Mammalian cells repair DNA double-strand breaks (DSBs) through either homologous recombination or non-homologous end joining (NHEJ). V(D)J recombination, a cut-and-paste mechanism for generating diversity in antigen receptors, relies on NHEJ for repairing DSBs introduced by the Rag1-Rag2 protein complex. Animals lacking any of the seven known NHEJ factors are therefore immunodeficient. Nevertheless, DSB repair is not eliminated entirely in these animals: evidence of a third mechanism, 'alternative NHEJ', appears in the form of extremely rare V(D)J junctions and a higher rate of chromosomal translocations. The paucity of these V(D)J events has suggested that alternative NHEJ contributes little to a cell's overall repair capacity, being operative only (and inefficiently) when classical NHEJ fails. Here we find that removing certain portions of murine Rag proteins reveals robust alternative NHEJ activity in NHEJ-deficient cells and some alternative joining activity even in wild-type cells. We propose a two-tier model in which the Rag proteins collaborate with NHEJ factors to preserve genomic integrity during V(D)J recombination.

For nearly two decades it has been suspected that the cutaneous T cell lymphoma, mycosis fungoides (MF), and its leukemic variant, the Sezary syndrome, are caused by the human T lymphotropic virus (HTLV-I/ll). Arguments against this concept included the finding that only a small number of MF patients have antibodies to HTLV-I/ll and that attempts to detect proviral sequences by mere Southern hybridization of extracted DNA usually met with failure. However, we have reported repeatedly that HTLV-like particles emerge in blood mononuclear cell (PBMC) cultures of practically all patients with this disease. In several instances, the particles were identified as HTLV by immunoelectron microscopy as well as biomolecular analysis. With the assumptions that the virus in MF patients may have become defective and that the remaining signal may be too weak for detection by Southern hybridization alone, the extracts of freshly isolated PBMC of 50 consecutive patients were subjected to combined PCR/Southern analysis. Here we report the presence of HTLV pol and/or tax proviral sequences in 46 out of 50 (92%) of the patients tested. In addition, five of the patients, who lacked antibodies to HTLV-I/ll structural proteins, were found to be seropositive for tax. It thus seems reasonable to conclude that MF/ Sezary syndrome is an HTLV-associated disease and that lack of an immune response does not preclude infection with this type of virus. (J. Clin. Invest. 1995. 95:547-554.)

In the United States, blood donors are being screened for infection with human T cell lymphotropic viruses I and II (HTLV-I͞II) by serologic means, which detect antibodies to the structural proteins of these viruses. Because patients with mycosis fungoides (MF) usually do not have such antibodies even though their cells harbor HTLV-I Tax and͞or pol proviral sequences, it was questioned whether the prevalence of HTLV infection among healthy blood donors may also be underestimated by current means of testing. To examine this possibility, a study on specimens of relatives of mycosis fungoides patients (MFR) was begun. In addition, to collect data more expeditiously, a cohort of former injection drug users (IDUs) was tested by routine serologic methods, as well as by PCR͞Southern blot analysis for Tax, pol, and gag proviral sequences and Western blot analysis for antibodies to the Tax gene product. To date, 6͞8 MFRs and 42͞81 (51.8%) of HIV-negative IDUs proved to be positive for HTLV, whereas routine serology identified none of the MFR and only 18͞81 (22.2%) of the IDUs. Among the latter test subjects, the incidence of HTLV-I also proved to be 10 times higher than expected. Therefore, it is likely that among healthy blood donors infection with HTLV-I͞II is more prevalent than is currently assumed. Since Tax is the transforming sequence of HTLV-