Mitotic deletions of 11p15.5 in two different tumors indicate that the CALCA locus is distal to the PTH locus

1. Henry.’ S. Grandjouan .? F. Barichard.’ C. Huerre-Jeanpierre,’ and C. Junien

Institut National de la Sante et de la Recherche Medicale U73. Paris, and - Service de Gastro-enterologic. Hopital Cochin, Paris (France)

Abstract. We have compared the constitutional and tumor genotypes in two patients with Wilms tumor and adrenocortical carcinoma. The allelic distribution of chromosome 11-specific markers spanning chromosome 11 from pter to qter (HRASI- HBB-[CALCA/PTH]-FSHB-CAT-APOA1) and an approach combining RFLP analysis and gene copy number determination showed that a mitotic deletion had occurred in both tumors. The loss of one copy of the gene for a-calcitonin-gene-related poly-

peptide (CALCA). together with that of a more distal marker (HRASI or HBB), indicates that CALCA is distal to the gene for parathyroid hormone (PTH), which was not deleted in either tu- mor. These results suggest that mitotic deletion mapping may be as useful as meiotic deletion or recombination mapping in ordering closely linked markers, such as CALCA and PTH. for which other approaches, including physical mapping and multipoint linkage analysis, have failed to accurately identify the gene order.

Since the localization of the genes for human a-calcitonin- gene-related polypeptide (CALCA) and parathyroid hormone (PTH) to the short arm of chromosome 11 by somatic cell hybrid- ization, chromosome sorting, in situ hybridization. and subse- quent family studies using Ilp DNA markers (Mayer et al., 1983; Naylor et al., 1983: Przepiorka et al., 1984; Höppener et al., 1984; Lebo et al., 1985; Zabel et al., 1985), researchers have continued to attempt to locate these genes more precisely by ordering them with respect to each other and to other linked 11p markers. Data from several groups have provided discordant results as to the genetic distance between these two markers. According to Holm et al. (1985). no recombinants were observed in 127 informative meioses, while a recombination fraction of 8% was reported by two other groups (Gerhardt et al., 1987; Meyers et al., 1987). Mul- tipoint mapping studies also failed to determine the exact genetic order of these markers (Bonaïti-Pellié et al., 1986; Meyers et al., 1987).

As recently demonstrated by Scrable et al. (1987), the study of somatic chromosomal interchanges provides an alternative ap- proach to localizing a gene within a chromosomal region. These investigators have used mitotic recombination mapping to localize a gene involved in rhabdomyosarcoma to 11pter-p15.5. the smallest region overlapping somatic homozygosity in these tu- mors.

Using the same approach, we have mapped a gene involved in adrenocortical carcinoma and in Wilms tumor to the same region (Henry et al., 1989b). In two of the tumors analyzed, somatic chro-

mosomal events were identified and revealed a breakpoint be- tween CALCA and PTH. Comparison of constitutional and tu- mor genotypes with 1 lp markers allowed us to identify the mecha- nisms involved and to determine the gene order of CALCA and PTH.

Materials and methods

Patients. Patient BAN has already been described elsewhere. Briefly. this patient. a boy. presented with WAGR syndrome. nephroblastoma, and a de- letion in chromosome 11 in the region 11p13-+p12 (Huerre et al., 1983), which includes the genes for catalase (CAT) and follicle stimulating hormone. ß poly- peptide (FSHB) (Couillin et al., 1989).

Patient ZEM was the second child of a family in which there was evidence for a familial predisposition to several types of cancer. including osteosarcoma. breast cancer, adrenocortical carcinoma, and rhabdomyosarcoma. This patient has also been described elsewhere (Henry et al., 1989a).

Probes. Southern blots were hybridized to the following probes detecting RFLPs on llp (Kazazian and Junien, 1987): c-Ha-ras 1 (HRAS1), a 2.9-kb Sacl genomic fragment (Chang et al., 1982): B-hemoglobin (HBB), a 4.4-kb Psrl genomic fragment (Lawn et al., 1978); CALCA. a 0.58-kb PsrI cDNA fragment (Craig et al., 1982): PTH. a 2.5-kb Psrl/EcoRI cDNA fragment (Hendy et al .. 1981); FSHB. a 0.5-kb Hincll/Sacl cD)NA fragment (Watkins et al., 1982): CAT. a 1.0-kb PsrI cDNA fragment (Korneluk et al .. 1984): apolipoprotein A-I (APOA1). a 2.2-kb Psrl cDNA fragment (Kessling et al., 1985); and collagen. type II. al (COL2A1). a 3.7-kb EcoRI genomic fragment (Huerre-Jeanpierre et al .. 1986).

Southern blot experiments. DNA was prepared from lymphocytes, fibro- blasts, or lymphoblastoid cell lines as described by Sanders-Haigh et al. (1980). Aliquots of 15-30 µg of DNA were digested with restriction endonucleases ac- cording to the manufacturer’s (Promega Biotech) recommendations.”P-dCTP and the nick-translation kit were purchased from Amersham. DNA was electro- phoresed and then transferred to nitrocellulose or nylon membranes as de- scribed by Southern (1975). Hybridization with “P-labeled probes (specific ac- tivity. 2-6 × 10% cpm/µg) was carried out for 16 h using 10% dextran sulphate (Wahl et al., 1979). For determination of the gene copy number. the filters were freed of probe in alkali and then rehybridized with other probes used as internal controls. The intensity of the hybridization signal was measured with a SEBIA densitometer. and the ratio of the 1 lp probe signal intensity to that of the control

Supported by grants from the Institut National de la Sante et de la Recherche Medicale (INSERM). Ligue Nationale Contre le Cancer. Association pour la Recherche Contre le Cancer (ARC), and Faculte de Medecine Paris V

probe was calculated. Each experiment was performed in triplicate, and the data were subsequently normalized to the control ratio (1.00). Statistical analysis was performed using Student’s / test with at least three independent values obtained from the normal control and from the constitutional and tumor DNA of each patient.

Results and discussion

The constitutional DNAs from patients BAN (del 11p13) and ZEM (normal) were analyzed with probes detecting 25 RFLPs spanning chromosome 11 from pter to qter. Tumor DNAs were subsequently analyzed with probes that detected heterozygosity at the constitutional level. As shown in Table I, tumor DNAs showed losses of heterozygosity at polymorphic loci along the length of chromosome 11; BAN showed losses at HRASI and CALCA, and ZEM at HBB and CALCA. In contrast, BAN was hemizygous for

FSHB and CAT, and ZEM was heterozygous for PTH and APOAI.

To determine whether these losses of heterozygosity were due to a somatic deletion or mitotic recombination, gene copy number was determined for the markers flanking the region of the break- point in each patient, i.e., between CALCA and PTH. The probes used as internal controls in these experiments were chosen either on another chromosome for which the patient’s tumor remained heterozygous (CALCA/COL2A1 and PTH/COL2A1 in patient BAN) or in the region of chromosome 11 for which the patient was heterozygous (CALCA/PTH in patient ZEM). As shown in Fig. la and Table I, BAN’s tumor had only one copy of CALCA (ratio, 0.48) and two copies of PTH (ratio, 0.98), indicating a de- letion with a breakpoint between CALCA and PTH. As shown in Fig. 1b, there was a loss of allele 2 of CALCA in the tumor of patient ZEM. In contrast, both alleles of PTH remained in the

Fig. 1. RFLP and dosage analysis of CALCA and PTH on chromosome 11 in constitutional (C) and tumor (T) DNA from patients BAN (Wilms tumor) and ZEM (adrenocortical carcinoma). (a) Dosage analysis of CALCA and PTH in patient BAN, using COL2AI as an internal control. The curves below cach blot are densitometric tracings. (b) Representative Southern blots of specific 11p15 polymorphic markers flanking the breakpoint of the mitotic rearrange- ment observed in the tumor of patient ZEM. The alleles are designated to the left of each autoradiogram, and the size of the restriction fragments (in kb) is in- dicated on the right side of each blot.

a

BAN

Eco RI

C

T

C

T

CALCA

PTH

COL2A1

C

T

6.2

5.0

6.2

4.3

11 11 MM

b ZEM

Tag I N

Pst I N

T

T

CALCA

PTH

1

-8.0

-6.5

1

2.8

2

2

2.2

Fig. 2. Schematic representation of the deletion of chromosome 11 that oc- curred in hereditary cases of Wilms tumor (WT) and adrenocortical carcinoma (ADCC). To determine the limits and nature of the mitotic events. a combina- tion of RFLP analysis and gene copy number determination was used. Numbers in parentheses indicate the normalized ratio (11p15 versus internal control) ob- tained by densitometer scanning (1.00 = two copies, 0.5 = one copy); diploid = two different alleles were present in the tumor.

q

p13

p15

WT

PTH

CALCA

BAN

(0.98)

(0,48)

ADCC

PTH

CALCA (0,55)

ZEM

diploid

Table I. Comparison of constitutional (C) and tumor (T) genotypes in patients BAN and ZEM by RFLP and dosage analyses2

PatientMethodChromosome Il probe
11pter HRASIHBBCALCAPTHFSHBCATAPOA1Ilqter
c'Tc'TCTCTcTCTCT
BANRFLP1/21/-1/211-1/01/01/0170
Dosage
T/C0.480.98
ZEMRFLP1/2-/21/21/21/21/2
Dosage
T/C0.551.00

‘Results are given only for loci informative at the constitutional level. RFLP alleles were named I and 2 according to decreasing length: 1/- or-/2 indicate the alleles that remained in the tumor samples: 1/2 means that the tumor DNA remained heterozygous at this locus; and 1/0 indicates that the patient who carried a constitutional del 1 1p1 3 retained the unique copy of this marker in his tumor cells. For determination of the gene copy number in the tumor DNA (dosage T/C), the ratio T/C was calculated by dividing the | Ip15 marker/internal control ratio in the tumor DNA from the patient by that in the constitutional DNA from the same patient. The order of the markers follows Kazazian and Junien (1987).

tumor, showing the same intensity as in the constitutional DNA. Since there were two copies of PTH in the tumor. we used this probe as an internal control on a CALCA/Taql blot to measure the gene copy number for CALCA (data not shown). As shown in Table 1, the CALCA/PTH ratio was 0.55, indicating that only one copy of CALCA was present in the tumor, as compared to two copies of CALCA in the constitutional DNA, which suggests that the loss of heterozygosity in this patient was due to a mitotic de- letion, the breakpoint residing between CALCA and PTH.

The results presented in Figs. 1 and 2 show that CALCA, which was deleted together with HRASI in patient BAN and with HBB in patient ZEM, maps distal to PTH, for which there was no loss of alleles in both patients. These data conflict with the results report- ed by Bonaïti-Pellié et al. (1986), who tentatively placed PTH dis- tal to CALCA on the basis of a multipoint linkage analysis of the family data obtained by Meyers et al. (1987). Meyers et al. estimat- ed a recombination fraction of 8% between CALCA and PTH based on 30 informative meioses. In contrast, Holm et al. (1985) failed to detect a recombinant among 127 informative meioses and could not propose a more likely gene order for these tightly linked markers. In the present report, although the hypothesis of a more

complex chromosomal rearrangement cannot be formally reject- ed. it seems unlikely that both tumors underwent the same com- plex rearrangement, therefore suggesting the wrong gene order. The occurrence in two independent tumors of the same deletion between the same markers, CALCA and PTH. thus strongly sup- ports a gene order compatible with CALCA being distal to PTH.

The occurrence of mitotic recombination is thus useful not only for ordering genes but also to determine precisely the regions in- volved in tumorigenesis. In this respect, studies on patient BAN have shown that the site for which the loss of alleles was observed in the tumor (11p15.5) differs from the site of the germinal muta- tion (i.e., the constitutional del 11p13). This disparity was also observed in a second patient and could have a bearing on the se- quence of events involved in tumorigenesis (Henry et al., 1989a).

Acknowledgements

We are grateful to A. Bank. E. Chang. S. Humphries. I. Mcintyre. R. Korneluck. and H. Kronenberg, who kindly provided the probes used in this study. We also thank Drs. Chaussain and Landrieu for referring their patients.

References

Bonaiti-Pellic C. Martinez M. Clerget-Darpoux F: Mul- tilocus linkage analysis of markers located on the short arm of chromosome 11. Genet Epidem (Suppl) 1:129-134 (1986).

Chang EH. Gonda MA. Ellis RW. Scolnick EM. Lowy DR: Human genome contains four genes homologous to transforming genes of Harvey and Kirsten murine sarcoma viruses. Proc natl Acad Sci. USA 79:4848-4852 (1982).

Couillin P. Azoulay M. Delcamp M. Ravise N. Grisard MC. Henry 1. Barichard F. Metezeau P. Junien C: Charac- terization of a panel of somatic cell hybrids for sub- regional mapping along Ilp and within band IIp13. Hum Genet (1989. in press).

Craig RK. Hall L. Edbrooke MR. Allison J. Mcintyre I: Partial nucleotide sequence of human calcitonin pre- cursors: mRNA nucleotide sequence of human cal- citonin. Nature 295:345-347 (1982).

Gerhard DS. Jones C. Morse HG. Handelin B. Weeks V. Housman DE: Analysis of human chromosome 11 by somatic cell genetics: reexamination of derivatives of human-hamster cell line JI. Somat Cell molec Genet 13:293-304 (1987).

Hendy GN. Kronenberg HM. Potts JT. Roch A: Nucleo- tide sequence of cloned cDNAs encoding human pre- proparathyroid hormone. Proc natl Acad Sci. USA 78:7365-7369 (1981).

Henry 1, Grandjouan S. Couillin P. Barichard F. Huerre- Jeanpierre C, Glaser T. Philip T. Lenoir G. Chaussain JL. Junien C: Tumor-specific loss of 11p15.5 alleles in dell Ip13 Wilms tumor and in familial adrenocortical carcinoma. Proc natl Acad Sci. USA (1989a, in press). Henry 1. Jeanpierre M. Couillin P. Barichard F. Serre JL. Journel H. Turleau C. Grouchy J de. Junien C: Mo- lecular definition of the Ilp region involved in Beck- with-Wiedemann syndrome and in predisposition to adrenocortical carcinoma. Hum Genet 81:273-277 (1989b).

Holm T. O’Connell P. Leppert M. Callahan P. Lalouel J M. White R : Parathyroid hormone and calcitoninaretightly

linked and have been placed on the genetic map of chro- mosome 11. Cytogenet Cell Genet 40:653-654(1985).

Höppener JWM. Steenberg PH. Zandberg J. Bakker E. Pearson PL. Geurt van Kessel AHM. Jansz HS. Lips CJM: Localization of the polymorphic human cal- citonin gene on chromosome 11. Hum Genet 66:309 -312 (1984).

Huerre C. Despoisse S. Gilgenkrantz S. Lenoir GM. Junien C: c-Ha-ras 1 is not deleted in aniridia- Wilms’ tumour association. Nature 305:638-641 (1983).

Huerre-Jeanpierre C, Mattei MG, Weil D), Grzeschick KH. Chu M L. Sangiorgi FO. Sobel M E. Ramirez F. Junien C: Further evidence for the dispersion of the human fibrillar collagen genes. Am J hum Genet 38:26-37 (1986).

Kazazian HH: Junien C: Report of the committee on the genetic constitution of chromosomes 10, 11. and 12. Ninth International Workshop on Human Gene Map- ping. Cytogenet Cell Genet 46:188-213 (1987).

Kessling AM. Horsthemke B. Humphries SR: A study of the DNA polymorphisms around the human apolipo- protein A I gene in hyperlipidemia and in normal in- dividuals. Clin Genet 28:296-306 (1985).

Korneluk RG, Quan F. Lewis WH, Guise KS, Willard HF. Holmes M. Gravel RA: Isolation of the human fibro- blast catalase cDNA clones: sequence of the clone derived from spliced and unspliced RNA. J biol Chem 259:13819-13823 (1984).

Lawn RM. Fritsch EF. Parker RC. Blacke G. Maniatis T: The isolation and characterization of linked ß-globin genes from a cloned library of human DNA. Cell 15:1157-1174 (1978).

Lebo R V. Cheung MC. Bruce BD. Riccardi VM. Kao FY. Kan Y W: Mapping parathyroid hormone. beta-globin. insulin, and LDH-A genes within the human chro- mosome 11 short arm by spot blotting sorted chro- mosomes. Hum Genet 69:316-320 (1985).

Mayer H. Breyel E. Bostock C. Schmidtke J: Assignment of the human parathyroid hormone gene to chromosome 11. Hum Genet 64:283-285 (1983).

Meyers DA. Beaty TH. Maestri NE, Kittur SD. An- tonarakis SE. Kazazian HH Jr: Multipoint mapping studies of six loci on chromosome 11. Hum Hered 37:94-101 (1987).

Naylor SL. Sakaguchi A Y. Szoka P. Hendy GN. Kronen- berg HM. Shows TB: Human parathyroid hormone gene (PTH) is on the short arm of chromosome II. Somat Cell Genet 9:609-616 (1983).

Przepiorka D. Baylin SB. McBride OW. Testa JR. Bustros A de. Nelkin DB: The human calcitonin gene is located on the short arm of chromosome 11 Biochem biophys Res Commun 120:493-499 (1984).

Sanders-Haigh L. Anderson F. Francke U: The ß-globin gene is on the short arm of human chromosome 11. Na- ture 283:397-400 (1980).

Scrable HJ. Witte DP. Lampkin BC. Cavenec WK: Chro- mosomal localisation of the human rhabdomyosar- coma locus by mitotic recombination mapping. Nature 329:645-647 (1987).

Southern EM: Detection of specific sequences among DNA fragments separated by gel electrophoresis. J molec Biol 98:503-517 (1975).

Wahl GM. Stern M. Stark GR: Efficient transfer of a large DNA fragment from agarose gels to diazobenzyloxy- methyl-paper and rapid hybridization by using dextran sulfate. Proc natl Acad Sci, USA 76:3683-3687 (1979).

Watkins PL. Eddy R. Beck AK. Vellucci V. Leverone B. Tanzi RE. Gusella JF. Shows TB: DNA sequence and regional assignment of the human follicle-stimulating hormone ß-suhunit gene to the short arm of human chromosome 11. DNA 6:205-212 (1982).

Zabel BU. Kronenberg HM. Bell G1. Shows T B: Chromo- some mapping of the genes on the short arm of human chromosome 11: parathyroid hormone gene is at 11p15 together with the genes for insulin, c-Harvey-ras I. and beta-hemoglobin. Cytogenet Cell Genet 39:200-205 (1985).