More on 13q14

I have been promising to write something more about the 13q14 deletion in CLL. I have already visited this subject several times and claim some ownership of it since it was in my lab that it was first discovered by David Oscier. There is a good review by Ulf Klein and Riccardo Dalla-Favera in Seminars in Cancer Biology published this month and if you can get hold of a copy I would recommend reading it. Unfortunately it is behind a paywall so I will summarise what it says about 13q14.

Based on a large number of cases of CLL it has been discovered that there is a minimally deleted region (MDR) that is always gone on at least one of the chromosomes. The amount that is missing is very variable, but this MDR is constant.

So what is it that goes missing?

1] A long non-coding RNA called deleted in leukemia (DLEU)-2. This bears no similarity to any other long non-coding RNA, but it is conserved among all vertebrates.
2] The first exon of the DLEU-1 gene; another sterile transcript, but unlike DLEU2 this gene is not evolutionarily conserved.

It was later discovered by Croce and Calin that two microRNAs, miR-15a nd miR-16-1 are found within intron 4 of DLEU2 and indeed require the DLEU2 promoter region for expression.

There are two adjacent genes that have been considered as possibly being involved in CLL, DLEU5 and KCNRG, but since these are not missing in all cases and KCNRG is not expressed in B cells, this seems unlikely.

Current thinking is that the two miR genes are implicated in the pathogenesis of CLL, either directly, or through the loss of DLEU2 promoter regions or perhaps in other cases by an epigenetic switch-off. miRs work by targeting mRNA and there is no doubt from biostatistical algorithms that there are miR-15a/16-1 binding sites in a number of mRNAs involved in proliferation and apoptosis; prominent among them have been the cyclins, CCND1 and CCND3, the cyclin-dependent kinase, CHK6, and the apoptotic gene, BCL2. After a lot of technical dispute these genes may all still be candidates, and the miR genes may target multiple regulators controling the transition of G0/G1 to S in the cell cycle.

To help sort out this problem, Klein and Dalla-Favera have created a mouse model. Mouse models have been a problem for CLL and I have already pointed out the problems with the TCL-1 mouse model produced by Croce's lab. The disease that these mice produce is like an aggressive CLL, but certainly not the indolent disease that we see in most humans.

Klein and Dalla-Favera have developed two transgenic lines, one mimicking the loss of the entire MDR and the second containing a specific deletion of miR-15a/16-1 within intron 4 of the DLEU2 gene. Both mouse lines developed clonal lymphoproliferative conditions late in life - the former with a penetrance of 40% and the second with a penetrance of 25%. Most were MBL or CLL, but occasionally CD5-negative proliferations were seen including diffuse large B-cell lymphoma.

There are two points to make obout the mouse models: first they showed the same sort of sterotypy that CLL shows and second the line with the larger deletion had the more aggressive leukemias.

Another clue that they are on the right lines comes from the CLL-like disease that is sometimes seen in New-Zealand Black (NZB) mice. These CD5+ lymphoproliferations occur at a greater age than in the miR mice and are less aggressive. Three loci have been linked with their development including a point mutation close to miR-16-1.

What this all suggests to me is that the basic lesion is a loss of control of the transition from G0/G1 to S in the cell cycle because of the loss of the genes that downregulate the entry into the cell cycle. This therefore expands the mature B-cell pool. Normally, entry into cell cycle would be in response to strong antigenic stimulation, but with loss of control, weak antigens or even self-antigens might be enough. Oligoclonality leads to monoclonality and a relatively small pool of such initiating antigens leads to stereotypy. The amount of chromosomal loss determines the aggressiveness of the proliferation. The basic miR los is sufficient to set the process going, but loss of other genes (such as DLEU5 or, further away on the chromosome, Rb1 or indeed anything in between) adds aggression to the tumor. I would predict therefore that we might see a difference between CLL with the MDR alone and CLL with a larger deletion.