Int. J. Cancer: 52, 130-136 (1992) 0 1992 Wiley-Liss, Inc.
'
Publicationof the InternationalUnion Against Cancer Publicationde I'Unlon lnternationaleContre le Cancer
JUMPING TRANSLOCATIONS ORIGINATE CLONAL REARRANGEMENTS IN SV4O-TRANSFORMED HUMAN FIBROBLASTS F. HOFFSCHIR', M. R I C O U LN. ~ , LEMIEUX-', S. ESTRADE?, R. CASSINGENA? and B. DUTRILLAUX'.~ 'DSV-DPTE-LCG, Centre d'Etudes Nidkaires de Fontenay-am-Roses, 68 avenue de la Division Leclerc, 92265 Fontenay-aiu-Roses; 'UPR 42, Institut de Recherches Scientijiqicessur le Cancer, B P No8, 94802 Vrllejuif Ckdex; and -'URA 620 CNRS, Institut Curie, Section de Biologie, 26 n4e d'Ulm, 75231 Paris Cedex 05, France. A comparative study of chromosomalrearrangements occurring in 4 independent clones obtained from SV40-transformed cornea and skin human fibroblasts was performed. Rearrangements principally affect some constitutive heterochromatin and, to a lesser degree, telomeric regions. This results in multiple exchanges between a limited number of chromosome structures, i.e., in jumping translocations. Such rearrangements occur even at early passages and some of them give rise to clonal rearrangements that accumulate at late passages. This process is responsible for progressive modification of the karyotypes, principally characterized by deletion of a number of chromosome segments. Thus, clonal rearrangements are selected among many others not occurring at random. The selective pressure retaining clonal rearrangement seems to be similar for the 4 independent clones, since selection of the derivative chromosomes leads to the same imbalances, whatever the origin of the clone. This sequence of events recalls that of human solid tumors, since jumping rearrangements are generally observed in pre-malignantconditions or in low-grade malignancies, whereas clonal rearrangements leading to typical imbalances are detected in more advanced malignant tumors. o 1992 Wiley-Liss, Inc.
In addition to hematological and a limited number of soft-tissue malignancies, in which specific chromosomal rearrangements are frequently observed as the sole anomaly, in vim-transformed or solid-tumor cells carry multiple structural and numerical anomalies. The complexity of the chromosomal changes was for long interpreted as being the result of gross disorganization of the genome, without real biological meaning. However, the recent development of cytogenetic and molecular studies indicated the highly recurrent character of several anomalies in a given tumor type. This results in a characteristic pattern of imbalances permitting new possibilities of expression of recessive mutations. For instance, nonexpression of the RB1 gene in retinoblastoma and overexpression of a mutated TP53 gene in colorectal carcinoma are correlated with deletions of chromosomes 13 and 17 (Cavenee et al., 1983; Baker et al., 1989; Muleris et al., 1985; Strong et al., 1981) on which these genes are mapped respectively. In some cases, it was possible to reconstruct approximately the sequence of chromosome changes from diploidy to aneuploidy, but the interpretation always remained speculative and simplified (Muleris et al., 1990). To understand the dynamics of chromosome modifications, in vztro systems of cell transformation may provide interesting information, since they also progressively lead to complex karyotypie changes. As for human cancer cells, these changes were initially considered to occur at random, but more cautious analyses indicated that this was not the case. For instance, SV40-transformed human epithelial cells or fibroblasts acquire characteristic patterns of chromosome imbalances (Rodgers et af., 1983; Christian et al., 1987; Hoffschir et al., 1988; Wu et al., 1988; Stacey et al., 1990). We developed a prospective study of karyotypic modifications of 4 clones of SV40-transformed human fibroblasts to reconstruct the sequence of the progressive change of their karyotypes. It indicates that the characteristic patterns of imbalances, observed after many cell passages (Hoffschir et al., 1988), are preceded by non-random occurrence of chromosome rearrangements at early passages. These rearrange-
ments, preferentially affecting some telomeric and centromeric regions, and resulting in multiple translocations of some chromosomes in a limited number of combinations, correspond to jumping translocations as originally defined by Lejeune ef al. (1979). They were described as either telomeric associations or jumping translocations in various non-malignant (Dutrillaux ef al., 1978), pre-malignant (Hayashi and Schmid, 1975) and malignant (Mandahl et al., 1985; Kovacs et al., 1987; Raimondi ef al., 1987; Aledo et al., 1989) conditions.
MATERIAL AND METHODS
Primary human fibroblast cultures, obtained from cornea (CHS) and skin (DHS), were transformed upon transfection with pSVHBl recombinant plasmid D N A harboring a complete but replication-defective SV40 genome (Saint-Ruf et al., 1989). Briefly, semi-confluent primary cultures, grown in DMEM supplemented with 10% newborn-calf serum in 25cm? Falcon (Oxnard, CA) plastic flasks, were transfected with 0.5 ml per culture of DNA-calcium-phosphate co-precipitate containing 20 Fg of carrier calf thymus DNA plus 2.5 Fg of pSVHBl DNA. After 6 hr at 3 7 T , the medium was removed and 5 ml of fresh medium per flask were added. The cultures were re-fed every 3 o r 4 days. Morphologically transformed foci appeared after 3 to 4 weeks' incubation, at which time they were cloned and regularly sub-cultured once a week, with 1 or 2 medium changes between each sub-culture. The expression of intranuclear SV40-specific T-antigen in 100% of transformed cells was assessed, at different cell passage levels, by indirect immunofluorescence (Wicker and Avrameas, 1969) with Syrian hamster anti-SV40 tumor serum and fluoresceinisothiocyanate-conjugated rabbit anti-serum to hamster y-globulin. Two independent clones from each transfection assay were retained and cultured for 38 to 61 passages: CHSV3 and CHSV4 (population doubling -56 hr); DHSV2 and DHSV4 (population doubling approx. 48 hr). Cytogenetic studies were performed at the various passages indicated in Figure 1. For each passage studied, 10 to 20 R-banded karyotypes were established, using our usual methods (Dutrillaux and Couturier, 1981). In most instances, the origin of all derivative chromosomes could be identified. All karyotypes were compared with each other in order to differentiate clonal from non-clonal rearrangements. We defined as clonal a rearrangement or a gain observed in 2 and a chromosome loss observed in 4/10 metaphases at least. Clonal rearrangements and chromosome gains or losses were used to reconstruct the chromosomal evolution of each clone. When a single chromosome band was involved in different rearrangements at the same passage in one third of the karyotypes at least, it was considered as being involved in a jumping rearrangement. A single chromosome band could be involved in both jumping and clonal rearrangements.
Received: February 17, 1992 and in revised form April 16,1992.
131
JUMPING REARRANGEMENTS IN SV4O-TRANSFORMED FIBROBLASTS CHSV3 ~asssge~
CHSV4
Clonal E v ~ l ~ f i o n
Clonal EVOlUIlO"
DHSV2 Passages
Clonal E~oluLlOn
170-
8
5
17
r - 2
12
18
20
31
36 21
48
34 52 31
58
44
61
FIGURE1 - Cytogenetic information and reconstruction of chromosomal evolution at various passages, indicated at the left of each scheme corresponding to CHSV3, CHSV4, DHSV2 and DHSV4 cultures. The schemes reconstitute the progressive accumulation of clonal anomalies, the number of karyotyped descendant cells with a given clonal anomaly being indicated in small circles. Endoreduplications are represented by a cross in a circle. RESULTS
Most results are summarized in Figure 1, which needs extensive description. For the 4 clones, 304 karyotypes were established: 110 for CHSV3; 94 for CHSV4 and 50 for each DHSV2 and DHSV4. The numbers of karyotypes characterized by clonal chromosomal anomalies are indicated, at each passage. in open circles. Since the earliest passages studied (4 to 8 depending on the clones), all the karyotypes studied exhibited chromosome anomalies. All karyotypes differed from cach other by one anomaly at least. Most rearrangements were unbalanced. and resulted in the formation of one derivativc from 2 normal chromosomes. This also resulted in a decrease of chromosome numbers. This decrease was progressive and paralleled the occurrence of clonal rearrangements. For clone CHSV3, it reached about 35 chromosomes at passage 36 (Table I) and was followed by endoreduplication between passages 36 and 48. The decrease of chromosome number continued after endoreduplication, reaching 55 chromoiomes at passage 61. A similar evolution occurred for clone DHSV?. whereas it was slower for clones CHSV4 and DHSV4, which did not yet undergo endoreduplication at passage 44. For each clone, we have reconstructed the evolution of the karyotypes from the progressive accumulation of clonal rearrangements and chromosome losses or gains. They are indicated in the branches of Figure 1. The average numbers of non-clonal chromosome rearrangements acquired between 2 passages for each metaphase are also given in Table I (N Rea). They ranged from 0.9 to 9 depending on the passage studied. In clones CHSV3, CHSV4 and DHSV4 there was a large proportion of tetraploid cells at passages 17-20, i.e., just before
crisis, but in all 3 it was a near-diploid line which escaped the crisis. For clone CHSV3, a second crisis occurred after passage 30.
Clonal anomalies After the comparison between CHSV3 and CHSV4, on the one hand, and between DHSV2 and DHSV4, on the other hand (Fig. 2), no identical clonal rearrangements were detectable. Thus, rearrangements were probably formed after the establishment of the independent cultures. However, the same chromosome bands tended to be affected by clonal rearrangements, and this similarity is approximately as close between CHSV and DHSV clones as between clones of CHS or DHS origin. This similarity may be related to the limited number of structures affected, in particular constitutive heterochromatin and, to a lesser degree, telomeric regions: among the 118 breakpoints from clonal rearrangements, 74 were located in constitutive heterochromatin, including 22 short arms of acrocentrics, 19 in telomeric regions and only 25 in the intercalary posit ion. Jumping rearrangements As for clonal rearrangements, the same structures of the same chromosomes had a very strong tendency to be affected from passage to passage and from clones DHSV to CHSV (Figs. 2 and 3). Among the 137 structures involved in jumping rearrangement at one passage, 101 were constitutive heterochromatin, including 43 short arms of acrocentrics, 26 telomeric regions and 10 intercalary regions. Almost all rearrangements were of chromosome type and unbalanced. They were most frequently observed at passages around the crisis.
r-!
4 tl 01 PI
-
a
-
N N
u^ m
t r-
c
N
5
T
00
- -m c
r-I
4
133
JUMPING REARRANGEMENTS IN SV40-TRANSFORMED FIBROBLASTS
1
13
9
lo
14
15
0.
11
12
21
22
13
Y
FIGURE 2 -Localization of the breaks resulting in jumping and clonal rearrangements, observed in the 4 clones (CHSV2, CHSV4, DHSV? and DHSV4). 0,jumping rearrangements, the breakpoint of which is apparently identical to that of a clonal rearrangement other jumping rearrangements; 0,clonal rearrangements the occurring at either the same or preceding or following passage studied; 0, breakpoint of which corresponds to that of jumping rearrangements occurring at either the same or preceding or following passage studied; A, other clonal rearrangements. Numbers inside symbols indicate the number of rearrangements of a same category affecting the same band.
134
HOFFSCHI R E T A L .
FIGURE 3 -Jumping translocations involving chromosomes 3, 10, 11, 15, 18 and 22 selected from 10 karyotypes of CHSV4. at passage 12.
JUMPING REARRANGEMENTS IN SV40-TRANSFORMED FIBROBLASTS
Comparison of clonal and jumping rearrangements Both types of rearrangement shared a number of characteristics: involvement of the same structures of the same chromosomes, with an excess of constitutive heterochromatin and telomeric regions; imbalances leading most frequently to deficiencies; occurrence varying with passages, being low at early and late passages and high around the 20th (crisis). They also differed in a number of characteristics: jumping translocations occurred slightly earlier than clonal rearrangements, on the average; the same was true for the involvement of a given structure, which, when involved in jumping rearrangements at a givcn passagc, was frequently involved in a clonal rearrangement at a later passage; involvement of telomeric regions was significantly more frequent in jumping (261137 = 19%) than in clonal (131104 = 8.3%) rearrangements ( x 2 = 5.53, v = 1, p < .02); intercalary regions were less frequently involved in jumping (101137 = 7.3%) than in clonal (251108 = 23.1%) rearrangements (x’ = 9.14, v = 1 , p < ,005). Finally, the relationship between jumping and clonal rearrangements appears quite obvious. Jumping rearrangements occur first, and a proportion of them give rise to the clonal rearrangements. Some of these clonal rearrangements can spread among all descendent cells, others among clones with variable extension and lifespan. Most frequently, when a clonal rearrangement appears, the jumping character is lost at later passages. DISCUSSION
The numerous anomalies occurring in SV40-transformed human cells clearly do not occur at random. In fibroblasts, a very typical pattern of chromosome imbalances has been already described (Hoffschir et af., 1988). In epithelial cells of mammary, skin, intestinal or urinary-tract origin, recurrent anomalies have also been described, and it was even noticed that similarities with cancerous cells of the same origin might exist (Rodgers et al., 1983; Meisner et al., 1988; Stacey et al., 1990). In SV40-transformed fibroblasts, we found a large excess of unbalanced rearrangements resulting from breakage in constitutive heterochromatin. The same observations were reported for epithelial human tumors such as colorectal, breast and lung carcinomas (Muleris et al., 1985; Dutrillaux et al., 1990; Viegas-Pequignot et al., 1990). Indeed. these rearrangements, which affect many juxtacentromeric regions are not specific for a givcn chromosome. They are not supposed to be involved in oncogene activation, as in a number of hematological malignancies. but merely reflect accidents of tumor progression. In human fresh tumors, all intermediate situations between the absence of clonal chromosome anomalies and the presence of clonal rearrangements shared by all metaphases exist. Multiple miniclones were sometimes observed, as for instance in squamous-cell carcinomas of the skin or of the oral cavity (Alcdo rt al., 1989; Heim et af., 1988; Jin et al., 1990). Jumping
135
translocations were also observed, and it was noticed that telomeric and heterochromatic regions were frequently involved (Aledo et al., 1989). End-to-end translocations were also described in pre-malignant conditions such as ataxia telangiectasia and in various malignancies (Kovacs et al., 1987; Mandahl et af., 1985; Pathak et af., 1988; Fitzgerald and Morris, 1984; Raimondi et aL, 1987). Finally, jumping rearrangements and end-to-end dicentrics (which frequently correspond to jumping rearrangements) seem to occur principally in premalignant conditions and in low-grade malignancies, whereas clonal rearrangements are rather characteristic of more advanced malignancies. Data on SV40-transformed cells strongly suggest that jumping gives rise to clonal rearrangements: jumping rearrangements, indicating hot spots of chromosome rearrangements, already occur at very early passages, become very frequent during the period of crisis, and less frequent when the cells are immortalized. Clonal rearrangements follow this scheme with a short delay and the fact that the same chromosome structures are involved in both events highly suggests that jumping originates clonal rearrangements, probably giving rise to multiple miniclones, of which one or a limited number is finally selected to form major clones. If this interpretation is correct, it means that clonal rcarrangements identify chromosomal segments where a high rate of rearrangement has occurred at an early stage of the malignant process. The causes of this chromosome instability remain to be elucidated, but our results with in situ hybridization of a SV40 probe (data not shown) and others indicate that the instability is obviously not limited to the site of insertion of the viral genome (Hara and Kaji, 1987; Romani et a/., 1990). Following this stage, cells would be submitted to a strong selective pressure, favoring replacement of one karyotypic formula by another in a relatively few cell generations. Indeed, this selection would have the tendency to retain the same imbalances, i.e., the same combinations of derivative chromosomes, explaining why the patterns of imbalance are quite similar from cell line to cell line, although the rearrangements differ. This interpretation is at contrast with that of Goolsby et al. (1991) who found unbalanced karyotypes in SV40transformed human fibroblasts after implantation in nude mice only. Our results show that unbalances can occur at early passages after SV40 infection, and that they are not always late events, as suggested by Meisner et al. (1988), Wu et af. (1988) and Goolsby et al. (1991). Deficiencies for 8p and l l p arms observed in 314 and of 6q, lop, 13q, 17p, 18p observed in 214 clones in this study were also recurrently observed in SV40transformed fibroblasts old cell lines, for which we postulated that the selective pressure might be related to early metabolic deviations (Hoffschir et aL, 1988) as in solid tumors (Dutrillaux and Couturier, 1981). This hypothesis has been partially demonstrated for colorectal adenocarcinoma (Bardot et al., 1991) and for SV40-transformed cells (Bravard et a/., 1992).
REFERENCES Ai.rno, R., ALIRIAS,A,, AVRIL.M.F. and DUTRILLAUX, B.. Jumping end-to-end dicrntrics in a case of sauamous-cell carcinoma from a patient with Xeroderma pigmentosum. Cancer Genet. Cyrogener., 40.
relation to their chromosomal patterns. Int. J. Cuncer, 47, 670-674 (1991). BRAVARD, A,, SABATIER, L., HOFFSCHIR, F., RICOUL, M., LLICCIONI, C. 95-103 (1989). and DUTRILLAUX, B., SOD2: a new type of tumor-suppressor gene? BAKER,S.J., F ~ A R O N E.R., , NIGRO,J.M.. HAMILTON, S.R., PREIS- Inf. J. Cancer, 52, 1-5 (1992). I N G E R , A.C., JESSUP. J.M., VANTUINEN. P., LEDBETTER. D.H., BARKER, CAVENEE. W.K., DRYJA,T.P., PHII.IPPS.R.A., BENEDICT, W.F., GODD.F., NAKAMURA, Y., WHITE,R. and VOGELSTEIN, B., Chromosome 17 BOUT, R., GALLIE,B.L., MURPHREE, A.L.. STRONG. L.C. and WHITE, deletions and p53 gene mutations in colorectal carcinomas. Science, R.L., Expression of recessive alleles by chromosomal mechanisms in 244,217-220 (1989). retinoblastoma. Nature (Lond.),305,779-783 (1983). BARDOT. V., LUCCIONI, C., LEFRANCOIS, D.. MULERIS, M. and DUTRIL- CHRISTIAN, B.J., LORETZ,L.J., OBERLY, T.D. and REZNIKOFF, C., A LAUS, B., Activity of thymidylate synthetase, thymidine kinase and characterization of human epithelial cells immortalized by simian virus galactokinase in primary and xenografted human colorectal cancers in 40. CuncerRes., 47,6066-6073 (1987).
136
HOFFSCHIR ET AL.
DUTRILLAUX, B. and COUTURIER, J., La pratique de I'analyse chromosomiquc. Masson, Paris (1981). E., AURIAS, DUTRILLAUX, B., CROQUETTE, M.F., VIEGAS-PEQUIGNOT, A,, COGET,J., COUTURIER, J. and LEJEUNE,J., Human somatic chromosome chains and rings. Cytogenet. CeN Genet., 20,70-77 (1978). DLJTRILLAUX, B., GERBAULT-SEUREAU, M. and ZAFRANI, B., Characterization of chromosomal anomalies in human breast cancer. A comparison of 30 paradiploid cases with few chromosome changes. Cancer Genet. Cytogenet.,49,203-217 (1990). FITZGERALD, P.H. and MORRIS, C.M., Telomeric association of chromosomes in B-cell lymphoid leukemia. Hum. Genet., 67, 385-390 (1984). GOOLSBY, C.L., WILEY,J.E., STEINER, M., BARTHOLDI, M.F., CRAM, L.S. and KRAEMER, P.M., Karyotype evolution in a simian-virus-40transformed tumorigenic human cell line. Cancer Genet. Cytogenet.,49, 231-248 (1991). HARA, H. and M I , H., Random integration of SV40 in SV40transformed, immortalized human fibroblasts. Exp. Cell Res., 168, 531-538 (1987). HAYASHI, K. and SCHMID, W., Tandem duplication q14 and dicentric formation by end to end chromosome fusions in ataxia telangiectasia (AT). Hum. Genef., 30,135-141 (1975). HEIM,S., JIN, Y.,MANDAHL, N., BIOKLUND, A,, WENNERBERG, J., JONSSON, N. and MITELMAN, F., Multiple unrelated clonal chromosome abnormalities in an in situ squamous cell carcinoma of the skin. Cancer Genef. Cytogenet.,36,149-153 (1988). HOFFSCHIR, F., RrcouL, M. and DUTRILLAUX, B.. SV-40-transformed human fibroblasts exhibit a characteristic chromosomal pattern. Cytogenet. Cell Genet., 49,264268 (1988). JIN,Y., HEIM,S., MANDAHL, N., BIOKLUND, A,, WENNERBERG, J. and MITELMAN, F., Unrelated clonal chromosomal aberrations in carcinomas of the oral cavity. Genes, Chromosotnes arid Cancer, 1. 209-215 (1990). G., MULLER-BRECHLIN. R. and SZUCS,S.. Telomeric associaKOVACS. tion in two human renal tumors. Cancer Genet. Cytogmef.,28,363-366 (1987). J., MAUNOURY, C., PRIEUR,M. and VANDEN AKKER,J., LEJELNE, Translocation sauteuse (5p;15q), (8q;15q), (12q:15q).Ann. Gknet., 22, 210-213 (1979). V., MITELMAN, F., ROOSER, MANDAHL, N., HEIM,S., KRISTOFFERSON, B., RYDHOLM, A. and WILLEN, H., Telomeric association in a malignant fibrous histiocytoma. Hum. Genet., 71,321-324 (1985). MEISNER, L.F., WU, S., CHRISTIAN, B.J. and REZNIKOFF, C., Cytogenetic instability with balanced chromosome changes in an SV40-
transformed human uroepithelial cell line. Cancer Res., 48,3215-3220 (1988). MULERIS, M., DELATTRE, O., OLSCHWANG, S., DUTRILLAUX, A.M., Y., SALMON,R.J., THOMAS,G. and DUTRILLAUX, B., REMVIKOS, Cytogenetic and molecular approaches of polyploidization in colorectal adenocarcinomas. Cancer Genet. Cytogt.net.,44, 107-1 18 (1990). MULERIS, M., SALMON, R.J., ZAFRANI, B., GIRODET, J. and DUTRILLAUX, B., Consistent deficiencies of chromosome 18 and of the short arm of chromosome 17 in 11 cases of human large-bowel cancer: a possible recessive determinism. Ann. Ginit., 28,206-213 (1985). PATHAK, S., WANG,Z., DHALIWAL, M.K. and SAKS,P.C., Telomeric association: another characteristic of cancer chromosomes'? Cytogenet. Cell Genet., 47,227-229 (1988). RAIMONDI, S.C., RAGSTALE, S.T., BEHM,F., RIVERA, G. and WILLIAMS, D.L.. Multiple telomeric associations of a trisomic whole q arm of chromosome 1 in a child with acute lymphoblastic leukemia. Cancer Genef.Cytogenef..24,87-93 (1987). RODGERS, C.S., SALLY,M.H., HULTEN.M.A.. CHANG,S.E., KEEN,J. and PAPADIMITRIOU, J.T., Cytogenetic analysis of SV40-transformed human breast epithelial cells. Cancer Genet. Cytogenet., 8, 213-221 (1983). ROMANI,M., DE AMBROSIS, A,, ALHADEFF,B., PURRELLO, M., GLUZMAN. Y. and SINISCALCO. M., Preferential integration of the Ad5iSV40 hybrid virus at the highly recombinogenic human chromosomal site lp36. Gene, 95,231-241 (1990). SAINT-RUF. C., NARDEUX. P., CEBRIAN, J., LACASA, M., LAVIALLE, C. and CASSINGENA, R.. Molecular cloning and early-region sequence analysis of replication-defective SV40 DNA from human HBL- 100 cells. Int. J. Cancer, 44,367-372 (1989). STACEY, M., GALLIMORE,P.H.. McCoNviLLt, C. and TAYLOR, A.M.R., Rearrangement of the same chromosome regions in different SV40transformed skin keratinocytes lines is associated with tumorigenicity. Oncogene, 5,727-739 (1990). STRONG,L.C., RICCARDI, V.M.. FERRELL,R.E. and SPARKES,R.S., Familial retinoblastoma and chromosome 13 deletion transmitted via an insertional translocation. Science, 213, 1501-1503 (1981). VIEGAS-PEQUIGNOT, E., FLURY-HERARD, A,, DE CREMOUX,H., CHLECQ, C., BIGNON, J. and DUTRILLAUX, B., Recurrent chromosome aberrations in human lung squamous-cell carcinomas. Cancer Genet. Cytogenet., 49,37-49 (1990). WICKER, R. and AVRAMEAS, S., Localization of virus antigens by enzyme-labelled antib0dies.J. gen. Virol.,4,465-471 (1969). Wu, S., CHRISTIAN, B.J., REZNIKOFF, C. and MEISNER, L., Marker chromosome stability associated with neoplastic transformation of human uroepithelial cells. Cancer Genet. Cytogenet., 36, 77-87 (1988).
'
Publicationof the InternationalUnion Against Cancer Publicationde I'Unlon lnternationaleContre le Cancer
JUMPING TRANSLOCATIONS ORIGINATE CLONAL REARRANGEMENTS IN SV4O-TRANSFORMED HUMAN FIBROBLASTS F. HOFFSCHIR', M. R I C O U LN. ~ , LEMIEUX-', S. ESTRADE?, R. CASSINGENA? and B. DUTRILLAUX'.~ 'DSV-DPTE-LCG, Centre d'Etudes Nidkaires de Fontenay-am-Roses, 68 avenue de la Division Leclerc, 92265 Fontenay-aiu-Roses; 'UPR 42, Institut de Recherches Scientijiqicessur le Cancer, B P No8, 94802 Vrllejuif Ckdex; and -'URA 620 CNRS, Institut Curie, Section de Biologie, 26 n4e d'Ulm, 75231 Paris Cedex 05, France. A comparative study of chromosomalrearrangements occurring in 4 independent clones obtained from SV40-transformed cornea and skin human fibroblasts was performed. Rearrangements principally affect some constitutive heterochromatin and, to a lesser degree, telomeric regions. This results in multiple exchanges between a limited number of chromosome structures, i.e., in jumping translocations. Such rearrangements occur even at early passages and some of them give rise to clonal rearrangements that accumulate at late passages. This process is responsible for progressive modification of the karyotypes, principally characterized by deletion of a number of chromosome segments. Thus, clonal rearrangements are selected among many others not occurring at random. The selective pressure retaining clonal rearrangement seems to be similar for the 4 independent clones, since selection of the derivative chromosomes leads to the same imbalances, whatever the origin of the clone. This sequence of events recalls that of human solid tumors, since jumping rearrangements are generally observed in pre-malignantconditions or in low-grade malignancies, whereas clonal rearrangements leading to typical imbalances are detected in more advanced malignant tumors. o 1992 Wiley-Liss, Inc.
In addition to hematological and a limited number of soft-tissue malignancies, in which specific chromosomal rearrangements are frequently observed as the sole anomaly, in vim-transformed or solid-tumor cells carry multiple structural and numerical anomalies. The complexity of the chromosomal changes was for long interpreted as being the result of gross disorganization of the genome, without real biological meaning. However, the recent development of cytogenetic and molecular studies indicated the highly recurrent character of several anomalies in a given tumor type. This results in a characteristic pattern of imbalances permitting new possibilities of expression of recessive mutations. For instance, nonexpression of the RB1 gene in retinoblastoma and overexpression of a mutated TP53 gene in colorectal carcinoma are correlated with deletions of chromosomes 13 and 17 (Cavenee et al., 1983; Baker et al., 1989; Muleris et al., 1985; Strong et al., 1981) on which these genes are mapped respectively. In some cases, it was possible to reconstruct approximately the sequence of chromosome changes from diploidy to aneuploidy, but the interpretation always remained speculative and simplified (Muleris et al., 1990). To understand the dynamics of chromosome modifications, in vztro systems of cell transformation may provide interesting information, since they also progressively lead to complex karyotypie changes. As for human cancer cells, these changes were initially considered to occur at random, but more cautious analyses indicated that this was not the case. For instance, SV40-transformed human epithelial cells or fibroblasts acquire characteristic patterns of chromosome imbalances (Rodgers et af., 1983; Christian et al., 1987; Hoffschir et al., 1988; Wu et al., 1988; Stacey et al., 1990). We developed a prospective study of karyotypic modifications of 4 clones of SV40-transformed human fibroblasts to reconstruct the sequence of the progressive change of their karyotypes. It indicates that the characteristic patterns of imbalances, observed after many cell passages (Hoffschir et al., 1988), are preceded by non-random occurrence of chromosome rearrangements at early passages. These rearrange-
ments, preferentially affecting some telomeric and centromeric regions, and resulting in multiple translocations of some chromosomes in a limited number of combinations, correspond to jumping translocations as originally defined by Lejeune ef al. (1979). They were described as either telomeric associations or jumping translocations in various non-malignant (Dutrillaux ef al., 1978), pre-malignant (Hayashi and Schmid, 1975) and malignant (Mandahl et al., 1985; Kovacs et al., 1987; Raimondi ef al., 1987; Aledo et al., 1989) conditions.
MATERIAL AND METHODS
Primary human fibroblast cultures, obtained from cornea (CHS) and skin (DHS), were transformed upon transfection with pSVHBl recombinant plasmid D N A harboring a complete but replication-defective SV40 genome (Saint-Ruf et al., 1989). Briefly, semi-confluent primary cultures, grown in DMEM supplemented with 10% newborn-calf serum in 25cm? Falcon (Oxnard, CA) plastic flasks, were transfected with 0.5 ml per culture of DNA-calcium-phosphate co-precipitate containing 20 Fg of carrier calf thymus DNA plus 2.5 Fg of pSVHBl DNA. After 6 hr at 3 7 T , the medium was removed and 5 ml of fresh medium per flask were added. The cultures were re-fed every 3 o r 4 days. Morphologically transformed foci appeared after 3 to 4 weeks' incubation, at which time they were cloned and regularly sub-cultured once a week, with 1 or 2 medium changes between each sub-culture. The expression of intranuclear SV40-specific T-antigen in 100% of transformed cells was assessed, at different cell passage levels, by indirect immunofluorescence (Wicker and Avrameas, 1969) with Syrian hamster anti-SV40 tumor serum and fluoresceinisothiocyanate-conjugated rabbit anti-serum to hamster y-globulin. Two independent clones from each transfection assay were retained and cultured for 38 to 61 passages: CHSV3 and CHSV4 (population doubling -56 hr); DHSV2 and DHSV4 (population doubling approx. 48 hr). Cytogenetic studies were performed at the various passages indicated in Figure 1. For each passage studied, 10 to 20 R-banded karyotypes were established, using our usual methods (Dutrillaux and Couturier, 1981). In most instances, the origin of all derivative chromosomes could be identified. All karyotypes were compared with each other in order to differentiate clonal from non-clonal rearrangements. We defined as clonal a rearrangement or a gain observed in 2 and a chromosome loss observed in 4/10 metaphases at least. Clonal rearrangements and chromosome gains or losses were used to reconstruct the chromosomal evolution of each clone. When a single chromosome band was involved in different rearrangements at the same passage in one third of the karyotypes at least, it was considered as being involved in a jumping rearrangement. A single chromosome band could be involved in both jumping and clonal rearrangements.
Received: February 17, 1992 and in revised form April 16,1992.
131
JUMPING REARRANGEMENTS IN SV4O-TRANSFORMED FIBROBLASTS CHSV3 ~asssge~
CHSV4
Clonal E v ~ l ~ f i o n
Clonal EVOlUIlO"
DHSV2 Passages
Clonal E~oluLlOn
170-
8
5
17
r - 2
12
18
20
31
36 21
48
34 52 31
58
44
61
FIGURE1 - Cytogenetic information and reconstruction of chromosomal evolution at various passages, indicated at the left of each scheme corresponding to CHSV3, CHSV4, DHSV2 and DHSV4 cultures. The schemes reconstitute the progressive accumulation of clonal anomalies, the number of karyotyped descendant cells with a given clonal anomaly being indicated in small circles. Endoreduplications are represented by a cross in a circle. RESULTS
Most results are summarized in Figure 1, which needs extensive description. For the 4 clones, 304 karyotypes were established: 110 for CHSV3; 94 for CHSV4 and 50 for each DHSV2 and DHSV4. The numbers of karyotypes characterized by clonal chromosomal anomalies are indicated, at each passage. in open circles. Since the earliest passages studied (4 to 8 depending on the clones), all the karyotypes studied exhibited chromosome anomalies. All karyotypes differed from cach other by one anomaly at least. Most rearrangements were unbalanced. and resulted in the formation of one derivativc from 2 normal chromosomes. This also resulted in a decrease of chromosome numbers. This decrease was progressive and paralleled the occurrence of clonal rearrangements. For clone CHSV3, it reached about 35 chromosomes at passage 36 (Table I) and was followed by endoreduplication between passages 36 and 48. The decrease of chromosome number continued after endoreduplication, reaching 55 chromoiomes at passage 61. A similar evolution occurred for clone DHSV?. whereas it was slower for clones CHSV4 and DHSV4, which did not yet undergo endoreduplication at passage 44. For each clone, we have reconstructed the evolution of the karyotypes from the progressive accumulation of clonal rearrangements and chromosome losses or gains. They are indicated in the branches of Figure 1. The average numbers of non-clonal chromosome rearrangements acquired between 2 passages for each metaphase are also given in Table I (N Rea). They ranged from 0.9 to 9 depending on the passage studied. In clones CHSV3, CHSV4 and DHSV4 there was a large proportion of tetraploid cells at passages 17-20, i.e., just before
crisis, but in all 3 it was a near-diploid line which escaped the crisis. For clone CHSV3, a second crisis occurred after passage 30.
Clonal anomalies After the comparison between CHSV3 and CHSV4, on the one hand, and between DHSV2 and DHSV4, on the other hand (Fig. 2), no identical clonal rearrangements were detectable. Thus, rearrangements were probably formed after the establishment of the independent cultures. However, the same chromosome bands tended to be affected by clonal rearrangements, and this similarity is approximately as close between CHSV and DHSV clones as between clones of CHS or DHS origin. This similarity may be related to the limited number of structures affected, in particular constitutive heterochromatin and, to a lesser degree, telomeric regions: among the 118 breakpoints from clonal rearrangements, 74 were located in constitutive heterochromatin, including 22 short arms of acrocentrics, 19 in telomeric regions and only 25 in the intercalary posit ion. Jumping rearrangements As for clonal rearrangements, the same structures of the same chromosomes had a very strong tendency to be affected from passage to passage and from clones DHSV to CHSV (Figs. 2 and 3). Among the 137 structures involved in jumping rearrangement at one passage, 101 were constitutive heterochromatin, including 43 short arms of acrocentrics, 26 telomeric regions and 10 intercalary regions. Almost all rearrangements were of chromosome type and unbalanced. They were most frequently observed at passages around the crisis.
r-!
4 tl 01 PI
-
a
-
N N
u^ m
t r-
c
N
5
T
00
- -m c
r-I
4
133
JUMPING REARRANGEMENTS IN SV40-TRANSFORMED FIBROBLASTS
1
13
9
lo
14
15
0.
11
12
21
22
13
Y
FIGURE 2 -Localization of the breaks resulting in jumping and clonal rearrangements, observed in the 4 clones (CHSV2, CHSV4, DHSV? and DHSV4). 0,jumping rearrangements, the breakpoint of which is apparently identical to that of a clonal rearrangement other jumping rearrangements; 0,clonal rearrangements the occurring at either the same or preceding or following passage studied; 0, breakpoint of which corresponds to that of jumping rearrangements occurring at either the same or preceding or following passage studied; A, other clonal rearrangements. Numbers inside symbols indicate the number of rearrangements of a same category affecting the same band.
134
HOFFSCHI R E T A L .
FIGURE 3 -Jumping translocations involving chromosomes 3, 10, 11, 15, 18 and 22 selected from 10 karyotypes of CHSV4. at passage 12.
JUMPING REARRANGEMENTS IN SV40-TRANSFORMED FIBROBLASTS
Comparison of clonal and jumping rearrangements Both types of rearrangement shared a number of characteristics: involvement of the same structures of the same chromosomes, with an excess of constitutive heterochromatin and telomeric regions; imbalances leading most frequently to deficiencies; occurrence varying with passages, being low at early and late passages and high around the 20th (crisis). They also differed in a number of characteristics: jumping translocations occurred slightly earlier than clonal rearrangements, on the average; the same was true for the involvement of a given structure, which, when involved in jumping rearrangements at a givcn passagc, was frequently involved in a clonal rearrangement at a later passage; involvement of telomeric regions was significantly more frequent in jumping (261137 = 19%) than in clonal (131104 = 8.3%) rearrangements ( x 2 = 5.53, v = 1, p < .02); intercalary regions were less frequently involved in jumping (101137 = 7.3%) than in clonal (251108 = 23.1%) rearrangements (x’ = 9.14, v = 1 , p < ,005). Finally, the relationship between jumping and clonal rearrangements appears quite obvious. Jumping rearrangements occur first, and a proportion of them give rise to the clonal rearrangements. Some of these clonal rearrangements can spread among all descendent cells, others among clones with variable extension and lifespan. Most frequently, when a clonal rearrangement appears, the jumping character is lost at later passages. DISCUSSION
The numerous anomalies occurring in SV40-transformed human cells clearly do not occur at random. In fibroblasts, a very typical pattern of chromosome imbalances has been already described (Hoffschir et af., 1988). In epithelial cells of mammary, skin, intestinal or urinary-tract origin, recurrent anomalies have also been described, and it was even noticed that similarities with cancerous cells of the same origin might exist (Rodgers et al., 1983; Meisner et al., 1988; Stacey et al., 1990). In SV40-transformed fibroblasts, we found a large excess of unbalanced rearrangements resulting from breakage in constitutive heterochromatin. The same observations were reported for epithelial human tumors such as colorectal, breast and lung carcinomas (Muleris et al., 1985; Dutrillaux et al., 1990; Viegas-Pequignot et al., 1990). Indeed. these rearrangements, which affect many juxtacentromeric regions are not specific for a givcn chromosome. They are not supposed to be involved in oncogene activation, as in a number of hematological malignancies. but merely reflect accidents of tumor progression. In human fresh tumors, all intermediate situations between the absence of clonal chromosome anomalies and the presence of clonal rearrangements shared by all metaphases exist. Multiple miniclones were sometimes observed, as for instance in squamous-cell carcinomas of the skin or of the oral cavity (Alcdo rt al., 1989; Heim et af., 1988; Jin et al., 1990). Jumping
135
translocations were also observed, and it was noticed that telomeric and heterochromatic regions were frequently involved (Aledo et al., 1989). End-to-end translocations were also described in pre-malignant conditions such as ataxia telangiectasia and in various malignancies (Kovacs et al., 1987; Mandahl et af., 1985; Pathak et af., 1988; Fitzgerald and Morris, 1984; Raimondi et aL, 1987). Finally, jumping rearrangements and end-to-end dicentrics (which frequently correspond to jumping rearrangements) seem to occur principally in premalignant conditions and in low-grade malignancies, whereas clonal rearrangements are rather characteristic of more advanced malignancies. Data on SV40-transformed cells strongly suggest that jumping gives rise to clonal rearrangements: jumping rearrangements, indicating hot spots of chromosome rearrangements, already occur at very early passages, become very frequent during the period of crisis, and less frequent when the cells are immortalized. Clonal rearrangements follow this scheme with a short delay and the fact that the same chromosome structures are involved in both events highly suggests that jumping originates clonal rearrangements, probably giving rise to multiple miniclones, of which one or a limited number is finally selected to form major clones. If this interpretation is correct, it means that clonal rcarrangements identify chromosomal segments where a high rate of rearrangement has occurred at an early stage of the malignant process. The causes of this chromosome instability remain to be elucidated, but our results with in situ hybridization of a SV40 probe (data not shown) and others indicate that the instability is obviously not limited to the site of insertion of the viral genome (Hara and Kaji, 1987; Romani et a/., 1990). Following this stage, cells would be submitted to a strong selective pressure, favoring replacement of one karyotypic formula by another in a relatively few cell generations. Indeed, this selection would have the tendency to retain the same imbalances, i.e., the same combinations of derivative chromosomes, explaining why the patterns of imbalance are quite similar from cell line to cell line, although the rearrangements differ. This interpretation is at contrast with that of Goolsby et al. (1991) who found unbalanced karyotypes in SV40transformed human fibroblasts after implantation in nude mice only. Our results show that unbalances can occur at early passages after SV40 infection, and that they are not always late events, as suggested by Meisner et al. (1988), Wu et af. (1988) and Goolsby et al. (1991). Deficiencies for 8p and l l p arms observed in 314 and of 6q, lop, 13q, 17p, 18p observed in 214 clones in this study were also recurrently observed in SV40transformed fibroblasts old cell lines, for which we postulated that the selective pressure might be related to early metabolic deviations (Hoffschir et aL, 1988) as in solid tumors (Dutrillaux and Couturier, 1981). This hypothesis has been partially demonstrated for colorectal adenocarcinoma (Bardot et al., 1991) and for SV40-transformed cells (Bravard et a/., 1992).
REFERENCES Ai.rno, R., ALIRIAS,A,, AVRIL.M.F. and DUTRILLAUX, B.. Jumping end-to-end dicrntrics in a case of sauamous-cell carcinoma from a patient with Xeroderma pigmentosum. Cancer Genet. Cyrogener., 40.
relation to their chromosomal patterns. Int. J. Cuncer, 47, 670-674 (1991). BRAVARD, A,, SABATIER, L., HOFFSCHIR, F., RICOUL, M., LLICCIONI, C. 95-103 (1989). and DUTRILLAUX, B., SOD2: a new type of tumor-suppressor gene? BAKER,S.J., F ~ A R O N E.R., , NIGRO,J.M.. HAMILTON, S.R., PREIS- Inf. J. Cancer, 52, 1-5 (1992). I N G E R , A.C., JESSUP. J.M., VANTUINEN. P., LEDBETTER. D.H., BARKER, CAVENEE. W.K., DRYJA,T.P., PHII.IPPS.R.A., BENEDICT, W.F., GODD.F., NAKAMURA, Y., WHITE,R. and VOGELSTEIN, B., Chromosome 17 BOUT, R., GALLIE,B.L., MURPHREE, A.L.. STRONG. L.C. and WHITE, deletions and p53 gene mutations in colorectal carcinomas. Science, R.L., Expression of recessive alleles by chromosomal mechanisms in 244,217-220 (1989). retinoblastoma. Nature (Lond.),305,779-783 (1983). BARDOT. V., LUCCIONI, C., LEFRANCOIS, D.. MULERIS, M. and DUTRIL- CHRISTIAN, B.J., LORETZ,L.J., OBERLY, T.D. and REZNIKOFF, C., A LAUS, B., Activity of thymidylate synthetase, thymidine kinase and characterization of human epithelial cells immortalized by simian virus galactokinase in primary and xenografted human colorectal cancers in 40. CuncerRes., 47,6066-6073 (1987).
136
HOFFSCHIR ET AL.
DUTRILLAUX, B. and COUTURIER, J., La pratique de I'analyse chromosomiquc. Masson, Paris (1981). E., AURIAS, DUTRILLAUX, B., CROQUETTE, M.F., VIEGAS-PEQUIGNOT, A,, COGET,J., COUTURIER, J. and LEJEUNE,J., Human somatic chromosome chains and rings. Cytogenet. CeN Genet., 20,70-77 (1978). DLJTRILLAUX, B., GERBAULT-SEUREAU, M. and ZAFRANI, B., Characterization of chromosomal anomalies in human breast cancer. A comparison of 30 paradiploid cases with few chromosome changes. Cancer Genet. Cytogenet.,49,203-217 (1990). FITZGERALD, P.H. and MORRIS, C.M., Telomeric association of chromosomes in B-cell lymphoid leukemia. Hum. Genet., 67, 385-390 (1984). GOOLSBY, C.L., WILEY,J.E., STEINER, M., BARTHOLDI, M.F., CRAM, L.S. and KRAEMER, P.M., Karyotype evolution in a simian-virus-40transformed tumorigenic human cell line. Cancer Genet. Cytogenet.,49, 231-248 (1991). HARA, H. and M I , H., Random integration of SV40 in SV40transformed, immortalized human fibroblasts. Exp. Cell Res., 168, 531-538 (1987). HAYASHI, K. and SCHMID, W., Tandem duplication q14 and dicentric formation by end to end chromosome fusions in ataxia telangiectasia (AT). Hum. Genef., 30,135-141 (1975). HEIM,S., JIN, Y.,MANDAHL, N., BIOKLUND, A,, WENNERBERG, J., JONSSON, N. and MITELMAN, F., Multiple unrelated clonal chromosome abnormalities in an in situ squamous cell carcinoma of the skin. Cancer Genef. Cytogenet.,36,149-153 (1988). HOFFSCHIR, F., RrcouL, M. and DUTRILLAUX, B.. SV-40-transformed human fibroblasts exhibit a characteristic chromosomal pattern. Cytogenet. Cell Genet., 49,264268 (1988). JIN,Y., HEIM,S., MANDAHL, N., BIOKLUND, A,, WENNERBERG, J. and MITELMAN, F., Unrelated clonal chromosomal aberrations in carcinomas of the oral cavity. Genes, Chromosotnes arid Cancer, 1. 209-215 (1990). G., MULLER-BRECHLIN. R. and SZUCS,S.. Telomeric associaKOVACS. tion in two human renal tumors. Cancer Genet. Cytogmef.,28,363-366 (1987). J., MAUNOURY, C., PRIEUR,M. and VANDEN AKKER,J., LEJELNE, Translocation sauteuse (5p;15q), (8q;15q), (12q:15q).Ann. Gknet., 22, 210-213 (1979). V., MITELMAN, F., ROOSER, MANDAHL, N., HEIM,S., KRISTOFFERSON, B., RYDHOLM, A. and WILLEN, H., Telomeric association in a malignant fibrous histiocytoma. Hum. Genet., 71,321-324 (1985). MEISNER, L.F., WU, S., CHRISTIAN, B.J. and REZNIKOFF, C., Cytogenetic instability with balanced chromosome changes in an SV40-
transformed human uroepithelial cell line. Cancer Res., 48,3215-3220 (1988). MULERIS, M., DELATTRE, O., OLSCHWANG, S., DUTRILLAUX, A.M., Y., SALMON,R.J., THOMAS,G. and DUTRILLAUX, B., REMVIKOS, Cytogenetic and molecular approaches of polyploidization in colorectal adenocarcinomas. Cancer Genet. Cytogt.net.,44, 107-1 18 (1990). MULERIS, M., SALMON, R.J., ZAFRANI, B., GIRODET, J. and DUTRILLAUX, B., Consistent deficiencies of chromosome 18 and of the short arm of chromosome 17 in 11 cases of human large-bowel cancer: a possible recessive determinism. Ann. Ginit., 28,206-213 (1985). PATHAK, S., WANG,Z., DHALIWAL, M.K. and SAKS,P.C., Telomeric association: another characteristic of cancer chromosomes'? Cytogenet. Cell Genet., 47,227-229 (1988). RAIMONDI, S.C., RAGSTALE, S.T., BEHM,F., RIVERA, G. and WILLIAMS, D.L.. Multiple telomeric associations of a trisomic whole q arm of chromosome 1 in a child with acute lymphoblastic leukemia. Cancer Genef.Cytogenef..24,87-93 (1987). RODGERS, C.S., SALLY,M.H., HULTEN.M.A.. CHANG,S.E., KEEN,J. and PAPADIMITRIOU, J.T., Cytogenetic analysis of SV40-transformed human breast epithelial cells. Cancer Genet. Cytogenet., 8, 213-221 (1983). ROMANI,M., DE AMBROSIS, A,, ALHADEFF,B., PURRELLO, M., GLUZMAN. Y. and SINISCALCO. M., Preferential integration of the Ad5iSV40 hybrid virus at the highly recombinogenic human chromosomal site lp36. Gene, 95,231-241 (1990). SAINT-RUF. C., NARDEUX. P., CEBRIAN, J., LACASA, M., LAVIALLE, C. and CASSINGENA, R.. Molecular cloning and early-region sequence analysis of replication-defective SV40 DNA from human HBL- 100 cells. Int. J. Cancer, 44,367-372 (1989). STACEY, M., GALLIMORE,P.H.. McCoNviLLt, C. and TAYLOR, A.M.R., Rearrangement of the same chromosome regions in different SV40transformed skin keratinocytes lines is associated with tumorigenicity. Oncogene, 5,727-739 (1990). STRONG,L.C., RICCARDI, V.M.. FERRELL,R.E. and SPARKES,R.S., Familial retinoblastoma and chromosome 13 deletion transmitted via an insertional translocation. Science, 213, 1501-1503 (1981). VIEGAS-PEQUIGNOT, E., FLURY-HERARD, A,, DE CREMOUX,H., CHLECQ, C., BIGNON, J. and DUTRILLAUX, B., Recurrent chromosome aberrations in human lung squamous-cell carcinomas. Cancer Genet. Cytogenet., 49,37-49 (1990). WICKER, R. and AVRAMEAS, S., Localization of virus antigens by enzyme-labelled antib0dies.J. gen. Virol.,4,465-471 (1969). Wu, S., CHRISTIAN, B.J., REZNIKOFF, C. and MEISNER, L., Marker chromosome stability associated with neoplastic transformation of human uroepithelial cells. Cancer Genet. Cytogenet., 36, 77-87 (1988).
