| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Centenary Institute of Cancer Medicine and Cell Biology, Newtown, Australia
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1–10 kb in length and G:C rich, contain many semipalindromic and partially conserved repeats from 25 to 80 bp in length, and are located upstream of each H chain C region gene. Ig somatic mutation and class switching are mechanistically related because they both depend absolutely on the cytidine deaminase encoded by the activation-induced cytidine deaminase (AID) gene (2). AID initiates mutation by directly deaminating cytidine bases (C), causing transitions to uracil (U). Replication of uracils causes G to A transitions in the first round and C to T transitions in the second round (3). In addition, repair mediated by uracil-DNA glycosylases (mainly UNG) creates abasic sites, which lead to transitions and tranversions at G:C base pairs. Deamination by AID also appears to stimulate mismatch repair using error-prone DNA polymerases, which introduce flanking mutations preferentially targeted to A/T base pairs (4, 5, 6, 7, 8). In the case of switching, direct deamination of S region DNA by AID and subsequent base excision by UNG are clearly involved (5), but it is not yet apparent exactly how this leads to recombination rather than merely point mutation. One possible difference between switching and somatic mutation is the role of DNA double-strand breaks (DSBs). DSBs have been detected in the V regions of mutating cells, but only in S and G2 phases of the cell cycle, so they may represent the stalling of replication forks at single-strand nicks. In contrast, DSBs are detectable in S regions of switching cells in G1 phase (L.O., A.J.L.C., and C.J.J., manuscript in preparation) (9, 10, 11). The cause/effect relationship between DSBs and AID activity is not clear (11, 12, 13), but the circularization of DNA excised by switching (14, 15, 16) strongly implicates the involvement of DSBs without proving the case.
Repair of DSBs in eukaryotic cells occurs by homology-directed repair or by nonhomologous end joining (NHEJ) (for reviews, see Refs. 17 and 18). NHEJ joins dsDNA ends by direct abutment (e.g., V(D)J recombination signal joints), or by very small (typically 1 to 3 bp) sequence overlaps (e.g., V(D)J coding joints), thus rendering the process error prone. The NHEJ pathway is initiated by the binding of Ku70/Ku80 as a heterodimer (hereafter referred to as Ku) to broken DNA ends. Analysis of the role of NHEJ in switching is complicated by the fact that NHEJ is essential for V(D)J recombination, but the block in B cell development in NHEJ-deficient mice can be overcome by the introduction of homologously integrated Ig transgenes encoding rearranged Ab sequences. Mature B cells rescued in this manner in Ku70-/- and Ku80-/- mice failed to undergo any detectable switching when stimulated in vivo or in vitro, implying the complete dependence of switching to all isotypes on NHEJ (19, 20, 21).
In vertebrates (but not yeast), NHEJ involves the DNA-dependent protein kinase catalytic subunit (DNA-PKcs) in addition to Ku. DNA-PKcs is an extremely large (470-kDa) serine kinase that is spontaneously mutated in SCID mice (22). Recruitment of DNA-PKcs to DNA-bound Ku leads to the formation of the DNA-PK holoenzyme and induces the kinase activity of DNA-PKcs (23). DNA-PKcs increases both the rate and fidelity of NHEJ (24, 25), possibly by increasing the recruitment of XRCC4 and DNA ligase IV to the broken DNA ends (26), although its role in NHEJ is certainly not limited to this function (23). Alt and colleagues (27) have reported that DNA-PKcs is absolutely required for switching to all isotypes other than IgG1, which surprisingly was still readily detectable. In contrast, Bosma et al. (28) found that switching to all isotypes was affected only a little, or not at all, in DNA-PKcs mutant (SCID) B cells. A confounding factor in the experiments of both Manis et al. and Bosma et al. (27, 28) was that B cells from V(D)J recombination-proficient Ig-transgenic mice were used as controls. Extensive V gene replacement in both the H and L chain loci of the control mice would almost certainly have occurred, producing polyclonal populations of recirculating B cells as a result of receptor editing. Such editing cannot occur in DNA-PKcs-/- or SCID mice, which would have thus maintained monoclonal B cell populations with a specificity conferred by the Ig transgenes used. Thus, the test cells expressing monospecific B cell receptors (BCRs) were compared with polyclonal control populations, which may have obscured the influence of DNA-PKcs deficiency on switching. The need to take the specificity of the BCR into account is illustrated by the fact that it greatly alters the kinetics of both B cell proliferation and switching, independently of exposure to cognate Ag (29). Thus, the exact role of DNA-PKcs in switching remains uncertain.
To examine the role of the SCID mutation in switching in a more definitive way, we elected to use RAG1-/- B cells as controls because they expressed a monoclonal BCR specificity identical with that of the DNA-PKcs-deficient test B cells. Thus, the B cell compartments of both SCID and RAG1-/- mice were rescued by crossing them with Ig transgenic (SWHEL) mice carrying B cells specific for hen egg lysozyme (HEL) (29). As expected, the Ig constructs used to create these mice did not undergo any receptor editing in SCID or RAG1-/- mice. When cells expressing identical BCRs were tested, DNA-PKcs was shown to be involved in the mechanism of switching to all isotypes, and conversely, the SCID mutation did not completely block switching to any isotype. Strikingly, mutation in recombined S regions of base pairs with A, G, or T bases on the sense strand was unaffected by the SCID mutation, whereas mutation of base pairs with cytidines on the sense strand was reduced 7-fold. In addition, the frequency of junction-proximal mutations was reduced, and the degree of identity between S region donor and acceptor DNA ends was slightly increased, relative to control RAG1-/- cells. This suggests that the processing of DNA ends involved in class switching is altered by the SCID defect.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All mice were maintained on a pure C57BL/6 background at the Centenary Institute Animal Facility. SCID (PrkdcSCID/SCID) and RAG1-/-mice on a C57BL/6 background were kindly provided by B. Fazekas (Centenary Institute, Newtown, Australia) (30). The production of SWHEL mice has been described elsewhere (29). SWHEL mice carry rearranged IgH and Ig
transgenes cloned from the HyHEL10 hybridoma specific for HEL. The IgH rearrangement (VH10tar) was targeted to replace the most 3' D element and all of the J elements present in C57BL/6 embryonic stem cells, allowing the transgene to undergo efficient switch recombination to all Ig isotypes (29). The Ig transgene (LC2) was randomly integrated into the genome in low copy number. Crossing of VH10tar and LC2 transgenic mice produced SWHEL mice carrying a large population of B cells specific for HEL. All SWHEL mice used in our experiments were heterozygous for VH10tar and either heterozygous or homozygous for LC2.
Abs and other reagents
The following mAbs to murine Ags were all purchased from BD PharMingen (San Diego, CA): anti-CD40 (HM40-3), anti-IgM biotin (R6-60.2), anti-IgG1 biotin (A85-1), anti-IgG2a biotin (R19-15), anti-IgG2b biotin (R12-3), anti-IgG3 biotin (R40-82), anti-IgE biotin (R35-118), and anti-IgA biotin (C10-1). Streptavidin PerCP and annexin V Cy5 were also purchased from BD PharMingen. Anti-mouse IgM PE (1B4B1), anti-mouse IgD FITC (11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26), and biotin-conjugated goat polyclonal Abs to mouse IgG1, IgG2a, IgG2b, IgG3, and IgA were purchased from Southern Biotechnology Associates (Birmingham, AL). PE- and allophycocyanin-conjugated anti-mouse B220 (CD45R) (RA3-6B2) and anti-CD4 PE (CT-CD4) were purchased from Caltag Laboratories (Burlingame, CA). Anti-mouse CD80 (16-10-H1 (31)) and anti-mouse CD38 (NIMR-R5 (32)) mAbs were used as hybridoma supernatants. HEL was purchased from Sigma-Aldrich (St. Louis, MO) and conjugated to FITC, as previously described (33). Mouse rIFN-
was purchased from Genzyme (Cambridge, MA), and human TGF-
1 from Boehringer Mannheim (Mannheim, Germany). Cell membranes expressing murine CD40L (CD154) were prepared, as previously described (34), from HEK293 cells stably transfected to express mouse CD40L (a generous gift of A. Grech, Centenary Institute). Mouse rIL-4 was kindly provided by P. Hodgkin (Walter and Eliza Hall Institute, Melbourne, Australia). Salmonella typhosa LPS was purchased from Sigma-Aldrich.
Serum ELISAs
Anti-HEL Ab in serum samples was detected by direct ELISA, essentially as described (33). Briefly, 96-well polystyrene plates (Nunc, Roskilde, Denmark) were coated overnight at 4°C with HEL at 10 µg/ml, following which wells were blocked with 100 µl of 1% BSA/PBS and serial dilutions of sera were added. Biotinylated isotype-specific mAbs in 0.1% BSA/1% skim milk powder/PBS were used to detect bound Ab. Streptavidin-alkaline phosphatase (Roche Diagnostics, Mannheim, Germany) was then added and visualized with the substrate p-nitrophenyl phosphate (ICN Biomedicals, Aurora, OH). Absorbance at 405 nm was read.
Isolation and culture of B cells
B cells were purified from RBC-depleted single-cell suspensions of splenocytes using anti-CD19-conjugated MACS beads (Miltenyi Biotec, Gladbach, Germany). If cells were to be labeled with CFSE (Molecular Probes, Eugene, OR), it was done as originally described by Lyons and Parish (35). Cells were cultured with various combinations of cytokines and mitogens in 48- or 24-well flat-bottom plates (Falcon; BD Biosciences, San Jose, CA) at a starting density of 2 x 105 to 106 cells/ml and incubated at 37°C in a humidified atmosphere of 5% CO2 in B cell medium: RPMI 1640 medium supplemented with 2 mM L-glutamine, 0.1 mM nonessential amino acids, 10 mM HEPES (pH 7.4), 100 µg/ml streptomycin, 100 U/ml penicillin, 5 x 10-5 M 2-ME (all from Sigma-Aldrich), 1 mM sodium pyruvate, and 10% heat-inactivated FCS (both from Life Technologies). The concentrations of mitogens and cytokines added at the start of culture were: LPS (10 µg/ml), CD40L (1:100), anti-CD40 (5 µg/ml), IL-4 (500 or 100 U/ml), anti-CD38 supernatant (1:20), and anti-CD80 supernatant (1:20). IFN- (500 ng/ml) and TGF-1 (1 ng/ml) were added to cultures after 24–36 h.
Flow cytometry
For phenotyping, PBLs or purified B cells were incubated with allophycocyanin-, PE-, and FITC-conjugated Abs or FITC-conjugated HEL in PBS containing 0.2% BSA and 0.02% NaN3 (PBA) for 30 min on ice, then washed. Propidium iodide (PI) was added immediately before flow cytometry to enable gating out of dead cells. To enumerate cells present at the end of culture periods, standard numbers of Calibrite beads (BD Biosciences) were added to culture wells immediately before harvest. Cells and beads were then collected into 1-ml tubes in arrays of 96 (Bio-Rad, Hercules, CA); PI was added to 1 µg/ml to label dead cells; and flow cytometry was performed immediately. For switched isotype staining, cells were washed with annexin V buffer (0.1 M HEPES, 0.14 M NaCl, 2.5 mM CaCl2, pH 7.4), then stained with annexin V Cy5 (1:100 in same) for 15 min at room temperature. This step labeled dead and apoptotic cells. After washing with annexin V buffer, cells were fixed in 0.1 ml 2% paraformaldehyde in PBS for 10 min at room temperature. They were then washed twice in PBA and incubated at room temperature in PBA + 1% saponin (Sigma-Aldrich) for 40 min with HEL-FITC (if not previously labeled with CFSE), plus anti-IgM PE, and biotinylated goat anti-IgG1, anti-IgG2a, anti-IgG2b, anti-IgG3, or anti-IgA, or monoclonal anti-IgE biotin. Biotinylated Ab binding was revealed by subsequent incubation with streptavidin PerCP. After each incubation, cells were washed twice with 0.5 ml of PBA. Four-color flow cytometry was performed on a dual-laser FACSCalibur flow cytometer (BD Biosciences) and analyzed with FlowJo version 4.2 software (Tree Star, Stanford, CA). Calibrite beads were counted in the appropriate forward and side scatter gate. Cell division number was determined from the CFSE peaks, and gates were drawn around each peak to allow division slicing (36).
PCR amplification and sequencing of hybrid S regions
DNA was extracted from cultured cells using Wizard genomic DNA purification kits (Promega, Madison, WI). To amplify hybrid S regions, aliquots of DNA were used to prime TripleMaster (Eppendorf, Westbury, NY) long-range hot start PCR consisting of 35 cycles of 93°C for 1 min, 65°C for 0.5 min, and 68°C for 10 min. All reactions contained a sense primer positioned 5' to Sµ (CTTCT AGAAT TCGCT AAACT GAGGT GATTA CTCTG AGG) and an antisense primer positioned 3' to one of the other S regions (S3, TGTGC ATGCT CCATC TGTCC TATCT ACTTG GTTCC; S1, CTTTC TAGAG CTTCC CTATG AGTAG CCAGC ATGCAC; S2b, TCAGG ATCCG AAGCT TTCCA TGAGC AGGTT TGCTA CC; S2a, CCCAG GATCC TGACA TCAGC AACTC CCATG AG; S
, GAAGG ATCCT TGATG CTCTA GAGTG TCCAC ATGCC C; S
, CAAGC AGTGG ATCCA AAGAC GAGAG GACTC). Primer sequences were based on GenBank accessions AC079180, AC079272, and AC110179, and on Celera Discovery System gene entry mCG127237. Reaction products were resolved on 0.8% agarose gels and blotted onto Hybond-N+ membranes (Amersham Biosciences, Little Chalfont, U.K.) by alkaline capillary transfer. Specific PCR products were detected by hybridization with a 32P-labeled oligonucleotide (GCCTA CACTG GACTG TTCT) positioned immediately downstream of the Sµ PCR primer. For sequencing, hybrid S region PCR products were purified using High Pure PCR Product Purification kits (Roche Diagnostics) and digested with EcoRI plus BamHI (Sµ/1, Sµ/), EcoRI plus KpnI (Sµ/2b), or EcoRI plus NlaIV (Sµ/3). Digested PCR products were cloned into pBluescriptII-KS+ (Stratagene, La Jolla, CA), and inserts in plasmids isolated from individual colonies were sequenced by the Australian Genome Research Facility (Brisbane, Australia).
PCR amplification of mRNA transcripts
mRNA was purified from cultured cells using an RNeasy kit (Qiagen, Hilden, Germany) and reverse transcribed (Enhanced Avian RT-PCR kit; Sigma-Aldrich). Aliquots of reverse-transcription reactions (± reverse transcriptase) equivalent to 10 ng of mRNA were subjected to 30 cycles of PCR using either -actin primers (ACG GCC TCA TCA CTA T and CTG CTT GCT GAT CCA CAT C) or AID primers (CAA CAA GTC TGG CTG CCA C and CGT ACA AGG GCA AAA GGA TGC).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
SCID mice (PrkdcSCID/SCID) carry a recessive spontaneous mutation that results in the loss of the last 83 aa of DNA-PKcs (22, 37, 38). This decreases the levels of DNA-PKcs protein in the nucleus to below the limit of detection, reduces cytoplasmic levels 10-fold, and removes all detectable DNA-dependent serine kinase activity (22). To allow analysis of switch recombination in mature B cells from mice in which V(D)J recombination was blocked by the SCID mutation, the B cell compartment was rescued using V(D)J-rearranged H and L chain Ig transgenes. Conventional IgH transgenes cannot undergo normal switching, necessitating the use of a transgene, VH10tar, previously described by Phan et al. (29), which was targeted to the endogenous IgH locus by means of homologous recombination in C57BL/6 embryonic stem cells. The VH10tar transgene encodes the VH10/D/JH3 rearrangement from the HyHEL10 hybridoma specific for HEL. For the L chain, the LC2 transgene was used (29), which carries a V10/J2 rearrangement also cloned previously from the HyHEL10 hybridoma (33), and was randomly integrated into the C57BL/6 genome in low copy number. To create a functional B cell Ag receptor, double-transgenic (SWHEL) mice expressing both the targeted VH10tar gene and the randomly integrated LC2 transgene were produced. SWHEL mice used in our experiments were always heterozygous for the targeted IgH transgene (VH10tar+/-) and either heterozygous or homozygous for the LC2 transgene. A prominent population of B cells in these mice is specific for HEL and can switch efficiently to all Ig isotypes while retaining the ability to bind HEL (29), thereby demonstrating that the VH10tar transgene can undergo normal switch recombination. The remaining B cells in SWHEL mice fail to bind HEL due to V gene replacement in the bone marrow (29).
The role of DNA-PKcs in switching was investigated by crossing SWHEL mice with C57BL/6 SCID (PrkdcSCID/SCID) mice. Because neither the RAG1 nor RAG2 protein is involved in Ig switch recombination (39, 40), we also bred the SWHEL transgenes into congenic RAG1-deficient mice as controls. Given that both V(D)J recombination of TCR and Ig genes and subsequent receptor editing of rearranged IgH genes depend on wild-type RAG1 and SCID proteins (41, 42, 43), SWHELSCID and SWHELRAG1-/- mice lacked mature T cells, as expected. The B cells present in these lines expressed high levels of IgM and displayed uniform (96%+) specificity for HEL (data not shown). This phenotype contrasted with that of wild-type SWHEL mice, in which only one-half of the B cells are HEL specific (as a consequence of V gene replacement; see Ref.29), and which expressed variable levels of IgM (data not shown).
Anti-HEL-specific isotypes present in vivo vary between SWHEL mice on wild-type, RAG1-/-, and SCID backgrounds
As a prelude to screening for isotype switching, ELISAs were performed to detect spontaneous production of HEL-binding Abs in the sera of age- and sex-matched SWHEL, SWHELSCID, and SWHELRAG1-/- mice (Fig. 1). HEL-specific IgM was much more abundant in the serum of SWHEL mice bred on either the SCID or RAG1-/- background than it was in wild-type SWHEL mice. Such a difference may have been due to the fact that the HyHEL10 Id is poorly selected into the B1 compartment of V(D)J recombination-proficient (SWHEL) mice (29), and thus makes only a minor contribution to the total IgM Ab pool. In contrast, the HyHEL10 Id was the only specificity available for natural Ab production in mice deficient for V(D)J recombination.
|
Switching to all isotypes (except IgA) occurred in in vitro stimulated SWHELSCID B cells
To quantify isotype switching in vitro, B cells were purified from spleens of SWHELSCID and SWHELRAG1-/- mice using anti-CD19 MACS beads. Cells from both lines of mice when purified on anti-CD19 beads were 95–98% B220+ve, of which 96–99% bound HEL (data not shown). Equal numbers of purified cells were grown separately for 4–5 days in triplicate or quadruplicate cultures supplemented with the agonists appropriate to induce switching (Fig. 2). At the end of the culture period, cells were harvested, stained with annexin V (to improve exclusion of cells apoptotic or dead at the time of harvest), and were then stained for IgM and switched Ig isotypes. Fixing and permeabilization were used to ensure that all switched cells could be detected, regardless of surface Ig expression. A number of independent experiments were conducted on cells from different pairs of age-matched SWHELSCID and SWHELRAG1-/- mice, although switching to all isotypes was not assessed in every experiment. As shown by the data presented in Fig. 2A, in vitro switching to all isotypes, except IgA, was detected by FACS analysis of SWHELSCID B cells. The specificity of isotype detection was confirmed by staining cells from cultures in which the isotype of interest was likely to be absent or reduced (Fig. 2B). For instance, IgG1 was detected in B cells from cultures supplemented with CD40L + IL-4 (±IFN-), but not in cells from cultures supplemented with LPS or LPS + TGF-.
|
for 5 days, then stained for HEL specificity in addition to IgM and IgA expression. As shown in Fig. 2C, switching to IgA could be induced under these culture conditions, but was only detectable in cells that had altered their BCR specificity by V gene replacement and thus no longer bound HEL. This result clearly underlined the need to exclude receptor-edited cells from experiments and to compare switching only in cells with identical BCRs. The reason for the failure of HEL-binding cells to undergo significant switching to IgA in response to LPS + TGF- in vitro may have been due to the lack of these cells in the splenic B cell subset most amenable to IgA switching. For instance, the IgMlowDhigh subset of splenic B cells, which represents the most mature B cells, does not include anti-HEL B cells (see Ref.29).
Confirmation of the occurrence of switching events in SCID B cells was sought by identifying recombination junctions between Sµ and downstream S regions. This involved PCR amplification of hybrid Sµ/1, Sµ/2b, Sµ/3, Sµ/2a, and Sµ/ junctions using DNA extracted from cultured B cells obtained from SWHELSCID vs SWHELRAG1-/- donors. Junctions confirming switching to each isotype were only detected in DNA samples when they had been extracted from cells cultured under the appropriate conditions (vide supra). For instance, Sµ/1 junctions were detected in cells cultured with CD40L + IL-4, but not in cells cultured with LPS or LPS + TGF- (Fig. 3). Similarly, Sµ/2a junctions were detectable in cells cultured with CD40L + IFN- + IL-4 (Fig. 3), but not in cells cultured with either LPS (Fig. 3), or CD40L + IL-4 (data not shown). Hybrid S region PCR amplification products from both SCID and control cells ranged in size from 300 bp to >8 kbp, with a majority lying between 1 and 4 kbp (Fig. 3). These sizes concur with the potential size range of hybrid S regions in mouse B cells. In agreement with the flow cytometric data, hybrid Sµ/ junctions could not be consistently amplified from cultures of either SWHELRAG1-/- or SWHELSCID splenic B cells (data not shown).
|
The FACS data presented in Fig. 2 not only indicated that switching to most isotypes occurred in SWHELSCID B cells, but suggested that when compared with the level in control SWHELRAG1-/- B cells it was always reduced. This reduction was reproducible in multiple independent experiments (Fig. 4), the difference in switching being significant for IgG1, IgG3, IgG2b, and IgE (p < 0.005, 0.025, 0.05, and 0.005, respectively; paired Student’s t test).
|
Although switching at low levels does occur during the first cell division after activation, efficient switching to all isotypes requires at least three cell divisions in normal B cells (44, 45, 46). The reduction in switching observed in SWHELSCID B cells could therefore have been due to a decrease in proliferation. This possibility was tested by culturing SWHELSCID and SWHELRAG1-/- B cells under switching conditions (CD40L + IL-4) and nonswitching conditions (CD40L + anti-CD38 + anti-CD80). AID message was undetectable by RT-PCR amplification of mRNA extracted from cells cultured under the nonswitching conditions (Fig. 5). The number of viable SWHELSCID B cells remaining at the end of the culture period was always less than the number of viable SWHELRAG1-/- B cells (Fig. 5A), whereas the number of dead and apoptotic (annexin V-staining) cells was always greater (data not shown). Much of the decreased viability of SCID B cells was unrelated to switching because it occurred even in cultures in which proliferation was induced in the absence of switching (Fig. 5, A and C). The decreased viability of cultured SCID B cells raised the possibility that the SCID mutation did in fact reduce switching by reducing the number of cell divisions that B cells underwent in culture. However, CFSE labeling showed that SWHELSCID B cells in both switching (e.g., CD40L + IL-4) and nonswitching (CD40L + anti-CD80 + anti-CD86) cultures progressed through almost identical numbers of cell divisions as did control SWHELRAG1-/- cells (Fig. 5B). Thus, the SCID mutation caused more cells to die per cell division (see supplemental Fig. 1), but had little effect on the number of times that cells divided in culture. When switching was analyzed in terms of division number, the SCID mutation was associated with reduced switching in all cell divisions (Fig. 5C). Taken together, the data showed that reduced switching in SCID B cells is not an artifact of reduced proliferation and that DNA-PKcs is likely to be involved mechanistically in switching.
|
To determine whether the SCID mutation qualitatively altered the mechanism by which switching occurred, hybrid S regions were PCR amplified from cultured B cells and cloned into plasmids, and individual recombination junctions were sequenced. PCR products were cloned with EcoRI (which cleaved in the 5' Sµ primer) plus a restriction enzyme that cut the hybrid S regions at multiple sites in the 3' S region (conferred by the repetitive nature of S regions). This strategy trimmed much of the 3' S region sequence from the hybrid S region PCR products before cloning, but left the recombination junction intact and was devised to minimize preferential sequencing of small PCR products. The precise sites of sequenced recombination junctions were determined by BLAST (47) alignment with the National Center for Biotechnology Information and Celera C57BL/6 mouse genome databases. Sµ/1, Sµ/2b, Sµ/3, and Sµ/ recombination junctions (and two S/ recombination junctions) that were nonredundant (30 from SCID B cells and 28 from control B cells) are illustrated in supplemental Fig. 2. A small increase in base identity in the 8–9 bp centered around the switch recombination junctions was evident in the SCID cells (mean = 3.4) compared with the control cells (mean = 2.3), but this increase was much less than that previously observed in PMS2, MLH1, or ATM mutants (48, 49, 50), and was not significant if only junctions of the same isotype were analyzed.
Decreased mutation in hybrid switch junctions from SCID B cells
Somatic mutations were detected in hybrid S regions (Table I) at a frequency above that predicted to occur by PCR error alone (1 mutation/1000 bp). Like V region mutation, S region mutation is known to be AID dependent and probably occurs by a similar mechanism (11, 51), although the role of mismatch repair in S region mutation appears to be quite different. In Ig V regions, mismatch repair functions to increase mutation, probably because it introduces mutations (preferentially at A:T base pairs) while removing AID-induced G:U mismatches by short patch repair (5, 52, 53, 54), whereas in hybrid S regions mismatch repair instead appears to remove mutations generated by the recombination process (55). The much stronger preference for mutation at G:C base pairs over A:T base pairs in hybrid S regions from control cells observed by us (Table I) and by Honjo and colleagues (51) is consistent with a minor role for mismatch repair in introducing S region mutations. The frequencies of mutation of base pairs in which A, G, or T was on the sense strand were very comparable between hybrid S regions from SCID and RAG1-/- control cells, whereas the frequency of mutation of base pairs in which C was on the sense strand was reduced 7-fold in SCID cells (Table I, after correction for the difference in total number of base pairs in the two databases 
2 = 15.1, p = 0.002). It seems unlikely that the reduced mutation of C in SCID cells shown in Table I is a PCR artifact, because the most common error generated by PCR is A:T to G:C transition (56), which is not overrepresented in the SCID database.
|
2 = 19.1, p = 0.00001). This difference is highly unlikely to be a PCR artifact and suggests that the processing of DNA ends created during switching was quite different in SCID cells compared with control RAG1-/- cells. One explanation for the altered processing is that recruitment of DNA ligase IV by DNA-PKcs protects the DNA ends from nucleases (26). DNA ends are also more processed in SCID cells during VDJ-signal joint formation, and this increased processing is associated with an increase in donor-acceptor microhomology at the joints formed (25). |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
8sd BCR, whereas these variations were eliminated from our experiments by the choice of RAG1-/- mice as controls. Receptor editing may also explain the differences between our data and that of Manis et al. (27), but much less plausibly. Thus, the lack of detectable switching to isotypes other than IgG1 in DNA-PKcs-/- B cells (27) could have been due to a combination of the effects of DNA-PKcs deficiency per se and further suppression of switching inherent in the particular transgenic BCR used. An alternative explanation is that residual DNA-PKcs activity is present in SCID B cells, whereas it is absent in DNA-PKcs-/- cells. Although DNA-dependent protein kinase activity has been undetectable in all SCID cells examined to date (22, 28, 37, 57, 58, 59), the SCID mutation does not remove every trace of DNA-PKcs polypeptide (22, 37, 57, 58, 59). In contrast, most evidence indicates that this residual SCID polypeptide is nonfunctional. For instance, V(D)J recombination, including the nature of the signal joints, is indistinguishable in DNA-PKcs-/- and SCID cells (25, 60, 61, 62), and overexpression of the SCID form of DNA-PKcs in DNA-PKcs-/- cells does not significantly alter this phenotype (25). If the SCID protein does participate in the switch recombination process, isotype switching is the first form of DNA repair involving DNA-PKcs in which its kinase activity is dispensable.
It is improbable that DNA-PKcs plays a direct role in S region point mutation because the SCID defect does not affect mutation of Ig V regions (63). This contrasts with the 7-fold reduction in mutation at base pairs in which C was on the sense strand evident in recombined S regions from SCID cells (Table I). Recombination of S regions depends on both AID and UNG (5), implying that deamination of cytidines in S regions and their subsequent removal by base excision repair are critical events in class switch recombination. Thus, a simple explanation for the reduced frequency of mutation of cytidines shown in Table I is that deamination of cytidines on the sense strand is frequently lethal in SCID B cells, while deamination of cytidines on the antisense strand is usually benign (reflected by the preservation in SCID cells of mutation at G in Table I). This conclusion is consistent with the higher death rate of cultured SCID B cells (Fig. 5), although it has not yet been possible to establish that switching per se contributes to this death rate (see supplemental Fig. 1). Because the most likely trigger of cell death in SCID cells is an unresolved DSB, we speculate that deamination of the sense strand of S regions may usually cause the formation of a DSB, whereas deamination of the antisense strand usually does not. This raises the question of how DSBs arise during switching. Staggered DSBs certainly appear to be involved in switching (10), and although AID prefers to deaminate the nontemplate (sense) strand, it does deaminate the template strand as well. In concert with UNG and apurinic/apyrimidinic endonuclease 1, AID could thus generate staggered nicks (64, 65, 66), but why should AID-initiated nicking of the nontemplate strand produce a DSB more frequently than nicking of the template strand, as we speculate? One possibility is that nicks have already occurred on the template strand independently of AID and are only converted to DSBs that involve DNA-PKcs in the repair process when AID-induced nicking takes place on the nontemplate strand. Perhaps the recruitment of nucleases to S region R loops (64) produces template strand nicking that is independent of AID.
We do not yet know what fraction of G:C mutations in S regions represents direct cytidine deamination sites. Thus, another explanation for the effect of the SCID defect on mutation of hybrid S regions is that absence of DNA-dependent kinase activity alters the error bias of the DNA polymerases involved in repairing S region DSBs. However, it is difficult to envisage how such an alteration in the error bias could produce the strand bias evident in Table I. Clearly, the skewed loss of mutations from hybrid S regions in SCID cells warrants further investigation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: The Children’s Medical Research Institute, 214 Hawkesbury Road, Westmead, Sydney, NSW 2145, Australia.
3 Address correspondence and reprint requests to Dr. Christopher J. Jolly, The Centenary Institute of Cancer Medicine and Cell Biology, Locked Bag 6, Newtown, NSW 2042, Australia. E-mail address: c.jolly{at}centenary.usyd.edu.au
4 Abbreviations used in this paper: S region, switch region; AID, activation-induced cytidine deaminase; BCR, B cell receptor; DNA-PKcs, DNA-dependent protein kinase catalytic subunit; DSB, double-strand break; HEL, hen egg lysozyme; NHEJ, nonhomologous end joining; PI, propidium iodide; UNG, uracil-DNA glycosylase.
Received for publication May 21, 2003. Accepted for publication October 3, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
is an A-T mutator in somatic hypermutation of immunoglobulin variable genes. Nat. Immun. 2:537.[Medline]
plays a major role in Ig and bcl-6 somatic hypermutation. Immunity 14:643.[Medline]
. Nature 419:944.[Medline]
3 DNA-specific double strand breaks are induced in mitogen-activated B cells and are implicated in switch recombination. J. Immunol. 159:4139.[Abstract]
-H2AX focus formation and mutations at sites of class switching. Nature 414:660.[Medline]
heavy chain class switching. Immunity 5:319.[Medline]
This article has been cited by other articles:
|
|
 |
|
  A. Kotnis, L. Du, C. Liu, S. W Popov, and Q. Pan-Hammarstrom Non-homologous end joining in class switch recombination: the beginning of the end Phil Trans R Soc B, March 12, 2009; 364(1517): 653 - 665. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Franco, M. M. Murphy, G. Li, T. Borjeson, C. Boboila, and F. W. Alt DNA-PKcs and Artemis function in the end-joining phase of immunoglobulin heavy chain class switch recombination J. Exp. Med., March 17, 2008; 205(3): 557 - 564. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. Jolly, A. J.L. Cook, and J. P. Manis Fixing DNA breaks during class switch recombination J. Exp. Med., March 17, 2008; 205(3): 509 - 513. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Soulas-Sprauel, G. Le Guyader, P. Rivera-Munoz, V. Abramowski, C. Olivier-Martin, C. Goujet-Zalc, P. Charneau, and J.-P. de Villartay Role for DNA repair factor XRCC4 in immunoglobulin class switch recombination J. Exp. Med., July 9, 2007; 204(7): 1717 - 1727. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Kiefer, J. Oshinsky, J. Kim, P. B. Nakajima, G. C. Bosma, and M. J. Bosma The catalytic subunit of DNA-protein kinase (DNA-PKcs) is not required for Ig class-switch recombination PNAS, February 20, 2007; 104(8): 2843 - 2848. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Paus, T. G. Phan, T. D. Chan, S. Gardam, A. Basten, and R. Brink Antigen recognition strength regulates the choice between extrafollicular plasma cell and germinal center B cell differentiation J. Exp. Med., April 17, 2006; 203(4): 1081 - 1091. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Rooney, F. W. Alt, J. Sekiguchi, and J. P. Manis Artemis-independent functions of DNA-dependent protein kinase in Ig heavy chain class switch recombination and development PNAS, February 15, 2005; 102(7): 2471 - 2475. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. A. Dunnick, J. Shi, K. A. Graves, and J. T. Collins Germline Transcription and Switch Recombination of a Transgene Containing the Entire H Chain Constant Region Locus: Effect of a Mutation in a STAT6 Binding Site in the {gamma}1 Promoter J. Immunol., November 1, 2004; 173(9): 5531 - 5539. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Arudchandran, R. M. Bernstein, and E. E. Max Single-Stranded DNA Breaks Adjacent to Cytosines Occur during Ig Gene Class Switch Recombination J. Immunol., September 1, 2004; 173(5): 3223 - 3229. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Li, S. J. Scherer, D. Ronai, M. D. Iglesias-Ussel, J. U. Peled, P. D. Bardwell, M. Zhuang, K. Lee, A. Martin, W. Edelmann, et al. Examination of Msh6- and Msh3-deficient Mice in Class Switching Reveals Overlapping and Distinct Roles of MutS Homologues in Antibody Diversification J. Exp. Med., July 6, 2004; 200(1): 47 - 59. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Lahdesmaki, A. M. R. Taylor, K. H. Chrzanowska, and Q. Pan-Hammarstrom Delineation of the Role of the Mre11 Complex in Class Switch Recombination J. Biol. Chem., April 16, 2004; 279(16): 16479 - 16487. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Zeng, G. A. Negrete, C. Kasmer, W. W. Yang, and P. J. Gearhart Absence of DNA Polymerase {eta} Reveals Targeting of C Mutations on the Nontranscribed Strand in Immunoglobulin Switch Regions J. Exp. Med., April 5, 2004; 199(7): 917 - 924. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), RAG1-/- (
), and SCID (
) backgrounds. Serum samples were incubated with HEL-coated ELISA plates at a dilution of 1/40 (1/200 for IgM), and bound Ab were detected using biotinylated isotype-specific mAbs and avidin-conjugated HRP. The mean absorbance of triplicates (±SEM) is shown.
, 0.05 < p < 0.01.
) mice were cultured with (-) CD40L + IL-4 or (- - -) CD40L + anti-CD38 + anti-CD80. The fractions of cells in each division that had undergone switching (identified as IgM-ve cells) are plotted (large symbols). Division cohorts were identified using CFSE fluorescence, as shown in B, Data are for triplicate cultures from one typical experiment. D, AID expression in cultured cells. Reverse transcription and PCR were used to detect AID and