Vous êtes sur la page 1sur 5

Monoclonal Antibody and Enzymatic Profiles of Human Malignant T-Lymphoid Cells and Derived Cell Lines

Stephen D. Smith, Margaret Shatsky, Pamela S. Cohen, et al. Cancer Res 1984;44:5657-5660.

Updated Version

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/44/12_Part_1/5657

Citing Articles

This article has been cited by 34 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/44/12_Part_1/5657#related-urls

E-mail alerts Reprints and Subscriptions Permissions

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at pubs@aacr.org. To request permission to re-use all or part of this article, contact the AACR Publications Department at permissions@aacr.org.

Downloaded from cancerres.aacrjournals.org on September 20, 2011 Copyright 1984 American Association for Cancer Research

[CANCER

RESEARCH

44,5657-5660,

December 1984]

Monoclonal Antibody and Enzymatic Profiles of Human Malignant T-Lymphoid Cells and Derived Cell Lines1
Stephen D. Smith,2 Margaret Shatsky, Pamela S. Cohen, Roger Warnke, Michael P. Link, and Berti! E. Glader
Department o1 Pediatrics and Pathology, Stanford University, School of Medicine, Stanford, California 94305 [S. D. S., P. S. C., R. W., M. P. L, B. E. G.J. and Division of HematologylOncology, Children's Hospital at Stanford, Palo Alto, California 94304 S.D. S., M. S., P. S. C., M. P. L, B. E. G.J

ABSTRACT Recently, four distinct cell lines were established from patients whose malignancies had been defined by immunological and biochemical markers. Each patient had a distinct subtype of a Tcell cancer, and each possessed elevated adenosine deaminase and reduced nucleoside phosphorylase activity. Cell lines cul tured in vitro possessed the same basic immunophenotype and biochemical enzyme activity as the patients' original malignant cells. In a direct comparison of the immunophenotype of the cell lines and the patients' malignant cells, full concordance existed for 48 of 52 paired antibody tests performed. However, when compared to the corresponding patient's sample, each cell line showed some minor changes in antigen expression or enzyme level. Antigen loss, de novo antigen expression, or elevated adenosine deaminase levels occurred in the cell lines, and these changes were stable on repeated analysis. While there was good general concordance between the patient's cancer and the es tablished cell line, minor biological differences in the cell lines may reflect cellular maturation or subpopulation selection in vitro.

of a cell line derived from a single colony (21, 22). In this assay, single cells divide in agar to form colonies, and single colonies are plucked and passed in liquid media for continued growth. The establishment of a cell line from a single unscreened colony could result in continued growth of cells that represent only a small subpopulation of the patient's cancer. In order to determine if such selection occurred, we compared the cell surface antigen and biochemical enzyme profile of the patient's malignant cells to the corresponding cell line. T-cell lymphoid cancers were studied because specific cell surface antigen and enzymatic markers are present in T-cells, and changes in these features occur with T-cell differentiation (6, 10, 15, 16). Four cell lines have been studied which showed good general concordance of antigen expression and enzyme levels between the patient's malignant cells and the corresponding cell line. However, minor changes in antigen expression (both antigen gain and loss) and enzyme levels occurred indicating that subpopulation selection or cellular differentiation had occurred in culture.

MATERIALS AND METHODS INTRODUCTION In vitro assay systems have been developed that support growth of human tumor cells from a variety of cancers including leukemia (3, 7,12,14). Both growth of tumor stem cells and the establishment of cell lines have provided much information about the basic biology and drug sensitivity profiles of malignant cells (19, 20). In the human tumor stem cell assay, single cells (suspended in agar) divide 6 or more times to form a colony (cell group with >40 cells) in 14 to 21 days. Tumor cell growth is usually suc cessful, and up to 60% of specific cancers (i.e., malignant melanoma) will grow in vitro (25). However, this assay has some limitations because the plating efficiency is low (often <0.01%), and the colony cell survival is short (usually <3 weeks). Malignant cell lines can be grown from about 1% of malignant samples, but once established, an unlimited number of cells become available for in vitro and in vivo studies. Cell lines can be stored and studied repeatedly under defined experimental conditions. However, growth in liquid media favors overgrowth by the most rapidly dividing cell, and cell mutations can occur rapidly in liquid culture media. Recently, we developed an in vitro assay that supports primary malignant lymphoid colony formation as well as continued growth
1Supported by USPHS Grants CA34710 and CA34233 awarded by the National Cancer Institute, Department of Health and Human Services, and American Cancer Society Grant CH182A. 2 Scholar of the Leukemia Society of America. To whom requests for reprints should be addressed, at Children's Hospital at Stanford, 520 Willow Road, Palo Alto, CA 94304. Received June 7,1984; accepted August 21,1984.

Source of Malignant T-Cells. The data discussed in this report were derived from 4 patients with T-cell non-Hodgkin's lymphoma or T-cell acute lymphoblastic leukemia. The protocol procedures were approved by the Medical Committee for the Use of Human Subjects in Research and informed consent was obtained from the parents or patients. The clinical and pathological data on each patient are presented in Table 1. Malignant cells were collected from pleural effusions, bone marrow, or peripheral blood and anticoagulated with preservative-free heparin. Within 4 hr of collection, malignant lymphoid cells were sepa rated by Ficoll-Hypaque density sedimentation and washed with "com plete medium" consisting of McCoy's 5A medium supplemented with 15% fetal calf serum, penicillin (50 units/ml) and streptomycin (50 /<g/ ml). After a second washing, the malignant cells were separated into aliquots for cell culture experiments, immunophenotypic analysis, and enzymatic studies. Establishment and Maintenance of Cell Lines. The technique for culturing malignant T-cells was basically a modification of our previously reported methods (21, 22). Briefly, separated malignant cells were warmed to 37in complete medium and pipeted repeatedly to break up cell aggregates. The resultant single-cell suspension next was mixed with agar (0.3%) and plated onto 35-mm Retri dishes containing a previously prepared feeder layer. The feeder layer contained a combi nation of complete medium with 10% normal human serum and 0.5% agar. The Retri dishes then were incubated at 37in a Heraeus incubator gassed with 5% O2, 6% COz, and 89% N2. Colonies were plucked after 18 days of growth, suspended in complete medium (without agar), and transferred to Retri dishes with a freshly prepared feeder layer. Cells were observed every 3 to 5 days and were passaged when the medium appeared acidic (i.e., phenolphthalein indicator turned yellow) or when the cell number approached 5 x 105 cells/ml. Of particular importance throughout the complete incubation period was that the assay system contained no added mitogens, thiols, conditioned media, or interieukins.

CANCER

RESEARCH

VOL. 44 DECEMBER 5657

1984

Downloaded from cancerres.aacrjournals.org on September 20, 2011 Copyright 1984 American Association for Cancer Research

PROFILES

OF PATIENTS'

CELLS AND DERIVED

CELL LINES

Table 1
Clinical data on 4 patients with malignant T-cell disease Each patient had a medias t mal mass at diagnosis. Patient1 2 34Age (yr)8 statusRelapse analyzedPteural effusion Bone marrow Peripheral blood Pleural effusion acute lymphoWastic

At diagnosis 52 T-ALL Relapse Male 12 24SexMateFemale T-NHLClinicalRelapseMaterial MateDiagnosisT-NHL"T-ALL T-NHL,T-cenon-HcOgkin's lymphoma; T-ALL, T-cell leukemia.

Studies to Characterize Surface Antigens on Malignant Cells from Patients and Cell Lines. Monoclonal antibodies to Leu-1 (Pan T), Leu2a (T-cytotoxic-suppressor), Leu-3a (T-helper), Leu-4 (Pan T), Leu-5 (erythrocyte-fonming rosette receptor), Leu-6 (thymocyte), Leu-9 (Pan T), CALLA,3 and HLA-DR were generously provided by the Becton-Dickinson Corporation (Mountain View, CA). OKT9 and OKT10 were provided by Ortho Pharmaceuticals (Raritan, NJ). Surface immunoglobulin was detected by the F(ab')z fragment of fluorescein isothiocyanate-conjugated goat anti-human immunoglobulin (Tago. Inc., Burlingame. CA). Cell surface antigens were identified by the binding of monoclonal antibody as detected by indirect immunoftuorescence. Separated lymphoblasts (10*) were placed in plastic tubes and incubated with 1 ^g of purified monoclonal antibody at 4.After 20 min incubation, cells were washed twice in cold phosphate-buffered saline with 0.02% sodium azide. Next, the washed cells were incubated (at 4for 20 min) with 100 ft\ of fluorescein isothiocyanate-conjugated goat anti-mouse immuno globulin (Tago), then washed twice, fixed in 1% formaldehyde in phos phate-buffered saline and analyzed for fluorescent staining with a fluo rescence-activated cell sorter (FACS IV; Becton-Dickinson Electronics Laboratory, Mountain View, CA). The intensity of fluorescence for 10,000 cells in each sample was determined. The results were expressed as a percentage of positive cells compared to background fluorescence when nonreactive myeloma protein was utilized in place of specific antibody. In addition, malignant cells were evaluated for antigen expression by the IHCS technique (28). This evaluation added supplemental data to the fluorescence-activated cell sorter analysis because the IHCS technique detects the presence of cytoplasmic as well as cell surface antigens. This technique was utilized when antigen discordance between the patient's sample and the corresponding cell line was found. Thawed aliquots of patients' cells (previously frozen at -70) and cell lines were cytospun onto glass slides, fixed with acetone, and stained according to the standard technique (28). Also, the ability of each cell line to rosette with sheep RBC was determined (4). Enzyme Analyses on Malignant Cells from Patients and Cell Lines. Prior to enzyme analysis, the previously separated cells were washed in 150 M potassium phosphate buffer (pH 7.1). Following a second wash, the cells were resuspended at a concentration of 0.5 to 1.5 x 107 cells/ ml in the same buffer. Next a lysate of the cell suspension was made by alternate freezing and thawing (3 times), and cell membranes were removed by centrifugation. Protein determinations were made on the cell lysate utilizing the method of Lowry et al. (11). Adenosine deaminase (8, 24) and nucleoside phosphorylase (9) activities were determined accord ing to established methods. Enzyme activity was expressed in units defined as AA over 1 min divided by the protein concentration. In addition to these purine enzymes, TdT activity was assayed utilizing a test system from Bethesda Research Laboratories, Bethesda, MD.

available for analysis in each case possessed enough malignant cells (5 x 107 cells) for immunophenotyping, enzyme determi nation, and in vitro culture. Cells were cultured as a single-cell suspension and grown in an hypoxic environment (5% O2, 6% CO2 and 89% N2). After 18 days in culture, individual colonies were aspirated from the agar with a Pasteur pipet and transferred to a Retri dish containing a freshly prepared feeder layer. Aspiration of a single colony was not difficult for only a few colonies grew on each plate (cloning efficiency, 0.005%). Only one colony was pfaced on each new feeder, thus assuring that any continued growth arose from a single colony. Most colony cells continued to proliferate, and a cell line was established from each sample. These cell lines have been growing for more than 8 to more than 21 months (Table 2). The cell lines were designated SUP- in recognition of their development at the Stanford University Pediatrie Department. Multiple T-cell antigens were present on the cell surfaces of both the patient's malignant cells and the cell lines grown from these cells. SUP-T1 did not form rosettes with sheep RBCs and lacked the sheep RBC receptor, Leu 5. SUP-T2, 3, and 4, however, were Leu-5+, and greater than 50% of the cells formed heatstable rosettes with sheep RBC. A comparison of the immunophenotype of the patient's malig nant T-cells and the corresponding cell lines is listed in Table 2. All of the primary samples and corresponding cell lines expressed the pan T-cell antigens Leu 1 and Leu 9 and lacked HLA-DR and surface immunoglobulin. In addition to these T-cell features, multiple other T-cell antigens were present on both the primary samples and the cell lines. In fact, full concordance existed for 48 of the 52 specific paired antibody tests performed. However, in 4 cases there were changes in the expression of the antigens on the cell lines. The SUP-T1 cell line lacked CALLA, an antigen present on 21% of the patient's malignant effusion. Since 98% of the effusion cells were lymphoblasts, the CALLA-positive cells probably represent an immature T-cell subpopulation and not normal cells present in the effusion. The SUP-T2 cell line ex pressed Leu 2a (T-cytotoxic-suppressor) and OKT9 (transfemn receptor), antigens which were not present on the patient's bone marrow cells. The cell line SUP-T4 expressed OKT10, an antigen not present on the cells from the patient's malignant effusion. The OKT10 antigen is detected on rapidly dividing cells and the expression of this antigen may be secondary to a change in the proliferation rate of the cell line. The patient's malignant cells and the cell lines were also evaluated for antigen expression by IHCS. The results showed agreement in antigen expression and intensity except for malig nant cells from Patient 2 (Table 3). While greater than 60% of the malignant cells expressed OKT9 by IHCS, less than 20% of these cells were positive by fluorescence-activated cell sorter analysis. Such results are commonly observed in our laboratory and reflect the ability of IHCS to detect cytoplasmic antigens. The absence of OKT9 could be secondary to antigen modulation on the bone marrow cells for the corresponding cell line ex pressed both cell surface and cytoplasmic OKT9. By the same token, absence of CALLA from the cell line SUP-T1 was not due to modulation because the cell line lacked cytoplasmic CALLA (immunohistochemical stain negative). A comparison of the purine enzymatic activity between the patient's cells and the cell lines are shown in Table 4. In each of the 4 cases, the patient's malignant cells at presentation manifest

RESULTS Recently, 4 samples from patients with malignant T-cell dis ease were successfully cultured in our laboratory. The material
3 The abbreviations used are: CALLA, common acute lymphoblastic leukemia antigen; IHCS, immunohistochemical staining; TdT, terminal deoxynudeotidyl transferase; ADA, adenosine deaminase; NP, nucleoside phosphorylase.

CANCER

RESEARCH

VOL. 44 DECEMBER 5658

1984

Downloaded from cancerres.aacrjournals.org on September 20, 2011 Copyright 1984 American Association for Cancer Research

PROFILES

OF PATIENTS'

CELLS AND DERIVED CELL LINES

Table 2
Reactivity of monoclonal antibodies with the patients' malignant cells and the corresponding Patientl B..M Leu1 Leu 2a Leu Leu Leu Leu Leu 3a 4 5 6 9 I +++ !_ZL SUP-T1 Patient 2 SUP-T2 Patients SUP-T3 Patient 4 cell lines SUP-T4

-H-++
I ++I

OKT9 OKT10 CALLA HLA-DR Slg6

" -, <20% of cells stain positive; +, 20 to 40% of cells stain positive; ++, 40 to 60% of cells stain positive; +++, 60 to 80% of cells stain positive; ++++, 6 Slg, surface immunoglobulin. >80% of cells stain positive.

Tables
Reactivity of monoclonal antibodies determined by fluorescence-activated sorter or immunohistochemical staining IHCSPatient 1 SUP-T1 Patient 2 SUP-T2 Patient 2 SUP-T2 Patient 4 SUP-T4CALLA e -, <20% of cells stain Antibody Fluorescence-activated cell sorter +' cell

DISCUSSION The present study was undertaken to determine the concord ance between the patient's cancer and a cell line established in

CALLA Leu 2a Leu 2a ++ + OKT9 +++ OKT9 -H-+ ++++ OKT10 OKT10 ++++ ++++ positive; +, 20 to 40% of cells stain positive; ++, 40 to 60 to 80% of cells stain positive; ++++, >80%

60% of cells stain positive; +++, of cells stain positive.

vitro from a single colony. The methods utilized in our laboratory routinely result in proliferation of only a small percentage of the tumor cells plated (21, 22). The initial cloning efficiency ap proached 0.005%, and individual colonies were plucked and passed on agar until growth in media could be sustained. With such culturing techniques, minor subpopulations of tumor cells which possess different, biological features from the majority of the patient's malignant cells could selectively grow In vitro. Selective growth (or cellular differentiation) has occurred in other In vitro assays because chromosomal, enzymatic, and antigenic shifts have been documented (1,18, 23, 26). Four distinct cell lines have been established which have been growing in vitro for more than 8 to more than 21 months. These cell lines possess the same basic immunophenotype and bio chemical enzyme levels as the patient's original tumor. Prelimi

Table 4
ADA and Durine NP activity in patients ' malignant cells and corresponding lines Enzyme units/mg protein cell

nary results show that antigen expression have been stable, and no spontaneous changes occurred when repeat testing was performed. Enzyme analysis showed that the cultured cells were 3SUP-T3Patient TdT positive and possessed high ADA and low NP activity. This pattern of enzyme activity is similar to an immature thymocyte 4SUP-T4Normal and is useful in characterizing the maturation levels of T-lympho1.2sNP0.150.180.160.140.150.190.190.160.36 0.05' lymphocytesADA1272826169531662372585.4 cytes (27). On repeat enzyme analysis, TdT remained strongly Mean S.D. positive, and ADA and NP activity varied only slightly. These results help substantiate the fact that the established biochemical features of T-lymphoid cells. ADA activity was 10- cell lines represent a significant part of the patient's cancer as
2SUP-T2Patient

1SUP-T1Patient Patient

to 50-fold greater than that seen in normal lymphoid cells while NP activity was significantly less than that observed in normal lymphocytes. This pattern of markedly increased ADA with de creased NP activity biochemically defines an immature T-lymph oid cell (5, 27). Of particular importance for the present study, however, was the concordance between the purine enzyme activity in the patient's malignant cells and the derived cell lines. Repeat enzyme analysis showed the same general results, and the values expressed in Table 4 are the average of 3 separate analyses each performed at least 2 months apart. In addition to the purine enzymes, TdT activity was present in greater than 90% of the cells of each cell line.
CANCER RESEARCH

determined by cell surface antigen and enzyme evaluations. The agreement between the cell lines established from a single colony (0.005% of tumor cells) and the cells that did not proliferate in vitro (>99% of tumor cells) was striking but not complete. Specific changes in antigen expression or enzyme levels oc curred in each cell line. Both de novo expression (Leu 2a:SUPT2; OKT9:SUP-T2; OKT10:SUP-T4) and loss of antigens (CALLA:SUP-T1) on the cell lines occurred. In addition, the ADA level in 2 cell lines was over twice that found in the patient's malignant cells. Regardless of the possible mechanisms of these changes, it does not appear to be due to spontaneous changes in the cultured cells since repeated analysis of each cell line
1984

VOL. 44 DECEMBER 5659

Downloaded from cancerres.aacrjournals.org on September 20, 2011 Copyright 1984 American Association for Cancer Research

PROFILES

OF PATIENTS'

CELLS AND DERIVED CELL LINES Growth of human tumour cell colonies from biopsies using the two soft-agar techniques. Br. J. Cancer, 38: 77-81,1978. 4. Froland,S. S. Bindingof sheeperythrocytes to human lymphocytes.A probable marker of T-lymphocytes. Scand. J. Immunol., 1: 269,1972. 5. Glader, B. E., Unk, M. P., Amylon, M. D., Penine, S. P., and Backer, K. Heterogeneityof acute lymphoblasticleukemia(ALL) as revealedby adenosine deaminase(ADA) and purine nudeoside phosphorylase(PNP)activity. Blood, 60: 126a, 1982. 6. Greaves, M. F. Analysis of the clinical and biological significance of lymphoid phenotypes in acute leukemia. Cancer Res., 41:4752-4766,1981. 7. Hamburger, A. W., and Salmon, S. F. Primary bioassay of human tumor stem cells. Science (Wash DC), 797: 461-463,1977. 8. Hopkinson, P. J. Further data on the adenosine deaminase(ADA) polymorph ism and a report of a new phenotype. Ann. Hum. Genet., 32: 361,1969. 9. Katekar, H. N. Differential spectropnotometry of purine compounds by means of specific enzymes. J. Biol. Chem., 767: 427,1947. 10. Koziner, B., Gebhard, D., Denny, T., McKenzie, S., Clarkson, B., Miller, D., and Evans, R. Analysis of T cell differentiation antigens in acute lymphatic leukemia using monoclonalantibodies. Blood, 60: 752-757,1982. 11. Lowry, 0. H., Rosebrough, N. J., Fair, A. L., and Randall, R. J. Protein measurement with the Folin phenol reagent. J. Biol. Chem., 793: 265-275, 1951. 12. Minowada, J. Immunologyof leukemiccells. In: F. M. Gunz and E. S. Hender son (eds.), Leukemia, Ed. 4, pp. 119-139. New York: Grue Stratton, 1983. & 13. Minowada, J., Koshiba, H., Sagawa, K., Kubonishi, l., Lok, M. S., Tatsumi, E., Han,T., Srivastava,B. I. S., and Ohnuma,T. Marker profilesof human leukemia and lymphoma cell lines. J. Cancer Res. Clin. Oncol., 707:91-100,1981. 14. Park. C. H., Savin, M. A., Hoogstraten, B., Amare, M., and Hathaway, P. Improved growth of in vitro colonies in human acute leukemia with the feeding culture method. Cancer Res., 37:4595-4601,1977. 15. Reinherz. E. L., Kung, P. C., Goldstein, G., Levey, R. H., and Schlossman,S. F. Discrete stage of human ntrathymic ifferentiation: analysis of normal d thymocytes and leukemic lymphoblastsof T-cell lineage.Proc. Nati. Acad. Sci. USA, 77:1588-1592,1980. 16. Reinherz, E. L., and Schlossman, S. F. Derivation of human T-cell leukemias. Cancer Res., 41: 4767-4770,1981. 17. Ritz, J., Pesando,J. M., Notis-McConarty,J., and Schtosman,S. F. Modulation of human acute lymphoblastic leukemia antigen induced by monoclonal anti body in vitro. J. Immunol., 725:1506-1514,1980. 18. Rockwell, S. In vivo-in vitro tumour cell lines: characteristics and limitations as models for human cancer. Br. J. Cancer, 41 (Suppl. 4V 118-122,1980. 19. Sachs, L. Control of growth and normal differentiation in leukemic cells: regulation of the normal developmentalprogram and restoration of the normal phenotype in myeloid leukemia. J. Cell Physiol., 7:151-164,1982. 20. Salmon, S. E., Hamburger, A. W., Soehnten, B., Durie, B. G. M., Alberts, D. S., and Moon, T. E. Quantitten differentialsensitivity of human-tumorstem of cells to anti-cancer drugs. N. Engl. J. Med., 298:1322-1327,1978. 21. Smith, S. D., and Rosen, D. Establishment and characterization of a human null-cell lymphoblastic lymphoma cell line (K-LL-3). Int. J. Cancer, 23: 494503,1979. 22. Smith, S. D., Wood, G. W., Fried, P., and Lowman, J. T. In vitro growth of lymphoma colonies from children with non-Hodgkin's lymphoma. Cancer (Phila.),48:2612-2623,1981. 23. Sundstrom, C., and Nilsson, K. Cytochemical profile of human haematopoietic biopsy cells and derived cell lines. Br. J. Haematol..37: 489-501,1977. 24. Smyth, J. F., Poplack, D. G., Holiman, B. J., Leventhal, B. G., and Yarbro, G. Correlation of adenosinedeaminaseactivity with cell surface markers in acute lymphoblastic leukemia.J. Clin. Invest., 62: 710-712,1978. 25. Tveit, K. M., Fodstad, O., Lotsberg, J., Vaage, S., and Pihl, A. Colony growth and chemosensitivity in vitro of human melanoma biopsies, relationship to clinical parameters. Int. J. Cancer, 29: 533-538,1982. 26. Tveit, K. M., and Pihl, A. Do cell lines in vitro reflect the properties of the tumours of origin? A study of lines derived from human melanomaxenografts. Br. J. Cancer, 44:775-786,1981. 27. van de Griend, R. J., van der Reijden, H. J., Bolhuis, R. L., Melief, C. J., von dem Borne. A. E., and Roos, D. Enzyme analysis of lymphoproliferative diseases: a useful addition to cell surface phenotyping. Blood, 62: 669-676, 1983. 28. Wood, G. S., and Warnke. R. Suppression of endogenous avidin-bmding activity in tissues and its relevance to btotin-avidin detection systems. J. Histochem. Cytochem., 29:1196-1204,1981.

showed the same antigenic and enzymatic expression. Such changes may be due to spontaneous maturation of the T-cell lines in vitro. Thymocyte maturation is associated with a pattern of specific T-cell antigen and enzyme changes, and these changes can be induced in vitro by phorbol esters or other compounds (2,15,16). However, no known differentiation agent was used in this assay and no recognized pattern of T-cell differentiation was observed. Indeed, cellular changes associated with both thymocyte differentiation (CALLA loss, Leu 2a expres sion) and dedifferentiation (higher ADA levels, OKT9, and OKT10 expression) were observed (13,15, 27). Antigen discordance between the patient's malignant cells and the cell lines might be secondary to antigen modulation. Such modulation can occur in the presence of specific antibody or agents which induce differentiation (2,17). T-lymphoblasts from Patient 2 lacked cllsurface OKT9, but this antigen was found within the cytoplasm of these cells. Thus, the presence of cell surface OKT9 on the SUP-T2 cell line could have been secondary to extemalization of modulated antigens. The significance of this phenomenon in the patient's malignant cells is not known. Antigen and enzyme differences between the cell lines and the patient's tumor could be due to selection in vitro of a subpopu lation of cells (the colony-forming cells) which had a different biology from the majority of the patient's cells. These cells would be undetectable in the patient's tumor sample because such subpopulations (1 in 20,000 cells) are well below the sensitivity limits of the fluorescence-activated cell sorter and enzyme tech niques. Experiments are in progress to establish multiple cell lines (each from a single colony) from a single malignant sample and to compare the enzyme and antigenic profiles. The results of such studies may define better subpopulations within Tlymphoid cancers. An additional objective of this study was to determine whether a common antigenic profile exists for the malignant T-lymphoid colony-forming cell. In the 4 cell lines established, no single antigen or enzyme profile was found which could be used to select the malignant T-lymphoid colony-forming cell from nonproliferating cells in the patient's tumor sample. In conclusion, the present report established good agreement between in vitro growth of T-cells and the uncultured tumor cells as measured by ADA and NP levels and analysis of cellular antigens. However, the cell lines exhibited stable antigen and enzyme level changes which suggest that subpopulation selec tion or differentiation of the patient's cancer occurred in vitro.

REFERENCES
1. Bander, N. H. Comparison of antigen expression of human renal cancers in vitro. Cancer (Phila.),53:1235-1239,1984. 2. Cassel. D. L., Hoxie, J. A., and Cooper, R. A. Pnorbol ester modulation of Tcell antigens in the Jurkat lymphoblastic leukemia cell line. Cancer Res., 43. 4582-4586,1983. 3. Courtenay, V. D., Selby, P. J., Smith, I. E., Mills, J.. and Peckham, M. S.

CANCER

RESEARCH

VOL. 44 DECEMBER 5660

1984

Downloaded from cancerres.aacrjournals.org on September 20, 2011 Copyright 1984 American Association for Cancer Research