Mechanisms of chromosomal rearrangement in the human genome

Abstract

Many human cancers are associated with characteristic chromosomal rearrangements, especially hematopoietic cancers such as leukemias and lymphomas. The first and most critical step in the rearrangement process is the induction of two DNA double-strand breaks (DSB). In all cases, at least one of the two DSBs is generated by a pathologic process, such as (1) randomly-positioned breaks due to ionizing radiation, free radical oxidative damage, or spontaneous hydrolysis; (2) breaks associated with topoisomerase inhibitor treatment; or (3) breaks at direct or inverted repeat sequences, mediated by unidentified strand breakage mechanisms. In lymphoid cells, one of the two requisite DSBs is often physiologic, the result of V(D)J recombination or class switch recombination (CSR) at the lymphoid antigen receptor loci. The RAG complex, which causes the DSBs in V(D)J recombination, can cause (4) sequence-specific, pathologic DSBs at sites that fit the consensus of their normal V(D)J recombination signal targets; or (5) structure-specific, pathologic DSBs at regions of single- to double-strand transition. CSR occurs specifically in the B-cell lineage, and requires (6) activation-induced cytidine deaminase (AID) action at sites of single-stranded DNA, which may occur pathologically outside of the normal target loci of class switch recombination regions and somatic hypermutation (SHM) zones. Recent work proposes a seventh mechanism: the sequential action of AID and the RAG complex at CpG sites provides a coherent model for the pathologic DSBs at some of the most common sites of translocation in human lymphoma - the bcl-2 gene in follicular lymphoma and diffuse large B-cell lymphoma, and the bcl-1 gene in mantle cell lymphoma.

Figure 1

Features of some neoplastic chromosomal translocations. Chromosomal translocations in somatic cells are most often encountered in the context of neoplasia. Constitutional translocations arise during gametogenesis or early divisions of the fertilized egg. It is useful to dissect the chromosomal translocation process into two phases: the causes of the breaks and then the rejoining of the four broken ends. [Most chromosomal translocations involve two DSBs, though some can occur during DNA replication (see text).] In neoplastic translocations, the causes of the two DSBs can be different. In this example, one chromosome break is of the CpG-type (see text), and the other can be a V(D)J-type or a random DSB. In addition, breaks can be of the class switch type (CSR-type) or somatic hypermutation type (SHM-type). The V(D)J-type, CSR-type, SHM-type, and CpG-type are limited to lymphoid cells, and further limited by stage of differentiation (see text). Random breaks are thought to be due to oxidative free radicals (reactive oxygen species), ionizing radiation (IR), or topoisomerase failures. The joining of the four DNA ends is done by nonhomologous DNA end joining (NHEJ) in most cases. The causes of the breaks all generate heterogeniety at the break site, and NHEJ creates additional heterogeneity at the joining site. Hence, even when both translocation junctions are sequenced, one can only trace the site or boundary of the original breaks to a zone of at least several nucleotides (often longer, such as 10-20 bp). Less commonly, no nucleotides are lost from either DSB site, and this permits determination of the specific phosphodiester bonds at which the original DSBs occurred. This proved useful for defining the CpG-type break (see text). Some of the specific translocations discussed here have been previously diagrammed in more detail [26-28]. The normal V(D)J and class switch recombination processes have been previously diagrammed [5-9] .