| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Center for Human Genetics
* and the Department of Pathology,
¶ Catholic University Leuven, Leuven, Belgium; the Department of Haematology,
Belfast City Hospital, Belfast, Northern Ireland; the School of Biomedical Sciences,
University of Ulster, Coleraine, Northern Ireland; DakoCytomation Denmark A/S,
Glostrup, Denmark; and the Department of Pathology,
|| Royal Bournemouth Hospital, Bournemouth, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The IGH-mediated translocations are relatively rare in B cell chronic lymphocytic leukemia (CLL),3, 4, 5 which is the most common form of leukemia in adults and shows a highly variable clinical course. The CLL cells display a phenotype of mature activated B lymphocytes expressing CD19, CD5, and CD23 and having reduced levels of membrane IgM, IgD, and CD79b. The co-expression of CD19 and CD5, however, is also characteristic for mantle cell lymphoma (MCL), which, in contrast to CLL, is usually an aggressive disease hallmarked by the t(11;14)(q13;q32)/IGH-CCND1 rearrangement. Differential diagnosis between CLL and leukemic MCL is sometimes challenging.5 Given that up to 30% of MCL cases have immunophenotypic features characteristic of B-CLL, albeit usually with atypical, pleomorphic morphology,6 immunophenotyping alone is insufficient to exclude a diagnosis of MCL. Therefore, in the Belfast City Hospital (Belfast, Northern Ireland) all suspected CLL cases have been routinely examined by rapid interphase FISH with the LSI IGH/CCND1 assay to identify t(11;14)-positive MCL cases, in addition to examination for CLL typical cytogenetic aberrations. During this analysis, a subset of CLL cases with IGH signal pattern suggestive of IGH translocation different from t(11;14) has been identified. Fourteen of these cases were subjected to cytogenetic, further FISH, and VH mutation analysis. Results are discussed below.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cytogenetic and Fluorescence in Situ Hybridization (FISH) Analysis
Cytogenetic analysis was performed on peripheral blood cells cultured for 72 hours after stimulation with 12-O-tetradecanoyl phorbol-13-acetate. In each case, 3 to 26 metaphase cells were analyzed. Karyotypes were described according to ISCN.7
Probes used for FISH analysis included the LSI IGH/CCND1, IGH/CCND1 XT, IGH, and IGH/BCL2 assays (Vysis, Downers Grove, IL), 14q32 subtelomeric probe (P1 60H4),8 and a set of nine bacterial artificial chromosome (BAC) clones covering IGH (11771, 417P24, 676G2, 965B13, 112H5, 101G24, 12F16, 47P23, and 2548B8).9 Noncommercial probes were directly labeled with Spectrum Orange- and Spectrum Green-dUTP (Vysis) using nick translation. Two IGH BAC clones, 676G2 and 965B13, co-hybridized with 15q11, and 2548B8 co-hybridized with the 16p11.2 region.9 Co-hybridization signals, however, were weaker and did not significantly perturb interphase FISH analysis.
FISH experiments were evaluated using an Axioplan 2 fluorescence microscope equipped with a charge-coupled device Axiophot 2 camera (Carl Zeiss Microscopy, Jena, Germany) and a MetaSystems Isis imaging system (MetaSystems, Altlussheim, Germany). One to 18 abnormal metaphases and/or 200 interphase cells were evaluated in each FISH experiment.
Fluorescence immunophenotyping and interphase cytogenetics as a tool for the investigation of neoplasm (FICTION) analysis with CD20 antibody (DakoCytomation, Glostrup, Denmark) and two pairs of IGH BAC clones performed on smears from normal peripheral blood and reactive lymph node followed previously published protocol.10
VH Mutation Analysis
Mutational status of VH was determined using the BIOMED-2 method as previously described.11, 12
Mutations and VH segment usage were identified by comparison with the germline sequence using Ig Blast database (http://www.ncbi.nlm.nih.gov/igblast/) and IMGT (http://imgt.cines.fr/). Sequences containing more than 2% deviation from the germline were considered as somatically mutated rather than genetic polymorphisms.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Given that the accompanying normal cells in these samples showed either two separated green signals (Figure 1B)
or two double green signals (together four green signals) (Figure 1C)
, molecular events underlying this aberrant signal constellation (2R3G) were not clear. We presumed that three IGH/green signals arose by IGH breakage due to chromosome translocation (split of one of two green signals in Figure 1
, B cells), but a partial deletion of IGH (loss of one of four green signals in Figure 1 C
cells) or a fragmental excision of the IGH locus followed by its insertion into another gene locus or a trisomy 14q32 could not be excluded. To clarify these intriguing FISH findings, 14 of these CLL cases (further referred to as index cases) with available peripheral blood cells at time of the study were subjected to cytogenetic and further detailed FISH analysis.
|
. The involvement of 14q32 harboring the IGH locus was detected in case 1 with a del(14)(q24q32) and in case 13 with a t(14;18)(q32;q21). The typical CLL abnormalities, –13/del(13q), trisomy (or tetrasomy) 12, del(6q), and loss of chromosome 17/17p, were found in six, three, two, and two cases, respectively.
|
FISH analysis with LSI IGH was performed on interphase and, when applicable, on metaphase cells. This analysis showed aberrant signal constellations in 10 to 88% of analyzed cells from all 14 cases (Table 1)
. The patterns included loss of the 3' IGHFP red signal (1F1G) detected in case 1 with del(14)(q24q32) (Figure 1D)
, complete or partial loss (diminished) of the IGHV/green signal (1F1R or 1F1RGd, respectively) found in 11 cases (nos. 2 to 12) (Figure 1 E and F)
, and the split 1F1R1G pattern observed in case 13 with a t(14;18) and in case 14 without cytogenetically detectable t(14q32) (LSI IGH/BCL2 negative). As a control, we examined nine new CLL cases recently registered in our laboratory. In three of them complete or partial loss of the IGHV/green signal was also found (data not shown). In control experiments performed on four blood samples from normal individuals (200 interphase evaluated in each experiment), the cut-off level of diminished/lost IGH signals was 1%. Further evidence of the predominant IGHV deletions underlying the aberrant "three green" IGH signal pattern in the present cases was provided by metaphase FISH analysis with LSI IGH/CCND1, LSI IGH/CCND1 XT, and LSI IGH/BCL2. In contrast to LSI IGH/CCND1, the IGH probe included in the latter assays is made up of a single 1.5-Mb probe covering essentially the entire locus. FISH analysis with these three assays performed on four cases (nos. 3, 8, 9, 10) showed the same results on metaphase cell level: normal green signal on one chromosome 14 and diminished green signal on a second chromosome 14. The difference was noticed at the interphase cell level because only LSI IGH/CCND1 produced a two double green signals pattern in a significant number of cells. Normal interphase cells hybridized with LSI IGH/CCND1 XT and LSI IGH/BCL2 displayed two regular green signals, and abnormal cells were hallmarked by the 1F1RGd pattern, supporting our previous conclusion.
The IGHV deletions detected in the index cases were further mapped by FISH using nine BAC clones covering the entire IGH locus labeled with either Spectrum Orange- (SO) or Spectrum Green (SG)-dUTP (Figure 2)
.9
This analysis performed on 11 cases (nos. 2 to 12), with loss of the IGH-"green signal," showed presence of the IGHC sequences covered by 417P24 and 11771 and loss of signal from various IGHV BAC clones (Figure 2)
. Loss of signals from three, four, and five IGHV clones was detected in three, three, and six cases, respectively. All of them showed normal presence of two 47P23 (IGHV terminal) signals. Cases 8 and 9 were additionally analyzed with the 14q subtelomeric probe (P1 60H4), which in both cases gave two regular signals. Unexpectedly, loss of signals from two or three IGHV BAC clones were also found in two control cases without LSI IGH-"green signal" losses, including case 1 with del(14)(q24q32) (1F1G-"red signal" deletion) and case 13 with a t(14;18) (1F1R1G-split signals) (Figure 2
, bottom panel).
|
. The most frequently used was VH1–69, identified in six cases (2 m/4 um). The VH3–48 was used in two cases (m), and single cases used the VH1–2 (m), VH3–30 (um), VH3–53 (m), VH4–31 (um), VH4–34 (m), and VH4–39 (um) genes.
Comparison of fine VH FISH mapping data available in 13 cases with results of the VH analysis showed that in all studied cases a distal breakpoint of VH correlated with the genomic localization of the used VH gene (Figure 2)
. This finding suggested that the detected deletions represent physiological loss of genomic sequences associated with a somatic V-D-J assembly in precursor B cells.
To confirm this supposition, we further examined by FISH the status of the VH region in a cohort of B cell, T cell, and myeloid malignancies. Given a common deletion of 965B13 (the most proximal VH BAC clone) and constant presence of 47P23 (the most terminal VH BAC clone) in the index CLL cases (Figure 2)
, we used these two clones labeled with SO- and SG-dUTP, respectively, for FISH analysis of a further 17 B cell leukemias/lymphomas, six T cell leukemias/lymphomas, and nine cases of myeloid malignancies listed in Table 2
. The balanced co-localized signals (2–4) were detected in two B cell lymphomas and in all analyzed myeloid and T cell malignancies, including HDML2, the Hodgkin’s lymphoma cell line of T cell origin. The remaining 13 B cell leukemia/lymphoma cases revealed aberrant FISH patterns manifested by either the diminished 965B13 signal (2/17) or by loss of one (10/17) or two (1/17) 965B13 signals. The two additional Hodgkin’s lymphoma cell lines of B cell origin included in this study, L-1236 and KM-H2, displayed either a total loss of 965B13 signals (L-1236) or complex unbalanced (1R4–8G) FISH pattern (KM-H2). Fifteen of these B cell leukemia/lymphoma cases were documented by various known 14q32/IGH translocations including t(3;14)(q27;q32) (3x), t(8;14)(q24;q32) (3x), t(11;14)(q13;q32) (3x), t(14;18)(q32;q21) (3x), and t(14;19)(q32;q13) (3x) (Table 2)
. Metaphase FISH analysis of these cases showed a common loss of the VH proximal signal from chromosome 14 not involved in translocation and presence of normal co-localized VH signal on the derivative partner chromosome (eg, der(3)t(3;14); Figure 1G
). In one case with t(8;14), however, loss of both proximal (965B13) signals from normal chromosome 14 and derivative chromosome was found.
|
Finally, we examined the occurrence of VH deletions in normal B cells. For this purpose, FICTION analysis using CD20 antibody and the 47P23-SG/2548B8-SO FISH assay were performed on two samples of normal peripheral blood and reactive lymph node. In these experiments a subset of CD20-positive cells with a 1F1G/1F1GRd signal pattern, similar to that seen in B cell leukemia/lymphoma, was identified (Figure 1H)
. In contrast, CD20-negative cells showed normal 2F signal configuration.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, center panel). The discontinuity of this probe and different levels of chromatin condensation in interphase cells possibly account for occurrence of either two or four (two double) green IGH signals in normal cells (Figure 1 B and C)
. Our further studies clearly showed that these three green signals found in interphase CLL cells mostly represented loss of one of four normal (two double) signals due to the partial IGH deletions (VH in 11/14 and CH in 1/14 cases) and only rarely was split of one of two green signals due to the initially expected t(14q32/IGH) (2/14 cases). In no case was the aberrant IGH FISH pattern caused by a cryptic excision/insertion of IGH or trisomy of 14q32/IGH. Our supposition was supported by metaphase FISH analysis of these CLL cases with LSI CCND1, LSI IGH, and additionally with LSI CCND1 XT and LSI IGH/BCL2, which constantly generated one regular and one diminished (or lost) green signal on chromosomes 14. Further evidences supporting this model were provided by cytogenetic, detailed FISH and molecular VH analysis of 14 cases representing this subset of CLL and by additional FISH studies of the VH region performed on approximately 40 B cell leukemia/lymphoma cases with known and unknown VH usage. All VH deletions detected in these cases were submicroscopic, interstitial, varying in size but constantly involving the most proximal VH sequences adjacent to the D-J region, and in all of them the distal breakpoint strictly correlated with a genomic localization of the used VH gene. They were detected exclusively in B cell neoplasms and not found in malignant proliferations of T cell and myeloid origin. Interestingly, these cryptic VH deletions were also demonstrated in both Hodgkin’s lymphoma cell lines of B cell origin (but not in HDML2 of T cell origin), providing additional evidence of somatic IGH rearrangements in typical Reed-Sternberg cells. Moreover, using FICTION we documented VH deletions in a subset of normal CD20+ B cells from the peripheral blood and reactive tonsil but not in CD20-negative cells.
It is also worth noting that our metaphase FISH studies, performed on several t(14;var)(q32;var), demonstrated that VH deletions constantly affected the normal chromosome 14 expected to carry a functionally rearranged IGH locus (Figure 1G)
, and only rarely chromosome 14 involved in translocation. These findings are fully concordant with the known mechanisms of somatic V-D-J recombination and allelic exclusion resulting in expression of only single productive immunoglobulin heavy and light chain rearrangements in a given lymphocyte.13, 14, 15, 16
According to this concept, B cells with failures in a primary V-D-J assembly (VDJ–/VDJ0), achieve productive rearrangement by ongoing independent recombination of IGH segments on the other chromosome 14 (VDJ–/VDJ+). We believe that biallelic VH deletions found by FISH in a few analyzed B cell leukemia/lymphoma cases reflect either this latter mechanism or a defective immunoglobulin allelic exclusion leading to expression of two functional IGH alleles (VDJ+/VDJ+), as detected in approximately 6% of CLL cases.17, 18
Moreover, we presume that analogous deletions in the CH region might be detected by FISH in malignancies originating from mature B cells subjected to immunoglobulin class switch recombination, as shown using fiber-FISH approach19
and array CGH.20
This mechanism, however, was not operating in the index CLL cases expressing IgM (data not shown).
Genomic losses from the 14q32/IGH region have been occasionally detected in CLL and other B cell lymphomas,4, 21, 22, 23 generating difficulties in evaluation of FISH results.24 Recently, Fink et al22 postulated that loss of the IGHV signal in CLL represents usage of a telomeric IGHV gene segment, as 9 of 11 of their cases expressed distally located VH 1–69, VH 3–72, and VH 3–74. In contrast to our study, however, they noted a low incidence of IGHV deletions in other B cell malignant disorders (1/6). This discordance (17% versus 87% in the present series) is caused by a low resolution of applied LSI IGH assay when compared with our BAC FISH strategy.
Finally, we want to stress that the present IGH alterations were identified by chance during routine interphase FISH analysis of CLL cases with LSI IGH/CCND1, and in this particular condition VH deletions were manifested by the unusual "three green signal" pattern in interphase CLL cells. In our daily practice, however, diminished or lost VH signal has been recurrently observed in a wide spectrum of B cell malignancies routinely analyzed with LSI IGH and other FISH assays containing IGH probe. Although the phenomenon of physiological VH deletions in B cells is well known, practical implications of these cryptic genomic losses for routine FISH analysis of B cell malignancies are often overlooked. As shown here, these somatic events may seriously complicate molecular cytogenetic analysis of B cell leukemia and lymphoma and lead to false conclusions. Therefore, to avoid misinterpretation of FISH data, FISH users should be aware not only of different molecular characteristics of commercial IGH probes routinely used for diagnostic purposes but also of cryptic VH deletions that in B cell disorders do not manifest oncogenic events but normal physiological features of precursor B cells. To unravel IGH abnormalities generating aberrant FISH patterns in B cell malignancies, we propose two slightly different decision trees, one for users of LSI IGH/CCND1 (Figure 3A)
and one for users of LSI IGH/CCND1 XT, -/BCL2, and -/CMYC (Figure 3B)
. Given the different characteristics of IGH probes included in these assays (discontinuous in LSI IGH/CCND1 and continuous in the remaining assays), these probes generate different aberrant interphase FISH patterns (Figure 3 A and B)
. In any case, however, when an aberrant (not specific for the applied assay) IGH signal pattern is seen, we recommend use of the LSI IGH dual-color break-apart rearrangement probe to distinguish between true 14q32/IGH-associated translocation and other acquired IGH aberrations (eg, nonreciprocal translocation, interstitial 14q deletion) or somatic VH deletions. The latter events may affect one or both chromosomes 14 (biallelic VH deletions are marked by asterisk in Figure 3
) hallmarked by either a diminished or lack of IGVH green signal. Note that the 1F1R pattern is produced by cells with either a nonreciprocal t(14;var)(q32;var) (loss of a partner derivative chromosome) or an extended somatic VH deletion that can be distinguished by further FISH analysis with, eg, 14q subtelomeric probe (loss of one signal in the former case). Likewise, the 1F1G signal pattern is produced by cells with either a nonreciprocal t(14;var)(q32;var) [loss of der(14)] or interstitial 14q deletion involving IGH that can be further distinguished by analysis with, eg, a near-centromeric 14q11 probe (loss of one signal in the former case). Of course, all IGH-associated problems could be avoided by direct use of gene-specific dual-color break-apart rearrangement probes (LSI CCND1, LSI BCL2, LSI CMYC). However, their thresholds are much higher than those of IGH-related dual-fusion assays and these characteristics limit their application in cases with a low incidence of abnormal cells. Finally, we recommend replacement of LSI IGH/CCND1 by LSI IGH/CCND1 XT, which, like LSI IGH/BCL2 and /CMYC, comprises a single large IGH probe. The evaluation of FISH experiments performed with these latter assays is much easier, and results are less misleading because their IGH probes are not producing a two double green signal pattern in normal cells.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Supported in part by the Fund for Scientific Research of Flanders (FWO-Vlaanderen), Grant G.0338.01, K.U. Leuven Research Foundation (BIL03/12), and by the Belfast City Hospital Trust Haematology Research Funds. C.M. was funded by the Northern Ireland Department of Education and Learning. V.V. is a fellow of the Belgian Foundation against Cancer, and P.V. is a Clinical Investigator of FWO-Vlaanderen. This work presents research results of the Belgian Programme of Interuniversity Poles of Attraction initiated by the Belgian State, Prime Minister’s Office, Science Policy Programming.
Accepted for publication August 18, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. R. Gunn, M. S. Mohammed, M. E. Gorre, P. D. Cotter, J. Kim, D. W. Bahler, S. N. Preobrazhensky, R. A. Higgins, A. R. Bolla, S. H. Ismail, et al. Whole-Genome Scanning by Array Comparative Genomic Hybridization as a Clinical Tool for Risk Assessment in Chronic Lymphocytic Leukemia J. Mol. Diagn., September 1, 2008; 10(5): 442 - 451. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Bagg Malleable Immunoglobulin Genes and Hematopathology - The Good, the Bad, and the Ugly: A Paper from the 2007 William Beaumont Hospital Symposium on Molecular Pathology J. Mol. Diagn., September 1, 2008; 10(5): 396 - 410. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Davidsson, A. Andersson, K. Paulsson, M. Heidenblad, M. Isaksson, A. Borg, J. Heldrup, M. Behrendtz, I. Panagopoulos, T. Fioretos, et al. Tiling resolution array comparative genomic hybridization, expression and methylation analyses of dup(1q) in Burkitt lymphomas and pediatric high hyperdiploid acute lymphoblastic leukemias reveal clustered near-centromeric breakpoints and overexpression of genes in 1q22-32.3 Hum. Mol. Genet., September 15, 2007; 16(18): 2215 - 2225. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
