P53-Binding Protein 1 Is Fused to the Platelet-Derived Growth Factor

[CANCER RESEARCH 64, 7216–7219, October 15, 2004] Advances in Brief p53-Binding Protein 1 Is Fused to the Platelet-Derived Growth Factor Receptor ␤ in a Patient with a t(5;15)(q33;q22) and an Imatinib-Responsive Eosinophilic Myeloproliferative Disorder

Francis H. Grand,1 Sonja Burgstaller,2 Thomas Ku¨hr,2 E. Joanna Baxter,1 Gerald Webersinke,3 Josef Thaler,2 Andrew J. Chase,1 and Nicholas C. P. Cross1 1Wessex Regional Genetics Laboratory, Salisbury and Human Genetics Division, University of Southampton, Southampton, United Kingdom; 2Department of Internal Medicine, General Hospital Wels, Wels, Austria; and 3Department of Internal Medicine, Hospital of the Sisters of Mercy, Linz, Austria

Abstract Materials and Methods

We describe the fusion of TP53BP1 to PDGFRB in a patient with a Patient Details. A 79-year-old man was admitted to hospital in September chronic myeloid leukemia-like disorder associated with eosinophilia and a 2001 with dyspnea, fatigue, and weight loss. He presented with a generalized t(5;15)(q33;q22). TP53BP1 encodes 53BP1, a p53-binding protein that erythematous rash, leukocytosis, and splenomegaly. The peripheral blood plays a role in cellular responses to DNA damage. The 53BP1-PDGFR␤ counts at this time were: white blood cells 138 ϫ 109/L, hemoglobin 4.0 g/dL, fusion protein is predicted to retain the kinetochore-binding domain of and platelets 92 ϫ 109/L. A diagnosis of CML was made after bone marrow 53BP1 fused to the transmembrane and intracellular tyrosine kinase aspiration and biopsy, which showed granulocytic hyperplasia with a marked domain of PDGFR␤. The presence of the fusion was confirmed by two- eosinophilia (45% of granulocytic cells) and a reduction of the erythroid and color fluorescence in situ hybridization, reverse transcription-PCR, and megakaryocytic series. In addition, histologic assessment of various skin by characterizing the genomic breakpoints. The reciprocal fusion, which biopsies exhibited a leukemic infiltrate with a preponderance of eosinophils. would contain the p53-binding 53BP1 BRCA1 COOH-terminal domains, Bone marrow cytogenetics and reverse transcription-PCR analysis were neg- was not detectable by fluorescence in situ hybridization or nested PCR. ative for the Philadelphia (Ph) chromosome and BCR-ABL rearrangement, Imatinib, a known inhibitor of PDGFR␤, blocked the growth of patient respectively; however, a t(5;15)(q33;q22) was seen in 21 of 21 metaphase colony-forming unit, granulocyte-macrophage in vitro and produced a cells. The possibility of a constitutional translocation was eliminated by clinically significant response before relapse and subsequent death with karyotyping a phytohemagglutinin-stimulated peripheral blood sample. After imatinib-resistant disease. We conclude that TP53BP1-PDGFRB is a novel diagnosis, the patient was treated with various agents (hydroxyurea, interferon imatinib target in atypical chronic myeloid leukemia. ␣, busulfan, and mercaptopurine). Although normalization of the leukocyte count and a decrease in the severity of skin infiltrations was achieved initially, Introduction the responses were transient. Treatment with imatinib was initiated in June 2002 after identification of a The molecular characterization of reciprocal chromosomal translo- PDGFRB rearrangement and demonstration of sensitivity to imatinib in vitro cations in myeloproliferative disorders has led to the identification of (see below). At this time the predominant clinical issue was a grade 4 diverse tyrosine kinase fusion genes. These fusions encode chimeric transfusion-dependent thrombocytopenia (platelets ranged between 6 and 9 proteins composed of an NH2-terminal partner joined in frame to the 18 ϫ 10 /L) and grade 4 anemia. After a starting dose of 400 mg of imatinib entire catalytic domain and COOH-terminal region of the kinase (1). daily, the white blood cell count declined from 27 ϫ 109/L to normal range at Typically, the partner proteins contain dimerization or oligomeriza- day 9, and the spleen size decreased from 20 cm to 15 cm. On day 36 treatment tion motifs, such as helix-loop-helix, leucine zipper, or coiled-coil was interrupted due to development of grade 4 neutropenia. Because this domains, that are essential for the observed constitutive tyrosine adverse event did not resolve within 2 weeks, the dose of imatinib was reduced kinase activity of the fusions and downstream signaling through to 300 mg/day. Cytopenia caused four interruptions in the imatinib adminis- growth-stimulatory and antiapoptotic pathways (1). Broadly, tyrosine tration. During the course of imatinib treatment the platelet counts gradually increased, and the platelet transfusion requirement reduced from alternate days kinase fusions are believed to deregulate hemopoiesis in a manner to once weekly. However, after ϳ5 months, resistance to imatinib developed, analogous to BCR-ABL in chronic myeloid leukemia (CML). From a which could not be overcome by dose escalation. Grade 4 transfusion-resistant clinical standpoint, identification of patients with PDGFRA, PDG- thrombocytopenia developed, and the patient died of an intracerebral hemor- FRB, or ABL fusions is particularly important, because these individ- rhage. uals are responsive to imatinib mesylate (2–4). To date, eight fusion In vitro Imatinib Sensitivity Assay. Imatinib sensitivity was tested on partners for PDGFRB have been reported: TEL/ETV6, CEV14, HIP1, colony-forming unit, granulocyte-macrophage (CFU-GM) growth before ima- H4, RABEP1, PDE4DIP, HCMOGT-1, and NIN (refs. 5–8 and ref- tinib therapy. Peripheral blood mononuclear cells were separated using lym- erences therein). Molecular cytogenetic evidence suggests that several phoprep (Axis-Shield, Oslo, Norway) and grown in methylcellulose supple- additional partner loci remain to be identified (9). Here we describe mented with growth factors (Stem Cell Technologies Ltd., Vancouver, British ϫ 5 the identification of one such locus, TP53BP1 at 15q22, and document Columbia, Canada) at a cell density of 2 10 cells/mL in 2.5-cm Petri dishes. Imatinib (Novartis, Basel, Switzerland) was added to final concentrations of 0, the response of cells harboring the TP53BP1-PDGFRB fusion to 1, and 5 ␮mol/L. Colony numbers were scored at days 7 and 14 from triplicate imatinib in vitro and in vivo. plates, and the numbers compared with those obtained from normal individuals and patients with BCR-ABL-positive CML. Received 6/7/04; revised 7/28/04; accepted 8/22/04. Fluorescence In situ Hybridization Analysis. Fluorescence in situ hybrid- Grant support: Leukemia Research Fund of the United Kingdom. ization for PDGFRB was performed with flanking cosmid probes as described 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 previously (9). Other clones were identified using the University of California 4 18 U.S.C. Section 1734 solely to indicate this fact. Santa Cruz genome browser and obtained from the Sanger Institute (Hinxton, Requests for reprints: Prof. Nicholas Cross, Wessex Regional Genetics Laboratory, United Kingdom). Bacterial artificial chromosome DNA was grown, extracted, Salisbury District Hospital, Salisbury SP2 8BJ, United Kingdom. Phone: 44-1722- 429080; Fax: 44-1722-338095; E-mail: [email protected]. ©2004 American Association for Cancer Research. 4 Internet address: http://genome.ucsc.edu/ 7216

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 2004 American Association for Cancer Research. A t(5;15)(q33;q22) AND TP53BP1-PDGFRB FUSION and labeled by nick translation with alkali stable digoxigenin or biotin-16–2Ј- deoxy-uridine-5Ј-triphosphate (Roche, Mannheim, Germany) and hybridized as described previously (9). PCR Methods. The partner gene was identified by 5Ј rapid amplification of cDNA ends using the GeneRacer kit (Invitrogen, Paisley, United Kingdom). Briefly, ϳ5 ␮g of total RNA extracted using the Qiagen RNeasy kit (Qiagen Ltd., Boundary Court, United Kingdom) was dephosphorylated, decapped, and ligated to the GeneRacer RNA oligo according to the manufacturer’s instructions. The ligated RNA was reverse transcribed using Superscript II reverse transcriptase (Invitrogen) and random primers (100 ng). The first step 5Ј rapid

Fig. 2. CFU-GM growth of blood mononuclear cells from (A) normal control, (B)a CML patient, and (C) the t(5;15)(q33;q22)-positive patient; bars, ϮSEM.

amplification of cDNA ends PCR was performed using the 5Ј GeneRacer primer from the kit in combination with a reverse primer from PDGFRB exon 15 (2R: 5Ј-TGCTGCAGGAAGGTGTGTTTGTTG-3Ј). The PCR cycles were designed to amplify fragments up to 5 kb, with an annealing temperature of 66°C using the High Fidelity PCR Master kit (Roche) according to the manufacturer’s instructions. First-step products were diluted 1:200 before being used as template for the second-step PCR. Second step PCR was performed with the 5ЈGeneRacer nested primer and a reverse primer derived from PDGFRB exon 15 (1R: 5Ј-AGGTAGTCCACCAGGTCTCCGTA-3Ј) under the same conditions as the first-step PCR. Products were cloned with the TOPO TA cloning kit for sequencing (Invitrogen) and sequenced. The presence of TP53BP1-PDGFRB mRNA was confirmed on random hexamer reverse-transcribed cDNA (8) using primers to TP53BP1 (3F: 5Ј-GGG- GAACTGTACTACAGCATTGA-3Ј) plus primer PDGFRB-1R. The genomic breakpoint was amplified using a forward genomic primer from TP53BP1 intron 23 (1F: 5Ј-ACCCGAAAGATCACAGAAAGTCC-3Ј) and a reverse primer from PDGFRB intron 11 (int1R: 5Ј-GAGAGCAGGCCATGAG- CAAAC-3Ј.

Results and Discussion Fig. 1. Fluorescence in situ hybridization analysis. A, a t(5;15) metaphase hybridized Involvement of PDGFRB. To our knowledge, the t(5;15)(q33; to differentially labeled cosmid probes flanking PDGFRB demonstrating disruption of this gene; B, a t(5;15) metaphase hybridized with bacterial artificial chromosome RP11– q22) has not been described before; however, the breakpoint on 114F23 (red), which contains TP53BP1 and spans the chromosome 15 breakpoint, in chromosome 5 in combination with the clinical phenotype suggested combination with the bacterial artificial chromosome clone RP11–100O5 (green; downstream of PDGFRB). A fusion signal is seen on the der(5) as expected, but no signal is that PDGFRB might be involved. To test this possibility we per- seen on the der(15). formed two-color fluorescence in situ hybridization using flanking 7217

Fig. 3. A, the mRNA fusion of TP53BP1 exon 23 to PDGFRB exon 11; B, amplification of the TP53BP1-PDGFRB fusion by single step reverse transcription-PCR from the t(5;15) patient but not normal controls (N1-N3). B1 and B2 are blanks. C, sequences surrounding the genomic breakpoints in TP53BP1 (underlined) and PDGFRB. The exact position of the breakpoint is ambiguous due to a short stretch of homology (GGG) between the two genes. cosmid probes that immediately flank this locus (9). These probes performed additional fluorescence in situ hybridization analysis. both hybridized to band q33 on the normal copy of chromosome 5 but Bacterial artificial chromosome RP11–114F23 contains the entire hybridized differentially to the der(5) and der(15), indicating that the TP53BP1 gene, and bacterial artificial chromosome RP11–100O5 translocation did indeed target PDGFRB (Fig. 1A). is downstream (centromeric) of PDGFRB.4 Cohybridization of Response to Imatinib In vitro and In vivo. To test if the presump- these two clones to t(5;15) metaphases revealed the expected green tive PDGFRB fusion gene was responsive to imatinib, we analyzed and red signals on the normal chromosomes 5 and 15, respectively, CFU-GM production in the presence of 0 ␮mol/L, 1 ␮mol/L, or 5 plus a fusion signal corresponding to TP53BP1-PDGFRB on the ␮mol/L imatinib. Peripheral blood from a normal control individual der(5; Fig. 1B). The expected TP53BP1 signal was not seen on the showed a nonspecific reduction in colony numbers of ϳ50% with 1 der(15), suggesting this sequence had been deleted. A deletion at ␮mol/L imatinib at both days 7 and 14. An additional 10% reduction the breakpoint explains why reciprocal PDGFRB-TP53BP1 fusion was seen with 5 ␮mol/L imatinib (Fig. 2A), and similar results were transcripts were not detected and may also account for the fact that seen with an additional 10 normal controls (data not shown). The the breakpoint was assigned cytogenetically to 15q22 but reduction was much greater for CFU-GM grown from the CML TP53BP1 maps to 15q15. patient, used as a positive control, with ϳ90% growth inhibition in the Structure and Function of 53BP1 and the 53BP1-PDGFR␤. presence of 1 ␮mol/L and 5 ␮mol/L imatinib at both days 7 and 14 53BP1 is a component of the cellular response to DNA damage that (Fig. 2B). Similarly, marked inhibition was seen for CFU-GM grown was initially identified as a protein that binds to wild-type but not from the t(5;15) patient, with Ͼ95% reduction in colony numbers at mutant p53 (10). Subsequently, it has been shown that the BRCA1 day 7 and 80% to 90% reduction at day 14 (Fig. 2C). These data COOH-terminus domains of 53BP1 bind to the central DNA-binding suggested that the t(5;15)-associated disease was indeed sensitive to domain of p53 and enhance p53-mediated transcriptional activation imatinib, and consequently the patient was treated with this com- (11, 12). After irradiation, 53BP1 is hyperphosphorylated in an ATM- pound. As described above, the leukocyte count and eosinophilia responded to imatinib, but the patient subsequently relapsed with imatinib-resistant disease. Unfortunately, no clinical material was available to investigate the molecular basis for imatinib resistance. Identification of the TP53BP1-PDGFRB Fusion. All of the PDGFRB fusions reported to date result in an mRNA junction with the partner gene sequence spliced to PDGFRB exons 11 or 12. Therefore, to identify the t(5;15) partner we performed 5Ј-rapid amplification of cDNA ends PCR. Sequencing of the products revealed several clones in which exon 23 of TP53BP1 (accession no. AF078776) was fused in frame to PDGFRB exon 11 (Fig. 3A). The presence of the TP53BP1-PDGFRB fusion was confirmed initially by reverse transcription-PCR. As shown in Fig. 3B, the fusion was specifically amplified from t(5;15) cells by single-step PCR but was not detectable in normal controls. The reciprocal PDGFRB-TP53BP1 product was not detectable by single-step or nested PCR. Amplifica- tion and sequencing of the genomic breakpoint confirmed the fusion of intron 23 of TP53BP1 to intron 10 of PDGFRB (Fig. 3C). TP53BP1-PDGFRB is predicted to be translated into a 247 KDa ␤ 53BP1-PDGFR fusion protein that retains the NH2-terminal region and kinetochore binding domain from 53BP1 fused to the transmembrane and entire cytoplasmic domain of PDGFR␤. Confirmation of the Presence of TP53BP1-PDGFRB. To addi- Fig. 4. Schematic representation of 53BP1, PDGFR␤, and the 53BP1-PDGFR␤ fusion tionally confirm the presence of a TP53BP1-PDGFRB fusion, we protein. 7218

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 2004 American Association for Cancer Research. A t(5;15)(q33;q22) AND TP53BP1-PDGFRB FUSION dependent manner and colocalizes with H2AX and other factors at References double-stranded DNA breaks (13, 14) in a manner dependent on the 1. Cross NCP, Reiter A. Tyrosine kinase fusion genes in chronic myeloproliferative conserved 53BP1 Tudor and Myb domains (15). TP53BP1(Ϫ/Ϫ) diseases. Leukemia (Baltimore) 2002;16:1207–12. mice are cancer prone, indicating that this gene functions as a tumor 2. Apperley JF, Gardembas M, Melo JV, et al. Response to imatinib mesylate in patients with chronic myeloproliferative diseases with rearrangements of the platelet-derived suppressor and manifests defects in DNA damage response and cell growth factor receptor beta. New Engl J of Med 2002;347:481–7. cycle checkpoint control (16). 3. Cools J, DeAngelo DJ, Gotlib J, et al. A tyrosine kinase created by fusion of the The t(5;15) is predicted to generate a chimeric protein that is PDGFRA and FIP1L1 genes as a therapeutic target of imatinib in idiopathic hyper- structurally similar to other tyrosine kinase fusions. Therefore, it is eosinophilic syndrome. N Engl J Med 2003;348:1201–14. 4. Pardanani A, Tefferi A. Imatinib targets other than bcr/abl and their clinical relevance very likely that 53BP1-PDGFR␤ is a constitutively active transform- in myeloid disorders. Blood Epub 2004 May 27. ing oncogene, a notion strongly supported by the fact that patient 5. Steer EJ, Cross NCP. Myeloproliferative Disorders with Translocation of Chromo- CFU-GM were inhibited by imatinib and the observed clinical re- somes 5q31–35: Role of the Platelet-Derived Growth Factor Receptor Beta. Acta Haematol 2002;107:113–22. sponse. It is conceivable, however, that the t(5;15) also interferes with 6. Wilkinson K, Velloso ER, Lopes LF, et al. Cloning of the t(1;5)(q23;q33) in a the DNA damage response, either by haploinsufficiency of TP53BP1 myeloproliferative disorder associated with eosinophilia: involvement of PDGFRB or through a dominant-negative effect of the fusion protein, which and response to imatinib. Blood 2003;102:4187–90. 7. Morerio C, Acquila M, Rosanda C, et al. HCMOGT-1 is a novel fusion partner to retains the Tudor and Myb domains but not the BRCA1 COOH- PDGFRB in juvenile myelomonocytic leukemia with t(5;17)(q33;p11.2). Cancer Res terminal domain (Fig. 4). 2004;64:2649–51. As described above, constitutive activation of fusion tyrosine 8. Vizmanos JL, Novo FJ, Roman JP, et al. NIN, a gene encoding a CEP110-like centrosomal protein, is fused to PDGFRB in a patient with a t(5;14)(q33;q24) and an kinases is mediated by an oligomerization domain in the partner imatinib-responsive myeloproliferative disorder. Cancer Res 2004;64:2673–6. protein. We did not identify any helix-loop-helix or leucine zipper 9. Baxter EJ, Kulkarni S, Vizmanos JL, et al. Novel translocations that disrupt the motifs in 53BP1, but the coiled-coil prediction program Coils v2.1 platelet-derived growth factor receptor beta (PDGFRB) gene in BCR-ABL negative 5 chronic myeloproliferative disorders. Br J of Haematol 2003;120:251–6. (17) indicated the presence of three potential coiled-coil domains: 10. Iwabuchi K, Bartel PL, Li B, Marraccino R, Fields S. Two cellular proteins that bind amino acids 133 to 150 (window ϭ 14; probability Ͼ94%); amino to wild-type but not mutant p53. Proc Natl Acad Sci USA 1994;91:6098–102. acids 731 to 744 (window ϭ 14; probability Ͼ92%) and amino 11. Iwabuchi K, Li B, Massa HF, Trask BJ, Date T, Fields S. Stimulation of p53-mediated ϭ Ͼ transcriptional activation by the p53-binding proteins, 53BP1 and 53BP2. J Biol acids 798 to 818 (window 21; probability 97%). It is possible Chem 1998;273:26061–8. that one or more of these domains is responsible for oligomeriza- 12. Derbyshire DJ, Basu BP, Serpell LC, et al. Crystal structure of human 53BP1 BRCT tion of 53BP1-PDGFR␤. domains bound to p53 tumour suppressor. EMBO J 2002;21:3863–72. 13. Anderson L, Henderson C, Adachi Y. Phosphorylation and rapid relocalization of Concluding Remarks. In summary, we have identified 53BP1- 53BP1 to nuclear foci upon DNA damage. Mol Cell Biol 2001;21:1719–29. PDGFR␤ fusion as a consequence of a t(5;15)(q33;q22) in a patient 14. DiTullio RA Jr., Mochan TA, Venere M, et al. 53BP1 functions in an ATM- with an atypical, CML-like myeloproliferative disorder. TP53BP1 is dependent checkpoint pathway that is constitutively activated in human cancer. Nat Cell Biol 2002;4:998–1002. the ninth PDGFRB fusion partner to be identified and further expands 15. Iwabuchi K, Basu BP, Kysela B, et al. Potential role for 53BP1 in DNA end-joining the repertoire of imatinib-responsive abnormalities in hematologic repair through direct interaction with DNA. J Biol Chem 2003;278:36487–95. disorders. 16. Ward IM, Minn K, van Deursen J, Chen J. p53 Binding protein 53BP1 is required for DNA damage responses and tumor suppression in mice. Mol Cell Biol 2003;23:2556–63. 17. Lupas A, Van Dyke M, Stock J. Predicting coiled coils from protein sequences. 5 Internet address: http://www.ch.embnet.org/software/COILS_form.html. Science (Wash DC) 1991;252:1162–64.

7219

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 2004 American Association for Cancer Research. p53-Binding Protein 1 Is Fused to the Platelet-Derived Growth Factor Receptor β in a Patient with a t(5;15)(q33;q22) and an Imatinib-Responsive Eosinophilic Myeloproliferative Disorder

Francis H. Grand, Sonja Burgstaller, Thomas Kühr, et al.

Cancer Res 2004;64:7216-7219.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/64/20/7216

Cited articles This article cites 15 articles, 10 of which you can access for free at: http://cancerres.aacrjournals.org/content/64/20/7216.full#ref-list-1

Citing articles This article has been cited by 8 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/64/20/7216.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/64/20/7216. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.