Forty-four malignant fibrous histiocytomas (MFHs) were studied by comparative genomic hybridization. Among the observed imbalances,losses of the 13q14–q21 region were observed in almost all tumors(78%), suggesting that a gene localized in this region could act as a tumor suppressor gene and that its inactivation could be relevant for MFH oncogenesis and/or progression. We determined by CA repeat analyses a consensus region of deletion focusing on the RB1region. The RB1 gene was then analyzed by protein truncation test, direct sequencing, fluorescence in situhybridization, Southern blotting, and immunohistochemistry. RB1 mutations and/or homozygous deletions were found in 7 of the 34 tumors analyzed (20%). Among the 35 tumors with comparative genomic hybridization imbalances analyzed by immunohistochemistry, 30 (86%) did not exhibit significant nuclear labeling. The high correlation between chromosome 13 losses and absence of RB1 protein expression and the mutations detected strongly suggest that RB1 gene inactivation is a pivotal event in MFH oncogenesis. Moreover, the observation of a high incidence of MFH in patients previously treated for hereditary retinoblastoma fits well this hypothesis.

MFH3is the most frequent soft tissue malignancy of late adult life. Its high incidence in patients previously treated for hereditary retinoblastoma suggests that both tumors could share a common oncogenic pathway. Despite extensive cytogenetic analyses, no consistent chromosome abnormalities have been described for this tumor. In a previous CGH study of a series of 30 MFHs, we have shown that these tumors are associated with recurrent profiles of DNA copy number imbalances (1). Among the observed imbalances, losses of the 13q14–q21 region were observed in almost all tumors, suggesting that a gene localized in this region could act as a tumor suppressor gene and that its inactivation could be relevant for MFH oncogenesis and/or progression. We have now confirmed on a larger series that these imbalance profiles are highly recurrent in MFH and that regional chromosome 13 losses are the most frequent chromosome abnormality in MFH. In an attempt to further characterize these chromosome 13 imbalances, we have undertaken a molecular analysis of a series of 44 MFHs using microsatellite loci dispersed along chromosome 13. A consensus region of deletion focusing on the RB1 region was thus delineated. The RB1 gene localized in this region was then analyzed by PTT, direct sequencing, FISH on nuclei, Southern blotting, and immunohistochemistry.

Tumor Samples.

Forty-four tumor samples were obtained from different pathology laboratories: 29 tumors were classified as storiform pleomorphic; 6 were classified as myxoid; and 1 was classified as inflammatory. For 8 tumors, a mixed phenotype or a more undifferentiated phenotype was observed. According to the recommendations of the FFCLCC, the grading was established for each of the 25 primary tumors (from grade I to grade III), but not for the 19 recurrences. Of the 25 primary tumors, 1 was grade I, 11 were grade II, and 13 were grade III. Eight tumors were obtained after an initial treatment. The main clinical and histopathological data are listed in Table 1. Twenty-four tumors were described previously (1). These cases are indicated in bold characters in Tables 1 and 2 and in Fig. 1. In four instances (cases 1, 3, 15, and 22), tumor-derived cell lines were obtained, and these cell lines are marked with an asterisk in Table 1. In two other instances (cases 18 and 40), normal fibroblasts were available.

CGH and Image Analysis.

CGH was performed as described previously (2). Tumor DNAs were labeled directly by nick translation using a FITC-dUTP nucleotide(DuPont New England Nuclear, Boston, MA). The nick translation conditions were adjusted to obtain DNA fragments ranging from 500-2000 bp. Spectrum Red normal male reference DNA (Vysis, Downers Grove, IL)was used as control DNA. Tumor DNA (400 ng), control DNA (400 ng), and Cot-1 DNA (80 μg; Life Technologies, Inc., Gaithersburg, MD) were mixed in 12 μl of hybridization buffer (50% formamide, 40 mm NaH2PO4,0.1% SDS, 10% dextran sulfate, and 2× SSC).

CGH pictures from 10 metaphases were captured using a Leica DMRB fluorescence microscope and a Photometrics Nu200 charge-coupled device camera and analyzed with Quips XL software (Vysis).

Two levels of imbalances were taken into account. The first level of imbalance corresponds to a range of 1.2–1.3 and 0.7–0.8 for gains and losses, respectively. The second level corresponds to a ratio of >1.3 for gains and <0.7 for losses. These levels are represented by two different thicknesses in Fig. 1.

FISH.

The probes used correspond to YAC 122F2, previously described as overlapping the RB1 genomic sequence (3), and to cosmids C108C0460, C108A03202, C108E0135 and C108H0170 precisely mapped in the RB1 genomic region (4). Two further YACs, 951A3 and 950D10, localized in bands 13q12.2 and 13q33,respectively, were used as control probes. YAC DNA was prepared by Alu-PCR amplification as described previously (5). Slides were denatured in 70% formamide/2× SSC at 70°C for 2 min and hybridized for 15 h at 37°C. They were washed in 50%formamide/2× SSC and then in 2× SSC at 42°C. Detection was achieved with a rhodamine-coupled anti-digoxigenin antibody (Boehringer,Mannheim, Germany) or a FITC-coupled anti-biotin antibody (Vector Laboratories, Burlingame, CA). Nuclei were counterstained using a 4′,6-diamidino-2-phenylindole/antifade solution.

Analysis of Microsatellite Markers.

Loss of the chromosome 13 long arm was investigated using microsatellite loci distributed along this chromosome. Microsatellite markers with a high percentage of heterozygosity were selected from the Généthon genetic map of human chromosome 13 established by linkage analysis (6). In the first step, tumor DNA was amplified by PCR with specific sets of primers at 24 microsatellite loci situated on the long arm of chromosome 13, including D13S153,which is localized in RB1 intron 2. In a second step, the minimal region of interest was refined on a subset of eight potentially informative tumors with four new microsatellite loci. In brief, 25 ng of DNA were amplified in a volume of 20 μl using 6.6 pmol of each primer, 25 μm of each deoxynucleotide triphosphate, and 0.4 unit of Taq DNA polymerase (Perkin-Elmer Cetus,Branchburg, NJ). PCR products were analyzed on a 6% denaturing polyacrylamide gel containing 7 m urea and 32%formamide that was transferred to a nylon N+membrane (Amersham, Les Ulis, France) and revealed by hybridization with a (CA)12 radioactive probe as described previously (7). Blots were analyzed by two independent observers and on two independent occasions. In case of discrepancy for a given locus, the result was considered as noninterpretable.

Southern Blot Analysis.

High molecular weight DNA was extracted from tumor samples using standard procedures and digested with EcoRI restriction enzyme, and the resulting fragments were separated according to size by electrophoresis on 0.8% agarose gels. After alkaline transfer to Hybond N+ membranes (Amersham), samples were hybridized with 32P-radiolabeled probes. RB1 probes were prepared by reverse transcription-PCR amplification of exons 1–12 (probe 1), 8–21 (probe 2), and 20–27(probe 3) using the following primer pairs (forward and reverse,respectively, 5′ to 3′): (a) AAAGGCGTCATGCCGCCC and CCTGACAATACTTGTGATAGG (probe 1); (b) AACAGGAGTGCACGGATAGC and CCGTATACGTTTCACTTCTT (probe 2); and (c)TGCAGAATGAGTATGAACTC and CGTATATCCACTACAAACG (probe 3).

A control probe corresponding to a 3.4-kb genomic segment from cosmid 29G2 (LL22NCO3 cosmid bank) and recognizing a 23-kb EcoRI band was systematically hybridized on the same blots.

PTT.

Total RNA was extracted from human tumor samples using Trizol (Life Technologies, Inc., Rockville, MD) according to the manufacturer’s specifications. First-strand cDNAs were generated using the First-Strand cDNA Synthesis kit according to the instructions of the manufacturer (Pharmacia, Orsay, France). The coding sequence of the RB1 gene was divided into three overlapping regions:(a) region 1 (codons 1–381); (b) region 2(codons 257–715); and (c) region 3 (codons 614–928). Forward primers were designed to include a T7 promoter sequence for the initiation of transcription by T7 RNA polymerase, as well as a translation initiation site. The primer sequences were 5′-AAAGGCGTCATGCCGCCC-3′ (forward) and 5′-GGATAGTGTTCATAACAGTCC-3′(reverse) for region 1, 5′-AACAGGAGTGCACGGATAGC-3′ (forward) and 5′-TTCTTCACTTTGCATATGCC-3′ (reverse) for region 2, and 5′-AAAGGTTCAACTACGCGTG-3′ (forward) and 5′-GCAAACATCACCTATATGC-3′(reverse) for region 3. PCRs were carried out in a total reaction volume of 25 μl containing 1× PCR buffer (Perkin-Elmer, Norwalk,CT), 0.2 mm deoxynucleotide triphosphates, 7.5 pmol of each primer, 1.5 mmMgCl2, 1 unit of AmpliTaq polymerase(Perkin-Elmer), and 1 μl of reverse transcription reaction. For region 1, 5% (v/v) DMSO was added. Initial denaturation was performed at 94°C for 2 min. All PCR cycles had a denaturation step at 94°C for 1 min, an annealing step for 1 min, and an extension step at 72°C for 2 min. Cycles were repeated 35 times. The PTT assay was performed directly from the PCR products using the TNT T7 Coupled Reticulocyte Lysate System as recommended by the manufacturer (Promega, Lyon,France). 35S-labeled protein products were separated on a 12% SDS-polyacrylamide gel and detected by autoradiography. Prestained protein markers were supplied by Bio-Rad(Hercules, CA) to size the protein products.

Sequencing.

PCR products showing an abnormal pattern in the PTT analysis were sequenced directly using the dRhodamine Terminator Cycle Sequencing kit on an ABI 377 automatic sequencer (PE Applied Biosystems, Foster City,CA). The RB1 reference sequence corresponds to GenBank accession number L11910.

Immunohistochemistry.

Tissue sections were obtained from paraffin blocks. They were deparaffinized in xylene and then rehydrated in a series of ethanol baths. An antigen retrieval technique was used: slides were incubated in 10 mmol of citrate buffer (pH 6.1) for 20 min in a microwave. Rb 1F8, a mouse monoclonal antibody against RB1 protein (Zymed, San Francisco, CA), was applied at a dilution of 1:50. Immunohistochemistry was performed with an avidin-biotin peroxidase complex (Vectastain Elite ABC; Vector Laboratories) using biotinylated goat antimouse antibody and 3,3′-diaminobenzidine as chromogen. Slides were counterstained with Mayer’s hematoxylin (RAL) and dehydrated. Normal tissue, which is known to express the RB1-encoded protein, was used as control.

Chromosome 13 Imbalances Detected by CGH.

Eight tumors exhibit normal CGH profiles (cases 37, 38, 39, 40, 41, 42,43, and 44), suggesting either a normal or balanced karyotype or a large amount of contaminating normal cells. The 36 other tumors show abnormal CGH profiles (data not shown); 28 of these tumors had significant chromosome 13 loss (78%). Five of these tumors exhibit both loss of the proximal part and gain of the distal part of the chromosome 13 long arm (Fig. 1). Among these 28 tumors, 26 exhibit loss of the 13q14 and/or 13q21 bands. The low resolution of the CGH technique (one chromosome band) led us to hypothesize that a common region of losses could exist in between these two contiguous bands.

Microsatellite Marker Analysis.

MFH samples are frequently contaminated by various amounts of infiltrating nontumor cells. As shown in Fig. 2,A, in these tumors, deletion or amplification of one allele appears as allelic imbalance. It is also noticeable that in such contaminated tumors, homozygous deletions can mimic a balanced heterozygous status (due to the amplification of DNA from the normal contaminating cells; Fig. 2,B). For most of the cases, we do not have DNA from nontumor cells. Several allele patterns were thus observed: (a) noninformative homozygosity of the microsatellite marker, corresponding either to a deletion of one allele(in noncontaminated samples) or to alleles of similar size;(b) balanced intensity of the two alleles, indicating neither loss nor gain of DNA at this level; (c) unbalanced intensity of the alleles, reflecting either an overrepresentation (in contaminated and noncontaminated samples) or a deletion of one allele(in contaminated samples); and (d) noninterpretable images due to weak alleles, superimposed ladders, or the presence of additional bands in the tumor DNA. For four cases, tumor cell lines were obtained. Comparison between matched tumor and cell line samples allowed us to confirm that the allelic imbalances observed in tumor samples actually corresponded to losses of heterozygosity in the corresponding cell lines. In two instances, normal fibroblasts were available, allowing us to demonstrate that some homozygous status observed in noncontaminated tumor samples did indeed correspond to LOH. The overall results obtained on the 44 samples at 24 microsatellite loci are presented in Table 1. Seven tumors do not exhibit any allelic imbalance or contiguous regions of homozygosity suggestive of chromosome 13 deletion. Six additional tumors show allelic imbalance at a single locus. It is noticeable that among these 13 tumors, 8 correspond to cases with normal CGH profiles, and 4 correspond to normal chromosome 13 profiles. All other tumors and cell lines exhibit clear allelic imbalances or LOH, respectively, at two or more contiguous loci, allowing the delineation of a common region of interest including D13S164, D13S153, and D13S284 (allelic imbalance in 68%, 72%, and 68% of the informative data, respectively). This small region, which contains the RB1 genomic sequence, was further refined by the analysis in eight particularly informative tumors of four other microsatellite loci mapping to this region (Table 2). Four tumors (cases 8, 11, 12, and 25) exhibit two noncontiguous regions of allelic imbalance, suggesting either the presence of two noncontiguous deletions or the occurrence of independent deletions on both chromosomes 13, with a common region of overlap leading to a homozygous deletion at the corresponding loci. Nevertheless, the smallest region of overlap of imbalances is between D13S153 and D13S284, a region of 0 cM (lod score, 35; τ = 0; CEPH database of genotypes4).

FISH.

Fourteen cases were studied by FISH to confirm CGH data and to refine a small region of imbalances. Two cases (cases 3 and 15) were studied on tumoral metaphases, 4 cases (cases 7, 20, 31, and 32) were studied on interphase nuclei isolated with a Dounce homogenizer, and 8 cases(cases 1, 8, 10, 11, 18, 19, 25, and 27) were studied on 5-μm-thick frozen tissue sections. For 6 cases (cases 1, 8, 10, 19, 31, and 32),no copy number differences were observed between the RB1locus and control loci (corresponding to either 951A3 or 950D10). Heterozygous deletions suspected from CGH profiles or observed on CA repeat analyses were confirmed in cases 3, 15, 18, 20, and 27. In case 7, two signals were observed for the proximal control locus (951A3) and for RB1 (122F2; Fig. 3,a), and six signals were observed for the distal control locus (950D10; Fig. 3,b). This result correlates perfectly with the CGH data. Thus, allelic imbalances observed for this tumor in CA repeat analysis on the whole chromosome 13 long arm correspond to losses for the proximal part and to gains for the distal part. Finally,two homozygous deletions of the RB1 locus (122F2 probe) were detected in cases 11 (Fig. 3,a) and 25 (Fig. 3 b). These homozygous deletions, which has been suspected in the CA repeat analyses, were confirmed in three independent FISH experiments.

Southern Blot.

In case 21, an abnormal higher band containing exon 20 was detected with probe 3. For case 25, bands detected by the three probes appear slightly fainter than those observed in other tumors, as compared with the bands detected by the control probe (Fig. 4). The DNA corresponding to case 11 was of too poor a quality for Southern blot analysis. No further DNA was available for this tumor.

RB1 Mutations Detected by PTT.

Among the 44 cases, we could not perform the PTT analysis in 9 cases(absence of tumor sample or low quality RNA). The RB1 coding sequence was divided in three overlapping regions. Abnormal patterns were detected in five cases: case 10 in region 1; cases 3, 21, and 36 in region 2; and cases 7 and 21 in region 3 (Fig. 5).

Case 10 showed an altered size protein product from fragment 1, as well as the normal-sized protein. The reverse transcription-PCR fragments were sequenced and showed a deletion of 1 bp in exon 7 (codon 216-delT). The frameshift created by this deletion led to a stop codon 6 bases downstream. The predicted protein was a truncated protein of 216 aa, plus 2 aa added after the frameshift (about Mr 25,000 versusMr 44,000 for the normal protein product).

Cases 3 and 36 showed the same abnormal pattern of PTT in region 2,with an approximately Mr 49,000 product compared with the expected Mr53,000 protein. The amplified cDNA fragment corresponding to this region was abnormally short, and its sequence showed a 117-bp deletion encompassing exon 13. This deletion created no frameshift and no abnormal codon, and the predicted protein, which was 39 aa shorter than the wild-type protein, was compatible with the PTT result. For both cases, the region containing exon 13 was amplified from genomic DNA and sequenced. Case 3 showed a single G to T transversion of the last nucleotide of exon 13. This transversion could interfere with the splice donor site and explain the observed exon skipping. For case 36,a deletion of 25 bp encompassing the splice donor site was detected,resulting in exon skipping.

A protein product slightly smaller than normal (about Mr 32,000 versusMr 36,000) was observed in region 3 for tumor 7. Direct sequencing of PCR products showed a mixture of cDNA species: in addition to the normal sequence, one product corresponds to a splicing of exon 22 (deletion of 114 bp); and the second corresponds to a deletion of the first 25 bp of exon 22. The sequencing of the relevant genomic region showed a single A to C transversion in position −2 of the exon 22 acceptor site. This mutation putatively explains the observed exon skipping and is compatible with the PTT result. On the other hand, the 25-bp deletion in exon 22 generated a truncated protein of 130 aa corresponding to a product of Mr 15,000, which could not be detected in a PTT assay. At the genomic level, this shorter cDNA could correspond to the involvement of a potential acceptor site located 25 bp downstream of the normal site.

Case 21 showed an altered pattern of PTT in regions 2 and 3, suggesting that a mutation occurred in the overlapping region. PCR products corresponding to region 2 showed the normal band and a higher band. After elution and purification, this abnormal band was sequenced,showing a duplication of exons 18 and 19. The frameshift created by this insertion led to a stop codon 15 bases downstream. The predicted protein was a truncated protein of 397 aa, plus 5 aa added after the frameshift. This protein product for region 2 was compatible with the PTT result (about Mr 44,000 versusMr 53,000 for the normal protein). On the other hand, interpretation of the PTT result for region 3 was more difficult.

Immunohistochemistry.

Among the 36 tumors with abnormal CGH profiles, 35 were studied by immunohistochemistry for RB1 protein expression using a monoclonal antibody. Only tumor cells were taken into consideration. Thirty of the tumors (86%) do not exhibit significant nuclei labeling(no more than 10% of positive nuclei). Five tumors were positive in various percentages of cells (100%, 90%, 70%, 50%, and 30% for cases 35, 30, 15, 28, and 32, respectively). Among the eight tumors with normal CGH profiles, seven were analyzed. One of them was strongly positive but corresponded to normal tissue (case 44) and was thus classified as ND in Table 1. Four of them were negative for RB1 staining, and one was positive in 30% of nuclei (case 40). Case 37 exhibits a signal in both the nucleus and cytoplasm. One positive tumor (case 30) and one negative tumor (case 36) are shown in Fig. 6.

Chromosome 13 loss is the most common genomic imbalance in MFH, as detected by CGH analysis in 78% of tumors (Ref. 1and the present study). In an attempt to further characterize the smallest region of overlap of the chromosome 13 losses detected in our MFH series, we performed a CA repeat analysis of these tumors. Among the 36 tumors with abnormal CGH profiles, 33 exhibit allelic chromosome 13 imbalances or LOH, delineating a region of interest including markers D13S164, D13S153, and D13S284. The region was secondarily refined by the analysis of four additional polymorphic markers on eight informative tumors. Finally, imbalance at the D13S153 locus, localized in RB1 intron 2, appears as the most frequent event. Some tumors show discontinuous regions of allelic imbalances. Two possibilities could account for these data: (a)noncontiguous regions of deletion and/or amplification; and(b) homozygous deletions in contaminated samples. The first possibility was demonstrated by FISH in case 7, in which the proximal part of chromosome 13 was underrepresented, and the distal part was overrepresented (Fig. 3). The second possibility was demonstrated in tumors 11 and 25, for which no FISH signals were detected in tumor cells with the RB1-containing probe 122F2,whereas it was present in small non-tumor cells (Fig. 3). This result led us to conclude that, in these tumors, heterozygosity retention at the D13S153 locus does indeed correspond to homozygous deletions. To demonstrate that the inactivation of the RB1gene was the target of chromosome 13 losses, we searched for inactivating mutations of the second allele by the PTT. Five abnormal PTT patterns were detected among the 33 tumors analyzed with abnormal CGH profiles. Sequencing of the cDNA PCR products and of the genomic DNAs confirmed that these abnormal patterns corresponded to truncating mutations. These mutations are spread over the RB1 coding sequence, except for two of them that result in exon 13 skipping. Thus,PTT and FISH analyses led us to detect RB1 inactivation in 20% of the tumors (7 of 34 tumors). RB1 mutations and/or homozygous deletions have been observed previously in small MFH series(8, 9). This low mutation rate in a series of tumors that almost all exhibit CGH losses and/or allelic imbalances at the RB1 locus prompted us to analyze RB1 protein expression by immunohistochemistry on tumor sections. As shown in Table 1 and Fig. 6, a very large proportion of tumors do not exhibit significant RB1 protein expression in tumor cells, as reported previously in small MFH series (10, 11). Of the 35 tumors with CGH imbalances analyzed by immunohistochemistry, only 5 (cases 15,28, 30, 32, and 35) show expression of the protein. Among them, loss of chromosome 13 was detected by both CGH and allelic imbalance at the RB1 locus for case 15; case 28 does not exhibit chromosome 13 loss as detected by CGH, and cases 30, 32, and 35 have neither CGH loss nor allelic imbalance. This high correlation between chromosome 13 losses and absence of RB1 protein expression suggests that the mutation frequency could be higher than detected by our approach or that the second inactivating event is not a mutation. Previous analyses performed on retinoblastoma, a tumor for which RB1 inactivation is the oncogenic event, show that both possibilities are observed: on one hand, the mutation rate detected in retinoblastoma (12) is also relatively low and in the same range as the one we detected in MFH; and on the other hand, RB1 inactivation by hypermethylation has also been reported in 10–13% of tumors (13, 14). The latter possibility could explain the very low RB1 protein expression (≤10% of positive cells) observed in a proportion of tumors, as compared to the complete absence in tumors with inactivating mutations (Fig. 1). It is also of importance to note that MFHs are often contaminated by nontumor cells, a situation that can complicate CGH, Southern blotting, PTT, and LOH analyses. It is noteworthy to observe that among the tumors with normal CGH profiles and without allelic imbalances (and thus supposedly containing a large amount of contaminating cells), five tumors do not express a normal RB1 protein pattern in tumor cells.

All these data strongly suggest that the RB1 locus is the target of chromosome 13 losses in MFH and that biallelic inactivation of this gene is a major (if not the primary) event in the oncogenesis of these tumors. The high incidence of MFH in patients previously treated for hereditary retinoblastoma in irradiated or possibly in nonirradiated areas clearly reinforces this hypothesis (15, 16). The detection of a RB1 germ-line mutation in a patient with a MFH and without retinoblastoma would be a definitive demonstration of the primary involvement of RB1 inactivation in MFH oncogenesis.

Fig. 1.

CGH results of chromosome 13 imbalances. Cases in bold correspond to previously published tumors.

Fig. 1.

CGH results of chromosome 13 imbalances. Cases in bold correspond to previously published tumors.

Close modal
Fig. 2.

Examples of microsatellite analyses. A, T, tumor DNA; L, cell line DNA; F, normal fibroblast DNA; •, LOH; d,allelic desequilibrium. Allelic imbalances (d) of tumors 1 and 15 are confirmed by the analysis of corresponding tumor cell lines exhibiting complete LOH. In cases 18 and 40, allelic imbalances observed in tumor samples are compared with the heterozygous status of corresponding normal fibroblasts. B, multiplex PCR amplification of markers D13S153 (RB1 intron 2) and D13S1265 (CEPH map section VIII) suggesting the occurrence of a homozygous deletion of the D13S153 locus in tumor 25. Faint bands(arrow) correspond to amplification of contaminating DNA from normal cells.

Fig. 2.

Examples of microsatellite analyses. A, T, tumor DNA; L, cell line DNA; F, normal fibroblast DNA; •, LOH; d,allelic desequilibrium. Allelic imbalances (d) of tumors 1 and 15 are confirmed by the analysis of corresponding tumor cell lines exhibiting complete LOH. In cases 18 and 40, allelic imbalances observed in tumor samples are compared with the heterozygous status of corresponding normal fibroblasts. B, multiplex PCR amplification of markers D13S153 (RB1 intron 2) and D13S1265 (CEPH map section VIII) suggesting the occurrence of a homozygous deletion of the D13S153 locus in tumor 25. Faint bands(arrow) correspond to amplification of contaminating DNA from normal cells.

Close modal
Fig. 3.

a, chromosome localization of probes 122F2(RB1 locus; red) and control probes 950D10 (green). FISH analysis of tumor 7 illustrating the overrepresentation of the distal part of chromosome 13, and FISH analysis of tumor 11 demonstrating the homozygous deletion of RB1 locus. Inset, a contaminating normal cell exhibiting both RB1 and control signals. b, chromosome localization of probes 122F2(RB1 locus; red) and control probes 951A3(green). FISH analysis of tumor 7 illustrating the presence of two copies of the proximal part of chromosome 13, and FISH analysis of tumor 25 demonstrating homozygous deletions of RB1 locus. Inset, a contaminating normal cell exhibiting both RB1 and control signals.

Fig. 3.

a, chromosome localization of probes 122F2(RB1 locus; red) and control probes 950D10 (green). FISH analysis of tumor 7 illustrating the overrepresentation of the distal part of chromosome 13, and FISH analysis of tumor 11 demonstrating the homozygous deletion of RB1 locus. Inset, a contaminating normal cell exhibiting both RB1 and control signals. b, chromosome localization of probes 122F2(RB1 locus; red) and control probes 951A3(green). FISH analysis of tumor 7 illustrating the presence of two copies of the proximal part of chromosome 13, and FISH analysis of tumor 25 demonstrating homozygous deletions of RB1 locus. Inset, a contaminating normal cell exhibiting both RB1 and control signals.

Close modal
Fig. 4.

Southern blot analysis of the RB1 gene. A, probe 2 hybridization. Tumor 25 (homozygous deletion)is compared with two other tumors and two control DNAs. Faint bands correspond to contaminating normal DNA. B,probe 3 hybridization. Tumor 21 exhibits an abnormal band, as compared with two other tumors and two control DNAs.

Fig. 4.

Southern blot analysis of the RB1 gene. A, probe 2 hybridization. Tumor 25 (homozygous deletion)is compared with two other tumors and two control DNAs. Faint bands correspond to contaminating normal DNA. B,probe 3 hybridization. Tumor 21 exhibits an abnormal band, as compared with two other tumors and two control DNAs.

Close modal
Fig. 5.

RB1 mutations detected by PTT. Top,schematic representation of the RB1 protein with the three fragments analyzed by PTT and their corresponding size in kilodaltons. Bottom, PTT analysis of the five cases showing abnormal patterns. CTL, control fibroblast sample showing a normal pattern. Cases 10, 3, 36, 21, and 7, MFH tumor samples showing truncated proteins. Dashes indicate wild-type full-length products, and arrows indicate abnormal bands.

Fig. 5.

RB1 mutations detected by PTT. Top,schematic representation of the RB1 protein with the three fragments analyzed by PTT and their corresponding size in kilodaltons. Bottom, PTT analysis of the five cases showing abnormal patterns. CTL, control fibroblast sample showing a normal pattern. Cases 10, 3, 36, 21, and 7, MFH tumor samples showing truncated proteins. Dashes indicate wild-type full-length products, and arrows indicate abnormal bands.

Close modal
Fig. 6.

Immunohistochemistry of RB1 protein on tumor 36(negative) and tumor 30 (positive).

Fig. 6.

Immunohistochemistry of RB1 protein on tumor 36(negative) and tumor 30 (positive).

Close modal

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

Supported by grants from INSERM, the Ligue Nationale Contre le Cancer (Comité de Paris and Comité de l’Oise), and the FEGEFLUC.

3

The abbreviations used are: MFH, malignant fibrous histiocytoma; FISH, fluorescence in situhybridization; CGH, comparative genomic hybridization; YAC, yeast artificial chromosome; PTT, protein truncation test; LOH, loss of heterozygosity; aa, amino acid(s).

4

World Wide Web address:http://www.cephb.fr/cephdb.

Table 1

Main clinical data, allelotyping results, RB1 mutations,immunohistochemistry, and FISH results on 44 MFH tumorsa

Main clinical data, allelotyping results, RB1 mutations,immunohistochemistry, and FISH results on 44 MFH tumorsa
Main clinical data, allelotyping results, RB1 mutations,immunohistochemistry, and FISH results on 44 MFH tumorsa
a

Tumors with data indicative of RB1 involvement are in gray squares. Bold cases correspond to previously published tumors. Clinical data: SP, storiform pleomorphic; MP, mixed phenotype; M, myxoid; I, inflammatory; PT,primary tumor; R, relapse; Tt, initial treatment; n, no treatment; CT,chemotherapy; NA, not available. Allelotyping: polymorphic markers are classified according to the CEPH-Généthon microsatellite map sections (from I to IX). •, loss of heterozygosity; ○,retention of heterozygosity; d, allelic disequilibrium; dash,noninformative; ni, noninterpretable. Gray squares correspond to LOH or allelic imbalance at the corresponding loci. The region of interest is boxed. RB1 mutations (as detected by PTT, Southern blotting,and sequencing): dash, no mutation detected; +, mutation; ND, not done. Immunohistochemistry: percentages of positive tumor cells. ND, not done. FISH results: del, hemizygous deletion; del/del, homozygous deletion; n del, no deletion. For these last three techniques, gray squares correspond to data indicative of RB1 inactivation.

Table 2

Polymorphic markers are classified according to the CEPH-Généthon microsatellite map sections (from II to V)a

Polymorphic markers are classified according to the CEPH-Généthon microsatellite map sections (from II to V)a
Polymorphic markers are classified according to the CEPH-Généthon microsatellite map sections (from II to V)a
a

○, retention of heterozygosity; •, loss of heterozygosity; d, allelic disequilibrium;dash, noninformative; ni, noninterpretable. Gray squares correspond to LOH or allelic imbalance at the corresponding loci.

We thank J. Lange for critical review of the manuscript.

1
Mairal A., Terrier P., Chibon F., Sastre X., Lecesne A., Aurias A. Loss of chromosome 13 is the most frequent genomic imbalance in malignant fibrous histiocytomas.
A comparative genomic hybridization analysis of a series of 30 cases. Cancer Genet. Cytogenet.
,
11
:
134
-138,  
1999
.
2
Kallionemi A., Kallionemi O. P., Sudar D., Rutovitz D., Gray J. W., Waldman F., Pinkel D. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors.
Science (Washington DC)
,
258
:
818
-821,  
1992
.
3
Cayanis E., Russo J. J., Kalachikov S., Ye X., Park S. H., Sunjevaric I., Bonaldo M. F., Lawton L., Venkatraj V. S., Schon E., Soares M. B., Rothstein R., Warburton D., Edelman I. S., Zhang P., Efstratiadis A., Fischer S. G. High-resolution YAC-cosmid-STS map of human chromosome 13.
Genomics
,
47
:
26
-43,  
1998
.
4
Cowell J. K., Jaju R., Kempski H. Isolation and characterization of a panel of cosmids which allows unequivocal identification of chromosome deletions involving the RB1 gene using fluorescence in situ hybridization.
J. Med. Genet.
,
31
:
334
-337,  
1994
.
5
Rosty C., Peter M., Zucman J., Validire P., Delattre O., Aurias A. Cytogenetic and molecular analysis of a t(1;22)(p36;q11.2) in a rhabdoid tumor with a putative homozygous deletion of chromosome 22.
Genes Chromosomes Cancer
,
21
:
82
-89,  
1998
.
6
Dib C., Faure S., Fizames C., Samson D., Drouot N., Vignal A., Millasseau P., Marc S., Hazan J., Seboun E., Lathrop M., Gyapay G., Morissette J., Weissenbach J. A comprehensive genetic map of the human genome based on 5,264 microsatellites.
Nature (Lond.)
,
380
:
152
-154,  
1996
.
7
Schleiermacher G., Peter M., Michon J., Hugot J. P., Vielh P., Zucke J. M., Magdelenat H., Thomas G., Delattre O. Two distinct deleted regions on the short arm of chromosome 1 in neuroblastoma.
Genes Chromosomes Cancer
,
10
:
275
-281,  
1994
.
8
Friend S. H., Horowitz J. M., Gerber M. R., Wang X. F., Bogenmann E., Li F. P., Weinberg R. A. Deletions of a DNA sequence in retinoblastoma and mesenchymal tumors: organization of the sequence and its encoded protein.
Proc. Natl. Acad. Sci. USA
,
84
:
9059
-9063,  
1987
.
9
Wunder J. S., Czitrom A. A., Kandel R., Andrulis I. L. Analysis of alterations in the retinoblastoma gene and tumor grade in bone and soft-tissue sarcomas.
J. Natl. Cancer Inst.
,
83
:
194
-200,  
1991
.
10
Karpeh M. S., Brennan M. F., Cance W. G., Woodruff J. M., Pollack D., Casper E. S., Dudas M. E., Latres E., Drobnjak M., Cordon-Cardo C. Altered patterns of retinoblastoma gene product expression in adult soft-tissue sarcomas.
Br. J. Cancer
,
72
:
986
-991,  
1995
.
11
Cohen J. A., Geradts J. Loss of RB and MTS1/CDKN2 (p16) expression in human sarcomas.
Hum. Pathol.
,
28
:
893
-898,  
1997
.
12
Blanquet V., Turleau C., Gross-Morand M. S., Senamaud-Beaufort C., Doz F., Besmond C. Spectrum of germline mutations in the RB1 gene: a study of 232 patients with hereditary and non-hereditary retinoblastoma.
Hum. Mol. Genet.
,
4
:
383
-388,  
1995
.
13
Greger V., Debus N., Lohmann D., Hopping W., Passarge E., Horsthemke B. Frequency and parental origin of hypermethylated RB1 alleles in retinoblastoma.
Hum. Genet.
,
94
:
491
-496,  
1994
.
14
Ohtani-Fujita N., Dryja T. P., Rapaport J. M., Fujita T., Matsumura S., Ozasa K., Watanabe Y., Hayashi K., Maeda K., Kinoshita S., Matsumura T., Ohnishi Y., Hotta Y., Takahashi R., Kato M. V., Ishizaki K., Sasaki M. S., Horsthemke B., Minoda K., Sakai T. Hypermethylation in the retinoblastoma gene is associated with unilateral, sporadic retinoblastoma.
Cancer Genet. Cytogenet.
,
98
:
43
-49,  
1997
.
15
Cole C. H., Magee J. F., Gianoulis M., Rogers P. C. Malignant fibrous histiocytoma in childhood.
Cancer (Phila.)
,
71
:
4077
-4083,  
1993
.
16
Imhof S. M., Moll A. C., Hofman P., Mourits M. P., Schipper J., Tan K. E. Second primary tumours in hereditary and nonhereditary retinoblastoma patients treated with megavoltage external beam irradiation.
Doc. Ophthalmol.
,
93
:
337
-344,  
1997
.