Rearrangement of the BCL2 gene is an important parameter for the differential diagnosis of non-Hodgkin lymphomas. Although a relatively large proportion of breakpoints is clustered, many are missed by standard PCR. A FISH assay is therefore desired. Up to now, a lack of probes flanking the BCL2 gene has limited the possibilities for a FISH assay to an approach based on colocalization of probes for BCL2 and the immunoglobulin heavy chain (IGH) locus. Intrinsically high rates of false positive nuclei and high interobserver variability make such assays unsuitable for use on lymphoma tissue samples, where tumor cells often form only a minority of the cell population. Using YAC end cloning techniques and screening of a PAC library, we have isolated PAC clones flanking the BCL2 gene. Using these PACs, and several cosmid clones in the second BCL2 intron, we developed a segregation-based interphase FISH assay with two probe combinations enabling separate detection of 5' and 3' (mbr/mcr) breakpoints. The assay was applied to a series of 40 follicular lymphomas. To evaluate the results, the same lymphomas were analyzed by DNA fiber FISH with a 600-kb set of BCL2 DNA clones labeled in alternating colors in combination with a color barcode covering the IGH locus. This approach allowed precise mapping of BCL2 breakpoints, and simultaneously showed juxtaposition of IGH genes to BCL2. Comparison of the results of interphase and fiber FISH showed complete correlation. Five cases were negative with both FISH techniques as well as with Southern blotting. Interestingly, all of these 5 cases lacked BCL2 overexpression as determined by immunohistochemistry, against 3 of 35 rearrangement-positive follicular lymphomas. Furthermore, absence of t(14;18) seemed to be correlated with a higher histologic grade (grades 2 and 3 according to Berard). These data indicate that the segregation-based interphase FISH assay detects 100% of BCL2 rearrangements. Because interpretation of the results is straightforward and requires no extensive experience, this assay may be the best available diagnostic test for BCL2 rearrangement. Genes Chromosomes Cancer 27:85-94, 2000.

We have applied two-colour fluorescence in situ hybridization (FISH) to DNA fibers and combined it with digital imaging microscopy for the mapping of large cosmid contigs. The technique was validated using a set of unique plasmids and a cosmid contig both originating from the thyroglobulin (Tg) gene and previously mapped by restriction analysis. The resolution proved to be close to the theoretical lower limit of approximately 1 kb, ranging > or = 400 kb. Subsequently a 400 kb cosmid contig derived from a DMD-YAC was directly mapped by Fiber-FISH. The resulting map is in full agreement with the restriction map. Two-colour Fiber-FISH mapping thus showed to be capable for accurately sizing gaps and overlaps, and to identify chimeric or repeat sequence containing cosmids across a 400 kb region at once. The generated 400 kb 'colour bar-code' was subsequently used to map two DMD deletion breakpoints in patient DNA with an accuracy of 1-2 kb. The results underscore the value of this method for the delineation of chromosomal rearrangements for positional cloning and single patient clinical studies.

Several hematologic malignancies are associated with specific chromosomal translocations. Because of the dispersed distribution, chromosomal breakpoints may be difficult to detect using molecular techniques. We present a new application of a recently developed method, DNA fiber fluorescence in situ hybridization (fiber FISH), which allows direct visualization and mapping of chromosomal breakpoints. We tested this method for detection of the t(11;14)(q13;q32) translocation in mantle cell lymphoma. In DNA fiber FISH, a series of fluorochrome-labeled DNA probes covering several hundreds of kilobasepairs is hybridized to linear DNA molecules (or fibers) prepared from frozen tissue or intact cells. By using alternate fluorescent colors, a potential breakpoint region is stained in a color barcode pattern. Breaks in this region will split the barcode in two complementary parts, from which the breakpoint position can be derived. We used a 250-kb barcode covering the BCL-1 locus to detect 11q13 breakpoints in 20 well-characterized mantle cell lymphomas. A t(11;14) was shown by cohybridization of these probes with probes for the Ig heavy chain locus at 14q32. In 18 of 20 mantle cell lymphomas, a breakpoint within the 11q13/BCL-1 barcode was shown by the presence of multiple, complementary translocation products. Fusion of 11q13 and 14q32 sequences on single fibers indicating t(11;14)(q13;q32) was found in all 18 breakpoint-positive mantle cell lymphomas. In one additional case, fusion of an intact 11q13 barcode with 14q32 sequences indicated a breakpoint 100 kb centromeric of the major translocation cluster of BCL-1. Within the 120-kb region of BCL-1, breakpoints were widely scattered. This explains why, so far, a BCL-1 breakpoint had been detected by Southern blot analysis in only 10 of 19 cases. DNA fiber FISH analysis showed a t(11;14) in 95% of mantle cell lymphoma. The results indicate that DNA fiber FISH is a rapid, simple, and equally powerful method for detection of clustered and dispersed translocation breakpoints.

Immunoglobulin class switching usually involves deletion of part of the immunoglobulin CH region. By DNA fiber fluorescence in situ hybridization (FISH) with a barcode of probes covering the DH, JH, and CH genes, the configuration of the entire CH region can be visualized on single DNA molecules. Using this technique, we have studied class switching in three types of B-cell neoplasia, mantle-cell lymphoma (MCL), follicular lymphoma (FL) and hairy cell leukemia (HCL), representing B cells in, respectively, pregerminal center, germinal center, and postgerminal center stages of development. In MCL and FL, simultaneous detection of the t(11;14) and t(14;18) breakpoint with probes for the BCL-1 and BCL-2 loci, respectively, allowed differentiation between productive and nonproductive alleles. In none of 10 MCL cases was class switching detected. In 21 HCL, all nonimmunoglobulin M (IgM) cases had class-switch deletion consistent with the expressed isotype on at least one allele. In FL, however, a peculiar pattern of CH rearrangement was observed. In IgM expressing FL, the translocated alleles had switched in 11 of 13 cases, and the nontranslocated allele showed complex rearrangements downstream from the Cμ-Cδ genes in 9 of 13 cases. These downstream rearrangements may reflect tumor-specific deregulation of the class-switch machinery. All seven immunoglobulin G (IgG) expressing FL showed class switching on both alleles. Fiber FISH analysis also showed several polymorphisms. The most frequent one, present on 38% of all analyzed alleles, consisted of an extra Cγ gene or pseudogene in the 3′ cluster. © 1998 by The American Society of Hematology.