Biased T-cell receptor delta element recombination in scid thymocytes

Distinct effects of DNA-PKcs and Artemis inactivation on signal joint formation in vivo

The assembly of functional immune receptor genes via V(D)J recombination in developing lymphocytes generates DNA double-stranded breaks intermediates that are repaired by non-homologous end joining (NHEJ). This repair pathway requires the sequential recruitment and activation onto coding and signal DNA ends of several proteins, including the DNA-dependent protein kinase and the nuclease Artemis. Artemis activity, triggered by the DNA-dependent protein kinase, is necessary to process the genes hairpin-sealed coding ends but appears dispensable for the ligation of the reciprocal phosphorylated, blunt-ended signal ends into a signal joint. The DNA-dependent protein kinase is however present on signal ends and could potentially recruit and activate Artemis during signal joint formation. To determine whether Artemis plays a role during the resolution of signal ends during V(D)J recombination, we analyzed the structure of signal joints generated in developing thymocytes during the rearrangement of T cell receptor genes in wild type mice and mice mutated for NHEJ factors. These joints exhibit junctional diversity resulting from N nucleotide polymerization by the terminal nucleotidyl transferase and nucleotide loss from one or both of the signal ends before they are ligated. Our results show that Artemis participates in the repair of signal ends in vivo. Furthermore, our results also show that while the DNA-dependent protein kinase complex protects signal ends from processing, including deletions, Artemis seems on the opposite to promote their accessibility to modifying enzymes. In addition, these data suggest that Artemis might be the nuclease responsible for nucleotide loss from signal ends during the repair process.

V(D)J recombination of chromosomally integrated, wild-type deletional and inversional substrates occur at similar frequencies with no preference for orientation

Efficient and correct recombination of V(D)J substrates results in the generation of antibodies. The RSS substrates are oriented in two directions with respect to each other: deletional and inversional. Deletional recombination results in the formation of the coding joint and excision of the intervening sequences. Inversional recombination retains all the genomic sequences and forms both a coding joint and a signal joint. A bias for deletional recombination has been characterized with specific loci in vivo and recapitulated in experiments using extrachromosomal substrates. We constructed retroviral substrates of RSS in the deletional and inversional orientation. We introduced the substrates into wild-type and scid pre-B cells and measured the frequency of functional recombination in addition to open/shut recombination. We also mutated the RSSs to determine whether mutated sequences influenced orientation bias. We show that pre-B cells recombine the wild-type substrates at a 1.6 ratio of deletion:inversion. Nonamer mutated substrates recombined with a deletional bias whereas heptamer mutated substrates recombined with an inversional bias. A spacer length mutation and drastic mutations in the RSS abolish all recombination. These results suggest that there is no orientation bias with wild-type RSSs but that orientation bias occurs when RSSs are mutated.

Cyclin A/CDK2 regulates V(D)J recombination by coordinating RAG-2 accumulation and DNA repair

Accumulation of the V(D)J recombinase protein RAG-2 is restricted to G0/G1 cells by phosphorylation-mediated degradation at the G1-S boundary. Here cyclin A/CDK2 is shown to oppose RAG-2 accumulation; conversely, RAG-2 is induced by p27Kip1 and related CDK inhibitors. Coinduction of RAG-2 and G1 delay by p27Kip1 is accompanied by strong stimulation of V(D)J recombination. Unexpectedly, induction of RAG-2 accumulation in the absence of G1 delay has no effect on recombination frequency. p27Kip1 may stimulate V(D)J recombination by coordinating accumulation of RAG-2 with prolongation of G1, when nonhomologous end joining is preferentially active. Consistent with this, enforced expression of RAG-2 throughout cell cycle is associated with accumulation of aberrant recombination products reminiscent of those formed in the absence of nonhomologous end joining.

Targeted disruption of the catalytic subunit of the DNA-PK gene in mice confers severe combined immunodeficiency and radiosensitivity

The DNA-dependent protein kinase is a mammalian protein complex composed of Ku70, Ku80, and DNA-PKcs subunits that has been implicated in DNA double-strand break repair and V(D)J recombination. Here, by gene targeting, we have constructed a mouse with a disruption in the kinase domain of DNA-PKcs, generating an animal model completely devoid of DNA-PK activity. Our results demonstrate that DNA-PK activity is required for coding but not for signal join formation in mice. Although our DNA-PKcs defective mice closely resemble Scid mice, they differ by having elevated numbers of CD4+CD8+ thymocytes. This suggests that the Scid mice may not represent a null phenotype and may retain some residual DNA-PKcs function.

V(D)J recombination in Ku86-deficient mice: distinct effects on coding, signal, and hybrid joint formation

Ku, a heterodimer of 70 and 86 kDa subunits, plays a critical but poorly understood role in V(D)J recombination. Although Ku86-deficient mice are defective in coding and signal joint formation, rare recombination products have been detected by PCR. Here, we report nucleotide sequences of 99 junctions from Ku86-deficient mice. Over 90% of the coding joints, but not signal or hybrid joints, exhibit short sequence homologies, indicating that homology is required to join coding ends in the absence of Ku86. Our results suggest that Ku86 may normally have distinct functions in the formation of these different types of junctions. Furthermore, Ku86(-/-) joints are unexpectedly devoid of N-region diversity, suggesting a novel role for Ku in the addition of N nucleotides by terminal deoxynucleotidyl transferase.

Characterization of coding ends in thymocytes of scid mice: implications for the mechanism of V(D)J recombination

We previously identified possible intermediates in V(D)J recombination at the TCR delta locus and characterized molecules with signal ends and with covalently sealed (hairpin) coding ends in thymocytes of scid mice by Southern blotting. Here, we use a sensitive ligation-mediated PCR assay to demonstrate that all coding ends detected in scid thymocytes are covalently sealed. Neither coding nor signal ends exhibit loss or addition of nucleotides. These data imply that hairpin formation is coupled to the initial cleavage at the signal/coding border, and that the cleavage step in V(D)J recombination is conservative. In scid/+ or wild-type thymocytes, hairpin coding ends are at least 1000-fold less abundant than signal ends. These results provide insight into the mechanism of V(D)J recombination.

The SCID mouse: relevance as an animal model system for studying human disease

The simultaneous description some 5 years ago of two methods for the partial reconstitution of a human immune system in severe combined immune-deficient (SCID) mice (collectively, human:SCID mice) was met with great enthusiasm. At the time, it was hoped that human:SCID mice would provide experimental animal model systems for studying human disease and the human immune system. Many of these hopes have been borne out. Importantly, the experimental results obtained from these chimeric human/animal studies appear to be relevant to human disease and immune function. In spite of these glowing achievements, the SCID mouse may not represent the optimal experimental system with which to address these questions. The incomplete penetrance ("leakiness") of the scid mutation and the recent discovery that the mutation is not lymphoid specific, but rather affects a general DNA repair pathway, will only serve to complicate the interpretation of already complex biological interactions. Recently other immune-deficient mice have been described that appear to overcome one or both of these problems and thus these mice could represent improved hosts for the adaptive transfer of a human immune system. The current status of the SCID mouse in light of these new findings is discussed.