| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Biology and Genetics |
Departments of Pathology [B. J. Q., S. W., P. D. C., C. C. M.] and Obstetrics, Gynecology, and Reproductive Biology [D. M. N., C. C. M.], Brigham and Women’s Hospital and Harvard Medical School [B. J. Q., S. W., P. D. C., C. C. M.], Boston, Massachusetts 02115; Sezione di Biologia e Genetica, Dipartimento Scienze Applicate ai Biosistemi, University of Cagliari, 09042 Monerrato (Cagliari), Italy [R. V.]; and Connecticut State Police Forensic Science Laboratory, Meriden, Connecticut 06451 [C. L.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
80% of females of reproductive age, arguably making them the most frequent human neoplasm. Approximately 40% of leiomyomata have cytogenetic abnormalities. One of the most frequent rearrangements is t(12;14)(q15;q23-24). This translocation is present in 59 of 383 (or 15.4%) uterine leiomyomata at the Mitelman Database of Chromosome Abnormalities web site (August 14, 2002 release).3
Rearrangements of 12q15 are also found in a number of other benign mesenchymal tumors. The chromosome 12 breakpoint region has been mapped by FISH4 in many of the involved tumor types. In each case, it is located in or near the DNA architectural factor HMGA2 (formerly HMGI-C), suggesting that HMGA2 participates in a wide spectrum of benign mesenchymal tumors. In lipomas, balanced translocations create fusion genes between HMGA2 and translocation partners such as lipoma-preferred partner and lipoma HMGIC fusion partner (1 , 2) . Three DNA binding domains encoded in the first three exons of HMGA2 are present in these fusion transcripts.
Involvement of 14q23-24 in rearrangements with 12q15 is relatively specific for those mesenchymal tumors with full or partial smooth muscle differentiation. Rearrangements of 14q23-24 are found in endometrial polyps and pulmonary chondroid hamartomata. The chromosome 14 breakpoint region in uterine leiomyomata has been mapped within the DNA repair gene RAD51L1, which spans a large genomic region of 680 kb. It has been hypothesized that a HMGA2-RAD51L1 fusion transcript, the reciprocal fusion transcript, or both, play an important role in leiomyomata (3, 4, 5)
. We mapped the chromosome 12 breakpoints relative to the HMGA2 locus in a series of 38 uterine leiomyomata with 12q15 rearrangements and found that 12q15 rearrangements mapped predominantly in the 5' region of HMGA2. In a subset, we tested for the presence of reciprocal fusion transcripts involving HMGA2 and RAD51L1.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
clones H121, H206, H2, H4, H311, H738, H501, H111, H528, H403, and H409 has been reported previously (8)
.
Reverse Transcription and PCR Analysis.
Total RNA from each sample was first transcribed with Superscript II reverse transcriptase (Invitrogen) and an oligo-dT primer. After confirmation of RNA integrity by RT-PCR for GAPDH, cDNA samples were subjected to a variety of PCR assays. Primers for HMGA2 and RAD51L1 were named according to their starting nucleotide in either of GenBank U28749 for HMGA2 or GenBank U92074 for RAD51L1, the letter U or L to denote whether the primer was derived from the upper or lower strand, and the primer length in bases, respectively. Primer sequences also may be found online.5
Assays for HMGA2 transcripts were performed with PCR conditions using 1 min of denaturation at 94°C, followed by 30 cycles of amplification (45 s at 94°C and 90 s at 68°C). Conditions for amplifying RT-PCR products from RAD51L1 were modified to include a 30 s step at 60°C for annealing.
Two sets of assays to detect a fusion transcript involving HMGA2 and RAD51L1 were performed. The first reaction involved using 5' RAD51L1 upper primers 20U25 or 536U23 with 3' HMGA2 lower primers 1115L22 or 1047L25. After this, products were subjected to nested amplification using 5' HMGA2 upper primers 712U24 or 959U25 with 3' RAD51L1 lower primers 976L24 or 1040L24. All reactions were performed using the PCR conditions described above for amplification of RAD51L1 RT-PCR products.
3' RACE.
Total RNA isolated from tumor 105274 was used for 3' RACE (Invitrogen). The 5' HMGA2 primer in the first round of amplification was 849U19. The hot start PCR profile was as follows: 3 min at 94°C for denaturation; 7 min at 80°C to allow for addition of eLONGgase polymerase mix (Invitrogen); 5 cycles of 30 s at 94°C and 4 min at 70°C; 5 cycles of 30 s at 94°C, 30 s at 65°C and 4 min at 70°C; and 30 cycles of 30 s at 94°C, 30 s at 60°C and 4 min at 70°C. The nested 5' HMGA2 primer in the second round was 909U19. The second round product was cloned into pCR2.1-TOPO vector (Invitrogen).
Radiation Hybrid Mapping.
The novel sequence in the 3' RACE product B11 was selected for radiation hybrid mapping. The sequences for the upper (ect1U) and lower (ect1L) primers were 5'-ATTACTGCTTAAAATCTCACACTG-3' and 5'-CAAACCATATGTAGAAAATGCTC-3', respectively. DNAs from monochromosomal hybrid cell lines for chromosomes 12 (GM10868) and 14 (GM10479 and GM11535) were obtained from the National Institute of General Medical Sciences Human Genetic Cell Repository (Coriell Cell Repositories). Radiation hybrid mapping of the amplicon was performed using the GeneBridge 4 radiation hybrid mapping panel (Invitrogen), and results were submitted to the Whitehead Institute Center for Genome Research Mapping Server for chromosomal assignment.6
Isolation and Characterization of BAC 475C2.
A BAC clone containing genomic DNA corresponding to the novel chromosome 14-specific sequence in plasmid B11 was isolated by screening the CITB Human BAC DNA Pools (B&C Libraries), Release III (Invitrogen). The PCR primers and profile used to screen the pools were the same as that for radiation hybrid mapping described above. Partial sequencing of this clone was performed by high throughput sequencing of subclones generated by digestion with Sau3A in the Collis Genome Laboratory at Brigham and Women’s Hospital.
STR Analysis.
DNA samples isolated from tumors 166931 and 105274 were typed for 13 unique polymorphic STR loci using the AmpFlSTR Profiler Plus and COfiler kits (Applied Biosystems) and the amelogenin locus. STR polymorphisms were resolved on an ABI 377 DNA Sequencer and analyzed using GeneScan and Genotyper software packages.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Most Chromosome 12 Breakpoints Are 5' (Centromeric) in the HMGA2 Locus.
Mapping of myomas with chromosome 12 rearrangements identified a breakpoint located 5' (centromeric) in the HMGA2 locus in 24 of 38 (63%) cases. In all of these tumors, both 5' and 3' hybridization signals were colocalized on the derivative chromosome (e.g., the der(14)t(12;14)(q15;q23-24) in 22 of 27 myomas), indicating that HMGA2 had remained intact during rearrangement (Figs. 1, A and B
, and 2)
. The distance between the 5' untranslated region of HMGA2 and the chromosome 12 breakpoint was variable relative to the 5' end of HMGA2. In many myomas (e.g., 112427 and 153759), the chromosome 12 breakpoint was centromeric to the most 5' clone tested by FISH, indicating that the 5' breakpoint was >100 kb upstream. In five cases, the breakpoint was localized to within a genomic clone or between clones (Fig. 2)
. Our mapping of ST90-194 10 kb 5' of HMGA2 has been confirmed independently (9)
, representing the tumor with a 5' breakpoint closest to the 5' end of HMGA2 in this series.
|
|
and 2)
. Specifically, the chromosome 12 breakpoint was mapped between clones 145E1 and 142H1 in 90% of cells (top and middle rows in Fig. 1B
), whereas the rearrangement was mapped within clone 27E12 in the remaining 10% of cells (bottom row in Fig. 1B
). These results indicated that this myoma contained two lineages, each with rearrangements at opposite ends of the HMGA2 locus. To exclude the unlikely possibility of admixture with cells from another tumor, STR polymorphisms were compared from myomas 166931 and 105274, which both have a 3' rearrangement. Results of DNA profiling showed at most two alleles for each polymorphism tested. Therefore, the mapping results in myoma 166931 are due to two independent chromosomal translocations flanking HMGA2. Whether this case represents two independent tumors that grew together or the result of secondary rearrangements after neoplastic transformation is unknown.
Some Chromosome 12 Breakpoints Are 3' (Telomeric) in the HMGA2 Locus.
Nine myomas, including four with simple t(12;14) rearrangements, were mapped to the 3' region of HMGA2 (Figs. 1C
and 2)
. Five myomas had breakpoints clearly 3' of HMGA2 because the breakpoint mapped telomeric to cosmid 27E12. Three of these breakpoints were located even further 3' because they were also telomeric to cosmid 50D7. Of note, the chromosome 12 breakpoint was located within clone 27E12 in four myomas. Cosmid 27E12 includes a small portion of intron 3 through the end of the 3' untranslated region of HMGA2, and breakpoints mapping within this clone potentially may disrupt HMGA2.
A Minority of Uterine Leiomyomata Harbor more Complex Rearrangements of 12q15.
Mapping of eight cases with more complex chromosomal rearrangements suggests that two breakpoints involved the HMGA2 locus. One of these breakpoints mapped centromeric to the 5' region of HMGA2 and the other breakpoint mapped within or telomeric to the 3' region.
Myoma ST93-220 is an example of such a tumor with a complex chromosomal rearrangement. The karyotype of ST93-220, 46,XX,t(12;14)(q13;q32), was notable because the breakpoints were located in distinctly different bands on both derivatives in comparison to the typical t(12;14)(q15;q23-24). Hybridization with whole chromosome painting probes wcp12 and wcp14 revealed a complex rearrangement and indicated four breakpoints [two on der(12) and two on der(14)]. As shown in Fig. 1D
, a segment of chromosome 12 was inserted into chromosome 14 at q23-24, and the remaining part of chromosome 12 (q15-qter) was reciprocally translocated to 14q32. FISH with probes from the 5' and 3' regions of HMGA2 (Fig. 1D
, inset) indicated that the gene was uninterrupted during this complex rearrangement. In another FISH experiment (data not shown), BAC 475C2 colocalized with cosmid 27E12 on the der(14).
Another example of a complex rearrangement involving chromosomes 10, 12, and 14 was revealed by FISH of myoma 190464. Whole chromosome painting with wcp12, wcp14, and wcp10 confirmed insertion of a segment of chromosome 12 into 14q23-24 (Fig. 1E)
and revealed a reciprocal translocation between the der(14) and the short arm of chromosome 10 (Fig. 1F)
. Interestingly, two sets of signals from cosmid 27E12 were detected on the der(14) at the insertion site of the chromosome 12 material (Fig. 1F
, inset) and indicated duplication of HMGA2 at either end of the inserted segment. Another experiment with clone 142H1 (data not shown) confirmed that HMGA2 remained intact during this complex rearrangement.
Myoma 119620 also had a complex rearrangement involving chromosome 12 but not chromosome 14. FISH with wcp12 was consistent with a simple translocation (Fig. 1G)
. FISH with single copy probes, however, suggested a more complex rearrangement (Fig. 1H)
. Cosmid 27E12, which corresponds to the 3' region of HMGA2, hybridized to both der(1) and der(12) chromosomes, indicating that a breakpoint occurred in the 3' region of HMGA2. Paradoxically, the most telomeric probe (cosmid 50D7) hybridized to the der(12) chromosome, whereas the most centromeric probes (cosmids 142H1 and 145E1) hybridized to the der(1) chromosome. This hybridization pattern could be explained by paracentric inversion of chromosome 12 with breakpoints flanking HMGA2 centromeric to 145E1 and telomeric to 50D7, followed by a reciprocal translocation to chromosome 1 with a break of the inv(12)
in 27E12 (Fig. 1I)
. GTG-banding and FISH with BAC 475C2 (data not shown) indicated that RAD51L1 was not rearranged.
In the tumors with complex rearrangements, we never observed a hybridization pattern in which 142H1 and 27E12 (5' and 3' probes in the HMGA2 locus, respectively) mapped to wholly different derivative chromosomes. Such an observation would have been clear evidence that the breakpoint had fallen within HMGA2.
Analysis of a Uterine Leiomyoma with a Breakpoint 3' in HMGA2 by 3' RACE.
FISH mapping of myoma 105274 showed that the chromosome 12 breakpoint was within cosmid 27E12, which contains a portion of intron 3 through the 3' untranslated region of HMGA2. This finding raised the possibility that the exons encoding the DNA binding domains were fused to another gene such as RAD51L1. 3' RACE was performed to search for such transcripts. Of five cloned RACE products characterized (Fig. 3A)
, none were full-length transcripts of HMGA2, and all contained unusual (ectopic) sequence replacing either the fifth (last) or fourth and fifth exons. Thus, these transcripts included the three DNA binding domains. Novel sequences fused to HMGA2 were analyzed by BLAST and four (B1, B3, B8, and B12) were identical to intronic sequence in HMGA2. One 174-bp sequence (B1) was identical to a transcript (GenBank U29115) previously detected in a different leiomyoma cell line (LM-30.1/SV40; Ref. 3
). A second 763-bp sequence (B3, GenBank AF533652) contained a 271-bp segment identical to another previously reported novel transcript (GenBank U29113) from leiomyoma LM-30.1/SV40 (3)
. In addition, the first 357-bp of the 467-bp ectopic sequence in transcript B8 (GenBank AF533652) was identical to another previously reported sequence (GenBank U29112) found in leiomyoma LM-538/SV40 (3)
. The sequences in transcripts B3 and B8, however, extend an additional 492 and 110 bp, respectively. The longer ectopic sequences in B3 and B8 and their correspondence with genomic sequence suggested that both transcripts were the result of alternative splicing and use of different polyadenylation sites compared with those described previously. The ectopic sequences B1, B3, B8, and B12 would result in translation of an additional 24, 13, 23, and 1 novel amino acid residues each, respectively, beyond HMGA2 exon 4 for B1 and exon 3 for B3, B8, and B12 (see web site for translations).5
|
. This transcript consequently leaves HMGA2 coding sequence uninterrupted but truncated the 3' untranslated region. Amplification using primers ect1U and ect1L derived from the 241-bp novel sequence from transcript B11 and genomic DNA from monochromosomal hybrid cell lines indicated that the novel sequence was present on chromosome 14 (data not shown). This amplicon was mapped with a LOD score of >3.0 to chromosome 14 between WI-5202 and D14S277 with the GeneBridge 4 panel. BAC 475C2 was identified by screening pooled BAC clones and hybridized to the normal chromosome 14 and both derivative chromosomes in cell line 105274 (Fig. 1C)
. Subsequently, it was determined that BAC 475C2 is located within the RAD51L1 locus. The breakpoint on chromosome 14 maps between exons 7 and 8 in an 405-kb intron based on partial sequencing of the BAC clone and comparison to genomic sequence.
Assessment of HMGA2-RAD51L1 and RAD51L1-HMGA2 Fusion Transcripts.
Although a fusion transcript between HMGA2 and RAD51L1 was not detected in 105274 by a generic approach (i.e., 3' RACE), aberrant splicing using apparently cryptic sites within HMGA2 raised the possibility that other patterns of aberrant splicing might occur. We developed a panel of RT-PCR reactions with paired gene-specific primers to evaluate directly mRNA from 10 leiomyomata with t(12;14). Fig. 4
shows the primer pairs used to test for HMGA2, RAD51L1, RAD51L1-HMGA2, and HMGA2-RAD51L1 transcripts. In all myomas, transcripts for HMGA2 and RAD51L1 were detected regardless of the location of the chromosome 12 breakpoint. Fusion transcripts were found in only four samples. HMGA2-RAD51L1 fusion transcripts were detected in both myomas with rearrangements in the 3' region of HMGA2. As noted above, both tumors 105274 and 166931 also expressed transcripts that included the complete coding sequence of HMGA2. The reciprocal fusion transcript (i.e., RAD51L1-HMGA2) was not detected in three different reactions for either tumor. From the RT-PCR product obtained from myoma 166931, an additional product (650 bp) of larger size than expected (500 bp) was observed in addition to the expected product (data not shown). This larger product was only detected by Southern blotting (data not shown). Both products were cloned from a nested PCR and subsequently sequenced. The expected product (D30 in Fig. 3B
, GenBank AY138858) showed a fusion product joining exon 3 of HMGA2 to exon 8 of RAD51L1 in the correct reading frame. The novel 381-bp product (D15, GenBank AY138860; Fig. 3B
) joined a truncated 3' UTR of HMGA2 to exon 8 of RAD51L1. Similar to myoma 105274, this latter transcript would leave the coding sequence of HMGA2 intact and suggests that one of the chromosome 12 breakpoints in tumor 166931 resides within the 3' UTR.
|
. Products of nested reactions were cloned and sequenced (Fig. 3C)
, showing that they were identical to D30, which was isolated from tumor 166931. |
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Interestingly, fusion transcripts in which the 5' region of HMGA2 was joined to the 3' region of RAD51L1 were detected in both tumors tested in which an uncommon 12q15 breakpoint was 3' to HMGA2. Sequence derived from novel transcripts suggests that the breakpoints reside within the 3' untranslated region of HMGA2, and fusion transcripts containing the full coding sequence of HMGA2 were detected.
Fusion transcripts involving HMGA2 and RAD51L1 were not detected in six of eight (75%) myomas with 5' breakpoints, but HMGA2 and RAD51L1 transcripts were present in all tumors tested with t(12;14). With the breakpoint 5' (centromeric) relative to HMGA2, the entire gene should be translocated to the der(14), and all of the sequence telomeric to HMGA2 should be derived from chromosome 12. Exons 8–11 of RAD51L1 included in a potential HMGA2-RAD51L1 fusion product would be located either upstream (i.e., centromeric) to HMGA2 on the der(14) or on the der(12) remote from HMGA2, depending on the position of the chromosome 14 breakpoint. HMGA2 and RAD51L1 would be present in an orientation that precludes transcription of fusion products in either case. Yet we detected such fusion transcripts in two different leiomyomata (ST94-114 and ST98-773). Aside from technical factors, we considered three possible explanations. First, cytogenetic mosaicism can produce populations of cells with breakpoints both 5' and 3' in the HMGA2 locus within a single tumor. An example of this mechanism is illustrated by tumor 166931. However, no evidence of such mosaicism was noted in FISH experiments in either myoma. Nevertheless, a very low level of mosaicism might be detectable by RT-PCR. Second, an additional rearrangement might position the two genes in the requisite orientation. We found evidence of multiple breakpoints in 7 of 38 (18.4%) myomas with rearrangements involving 12q15. There was, however, no evidence of secondary rearrangements in either tumor. Lastly, transcripts from HMGA2 and RAD51L1 may combine by trans-splicing (10) . On the basis of the available data, it is not possible to make a compelling argument for one mechanism over another. Despite these exceptions, formation of fusion transcripts in leiomyomata with the most frequent rearrangement of 12q15 [i.e., breakpoints 5' in HMGA2 in t(12;14)] is not the most common finding. FISH mapping of karyotypically complex cases reveals two breakpoints on chromosome 12 flanking HMGA2 and raises for consideration the biological consequence of inserting HMGA2 into a wholly new chromosomal context.
Previously, fusion transcripts between HMGA2 and RAD51L1 have been reported by several groups. Some have hypothesized that the key pathobiological mechanism operating in leiomyomata with t(12;14) involve fusion products between HMGA2 and RAD51L1. Schoenmakers et al. (3) first recognized that the chromosome 14 breakpoint fell within the large RAD51L1 locus and detected reciprocal fusion products between the two genes in five SV40 transformed leiomyoma cell lines. In contrast, Ingraham et al. (4) did not find fusion transcripts in leiomyoma GM10964, which is the same tumor as ST90-194. Of note, HMGA2 is uncommonly fused to several other sequences, including COX6C on chromosome 8, a RTVL-H 3' LTR and mitochondrial dehydrogenase gene (ALDH2) on chromosome 12, and enhancer of invasion 10 (HEI10) on chromosome 14 at band q11 (11, 12, 13, 14) . Several other leiomyomata in our series involved other chromosomal partners, suggesting that various genes potentially may infrequently be fused with HMGA2.
It is possible that such rearrangements could still form a fusion transcript in which 5' exons from RAD51L1 could be spliced to exons 2–5 of HMGA2, skipping over exon 1 of HMGA2. Indeed, Takahashi et al. (5)
have detected such transcripts in 2 of 81 myomas. Tumors in that series were not karyotyped but only 10% would have been expected to have had 12q15 rearrangements. The resulting polypeptides would lack either the first or first and second AT hooks of the HMGA2 protein. HMGA2 transcripts, however, were detected in both leiomyomata, raising a question about the pathobiological significance of truncated HMGA2 or RAD51L1-HMGA2 transcripts. Interestingly, HMGA2-RAD51L1 fusion transcripts were not reported in that series. In myomas specifically selected for presence of t(12;14), we did not detect such anomalous splicing in RT-PCR reactions designed to detect RAD51L1-HMGA2 fusion transcripts. The most frequent finding in leiomyomata with chromosomal rearrangements involving 12q15 is aberrant expression of transcripts containing the full coding sequence for HMGA2. The resultant polypeptides would be indistinguishable from normal by size and presence of specific epitopes, as has been shown by Klotzbucher et al. (Ref. 15
and data not shown). Furthermore, dysregulated expression of the HMGA2 polypeptide also is the expected molecular mechanism underlying myomas with trisomy 12.
In addition to fusion transcripts containing the first three exons of HMGA2 joined to RAD51L1, we observed several truncated HMGA2 transcripts lacking the fourth and fifth exons or only the fifth exon. All of these transcripts were found in the same tumor (105274), which also made transcripts capable of encoding a complete HMGA2 polypeptide. Two of the fusion transcripts were similar to those previously reported by Schoenmakers et al. (3) . All of these truncated transcripts included exons 1–3, encoding the AT-hook DNA binding domains translated in the HMGA2 polypeptide (3 , 16 , 17) . The 3' sequence of these transcripts was found in chromosome 12, suggesting the possibility that they are cryptic exons hidden in the large third or smaller fourth introns or illegitimately spliced products (18) . These sequences contribute as little as one and as many as 23 additional amino acid residues to the truncated polypeptide. The biological activity of HMGA2 polypeptides in which the acidic tail has been replaced by a variable number of seemingly random amino acid residues, either in the presence or the absence of the normal polypeptide, is not understood presently in human uterine leiomyomata (15) . Overexpression of similar transcripts in transgenic mice, however, is not associated with leiomyomata but rather is associated with atypical lipomatous proliferations (19) .
Two molecular mechanisms might account for aberrant expression of HMGA2 after t(12;14) in uterine leiomyomata. First, a regulatory element that ordinarily negatively regulates HMGA2 expression in adult myometrial tissue might be separated from the gene by chromosomal rearrangement. This hypothetical mechanism would be plausible if the regulatory element was located 5' to HMGA2. Analysis of the genomic sequence 5' to HMGA2 does not indicate such an element, but gaps persist in the genomic sequence and the features by which such an element might be recognized are not well defined. A more serious criticism of this model is that it is not comprehensive. Specifically, simple or complex rearrangements with breakpoints 5' with respect to HMGA2 would remove the gene from the putative regulator, but the infrequent rearrangements located 3' relative to HMGA2 would not alter the relationship between the gene and its regulator. It is unlikely that rearrangements on either side of the gene could abrogate a single regulatory element.
A second mechanism would be the placement of a foreign regulatory element near HMGA2 that overcomes silencing of this gene observed in normal adult tissues. Such regulatory elements might include an enhancer of gene expression. The paradigm for this mechanism is MYC and the reciprocal translocations between chromosome 8 and chromosomes that harbor immunoglobulin loci expressed in hematopoietic cells. In the case of MYC, modest increases in mRNA can have profound biological consequences (20) . Another interesting paradigm is the t(3;8) found in pleomorphic adenoma of the salivary gland. This rearrangement occurs in the 5' noncoding regions of PLAG1 and ß-catenin (CTNNB1) and results in exchanging promoters and upstream elements between a developmentally regulated gene (PLAG1) and a constitutively expressed gene (CTNNB1; Ref. 21 ). Consequently, PLAG1 is inappropriately expressed. Such promoter swapping is unlikely in uterine leiomyomata because 3' rearrangements, even if uncommon, would not involve the promoter for HMGA2. FISH mapping of 14q23-24 breakpoints in leiomyomata with 12q15 breakpoints 5' to HMGA2 also suggests that the interval between the RAD51L1 promoter and the HMGA2 transcriptional start site is impressively large for promoter swapping. A mechanism invoking an enhancer, however, would not be constrained by variation in the orientation or distance between the chromosome 12 breakpoint and HMGA2. This suspected enhancer may potentially be a specific discrete feature contained within RAD51L1 before rearrangement. Examination of the genomic sequence of RAD51L1 does not reveal presence of another gene within RAD51L1 that might supply an enhancer. Alternatively, enhancement of HMGA2 after translocation may be a result of altered chromatin structure transmitted from the translocation partner.
RAD51L1 per se does not play an exclusive pathobiological role in uterine leiomyomata. In most cases, the 12q15 breakpoint occurs 5' relative to HMGA2, precluding formation of fusion genes with RAD51L1 excluding exon skipping. We detected evidence of novel fusion transcripts in two cases, both of which had uncommon 12q15 breakpoints located 3' relative to HMGA2. Fusion transcripts were detected in two of eight tumors with typical rearrangements 5' in the HMGA2 locus and their origin is unclear. Although misspliced products containing exons 1–3 may contribute in the pathobiology of uterine leiomyomata, our findings suggest that transcripts from an intact dysregulated HMGA2 gene are the primary molecular result of the 12;14 translocation in these tumors, distinct from the fusion transcripts produced by diverse translocations involving 12q15 in lipoma. Investigations into the molecular mechanisms underlying dysregulated expression of HMGA2 in t(12;14) and in other genomic rearrangements involving HMGA2 in uterine leiomyomata will be necessary to gain an understanding of the complex biology underlying the genesis, growth, and development of these common benign neoplasms.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by the National Cancer Institute, NIH Grants R01 CA078895 (to C. C. M.) and K08 CA72594 (to B. J. Q.) and a grant from Assessorato Igiene e Sanitá Regione Autonoma Sardegna (to R. V.). 
2 To whom requests for reprints should be addressed, at Department of Pathology, Brigham and Women’s Hospital, 75 Francis Street, Boston, MA 02115. Phone: (617) 732-7980; Fax: (617) 738-6996; E-mail: cmorton{at}partners.org
3 Internet address: cgap.nci.nih.gov/Chromosomes/Mitelman.
4 The abbreviations used are: FISH, fluorescence in situ hybridization; RT-PCR, reverse transcription-PCR; RACE, rapid amplification of cDNA ends; BAC, bacterial artificial chromosome.
5 Internet address: fibroids.net/research/hmga2_supplement.htm.
6 Internet address: www-genome.wi.mit.edu/cgi-bin/contig/rhmapper.
Received 9/23/02. Accepted 1/31/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
locus in Burkitt. Proc. Natl. Acad. Sci. USA, 80: 6922-6926, 1983.This article has been cited by other articles:
|
|
 |
|
  Y. Peng, J. Laser, G. Shi, K. Mittal, J. Melamed, P. Lee, and J.-J. Wei Antiproliferative Effects by Let-7 Repression of High-Mobility Group A2 in Uterine Leiomyoma Mol. Cancer Res., April 1, 2008; 6(4): 663 - 673. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. S. Lee and A. Dutta The tumor suppressor microRNA let-7 represses the HMGA2 oncogene Genes & Dev., May 1, 2007; 21(9): 1025 - 1030. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Hodge and C. C. Morton Genetic heterogeneity among uterine leiomyomata: insights into malignant progression Hum. Mol. Genet., April 15, 2007; 16(R1): R7 - R13. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Date, M. Katsura, M. Ishida, T. Yoshihara, A. Kinomura, T. Sueda, and K. Miyagawa Haploinsufficiency of RAD51B Causes Centrosome Fragmentation and Aneuploidy in Human Cells. Cancer Res., June 15, 2006; 66(12): 6018 - 6024. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. D. P. Moore, S. R. Herrick, T. A. Ince, M. S. Kleinman, P. D. Cin, C. C. Morton, and B. J. Quade Uterine Leiomyomata with t(10;17) Disrupt the Histone Acetyltransferase MORF Cancer Res., August 15, 2004; 64(16): 5570 - 5577. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Lehtonen, M. Kiuru, S. Vanharanta, J. Sjoberg, L.-M. Aaltonen, K. Aittomaki, J. Arola, R. Butzow, C. Eng, K. Husgafvel-Pursiainen, et al. Biallelic Inactivation of Fumarate Hydratase (FH) Occurs in Nonsyndromic Uterine Leiomyomas but Is Rare in Other Tumors Am. J. Pathol., January 1, 2004; 164(1): 17 - 22. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
