Chromosome 12 in human testicular cancer: Dosage changes and their parental origin

Chromosome 12 in Human Testicular Cancer: Dosage Changes and Their Parental Origin P iivi Peltom iki, Ragnhild A. Lothe, Anne-Lise Borresen, Sophie D. FossA, Anton Brogger, and Albert de la Chapelle

ABSTRACT: Cytogenetically, a marker chromosome interpreted as i(12p) is present in most testicular tumors of germ cell origin. In this study, 22 patients with testicular germ-cell tumors were investigated by Southern blot hybridization to characterize changes in chromosome 12. In comparison with normal DNA, tumor DNA of 18 patients showed increased dosages of 12p accompanied by a comparable or smaller increase or no change in the dosage of centromeric sequences of chromosome 12. A likely interpretation was that most testicular tumors had one or several isochromosomes for 12p that were formed by somatic division of the centromere and that the points of breakage and reunion in the centromeric region were different in different tumors. Allelic 12p fragments showing increased intensity were paternal in four and maternal in three of seven informative cases. Thus, there was no evidence of sex-limited parental imprinting. Furthermore, the observed patterns of allelic fragments suggested that the marker was an i(12p) formed by sister chromatids of one homolog number 12 rather than the result of interchange of genetic material between different homologues.

INTRODUCTION A marker interpreted as i(12p) has been found in more than 80% of testicular germ cell tumors by cytogenetic techniques [1, 2] and is considered specific for tumors of germ-cell origin [3]. How the marker arises and what genetic consequences it may have is not known. Alleles can function differently depending on whether they originate from the mother or the father. This is due to an epigenetic marking process called imprinting [4]. Several mouse chromosomes contain imprinted genes [5]. Human chromosome 12 has regions homologous with mouse chromosomes 5, 6, 10, and 15 [6]. Genes with homology to loci on human chromosome 12 have not yet been described as being imprinted. Nonrandom loss of maternal alleles in some human tumors [7-9] suggests inactivation, by imprinting, of the paternal allele of the tumor suppressor gene in question. Parental origin may have a profound effect on the clinical features of a disease, as is evident in the PraderWilli and Angelman syndromes [10]. In the present study, chromosome 12-specific changes associated with testicular germ cell malignancy were characterized by comparing the involvement of the short arm, From the Department of Medical Genetics (P. P., A. de la C.), University of Helsinki, Helsinki, Finland, Department of Genetics (R. A. L., A.-L. B., A. B.), Institute for Cancer Research, and Department of Medical Oncology and Radiotherapy (S. D. F.), the Norwegian Radium Hospital, Oslo, Norway. Address reprint requests to: Dr. P~iivi PeltomC~ki, Department of Medical Genetics, University of Helsinki, Haartmaninkatu 3, 00290 Helsinki, Finland. Received April 27, 1992; accepted July 17, 1992.

the centromeric region, and the long arm in tumors by Southern blot hybridization. Whenever possible, the parental origin of allelic fragments representing a locus on 12p was determined to discover whether genomic imprinting played a role.

MATERIALS AND METHODS Samples of primary tumor tissue and heparinized venous blood were obtained from 22 Norwegian testicular cancer patients. Data on histology and stage are summarized in Table 1. Tumor samples were representative of tumor tissue, as shown by histologic evaluation [11, 12]. Blood samples were available from the parents of 12 patients. Three Finnish patients and their parents from a series of 14 patients described previously [13] were included in the study of parental origin. The individuals were those who proved informative in the analysis. High-molecular-weight DNA was isolated from blood and tissue samples and aliquots were analyzed as TaqI or EcoRI digests by Southern blot hybridization as described [14]. Dosage quantitation was based on densitometric measurements as described previously [12]. Probes pPRP112.2RP (loci PRB1-PRB4) [15], pSP12-1 (D12Z3) [16], and pDL32B (D12S7) [17] were used to represent 12p, 12 centromeric region, and 12q, respectively. Membranes hybridized with chromosome 12-specific probes were rehybridized with probes from chromosome 1, pSDI-1 (DIZ5) [18], and/or pl-79 (DIZ2) [19] to obtain quantitative references. The intensities of chromosome 1-specific hybridization signals correlated well with those obtained with probes 21

© 1992 Elsevier Science P u b l i s h i n g Co., Inc. 655 A v e n u e of the Americas, N e w York, NY 10010

Cancer Genet Cytogenet 6 4 : 2 1 - 2 6 (1992} 0165-4608/92/$05.00

22

Table 1

P. Peltom~iki et al.

Histological classification, stage, and some molecular characteristics of testicular germ cell tumors Dosage of 12pa

Dosage of 12cen °

6.1 2.9 2.1 3.6 2.7 4.5 2.4 2.7 5.2 2.8 3.7 5.3 2.6 3.5 1.7 1.7 2.3 N

N N N N N N N N N 1.9 1.8 3.4 1.6 4.9 2.9 2.4 2.3 N

I I I

N N N

N N N

I

2.4

ND

Patient

Histology

Stage

5 7 8 10 20 21 22 24 25 12 18 23 26 14 17 28 29 9 11 13 15 19

Seminoma EmCa, IMT, MT EmCa, CIS Seminoma EST, CIS Seminoma Seminoma EmCa, CIS Seminoma EmCa, CIS Seminoma Seminoma Seminoma Seminoma MT EST Seminoma Seminoma Seminoma Seminoma Seminoma Seminoma

II II ! III I I I I III I II I I I

III II IV I

X/Y or pseudoautosomal dosage changeb No No Yes Yes No Yes Yes No Yes No No Yes No No No Yes No Yes No No

No Yes

Abbreviations: EmCa, embryonalcarcinoma;MT, mature teratoma;EST, endodermalsinus tumor; N, no change (dosage ratio 1 -+ 0.4); ND, not determined. a Calculated accordingto the followingformula: [12p (or 12cen)-specificsignal/referencesignal in tumor DNA]/[12p (or 12cen)-specificsignal/reference signal in normal DNA]. bThere was a decrease in dosage of Y, sometimesaccompaniedby an increase in X-chromosomaldosage. Differentkinds of dosage changes occurred in the pseudoautosomal region [12].

from chromosomes 2 and 4, which were used as additional references. Furthermore, allelic losses on chromosome 1 were not detected in a previous study of this tumor series [11]. RESULTS AND DISCUSSION

The characteristic features of most tumor DNA samples in this study were 1) 12p dosage e n h a n c e m e n t without a comparable increase in 12q dosage, and 2) a difference between 12p and 12cen dosages that varied in different tumors. The dosage of 12p was increased in tumor DNA of 18 of 22 patients (82%) as compared with normal DNA of the same i n d i v i d u a l s (Table 1, Fig. 1). The increase was 1.7 to 6.1-fold relative to a reference locus from chromosome 1. Probe pPRP112.2RP from the salivary proline-rich protein gene was considered a good representative of 12p in dosage determinations because this locus and loci from other parts of 12p exhibited equal increases in signal intensity in a previous experiment ( u n p u b l i s h e d observations). In addition to 12p, sequences specific for 12cen and 12q were studied in 21 and 14 cases, respectively. The patterns of dosage changes were interpreted to suggest that there was one or more copies of the characteristic 12p derived marker chromosome (described below). Although not directly comparable, the frequency with which the dosage of 12p was enhanced in the present tumor series is close to the

cytogenetically observed occurrence rate of i(12p) in testis tumors [1, 2]. There was no concomitant increase in the dosage of sequences specific for the centromere of chromosome 12 recognized by probe pSP12-1 in n i n e patients (patients 5, 7, 8, 10, 20, 21, 22, 24, and 25). In tumor DNA from four patients (patients 12, 18, 23, and 26) the increase in 12p dosage was accompanied by a parallel but smaller change in the dosage of the centromeric region (dosage ratios approximately 1 . 5 - 2 : 1 ) . Five patients (patients 12, 18, 21, 23, and 24) showed intensity changes between pSP12-1 fragments (constant and allelic). Based on hybridization with a 12q-specific probe, the changes described above could not be explained by a n e u p l o i d y of entire chromosomes 12. Therefore, our results suggested the presence of chromosome 12p-derived marker chromosomes carrying different amounts of centromere-specific DNA. Tumor DNA from patients 14, 17, 28, and 29 displayed increases in dosage that were approximately equal or even greater for the centromeric region of chromosome 12 as compared with the short arm of this chromosome, Possible explanatory m e c h a n i s m s i n c l u d e duplication of entire chromosomes 12 and the occurrence of dicentric 12p-derived marker chromosomes. Hybridization results obtained with probe pDL32B from 12q suggested an increased copy n u m b e r of chromosome 12 in tumor tissue from patient 14 (Fig. 2). In contrast, a m e c h a n i s m other than duplication of

)SD1-1

;)SP12-1

B

....

ii

I

ii

T

T

I!i

E

Wi'

3

18

i

i

i

¸

~

T

U

~i~ ¸ ~

~

~:

i I

B

23 i

T

6 m

:~ ~i~I i~

: iii~

3

14 B

17 T

II

iji! ¸ :ii!!i~¸

B

)

Figure 1 TaqI-digested blood (B) and tumor (T) DNA samples from patients 7, 9, 11, 14, 17, 18, 21, and 23 ~howing different patterns of chromosome 12-specific changes (Table 1). The autoradiographs represent successive lybridizations of the membranes using probes pPRP112.2RP (12p), pSP12-1 (12cen), and pSDI-1 (quantitative :eference from chromosome 1). Probe pPRP112.2RP detects an insertion/deletion polymorphism with fragments Jetween 7 and 1.2 kilobases (kb) and regions of allelic variation are indicated by brackets. Probe pSP12-1 recognizes i constant fragment of 1.4 kb and an allelic fragment of 1.1 kb (short line) that is either present or absent. Probe jSDlol hybridizes strongly to a fragment of approximately 1.1 kb and to additional fragments of lower homology.

-OO

)PRP112.2RP~-:

T

21

~::~i~i~ii!,,~:! i!~:i~i'~iiiiiii~ !iiii'I,!~!~,~,'i

3

7 T

:~

B

T

~ ¸¸¸iLl:/¸

11

24

P. Peltom/iki et al.

29 B T

~)---gt

pPRP112.2RP

Q

ie i 14 B T

pSP12-1

a/b

pSD1-1

F-

b/

¸¸ iiii ii!ii 'ii pDL32B

;Q

L

Figure 2 TaqI-digested DNA samples from patients 14 and 29 representing blood (B) and tumor (T). The samples from patient 14 were hybridized with probe pDL32B (12q) and rehybridized with pSDI-1 (quantitative reference from chromosome 1). The dosage analysis of 12p and 12cen in this patient is shown in Figure 1. The membrane containing samples from patient 29 was hybridized successively with probes pPRP112.2RP (12p), pSP121 (12cen), pSDI-1, and pDL32B. (a and b) Allelic fragments detected by the probes (polymorphism shown by pSP12-1 involves a fragment that is either present or absent). Probe pDL32B recognizes allelic fragments of 6.8 kilobases (kb) (a) and 4.1 + 2.7 kb (b). Figure I shows the sizes of fragments detected by the other probes. Autoradiographs of patients 14 and 29 represent different electrophoretic runs.

chromosomes number 12 was likely in patient 29. Tumor DNA from this patient exhibited a partial loss of one allelic fragment (b) detected by pDL32B (Fig. 2). Because there was also partial loss of a pPRP112.2RP allele, accompanied by multiplication of the other allele, the presence of 12pderived marker chromosome(s) originating from one chromosome 12 concomitant with loss of the other homolog (e.g., by mitotic nondisjunction) was suggested. Finally, hybridization with probes pPRP112.2RP, pSP12-1, and pDL32B showed no changes in tumor DNA from four patients (patients 9, 11, 13, and 15). Thus, evidence for any major chromosomal changes affecting chromosome 12 in these tumors was not obtained in this study. (In patient 19, the ratio of dosages of 12p to 12cen could not be determined.) Our study suggested that breakpoints on chromosome 12 vary relative to the centromeric marker (pSP12-1) used. Molecular heterogeneity is well documented in the pericentromeric structure of h u m a n Xq isochromosomes [20]. Using fluorescence in situ hybridization, Mukherjee et al. [21] detected size variation in centromeric signals originating from the characteristic marker chromosomes of testis tumors. Although direct confirmation by other techniques is necessary, our findings are in accordance with these observations. Probe pPRP112.2RP detected allelic imbalances in tumor DNA from five of seven informative patients with an increase in 12p dosage. The parental origin of altered fragments could be determined in four. In addition, three Finnish patients were included in this part of the study. Allelic fragments with increased intensity originated in four cases from the father and in three cases from the mother, suggesting random distribution with respect to parental sex. Therefore, there was no evidence for genomic imprinting. Figure 3 shows examples of the study of parental origin. Increased 12p alleles apparently originated from only one parent in each case (Fig. 3), compatible with the notion that the putative marker is an i(12p) formed by sister chromatids of one chromosome 12. Our findings thus contrast with those of Mukherjee et al. [21], who suggested that the cytogenetic marker most likely represented an i(12p) formed by (nonreciprocal) centromeric interchanges between nonsister chromatids of homologous chromosomes. This suggestion was based on fluorescence in situ hybridization results combined with analysis of restriction fragment length polymorphisms from 12q. In addition, loss of heterozygosity in 12q was a rare occurrence in our study, suggesting that i(12p) arose after aneuploidization, as proposed by van Kessel et al. [22]. Castedo et al. [1] proposed that i(12p) - testicular germ cell tumors might constitute a distinct subgroup associated with a missing Y chromosome as an accompanying abnormality. The association suggested by Castedo et al. [1] was not supported by our results since only one of four tumors with no increase in the relative 12p dosage (25%) showed a decrease in Y-chromosomal dosage whereas Y dosage was diminished in 8 of 18 tumors with an increase in 12p dosage (44%) (Table 1). This conclusion can be made only within

Chromosome 12 in Testis Cancer

25

14 M

5 M

P

B

P

B

T

7 T

M

P

E

T

I

~I

i~ :

~

i

,i i

EcoR I

....~--" - - Dii~i!!

~

*-- P

g o i i

~

ii~!~il

-

.-p

EcoRI

I

i

i • Taq I Figure 3 Parental origin of 12p fragments showing intensity changes in Finnish patients 5, 7, and 14. Normal (B, blood; E, epididymis) and tumor (T) DNA samples from the patients as well as blood DNA samples from their parents (M, maternal; P, paternal) were digested with TaqI or EcoRI and hybridized with probe pPRP112.2RP. Regions polymorphic for EcoRI are indicated by brackets (TaqI polymorphisms are shown in Fig. 1). Allelic fragments allowing differentiation between maternal and paternal origin (short lines); fragments with increased intensity (arrows).

the limits of detection a l l o w e d by the quantitative Southern blot h y b r i d i z a t i o n technique. Tumors in w h i c h the dosage of 12p was u n a l t e r e d relative to the quantitative reference used were all s e m i n o m a s and represented clinical stage I. M a n y tumors w i t h the most p r o m i n e n t increases in 12p dosage were also seminomas, however. Thus, there is no association b e t w e e n s e m i n o m a and 12p enhancement. On the other hand, because all nonseminomas d i d show an increase in 12p dosage a correlation b e t w e e n 12p dosage and histologic t y p e s h o u l d not be excluded. Furthermore, an association m a y exist b e t w e e n clinical stage and the occurrence a n d / o r m a g n i t u d e of 12p dosage increase. Not only d i d all four tumors w i t h n o r m a l relative dosages of 12p represent stage I (described above), but there was also a p r e d o m i n a n c e of more a d v a n c e d clinical stages among tumors w i t h the highest increases in 12p dosage (Table 1), w h i c h is in a c c o r d a n c e w i t h the findings of other investigators showing that a high c o p y n u m b e r of i(12p) could be associated with a d v a n c e d clinical stage a n d / o r resistance to c h e m o t h e r a p y [23, 24].

The authors thank H. F. Willard for probes pSP12-1 and pSDI-1, L. C. Tsui for probe pDL32B, and M. Litt for probe pl-79. Plasmid pPRP112.2RP was purchased from American Type Culture Collection. They also thank A. E. Stenwig for histologic evaluation of the Norwegian samples and K. Heimdal, A. Halme, and S. Lindh for collecting parental blood samples. This work was supported by the Academy of Finland, the Finnish Cancer Society, the Folkhiilsan Institute of Genetics, and the Sigrid Juselius Foundation. R. A. L. is a postdoctoral fellow of the Norwegian Research Council for Science and Humanities.

REFERENCES 1. Castedo SMMJ, de Jong B, Oosterhuis JW, Seruca R, Idenburg VJ, Bruist J, Sleijfer DT (1988): i(12p)-negative testicular germ cell tumors. A different group? Cancer Genet Cytogenet 35:171-178. 2. Samaniego F, Rodriguez E, Houldsworth J, Murty VVVS, Ladanyi M, Lele KP, Chert Q, Dmitrovsky E, Geller NL, Reuter V, Jhanwar SC, Bosl GJ (1990): Cytogenetic and molecular analysis of human male germ cell tumors: Chromosome

26

3.

4. 5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

P. P e l t o m ~ k i et al.

12 abnormalities and gene amplification. Genes Chromosomes Cancer 1:289-300. Castedo SMMJ, de Jong B, Oosterhuis JW, Seruca R, Idenburg VJS, Dam A, te Meerman G, Koops HS, Sleijfer D (1989): Chromosomal changes in h u m a n primary testicular nonseminomatous germ cell tumors. Cancer Res 49:5696-5701. Solter D (1987): Inertia of the embryonic genome in mammals. Trends Genet 3:23-27. Hall J (1989): Potential importance of imprinting in explaining common disorders, In: From Phenotype to Gene in Common Disorders, K Berg, N Retterstol, S Refsum (eds.). Oslo, pp. 9-36. O'Brien SJ, Marshall Graves JA (1991): Report of the committee on comparative gene mapping. Human Gene Mapping 11. Cytogenet Cell Genet 58:1124-1151. Schroeder WT, Chao L-Y, Dao DD, Strong LC, Pathak S, Riccardi V, Lewis WH, Saunders GF (1987): N o n r a n d o m loss of maternal chromosome 11 alleles in Wilms tumors. Am J Hum Genet 40:413-420. Scrable H, Cavenee W, Ghavimi F, Lovell M, Morgan K, Sapienza C (1989): A model for embryonal rhabdomyosarcoma tumorigenesis that involves genome imprinting. Proc Natl Acad Sci USA 86:7480-7484. Toguchida J, Ishizaki K, Sasaki MS, Nakamura Y, Ikenaga M, Kato M, Sugimoto M, Kotoura Y, Yamamoto T (1989): Preferential mutation of a maternally derived RB gene as the initial event in sporadic osteosarcoma. Nature 338:156-158. Nicholls RD, Knoll JHM, Butler MG, Karam S, Lalande M (1989): Genetic imprinting suggested by maternal heterodisomy in nondeletion Prader-Willi syndrome. Nature 342: 281-285. Lothe RA, Foss~ D, Stenwig AE, Nakamura Y, White R, Borresen A-L, Brogger A (1989): Loss of 3p or 11p alleles is associated with testicular cancer tumors. Genomics 5:134-138. Peltom~ki P, Lothe R, Borresen A-L, Foss~ SD, Brogger A, de la Chapelle A (1991): Altered dosage of the sex chromosomes in h u m a n testicular cancer: A molecular genetic study. Int J Cancer 47:518-522. Peltom~ki P, Alfthan O, de la Chapelle A (1991): Oncogenes in h u m a n testicular cancer: DNA and RNA studies. Br J Cancer 63:851-858. Peltom~ki P, Halme A, de la Chapelle A. (1990): Human testic-

ular cancer: Changes in autosomal dosage. Cancer Genet Cytogenet 48:1-12. 15. Lyons KM, Kim HS, Maeda N, Smithies O (1985): Insertion/ deletion polymorphisms in h u m a n salivary proline-rich protein genes. Genetics 110:31. 16. Greig GM, Parikh S, George J, Powers VE, Willard HF (1991): Molecular cytogenetics of a satellite DNA from chromosome 12: Fluorescence in situ hybridization and description of DNA array length polymorphism. Cytogenet Cell Genet 56:144-148. 17. Tsui LC, Buchwald M, Markiewicz D, Kao FT, Cai GY, Lau ML (1985): Identification of a polymorphic DNA marker pDL32B (D12S7) and its localization to the long arm of chromosome 12, region q14.3-qter. Cytogenet Cell Genet 40:763-764. 18. Waye JS, Durfy SJ, Pinkel D, Kenwrick S, Patterson M, Davies KE, Willard HF (1987): Chromosome-specific a l p h a satellite DNA from c h r o m o s o m e 1: Hierarchical structure and genomic organization of a p o l y m o r p h i c d o m a i n s p a n n i n g several h u n d r e d kilobase pairs of centromeric DNA. Genomics 1:43-51. 19. Magenis RE, Sheehy R, Litt M (1987): Hypervariable repetitive sequence DIZ2 on distal short arm of chromosome 1 is not telomeric. Cytogenet Cell Genet 46:654. 20. Sharp CB, Bedford HM, Willard HF (1990): Pericentromeric structure of h u m a n X "isochromosomes": Evidence for molecular heterogeneity. Hum Genet 85:330-336. 21. Mukherjee AB, Murty VVVS, Rodriguez E, Reuter VE, Bosl GJ, Chaganti RSK (1991): Detection and analysis of origin of i(12p), a diagnostic marker of h u m a n male germ cell tumors, by fluorescence in situ hybridization. Genes Chromosomes Cancer 3:300-307. 22. Van Kessel AG, van Drunen E, de Jong B, Oosterhuis JW, Langeveld A, Mulder MP (1989): Chromosome 12q heterozygosity is retained in i(12p)-positive testicular germ cell tumor cells. Cancer Genet Cytogenet 40:129-134. 23. Delozier°Blanchet CD, Walt H, Engel E, Vuagnat P (1987): Cytogenetic studies of h u m a n testicular germ cell tumors. Int J Androl 10:69-77. 24. Bosl GJ, Dmitrovsky E, Reuter VE, Samaniego F, Rodriguez E, Geller NL, Chaganti RSK (1989): Isochromosome of chromosome 12: Clinically useful marker for male germ cell tumors. J Natl Cancer Inst 81:1874-1878.