. The tissue was mechanically dissociated into a single cell suspension as described previously [I]; cells were cryopreserved in 1% dimethylsulfoxide in Ham's F1 medium and 1% heatinactivated fetal calf serum (FCS). We have previously described the culture system which is a simplification [l] of the bilayer agar system developed by Hamburger and Salmon [2]. Briefly, the underlayer was 1. ml of.5% agar (Bacto, Difco Lab, Detroit, MI) in Ham's F1 medium which contained 1% FCS, penicillin (1 pg/ml) and streptomycin (1 unitdml). The plating layer contained 5, nucleated cells in 1. ml of.3% agar in the same medium as the underlayer and was supplemented with insulin (1.54 units/ml), glutamine (.45 pg/ml), pyruvate (.34 pg/ml) and mercaptoethanol (.77 mm). The plates were cultured for 1421 days in a humidified 5% COP, 95% air atmosphere at 37 C. In this study colonies were defined as groups of cells greater than, or equal to, 6 pm in diameter while clusters were defined as groups of cells between 3 and 6 pm in diameter. Clusters and colonies were enumerated using the Omnicon FAS I1 (Bausch and Lomb, Rochester, NY) image analysis system [6]. An inverted microscope equipped with a calibrated reticle was used to obtain measurement of the diameters of clusters and colonies. Preparation of Agar Stock 3% agar (Bacto) solutions were prepared by adding.75 g of agar to 25. ml of distilled HzO and sterilized by two different methods to compare their effect on colony formation. The standard method was to autoclave the stock solutions of agar at 121OC and 15 PSI for 2 min, store at 4OC, and boil immediately before use for 2 to 25 min. Alternatively, the stock solution of agar was boiled immediately before use for 2 to 25 min without prior autoclaving. Results and Discussion The frequency of colony formation using agar sterilized by boiling or autoclaving agar was determined using cells from 17 different biopsies of human malignant melanoma (Table I). Cells from eight samples (47%) produced significantly more colonies when boiled rather than autoclaved agar was used. Cells from eight samples (47%) produced similar numbers of colonies in both agars, including 3 samples which did not grow in either condition; only one sample (6%) had fewer colonies using boiled agar. The median increase in colony formation using boiled agar was 66%. We also counted all the clusters and colonies equal to or greater than 3 micrometers in diameter in 4 of the 8 samples which had significant in

3 Improvement of Melanoma Colony Formation 87 Table I. Effect of autoclaved or boiled agar on frequency of melanoma colony formation from 17 biopsies of human melanoma Patient code No. colonies 1 5 pm 76 change pvaluea diameter/plate Autoclaved Boiled *4 81 *44 81 =46A 8146C 81 * A 82*7B f 21b 119f6 57f9 172 f 2 16f3 13f f f f8 25f8 776 f f f f f f4 87 f f f9 2f I 476 f f13 21 k f f f f a Student s ttest; =not significant; p>.5. b Mean k S.E. creases in the frequency of colony formation to determine whether the increase was secondary to more clusters developing into colonies or represented de novo growth. If the effect of boiled agar was only to allow more clusters to develop into colonies without de now growth, one would expect a decrease in the number of clusters with a concomitant increase in the number of colonies. However, the data in Figure 1 show that boiled agar increased the number of both clusters and colonies. For example, if the increase in the number of colonies for patient B (Fig. 1B) was due solely to an increase in the growth of clusters into the colony size class, most of the clusters grown with autoclaved agar (Fig. 1B) would not be present when boiled agar was used. However, many clusters were present when boiled agar was used (Fig. lb), favoring the explanation that the effect of boiled agar was to allow more de novo growth. Further support is provided by the

4 Thornson/ Wright/ Meyskens 88 Fig. 1. Size distribution of clusters and colonies grown using boiled or autoclaved agar. The dashed line is at a diameter of 6 micrometers and by our definition divides clusters and colonies. Patient A (No ), B (No A), C (No ), and D (No ) observation that boiled agar did not increase the size of colonies formed (Fig. 1AD), a result one would expect if the only effect of boiled agar was to increase the size of groups of cells once they started to'proliferate. These results favor the explanation that the effect of boiled agar was to allow more de novo growth of cells into clusters and colonies rather than to increase the size of clusters after they have begun to proliferate. There were also qualitative morphological differences between the colonies formed in boiled or autoclaved agar. These qualitative differences were noted in 4 of the 8 samples which had quantitative increases in colony formation. The colonies formed in boiled agar typically contained cells that were very smooth edged and rounded (Fig. 2A) while the colonies which formed in autoclaved agar usually contained cells that were ir

5 Improvement of Melanoma Colony Formation 89 Fig. 2. Qualitative morphological differences between colonies formed in boiled and autoclaved agar. Patient No (A) Note the rounded, smoothedged cells and the lack of extracellular material in the colony formed in boiled agar. (B) The colony grown in autoclaved agar contained irregular shaped cells and increased extracellular debris which prevented precise focusing on the cells of the colony. Bar = 5 micrometers.

6 Thomson / W right /Meyskens 9 regularly shaped (Fig. 2B). Additionally, the agar immediately adjacent to the colonies formed in autoclaved agar contained large amounts of extracellular debris and granular material. The colonies which formed in boiled agar were structurally and phenotypically similar to those which developed in autoclaved agar and did not morphologically resemble granulocyte or macrophage clusters or colonies. Previous detailed morphological studies of colonies formed from cells from several different biopsies of melanoma have shown that colonies of nonmelanoma cell types are not present [7]. Additionally, increases in colony growth have been noted from four melanoma cell lines using boiled agar (unpublished observations). These data suggest that the increases in colony formation using boiled agar represent an improvement in melanoma colony growth rather than of other cell types. Our results demonstrate that human melanoma colony formation was improved when boiled rather than autoclaved agar was used. Our data extends the finding of Dixon et al. [5] who demonstrated that autoclaved agar contained an inhibitor of granulocytemacrophage colony formation to include a detrimental effect of autoclaved agar on the growth of melanoma colonies. Our data and those of Dixon et al. [5] suggest that the effect of boiled and autoclaved agar on colony formation of other tumor types should also be compared systematically. Acknowledgments This work was supported in part by a grant from the American Cancer Society (PDT184) and the National Cancer Institute (CA2752, CA1794). We thank L. Kimball for the fine secretarial assistance and preparation of the graphics. References Meyskens, F.L., Jr.; Soehnlen, B.J.; Saxe, D.F.; Casey, W.J.; Salmon, S.E.: In vitro clonal assay for human metastatic malignant melanoma. Stem Cells I: 6172 (1981). Hamburger, A.W.; Salmon, S.E.: Primary bioassay of human tumor stem cells. Science 197: (1977). Courtenay, V.D.; Mills, J.: An in vitro colony assay for human tumors grown in immunesuppressed mice and treated in vivo with cytotoxic agents. Br J Cancer 37: (1978). Tveit, K.M.; Fodstad,.; Pihl, A.: Cultivation of human melanomas in soft

7 Improvement of Melanoma Colony Formation 91 agar. Factors influencing plating efficiency and chemosensitivity. Int J Cancer (1981). 5 Dixon, R.A.; Linch, D.; Baines, P.; Rosendaal, M.: Autoclaved agar contains an inhibitor of granulocytemacrophage colony growth in vitro. Exp Cell Res 132: (1981). 6 Kressner, B.E.; Morton, R.A.; Martens, A.E.; Salmon, S.E.; Von Hoff, D.D.; Soehnlen, B.: Use of an image analysis system to count colonies in stem cell assays of human tumors; in Salmon, Cloning of Human Tunor Stem Cells, pp (Liss, New York 198) 7 Persky, B.; Thomson, S.P.; Meyskens, F.L., Jr.; Hendrix, M.J.C.: Methods for evaluating the morphological and immunohistochemical properties of human tumor colonies grown in soft agar. In Vitro Z8: (1982). Accepted: March 9, 1983 Dr. Frank L. Meyskens, Jr., Department of Internal Medicine and Cancer Center, University of Arizona, Tucson, AZ (USA)