- Email: [email protected]
Independent clonal origin of multiple uterine leiomyomas that was determined by X chromosome inactivation and microsatellite analysis
American Journal of Obstetrics and Gynecology (2005) 193, 1395–403
www.ajog.org
Independent clonal origin of multiple uterine leiomyomas that was determined by X chromosome inactivation and microsatellite analysis Renata A. Canevari, PhD,a Anaglo´ria Pontes, MD, PhD,b Fabı´ola E. Rosa, MSc,a Cla´udia A. Rainho, PhD,a Silvia R. Rogatto, PhDc,* Department of Genetics, Institute of Biosciences,a Department of Obstetrics and Gynecology, Faculty of Medicine,b Department of Urology, Faculty of Medicine,c UNESP–University of Sa˜o Paulo State, Botucatu, Sa˜o Paulo, Brazil Received for publication October 21, 2004; revised January 27, 2005; accepted February 15, 2005
KEY WORDS Loss of heterozygosity X chromosome inactivation Clonality Uterine leiomyoma
Objective: In an attempt to clarify the clonality and genetic relationships that are involved in the tumorigenesis of uterine leiomyomas, we used a total of 43 multiple leiomyomas from 14 patients and analyzed the allelic status with 15 microsatellite markers and X chromosome inactivation analysis. Study design: We have used a set of 15 microsatellite polymorphism markers mapped on 3q, 7p, 11, and 15q by automated analysis. The X chromosome inactivation was evaluated by the methylation status of the X-linked androgen receptor gene. Results: Loss of heterozygosity analysis showed a different pattern in 7 of the 8 cases with allelic loss for at least 1 of 15 microsatellite markers that were analyzed. A similar loss of heterozygosity findings at 7p22-15 was detected in 3 samples from the same patient. X chromosome inactivation analysis demonstrated the same inactivated allele in all tumors of the 9 of 12 informative patients; different inactivation patterns were observed in 3 cases. Conclusion: Our data support the concept that uterine leiomyomas are derived from a single cell but are generated independently in the uterus. Loss of heterozygosity findings at 7p22-15 are consistent with previous data that suggested the relevance of chromosomal aberrations at 7p that were involved in individual uterine leiomyomas. Ó 2005 Mosby, Inc. All rights reserved.
Supported by the Fundac¸a˜o de Amparo a` Pesquisa do Estado de Sa˜o Paulo (FAPESP) and CNPq (Conselho Nacional de Pesquisa), Brazil. * Reprint requests: Silvia Regina Rogatto, PhD, NeoGene Laboratory, Departamento de Urologia, Faculdade de Medicina - UNESP, Botucatu, Sa˜o Paulo, Brazil, CEP: 18618-000. E-mail: [email protected] 0002-9378/$ - see front matter Ó 2005 Mosby, Inc. All rights reserved. doi:10.1016/j.ajog.2005.02.097
Uterine leiomyomas are the most frequently encountered neoplasm of the uterus, with an estimated prevalence of 20% to 30% in women who are O30 years old. On the basis of a study of serial sections of uteri, it has been estimated that up to 77% of reproductive-aged women have uterine leiomyomas.1 Although leiomyomas can be asymptomatic, approximately 20% to 50% are symptomatic and lead patients to seek therapy. The symptoms that are associated with
1396 uterine leiomyomas vary considerably and include abnormal uterine bleeding, pelvic pain or pressure, urinary dysfunction, spontaneous abortion, premature delivery, and infertility.2 These tumors are the most frequent indication for hysterectomy.3,4 The incidence and severity of symptoms is related directly to the number, size, and location of the tumors.4 Approximately two-thirds of women have multiple myomas of various sizes.5 These leiomyomas can occur in any part of the uterus and may be intramural, submucosal, or subserosal. The pathogenesis and evolution of uterine leiomyomas are not well understood. Some authors have suggested that residual undifferentiated cells in the myometrium are the progenitor cells of leiomyomas.6 Other authors have assumed that acquired chromosomal aberrations play a pathogenetic role in uterine leiomyomas.7-9 In addition to gene identification in chromosomally abnormal tumors, the clonality of uterine leiomyomas has been studied to understand the genetic mechanisms that are involved in the genesis and growth of these tumors. Most human tumors are believed to be cytogenetically abnormal clones of cells that are derived from a single progenitor cell that has undergone a mutation. Although an initial mutation may trigger the formation of the tumor, additional mutations may occur during tumor progression, thereby creating subclones with a selective growth advantage and the potential for clonal expansion. If leiomyomas arise from somatic mutations in a myometrial cell that result in the dysregulation of the mechanisms that control cell growth and confer a growth advantage, one would expect to observe clonal expansion. A frequently used method for the evaluation of clonality is based on X-chromosome inactivation. Clonality of tissue samples can be detected by digestion with a methylation-sensitive enzyme and amplification of the remaining allele by polymerase chain reaction (PCR) at a polymorphic X chromosome-linked marker. Cytogenetic and X chromosome inactivation studies strongly suggest that leiomyomas are clonal.8,10,11 As an alternative approach, loss of heterozygosity (LOH) analysis with multiple microsatellite markers has been used to study clonality in bilateral ovarian carcinomas and bilateral breast cancer.12,13 LOH analysis at multiple marker loci can detect a specific allelic deletion pattern. Identical patterns suggest a monoclonal origin of the tumors, whereas polyclonal or tumors with independent clonal origin reveal a completely different pattern with a lack of common alterations. By an investigation of chromosomal regions that were deleted very early in carcinogenesis, an identical LOH pattern is a strong indicator for monoclonality of multiple lesions. However, a lack of common deletion between the tumors that are compared could be due to the omission of the right marker locus in LOH analysis.14 In the present study, LOH analysis and the
Canevari et al pattern of X chromosome inactivation, which is identified by the differential methylation of a site close to the polymorphic CAG repeat in the androgen receptor (AR) gene, were used to determine the clonal status of 43 uterine leiomyomas from 14 patients. To date, this is the first report to analyze the clonality of uterine leiomyomas with the use of both LOH and X chromosome inactivation status.
Material and methods Tumor samples and DNA preparation In 64 patients with uterine leiomyomas, only 14 presented multiple uterine leiomyomas and were analyzed by microsatellite analysis and X chromosome inactivation. The uterine leiomyomas and corresponding peripheral blood leukocytes and normal myometrium from 1 patient were collected in the Department of Obstetrics and Gynecology, Faculty of Medicine, UNESP, Sa˜o Paulo, Brazil. The tissues were obtained from symptomatic women who underwent abdominal hysterectomy or myomectomy for medically indicated reasons. Patients for whom there were available samples of the multiple tumors were selected to do this investigation. Tissue donors were premenopausal, had regular menstrual cycles, and had received no exogenous hormones or hormone suppression therapy before surgery. Women with endometriosis, adenomyosis, or other diseases of the reproductive system were excluded from the study. Connective tissue and necrotic tissue were removed from specimens. Informed consent was obtained from all of the patients before sampling was performed; this study was approved by the Brazilian Ethics Committee (CONEP 334/2001). Genomic DNA from uterine tissues and corresponding blood leukocytes was prepared by standard sodium dodecyl sulfate/proteinase K digestion, followed by phenol and chloroform extraction and ethanol precipitation. DNA was stored at 20(C until amplification by PCR. The medical records of all of the patients were examined to retrieve clinical and histopathologic data. All of the tumors were diagnosed histopathologically to be usual uterine leiomyomas. The clinical and personal data of the patients are shown in Table I. The family history of cancer considered first- and second-degree relatives with cancer to be informative. Where possible, all assessments were based on documented medical records as evidence of cancer. For some families, a history of cancer was ascertained by a cancer geneticist who interviewed the patients and their relatives.
PCR and microsatellite LOH analysis Matched pairs of normal and tumor tissue were PCR amplified at 15 microsatellite loci distributed on 3q263q29, 7p22-7p15, 11p15/q23-25 and 15q26. The primer
Canevari et al
1397
Table I
Clinical and histopathologic data of the 14 patients with multiple uterine leiomyomas
Patient
Age (y)
Location
Familial/personal history
29 42 40 59
Intramural, submucosal Intramural Intramural Intramural, subserosal
40 50 49 43 46 39 49 51 Not determined 48
Subserosal, submucosal Intramural Intramural Intramural Intramural Intramural (one of them giant) Intramural, subserosal, and submucosal Submucosal Intramural Intramural
Not determined/not determined Three sisters with uterine leiomyoma; 1 sister with melanoma/-/Twin sister with uterine leiomyoma; 2 sisters with endometrial carcinoma/skin carcinoma Three sisters with uterine leiomyoma; sister melanoma/Not determined/A sister with uterine leiomyoma/-/Mother with uterus carcinoma/Brother with leukemia/Sister with uterine leiomyoma/Mother with kidney carcinoma/-/-/-
1 2 3 4 5 6 7 8 9 10 11 12 13 14
sequences and cytogenetic maps for the microsatellite markers were retrieved from the Genome Database (http://www.gdb.org) and from The Wellcome Trust Sanger Institute (http://www.ensembl.org), respectively. All of these markers were dinucleotide repeats. Forward primers were synthesized with a fluorescent tag (FAM, NED, or HEX) on the 50 end. Fluorescent PCR amplification was performed in a thermal cycle (Peltier Thermal Cycler, PTC-200; MJ Research, Waltham, MA) in a 15 mL reaction volume that contained 100 ng of template DNA, 10 mmol/L Tris-HCl, pH 8.3, 50 mmol/L KCl, 2.5 mmol/L MgCl2, 200 mmol/L of each dNTP, 0.5 mmol/L of each primer and 1 unit of AmpliTaq Gold DNA polymerase (Applied Biosytems, Foster City, CA). The PCR conditions consisted of an initial denaturation at 95(C for 10 minutes, with subsequent amplification for 30 cycles of 45 seconds at 94(C, 45 seconds at 58( to 60(C, and 1 minute at 72(C. All of the PCR reactions were performed in duplicate to exclude technical errors and to confirm the findings. The fluorescent PCR products were analyzed on 5% polyacrylamide-urea (Long Ranger Singel Packs/FMC, Rockland, ME) denaturing gels with the use of a DNA automated sequencer (ABI Prism 377; Applied Bioystems). The fluorescent gel data were collected and stored with data collection software and then analyzed with Genescan and Genotyper softwares (Applied Biosystems), which quantified the size (in base pairs), height, and area of the allelic peak. An imbalance factor was defined, with both peak areas and heights of the two alleles in the normal and tumor samples as parameters. The formula that was used for the calculation was T1:T2/N1:N2, where T1 and N1 were the peak areas for the shorter length allele products of the tumor and normal samples, respectively, and T2 and N2 were the peak areas for the longer length
allele products of the tumor and normal samples, respectively. For ratios of O1.00, the ratio was converted to 1/AR to give a value between 0.00 and 1.00. A ratio of %0.60 was considered to be indicative of LOH. Comparisons of the ratios with the peak height as the parameter was done with the equation N2:N1/T2:T1, where N2 and T2 were the peak heights of the longer length allele product for the normal and tumor samples, respectively, and N1 and T1 were the peak heights of the shorter length allele product for the normal and tumor samples, respectively. A probability value of %.60 was considered to be indicative of the loss of the shorter length allele, and a probability value R1.60 was indicative of the loss of the longer length allele. Allelic losses were assigned only after the results were found consistently in all 3 experiments that were considering both peak areas and heights.
Analysis of X chromosome inactivation at the AR locus and clonality Clonality was assessed on the basis of the methylation status of the AR gene. This gene is located on Xq11-q12 and contains a region with a CAG polymorphic repeat. The methylation status of 2 HpaII sites located in the CAG trinucleotide repeat AR locus has been shown to correlate with X-chromosome inactivation. To determine the inactivation status of the AR alleles, genomic DNA was digested with the restriction endonuclease HpaII, a methylation-sensitive enzyme that cleaves unmethylated DNA (active X),15 followed by PCR amplification of the CAG polymorphic region. The inactive allele is resistant to digestion with this endonuclease and will amplify a detectable product after PCR. To examine methylation of the AR gene, each DNA sample was digested in a 15 mL reaction volume with 10 U of restriction enzyme RsaI
1398
Canevari et al
Table II Results of microallelotyping analysis with the use of 15 microsatellite markers and clonality for X chromosome inactivation in 14 patients with multiple uterine leiomyomas Case
Tumor
D3S1614
D3S1565
D3S1601
D3S1311
D7S517
D7S513
D7S507
D7S493
1
TA TB TC TD
-
-
-
-
-
-
-
-
2
TA TB TC
-
-
-
-
-
-
3
TA TB TC
-
-
-
-
: : :
: : :
: : :
: : :
4
TA TB
-
-
-
-
-
:
5
TA TB
-
-
-
-
:
-
-
6
TA TB TC TD TE TF
-
-
-
-
-
-
-
-
7
TA TB TC TD
-
-
-
-
-
8
TA TB TC TA TB
-
-
-
-
-
10
TA TB TC
-
-
: -
-
-
-
-
11
TA TB
-
:
-
-
-
-
12
TA TB TC
ND ND ND
-
:
-
-
-
-
13
TA TB TC
-
-
-
-
-
-
14
TA TB TC
-
-
-
-
-
-
-
9
-, Allelic retention;
-
, not informative; :, LOH; ND, not determined.
alone (10 U/ml) and 10 U of RsaI and 60 U of HpaII (10 U/ mL). After 16 hours at 37(C, the samples were incubated at 70(C for 10 minutes to heat-inactivate the enzymes. Genomic DNA was amplified with fluorescent labeled primers, according to Bharaj et al.16 The forward and reverse primer sequences that were used to amplify
the AR gene were 50 -TCCAGAATCTGTTCCAGAGCGTGC-30 and 50 -GCTGTGAAGGTTGCTGTTCCTCAT-30 . One hundred nanograms of digested DNA was amplified in a reaction mixture that contained 125 mmol/L of each deoxynucleotide triphosphate, 0.2 mmol/L of each primer, 10 mmol/L Tris-HCl, pH 8.3, 50 mmol/L
Canevari et al Table II
1399
(Continued) Different LOH/ informative loci(%)
D11S1338
D11S902
D11S925
D11S968
D15S127
D15S120
D15S87
AR allele inactive
-
-
-
-
-
-
0/14 (0)
-
-
-
0/10 (0)
2 2 1
-
-
-
-
-
-
0/14 (0)
1 ND 1
ND
:
-
-
2/10 (20.0)
2 2
-
-
-
-
: :
1/10 (10.0)
1 1
-
-
-
0/12 (0)
2 2 1 2 1 2 1 1 1 1
-
-
-
-
-
-
-
-
0/12 (0)
-
-
: -
-
-
: : -
(16.7)
-
-
-
-
-
(9.0)
1 1
ND -
ND -
ND -
ND -
-
-
-
0/13 (0)
1 1 1
:
-
-
: :
2/10 (20.0)
2 2
-
-
-
1/9 (11.1)
1 1 2
-
-
-
-
-
: : -
1/12 (8.3)
-
-
-
-
-
-
-
0/14 (0)
1 ND ND
-, Allelic retention;
2/12
1/11
2 2 2
, not informative; :, LOH; ND, not determined.
KCl, 1.5 mmol/L MgCl2, and 1 U of Taq DNA polymerase (Invitrogen, Carlsbad, CA). After initial denaturation at 94(C for 10 minutes, PCR was performed with a thermal profile of 30 seconds at 95(C, 1 minute at 65(C, and 5 minutes at 72(C, for a total of 35 cycles. The PCR products were separated by electrophoresis on 5% poly-
acrylamide denaturing gels, and the size of the amplified product was determined in an ABI Prism 377 DNA automatic sequencer, as described earlier. All of the PCR reactions were performed in duplicate to exclude technical errors and to confirm the size of the CAG repeats of each AR allele.
1400 For informative heterozygous cases, clonality was assessed as described in Lucas et al17 and Siu et al.18 Each digested DNA sample was read separately, and the intensity and area of each peak (allele) were measured. LOH analysis was done as described earlier. Only cases without LOH at the AR locus were included in this investigation. The allele amplification ratio for each sample before digestion (RU) with HpaII was calculated by dividing the area of the longer allele peak (allele 2) by that of the shorter allele peak (allele 1). The RU had to be O0.60 to confirm equal amplification of both alleles. Similarly, the RH (allele amplification ratio after digestion with HpaII) was calculated by obtaining the ratio of the longer and shorter allele peaks after digestion. These ratios were divided such that the longer ratio was the denominator: if R was !0.40, the sample was considered to be monoclonal, but if R was O0.40, then the sample was considered to be polyclonal.17
Results Clonality on the basis of the LOH pattern Intratumor heterogeneity of the LOH pattern was assessed with the use of 15 microsatellite markers in 43 samples from 14 patients with uterine leiomyomas. The details of the LOH analysis with the markers that were used are given in Table II, and representative examples of LOH are shown in the Figure, A. In 8 cases, LOH was demonstrated for at least 1 microsatellite marker and at least 1 of the tumors. An identical pattern of LOH was observed in 1 unique patient (case 3) in the microsatellite markers that were mapped on chromosome 7p. Seven cases (cases 4, 5, 8, 9, 11, 12, and 13) showed differential patterns of allelic loss in multiple tumors from the same patient (Figure, A). The percentage of different LOH alleles (number of different LOH loci/number of informative loci ! 100) in 7 cases with different LOH patterns ranged from 8.3% to 20%.
Clonal analysis Clonality was assessed by the determination of the pattern of X chromosome inactivation for the AR gene in all uterine leiomyomas from 14 patients; 7 leiomyomas from 2 patients showed homozygous AR alleles (not informative). The results for 36 uterine leiomyomas from 12 informative cases are summarized in Table II. Each leiomyoma sample that was analyzed consisted of only a single type of inactive allele, which is consistent with a monoclonal neoplastic population of cells. In 3 patients (cases 2, 6, and 12), different X chromosomes were inactivated in multiple tumors, which confirmed the independent clonal origin of each tumor. An analysis of X chromosome inactivation in patient 12 revealed the presence of independent clones, with 1 allele inactivation
Canevari et al in 2 tumors (12A and 12B) and 2 allele inactivation in 1 tumor (12C). In the same case, LOH analysis revealed allelic loss at the D3S1311 marker in tumor 12C but not in the other tumors (12A and 12B; Table II). Although the analysis of X chromosome inactivation in the AR gene revealed independent clones in patients 2 and 6, no allelic losses were found in the autosomal microsatellite markers that were used here. The myometrium (normal control) of patient 6 showed both AR alleles, which is indicative of a mosaic status, whereas each tumor had a monoclonal origin with different AR alleles inactivated, which suggests that they had originated independently (Figure, B). Nine patients who were heterozygous for the AR gene polymorphism showed the same pattern of inactivation; different patterns of allelic loss in the autosomal polymorphic markers of the tumors were analyzed in 5 of these patients (cases 4, 5, 8, 9, and 11). Three of the 36 uterine leiomyomas (3TB, 14TB, and 14TC) were eliminated from the study because the PCR amplification products for the 2 alleles were not approximately equal before digestion with HpaII, nor was there any detectable amplification after digestion with HpaII.
Comment Several leiomyomas often occur in the same uterus, which suggests that there may be an inherited genetic abnormality that predisposes some women to the development of these tumors.19 In the present study, 5 of 13 patients had a familial history of uterine leiomyomas in first-degree relatives. Epidemiologic data suggest that there is a familial risk of the development of uterine leiomyomas in first-degree relatives of affected individuals and that there is a heritable predisposition.20 Vikhlyaeva et al20 observed that uterine leiomyomas were diagnosed 2.2-fold more frequently in families with R2 affected individuals. However, the nature of this familial predisposition is ill defined. An inherited factor in the cause of leiomyoma was also suggested by family studies of Reed’s syndrome.21 The genetic locus for this syndrome was mapped to chromosome 1q42.3-43.22,23 It was shown that germline mutations of fumarate hydratase (FH gene) are responsible for the inherited syndromes of multiple cutaneous and uterine leiomyomas (MCUL, OMIM 150800) and of hereditary leiomyomatosis and renal cell cancer (HLRCC, OMIM 605839).24 These syndromes are inherited as autosomal dominant disorders with reduced penetrance and are associated with cutaneous leiomyomas and, in affected female patients, with uterine leiomyomas. Recently, Gross et al25 showed evidence that was suggestive of linkage to the FH region in a subset of nonsyndromic patients with uterine leiomyomas. We have no patients in our
Canevari et al
1401
Figure A, Allelotypes of a normal tumor sample in which allele loss was detected in one of the uterine leiomyomas of the same patient for the marker D7S493 (case 4) and for the marker D7S513 (case 5). The arrows indicate an LOH. B, Electropherograms show different patterns of X chromosome inactivation in case 6. The arrows indicate the AR allele digested with HpaII (activated allele).
1402 study with MCUL and HLRCC syndromes, as determined by clinical history. Another explanation for the existence of multiple leiomyomas is that an existing tumor may spread to other regions of the uterus. In this case, all of the tumors would be related clonally. However, several studies of X chromosome inactivation in multiple leiomyomas in the same uterus have indicated that the various tumors are not related clonally but rather initiate independently of each other and progress by discrete and separate somatic mutations.10,11,26-28 The independent clonal origin of multiple leiomyomas was determined initially by glucose-6-phosphate dehydrogenase (G6PD) isoenzyme analysis.10 The latter study showed that either the A or B type allele, but not both, was present in each tumor from a series of 7 women who were found to be heterozygous for this enzyme. Furthermore, both A and B type tumors occurred in the same individual. This observation indicated a random pattern of inactivation among multiple tumors, with individual tumors exclusively expressing one or the other G6PD type and also revealed that multiple tumors in the same uterus most likely arise in situ from different myometrial cells and not from a single mutated myometrial cell. However, a low degree of G6PD isoenzyme polymorphism among most female patients across various ethnic groups limited the use of this approach. Subsequent studies that involved CAG trinucleotide repeats in the X-linked AR gene confirmed the independent origin of uterine leiomyomas.11,27 These studies involved the digestion of DNA with methylationsensitive restriction enzymes to assess the differences in methylation between active and inactive X chromosomes at the AR locus in uterine leiomyomas. The cytogenetic studies of multiple uterine leiomyomas from the same uterus have also suggested that each tumor arises independently; hence, these tumors often exhibit different chromosomal alterations.9,29-31 In our study, 7 of 8 cases with allelic loss showed a different pattern of LOH between multiple leiomyomas. These data suggest that uterine leiomyomas can be derived from entirely independent clones or, alternatively, that the abnormality is not primary and could have occurred as a result of clonal evolution. The independent monoclonal origin of these tumors was confirmed here by the X chromosome inactivation analysis that revealed alternate patterns of inactivation in 3 of the 12 informative cases that were studied. The use of X chromosome inactivation to analyze clonality is based on the occurrence of random X chromosome inactivation in early embryogenesis. Because, for a specific progenitor, all daughter cells will inherit the same inactivated X chromosome, this can be used as a marker of clonality.32 Although the initial determination of which of the X chromosomes is inactivated is made at
Canevari et al random, the pattern of inactivation is permanent and is constant throughout subsequent cell divisions.33 Thus, normal tissues of adult women consist of a mosaic of cell types, which differ in whether the maternally or paternally derived X chromosome has been inactivated. In contrast, clonal cellular proliferation maintains the same pattern of X chromosome inactivation that was determined in the single cell of origin.34 However, the reliability of X-chromosome inactivation analysis for clonality analysis in tumors has been challenged.35 One problem is that tumors may show altered DNA methylation patterns.36 As an alternative, LOH analysis with multiple microsatellite markers has been used to study clonality. In this study, the data for 9 of 12 informative patients showed the same pattern of inactivation in different tumors. Because there is a 50% chance of inactivating the same X chromosome in genetically unrelated tumors from different clones, no conclusions could be drawn from these 9 patients. However, 5 of these 9 cases showed different patterns of allelic loss, which suggested an independent evolution of these tumors. Our results agreed with those of cytogenetic and X chromosome inactivation studies that demonstrated the independent origin of multiple uterine leiomyomas. However, of the 8 cases in our study that showed LOH, 1 patient (case 3) had the same LOH pattern in all of the leiomyomas that were present in the uterus and also showed inactivation of the same X chromosome. It has been considered that some leiomyomas develop from a common precursor cell, which suggests that multiple tumors may develop by spreading from a single primary tumor. The discovery of identical cytogenetic alterations in multiple leiomyomas of the same patient favors this developmental pathway.8,37 Alternatively, these changes may represent recurrent chromosomal aberrations that are involved in the tumorigenesis of uterine leiomyomas. This hypothesis is supported by a report of a patient with 2 independent leiomyomas, each showing different patterns of X chromosome inactivation but identical del(7)(q21.2q31.2) derivative chromosomes.30 We recently reported the frequent involvement of allelic loss at 7p22-15 in uterine leiomyoma.38 In addition to the previous reports, the same allelic losses that were found at 7p in all 3 uterine leiomyomas from the same patient suggest their relevance in a subgroup of uterine leiomyomas. In conclusion, clonal analysis of multiple uterine leiomyomas that are based on the patterns of LOH and X chromosome inactivation indicated that the leiomyomas that were studied were unicellular in origin but are independently generated in the uterus. The chromosomal and molecular heterogeneity and the different patterns of X chromosome inactivation that were found in these tumors may provide more accurate information on the clonal origin of tumors and may
Canevari et al explain clinicopathologic differences that are seen in uterine leiomyomas.
Acknowledgment We thank Rodrigo Mattos dos Santos for expert technical assistance.
References 1. Cramer S, Patel A. The frequency of uterine leiomyomas. Am J Clin Pathol 1990;94:435-8. 2. Quade BJ. Pathology, cytogenetics and molecular biology of uterine leiomyomas and other smooth muscle lesions. Curr Opin Obstet Gynecol 1995;7:35-42. 3. Ligon AH, Morton CC. Genetics of uterine leiomyomata. Genes Chromosomes Cancer 2000;28:235-45. 4. Buttram VC, Reiter RC. Uterine leiomyomata: etiology, symptomatology and management. Fertil Steril 1981;36:433-45. 5. Zaloudek CJ, Norris HJ. Mesenchymal tumors of the uterus. In: Kurman RJ, editor. Blausteins’s pathology of the female genital tract. 4th ed. New York: Springer-Verlag; 1994. p. 487-94. 6. Hoang Ngoc M, Smadja A, Herve de Sigalony JP. A classification of uterine tumors based upon embryogenesis. Arch Anat Cytol Pathol 1991;39:94-102. 7. Mark J, Havel G, Grepp C, Dahlenfors R, Wedell B. Chromosomal patterns in human benign uterine leiomyomas. Cancer Genet Cytogenet 1990;44:1-13. 8. Nilbert M, Heim S. Uterine leiomyoma cytogenetics. Genes Chromosomes Cancer 1990;2:3-13. 9. Rein MS, Friedman AJ, Barbieri RL, Pavelky K, Fletcher JA, Morton CC. Cytogenetic abnormalities in uterine leiomyomata. Obstet Gynecol 1991;77:923-6. 10. Townsend DE, Sparkes RS, Baluda MC, McClelland G. Unicellular histogenesis of uterine leiomyomas as determined by electrophoresis of glucose-6-phosphate dehydrogenase. Am J Obstet Gynecol 1970;107:1168-73. 11. Mashal RD, Schoenberg ML, Friedman AJ, Nowak RA, Rein MS, Morton CC, et al. Analysis of androgen receptor DNA reveals the independent clonal origins of uterine leiomyomata and the secondary nature of cytogenetic aberration in the development of leiomyomata. Genes Chromosomes Cancer 1994;7:1-6. 12. Abeln ECA, Kuipers-Dijkshoorn NJ, Berns EMJJ, HenzenLogmans SC, Fleuren GJ, Cornelisse CJ. Molecular genetic evidence for unifocal origin of advanced epithelial ovarian cancer and for minor clonal divergency. Br J Cancer 1995;72:1330-6. 13. Imyanitov EN, Suspitsin EN, Grigoriev MY, Togo AV, Kuligina ES, Belogubova EV, et al. Concordance of allelic imbalance profiles in synchronous and metachronous bilateral breast carcinomas. Int J Cancer 2002;100:557-64. 14. Hafner C, Knuechel R, Stoehr R, Hartmann A. Clonality of multifocal urothelial carcinomas: 10 years of molecular genetic studies. Int J Cancer 2002;101:1-6. 15. Allen RC, Zoghbi HY, Moseley AB, Rosenblatt HM, Belmont JW. Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in the human androgen-receptor gene correlates with X-chromosome inactivation. Am J Hum Genet 1992;51:1229-39. 16. Bharaj BS, Vassilikis EJK, Diamandis EP. Rapid and accurate determination of (CAG) repeats in the androgen receptor gene using polymerase chain reaction and automated fragment analysis. Clin Biochem 1999;32:327-32. 17. Lucas DR, Shroyer K, McCarthy P, Marklam N, Fujita M, Enomoto T. Desmoid tumor is a clonal cellular proliferation: PCR amplification of HUMARA for analysis of patterns of Xchromosome inactivation. Am J Surg Pathol 1997;21:306-11.
1403 18. Siu IM, Robinson DR, Schwartz S, Kung HJ, Pretlow TG, Petersen RB, et al. The identification of monoclonality in human aberrant crypt foci. Cancer Res 1999;59:63-6. 19. Mao X, Barfoot R, Hamoudi RA, Easton DF, Flanagan AM, Stratton MR. Allelotype of uterine leiomyomas. Cancer Genet Cytogenet 1999;114:89-95. 20. Vikhlyaeva EM, Khodzhaeva ZS, Fantschenko ND. Familial predisposition to uterine leiomyomas. Int J Gynaecol Obstet 1995;51:127-31. 21. Reed WB, Walker R, Horowitz R. Cutaneous leiomyomata with uterine leiomyomata. Acta Derm Venereol 1973;53:409-16. 22. Alam NA, Bevan S, Churchman M, Barclay E, Barker K, Jaeger EE, et al. Localization of a gene (MCUL1) for multiple cutaneous leiomyomata and uterine fibroids to chromosome 1q42.3-q43. Am J Hum Genet 2001;68:1264-9. 23. Martinez-Mir A, Gordon D, Horev L, Klapholz L, Ott J, Christiano AM, et al. Multiple cutaneous and uterine leiomyomas: Refinement of the genetic locus for multiple cutaneous and uterine leiomyomas on chromosome 1q42.3-q43. J Invest Dermatol 2002;118:876-80. 24. Tomlinson IP, Alam NA, Rowan AJ, Barclay E, Jaeger EM, Kelsell D, et al. Germline mutations in FH predispose to dominantly inherited uterine fibroids, skin leiomyomata and papillary renal cell cancer. Nat Genet 2002;30:406-10. 25. Gross KL, Panhuysen CM, Kleinman MS, Goldhammer H, Jones ES, Nassery N, et al. Involvement of fumarate hydratase in nonsyndromic uterine leiomyomas: Genetic linkage analysis and FISH studies. Genes Chromosomes Cancer 2004;41:183-90. 26. Fialkow PJ. Use of genetic markers to study cellular origin of tumors in human females. Adv Cancer Res 1972;15:191-226. 27. Hashimoto K, Azuma C, Kamiura S, Kimura T, Nobunaga T, Kanai T, et al. Clonal determination of uterine leiomyomas by analyzing differential inactivation of the X-chromosome-linked phosphoglycerokinase gene. Gynecol Obstet Invest 1995;4093:204-8. 28. Bonatz G, Frahm SO, Andre´as S, Heidom K, Jonat W, Parwaresch R. Telomere shortening in uterine leiomyomas. Am J Obstet Gynecol 1998;179:591-6. 29. Nilbert M, Heim S, Mandahl N, Floderus UM, Willen H, Mitelman F. Different karyotypic abnormalities, t(1;6) and del(7), in two uterine leiomyomas from the same patient. Cancer Genet Cytogenet 1989;42:51-3. 30. Nilbert M, Stro¨mbeck B. Independent origin of uterine leiomyomas with karyotypically identical alterations. Gynecol Obstet Invest 1992;33:246-8. 31. Mantovani MS, Neto JB, Philbert PM, Casartelli C. Multiple uterine leiomyomas: cytogenetic analysis. Gynecol Oncol 1999;72:71-5. 32. Lyon MF. X-chromosome inactivation and developmental patterns in mammals. Biol Rev 1972;47:1-35. 33. Gartler SM, Riggs AD. Mammalian X-chromosome inactivation. Annu Rev Genet 1983;17:155-90. 34. Fialkow PJ. Clonal origin of human tumors. Biochim Biophys Acta 1976;458:283-321. 35. Buller RE, Sood AK, Lallas T, Buekers T, Skilling JS. Association between nonrandom X-chromosome inactivation and BRCA1 mutation in germline DNA of patients with ovarian cancer. J Natl Cancer Inst 1999;91:339-46. 36. Yang DH, Smith ER, Cohen C, Wu H, Patriotis C, Godwin AK, et al. Molecular events associated with dysplastic morphologic transformation and initiation of ovarian tumorigenicity. Cancer 2002;94:2380-92. 37. Nilbert M, Heim S, Mandahl N, Floderus UM, Willen H, Mitelman F. Characteristic chromosome abnormalities, including rearrangements of 6p, del(7q), C12, and t(12;14), in 44 uterine leiomyomas. Hum Genet 1990;85:605-11. 38. Canevari RA, Pontes A, Rogatto SR. Microallelotyping defines novel regions of loss of heterozygosity in uterine leiomyomas. Mol Carcinog 2005;42:177-82.
www.ajog.org
Independent clonal origin of multiple uterine leiomyomas that was determined by X chromosome inactivation and microsatellite analysis Renata A. Canevari, PhD,a Anaglo´ria Pontes, MD, PhD,b Fabı´ola E. Rosa, MSc,a Cla´udia A. Rainho, PhD,a Silvia R. Rogatto, PhDc,* Department of Genetics, Institute of Biosciences,a Department of Obstetrics and Gynecology, Faculty of Medicine,b Department of Urology, Faculty of Medicine,c UNESP–University of Sa˜o Paulo State, Botucatu, Sa˜o Paulo, Brazil Received for publication October 21, 2004; revised January 27, 2005; accepted February 15, 2005
KEY WORDS Loss of heterozygosity X chromosome inactivation Clonality Uterine leiomyoma
Objective: In an attempt to clarify the clonality and genetic relationships that are involved in the tumorigenesis of uterine leiomyomas, we used a total of 43 multiple leiomyomas from 14 patients and analyzed the allelic status with 15 microsatellite markers and X chromosome inactivation analysis. Study design: We have used a set of 15 microsatellite polymorphism markers mapped on 3q, 7p, 11, and 15q by automated analysis. The X chromosome inactivation was evaluated by the methylation status of the X-linked androgen receptor gene. Results: Loss of heterozygosity analysis showed a different pattern in 7 of the 8 cases with allelic loss for at least 1 of 15 microsatellite markers that were analyzed. A similar loss of heterozygosity findings at 7p22-15 was detected in 3 samples from the same patient. X chromosome inactivation analysis demonstrated the same inactivated allele in all tumors of the 9 of 12 informative patients; different inactivation patterns were observed in 3 cases. Conclusion: Our data support the concept that uterine leiomyomas are derived from a single cell but are generated independently in the uterus. Loss of heterozygosity findings at 7p22-15 are consistent with previous data that suggested the relevance of chromosomal aberrations at 7p that were involved in individual uterine leiomyomas. Ó 2005 Mosby, Inc. All rights reserved.
Supported by the Fundac¸a˜o de Amparo a` Pesquisa do Estado de Sa˜o Paulo (FAPESP) and CNPq (Conselho Nacional de Pesquisa), Brazil. * Reprint requests: Silvia Regina Rogatto, PhD, NeoGene Laboratory, Departamento de Urologia, Faculdade de Medicina - UNESP, Botucatu, Sa˜o Paulo, Brazil, CEP: 18618-000. E-mail: [email protected] 0002-9378/$ - see front matter Ó 2005 Mosby, Inc. All rights reserved. doi:10.1016/j.ajog.2005.02.097
Uterine leiomyomas are the most frequently encountered neoplasm of the uterus, with an estimated prevalence of 20% to 30% in women who are O30 years old. On the basis of a study of serial sections of uteri, it has been estimated that up to 77% of reproductive-aged women have uterine leiomyomas.1 Although leiomyomas can be asymptomatic, approximately 20% to 50% are symptomatic and lead patients to seek therapy. The symptoms that are associated with
1396 uterine leiomyomas vary considerably and include abnormal uterine bleeding, pelvic pain or pressure, urinary dysfunction, spontaneous abortion, premature delivery, and infertility.2 These tumors are the most frequent indication for hysterectomy.3,4 The incidence and severity of symptoms is related directly to the number, size, and location of the tumors.4 Approximately two-thirds of women have multiple myomas of various sizes.5 These leiomyomas can occur in any part of the uterus and may be intramural, submucosal, or subserosal. The pathogenesis and evolution of uterine leiomyomas are not well understood. Some authors have suggested that residual undifferentiated cells in the myometrium are the progenitor cells of leiomyomas.6 Other authors have assumed that acquired chromosomal aberrations play a pathogenetic role in uterine leiomyomas.7-9 In addition to gene identification in chromosomally abnormal tumors, the clonality of uterine leiomyomas has been studied to understand the genetic mechanisms that are involved in the genesis and growth of these tumors. Most human tumors are believed to be cytogenetically abnormal clones of cells that are derived from a single progenitor cell that has undergone a mutation. Although an initial mutation may trigger the formation of the tumor, additional mutations may occur during tumor progression, thereby creating subclones with a selective growth advantage and the potential for clonal expansion. If leiomyomas arise from somatic mutations in a myometrial cell that result in the dysregulation of the mechanisms that control cell growth and confer a growth advantage, one would expect to observe clonal expansion. A frequently used method for the evaluation of clonality is based on X-chromosome inactivation. Clonality of tissue samples can be detected by digestion with a methylation-sensitive enzyme and amplification of the remaining allele by polymerase chain reaction (PCR) at a polymorphic X chromosome-linked marker. Cytogenetic and X chromosome inactivation studies strongly suggest that leiomyomas are clonal.8,10,11 As an alternative approach, loss of heterozygosity (LOH) analysis with multiple microsatellite markers has been used to study clonality in bilateral ovarian carcinomas and bilateral breast cancer.12,13 LOH analysis at multiple marker loci can detect a specific allelic deletion pattern. Identical patterns suggest a monoclonal origin of the tumors, whereas polyclonal or tumors with independent clonal origin reveal a completely different pattern with a lack of common alterations. By an investigation of chromosomal regions that were deleted very early in carcinogenesis, an identical LOH pattern is a strong indicator for monoclonality of multiple lesions. However, a lack of common deletion between the tumors that are compared could be due to the omission of the right marker locus in LOH analysis.14 In the present study, LOH analysis and the
Canevari et al pattern of X chromosome inactivation, which is identified by the differential methylation of a site close to the polymorphic CAG repeat in the androgen receptor (AR) gene, were used to determine the clonal status of 43 uterine leiomyomas from 14 patients. To date, this is the first report to analyze the clonality of uterine leiomyomas with the use of both LOH and X chromosome inactivation status.
Material and methods Tumor samples and DNA preparation In 64 patients with uterine leiomyomas, only 14 presented multiple uterine leiomyomas and were analyzed by microsatellite analysis and X chromosome inactivation. The uterine leiomyomas and corresponding peripheral blood leukocytes and normal myometrium from 1 patient were collected in the Department of Obstetrics and Gynecology, Faculty of Medicine, UNESP, Sa˜o Paulo, Brazil. The tissues were obtained from symptomatic women who underwent abdominal hysterectomy or myomectomy for medically indicated reasons. Patients for whom there were available samples of the multiple tumors were selected to do this investigation. Tissue donors were premenopausal, had regular menstrual cycles, and had received no exogenous hormones or hormone suppression therapy before surgery. Women with endometriosis, adenomyosis, or other diseases of the reproductive system were excluded from the study. Connective tissue and necrotic tissue were removed from specimens. Informed consent was obtained from all of the patients before sampling was performed; this study was approved by the Brazilian Ethics Committee (CONEP 334/2001). Genomic DNA from uterine tissues and corresponding blood leukocytes was prepared by standard sodium dodecyl sulfate/proteinase K digestion, followed by phenol and chloroform extraction and ethanol precipitation. DNA was stored at 20(C until amplification by PCR. The medical records of all of the patients were examined to retrieve clinical and histopathologic data. All of the tumors were diagnosed histopathologically to be usual uterine leiomyomas. The clinical and personal data of the patients are shown in Table I. The family history of cancer considered first- and second-degree relatives with cancer to be informative. Where possible, all assessments were based on documented medical records as evidence of cancer. For some families, a history of cancer was ascertained by a cancer geneticist who interviewed the patients and their relatives.
PCR and microsatellite LOH analysis Matched pairs of normal and tumor tissue were PCR amplified at 15 microsatellite loci distributed on 3q263q29, 7p22-7p15, 11p15/q23-25 and 15q26. The primer
Canevari et al
1397
Table I
Clinical and histopathologic data of the 14 patients with multiple uterine leiomyomas
Patient
Age (y)
Location
Familial/personal history
29 42 40 59
Intramural, submucosal Intramural Intramural Intramural, subserosal
40 50 49 43 46 39 49 51 Not determined 48
Subserosal, submucosal Intramural Intramural Intramural Intramural Intramural (one of them giant) Intramural, subserosal, and submucosal Submucosal Intramural Intramural
Not determined/not determined Three sisters with uterine leiomyoma; 1 sister with melanoma/-/Twin sister with uterine leiomyoma; 2 sisters with endometrial carcinoma/skin carcinoma Three sisters with uterine leiomyoma; sister melanoma/Not determined/A sister with uterine leiomyoma/-/Mother with uterus carcinoma/Brother with leukemia/Sister with uterine leiomyoma/Mother with kidney carcinoma/-/-/-
1 2 3 4 5 6 7 8 9 10 11 12 13 14
sequences and cytogenetic maps for the microsatellite markers were retrieved from the Genome Database (http://www.gdb.org) and from The Wellcome Trust Sanger Institute (http://www.ensembl.org), respectively. All of these markers were dinucleotide repeats. Forward primers were synthesized with a fluorescent tag (FAM, NED, or HEX) on the 50 end. Fluorescent PCR amplification was performed in a thermal cycle (Peltier Thermal Cycler, PTC-200; MJ Research, Waltham, MA) in a 15 mL reaction volume that contained 100 ng of template DNA, 10 mmol/L Tris-HCl, pH 8.3, 50 mmol/L KCl, 2.5 mmol/L MgCl2, 200 mmol/L of each dNTP, 0.5 mmol/L of each primer and 1 unit of AmpliTaq Gold DNA polymerase (Applied Biosytems, Foster City, CA). The PCR conditions consisted of an initial denaturation at 95(C for 10 minutes, with subsequent amplification for 30 cycles of 45 seconds at 94(C, 45 seconds at 58( to 60(C, and 1 minute at 72(C. All of the PCR reactions were performed in duplicate to exclude technical errors and to confirm the findings. The fluorescent PCR products were analyzed on 5% polyacrylamide-urea (Long Ranger Singel Packs/FMC, Rockland, ME) denaturing gels with the use of a DNA automated sequencer (ABI Prism 377; Applied Bioystems). The fluorescent gel data were collected and stored with data collection software and then analyzed with Genescan and Genotyper softwares (Applied Biosystems), which quantified the size (in base pairs), height, and area of the allelic peak. An imbalance factor was defined, with both peak areas and heights of the two alleles in the normal and tumor samples as parameters. The formula that was used for the calculation was T1:T2/N1:N2, where T1 and N1 were the peak areas for the shorter length allele products of the tumor and normal samples, respectively, and T2 and N2 were the peak areas for the longer length
allele products of the tumor and normal samples, respectively. For ratios of O1.00, the ratio was converted to 1/AR to give a value between 0.00 and 1.00. A ratio of %0.60 was considered to be indicative of LOH. Comparisons of the ratios with the peak height as the parameter was done with the equation N2:N1/T2:T1, where N2 and T2 were the peak heights of the longer length allele product for the normal and tumor samples, respectively, and N1 and T1 were the peak heights of the shorter length allele product for the normal and tumor samples, respectively. A probability value of %.60 was considered to be indicative of the loss of the shorter length allele, and a probability value R1.60 was indicative of the loss of the longer length allele. Allelic losses were assigned only after the results were found consistently in all 3 experiments that were considering both peak areas and heights.
Analysis of X chromosome inactivation at the AR locus and clonality Clonality was assessed on the basis of the methylation status of the AR gene. This gene is located on Xq11-q12 and contains a region with a CAG polymorphic repeat. The methylation status of 2 HpaII sites located in the CAG trinucleotide repeat AR locus has been shown to correlate with X-chromosome inactivation. To determine the inactivation status of the AR alleles, genomic DNA was digested with the restriction endonuclease HpaII, a methylation-sensitive enzyme that cleaves unmethylated DNA (active X),15 followed by PCR amplification of the CAG polymorphic region. The inactive allele is resistant to digestion with this endonuclease and will amplify a detectable product after PCR. To examine methylation of the AR gene, each DNA sample was digested in a 15 mL reaction volume with 10 U of restriction enzyme RsaI
1398
Canevari et al
Table II Results of microallelotyping analysis with the use of 15 microsatellite markers and clonality for X chromosome inactivation in 14 patients with multiple uterine leiomyomas Case
Tumor
D3S1614
D3S1565
D3S1601
D3S1311
D7S517
D7S513
D7S507
D7S493
1
TA TB TC TD
-
-
-
-
-
-
-
-
2
TA TB TC
-
-
-
-
-
-
3
TA TB TC
-
-
-
-
: : :
: : :
: : :
: : :
4
TA TB
-
-
-
-
-
:
5
TA TB
-
-
-
-
:
-
-
6
TA TB TC TD TE TF
-
-
-
-
-
-
-
-
7
TA TB TC TD
-
-
-
-
-
8
TA TB TC TA TB
-
-
-
-
-
10
TA TB TC
-
-
: -
-
-
-
-
11
TA TB
-
:
-
-
-
-
12
TA TB TC
ND ND ND
-
:
-
-
-
-
13
TA TB TC
-
-
-
-
-
-
14
TA TB TC
-
-
-
-
-
-
-
9
-, Allelic retention;
-
, not informative; :, LOH; ND, not determined.
alone (10 U/ml) and 10 U of RsaI and 60 U of HpaII (10 U/ mL). After 16 hours at 37(C, the samples were incubated at 70(C for 10 minutes to heat-inactivate the enzymes. Genomic DNA was amplified with fluorescent labeled primers, according to Bharaj et al.16 The forward and reverse primer sequences that were used to amplify
the AR gene were 50 -TCCAGAATCTGTTCCAGAGCGTGC-30 and 50 -GCTGTGAAGGTTGCTGTTCCTCAT-30 . One hundred nanograms of digested DNA was amplified in a reaction mixture that contained 125 mmol/L of each deoxynucleotide triphosphate, 0.2 mmol/L of each primer, 10 mmol/L Tris-HCl, pH 8.3, 50 mmol/L
Canevari et al Table II
1399
(Continued) Different LOH/ informative loci(%)
D11S1338
D11S902
D11S925
D11S968
D15S127
D15S120
D15S87
AR allele inactive
-
-
-
-
-
-
0/14 (0)
-
-
-
0/10 (0)
2 2 1
-
-
-
-
-
-
0/14 (0)
1 ND 1
ND
:
-
-
2/10 (20.0)
2 2
-
-
-
-
: :
1/10 (10.0)
1 1
-
-
-
0/12 (0)
2 2 1 2 1 2 1 1 1 1
-
-
-
-
-
-
-
-
0/12 (0)
-
-
: -
-
-
: : -
(16.7)
-
-
-
-
-
(9.0)
1 1
ND -
ND -
ND -
ND -
-
-
-
0/13 (0)
1 1 1
:
-
-
: :
2/10 (20.0)
2 2
-
-
-
1/9 (11.1)
1 1 2
-
-
-
-
-
: : -
1/12 (8.3)
-
-
-
-
-
-
-
0/14 (0)
1 ND ND
-, Allelic retention;
2/12
1/11
2 2 2
, not informative; :, LOH; ND, not determined.
KCl, 1.5 mmol/L MgCl2, and 1 U of Taq DNA polymerase (Invitrogen, Carlsbad, CA). After initial denaturation at 94(C for 10 minutes, PCR was performed with a thermal profile of 30 seconds at 95(C, 1 minute at 65(C, and 5 minutes at 72(C, for a total of 35 cycles. The PCR products were separated by electrophoresis on 5% poly-
acrylamide denaturing gels, and the size of the amplified product was determined in an ABI Prism 377 DNA automatic sequencer, as described earlier. All of the PCR reactions were performed in duplicate to exclude technical errors and to confirm the size of the CAG repeats of each AR allele.
1400 For informative heterozygous cases, clonality was assessed as described in Lucas et al17 and Siu et al.18 Each digested DNA sample was read separately, and the intensity and area of each peak (allele) were measured. LOH analysis was done as described earlier. Only cases without LOH at the AR locus were included in this investigation. The allele amplification ratio for each sample before digestion (RU) with HpaII was calculated by dividing the area of the longer allele peak (allele 2) by that of the shorter allele peak (allele 1). The RU had to be O0.60 to confirm equal amplification of both alleles. Similarly, the RH (allele amplification ratio after digestion with HpaII) was calculated by obtaining the ratio of the longer and shorter allele peaks after digestion. These ratios were divided such that the longer ratio was the denominator: if R was !0.40, the sample was considered to be monoclonal, but if R was O0.40, then the sample was considered to be polyclonal.17
Results Clonality on the basis of the LOH pattern Intratumor heterogeneity of the LOH pattern was assessed with the use of 15 microsatellite markers in 43 samples from 14 patients with uterine leiomyomas. The details of the LOH analysis with the markers that were used are given in Table II, and representative examples of LOH are shown in the Figure, A. In 8 cases, LOH was demonstrated for at least 1 microsatellite marker and at least 1 of the tumors. An identical pattern of LOH was observed in 1 unique patient (case 3) in the microsatellite markers that were mapped on chromosome 7p. Seven cases (cases 4, 5, 8, 9, 11, 12, and 13) showed differential patterns of allelic loss in multiple tumors from the same patient (Figure, A). The percentage of different LOH alleles (number of different LOH loci/number of informative loci ! 100) in 7 cases with different LOH patterns ranged from 8.3% to 20%.
Clonal analysis Clonality was assessed by the determination of the pattern of X chromosome inactivation for the AR gene in all uterine leiomyomas from 14 patients; 7 leiomyomas from 2 patients showed homozygous AR alleles (not informative). The results for 36 uterine leiomyomas from 12 informative cases are summarized in Table II. Each leiomyoma sample that was analyzed consisted of only a single type of inactive allele, which is consistent with a monoclonal neoplastic population of cells. In 3 patients (cases 2, 6, and 12), different X chromosomes were inactivated in multiple tumors, which confirmed the independent clonal origin of each tumor. An analysis of X chromosome inactivation in patient 12 revealed the presence of independent clones, with 1 allele inactivation
Canevari et al in 2 tumors (12A and 12B) and 2 allele inactivation in 1 tumor (12C). In the same case, LOH analysis revealed allelic loss at the D3S1311 marker in tumor 12C but not in the other tumors (12A and 12B; Table II). Although the analysis of X chromosome inactivation in the AR gene revealed independent clones in patients 2 and 6, no allelic losses were found in the autosomal microsatellite markers that were used here. The myometrium (normal control) of patient 6 showed both AR alleles, which is indicative of a mosaic status, whereas each tumor had a monoclonal origin with different AR alleles inactivated, which suggests that they had originated independently (Figure, B). Nine patients who were heterozygous for the AR gene polymorphism showed the same pattern of inactivation; different patterns of allelic loss in the autosomal polymorphic markers of the tumors were analyzed in 5 of these patients (cases 4, 5, 8, 9, and 11). Three of the 36 uterine leiomyomas (3TB, 14TB, and 14TC) were eliminated from the study because the PCR amplification products for the 2 alleles were not approximately equal before digestion with HpaII, nor was there any detectable amplification after digestion with HpaII.
Comment Several leiomyomas often occur in the same uterus, which suggests that there may be an inherited genetic abnormality that predisposes some women to the development of these tumors.19 In the present study, 5 of 13 patients had a familial history of uterine leiomyomas in first-degree relatives. Epidemiologic data suggest that there is a familial risk of the development of uterine leiomyomas in first-degree relatives of affected individuals and that there is a heritable predisposition.20 Vikhlyaeva et al20 observed that uterine leiomyomas were diagnosed 2.2-fold more frequently in families with R2 affected individuals. However, the nature of this familial predisposition is ill defined. An inherited factor in the cause of leiomyoma was also suggested by family studies of Reed’s syndrome.21 The genetic locus for this syndrome was mapped to chromosome 1q42.3-43.22,23 It was shown that germline mutations of fumarate hydratase (FH gene) are responsible for the inherited syndromes of multiple cutaneous and uterine leiomyomas (MCUL, OMIM 150800) and of hereditary leiomyomatosis and renal cell cancer (HLRCC, OMIM 605839).24 These syndromes are inherited as autosomal dominant disorders with reduced penetrance and are associated with cutaneous leiomyomas and, in affected female patients, with uterine leiomyomas. Recently, Gross et al25 showed evidence that was suggestive of linkage to the FH region in a subset of nonsyndromic patients with uterine leiomyomas. We have no patients in our
Canevari et al
1401
Figure A, Allelotypes of a normal tumor sample in which allele loss was detected in one of the uterine leiomyomas of the same patient for the marker D7S493 (case 4) and for the marker D7S513 (case 5). The arrows indicate an LOH. B, Electropherograms show different patterns of X chromosome inactivation in case 6. The arrows indicate the AR allele digested with HpaII (activated allele).
1402 study with MCUL and HLRCC syndromes, as determined by clinical history. Another explanation for the existence of multiple leiomyomas is that an existing tumor may spread to other regions of the uterus. In this case, all of the tumors would be related clonally. However, several studies of X chromosome inactivation in multiple leiomyomas in the same uterus have indicated that the various tumors are not related clonally but rather initiate independently of each other and progress by discrete and separate somatic mutations.10,11,26-28 The independent clonal origin of multiple leiomyomas was determined initially by glucose-6-phosphate dehydrogenase (G6PD) isoenzyme analysis.10 The latter study showed that either the A or B type allele, but not both, was present in each tumor from a series of 7 women who were found to be heterozygous for this enzyme. Furthermore, both A and B type tumors occurred in the same individual. This observation indicated a random pattern of inactivation among multiple tumors, with individual tumors exclusively expressing one or the other G6PD type and also revealed that multiple tumors in the same uterus most likely arise in situ from different myometrial cells and not from a single mutated myometrial cell. However, a low degree of G6PD isoenzyme polymorphism among most female patients across various ethnic groups limited the use of this approach. Subsequent studies that involved CAG trinucleotide repeats in the X-linked AR gene confirmed the independent origin of uterine leiomyomas.11,27 These studies involved the digestion of DNA with methylationsensitive restriction enzymes to assess the differences in methylation between active and inactive X chromosomes at the AR locus in uterine leiomyomas. The cytogenetic studies of multiple uterine leiomyomas from the same uterus have also suggested that each tumor arises independently; hence, these tumors often exhibit different chromosomal alterations.9,29-31 In our study, 7 of 8 cases with allelic loss showed a different pattern of LOH between multiple leiomyomas. These data suggest that uterine leiomyomas can be derived from entirely independent clones or, alternatively, that the abnormality is not primary and could have occurred as a result of clonal evolution. The independent monoclonal origin of these tumors was confirmed here by the X chromosome inactivation analysis that revealed alternate patterns of inactivation in 3 of the 12 informative cases that were studied. The use of X chromosome inactivation to analyze clonality is based on the occurrence of random X chromosome inactivation in early embryogenesis. Because, for a specific progenitor, all daughter cells will inherit the same inactivated X chromosome, this can be used as a marker of clonality.32 Although the initial determination of which of the X chromosomes is inactivated is made at
Canevari et al random, the pattern of inactivation is permanent and is constant throughout subsequent cell divisions.33 Thus, normal tissues of adult women consist of a mosaic of cell types, which differ in whether the maternally or paternally derived X chromosome has been inactivated. In contrast, clonal cellular proliferation maintains the same pattern of X chromosome inactivation that was determined in the single cell of origin.34 However, the reliability of X-chromosome inactivation analysis for clonality analysis in tumors has been challenged.35 One problem is that tumors may show altered DNA methylation patterns.36 As an alternative, LOH analysis with multiple microsatellite markers has been used to study clonality. In this study, the data for 9 of 12 informative patients showed the same pattern of inactivation in different tumors. Because there is a 50% chance of inactivating the same X chromosome in genetically unrelated tumors from different clones, no conclusions could be drawn from these 9 patients. However, 5 of these 9 cases showed different patterns of allelic loss, which suggested an independent evolution of these tumors. Our results agreed with those of cytogenetic and X chromosome inactivation studies that demonstrated the independent origin of multiple uterine leiomyomas. However, of the 8 cases in our study that showed LOH, 1 patient (case 3) had the same LOH pattern in all of the leiomyomas that were present in the uterus and also showed inactivation of the same X chromosome. It has been considered that some leiomyomas develop from a common precursor cell, which suggests that multiple tumors may develop by spreading from a single primary tumor. The discovery of identical cytogenetic alterations in multiple leiomyomas of the same patient favors this developmental pathway.8,37 Alternatively, these changes may represent recurrent chromosomal aberrations that are involved in the tumorigenesis of uterine leiomyomas. This hypothesis is supported by a report of a patient with 2 independent leiomyomas, each showing different patterns of X chromosome inactivation but identical del(7)(q21.2q31.2) derivative chromosomes.30 We recently reported the frequent involvement of allelic loss at 7p22-15 in uterine leiomyoma.38 In addition to the previous reports, the same allelic losses that were found at 7p in all 3 uterine leiomyomas from the same patient suggest their relevance in a subgroup of uterine leiomyomas. In conclusion, clonal analysis of multiple uterine leiomyomas that are based on the patterns of LOH and X chromosome inactivation indicated that the leiomyomas that were studied were unicellular in origin but are independently generated in the uterus. The chromosomal and molecular heterogeneity and the different patterns of X chromosome inactivation that were found in these tumors may provide more accurate information on the clonal origin of tumors and may
Canevari et al explain clinicopathologic differences that are seen in uterine leiomyomas.
Acknowledgment We thank Rodrigo Mattos dos Santos for expert technical assistance.
References 1. Cramer S, Patel A. The frequency of uterine leiomyomas. Am J Clin Pathol 1990;94:435-8. 2. Quade BJ. Pathology, cytogenetics and molecular biology of uterine leiomyomas and other smooth muscle lesions. Curr Opin Obstet Gynecol 1995;7:35-42. 3. Ligon AH, Morton CC. Genetics of uterine leiomyomata. Genes Chromosomes Cancer 2000;28:235-45. 4. Buttram VC, Reiter RC. Uterine leiomyomata: etiology, symptomatology and management. Fertil Steril 1981;36:433-45. 5. Zaloudek CJ, Norris HJ. Mesenchymal tumors of the uterus. In: Kurman RJ, editor. Blausteins’s pathology of the female genital tract. 4th ed. New York: Springer-Verlag; 1994. p. 487-94. 6. Hoang Ngoc M, Smadja A, Herve de Sigalony JP. A classification of uterine tumors based upon embryogenesis. Arch Anat Cytol Pathol 1991;39:94-102. 7. Mark J, Havel G, Grepp C, Dahlenfors R, Wedell B. Chromosomal patterns in human benign uterine leiomyomas. Cancer Genet Cytogenet 1990;44:1-13. 8. Nilbert M, Heim S. Uterine leiomyoma cytogenetics. Genes Chromosomes Cancer 1990;2:3-13. 9. Rein MS, Friedman AJ, Barbieri RL, Pavelky K, Fletcher JA, Morton CC. Cytogenetic abnormalities in uterine leiomyomata. Obstet Gynecol 1991;77:923-6. 10. Townsend DE, Sparkes RS, Baluda MC, McClelland G. Unicellular histogenesis of uterine leiomyomas as determined by electrophoresis of glucose-6-phosphate dehydrogenase. Am J Obstet Gynecol 1970;107:1168-73. 11. Mashal RD, Schoenberg ML, Friedman AJ, Nowak RA, Rein MS, Morton CC, et al. Analysis of androgen receptor DNA reveals the independent clonal origins of uterine leiomyomata and the secondary nature of cytogenetic aberration in the development of leiomyomata. Genes Chromosomes Cancer 1994;7:1-6. 12. Abeln ECA, Kuipers-Dijkshoorn NJ, Berns EMJJ, HenzenLogmans SC, Fleuren GJ, Cornelisse CJ. Molecular genetic evidence for unifocal origin of advanced epithelial ovarian cancer and for minor clonal divergency. Br J Cancer 1995;72:1330-6. 13. Imyanitov EN, Suspitsin EN, Grigoriev MY, Togo AV, Kuligina ES, Belogubova EV, et al. Concordance of allelic imbalance profiles in synchronous and metachronous bilateral breast carcinomas. Int J Cancer 2002;100:557-64. 14. Hafner C, Knuechel R, Stoehr R, Hartmann A. Clonality of multifocal urothelial carcinomas: 10 years of molecular genetic studies. Int J Cancer 2002;101:1-6. 15. Allen RC, Zoghbi HY, Moseley AB, Rosenblatt HM, Belmont JW. Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in the human androgen-receptor gene correlates with X-chromosome inactivation. Am J Hum Genet 1992;51:1229-39. 16. Bharaj BS, Vassilikis EJK, Diamandis EP. Rapid and accurate determination of (CAG) repeats in the androgen receptor gene using polymerase chain reaction and automated fragment analysis. Clin Biochem 1999;32:327-32. 17. Lucas DR, Shroyer K, McCarthy P, Marklam N, Fujita M, Enomoto T. Desmoid tumor is a clonal cellular proliferation: PCR amplification of HUMARA for analysis of patterns of Xchromosome inactivation. Am J Surg Pathol 1997;21:306-11.
1403 18. Siu IM, Robinson DR, Schwartz S, Kung HJ, Pretlow TG, Petersen RB, et al. The identification of monoclonality in human aberrant crypt foci. Cancer Res 1999;59:63-6. 19. Mao X, Barfoot R, Hamoudi RA, Easton DF, Flanagan AM, Stratton MR. Allelotype of uterine leiomyomas. Cancer Genet Cytogenet 1999;114:89-95. 20. Vikhlyaeva EM, Khodzhaeva ZS, Fantschenko ND. Familial predisposition to uterine leiomyomas. Int J Gynaecol Obstet 1995;51:127-31. 21. Reed WB, Walker R, Horowitz R. Cutaneous leiomyomata with uterine leiomyomata. Acta Derm Venereol 1973;53:409-16. 22. Alam NA, Bevan S, Churchman M, Barclay E, Barker K, Jaeger EE, et al. Localization of a gene (MCUL1) for multiple cutaneous leiomyomata and uterine fibroids to chromosome 1q42.3-q43. Am J Hum Genet 2001;68:1264-9. 23. Martinez-Mir A, Gordon D, Horev L, Klapholz L, Ott J, Christiano AM, et al. Multiple cutaneous and uterine leiomyomas: Refinement of the genetic locus for multiple cutaneous and uterine leiomyomas on chromosome 1q42.3-q43. J Invest Dermatol 2002;118:876-80. 24. Tomlinson IP, Alam NA, Rowan AJ, Barclay E, Jaeger EM, Kelsell D, et al. Germline mutations in FH predispose to dominantly inherited uterine fibroids, skin leiomyomata and papillary renal cell cancer. Nat Genet 2002;30:406-10. 25. Gross KL, Panhuysen CM, Kleinman MS, Goldhammer H, Jones ES, Nassery N, et al. Involvement of fumarate hydratase in nonsyndromic uterine leiomyomas: Genetic linkage analysis and FISH studies. Genes Chromosomes Cancer 2004;41:183-90. 26. Fialkow PJ. Use of genetic markers to study cellular origin of tumors in human females. Adv Cancer Res 1972;15:191-226. 27. Hashimoto K, Azuma C, Kamiura S, Kimura T, Nobunaga T, Kanai T, et al. Clonal determination of uterine leiomyomas by analyzing differential inactivation of the X-chromosome-linked phosphoglycerokinase gene. Gynecol Obstet Invest 1995;4093:204-8. 28. Bonatz G, Frahm SO, Andre´as S, Heidom K, Jonat W, Parwaresch R. Telomere shortening in uterine leiomyomas. Am J Obstet Gynecol 1998;179:591-6. 29. Nilbert M, Heim S, Mandahl N, Floderus UM, Willen H, Mitelman F. Different karyotypic abnormalities, t(1;6) and del(7), in two uterine leiomyomas from the same patient. Cancer Genet Cytogenet 1989;42:51-3. 30. Nilbert M, Stro¨mbeck B. Independent origin of uterine leiomyomas with karyotypically identical alterations. Gynecol Obstet Invest 1992;33:246-8. 31. Mantovani MS, Neto JB, Philbert PM, Casartelli C. Multiple uterine leiomyomas: cytogenetic analysis. Gynecol Oncol 1999;72:71-5. 32. Lyon MF. X-chromosome inactivation and developmental patterns in mammals. Biol Rev 1972;47:1-35. 33. Gartler SM, Riggs AD. Mammalian X-chromosome inactivation. Annu Rev Genet 1983;17:155-90. 34. Fialkow PJ. Clonal origin of human tumors. Biochim Biophys Acta 1976;458:283-321. 35. Buller RE, Sood AK, Lallas T, Buekers T, Skilling JS. Association between nonrandom X-chromosome inactivation and BRCA1 mutation in germline DNA of patients with ovarian cancer. J Natl Cancer Inst 1999;91:339-46. 36. Yang DH, Smith ER, Cohen C, Wu H, Patriotis C, Godwin AK, et al. Molecular events associated with dysplastic morphologic transformation and initiation of ovarian tumorigenicity. Cancer 2002;94:2380-92. 37. Nilbert M, Heim S, Mandahl N, Floderus UM, Willen H, Mitelman F. Characteristic chromosome abnormalities, including rearrangements of 6p, del(7q), C12, and t(12;14), in 44 uterine leiomyomas. Hum Genet 1990;85:605-11. 38. Canevari RA, Pontes A, Rogatto SR. Microallelotyping defines novel regions of loss of heterozygosity in uterine leiomyomas. Mol Carcinog 2005;42:177-82.