[CANCER RESEARCH 55. 3865-3872.
September 1, 1995]
Analysis of Meningiomas by Methylation- and Transcription-based
Clonality
Assays1 Jiguang Zhu, Matthew P. Frosch, Lambert Busque,2 Alan H. Beggs, Kathleen Dashner, D. Gary Gilliland, and Peter M. Black3 Neurosurgical Laboratories ami Brain Tumor Center, Brigham & Women's Hospital, Department of Surgery, Children's Hospital, Dana-f-'arber Cancer Institute ¡J.Z. K. D., P. M. B.I; Division of Hematology/Oncology, Brigham & Women's Hospital, Department of Medicine ¡L B., D. G. G.J; Department of Pathology, Brigham & Women's Hospital ¡M.P. F.¡;and Genetics Division, Children's Hospital ¡A.H. B.¡,Harvard Medical School, Boston, Massachusetts 02115
Although this method provides valuable and definitive data on a number of diseases, including clonal origin determination of myelo-
ABSTRACT The clonal derivation of tumors can be determined by X chromosome inactivation analysis based on differential expression of genes or differ ential methylation of cytosine residues in CpG islands near polymorphic loci. In this report, we compared a transcription-based RNA analysis with a methylation-based DNA assay to determine clonality of meningiomas. Both clonality assays use PCR-based analysis at the human androgen-
proliferative disorders (4), it is limited by the low frequency of heterozygosity in females of many ethnic populations. Vogelstein et al. (5) introduced a molecular genetic approach to assess the clonal origin of tumors using DNA RFLP and the differ ential methylation of nearby cytosine residues on the X chromosomes. Combining RFLPs occurring in the X-linked genes HPRT and PGK, this methylation-dependent method is informative in over 50% of
receptor gene (HUMARA ) on the X chromosome. Among 23 meningiomas from female patients, 19 were informative hétérozygotes at this locus (83%). The patterns of X chromosome inactivation in four patients were extremely skewed towards one alÃ-ele in blood (unequal Lyonization), which precluded clonality determination in the tumor samples. Concord ant clonality results with methylation- and transcription-based clonality assays were demonstrated in 9 of 13 informative tumors expressing the androgen receptor. Seven meningiomas were monoclonal, but surpris ingly, two pathologically documented cases of meningiomas were polyclonal. There was disparity in 4 of 13 tumor specimens that were polyclonal by the methylation-based assay but monoclonal by the transcription assay. Clonality examination of these tumors by the methylation-based
human tumors in females (6). Another clonality assay that uses a highly informative (90%) variable number tandem repeat marker, M27ßat the DXS255 locus, has expanded informativeness to include most females (7). However, hypermethylation at the M27ßlocus limits its application (8, 9). In addition, HPRT, PGK, and M27ß techniques are limited by the requirement of large amounts of DNA (5-10 /¿g)for Southern blot analysis. A modified clonality assay at the PGK locus has been devised to study clonality of myeloproliferative disorders with small numbers of cells by means of PCR (10). The validity of this method was confirmed by van Kamp et al. (11) through comparative clonal analysis of hematopoietic cells by the PCR-based assay with that of Southern blot analysis. Recently, Allen et al. (12) developed a methylation-based clonality assay that takes advantage of a highly polymorphic trinucleotide repeat (CAG) in the coding region of the first exon of the human androgen receptor gene on the X chromosome (HUMARA ). This CAG repeat is closely linked to four methylation sites that have served as the basis for studying patterns of X chromosome inactivation in female carriers. The advantages of this clonality assay include high frequency of heterozygosity (90%), reliable methylation patterns, and that it is PCR-based so that a small amount of tumor specimen from
phosphoglycerate kinase assay provided identical results to the methyla tion-based analysis at the HUMARA locus. In addition, loss of heterozygosity (LOH) studies of chromosome 22, which is frequently deleted in meningiomas, showed that four of four informative samples of the six polyclonal tumors had partial LOH in tumor tissues. However, complete LOH was observed in primary cultured cells, which were also monoclonal by the methylation assay. Taken together, these data suggest that the disparity of the two assays in these four cases may be due to differences in the level of expression of the androgen receptor gene in tumors. Therefore, we conclude that: (a) clonal derivation of meningiomas determined by both transcription- and methylation-based clonality assays are in full agreement in many (9 of 13) but not all cases (4 of 13); and (In most meningiomas (9 of 15) are monoclonal in origin, whereas some meningioma samples (6 of 15) are polyclonal or may contain heterogeneous components.
INTRODUCTION
Characterization of clonal derivation of human neoplasms has provided important information about etiology and pathogenesis and has practical implications for both diagnosis and subsequent studies of disease progression. Analysis of clonality in females heterozygous for specific genes or polymorphic markers on the X chromosome has been widely used for determination of tumor origin (1). The tradi tional method of clonality analysis developed by Fialkow (2, 3) used isozyme expression of the X chromosome-linked G6PD4 gene. Received 3/20/95; accepted 7/6/95. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1This work was supported by the Boston Neurosurgical Foundation and Brain Tumor Society. L. B. is supported by the National Cancer Institute of Canada. " Present address: 5415 Assomption Boulevard. Montreal, Quebec, HIT 4B3, Canada. ' To whom requests for reprints should be addressed, at Brigham & Women's Hospital, Neurosurgcry, 75 Francis Street, Boston. MA 02115. 4 The abbreviations used are: G6PD, glucose-o-phosphate hypoxanthine phosphoribosyl transferase; PGK, phosphoglycerate ratio; LOH. loss of hetero/ygosity; RT, reverse transcripta.se.
dehydrogenase;
HPRT,
kinase: CR, clonality
OCT or paraffin can be successfully analyzed (13). By using this technique, monoclonality in Langerhans cell histocytosis has been documented, and the results were confirmed with the PGK and M27ß assays (14). However, there has been concern that differential methylation at these loci is not a reliable marker for the state of activation of the X chromosome or that variable methylation may occur in association with malignancy (15-19). This is well exemplified at the DXS255 locus analyzed with the M27ßprobe, which is hypermethylated in a significant proportion of acute myelogenous leukemia blast popula tions (8, 9). Inconsistent data have been reported by groups using the methylation and expression assays. For example, parathyroid adeno mas were determined as multicellular in origin by an expression-based clonality assay (20), whereas the same type of tumor was shown to be monoclonal by methylation-based clonality techniques (21, 22). This discrepancy may be due to several causes, including the complexity of tumor, normal tissue contamination, abnormal methylation, or vari able expression of the G6PD gene between tumor and normal tissues. Therefore, it is very important to conduct comparative clonality de terminations in the same tumor using both expression- and methyla tion-based clonality assays. Until recently, no single locus has had a coding polymorphism coupled with differential methylation sites that
3865
Downloaded from cancerres.aacrjournals.org on February 13, 2019. © 1995 American Association for Cancer Research.
CLONALITY
ANALYSIS
would allow comparison of both the expression- and methylationbased assays at the same locus. Such a combination would allow validation of methylation-based assays in comparison with expression analysis at the same locus. A transcription assay was recently developed at the HUMARA locus using a RT-PCR strategy that allows clonality determination without relying on differential methylation of the X chromosomes (23). It permits direct comparison of methylation-based clonality assay with transcription-based assay at the same locus. However, the technique is limited by the level of transcription of the androgen receptor gene in the tissue of investigation (24). In order to validate this new assay and to compare the results of transcription- and methylation-based clonal ity assays, we decided to investigate the origin of clonality in a large number of solid tumors with both methylation- and transcriptionbased clonality techniques. Meningiomas are one of the most common primary intracranial tumors and are thought to arise from the arachnoid cells of the meninges. They occur frequently in middle to late years of life as sporadic solitary tumors with slow growth and rare infiltration of surrounding tissues. They also occur as heritable tumors associated with bilateral acoustic neurofibromatosis (NF2) (25, 26). Jacoby et al. (27) have shown that meningiomas are monoclonal in origin by methylation-based method with PGK and HPRT probes. Cytogenetic and molecular genetic analyses have demonstrated that loss of part or a copy of chromosome 22 (monosomy) is a predominant genetic abnormality in meningiomas (28, 29). Recently, it has been shown that the NF2 tumor suppressor gene, on chromosome 22 is inactivated in 60% of sporadic meningiomas (30, 31). There is some evidence that meningiomas may be stimulated by hormonal components. It is known that the incidence of meningioma is twice as high in women than men and the tumors enlarge during pregnancy (25, 26). There is also association between meningioma and breast cancer (32, 33). Many studies have shown an increased level of expression and nuclear localization of steroid hormone re ceptors, including progesterone receptor and androgen receptor, in meningiomas (34, 35). Based on the results that the androgen receptor is expressed in most meningiomas (35) and the availability of a large number of tumor specimens from female patients, we decided to directly compare the two clonality assays through analysis of meningiomas. We examined 19 informative sporadic meningiomas from female patients using both methylation- and transcription-based clonality assays at the human androgen receptor gene locus (HUMARA). Our data demonstrate that clonality results obtained with the two techniques are concordant in most cases and confirm that most meningiomas are monoclonal, although some samples are polyclonal. MATERIALS
OF MENINGIOMAS
evaluation by a neuropathologist (Fig. 3). The adjacent sections were subjected to DNA isolation. Patient records were reviewed, and tumors were classified clinicopathologically (Table 1) on the basis of presenting clinical signs, symp toms, and pathological examinations. Approval for this study was granted by the Human Research Committee of the Brigham & Women's Hospital. Isolation of Blood DNA, Tumor DNA and RNA from Frozen Tissues, and DNA from OCT Specimens (Cryostat Sections). Blood DNA was extracted from peripheral blood leukocytes by SDS/Proteinase K digestion followed by phenol and chloroform extraction (36). Tumor DNA and RNA were extracted simultaneously from the same fresh frozen tissue by a modified procedure as described previously (37, 38). Isolation of tumor DNA embedded in OCT compound was performed according to an existing protocol (39). Briefly, one 6-jj.m and 4-5 10-fim sections were cut serially at -20°C in a cryostat. The 6-/j.m section was stained with hematoxylin
and eosin for
pathological examination, and the adjacent sections were subjected to DNA isolation with proteinase K digestion followed by phenol/chloroform extractions (39). Methylation-based Clonality Assay at the Androgen Receptor Locus. The PCR-based clonality analysis at the human androgen receptor gene (HU MARA) was performed as described previously with the following modifica tions (Fig. 1A; Ref. 12). In a 20-^1 volume, 200 ng of genomic DNA were pretreated with or without 12 units Hpa\\ New Haven, CT) in 10X PCR buffer [0.5 BSA, and 15 mM MgCl,]. After digestion, mixture of 10 /il final volume containing DuPont)
end-labeled
(International Biotechnologies, Inc., M KC1, 0.2 M Tris (pH 8.6), l /xg/^l 5 /j.1was taken and added to a PCR [y-32P]ATP (3000 Ci/mmol; NEN,
primer, and the PCR reaction
was then performed
according to conditions described (12, 38). The PCR products were mixed with an equal volume of formamide loading buffer (95% formamide, 20 mM EDTA, 0.05% bromophenol blue, and 0.05% xylene cyanol), denatured at 95°Cfor 5 min, and cooled on ice. About 2 ¡i.\ of each sample were loaded onto a 6% denaturing polyacrylamide gel containing 8 Murea, and the gel was run at 100 W for 4 h. Subsequently, the gel was fixed, dried, and exposed to XAR-5 (Kodak) film with an intensifying screen at -80°C for 1-3 days. Transcription-based The transcription-based
Clonality Assay at the Androgen Receptor Locus. clonality assay was conducted as reported (Fig. Iß;
Ref. 23) with the following changes. The RNA samples (1 /xg) were first Table 1 Clonaliiy results and clinicopathological
Methylation
features of meningionuts
Expression
Cases"KidnevNB1NB2N.
»ft»-(yrs)60496331392443514072503739707869793569Tumor
typesSclerotic
DNDND14.1CNDNDNDNDNDNDND.3.3.1.3.32.6NDNDNDNDTissue1.5"1.51.0ND3.6N AraMlM2M3M4M5MhM7MHM9MIOMilM12M13M14M15M16M17M18M19Tissue1.1*1.61.32.63.4'19.66.25.16. meningiomaMeningotheliamatousSyncytialT
AND METHODS
Collection of Tissue and Blood Samples, Pathological Evaluation of Tissue, and OCT Tumor Specimens. Meningioma specimens were obtained at the time of surgery. Tumor tissue was evaluated by a staff neuropathologist, and a small portion of each specimen was excised for pathological examina tion. After sampling, fresh tissues were snap-frozen and stored in liquid nitrogen, and a small portion of the same tissue was embedded in OCT compound (Baxter Scientific. Bedford. MA) and stored at -70°C. Normal brain tissue was obtained from patients with medically intractable epilepsy following temporal corticectomy. Normal kidney, where the androgen receptor gene is shown to be expressed at high levels in mammals, was obtained at autopsy from a female and served as a control specimen in the transcription-
" NB, normal brain; N. Ara, normal arachnoid; NE, no detectable expression of the
based clonality analysis. Peripheral venous blood samples were obtained at the time of tumor resection from patients. Sectioning of OCT blocks was carried out in a cryostat at — 20°C.A 6-/xm
androgen receptor gene; NI, not informative; Mono, monoclonal; Poly, polyclonal; ND, not done. h Corrected ratio for normal tissue only. ' CR (see "Materials and Methods"). d Ratio of two alÃ-elesin RT+ sample. ' Extremely skewed towards one alÃ-elein lymphocyte DNA similar to the skewing
section was cut and stained with hematoxylin
pattern in tumor DNA (see Ref. 13 for details).
and eosin for histológica! 3866
Downloaded from cancerres.aacrjournals.org on February 13, 2019. © 1995 American Association for Cancer Research.
CLONAUTY
A
ANALYSIS
Genomic DNA Restriction digestion with Hpa n or Hha I (CAG)n 3 and 6; samples were polyclonal with CRs < 3 (Fig. 2; Table 1). To conditions for PCR reaction, gel separation, and autoradiography are the same assess the possibility that normal tissues mixed with the tumor spec as described above. Methylation-based Clonality Assay at the PGK Locus. Screening for imens were amplified by PCR, resulting in polyclonal patterns in the heterozygosity at the PGK locus was carried out as described (10). To deter six tumors, we repeated the clonality analysis on DNAs extracted mine clonality, 100 ng of DNA were digested with or without 12 units Hpall from OCT samples where the presence of tumor cells had been (International Biotechnologies, Inc., New Haven, CT) and then amplified with confirmed by histopathological examination (Fig. 3). The histological conditions as described previously (10), except one primer was end-labeled examination of adjacent serial section confirmed that the tissue sam with [7-12P]ATP (3000 Ci/mmol; NEN, DuPont). The Bs/XI-digested PCR ples came from the lesionai tissue. Although nonneoplastic elements products were separated on a 1.5% agarose gel, and the gel was Southern are present in all tumors, these elements appeared to represent a transferred to a nylon filter and subjected to autoradiography at -80°C minority of these regions chosen under investigation (Fig. 3). The overnight. methylation-based clonality analysis demonstrated that all six samples Cell Culture, DNA Extraction, and LOH Analysis of Chromosome 22. from OCT are still polyclonal in origin (Fig. 2B), consistent with Primary cultured cells were obtained from one tumor tissue (M12). After examination by a neuropathologist, a portion of the fresh tumor specimen was clonality data obtained from frozen tumor tissues (Fig. 2). The CRs of 3867 Polyclonal
Monoclonal
Monoclonal
Downloaded from cancerres.aacrjournals.org on February 13, 2019. © 1995 American Association for Cancer Research.
CLONALITY ANALYSIS OF MLNINGIOMAS
A
MlNormal MS „• ,-—~-—^
Fig. 2. Methylation-hased clonality analysis of fresh frozen samples (A) and frozen tissue and OCT specimens (B) at the HUMARA locus on the X chromosome. The marker is a trinucleotide tandem repeat (CAG). and its PCR products are around 280 bp. A, parallel samples tor blood (B) and tumor (7") were predigested with ( + ) or without (-) Hpall restriction enzyme. The normal arachnoid was served as a control for polyclonality with a cor rected ratio of 2.6. The CRs for M5, Ml, M2, M3, and M13 are 6.0, 3.4, 19.6, 6.2 and 1.1, respec tively. M19 was precluded from clonality determi nation due to extreme skewing towards one alÃ-ele in the blood (Lyonization). 8. both frozen and OCT samples were from the same tumor specimens. M2 was identified as monoclonal in origin (A ) and was served as a positive control for complete Hpall digestion of DNA extracted from OCT. The CRs for both frozen tissues and OCTs are listed in Table 1. R, blood; 7",frozen tissue; OCT. tumor slices cut out of OCT compound; -, no Hpall; +. with
,
*
M34^^-.
M19k
JL-'•iMÕ3 ^ *
MAM2
M2
both frozen tissues and OCT tumors are shown in Table 1. It is unlikely that the polyclonal patterns observed here resulted from incomplete Hpa\\ digestion, since one master mix was prepared for all Hpa\\ digestions and a positive control included in all experiments was always monoclonal (Fig. 2, M2). In addition, each experiment was repeated at least two times, and the results were the same as shown in Fig. 2. Analysis of Tumor Tissues by Transcription-based Clonality Assay. Compared to the methylation-based assay, the transcriptionbased clonality analysis uses RT of RNA to cDNA, which obviates the
need to use methylation status at a given locus as a surrogate for X chromosome status (Fig. 1). That is, androgen receptor transcripts will only be expressed from an active X chromosome. Total RNA used for this analysis was isolated from the same tumor tissue where DNA was obtained for methylation clonality assay using a modified procedure to extract DNA and RNA simultaneously (38). One kidney and two normal brain samples (NB1 and NB2) expressing detectable levels of androgen receptor message were included as polyclonal controls. Methylation-based clonality analysis of DNA from these normal tissues was performed for comparison with transcription assay results
m Fig. 3. Representative pathological examination of tumor specimens from OCT sections. Sectioning of OCT blocks was carried out in a cryostat at -20°C. One 6-^im section was cut and stained with hematoxylin and eosin. A, meningiotheliamatous meningioma from M 14.
meningioma from M2; B, transitional meningioma from M12; C transitional meninginma from M15; D, secretory 3868
Downloaded from cancerres.aacrjournals.org on February 13, 2019. © 1995 American Association for Cancer Research.
C'LONALITY
Kidney DNR
B
NB1 RT
M9
M12 —-—. B
RT
DNR
•*••• -
B
RT
NB2 RT
M13 .—-— B RT
DISH
MS RT
Mie * B
ANALYSIS
,
RT
MIS
Mil RT
T
RT
• • -! Fig. 4. Transcription-based clonality assay at HUMARA in kidney, normal brains, and mcningioma samples (A and ß).DNAs from human kidney and two normal brains (NB1 and NB2). predigested with ( + ) or without (-) Hpa\\, were subjected to methylation clonality analysis. For meningiomas, blood DNAs (H) were amplified directly with the HUMARA primers and used as markers for locations of two alÃ-elesfor RT-PCR tran scription clonality analysis. Total RNAs from normal tissues and frozen tumor tissues were reverse transcribed with ( + ) or without ( — ) RT followed by PCR amplification with HUMARA.
OF MENINGIOMAS
tumors, including MIO and M12, which had discordant results in the previous assays. Representative results are shown in Fig. 6. The intensity ratio of the 530-bp band to the 433-bp band was approxi mately 3:1 for polyclonal samples. This is due to the formation of heteroduplex (50%) after PCR amplification, which resists the B.\t\l digestion. Therefore, only the predicted 25% PCR products that have the AsrXI site on each strand of a homoduplex would be digested (10). Based on this PGK assay, M4 and M7 are monoclonal, MIO (data not shown) and M12 are polyclonal, and M18 is precluded from clonality determination due to extreme Lyonization in the blood (Fig. 6; Table 1). LOH Analysis of Chromosome 22 and Clonality Analysis of Primary Cultured Cells by the Methylation Assay. To further study the biology of one tumor with discordant results, we took advantage of the primary cultured cells available from one tumor tissue (Ml2) prepared previously for other experiments. A portion of the original tumor specimen (Ml2) was excised and subjected to cell culture upon receipt of the tissue from the operating room. After two and three passages, the cells were harvested, and DNAs were isolated. LOH analysis of tumor DNA and DNA from cultured cells was performed with a polymorphic marker, F8VWFP. on chromosome 22 (41). Partial loss (incomplete loss) of chromosome 22 was detected in frozen tissue of all four informative samples (Ml2, M13, MIO, and M15) of the six tumors, which are polyclonal in origin by methylation assay (Fig. 7/4). However, complete loss of chromosome 22 was observed in DNA from second and third passages of primary cultured cells (Fig. 7A, M12). Methylation-based clonality analysis (HU MARA) of DNA from cultured cells demonstrated that these cells are monoclonal in origin with a CR of 6.5 (Fig. IB), although polyclonality was detected in the parallel tumor tissue (Fig. IB) and in the OCT sample from the same patient (Fig. 2B).
and for the positions of the two alÃ-eles.For each RT, two identical reactions with (+) and without (-) RT were carried out, where RT-
The CRs for the kidney and two normal brains were 1.5, 1.5, and 1.0, respectively, indicating that they are polyclonal in origin by the transcription assay (Fig. 4/4; Table 1). No expression of the androgen receptor gene was detected in two meningiomas (Fig. 4B. M2 and A/9; Table 1). The absence of RT-PCR products from M2 and M9 likely results from a limited or lack of gene expression from the androgen receptor rather than degradation of the RNA during DNase I treat ment, since a master solution was used for all the samples and RT-PCR amplifications were observed in most specimens (Table 1; Fig. 4). In addition, RT-PCR products were observed for other genes, such as dopamine 1 and 2 receptors, from the M2 RNA and other meningiomas using the same batch of DNase I (data not shown). Application of this assay to the remaining 13 meningiomas that express the androgen receptor gene revealed that 11 samples are monoclonal in origin with CRs over 3.0, whereas 2 samples are polyclonal with CRs of 1.7 and 1.2 (Figs. 4 and 5; and Table 1). The results for four apparently monoclonal tumors (M10-M13) are at variance with the previous data derived from the methylation assay (Table 1). Analysis of OCT Specimens by Methylation-based Clonality Assay at the PGK Locus. To resolve the inconsistent results of these four tumors that were shown to be polyclonal by methylation assay but monoclonal by transcription assay, we further studied clonality in these samples by the PCR-based PGK method, which is independent of HUMARA locus and has been shown to be a reliable assay (10, 11, 44). Heterozygosity for BstXl digestion was identified in 5 of 19
MI2
MS
reactions served as controls for possible DNA contamination, which would produce two bands. In each of the three normal tissues, two bands corresponding to the ones amplified from DNA were observed in RT+ lanes, while no band was detected in RT—lanes (Fig. 4,4).
B
T
RT
••i
T
M14 HT
T
RT
fill Bfc
Fig. 5. Direct comparison of methylation- .imi transcription-based clonality assay at HUMARA. For each patient, the first four lanes are results of methylation clonality analysis of DNA. R. blood; T, tumor from Im/ai tissue; —.without ///will; + , with HfxiU. The last two lanes for each patient are transcription clonality analysis of RNAs from the same tissues. RT, reverse transcriptase; -, without RT; +, with RT. The result shown in Lune M14. RT+ was a longer exposure of 7 days in — 80°C,and the two bands of equal intensity are evident, while no band is ilctccted in the RT- lane of the same autoradiogram (data not shown).
M7
M12
M18
530 bp 433 bp
Fig. h. PCR-based PGK clonalily analysis of meningioiila v Blood (B) and tumor ('/') DNA from OCTs predigested with ( + ) or without (-) Hpa\\ were PCR amplified with one of the two primers end-labeled by •'"P. The PCR products were then digested with ÄvfXI. and the mixtures were separatedon an agaroscgel, transferred to a nylon filter, and visualized by auloradiography.
3869
Downloaded from cancerres.aacrjournals.org on February 13, 2019. © 1995 American Association for Cancer Research.
U.ONALITY
M12
M13
MIO
ANALYSIS
B
BTOCC'BTBTO
Fig. 7. A. loss of heterozygosity analysis of chromosome 22 in tumors and primary cultured cells. The intensity of each alÃ-eleis quantified by Phosphorlmager, and the ratios of the upper alÃ-eleover the lower one of each lane are 0.85, 2.8, 2.7, 13.2, and 10.2 for B. T, O, C, and C' in M12, respectively; 1.0 and 1.7 for B and Tin M13, respectively; and 1.0, 2.2. and 2.5 for B. T, and O in MIO, respectively, ß.blood; T. fresh-frozen tumor; O, OCT tumor; C and C'. second and third passages of primary cultured cells from the same tissue specimen, respectively. B. methylation-based clonality analysis of primary cultured cells from MI2 at HUMARA. Tu, tumor; Cell, second passage of primary cultured cells; -, without Hpatt; +. with Hpa\\.
DISCUSSION Expression-based clonality assays, such as the G6PD assay (3), have been considered as the "gold standard" for determination of
OF MCMNCilOMAS
are also evident for these four cases. For example, the upper alÃ-elein T+ lane of M13 (Fig. 2/4) by the methylation assay is lower in signal intensity than the lower alÃ-ele,suggesting that the upper alÃ-eleis unmethylated and from the active X chromosome. In transcription clonality analysis of the same tumor, the upper alÃ-elewas identified expressing the androgen receptor gene as shown in the RT+ lane of M13 in Fig. 4B. The same relationship of unmethylation in DNA corresponding to transcription in RNA was also observed in the other three cases (MIO, Ml 1, and M12). These data suggest that the results of clonality analysis by both methods are consistent in all cases, and the apparent disparity in the four cases may be due to a small proportion of clonally derived cells expressing relatively high levels of the androgen receptor gene. Polyclonality shown by methylation clonality assays in six cases is a striking finding and is unlikely to be explained by experimental artifacts. Incomplete digestion of DNA from either fresh-frozen tissue or OCT by Hpall was possible but unlikely, since complete digestion was observed in monoclonal samples (such as M5 and M2 in Fig. 2), which shared the master mixture for Hpall digestion with other polyclonal samples. In addition, the same results were obtained in at least two repetitive experiments, which included a positive control of known monoclonality in every experiment. Significant contamination by normal tissue or nontumor tissue intermixed with tumor mass was not likely either, since meningiomas are benign solitary tumors with rare infiltration of adjacent tissues. The tumor samples subjected to investigation were usually taken from the center of the tumor masses by neurosurgeons and were always examined by staff neuropatholo gists. Furthermore, the clonality results of DNA extracted from OCT sections, which were reviewed by histological examination of a neu ropathologist to exclude bulk contamination (Fig. 3), are in full agreement with the clonality data from fresh-frozen tissues. Abnormal methylation at the HUMARA locus is highly unlikely, since PCRbased clonality analysis at the PGK locus of informative tumor specimens provides concordant results as HUMARA methylation as say (Fig. 6). Based on our data, one might speculate that heterogeneity of the tumor mass at early stages of clonal expansion may explain polyclonality. The monoclonality of the four tumors demonstrated by the transcription-based clonality assay suggests that these tumors are monoclonal in origin, or they resulted from differential levels of transcription of androgen receptor between clonal tumor and other components. Since the androgen receptor gene is not a housekeeping gene and its level of expression varies in different tissues (24), including high levels of expression in meningiomas (35), it is possible that the small population of clonal tumor cells expresses a signifi cantly higher level of the androgen receptor in comparison to reactive or another population of clonal cells. This will bias the assay in the favor of this small population of clonally derived cells. We, therefore, view the assay as complementary, with the DNA methylation-based assay giving an estimation of the number of clonal cells in the sample and the transcription assay in this setting confirming the presence of a clonal population of cells. Results obtained with LOH analysis of chromosome 22 support this explanation. Incomplete LOH of chromosome 22 in all four inform ative cases of the six tumors provides evidence that there are two or more different population of cells within the tumor mass (Fig. 7/4). Complete LOH detected in primary cultured cells derived from the same tumor suggests that the clonal population of cells has a proliferative advantage during cell culture, as would be expected for neo-
clonality. However, their application is limited by low frequency of heterozygosity in many ethnic groups (45). The development of a transcription assay at the highly polymorphic HUMARA locus using a RT-PCR strategy allows clonality determination without relying on differential methylation of X chromosomes (23). We have compared the results of a transcription assay through analysis of RNA (23) with that of a methylation assay of DNA (12) from the same tumor at the same locus in meningiomas. Twenty-three sporadic meningioma specimens from female pa tients were collected, of which 19 were heterozygous at the HUMARA locus (83%). Four cases (21%) were excluded from clonality deter mination due to severe skewing towards one alÃ-elein //pall-digested lymphocyte DNA (Fig. 2A. M19; Table 1). The frequency (4 of 19) of extreme Lyonization observed in our experiments (21%) is consistent with observations in other clonality studies (8, 43). Expression of the androgen receptor gene was not detected in two cases (Fig. 45. M2 and M9), and clonal origin of these two tumors was demonstrated by methylation-based clonality assay only (Fig. 2A; Table 1). Clonality analysis of the remaining 13 informative meningiomas with both transcription and methylation assays has revealed that the results generated by the two techniques are concordant in nine cases (69%; Table 1). Seven of the nine tumors were identified as monoclonal in origin by both assays with CRs above 3.0 (Table 1). In the methyla tion-based assay, Hpall cleaves the active alÃ-elesuch that only the inactive alÃ-eleis amplified, whereas in the transcription-based assay, only the alÃ-eletranscribed from the active X chromosome is amplified. Therefore, opposite skewings of the alÃ-elesbetween Hpal I-treated tumor lanes in the methylation assay and RT+ lanes in the transcrip tion assay are expected. It provides visual assurance of monoclonality and supports consistency between the two methods (Fig. 5, MS). In two cases (M14 and M15), polyclonality was demonstrated by both methods with CRs below 3.0 (Table 1). However, differences in band intensities between the two alÃ-elesin both methylation (Fig. 2B, MI5) and transcription clonality results (Fig. 4B, MI5) suggest that small proportions of clonal cells are present in this tumor. Surprisingly, four cases (M10-M13) yielded inconsistent results, where polyclonality was observed by the methylation assay, but monoclonality was detected by transcription-based assay. However, careful analysis of the skewing patterns of the same patient between plastic cells (Fig. 7/4. MI2). In support of this hypothesis, cultured T+ lanes of methylation data (Fig. 2) and RT+ lanes in transcription meningioma cells derived from polyclonal tissue were monoclonal by the methylation assay (Fig. IB). Northern blot analysis of RNA from results (Fig. 4B) indicate that the two sets of data are also reconcil such cells with androgen receptor probe demonstrated transcription of able. Similar opposite skewing of alÃ-elesbetween Hpall precut tumor DNA and RT+ RNA lanes observed in the seven monoclonal samples androgen receptor at similar levels as detected in tumor tissue (Ml2; 3870
Downloaded from cancerres.aacrjournals.org on February 13, 2019. © 1995 American Association for Cancer Research.
CLONALITY
ANALYSIS OF MEMNOIOMAS
data not shown). Based on these data, we speculate that these six tumors may be heterogeneous in composition containing clonal tumor cells and another polyclonal component, which most likely represents normal tissue due to hormonal stimulation, since it has been shown that many hormone receptors, including progesterone receptor and androgen receptor, are expressed in meningiomas (34, 35). Jacoby et al. (27) have shown that meningiomas are monoclonal in 9 informative patients by an HPRT methylation assay. However, in four of the nine tumors (44%), the CRs of the four tumors were less than 3.0, including two cases at the border line between monoclonal and polyclonal, which is consistent with our results (27). Recently, Ruttledge et al. (46) reported that three tumor fragments randomly separated from the same tumor mass (case 119) with no difference from one another pathologically were subjected to LOH analysis of chromosome 22. They found LOH in one of the fragments but not the other two fragments. This suggests that the tumor is not homogeneous at the DNA level, although they are indistinguishable by pathological means. Similar observations have been reported in thyroid nodules, where inconsistent data were obtained between molecular analysis and histomorphological examination of 39 thyroid nodules (47). Genetic heterogeneity in colorectal adenomas, but not in adenocarcinomas, has been re ported by Shibata et al. (48). They found c-K-ra.v gene mutations
REFERENCES 1. Wainscoal, J. S.. and Fey, M. F. Assessment of clonality in human tumors: a review. Cancer Res., 50.- 1355-1360. 1990. 2. Fialkow, P. J. Use of genetic markers to study cellular origin and development of tumors in human females. Adv. Cancer Res., /5: 191-226. 1972. 3. Fialkow. P. J. Clonal origin of human tumors. Biochim. Biophys. Acta, 45ft: 283-321, 1976. 4. Fialkow. P. J.. Gartlcr. S. M.. and Yoshida. A. Clonal origin of chronic myelocytic leukemia in man. Proc. Nati. Acad. Sci. USA. 5«:1468-1471, 1967. 5. Vogelstein, B„Fearon. E. R.. Hamilton. S. R., and Feinherg. A. P. Use of restriction fragment length polymorphisms to determine the clonal origin of human tumors. Science (Washington DC), 227: 642-645, 1985. 6. Vogelstein, B., Fearon. E. R., Hamilton, S. R.. Precinger. A. C.. Willard, H. F., Michelson, A. M., Riggs, A. D., and Orkin. S. H. Clonal analysis using recombinant DNA probes from the X-chromosome. Cancer Res., 47: 4806-4813. 1987. 7. Abrahamson, G., Fraser. N. J., Boyd, J., Craig, 1., and Wainscoat, J. S. A highly informative X-chromosome probe. M27 beta, can be used for the determination of tumour clonality. Br. J. Haematol., 74: 371-372, 1990. 8. Gale, R. E.. Wheadon, H.. and Linch, D. C. Assessment of X-chromosome inactivation patterns using the hypervariable probe M27 beta in normal hemopoietic cells and acute myeloid leukemic blasts. Leukemia (Baltimore). 6: 649-55. 1992. 9. Hodges. E., Howell, W. M., Boyd, Y.. and Smith. J. L. Variable X-chromosome DNA methylation patterns detected with probe M27 beta in a scries of Ivmphoid and myeloid malignancies. Br. J. Haematol., 77: 315-322. 1991. 10. Gilliland. D. G.. Blanchard. K. L.. Levy, J.. Perrin. S.. and Bunn, H. F. Clonality in myeloproliferative disorders: analysis by means of the polymerase chain reaction. Proc. Nati. Acad. Sci. USA, 88: 6848-6852, 1991. 11. van Kamp. H.. Jansen. R.. Willemze. R., Fibhe. W. E., and Landcgcnt. J. E. Studies on clonality by PCR analysis of the PGK-1 gene. Nucleic Acids Res.. 19: 2794, 199]. 12. Allen, R. C., Zoghbi, H. Y.. Moseley, A. B., Rosenblatt. H. M.. and Belmont, J. W. Methylation of Hpu\\ and llhul sites near the polymorphic CAG repeat in the human androgen-receptor gene correlates with X chromosome inactivation. Am. J. Hum. Genet., 51: 1229-1239. 1992. 13. Mashal. R. D.. Lester, S. C., and Sklar, J. Clonal analysis by study of X chromosome inactivation in formalin-fixed paraffin-embedded tissue. Cancer Res., 53: 4676-4679, 1993. 14. Willman, C. L, Busque. L., Griffith, B. B., Favara, B. E.. McCIain. K. L., Duncan, M. H., and Gilliland. D. G. Langerhans'-ccll histiocytosis (histiocytosis X)—a clonal
only at discrete portions in four of seven adenomas, whereas c-K-ra.v mutations were detected throughout the tumor specimens in three of seven adenomas and all seven adenocarcinomas. Interestingly, a similar frequency of polyclonality in meningiomas has been observed recently by Wu et al.5 They found that 6 of 17 (35%) of
proliferative disease. N. Engl. J. Med., 331: 154-160. 1994. 15. Sakai. T., Toguchida. J.. Ohlani. N.. Yandell. D. W.. Rapaport. J. M., and Dryja, T. P. Allelc-specific hypermethylation of the relinohlastoma tumor-suppressor gene. Am. J. Hum. Genet.. 48: 880-888, 1991. 16. Makos, M.. Nclkin, B. D., Chazin, V. R., Cavenee, W. K., Brodeur, G. M„and Baylin. S. B. DNA hypermethylation is associated with 17p allelic loss in neural tumors. Cancer Res., 53: 2715-2718, 1993. 17. Goelz, S. E., Vogelstein, B.. Hamilton, S. R.. and Feinherg. A. P. Hypomelhylation of DNA from benign and malignant human colon neoplasms. Science (Washington DC), 22«:187-190, 1985. 18. Baylin. S. B.. Makos, M.. Wu. J. J., Yen, R. W., de Bustros, A., Vertino, P.. and Nelkin, B. D. Abnormal patterns of DNA methylation in human neoplasia: potential consequences for tumor progression. Cancer Cells, 3: 383-390. 1991. 19. Baylin. S. B. Abnormal regional hypermethylation in cancer cells. Aids Res. Hum. Retroviruses, 8: 811-820, 1992. 20. Fialkow, P. J., Jackson, C. E., Block, M. A., and Grecnawald, K. A. Multicellular origin of parathyroid "adenomas". N. Engl. J. Med., 297: 696-698, 1977.
informative meningioma tumors are polyclonal by the M27ßprobe, a methylation-based clonality assay at the DXS255 locus.5 In 1976, Nowell (49) proposed a model of clonal evolution of tumor cell populations from precancerous cells that acquired one or more genetic changes and exhibited growth advantage over adjacent cells. This model suggests that in the early stages of tumorigenesis, such as benign tumors, the cell mass may be heterogeneous. Tumor hetero geneity within a single neoplasm is well recognized (50, 51) and has been described in many types of tumors, including human gliomas (52), where 3 to 21 subpopulations were observed karyotypically in each of 8 glioma tissue specimens studied. Our clonality results demonstrate that some meningiomas contain mostly clonally ex panded tumor cells that are monoclonal, whereas others may be polyclonal, consisting of reactive cells by hormonal stimulation or several distinct clonal populations of tumor cells. In summary, we have shown that methylation- and transcriptionbased clonality assays are often but not always concordant. Our data also demonstrate that most meningiomas are monoclonal, but some are apparently mixed or contain polyclonal cells, suggesting that heterogeneity or reactive cells due to hormonal stimulations may be present in a high proportion in tumors, which may represent early stages of clonal revolution during tumorigenesis.
ACKNOWLEDGMENTS We thank Drs. Steve Leon and RoñaCarroll for stimulating discussions and Dr. Louis Kunkel for permission to use the oligonucleotide synthesizer, Phosphorlmager, and for helpful advice. We are indebted to Nancy Olsen, RN for aiding in the collection of hlood specimens. 5J. K. Wu, M. MacGillavry, communication.
C. Kcssaris. B. Verheul, and B. T. Darras, personal
21. Arnold, A., Staunton, C. E., Kim. H. G., Gaz, R. D., and Kronenberg, H. M. Monoclonality and abnormal parathyroid hormone genes in parathyroid adenomas. N. Engl. J. Med., 318: 658-662, 1988. 22. Noguchi. S., Motomura. K.. Inaji, H.. Imaoka. S., and Koyama, H. Clonal analysis of parathyroid adenomas by means of the polymerase chain reaction. C'ancer Lett., 78: 93-97. 1994. 23. Busque. L, Zhu. J. G.. Dehart. D., Griffith. B.. Willman, C., Carroll. R., Black, P. M.. and Gilliland. D. G. An expression based clonality assay at the human androgen receptor locus (HUMARA) on chromosome X. Nucleic Acids Res., 22: 697-698. 1994. Luhahn. D.. Joseph. D.. Sullivan. P.. Willard, H., French. F.. and Wilson, E. Cloning of human androgen receptor complementary DNA and localization to the X chromo some. Science (Washington DC), 240: 327-330. 1988. Black, P. M. Brain tumors. N. Engl. J. Med., 324: 1555-1564, 1991. Black. P. M. Brain tumors. N. Engl. J. Med.. 324: 1471-1476. 1991. Jacoby, L. B., Pulaski. K.. Rouleau, G. A., and Martuza. R. L. Clonal analysis of human meningiomas and schwannomas. Cancer Res.. 50: 6783-6786. 1990. Seizinger. B. R., de la Monte, S., Atkins. L., Gusella, J.. and Martuza, R. L. Molecular genetic approach to human meningioma: loss of genes on chromosome 22. Proc. Nati. Acad. Sci. USA, 84: 5419-5423, 1987. Dumanski. J. P., Rouleau. G. A.. Nordenskjöld. M.. and Collins. V. P. Molecular genetic analysis of chromosome 22 in 81 cases of meningioma. Cancer Res., 50: 5863-5867, 1990. Trofattcr. J. A.. MacCollin. M. M.. Rutter. J. L. Mimeli. J. R., Duyao, M. P., Parry, D. M., Eldridge, R., Kley. N., Menon, A. G., Pulaski, K., Haase, V. H.. Ambrose, C. M., Munroe, D.. Bove, C., Haines, J. L.. Martuza. R. I... MacDonald. M. E., Seizinger. B. R., Short, M. P.. Buckler. A. J.. and Gusella, J. F. A novel moesin-, ezrin-, radixin-like gene is a candidate for the neurofibromatosis 2 tumor suppressor. Cell, 72: 791-8(XI. 1993. 31. Ruttledge, M. H., Sarrazin, J., Rangaratnam, S.. Phclan, C. M., Twist, E., Merei, P., Delatlrc. O.. Thomas. G.. Nordenskjöld. M.. Collins. V. P.. Dumanski. J. P., and
3871
Downloaded from cancerres.aacrjournals.org on February 13, 2019. © 1995 American Association for Cancer Research.
CLONALITY
ANALYSIS
Rouleau, G. A. Evidence for the complete inactivation of the n/2 gene in the majority of sporadic meningiomas. Nat. Genet., 6.- 180-184, 1994. 32. Schonberg, B. S., Christine, B. W., and Wisnant, J. P. Nervous system neoplasms and primary malignancies of other sites: the unique association between meningiomas and breast cancer. Neurology, 25: 705-712, 1975. 33. Mehta, D., Khatib, R., and Palei, S. Carcinoma of the breast and meningioma: association and management. Cancer (Phila.), 51: 1937-1940, 1983. 34. Carroll, R. S., Glowacka, D., Dashner, K., and Black, P. M. Progesterone receptor expression in meningiomas. Cancer Res., 53: 1312-1316, 1993. 35. Carroll, R. S., Zhang, J. P., Dashner, K., Sar, M., Wilson, E. M., and Black, P. M. Androgen receptor expression in meningiomas. J. Neurosurg., 82: 453-460, 1994. 36. Seizinger, B. R., Martuza, R. L., and Gusella, J. F. Loss of genes on chromosome 22 in tumorigenesis of human acoustic neuroma. Nature (Lond.), 322: 644-647, 1986. 37. Chirgwin, J. M., Przbyla, A. E., MacDonald, R. J., and Rutter, W. J. Isolation of biologically active ribonucleic acid from sources enriched with ribonuclease. Bio chemistry, J8: 5294-5299, 1979. 38. Zhu, J., Leon, S. P., Beggs, A. H., Busque, L., Gilliland, D. G., and Black, P. M. Human pituitary adenomas show no loss of heterozygosity at the retinoblastoma gene locus. J. Clin. Endocrinol. Metab., 78: 922-927, 1994. 39. Sun, Y., Hegamyer, G., and Colburn, N. H. A simple method using PCR for direct sequencing of genomic DNA from frozen tumor tissue embedded in optimal cutting temperature compound. Biotechniques, 12: 639-640, 1992. 40. Jimenez-Hakim, E., El-Azouzi, M., and Black, P. The effect of prolactin and bombesin on the growth of meningioma-derived cells in monolayer culture. J. Neuro.Oncol., 16: 185-190, 1993. 41. Buetow, K. H., Duggan, D., Yang, B., Ludwigsen, S., Puck, J., Porter, J., Budarf, M., Spielman, R., and Emanuel, B. S. A microsatellite-based multipoint index map of human chromosome 22. Genomics, IS: 329-339, 1993.
OF MENINGIOMAS
42. Lyon, M. F. Gene action in the X-chromosome of the mouse. Nature (Lond.), 190: 372-373, 1961. 43. Gale, R. E., Wheadon, H., and Linch, D. C. X-chromosome inactivation patterns using HPRT and PGK polymorphisms in haematologically normal and post-chemo therapy females. Br. J. Haematol., 79: 193-197, 1991. 44. Noguchi, S., Motomura, K., Inaji, H., Imaoka, S., and Koyama, H. Clonal analysis of human breast cancer by means of the polymerase chain reaction. Cancer Res., 52: 6594-6597, 1992. 45. Busque, L., and Gilliland, D. G. Clonal evolution in acute myeloid leukemia. Blood, 82: 337-342, 1993. 46. Ruttledge, M. H., Xie, Y. G., Han, F. Y., Peyrard, M., Collins, V. P., Nordenskjöld, M., and Dumanski, J. P. Deletions on chromosome 22 in sporadic meningioma. Genes Chromosomes Cancer, 10: 122-130, 1994. 47. Aeschimann, S., Kopp, P. A., Kimura, E. T., Zbaeren, ]., Tobler, A., Fey, M. F., and Studer, H. Morphological and functional polymorphism within clonal thyroid nod ules. J. Clin. Endocrinol. Metab., 77: 846-851, 1993. 48. Shibata, D., Schaeffer, J., Li, Z. H., Capella, G., and Perucho, M. Genetic heteroge neity of the c-K-ras locus in colorectal adenomas but not in adenocarcinomas. J. Nati. Cancer Inst., 85: 1058-1063, 1993. 49. Nowell, P. C. The clonal evolution of tumor cell population. Science (Washington DC), 194: 23-28, 1976. 50. Heppner, G. H. Tumor heterogeneity. Cancer Res., 44: 2259-2265, 1984. 51. Heppner, G. H. Cancer cell societies and tumor progression. Stem Cells, //: 199203, 1993. 52. Shapiro, J. R., Yung, W. K. A., and Shapiro, W. R. Isolation, karyotype, and clonal growth of heterogeneous subpopulations of human malignant gliomas. Cancer Res., 41: 2349-2359, 1981.
3872
Downloaded from cancerres.aacrjournals.org on February 13, 2019. © 1995 American Association for Cancer Research.
Analysis of Meningiomas by Methylation- and Transcription-based Clonality Assays Jiguang Zhu, Matthew P. Frosch, Lambert Busque, et al. Cancer Res 1995;55:3865-3872.
Updated version
E-mail alerts Reprints and Subscriptions Permissions
Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/55/17/3865
Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at
[email protected]. To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/55/17/3865. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.
Downloaded from cancerres.aacrjournals.org on February 13, 2019. © 1995 American Association for Cancer Research.