Detection of Integrated Murine Leukemia Viruses in a Mouse Model of Acute Myeloid Leukemia by Fluorescence In Situ Hybridization Combined with Tyramide Signal Amplification

Detection of Integrated Murine Leukemia Viruses in a Mouse Model of Acute Myeloid Leukemia by Fluorescence In Situ Hybridization Combined with Tyramide Signal Amplification Hasan Acar, Neal G. Copeland, Debra J. Gilbert, Nancy A. Jenkins, and David A. Largaespada

ABSTRACT: This study was undertaken to develop a reliable method to enumerate and map somatically acquired, clonal, murine leukemia virus (MuLV) proviral insertions in acute myeloid leukemia (AML) cells from the BXH-2 mouse strain. This was achieved by using fluorescence in situ hybridization combined with tyramide signal amplification (FISH-TSA) and an 8.8 kilobase pair (kb) full-length ecotropic MuLV or 2.0 kb MuLV envelope (env) gene probe. Two-color FISH was utilized combining chromosome-specific probes for regions near the telomere and/or centromere and the MuLV probes. The technique reliably detected germline and somatically acquired, tumor-specific, MuLV proviruses in BXH-2 AML cell lines. It was possible to readily verify homozygous insertions at endogenous ecotropic MuLV loci, Emv1 (chromosome 5), Emv2 (chromosome 8) and a BXH-2 strain-specific locus (chromosome 11). This strategy also verified the presence of molecularly cloned proviral insertions within the mouse Nf1 gene and another locus on distal chromosome 11, as well as on chromosome 7 and chromosome 9 in BXH-2 AML cell line B117. The technique was also used to detect several new tumor-specific, proviral insertions in BXH-2 AML cell lines. © 2000 Elsevier Science Inc. All rights reserved.

INTRODUCTION The identification of recurrent, clonal, proviral insertions in murine leukemia virus (MuLV)-induced hematologic malignancy has lead to the identification of many new cancer genes [1], including genes also involved in human leukemia [2, 3]. These models include forms of B and T cell malignancy, as well as myeloid leukemia [4–6]. In these disease models, the MuLV cause disease by acting as somatic cell, insertional mutagens. In rare somatic cells, the integrated provirus disrupts a tumor suppressor gene or activates a proto-oncogene. This leads to the clonal outgrowth of the affected cell, acquisition of additional onco-

From the University of Minnesota Cancer Center, Department of Genetics, Cell Biology and Development, University of Minnesota (H. A., D. A. L.), Minneapolis, Minnesota; the Department of Medical Genetics, Selcuk University (H. A.), Konya, Turkey; and the Mouse Cancer Genetics Program, NCI-Frederick Cancer Research and Development Center (N. G. C., D. J. G., N. A. J.), Frederick, Maryland, USA. Address reprint requests to: Dr. David A. Largaespada, University of Minnesota Cancer Center Research Building, 425 East River Road, Minneapolis, MN 55455. Received December 9, 1999; accepted February 11, 2000. Cancer Genet Cytogenet 121:44–51 (2000)  2000 Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010

genic proviral insertions, and eventual establishment of a highly malignant clone. Thus, the somatically acquired, clonal, proviral insertions in these tumors serve as valuable sign posts and molecular tags for the identification of cancer genes, not unlike the chromosomal abnormalities observed in human hematologic malignancies. For some models of MuLV-induced cancer, it has become clear that many different genes can be mutated by proviral insertion to give rise to disease. One such model is the BXH-2 mouse strain, which develops spontaneously a very high frequency of AML due to chronic infection with a horizontally-transmitted B-ecotropic MuLV [4]. In such cases, FISH-TSA-based detection of MuLV-insertions could be used to rapidly determine the number and approximate map position of proviruses among a series of tumors. In this way, common sites of proviral insertions could be revealed, prior to molecular cloning of insertion sites. Furthermore, the extensive synteny between the mouse and human genomes means that the map position of recurrent proviral insertions could be compared with the map position of recurrent chromosomal abnormalities in human malignancies that have been cataloged in large public databases [7]. In order to achieve these goals, a probe with sufficient specificity and signal strength is required. In this

0165-4608/00/$–see front matter PII S0165-4608(00)00232-6

Mapping MuLV by FISH-TSA report, we describe the utilization of FISH-TSA with the combination of retroviral probes and also chromosomespecific probes, in order to reliably detect the three endogenous MuLV present in BXH-2 strain mice and somaticallyacquired, tumor-specific MuLV in BXH-2 AML cell lines. METHODS Molecular Cloning of MuLV and MuLV Insertions Fifty micrograms of B117 genomic DNA were digested to completion with EcoRI and electrophoresed on a 1.0% Ultrapure LMP agarose gel (GIBCO-BRL, Gaithersburg, MD), and fragments of the appropriate size were cloned into EMBL4 lambda phage (Stratagene, La Jolla, CA). Positive plaques were identified with an ecotropic MuLV-specific probe (0.4 kb SmaI fragment from pEco, [8]) and DNA from positive clones subcloned into pBluescript (Stratagene). Clones were restriction enzyme mapped, and repeat-free probes flanking the proviruses were generated for each of four different clones. These probes were used in Southern blotting experiments, to detect rearrangements in B117 and for mapping the insertion sites by interspecific backcross analysis, performed as described before [9]. Probes were labeled by random priming [10], using the Prime-It-II labeling kit (Stratagene). The Southern blot probes derived from each clone are listed: B117c: 1.5 kb EcoRI-HindIII fragment B117d: 1.0 kb EcoRI-XbaI fragment B117e: 0.6 kb PstI fragment B117l: 1.5 kb PstI fragment Interspecific Backcross Mapping Interspecific backcross progeny were generated by mating (C57BL/6J ⫻ Mus spretus)F1 females and C57BL/6J males as described [11]. A total of 205 N2 mice were used to map the B117 loci (see text for details). DNA isolation, restriction enzyme digestion, agarose gel electrophoresis, Southern blot transfer, and hybridization were performed essentially as described [9]. The B117c probe, a 苲1.5 kb EcoRI/ HindIII fragment of mouse genomic DNA, detected a fragment of 3.6 kb in BamHI digested C57BL/6J (B) DNA and a fragment of 3.2 kb in M. spretus (S) DNA. The B117d probe, a 苲1.0 kb EcoRI/XbaI fragment of mouse genomic DNA, detected HindIII fragments of 4.5 kb (B) and 6.6 kb (S). The B117e probe, a 苲0.6 kb PstI fragment of mouse genomic DNA, detected SphI fragments of 8.0 kb (B) and 12.0 kb (S). The B117l probe, a 苲1.5 kb PstI fragment of mouse genomic DNA, detected PstI fragments of 9.4 kb (B) and 5.4 and 3.5 kb (S). Probes were labeled with ␣32P using random priming (Stratagene) or nick translation (Boehringer Mannheim) labeling kits. Washing was done to a final stringency of 0.25 to 1.0 ⫻ SSCP, 0.1% SDS, 65⬚C. The presence or absence of the M. spretus-specific fragment was followed in backcross mice. A description of the probes and restriction fragments length polymorphisms (RFLPs) for loci linked to B117 loci including En2, IL6, and Fgfr3, chromosome 5 [12, 13], Htrlb and Trf, chromosome 9 [14, 15], Snrpn and Igfrl,

45 chromosome 7 [16], and Cola1, Thra, and Wnt3, chromosome 11 [17] has been reported previously. One locus has not been reported for our interspecific backcross. A RASGRF1 probe, 苲600 bp EcoRI/SacI fragment of mouse cDNA, kindly provided by David Broek, detected SphI fragments of 8.6, 4.2, and 1.8 kb (B) and 4.0, 2.1, and 1.8 kb (S). The presence or absence of the 2.1 kb SphI M. spretusspecific fragment was followed in backcross mice. Recombination distances were calculated using Map Manager, version 2.6.5. Gene order was determined by minimizing the number of recombination events required to explain the allele distribution patterns. Mice, Bone Marrow Cells, and BXH-2 AML Cell Lines C57BL/6J mice were obtained from the Jackson Laboratory (Bar Harbor, ME) and housed in a specific pathogen-free facility. The establishment of BXH-2 AML cells lines and their culture conditions have been described before [2]. For optimization of the FISH-TSA technique, C57BL/6J and normal BXH-2 bone marrow cells were obtained from the femurs of adult mice and made into a single cell suspension by passage through a 27-gauge needle. The cells were plated in 10% fetal bovine serum-RPMI-1640 prior to generation of metaphase spreads. Exponentially growing cell lines B117, B132, and B186 were used to screen for endogenous and somatically acquired proviral insertions. Metaphase chromosome spreads were prepared by standard cytogenetic techniques. Fluorescence In Situ Hybridization-Tyramide Signal Amplification Slides were aged at room temperature (RT) for from three days to two weeks. Before hybridization, slides were treated with 5mM DTT (dithriotheitol) at RT for 5 to 10 minutes to decondense chromosomal DNA. Washing in 2 ⫻ SSC at RT stopped DTT treatment. The slides were dehydrated in washes of 70, 90, and 100% ethanol for 3 minutes each, and air-dried. The slides were finally treated with acetone for 10 minutes at RT. Slides were denatured in 70% formamide/2 ⫻ SSC for 2–4 minutes at 65⬚C, chilled with ice-cold 70% ethanol and dehydrated with sequential washes in 90 and 100% ethanol. For the verification of the chromosomes having endogenous or somatically acquired proviral insertions, P1 (PAC) probes near the centromeres or telomeres of chromosomes 1, 5, 7, 8, 9, and 11 (Genome Systems [18]), were used. The P1 DNAs were directly labeled with fluorescein-dUTP (Boehringer Mannheim) using a nick translation system (Gibco BRL). For the identification of integrated ecotropic MuLV, the entire pB117c plasmid probe was used. Alternately, primer pairs were designed, from near the 5⬘ (5⬘-GCT CTC AAG GTA GAC GGC ATC-3⬘) and 3⬘ ends of the Env gene (5⬘-GGG CCT GCA GTA CCG AAA T-3⬘), and used to amplify and clone a 2.0 kb fragment for use as a FISH probe. Both MuLV probes were labeled with either biotin-dATP (GIBCO BRL) or digoxigenin-dUTP (Boehringer Mannheim) using a nick translation system (GIBCO BRL). The probe mixtures contained 150 ng of chromosomespecific P1 probe, 100 ng of pB117c plasmid probe, 5 ␮g

46

H. Acar et al.

unlabeled salmon sperm DNA, and 6 ␮g unlabeled mouse Cot-1 DNA in 15 ␮l 50% formamide/2 ⫻ SSC/10% dextran sulfate. The probe mixtures were denatured at 65⬚C for 10–15 minutes and then incubated at 37⬚C for one hour. Then the probe mixtures were added to denatured slides and coverslipped. Slides were sealed with rubber cement, and the hybridization occurred, in a water bath at 42⬚C, for at least 22 hours. After hybridization, washes were done in 2 ⫻ SSC, 50% formamide/2 ⫻ SSC twice

and 2 ⫻ SSC, respectively, at 42⬚C for 5 minutes each. The slides were then rinsed in washing solution, TNT (0.1M Tris-HCI pH ⫽ 7.5, 0.15 M NaCl and 0.05% Tween-20), followed by incubation in blocking reagent (NEN Life Science product, Inc.) at RT for 30 minutes. Slides were then incubated with 100 ␮l of SA-HRP (streptavidin-horse radish peroxidase) (NEN Life Science product, Inc.) or antidigoxigenin-HRP (Boehringer Mannheim), depending on the labeled probe. After HRP incubation, the slides were

Figure 1 Murine chromosomal location of the B117 loci. Partial chromosome 5, 7, 9, and 11 linkage maps showing the location of the 117 loci in relation to linked genes are shown. The number of recombinant N2 animals over the total number of N2 animals typed plus the recombination frequency, expressed as genetic distance in centimorgans (⫹/⫺S.E.) is shown for each pair of loci to the left of each chromosome map. When no recombination was detected between loci, the upper 95% confidence limit of the recombination distance is given in parentheses. The positions of loci in human chromosomes, where known, are shown to the right. References for the human map positions of loci cited in this study can be obtained from GDB (Genome Data Base), a computerized database of human linkage information maintained by The William H. Welch Medical Library of The Johns Hopkins University (Baltimore, MD, USA).

Mapping MuLV by FISH-TSA washed in TNT solution three times for 5 minutes each at RT with agitation. Slides were treated with a 1:50 dilution of tetramethylrhodamine-tyramide (NEN Life Science product, Inc.) in PBS/H2O2, for 6–9 minutes at RT, which was then followed by three washes with TNT for 5 minutes each at RT. Slides were counterstained with 4⬘,6⬘-diamidino-2-phenyloindole (DAPI) in Antifade solution (Vysis). Microscopic analysis was performed on an Olympus BX60 microscope equipped with a digital imaging system (Applied Imaging). For each sample, and for each chromosome tested, at least 25 metaphase spreads were examined. RESULTS Genomic DNA from BXH-2 AML cell line, B117, was used to construct an EcoRI-digested, genomic lambda library and was screened for integrated ecotropic MuLV. Four lambda clones that hybridized with the ecotropic-specific envelope gene probe [8] were plaque purified. Each clone was subcloned into pBluescript; restriction enzyme mapped and probes were derived, suitable for use in Southern blotting experiments. The results of Southern blotting experiments on normal BXH-2 tissue DNA and B117 cell line DNA, as well as interspecific backcross mapping using these probes, showed that clone pB117c was a molecular clone of the Emv1 locus, and that clones pB117d, pB117l, and pB117e represented clonal, tumor-specific, ecotropic MuLV proviral insertions in cell line B117 (data not shown). B117 also has at least two other uncloned proviral insertions that can be detected by Southern blotting (data not shown). Proviral insertions at the B117e locus have been detected in another BXH-2 AML (data not shown) and this locus is the subject of ongoing analysis to determine if a cancer gene resides near this site. The mouse chromosomal location of B117 loci was determined by interspecific backcross analysis using informative restriction fragment length polymorphisms (RFLPs) from progeny derived from matings of [(C57BL/6J ⫻ M. spretus)F1 ⫻ C57BL/6J] mice as described [11]. The strain distribution pattern for each polymorphism in the interspecific backcross mice was determined and used to position the B117 loci on the interspecific map. B117c mapped to the proximal region of mouse chromosome 5 and did not recombine with Il6 in 189 mice typed in common, suggesting that the two loci are within 1.6 cM of each other (upper 95% confidence limit). This cluster is 2.5 cM distal of En2 and 2.1 cM proximal of Fgfr3 (Fig. 1). This is also the map position of the Emv1 locus (Mouse Genome Informatics, http://www.informatics.jax.org). The B117c locus probe detected no heterozygous rearrangements in B117 DNA, but instead recognizes RFLPs by Southern blotting, specific to Emv1 positive strains (data not shown). This strongly suggests that pB117c is a molecular clone of Emv1. B117d mapped to the distal region of mouse chromosome 9 and did not recombine with Rasgrf1 in 108 mice typed in common, suggesting that the two loci are within 2.8 cM of each other. These genes are 4.2 cM distal of Htr1b and 5.4 cM proximal of Trf (Fig. 1). B117e mapped to the central region of mouse chromosome 7, 3.4 cM distal of Snrpn and 3.3 cM proximal of Igf1r (Fig. 1).

47 Finally, B117l mapped to the distal region of mouse chromosome 11 and did not recombine with Thra in 141 mice typed in common, suggesting the two loci are within 2.1 cM of each other. This cluster is 0.7 cM distal to Cola1 and 3.9 cM proximal of Wnt3. The pB117c plasmid, which contains a full-length ecotropic provirus and 苲3.2 kb of flanking mouse genomic DNA, was used as a probe to detect integrated MuLV. Alternately, primer pairs were designed, from near the 5⬘ and 3⬘ ends of the Env gene, and used to amplify and clone a 2.0 kb fragment for use as a FISH probe (Materials and Methods). The Env gene probe gave a similar and informative, but weaker signal (data not shown). Hybridization of C57BL/6J normal bone marrow metaphase spreads revealed a signal with the pB117c probe on the distal tip of chromosome 8, where Emv2 has been shown to be located (data not shown). The C57BL/6J strain harbors no other ecotropic MuLV in its genome. No consistent signals were detected on other chromosomes in C57BL/6J metaphase spreads. Thus, we tested the pB117c plasmid probe for hybridization of BXH-2 AML cell line metaphase spreads to determine if it could be used to reliably detect ecotropic MuLV in these cells. BXH-2 AML cell lines B117, B132, and B186 were screened by FISH-TSA with the pB117c plasmid probe and centromere-specific probes for mouse chromosomes 1, 5, 7, 8, 9, and 11 [18]. As shown in Figure 2, hybridization of the BXH-2 AML cell line B117 metaphase spread with the pB117c probe results in the detection of a fairly large number of signals. We decided to determine if these signals might represent bona fide ecotropic MuLV integrations using two-color FISH. It is not possible to determine with certainty if all the signals detected represent true ecotropic MuLV integrations because not all of the proviral insertion sites from the B117 cell line have been molecularly cloned and mapped by interspecific backcross analysis. However, we sought to test specific chromosomes that we expect to have proviral integrations. For example, B117 has proviral integrations on mouse chromosome 11 near Thra (B117l) and within the Nf1 gene as well as a BXH-2-specific endogenous, ecotropic MuLV closer to the centromere [2, 19]. Two-color FISH using a FITC-labeled P1 clone specific for the centromere region of chromosome 11 [18] and the TSA-amplified pB117c probe shows that signals consistent with these proviral integrations are seen in B117 (Fig. 2, Fig. 3F). Three signals for the endogenous MuLV were detected on both homologs of chromosomes 5, 8, and 11, in normal BXH-2 bone marrow cells, and also in each of the BXH-2 AML we have examined (Fig. 3 (B), (D), (F)). These endogenous MuLV in the BXH-2 strain are Emv1 inherited from the C3H/HeJ parental strain on mouse chromosome 5 and Emv2 inherited from the C57BL/6J parental strain, and a newly acquired, germline, ecotropic MuLV on chromosome 11 [2, 19]. The FISH-TSA signal, using the pB117c probe, on chromosome 8 is distal to the telomere-region P1 probe (P1#5705, Genome systems, [18]) that was used (Fig. 3D). This is consistent with the fact that this P1 probe contains the D8MIT42 marker, estimated to be at a position 62 cM from the centromere, while Emv2 is esti-

48

H. Acar et al.

Figure 2 Two-color FISH. FISH-TSA using pB117c ecotropic digoxigenin-labeled whole MuLV probe (red signals in panel A) and FITC-dUTP-labeled P1 DNA probe specific for the centromere region of chromosome 11 (green signals in panel B). Results shown are from the B117 cell line. DAPI counterstain.

mated to be at a position 67 cM from the centromere (mapping data is available over the internet from the Mouse Genome Informatics, http://www.informatics.jax.org/searches/ crossdata form.shmtl). This result also suggests that sig-

nals generated using FISH-TSxA with pB117c can be resolved from signals generated using conventional FISH that are separated by roughly 5 cM on normal metaphase chromosome spreads. Homozygous signals in the cen-

49

Mapping MuLV by FISH-TSA tromere region of chromosome 5 and 20% down the length of chromosome 11 are consistent with the location of Emv1 and the BXH-2-specific endogenous ecotropic MuLV, respectively (Fig. 3B and 3F). In BXH-2 AML cell line B132, a signal was detected near the distal tip of one chromosome 5 homolog, as well as near the centromere where Emv1 maps (Fig. 3B). In the BXH-2 AML cell line B117, the pB117c probe also gave signals, in the distal region, on both chromosome 11 homologs (Fig. 3F). This is consistent with detection of an insertion within the Nf1 locus on one homolog and in the B117l locus on the other homolog. Signals were also observed on single homologs of chromosome 7 and chromosome 9, in B117 metaphase spreads, consistent with detection of insertions at the B117e and B117d loci, respectively (Figs. 3C and E). While the FISH signal on chromosome 9 is in the proximal half of chromosome 9, the map position of the B117d insertion by interspecific backcross analysis is in the distal half of this chromosome. This indicates either an inversion or other rearrangement involving chromosome 9, or that the B117d insertion has not been detected but rather another insertion, not yet molecularly cloned, has been detected. The loss and acquisition of proviral insertions in BXH-2 AML cell lines has been observed before [2]. It is possible that the B117d insertion has been lost during extended culture of B117. In addition to these signals, the BXH-2 AML cell line also showed a strong signal on the distal tip of an isochromosome 1 (Fig. 3A). The normal chromosome 1 homolog did not show any signal. Neither this chromosomal abnormality nor the MuLV-signal was detected on chromosome 1 in B132 or B186. Thus, B117 possesses at least one clonal, karyotypic abnormality. Cytogenetic studies on BXH-2 AML have not been performed, and it is not clear whether they characteristically display a high frequency of karyotypic abnormality.

DISCUSSION In this work, we have demonstrated that it is possible to screen MuLV-induced tumor cells for the position and number of proviral insertions using fluorescence in situ hybridization. The use of tyramide signal amplification makes this possible with the relatively small MuLV probes we used. Furthermore, two-color FISH, using chro-

mosome-specific P1 and MuLV probes, makes it feasible to make chromosomal assignments easily. One previous report utilized FISH-TSA to detect integrated Human Immunodeficiency Virus in two infected cell lines [20]. Our report is the first to use these methods to detect and map MuLV insertions in leukemic cells. Our technique utilizes hybridization with a full-length ecotropic MuLV probe and amplification of the signal using tyramide signal amplification. Because the ecotropic MuLV are well conserved at the nucleotide sequence level, it is likely that the same probe, pB117c, can be used to detect proviruses in other models of MuLV-induced cancer. Examples include Moloney MuLV-induced thymoma and Gross MuLV-induced myeloid leukemia. Furthermore, molecular clones are available for most of the MuLV, as well as other types of oncogenic murine retroviruses (e.g., Mouse Mammary Tumor Virus), which cause cancer in mice. These molecular clones also could be used for FISH-TSA to detect and map proviral insertions in cancer cells. The dependence of progress, in understanding the complexity of MuLV-induced cancer, on the molecular cloning of proviral insertion sites followed by mapping using mouse crosses can potentially be overcome by using FISH-TSA-based mapping as shown in the present work. The appearance of proviruses in the same chromosomal region in a series of tumors would indicate the possible presence of a common site of proviral insertion. Cloning projects that focus on just the involved set of tumors could then follow this result. In some cases, it is clear that proviral insertions that occur quite a distance from a gene can deregulate its expression [1]. In these cases, molecular cloning of an insertion site, followed by screening of other tumors for insertions in this region by Southern blot, may not reveal insertions that are near enough to deregulate the gene but not to be detected at the level of rearranged bands on a Southern blot. The use of FISH mapping would still allow the detection of such events. Another recently described technique for large-scale analysis of proviral insertion sites in MuLV-induced leukemia has recently been published [21]. This technique depends upon the use of long-template inverse PCR to clone and then sequence large numbers of proviral insertion sites from MuLV-induced leukemia samples. This approach proved to be a powerful survey method for discovering many of the genes that may be mutated by proviral

Figure 3 FISH-TSA mapping of ecotropic MuLV. Two-color FISH with FITC-dUTP-labeled P1 DNA probes specific for regions near the telomere or centromere of chromosomes 1, 5, 7, 8, 9, and 11 (green signals) and biotinylated or digoxigenin-labeled whole MuLV probe, pB117c (red signals). DAPI counterstain. (A) An isochromosome 1, showing a somatically acquired proviral insertion in the telomere region of the B117 cell line. (B) Chromosome 5, showing signals consistent with MuLV insertions representing endogenous ecotropic locus, Emv1, and a provirus near the telomere in the B132 cell line. (C) Chromosome 7, showing a somatically acquired proviral insertion in the B117 cell line (consistent with the location of B117e). (D) Chromosome 8, showing the endogenous ecotropic provirus, Emv2, from B186. (E) Chromosome 9, showing a somatically acquired proviral insertion from the B117 cell line (possibly B117d). (F) Chromosome 11, showing both somatically acquired proviral insertion in the terminal region (consistent with proviral insertions at the Nf1 locus and at B117l) and the BXH-2-specific endogenous ecotropic provirus near the centromere from the B117 cell line.

50 insertion to cause AML in BXH-2 mice and leukemia in AKXD strain mice [21]. In comparison with the method described in this publication, inverse PCR cloning clearly gives a huge advantage for actually identifying specific proviral insertion sites at the molecular level. However, this approach is very expensive and until the mouse genome has been sequenced entirely the map position of many proviral insertions remains unknown. FISH mapping of proviral insertion sites has the potential to allow a faster screen of a series of leukemia samples for proviral insertions within a given chromosomal region. From a practical standpoint, FISH mapping allows the quickest screen of a series of tumors, for insertions in regions of interest because they are thought to contain a cancer gene based on recurrent chromosomal rearrangements detected in the syntenic human region. Of particular interest for BXH-2 leukemia would be proviral insertions in mouse regions syntenic with important putative myeloid tumor suppressor genes that are not yet molecularly identified. The B132 MuLV insertion near the telomere of mouse chromosome 5 represents just such a site, since this region is syntenic with human chromosome 7q22, which is deleted in many cases of myeloid leukemia [22]. A further modification of this idea, to find MuLV insertions that may be in tumor suppressor genes, would be to use embryonic stem cell technology to create mice hemizygous for large chromosomal regions that cover putative tumor suppressor genes. These deletion chromosomes would then be crossed onto a MuLV susceptible strain, such as BXH-2. FISH-TSA could then be used to rapidly screen resultant tumors for MuLV insertions in the region covered by the deletion on the wild-type chromosome homolog. The limitations of FISH-TSA, for the analysis of MuLV insertions, are that not all proviruses may be detected, and some signals detected by MuLV FISH-TSA may not be true somatically-acquired tumor-specific proviruses. Indeed, we have observed some variation in signal strength between different insertions. Also, in some models of MuLV-induced cancer, non-ecotropic MuLV may be generated. Another limitation is the resolution that can be obtained by FISH. Our experience with the Emv2 locus near the distal tip of chromosome 8 shows that the resolution, practically obtained, is roughly 5 cM. This means that some spurious associations may be made between insertion sites in different AML. In practice, these are not likely to be serious problems. The strength of the FISH-TSA method is the ability to screen a series of AML at the genomic scale for recurrent and interesting alterations induced by MuLV insertional mutagenesis. For BXH-2 AML cell lines B117 and B132, several potentially interesting observations have been made in this preliminary work. We detected a proviral insertion on the distal tip of an isochromosome 1 in B117 cell line. The proviral insertion is likely to have occurred prior to the generation of the isochromosome. The distal tip of mouse chromosome 1 is syntenic with human chromosome 1q3–q4, where several alterations in human cancer have been observed (Mouse Genome Informatics, http://www.informatics.jax.org) [7]. We detected a proviral insertion near the

H. Acar et al. centromere of chromosome 5 in cell line B132. This region contains a cluster of genes which map to human chromosome 7q22 in a region thought to contain an important myeloid tumor suppressor gene [22]. We thank LeAnn Oseph and Betsey Hirsch for helpful discussion, technical assistance, and reviewing the manuscript. H.A. is supported by an I.D.B. fellowship. This research was sponsored by the National Cancer Institute, DHHS, in part, under contract UO1-CA84221.

REFERENCES 1. Jonkers J, Berns A (1996): Retroviral insertional mutagenesis as a strategy to identify cancer genes. Biochim Biophys Acta 1287:29–57. 2. Largaespada DA, Shaughnessy Jr JD, Jenkins NA, Copeland NG (1995): Retroviral integration at the Evi2 locus in BXH2 myeloid leukemia cell lines disrupts Nf1 expression without changes in steady-state Ras-GTP levels. J Virol 69:5095–5102. 3. Nakamura T, Largaespada DA, Shaughnessy JD, Jenkins NA, Copeland NG (1996): Cooperative activation of Hoxa and Pbx1-related genes in murine myeloid leukaemias. Nat Genet 12:149–153. 4. Bedigian HG, Johnson DA, Jenkins NA, Copeland NG, Evans R (1984): Spontaneous and induced leukemias of myeloid origin in recombinant inbred BXH mice. J Virol 51:586–594. 5. Mucenski ML, Taylor BA, Jenkins NA, Copeland NG (1986): AKXD recombinant inbred strains: models for studying the molecular genetic basis of murine lymphomas. Mol Cell Biol 6:4236–4243. 6. Gilbert DJ, Neumann PE, Taylor BA, Jenkins NA, Copeland NG (1993): Susceptibility of AKXD recombinant inbred mouse strains to lymphomas. J Virol 67:2083–2090. 7. Mitelman F, Mertens F, Johansson BA (1997): Breakpoint map of recurrent chromosomal rearrangements in human neoplasia. Nat Genet 15:417–474. 8. Chattopadhyay S, Lander MS, Rands E, Lowy DR (1980): Structure of endogenous murine leukemia virus DNA in mouse genomes. Proc Natl Acad Sci USA 77:5774–5778. 9. Jenkins NA, Copeland NG, Taylor BA, Bedigian HG, Lee BK (1982): Ecotropic murine leukemia virus DNA content of normal and lymphomatous tissues of BXH-2 recombinant inbred mice. J Virol 42:379–388. 10. Feinberg AP, Vogelstein BA (1983): Technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 132:6–13. 11. Copeland NG, Jenkins NA (1991): Development and applications of a molecular genetic linkage map of the mouse genome. Trends Genet 7:113–118. 12. Avraham KB, Givol D, Avivi A, Yayon A, Copeland NG, Jenkins NA (1994): Mapping of murine fibroblast growth factor receptors refines regions of homology between mouse and human chromosomes. Genomics 21:656–658. 13. Mock BA, Nordan RP, Justice MJ, Kozak C, Jenkins NA, Copeland NG, Clark SC, Wong GG, Rudikoff S (1989): The murine IL-6 gene maps to the proximal region of chromosome 5. J Immunol 142:1372–1376. 14. Hasson T, Skowron JF, Gilbert DJ, Avraham KB, Perry WL, Bement WM, Anderson BL, Sherr EH, Chen ZY, Greene LA, Ward DC, Corey DP, Mooseker MS, Copeland NG, Jenkins NA (1996): Mapping of unconventional myosins in mouse and human. Genomics 36:431–439. 15. Kingsley DM, Jenkins NA, Copeland NG (1989): A molecular linkage map of mouse chromosome 9 with regional localiza-

Mapping MuLV by FISH-TSA

16.

17.

18.

19.

tions for the Gsta, T3g, Ets-1, and Ldlr loci. Genetics 123:165–172. Leff SE, Brannan CI, Reed ML, Ozcelik T, Franke U, Copeland NG, Jenkins NA (1992): Maternal imprinting of the mouse Snrpn gene and conserved linkage homology with the human Prader-Willi syndrome region. Nature Genet 2:259– 264. Buchberg AM, Brownell E, Nagata S, Jenkins NA, Copeland NG (1989): A comprehensive genetic linkage map of murine chromosome 11 reveals extensive linkage conservation between mouse and human. Genetics 122:153–161. Shi Y-P, Mohapatra G, Miller J, Hanahan D, Lander E, Gold P, Pinkel D, Gray J (1997): FISH probes for mouse chromosome identification. Genomics 45:24–47. Cho BC, Shaughnessy JD Jr, Largaespada DA, Bedigian HG,

51 Buchberg AM, Jenkins NA, Copeland NG (1995): Frequent disruption of the Nf1 gene by a novel murine AIDS virusrelated provirus in BXH-2 murine myeloid lymphomas. J Virol 69:7138–7146. 20. Deichmann M, Bentz M, Haas R (1997): Ultra-sensitive FISH is a useful tool for studying chronic HIV-1 infection. J Virol Meth 65:19–25. 21. Li J, Shen H, Himmel KL, Dupuy AJ, Largaespada DA, Nakamura T, Shaughnessy Jr JD, Jenkins NA, Copeland NG (1999): Leukemia disease genes: large scale cloning and pathway predictions. Nature Genet 23:348–353. 22. Liang H, Fairman J, Claxton DF, Nowell PC, Green ED, Nagarajan L (1998): Molecular anatomy of chromosome 7q deletions in myeloid neoplasms: evidence for multiple critical loci. Proc Natl Acad Sci USA 95:3781–3785.