Search

SummaryWhereas chromosomal translocations are common pathogenetic events in cancer, mechanisms that promote them are poorly understood. To elucidate translocation mechanisms in mammalian cells, we developed high-throughput, genome-wide translocation sequencing (HTGTS). We employed HTGTS to identify tens of thousands of independent translocation junctions involving fixed I-SceI meganuclease-generated DNA double-strand breaks (DSBs) within the c-myc oncogene or IgH locus of B lymphocytes induced for activation-induced cytidine deaminase (AID)-dependent IgH class switching. DSBs translocated widely across the genome but were preferentially targeted to transcribed chromosomal regions. Additionally, numerous AID-dependent and AID-independent hot spots were targeted, with the latter comprising mainly cryptic I-SceI targets. Comparison of translocation junctions with genome-wide nuclear run-ons revealed a marked association between transcription start sites and translocation targeting. The majority of translocation junctions were formed via end-joining with short microhomologies. Our findings have implications for diverse fields, including gene therapy and cancer genomics.

We describe novel products of V(D)J recombination in which signal sequences become joined to coding elements, in contrast to the standard reaction whose products are junctions of two signal sequences or two coding elements. In this variant reaction, the recombination machinery evidently recognizes signal sequences and introduces strand breaks at the normal positions, but then connects the elements in unusual combinations. The lack of fixed directionality indicates that recombination sites are not uniquely aligned when strand exchange occurs. The discovery of these variant junctions suggests a model for the evolution of the antigen receptor loci.

A cell line transformed by Abelson murine leukemia virus, called PD, is capable of carrying out kappa-gene rearrangement while growing in culture. Subclones of PD have diverse kappa-gene structures, and some derivatives show evidence of continued joining activity after as many as three subclonings. Analysis of PD sublineages has shown that a rearranged chromosome can undergo secondary kappa-gene rearrangements, producing either a new rearrangement or a deletion of C kappa. Although the PD line actively rearranges its kappa genes, its rearranged heavy-chain genes show little variation, and there is no rearrangement of lambda genes. In PD subclones, DNA fragments representing the reciprocal product of kappa-gene rearrangement are often evident, and they may undergo either further rearrangement or deletion. The implications of multiple rearrangements on a single chromosome and of the maintenance of reciprocal fragments are considered in the context of a model that postulates that the V kappa and J kappa segments are not all organized in the DNA in the same transcriptional direction, leading to inversions rather than deletions during joining.

The structure of immunoglobulin-related gene was analyzed in individual Abelson murine leukemia virus (A-MuLV)-transformed lymphoid cell lines. Essentially all of these lines, whether immunoglobulin-containing or null, had DNA rearrangements in the vicinity of the JH regions on both chromosomes as well as deletions of at least 5 kb of DNA 5' to JH. None of these lines, however, except rare light chain producers, had detectable rearrangement at either their kappa or lambda light chain loci. In contrast to A-MuLV-transformants derived from bone marrow. Those from early fetal liver frequently contained more than two and sometimes 12 or more distinct, rearranged JH-containing fragments. Cellular subclones derived from these lines had a subset, usually two, of the fragments found in the parent line. Therefore, heavy chain gene rearrangement appears to precede that of light chain gene rearrangement and is still continuing in certain cultured A-MuLV transformants.

It may be significant that the 12/23 rule distinguishes V(D)J recombination from both TPN and CSSR. The crucial and integral nature of the 12/23 rule is underscored by several observations. One is that the rearranging loci analyzed in modern examples of ancient vertebrate radiations always contain both versions of the V(D)J joining signal (reviewed inLewis 1994xLewis, S.M. Adv. Immunol. 1994; 56: 27–150Crossref | PubMedSee all ReferencesLewis 1994). Another is that during in vivo recombination, the 12/23 rule is enforced early, at the cleavage step (Steen et al. 1996xSteen, S.B, Gomelsky, L, and Roth, D.B. Genes to Cells. 1996; 1: 543–553Crossref | PubMedSee all ReferencesSteen et al. 1996). This constraint has been reconstructed in vitro with purified RAG1 and 2 alone (van Gent et al. 1996bxvan Gent, D.C, Ramsden, D.A, and Gellert, M. Cell. 1996; 85: 107–113Abstract | Full Text | Full Text PDF | PubMed | Scopus (225)See all Referencesvan Gent et al. 1996b).Why don't any of the analyzed CSSR or transposition systems have a 12/23 rule, or its equivalent? A teleological view is that the 12/23 rule is found in V(D)J recombination because V(D)J recombination is required to accurately manipulate hundreds, not just two or three, potential recombination sites. The 12/23 rule has the appearance, at least, of a particularly economical solution to the problem of avoiding unprofitable combinations in the case of a recombination system that is confronted with multi-site arrays.Perhaps at this juncture, a hunt for origins would best be served by looking among multi-site recombination systems for one that more clearly shares basic attributes with V(D)J recombination. Conceivably such a system could be transpositional, where a transposase is required to handle numerous recombination targets created by high-copy transposon integration. Alternatively, relatives of known CSSR systems (as for example Min; Figure 3Figure 3; Iida et al. 1990xIida, S, Sandmeier, H, Huber, P, Giestand-Nauer, R, Schneitz, K, and Arber, W. Mol. Microbiol. 1990; 4: 991–997Crossref | PubMed | Scopus (11)See all ReferencesIida et al. 1990) that serve to rearrange a multi-site array might bear closer scrutiny. There is also the possibility that further afield, recombination systems that enable RNA to wander, “integrons” to move, or DNA to exit en masse from a genome may provide key insights (see, for example,5xHall, R.M and Collis, C.M. Mol. Microbiol. 1995; 15: 593–600Crossref | PubMed | Scopus (474)See all References, 9xKlobutcher, L.A and Herrick, G. Nucl. Acids Res. 1995; 23: 2006–2013Crossref | PubMed | Scopus (80)See all References, 13xPyle, A.M. Nature. 1996; 381: 280–281Crossref | PubMed | Scopus (7)See all References). In the meantime, there is no question that the comparative molecular analyses carried out to date have been extremely informative (2xDifilippantonio, M.J, McMahan, C.J, Eastman, Q.M, Spanopoulou, E, and Schatz, D.G. Cell. 1996; 87: 253–262Abstract | Full Text | Full Text PDF | PubMed | Scopus (165)See all References, 16xSpanopoulou, E, Zaitseva, F, Wang, F.-H, Santagata, S, Baltimore, D, and Panaoytou, G. Cell. 1996; 87: 263–276Abstract | Full Text | Full Text PDF | PubMed | Scopus (189)See all References, 19xvan Gent, D.C, Mizuuchi, D, and Gellert, M. Science. 1996; 271: 1592–1594Crossref | PubMedSee all References). At the present pace, it would not be surprising to see the V(D)J recombination system soon begin to rival the more classical site- directed recombination systems both in terms of the rich detail with which it is understood, and as a paradigm for revealing fundamental features of protein–DNA interactions.

Pre-B and pre-T cell lines from mutant mice with severe combined immune deficiency (scid mice) were transfected with plasmids that contained recombination signal sequences of antigen receptor gene elements (V, D, and J). Recovered plasmids were tested for possible recombination of signal sequences and/or the adjacent (coding) sequences. Signal ends were joined, but recombination was abnormal in that half of the recombinants had lost nucleotides from one or both signals. Coding ends were not joined at all in either deletional or inversional V(D)J recombination reactions. However, coding ends were able to participate in alternative reactions. The failure of coding joint formation in scid pre-B and pre-T cells appears sufficient to explain the absence of immunoglobulin or T cell receptor production in scid mice.

Division of Biology California Institute of Technology Pasadena, California 91125 f Laboratory of Molecular Biology National Institute of Diabetes and Digestive and Kidney Diseases National Institutes of Health Bethesda, Maryland 20892 Much of the variety in binding specificities of the immune system is generated when subgenic DNA elements are assembled into antigen receptor genes during T and B cell differentiation. This process has attracted interest both because of its important role in the differentiation of the immune system and because it is the only site-specific recombination system known to operate in vertebrates. The natural substrates for immunoglobulin and T cell receptor gene rearrangement are complex loci (seven have been described in mice) composed of multiple cop- ies of V, D in some cases, and J segments. A complete an- tigen receptor gene is created by fusion of one V to one J, or one V to one D to one J. The basic operation (which we call V-J joining whether it involves joining V to J, V to D, or D to J), is represented in the left column of the figure. Two coding segments (shown as boxes) are detached from their joining signal sequences (shown as triangles) and then attached to one another to form a contiguous coding sequence. The junction is called a coding joint. A recipro- cal fusion of the signal sequences leads to a signal joint. (General principles are reviewed in Tonegawa, 1983.) The details of this process can be probed using recom- bination substrates that are constructed to contain a de- fined pair (or pairs) of gene segments. These substrates are designed either to become integrated into the genome of the recipient cells or to remain extrachromosomal. In ei- ther case, when introduced into recombinationally active cell lines, the gene segments are targeted by the recombi- nation machinery and become rearranged in the same fashion as their endogenous counterparts. The ease of experimentation afforded by this methodology has led to a rapid increase in available information about V-J recom- bination. Here we will review some of the key features of the process and present a current model of how it may work. One clearcut result is that the joining signals alone are the necessary and sufficient targets for recombination. Joining signals are simple DNA sequence motifs consist- ing of conserved heptamer and nonamer sequences separated by a nonconserved spacer of 12 or 23 bp. With- out joining signals, no recombination occurs. On the other hand, if V, D, or J coding elements are replaced by other DNA, the new sequences abutting the joining signals are still recombined site specifically (Lewis et al., 1985; Akira et al., 1987; Hesse et al., 1987). Thus, of the four elements