BCL6 and Hypermutation

1.5.1. Somatic Hypermutation in Normal B-Cells

The increase in affinity of serum for antigen on repeated immunization is accounted for both by changes in immunoglobulin V gene usage and by accumulation of point mutations in the V genes (50). The characteristics and requirement for cis-acting control elements are well recognized (51) and, more recently, an enzyme (activation induced deaminase [52,53]) has been found to be essential not only for this process but also for immunoglobulin class switching and gene conversion (54-56). Somatic hypermutation occurs within the lymph node germinal center (57), which is also the site of BCL-6 expression. The BCL-6 gene from normal germinal center B-cells has been sequenced to find out whether all actively transcribed genes in this location bear point mutations. Surprisingly, BCL-6 was found to be mutated but other genes, for example, c-MYC and alpha-fetoprotein, were not (58,59). However, it has since been shown that several proto-oncogenes as well as BCL-6 can be mutated in normal B-cells (60). Although calculation of the rate of mutation is complicated by the low absolute numbers of mutations, it may be that BCL-6 mutates at a lower rate than immunoglobulin genes (58,59) in normal B-cells.

1.5.2. BCL-6 Mutationin Lymphomas

Mutation clearly could provide a mechanism for dysregulation of BCL-6 expression, which could play a role in lymphomagenesis. Analysis of material from two patients and a B-cell line that expresses BCL-6 (61) revealed deletions of varying extent but with a shared common region of about 200 bp (Fig. 1) in the first intron of BCL-6. This study suggests that there may be negative control elements whose function is disrupted and leads to constitutive BCL-6 expression. Indirect support for this view is provided by the finding of a 184-bp region of homology between mouse and human sequence that overlaps the deleted region. Such regions of intronic homology are likely to be

37695

tcatgatca ttattttacc ttttaattct tettttttcc

A 37741 gctcttgcca aatgctttgg ctccaagttt tctatgtgta tctattgata taaatgtata B 37741 gctcttgcca aatgctttgg ctccaagttt tctatgtgta tctattgata taaatgtata C 37741 gctcttgcca aatgctttgg ctccaagttt tctatgtgta tctattgata taaatgtata]

A 37301 tatttattta ttctagctgt caggtgttaa aataaatgcc gaagattagt cccacgtctc

B 37801 tatttattta ttctagctgt caggtgttaa aataaatgcc gaagattagt cccacgtctc C 37801 jtatttattta ttctagctgt caqgtgjt taa aataaatgcc gaagattagt cccacgtccc

37861 tcccaccata ggatatacfat TGTTATGTAT TTATTATTAT TATTGTTGTC TTTGAGTGAA

37921 tcgggcggtt tggggaggct tttgccacgc tccgttgtgt tgttttggtt tttggaaagg

37931 AGGTGGAGGA GAGGAAGGAG GGGAATTAGG GGGCGGCCGG AGCA.GAGAGG AGGAGACAGT

a 3S041 GCTTGGGGGG TGATTCGGGC TAGTCTGGGG GCTGTCTGGC CCCAGACCGC GGAGAGGACG

B 38041 GCTTGGGGGG TGATTCGGGC TAGTCTGGGG GCTGTCTGGC CCCAGACCGC GGAGAGGACG C 33041 GCTTGGGGGG TGATTCGGGC TAGTCTGGGG GCTGTCTGGC CCCAGACCGC GGAGAGGACG

3 22111 122 31 2 A 38101 CGCGCTCGCG CTCTCCCTCT TTCTGCTGCT GCTTGCGTAC GGCTTGTGAT CTCTCTGGAT 212111 11 122B112 1 2 111 1 4 11111 Ï 111 TT

B 33101 CGCGCTCGCG CTCTCCCTCT TTCTGCTGCT GCTTGCGTAC GGCTTGTGAT CTCTCTGGAT C 38101 CGCGCTCGCG CTCTCCCTCT TTCTGCTGCT GCTTGCGMC GGCTTGTGAT CTCTCTGGAT

a 3 3161 TCGTGCGGCT GTGTTTTTTC CCTCTTTTCT CGCTTGCAAA CTGCTTTCCT TGCTCCG 11 22 2 1 1 1 11 1 11 1 B 38161 TCGTGCGGCT GTGTTTTTTC CCTCTTTTCT CGCTTGCAAA CTGCTTTCCT TGCTCCG C 3 8161 TCGTGCGGCT GTGTTTTTTC CCTCTTTTCT CGCTTGCAAA CTGCTTTCCT TGCTCCG

Fig. 1. Compilation of data on the first intron of human BCL-6 showing the relationships between mutations and regions of interest. The nucleotide numbering is from a human BAC clone (Homo sapiens 3 BAC RP11-211G3 Roswell Park Cancer Institute Human BAC Library) accession number AC072022. A region of human and mouse homology (37702-37880) is in plain lower case. Such regions are often the site of control elements (54). The region italicized in lower case is a putative negative control or silencer region as identified by Kikuchi et al. (55). The underlined nucleotides form the core region of deletions found by Nakamura et al. (53), either in the malignant cells of patients with diffuse large cell lymphoma or in a BCL-6 expressing cell line. The mutation data are taken from three sources. Sequence A derived from Capello et al. (58) shows a 26-bp block of nucleotides (37800-37826) with numbers of mutations per nucleotide presented above. Sequence B is from Lossos et al. (56,60) and again the number of mutations found at each position is presented above the sequence. Sequence C describes two boxed stretches of nucleotides (37800-37826 and 38117-38138) which are short "mutational hotspots" discovered by Artiga et al. (59). The two areas shaded in gray show the clear correspondence between the different data sets. There is an RGYW motif at position 37815. The breakpoint cluster defined by Akasaka et al. (36) is italicized and in upper case. It is apparent that this 521-bp sequence contains regulatory elements, flanked by mutational hotspots and is also the site of breakpoint clusters.

involved in transcriptional control (62). More direct evidence for the existence of negative transcriptional control elements or silencers comes from experiments in which luciferase reporter constructs were driven by sections of the first intron of BCL-6 (63). Although there is a control region within the first exon, there also appear to be control elements (in the mouse/human homology and deleted in lymphoma sections) within the first intron of BCL-6.

Furthermore, gel mobility shift assays have shown the presence of specific protein binding complexes (61,64), again supporting the importance of the first intron as a focus for transcription factor binding. Thus, one possibility is that disruption of the core region by hypermutation leads to loss of the normal mechanisms that "turn off" BCL-6. To pursue this concept, sequencing from cases of lymphoma has been conducted by various groups.

1.5.3. Location of Mutations in Lymphoma and Effects on BCL-6 Expression

Great consistency has been seen in results obtained from sequencing the first intron of BCL-6 in lymphomas (Fig. 1). There are two mutational hotspots (65-68), one of which contains an RGYW motif, which is a particular focus of targeting by the mutational mechanism (69) and, overall, approx 50% of DLCL may bear mutations (70). Thus, there are indications that the mechanism of BCL-6 hypermutation and immunoglobulin gene hypermutation are the same, but some evidence that this is not the case is provided by an analysis of chronic lymphocytic leukemia, a disease that can be divided into two subgroups depending on the presence of immunoglobulin gene hypermutation. Some of the cases of chronic lymphocytic leukemia lacking immunoglobulin gene hypermutation did have BCL-6 changes (71). Others, however, have not confirmed this (72), although they have demonstrated a constant association between immunoglo-bulin gene and BCL-6 intron mutations. Lymphocyte predominant Hodgkin's disease has recently been recognized to have a germinal center origin, and sequencing of material derived from single cells has shown that all cases bear mutations (73).

An unresolved question is: can mutations indeed lead to increased BCL-6 expression? Despite the fact that lymphomas, like other malignant diseases, are clonal the expression of BCL-6 is not uniform within tissue samples, and this is a surprising result in itself. For example, one series found that 10% of cases had fewer than 20% of BCL-6-positive cells on immunohistochemistry, and 25% of cases had greater than 80% positive cells whilst most cases had intermediate numbers of positive cells (67). Using the proportion of positive cells as an index of protein expression, no correlation was found with the presence of mutations. However, such an analysis has the weakness that variations in BCL-6 expression during the cell cycle are not taken into account, and nor is an assessment of how amounts of BCL-6 within individual positive cells compares to that within normal germinal center B-cells.

More recently, it has been specifically demonstrated that there are two auto-regulatory BCL-6 binding sites within the first exon of the gene. Occupancy of these sites by BCL-6 causes repression of BCL-6 transcription. Not only is this autoregulatory region removed in the majority of translocations (77), but a subset of point mutations of this region found in diffuse large cell lymphoma also deregulate BCL-6 expression (75).

1.5.4. Correlation With Clinical Outcome

The presence of mutations may be associated with bulkier disease at presentation as assessed by lactate dehydrogenase level (70). No consensus exists on the effects of intronic mutations on survival. One series of cases suggests that mutation of the second hotspot (Fig. 1; 38117-38138) confers a better overall and disease free survival compared with those having a wild-type sequence (67). However, others (70) found that although there was no effect on overall survival, patients with mutated BCL-6 introns did have an improved 5-yr disease-free survival.

Acquisition of mutations during the course of evolution of follicular lymphoma has been examined (76). Although relapsed follicular lymphomas did not have new intronic mutations, most transformed cases did bear mutations not seen at presentation or relapse. This association may mean that dysregulated BCL-6 has a role in the transformation of follicular lymphoma or may be a reflection of an altered pattern of gene expression in which mutation is more likely.

Was this article helpful?

0 0

Post a comment