sign in sign up
<<
Home , Chromosome 22

Proc. Nati. Acad. Sci. USA Vol. 84, pp. 5419-5423, August 1987 Medical Sciences Molecular genetic approach to human : Loss of on 22 (brain tumors/neurofibromatosis/recessive oncogenes/DNA markers/inherited disease genes) BERND R. SEIZINGER*t, SUZANNE DE LA MONTEt, LEONARD ATKINS§, JAMES F. GUSELLA*t, AND ROBERT L. MARTUZA¶ *Neurogenetics Laboratory, tNeuropathology Laboratory, §Department of Pathology, and fDepartment of Surgery (Neurosurgery), Massachusetts General Hospital, and tDepartment of Genetics, Harvard Medical School, Boston, MA 02114 Communicated by Richard L. Sidman, March 30, 1987

ABSTRACT A molecular genetic approach employing study collected data. In view of our observation that acoustic polymorphic DNA markers has been used to investigate the role neuromas lose genes on (7) and the striking of chromosomal aberrations in meningioma, one of the most clinical association ofmeningiomas and acoustic neuromas in common tumors of the human nervous system. Comparison of patients with bilateral acoustic neurofibromatosis (BANF) the alleles detected by DNA markers in tumor DNA versus (8-11), this finding suggests a similar pathogenic mechanism DNA from normal tissue revealed chromosomal alterations in both tumor types. present in primary surgical specimens. In agreement with cytogenetic studies ofcultured , the most frequent alteration detected was loss of heterozygosity on chromosome MATERIALS AND METHODS 22. Forty of 51 patients were constitutionally heterozygous for DNA of high relative molecular weight was isolated from at least one chromosome 22 DNA marker. Seventeen of the 40 primary intracranial meningiomas and corresponding normal constitutionally heterozygotic patients (43 %) displayed tissue (peripheral leukocytes) as described (7). With the hemizygosity for the corresponding marker in their meningi- exception of tumor M10, which was irradiated, all specimens oma tumor tissues. Loss of heterozygosity was also detected at were obtained prior to any radiotherapy or chemotherapy. a significantly lower frequency for markers on several other Approximately 5 ,ug of normal and tumor DNA was digested . In view of the striking association between acoustic to completion with appropriate restriction enzyme, fraction- neuroma and meningioma in bilateral acoustic neurofi- ated by agarose gel electrophoresis, transferred to nylon bromatosis and the discovery that acoustic neuromas display membrane, and hybridized to 32P-labeled probe DNA (7). The specific loss of genes on chromosome 22, we propose that a following probes known to reveal restriction fragment length common mechanism involving chromosome 22 is operative in polymorphisms (RFLPs) in human genomic DNA for loci on the development of both tumor types. Fine-structure mapping several different were used: Psisl (Oncor) to reveal partial deletions in meningiomas may provide the (SlS) (12, 13); pMS3-18 (D22S1) (14); p22/34 (D22S9) (15); means to clone and characterize a (or genes) ofimportance HuXC-2 (IGLC) (15); N8C6 (NGFB) (16, 17); G8 (D4S10) (18, for tumorigenesis in this and possibly other clinically associated 19); Dry5-1 (DiOS1) (15); 1-101 (D1OSPDX1) (20); pEJ6.6 tumors of the human nervous system. (HRASI) (21); pJ19.4 (D11S17) (15); pHINS 321 (INS) (22); p640 (KRAS2) (20); p7F12 (D13S1) (23); pIE8 (D13S4) (23); Meningiomas, which arise from arachnoidal cells surround- pHUB8 (D13S5) (24); pHU26 (D13S7) (24); pAW101 (D14S1) ing the brain, are one of the most common tumors of the (25); CH800 (GHI) (26); B74 (D18S3) (15); pC3 (C3) (27); human nervous system and are a frequent cause of seizures pGSH8 (D21S17) (28). and other neurological dysfunction (1). Although generally Autoradiograms were analyzed by scanning densitometry considered to be benign neoplasms, meningiomas may de- with a LKB Ultrascan XL. The peak areas corresponding to velop aggressive or malignant behavior with cortical invasion each hybridization signal were calculated by electronic inte- and multiple recurrences leading to progressive neurological gration. To determine whether loss of one allele for chromo- disability and death (2, 3). The pathogenic mechanisms some 22 in the tumor tissue was associated with duplication underlying tumor initiation, progression, malignancy, and of the remaining allele, the hybridization signals for chromo- recurrence are not yet understood. some 22 probes were normalized to those obtained when the studies have that the same Southern blots were rehybridized with probes for loci Cytogenetic suggested (4-6) majority on other chromosomes. The sequence D4S10 on chromo- of meningiomas are associated with various chromosomal some 4p was used as a control for the tumors M8, M10, abnormalities, most frequently the loss of one copy of M17, M19, and M25; sequence D21S17 on chromosome 21q chromosome 22. However, virtually all of these investiga- was used for the tumors M5, M29, M30, M33, and M40; tions have been performed on cultured tumor cells, since the NGFB on chromosome lp was used for the tumors M15, original tumor usually contains too few metaphases for direct M20, M31, and M39; and sequence D13S5 on chromosome karyotyping. To avoid the potential difficulty of chromo- 13q was used for the tumors M2, M11, and M38. Het- somal alterations occurring under the selective pressure of in erozygosity for RFLPs at these control loci frequently vitro propagation, we have employed a molecular genetic provided clear information that they were not deleted in the approach with polymorphic DNA markers for several chro- tumor DNA. For each tissue sample the hybridization signals mosomes to investigate the genetic constitution of primary specific to chromosome 22 were normalized to those for surgical tumor specimens. Loss of genes on chromosome 22 control chromosome probes. Then, a ratio of the normalized was detected in 43% of the meningiomas from which our values for each tumor/normal tissue pair was obtained (for details see ref. 7). To validate our approach of normalizing The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" Abbreviations: BANF, bilateral acoustic neurofibromatosis; RFLP, in accordance with 18 U.S.C. §1734 solely to indicate this fact. restriction fragment length polymorphism.

5419 Downloaded by guest on September 29, 2021 5420 Medical Sciences: Seizinger et al. Proc. Natl. Acad. Sci. USA 84 (1987) the hybridization signals for chromosome 22 probes, we Table 1. Loss of heterozygosity at loci on chromosome 22 compared (7) hybridization signals for pairs of markers for in meningiomas other chromosomes. These invariably gave constant ratios Marker/enzyme for tumor and normal DNA. For cytogenetic studies, meningiomas were dispersed with SIS/ D22S1/ D22S9/ IGLC/ collagenase II (Cooper Biomedical, Malvern, PA) cultured in Patient hfindIll Bgl II Taq I EcoRI F-10 medium (GIBCO) with 15% (vol/vol) fetal calf serum Ml 12 - 12 (GIBCO) plus antibiotics at 370C in an atmosphere of 95% M2 air/5% CO2. After 3 days to 2 weeks in culture, Colcemid M3 12 (CIBA-Geigy) (10 mg/ml) was added for 6 hr. The cells were M4 12 12 12 dispersed with trypsin/EDTA (GIBCO), washed with M5 Hanks' balanced salt solution without calcium or magnesium M6 -0 12 12 (GIBCO) and centrifuged. The supernatant was decanted, M7 12 12 and the pellet was incubated in 4 ml of hypotonic solution M8 0 (KCl at 3 g/liter; EDTA at 0.2 g/liter; Hepes at 4.8 g/liter, pH M9 12 12 7.4) with 1500 units of streptokinase (Hoechst Pharmaceuti- M1o 12 012 cals, Hounslow, Middlesex, U.K.) for 12 min, at 370C, and Mul 0 fixed with methanol/glacial acetic acid, 3:1 (vol/vol). After M12 12 12 24 hr at 40C, air-dried slides were prepared using a modifi- M13 12 12 cation ofa standard technique (29). The slides were incubated M14 - 12 12 at 650C overnight and G-banded with trypsin the following M15 day. Slides were microscopically evaluated, 20 metaphases and at least 5 metaphases of M16 46, XY were counted in most instances, M17 45, XX, -22 each specimen were karyotyped in detail. M18 12 12 M19 46, XY RESULTS M20 43, X, -Y, -14, -22 - 12 To detect somatic loss of chromosome 22 sequences in the M21 M22 12 - 12 tumors, DNA was typed with four different polymorphic M23 12 12 12 DNA loci: SIS, the platelet-derived growth factor p-chain 12 12 12 locus homologous to the sis oncogene mapping to 22q12.3- M24 12 46, XX M25 ©2 - 45, XX, -22 13.1; D22S1, an anonymous DNA segment at 22q12-13; 12 12 M26 12 12 12 46, XY/46, XY, lp+ D22S9, an anonymous DNA segment at 22q11; and IGLC, M27 12 the X-chain immunoglobin constant-region locus at 22q11. All 12 12 12 four markers, which were the only polymorphic DNA mark- M28 ers available, map to the long arm of chromosome 22. M29 Cytogenetic studies on meningiomas show (4-6) either loss of M30 0 45, XX, -22 the entire chromosome 22 or partial deletions on the long arm M31 12 12 0 but not on the short arm. Forty ofthe 51 meningioma patients M32 12 12 46, XX investigated were heterozygous in their normal tissue for at M33 12 45, XX, -22 least one of the four polymorphic DNA markers and conse- M34 46, XX quently could be used to determine whether loss of consti- M35 12 tutional heterozygosity had occurred in their respective M36 12 - tumor tissue. The results of this analysis are presented in M37 12 - 12 46, XX/45, XX, -1, Table 1. Tumor DNA from 17 ofthe 40 heterozygous patients -4, -11, -13, (43%) displayed a loss or marker reduction in one of the two +M1+M2+M3, chromosome 22 alleles (Fig. 1). The remaining hybridization Inv(2) signals corresponding to the deleted alleles in some tumor M38 -0@ 46, XX samples might be due to contaminating nontumor cells M39 2 - 0D 0 (vascular or connective tissue) that might be present in M40 45, XX, -4, -5, -?8 primary biological specimens (7, 30). Alternatively, these (?9), +15, -22, -22, meningiomas might consist of a of cells with normal +4 TO 7 markers karyotype and cells with for chromosome 22. This situation has been observed in cytogenetic studies ofcultured DNA from primary surgical tumor specimens and from correspond- meningiomas (4). ing normal peripheral leukocytes was isolated, digested with appropri- Several abnormal mitotic events could ate restriction enzymes known to reveal RFLPs, fractionated by types of potentially agarose gel electrophoresis, transferred to nylon membranes, and account for somatic loss of constitutional heterozygosity in hybridized to 32P-labeled probe DNA for polymorphic loci on chromo- tumors (31-36). To distinguish among the possible mecha- some 22. nisms, we compared the intensity of the hybridization signals The phenotype observed in the tumor tissue is shown for every case, in Fig. 1 with those obtained by rehybridizing the same filters where the blood DNA displayed heterozygosity. 12, heterozygosity with several probes for loci on other chromosomes. The ratio (even though different pairs of alleles may be present for certain between tumor and normal tissue of the copy number of loci multiallele markers); 1, continued presence of the larger allele restric- on chromosome 22 was -1:2 in all 17 patients (Table 2). The tion fragment and loss of the smaller allelic fragment relative to normal densitometric analyses were, therefore, consistent in all 17 tissue DNA; and 2, continued presence of the smaller allelic restriction cases with a one rather than fragment and loss of the larger fragment. Where the normal DNA was loss of copy of chromosome 22, tested but was uninformative because it did not display heterozygosity, with mitotic recombination or with chromosomal loss and a minus sign (-) has been entered to simplify consideration of the data. subsequent reduplication of the remaining homolog. We are The absence ofany entry indicates that a marker was not tested because presently unable to distinguish with certainty between loss of of limited availability of DNA or did not give a readable result for the the entire chromosome 22 and partial deletions of only particular tumor. Cases of hemizygosity in the meningioma tumor regions on chromosome 22 containing the four marker loci. tissues are circled. were available for 14 meningiomas. Downloaded by guest on September 29, 2021 Medical Sciences: Seizinger et al. Proc. Natl. Acad. Sci. USA 84 (1987) 5421

M8 MIC M17 M19 M25 M39 M40 N T N T N T NT N T N T N T

- '- ' -fF - .. .. I- - 1I- I _* 0 is T 2- - 2- -- s- 2- 4 2- 2 2- 2

Sis - oncogene 2- ** 2- i 2- _2. 2-2_ - 2- - 2 - -

M5 M8 M!o M15 M29 M33 N T N T N T N T N T N T NT N T

I- _ I- i- _ I- - I- w4 I- _I - 2I- D22S1 2-- 2- 2-- 2- - 2- .w 2- -0 - 2- 40 2- _0

M8 M17 t2 v MNtr M39 N T N T N T N T ! T . - - N I. 1- qw .# f- _, I- *l_ -l

2 - AW D22S9 2-b , 4

M2 M!O M1I M20 M30 M31 M38 139 N T N T N T N T N T N T N T N T

2- 21- @@ 1-_* -$§2 i~-*, 2- 40 IGLC 2 2- **It 2-- 2-9w 2-*9 2- 40 - 2- 41 - .-.

FIG. 1. Loss of constitutional heterozygosity at loci on chromosome 22 in meningioma tumor tissue. Patient designations are shown above the autoradiograms. Numbers on the left indicate the observed alleles, with 1 and 2 referring to the larger and smaller allelic restriction fragments, respectively. DNA was isolated from tumor specimens and corresponding normal tissue (peripheral leukocytes), digested with appropriate restriction enzymes, fractionated by agarose gel electrophoresis, and transferred to nylon membrane. Southern blots were hybridized to 32P-labeled DNA probes for the following loci on chromosome 22: Psisl (Oncor) (SIS); pMS3-18 (D22S1); p22/34 (D22S9); HuXC2 (IGLC). Psisl reveals a RFLP with fragments of 21 kilobases (kb) and 14.5 and 6.5 kb in HindIII-digested human genomic DNA (12, 13). The pMS3-18 probe detects a Bgi II RFLP with fragments of 9.5 kb and 6.5 kb (14). The probe p22/34 reveals allelic fragments of 5.8 kb and 3.2 kb in Taq I-digested DNA (15). HuXC2 detects a multiallele RFLP with EcoRI fragments of 8 kb, 13 kb, 18 kb, and 23 kb. N, DNA from normal tissue (peripheral leukocytes); T, DNA from meningioma tumor tissue.

Note, however, that in all seven meningiomas where the was observed in 8 meningiomas and involved several differ- patient was constitutionally heterozygous for more than one ent autosomes: (M40; 1 of 22 informative marker, the corresponding tumor lost heterozygosity for all patients), (Ml, M2, and M10; 3 of 21 chromosome 22 loci tested. informative patients), (M4 and M10; 2 of 31 Karyotype data, which were available for 14 of the 40 informative patients), (M19; 1 of 31 infor- meningiomas in Table 1, provided support for the view that mative patients), (M8, M10, and M29; 3 of 28 the loss usually involved the entire chromosome. In 6 of 7 informative patients), and (M10; 1 of 4 meningiomas with abnormal karyotypes, one complete copy informative patients). Investigations of markers for chromo- of chromosome 22 was found to be lost. Tumor M40 had some 1 (17 informative patients), (6 infor- multiple aberrations with apparent loss of both copies of mative patients), (20 informative patients), chromosome 22 and several marker chromosomes (Table 1). (12 informative patients), and chromosome (A is a chromosome or (15 informative patients) did not reveal any alterations. fragment whose origin cannot be cytogenetically defined). Statistical analysis (exact Wilcoxon test for ordering, ref. 37) The molecular genetic analysis ofthis meningioma, however, of the marker losses indicated a significantly higher frequen- suggested that only one copy of chromosome 22 was lost cy of reduction to hemizygosity for loci on chromosome 22 (P indicating that the other copy might be included in one of the < 0.01) as compared to any other chromosome. It should be marker chromosomes seen in the corresponding karyotype. noted that the list of meningiomas with loss of chromosome With the exception of tumors M19 and M38 that displayed 22 (Table 1) overlaps only partially with that showing losses normal karyotypes but showed loss of genes on chromosome on other chromosomes. While several meningiomas exhibit- 22 in the molecular genetic analysis, the karyotype data were ed alterations only with chromosome 22 markers, two tumors consistent with those obtained with polymorphic DNA mark- (Ml and M4) without apparent loss of chromosome 22 were ers for this chromosome (Table 1). found to have lost heterozygosity for loci on other chromo- To explore the specificity of the changes involving chro- somes. Certain chromosomal aberrations seen in the mosome 22, we typed DNA from the 40 meningiomas and karyotypes, such as the loss of chromosome 14 in tumor M20 corresponding normal tissue with a panel of 17 randomly and the loss ofchromosomes 4, 11, and 13 in tumor M37, were chosen additional molecular probes for 11 other chromo- not detected with informative DNA probes for these chro- somes. Loss of heterozygosity for one or more of these loci mosomes. Potential explanations for this apparent discrep- Downloaded by guest on September 29, 2021 5422 Medical Sciences: Seizinger et al. Proc. Natl. Acad. Sci. USA 84 (1987) Table 2. Quantitative densitometry of probe hybridization for neurofibromatosis according to National Institutes of Health meningiomas with loss of heterozygosity on chromosome 22 criteria (39). Unfortunately, acoustic neuroma tissue was not Normalized hybridization signals available for this patient. Among the nonfamilial cases, loss of alleles on chromosome 22 was seen in both benign and Chromosome 22/ Tumor/ "malignant" meningiomas. Five of the 40 tumors that were Patient Patient control chromosome normal informative for chromosome 22 markers represented recur- M2 Meningioma 1.76 0.45 rent cases of atypical or malignant meningiomas showing Leukocytes 3.95 aggressive clinical behavior. In the following cases, tumor M5 Meningioma 0.70 0.55 recurrence necessitated multiple operations subsequent to Leukocytes 1.27 the initial surgery: M2, two operations in 3 years; M8, two M8 Meningioma 0.09 0.51 operations in 3 years; M10, three operations in 3 years; M11, Leukocytes 0.17 four operations in 3 years; and M19, two operations in 1 year. M10 Meningioma 0.69 0.52 For this investigation, only tissue from the most recent Leukocytes 1.33 operation was available from each of these cases. All five M1l Meningioma 0.95 0.45 malignant meningiomas displayed loss of genes on chromo- Leukocytes 2.13 some 22. Among the other 35 meningiomas, 12 tumors M15 Meningioma 0.72 0.51 showed loss of chromosome 22 alleles, but these patients Leukocytes 1.43 have not been followed for a sufficient period (all <2 years) M17 Meningioma 0.27 0.40 to exclude the possibility of tumor recurrence. The appar- Leukocytes 0.67 ently higher frequency of chromosome 22 loss in malignant M19 Meningioma 0.23 0.61 meningiomas raises the possibility of a role for this genetic Leukocytes 0.38 aberration in the development of aggressive features or M20 Meningioma 0.85 0.45 recurrence in this tumor type. However, investigation of a Leukocytes 1.88 much larger number of tumors from patients followed over a M25 Meningioma 0.51 0.51 longer period will be needed to clarify this important issue. Leukocytes 1.00 M29 Meningioma 0.59 0.52 DISCUSSION Leukocytes 1.14 M30 Meningioma 0.51 0.56 We have used a molecular genetic approach to show that Leukocytes 0.93 meningiomas frequently display loss of genes on chromo- M31 Meningioma 1.01 0.46 some 22, although this is not the only genetic abnormality Leukocytes 2.21 detected. Our data indicate, in general agreement with M33 Meningioma 0.40 0.40 karyotype studies on cultured tumor cells (4-6), that loss of Leukocytes 1.00 genes on chromosome 22 is significantly more frequent than M38 Meningioma 2.10 0.48 alterations on other chromosomes in primary tumor speci- Leukocytes 4.40 mens. Thus, the reduction to hemizygosity on chromosome M39 Meningioma 1.62 0.56 22 may be one important step in tumorigenesis of menin- Leukocytes 2.91 gioma. Since it is observed in both benign and malignant M40 Meningioma 0.47 0.42 meningiomas, this mechanism might operate at the primary Leukocytes 1.11 level of tumor initiation. Whether the additional chromo- somal alternative causes of tumor Southern blots, which had been hybridized to probes for chromo- changes might represent some 22 (see Fig. 1), were freed of these probes in distilled water for formation or contribute to progression or malignancy remains 2 hr at 650C and rehybridized with probes for polymorphic loci on to be determined. Similarly, investigation of meningiomas other chromosomes (control chromosomes). Heterozygosity for with probes for other chromosomes not included in the RFLPs at these control loci usually provided clear evidence that present investigation might reveal additional chromosome these were not deleted in the tumor DNA (data not shown). The loss relevant to tumor development. hybridization signals on the autoradiograms were analyzed by The majority of the tumors in the present study were quantitative densitometry. Hybridization signals specific to chromo- sporadic meningiomas from individuals with no evident some 22 were normalized to hybridization signals for control chro- family history of nervous system tumors. However, one of mosome probes in the same sample. Then, a ratio of the normalized 22 values for each tumor/normal tissue pair was calculated. The the meningiomas that lost heterozygosity on chromosome reliability of this approach has been demonstrated (7). was obtained from a patient with BANF, an inherited neurological disorder characterized by neoplasia of cells of origin. Its hallmark is the bilateral formation of ancy are that these changes have been present in a small acoustic neuromas and a high susceptibility to a variety of subset of the tumor cells that display a growth advantage in other nervous system tumors, particularly to meningioma culture or that they may have occurred in vitro. There was no (8-11). We have shown (7, 40) that both familial and sporadic evidence for gene amplifications in the genetic analysis ofany cases of acoustic neuroma display specific loss of genes on of the 40 meningiomas, nor did any of the karyotypes show chromosome the of a common double minutes or homogeneously staining regions. 22, indicating possibility Ofthe 40 meningioma patients, 13 (33%) were male, and 27 mechanism of tumorigenesis for meningioma and acoustic (67%) were female. Ages ranged from 19 to 77 years (mean ± neuroma. Furthermore, this finding suggests that the defec- SEM, 52 + 3.2 years). No correlation was found between the tive gene causing BANF may reside on chromosome 22. We age and sex of the patient and any chromosomal alterations have identified (40) acoustic neuromas in which only a including those on chromosome 22. Furthermore, there was portion of chromosome 22 was deleted, narrowing the pos- no correlation between any chromosomal changes and the sible location of the gene for BANF to the region distal to classical histological subtypes of meningiomas (38). With the band 22q11 of the D22S9 locus. The identification of pro- exception of M31, all meningiomas represented sporadic gressively smaller deletions on chromosome 22 in tumors cases without apparent family history of similar tumors. associated with BANF, including acoustic neuromas and Patient M31, whose meningioma tumor tissue showed loss of meningiomas, may well provide a means to clone and chromosome 22, was diagnosed with bilateral acoustic characterize the defect. Downloaded by guest on September 29, 2021 Medical Sciences: Seizinger et al. Proc. Nati. Acad. Sci. USA 84 (1987) 5423 It is noteworthy that two meningiomas with loss of genes 11. Huson, S. M. & Thrush, D. C. (1985) Q. J. Med. 218, 213-224. on chromosome 22 in the molecular genetic analysis (M19 12. Barker, D. & White, R. (1984) Cytogenet. Genet. 37, 250. 13. Julier, C. M. L., Lalouel, J. M., Reghis, A., Szajnert, M. F. & and M38) displayed a normal karyotype. This might be Kaplan, J. C. (1985) Cytogenet. Cell Genet. 40, 664. explained by a selective growth of cell clones with normal 14. Barker, D., Schafer, M. & White, R. (1984) Cell 36, 131-138. karyotype over those with monosomy 22 or overgrowth of 15. Willard, H., Scolnick, M. H., Pearson, P. L. & Mandel, J. L, nontumor cells, such as fibroblasts under in vitro cell culture (1985) Cytogenet. Cell Genet. 40, 360-489. conditions prior to karyotyping. Alternatively, these two 16. Breakefield, X. O., Orloff, G., Castiglione, C., Coussens, L., meningiomas might well represent cases with small deletions Axelrod, F. B. & Ullrich, A. (1984) Proc. Natl. Acad. Sci. USA 81, 4213-4216. on chromosome 22, undetectable by cytogenetic analysis. 17. Darby, J. K., Feder, J., Selby, M., Riccardi, V., Ferrell, R., Siao, The application of additional high-quality DNA markers from D., Goslin, K., Rutter, W., Shooter, E. M. & Cavalli-Sforza, L. L. many different regions on chromosome 22 will facilitate the (1985) Am. J. Hum. Genet. 37, 52-59. detection of such tumors and help to define the limits of the 18. Gusella, J. F., Wexler, N. S., Conneally, P. M., Naylor, S. L., deleted regions. Anderson, M. A., Tanzi, R. E., Watkins, P. C., Ottina, K., Wal- contains a number of loci where lace, M. R., Sakaguchi, A. Y., Young, A. B., Shoulson, I., The human clearly Bonilla, E. & Martin, J. B (1983) Nature (London) 306, 234-238. or deletions predispose to the development of 19. Gusella, J. F., Tanzi, R. E., Bader, P. I., Phelan, M. C., certain sets of tumors (41). They have been termed "reces- Stevenson, R., Hayden, M. R., Hofman, K. J., Faryniarz, A. G. & sive oncogenes" or "anti-oncogenes" (41). For example, the Gibbons, K. (1985) Nature (London) 318, 75-78. RBI gene on chromosome 13q is probably not only important 20. Litt, M. & Howell-Litt, R. (1986) Am. J. Hum. Genet. Suppl. 39, for tumorigenesis of retinoblastoma but also for that of 476. 21. Bell, G. I., Horita, S. & Karam, J. H. (1984) Diabetes 33, 176-183. osteosarcoma, which commonly arises as a second cancer in 22. Barker, D., McCoy, M., Weinberg, R., Goldfarb, M., Wigler, M., children who survive retinoblastoma (42, 43). Three embry- Burt, R., Gardner, 1. & White, R. (1983) Mol. Biol. Med. 1, onal cancers, Wilms' tumor, rhabdomyosarcoma, and hepato- 199-206. blastoma have been shown to cluster in the Beckwith-Wiede- 23. Cavenee, W., Leach, R., Mohandas, T., Pearson, P. & White, R. man syndrome and were found to be associated with loss of (1984) Am. J. Hum. Genet. 36, 10-24. genes on chromosome 11p, suggesting a common pathogenic 24. Dryja, T., Rapaport, J. M., Weichselbaum, R. & Bruns, G. A. P. (1984) Hum. Genet. 65, 320-324. mechanism (44). However, molecular genetic studies with 25. Wyman, A. R. & White, R. L. (1980) Proc. Natl. Acad. Sci. USA additional markers have revealed (45) that the deleted regions 77, 6754-6758. on chromosome 11p in rhabdomyosarcoma are in close 26. Chakravarti, A., Phillips, J. A., 111, Mellits, K. H., Buetow, K. H. proximity, but not identical, to the Wilms' tumor locus. & Seeburg, P. H. (1984) Proc. Natl. Acad. Sci. USA 81, Despite their strikingly frequent clinical association in 6085-6089. BANF, it is not yet clear whether meningioma and acoustic 27. Davies, K. E., Jackson, J., Williamson, R., Harper, P. S., Ball, S., Sarfarazi, M., Meredith, L. & Fey, G. (1983) J. Med. Genet. 20, neuroma arise from a or at the same locus 259-263. on chromosome 22 or whether they result from alterations at 28. Stewart, G. D., Harris, P., Gait, J. & Ferguson-Smith, M. A. two distinct loci. Fine structure molecular genetic analysis of (1985) Nucleic Acids Res. 13, 4125-4132. chromosome 22 deletions in both tumor types will be needed 29. Moorhead, P. S., Nowell, P. C., Mellman, W. J., Battips, D. M. & to settle this issue. Such studies could ultimately lead to the Hangerford, D. A. (1960) Exp. Cell. Res. 20, 613-616. 30. Fearon, E. R., Feinberg, A. P., Hamilton, S. H. & Vogelstein, B. identification of the specific gene or genes responsible for (1985) Nature (London) 318, 377-380. development ofthese tumors, thereby providing a window on 31. Koufos, A., Hansen, M. F., Lampkin, B. C., Workman, M. L., the fundamental genetic mechanisms controlling both normal Copeland, N. G., Jenkins, N. A. & Cavenee, W. K. (1984) Nature and abnormal development of the human nervous system. (London) 309, 170-172. 32. Orkin, S. H., Goldman, D. S. & Sallan, S. E. (1984) Nature We thank Dr. N. T. Zervas, Dr. R. 6. Ojemann, and the (London) 309, 176-178. Neurosurgical Stafffor providing specimens; and A. Lane, K. Riley, 33. Reeve, A. E., Housiaux, P. J., Gardner, R. J., Chewings, W. E., K. Dashner, and J. Logan for technical assistance. We are grateful Grindley, R. M. & Millow, L. J. (1984) Nature (London) 309, to the following scientists for providing DNA probes: Drs. A. 174-176. 34. Fearon, E. R., Vogelstein, B. & Feinberg, A. P. (1984) Nature Ullrich, J. Habener, T. Dryja, G. Bell, C. Shih, R. Weinberg, R. (London) 309, 176-178. White, B. White, H. Goodman, J. L. Mandel, G. Fey, G. Stewart, 35. Cavenee, W. K., Dryja, T. P., Phillips, R. A., Benedict, W. F., M. Litt, and P. Leder. B.R.S. is supported by a fellowship and a grant Godboit, R., Gallie, B. L., Murphree, A. L., Strong, L. C. & from the National Neurofibromatosis Foundation. J.F.G. is a Searle White, R. (1983) Nature (London) 305, 779-784. Scholar of the Chicago Community Trust. R.L.M. is recipient of the 36. Dryja, T. P., Cavenee, W., White, R., Rapaport, J. M., Peterson, National Institute of Neurological and Communicative Disorders and R., Albert, D. M. & Bruns, G. A. (1984) N. Engl. J. Med. 310, Stroke Teacher-investigator Career Development Award NS00654. 550-553. Funds for this work were provided by Grants NS00654, NS22224, 37. Mehta, C. R., Patel, N. R. & Tsiatus, A. A. (1984) Biometrics 40, and NS20025 from the National Institute of Neurological and 819-824. Communicative Disorders and Stroke and by a grant from the 38. Zulch, K. J., ed. (1979) World Health Organization Histological McKnight Foundation. Typing ofthe Central Nervous System (World Health Organization, Geneva). 1. Rubenstein, L. J. (1972) Tumors of the Central Nervous System 39. Mulvihill, J., ed. (1986) Neurofibromatosis Res. Newsl. 2, 1. (Armed Forces Institute of Pathology, Washington, DC). 40. Seizinger, B. R., Rouleau, G., Ozelius, L. J., Lane, A. H., St. 2. New, P. F., Hesselink, Y. R., O'Carroll, C. P. & Kleinman, G. M. George-Hyslop, P., Huson, S., Gusella, J. F. & Martuza, R. L. (1982) Am. J. Neuroradiol. 3, 267-276. (1987) Science 236, 317-319. 3. Mirimanoff, R. O., Dosoretz, D. E., Linggood, R. M., Ojemann, 41. Knudson, A. 6. (1985) Cancer Res. 45, 1437-1443. R. G. & Martuza, R. L. (1985) J. Neurosurg. 62, 18-24. 42. Hansen, M. F., Koufos, A., Gallie, B. L., Phillips, R. A., Fodstad, 4. Zang, K. D. (1982) Cancer Genet. Cytogenet. 6, 249-274. 0., Br0gger, A., Gedde-Dahl, T. & Cavenee, W. K. (1985) Proc. 5. Mark, J. (1977) Adv. Cancer Res. 124, 165-222. Natl. Acad. Sci. USA 82, 6216-6220. 6. Zankl, H. & Zang, K. D. (1980) Cancer Genet. Cytogenet. 1, 43. Friend, H. S., Bernards, R., Rogeli, S., Weinberg, R. A., 351-356. Rapaport, J. M., Albert, D. M. & Dryja, T. P. (1986) Nature 7. Seizinger, B. R., Martuza, R. L. & Gusella, J. F. (1986) Nature (London) 323, 643-646. (London) 322, 644-647. 44. Koufos, A., Hansen, M. F., Copeland, N. G., Jenkins, N. A., 8. Eldridge, R. (1981) Adv. Neurol. 29, 57-65. Lampkin, B. C. & Cavenee, W. K. (1985) Nature (London) 316, 9. Martuza, R. L. & Ojemann, R. G. (1982) Neurosurgery 10, 1-12. 330-334. 10. Kanter, W. R., Eldridge, R., Fabricant, R., Allen, J. C. & 45. Scrable, H., Witte, D., Wang-Wuu, F., Koufos, A. & Cavenee, Koerber, T. (1980) Neurology 30, 851-859. W. K. (1986) Am. J. Hum. Genet. Suppl. 39, 117. Downloaded by guest on September 29, 2021