ALBERT

Description: Activating mutations and deletions affecting specific NOTCH1 protein domains have been recently shown to occur widely in T-cell neoplasia, e.g. in T-acute lymphoblastic leukemia (T-ALL). However, knowledge of NOTCH1 chromosomal alterations is largely based on a single cell line model (SUP-T1) with t(7;9)(q35;q34) in which NOTCH1 truncated at exon 24 is juxtaposed with TCRB. We describe the characterization of a novel rearrangement, t(9;14)(q34.3;q11) in two T-cell lymphoma cell lines, HD-MAR and HT-1. FISH analysis using fosmid clones and sequencing of fragments identified by long distance inverse PCR showed that in both cases t(9;14) effected tail-to-tail juxtaposition of intron 27 of NOTCH1 with TCRA genes, namely 5′-TRAV40 in HD-MAR, and intron 2 of TRAV5 in HT-1. Thus, in both cell lines t(9;14) places NOTCH1, truncated immediately 3′ of the HD-domain, under transcriptional control of TCRA. The 14q11.2 breakpoints in HD-MAR and HT-1 lie, respectively, near the proximal E-delta enhancer and amid a cryptic enhancer region represented by a cluster of T-cell specific DNase-I hypersensitive sites. Western blotting revealed prominent expression of truncated activated NOTCH1 polypeptides, ranging in size from 100 to 115 kDa in both cell lines. Antibodies recognizing ANK and TAD domains, believed essential for inducing T-ALL, detected the aberrant polypeptides. Moreover, treatment with gamma-secretase inhibitor (GSI) altered expression patterns of NOTCH1 polypeptides and induced growth inhibition due to G0/G1 cell cycle arrest in both t(9;14) cell lines, in stark contrast to GSI-resistant SUP-T1 cells wherein truncation occurs before the heterodimerization (HD) domain. (Another recently described t(7;9) cell line (CUTLL1) which is GSI-sensitive also carries a NOTCH1 breakpoint at intron 27.) The same protein species were not detectable by antibodies recognizing the transmembrane domain of NOTCH1 which requires GS for exposure suggesting nuclear access requires GS-cleavage. Immunostaining confirmed extranuclear blocking of NOTCH1 in response to GSI in HD-MAR/HT-1 but not in SUP-T1. In contrast, repression of HES1 occurred in response to GSI irrespective of NOTCH1 breakpoint location, suggesting its non-involvement in growth signaling. In addition to providing cell line models for a new NOTCH1 disease translocation, these data suggest that the sensitivities of T-cell neoplasias bearing NOTCH1 translocations may critically depend on whether 9q34 breakpoints lie upstream or downstream of the HD domain.

Description: BCL6 (at 3q27) is highly expressed in mature B-cells within the germinal center where immunoglobulin (IG) genes undergo somatic hypermutation and class switch recombination. BCL6 is believed to downregulate apoptotic damage responses by p53 induced by physiological DNA breakage therein. In germinal-center-derived B-cell non-Hodgkin lymphomas (B-NHL), such as follicular lymphoma and diffuse large B-cell lymphoma, BCL6 may be deregulated by juxtaposition with IG loci, or by promoter exchange with a variety of other (usually) hematopoietic genes of which 〉20 have been described hitherto. FISH analysis of well characterized cell lines from patients with B-NHL revealed 9/20 (45%) with previously undetected, exclusively non-IG, BCL6 rearrangements. These included 6 cell lines with breakage at the canonical breakpoint cluster region (BCR), 5 of which involved new partner loci, at 1q42, 3q12, 7q32, 8q24, 12p11, the sixth (MBNL1 at 3q25) having been described once previously. Fine-mapping with tile-path fosmid clones revealed that translocations involving the BCL6 BCR with partner loci at 1q42, 3q12, 7q32, and 8q24 were unlikely to involve classical promoter exchange, as no strand-compatible partner genes or transcripts map to these breakpoints. This was tested both by RT-PCR using primers for BCL6 BCR together with all known compatible ESTs and in silico genes at the breakpoint loci (according to UCSC May 2004 assembly), and by 5′-RACE. Intriguingly, the majority of these partner breakpoints mapped to common fragile sites, including FRA1H (1q42), FRA7H (7q32), and FRA8C (8q24). Treatment of the t(3;7)(q27;q32) cell line with aphidicolin preferentially induced breakage within the FRA7H breakpoint on the non-rearranged homolog, implying a stabilizing effect for this translocation. This cell line thus affords a unique model for chromosome fragility contributory to hemopoietic neoplasia. The break at 8q24 was effected by formation of a non-reciprocal der(8)t(3;8)(q27;q24) which generated head-to-head juxtaposition of the BCL6 regulatory region with that of MYC which was over-expressed in the cell line involved. Similarly, 1/3 remaining B-NHL cell lines which evidenced BCL6 breakage outwith the BCR also involved juxtaposition with MYC which again was over-expressed. A further non-BCR B-NHL cell line exhibited hemizous BCL6 loss concomitant with gross somatic hypermutation of the residual upstream regulatory region and BCL6 overexpression. In summary, our data show: 1) an unexpected variety of novel BCL6 rearrangement at the apparent expense of IG locus juxtaposition; 2) a novel (third) category of BCL6 partnership involving the BCR region apparently together with non-coding DNA; 3) the likely involvement of BCL6 in MYC cis-regulation, and 4) the contribution of chromosomal fragility to lymphomagenic BCL6 rearrangement. Future studies will address the contribution of DNA stress-stability at BCL6 translocation breakpoints.