| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,§
*
Harold C. Simmons Arthritis Research Center and Departments of Internal Medicine and Microbiology, University of Texas Southwestern Medical Center, Dallas, TX 75235;
Department of Molecular Medicine, Institute for Biotechnology, University of Texas Health Science Center, San Antonio, TX 78245; and
Department of Microbiology and Immunology and
§
Howard Hughes Medical Institute, Baylor College of Medicine, Houston, TX 77030
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In recent years, considerable progress has been made in understanding mechanistic details of the V(D)J recombination reaction. Recombination is initiated by the lymphocyte specific endonuclease, composed of recombination activating gene-1 (RAG-1)3 1 and RAG-2, leaving hairpinned coding end intermediates and blunt phosphorylated signal end intermediates (9, 10, 11, 12, 13, 14, 15, 16, 17). Resolution of recombination intermediates is mediated by ubiquitously expressed factors of the DNA double strand break repair pathway (18, 19). Four of these double strand break repair factors have been defined: Ku70, Ku86, DNA-PKCS, and XRCC4 (20, 21, 22, 23, 24, 25, 26). The heterodimeric DNA end binding factor, Ku, has been shown to direct DNA-PKCS (the SCID factor) to damaged DNA, thus generating a DNA-dependent protein kinase (DNA-PK; Refs. 27, 28). Although Ku and DNA-PKCS are clearly required for V(D)J recombination, there are currently no biochemical data clarifying what their specific role(s) are. The fact that DNA-PK is activated by damaged DNA has led to the hypothesis that DNA-PK may activate mediators of DNA repair via phosphorylation. To date, this model has not been tested, and other functions for DNA-PK (such as a role in scaffolding to facilitate assembly of the repair complex) have also been suggested (reviewed in Refs. 29, 30, 31). We have recently demonstrated that XRCC4 is both a target for and complexes with DNA-PK in vitro, suggesting that DNA-PK and XRCC4 may function in concert in resolving DNA breaks (32). In addition, XRCC4 has been shown to interact with and stimulate DNA ligase IV activity (33, 34). These data suggest a model whereby XRCC4 may target cellular ligase activities to damaged DNA bound by DNA-PK.
It has been appreciated for more than a decade that the resolution of coding ends and signal ends are fundamentally different (35). Signal ends are largely unmodified before ligation, whereas coding joints exhibit both nucleotide loss and gain at the site of joining (2, 35). It has been established that nucleotide loss from coding ends is an intrinsic characteristic of the recombination reaction. The degree of nucleotide loss is largely dependent on the fine structure of the coding ends and is not differentially regulated (7, 8, 36); the exonuclease(s) responsible for coding modification has not been identified (37). In contrast, nucleotide addition is not an intrinsic feature of the recombination reaction in that N segments are not observed in certain coding joints (for example in mice, Ig light chains and fetal and neonatal immune receptors generally lack N segments (7, 38, 39, 40, 41, 42, 43)). This is explained by the regulated expression of TdT, both during development (absent in fetal and neonatal lymphocytes) and during B cell ontogeny (expressed in pro B cells but absent in pre B cells; reviewed in Ref. 44). However, regulated expression of TdT does not explain the relative lack of N segments in signal joints.
Recently, we reported that rare coding joints from
Ku86-/--deficient mice have virtually no N
segments even though TdT mRNA is expressed (45). In fact,
in Ku86-/- animals, the number of
rearrangements with N segments is comparable with that observed in
TdT-/- animals (Table I
; Refs. 46, 47).
Furthermore, a less complete (but still severe) deficit in N segment
addition is observed in coding joints isolated from C.B-17 SCID mice
(48); these animals are defective in
DNA-PKCS expression because of a premature stop
codon in the coding region resulting in a small C-terminal deletion
(49, 50). Although the mutation in
DNA-PKCS in SCID mice deletes only the C-terminal
83 aa, the expressed protein is not stable and
DNA-PKCS levels are severely diminished (0–10%
of wild-type levels, depending on the cell type analyzed)
(50). Together, these findings (summarized in Table I)
suggest an unsuspected role for DNA-PK in addition of N nucleotides to
DNA ends during V(D)J recombination. Thus, in this report we
investigated interactions between DNA-PK and TdT in vitro. We present
the first evidence that TdT can form a stable complex with DNA-PK bound
to DNA.
|
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Sequences of the oligonucleotides used in these studies are as follows: 50 mers: top, 5'-GATCTGGCCTGTCTTACACAGTGCTACAGACTGGAACApAAAACCCTGCAG-3'; bottom, 5'-CTGCAGGGTTTTTGTTCCAGTCTGTAGCACTGTGTAAGACAGGCCAGATC-3'; bottom 5' overhang, 5'-TCCTGCAGGGTTTTTGTTCCAGTCTGTAGCACTGTGTAAGACAGGCCAGA; bottom 3' overhang, 5'-GCAGGGTTTTTGTTCCAGTCTGTAGCACTGTGTAAGACAGGCCAGATCCT. 30 mers: top, 5'-TACGCTGGGAATTCGGGAAAGGATCCGGCC-3'; bottom, 5'-GGCCGGATCCTTTCCCGAATTCCCAGCGTA-3'; bottom 3' overhang, 5'- GGATCCTTTCCCGAATTCCCAGCGTA.
The 50 mers used in Figs. 1, 3, 5, and 6 contain a V(D)J RSS; however,
completely analogous results were obtained with other oligonucleotides
lacking RSS. Single-stranded oligonucleotides were labeled at the 5'
termini with T4 PNK (New England Biolabs; Beverly, MA) and
[
-32P]ATP and were annealed with
complementary oligonucleotides to generate double-stranded
oligonucleotides with blunt termini, or termini with either 5' or 3'
extensions. After annealing, oligonucleotides were purified on native
polyacrylamide gels.
|
|
|
|
EMSAs were done by the method of Landolfi et al. (55) with the following modifications. Binding reactions were done in the following buffer C (10 mM HEPES (pH 7.9), 12.5% glycerol, 210 mM NaCl, and 0.75 mM MgCl2) instead of buffer D, nonspecific competitor was not added, and BSA was included at a concentration of 3 mg/ml. DNA-PK purified from HeLa cells (Promega, Madison, WI) was used in conjunction with either recombinant bovine TdT (Life Technologies, Gaithersburg, MD) or bovine TdT purified from calf thymus (Boehringer Mannheim, Indianapolis, IN). Binding reactions were assembled on ice and then incubated at 30°C for 30 min. Protein/DNA complexes were resolved on 4% native polyacrylamide gels in 0.5x TBE with 2% glycerol.
Purification of DNA-PKCS from Raji cell extracts
Clarified nuclear extract (fraction I), prepared as described
(56, 57) from 10 ml of human Burkitt’s lymphoma cell
pellet (Raji cells; purchased from the National Cell Culture Center,
Minneapolis, MN) obtained from 10 liters culture, was treated with
ammonium sulfate at 0.21 g/ml. The supernatant obtained after
centrifugation (20,000 x g; 30 min) was treated with
additional ammonium sulfate at 0.08 g/ml, and the protein precipitate
was collected by centrifugation (20,000 x g, 30 min),
dissolved in 20 ml buffer T (20 mM Tris-HCl (pH 7.4), 10% glycerol,
0.5 mM EDTA, and 1 mM DTT) with protease inhibitors (2 µg/ml each of
aprotinin, chymostatin, leupeptin, and pepstatin A), and dialyzed for
3 h against buffer T with 50 mM KCl. The dialysate (fraction II)
was applied onto a column of Q-Sepharose (3 ml total) and the bound
proteins were eluted with a 40-ml gradient of 50–330 mM KCl in buffer
T, collecting 40 fractions. The peak of DNA-PKCS
(fraction III, 4.5 ml), eluting at 
200 mM KCl, was identified by
immunoblotting and applied directly onto a 1-ml Macrohydroxyapatite
(Bio-Rad, Richmond, CA) column, which was developed with a 24-ml
gradient of 100–300 mM
KH2PO4 in buffer K (20 mM
KH2PO4, 10% glycerol, 0.5
mM EDTA, 0.5 mM DTT, and 0.01% Nonidet P-40) with 50 mM KCl,
collecting 24 fractions. The DNA-PKCS pool
(fraction IV, 3 ml), which eluted at 200 mM
KH2PO4, was concentrated to
0.5 ml in a Centricon-30 microconcentrator (Amicon, Beverly, MA) and
subjected to gel filtration in a 22 ml Sephacryl S300 column in buffer
K with 100 mM KCl. The S300 pool of DNA-PKCS
(fraction V, 2.4 ml) was fractionated in a MiniS column (Pharmacia,
Piscataway, NJ) with an 8-ml gradient of 50–525 mM KCl in K buffer,
collecting 32 fractions. The DNA-PKCS peak
(275 mM KCl, fraction VI, 1.0 ml) was diluted with two volumes of
buffer K and further fractionated in a MiniQ column (Pharmacia) with a
6-ml gradient of 50–330 mM KCl in buffer K. Nearly homogeneous
DNA-PKCS (fraction VII, 200 mM KCl) was
concentrated to a small volume and stored at -70°C in small
portions.
Purification of recombinant Ku
The Ku86 and Ku70 baculoviruses were the generous gifts of Dr.
J. Donald Capra (Oklahoma Medical Research Foundation, Oklahoma City,
OK) and have been described previously (58). For protein
purification, 3.5 x 108 Sf9 cells which had
been infected with either equal ratios of Ku86 and Ku70 encoding
viruses or wild-type AcMNPV 60 h earlier were collected and washed
with PBS. Cell pellets were frozen and thawed three times in liquid
nitrogen in 25 ml of the following buffer: 20 mM HEPES (pH 7.9), 0.42 M
NaCl, 25% glycerol, 1.5 mM MgCl2, 0.2 mM EDTA, 5 mM
imidazole, and 0.5 mM PMSF. The lysates were centrifuged for 30 min at
18,000 x g. Lysates were incubated for 1.5 h at
4°C with 1 ml Ni-NTA resin (Qiagen, Chatsworth, CA), and the resin
washed three times in the same buffer with the addition of 50 mM
imidazole. Proteins were eluted in the same buffer with 500 mM
imidazole and then dialyze into the starting buffer without imidazole.
Contaminating Sf9 proteins can comprise a significant percentage of the
total protein in crude Ni+ fractions; for this
reason, control Ni+ preparations of wild-type
infected Sf9 cells were also prepared and used as controls in EMSA
assays and TdT assays.
TdT assays
32P-end labeled double-stranded oligonucleotides were used as TdT substrates. When using cobalt as a source of divalent cations, TdT activity was assessed in the following buffer: 100 mM potassium cacodylate (pH 7.2), 1 mM CoCl2, 1 mM dATP, 1 mM dCTP, 1 mM dGTP, 1 mM dTTP, and 1.25 mg/ml BSA. When using magnesium as a source of divalent cations, TdT activity was assessed in the following buffer: 100 mM Na cacodylate (pH 7.5), 1 mM dATP, 1 mM dCTP, 1 mM dGTP, 1 mM dTTP, 1 mM MgCl2, and 3 mg/ml BSA. In some experiments, the nucleotide and/or divalent cation concentrations are altered as indicated in each figure legend. TdT reactions were analyzed on 10% sequencing gels.
Kinase assays
Recombinant murine XRCC4 (1 µg) (32) was
incubated with 300 ng DNA-PK (Promega), with or without 100 ng of
linearized pTZ19R plasmid DNA (Pharmacia Biotech) in a total volume of
20 µl Z buffer (25 mM HEPES/KOH (pH 7.9), 50 mM KCl, 10 mM
MgCl2, 20% glycerol, 0.1% Nonidet P-40, and 1
mM DTT) (22). The mixtures were incubated with 37.5 µCi
[-32P]ATP for 30 min at
30oC, then mixed with an equal volume of 2x
SDS-PAGE buffer and loaded onto a 10% SDS-polyacrylamide gel.
Following electrophoresis, gels were dried and exposed to film.
Wortmannin (Sigma, St. Louis, MO) was included at final concentrations
of 0.05, 0.5, or 5 µM, as indicated.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Interactions between TdT and DNA-PK were first examined using
EMSAs using commercially available complete DNA-PK (i.e., containing Ku
+ DNA-PKCS isolated from HeLa cells), baculovirus expressed Ku, and
baculovirus expressed bovine TdT (Fig. 1
A). One distinct species is
apparent when recombinant Ku is incubated with the radiolabeled 30 mer
corresponding to Ku + DNA (lane 2), whereas an
additional weaker species is apparent when purified DNA-PK is incubated
with the probe (corresponding to Ku + DNA-PKCS +
DNA, lane 7). Under the conditions employed in Fig. 1A, stable TdT + DNA complexes are not observed
(lane 6). However, TdT has been shown to stably
complex with DNA in vitro (53, 54), but the binding of TdT
to DNA was shown to be cooperative, and these interactions are best
demonstrated with large DNA fragments (the probe in Fig. 1 is only 30
bp). Thus, under the conditions employed, the interactions of TdT with
DNA do not interfere with the assessment of the potential interactions
of TdT with DNA-PK. In the presence of 1.5-, 3-, and 6-fold molar
excess of recombinant TdT (lanes 8–10), the Ku +
DNA-PKCS + DNA species is completely converted to
a slower migrating form, suggesting that TdT can complex with complete
DNA-PK bound to DNA. This slower migrating form is not observed when
the probe is incubated with DNA-PK and heat-inactivated TdT
(lane 11). In the presence of TdT, the complex
containing Ku + DNA-PKCS retards significantly
more probe and at higher TdT concentrations, the Ku alone complex is
diminished suggesting that the TdT dependent complex is more stable
than Ku + DNA-PKCS alone. The fact that the
DNA-PK complex is completely converted to a slower migrating form with
only 1.5-fold molar excess of recombinant TdT suggests that TdT (as
opposed to contaminants in the TdT preparation) is complexing with
complete DNA-PK bound to DNA ends. Indeed, silver staining of this
preparation demonstrates that the recombinant TdT preparations are
extremely pure (lanes 12 and 13)
containing only the 58-kDa and 42-kDa forms of TdT.
In contrast, the commercial preparation of DNA-PK is only 70% pure;
silver staining of this preparation demonstrates several contaminants
(Fig. 1B, lane 13). To rule out the possibility
that the DNA-protein complexes observed contain proteins other than
DNA-PKCS, Ku + TdT, DNA-PK-TdT interactions were
reassessed using recombinant Ku, and DNA-PKCS
purified from HeLa cells to near homogeneity (as ascertained by silver
staining, lane 14). Because the Ni+
agarose fractions of recombinant Ku are not completely pure
(lane 15) a Ni+ agarose
fraction of Sf9 cells infected with wild-type AcMNPV was used as a
control (lanes 5–7). As can be seen in Fig. 1B, coincubation of 30 ng recombinant Ku with a radiolabeled
30 mer generates a single predominate retarded species representing Ku
+ DNA and a minor species which represents two Ku heterodimers bound to
the DNA (lane 1). The addition, of increasing amounts
of highly purified DNA-PKCS (60, 120, or 240 ng,
lanes 2–4) generates a slower migrating complex which
exactly comigrates with the slower migrating complex observed using
preparations of complete DNA-PK (lane 11). The
Ku + DNA protein complex is the predominate complex observed when
recombinant TdT is coincubated with Ku + DNA (lane
8). However, when recombinant TdT is coincubated with highly
purified DNA-PKCS and recombinant Ku
(lane 9) or complete DNA-PK (lane
12), the slower migrating complex representing Ku + DNA +
DNA-PKCS is completely converted to a slower
migrating form. In lanes 4, 8, and 9,
a minor complex which migrates slightly slower than Ku + DNA alone but
faster than complete DNA-PK + DNA or the 2x Ku complex is observed
suggesting a possible weak interaction between TdT and Ku. As in Fig. 1A, in the presence of TdT, complexes containing complete
DNA-PK (lanes 9 and 12) retard
significantly more probe than in the absence of TdT; in addition, in
the presence of TdT, the Ku alone complex is diminished (in lane
9 there is a 2.5 molar excess of DNA-PKCS
over Ku). This finding suggests that the TdT + Ku +
DNA-PKCS + DNA complex is more stable than
DNA-PKCS + Ku + DNA alone. As discussed above,
stable TdT + DNA complexes are not observed under these conditions
(lane 7). Although DNA-PKCS has
been shown to bind DNA in the absence of Ku, this only occurs in
conditions employing less than 100 mM NaCl (59). In this
experiment (with 200 mM NaCl), when highly purified
DNA-PKCS is incubated with radiolabeled 30 mer
alone, no DNA-protein complexes are observed (lane
5). However, when the probe is coincubated with both recombinant
TdT and highly purified DNA-PKCS, a weak complex
is consistently observed which migrates similarly to the Ku +
DNA-PKCS complex, suggesting that TdT and
DNA-PKCS can bind DNA weakly in the absence of Ku
(lane 6). Finally, we also assessed TdT interactions
with DNA-PK by EMSA using low salt concentrations and the same buffer
used to assay TdT catalysis (25 mM sodium cacodylate; see below) with
completely analogous results (data not shown). In sum, these data
demonstrate that TdT interacts with complete DNA-PK bound to DNA ends
and suggest that DNA-PK may play a role in targeting TdT to DNA ends
during V(D)J recombination.
We next assessed DNA-PK interactions with TdT isolated from calf
thymus. As can be seen in Fig. 1C, when calf thymus TdT is
coincubated with complete DNA-PK (lanes 6 and
7), the slower migrating complex representing Ku + DNA +
DNA-PKCS is completely converted to an even
slower migrating form completely analogous to what is observed when
recombinant TdT is coincubated with complete DNA-PK and DNA. However,
an additional complex is also observed which migrates just slower than
the Ku + DNA complex, suggesting an interaction between Ku and TdT.
Similarly, when calf thymus TdT is coincubated with recombinant Ku, an
additional complex is observed which migrates more slowly than the Ku +
DNA complex (lane 3). TdT isolated from normal cells
is extremely heterogeneous consisting of 58-, 44-, 42-, and 32-kDa
forms that result from proteolysis of the full-length (58-kDa) form. It
is not entirely clear whether these catalytically active degradation
products are physiologically relevant. Many commercially available
preparations of TdT from calf thymus contain only the smaller
proteolytic fragments of TdT. The impurity of the calf thymus
preparations which contain predominately the 44-kDa and 32-kDa forms
(lane 8) precludes concluding that Ku alone interacts
with TdT. However, numerous coprecipitation experiments conducted using
untagged and His-tagged forms of both Ku and full-length TdT expressed
in Sf9 cells consistently failed to demonstrate interactions between
full-length TdT and Ku (data not shown). The question of whether Ku
interacts with proteolytic fragments of TdT is under investigation.
DNA-PK inhibits TdT catalysis
In vitro TdT catalyzes template-independent addition of long
extensions (up to several kilobases; Ref. 51) to the 3'
ends of DNA. The properties exhibited by purified TdT in vitro do not
correlate well with the characteristics of N regions produced in vivo
(44). However, in vitro, the extent of modification varies
dramatically under different conditions; for instance, the addition of
cobalt ions has been demonstrated to result in higher TdT activity than
magnesium or manganese ions. (The physiologic divalent cofactor for TdT
is not known.) The relative concentrations of divalent cation and free
nucleotides have also been reported to alter TdT catalysis
(51). In vivo, the concentration of divalent cations
exceeds that of free nucleotide (in lymphocytes, free nucleotide
concentrations have been estimated to be 30–90 µM,
(60)). We investigated the effect of altering the ratio of
nucleotide and divalent cation concentrations to determine whether
under more physiologic nucleotide concentrations, TdT catalysis more
closely mimicked authentic N segment addition (Fig. 2A). Conditions normally used
in TdT assays include 1 mM cobalt and 1 mM dNTPS. Under these
conditions, extremely long TdT additions are observed
(lanes 2 and 10). When the concentration
of divalent cation is increased (so the concentration of cations
exceeds nucleotide concentration, lanes 3–5) no major
effect on catalysis is observed. In contrast, when the nucleotide
concentration is decreased to more physiologic levels, the average
length of TdT additions is actually markedly enhanced
(lanes 11–13); whereas increasing free nucleotide
levels above 1 mM strongly inhibits activity (lanes
6–8). Thus, it seems possible that the activity of TdT in vivo
might be in some way regulated, and we next examined the effect of
DNA-PK on the polymerase activity of TdT. Because of the extreme
variability of TdT catalysis in vitro, we performed these
experiments under a variety of conditions. In the presence of cobalt,
TdT catalyzes the addition of several hundred nucleotides (Fig. 2B, lane 2), whereas in the presence of
magnesium, TdT catalyzes the addition of 1–20 nt to the
oligonucleotide substrate (Fig. 2B, lane 8).
However, regardless of the divalent cation used, virtually all of the
substrate is modified to some extent by TdT. Under both experimental
conditions, the activity of TdT is altered by DNA-PK in two ways: less
of the substrate is modified and shorter nucleotide extensions are
observed (only 1–2 nt are added in the presence of DNA-PK and
magnesium (Fig. 2B, lane 9); and 1–20 nt are
added in the presence of DNA-PK and cobalt (Fig. 2B,
lane 3)]. Furthermore, the same results were obtained using
either recombinant TdT or TdT purified from calf thymus. As expected,
heat inactivation ablates the inhibition of TdT activity of DNA-PK
(Fig. 2B, lanes 4 and 10).
Furthermore, TdT catalysis was not altered by the addition of several
other DNA binding proteins (histone H1, topoisomerase I, data not
shown). To more formally demonstrate that DNA-PK is responsible for the
observed inhibition of TdT activity, we also assessed TdT activity in
the presence of DNA-PK which had been pre-incubated with recombinant
caspase-3. (It has been demonstrated previously that
DNA-PKCS is specifically cleaved by caspase 3
during apoptosis (61, 62, 63, 64).) Thus, DNA-PK was preincubated
with or without recombinant caspase-3; DNA-PKCS
cleavage was confirmed by immunoblot analysis (data not shown). As can
be seen in Fig. 2C, the inhibition of TdT catalysis for
DNA-PK is substantially relieved by proteolytic cleavage of the
catalytic subunit (compare lanes 3 and 4). Thus,
we conclude that DNA-PKCS limits TdT
catalysis in vitro. We next assessed the effect of both Ku and
DNA-PKCS alone on TdT catalysis. As can be seen
in Fig. 2D, TdT catalysis is not affected by 30 ng
recombinant Ku (2x molar excess, lane 4) as compared
with control Ni+ fractions of wild-type AcMNPV
infected Sf9 cells (lane 3). When increasing amounts
of highly purified DNA-PKCS (60 ng, equal molar
ratio; 120 ng, 2x molar ratio; 240 ng, 4x molar ratio; lanes
5–7) as well as Ku are included in the reaction, TdT catalysis is
increasingly inhibited; the inhibition observed is roughly analogous to
that observed when increasing concentrations of complete DNA-PK are
included (lanes 11–13). Surprisingly, comparable
inhibition is obtained with highly purified
DNA-PKCS alone (lanes 8–10).
Thus, we conclude that the catalytic subunit of DNA-PK inhibits TdT
catalysis in vitro; the net result of this inhibition is that TdT
extensions catalyzed in the presence of DNA-PK more closely resemble N
segments in authentic immune receptors.
|
It is well established that N segment addition is generally
limited to coding joints though short N segment additions are
occasionally observed in signal joints (65, 66). If DNA-PK
preferentially directs TdT to coding ends, this presents a paradox in
that both coding and signal end resolution require DNA-PK (18, 67, 68, 69, 70, 71). In fact, both Ku and DNA-PKCS
have been shown to complex with both coding and signal recombination
intermediates in vitro (72). How then is TdT excluded from
modifying signal ends bound by DNA-PK? One possibility is that in vivo
blunt signal ends are poor substrates for TdT. Thus, we considered that
DNA-PK bound to DNA probes with different termini might interact
differently with TdT, and we compared the capacity of TdT to interact
with DNA-PK bound to probes with 5' or 3' extensions and blunt termini.
As can be seen in Fig. 3, when complete
DNA-PK is coincubated with TdT and the DNA probe, virtually all of the
DNA-PK + DNA complexes are shifted to a slower migrating form
regardless of the structure of the DNA terminus. Thus, we conclude that
TdT can interact with DNA-PK complexed to a variety of DNA end
structures.
It has been reported that TdT prefers DNA termini with 3' or 5'
extensions over blunt DNA ends though these preferences are somewhat
controversial (51, 73, 74, 75). Thus, we next assessed DNA-PK
inhibition of TdT activity using oligonucleotide substrates with
different DNA termini. In these experiments, a single oligonucleotide
was labeled and then annealed to two different complementary
oligonucleotides to generate ends with either blunt or 3' extensions;
TdT activity was subsequently assessed using these oligonucleotides in
the presence of either cobalt or magnesium (Fig. 4). In both situations, the blunt DNA
oligonucleotide is a poorer substrate for TdT than the oligonucleotide
with the 3' extension, although considerable nucleotide addition to the
blunt substrate is observed in the cobalt containing reactions.
Furthermore, in the presence of DNA-PK, nucleotide addition by TdT is
markedly inhibited for both substrates. However, in the presence of
DNA-PK in the cobalt containing reactions (compare lanes 11
and 12), the relatively efficient (as compared with
magnesium containing reactions) nucleotide addition to the blunt
substrate is virtually completely ablated, whereas nucleotide addition
to the 3' overhang substrate still occurs to a considerable degree
(compare lanes 8 and 9). Furthermore, in the
presence of DNA-PK, the lengths of TdT catalyzed additions to the 3'
overhang substrate are dramatically shorter and more closely resemble
authentic N segments. In sum, the inefficiency with which TdT catalyzes
nucleotide additions to blunt vs overhang termini in the presence of
DNA-PK is observed 1) with oligonucleotide substrates of different
lengths and unrelated sequence, 2) in the presence of either magnesium
or cobalt, and 3) when comparing DNA termini with either 3' or 5'
extensions to blunt termini. These data support previous conclusions
that blunt termini are the poorest templates for TdT additions;
furthermore, this bias against blunt substrates is accentuated in the
presence of DNA-PK. Thus, although other explanations cannot be
excluded, the blunt termini of signal end intermediates may contribute
to the relative lack of N segment addition to signal joints during
V(D)J recombination.
|
It is well appreciated that TdT demonstrates certain nucleotide
preferences (51, 74); still, in vitro, dATP, dCTP, dGTP,
and dTTP are all added relatively efficiently. In vivo, N segment
addition has a pronounced bias for GC over AT additions (the GC:AT
ratios range from 2 to 3:1, (38, 41). Thus, we next
assessed the capacity of DNA-PK to inhibit the addition of TdT of only
dATP, dCTP, dGTP, or dTTP (Fig. 5). In
the presence of dATP and dTTP and magnesium, TdT catalyzes very long
extensions to the oligonucleotide (Fig. 5, lanes 4 and
10); furthermore, virtually all of the substrate is
modified. These long nucleotide extensions were observed with a variety
of different oligonucleotide substrates (data not shown). In the
presence of dCTP and dGTP, although all of the substrate is modified,
the extensions terminate at distinct lengths (Fig. 5, lanes
6 and 8), possibly because of secondary structures
generated by the C or G homopolymer tails. This was also observed with
a variety of different oligonucleotide substrates. In the presence of
DNA-PK, addition of all four nucleotides is inhibited; however,
addition of dGTP is consistently less inhibited than the other
nucleotides. There is more substrate modified and longer extensions in
the presence of dGTP (Fig. 5, lane 9) than dATP, dCTP, or
dTTP (Fig. 5, lanes 5, 7, and 11). Although this
difference in inhibition of dGTP is consistent with a variety of
different oligonucleotide substrates, it is not seen in TdT reactions
done in the presence of cobalt (data not shown). Thus, under certain
experimental conditions, in the presence of DNA-PK, TdT polymerizes
dGTP more efficiently than dATP, dCTP, or dTTP. This provides a
potential explanation for the relatively high G/C content in Ig and TCR
N segments.
Kinase activity is not required for the interaction of DNA-PK and TdT or for the inhibition of TdT activity
We considered that the regulation of the activity of TdT by DNA-PK
could be due to phosphorylation of TdT; in fact, it has been
established that TdT is phosphorylated in vivo (51, 52).
Thus, we also assessed TdT as a DNA-PK substrate. However, we found no
evidence of TdT phosphorylation by DNA-PK in vitro (data not shown).
This was true in assessing phosphorylation of calf thymus TdT (which
contains the 42- and 32-kDa forms) or recombinant TdT (which contains
the 58- and 42-kDa forms). To more formally address whether the DNA-PK
kinase activity contributes to its regulation of TdT activity, we next
analyzed the effect of wortmannin on the capacity of DNA-PK to complex
with TdT and to modulate TdT catalysis. Wortmannin inhibits members of
the PI3K family by covalently (and irreversibly) binding to the
catalytic site of the kinase (76). Although it has been
demonstrated that wortmannin inhibits the kinase activity of DNA-PK
(20), the effects of wortmannin on DNA-PK assembly have
not been examined. Thus, we initially assessed whether concentrations
of wortmannin which have previously been shown to ablate DNA-PK
activity altered DNA-PK assembly in EMSA analyses. As can be seen,
wortmannin concentrations as high as 5 µM had no demonstrable effect
on the ability of either Ku or complete DNA-PK to complex to the DNA
probe (Fig. 6B). In contrast
(as shown previously (32), DNA-dependent phosphorylation
of recombinant murine XRCC4 (Fig. 6A) is partially inhibited
at wortmannin concentrations as low as 0.05 µM and completely
inhibited by 0.5 µM wortmannin. As with DNA-PK assembly, wortmannin
concentrations of 5 µM had no affect on the ability of TdT to complex
with DNA-PK (Fig. 6C, compare lanes 1 and
3). Finally, the ability of DNA-PK to inhibit the extent of
TdT catalysis in vitro was also not altered by the addition of
wortmannin (Fig. 6D, lanes 6 and 7). Thus, we
conclude that modulation of TdT activity by the DNA-dependent protein
kinase is not mediated through the kinase activity of DNA-PK.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The fact that TdT interacts with DNA-bound DNA-PK is consistent with the possibility that DNA-PK targets TdT to V(D)J recombination intermediates and suggests an explanation for the deficits in N segment addition observed in Ku-/- and C.B-17 SCID mice. Recent work has shown that V(D)J recombination intermediates are present in a stable complex containing the RAG proteins after cleavage (72). We have previously suggested that Ku and DNA-PK might be involved in remodeling these postcleavage complexes, removing the RAG proteins, and/or recruiting end processing and joining factors such as ligases and hairpin opening enzymes (77). Furthermore, in the absence of DNA-PK, it has been proposed that inefficient resolution of coding joints is mediated by an alternative DNA end joining pathway (77). The data presented here strongly support a role for DNA-PK in targeting processing factors such as TdT to coding end intermediates. In addition, this model predicts that the alternative DNA end joining pathway which inefficiently resolves coding ends in Ku-/- and SCID mice excludes TdT.
There is evidence that SCID rearrangements are aberrantly processed in that unusually long P segments are often observed. Another potential explanation for the relative diminution of rearrangements containing N segments in SCID cells is that the unusually long P segments in some way preclude N addition. However, the fact that rearrangements from Ku86 deficient mice do not have excessive P nucleotide additions, but also lack N segments, argues against this possibility.
We also considered an alternative (and not mutually exclusive) explanation for the lack of N segments in Ku86-/- and SCID mice: that N nucleotides might be added but fail to appear in the junctions (45). Here we have shown that in vitro DNA-PK substantially limits the length of N segments added by TdT. Thus, even if TdT could access recombination intermediates in the absence of DNA-PK, our data suggest that unregulated additions of extremely long N segments would occur that might block the joining reaction.
TdT catalyzes template-independent addition of long extensions (up to several kilobases, (51) to the 3' ends of DNA in vitro; furthermore, virtually all 3' ends (i.e., single stranded or double stranded with either blunt ends or termini with 5' or 3' extensions) can be modified. The actions of TdT are dramatically different in vivo (44) in that the average N segment length is only 2–5 bp (depending on which immune receptor rearrangements are analyzed (38, 41, 44), and the sequence composition is markedly biased toward G:C pairs. The nonprocessive nature of TdT synthesis may in part explain the difference between TdT additions in vitro and in vivo. In vitro, under optimal conditions, the catalysis of TdT is unimpeded and can result in very long additions. In vivo, for each nucleotide added, it seems likely that TdT must compete for free coding ends with repair factors; our data suggest that DNA-PK may be an important factor in regulating the access of TdT to coding ends during V(D)J recombination. In sum, although other explanations cannot be excluded, the fact that DNA-PK modulates both the length and content of TdT additions provides a plausible explanation for the dramatic differences observed between the products of TdT catalysis in vivo vs in vitro.
In addition, although TdT is often expressed during signal end resolution, N segments are rarely observed in signal joints. Here we demonstrate that in the presence of DNA-PK, the polymerase activity of TdT on blunt end substrates (as opposed to substrates with 3' or 5' extensions) is virtually completely inhibited. This finding provides one explanation for the low level of N additions in signal joints. Furthermore, in rearrangements derived from cells that express TdT, N segments are not found in all joints; for example, only 70–80% of Ig VH-DH and DH-JH joints from adult mice have N segment additions (38, 41). The significant proportion of coding joints lacking N segments could result from the nicking of hairpin coding ends to generate blunt termini, which are not efficiently modified by TdT.
The notion that DNA-PK may act to recruit joining factors is consistent with our recent data (32) showing that the ability of the XRCC4 protein to interact with DNA-PK correlates with its ability to complement the double-strand break repair defects in XRCC4-deficient cells. However, although XRCC4 was efficiently phosphorylated by DNA-PK, phosphorylation did not correlate with the functional activity of XRCC4. Similarly, the data presented here demonstrate that the kinase activity of DNA-PK is not essential for its capacity to either complex with TdT or modulate TdT activity. Together, these data suggest that an important aspect of the role of DNA-PK in V(D)J recombination may be to direct end joining and processing factors, such as XRCC4 and TdT, to the DNA ends.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Katheryn Meek, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75235-8884. E-mail address:
3 Abbreviations used in this paper; RAG, recombination activating gene; RSS, recombination signal sequence; DNA-PK, DNA-dependent protein kinase; DNA-PKCS, DNA-PK catalytic subunit; XRCC, x-ray cross complementation group.
Received for publication November 9, 1998. Accepted for publication May 10, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T cell lineages and for a novel intermediate of V-(D)-J joining. Cell 59:859.[Medline]
ß T cells in neonatal mice. EMBO J. 10:3647.[Medline]
locus of mouse thymocytes. Mol. Cell. Biol. 14:4455.: structure-activity analysis and interactions with wortmannin. Biochem. J. 324:489.
This article has been cited by other articles:
|
|
 |
|
  J. M. Murray, J. P. O'Neill, T. Messier, J. Rivers, V. E. Walker, B. McGonagle, L. Trombley, L. G. Cowell, G. Kelsoe, F. McBlane, et al. V(D)J Recombinase-Mediated Processing of Coding Junctions at Cryptic Recombination Signal Sequences in Peripheral T Cells during Human Development J. Immunol., October 15, 2006; 177(8): 5393 - 5404. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Fujisaki, A. Sato, T. Toyomoto, T. Hayano, M. Sugai, T. Kubota, and O. Koiwai Direct binding of TReP-132 with TdT results in reduction of TdT activity Genes Cells, January 1, 2006; 11(1): 47 - 57. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. E. Repasky, E. Corbett, C. Boboila, and D. G. Schatz Mutational Analysis of Terminal Deoxynucleotidyltransferase- Mediated N-Nucleotide Addition in V(D)J Recombination J. Immunol., May 1, 2004; 172(9): 5478 - 5488. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Sandor, M. L. Calicchio, R. G. Sargent, D. B. Roth, and J. H. Wilson Distinct requirements for Ku in N nucleotide addition at V(D)J- and non-V(D)J-generated double-strand breaks Nucleic Acids Res., March 26, 2004; 32(6): 1866 - 1873. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Smith, E. Riballo, B. Kysela, C. Baldeyron, K. Manolis, C. Masson, M. R. Lieber, D. Papadopoulo, and P. Jeggo Impact of DNA ligase IV on the fidelity of end joining in human cells Nucleic Acids Res., April 15, 2003; 31(8): 2157 - 2167. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. N. Mahajan, S. A. Nick McElhinny, B. S. Mitchell, and D. A. Ramsden Association of DNA Polymerase {micro} (pol {micro}) with Ku and Ligase IV: Role for pol {micro} in End-Joining Double-Strand Break Repair Mol. Cell. Biol., July 15, 2002; 22(14): 5194 - 5202. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-B. Duvauchelle, L. Blanco, R. P. P. Fuchs, and A. M. Cordonnier Human DNA polymerase mu (Pol {micro}) exhibits an unusual replication slippage ability at AAF lesion Nucleic Acids Res., May 1, 2002; 30(9): 2061 - 2067. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. Purugganan, S. Shah, J. F. Kearney, and D. B. Roth Ku80 is required for addition of N nucleotides to V(D)J recombination junctions by terminal deoxynucleotidyl transferase Nucleic Acids Res., April 1, 2001; 29(7): 1638 - 1646. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. L. Benedict, S. Gilfillan, and J. F. Kearney The Long Isoform of Terminal Deoxynucleotidyl Transferase Enters the Nucleus And, Rather than Catalyzing Nontemplated Nucleotide Addition, Modulates the Catalytic Activity of the Short Isoform J. Exp. Med., January 1, 2001; 193(1): 89 - 100. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Bemark, J. E. Sale, H.-J. Kim, C. Berek, R. A. Cosgrove, and M. S. Neuberger Somatic Hypermutation in the Absence of DNA-Dependent Protein Kinase Catalytic Subunit (DNA-Pkcs) or Recombination-Activating Gene (Rag)1 Activity J. Exp. Med., November 20, 2000; 192(10): 1509 - 1514. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Tuaillon and J. D. Capra Evidence That Terminal Deoxynucleotidyltransferase Expression Plays a Role in Ig Heavy Chain Gene Segment Utilization J. Immunol., June 15, 2000; 164(12): 6387 - 6397. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
