TdT is a nuclear enzyme that catalyzes the addition of random nucleotides at Ig and TCR V(D)J junctions. In this paper we analyze human IgH rearrangements generated from transgenic minilocus mice in the presence or absence of TdT. In the absence of TdT, the pseudo-VH gene segment present in the minilocus is rearranged dramatically more frequently. Additionally, JH6 gene segment utilization is increased as well as the number of rearrangements involving only VH and JH gene segments. Thus, the recombination of IgH gene segments that are flanked by 23-nt spacer recombination signal sequences may be influenced by TdT expression. Extensive analysis indicates that these changes are independent of antigenic selection and cannot be explained by homology-mediated recombination. Thus, the role played by TdT may be more extensive than previously thought.

The variable regions of Ig and TCR chains are produced by combining different separate germline elements at the DNA level through a cell specific process termed V(D)J recombination (1). The rearrangement is directed by recombination signal sequences (RSS)4 flanking each coding gene segment (2, 3, 4). During recombination, coding gene segments are joined in an imprecise manner. Coding joints include nucleotide deletions as well as non-germline-encoded (N) nucleotide additions by TdT and germline-encoded (P, palindromic) nucleotide additions (5, 6, 7).

TdT is a nuclear enzyme that catalyzes the addition of deoxynucleoside triphosphates (preferably guanine and cytosine) onto the DNA 3′ OH terminus without template (8). In the fetus, TdT is expressed in the thymus. After birth, it is expressed in the bone marrow and in the germinal center (9, 10, 11). The level of N nucleotide addition in vivo and in cell lines correlates with the level of TdT expression. However, some cell lines can add N nucleotides without showing detectable levels of TdT activity suggesting that other mechanisms may be involved (5, 12, 13). Data suggest that TdT adds nucleotides during the joining phase of the recombination reaction (14). The murine fetal repertoire is thought to be more restricted than the adult repertoire, in part due to the absence of TdT during recombination. The fetal repertoire is characterized by a low frequency of N nucleotides and by a high frequency of homology-mediated recombination events. These homology-mediated rearrangements contain 1–6 nt that could have been encoded by either of the two joined gene segments. In T cells, homology-mediated recombination is involved in the over-representation of some junctions (15, 16, 17, 18, 19).

Genetically modified mice have been engineered that either over-express TdT or that are TdT deficient (TdT/−). The constitutive expression of TdT in transgenic mice results in the addition of N nucleotides in Ig light chain rearrangements (20). In TdT/− mice, however, no N nucleotide are added to IgH and TCRγ, -δ, and -β rearrangements. Furthermore, the frequency of homology-mediated recombination is increased in the adult repertoire. TdT/− mice respond effectively to antigenic stimulation (21, 22, 23, 24, 25, 26).

To study the mechanisms of Ig VHDJH recombination, we engineered mice transgenic for a human IgH minilocus, pHC1 (Fig. 1). Several mechanisms of recombination have been extensively studied in this system (27, 28, 29, 30, 31, 32). In this paper we analyze human IgH rearrangements generated at the DNA level in pHC1 minilocus transgenic mice, in the presence or absence of TdT. The human transgenic minilocus contains a limited number of well-characterized gene segments. Therefore, the analysis of gene segment utilization is straightforward.

FIGURE 1.

Schematic representation of the 60-kb-long human Ig heavy chain transgenic minilocus pHC1. VH, D, and JH gene segments are represented with shaded, gray, and hatched boxes, respectively. VH, D, and JH regions are enlarged. The human Cμ and Cγ1 regions are indicated with dotted boxes. The human JH-C intronic enhancer and the rat 3′ enhancer are indicated with filled and open circles, respectively.

FIGURE 1.

Schematic representation of the 60-kb-long human Ig heavy chain transgenic minilocus pHC1. VH, D, and JH gene segments are represented with shaded, gray, and hatched boxes, respectively. VH, D, and JH regions are enlarged. The human Cμ and Cγ1 regions are indicated with dotted boxes. The human JH-C intronic enhancer and the rat 3′ enhancer are indicated with filled and open circles, respectively.

Close modal

The construct pHC1 used to establish the transgenic founder line 119 contains 2 VH gene segments (VH5–251 and ψVH3–105), 10 D, and 6 JH, Cμ, and Cγ1 human gene segments, the human heavy chain intronic enhancer, and the rat heavy chain 3′ enhancer. The construct is 60 kb long and has been described (Fig. 1; Refs. 27, 28, 29, 30, 31, 32). TdT/− mice have been described (21, 22, 23). pHC1 transgenic TdT/− mice were obtained by cross-breeding and selection by Southern filter hybridization (32). The resulting mice are on a mixed C57BL/6 × 129 background. Both TdT+/+ and TdT/− mice are heterozygous for the minilocus transgene. To minimize differences between the two mice, TdT+/+ and TdT/− mice were analyzed in parallel: mice were kept in the same conditions in the same room and were of the same age at the time of analyses. In all of these mice, Ig light chains are of endogenous murine origin.

Genomic DNA was isolated from the spleens of 2-mo-old transgenic TdT+/+ or TdT/− mice by phenol-chloroform extraction after proteinase K digestion. Alternatively, genomic DNA was isolated from the liver of 14-day-old transgenic embryos. The embryos used in this study were all from the original transgenic breeding, not from the TdT/− cross-breeding. These embryos are on the C57BL/6 background. The DNA was resuspended in Tris-EDTA buffer containing DNase-free RNase A.

The amplification of 0.5 μg of DNA was performed with 40 cycles of 1-min denaturation at 94°C, 2-min annealing at 52°C, and 2-min elongation at 72°C using one unit of Taq polymerase. The final cycle was completed by 7-min elongation at 72°C. VHDJH amplifications were performed using a VH oligonucleotide complementary to both human VH gene segments (5′-AGGTGCAGCTGGTGSAGTCTG-3′) and a human JH consensus oligonucleotide (5′-ACCTGAGGAGACGGTGACCAGGGT-3′). The human JH consensus oligonucleotide amplifies efficiently every human JH gene segment except JH3. Ten picomol of each primer was used. The PCR products were purified using Microcon 50 (Amicon, Beverly, MA) as described by the supplier.

The purified PCR products were blunt-end ligated into EcoRV digested pBluescript KS+ plasmids (Stratagene, La Jolla, CA), and the ligation mixture was used to transform Escherichia coli (XL1-Blue) competent cells. The resulting colonies were screened using three different 32P-labeled oligonucleotides: 1) VHint (5′-TGTATTACTGTGYGAGA-3′) corresponding to the 3′ end of both VH gene segment coding sequence, 2) a VH5–251-specific oligonucleotide (5′-CTATCCTGGTGACTC-3′), and 3) a ΨVH3–105 specific oligonucleotide (5′-AAAGTGTGACGGAAG-3′). The number of colonies positive with each primer were counted to approximate the VH gene segment utilization frequencies. Alternatively, dsDNA was prepared from randomly picked colonies positive with the VHint oligonucleotide and sequenced with an ABI Prism 377 automated DNA sequencer (Applied Biosystems, Foster City, CA).

Murine and human Abs were detected in murine blood by ELISA. Blood was collected from nonimmunized 3-mo-old mice. Microtiter plates were coated with the serum diluted 100 times in 0.1 M carbonate buffer overnight at 37°C. Blocking of remaining sites on the plastic was achieved by incubation for 30 min at 37°C with 10 mg/ml BSA in PBS and 0.05% Tween 20. After repeated washings, Abs were detected by incubation with rat anti-mouse IgM, rat anti-mouse IgG1, mouse anti-human IgM, or mouse anti-human IgG1 conjugated to biotin (PharMingen, San Diego, CA) for 1 h at 37°C. Positive reactions were detected by incubation with streptavidin-HRP (Life Technologies, Grand Island, NY) for 1 h at 37°C. The final reaction was visualized by addition of 3,3′,5,5′-tetramethylbenzidine (TMB; Life Technologies) for 15 min at room temperature. The reaction was stopped by addition of 1 N H2SO4, and absorbance was measured at 450 nm. Purified mouse and human Abs (Sigma, St. Louis, MO) were used as controls.

Murine splenic and bone marrow cells were harvested from 2-mo-old mice. Staining experiments were performed in HBBS containing 3% FCS. Murine cells were stained using allophycocyanin-conjugated anti-B220 (PharMingen), human adsorbed R-PE-conjugated anti-mouse IgM (Southern Biotechnology Associates, Birmingham, AL), and biotin-conjugated anti-human IgM (PharMingen) Abs. The presence of human IgM Abs was further detected using streptavidin-FITC (Zymed Laboratories, San Francisco, CA). Alternatively, isotype control Abs were used as appropriate. Samples were analyzed by flow-cytometry using a FACSCalibur (Becton Dickinson, San Jose, CA).

Statistical analyses were performed using the test for independence: the χ2 test. The χ2 test is used to determine whether hypothesized results are verified by an experiment. It uses the formula χ2 = Σ(f - fi)2/fi, where f is the actual value observed in TdT+/+ mice and fi is the expected value observed in TdT/− mice. Statistical analyses were performed using Microsoft Excel.

Both VH gene segments are rearranged in minilocus transgenic mice. To estimate the relative rearrangement frequency of these two transgenic VH gene segments, we amplified and cloned human VHDJH rearrangements from murine fetal liver and adult spleen. At least four mice per group were used. The recombination frequencies of the two VH gene segments were then determined using gene-specific oligonucleotides. Only 2% of the rearrangements (from 397 colonies) utilize the ΨVH3–105 gene segment in adult pHC1 transgenic mice (Fig. 2). In the fetal repertoire, however, 27% of the rearrangements (from 238 colonies) utilize the ΨVH3–105 gene segment (Fig. 2). To determine whether TdT expression was linked to this difference in gene segment recombination, the experiment was performed in adult minilocus transgenic TdT/− mice. In these mice, 28% of the rearrangements (from 97 colonies) utilize the ΨVH3–105 gene segment (p < 0.001; Fig. 2). These data clearly show that TdT expression influences VH gene segment utilization.

FIGURE 2.

Human Ig VH gene segment utilization in the adult spleen and in the fetal liver of transgenic TdT+/+ mice as well as in the adult spleen of transgenic TdT/− mice. Rearrangements using the nonfunctional ΨVH3–105 and the functional VH5–251 gene segments are indicated in gray and black, respectively.

FIGURE 2.

Human Ig VH gene segment utilization in the adult spleen and in the fetal liver of transgenic TdT+/+ mice as well as in the adult spleen of transgenic TdT/− mice. Rearrangements using the nonfunctional ΨVH3–105 and the functional VH5–251 gene segments are indicated in gray and black, respectively.

Close modal

Over 300 splenic rearrangements were sequenced from the colony counting experiments as well as from additional cloning experiments. A total of 104 and 73 unique human VHDJH rearrangements using the functional VH5–251 and the ΨVH3–105 gene segments, respectively, have been sequenced from splenic DNA of three pHC1 transgenic mice (Fig. 3, A and B). In parallel, 109 and 35 unique human VHDJH rearrangements using the functional VH5–251 and the ΨVH3–105 gene segments, respectively, were sequenced from splenic DNA of five pHC1 transgenic TdT/− mice (Fig. 3, C and D). Rearrangements generated from the minilocus gene segments are extremely diverse. Additionally, every human gene segment is utilized. In minilocus transgenic TdT+/+ mice, 12 rearrangements (12%) utilizing the functional VH5–251 gene segment and 20 rearrangements (27%) utilizing the nonfunctional ΨVH3–105 gene segment are in-frame. In minilocus transgenic TdT/− mice, 28 rearrangements (26%) utilizing the functional VH5–251 gene segment and 8 rearrangements (23%) utilizing the nonfunctional ΨVH3–105 gene segment are in-frame. The addition of N nucleotides in TdT+/+ mice results in a lower frequency of in-frame productive rearrangements (utilizing the VH5–251 gene segment), suggesting that these rearrangements may be counter-selected.

FIGURE 3.

CDR3 of VHDJH rearrangements from minilocus transgenic TdT+/+ and TdT/− mice. The VH gene segment 3′ end, the D gene segment, and the JH gene segment are indicated. Rearrangements by inversion are indicated with an “i” in front of the D gene segment. For each clone, homologies with the germline gene segments are indicated by dashes. Double lines indicate nucleotides that can derive from recombination at sites of short homologies. P and N nucleotides are indicated. The CDR3 length is indicated in parenthesis when the rearrangement is productive, whereas nonproductive rearrangements are indicated by a dash. A, CDR3 of 104 VHDJH rearrangements using the VH5–251 gene segments obtained from minilocus transgenic TdT+/+ mice. B, CDR3 of 73 VHDJH rearrangements using the ΨVH3–105 gene segments obtained from minilocus transgenic TdT+/+ mice. C, CDR3 of 109 VHDJH rearrangements using the VH5–251 gene segments obtained from minilocus transgenic TdT/− mice. D, CDR3 of 35 VHDJH rearrangements using the ΨVH3–105 gene segments obtained from minilocus transgenic TdT/− mice.

FIGURE 3.

CDR3 of VHDJH rearrangements from minilocus transgenic TdT+/+ and TdT/− mice. The VH gene segment 3′ end, the D gene segment, and the JH gene segment are indicated. Rearrangements by inversion are indicated with an “i” in front of the D gene segment. For each clone, homologies with the germline gene segments are indicated by dashes. Double lines indicate nucleotides that can derive from recombination at sites of short homologies. P and N nucleotides are indicated. The CDR3 length is indicated in parenthesis when the rearrangement is productive, whereas nonproductive rearrangements are indicated by a dash. A, CDR3 of 104 VHDJH rearrangements using the VH5–251 gene segments obtained from minilocus transgenic TdT+/+ mice. B, CDR3 of 73 VHDJH rearrangements using the ΨVH3–105 gene segments obtained from minilocus transgenic TdT+/+ mice. C, CDR3 of 109 VHDJH rearrangements using the VH5–251 gene segments obtained from minilocus transgenic TdT/− mice. D, CDR3 of 35 VHDJH rearrangements using the ΨVH3–105 gene segments obtained from minilocus transgenic TdT/− mice.

Close modal

The human transgenic minilocus contains 10 well-characterized D gene segments (at least one member of each D gene segment families). DHQ52 is preferentially utilized. It is used in 34% and 26% of the rearrangements in minilocus transgenic TdT+/+ and TdT/− mice, respectively. The DN1 and DXP′1 gene segments are also frequently used. D gene segment recombination is very similar between minilocus TdT+/+ and TdT/− mice (Fig. 4; p = 0.99). The only significant difference is in DIR2 gene segment utilization. This particular gene segment is more frequently used in TdT+/+ mice (11% of the rearrangements) than in TdT/− mice (6% of the rearrangements; p < 0.001). However, the possibility that N nucleotides have been attributed to the DIR gene segment cannot be ruled out in TdT+/+ mice. Additionally, in minilocus transgenic TdT/− mice 17 rearrangements (12%) can be explained by invoking VH and JH gene segments only. Although we occasionally note these in TdT+/+ mice, they are more frequent in TdT/− mice (p = 0.001). In the absence of TdT, D gene segment DNA ends may be less protected against exonuclease activity, resulting in a larger number of rearrangements without apparent D gene segments. Alternatively, VH to JH rearrangements may occur more frequently in the absence of TdT in vivo.

FIGURE 4.

D gene segment utilization in human IgH minilocus transgenic TdT+/+ (shaded bars) and transgenic TdT/− (filled bars) mice. D gene segments are represented from the most 5′ (DLR4) to the most 3′ (DHQ52) gene segments. Rearrangements in which the D gene segment could not be identified (too short) as well as rearrangements that do no require a D gene segment and rearrangements with D-D fusion are also indicated.

FIGURE 4.

D gene segment utilization in human IgH minilocus transgenic TdT+/+ (shaded bars) and transgenic TdT/− (filled bars) mice. D gene segments are represented from the most 5′ (DLR4) to the most 3′ (DHQ52) gene segments. Rearrangements in which the D gene segment could not be identified (too short) as well as rearrangements that do no require a D gene segment and rearrangements with D-D fusion are also indicated.

Close modal

The human minilocus contains the complete JH region. Every human JH gene segment is observed in VHDJH rearrangements (Fig. 5). JH4 is preferentially utilized (68% and 40% of the rearrangements in TdT+/+ and TdT/− mice, respectively). JH2 (21% and 20% of the rearrangements in TdT+/+ and TdT/− mice, respectively) is also frequently used in both backgrounds (p = 0.98). However, JH6 (1% and 30% of the rearrangements in TdT+/+ and TdT/− mice, respectively) is more frequently rearranged in the absence of TdT (p < 0.01). In this study, the utilization of JH3 is not representative as this gene segment is poorly amplified using the JH consensus primer. The presence or absence of TdT influences JH gene segment utilization. In particular, JH6 gene segment utilization is favored in the absence of TdT.

FIGURE 5.

JH gene segment utilization in human IgH minilocus transgenic TdT+/+ (shaded bars) and transgenic TdT/− (filled bars) mice. JH gene segments are represented from the most 5′ (JH1) to the most 3′ (JH6) gene segments.

FIGURE 5.

JH gene segment utilization in human IgH minilocus transgenic TdT+/+ (shaded bars) and transgenic TdT/− (filled bars) mice. JH gene segments are represented from the most 5′ (JH1) to the most 3′ (JH6) gene segments.

Close modal

The repertoire of N segments in VHDJH rearrangements from minilocus transgenic TdT+/+ mice is extensive. In minilocus transgenic TdT+/+ mice, 56% and 70% of the rearrangements contain 1–13 additional nucleotides at the VHD and DJH junctions, respectively (Fig. 3, A anbd B). In contrast, rearrangements from minilocus transgenic TdT/− mice are characterized by a low frequency of N regions. Only 3% of the rearrangements show evidence of 1–2 additional nucleotides at the VHD or D-JH junction (Fig. 3, C and D). In minilocus transgenic TdT/− mice, a total of 12 extra-nucleotides are found either at the VHD or the DJH junction. The majority (75%) of the extra nucleotides are T and A (Fig. 3, C and D). The frequency of P nucleotides is similar in TdT+/+ and TdT/− mice. These data show that in the absence of TdT, the number of nonencoded nucleotides in transgenic rearrangements is limited. These nucleotides are added by other mechanisms.

Recombination at sites of short homology is rare in minilocus transgenic TdT+/+ mice. In these mice, 8% of the rearrangements occur at sites of short homology between the VH and D gene segments (1–2 nt) and 6% of the rearrangements occur at sites of short homology between the D and JH gene segments (1–7 nt). In contrast, in minilocus transgenic TdT/− mice, 53% of the rearrangements occur at sites of short homology between the VH and D gene segments (1–3 nt) and 33% of the rearrangements occur at sites of short homology between the D and JH gene segments (1–6 nt). Every VH, D, and JH gene segment is observed at least once in homology-mediated rearrangements. In minilocus transgenic TdT/− mice, 50 (53%) and 16 (59%) rearrangements utilizing the VH5–251 and the ΨVH3–105 gene segments, respectively, occur at sites of short homology at the VHD junction. Rearrangements that can be explained by invoking VH and JH gene segments only also occur by short homology. This is especially true for VH5–251 rearrangements. Five (50%) and one (14%) rearrangements utilizing the VH5–251 and the ΨVH3–105 gene segments, respectively, occur at sites of short homology at the VHJH junction (Fig. 3, C and D). These data show that rearrangements at sites of short homologies occur in the minilocus transgene. However, these types of rearrangements involve, equally, VH5–251 and ΨVH3–105 gene segments. It is noteworthy that this appears to be true whether 1-nt homologies are included or not. Therefore, homology-mediated rearrangements are not the only factor responsible for the changes in gene segment recombination frequencies.

Ig containing human μ and γ1 heavy chains can be detected in the sera of pHC1 transgenic TdT+/+ and TdT/− mice by ELISA. In the absence of immunization, the level of serum human IgH is similar in TdT/− and in TdT+/+ minilocus transgenic mice. Additionally, the minilocus transgenic mice continue producing normal amounts of serum murine IgH (data not shown). Approximately 8% of the splenic B220+ cells express surface human IgH in addition to surface murine heavy chains in both minilocus transgenic TdT+/+ and TdT/− mice (data not shown). These data indicate that the minilocus transgene is expressed without xenotypic exclusion in both mice. TdT expression does not influence endogenous nor transgenic Ig serum and surface levels.

TdT contributes to Ig and TCR diversity by adding random nucleotides at gene segment junctions. The absence of TdT in the murine fetus or in genetically deficient mice results in a low level of N nucleotides and in a high level of homology-mediated rearrangements (15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25). In 1992, we developed mice transgenic for a human IgH minilocus termed pHC1. This model was developed to analyze the mechanisms of V(D)J recombination, somatic mutation, as well as class-switching (27, 28, 29, 30, 31, 32). The major advantage of the human minilocus is that it contains a limited number of well-characterized gene segments, thus allowing straightforward repertoire analyses. Studies in pHC1 transgenic mice previously indicated that the human repertoire generated is extremely diverse. By transferring the human IgH minilocus pHC1 onto the TdT/− background, we were able to analyze Ig expression as well as gene segment recombination to determine whether TdT expression influences repertoire development.

The level of human μ heavy chain expression in the serum and on the B cell surface is similar in minilocus transgenic TdT+/+ and TdT/− mice. Thus, our data clearly indicate that TdT expression does not influence the level of rearrangement, transcription, and translation of the human transgenic gene segments. This also suggests that although genetic differences exist between minilocus transgenic TdT+/+ and TdT/− mice, the effect of these differences on the transgenic minilocus rearrangement cannot be solely responsible for the data.

The absence of TdT during fetal development or in TdT-deficient mice has been reported by several authors to be associated with the production of a more restricted but functional repertoire (15, 16, 17, 18, 20, 21, 22, 23, 24, 25, 26). The more restricted repertoire has been linked to the absence of N nucleotides but also to an increase of homology-mediated recombination (15, 16, 17, 18) and in some circumstances to selection (33). The number of additional nucleotides at the VH-D-JH junctions in minilocus transgenic TdT/− mice is extremely low. It is comparable to the level found in endogenous rearrangements of TdT/− mice (21, 22, 23, 24, 25).

It has been suggested that selection can be responsible for the restricted fetal repertoire (33). The issue of selection in our model is complex. In our system, selection cannot be ruled out but is unlikely to account for the differences in gene segment recombination. Selection could only occur on rearrangements involving the functional VH5–251 gene segment. Our strongest argument against selection is the frequency of productive rearrangements. Without selection, ∼25% of the rearrangements are in-frame (rearrangements using the pseudo-VH gene segment). In TdT/− mice, this percentage is observed with the functional gene segment as well, suggesting an absence of selection. In TdT+/+ mice, this percentage is lower, suggesting counter-selection. Additionally, we think that selection is limited in these mice as no specific Ab using human heavy chains has been found despite the large number of Ags tested by us and by others. VH5–251 rearrangements are more frequent in the TdT+/+ background. However, only 12% of these rearrangements are in-frame. Therefore, in these mice, the increase in VH5–251 gene segment rearrangement is not due to higher positive selection of productive rearrangements. Minilocus transgenic TdT/− mice exhibit an increase in ΨVH3–105 gene segment recombination compared with minilocus transgenic TdT+/+ mice (Fig. 2). This increase cannot be due to antigenic selection because these rearrangements cannot be translated into functional proteins. Furthermore, 50% of the rearrangements utilizing the JH6 gene segment also use the ΨVH3–105 gene segment, meaning that increased JH6 utilization is independent of antigenic selection. It is noteworthy that selection could not be responsible for our data in the fetus, as our analyses were performed at embryonic day 14 before B cells can be detected. In addition, when productive rearrangements are removed from the study, there is still an increase in JH6 gene segment utilization in TdT/− mice. The high frequency of ΨVH3–105 gene segment recombination in TdT/− mice is not simply due to a different murine genetic background. Indeed, the same increment in ΨVH3–105 rearrangements is observed in minilocus transgenic TdT+/+ fetuses, which are on the C57BL/6 background and not on a mix C57BL/6 × 129 background (Fig. 2). This shows that the increase in ΨVH3–105 gene segment recombination is likely related to the absence of TdT during the recombination process regardless of the murine genetic background.

Homology-mediated recombination is an important event in the absence of TdT. In our system, it is involved in more than 50% of the rearrangements involving both the ΨVH3–105 and the functional VH5–251 gene segment. However, only 27% of the rearrangements utilizing JH6 occur by homology. This is not sufficient to explain the dramatic increase in ΨVH3–105 and JH6 gene segment recombination in the absence of TdT. Additionally, as shown in the absence of a D gene segment, homology-mediated rearrangement favors the VH5–251 gene segment. Indeed, 50% of VH5–251-JH rearrangements occur at sites of short homologies vs only 14% of ΨVH3–105-JH rearrangements occurring at sites of short homologies. Therefore, homology-mediated recombination cannot account for the increase in ΨVH3–105 gene segment utilization detected in these type of rearrangements. These data suggest that while homology-mediated recombination may play an important role in repertoire development in the absence of TdT, it is not the only factor that modulates gene segment utilization.

The fact that we observe differences in gene segment utilization in the presence or absence of TdT is surprising. TdT is known to add nucleotides during the joining phase of the recombination reaction (14). Our data suggest that changes in repertoire development in the absence of TdT are gene segment specific and involve mainly gene segments with 23-bp spacer RSS. TdT has been shown to interact with the DNA-dependent protein kinase (DNA-PK) and in particular with the Ku proteins (34, 35). Therefore, we hypothesize that TdT expression may modulate gene segment recombination at the ligation stage. We propose that TdT may interact with DNA-PK through the Ku proteins. TdT then modifies DNA ends by adding random nucleotides and therefore changes their ligation efficiency. The ligation is then mediated by the DNA ligase IV (36). Previous data clearly indicate that coding end sequences play an important role in V(D)J recombination. In particular, coding end sequences appear to influence the formation of DNA double-strand breaks (37, 38, 39, 40). Our data suggest that DNA end may also influence recombination efficiency at the ligation step. Indeed, the N segment addition appears to affect gene segment ligation efficiency. The addition of N nucleotides appears to favor the functional VH gene segment in providing different coding ends, more efficient for the ligation step. The pseudo-VH gene segment, however, is more efficiently recombined when the coding ends are not modified by N nucleotide addition. This effect on ligation efficiency appears to be independent of homology-mediated recombination. Our data also indicate that TdT may affect differently gene segments flanked with 12- or 23-bp spacers. This is in agreement with previous data indicating that requirements for coding end sequences are different for VH, D, and JH gene segments (38, 39, 40).

These studies clearly show that the IgH repertoire may be modulated by the presence or absence of TdT. Repertoire changes appear to be gene segment and RSS specific and to involve 23-nt spacer RSS. The changes are independent of antigenic selection and cannot exclusively be explained by an increase in homology-mediated recombination. The presence or absence of TdT expression modulates CDR3 repertoires. In this paper we show that it may also modulate gene segment utilization by influencing ligation efficiency. Thus, the role played by TdT may be more extensive than previously thought.

We are grateful to N. Lonberg for the pHC1 transgenic mice and to S. Gilfillan and D. Mathis for the TdT/− mice. We thank K. Kamm, K. Wilson, and A. Robben for their technical assistance. We thank R. van der Heijden for his careful review of the manuscript.

1

This work is supported by grants from the American Heart Association Oklahoma Affiliate (9805506S) and from the National Institutes of Health (AI-12127).

4

Abbreviations used in this paper: RSS, recombination signal sequences; N, non-germline-encoded; P, palindromic; CDR, complementarity-determining region.

1
Weaver, D. T..
1995
. V(D)J recombination and double-strand break repair.
Adv. Immunol.
58
:
29
2
Early, P., H. Huang, M. Davis, K. Calame, L. Hood.
1980
. An immunoglobulin heavy chain variable region is generated from three segments of DNA: VH, D, and JH.
Cell
19
:
981
3
Sakano, H., R. Maki, Y. Kurosawa, W. Roeder, S. Tonegawa.
1980
. Two types of somatic recombination are necessary for the generation of complete heavy chain genes.
Nature
286
:
676
4
Hesse, J. E., M. R. Lieber, K. Mizuuchi, M. Gellert.
1989
. V(D)J recombination: a functional definition of the joining signals.
Genes Dev.
3
:
1053
5
Landau, R., D. G. Schatz, M. Rosa, D. Baltimore.
1987
. Increased frequency of N-region insertion in a murine pre-B cell line infected with a terminal deoxynucleotidyltransferase retroviral expression vector.
Mol. Cell. Biol.
7
:
3237
6
Lafaille, J. J., A. DeCloux, M. Bonneville, Y. Takagaki, S. Tonegawa.
1989
. Junctional sequences of T-cell receptor γδ genes: implications for γδ T-cell lineages and for a novel intermediate of V(D)J joining.
Cell
59
:
859
7
Meier, J. T., S. M. Lewis.
1993
. P nucleotides in V(D)J recombination: a fine-structure analysis.
Mol. Cell. Biol.
13
:
1078
8
Bollum, F. J..
1974
. Terminal deoxynucleotidyltransferase. P. Boyer, ed. In
The Enzymes
Vol. 10
:
145
-171. Academic Press, NY.
9
Silverstone, A. E., H. Cantor, G. Goldstein, D. Baltimore.
1976
. Terminal deoxynucleotidyltransferase is found in pro-thymocytes.
J. Exp. Med.
144
:
544
10
Bogue, M., C. Gilfillan, C. Benoist, D. Mathis.
1992
. Regulation of N-region diversity in antigen receptors through thymocyte differentiation and thymus ontogeny.
Proc. Natl. Acad. Sci. USA
89
:
11011
11
Meffre, E., F. Papavasiliou, P. Cohen, O. Bouteiller, D. Bell, H. Karasuyama, C. Schiff, J. Banchereau, Y. J. Liu, M. Nussenzweig.
1998
. Antigen receptor engagement turns off the V(D)J recombination machinery in human tonsil B cells.
J. Exp. Med.
188
:
765
12
Desiderio, S. V., G. D. Yancopoulos, M. Paskind, E. Thomas, M. A. Boss, N. Landau, F. W. Alt, D. Baltimore.
1984
. Insertion of N region into heavy chain genes is correlated with expression of terminal deoxynucleotidyltransferase in B cells.
Nature
311
:
752
13
Lieber, M. R., J. E. Hesse, K. Mizuuchi, M. Gellert.
1988
. Lymphoid V(D)J recombination: nucleotide insertion at signal joints as well as coding joints.
Proc. Natl. Acad. Sci. USA
85
:
8588
14
Grawunder, U., M. R. Lieber.
1997
. A complex of RAG1 and RAG2 proteins persists on DNA after single-strand cleavage at V(D)J recombination signal sequences.
Nucleic Acids Res.
25
:
1375
15
Feeney, A. J..
1992
. Predominance of VH-D-JH junctions occurring at sites of short sequence homology results in limited junctional diversity in neonatal antibodies.
J. Immunol.
149
:
222
16
Feeney, A. J..
1993
. Junctional diversity in the absence of N region: neonatal T cell receptor β chain junctional sequences are more heterogeneous than neonatal T cell receptor γδ or IgH junctions.
J. Immunol.
151
:
3094
17
Chukwuocha, R. U., B. Nadel, A. J. Feeney.
1995
. Analysis of homology-directed recombination in VDJ junctions from cytoplasmic Ig-pre-B cells of newborn mice.
J. Immunol.
154
:
1246
18
Zhang, Y., D. Cado, D. M. Asarnow, T. Komori, F. W. Alt, D. H. Raulet, J. P. Allison.
1995
. The role of short homology repeats and TdT in generation of the invariant γδ antigen receptor repertoire in the fetal thymus.
Immunity
3
:
439
19
Kepler, T. B., M. Borrero, B. Rugerio, S. K. McCray, S. H. Clarke.
1996
. Interdependence of N nucleotide addition and recombination site choice in V(D)J rearrangement.
J. Immunol.
157
:
4451
20
Bentolila, L. A., G. E. Wu, F. Nourrit, M. Fanton-d’Andon, F. Rougeon, N. Doyen.
1997
. Constitutive expression of TdT in transgenic mice is sufficient for N region diversity to occur at any Ig locus throughout differentiation.
J. Immunol.
158
:
715
21
Gilfillan, S., A. Dierich, M. Lemeur, C. Benoist, D. Mathis.
1993
. Mice lacking TdT: mature animals with an immature lymphocyte repertoire.
Science
261
:
1175
22
Gilfillan, S., M. Bachmann, S. Trembleau, L. Adorini, U. Kalinke, R. Zinkernagel, C. Benoist, D. Mathis.
1995
. Efficient immune responses in mice lacking N-region diversity.
Eur. J. Immunol.
25
:
3115
23
Gilfillan, S., C. Benoist, D. Mathis.
1995
. Mice lacking TdT: adult mice with fetal antigen receptor repertoire.
Immunol. Rev.
148
:
201
24
Komori, T., A. Okada, V. Stewart, F. W. Alt.
1993
. Lack of N regions in antigen-receptor variable region genes of TdT-deficient lymphocytes.
Science
261
:
1171
25
Komori, T., L. Pricop, A. Hatakeyama, C. A. Bona, F. W. Alt.
1996
. Repertoires of antigen receptors in TdT congenitally deficient mice.
Int. Rev. Immunol.
13
:
317
26
Weller, S., C. Conde, A. M. Knapp, H. Levallois, S. Gilfillan, J. L. Pasquali, T. Martin.
1997
. Autoantibodies in mice lacking TdT: evidence for a role of N region addition in the polyreactivity and in the affinities of anti-DNA antibodies.
J. Immunol.
159
:
3890
27
Tuaillon, N., L. D. Taylor, N. Lonberg, P. W. Tucker, J. D. Capra.
1993
. Human immunoglobulin heavy chain minilocus recombination in transgenic mice: gene segment utilization in μ and γ transcripts.
Proc. Natl. Acad. Sci. USA
90
:
3720
28
Tuaillon, N., A. B. Miller, N. Lonberg, P. W. Tucker, J. D. Capra.
1994
. Biased utilization of DHQ52 and JH4 gene segments in a human immunoglobulin transgenic minilocus is independent of antigenic selection.
J. Immunol.
152
:
2912
29
Tuaillon, N., A. B. Miller, P. W. Tucker, J. D. Capra.
1995
. Analysis of direct and inverted D-JH rearrangements in a human immunoglobulin heavy chain transgenic minilocus.
J. Immunol.
154
:
6453
30
Tuaillon, N., P. W. Tucker, J. D. Capra.
1995
. VHDJH recombination in a human immunoglobulin heavy chain transgenic minilocus.
Immunologist
3
:
269
31
Tuaillon, N., J. D. Capra.
1995
. V(D)J recombination, class-switching and somatic mutation in human immunoglobulin minilocus transgenic mice. In.
The Antibodies. J. D. Capra and M. Zanetti, eds. Gordon & Breach, New York, pp.
:
1
-39.
32
Tuaillon, N., J. D. Capra.
1998
. Utilization of DIR gene segments in TdT+/+ and TdT−/− mice carrying a human immunoglobulin heavy chain transgenic minilocus.
Proc. Natl. Acad. Sci. USA
95
:
1703
33
Pandey, A., L. W. Tjoelker, C. B. Thompson.
1993
. Restricted immunoglobulin junctional diversity in neonatal B cells results from developmental selection rather than homology-based V(D)J joining.
J. Exp. Med.
177
:
329
34
Mickelsen, S., C. Snyder, K. Trujillo, M. Bogue, D. B. Roth, K. Meek.
1999
. Modulation of terminal deoxynucleotidyltransferase activity by the DNA-dependent protein kinase.
J. Immunol.
163
:
834
35
Mahajan, K. N., L. Gangi-Peterson, D. H. Sorscher, J. Wang, K. N. Gathy, N. P. Mahajan, W. H. Reeves, B. S. Mitchell.
1999
. Association of terminal deoxynucleotidyltransferase with Ku.
Proc. Natl. Acad. Sci. USA
96
:
13926
36
Grawunder, U, D. Zimmer, S. Fugmann, K. Schwarz, M. R. Lieber.
1998
. DNA ligase IV is essential for V(D)J recombination and DNA double-strand break repair in human precursor lymphocytes.
Mol. Cell
2
:
477
37
Gerstein, R. M., M. R. Lieber.
1993
. Coding end sequence can markedly affect the initiation of V(D)J recombination.
Genes Dev
7
:
1459
38
Ezekiel, U. R., P. Engler, D. Stern, U. Storb.
1995
. Asymmetric processing of coding ends and the effect of coding end nucleotide composition on V(D)J recombination.
Immunity
2
:
381
39
VanDyk, L. F., T. W. Wise, B. B. Moore, K. Meek.
1996
. Immunoglobulin D(H) recombination signal sequence targeting: effect of D(H) coding and flanking regions and recombination partner.
J. Immunol.
157
:
4005
40
Ezekiel, U. R., T. Sun, G. Bozek, U. Storb.
1997
. The composition of coding joints formed in V(D)J recombination is strongly affected by the nucleotide sequence of the coding ends and their relationship to the recombination signal sequences.
Mol. Cell. Biol.
17
:
4191