Detection of Translocations Affecting the BCL6 Locus in B Cell

Detection of translocations affecting the BCL6 locus in B cell non-Hodgkin’s lymphoma by interphase ﬂuorescence in situ hybridization D Sanchez-Izquierdo1, R Siebert2, L Harder2, I Marugan1, A Gozzetti3, HP Price4,SGesk2, JM Hernandez-Rivas5, I Benet1, F Sole´6, T Sonoki4, MM Le Beau3, B Schlegelberger2, MJS Dyer4, J Garcia-Conde1 and JA Martinez-Climent1

1Department of Hematology and Oncology, Hospital Clinico, University of Valencia, Spain; 2Institute of Human Genetics, University Hospital Kiel, Germany; 3Section of Hematology and Oncology, University of Chicago, USA; 4Academic Dept of Haematology and Cytogenetics, Institute of Cancer Research, Sutton, UK; 5Department of Hematology, Hospital Clı´nico, University of Salamanca, Spain; and 6Department of Hematology, Hospital del Mar, Barcelona, Spain

Structural alterations in 3q27 affecting the BCL6 locus are potential in their own right including BOB-1, TTF and PIM1. among the most frequent changes in B-NHL. The aim of the Breakpointsat 3q27 are predominantly located in the 5 Ј present study was to establish an interphase-FISH assay for the detection of all diverse BCL6 translocations in B-NHL. Two untranslated region of BCL6, including the promoter and the different approaches were tested, one using a PAC-clone span- first non-coding intron; this region has been termed the major ning the major breakpoint region (MBR) of BCL6 (span-assay), breakpoint region (MBR). In such translocations, the BCL6 and another using two BAC clones flanking the MBR (flank- coding exons are juxtaposed downstream to the promoter assay). Interphase FISH with the span-assay detected the vari- derived from the reciprocal chromosomal partner, resulting in ous BCL6 translocations in seven B-NHL cell lines. The dual- deregulated BCL6 expression.9,10 However, there hasbeen no color flank-assay was evaluated in two laboratories independently: in normal controls, the cutoff level for false-positive sig- clear link between the presence of BCL6 chromosomal trans- nals was 2.6%, whereas the cutoff level for false-negatives in locationsand deregulated BCL6 expression and it is possible the seven cell lines was 7.5%. To test the feasibility of the FISH that clonal, somatic mutations and deletions within the same strategies, 30 samples from patients with B-NHL with cytog- regulatory region of the BCL6 gene may also contribute to enetic abnormalities of 3q27 were evaluated with both assays. deregulated expression.8,10 In 21 cases, the span-assay indicated a BCL6 rearrangement. Chromosomal translocations involving 3q27 or rearrange- In 18 of the 21 cases, the dual-color flank-assay confirmed the translocation including 12 different partner chromosomal loci. mentsof the BCL6 gene have been identified in about 40% The three false-positive cases detected with the span-assay of diffuse large B-NHL (DLBCL), but they have also been showed trisomy of chromosome 3 by cytogenetic analyses, found at lower frequency in other subtypes of B-NHL.2,7,11,12 and they were correctly classified as non-rearranged with the Whether translocations affecting 3q27 or BCL6 rearrange- flank-assay. In summary, our FISH strategy using two differ- ments are of prognostic value in DLBCL remains controversial. ently labeled flanking BCL6 BAC probes provides a robust, One study reported a favorable prognosis for patients with sensitive, and reproducible method for the detection of com- 11 mon and uncommon abnormalities of BCL6 gene in interphase BCL6 rearrangementsdetected by Southern blot. However, 7,13 nuclei. The routine application of this assay to patients with B- these data have not been confirmed. Resolution of this NHL will allow the assessment of the diagnostic and prognostic point demands an assay that will allow the routine detection significance of BCL6 rearrangements. Leukemia (2001) 15, of all BCL6 translocations. This has been hampered by a num- 1475–1484. ber of different factors. Conventional cytogenetic analysis is Keywords: BCL6 translocations; FISH; B cell lymphoma limited by the location of BCL6 in the terminal portion of chromosome 3q and by the presence of complex three-way translocations within the BCL6 locus.14 Additionally, cyto- Introduction genetically detectable breakpointsin 3q27 might not affect the BCL6 gene. For a regular polymerase chain reaction (PCR) Reciprocal chromosomal translocations involving band 3q27 assay, a complex set of primers for each corresponding are frequent in B cell non-Hodgkin’slymphomas(B-NHL). 1,2 rearrangement would need to be designed to detect all poss- The involved gene in most cases is a zinc finger gene, BCL6, ible rearrangements.8 Currently, the only available techniques and hasbeen found juxtaposednot only to the immunoglob- for the screening of most BCL6 translocations are Southern ulin (IG) genes, but also to a variety of other genes, some of blot analyses and long-distance inverse PCR, but these which have been characterized by molecular cloning.3–8 methods are limited by the necessity for high molecular These genes include not only constitutively expressed genes weight DNA, which isnot alwaysavailable from every patient such as histone H4, but also genes with possible oncogenic and by the technical difficulties associated with these methods.7,8,11–13 Furthermore, the frequent presence of deletionsand mutationswithin the BCL6 MBR may confound Correspondence: JA Martıńez-Climent, Department of Hematology results of Southern blot analysis. Additionally, the expression and Medical Oncology, Hospital Clińico Universitario, University of Valencia, Avda Blasco Iban˜ez, 17, 46010 Valencia, Spain; Fax: 34- of BCL6 protein doesnot correlate with BCL6 gene rearrange- 96-362-2238 ment and therefore immunohistochemical studies cannot be Received 12 January 2001; accepted 9 May 2001 used as a surrogate marker for BCL6 translocations.15 Detection of BCL6 translocations by FISH D Sanchez-Izquierdo et al 1476 Fluorescence in situ hybridization (FISH) is a sensitive Probes used for FISH method for the routine identification of leukemia and lymphoma-associated chromosomal translocations. Using large Two PAC clonesRMC03P056 and RMC03P061 containing DNA probes such as yeast artificial chromosome (YAC), P1 BCL6 intron 1 were kindly provided by WL Kuo and JW Grey artificial chromosome (PAC) or bacterial artificial chromo- (University of California, San Francisco; http://rmc- some (BAC) spanning common breakpoints, it is possible to www.lbl.gov). BAC clonesRPCI-11 211G3, 528E8 and 690C8 detect most translocations within a specific gene, despite a from the contig HPFCctg 13301 covering the BCL6 locus large number of partner chromosomes.16 However, the sys- (Washington University, http://genome.wustl.edu/) were tematic evaluation of any single color assay is technically dif- obtained from Research Genetics (Huntsville, AL, USA; ficult on patient samples. Thus, for diagnostic purposes, a http://www.resgen.com). more specific approach using two differentially labeled probes Other DNA probes were also used in selected cell lines and from both partner chromosomes leading to the detection of patientsfor the detection of BCL6 translocations to IGH, IGL, the specific fusion signals on interphase nuclei, such as in the and C-MYC genes. For IGH, cosmid c␣1-IGH containing the t(8;14)(q24;q32), t(14;18)(q32;q21) or t(11;14)(q13;q32) trans- C␣1 gene wasused, 17 aswell asPAC 1098L17, kindly pro- locations, is commonly used.17–20 In the case of promiscuous vided by Dr Mariano Rocci (University of Bari, Italy; genes, such as MLL or ETV6, these assays are not of value, http://www.bioserver.biologia.uniba.it), which contains the but translocations can be detected using FISH probes flanking variable region of IGH. For IGL, PAC 1019H10 (also from Dr the MLL or ETV6 breakpoints.21 Few previousstudieshave Rocci) wasselected;thisPAC containsexon 3 of the IGL join- attempted the detection of BCL6 rearrangementsusingFISH. ing segment. By FISH, it was mapped to 22q11 and was fused Two of these reports established assays which were only to C-MYC and BCL2 in several patients and cell lines with applicable to metaphase cells,22,23 whereasanother studyper- t(8;22)(q24;q11) and t(18;22)(q21;q11), respectively (Sanchez formed on interphase cells was limited to the identification of et al, manuscript in preparation). Probes for the C-MYC gene BCL6 translocations involving IGH.24 To date, there hasnot have been described recently.17 DNA from PACs, BACs and been a FISH method available for the routine detection of all cosmids was prepared using the Qiagen (Courtabouef, France) BCL6 translocations. kit system. We report a novel strategy using two differentially labeled flanking BCL6 BAC probes, which provides a robust, sensitive, and reproducible method for the detection of all common and FISH methods uncommon abnormalitiesof BCL6 in B-NHL by interphase FISH. FISH studies were performed on fixed cells from cytogenetic samples. In three patients without available material, FISH was performed on G-banded slides stored at room tempera- ture between 1 and 8 years. DNA from the probes was directly labeled with Spectrum Red (SR) and Spectrum Green (SG) Materials and methods (Vysis, Downers Grove, IL, USA) by nick translation. FISH pro- cedure wasperformed asreported previously. 21 For each experiment, at least 200 interphase nuclei, as well as 20 meta- Cell lines and patients with B-NHL phase cells were examined by two independent observers. The samples were examined in Valencia with a Nikon Eclipse 400 fluorescence microscope (Nikon Corp, Tokyo, Japan) and Seven B-NHL cell lineswith different BCL6 translocations in Kiel with a Zeiss Axioskop 2 fluorescence microscope previously characterized at the cytogenetic and/or molecular (Zeiss, Oberkochen, Germany) equipped with appropriate fil- levels were selected for this study: VAL,25 MD903,26 Oci- ter sets and documented using the QuipsXL Genetic Workst- Ly8,14,27 MD901,6 Karpas231, 28,29 YM30 and the B-NHL cell ation (Vysis) and the ISIS3 software (MetaSystems, Altluss- line CTB-1.31 The CTB-1 cell line hasbeen reported on the heim, Germany), respectively. basis of cytogenetic analysis alone to have a t(14;22)(q32;q11); however, we have found by FISH, molecular methods, and cross-species color banding (RxFISH) an Molecular studies exchange of chromosomal material from 3q to 14q, resulting in a complex four way translocation at the IGH locus, which Screening of genomic libraries for BCL6 intron 1-containing will be described in greater detail elsewhere (Sonoki et al, clones: PCR for detection of intron 1 of BCL6 corres- manuscript in preparation). All cell lines were grown in RPMI- ponding to the MBR region wasperformed usingthe follow- 1640 medium supplemented with 10% fetal calf serum. ing primers: 5Ј CTCTTGCCAAATGCTTTG 3Ј and 5Ј Samplesfrom 30 patientswith B-NHL (21 DLBCL, seven AAGCAGTTTGCAAGCGAG 3Ј, generating a 465 bp fragment follicular lymphoma, one MALT lymphoma, one B-NHL of in the wild-type gene. After an initial denaturation cycle at indeterminate histological subtype) with cytogenetic abnor- 94°C for 8 min, 30 cyclesof denaturation (94 °C for 30 s), malities of 3q27 were also evaluated. Twenty-five samples annealing (60°C for 30 s) and extension (72°C for 30 s), and were from lymph node, two were from bone marrow, two a final extension cycle at 72°C for 5 min, the samples were were from peripheral blood, and one wasfrom a pleural loaded on to a 1% agarose gel with ethidium bromide. effusion. Cytogenetic studies on samples from the patients and cell lines were performed and described according to standard methods using either G- or R-banding techniques.32 Addition- Southern blot and IGH LDI-PCR to detect BCL6 rearrange- ally, cross-species color karyotyping (RxFISH) was performed ments: In cases where sufficient high molecular weight in OCI-Ly8 and CTB-1 cell linesasreported. 33 Results of cyto- DNA wasavailable, Southern blot for BCL6 gene rearrange- genetic and FISH analyses are summarized in Table 1. mentswasperformed asdescribed. 34 DNA wasdigestedwith

Leukemia Detection of BCL6 translocations by FISH D Sanchez-Izquierdo et al 1477

assay assay

FISH Span- Flank- R t(3;22)(q27;q11) 10% 12% 9 [2]/ R t(3;13)(q27;q14) 70% 90% 15, R t(3;16)(q27;q22) 21% 44% − − 21 + X [3]/ R t(2;3)(p11;q27) 47% 63% 6 mar [cp15] R (t(3;14)(q27;q32) 88% 41% + 13, ෂ ,add(4)(q21), R t(3;22)(q27;q11) 79% 91% − ,add(9)(p21),del(13) R t(3;14)(q27;q32) 89% 93% 2 + X [1] 11, + − 21, + 13,i(13)(q10),t(16;22)(p12;q11), R t(3;12)(q27;q13) 89% 94% 7,del(8)(q13q21)x3, R t(3;14)(q27;q32) 81% 92% 24), 22 [9] R t(3;14)(q27;q32) 89% 95% − + ෂ del(3)(q27) − 7, 10, + − mar [3]/50,idem, 7, t(3;22)(q27;q11) + 21 [1]/47,XY, + + 22;q23 mar [9]/46,XY [11] R t(3;14)(q27;q32) 85% 95% ෂ + 21 [5]/46,XX [6] r, + ,der(18)t(14;18)(q32;q21) R t(3;22)(q27;q11) 35% 38% + 22,der(22)t(3;22)(q27;q11)[1]/ 18 + + 3, + 9 [1] 25),t(10;15;22)(q24;q26.1;q13) [1]/46,XX [3] + 2, ෂ + ,dic(6;17)(p25;p12),add(9)(p21), 12,t(14;18)(q32;q21), 22,der(22)t(3;22)(q27;q11) [14]/48,idem, + 8,der(12)t(?8;12), + 21 [1]/46,XY[13] − + 2, + ,del(6)(q13q21),t(14;18)(q32;q21) [8]/46,XX [3] 14[2]/47,idem, ,der(8)t(8;8)(p21;q31), 7,t(14;18)(q32;q21), 18,der(18)t(14;18)(q32;q21) [29]/47,idem,del(5)(q13) [1]/ R t(3;16)(q27:p13) 88% 95% + der(18;22)(q10;q10),add(19)(q13), + ,del(8)(p21),del(9)(q21q31–33), ,t(14;18)(q32;q21) [6]/46,XY [14] R t(3;6)(q27;q15) 44% 35% + + [17]/46,XY [3] R inv(3)(q13.2q27) 68% 88% dic(5;9)(q31;p13),del(6)(q12)x2, 5,i(6)(p10), + , + , inv(3)(p21q2?7) ,der(3)t(3;14)(q27;q32),der(8)t(8;8)(p21 t(3;13)(q27;q14) t(3;14)(q27;q32) mar[2]/47,idem, t(3;6)(q27;q15) ,del(6)(q13q25), ,t(14;18)(q32;q21) [1]/50,idem, ,del(6)(q13q25),t(14;18)(q32;q21),i(17)(q10), + mar [cp10] 3,t(3;12)(q27;q13) + [9]/46,X,t(X;7)(q22;q32) [5]/46,idem,t(1;5)(p2?2;q3?1) [1]/46,XX,t(1;7) R t(3;14)(q27;q32) 46% 53% [4]/46,XX [16] ,t(14;18)(q32;q21), ,del(6)(q1?5q2?3), ,?dup(6)(p23p24) [15] 46,XY [7] R t(3;14)(q27;q32) 65% 75% + ,add(12)(q24),t(14;18)(q32;q21) [4]/46,dem,add(16)(q22) R t(3;4)(q27;p11) 87% 94% t(2;3)(p11;q27) t(3;14)(q27;q32) 20, 1, der(1)t(1;5)(p13;p13),t(1;14)(q42;q32), del(2)(q32q34), + 3, 18,der(18)t(14;18)(q32;q21), + − + 20, 3, + + 2mar [5]/46,XY [15] + 2, + 3,t(3;14)(q27;q32) 1,del(1)(p21p36),del(6)(q13q21),del(13)(q13q14),t(14;18)(q32;q21) [7]/49,XX,idem, inv(3)(q13.2q27 or q12q29) + t(3;22)(q27;q11) + t(3;22)(q27;q11) t(3;22)(q27;q11) 1, X, X, 12, XX, t(3;14)(q27;q32) + t(3;22)(q27;q11) t(3;4)(q27;p11) t(3;16)(q27;p13) t(3;14)(q27;q32) + t(3;14)(q27;q32) + + mar [22] + 18, add(5)(p12),del(6)(q13q25),der(9)t(4;9)(q21;p22)del(4)(q31q33),der(10)t(5;10)(p12;q11), der(14)t(2;14)(q31;q32),der(18)t(14;18)(q32;q21), (q25;p15) [1]/46,XX,der(2)t(2;7)(p13;p11),t(4;11)(p11;q24 [4]/48,XY, 48,XY, (q31q33),der(14)t(8;14)(q21;p13),add(19)(p13.3) [20] [1]/47,idem,add(16)(q22), del(19)(p13)x2, 47,idem,t(10;19)(p14;q12) [1]/47,idem,t(12;14)(q13;q32) [1] 49,idem,der(12)t(?8;12)(q21;q24) [2]/50,idem add(18)(q23), 46,XX,der(1)t(1;1)(p36;q21), + [5]/51,idem, Results of cytogenetic and FISH analyses from 30 patients with B-NHL 1 DLBCL LN 46,XX, 2 B-NHL BM 48,XY,der(3)t(3;9)(p21;q13) 3 DLBCL4 LN DLBCL 47,X,add(Y)(p11), LN 90–95<4n>,XXXX, 5 DLBCL LN 48–52, 6 DLBCL LN 46,XY, 78 DLBCL LN FL 46,XX, LN 47,XY, 9 FL LN 47,XY,dup(1)(q12q25),der(2)dup(2)(p15p21 or p21p23)t(2;14)(q31;q32), Table 1 1011 MALT FL LN 46,XX, LN 46,XX, 12 DLBCL LN 47,XY, 13 FL LN 49,XX,trp(1)(q21q25), 1415 FL FL LN 48,Y,der(X)t(X;3)(p22;p23), LN 48,XX, Pt.No. Dx So Karyotype 1617 DLBCL LN FL18 47,XY, DLBCL LN LN 44,XY,inv(1)(p36q11),add(2)(q3?5), 47,XX,dup(1)(q21q32),

Leukemia Detection of BCL6 translocations by FISH D Sanchez-Izquierdo et al 1478 2% 2% 2% 2% 2% 2% 2% 2% 2% Ͻ Ͻ Ͻ Ͻ Ͻ Ͻ Ͻ Ͻ Ͻ 10% 10% 10% 10% 10% 10% 10% 10% 10% Ͻ Ͻ Ͻ Ͻ Ͻ Ͻ Ͻ Ͻ assay assay Ͻ marrow; PB, peripheral blood;

FISH Span- Flank- NR 10, NR add(X)(p11), NR − + 17, NR + mar [12]/46,idem, + 7[1]/92,idemx2 [2] NR 13,del(14)(q12), NR t(3;12)(q27;q12) + mar [16] + + 18, − mar [13]/45,XY, ,del(7)(q22q32), + 2mar [9]/46,XY [1] NR 18, + − 17, − 18, − 17, 9,del(13)(q14q22),add(17)(p13) [2]/ − − mar [2]/47,idem, 9,add(12)(q12), 5,add(5)(p13), NR i(3q),del(12p) 89% 1% + − + 8, inv(3)(p21q2?7) 6mar [16] 5, + ෂ + mar [9]/46,XY,del(10)(p11p12) [4]/46,XY [7] NR 31% 1% , 3 5, + + − 3, 4, + + 7,del(7)(q?22),der(14)t(14;18)(q32;q21),add(19)(q13), NR 51% 1% , + 9 [2]/47,XY,idem, − 5,del(5)(q31),add(11)(q14), − rearrangement are according to FISH results. 10,add(16)(q24) x2,del(22)(q11q13) x2, 6mar [3]/ 3,t(3;12)(q27;p11) 7 [4]/46,XX [10] NR − ,t(11;14)(q14;q32), + + t(1;3)(p34;q27), [11]/46,idem,del(6)(q?14q?23) [5] NR t(3;9)(q27;p21) − X, 10,

BCL6 + − 14 [30] 6,del(6)(q21q27), ,der(9)add(9)(p23)del(9)(q34),i(17)(q10), − + 1,dic(1;15)(p1?3;p11) x2,der(1)t(1;1)(p36;q25)del(1)(q23q42) x2, , − ,add(4)(p11), 13, − 4mar[3]/47,XY,idem, + 7 [2]/48,XX, + by FISH; NR, not rearrangement. ,der(6)t(6;7)(q13;q22),der(7)t(6;7)(q?21;q22),t(14;18)(q32;q21), 100<4n>,idem, ,der(6)t(6;8)(q1?3;q22),t(14;18)(q32;q21), , – 3,add(3)(q27) t(2;3)(p11;q27) + 11,del(11)(q23q25) x2 [3]/46,XY [4] mar[2]/96,XXYY, [10]/46,XX, [5] add(3)(q27) or der(3)t(1;3)(p35;q27) + BCL6 + 11, 13, + 18 [2]/97 − + 12) [4] (3)(q26) ෂ t(1;3)(p34;q27) del(3)(q21q27) t(3;22)(q27;q11) 3,der(3)del(3)(q14q23)hsr(3)(q14) X, 1,del(1)(p21), X, add(X)(p11),add(3)(p21),add(4)(p11),del(5)(q31),add(11)(q14), + del(3)(q27),der(9)t(3;9)(q27;p21) + + + + X,idem, X, − − mar [15]/46,XX [5] add(16)(q24),del(22)(q11q13), 47,Y, inv(3)(p21q2?7) x2,del(7)(q22q32) x2, 98,XXYY,idem, t(15;16)(q23;q23),der(?16)hsr(16)(q23)add(16)(q23),der(17)t(12;17)(q14;q22), der(9)del(9)(p22)t(1;9)(q13;q34), idic(17)(p11), + der(3)add(3)(p21)t(3;?5)(q27;q31) del(8)(p11 Continued In bold, cytogenetic alterations affecting 3q region are shown. Table 1 19 DLBCL LN 47,XX,add(1)(q43), Pt. No., patientPE, number; pleural Dx, effusion; diagnosis; R, So, rearrangement source of of samples; DLBCL, diffuse large B cell lymphoma; FL, follicular lymphoma; LN, lymph node; BM, bone In Span-assay and Flank-assay, the percentage of cells with 30 DLBCL LN 47,XY, 20 DLBCL LN 46,XX, 21 DLBCL LN 46,XY,dic(1;1)(p13;p34),t(2;5)(p13;p15.3), 22 DLBCL PB 49,XX, 23 DLBCL BM 51,XX, 24 DLBCL PB 48,XY,del(1)(q32q42), 25 DLBCL LN 46,X, 26 DLBCL LN 48,XX, 27 DLBCL PE 46,XY,add 28 DLBCL LN 46,XY, Pt.No. Dx So Karyotype 29 DLBCL LN 48,XY,der(1)t(1;1)(p36;q25)del(1)(q23q42), dic(1;15)(p1?3;p11),

Leukemia Detection of BCL6 translocations by FISH D Sanchez-Izquierdo et al 1479 three restriction enzymes (HindIII, XbaI and EcoRI; GibcoBRL, using BAC 211G3 were performed in VAL, CTB-1 and Oci- Gaithersburg, MD, USA) and probed with the BCL6 MBR gen- Ly8 cell lines, showing a partial translocation of the probe omic probe, F381.29 LDI-PCR wasperformed on the CTB-1 to the corresponding derivative chromosome. To confirm the cell line to investigate the t(14;22)(q32;q11) observed cyto- t(3;14)(q27;q32) and t(3;22)(q27;q11) in Oci-Ly8, CTB-1, genetically.32 LDI-PCR from the IGHJ segments and from 5Ј MD901, and MD903 cell lines with such rearrangements, S␮ to amplify the derivative partner chromosome was perfor- IGH and IGL gene probeswere co-hybridized with PAC med asdescribed. 35,36 LDI-PCR productswere cloned into 03P056 and PAC 03P061, respectively. In all cases, the fusion pTOPO (Invitrogen, San Diego, CA, USA) and sequenced. signals were identified on metaphase spreads and interphase cells, thus confirming the corresponding translocation. To validate the flank-assay assay, the two differently labeled Results BACs528E8 (red) and 690C8 (green) hybridizing centromeric and telomeric to the BCL6 gene, respectively, were also tested Design of FISH assays and selection of DNAprobes in four normal donors and in the seven cell lines. The normal signal pattern corresponding to two fused (overlapping) or col- Two different FISH approaches were evaluated. The first assay ocalized (defined as estimated signal distance less than once used PAC and BAC clones spanning the non-coding intron 1 the signal diameter) red–green signals were observed in the of BCL6 including the MBR (span-assay). For this purpose, sev- normal metaphase and interphase cells. The expected eral DNA clonescontaining BCL6 gene sequences from differ- rearranged pattern for a simple balanced BCL6 translocation ent genomic librarieswere mapped and testedby PCR for the was defined as one fused or colocalized red–green signal cor- presence of intron 1. Three clones mapping to 3q27 in normal responding to the normal chromosome 3, and two separated donorsand containing BCL6 intron 1 were selected: two red and green signals (one signal diameter minimum distance) PACs, RMC03P056 and RMC03P061, and RPCI-11 BAC from the two derivative chromosomes (Figure 2b). In case of 211G3. The latter clone containsthe entire BCL6 gene complex karyotypes, gains or losses of normal or derivative according to sequence analyses using the BLAST algorithm chromosomes could result in variant rearranged patterns, in and ispart of the contig HPFCctg 13301 which the presence of one or more isolated red or green sig- (http://genome.wustl.edu/). The second assay used two BAC nalsindicatesa break in the BCL6 locus. All the experiments clones flanking the MBR (flank-assay), also selected from the on positive and negative controls were evaluated in two lab- contig HPFCctg 13301 (RPCI-11 528E8 and 690C8). Accord- oratories independently on the same slides. In normal con- ing to sequence comparisons and physical mapping data, trols, 0 to 2% (mean 0.44%; s.d. 0.73%) of interphase nuclei neither probescontained BCL6 sequences and therefore were matched the rearranged pattern and thus, were classified as negative for the presence of intron 1. BAC 528E8 was cen- false-positives (cutoff level, mean +3 s.d., 2.6%). In 98 to tromeric, and 690C8 telomeric to the MBR, and the distance 100% of interphase nuclei, a normal pattern (consisting of between them was estimated at about 200 kb. A schematic 90% of fused or touching signals, and 10% of signals with a representation of the probes in the BCL6 region isshownin distance of less than one signal diameter) was found. In all Figure 1. All probeswere shownto map to 3q27 and to be the seven cell lines, irrespective of the BCL6 partner chromo- non-chimeric by FISH on metaphase cells from four donors some, 94 to 100% (mean 97.5%; s.d. 1.69%) of nuclei with a normal karyotype. matched the rearrangement pattern (cutoff level for false-negatives, 7.5%). Interobserver variability, defined as the difference in the percentage of positive and negative rearranged cells in Evaluation of the FISH assays in cell lines normal controls and cell lines detected by the two observers, was lower than 1%. Therefore, our results demonstrate our To validate the span-assay, the three BCL6 probescontaining FISH assay to be a sensitive and reproducible method for the the BCL6 MBR were tested on metaphase and interphase cells detection of different translocations of BCL6. from the seven B-NHL cell lines with various BCL6 translocations. PAC 03P056 was split and translocated to the corresponding derivative chromosome in all of the cell lines, thus Evaluation of the FISH assays in samples from patients giving a three signal pattern on interphase nuclei, two corre- with B-NHL sponding to the split probe and one to the normal allele (Figure 2a). On the contrary, PAC 03P061 wasnot translo- To test the clinical feasibility of the two FISH strategies, cated and therefore remained in the derivative chromosome samples from 30 patients with B cell NHL with cytogenetic 3 on metaphase spreads; a pattern corresponding to two sig- abnormalities of 3q27 were evaluated with both assays. In 21 nals was observed on interphase nuclei. Additional studies cases, the span-assay detected the translocation in 10–89% of interphase nuclei. In 18 cases, the dual-color flank-assay confirmed the BCL6 rearrangementsin 12–95% of interphase nuclei (Figure 3). Nevertheless, the three false-positive cases identified with the span-assay showed a total or partial trisomy 3 by cytogenetic analysis, and they were correctly classified as non-rearranged with the flank-assay (Figure 4). Considering the patients and cell lines together, 12 different partner chromosomal loci translocated to BCL6 were detected by the two assays involving 2p11, 3q13, 4p11, 6q15, 11q23, Figure 1 Schematic map of 3q27 region including MBR break- 12q13, 13q14, 14q32, 16p11, 16p13, 16q22 and 22q11. In point. PAC and BAC clones used for FISH assays are indicated. PAC 03P056 and PAC 03P061 both labeled in red were used in the span- five of these cases where sufficient high molecular weight assay. BAC 690C8 labeled in red and BAC 528E8 labeled in green DNA wasavailable, the involvement of the BCL6 MBR was were used in the flank-assay. confirmed by Southern blot analysis (data not shown). Apart

Leukemia Detection of BCL6 translocations by FISH D Sanchez-Izquierdo et al 1480 a b

Figure 2 Fluorescence in situ hybridization assays for the detection of BCL6 translocations on metaphase chromosomes and interphase nuclei. (a) Span-assay. On the left side, normal localization of the BCL6 PAC 03P056 in band 3q27, with the corresponding normal pattern consisting in two signals on nuclei cells. On the right side, a translocation involving BCL6 with a chromosome termed A is represented: the PAC is split by the translocation, giving a three signal pattern on interphase nuclei, two corresponding to the split probe and one to the normal allele. (b) Flank-assay. Normal colocalization of the two differentially labeled BACs RPCI-11 528E8 (red) and 690C8 (green), hybridizing centromeric and telomeric to the BCL6 gene, respectively. The normal hybridization pattern corresponding to two fused (overlapping) or colocalized (touching) signals was observed in the normal metaphase and interphase cells (left). The expected rearranged pattern in a simple balanced BCL6 translocation was deﬁned as one fused or colocalized red–green signal corresponding to the normal chromosome 3, and two separated red and green signals (one signal diameter minimum distance) from the two derivative chromosomes.

a b

Figure 3 Detection of BCL6 rearrangements in samples from patients with B-NHL by FISH. (a) Example of span-assay using PAC 03P056 labeled in red. Metaphase from a patient with DLBCL (case 16) with inv(3)(q13.2q27 or q12q29) according to cytogenetic analysis. The PAC is split by the inversion leading to two hybridization signals on the same chromosome 3 (arrowheads), conﬁrming the inv(3)(q13.2q27) with BCL6 rearrangement. (b) Example of ﬂank-assay using the BACs RPCI-11 528E8 (red) and 690C8 (green) in metaphase cell and interphase nuclei from a patient with FL (case 9) with a t(3;22)(q27;q11). The red–green double signal corresponds to the normal chromosome 3, whereas the single hybridization red signal corresponds to BAC 528E8 on the derivative chromosome 3, and the single hybridization green signal corresponds to BAC 690C8 translocated to the derivative chromosome 22.

from the previously reported translocations most of which meric breakpoint to MBR within BCL6 wasidentified, asboth have now been cloned,8,25,26,30 we identified two novel part- PAC 03P061 and PAC 03P056 were shown to be split. ner chromosome loci at 3q13 and 12q13. In the Seven patientswith t(3;14)(q27;q32) and three with inv(3)(q13q27) (case 16), PAC 03P056 was split whereas PAC t(3;22)(q27;q12) were also studied using IGH and IGL gene 03P061 wascompletely translocatedto 3q13, indicating that probescombined with BCL6 PACs, respectively. In all cases, the breakpoint wassituatedat the common region within MBR a fusion signal was also identified on interphase nuclei. These (Figure 3). In the t(3;12)(q27;q13) (case 18), a more centro- cases included a MALT lymphoma (case 10) with 12% of

Leukemia Detection of BCL6 translocations by FISH D Sanchez-Izquierdo et al 1481 vious results.14 In the CTB-1 cell line,31 which shows by standard cytogeneticsa t(14;22)(q32;q11), a BCL6-IGH gene fusion was identiﬁed by LDI-PCR. Cloning of an illegitimate switch rearrangement using primers situated in the 5Ј region of S␮ showed a translocation involving the ﬁrst intron of BCL6. The sequence of this breakpoint is shown in Figure 5a. RxFISH detected a t(8;14)(q24;q32), but no involvement of chromosome 3 was found (Figure 5b). However, a t(3;14;8)(q27;q32;q24) similar to that present in the Oci-Ly8 cell line, involving BCL6, IGH and C-MYC, wasdetected by FISH (Figure 5c).

Discussion

All currently available techniquesfor the identification of BCL6 translocations, such as conventional cytogenetic analysis, Southern blot analysis, or LDI-PCR, have significant technical or methodological limitations. Moreover, previously reported FISH strategies for the screening of BCL6 translocations also had important restrictions, and could only be applied to metaphase cells,22,23 or to cases with t(3;14)(q27;q32) on interphase nuclei.24 We report here a Figure 4 Diverse normal and rearranged patterns observed with the span-assay and the flank-assay, respectively, for the detection of novel FISH strategy using two differently labeled flanking BCL6 translocations in patients with B-NHL. (a) Normal pattern BCL6 BAC probes which provides a robust, sensitive, and observed in normal controls and non-rearranged cases. (b) Pattern of reproducible method for the detection of all common and BCL6 translocation in case 1, corresponding to one red–green double uncommon translocations involving the BCL6 gene in signal and two separated single colored signals corresponding to the interphase nuclei. derivative chromosome 3 (red) and to any derivative partner chromo- We investigated the sensitivity of our two-color interphase some (green); (c) An interphase cell from case 23 with a suspected add(3)(q27) involving BCL6. The span-assay identified a three signal FISH in normal controlsand cell lines.The cutoff level for pattern on interphase nuclei (data not shown) indicating possibly a false-positive signals was as low as 2.6%. In most other dual- BCL6 rearrangement, which was correctly classified as non- color flanking assays similar to ours, the reported cutoff levels rearranged with the dual-color flank-assay. The extra signal corre- have been considerably higher, ranging from 5 to 11%.37–39 sponds to a normal chromosome 3; (d) An interphase cell from case The incidence of false negatives will depend upon the normal 8 showing a BCL6 translocation pattern (red and green separated variance of probe separation within interphase nuclei, which signals) corresponding to the t(3;22)(q27;q11), with two additional red–green signals corresponding to two normal chromosomes 3. will be increased with widely spaced probes. The high sensitivity of our FISH assay may be explained by the short distance between the two probes, estimated at about 200 kb. In 90% interphase cells showing IGL-BCL6 gene fusion, thus of control nuclei the differentially labeled signals of the flank- confirming the FISH studies with the flanking probes. assay were fused or touching. Moreover, in the majority of the remaining control nuclei, the signals were separated by less than one signal diameter. Defining these colocalisations as a Rearrangements of chromosome 3q27 not involving normal pattern significantly increased the sensitivity of the BCL6 assay, and 200 cells per sample could be rapidly and easily evaluated. In other reported dual-color FISH assays, a translo- Twelve of the 30 cases (patients 19–30 in Table 1) with cyto- cation, eg a t(8;14)(q24;q32), t(14;18)(q32;q21), t(11;14) genetic abnormalitiesof 3q27 did not show BCL6 rearrange- (q13;q32), or t(11;18)(q21;q21), is indicated by two fusion sig- ment by FISH, including two translocations involving the IG nals (touching or superimposed yellow signal).17–20,37 Theor- gene loci, and other rare rearrangements. None of these cases etically, the highest sensitivity of a double-color FISH assay showed BCL6 gene rearrangement using either FISH assays. can be achieved by using probes spanning the breakpoint In two cases with a t(3;9)(q27;p21) or a t(3;12)(q27;q12), the regions of both translocation partners leading to two fusion flanking BACs528E8 and 690C8 were completely translo- signals in case of a translocation. Nevertheless, for routine use cated to the derivative chromosome, indicating a centromeric thisapproach isnot applicable to the BCL6 gene due to its breakpoint outside BCL6 at 3q27. In the other cases, the FISH promiscuity and different translocation partners. analysis did not confirm the BCL6 involvement asthe FISH We also validated our dual-color assay on seven cell lines probesremained on the derivative chromosome3. with diverse BCL6 translocations, demonstrating the cutoff level of false-negatives to be 7.5%. Most FISH reports lack such validation of the probes on a substantial number of cell Complex BCL6 rearrangements detected by FISH lines.17–24,37–39 The low percentage of false-negative cells indi- cates the high specificity of our assay, demonstrated by the Complex rearrangementswithin the BCL6 locuswere also extensive control studies performed. Our FISH experiments detectable using the flank assay. In the Oci-Ly8 cell line, a were also evaluated in two different laboratories on the same complex three-way translocation t(3;14;8)(q27;q32;q24) slides. Interobserver variability was very low, indicating that involving the BCL6, IGH and C-MYC genes, was confirmed the evaluation issimpleand reproducible. by standard cytogenetic analysis and RxFISH, confirming pre- Both FISH assays were compared on 30 patients with varied

Leukemia Detection of BCL6 translocations by FISH D Sanchez-Izquierdo et al 1482 a 3q27 translocations. The flank-assay was a superior tech- nique, which showed a lower false-negative rate and was much easier to analyze on interphase nuclei than the span- assay. Therefore, the flanking strategy can be easily applied to patient samples without the necessity of metaphase cells. Moreover, we successfully analyzed with the flanking strategy three G-banded slides, showing BCL6 rearrangements. There- fore our assay could be applied routinely to most patients and b this is one of its major advantages. Because partial or total trisomy of chromosome 3 is commonly seen in up to 10% of B cell NHL,1,2 we were also concerned that our assays be capable of differentiating cases with trisomy 3 from those carrying true BCL6 translocations. In 18 cases, a translocation involving BCL6 was confirmed by both assays, but three other rearrangements were only detected with the spanning probes. Based on the single-color assay, the presence of three signals on interphase nuclei may reflect either trisomy or a gene translocation. Our flank-assay confirmed these three false-positive results to be caused by trisomy 3. Our flanking strategy has been demonstrated to be of value in detecting 12 different BCL6 translocations, including common and uncommon chromosomal partners. According to extensive molecular studies, more than 95% of BCL6 translo- cationsclusterwithin the MBR. 8–10 Therefore, our dual-color assay will detect all these rearrangements. It is remarkable that in this series, approximately one-third of cases with 3q27 translocations defined cytogenetically did not show BCL6 c involvement. Thisisin agreement with previousreports. 22,23 The results obtained in two of our cases, where the two flanking probes were translocated to the partner chromosomes, as well as in a case with a t(1;3)(q21;q27) reported by Wlodraska et al,22 suggest that breakpoints centromeric to the BCL6 gene may be involved in some of these patients. In the other cases without BCL6 involvement according to our FISH assays, we could not rule out the possibility that some of these cases involve the alternative breakpoint region (ABR) located 200– 270 kb telomeric to MBR, asreported by Chen et al40 in a t(2;3)(q21;q27). Unfortunately, we did not have access to any cases containing such t(2;3) translocations and therefore were not able to validate our assays for the detection of breaks in the ABR. The true prognostic significance of BCL6 rearrangementsin Figure 5 Molecular and cytogenetic characterization of CTB-1 cell DLBCL is an issue that remains controversial and could be line. (a) Sequence of a BCL6-IGH gene fusion identified by LDI-PCR addressed with the application of our dual-color assay. An in the CTB-1 cell line. Cloning of an illegitimate switch rearrangement 11 was performed by LDI-PCR using primers situated in the 5Ј region initial report by Offit et al in 1994 indicated that BCL6 gene of S␮ and showed a translocation involving the first intron of BCL6. rearrangement in DLBCL correlated predominantly with Sequence derived from the CTB-1 cell line in black and underlined extranodal disease and better clinical outcome. Subsequent iscompared with the germline IGH S␮ sequence in red and the BCL6 studies using similar Southern blot analysis, however, have intron 1 sequence in blue. (b) Example of RxFISH study on metaphase not confirmed these results.7,13 These differences may be cell from the CTB-1 cell line. A t(8;14)(q24;q32) isshown,but no related to either difficultiesin the identification of all hetero- involvement of chromosome 3 was identified. This cell line was reported to have a t(14;22)(q32;q11) by standard cytogenetic analy- logous BCL6 gene rearrangementswith current molecular sis.31 By using RxFISH, the t(14;22) was correctly classified as a techniques (although all studies should have detected most reciprocal translocation between chromosomes 12p11 and 22q11, BCL6 rearrangements) or selection bias in the inclusion of with the involvement of chromosome 14q32 in the complex t(3;14;8) patientswith enough tumor DNA for Southern blot. Addition- (q27;q32;q24) shown below. (c) A complex t(3;14;8)(q27;q32;q24), ally, our results show that cytogenetically detectable breakpo- involving BCL6, IGH and C-MYC genes, was detected by FISH. In intsin 3q27 might not involve BCL6. The routine application the image, two probesflanking of the IGH locus(cosmidc ␣1-IgH centromeric to IGH in green and PAC 1098L17 telomeric to IGH in of our assay to patients with B-NHL could help to refine the red) were used. A red–green hybridization signal corresponds to the true incidence of BCL6 translocations and, thus, to clarify the normal chromosome 14; both probes are translocated and therefore prognostic role of BCL6 rearrangements in these neoplasms. not detected on the derivative chromosome 14 (marked with an However, there may be significant biological heterogeneity arrow): PAC 1098L17 (red) mapsto 3q27 fusingto BCL6 (data not among the BCL6 translocations themselves. Thus, in some ␣ shown) whereas cosmid c 1-IgH (green) mapsto 8q24, fusingto c- BCL6 translocations, it is clear that the involved gene may be MYC (data not shown). deregulated, aswell as BCL6 as a consequence of translocation. For example, in the cell line Karpas231, which exhib- itsa t(3;11)(q27;q23.1), both BCL6 and BOB-1/OBF-1,aB

Leukemia Detection of BCL6 translocations by FISH D Sanchez-Izquierdo et al 1483 cell-specific transcriptional activator, are expressed at high B, Lista F, Filippa DA, Rosenbaum A, Ladanyi M, Jhanwar S, level.29 The possible contribution of partner genes such as this Dalla- Favera R, Chaganti RSK. Rearrangement of the BCL-6 gene to the pathogenesis or evolution of B cell malignancies is not asa marker in diffuselarge-cell lymphoma. N Engl J Med 1994; 331: 74–80. known. During the preparation of thismanuscript,a report 12 Lo Coco F, Ye BH, Lista F, Corradini P, Offit K, Knowles DM, based on combined use of Southern blot and LDI-PCR Chaganti RS, Dalla-Favera R. Rearrangementsof the BCL6 gene approachesappeared, indicating that the involvement of BCL6 in diffuse large cell non-Hodgkin’s lymphoma. Blood 1994; 83: with non-IG loci translocation was associated with a signifi- 1757–1759. cantly worse outcome than those with BCL6/IG translo- 13 Muramatsu M, Akasaka T, Kadowaki N, Ohno H, Yamabe H, Eda- cations.41 Combined use of the BCL6 flanking probe set mura S, Dor S, Mori T, Okuma M, Fukuhara S. Rearrangement of the BCL6 gene in B-cell lymphoid neoplasms: comparison with described here with the IG probeswould allow thispoint to lymphomas associated with BCL2 rearrangement. Br J Haematol be assessed in a larger number of patients. 1996; 93: 911–920. 14 Chaganti SR, Rao PH, Chen W, Dyomin V, Jhanwar SC, Parsa NZ, Dalla-Favera R, Chaganti RS. Deregulation of BCL6 in non- Acknowledgements Hodgkin lymphoma by insertion of IgH sequences in complex translocations involving band 3q27. Genes Chromosom Cancer 1998; 23: 328–336. Thiswork wassupportedby grantsfrom the Fondo de 15 Skinnider BF, Horsman DE, Dupuis B, Gascoyne RD. Bcl-6 and Investigacioń Sanitaria FIS-98/0491 and BECE-99/6023, by the Bcl-2 protein expression in diffuse large B-cell lymphoma and fol- Generalitat Valenciana, Conselleria de Educacion y Ciencia licular lymphoma: correlation with 3q27 and 18q21 chromosomal POST-99/12–61, by the Deutsche Krebshilfe Grants 10–1556- abnormalities. Hum Pathol 1999; 30: 803–808. Schl4 and 10–1641-De1, and by the IZKF Kiel. We thank 16 Rowley JD, Dıáz MO, Espinosa R, Patel YD, Melle E, Ziemin S, Taillon-Miller P, Lichter P, EvansG, KerseyJH, Ward DC, Domer Gianluca Gaidano (Torino, Italy), for the OCI-Ly8 cell line; PH, Le Beau MM. Mapping chromosome band 11q23 in human Hitoshi Ohno (Kyoto, Japan), for the YM cell line; Wen-Lin acute leukemia with biotinylated probes: identification of 11q23 Kuo and Joe Grey (University of California, San Francisco) for translocation breakpoints with a yeast artificial chromosome. Proc PACsRMC03P056 and RMC03P061; Mariano Rocci (Bari, Natl Acad Sci USA 1990; 87: 9358–9362. Italy), for PACs1019H10 and 1098H17; and M Ordon˜ez, E 17 Siebert R, Matthiesen P, Harder S, Zhang Y, Borowski A, Zu¨lhke- Cervello, R Marques, M Botia, A Schneider, F Kahlert, L Jenisch R, Metzke S, Joos S, Weber-Matthiesen K, Grote W, Schle- Askari, C Becher, D Schuster and R Zu¨hlke-Jenisch for excel- gelberger B. Application of interphase fluorescent in situ hybridization for the detection of the Burkitt translocation t(8;14)(q24;q32) lent technical assistance. in B-cell lymphomas. Blood 1998; 91: 984–990. 18 Dierlamm J, BaensM, Stefanova-Ouzounova M, Hinz K, Wlodar- ska I, Maes B, Steyls A, Driessen A, Verhoef G, Gaulard P, Hage- References meijer A, Hossfeld DK, De Wolf-Peeters C, Marynen P. Detection of t(11;18)(q21;q21) by interphase fluorescence in situ hybridiz- 1 Rowley JD. Recurring chromosome abnormalities in leukemia and ation using API2 and MLT specific probes. Blood 2000; 96: lymphoma. Semin Hematol 1990; 27: 122–136. 2215–2218. 2 Offit K, Chaganti RSK. Chromosomal aberrations in non-Hodgkin’s 19 Siebert R, Matthiesen P, Harder S, Zhang Y, Borowski A, Zuhlke- lymphoma. Biological and clinical correlations. Hematol Oncol Jenisch R, Plendl H, Metzke S, Joos S, Zucca E, Weber-Matthiesen Clin North Am 1991; 5: 853–869. K, Roggero E, Grote W, Schlegelberger B. Application of 3 Ye BH, Lista F, Lo Coco F, Knowles DM, Offit K, Chaganti RS, interphase cytogenetics for the detection of t(11;14)(q13;q32) in Dalla-Favera R. Alterationsof a zinc finger-encoding gene, BCL- mantle cell lymphoma. Ann Oncol 1998; 9: 519–526. 6, in diffuse large-cell lymphoma. Science 1993; 262: 747–750. 20 Taniwaki M, Silverman GA, Nishida K, Horiike, Misawa S, Shima- 4 Kerckaert JP, Deweindt C, Tilly H, Quief S, Lecocq G, Bastard C. zaki C, Miura I, Nagai M, Abe M, Fukuhara S, Kashima K. Translo- LAZ3, a novel zinc-finger encoding gene, isdisruptedby recurring cationsand amplificationsof the BCL-2 gene are detected in chromosome 3q27 translocations in human lymphomas. Nat interphase nuclei of non-Hodgkin’s lymphoma by in situ hybridiz- Genet 1993; 5: 66–70. ation with yeast artificial chromosome. Blood 1995; 86: 1481– 5 Baron BW, Nucifora G, McCabe N, Espinosa R, Le Beau MM, 1486. McKeithan TW. Identification of the gene associated with the 21 Martinez-Climent JA, Thirman MJ, Espinosa R, Le Beau MM, Row- recurring chromosome translocations t(3;14)(q27;q32) and ley JD. Detection of 11q23/MLL rearrangementsin infant leuke- t(3;22)(q27;q11) in B-cell lymphomas. Proc Natl Acad Sci USA miaswith fluorescence in situ hybridization and molecular analy- 1993; 90: 5262–5266. sis. Leukemia 1995; 9: 1299–1304. 6 Miki T, Kawamata N, Arai A, Ohashi K, Nakamura Y, Kato A, 22 Wlodarska I, Mecucci C, Stul M, Michaux L, Pittaluga S, Hernań- Hirosawa S, Aoki N. Molecular cloning of the breakpoint for 3q27 dez JM, Cassiman JJ, De Wolf-Peeters C, Van der Berghe H. Fluor- translocation in B-cell lymphomas and leukemias. Blood 1994; escent in situ hybridization identifieschromosomalchanges 83: 217–222. involving 3q27 in non-Hodgkin’slymphomaswith BCL6/LAZ3 7 Bastard C, Deweindt C, Kerckaert JP, Lenormand B, Rossi A, Pez- rearrangement. Genes Chromosom Cancer 1995; 14: 1–7. zella F, Fruchart C, Duval C, Monconduit M, Tilly H. LAZ3 23 Roumier C, Galie`gue-Zouitina S, Bastard C, Soenen V, Lai JL, rearrangementsin non-Hodgkin’slymphoma correlation with his- Denis C, Buchonnet G, Kerckaert JP, Cosson A, Fenaux P, tology, immunophenotype, karyotype, and clinical outcome in Predhomme C. FISH analysis with a YAC probe improves detec- 217 patients. Blood 1994; 83: 2423–2427. tion of LAZ3/BCL6 rearrangement in non-Hodgkin’slymphoma. 8 Akasaka H, Akasaka T, Kurata M, Ueda C, Shimizu A, Uchiyama Hematol J 2000; 1: 117–125. T, Ohno H. Molecular anatomy of BCL-6 translocations revealed 24 Ueda Y, Nishida K, Miki T, Horiike S, Kaneko H, Yokota S, Mis- by long-distance polymerase chain reaction-based asssay. Cancer awa S, Abe T, Kshima K, Taniwaki M. Interphase detection of BCL- Res 2000; 60: 2335–2341. 6/IgH fusion gene in situ hybridization. Cancer Genet Cytogenet 9 Dalla-Favera R, Ye BH, Lo Coco F, Chaganti RSK. BCL-6 and the 1997; 99: 102–107. molecular pathogenesis of B-cell lymphoma. Cold Spring Harb 25 Dallery E, Galie`gue- Zouitina S, Collyn-d’Hooghe M, Quief S, Symp Quant Biol 1994; 59: 117–123. DenisC, Hildebrand MP, Lantoine D, Deweindt C, Tilly H, Bas- 10 Ye BH, Chaganti S, Chang CC, Niu H, Corradini P, Chaganti RSK, tard C, Kerckaert JP. TTF, a gene encoding a novel small G protein, Dalla-Favera R. Chromosomal translocations cause deregulated fuses to the lymphoma-associated LAZ3 gene by t(3;4) chromo- BCL6 expression by promoter substitution in B-cell lymphoma. somal translocation. Oncogene 1995; 10: 2171–2178. EMBO J 1995; 14: 6209–6217. 26 Kobayashi H, Miki T, Nakamura Y, Kaneko Y. Bcl-6 gene stays on 11 Offit K, Lo Coco F, Louie DC, Parsa NZ, Leung D, Portlock C, Ye the der(3) chromosome in a t(3;22)(q27;q11) or t(3;14)(q27;q32)

Leukemia Detection of BCL6 translocations by FISH D Sanchez-Izquierdo et al 1484 associated with non Hodgkin’s lymphoma. Genes Chromosom novel gene (BCL9) at chromosome 1q21. Blood 1998; 91: Cancer 1995; 13: 225–227. 1873–1881. 27 Tweeddale M, Lim B, Jamal N, Robinson J, Zaleberg J, Lockgood 35 Satterwhite E, Sonoki T, WillisTG, Siebert R, Nowak R, Arriola G, Mindem MD, Messner HA. The presence of clonogenetic cells EL, Liu H, Harder L, Price HP, Gesk S, Steinemann D, Oscier DG, in high-grade malignant lymphoma: a prognostic factor. Blood Schlegelberger B, Tucker PW, Dyer MJS. Deregulated expression 1987; 69: 1307–1314. of BCL11A,aBCL6-interacting transcriptional repressor, in lym- 28 Nacheva E, Dyer MJS, Fisher P, Stranks G, Heward JN, Marcus phomaswith abnormalitiesof chromosome2p13. Proc Natl Acad RE, Grace C, KarpasA. c-myc translocationsin the de novo B cell Sci USA (in press). lineage acute leukemiaswith t(14;18) (cell linesKarpas231 and 36 WillisTG, Jadayel DM, Coignet LJ, Abdul-Rauf M, Treleaven JG, 353). Blood 1993; 82: 231–240. Catovsky D, Dyer MJ. Rapid molecular cloning of rearrangements 29 Galiegue-Zouitina S, Quief S, Hildebrand MP, DenisC, of the IGHJ locus by long-distance inverse PCR. Blood 1997; 90: DetourmigniesL, Lai JL, Kerckaert JP. The B cell transcriptional 2456–2464. coactivator BOB1/OBF1 gene fuses to the LAZ3/BCL6 gene by 37 Rack KA, Salomon-Nguyen F, Radford-Weiss I, Gil MO, Schmitt t(3;11)(q27;q23.1) chromosomal translocation in a B cell leukemia C, Belanger C, Nusbaum S, Vekemans M, Valensi F, Macintyre line (Karpas231). Leukemia 1996; 10: 579–587. EA. FISH detection of chromosome 14q32/IgH translocations: 30 Yonetani N, Akasaka T, Aksaka H, Ohno H, Okuma M, Miura I, evaluation in follicular lymphoma. Br J Hematol 1998; 103: Takahashi N, Miyanishi S, Okumura A, Muramatsu M, Fukuhara 495–504. 38 MatutesE, Carrara P, Coignet L, Brito-Babapulle V, Villamor N, S. Molecular featuresof a new human lymphoma cell line carrying Wotherspoon A, Catovsky D. FISH analysis for BCL-1 rearrange- both BCL-2 and BCL-6 gene rearrangement. Oncogene 1998; 17: ments and trisomy 12 helps the diagnosis of atypical B cell leu- 971–979. kaemias. Leukemia 1999; 13: 1721–1726. 31 Uchida Y, Miyazawa K, Yaguchi M, Gotoh A, Iwase O, Ohyashiki 39 Taniwaki M, Matsuda F, Jauch A, Nishida K, Takashima T, Tagawa K, Toyama K. Establishment of a novel B-lymphoma cell line, CTB- S, Sugiyama H, Misawa S, Abe T, Kashima K. Detection of 14q32 I, with strong Fas antigen expression having chromosomal translo- translocation in B-cell malignancies by in situ hybridisation with cation t(14;22). Int J Oncol 1997; 10: 1103–1107. yeast artiﬁcial chromosome clones containing the human IgH 32 Schlegelberger B, Metzke S, Harder S, Siebert R. Classical and gene locus. Blood 1994; 83: 2962–2969. molecular cytogeneticsof tumor cells.In Wegner RD (ed.). Diag- 40 Chen W, Butler M, Rao PH, Chaganti SR, Louie DC, Dalla-Favera nostic Cytogenetics. Springer-Verlag: Berlin, 1999, pp 151–185. R, Chaganti RS. The t(2;3)(q21;q27) translocation in non-Hodg- 33 Espinet B, Sole F, Salido M, Lloveras E, Abella E, Besses C, Serrano kin’slymphoma displaysBCL6mutationsin the 5 Ј regulatory S, Woessner S, Florensa L. Application of cross-species color region and chromosomal breakpoints distant from the gene. banding (RxFISH) in the study of T-prolymphocytic leukemia. Oncogene 1998; 17: 1717–1722. Haematologica 2000; 85: 607–612. 41 Akasaka T, Ueda C, Kurata M, Akasaka H, Yamabe H, Uchiyama 34 WillisTG, Zalcberg IR, Coignet LJ, WlodarskaI, Stul M, Jadayel T, Ohno H. Nonimmunoglobulin (non-Ig)/BCL6 gene fusion in dif- DM, Bastard C, Treleaven JG, Catovsky D, Silva ML, Dyer MJ. fuse large B-cell lymphoma results in worse prognosis than Molecular cloning of the translocation t(1;14)(q21;q32) deﬁnes a Ig/BCL6. Blood 2000; 96: 2907–2909.

Leukemia