- Email: [email protected]
Clonal sublines of rat neurotumor RT4 and cell differentiation
DEVELOPMENTAL
BIOLOGY
104.240-246
(1984)
Clonal Sublines of Rat Neurotumor RT4 and Cell Differentiation VI. Chromosome
Analysis’
MARY M.HAAG,* SHIRLEY W. SoumP,t ANDNOBORU SUEOKA* *Department of Molecular, CeUular, and Develqwmental Biology, University of Colm&, Boulder, Colorado 80309,and tChi.!dren ‘s Hospital Research Foundation, Division of Human Genetics, Cincinnati, Ohio 45229 Received November 18, 1983;accepted in revised form Februnry 22, 198.4 The RT4 neurotumor cell system consists of clonally derived cell lines where a stem cell type segregates in vitro into three biochemically and morphologically different cell types, one glial and two neuronal types. This process has been termed cell-type conversion (M. Imada and N. Sueoka, 1978, Deu. BioL 66.9’7-108). Detailed cytogenetic analysis of the RT4 cell lines are described. Giemsa-banding analysis of 12 independent clonal isolates of the four different RT4 cell types showed a relatively stable karyotype. The stem cell line, RT4-AC, is diploid and most stable, and it has one 4q+ marker chromosome in place of a normal No. 4. This 4q+ marker was identified in all cell types of the RT4 system and was not observed in other cell lines of BDIX origin. The 4q+, therefore, is a chromosomal marker of the RT4 system. Consistent chromosome rearrangement was not found in any one of the cell-type conversions of the RTGAC cells into the three derivative cell types. The relative stability of the karyotype of the different clonal isolates gives the RT4 system an advantage in studies of genetic regulation and expression of cell-type conversion in vitro. Also the 4q+ marker can be used to identify RT4 cells in coculture experiments or to distinguish RT4 cells in cases of suspected cell-line contamination. INTRODUCTION
The RT4 peripheral nerve tumor was induced in a male BDIX rat by a single postnatal dose of the carcinogen ethylnitrosourea (ENU). Long-term culture of this tumor line and extensive cloning resulted in identification of a multipotential neural stem cell line. This cloned parental line (RTI-AC) spontaneously and repeatedly gives rise to three phenotypically distinct cell types (RT4-B, RTI-D, and RT4-E) with the frequency of 1 in about lo4 to ld RTI-AC cells. These three derivative cell lines are morphologically distinct from each other and from the parental stem cell line (RT4-AC). In addition, the derivative cell lines exhibit different biochemical properties with regard to neuronal and glial character (Imada and Sueoka, 1978; Imada et ah, 1978; Tomozawa and Sueoka, 1978). The stem cell line and the glial-type cell line maintain the ability to form tumors in host animals, while the two neuronal-type cell lines have lost this ability (Imada et ah, 1978). Properties of the four clonally derived cell types of the RT4 system have been well characterized and the segregation of phenotypes in culture has been termed cell-type conversion (Imada and Sueoka, 1978); the results so far obtained are summarized in Fig. 1. Cytogenetic studies of many human and animal tui Previous papers in this series: I (Imada and Sueoka, 19’7% II (Imada, Sueoka, and Rifkin, 1978), III (Tomozawa and Sueoka, 1978), IV (Imada and Sueoka, 1980), V (Tomozawa and Sueoka, submitted). 0012-1606194 $9.00 Copyright All rights
8 1984 by Academic Press, Inc. of reproduction in any form reserved.
240
mors and cell lines show a consistent, nonrandom involvement of specific chromosome abnormalities (Sandberg, 1980; Mitelman, 1980; Yunis, 1983). For the rat, cytogenetic studies of cell lines from ENU-induced tumors have shown a consistent involvement of chromosome 4 in numerical and/or structural abnormalities (Roscoe and Gibbs, 1974; Au et ah, 1977; Claisse et a& 1979; Haag and Soukup, 1984). A previous study of two phenotypically different cloned cell lines, which originated from the same ENU-induced rat glioma, suggested an association of chromosome 4 abnormalities with different cellular phenotypes in vitro (Claisse et cd, 1979). An earlier study of unbanded chromosome preparations of RT4 showed a predominantly normal 42-chromosome complement (Imada and Sueoka, 1978). We have since had the opportunity to perform Giemsa-banding analyses on several clonal isolates from the RT4 system to compare cell-type conversion and the possible involvement of acquired chromosome rearrangements. These G-banded results have shown that (1) examples for each derivative cell type have a diploid chromosome number in most or all cells of their population; (2) in place of one normal No. 4 there is a No. 4 chromosome marker designated 4q+ present in all cell lines from the RT4 system; (3) the stem cell line has a stable karyotype; and (4) there is no consistent difference in karyotype for the tumorigenic and nontumorigenic cell types. These results establish that a systematic chromosomal change does not account for cell-type conversion nor
HAAG, SOUKUP, AND SUEOKA
in warm 0.56% KC1 hypotonic solution for 14 min at 37”C, centrifuged, and fixed in a 3:l mixture of methanol and acetic acid. The metaphase spreads were prepared by dropping the cell suspension onto cold, wet slides about 20 hr after harvesting. Air-dried slides were Giemsa stained for examination of unbanded metaphases or Giemsa banded by pretreatment with trypsin then Giemsa staining (i.e., GTG banding) (Dutrillaux and Lejeune, 1975). Karyotypes were arranged according to the conventional methods of Levan (1974).
RT4-AC (Stem
cell
+ SlOOP, f + Na+ -influx
type) GFAP
Tumorigenic
/
RTI-B
(Glial-type) I
1
1
RT4-D
RT4-E
(Neuronal-type)
(Neuronal-type)
+ SlOOP, + GFAP - No’ -influx
- SlOOP, + No+ -influx
-
Non-tumorigenic
- SlOOP, + Na’ -influx + K’ -efflux
Excitable
Non-tumorigenic
K+-efflux
Tumorigenic
GFAP
FIG. 1. Cell-type conversion and conversion coupling in the RT4 family. In cell-type conversion, tumorigenicity is coupled with the expression of Sl99 protein (SloOP) and glial fibrillary acidic protein (GFAP), whereas cells with positive Na+ influx are not tumorigenic. The expression of these and morphological characteristics are “coupled” in that whenever one characteristic (e.g., the presence of Sl99P) is observed, tumorigenicity is observed. In all cases so far examined, when cells converted to neuronal type (i.e., absence of SloOP and GFAP and positive Na+ influx), tumorigenicity is negative. The coupling feature has been consistent in all cell lines of each cell type so far examined. In the RTI-AC line, the electrophysiological excitability and K+ efflux have not been analyzed. Electrophysiological excitability was indicated by a depolarization response to an anode break stimulation in cell lines tested. The diagram is based on our previous results (Imada and Sueoka, 1978; Imada et al, 1978; Tomozawa and Sueoka, 1978; Tomozawa and Sueoka, submitted).
does it correlate with reversion of tumorigenicity in the RT4 system. However, the 4q+ chromosome identified in this study serves as a valuable chromosomal marker of the RT4 lines. MATERIAL
RESULTS
GFAP
Excitable
, Non-excitable
241
Chromosome Analysis of RT4 Lines
Twelve samples of the four different RT4 clonal cell types were analyzed for chromosome constitution. Two to four chromosome preparations were made for each sample. Chromosome constitution for each cell line was based on at least 100 cells scored in the microscope, in the majority of cases, and on unbanded and Giemsabanded karyotypes. A summary of the results appears in Table 1. Tumorigenicity of cell lines is based on the data from Imada et al. (1978). The samples of the RTI-AC lines showed the chromosome constitution to be 42,XY,4q+, with occasional random loss of chromosomes. A minor subline was observed in RTIACl, with 6 cells out of 100 having 43,XY,+1,4q+. All banded karyotypes and cells showed a marker 4q+ in place of one normal No. 4 (Fig. 2). The extra chromosome material in the 4q+ marker could best be described as having a dark or positive band followed by a negative band at 4q42 (Fig. 3). We were not able to determine whether this band represents an insertion or a terminal duplication of chromosomal material and the source of the extra material could not be
TABLE 1 CHROMOSOME ANALYSIS OF RT4 TUMOR CELL LINES
AND METHODS
Chromosome analyses were carried out on cultures from frozen stocks of the cloned stem cell line, RT4-AC; the glial line, RT4-D; and the two neuronal cell lines, RT4-B and RT4-E. Each analysis represents findings from clonal isolates previously characterized for properties associated with cell-type conversion. Samples with different numbers represent different subclones from different passages of the lines (Imada and Sueoka, 1978; Imada et al, 1978; Tomozawa and Sueoka, 1978). For certain lines (RT4-AC3 and RT4-E5) duplicate samples of separately maintained lines were analyzed and the findings are presented individually. Metaphases were harvested 24 hr after subculture and feeding of cultures. Cells were treated with a final concentration of 0.1 pg/ml colcemid for 1 to 4 hr, trypsinized, and centrifuged. The cell pellet was suspended
NO.
No. of karyotypes
Chromosome abnormalities
Tumorigenic potential
Cell lines
Modal
counted
RT4-AC3 RT4-AC3 RT4-AC1
42 42 42/43’
120 100 101
17 6 11
a+ 4q+
+ +
4q+/+l
+
RT4-D3 RT4-D5 RT4-D6 RT4-D7
42 42 42143’ 42/43’
71 85 168
6 5 29 15
4q+ 4q+ 4q+/+4 4q+/+4
+ + + +
RT-E4 RT4-E5 RT4-E5
42 42 42/43d
101 100 170
5 15
4q+ ‘Q+ 4q+/+LwI,W
-
RT4-B7 RT4-B8
81-84’ 42f
8’7 103
2(4q+)/+l 4a+/la-.5a+
-
100
19 14 22
-
o Six percent of cells have 43; ‘25% of cells have 43; “43% of cells have 43; ‘20% of cells have 43, ‘36% of cells have +l; f30% of cells have del(1).
242
DEVELOPMENTALBIOLOGY
2
3
14
15
VOLUME104,19&i
6
-
25
16
17
18
SEX
HAAG, SOUKUP, AND SUEOKA
Chromosome Analysis of RTh Lines
identified. The 4q+ marker was subsequently seen in all the RT4 lines. This 4q+ marker was not observed after a review of six BDIX tumor cell lines from other sources (Haag and Soukup, 1984), nor in preparations of cultured BDIX fetal fibroblasts. This suggested that the 4q+ marker was not a constitutional abnormality nor polymorphism particular to this strain of rat. Four different clones of RTI-D, with the tumorigenic and glial phenotype, had 42,XY,4q+. No other abnormalities were consistently observed in this line (Fig. 4). However, in 25 and 43% of cells from RT4-D6 and RT4D7 respectively, there was mosaicism with sublines of 43,XY,4q+,+4. Three samples of the nontumorigenic neuronal clone of RTI-E had the 42,XY,4q+ karyotype (Fig. 5A). No mosaicism was seen in two samples. However, one sample, RT4-E5, had trisomy 1 in 60% of the cells and trisomy 1 plus a balanced translocation (2q,5q) in another 20% of the cells (Fig. 5B). Two clones of the RT4-B lines with the neuronal, nontumorigenic phenotype were analyzed. The first RTI-B clone, RT4-B7, was polyploid with a modal number of 81-84 and random loss of various chromosomes. Two copies of the 4q+ marker were observed. Examination of RTCB’I-banded karyotypes showed 5 out of 14 had pentasomy 1, and 4 others had a large unidentifiable marker. Approximately 10% of the second clone, RT4B8 was 42,XY,4q+ (Fig. 6A), and the other 90% of the cells in this line had a structural abnormality of one of the No. 1 chromosomes, characterized as an interstitial deletion of No. 1, del(l)(q12q22). A second rearrangement, identified as a 5q+ could only be characterized in a few banded karyotypes. It sometimes appeared to be t(5q,6q) (Fig. 6B). In summary, all cell lines had the marker 4q+, which was not observed in normal cultured rat tissue or in cell lines of other BDIX rat tumors. The tumorigenic lines had no further consistent abnormalities. The stem cell line, RTCAC showed remarkable stability of karyotype. In subpopulations of the RT4 lines examined, three lines had abnormalities notably of chromosomes 1, 2, and 4, and two lines had No. 5 involved in rearrangements in some cells. The nontumorigenic lines had more chromosome rearrangements in some of their cells, which persist after long-term culture; however, no consistent abnormality could be associated with the celltype conversion phenomenon.
243
The presence of the 4q+ marker in the RT4 lines has proved to be of practical value as a chromosomal marker. In studies of the effect of 5-azacytidine on the RT4 lines, myoblast-like cells were induced in the RT4-B cell type. The 4q+ marker was observed in these cells and established their RT4 identity ruling out cell-line contamination (Y. Tomozawa, unpublished). DISCUSSION
Cell culture of nervous system tumors has been a valuable means by which to study various aspects of neuronal and glial interaction and biochemical properties of the various types of cells. The mouse Cl300 neuroblastoma cell line has been extensively studied for various biochemical properties of neuronal cells (see Sato, 1973). However, the subclones of Cl300 have also developed heterogeneity of karyotype. The difference in karyotype as well as phenotype for the various subclones of the Cl300 makes it difficult to evaluate and compare developmental genetic regulation of these cells (Schubert, 1973). In contrast, ethylnitrosourea-induced rat numbers (Schubert et aL, 1974; Au, et al, 1977; Imada and Sueoka, 1978). In this report, we have been able to demonstrate a relatively stable, consistent karyotype for the four different phenotypes of the RT4 system by Giemsa-banding analysis. The stem cell line, RT4-AC, in particular, had a remarkably stable karyotype. The general occurrence and significance of karyotype stability in other stem cell lines, however, is currently not clear. This aspect of the different relatively stable subclones of the RT4 system, together with the multipotential nature of the stem cell line, RT4-AC, with regard to neuronal and glial properties, makes the RT4 lines significant for the study of the genetic regulation of branching of neuronal and glial-type cells in the nervous system. The clonal origin of most neoplasms has been demonstrated by cytogenetic, immunologic, and biochemical markers (Fialkow, 1974). During tumor development subpopulations arise resulting in genetic variants with different phenotypes; certain variants survive, while others are eliminated by immunological destruction or metabolic disadvantage (Nowell, 1976). Because there are many inherent difficulties in studying evolution of neoplasia in viva (Fogh, 1975), tumor lines in culture
FIG. 2. A representative karyotype of the RT4-AC cell line (the stem cell type). Arrow indicates the 4q+ marker, characteristic of all the RT4 cell lines. Otherwise the karyotype is normal 42,XY. FIG. 3. Pairs representing the normal No. 4 chromosome and the 4q+ marker from the RT4 lines. Arrows indicate the 4q+ marker of the pairs for comparison. FIG. 4. A representative karyotype of the RTI-D cell line (the tumorigenic glial type). Arrow indicates the 4q+ unique to the RT4 system, otherwise the karyotype is 42,XY.
244
DEVELOPMENTAL BIOLOGY
1
a
3
VOLUME 104,19&1
4A
5
6
FIG. 5. Representative karyotypes of RTI-E cells (one of the two neuronal type, nontumorigenic). (A) This is representative of typical pairs 1 through 6 of the RT4-E chromosomes. The arrow indicates the 4q+ marker of the RT4 lines. (B) This represents typical pairs 1 through 6 of RT4-E5 chromosomes in which an additional No. 1 was observed. A marker t(Zq,5q) and a missing normal No. 5 are indicated with small arrows; 4q+ is indicated with large arrow. FIG. 6. Representative karyotypes of RM-B cells (the second neuronal type, nontumorigenic). (A) This represents typical pairs 1 through 6 of RT4-B8 chromosomes. About 10% of the cells had only the 4q+ marker indicated with the arrow. (B) This represents chromosome pairs 1 through 6 of a subpopulation of RT4-B8 cells. Small arrows indicate secondary markers del(1) and a 5q+, large arrow indicates 4q+.
have been used to provide information about phenotype evolution and genetic variants (Yosida, 1983). The RT4 cell-line system appears to be an in vitro
counterpart to what may be occurring during tumor development in viva. The stem cell line, RTI-AC, spontaneously gives rise to other cell types with different
HAAG, SOUKUP, AND SUEOKA
Chromosome Analysis of RT4 Lines
morphological and biochemical phenotypes. The subclones, RTIB, RTGD, and RT4-E have differentiated properties (Fig. 1). However, these subclones maintain the ability to divide and form homogeneous populations. One cell type, RTI-D, which resembles the RT&AC parental line, maintains a tumorigenic phenotype and expresses differentiated glial functions (Imada et o& 1978). The other two cell types, RTI-B and RTGE, derived from the stem cell line by cell-type conversion, are revertant with respect to their tumorigenic potential and have differentiated neuronal properties which are also expressed to a lesser degree in RT4-AC (Tomozawa and Sueoka, 1978). Several reports have now shown the involvement of No. 4 in structural and/or numerical abnormalities in ENU-induced rat tumor cell lines (Roscoe and Gibbs, 1974;Au et uL, 19’77;Claisse et aL, 1979;Haag and Soukup, 1984). Our findings of a 4q+ marker consistently observed in all the RT4 lines provided additional evidence for No. 4 abnormalities in ENU-induced rat neurotumor lines. However, the 4q+ marker is present in both tumorigenic and nontumorigenic. lines of the RT4 system, suggesting no direct causative relationship of this chromosome abnormality to the tumorigenic phenotype for this cell system. The 4q-t marker was observed in all of the RT4-AC stem cell-line isolates and all of the other cell types of the RT4 syst,em so far examined. This suggests that this chromosomal rearrangement occurred early in tumor-cell evolution, and that the reversion of tumorigenicity of RT4-B and RT4-E by cell-type conversion of RT4-AC is a result of a change in some other regulatory mechanism in the RTCAC cells. Thus, tumorigenicity may have been obtained by an additional step after the initial effect of the carcinogen, ENU, on the genotype of the stem cell. Likewise, the reversion of tumorigenicity accompanying the cell-type conversion of RTCAC to RT4-B and RT4-E may have resulted by reversion of the second step by some secondary regulatory mechanism. It appears that the majority of the tumorigenic samples had normal karyotypes with a 4q+ marker. In a previous study of clones from an ENU-induced neurotumor cell line, the more tumorigenic clone was normal and the less tumorigenic line had +4 (Claisse et al, 1979). We had also observed subpopulations in some of the RTCAC and RT4-D samples with three copies of the No. 4 chromosome, one of which was the 4q+ marker. It seems that in the RT4 system, the 4q+ marker is the primary rearrangement, while other secondary rearrangements observed in our lines and those of other studies may confer a growth advantage to the cells under certain conditions, such as in culture (Mitelman et d, 1975; Claisse et al, 1979). This observation is partly in agreement with those of other studies showing that the karyotype of tumor cells with a regularly dividing cell
245
type is near diploid in the early stages of a tumor and that further chromosomal aberrations may be involved in tumor progression (Mitelman, 1974,198O;Wake et d, 1982; Yosida, 1983). Our study of the RT4-E and RT4-B lines show additional chromosome rearrangements in some clonal isolates which were not observed in RTI-AC lines. However, these rearrangements are not consistent or specific among the lines and therefore do not appear to be the cause of cell-type conversion of the RT4-AC line to the RTI-B and RT4-E cell types. Tendency of chromosome 1 rearrangements, which were observed in some of the RT4 lines, have been reported for other rat tumors and cell lines. The significance of this observation remains unknown (Levan, 1974; Mori and Sasaki, 1974; Kovi et ak, 1978; Wake et aL, 1982; Haag and Soukup, 1984). In conclusion, the cell-type conversions in the RT4 system cannot be explained by systematic chromosome alterations. The karyotype stability of the different clonal isolates gives the RT4 system an advantage in studies of genetic regulation and expression of cell-type conversion in vitro. Also, the 4q+ marker, unique to the RT4 lines, allows clear identification in studies using coculture and also allows us to distinguish RT4 cells from others in the case of suspected cell-line contamination. This work was supported by NIH Grant NS15304 and ACS Grant CD-l to N.S. and also by NC1 Grant ROl-CA18588 to S.S.; M.H. was supported by NIH F32-CA07120. REFERENCES Au, W., SOUKUP,S. W., and MANDYBUR,T. I. (1977).Excess chromosome #4 in ethylnitrosourea-induced neurogenic tumor lines of the rat. J. Nat. Cancer Inst. 59,1709-1716. CLAISSE,P. .I., ROSCOE,J. P., and LANTOS,P. L. (1979). Cellular heterogeneity in an ethylnitrosourea-induced glioma; malignancy, karyology and other properties of tumor cell types. Brit. J. Ezp. Pathd 60,209-224. DUTRILLAUX,B., and LEJEUNE,J. (1975). New techniques in the study of human chromosomes: Methods and applications. In “Advances in Human Genetics, 5” (H. Harris and K. Hirschhorn, eds.), pp. 119-156. Plenum, New York. FIALKOW,P. J. (1974). The origin and development of human tumors studied with cell markers. N. Engl J. Med 291,26-35. FOGH, J., ed. (1975). “Human Tumor Cells in vitro.” Plenum, New York. HAAG, M. M., and SOUICUP, S. W. (1984).The association of chromosome #4 abnormalities with ethylnitrosourea-induced neuro-oncogenesis in the rat. Cancer Res. 44,784-790. IMADA, M., and SUEOKA,N. (1978). Clonal sublines of rat neurotumor RT4 and cell differentiation. I. Isolation and characterization of cell lines and cell type conversion. Dev. Bid 66,97-108. IMADA, M., SUEOKA,N., and RIFKIN, D. (1978). Clonal sublines of rat neurotumor RT4 and cell differentiation. II. A conversion coupling of tumorigenicity and a glial property. Dev. Bid 66,109-116. IMADA, M., and SUEOKA,N. (1980). Cell surface proteins. IV. Den BioL 79,199-207. KOVI, J., KOM, E., MORRIS,H., and RAO,M. (1978).Chromosome banding
246
DEVELOPMENTALBIOLOGY
patterns and breakpoints of three transplantable bepatomas induced in rata by aromatic amines. J. Nat. Cancer Inst 61,495~506. LEVAN, G. (1974). Nomenclature for G-bands in rat chromosomes. Hem&as 77,37-52. MITELMAN, F. (1974). The rous sarcoma virus story: Cytogenetics of tumors induced by RSV. In “Chromosomes and Cancer” (J. German, ed.), pp. 675-686. Wiley, New York. MITELMAN, F. (1980). Cytogenetics of experimental neoplasms and non-random chromosome correlations in man. clinics HemutoL 9, 195-219. MITELMAN, F., LEVAN, G., and BRANDT,L. (1975). Highly malignant cells with normal karyotype in G-banding. Hereditas 80,219~293. MORI,M., and SASAIU,M. (1974). Chromosome studies on rat leukemias and lymphomas, with special attention to fluorescent karyotype analysis. J. Nat Cancer Inst. 52, 53-60. NOWELL,P. C. (1976). The clonal evolution of tumor cell populations. science 194,23-28.
ROSCOE, J. P., and GIBBS,B. E. (1974). Several changes associated with the acquisition of a single chromosome in rat glial tumor cells. J. Nat Caner Inst 63,581~583. SANDBERG,A. A. (1980). “The Chromosomes in Human Cancer and Leukemia.” Elsevier, New York.
VOLUME104, 1984
SATO,G., ed. (1973). “Tissue Culture of the Nervous System.” Plenum, New York. SCHUBERT,D., HARRIS,A. J., HEINEMANN,S., KIDOKORO,Y., PATRICK, J., and STEINBACH,J. H. (1973). Differentiation and interaction of clonal cell lines of nerve and muscle. In “Tissue Culture of the Nervous System” (G. Sato, ed.). Plenum, New York. SCHUBERT,D., HEINEMANN,S., CARLISLE,W., TARIKAS, H., KIMES, B., PATRICK,J., STEINBACK,J., CULP,W., and BRANDT,B. (1974). Clonal cell lines from the rat central nervous system. Nature (Londa) 249, 224-227. TOMOZAWA, Y., and SUEOICA,N. (1978). In uitrosegregation of different cell lines with neuronal and glial properties from a stem cell line of rat neurotumor RT4. Proc Nat Ami Sci USA 76,3305-6309 TOMOZAWA,Y., and SUEOKA,N. Distribution of neuronal and glial cell lines in Na+-influx and action potential. Submitted. WAKE, N., ISSACS,J., and SANDBERG, A. A. (1982). Chromosome changes associated with progression of the Dunning R-3327 rat prostatic adenocarcinoma system. Cancer Rea 42,4131-4142. YOSIDA,T. (1983). Karyotype evolution and tumor development. Cancer Genet.cytogen 8.153-179. YIJNIS,J. J. (1983).The chromosomal basis of human neoplasia. sciaoe 221,227-236.
BIOLOGY
104.240-246
(1984)
Clonal Sublines of Rat Neurotumor RT4 and Cell Differentiation VI. Chromosome
Analysis’
MARY M.HAAG,* SHIRLEY W. SoumP,t ANDNOBORU SUEOKA* *Department of Molecular, CeUular, and Develqwmental Biology, University of Colm&, Boulder, Colorado 80309,and tChi.!dren ‘s Hospital Research Foundation, Division of Human Genetics, Cincinnati, Ohio 45229 Received November 18, 1983;accepted in revised form Februnry 22, 198.4 The RT4 neurotumor cell system consists of clonally derived cell lines where a stem cell type segregates in vitro into three biochemically and morphologically different cell types, one glial and two neuronal types. This process has been termed cell-type conversion (M. Imada and N. Sueoka, 1978, Deu. BioL 66.9’7-108). Detailed cytogenetic analysis of the RT4 cell lines are described. Giemsa-banding analysis of 12 independent clonal isolates of the four different RT4 cell types showed a relatively stable karyotype. The stem cell line, RT4-AC, is diploid and most stable, and it has one 4q+ marker chromosome in place of a normal No. 4. This 4q+ marker was identified in all cell types of the RT4 system and was not observed in other cell lines of BDIX origin. The 4q+, therefore, is a chromosomal marker of the RT4 system. Consistent chromosome rearrangement was not found in any one of the cell-type conversions of the RTGAC cells into the three derivative cell types. The relative stability of the karyotype of the different clonal isolates gives the RT4 system an advantage in studies of genetic regulation and expression of cell-type conversion in vitro. Also the 4q+ marker can be used to identify RT4 cells in coculture experiments or to distinguish RT4 cells in cases of suspected cell-line contamination. INTRODUCTION
The RT4 peripheral nerve tumor was induced in a male BDIX rat by a single postnatal dose of the carcinogen ethylnitrosourea (ENU). Long-term culture of this tumor line and extensive cloning resulted in identification of a multipotential neural stem cell line. This cloned parental line (RTI-AC) spontaneously and repeatedly gives rise to three phenotypically distinct cell types (RT4-B, RTI-D, and RT4-E) with the frequency of 1 in about lo4 to ld RTI-AC cells. These three derivative cell lines are morphologically distinct from each other and from the parental stem cell line (RT4-AC). In addition, the derivative cell lines exhibit different biochemical properties with regard to neuronal and glial character (Imada and Sueoka, 1978; Imada et ah, 1978; Tomozawa and Sueoka, 1978). The stem cell line and the glial-type cell line maintain the ability to form tumors in host animals, while the two neuronal-type cell lines have lost this ability (Imada et ah, 1978). Properties of the four clonally derived cell types of the RT4 system have been well characterized and the segregation of phenotypes in culture has been termed cell-type conversion (Imada and Sueoka, 1978); the results so far obtained are summarized in Fig. 1. Cytogenetic studies of many human and animal tui Previous papers in this series: I (Imada and Sueoka, 19’7% II (Imada, Sueoka, and Rifkin, 1978), III (Tomozawa and Sueoka, 1978), IV (Imada and Sueoka, 1980), V (Tomozawa and Sueoka, submitted). 0012-1606194 $9.00 Copyright All rights
8 1984 by Academic Press, Inc. of reproduction in any form reserved.
240
mors and cell lines show a consistent, nonrandom involvement of specific chromosome abnormalities (Sandberg, 1980; Mitelman, 1980; Yunis, 1983). For the rat, cytogenetic studies of cell lines from ENU-induced tumors have shown a consistent involvement of chromosome 4 in numerical and/or structural abnormalities (Roscoe and Gibbs, 1974; Au et ah, 1977; Claisse et a& 1979; Haag and Soukup, 1984). A previous study of two phenotypically different cloned cell lines, which originated from the same ENU-induced rat glioma, suggested an association of chromosome 4 abnormalities with different cellular phenotypes in vitro (Claisse et cd, 1979). An earlier study of unbanded chromosome preparations of RT4 showed a predominantly normal 42-chromosome complement (Imada and Sueoka, 1978). We have since had the opportunity to perform Giemsa-banding analyses on several clonal isolates from the RT4 system to compare cell-type conversion and the possible involvement of acquired chromosome rearrangements. These G-banded results have shown that (1) examples for each derivative cell type have a diploid chromosome number in most or all cells of their population; (2) in place of one normal No. 4 there is a No. 4 chromosome marker designated 4q+ present in all cell lines from the RT4 system; (3) the stem cell line has a stable karyotype; and (4) there is no consistent difference in karyotype for the tumorigenic and nontumorigenic cell types. These results establish that a systematic chromosomal change does not account for cell-type conversion nor
HAAG, SOUKUP, AND SUEOKA
in warm 0.56% KC1 hypotonic solution for 14 min at 37”C, centrifuged, and fixed in a 3:l mixture of methanol and acetic acid. The metaphase spreads were prepared by dropping the cell suspension onto cold, wet slides about 20 hr after harvesting. Air-dried slides were Giemsa stained for examination of unbanded metaphases or Giemsa banded by pretreatment with trypsin then Giemsa staining (i.e., GTG banding) (Dutrillaux and Lejeune, 1975). Karyotypes were arranged according to the conventional methods of Levan (1974).
RT4-AC (Stem
cell
+ SlOOP, f + Na+ -influx
type) GFAP
Tumorigenic
/
RTI-B
(Glial-type) I
1
1
RT4-D
RT4-E
(Neuronal-type)
(Neuronal-type)
+ SlOOP, + GFAP - No’ -influx
- SlOOP, + No+ -influx
-
Non-tumorigenic
- SlOOP, + Na’ -influx + K’ -efflux
Excitable
Non-tumorigenic
K+-efflux
Tumorigenic
GFAP
FIG. 1. Cell-type conversion and conversion coupling in the RT4 family. In cell-type conversion, tumorigenicity is coupled with the expression of Sl99 protein (SloOP) and glial fibrillary acidic protein (GFAP), whereas cells with positive Na+ influx are not tumorigenic. The expression of these and morphological characteristics are “coupled” in that whenever one characteristic (e.g., the presence of Sl99P) is observed, tumorigenicity is observed. In all cases so far examined, when cells converted to neuronal type (i.e., absence of SloOP and GFAP and positive Na+ influx), tumorigenicity is negative. The coupling feature has been consistent in all cell lines of each cell type so far examined. In the RTI-AC line, the electrophysiological excitability and K+ efflux have not been analyzed. Electrophysiological excitability was indicated by a depolarization response to an anode break stimulation in cell lines tested. The diagram is based on our previous results (Imada and Sueoka, 1978; Imada et al, 1978; Tomozawa and Sueoka, 1978; Tomozawa and Sueoka, submitted).
does it correlate with reversion of tumorigenicity in the RT4 system. However, the 4q+ chromosome identified in this study serves as a valuable chromosomal marker of the RT4 lines. MATERIAL
RESULTS
GFAP
Excitable
, Non-excitable
241
Chromosome Analysis of RT4 Lines
Twelve samples of the four different RT4 clonal cell types were analyzed for chromosome constitution. Two to four chromosome preparations were made for each sample. Chromosome constitution for each cell line was based on at least 100 cells scored in the microscope, in the majority of cases, and on unbanded and Giemsabanded karyotypes. A summary of the results appears in Table 1. Tumorigenicity of cell lines is based on the data from Imada et al. (1978). The samples of the RTI-AC lines showed the chromosome constitution to be 42,XY,4q+, with occasional random loss of chromosomes. A minor subline was observed in RTIACl, with 6 cells out of 100 having 43,XY,+1,4q+. All banded karyotypes and cells showed a marker 4q+ in place of one normal No. 4 (Fig. 2). The extra chromosome material in the 4q+ marker could best be described as having a dark or positive band followed by a negative band at 4q42 (Fig. 3). We were not able to determine whether this band represents an insertion or a terminal duplication of chromosomal material and the source of the extra material could not be
TABLE 1 CHROMOSOME ANALYSIS OF RT4 TUMOR CELL LINES
AND METHODS
Chromosome analyses were carried out on cultures from frozen stocks of the cloned stem cell line, RT4-AC; the glial line, RT4-D; and the two neuronal cell lines, RT4-B and RT4-E. Each analysis represents findings from clonal isolates previously characterized for properties associated with cell-type conversion. Samples with different numbers represent different subclones from different passages of the lines (Imada and Sueoka, 1978; Imada et al, 1978; Tomozawa and Sueoka, 1978). For certain lines (RT4-AC3 and RT4-E5) duplicate samples of separately maintained lines were analyzed and the findings are presented individually. Metaphases were harvested 24 hr after subculture and feeding of cultures. Cells were treated with a final concentration of 0.1 pg/ml colcemid for 1 to 4 hr, trypsinized, and centrifuged. The cell pellet was suspended
NO.
No. of karyotypes
Chromosome abnormalities
Tumorigenic potential
Cell lines
Modal
counted
RT4-AC3 RT4-AC3 RT4-AC1
42 42 42/43’
120 100 101
17 6 11
a+ 4q+
+ +
4q+/+l
+
RT4-D3 RT4-D5 RT4-D6 RT4-D7
42 42 42143’ 42/43’
71 85 168
6 5 29 15
4q+ 4q+ 4q+/+4 4q+/+4
+ + + +
RT-E4 RT4-E5 RT4-E5
42 42 42/43d
101 100 170
5 15
4q+ ‘Q+ 4q+/+LwI,W
-
RT4-B7 RT4-B8
81-84’ 42f
8’7 103
2(4q+)/+l 4a+/la-.5a+
-
100
19 14 22
-
o Six percent of cells have 43; ‘25% of cells have 43; “43% of cells have 43; ‘20% of cells have 43, ‘36% of cells have +l; f30% of cells have del(1).
242
DEVELOPMENTALBIOLOGY
2
3
14
15
VOLUME104,19&i
6
-
25
16
17
18
SEX
HAAG, SOUKUP, AND SUEOKA
Chromosome Analysis of RTh Lines
identified. The 4q+ marker was subsequently seen in all the RT4 lines. This 4q+ marker was not observed after a review of six BDIX tumor cell lines from other sources (Haag and Soukup, 1984), nor in preparations of cultured BDIX fetal fibroblasts. This suggested that the 4q+ marker was not a constitutional abnormality nor polymorphism particular to this strain of rat. Four different clones of RTI-D, with the tumorigenic and glial phenotype, had 42,XY,4q+. No other abnormalities were consistently observed in this line (Fig. 4). However, in 25 and 43% of cells from RT4-D6 and RT4D7 respectively, there was mosaicism with sublines of 43,XY,4q+,+4. Three samples of the nontumorigenic neuronal clone of RTI-E had the 42,XY,4q+ karyotype (Fig. 5A). No mosaicism was seen in two samples. However, one sample, RT4-E5, had trisomy 1 in 60% of the cells and trisomy 1 plus a balanced translocation (2q,5q) in another 20% of the cells (Fig. 5B). Two clones of the RT4-B lines with the neuronal, nontumorigenic phenotype were analyzed. The first RTI-B clone, RT4-B7, was polyploid with a modal number of 81-84 and random loss of various chromosomes. Two copies of the 4q+ marker were observed. Examination of RTCB’I-banded karyotypes showed 5 out of 14 had pentasomy 1, and 4 others had a large unidentifiable marker. Approximately 10% of the second clone, RT4B8 was 42,XY,4q+ (Fig. 6A), and the other 90% of the cells in this line had a structural abnormality of one of the No. 1 chromosomes, characterized as an interstitial deletion of No. 1, del(l)(q12q22). A second rearrangement, identified as a 5q+ could only be characterized in a few banded karyotypes. It sometimes appeared to be t(5q,6q) (Fig. 6B). In summary, all cell lines had the marker 4q+, which was not observed in normal cultured rat tissue or in cell lines of other BDIX rat tumors. The tumorigenic lines had no further consistent abnormalities. The stem cell line, RTCAC showed remarkable stability of karyotype. In subpopulations of the RT4 lines examined, three lines had abnormalities notably of chromosomes 1, 2, and 4, and two lines had No. 5 involved in rearrangements in some cells. The nontumorigenic lines had more chromosome rearrangements in some of their cells, which persist after long-term culture; however, no consistent abnormality could be associated with the celltype conversion phenomenon.
243
The presence of the 4q+ marker in the RT4 lines has proved to be of practical value as a chromosomal marker. In studies of the effect of 5-azacytidine on the RT4 lines, myoblast-like cells were induced in the RT4-B cell type. The 4q+ marker was observed in these cells and established their RT4 identity ruling out cell-line contamination (Y. Tomozawa, unpublished). DISCUSSION
Cell culture of nervous system tumors has been a valuable means by which to study various aspects of neuronal and glial interaction and biochemical properties of the various types of cells. The mouse Cl300 neuroblastoma cell line has been extensively studied for various biochemical properties of neuronal cells (see Sato, 1973). However, the subclones of Cl300 have also developed heterogeneity of karyotype. The difference in karyotype as well as phenotype for the various subclones of the Cl300 makes it difficult to evaluate and compare developmental genetic regulation of these cells (Schubert, 1973). In contrast, ethylnitrosourea-induced rat numbers (Schubert et aL, 1974; Au, et al, 1977; Imada and Sueoka, 1978). In this report, we have been able to demonstrate a relatively stable, consistent karyotype for the four different phenotypes of the RT4 system by Giemsa-banding analysis. The stem cell line, RT4-AC, in particular, had a remarkably stable karyotype. The general occurrence and significance of karyotype stability in other stem cell lines, however, is currently not clear. This aspect of the different relatively stable subclones of the RT4 system, together with the multipotential nature of the stem cell line, RT4-AC, with regard to neuronal and glial properties, makes the RT4 lines significant for the study of the genetic regulation of branching of neuronal and glial-type cells in the nervous system. The clonal origin of most neoplasms has been demonstrated by cytogenetic, immunologic, and biochemical markers (Fialkow, 1974). During tumor development subpopulations arise resulting in genetic variants with different phenotypes; certain variants survive, while others are eliminated by immunological destruction or metabolic disadvantage (Nowell, 1976). Because there are many inherent difficulties in studying evolution of neoplasia in viva (Fogh, 1975), tumor lines in culture
FIG. 2. A representative karyotype of the RT4-AC cell line (the stem cell type). Arrow indicates the 4q+ marker, characteristic of all the RT4 cell lines. Otherwise the karyotype is normal 42,XY. FIG. 3. Pairs representing the normal No. 4 chromosome and the 4q+ marker from the RT4 lines. Arrows indicate the 4q+ marker of the pairs for comparison. FIG. 4. A representative karyotype of the RTI-D cell line (the tumorigenic glial type). Arrow indicates the 4q+ unique to the RT4 system, otherwise the karyotype is 42,XY.
244
DEVELOPMENTAL BIOLOGY
1
a
3
VOLUME 104,19&1
4A
5
6
FIG. 5. Representative karyotypes of RTI-E cells (one of the two neuronal type, nontumorigenic). (A) This is representative of typical pairs 1 through 6 of the RT4-E chromosomes. The arrow indicates the 4q+ marker of the RT4 lines. (B) This represents typical pairs 1 through 6 of RT4-E5 chromosomes in which an additional No. 1 was observed. A marker t(Zq,5q) and a missing normal No. 5 are indicated with small arrows; 4q+ is indicated with large arrow. FIG. 6. Representative karyotypes of RM-B cells (the second neuronal type, nontumorigenic). (A) This represents typical pairs 1 through 6 of RT4-B8 chromosomes. About 10% of the cells had only the 4q+ marker indicated with the arrow. (B) This represents chromosome pairs 1 through 6 of a subpopulation of RT4-B8 cells. Small arrows indicate secondary markers del(1) and a 5q+, large arrow indicates 4q+.
have been used to provide information about phenotype evolution and genetic variants (Yosida, 1983). The RT4 cell-line system appears to be an in vitro
counterpart to what may be occurring during tumor development in viva. The stem cell line, RTI-AC, spontaneously gives rise to other cell types with different
HAAG, SOUKUP, AND SUEOKA
Chromosome Analysis of RT4 Lines
morphological and biochemical phenotypes. The subclones, RTIB, RTGD, and RT4-E have differentiated properties (Fig. 1). However, these subclones maintain the ability to divide and form homogeneous populations. One cell type, RTI-D, which resembles the RT&AC parental line, maintains a tumorigenic phenotype and expresses differentiated glial functions (Imada et o& 1978). The other two cell types, RTI-B and RTGE, derived from the stem cell line by cell-type conversion, are revertant with respect to their tumorigenic potential and have differentiated neuronal properties which are also expressed to a lesser degree in RT4-AC (Tomozawa and Sueoka, 1978). Several reports have now shown the involvement of No. 4 in structural and/or numerical abnormalities in ENU-induced rat tumor cell lines (Roscoe and Gibbs, 1974;Au et uL, 19’77;Claisse et aL, 1979;Haag and Soukup, 1984). Our findings of a 4q+ marker consistently observed in all the RT4 lines provided additional evidence for No. 4 abnormalities in ENU-induced rat neurotumor lines. However, the 4q+ marker is present in both tumorigenic and nontumorigenic. lines of the RT4 system, suggesting no direct causative relationship of this chromosome abnormality to the tumorigenic phenotype for this cell system. The 4q-t marker was observed in all of the RT4-AC stem cell-line isolates and all of the other cell types of the RT4 syst,em so far examined. This suggests that this chromosomal rearrangement occurred early in tumor-cell evolution, and that the reversion of tumorigenicity of RT4-B and RT4-E by cell-type conversion of RT4-AC is a result of a change in some other regulatory mechanism in the RTCAC cells. Thus, tumorigenicity may have been obtained by an additional step after the initial effect of the carcinogen, ENU, on the genotype of the stem cell. Likewise, the reversion of tumorigenicity accompanying the cell-type conversion of RTCAC to RT4-B and RT4-E may have resulted by reversion of the second step by some secondary regulatory mechanism. It appears that the majority of the tumorigenic samples had normal karyotypes with a 4q+ marker. In a previous study of clones from an ENU-induced neurotumor cell line, the more tumorigenic clone was normal and the less tumorigenic line had +4 (Claisse et al, 1979). We had also observed subpopulations in some of the RTCAC and RT4-D samples with three copies of the No. 4 chromosome, one of which was the 4q+ marker. It seems that in the RT4 system, the 4q+ marker is the primary rearrangement, while other secondary rearrangements observed in our lines and those of other studies may confer a growth advantage to the cells under certain conditions, such as in culture (Mitelman et d, 1975; Claisse et al, 1979). This observation is partly in agreement with those of other studies showing that the karyotype of tumor cells with a regularly dividing cell
245
type is near diploid in the early stages of a tumor and that further chromosomal aberrations may be involved in tumor progression (Mitelman, 1974,198O;Wake et d, 1982; Yosida, 1983). Our study of the RT4-E and RT4-B lines show additional chromosome rearrangements in some clonal isolates which were not observed in RTI-AC lines. However, these rearrangements are not consistent or specific among the lines and therefore do not appear to be the cause of cell-type conversion of the RT4-AC line to the RTI-B and RT4-E cell types. Tendency of chromosome 1 rearrangements, which were observed in some of the RT4 lines, have been reported for other rat tumors and cell lines. The significance of this observation remains unknown (Levan, 1974; Mori and Sasaki, 1974; Kovi et ak, 1978; Wake et aL, 1982; Haag and Soukup, 1984). In conclusion, the cell-type conversions in the RT4 system cannot be explained by systematic chromosome alterations. The karyotype stability of the different clonal isolates gives the RT4 system an advantage in studies of genetic regulation and expression of cell-type conversion in vitro. Also, the 4q+ marker, unique to the RT4 lines, allows clear identification in studies using coculture and also allows us to distinguish RT4 cells from others in the case of suspected cell-line contamination. This work was supported by NIH Grant NS15304 and ACS Grant CD-l to N.S. and also by NC1 Grant ROl-CA18588 to S.S.; M.H. was supported by NIH F32-CA07120. REFERENCES Au, W., SOUKUP,S. W., and MANDYBUR,T. I. (1977).Excess chromosome #4 in ethylnitrosourea-induced neurogenic tumor lines of the rat. J. Nat. Cancer Inst. 59,1709-1716. CLAISSE,P. .I., ROSCOE,J. P., and LANTOS,P. L. (1979). Cellular heterogeneity in an ethylnitrosourea-induced glioma; malignancy, karyology and other properties of tumor cell types. Brit. J. Ezp. Pathd 60,209-224. DUTRILLAUX,B., and LEJEUNE,J. (1975). New techniques in the study of human chromosomes: Methods and applications. In “Advances in Human Genetics, 5” (H. Harris and K. Hirschhorn, eds.), pp. 119-156. Plenum, New York. FIALKOW,P. J. (1974). The origin and development of human tumors studied with cell markers. N. Engl J. Med 291,26-35. FOGH, J., ed. (1975). “Human Tumor Cells in vitro.” Plenum, New York. HAAG, M. M., and SOUICUP, S. W. (1984).The association of chromosome #4 abnormalities with ethylnitrosourea-induced neuro-oncogenesis in the rat. Cancer Res. 44,784-790. IMADA, M., and SUEOKA,N. (1978). Clonal sublines of rat neurotumor RT4 and cell differentiation. I. Isolation and characterization of cell lines and cell type conversion. Dev. Bid 66,97-108. IMADA, M., SUEOKA,N., and RIFKIN, D. (1978). Clonal sublines of rat neurotumor RT4 and cell differentiation. II. A conversion coupling of tumorigenicity and a glial property. Dev. Bid 66,109-116. IMADA, M., and SUEOKA,N. (1980). Cell surface proteins. IV. Den BioL 79,199-207. KOVI, J., KOM, E., MORRIS,H., and RAO,M. (1978).Chromosome banding
246
DEVELOPMENTALBIOLOGY
patterns and breakpoints of three transplantable bepatomas induced in rata by aromatic amines. J. Nat. Cancer Inst 61,495~506. LEVAN, G. (1974). Nomenclature for G-bands in rat chromosomes. Hem&as 77,37-52. MITELMAN, F. (1974). The rous sarcoma virus story: Cytogenetics of tumors induced by RSV. In “Chromosomes and Cancer” (J. German, ed.), pp. 675-686. Wiley, New York. MITELMAN, F. (1980). Cytogenetics of experimental neoplasms and non-random chromosome correlations in man. clinics HemutoL 9, 195-219. MITELMAN, F., LEVAN, G., and BRANDT,L. (1975). Highly malignant cells with normal karyotype in G-banding. Hereditas 80,219~293. MORI,M., and SASAIU,M. (1974). Chromosome studies on rat leukemias and lymphomas, with special attention to fluorescent karyotype analysis. J. Nat Cancer Inst. 52, 53-60. NOWELL,P. C. (1976). The clonal evolution of tumor cell populations. science 194,23-28.
ROSCOE, J. P., and GIBBS,B. E. (1974). Several changes associated with the acquisition of a single chromosome in rat glial tumor cells. J. Nat Caner Inst 63,581~583. SANDBERG,A. A. (1980). “The Chromosomes in Human Cancer and Leukemia.” Elsevier, New York.
VOLUME104, 1984
SATO,G., ed. (1973). “Tissue Culture of the Nervous System.” Plenum, New York. SCHUBERT,D., HARRIS,A. J., HEINEMANN,S., KIDOKORO,Y., PATRICK, J., and STEINBACH,J. H. (1973). Differentiation and interaction of clonal cell lines of nerve and muscle. In “Tissue Culture of the Nervous System” (G. Sato, ed.). Plenum, New York. SCHUBERT,D., HEINEMANN,S., CARLISLE,W., TARIKAS, H., KIMES, B., PATRICK,J., STEINBACK,J., CULP,W., and BRANDT,B. (1974). Clonal cell lines from the rat central nervous system. Nature (Londa) 249, 224-227. TOMOZAWA, Y., and SUEOICA,N. (1978). In uitrosegregation of different cell lines with neuronal and glial properties from a stem cell line of rat neurotumor RT4. Proc Nat Ami Sci USA 76,3305-6309 TOMOZAWA,Y., and SUEOKA,N. Distribution of neuronal and glial cell lines in Na+-influx and action potential. Submitted. WAKE, N., ISSACS,J., and SANDBERG, A. A. (1982). Chromosome changes associated with progression of the Dunning R-3327 rat prostatic adenocarcinoma system. Cancer Rea 42,4131-4142. YOSIDA,T. (1983). Karyotype evolution and tumor development. Cancer Genet.cytogen 8.153-179. YIJNIS,J. J. (1983).The chromosomal basis of human neoplasia. sciaoe 221,227-236.