This article is based on a presentation at the conference on Women's Health and the Environment: The Next Century--Advances in Uterine Leiomyoma Research held 7-8 October 1999 in Research Triangle Park, North Carolina, USA.
Address correspondence to C.C. Morton, Department of Pathology, Brigham and Women's Hospital, 75 Francis St., Boston, MA 02115 USA. Telephone: (617) 732-7980. Fax: (617) 738-6996. E-mail: ccmorton@bics.bwh.harvard.edu
The author thanks members and collaborators of the Morton laboratory. This research was supported by the National Cancer Institute (CA78895).
Received 23 February 2000; accepted 3 May 2000.
Multiple genetic approaches are now being used to discover genes involved in the pathogenesis and pathobiology of uterine leiomyomata (
1). One such method is the use of cytogenetics to proceed to positional cloning of genes, and there have been several recent successes using this approach. Another method, genetic linkage analysis, makes it possible to identify genes that might predispose women to develop uterine fibroids. Finally, an exciting innovative genetic technology known as transcriptional profiling is providing new insight into the molecular dissection of cellular pathways that contribute to the growth and development of uterine myomas.
Uterine leiomyomata are the most common pelvic tumor in women, occurring in a minimum of 20-25% of women of reproductive age. In fact, these tumors may be as prevalent as 70% of women in this age range (2); fortunately, not all of these women suffer from the sequelae of these tumors. Regardless, it is clear that fibroids are a major public health problem, accounting for greater than 200,000 hysterectomies in the United States annually (3). Myomas are associated with various clinical problems, including abnormal uterine bleeding, pelvic pain, urinary frequency, and constipation, that greatly decrease the quality of life for women. In addition, they have a biological role in impaired fertility and spontaneous abortion that remains to be elucidated. Although much remains to be known about the growth and development of these smooth-muscle tumors, it is clear that they are steroid-dependent tumors not seen prior to puberty, may increase in size during pregnancy, and can regress postmenopausally.
With regard to the pathology of uterine fibroids, these tumors are benign, with an estimate of less than 0.1% progression to their presumed malignant counterpart, uterine leiomyosarcoma. This rare event of progression is an intriguing observation because of the high frequency of these tumors, about six or seven tumors per uterus (2), and the high prevalence among women of reproductive age. It is reasonable to question whether these tumors are in the same genetic pathway to become uterine leiomyosarcomas. As more genetic markers are developed, it will be possible to address this question in a scientifically rigorous fashion. Myomas appear as well-differentiated, whorled bundles of smooth-muscle cells that comprise distinct nodules. Consistent with their benign nature, mitoses are scant and normal in appearance. In most instances, histopathology with hematoxylin and eosin staining clearly differentiates the tumor nodule from surrounding myometrium. Underlying molecular changes such as alterations in gene expression, and perhaps protein stability, undoubtedly account for the characteristic fibroid phenotype that can be detected even by this relatively simple microscopic analysis. Anatomically, fibroids can be found in intramural, subserosal, or submucosal locations, often deforming the shape of the uterus and expanding its size.
Concerning the genetics of uterine leiomyomata, our understanding can be partitioned into three broad categories: epidemiology, molecular/biochemical, and cytogenetic studies. Epidemiological studies can provide valuable clues to the contribution of heredity to a particular trait by ascertaining ethnic and racial differences in prevalence, familial aggregation, and twin studies. Some epidemiological studies have indicated uterine leiomyomata are 3-9 times more frequent in black women than in white women (4,5). Specifically, a recent analysis from the Nurses' Health Study has confirmed about a 3-fold increased frequency of myomas in African American women compared to Caucasian women after adjustment for known risk factors (5). Furthermore, various studies have reported familial predisposition to uterine leiomyomata in Russian populations. One such study (6) of 97 families consisting of 215 female patients revealed that fibroids were 2.2 times more frequent among first-degree female relatives in families with two or more verified leiomyoma cases. An interesting correlation for determining the genetic basis for a trait is the comparison of concordance rates in monozygotic twins, who are genetically identical, with that of dizygotic twins, who have, on average, one-half of their genes in common. The correlation for hysterectomy in monozygotic twins (about 0.6) is approximately twice that found in dizygotic twins (0.3) (7), exactly the expected rate for a genetically influenced trait. Although other genetically influenced conditions besides fibroids may underlie the medical need for hysterectomy, these findings in twins certainly suggest a genetic liability for fibroids, as these tumors are the most common indication for hysterectomy.
Biochemical studies that date back to the early 1970s examined whether fibroids are clonal tumors. Using glucose-6-phosphate dehydrogenase isozyme analysis, which is not a very polymorphic marker system, it was found that uterine leiomyomata are independent, clonal proliferations (8). More recently, it has been possible to confirm these findings by exploiting variation in triplet repeats at the androgen receptor locus, which is a much more informative marker system (9). The androgen receptor polymorphism assay is based on the fact that one X chromosome in human females is essentially inactivated and that methylation-sensitive restriction enzymes (such as HhaI) can be used to distinguish alleles on the active and inactive X chromosomes. Oligonucleotide primers spanning the androgen receptor repeats and encompassing a region containing an HhaI site are used to amplify by polymerase chain reaction (PCR) the DNA residing between the primer sites. If the genomic DNA is digested prior to performing PCR amplification, no product will be amplified from the active X chromosome, as HhaI will cut the DNA inside the primer sites. However, a DNA fragment will be amplified from the inactive X chromosome, as the restriction enzyme will not be able to digest methylated DNA. In addition to confirming previous data on clonality, this assay provided a method to study a group of chromosomally mosaic tumors (i.e., tumors that show a mixture of chromosomally normal and abnormal cells in the same culture). Results of this assay in one particular cytogenetic subgroup characterized by a deletion on chromosome 7 indicate that even though the tumors are mosaic, they appear to be clonal (9,10). An interpretation of this finding is that the cytogenetic event was most likely a secondary event in the karyotypic evolution of the tumor and not the primary tumorigenic event. The consistent observation of del(7q) in uterine leiomyomata indicates, nevertheless, that it is important in the pathobiology of the tumor.
With regard to cytogenetic investigations, tumor-specific chromosome aberrations have been detected in myomas. Interestingly, the majority of tumors are chromosomally normal; about 40% are abnormal. Characteristic cytogenetic aberrations have provided crucial landmarks for gene discovery and predict different genetic mechanisms for tumor growth. For example, one of the cytogenetic subgroups consists of trisomy 12; trisomies are generally thought mechanistically to involve an increase in gene dosage. Thus, an extra copy of chromosome 12 in the fibroid cells is likely to result in an increase in a gene product from chromosome 12 that participates in the growth and development of these tumors. Besides trisomy 12, other common aberrations are translocations involving 12. Translocations are associated with dysregulating gene products, abrogating expression of a particular gene, or formation of a chimeric gene transcript. Finally, chromosomal deletions, such as that of the characteristic del(7q), are considered to result in a loss of gene function and may uncover tumor-suppressor genes. The diversity of these predicted mechanisms is perhaps not surprising in view of the high prevalence of these tumors. One hypothesis that can be formulated from these observations is that there are many different genetic pathways for a fibroid to develop and grow.
Uterine leiomyomata can harbor more than one of the characteristic chromosomal rearrangements, such as a t(12;14) and a del(7q) rearrangement. The significance of the presence of two consistent abnormalities in the same tumor is unknown. Observations of either t(12;14) or del(7q) alone are common, so there does not appear to be an interdependence of these molecular events in the biology of the tumor. However, it will be important to address this possibility at a finer level of resolution, as there may be cryptic events involving either of the loci. For example, a tumor that cytogenetically has a del(7q) may already have a submicroscopic abnormality of chromosome 12 (e.g., point mutation) not apparent as a translocation or trisomy.
Uterine leiomyomata can be very heterogeneous with respect to their size, location, and clinical response to therapy, and provide a model system for understanding how different genetic events are related to the phenotype of the tumors. Do different cytogenetic abnormalities correlate with the size of the tumor, the location of the tumor, or how the tumor responds to therapy? Several collections of karyotyped tumors have been analyzed for an association between karyotype and size (11,12), and there is agreement among the reports that karyotypically abnormal tumors are significantly larger than karyotypically normal tumors. Furthermore, it appears that tumors with t(12;14) are larger than tumors with del(7q), suggesting that tumors with t(12;14) rearrangements may have some growth advantage.
The region of chromosome 12 in band q15 involved in t(12;14) in uterine leiomyomata is of particular interest. It is involved nonrandomly in these tumors and is also a frequent breakpoint region in a variety of mesenchymally derived tumors, including lipomas, pulmonary chondroid hamartomas, lipoleiomyomas, and others. Although lipomas often have rearrangements of 12q15, these rearrangements differ remarkably from those of myomas in that they rarely involve chromosome 14. In contrast, rearrangements in pulmonary chondroid hamartomas often involve the same region of chromosome 14. A hierarchy of chromosome rearrangements has come to be recognized within the group of benign mesenchymal tumors. At present the basis for this observation is unknown, but it is clear that it is not stochastic, because hundreds of these tumors have been karyotyped.
Using fluorescence in situ hybridization (FISH) and DNA clones mapping to 12q15, a yeast artificial chromosome (YAC) was identified that crossed the translocation breakpoint in a uterine leiomyoma, a lipoma and a pulmonary chondroid hamartoma with 12q rearrangements (13). This finding was a major breakthrough in this positional cloning effort, as it indicated that the gene involved in these benign mesenchymal tumors resided within this region of the genome. As a direct result of the genome project, it had been determined that the HMGIC gene mapped within this YAC, and it became an immediately attractive positional candidate. Besides the physical location of HMGIC within the YAC, other findings suggested that it was a likely candidate gene (e.g., it is large gene providing a big target for rearrangement, an Hmgic knockout mouse showed a significant reduction in adipose tissue, and expression in mesenchymal tissues was observed). Also of interest, there was a gene in the HMGIC family known as HMGIY that had been mapped to 6p21, another site of chromosomal rearrangement in mesenchymal tumors including uterine leiomyomata. Genomic clones from either end of the HMGIC gene were used as probes in FISH experiments to tumor metaphases with 12q15 rearrangements to determine whether there was any type of disruption in the gene at the genomic level. FISH experiments with metaphase spreads from three lipomas with rearrangements of chromosome 12 revealed that a genomic rearrangement occurred somewhere between the two probes encoding HMGIC (14). Subsequently, fusion transcripts were identified in the lipomas that joined the 5´ region of HMGIC, including its three AT hook DNA-binding motifs, to heterologous sequences from diverse partner chromosomes.
In contrast, in uterine leiomyomas with t(12;14), both probes were translocated to chromosome 14. This result indicated that a different molecular mechanism from that in lipomas might be operative in myomas (15). In FISH analyses of more than two dozen uterine leiomyomas, almost all breakpoints have been mapped 5´ of the HMGIC gene, leaving the coding region of HMGIC intact. A few variant rearrangements have been detected that rearrange 3´ of the gene but leave the entire coding region of the gene intact. This situation is reminiscent of the rearrangements that occur in Burkitt lymphoma in which translocations take place either 5´ or 3´ of MYC (16). In those tumors most translocations are t(8;14), involving MYC and the Ig heavy-chain locus, and the breakpoints are 5´ of MYC. However, variant translocations involving either chromosome 2 (Ig kappa) or 22 (Ig lambda) are seen, and in these translocations the breakpoints are 3´ of MYC. Proving the dysregulation of HMGIC in this situation is much more difficult than if a fusion transcript were readily detectable, as in lipomas. Expression studies of HMGIC in fresh tissues from myomas with t(12;14) rearrangements and matched myometrium revealed HMGIC expression only in myomas with t(12;14), providing evidence for dysregulation of HMGIC in at least a subgroup of myomas (17).
As mentioned above, there is also a cytogenetic subgroup among the benign mesenchymal tumors that involves rearrangements of chromosome 6 in band 6p21 where the HMGIY gene resides. Rearrangements involving chromosome 6 and the same region of 14 are seen frequently in endometrial polyps and pulmonary chondroid hamartomas but relatively infrequently in fibroids, although there are a few reported cases (18,19). One fibroid showed a pericentric inversion of chromosome 6 involving bands p21 and q15 (20). FISH with a P1 clone containing HMGIY was disrupted in this inv(6) chromosome; hybridization signals were detected at both 6p21 and 6q15 corresponding to the breakpoints of the inversion. Using an electrophoretic mobility shift assay (EMSA) it was possible to determine that HMGIY was highly expressed in that fibroid, and no HMGIY expression was detectable in the matched myometrium. Surprisingly, a survey of a number of uterine leiomyomata and matched myometrium that were either karyotypically normal or had chromosome rearrangements other than ones at 6p21 revealed that some tumors expressed HMGIY. No myometria were found to express HMGIY by EMSA. These results were unexpected, but may reflect in part the fact that HMGIY has a much wider distribution of expression than does HMGIC.
A more sensitive method to assess the presence of HMGIC and HMGIY transcripts, reverse transcriptase-polymerase chain reaction, was explored to address these findings (17). An absolute correlation of HMGIC expression in tumors that have t(12;14) rearrangements was found with no expression in matched myometrium, which is consistent with Northern blot results. Also consistent was the finding that HMGIY was expressed in tumors that were either karyotypically normal or with chromosome rearrangements other than ones involving 6p21. Thus, there is not a simple correlation of HMGIY expression with a particular karyotype like that observed for HMGIC.
The identification of the gene on chromosome 14 has been an area of immense interest due to its frequent involvement as a partner in rearrangements with either HMGIC or HMGIY. A recent paper reports identification of the disruption of a member of the rad 51 family of DNA repair genes in t(12;14) in fibroids (21). Further studies will doubtless be forthcoming about the role of this gene in the biology of uterine leiomyomata.
Is there a genetic liability to developing uterine fibroids? There are various clues that suggest that there might be a genetic liability including ethnic predisposition studies, twin studies, and familial aggregation studies discussed previously. Another potentially important finding that suggests a trait has a genetic basis is association of the trait with other disorders with a clear Mendelian pattern of inheritance. In fact, there have been reports of fibroids in association with inherited disorders such as Reed syndrome (
22), Bannayan-Zonana syndrome, Cowden syndrome, and multiple lipomatosis. Reed syndrome (MIM150800) is inherited as an autosomal dominant trait with reduced penetrance and is characterized by multiple leiomyomata of the skin believed to arise from erector pili muscles. In addition, this syndrome reportedly has an association with uterine myomata. Interestingly, a possible locus has been predicted to reside on the short arm of chromosome 18, based on studies of a patient with Reed syndrome who had an unbalanced chromosome rearrangement involving chromosomes 9 and 18 (
23). The rearrangement resulted in partial trisomy 9p and partial monosomy 18p, suggesting that this disorder may result from the loss of a gene sequence from 18p. Moreover, it is possible that genetic linkage analysis in a clearly autosomally inherited Mendelian disorder might provide valuable insight as to where predisposition genes for fibroids might reside, which is difficult to investigate because of their high prevalence in the population. Despite the otherwise inherent difficulties in genomewide scans of common diseases, the affected sib pairs method for identifying predisposition genes for uterine leiomyomata is a feasible approach, and peripheral blood specimens are being collected to undertake such a study (
24).
Transcriptional profiling methods using microarray technologies (e.g., oligonucleotide chips, cDNA microarrys on slides or filters) are making possible the study of the expression of thousands of genes at a particular time point in a tissue. There is no doubt that these types of experiments will give valuable insight into the global disturbances taking place in transformed cells and the molecular pathways involved. Uterine fibroids provide an interesting experimental design for using gene expression chips because multiple independent tumors are available from the same individual, providing an opportunity to address tumor-specific genetic events not confounded by environmental effects that might differ between patients with the same tumor type. In this context, environmental effects also include internal environmental effects such as host hormonal influences. Furthermore, genetic events that are recognizable by cytogenetics will be able to be assessed for their effects on all genes being expressed in that tumor. For example, are genes dysregulated in the t(12;14) subgroup providing those tumors with a growth advantage compared to karyotypically normal tumors? Scatter plots of four fibroids and the matched myometrium from the same patient show that about 1,500 genes of the approximately 6,800 genes on an Affymetrix gene expression chip are expressed differentially. It is a complex question to figure out which expression differences are important and which ones to investigate further. Of note, the scatter plots of the four fibroids reveals a similar pattern of expression. Thus, there are certainly similar genetic events occurring in these independent smooth-muscle tumors in addition to unique ones.
There are many different ways to manipulate the gene expression data. One analysis is a dendrogram, which looks at the relationship of the gene expression in the tumors with respect to each other and to the myometrium. The dendrogram reveals that the pattern of gene expression in the myometrium is the most different from the four fibroids. Another method of analysis is a K-means analysis, in which it is possible to cluster genes that are either upregulated or downregulated coordinately. Each cluster is made up of many genes, and the interpretation of these clusters is an evolving science that will clearly change our understanding of the complexity of the molecular events underlying dysregulated cellular growth.
The patients who will one day be served by the clinical and basic research ongoing in uterine leiomyomata are constant sources of inspiration to researchers. An example of this is an e-mail message received from a patient:
I read about your research in the Washington Post. I'm offering my mother and me as subjects if you wish. We both had fibroids that tipped over our uterus at the exact same age, 44. She had a hysterectomy; I had a myomectomy. We appreciate your research.
REFERENCES AND NOTES
1. Ligon AH, Morton CC. Genetics of uterine leiomyomata. Genes Chromosomes Cancer 28L235-245 (2000).
2. Cramer SF, Patel A. The frequency of uterine leiomyomas. Am J Clin Pathol 94:435-438 (1990).
3. National Center for Health Statistics. Hysterectomies in the United States 1965-84. Atlanta, GA:Centers for Disease Control, 1987.
4. Kjerulff KH, Guzinski GM, Langenberg PW, Stolley PD, Adler Moye NE, Kazandjian VA. Hysterectomy and race. Obstet Gynecol 82:757-764 (1993).
5. Marshall LM, Spiegelman D, Barbieri RL, Goldman MB, Manson JE, Colditz GA, Willett WC, Hunter DJ. Variation in the incidence of uterine leiomyomata among premenopausal women by age and race. Obstet Gynecol 90:967-973 (1997).
6. Vikhlyaeva EM, Khodzhaeva ZS, Fantschenko ND. Familial predisposition to uterine leiomyomas. Int J Gynecol Obstet 51:127-131 (1995).
7. Treloar SA, Martin NG, Dennerstein L, Raphael B, Heath AC. Pathways to hysterectomy: insights from longitudinal twin research. Am J Obstet Gynecol 167:82-88 (1992).
8. 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 107:1168-1173 (1970).
9. Mashal RD, Fejzo MS, Friedman AJ, Mitchner N, Nowak RA, Rein MS, Morton CC, Sklar J. Analysis of androgen receptor DNA reveals the independent clonal origins of uterine leiomyomata and the secondary nature of cytogenetic aberrations in the development of leiomyomata. Genes Chromosomes Cancer 11:1-6 (1994).
10. Xing YP, Powell WL, Morton CC. The del(7q) subgroup in uterine leiomyomata: genetic and biologic characteristics. Cancer Genet Cytogenet 98:69-74 (1997).
11. Rein MS, Powell WL, Walters FC, Weremowicz S, Cantor RM, Barbieri RL, Morton CC. Cytogenetic abnormalities in uterine myomas are associated with myoma size. Mol Hum Reprod 4:83-86 (1998).
12. Brosens I, Deprest J, Dal Cin P, Van den Berghe H. Clinical significance of cytogenetic abnormalities in uterine myomas. Fertil Steril 69:232-235 (1998).
13. Schoenberg Fejzo M, Yoon S-J, Montgomery K, Rein MS, Weremowicz S, Krauter KS, Dorman TE, Fletcher JA, Mao J, Moir DT, et al. Identification of a YAC spanning the translocation breakpoints in uterine leiomyomata, pulmonary chondroid hamartoma and lipoma: physical mapping of the 12q14-q15 breakpoint region in uterine leiomyomata. Genomics 26:265-271 (1995).
14. Ashar HR, Schoenberg Fejzo M, Tkachenko A, Zhou X, Fletcher JA, Weremowicz S, Morton CC, Chada K. Disruption of the architectural factor HMGI-C: DNA-binding AT hook motifs fused in lipomas to distinct transcriptional regulatory domains. Cell 82:57-65 (1995).
15. Schoenberg Fejzo M, Ashar HR, Krauter KS, Powell WL, Rein MS, Weremowicz S, Yoon S-J, Kucherlapati RS, Chada K, Morton CC. Translocation breakpoints upstream of the HMGIC gene in uterine leiomyomata suggest dysregulation of this gene by a mechanism different from that in lipomas. Genes Chromosomes Cancer 17:1-6 (1996).
16. Holt JT, Morton CC, Nienhuis AW, Leder P. Molecular Mechanisms of Hematological Neoplasms. Philadelphia:W.B. Saunders Co., 1987.
17. Gattas GJF, Quade BJ, Nowak RA, Morton CC. HMGIC expression in human adult and fetal tissues and in uterine leiomyomata. Genes Chromosomes Cancer 25:316-322 (1999).
18. Sornberger KS, Weremowicz S, Williams AJ, Quade BJ, Ligon AH, Pedeutour F, Vanni R, Morton CC. Expression of HMGIY in three uterine leiomyomata with complex rearrangements of chromosome 6. Genes Chromosomes Cancer 114:9-16 (1999).
19. Dal Cin P, Quade BJ, Weremowicz S, Morton CC, Pauwels P, Van Den Berghe H. Primary parauterine leiomyoma with a t(6;14). Genes Chromosomes Cancer 26:383-384 (1999).
20. Williams AJ, Powell WL, Collins T, Morton CC. HMGI(Y) expression in human uterine leiomyomata: involvement of another HMG architectural factor in a benign neoplasm. Am J Pathol 150:911-918 (1997).
21. Schoenmakers EFPM, Huysmans C, Van de Ven WJM. Allelic knockout of novel splice variants of human recombination repair gene RAD51B in t(12;14) in uterine leiomyomas. Cancer Res 59:19-23 (1999).
22. Reed WB, Walker R, Horowitz R. Cutaneous leiomyomata with uterine leiomyomata. Acta Dermatovener (Stockholm) 53:409-416 (1973).
23. Fryns J-P, Haspeslagh M, de Muelenaere A, Van den Berghe H. 9p trisomy/18p distal monosomy and multiple cutaneous leiomyomata: another specific chromosomal site (18pter) in dominantly inherited multiple tumors. Hum Genet 70:284-286 (1985).
24. Gross K, Morton CC. Available: http://www.fibroid.net [updated 12 September 2000].
Last Updated: October 2, 2000
