Scid mice express a truncated form of the catalytic subunit of the DNA-dependent protein kinase (DNA-PKcs) and are unable to properly rearrange their Ig and TCR genes, resulting in a severe combined immunodeficiency that is characterized by arrested differentiation of B and T lymphocytes. Treatment of scid mice with low doses of gamma irradiation rescues rearrangements at several TCR loci and promotes limited thymocyte differentiation. The machinery responsible for sensing DNA damage and the mechanism by which irradiation compensates for the scid defect in TCR recombination remain unknown. Because DNA-PKcs is present in scid thymocytes, it may mediate some or all of the irradiation effects. To test this hypothesis, we examined the effects of irradiation on DNA-PKcs-deficient (slip) mice. Our data provide the first evidence that DNA-PKcs is not required for limited rescue of thymocyte differentiation or TCR rearrangements.
Differentiation of T lineage lymphocyte precursors takes place in the thymus and can be conveniently monitored by following cell-surface expression of CD3 (a component of the TCR) and, for the αβ T cell lineage, the CD4 and CD8 coreceptor molecules (1). Early intrathymic precursors are CD4−CD8− double negative (DN)3 cells that typically do not express CD3; in these cells, TCRβ rearrangement is ongoing (2). Expression of a functional TCRβ-chain allows thymocytes to mature to the next stage, where both CD4 and CD8 are expressed, as well as low to intermediate levels of CD3. TCRα rearrangements occur in these double positive (DP) thymocytes. Following productive TCRα rearrangement and selection, thymocytes become mature single positive (SP) cells, expressing high levels of CD3 and either CD4 or CD8 at the cell surface.
The process responsible for TCR rearrangements, V(D)J recombination, can be divided into two steps: DNA cleavage and joining of the broken ends. Cleavage, mediated by the recombination-activating gene products, RAG-1 and RAG-2, produces two kinds of DNA termini: signal ends and coding ends. These intermediates join to form signal and coding joints, respectively. Several double-strand break repair factors are required for joining, including the components of the DNA-dependent protein kinase (DNA-PK). DNA-PK is made up of a catalytic subunit, DNA-PKcs, and two DNA binding subunits, Ku86 and Ku70 (3, 4).
Mice bearing mutations in any of the three genes encoding DNA-PK subunits exhibit defects in the joining phase of V(D)J recombination, resulting in a severe combined immunodeficiency characterized by prematurely arrested development of B and T lymphocytes. This was first demonstrated in the case of the scid mouse, a naturally occurring mutant (5) bearing a point mutation in the DNA-PKcs gene that causes the protein to be prematurely terminated 83 aa from the normal C terminus (6, 7). This truncated protein is expressed in scid thymocytes at about 10% of the levels found in wild-type thymocytes (7) and may possess kinase activity (8).
We and others have shown that treatment of scid mice with low doses of gamma irradiation partially rescues V(D)J rearrangements at the TCRγ, δ, and β loci and promotes thymocyte differentiation to the DP stage (9, 10, 11, 12, 13). Interestingly, irradiation does not promote significant rearrangement of TCRα and does not rescue differentiation or rearrangements in the B cell lineage (9, 10, 11, 12, 13). We have recently shown that irradiated scid bone marrow cells are capable of transferring the irradiation rescue effects to unirradiated host mice, indicating that very early lymphocyte precursors may be the irradiation targets (14). Irradiation also exerts a profound oncogenic effect on these animals, as all treated scid mice develop thymic lymphomas within a few months of treatment (9, 10), compared with a 15% incidence of spontaneous thymic lymphoma in nontreated scid mice (15).
The molecular mechanisms of these irradiation effects remain unknown. The completed, in-frame V(D)J rearrangements at the TCRβ locus observed in irradiated scid thymocytes are thought to promote thymocyte differentiation and proliferation (9). However, they are not required for these effects of irradiation, as RAG−/− thymocytes proliferate and differentiate to the DP stage in the absence of any TCR rearrangements (16). We proposed that in irradiated scid mice, both rearrangement-dependent and -independent mechanisms might be responsible for the appearance of DP thymocytes (11). In particular, we suggested that activation of a DNA damage response pathway may have two separate effects: 1) providing signals for growth and differentiation of thymocytes, perhaps by mimicking signals normally given through the TCR complex, and 2) facilitating joining of the stalled V(D)J recombination intermediates that accumulate in scid thymocytes (11).
This hypothesis requires two key components: a DNA damage sensor and a mechanism for facilitating the joining of V(D)J recombination intermediates. Because the scid mutation is not a null allele, either or both of these processes may require the presence of residual DNA-PK activity or DNA-PKcs protein. Recent experiments have attempted to address this question by examining the effects of irradiation on mice genetically deficient for another DNA-PK component, Ku86. Interestingly, while irradiation stimulated the appearance of DP thymocytes, no overall increase in thymic cellularity was observed (TCR rearrangements were not examined) (17). Thus, while a full irradiation rescue was not observed in these animals, the results suggest that DNA-PK activity is not required for the transmission of the differentiation signal. However, recent biochemical experiments have shown that DNA-PK activity can be induced by broken DNA ends in the absence of Ku (18, 19). Therefore, we cannot rule out the possibility that irradiation rescue in Ku-deficient mice was mediated by direct activation of DNA-PK.
The absence of irradiation-induced increases in thymic cellularity in the Ku86-deficient mice suggests that thymocyte proliferation may be impaired in the absence of DNA-PK. However, this interpretation is clouded by the observation that Ku is involved in several critical cell growth processes, including telomere length regulation and senescence (20, 21, 22, 23, 24); furthermore, Ku86- and Ku70-deficient mice exhibit dwarfism (22, 25), suggesting a serious problem with cellular growth regulation. None of these are features of scid or DNA-PKcs-null mice (5, 26, 27, 28, 29). Thus, the failure to observe thymocyte proliferation in response to irradiation in Ku86-deficient mice may reflect a specific requirement for Ku, rather than DNA-PK, in this process.
As the preceding discussion indicates, it is now clear that Ku and DNA-PKcs have nonoverlapping activities. Thus, the determination of whether DNA-PKcs is required for the irradiation rescue phenomenon necessitates examination of irradiation responses in DNA-PKcs-deficient mice. The first such model was provided by the slip mouse, in which the DNA-PKcs gene was inactivated by integration of a transgene within the first few hundred nucleotides of the gene, generating mice that fail to express detectable levels of DNA-PKcs mRNA (by RT-PCR) and have no detectable DNA-PK activity (29). More recently, three different targeted alterations of the DNA-PKcs gene have been generated in mice by homologous recombination. All of these mutations have virtually identical effects on V(D)J recombination (26, 27, 28, 30).
While all four DNA-PKcs-deficient mouse lines have essentially the same defect in V(D)J recombination as the classical scid mouse, some differences in the effects of these mutations on thymocyte differentiation have been observed. In two of the knockout lines, thymocyte differentiation is arrested at the DN stage, as in the scid mouse (26, 27). In the other knockout and in the slip mouse, some “leakiness” (progression to the DP stage) is observed (28, 29). This phenotypic variability could be due to effects of genetic background. At present, the interpretation of results from the knockout mice is clouded by the variable admixture of two different genetic backgrounds present in each animal.
Here we use slip mice as a model system to ask whether the irradiation rescue phenomenon requires the presence of the truncated version of the DNA-PKcs protein present in scid thymocytes. We found that, as in Ku86-deficient mice, irradiation rescues some thymocyte differentiation without substantial increases in thymic cellularity. Surprisingly, irradiation efficiently rescues rearrangements at the TCRα locus, a situation not observed in irradiated scid mice. These results provide the first evidence that residual DNA-PK activity is not required for irradiation-induced rescue of limited thymocyte differentiation or V(D)J recombination.
Materials and Methods
Slip mice (29) were maintained in the animal facility at Baylor College of Medicine. Newborn mice were irradiated (1 Gy) within 72 h of birth by exposure to a 137Cs source. Mice are reported by age (in days), regardless of treatment. Slip mice used here were generated on an FVB/N background (29). Age-matched BALB/c mice were used as controls.
Thymocyte cell suspensions and flow cytometry
Thymi were homogenized, and the cells were washed and counted. Each sample was subjected to flow cytometry and V(D)J rearrangement analysis (see below). Thymocytes were stained with anti-CD4 (RM4-4), and anti-CD8 (53-6.7) that were conjugated with CyChrome or FITC (PharMingen, San Diego, CA), respectively. Thymocytes were analyzed on an EPICS XL (Coulter, Palo Alto, CA).
DNA preparation and semiquantitative PCR
Thymocyte DNA was prepared as described (31). Vα8-Jα49 and Vβ7-Jβ2 rearrangements (coding joints) were amplified from genomic DNA in a total volume of 50 μl with 1 U Taq polymerase (Perkin-Elmer, Norwalk, CT) in sample buffer supplied by the vendor with 2 mM MgCl2 and 25 pmol of each primer (see below). Thirty cycles of amplification were performed in a GeneAmp PCR System 9700 (Perkin-Elmer), where one cycle was 95°C, 55°C, 72°C, each for 30 s. PCR products were separated on a 6% polyacrylamide gel, electrophoretically transferred onto GeneScreen Plus membranes (DuPont, Boston, MA), hybridized to internal oligonucleotide probes that were 32P end-labeled, and analyzed using a Molecular Dynamics Storm 860 Phosphorimager (Molecular Dynamics, Sunnyvale, CA).
Sequencing and assignment of nucleotides from V(D)J rearrangements
PCR products were ligated into pCR2.1 (Invitrogen, San Diego, CA). Inserts from plasmid preparations derived from transformed DH5α colonies that hybridized with internal oligonucleotide probes (see below) were subjected to cycle sequencing (manual sequencing with ThermoSequenase (USB, Cleveland, OH) detected by 33P-labeled dideoxynucleotides or automated sequencing (Applied Biosystems, Foster City, CA) performed by Lark Technologies, Houston, TX). Nucleotides were assigned based on reported germline sequences for Jα49 (EMBL/GenBank accession no. M64239), Jβ2 (32), and Dβ (33) gene segments. Germline nucleotides from Vβ7 and Vα8 were assigned with a high degree of confidence through sequence analysis of various cDNAs using these gene segments (34, 35, 36, 37, 38, 39, 40). Nucleotides were strictly assigned to categories in the following order: J coding region, J-associated P nucleotides, V coding region, V-associated P nucleotides, and (in the case of TCRβ rearrangements), D coding region, followed by D-associated P nucleotides. Ambiguous nucleotides are underlined and represent junctions that could have used short sequence homologies for joining. Nucleotides that did not fall into one of these categories were designated N nucleotides.
Oligonucleotides and sequence source
Oligonucleotides (obtained from Life Technologies, Gaithersburg, MD) and their respective sequence source are as follows: PCR primers: Vα8, CGCCACTCTCCATAAGAGCAGC (39); Jα49, CATGCCCATCAGTTGGTGTGAAAG (accession no. M64239); Vβ7, GCCATGAAACAATGTACTGGTATCG (34); Jβ2.6, GCCTGGTGCCGGGACCGAAGTA (32); probes: Jα49, GGACTCACTGTGAGCTTTGC (accession no. M64239); Jβ2.6, CCTATGAACAGTACTTCGGTCCCGG (32).
Rescue of thymocyte differentiation in irradiated slip mice
Following the protocol established previously for irradiation of newborn scid mice (9, 11), we administered 1 Gy of gamma irradiation to newborn slip mice within the first 72 h of life. Three weeks later, thymocytes were harvested and examined by flow cytometry for expression of CD4 and CD8. Two color profiles of unirradiated and irradiated slip thymocytes are shown in Fig. 1; a wild-type profile is shown for comparison. While DP thymocytes are present in the unirradiated control, it is evident that irradiation induces a substantial increase in the proportion of DP cells. However, in this experiment no increase in thymic cellularity was observed. Data from a number of experiments are summarized in Fig. 2. The proportion of DP thymocytes in unirradiated slip mice appears to increase with age from day 10 (roughly 25%) to day 21 (∼50%) after birth (Fig. 2 A, □). The presence of DP cells and the frequent appearance of DP thymic lymphomas in slip mice have been noted previously (29), raising the possibility that the DP cells arising spontaneously in unirradiated slip mice represent a preneoplastic population.
After irradiation, some increase in the DP population (up to 70–80%) was observed (Fig. 2,A, ▪) in younger animals (up to 14 days of age). In older mice, the spontaneous appearance of DP thymocytes apparently obscured the irradiation response. Unlike the situation in irradiated scid mice, no consistent increase in thymic cellularity was observed in irradiated slip mice compared with unirradiated controls (Fig. 2,B). Analysis of the DP population based on absolute numbers of DP thymocytes failed to reveal a significant increase in DP thymocytes after irradiation, even in younger animals (Fig. 2 C). Based on these data, we conclude that while irradiation can signal limited thymocyte differentiation in the absence of DNA-PKcs, either the proliferative response is diminished or other factors disfavor the accumulation of DP thymocytes in response to irradiation. This situation is similar to that described for irradiated Ku86-deficient mice (17).
Effects of irradiation on TCR rearrangements
To assess the ability of ionizing radiation to rescue V(D)J recombination, we examined the status of several TCR loci using semiquantitative PCR assays for rearrangements (12, 30). We first looked for rearrangements at the TCRβ locus, using primers specific for Vβ7-Jβ2.6 coding joints. As described previously for Vβ8 rearrangements (9), irradiation of scid mice substantially rescues Vβ7-Jβ2.6 coding joints (Fig. 3 A, compare lanes 5 and 6). In thymocytes from three individual unirradiated slip mice, we observed variable, low levels of spontaneous Vβ7-Jβ2.6 rearrangements (lanes 7–9). These rearrangements may be related to the appearance of DP thymocytes in these animals. While low levels of these rearrangements were also detected in two of three irradiated slip mice (lanes 10–12), we failed to observe consistent, robust rescue of rearrangements as seen in irradiated scid thymocytes. We also examined TCR Vγ2-Jγ1 rearrangements, which are efficiently rescued in irradiated scid thymocytes (11). In agreement with our analysis of TCRβ rearrangements, spontaneous TCRγ rearrangements were readily detected in thymocytes from three of four slip mice examined; after irradiation, no consistent increase in the levels of these rearrangements was observed (data not shown). Thus, unlike the situation in scid mice, irradiation does not efficiently rescue TCRβ or TCRγ rearrangements, an effect that may be related to the failure of irradiation to stimulate thymocyte proliferation.
Finally, we examined the status of the TCRα locus. Previous work has shown that recombination does not initiate at the TCRα locus in unirradiated scid mice (11, 12, 13). However, irradiation of scid mice promotes the appearance of DP thymocytes, which allows TCRα recombination to begin (11, 12, 13). Because DP thymocytes are already present in unirradiated slip mice, we were interested to see whether completed rearrangements at TCRα could be detected and whether these rearrangements could be stimulated by irradiation. We performed PCR analysis for TCRα rearrangements involving the Vα8 and Jα49 elements. We previously showed that these gene segments are frequently rearranged in wild-type thymocytes; moreover, recombination is initiated at these sites in DP thymocytes from irradiated scid mice (12). As shown in Fig. 3 B, Vα8-Jα49 rearrangements are not observed in unirradiated scid thymocytes (lane 5). After irradiation, only very low levels of TCRα rearrangements are observed (lane 6), in agreement with previous results (12). With one exception (lane 8), thymocytes from unirradiated slip mice showed only trace amounts of Vα8-Jα49 rearrangements. However, after irradiation, three of four slip thymi contained high levels of TCRα coding joints (lanes 11–14). Thus, unlike the situation in scid mice, irradiation efficiently rescues recombination at the TCRα locus in slip mice. Similar results were observed for Vα8-Jα50 rearrangements (data not shown).
Efficient rescue of normal, diverse TCRα rearrangements in irradiated slip mice
The appearance of abundant TCRα rearrangements after irradiation could result from efficient joining of broken TCRα V(D)J recombination intermediates. In this case, we would expect to find a diverse array of junction sequences in thymus DNA preparations from individual animals. Alternatively, irradiation might not promote efficient joining, but, instead, could stimulate proliferation of rare cells that contain successful joining events. The latter model predicts that individual thymi will not contain a diverse set of junction sequences. We were also interested in determining whether irradiation rescues formation of normal junctions, without the characteristic features of junctions formed in scid mice, such as excessive P nucleotide insertions and abnormal deletions (41, 42, 43).
To test these predictions, PCR products resulting from Vα8-Jα49 joining in several different individual thymi from irradiated and unirradiated slip mice were cloned and sequenced. Nucleotide sequences of these junctions are displayed in Fig. 4 along with sequences obtained from an age-matched wild-type (BALB/c) mouse, which are shown for comparison. Inspection of these junctions reveals several important features. First, junctions from unirradiated slip mice display unusually long P nucleotides, as previously reported in scid mice (42, 43). These are not observed in junctions from irradiated slip mice. Indeed, the structural characteristics of the junctions (number of nucleotides deleted, presence of extra nucleotides, appearance of short sequence homologies, frequency of junctions in-frame, and CDR3 length) appear indistinguishable from wild-type controls, indicating that irradiation stimulates normal joining. Furthermore, a highly diverse set of junctions is observed after irradiation in individual animals. These data indicate that irradiation promotes efficient coding joint formation by a mechanism that appears to be distinct from that operative in unirradiated slip mice.
Because Vβ7-Jβ2 rearrangements were detected at similar levels in both unirradiated and irradiated slip mice, we were curious to determine whether there are any differences in the fine structure of the junctions produced after irradiation. PCR products from Vβ7-Jβ2 joints from both irradiated and unirradiated slip mice were cloned and sequenced. As shown in Fig. 5, sequences from unirradiated slip mice revealed several abnormal features commonly associated with aberrant joining seen in scid mice. First, abnormally long P nucleotides are present at both V-D and D-J junctions that are not seen in junctions from age-matched BALB/c controls. Second, about half of the sequences from unirradiated slip mice have extensive, aberrant deletions that remove most or all of the D region. These could result from extensive deletions or from aberrant V-J joining. Third, none of the junctions contain N regions. However, after irradiation the sequences of junctions from slip mice indicate that there is some normalization of joining, as N regions are present, the D elements appear normal, and the proportion of in-frame junctions increases (although CDR3 lengths remain outside of the normal range, which is 9–12 aa in TCRs (44)). However, rearrangements from irradiated slip mice still contained abnormally long P nucleotide inserts, indicating that irradiation does not completely normalize the joining process. These results are consistent with the abnormal features evident in sequence sets from irradiated scid mice analyzed according to our nucleotide assignment rules (see Materials and Methods). These include excessively long P nucleotides in TCR Vδ4-Dδ1 junctions (13) and absence of D regions in TCR Vβ8-Jβ2 junctions (9).
DNA-PKcs is not required for rescue of thymocyte differentiation
Exposure of scid mice to low doses of ionizing radiation promotes proliferation and differentiation of thymocytes to the DP stage, resulting in a 10- to 20-fold increase in thymic cellularity. Irradiation-induced DNA damage appears to signal these differentiative events, as they are also observed with other DNA damaging agents, such as bleomycin (9). The identities of the proteins that recognize and transmit the DNA damage signal remain unknown. The presence of a truncated form of DNA-PKcs in scid thymocytes, which may retain kinase activity (8), has raised the possibility that DNA-PK itself could be required for transmission of the DNA damage signal or for irradiation-induced rescue of V(D)J rearrangements.
Our results indicate that DNA-PKcs is not required for rescue of limited thymocyte differentiation or V(D)J recombination in response to ionizing radiation. Thus, neither transmission of the DNA damage signal nor rescue of coding joint formation requires DNA-PK activity. These data are in agreement with the observation that irradiation of Ku86-deficient mice results in the appearance of DP cells (17). Together, these results rule out two of the most obvious candidates for the DNA damage sensing function. What, then, are the molecules responsible for signaling? We have shown that irradiation rescue in scid mice requires p53 (11), which can be activated by binding to DNA lesions (45, 46). Thus, p53 itself could be the damage sensor. Alternatively, sensing could be performed by molecules such as ATM, poly(ADP-ribose) polymerase, or other factors.
DNA-PK and thymocyte proliferation
The lack of thymocyte proliferation in response to DNA damage provides the first clear indication that DNA-PKcs may play a critical role in thymocyte proliferation. Previous analysis of Ku86-deficient mice yielded a similar finding (17), but it was not possible to attribute this defect to lack of DNA-PK activity for two reasons: First, DNA-PKcs can be activated in the absence of Ku (18, 19), and second, Ku86-deficiency has several effects on cellular growth regulation (22) that are not observed in scid or DNA-PKcs-deficient mice (27, 28, 29, 30). Thus, our data suggest that DNA-PKcs may play a role in growth control. Possible mechanisms could include effects on transcription, as DNA-PK can phosphorylate a number of transcription factors in vitro (4, 47, 48).
One notable feature of unirradiated slip mice as well as some lines of DNA-PKcs-deficient mice created by gene targeting is the spontaneous appearance of DP thymocytes as the mice age (28, 29). It is not clear why this phenomenon is only observed in some mouse lines. The apparent “leakiness” of some DNA-PK-deficient mouse strains may be due to effects of the genetic background, which in the case of the knockout animals is a variable mixture of the 129 and C57BL/6 strains. Support for this view is provided by our observation that some of the F1 progeny generated by mating scid/scid and slip/slip homozygotes exhibit spontaneous DP thymocytes (M.A.B. and D.B.R., unpublished observations). It will be interesting to see if this is also the case for lymphomas, which appear in virtually all slip mice (29) but are much more rare in scid mice (15) and in at least some lines of DNA-PKcs knockout mice (26). It is possible that the effects of background genes may modify the irradiation response in slip mice. Future experiments using animals that have been extensively backcrossed onto appropriate genetic backgrounds will be required to determine whether this is the case.
Our data show that in the slip mice, spontaneous thymocyte differentiation is accompanied by an increased abundance of completed TCR rearrangements. Appearance of DP thymocytes and completed V(D)J rearrangements has also been noted in Ku70−/− mice (25). It is not clear whether in the absence of DNA-PK activity the ability to bypass the blocks to both thymocyte differentiation and V(D)J recombination reflects a single mechanism or whether these events can occur separately. An increased frequency of productive TCR rearrangements could promote thymocyte differentiation; however, DP thymocytes can develop in the absence of rearrangements (16).
Irradiation-mediated rescue of V(D)J recombination in the absence of DNA-PKcs
Three aspects of the effect of irradiation on TCR rearrangements in slip mice are particularly striking. First, while irradiation normalizes several features of TCRβ coding joints, as seen in scid mice (9), the levels of rescued TCRβ and TCRγ rearrangements are much lower in slip mice than in scid mice. Two factors may contribute to the relatively small increases in levels of rearrangements at the β and γ loci after irradiation: the increased baseline levels of spontaneous rearrangements in unirradiated animals and the lack of irradiation-induced thymocyte proliferation. The increased baseline levels seen in the present study are probably due to the presence of DP thymocytes. It should be noted that in our previous analysis of TCR rearrangements in unirradiated slip mice, which focused exclusively on young slip mice lacking significant numbers of DP thymocytes, very low levels of coding joints were detected (30). Progression of thymocytes to the DP stage in the absence of proper TCRβ rearrangements may reflect failure of a DNA-PK-dependent checkpoint that normally functions to prevent inappropriate developmental progression without productive TCRβ rearrangement (28). This hypothesis is supported by the fact that <50% of the TCRβ sequences from irradiated slip mice are in frame, compared with >95% in wild-type and irradiated slip mice. Furthermore, loss of this DNA-PK-dependent checkpoint may be more evident in slip mice than in scid mice, which could retain partial DNA-PK function.
A second notable feature of our results is the rescue of abundant, normal-appearing, diverse TCRα rearrangements. Based on previous analyses of irradiated scid mice, we did not expect to see rescue of these rearrangements. We have previously shown that the same TCRα rearrangements studied here are rescued, but only very weakly, by irradiation of newborn scid mice (12). The basis for the ability of irradiated slip, but not scid, mice to rescue TCRα rearrangements remains unclear.
A third interesting feature is the differential effect of irradiation on rescue of rearrangements at the TCRα and β loci. While irradiation rescues TCRα rearrangements that appear normal by DNA sequence analysis, irradiation-induced TCRβ rearrangements retain some abnormal features, in particular excessively long P nucleotides. Because these extra nucleotides are thought to be derived from hairpin opening (49, 50), the presence of aberrant P nucleotide inserts suggests that the hairpin opening reaction may not proceed normally. One explanation for these differential effects is that irradiation might affect the two loci in subtly different ways. Alternatively, differences in the developmental stages of the thymocytes at the time of irradiation might affect the mechanism of rescue. For example, substantially more RAG RNA is expressed at the time of TCRα rearrangement in DP thymocytes than during TCRβ rearrangement in DN thymocytes (51). In agreement with this finding, we have shown that RAG-1 protein levels are substantially higher, on a per cell basis, in thymocytes from slip mice than in scid thymocytes (M. Purugganan and D.B.R., unpublished observations), an observation that presumably reflects the greater abundance of DP cells in slip thymi. Conceivably, this could affect the mechanism of rescue, as recent reports have indicated that the RAG proteins are capable of opening hairpins in vitro (52).
Although the mechanistic details remain unclear, the observation of a robust rescue of rearrangements at the TCRα locus in irradiated slip mice provides firm evidence that the residual DNA-PKcs protein present in scid thymocytes is not required for rescue of V(D)J recombination. This information allows further refinement of our understanding of the irradiation rescue phenomenon, as scenarios involving up-regulation of weak DNA-PK activity in scid mice can now be discounted.
We thank Jeff Lin for technical help and Mary Lowe for secretarial assistance. We thank the flow cytometry core facility staff at Baylor College of Medicine for analysis of data. Vicky Brandt, Leslie Huye, Mary Purugganan, and Heather Yarnall provided critical comments on the manuscript.
This work was supported by a grant (RPG-95-027-04 CIM) from the American Cancer Society (to D.B.R.).<. >
Abbreviations used in this paper: DN, double negative; DP, double positive; SP, single positive; RAG, recombination-activating gene; DNA-PK, DNA-dependent protein kinase; DNA-PKcs, catalytic subunit of DNA-PK; ATM, ataxia telangiectasia mutated.