Metabolism of Recombination Coding Ends
in scid Cells
Matthew L. Brown and Yung Chang
This information is current as
of July 31, 2017.
Subscription
Permissions
Email Alerts
This article cites 48 articles, 16 of which you can access for free at:
http://www.jimmunol.org/content/164/8/4135.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2000 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on July 31, 2017
References
J Immunol 2000; 164:4135-4142; ;
doi: 10.4049/jimmunol.164.8.4135
http://www.jimmunol.org/content/164/8/4135
Metabolism of Recombination Coding Ends in scid Cells1
Matthew L. Brown and Yung Chang2
U
nique to developing lymphocytes, the germline V, D,
and J gene segments are somatically recombined to generate a diverse array of Ag receptors (1, 2). V(D)J recombination mechanistically proceeds through two steps, a sitespecific cleavage to generate dsDNA breaks, which is then
followed by general end joining of these breaks. The proteins encoded by recombination-activating genes (RAG13 and RAG2) recognize recombination signal sequences flanking each gene segment and cleave the DNA at the junction of recombination signal
sequences and coding gene segments to produce two types of DNA
ends: blunt signal ends and covalently sealed hairpin coding ends
(3–5). Resolution of signal ends presumably occurs by direct ligation of blunt ends to form precise signal joints (6). In contrast,
resolution of hairpin coding ends is quite complex, but is believed
to proceed through nucleolytic processing, end modification, and
ultimately religation of the broken coding ends (7–9). This multistep process produces coding joints with considerable junction
variability, which significantly contributes to the diversity of the
immune repertoire. Signal ends are readily detectable and are persistent in cells undergoing V(D)J recombination, while coding
ends are only detectable on limited occasions (7–10). Thus, counterintuitively, the direct ligation of signal ends appears to be much
slower than the multistep joining of coding ends.
Department of Microbiology, Arizona State University, Tempe, AZ 85287
Received for publication November 24, 1999. Accepted for publication February
7, 2000.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported by National Institutes of Health Grant CA73857 to Y.C.,
and a fellowship to M.L.B. from Motorola.
2
Address correspondence and reprint requests to Dr. Yung Chang, Department of
Microbiology, Arizona State University, Tempe, AZ 85287-2701. E-mail address:
[email protected]
3
Abbreviations used in this paper: RAG, recombination-activating gene; Ab-MLV,
Abelson murine leukemia virus; ␤2m, ␤2-microglobulin; DNA-PK, DNA-dependent
protein kinase; DNA-PKcs, catalytic subunit of DNA-PK; LM-PCR, ligation-mediated PCR; MBN, mung bean nuclease; P, palindrome; s/⫹, scid heterozygous; s/s,
scid homozygous; ts, temperature sensitive.
Copyright © 2000 by The American Association of Immunologists
Genetic analyses of various mutant cells and mice have revealed
important roles for many proteins in the V(D)J recombination joining process, such as the DNA-end binding protein, Ku 70/80 heterodimer (11–13), the catalytic subunit of DNA-dependent protein
kinase (DNA-PKcs) (14 –16), XRCC4 (17–19), and ligase IV (20 –
23). Except for DNA-PKcs, all of these proteins are required for
the formation of both signal joints and coding joints (11, 19, 21,
24). Cells with a mutation in the DNA-PKcs gene, incurred in scid
mice, retain the ability to form signal joints, but are afflicted with
a defect in the formation of coding joints (25, 26). As such, the role
of DNA-PKcs may be related to the accelerated resolution of coding ends, even though it remains speculative as to how DNA-PKcs
might function in this regard.
Several other proteins have also been implicated in coding end
resolution. RAG1 and RAG2 proteins have been shown to possess
end-processing activities, including binding of recombination intermediates (27, 28), nicking synthetic hairpin ends (29, 30), and
rejoining cleaved signal ends to coding ends (31). Recently, it has
been demonstrated that the Mre11 protein, when complexed with
Rad50 and the Nijmegen breakage syndrome gene product, Nbs1,
exhibits several nuclease activities in vitro, such as hairpin nicking
and processing of opened ends (32). Thus, it is conceivable that the
Mre11/Rad50/Nbs1 complex may also participate in coding end
resolution in vivo. However, it is not clear how these different
protein complexes, DNA-PK, RAG1/2, and Mre11/Rad50/Nbs1,
interact and execute their function during the processes of nicking,
trimming, and joining.
It is assumed that opened coding ends are the intermediate products during the conversion of hairpin ends into coding joints. Although opened coding ends were detected in developing lymphocytes, direct evidence for their conversion from hairpin coding
ends is nonetheless lacking (9). In an effort to elucidate coding end
resolution, we have developed recombination-inducible cell lines
from both scid homozygous (s/s) and scid heterozygous (s/⫹) mice
by transforming B cell precursors with the temperature-sensitive
Abelson murine leukemia virus (ts-Ab-MLV). As we reported previously, scid ts-Ab-MLV cell lines exhibit a temperature-dependent ability to resolve recombination coding ends (33). In our current study, we examined the intermediate structures of coding ends
0022-1767/00/$02.00
Downloaded from http://www.jimmunol.org/ by guest on July 31, 2017
V(D)J recombination cleavage generates two types of dsDNA breaks: blunt signal ends and covalently sealed hairpin coding ends.
Although signal ends can be directly ligated to form signal joints, hairpin coding ends need to be opened and subsequently
processed before being joined. However, the underlying mechanism of coding end resolution remains undefined. The current study
attempts to delineate this process by analyzing various structures of coding ends made in situ from recombination-inducible pre-B
cell lines of both normal and scid mice. These cell lines were derived by transformation of B cell precursors with the temperaturesensitive Abelson murine leukemia virus. Our kinetic analysis revealed that under conditions permissive to scid transformants,
hairpin coding ends could be nicked to generate 3ⴕ overhangs and then processed into blunt ends. The final joining of these blunt
ends followed the same kinetics as signal joint formation. The course of this process is in sharp contrast to coding end resolution
in scid heterozygous transformants that express the catalytic subunit of DNA-dependent protein kinase, in which hairpin end
opening, processing, and joining proceeded very rapidly and appeared to be closely linked. Furthermore, we demonstrated that
the opening of hairpin ends in scid cells could be manipulated by different culture conditions, which ultimately influenced not only
the level and integrity of the newly formed coding joints, but also the extent of microhomology at the coding junctions. These
results are discussed in the context of scid leaky recombination. The Journal of Immunology, 2000, 164: 4135– 4142.
4136
in both scid and s/⫹ cell lines. We found that the scid cells are
capable of converting the newly produced hairpin coding ends into
opened ends that possess either a staggered 3⬘ overhang or a blunt
end. The quality and quantity of the end processing in these scid
cells can be manipulated by altering the culture conditions. However, this conditional resolution of coding ends is much slower
than the coding end resolution in the control s/⫹ cells. Therefore,
by comparing the kinetics of coding end resolution in scid and s/⫹
cells, we provide evidence for the role of DNA-PKcs in linking the
two steps of the recombination process: cleavage and resolution.
Materials and Methods
Cell culture
DNA preparation
DNA was prepared in agarose plugs, as previously reported (33, 34).
Briefly, cells resuspended in the agarose mixture were solidified to form
agarose plugs (Bio-Rad, Richmond, CA), and then deproteinized by proteinase-K treatment in 200 ␮l of lysis buffer (100 mM Tris, pH 8, 25 mM
EDTA, 1% Sarkosyl, 400 ␮g/ml proteinase K) at 50°C overnight. Multiple
washes were done over a 24-h period with TE/PMSF (10 mM Tris, pH 8,
1 mM EDTA, 0.5 mM PMSF) and TE only.
Probe
Blots were hybridized with 32P-labeled probes of: 1) pJ␬ plasmid (36) to
analyze the J␬-related PCR products (signal ends, coding ends, and coding
joints); 2) V␭1 insert to detect V␭1J␭1 signal joints; and 3) pActin plasmid
to reveal ␣-actin PCR products (36).
DNA sequence analysis of coding ends and coding joints
The LM-PCR products for J␬ coding ends were further amplified by YC25
and YC32 (5⬘-GAGCATGGTCTGAGCACCGAGT-3⬘) primers, and the
VJ␬ PCR products were amplified with MB46 and the J␬1 primer YC32
for sequence analysis. The amplified products were purified on a 2% NuSieve agarose gel. The isolated PCR fragments were cloned into a
TOPO-TA vector (Invitrogen, San Diego, CA) and sequenced on an automated DNA sequencer (ABI 377; Applied Biosystems, Foster City, CA).
The sequence from each clone was compared with the germline V␬ and J␬
regions by BLAST similarity from the GenBank database.
Western blot analysis
Cells were swelled on ice in hypotonic buffer (10 mM Tris-HCl, pH 7.9, 10
mM KCl, 1 mM MgCl2, 1 mM DTT) including proteinase inhibitors (1
␮g/ml aprotinin, 1 ␮g/ml leupeptin, 1 ␮g/ml pepstatin A, and 0.5 mM
PMSF), lysed by addition of Nonidet-P40 to 0.05%, and microcentrifuged
to pellet the nuclei. The nuclear proteins were eluted from the pellet in
extraction buffer (50 mM Tri-HCl, pH 7.9, 300 mM KCl, 12.5 mM MgCl2,
1 mM EDTA, 20% glycerol, 1 mM DTT, and proteinase inhibitors). Protein concentration of the extract was determined by Bradford assay (BioRad), and 50 ␮g of each sample was subjected to 10% SDS-PAGE and
transferred to NitroBind membrane (Fisher, Pittsburgh, PA). Membranes
were incubated with Abs to RAG2 (PharMingen, San Diego, CA), Mre11
(Novus Biologicals, Littlton, CO), or actin (Santa Cruz Biotechnology,
Santa Cruz, CA), followed by HRP-conjugated secondary Abs, and visualized by chemiluminescence detection (Pierce, Rockford, IL).
RNA preparation and RT-PCR
Results
RNA was prepared from various cell samples using CsCl ultracentrifugation method (35). RNA was reverse transcripted into cDNA with random
oligonucleotides. The cDNA was serially diluted and amplified for RAG2
and ␤2-microglobulin (␤2m) genes using the oligonucleotides described
previously (35). Amplification of the ␤2m message served as an internal
control for input cDNA. These PCR products were analyzed by Southern
blot. A probe for RAG2 was prepared by PCR amplification of RAG2
constructs (kindly provided by M. Gellert, National Institutes of Health).
The probe for ␤2m was a gel-purified PCR product made with primers
specific for ␤2m cDNA.
Opening of hairpin coding ends in scid-ts cells: quantity and
quality affected by in vitro culture conditions
Assessment of J␬ coding end heterogeneity by modified LM-PCR
One-third of the agarose plug was incubated with either 5 U of T4 DNA
polymerase (New England Biolabs) or 5 U of mung bean nuclease (MBN;
Boehringer Mannheim, Indianapolis, IN), followed by 5 U of T4 DNA
polymerase. The enzymes were either heat inactivated at 75°C for 10 min
(to inactivate T4 DNA polymerase) or by treatment with proteinase K (to
inactivate MBN). The samples were then subjected to linker ligation with
2 U of T4 DNA ligase (Boehringer Mannheim) at 16°C for 18 h. LM-PCR
was used to assess the level of dsDNA breaks resulting from ␬-chain gene
rearrangement. The linker was generated by annealing the two primers,
MB-216 (5⬘-CACGAATTCCC-3⬘) and YC-25 (5⬘-GCTATGTACTACCC
GGGAATTCGTG-3⬘), as described previously (35). The ligated reaction
was subjected to PCR reaction using the linker-specific primer (YC-25)
and the locus-specific primers, MB224 (5⬘-AGTGCCACTAACTGCTGA
GCCACCT-3⬘) for amplifying J␬ signal ends, and YC-40 (5⬘-CCAAG
CTTTCCAGCTTGGTCCCCCCTCCGAA-3⬘) for J␬1/2 coding ends. For
better quantitation, only one round of PCR with 28 cycles was applied to
amply the ligated ends (33). For comparison, VJ␬ coding joints and ␣-actin
were also amplified by PCR (35). The level of amplified ␣-actin products
served as a control for the amount of input DNA. PCR products were
separated by electrophoresis and analyzed by Southern blotting.
PCR analysis of signal joints
The V␭1J␭1 signal joints were amplified by PCR using oligomers specific
to 3⬘V␭1 (YC24, 5⬘-CAATGATTCTATGTTGTGCC-3⬘) and to 5⬘J␭1
(YC23, 5⬘-GCTGCATACATCACAGATGC-3⬘). To determine the structure of signal joints as well as signal ends, their corresponding PCR products (one-third of the PCR reaction) were digested with ApaLI (New England Biolabs, Beverly, MA), and analyzed by Southern blotting along
with the undigested PCR samples (one-sixth of the PCR reaction).
Recently, by using recombination-inducible cell lines transformed
with the ts-Ab-MLV mutant, we have shown that coding joint
formation in scid cells could be manipulated by changing the culture temperature (33). In contrast to the s/⫹-ts cells that can form
coding joints immediately after recombination induction (at 39°C),
scid-ts cells accumulate a substantial population of hairpin coding
ends. The resolution of these hairpin coding ends is dependent on
returning the cells to 33°C. To pinpoint the detailed steps in end
resolution, we have analyzed different structures of coding ends
under two culture conditions, specifically a 3-day culture at 39°C
(3-0), and a 2-day culture at 39°C, followed by a 1-day culture at
33°C (2-1). The blunt opened ends can be directly amplified by
ligation-mediated PCR (LM-PCR), while the staggered ends are
revealed by treatment with T4 DNA polymerase, followed by LMPCR. T4 DNA polymerase flushes staggered ends, as it has 3⬘ to
5⬘ exonuclease and 5⬘ to 3⬘ polymerase activities. The hairpin
ends, after being opened by MBN and flushed by T4 DNA polymerase, can also be amplified by LM-PCR. A summary of our
detection scheme is presented in Fig. 1A. To minimize any DNA
damage caused by the routine extraction procedure, DNA samples
were prepared in an agarose plug, as described by Schlissel (9).
Fig. 1B is a representative analysis of several LM-PCR experiments. It is clear that a large amount of blunt opened ends is
present in scid-ts cells cultured at 39°C, followed by an incubation
at 33°C (Fig. 1B, lane 6). As treatment with T4 DNA polymerase
did not significantly increase the amount of LM-PCR products
(Fig. 1B, lanes 4 and 6), staggered ends constitute a small fraction
of the coding ends. Pretreatment of DNA with MBN and T4 DNA
polymerase did not significantly increase the detection of the coding ends (Fig. 1B, lanes 2 and 6), indicating that only a few hairpin
ends remained under this culture condition. Although it is possible
that the level of hairpin ends is underestimated, the detection of a
Downloaded from http://www.jimmunol.org/ by guest on July 31, 2017
As described previously, temperature-sensitive pre-B cell lines, A-1 and
FL2-1, were derived from bcl-2 transgenic s/s and s/⫹ mice, respectively,
by transformation of fetal B cell precursors with ts-Ab-MLV (33). These
ts-Ab-MLV-transformed s/⫹ and s/⫹ cell lines are referred to as s/⫹-ts
and scid-ts cells, respectively. Cells were maintained at 33°C. To induce
V(D)J recombination, cells were incubated at 39°C. To facilitate recombination resolution, the cells cultured at 39°C were returned to 33°C for
various times.
DNA RECOMBINATION
The Journal of Immunology
significant amount of hairpin ends in other samples (Fig. 1B, lane
1) establishes the adequacy of our assay. Interestingly, the size of
blunt opened ends seems slightly smaller than the full length of
artificially nicked hairpin ends (Fig. 1B, lanes 1 and 6), indicating
a loss of several nucleotides at the ends during such conversion.
Thus, a substantial number of hairpin coding ends can be opened
and processed to blunt ends in scid-ts cells once they have been
returned to 33°C. Since the appearance of blunt coding ends occurred concurrently with the production of coding joints, it is possible that these blunt ends are readily ligatable and are the immediate precursors of coding joints.
In contrast, very few blunt opened coding ends were detected in
the cells cultured at 39°C for 3 days (Fig. 1B, lane 5). A strong
LM-PCR band did appear in the sample pretreated with MBN and
T4 DNA polymerase (Fig. 1B, lane 1). Thus, the vast majority of
the coding ends remained in a hairpin structure. The amount of
PCR products that represent opened ends increased when the DNA
sample was pretreated with T4 DNA polymerase, indicating the
presence of some staggered opened coding ends (Fig. 1B, lanes 1
and 3). The PCR products amplified from these staggered ends are
much more heterogeneous than the LM-PCR band without pretreatment of T4 DNA polymerase (Fig. 1B, lanes 3 and 5). The
faster moving bands suggest the presence of ends with extensive
loss of nucleotides. Hence, even though hairpin nicking may occur
FIGURE 2. A, Aberrant processing of the hairpin coding ends at the
nonpermissive temperature. scid-ts cells were cultured at 39°C for either 2
days (2-0), 4 days (4-0), or 4 days, followed by 1 day at 33°C (4-1). The
isolated DNA samples were treated with MBN/T4 DNA polymerase (T4
Pol) alone, or no treatment before being assayed by LM-PCR. B, Lack of
enhanced detection of coding ends by pretreatment with Klenow. The 4-0
DNA sample was pretreated with T4 Pol or Klenow before being subjected
to LM-PCR. Details for LM-PCR are described in Fig. 1.
at 39°C (3-0), this event is significantly less frequent and considerably more error prone than the hairpin opening in the shift-down
culture (Fig. 1B, compare lanes 3 and 4).
To further understand the processing of coding ends at the nonpermissive temperature, we compared the structures of coding
ends in cells cultured at 39°C for 2 days (2-0) vs 4 days (4-0). The
sample from cells cultured at 39°C for 4 days followed by 1 day at
33°C (4-1) was included as a control, which again showed predominantly blunt opened ends (Fig. 2A, lane 8). The 2-0 sample
contained mostly hairpin coding ends, and even though there was
a small number of staggered ends, no blunt opened ends were
present. The 4-0 sample, on the other hand, showed an accumulation of the staggered ends with a concurrent reduction in the
amount of hairpin ends. A small number of blunt opened ends also
emerged (Fig. 2A, lane 6), but at a level much lower than that in
the 4-1 control (Fig. 2A, lane 8). Thus, at 39°C, both blunt and
staggered ends are gradually increased over time (Fig. 2A, lanes 2
and 4). Strikingly, an increase in the level of staggered ends was
accompanied by a decrease in the size of these ends over the
course of 2, 3, and 4 days (Fig. 1B, lane 3, and Fig. 2A, lanes 2 and
5). These staggered ends are likely to bear a 3⬘ overhang structure
since treatment with Klenow exo⫺/⫺ (Promega), which fills in a 5⬘
overhang, did not increase the detection of the ends (Fig. 2B, lane
3). The size of the blunt ends remained relatively unchanged (Fig.
1B, lane 3, and Fig. 2, A, lane 6, and B, lane 1). This finding
indicates that blunt ends are relatively stable, whereas staggered
ends are susceptible to successive deletions during an extended
culture at 39°C.
A summary of the relative distribution of the possible coding
end species is presented in Fig. 3. After quantification of the coding end products (see Fig. 3 legend), it becomes clear that the
Downloaded from http://www.jimmunol.org/ by guest on July 31, 2017
FIGURE 1. Coding end resolution in scid ts-Abl-MLV cell line. A,
Schemetic illustration of the modified LM-PCR for detecting hairpin, staggered, and blunt ends. The DNA fragments and ligation linkers are represented by the thick and thin lines, respectively. B, Detection of coding ends
by modified LM-PCR. scid-ts cells were shifted to 39°C for either 3 days
(3-0) or 2 days, followed by 1 day at 33°C (2-1). DNA was prepared in
agarose plugs, as described in Materials and Methods. Coding ends (CE)
were analyzed by LM-PCR at the J␬1 locus. MBN and T4 DNA polymerase (T4 Pol) were used to create ligatable blunt ends. Coding joint (CJ)
formation was assayed by PCR using a V␬ primer and a J␬2 primer. Amplification of the actin gene was used as a control for the amount of DNA.
4137
4138
DNA RECOMBINATION
did lead to the addition of a P nucleotide. This indicates that nicking of endogenous hairpin ends is likely to generate 3⬘ overhangs,
but not 5⬘ overhangs. In addition, the structural comparison of
coding ends made at 39°C for 2 days vs 4 days reaffirms that the
longer the culture at 39°C, the more abnormal the ends.
Different culture conditions affect the quality of coding joints
hairpin ends are processed differently under the two culture conditions. Only 1 day after returning the cells to 33°C from 39°C,
virtually all hairpin ends were converted to blunt opened ends.
Even 2 extra days at 39°C (4 days total) only caused a slight
reduction in the number of hairpin ends. This limited hairpin opening also resulted in a preponderance of staggered ends over
blunt ends.
To gain further insight into the coding end structures, the LMPCR products were cloned for sequence analysis. As shown in
Table I, none of the staggered ends contained palindrome (P) nucleotides that would accompany ends that possessed a 5⬘ overhang.
On the other hand, the hairpin end opening made in vitro by MBN
Table I. Coding end structures under different culture conditions
a
Culture
Conditionsa
DNA
Treatment
Deletion at
J␬1 Locus
No. of
Isolates
3–0
MBN/T4
⫹1
⫺2
⫺3
⫺43
5
1
1
1
2–0
T4
⫺32
⫺54
⫺63
⫺74
⫺77
2
1
1
1
1
4–0
T4
⫺45
⫺49
⫺53
⫺60
⫺80
⫺146
⫺178
⫺270
⫺282
1
1
1
1
1
1
1
1
1
See legends to Figs. 1 and 2 for culture conditions.
Kinetic analysis of blunt end resolution in both scid-ts and
s/⫹-ts cells
We have shown in Figs. 1 and 2 that the majority of the coding
ends present at 33°C are blunt opened ends. Is it possible that these
blunt opened ends are subjected to further nucleotide deletions, as
Downloaded from http://www.jimmunol.org/ by guest on July 31, 2017
FIGURE 3. Distribution of the coding end structures is affected by culture temperatures. This graph is tabulated from Figs. 1 and 2A. Each column represents the percentage of each coding end structure in total coding
ends under various culture conditions. Note that for samples 2-0 and 4-0,
the three types of coding ends do not overlap in size. The amount of
relative coding ends, including hairpin (H), staggered (S), and blunt (B),
was quantitated using a phosphor imager and ImageQuant software, and
adjusted against the actin control (i.e., relative hairpin end level (H) ⫽
cpm ⫺ CE/cpm ⫺ actin). The total level of coding ends (TCE) is the
summation of the three types of coding ends (TCE ⫽ H ⫹ S ⫹ B). The
relative percentage of each coding end structure was derived by dividing
the relative coding ends by the TCE, so that the percentage of hairpin
ends ⫽ H/TCE ⫻ 100%. For the sample 3-0, the TCE is the sum of H and
S, and B is not included because S and B overlapped in size (Fig. 1, lanes
3 and 5). For the 2-1 sample (see Fig. 1), S is derived by subtracting the
relative activity (cpm ⫺ CE/cpm ⫺ actin) of lane 4 with B, while H is
calculated by subtracting the relative activity of lane 2 with S and B.
Because a longer incubation of scid cells at 39°C leads to an aberrant processing of coding ends, we reasoned that the fidelity of
coding joints could be increased by shortening the culture length at
this temperature. To that end, we set up two disparate culture conditions. The first condition was a 12-h incubation at 39°C, followed by a 1-day incubation at 33°C (0.5-1), while the second
condition was a straight 4-day incubation at 39°C (4-0).
The amount of coding joints is expected to be different because
the cells under these two culture conditions have different activities
in recombination cleavage (due to different level of RAG expression; unpublished observation) and end joining (attributed to different activity in end opening; Figs. 1 and 3). Instead, our objective
was to examine the quality of coding joints. The PCR products
were cloned and sequenced, and 24 junctions from the 0.5-1 sample and 20 junctions from the 4-0 sample are presented graphically
in Fig. 4. We mainly focused on the J␬1 region for comparison, as
the V␬ primer is only 100 bp upstream of the recombination signal
sequence and thus might lead to an underestimation of any excessive V␬ deletions. Consistent with our previous analysis (33), nontemplated nucleotide (N) addition was not detected among all of
the recovered joints. P addition with one nucleotide was identified
in three clones. Forty-two percent of the coding joints in the 0.5-1
sample contained 0 –10 nucleotide deletions. The two perfect
joints are independent clones, as their V␬ genes belong to two
different V␬ gene families (data not shown). Twelve junctions had
deletions ranging from 12 to 56 bp, and only two had lost more
than 100 bp. In sharp contrast, 19 of 20 junctions recovered from
the 4-0 sample had lost more than 40 nucleotides. Among them,
one-half contained deletions ranging from 131 to 228 bp. Thus,
scid-ts cells cultured at 39°C for extended time gave rise to more
deleted joining products than those cells cultured at 39°C for limited times. In light of the large deletions in the staggered ends from
the 4-0 sample (Fig. 2A and Table I), the deletions found at the junction of their coding joints are contributed mainly by the abnormal end
processing. The extensive loss of nucleotides at the ends could result
from either inappropriate nicking distal from the hairpin termini
and/or by successive deletions from the opened ends.
It is striking that 75% of the junctions recovered from the 0.5-1
sample contain 1- to 5-bp homology, whereas only 15% of the
junctions from the 4-0 sample have this homology. This could be
indicative of two different end joining mechanisms employed by
the cells under different culture conditions. As such, the high level
of microhomology in the 0.5-1 sample may reflect a preference for
homology-directed joining, in which the alignment of two coding
ends may be facilitated by the presence of homologous nucleotides. On the other hand, the end joining in the 4-0 sample seems
to be relatively independent of microhomology. It is possible that
the coding end conformation or the end/protein complex in the 4-0
sample may preclude homology-mediated alignment. Instead,
these extensively deleted coding ends might be joined through a
random ligation event.
The Journal of Immunology
4139
seen with the staggered ends at 39°C? Additionally, can these ends
be directly ligated similar to the formation of signal joints? To
address these questions, we performed a kinetic analysis of end
resolution for both signal ends and coding ends in s/⫹-ts and
scid-ts cells. We set up a 3-day culture at 39°C, followed by a
series of incubations at 33°C: 1 day (3-1), 3 days (3-3), and 6 days
(3-6).
In the s/⫹-ts cells, coding ends were virtually undetectable by
our LM-PCR assay (Fig. 5A, lane 1), and neither could hairpin or
staggered coding ends be detected (unpublished observation). Yet,
a high level of coding joints appeared upon recombination induction. We subjected the ligated DNA to one round of PCR with 28
cycles of amplification, in which coding ends could be readily
detected in scid-ts cells, but not in s/⫹-ts cells. The presence of
signal ends in the same cell samples indicates that the paucity of
coding ends detected is not due to technical limitations in our
LM-PCR. Rather, our data demonstrate a rapid conversion of hairpin coding ends to coding joints in s/⫹-ts cells.
In contrast, the blunt opened coding ends in scid-ts cells reached
a maximum level 1 day after shifting the cells from 39°C to 33°C
(similar finding to Fig. 1), indicating temperature-dependent end
opening. During the course of their resolution, the size of the coding ends remained relatively intact (Fig. 5A, lanes 6 and 7). Thus,
after returning to 33°C, the blunt opened ends are not susceptible
to gross nucleotide deletions. Interestingly, even in the presence of
an abundant amount of blunt coding ends, the accumulation of
coding joints occurred rather slowly over the extended incubation
at 33°C. Nonetheless, this was concurrent with the gradual disappearance of blunt coding ends and signal ends (Fig. 5A, lanes 5, 6,
7, and 8). Hence, even though scid-ts cells can eventually open all
the hairpin ends at 33°C, the joining of these ends is still much
slower than the coding joint formation from the uncleaved recombination locus in s/⫹-ts cells (Fig. 5A, lanes 1 and 5– 8). This
finding suggests that the recombination events, site-specific cleavage, hairpin end nicking, and end joining, are closely linked in
s/⫹-ts cells, but dissociated in scid-ts cells defective in
DNA-PKcs.
It has been known that the scid mutation also moderately affects
the resolution of signal ends (25). To further test this, we compared
the newly generated signal ends and signal joints in s/⫹-ts and
scid-ts cells. The integrity of signal ends and signal joints can be
assessed by ApaLI restriction digestion of amplified PCR products,
as an ApaLI restriction site is created in the perfect signal joints as
well as in the intact signal ends that are ligated to the artificial
primer. It is clear from Fig. 5B that signal ends made in both s/⫹-ts
and scid-ts cells are intact without nucleotide modifications, as
their LM-PCR products are sensitive to ApaLI digestion. Likewise,
the integrity of these ends is not altered in the V␭1 and J␭1 signal
ends (unpublished observation).
Due to the simplicity of the ␭-locus, we examined the formation
of V␭1J␭1 signal joints. Although some signal joints were found
in the s/⫹-ts cells cultured at 39°C, more signal joints appeared in
the cells returning to 33°C (Fig. 5B, lanes 1 and 3). The temperature-dependent resolution of signal ends is even more apparent in
scid-ts cells (Fig. 5B, lanes 5 and 7). Based on the sensitivity to
ApaLI digestion, it is clear that scid signal joints contain nucleotide
modifications (Fig. 5B, lanes 6 and 8), whereas signal joints made
in s/⫹-ts cells have perfect junctions (Fig. 5B, lanes 2 and 4).
These results are consistent with the previous report that the scid
mutation alters the efficiency and accuracy of signal end resolution
Downloaded from http://www.jimmunol.org/ by guest on July 31, 2017
FIGURE 4. Quality of coding joints is influenced
by the length of incubation at the nonpermissive temperature. scid-ts cells were cultured for 12 h at 39°C,
followed by 1 day at 33°C (0.5-1) or for 4 days at
39°C (4-0). Coding joints were amplified, purified,
cloned, and sequenced, as described in Materials and
Methods. Each box in the graphical display represents
20 bp, and the darker boxes represent the coding regions. The number of nucleotides lost during recombination is displayed on the side of each bar. One P
nucleotide (C) addition was identified in three clones,
designated as ⫹1 (P). The nucleotides with microhomology at the ends of each junction are presented.
4140
DNA RECOMBINATION
FIGURE 7. Mre11 expression is not altered by temperature changes.
Nuclear lysates were analyzed for Mre11 by Western blot with a polyclonal
Ab specific to Mre11. The same blot was stripped and analyzed for actin.
FIGURE 5. A, Delayed resolution of blunt coding ends in scid-ts cells.
scid-ts and s/⫹-ts cells were cultured for 3 days (3-0) at 39°C, then shifted
back to 33°C for either 1 (3-1), 3 (3-3), or 6 (3-6) days. Both coding ends
and signal ends were examined at the J␬1 locus by LM-PCR. Coding joints
were examined using a V␬ and a J␬2 primer. Actin was included as an
internal control for DNA in each reaction. B, Signal end resolution. J␬
signal ends and V␭1J␭1 signal joints were amplified by LM-PCR.
Genomic DNA was prepared from both s/⫹-ts and scid-ts cells that had
been cultured at 39°C for 3 days or 3 days at 39°C, followed by 1 day at
33°C. The PCR products treated with or without ApaLI were analyzed by
Southern blot. The slight down-shift of PCR products indicates a digestion
by ApaLI, i.e., the presence of the perfect signal joints or intact signal ends.
(25, 37). Despite the modified signal joints, the structure of the
signal ends remain intact (Fig. 5B and unpublished observation).
Therefore, the scid mutation seems to interfere with the signal end
resolution at the step of alignment and ligation.
Expression of RAG2 and Mre11
The biochemical nature of the conditional resolution of coding
ends remains to be defined. It has been shown that RAG1/2 proteins can nick artificial hairpin ends in a cell- free system (29, 30).
To determine whether these proteins are responsible for processing
hairpin ends in vivo, we analyzed RAG2 expression in various cell
samples. Similar to the previous report (38), a high level of RAG2
RNA and protein is induced during the culture at 39°C (Fig. 6, A
and B). Upon returning the cells to 33°C, both RAG2 RNA and
protein levels were substantially reduced (Fig. 6, A and B), during
which time coding ends were found to be processed and joined
(Fig. 1). Thus, coding end resolution occurs in the cells that express little RAG2.
Mre11, when complexed with Rad50 and Nbs1, was found to
possess nuclease activity, nicking fully paired artificial hairpin
ends in a cell-free system (32). To determine whether temperaturedependent end resolution could be attributed to the differential expression of Mre11, we examined the amount of Mre11 protein
expressed in the cells under our defined culture conditions by
Western blot. As shown in Fig. 7, Mre11 protein levels appear
comparable in s/⫹-ts and scid-ts cells, and, importantly, it does not
fluctuate during temperature changes. Because the Mre11 nuclease
is constitutively available to both s/⫹-ts cells and scid-ts for possible nicking and trimming, the temperature-dependent resolution
of scid coding ends may instead reflect different accessibility of the
ends to this enzyme.
Discussion
The resolution of hairpin coding ends must proceed through a multitude of steps, which include opening, processing, aligning, and
joining. The kinetic comparison on the resolution of newly generated hairpin ends in s/⫹-ts and scid-ts cells reveals two different
pathways for resolving coding ends: DNA-PKcs dependent and
DNA-PKcs independent. Upon recombination induction, DNAPKcs-proficient cells produce a great abundance of coding joints,
but very few detectable coding ends (Fig. 5A). This suggests that
recombination cleavage is coupled to end processing and joining.
This coupled event was also demonstrated in a cell-free assay, in
which coding joint formation was much more efficient when intact
recombination substrates rather than precleaved coding ends were
Downloaded from http://www.jimmunol.org/ by guest on July 31, 2017
FIGURE 6. Temperature-dependent regulation of RAG expression. A,
RNA was isolated from cells cultured at 33°C (33), 39°C for 2 days (2– 0),
or 39°C for 2 days, followed by 1 day at 33°C (2-1), and analyzed by
RT-PCR for RAG2 and ␤2m levels. B, Nuclear lysates prepared from the
above cell samples were examined for RAG2 protein by Western blot. The
same blot was stripped and analyzed for actin.
The Journal of Immunology
like in that it involves a DNA-PKcs-independent opening, which is
slow and uncoupled. This uncoupled recombination process would
predict that the unresolved coding ends should be available for
nonspecific end association, such as forming hybrid joints and interlocus recombination products. Indeed, we have recently shown
that scid-ts cells have a higher frequency of hybrid joints and V␬J␭
interlocus joints than their normal counterpart 4. Thus, the price for
a higher level of leaky scid recombination would be an increased
risk for chromosomal instability. This correlation was also demonstrated in those scid mice that were exposed to ionizing radiation: a high level of recombination coding joints was accompanied
by an elevated level of interlocus recombination products and an
accelerated development of thymoma (45). The ability to manipulate our cell lines for coding end resolution will help us to identify
the factors responsible for inducing or promoting the development
of the scid leaky phenotype, as well as those factors responsible for
triggering chromosomal instability.
It has been shown that the recombination intermediates are held
in a postcleavage complex that contains RAG1 and RAG2 proteins
(27, 28, 46). The purified RAG1/2 proteins were found to mediate
the reverse reaction of the cleavage to generate hybrid joints and
open-shut products (31). Recently, two laboratories have shown
that RAG1/2 proteins can nick hairpin ends (29, 30). Interestingly,
however, the resulting opened ends contained 5⬘ overhangs, which
differ from the opened coding ends made in vivo (Table I; 9). In
our studies, the resolution of hairpin coding ends occurs when the
RAG2 protein level is reduced. This finding suggests that hairpin
end processing in scid-ts cells is not likely to be mediated by the
RAG1/2 proteins. Instead, the reduction of RAG1/2 expression at
33°C (Fig. 6) may help to disassemble the postcleavage complex
and to facilitate formation of both coding and signal joints (Fig. 5,
A and B). Thus, the most logical explanation for the temperaturedependent resolution of coding ends and signal ends in scid-ts cells
is that both types of ends become accessible at 33°C to DNA
processing and/or joining machinery, as signal ends do not require
end processing.
It has also been demonstrated that Mre11 acts as a 3⬘-5⬘ dsDNA
exonuclease as well as an endonuclease (47). After being complexed with Rad50 and Nbs1, Mre11 can nick fully paired hairpin
ends to produce 3⬘ overhang ends in vitro (32). This complex can
also cleave 3⬘ overhangs at a double-/single-strand transition,
which would generate blunt ends (48). The dual activities of this
complex perfectly fit the characteristics of the opened coding ends
detected in our scid-ts transformants, in which both blunt opened
ends and staggered ends with 3⬘ overhangs are present (Table I).
The accumulation of unresolved hairpin ends in scid-ts cells at
39°C may reflect an inaccessibility of these ends to Mre11, presumably due to their binding by the RAG1/2 complex. Then, it
should not be unexpected that the level of Mre11 protein does not
fluctuate during temperature changes (Fig. 7). Whether or not the
Mre11 complex is the nicking enzyme in the DNA-PK-independent resolution of coding ends is currently under investigation.
Furthermore, Mre11 protein has been shown to facilitate joining of
mismatched ends presumably through its endo- and exonuclease
activity to reveal the homologous nucleotides for end aligning and
joining (48). In this context, different accessibilities of the ends to
Mre11 can also explain the different levels of microhomology observed at the coding junctions made under the two culture conditions (Fig. 3). Collectively, our results support the assumption that
4
S. Lew, D. Franco, and Y. Chang. Activation of V(D)J recombination induces the
formation of interlocus joints and hybrid joints in scid pre-B cell lines. Submitted for
publication.
Downloaded from http://www.jimmunol.org/ by guest on July 31, 2017
provided (39). In contrast, the resolution of coding ends in DNAPKcs deficient cells is a very protracted process, as several structures of coding ends could be identified: hairpin ends, staggered
opened ends, and blunt opened ends (Fig. 3). Furthermore, the
direct ligation of blunt coding ends was much slower than the
formation of coding joints from the uncut ␬-locus in s/⫹-ts cells
(Fig. 5). Thus, DNA-PKcs may be responsible for coupling those
processing events to facilitate coding joint formation. Without
functional DNA-PKcs, these events proceed separately and slowly.
The coding end resolution mediated by the DNA-PK-independent pathway can be manipulated in terms of its efficiency and
fidelity by altering the culture conditions, as shown in Fig. 1 and
Table I. This finding points to two different schemes utilized by
scid-ts cells to process coding ends. One, occurring at 33°C, can
convert essentially all hairpin ends into blunt opened ends that
possess limited nucleotide deletions. Otherwise, at 39°C, only a
small fraction of hairpin ends are converted into staggered ends,
which are characterized by gross deletions. Our sequence analysis
indicates that the staggered ends bear a 3⬘ overhang. These 3⬘
overhang ends could then be further processed into blunt ends by
resecting the ssDNA tails. A slight increase in LM-PCR products
following the pretreatment of DNA with T4 DNA polymerase implies that the ends with 3⬘ overhangs precede the blunt ends (Fig.
1, lanes 5 and 6). Schlissel (9) has reported that in normal recombination-active cells, the recovered coding ends bear a 3⬘ overhang
structure. Thus, the polarity of hairpin end nicking is not altered by
the scid mutation. The predominance of blunt opened ends in
scid-ts transformants at 33°C may reflect a higher nuclease activity
in removing 3⬘ overhangs or, more likely, an increased accessibility of the ends to nucleases at this condition. Experiments are
underway to examine these possibilities.
The recent finding of DJ coding joints in DNA-PKcs⫺/⫺ mutant
mice provides direct evidence for the role of a DNA-PKcs-independent pathway in resolving coding ends (16, 37, 40). It is plausible that this mechanism accounts for the rare production of functional recombination joints that promote the development of
oligoclonal B and T cells, known as leaky scid cells. In our scid
ts-Abl-MLV model, the cells cultured at 39°C possibly resemble
the majority of developing scid lymphocytes in vivo, as both
scid-ts cells and scid thymocytes contain unresolved haripin coding ends. On the other hand, the scid-ts cells after returning from
39°C to 33°C may mimic leaky scid lymphocytes, capable of resolving coding ends. Our data further indicate that the scid leaky
recombination can be manipulated by external factors.
The unusually high level of coding joints found in our ts-AbMLV-transformed scid cells might be attributable to the presence
of the bcl-2 transgene and/or the v-abl oncogene in our cell lines.
The expression of a bcl-2 transgene renders these recombinationactive cells resistant to apoptosis. Owing to their sustained life
span, the cells are given the opportunity to eventually resolve their
ends and make coding joints. This scenario was observed in bcl-2
transgenic scid mice (41). It is not quite certain how the temperature-dependent reactivation of v-abl might affect scid coding joint
formation. It has been speculated that indirect events induced by
v-abl reactivation such as cell cycle progression and/or RAG1/2
down-regulation may create conditions that favor the recombination joining activity (37, 42, 43).
Susceptibility of scid leaky recombination to manipulation is
also witnessed in scid thymocytes exposed to ionizing radiation, as
these cells increase both the level and the fidelity of their rearranged TCR coding joints (44). Analogous to the temperaturedependent resolution of coding ends in our scid-ts transformants,
ionizing radiation may somehow facilitate hairpin end opening and
processing in scid thymocytes. This conversion would still be scid-
4141
4142
the Mre11/Rad50/Nbs1 complex is involved in processing recombination intermediates in vivo.
Acknowledgments
We thank S. Bingham for DNA sequencing, and E. Birge and E. Grant for
their assistance in editing this manuscript.
References
22. Gao, Y., Y. Sun, K. M. Frank, P. Dikkes, Y. Fujiwara, K. J. Seidl,
J. M. Sekiguchi, G. A. Rathbun, W. Swat, J. Wang, et al. 1998. A critical role for
DNA end-joining proteins in both lymphogenesis and neurogenesis. Cell 95:891.
23. Grawunder, U., D. Zimmer, S. Fugmann, K. Schwarz, and 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.
24. Taccioli, G. E., G. Rathbu, E. Oltz, T. Stamato, P. A. Jeggo, and F. W. Alt. 1993.
Impairment of V(D)J recombination in double-strand break repair mutants. Science 260:207.
25. Lieber, M. R., J. E. Hessie, S. Lewis, G. C. Bosma, K. Mizuuchi, M. J. Bosma,
and M. Gellert. 1988. The defect in murine severe combined immune deficiency:
joining of signal segments but not coding segments in V(D)J recombination. Cell
55:7.
26. Kulesza, P., and M. R. Lieber. 1998. DNA-PK is essential only for coding joint
formation in V(D)J recombination. Nucleic Acids Res. 26:3944.
27. Hiom, K., and M. Gellert. 1997. A stable RAG1-RAG2-DNA complex that is
active in V(D)J cleavage. Cell 88:65.
28. Bailin, T., X. Mo, and M. J. Sadofsky. 1999. A RAG1 and RAG2 tetramer
complex is active in cleavage in V(D)J recombination. Mol. Cell. Biol. 19:4664.
29. Besmer, E., J. Mansilla-Soto, S. Cassard, D. J. Sawchuk, G. Brown, M. Sadofsky,
S. M. Lewis, M. C. Nussenweig, and P. Cortes. 1998. Hairpin coding end opening
is mediated by RAG1 and RAG2 proteins. Mol. Cell 2:817.
30. Shockett, P. E., and D. G. Schatz. 1999. DNA hairpin opening mediated by the
RAG1 and RAG2 proteins. Mol. Cell. Biol. 19:4195.
31. Melek, M., M. Gellert, and D. C. van Gent. 1998. Rejoining of DNA by the
RAG1 and RAG2 proteins. Science 280:301.
32. Paull, T. T., and M. Gellert. 1999. Nbs1 potentiates ATP-driven DNA unwinding
and endonuclease cleavage by the Mre11/Rad50 complex. Genes Dev. 13:1276.
33. Chang, Y., and M. L. Brown. 1999. Formation of coding joints in V(D)J recombination-inducible severe combined immune deficient pre-B cell lines. Proc.
Natl. Acad. Sci. USA 96:191.
34. Nakajima, P. B., and M. J. Bosma. 1997. Characterization of excised DNA intermediates associated with V(D)J recombination at the T-cell receptor ␦ locus.
Mol. Cell. Biol. 17:2631.
35. Chang, Y., and M. J. Bosma. 1997. Effect of different Ig transgenes on B cell
differentiation in scid mice. Int. Immunol. 9:373.
36. Chang, Y., G. C. Bosma, and M. J. Bosma. 1995. Development of B cells in scid
mice with immunoglobulin transgenes: implications for the control of V(D)J
recombination. Immunity 2:607.
37. Bogue, M. A., C. Jhappan, and D. B. Roth. 1998. Analysis of variable (diversity)
joining recombination in DNA-dependent protein kinase (DNA-PK)-deficient
mice reveals DNA-PK-independent pathways for both signal and coding joint
formation. Proc. Natl. Acad. Sci. USA 95:15559.
38. Chen, Y. Y., L. C. Wang, M. S. Huang, and N. Rosenberg. 1994. An active v-abl
protein tryrosine kinase blocks immunoglobulin light-chain gene rearrangement.
Genes Dev. 8:688.
39. Leu, T. M., Q. M. Eastman, and D. G. Schatz. 1997. Coding joint formation in
a cell-free V(D)J recombination system. Immunity 7:303.
40. Taccioli, G. E., A. G. Amatucci, H. J. Beamish, D. Gell, X. H. Xiang,
M. I. T. Arzayus, A. Priestley, S. P. Jackson, M. Rothstein, P. A. Jeggo, and
V. L. M. Jerrea. 1998. Targeted disruption of the catalytic subunit of the
DNA-PK gene in mice confers severe combined immunodeficiency and radiosensitivity. Immunity 9:355.
41. Strasser, A., A. W. Harris, L. M. Corcoran, and S. Cory. 1994. Bcl-2 expression
promotes B- but not T-lymphoid development in scid mice. Nature 368:457.
42. Chen, Y. Y., and N. Rosenberg. 1992. Lymphoid cells transformed by Abelson
virus require the v-abl protein-tyrosine kinase only during early G1. Proc. Natl.
Acad. Sci. USA 89:6683.
43. Ramsden, D. A., and M. Gellert. 1995. Formation and resolution of double-strand
break intermediates in V(D)J rearrangement. Genes Dev. 9:2409.
44. Danska, J. S., F. Pflumio, C. J. Williams, O. Hunter, J. E. Dick, and C. J. Guidos.
1994. Rescue of T cell-specific V(D)J recombination in scid mice by DNAdamaging agents. Science 266:450.
45. Lista, F., V. Bertness, C. J. Guidos, J. S. Danska, and I. R. Kirsch. 1997. The
absolute number of trans-rearrangements between the TCRG and TCRB loci is
predictive of lymphoma risk: a severe combined immune deficiency (scid) murine
model. Cancer Res. 57:4408.
46. Grawunder, U., and M. R. Lieber. 1997. A complex of RAG-1 and RAG-2 proteins persists on DNA after single-strand cleavage at V(D)J recombination signal
sequences. Nucleic Acids Res. 25:1375.
47. Haber, J. 1998. The many interfaces of Mre11. Cell 95:583.
48. Paull, T., and M. Gellert. 1998. The 3⬘ to 5⬘ exonuclease activity of Mre11
facilitates repair of DNA double-strand breaks. Mol. Cell 1:969.
Downloaded from http://www.jimmunol.org/ by guest on July 31, 2017
1. Gellert, M. 1992. Molecular analysis of V(D)J recombination. Annu. Rev. Genet.
26:425.
2. Lewis, S. M. 1994. The mechanism of V(D)J joining: lessons from molecular,
immunological, and comparative analyses. Adv. Immunol. 56:27.
3. McBlane, J. F., D. C. van Gent, D. A. Ramsden, C. Romeo, C. A. Cuomo,
M. Gellert, and M. A. Oettinger. 1995. Cleavage at a V(D)J recombination signal
requires only RAG1 and RAG2 proteins and occurs in two steps. Cell 83:387.
4. Difilippantonio, M. J., C. J. McManhan, Q. M. Eastman, E. Spanopoulou, and
D. G. Schatz. 1996. RAG1 mediates signal sequence recognition and recruitment
of RAG2 in V(D)J recombination. Cell 87:253.
5. Van Gent, D. C., J. F. McBlane, D. A. Ramsden, M. J. Sadofsky, J. E. Hesse, and
M. Gellert. 1996. Initiation of V(D)J recombination in a cell-free system by
RAG1 and RAG2 proteins. Curr. Top. Microbiol. Immunol. 217:1.
6. Hesse, J. E., M. R. Lieber, M. Gellert, and K. Mizuuchi. 1987. Extrachromosomal
DNA substrates in pre-B cells undergo inversion or deletion at immunoglobulin
V-(D)-J joining signals. Cell 49:775.
7. Roth, D. B., C. Zhu, and M. Gellert. 1993. Characterization of broken DNA
molecules associated with V(D)J recombination. Proc. Natl. Acad. Sci. USA 90:
10788.
8. Zhu, C., and D. B. Roth. 1995. Characterization of coding ends in thymocytes of
scid mice: implications for the mechanism of V(D)J recombination. Immunity
2:101.
9. Schlissel, M. 1998. Structure of nonhairpin coding-end DNA breaks in cells
undergoing V(D)J recombination. Mol. Cell. Biol. 18:2029.
10. Schlissel, M., A. Constantinescu, T. Morrow, M. Baxter, and A. Peng. 1993.
Double-strand signal sequence breaks in V(D)J recombination are blunt, 5⬘-phosphorylated, RAG-dependent, and cell cycle regulated. Genes Dev. 7:2520.
11. Zhu, C., M. A. Bogue, D.-S. Lim, P. Hasty, and D. B. Roth. 1996. Ku86-deficient
mice exhibit severe combined immunodeficiency and defective processing of
V(D)J recombination intermediates. Cell 86:379.
12. Nussenzweig, A., C. H. Chen, V. D. Soares, M. Sanchez, K. Sokol,
M. C. Nussenzweig, and G. C. Li. 1996. Requirement for Ku80 in growth and
immunoglobulin V(D)J recombination. Nature 382:551.
13. Gu, Y. S., K. J. Seidl, G. A. Rathbun, C. M. Zhu, J. P. Manis, N. vanderStoep,
L. Davidson, H. L. Cheng, J. M. Sekiguchi, K. Frank, et al. 1997. Growth retardation and leaky scid phenotype of ku70-deficient mice. Immunity 7:653.
14. Blunt, T., N. J. Finnie, G. E. Taccioli, G. C. Smith, J. Demengeot, T. M. Gottlieb,
R. Mizuta, A. J. Varghese, F. W. Alt, P. A. Jeggo, and S. P. Jackson. 1995.
Defective DNA-dependent protein kinase activity is linked to V(D)J recombination and DNA repair defects associated with the murine scid mutation. Cell 80:
813.
15. Blunt, T., D. Gell, M. Fox, G. E. Taccioli, A. R. Lehmann, S. P. Jackson, and
P. A. Jeggo. 1996. Identification of a nonsense mutation in the carboxyl-terminal
region of DNA-dependent kinase catalytic subunit in the scid mouse. Proc. Natl.
Acad. Sci. USA 93:10285.
16. Gao, Y., J. Chaudhuri, C. Zhu, L. Davidson, and F. W. Alt. 1998. A targeted
DNA-PKcs-null mutation reveals DNA-PK-independent functions for KU in
V(D)J recombination. Immunity 9:367.
17. Li, Z., T. Otevrel, Y. J. Gao, H. L. Cheng, B. Seed, T. D. Stamato, G. E. Taccioli,
and F. W. Alt. 1995. The XRCC4 gene encodes a novel protein involved in DNA
double-strand break repair and V(D)J recombination. Cell 83:1079.
18. Li, Z., and F. W. Alt. 1996. Identification of the XRCC4 gene: complementation
of the DSBR and V(D)J recombination defects of XR-1 cells. Curr. Top. Microbiol. Immunol. 217:143.
19. Kabotyanski, E. B., L. Gomelsky, J. O. Han, T. D. Stamato, and D. B. Roth. 1998.
Double-strand break repair in Ku86- and XRCC4-deficient cells. Nucleic Acids
Res. 26:5333.
20. Grawunder, U., M. Wilm, X. Wu, P. Kulesza, T. E. Wilson, M. Mann, and
M. R. Lieber. 1997. Activity of DNA ligase IV stimulated by complex formation
with XRCC4 protein in mammalian cells. Nature 388:492.
21. Frank, K. M., J. M. Sekiguchi, K. J. Seidl, W. Swat, G. A. Rathbun, H.-L. Cheng,
L. Davidson, L. Kangaloo, and F. W. Alt. 1998. Late embryonic lethality and
impaired VDJ recombination in mice lacking DNA ligase IV. Nature 396:173.
DNA RECOMBINATION