TY - JOUR
AU - Honjo, Tasuku
AB - Abstract In order to understand the specificity of sequences or structures recognized by a recombinase involved in class switch recombination (CSR), we examined the relative CSR efficiency of various switch sequences in artificial CSR constructs that undergo CSR in CH12F3-2 murine B lymphoma line. Since CSR recombination is not specific to switch regions of different isotypes or orientation of S sequences, we examined the efficiency of S sequences of non-mammalian species and artificial sequences which lack several characters of mammal switch sequences: chicken Sμ, Xenopus Sμ, telomere, multiple cloning site (MCS) and unrelated negative control sequence. CSR occurred in chicken Sμ and MCS with significantly higher efficiency than the negative control. A common character of these two sequences is that they are rich in palindrome and stem–loop structures. However, telomeres, which are G-rich and repetitive but not palindromic, could not serve as switch sequences at all. The AT-rich Xenopus Sμ sequence was inefficient but capable of CSR. CSR breakpoint distribution suggests that the cleavage may take place preferentially in the proximity of the junctions (neck) between the loop and stem in the secondary structure of the single-stranded S sequence, which can be formed by palindromic sequences. The results suggest that the secondary structure of S-region sequences which is transiently formed during transcription may be necessary for recognition by class switch recombinase. artificial switch constructs, B lymphoma cells, breakpoint sequence, S sequence, secondary structure AID activation-induced cytidine deaminase, CSR class switch recombination, EC extracellular portion, gμ chicken Sμ, GFP green fluorescent protein, GANC gancyclovir, HyTK hygromycin phosphotransferase–thymidine kinase fusion protein, HYG hygromycin, MCS multiple cloning site, NS Non-S, Post-Tr postswitch transcript from CSR substrate, Pre-Tr preswitch transcript from CSR substrate, TGF transforming growth factor, TM transmembrane, TEL telomere sequence, SHM somatic hypermutation, SS splice acceptor signal, Xμ Xenopus Sμ Introduction Three types of genetic alteration are involved in amplification of diversity of the Ig loci. (i) The V(D)J recombination assembles V, D and J segments into a variable region (V) exon by the RAG-1 and RAG-2 proteins, and contributes to diversification of the antigen-binding sites during differentiation of lymphocytes (1). (ii) Somatic hypermutation (SHM) introduces massive mutations in the rearranged V region exon upon antigen stimulation and generates a large number of progeny clones which are selected by high-affinity binding to given antigens (2). (iii) Class switch recombination (CSR) converts the constant region of the heavy chain (CH) from Cμ to CH of other classes or isotypes such as Cγ, Cα and Cε. Ig of different classes have distinct physiological functions that differ in their properties of polymerization, complement fixation, binding to class-specific Fc receptors, and site of secretion and deposition. Since class switching maintains the variable region (antigen binding site) unchanged, it is a mechanism to diversify Ig effector functions with the same antigen specificity (3,4). CSR introduces DNA breaks in the switch (S) region located upstream of each CH gene except for the Cδ gene. S regions of human and mice are composed of tandem repeats of G-rich motifs which contain variants of pentameric motifs of GAGCT and GGG(G/C)T, and are similar but distinct among different isotypes. Cleavage of the S region of Cμ (Sμ) and one of the downstream S regions (Sγ, Sε or Sα) is followed by circularization of the cleaved inter-S segment containing the Cμ gene, and juxtaposition of the V exon and the downstream CH exon. Although recent reports indicate that activation-induced cytidine deaminase (AID) plays a critical role in both SHM and CSR, the molecular mechanisms for SHM and CSR are still unknown (5,6). In particular, it is totally unknown how S regions are recognized by class switch recombinase (endonuclease) whose activity is probably regulated by AID. To address this issue, we have developed an artificial CSR system by combining a DNA construct containing the Sμ and Sα sequences, and CH12F3-2 B lymphoma line which almost exclusively switches to IgA by cytokine stimulation (7). Successful replacement of the Sα sequence with S segments of other isotypes in the construct indicates that CSR endonuclease in CH12F3-2 cells can recognize all of the murine Sγ, Sε and Sα sequences in spite of almost exclusive class switching to IgA in the endogenous locus. In addition, CSR endonuclease is indifferent to the orientation of S sequences. These observations suggest that CSR endonuclease may recognize not the primary sequence of S regions, but rather the secondary structure formed by S sequences. To explore this possibility, we have carried out more detailed analyses of S-region specificity using this artificial CSR substrate. We report here that murine class switch recombinase can recognize distantly related S sequences of other species such as chicken (C-rich plus G-rich) (8) and frog (AT-rich) (9), albeit less efficiently, but not vertebrate telomere sequences which contain repetitive sequences of G-rich motifs similar to mammalian S sequences. Class switch recombinase also recognized artificial sequences composed of a tandem array of restriction sites containing many short palindromes. These observations together with breakpoint distribution analyses led us to conclude that class switch recombinase may recognize a secondary structure of S regions, which is dependent on repeats of palindromes but not on repeats of simple G-rich sequences. Methods Constructs The extracellular portion (EC) and flanking introns of mouse CD8α (10) were amplified by PCR from genomic DNA of CH12F3-2 cells using the following primers: 5′-CCT AGA GCC CTA GCT TGA CCT AAG CTG C-3′ and 5′-TTG TCG ACC CCA GGC TAT CTG CTT ATC-3′. The amplified EC portion of CD8α was ligated to a splice acceptor signal (SS) from an expression vector, pcDL˙SRα (11), in pGEM-T vector (Promega, Madison, WI). The EC-SS fragment was then integrated into pEF-BOS (12), an expression vector with a strong promoter from the human EF-1α gene. From the resulting plasmid, a fragment containing the entire transcription unit was excised as BOS˙EC˙SS. The CD8α transmembrane (TM) exon with the 5′ intron was amplified by genomic PCR using the primers: 5′-TGG TCG ACT CCT GAT CTT GGA GGG AGA C-3′ and 5′-CTG GAT CCC TGT GGT AGC AGA TGA GAG TGA-3′. A green fluorescent protein (GFP)-coding sequence was amplified by PCR with the primers; 5′-CGG GAT CCA CCG GTC GCC ACC GTG GTG-3′ and 5′-GTC GCG GCC GCT TTA CTT GTA CAG CTC GTC CAT GC-3′ using pEGFP-N1 (Clontech, Palo Alto, CA) as a template. This PCR was designed to introduce a replacement of the first methionine of GFP with valine to eliminate expression of GFP without CSR. The mutated GFP sequence was cloned in pGEM-T vector and the amplified CD8α TM exon was inserted in-frame upstream of the GFP sequence. The TM–GFP sequence was excised and inserted with modified pcDL˙SRα devoid of the SS. We confirmed that transfection of COS7 cells with this plasmid resulted in transcription of TM–GFP. This plasmid was digested with HindIII and SalI to generate a fragment containing a whole transcription unit as SRα˙TM˙GFP. The BOS˙EC˙SS fragment and the SRα˙TM˙GFP fragment were connected on a plasmid with a neomycin-resistance gene cassette, SMC-11 (7), which is derived from pMC1neopolyA (Stratagene, La Jolla, CA). The recombinant contained the two transcription units in the same direction and was designated as SMC-24. A fragment containing the CMV promoter-driven hygromycin phosphotransferase–thymidine kinase fusion protein (HyTK) gene was isolated from ptgCMV/HyTK (13) as CMV˙HyTK. CMV˙HyTK was inserted between the two transcription units of SMC-24 to generate SCG(0,0). Two copies of FSα-2, a 1.2-kb fragment of the mouse Sα core sequence (7), were inserted into the intron of the SRα transcription unit of SCG(0,0) to give rise to SCG(0,2α). A single copy of FSμ-4, a 1.3-kb fragment of the mouse Sμ core sequence (7), was further inserted into the intron between the EC and SS exons to generate SCG(μ,2α) as shown in Fig. 1(a). The entire chicken Sμ region was amplified by PCR using the primers 5′-flanked with the SalI site (5′-ATG GTC GAC GCG GCG CTA ATT AAG GCG GTT AAT GAA GGT C-3′; 5′-ATG GTC GAC CGT CGT GGG ATG GAC TGG GAT GGA CTG-3′) from plasmid pE13-A containing the 13.5-kb EcoRI fragment of the chicken Sμ region (14). After SalI digestion, PCR products were ligated into the SalI site of SCG(0,2α) to generate SCG(gμ,2α) as shown in Fig. 1(b). The entire Xenopus Sμ region was amplified by PCR using the primers 5′-flanked with the SalI site (5′-ATG GTC GAC TTT TTG CCA AAA TGA GAT TGA GAT TGC TTT AAC-3′; 5′-ATG GTC GAC TGA CAA GAT GAG CAC TTC ACA GAA TGA ATT TA-3′) from genomic DNA of Xenopus laevis (9). After SalI digestion, the 3.1-kb fragment obtained was inserted into the SalI site of SCG(0,2α) to generate SCG(Xμ,2α). Specific amplification of the chicken and the Xenopus Sμ sequences was confirmed by restriction mapping and sequencing. To generate the telomere sequence, PCR was performed by using the primers with telomere repeats complimentary to each other, i.e. 5′-(GGGTTA)20-3′ and 5′-(CCCTAA)20-3′, without a template (15,16). A 120-bp fragment composed of nine repeats of the telomere motif was cloned into the SmaI site of pBSKS (Stratagene). Two copies of this fragment were concatemerized after SmaI excision, ligated with a ClaI-linker at both ends and inserted into SCG(μ,0) to generate SCG(μ,TEL). A multiple cloning site of pBluescript KS II (Stratagene) was isolated by BssHII digestion, and six copies of the fragment (170 bp) without the ClaI and SalI site were ligated into pGEM-T easy (Promega). After SalI and SphI digestion and blunting, two tandem copies of the blunted fragments were ligated into SCG(0,2α) to generate SCG(MCS,2α). As a negative control (NS), the 1.1-kb EcoRI–BamHI fragment of intron 1 of the mouse PD-1 gene (17) was isolated. After addition of the SalI site to both ends, the NS fragment was inserted into the SalI site of SCG(0,2α) to generate SCG(NS,2α). In order to increase stability of constructs, high-copy origins containing the 1792-bp Bst1107I–ScaI fragments of SCG(gμ,2α), SCG(Xμ,2α), SCG(MCS,2α) and SCG(NS,2α) were replaced with a low-copy origin containing the 1600-bp Bst1107I–ScaI fragment of pBR322. Cell culture and transfection The CH12F3-2 cell line was maintained as described (18). To induce CSR, cells were stimulated with transforming growth factor-β1 and IL-4 as described (18), and with monoclonal anti-mouse CD40 antibody, HM40-3 (a gift from Dr Yagita). Linearized DNA constructs were introduced into CH12F3-2 cells by electroporation. Stable transfectants were selected with G418. Single-copy transfectants confirmed by genomic Southern blotting were used for analyses. Flow cytometry Cells were stained with phycoerythrin-conjugated goat anti-mouse IgA polyclonal antibodies (Southern Biotechnology Associates, Birmingham, AL) and allophycocyanin-conjugated anti-mouse CD8α mAb 53-6.7 (rat IgG2a,κ; PharMingen San Diego, CA), and analyzed by FACSCalibur and CellQuest software (Becton Dickinson, Mountain View, CA). RT-PCR Total RNA was isolated using Trizol reagent (Gibco/BRL, Gaithersburg, MD). First-strand cDNA was synthesized from 1 μg of total RNA with SuperScript II reverse transcriptase (Gibco/BRL). From this cDNA, preswitch transcripts (Pre-Tr1 and Pre-Tr2) were amplified by 30 cycles of PCR using the primer pairs (BF, 5′-GGT TTG CCG CCA GAA CAC AG-3′; BR, 5′-GAT TTC TTG TCT CCC ACG T-3′) and (SF, 5′-CTC GAG GAA CTG AAA AAC CAG AAA G-3′; SR, 5′-GTG GTT TGT CCA AAC TCA TCA ATG T-3′), respectively (Fig. 1a). Postswitch transcript (Post-Tr) was amplified with the primers BF and SR. GAPDH and HPRT transcripts were amplified with the primer pairs (GF, 5′-ACC ACA GTC CAT GCC ATC AC-3′; GR, 5′-TCC ACC ACC CTG TTG CTG TA-3′) and (HF, 5′-CTC GAA GTG TTG GAT ACA GG-3′; HR, 5′-TGG CCT ATA GGC TCA TAG TG-3′) respectively. Determination of breakpoint sequences CD8α–GFP+ cells in stimulated SCG(gμ,2α) and SCG(MCS,2α) transfectants were sorted with anti-CD8α antibody-coupled magnetic beads (Miltenyi Biotec, Bergisch Gladbach, Germany) and those of stimulated SCG(Xμ,2α) transfectants isolated by limiting dilution. CSR breakpoints in isolated CD8α–GFP+ clones were recovered by genomic PCR using the primers (5′-TAA ATG CGG GCC AAG ATC TGC ACA CTG GTA TTT C-3′; 5′-TGA ACA GCT CCT CGC CCT TGC TCA CCA C-3′). The PCR products were directly sequenced using either upstream or downstream primer (5′-GCT CAG GTT AGG TGC TCT CA-3′ or 5′-CGT CTC CCG GTC CAG GTC TC-3′ respectively). Computer analysis for secondary structures in S sequences Folding of the single-stranded DNA in S sequences was analyzed by using the DNA folding program of Michael Zuker (Rensselaer Polytechnic Institute): http://mfold2.wustl.edu/~mfold/dna/form1.cgi. The folding conditions used were: temperature, 37°C; concentrations of [Na+] and [Mg2+], 150 and 0.5 mM respectively. The most stable structure of the ~100–200-base region surrounding the breakpoints which has the lowest free energy of duplex formation (▵G), was chosen in each switch sequence. Lengths and locations of folding sequences are described in Fig. 6. Results Development of FACS-detectable CSR substrate We modified the CSR substrate reported previously (7) so that CSR can be evaluated easily and quantitatively. The new substrate contains the genomic sequence of the EC exons of mouse CD8α, and its TM exon fused with the GFP gene in the upstream (Pre-Tr1) and downstream (Pre-Tr2) transcription units respectively (Fig. 1a), although the basic features of the prototype substrate are conserved: (i) the presence of two S sequences, (ii) constitutive transcription of the S sequences and (iii) splicing that removes S sequences from the transcripts. Deletion of a segment between two S sequences would generate a new transcript (Post-Tr), in which CD8α EC is spliced to the TM domain and anchored to the membrane. Between the two transcription units, the HyTK gene is inserted in order to select cells that underwent recombination in the substrate (Fig. 1a). SCG(μ,2α) harboring the murine Sμ and Sα sequences in the first and second transcription units respectively was stably introduced in CH12F3-2 cells and eight single-copy transfectants were selected by genomic Southern blotting (data not shown). Table 1 shows FACS and RT-PCR analyses of these transfectants after stimulation with cytokines and CD40 ligand. Such analyses of representative clones are shown in Fig. 2. #204 cells express no CD8α before stimulation. Three-day stimulation with IL-4, TGF-β1 and the anti-CD40 mAb(HM40-3) made 5.35% of the population CD8α+ cells, suggesting the occurrence of CSR in the substrate (Fig. 2a). To confirm deletion of the HyTK gene in CD8α+ cells, the stimulated cells were further cultured in the presence of either hygromycin (HYG) or gancyclovir (GANC). Surviving cells selected with HYG contained no CD8α+ cells, while the vast majority of GANC-selected cells expressed the CD8α–GFP fusion protein. This results strongly indicate deletion of the HyTK gene in CD8α+ cells. When RNAs transcribed from the substrate in clones #204, #209 and #212 were analyzed by RT-PCR, non-stimulated cells expressed preswitch transcripts Pre-Tr1 and Pre-Tr2, suggesting that the S regions are accessible. Cytokine stimulation induced expression of CD8α–GFP fusion transcripts (Post-Tr) in all three clones (Fig. 2b). Recombination breakpoints in the S region of substrate were identified by genomic PCR and genomic Southern blotting after cloning of CD8α+ cells (data not shown). When the Sμ sequence of this substrate was replaced by the non-repetitive Non-S (NS) sequence [(SCG(NS,2α)], induction of neither CD8α+ cells nor Post-Tr was observed, in agreement with the previous report (7) that two S sequences are essential for CSR (Fig. 2c and d, and Table 1). Nevertheless, the expression levels of Pre-Tr1 and Pre-Tr2 in SCG(NS,2α) were comparable to those in SCG(μ,2α) (Fig. 2e). The frequency of the CD8α–GFP+ cells before and after stimulation in the lines carrying SCG(NS,2α) was <0.01% (Table 1), indicating a very low background of the FACS assay. These analyses indicate that the SCG(μ,2α) construct serves as a good CSR substrate while the SCG(NS,2α) is a good negative control. Non-mammalian S sequences are recognized by mouse CSR machinery We examined whether non-mammalian S sequences from chicken and frog can be recognized by murine CSR recombinase (Table 1). The chicken Sμ sequence amplified by genomic PCR was inserted to replace the murine Sμ sequence of SCG(μ,2α) to generate SCG(gμ,2α) (Fig. 1b). This substrate was stably introduced in CH12F3-2 cells and 14 single-copy transfectants were selected after Southern blot analyses of genomic DNA (data not shown). Five out of 14 clones exhibited the appearance of CD8α+ cells and postswitch transcripts after 3-day stimulation (Table 2). FACS analysis of a representative clone #619, and RT-PCR analysis of three representative clones #608, #619 and #621 are shown in Fig. 3. CD8α+ cells were enriched from stimulated cells using anti-CD8α antibody-conjugated magnetic beads (Fig. 3a). Clones of CD8α+ cells were isolated by limiting dilution of the enriched population and their DNAs were analyzed by genomic PCR. The PCR products were of variable sizes but within a range expected from deletion between the S regions (9.5–2.0 kb) (Fig. 3c). In addition, recombination breakpoints were identified by sequencing these PCR products directly. Breakpoints were found not only in the G-rich part (Sμ1) but also in the C-rich part (Sμ2) (Fig. 4). These results indicate that the chicken Sμ sequence is recognized by murine class switch recombinase. Next, we examined whether the frog Sμ sequence can function in CH12F3-2 cells. We prepared a construct SCG(Xμ,2α), in which Xenopus Sμ replaced murine Sμ of SCG(μ,2α) (Fig. 1b). This substrate was stably introduced in CH12F3-2 cells and 14 single-copy transfectants were analyzed (Table 2). In spite of extensive analysis of 14 transfectants, only one clone #676 showed a low but significant level of CD8α+ cells after stimulation (Fig. 3a). RT-PCR analysis demonstrated induction of Post-Tr (Fig. 3b), and comparable levels of Pre-Tr1 and Pre-Tr2 expression (Fig. 3d). DNA analysis of CD8α+ clones from #676 demonstrated deletion of the inter-S segment and recombination breakpoints within the Xenopus S sequence (Figs 3c and 4). Thus the frog Sμ sequence appears to function in murine B cells, albeit much less efficiently as compared with the murine or chicken Sμ sequence. CSR does not depend on G-rich repeat sequences Mammalian S sequences rich in stretches of three or four guanines have been known to form in vitro a G-quartet structure, which is a parallel or antiparallel quadraplex of DNA stabilized by Hoogsteen bonding between guanine bases (19,20). The same structure is predicted to be formed by vertebrate telomere sequences composed of repeats of a hexamer unit TTAGGG (Table 2). To examine whether the telomere sequence can substitute for the S sequence, we generated a CSR substrate SCG(μ,TEL) containing the telomere sequence in place of the Sα sequence. None of nine single-copy CH12F3-2 transfectants with this substrate expressed CD8α on the surface or Post-Tr upon stimulation (Fig. 5a and b, and Table 2). Pre-Tr1 and Pre-Tr2 were comparably expressed in these transfectants as assessed by RT-PCR in the non-saturating PCR condition (Fig. 5d), indicating that the absence of CSR is not due to a decreased transcription of the target sequences (Figs 2e, 3d and 5d). It is therefore concluded that the telomere sequence cannot be recognized by class switch recombinase. Consequently, repeats of G-rich sequences are not sufficient as a target of CSR. Multiple restriction enzyme sites can be a target of CSR One feature of the S sequence conserved among mammals, chicken and frog is the presence of short inverted repeats or palindromes in their repetitive units (8,9). It has been suggested that the stem–loop structure predicted from those palindromic sequences may play some roles for recognition by recombinase (9,21). The MCS of a conventional plasmid vector pBluescript KS II (Stratagene) contains such palindrome-rich sequences. Twelve tandem copies of the 170-bp BssHII fragment containing MCS were inserted in place of Sμ of SCG(μ,2α) to generate SCG(MCS,2α) (Fig. 1b). Twelve single-copy transfectants of this construct were isolated and stimulated with cytokines. Four out of 12 transfectants that expressed comparable amounts of Pre-Tr showed stimulation-dependent appearance of CD8α–GFP+ cells by FACS and expression of CD8α–GFP transcripts by RT-PCR (Table 1, and Fig. 5a and b). CD8α+ clones isolated after magnetic beads sorting were shown to contain rearranged SCG(MCS,2α) as revealed by genomic PCR (Fig. 5c). Recombination breakpoints of CD8α+ clones were identified within the MCS sequence (Fig. 4). These results indicate that repeat sequences containing short palindromes can be a target of CSR. Breakpoint distribution on the secondary structure of single-stranded S sequences To confirm the importance of the palindromic structure for recognition by CSR recombinase, we mapped the recombination breakpoints obtained in the present study (Fig. 4) on the predicted secondary structure of the single-stranded S sequences (Fig. 6). An ~100–200-base sequence surrounding each breakpoint was analyzed. Overall, 23 of 26 breakpoints were located within 1 base from the junction (neck) between the single-stranded loop and double-stranded stem (Fig. 6 and Table 3). The other three breakpoints (one in mouse Sα and two in Xenopus Sμ) were located 3 or 8 bases away from the junction but in the single-stranded loop (Fig. 6 and Table 3). Since Xenopus Sμ had a low frequency of CSR, we might have selected an unusual population of switched clones, which may reflect the different distribution profile of Xenopus Sμ breakpoints on the secondary structure. These results suggest that some secondary structures are most likely a recognition target for CSR cleavage enzyme. Discussion We described here that the murine CSR machinery can target the chicken Sμ sequence and the frog Sμ sequence, albeit less efficiently. A previous study showing that transgenic mice with the human IgH locus can undergo CSR indicates the functional compatibility of human S sequences in mouse (22,23). This conclusion is not surprising because S sequences are highly conserved between human and mouse. On the other hand, the chicken and frog S sequences clearly diverge from G-rich mammalian S sequences: chicken Sμ consists of C-rich and G-rich regions (8), and frog Sμ of AT-rich regions (9), although some evolutional link is found between motifs of these S sequences and mammalian homologues (8). Together with our previous observation that CSR recombinase does not distinguish either the primary sequence of various murine S regions or their orientation (7), these results further strengthen our hypothesis that class switch recombinase may not recognize the DNA sequence per se. We are therefore forced to consider the possibility that CSR recombination recognizes the three-dimensional structure formed by the S region when it is actively transcribed. Then, what types of three-dimensional structure can be formed by the transcribed S sequence? Many stretches of three or four consecutive guanines found in mammalian S sequences are known to form tetraplex DNA called G-quartet or G4-DNA. (19,20). Since the same structure is implicated in vertebrate telomere sequences, we examined whether CSR can occur in the substrate containing the murine telomere sequence and demonstrated that the telomere sequence is not active for CSR, suggesting that the G-quartet structure may not be recognized as a target of CSR. Another possible structure is a stem and a loop structure formed by palindromes or inverted repeats in S sequences. The stem–loop structure has been implicated for essential functions of many RNAs such as transfer RNA, ribosomal RNA, ribozymes and RNA editing (24,25). Although less well known, the functional implication of stem–loop DNA is also suggested for the termination signal of transcription (26). The fact that palindromic motifs are universally present in the repeat units of S regions of mammals, birds and amphibians in spite of non-conservation of their primary sequences (Table 2) suggests that the stem–loop structure might play an important role in CSR (21). It is reported that CSR breakpoints are often located in the vicinity of the predicted single-stranded bulge or transition between the stem and loop, i.e. neck regions, in the single-stranded S sequences (9). This hypothesis is further supported by the present study showing that the artificial sequence containing many restriction enzyme sites or MCS was found to function as a recombination target. Furthermore, comparison of breakpoint distributions and predicted secondary structures of the S sequences revealed that 23 of 26 junctions in murine Sα, chicken Sμ, Xenopus Sμ and MCS were located in the proximity of the border nucleotide (neck) between the loop and stem regions (Fig. 6 and Table 3). Although the fraction of the neck regions in the total sequences of chicken Sμ, Xenopus Sμ, MCS and murine Sα is 63%, 88% of the breakpoints were located in the neck region, which is significantly higher than expected from a random distribution (P < 0.01). We have determined about 63 breakpoints in another type of CSR construct of murine S and found that all of them are located at the loop or neck structure (X. C. Chen et al., unpublished data). This conclusion is also consistent with our previous observation that the inverted S sequence is recognized by class switch recombinase (7) because inversion of a palindromic motif would give almost the same sequence and structure. The poor efficiency of CSR in a substrate with the Xenopus Sμ sequence might be explained by a low stability of the secondary structure formed by the AT-rich sequence in murine cells cultured at a higher temperature (37°C) than normal for cold-blooded vertebrates. How can the stem structure of S-region DNA be formed to be recognized by CSR? It is well established that CSR requires transcription of target S regions. We speculate that transcription transiently renders S-region DNA to become single-stranded, and to form the stem–loop structure on both template and non-template strands. Since positions of stem–loop formation are not always the same on both strands of DNA—endonucleolytic cleavages can occur at different positions of two DNA strands, resulting in the formation of staggered nicks on DNA. It is expected that repair joining of staggered-nicked DNA ends results in deletion or duplication with mutations in the proximity of recombination junctions, which are in fact frequently observed in CSR (27 and X. C. Chen et al., unpublished data). Cytokine stimulation of B cells has at least dual facets—activation of germline transcription and induction of AID (5,6). In AID-deficient mice and HIGM2 patients caused by AID mutations, not only CSR but also SHM is defective. Since the molecular mechanisms of CSR and SHM are not delineated, it is not clear why a single mutation in the AID gene affects two different DNA modifications involving S and V regions. However, there are several common features in CSR and SHM, and one possible but not exclusive link to connect these two events may be a common endonuclease that introduces DNA breaks in S and V regions because there is emerging evidence that DNA breaks are involved in SHM (28–31). Secondly, it has been reported that SHM also depends on transcription of the target V region (32–34). Thirdly, the array of restriction sites inserted in a V region was a hot spot of SHM when tested in a transgenic Ig locus (35) in parallel with our experiment using MCS as a target of CSR in the artificial substrate (Fig. 5). Furthermore, many V-region sequences can form a stem–loop structure as revealed by computer analysis (35). Based on these observations, it is tempting to speculate that AID activates a stem–loop-specific endonuclease that cleaves DNA at S and V regions when they are actively transcribed. This hypothesis can explain why the same endonuclease can recognize totally different sequences of V and S regions. Since palindromic targets appear less frequently in the V regions than S regions, the frequency of DNA cleavage may be less in the V regions than S regions, giving rise to apparently contrasting outcomes: mutations in the former and deletions in the latter, although deletions can also take place in V regions (28,29). It is important to point out that this putative enzyme potentially cleaves non-Ig DNA sequences which can form a stem–loop structure during active transcription. This may explain occasional chromosomal translocation involving Ig and non-Ig loci such as c-myc and hypermutation in the human bcl-6 gene in the progenies of germinal center B cells, in which both CSR and SHM are active (41,42). Table 1. Summary of target sequences used in artificial switch constructs Target  Length (kb)  Character  Short repeat unit  Master sequencea  GenBank accession no.  References  aMaster sequences and their derivatives are repeated in target sequences. Underline indicates the short repeat unit.  Murine Sμ(μ)  1.3  G-rich palindromic  CTGAG, CTGGG, GTGAG  GAG CTGAGCTGGGGTGAG CT (20 bp)  J00442-2  7,8,38  Sα(α)  1.2    CTGAG, CTGGG  ATGAGCTGGGATGAG CTGAG CT AGGCTGGAATAGG CTGGGCTGGG CTGGTGTGAG CTGGG TTAGG CTGAGCTGAG CTGGA (80 bp)  D11468  7,8,39    3.0  no repeats non-palindromic    part of intron1 in the mouse PD-1 gene    17  Chicken Sμ(gμ)  5.2              3.8 (Sμ1)  G-rich (Sμ1) palindromic  ACCAG, TATGG  GGGGGTYGCCGGGAGCTGTACT GGTTYKTACTGGTGCYGGT ACCAGTATGGACCAGTATGG (101 bp)  AB029075  8    1.4 (Sμ2)  C-rich (Sμ2) palindromic  CCCAG, TACAG  (CCCAGTACAG)10      Xenopus Sμ(Xμ)  3.1  AT-rich palindromic  AGCT, GTAC, CTNAG, GAATT  AAGCTCAGCT TGGTCTGGA AGCT ATGCACT GAATT ATGCTGTATAGG ACAC AGCT TGCTCAGGACTGCA AGCT ACAGTATGAAAT GAATT ATTCG GAATAATG CTTAGCT TGGTTTGGA AGCAATATCT GAATT AATTCTGTA TA (145 bp)      Telomere (TEL)  2.2  G-rich non-palindromic  TTAGGG  (TTAGGG)n  U84151  36,37  Multiple cloning site (MCS)  2.3  palindromic    12 concatemers of 170 bp multiple cloning site of BSIIKS  X52327  35  Target  Length (kb)  Character  Short repeat unit  Master sequencea  GenBank accession no.  References  aMaster sequences and their derivatives are repeated in target sequences. Underline indicates the short repeat unit.  Murine Sμ(μ)  1.3  G-rich palindromic  CTGAG, CTGGG, GTGAG  GAG CTGAGCTGGGGTGAG CT (20 bp)  J00442-2  7,8,38  Sα(α)  1.2    CTGAG, CTGGG  ATGAGCTGGGATGAG CTGAG CT AGGCTGGAATAGG CTGGGCTGGG CTGGTGTGAG CTGGG TTAGG CTGAGCTGAG CTGGA (80 bp)  D11468  7,8,39    3.0  no repeats non-palindromic    part of intron1 in the mouse PD-1 gene    17  Chicken Sμ(gμ)  5.2              3.8 (Sμ1)  G-rich (Sμ1) palindromic  ACCAG, TATGG  GGGGGTYGCCGGGAGCTGTACT GGTTYKTACTGGTGCYGGT ACCAGTATGGACCAGTATGG (101 bp)  AB029075  8    1.4 (Sμ2)  C-rich (Sμ2) palindromic  CCCAG, TACAG  (CCCAGTACAG)10      Xenopus Sμ(Xμ)  3.1  AT-rich palindromic  AGCT, GTAC, CTNAG, GAATT  AAGCTCAGCT TGGTCTGGA AGCT ATGCACT GAATT ATGCTGTATAGG ACAC AGCT TGCTCAGGACTGCA AGCT ACAGTATGAAAT GAATT ATTCG GAATAATG CTTAGCT TGGTTTGGA AGCAATATCT GAATT AATTCTGTA TA (145 bp)      Telomere (TEL)  2.2  G-rich non-palindromic  TTAGGG  (TTAGGG)n  U84151  36,37  Multiple cloning site (MCS)  2.3  palindromic    12 concatemers of 170 bp multiple cloning site of BSIIKS  X52327  35  View Large Table 2. CSR efficiency of various S-like sequences   Transcripta  CSR frequency    Pre-Tr1  Pre-Tr2  Post-Tr  CD8α–GFP+ (%)c  IgA (%) (endogenous CSR)c            Averaged  Induction folde    Averagef  aPre-Tr1, 2 and Post-Tr were detected by RT-PCR (+, expressed; –, not expressed).  bNumbers in parentheses indicate the number of transfectants analyzed by FACS before or after cytokine stimulation.  cAverages of the difference between stimulation + and stimulation – of two independent experimental values are shown. Average percentages of background (stimulation –) in CD8α–GFP+ and IgA+ was 0.01 and 0.20 respectively. The IgA percentage of each clone at the left is indicated at the corresponding position.  dAverages among clones which express Post-Tr.  eInduction index was obtained by division of CD8 α–GFP+ average percentages by the background (0.01).  fAverages among clones which express IgA.  CG (μ,2α)[8]b                  #204  +  +  +  8.38  2.52 ± 2.67  252  74.7  59.8 ± 15.5  #207  +  +  +  4.02      78.9    #209  +  +  +  3.70      71.4    #212  +  +  +  2.63      63.0    #221  +  +  +  0.72      63.6    #222,#216  +  +  +  0.29      55.4, 36.3    #210  +  +  +  0.10      34.4    SCG (NS,2α0[10]                  #710, #703, #708, #702  +  +  –  0.00  –  –  55.5, 34.2, 33.0, 32.2  26.9 ± 14.2  #707, #705, #711, #709  +  +  –  0.00      30.7, 28.9, 28.6, 13.2    #712, #713  +  +  –  0.00      7.6, 4.7    SCG (gμ,2α)[14]                  #621  +  +  +  0.75  0.56 ± 0.21  56  32.4  37.7 ± 13.3  #631  +  +  +  0.74      53.1    #619  +  +  +  0.66      34.2    #629  +  +  +  0.36      51.4    #608  +  +  +  0.27      30.4    #632, #637  +  +  –  0.01      37.5, 18.0    #615, #634, #638, #635  +  +  –  0.00      66.3, 46.5, 40.8, 38.0    #636, #633, #639  +  +  –  0.00      35.2, 31.6, 12.5    SCG(Xμ,2α)[14]                  #676  +  +  +  0.11  0.11  11  53.3  21.7 ± 8.6  #683  +  +  –  0.03      25.4    #684, #687, #662  +  +  –  0.02      45.1, 25.6, 18.7    #690, #652  +  +  –  0.01      27.8, 21.4    #665, #659, #658, #686  +  +  –  0.00      26.3, 22.5, 16.1, 14.9    #689, #685, #688  +  +  –  0.00      14.5, 13.4, 11.1    SCG (μ, TEL)[9]                  #802, #806, #805  +  +  –  0.04  –  –  49.7, 43.5, 43.3  37.8 ± 12.9  #801, #812  +  +  –  0.02      59.5, 23.4    #803, #811, #804, #813  +  +  –  0.00      42.7, 35.3, 27.2, 15.9    SCG (MCS,2α)[12]                  #756  +  +  +  0.40  0.25 ± 0.09  25  66.1  33.8 ± 15.0  #752  +  +  +  0.22      43.8    #753  +  +  +  0.20      40.8    #762  +  +  +  0.19      32.5    #751  +  +  –  0.08      50.4    #754  +  +  –  0.06      39.6    #766  +  +  –  0.01      13.3    #767, #768, #769, #770  +  +  –  0.00      32.6, 32.5, 20.6, 20.3    #765  +  +  –  0.00      12.9      Transcripta  CSR frequency    Pre-Tr1  Pre-Tr2  Post-Tr  CD8α–GFP+ (%)c  IgA (%) (endogenous CSR)c            Averaged  Induction folde    Averagef  aPre-Tr1, 2 and Post-Tr were detected by RT-PCR (+, expressed; –, not expressed).  bNumbers in parentheses indicate the number of transfectants analyzed by FACS before or after cytokine stimulation.  cAverages of the difference between stimulation + and stimulation – of two independent experimental values are shown. Average percentages of background (stimulation –) in CD8α–GFP+ and IgA+ was 0.01 and 0.20 respectively. The IgA percentage of each clone at the left is indicated at the corresponding position.  dAverages among clones which express Post-Tr.  eInduction index was obtained by division of CD8 α–GFP+ average percentages by the background (0.01).  fAverages among clones which express IgA.  CG (μ,2α)[8]b                  #204  +  +  +  8.38  2.52 ± 2.67  252  74.7  59.8 ± 15.5  #207  +  +  +  4.02      78.9    #209  +  +  +  3.70      71.4    #212  +  +  +  2.63      63.0    #221  +  +  +  0.72      63.6    #222,#216  +  +  +  0.29      55.4, 36.3    #210  +  +  +  0.10      34.4    SCG (NS,2α0[10]                  #710, #703, #708, #702  +  +  –  0.00  –  –  55.5, 34.2, 33.0, 32.2  26.9 ± 14.2  #707, #705, #711, #709  +  +  –  0.00      30.7, 28.9, 28.6, 13.2    #712, #713  +  +  –  0.00      7.6, 4.7    SCG (gμ,2α)[14]                  #621  +  +  +  0.75  0.56 ± 0.21  56  32.4  37.7 ± 13.3  #631  +  +  +  0.74      53.1    #619  +  +  +  0.66      34.2    #629  +  +  +  0.36      51.4    #608  +  +  +  0.27      30.4    #632, #637  +  +  –  0.01      37.5, 18.0    #615, #634, #638, #635  +  +  –  0.00      66.3, 46.5, 40.8, 38.0    #636, #633, #639  +  +  –  0.00      35.2, 31.6, 12.5    SCG(Xμ,2α)[14]                  #676  +  +  +  0.11  0.11  11  53.3  21.7 ± 8.6  #683  +  +  –  0.03      25.4    #684, #687, #662  +  +  –  0.02      45.1, 25.6, 18.7    #690, #652  +  +  –  0.01      27.8, 21.4    #665, #659, #658, #686  +  +  –  0.00      26.3, 22.5, 16.1, 14.9    #689, #685, #688  +  +  –  0.00      14.5, 13.4, 11.1    SCG (μ, TEL)[9]                  #802, #806, #805  +  +  –  0.04  –  –  49.7, 43.5, 43.3  37.8 ± 12.9  #801, #812  +  +  –  0.02      59.5, 23.4    #803, #811, #804, #813  +  +  –  0.00      42.7, 35.3, 27.2, 15.9    SCG (MCS,2α)[12]                  #756  +  +  +  0.40  0.25 ± 0.09  25  66.1  33.8 ± 15.0  #752  +  +  +  0.22      43.8    #753  +  +  +  0.20      40.8    #762  +  +  +  0.19      32.5    #751  +  +  –  0.08      50.4    #754  +  +  –  0.06      39.6    #766  +  +  –  0.01      13.3    #767, #768, #769, #770  +  +  –  0.00      32.6, 32.5, 20.6, 20.3    #765  +  +  –  0.00      12.9    View Large Table 3. Clustering of breakpoints in the proximity of the injunction (neck) between stem and loop regions in the secondary structure S sequence  Distance from neck (bases)a    0 or 1  ≥2    Secondary structureb  Breakpointsc  Secondary structureb  Breakpointsc  aDistance from the junction between the single-strand and double-strand regions (Fig. 6c).  bRelative value was obtained by division of nucleotides in the region defined by the total nucleotides examined.  cFrequency of breakpoints in the region. Numbers of breakpoints in parentheses.  dSignificance by Z-test (P<0.01)  Murine Sα  0.75  0.92 (12/13)  0.25  0.08 (1/13)  Chicken Sμ  0.51  1.00 (6/6)  0.49  0.00 (0/6)  Xenopus Sμ  0.52  0.33 (1/3)  0.48  0.67 (2/3)  MCS  0.72  1.00 (4/4)  0.28  0.00 (0/4)  Average frequency  0.63  0.88d  0.38  0.12d  S sequence  Distance from neck (bases)a    0 or 1  ≥2    Secondary structureb  Breakpointsc  Secondary structureb  Breakpointsc  aDistance from the junction between the single-strand and double-strand regions (Fig. 6c).  bRelative value was obtained by division of nucleotides in the region defined by the total nucleotides examined.  cFrequency of breakpoints in the region. Numbers of breakpoints in parentheses.  dSignificance by Z-test (P<0.01)  Murine Sα  0.75  0.92 (12/13)  0.25  0.08 (1/13)  Chicken Sμ  0.51  1.00 (6/6)  0.49  0.00 (0/6)  Xenopus Sμ  0.52  0.33 (1/3)  0.48  0.67 (2/3)  MCS  0.72  1.00 (4/4)  0.28  0.00 (0/4)  Average frequency  0.63  0.88d  0.38  0.12d  View Large Fig. 1. View largeDownload slide Schematic structure of CSR substrates used in the present study. (A) Basic structure of CSR substrate containing the murine Sμ and Sα sequences, and assay for CSR. Two promoters, EF1-α (shaded arrow) and SRα (dotted arrow), initiate two independent transcription units for S sequences upstream and downstream respectively. Preswitch transcripts (Pre-Tr1 and Pre-Tr2) and CSR postswitch transcript (Post-Tr) were spliced (V-shaped lines) and detected by primer pairs: (BF, BR), (SF, SR) and (BF, SR) as indicated. CSR between two S regions allows expression of the CD8α–GFP fusion protein on the cell surface as a TM-type protein which can be detected by FACS. Exons and S regions are indicated by rectangles and ovals respectively. (B) Variant switch substrates carrying various S and artificial sequences. Each variant was constructed as described in Methods. Fig. 1. View largeDownload slide Schematic structure of CSR substrates used in the present study. (A) Basic structure of CSR substrate containing the murine Sμ and Sα sequences, and assay for CSR. Two promoters, EF1-α (shaded arrow) and SRα (dotted arrow), initiate two independent transcription units for S sequences upstream and downstream respectively. Preswitch transcripts (Pre-Tr1 and Pre-Tr2) and CSR postswitch transcript (Post-Tr) were spliced (V-shaped lines) and detected by primer pairs: (BF, BR), (SF, SR) and (BF, SR) as indicated. CSR between two S regions allows expression of the CD8α–GFP fusion protein on the cell surface as a TM-type protein which can be detected by FACS. Exons and S regions are indicated by rectangles and ovals respectively. (B) Variant switch substrates carrying various S and artificial sequences. Each variant was constructed as described in Methods. Fig. 2. View largeDownload slide Deletion-type recombination in SCG(μ,2α) but not in SCG(NS,2α). (a) Cytokine stimulation-dependent appearance of CD8α–GFP+ cells in SCG(μ,2α) transfectants #204. FACS to assay CD8α–GFP+ cells was performed by staining with anti-CD8α mAb and anti-IgA polyclonal antibodies before or after 3-day cytokine stimulation. Stimulated cells were selected by either HYG or GANC for 6 and 9 days respectively. Percentages of CD8α–GFP+ cells are indicated below each plot. Stim, stimulation. (b) RT-PCR assay of Pre-Tr1, Pre-Tr2 and Post-Tr. Preswitch or postswitch transcripts were examined by RT-PCR before (–) or after (+) cytokine stimulation. (c) FACS analysis of SCG(NS,2α) transfectant #705 after cytokine stimulation. (d) Absence of Post-Tr in three SCG(NS,2α) transfectants after cytokine stimulation. (e) Both murine Sμ and NS sequences are transcribed with comparable efficiency. RT-PCR shows expression levels of Pre-Tr1 and Pre-Tr2 in transfectants of SCG(μ,2α) and SCG(NS,2α). The expression was compared by PCR cycles, at 25 or 30 cycles. The internal control, HPRT, was expressed at similar levels in all clones. Fig. 2. View largeDownload slide Deletion-type recombination in SCG(μ,2α) but not in SCG(NS,2α). (a) Cytokine stimulation-dependent appearance of CD8α–GFP+ cells in SCG(μ,2α) transfectants #204. FACS to assay CD8α–GFP+ cells was performed by staining with anti-CD8α mAb and anti-IgA polyclonal antibodies before or after 3-day cytokine stimulation. Stimulated cells were selected by either HYG or GANC for 6 and 9 days respectively. Percentages of CD8α–GFP+ cells are indicated below each plot. Stim, stimulation. (b) RT-PCR assay of Pre-Tr1, Pre-Tr2 and Post-Tr. Preswitch or postswitch transcripts were examined by RT-PCR before (–) or after (+) cytokine stimulation. (c) FACS analysis of SCG(NS,2α) transfectant #705 after cytokine stimulation. (d) Absence of Post-Tr in three SCG(NS,2α) transfectants after cytokine stimulation. (e) Both murine Sμ and NS sequences are transcribed with comparable efficiency. RT-PCR shows expression levels of Pre-Tr1 and Pre-Tr2 in transfectants of SCG(μ,2α) and SCG(NS,2α). The expression was compared by PCR cycles, at 25 or 30 cycles. The internal control, HPRT, was expressed at similar levels in all clones. Fig. 3. View largeDownload slide Chicken Sμ and Xenopus Sμ can function as targets of mammalian class switch recombinase. (a) FACS analysis shows induction of CD8α–GFP+ cells in SCG(gμ,2α) #619 and SCG(Xμ,2α) #676 after cytokine stimulation. CD8α–GFP+ cells in of SCG(gμ,2α) were sorted by MACS. Frequencies of CD8α–GFP+ cells were indicated below each plot. Stim, stimulation. (b) Stimulation-dependent expression of Post-Trs was shown in three lines of SCG(gμ,2α) transfectants and in one line of SCG(Xμ,2α) transfectants. Preswitch or postswitch transcripts were analyzed by RT-PCR before (–) or after (+) cytokine stimulation. (c) Genomic PCR of postswitch products in CD8a–GFP+ clones derived from stimulated SCG(gμ,2α) #619 and SCG(Xμ,2α) #676. The upper and lower brackets indicate the expected size range of the preswitch product and postswitch product in SCG(gμ,2α) and SCG(Xμ,2α). (d) Germline transcription of chicken Sμ and Xenopus Sμ. RT-PCR shows the comparison of Pre-Tr1 and Pre-Tr2 in each transfectant of SCG(gμ,2α) and SCG(Xμ,2α). The expression was compared by 25 or 30 cycles of PCR. The internal control, HPRT, was expressed at similar levels in all transfectants. Fig. 3. View largeDownload slide Chicken Sμ and Xenopus Sμ can function as targets of mammalian class switch recombinase. (a) FACS analysis shows induction of CD8α–GFP+ cells in SCG(gμ,2α) #619 and SCG(Xμ,2α) #676 after cytokine stimulation. CD8α–GFP+ cells in of SCG(gμ,2α) were sorted by MACS. Frequencies of CD8α–GFP+ cells were indicated below each plot. Stim, stimulation. (b) Stimulation-dependent expression of Post-Trs was shown in three lines of SCG(gμ,2α) transfectants and in one line of SCG(Xμ,2α) transfectants. Preswitch or postswitch transcripts were analyzed by RT-PCR before (–) or after (+) cytokine stimulation. (c) Genomic PCR of postswitch products in CD8a–GFP+ clones derived from stimulated SCG(gμ,2α) #619 and SCG(Xμ,2α) #676. The upper and lower brackets indicate the expected size range of the preswitch product and postswitch product in SCG(gμ,2α) and SCG(Xμ,2α). (d) Germline transcription of chicken Sμ and Xenopus Sμ. RT-PCR shows the comparison of Pre-Tr1 and Pre-Tr2 in each transfectant of SCG(gμ,2α) and SCG(Xμ,2α). The expression was compared by 25 or 30 cycles of PCR. The internal control, HPRT, was expressed at similar levels in all transfectants. Fig. 4. View largeDownload slide Breakpoint sequences in chicken Sμ, MCS and Xenopus Sμ. Alignment of recombination junctions in SCG(gμ,2α), SCG(Xμ,2α) and SCG(MCS,2α). Each breakpoint is represented by three sequences: upstream S sequence (top), recombinant S sequence (middle) and downstream S sequence (Sα, bottom). Pairs of identical sequences are shadowed, while mutations and deletions are indicated by a gap and an asterisk respectively. Each breakpoint is represented as a transition of the' inverted' area from the top to the bottom. Numbers to the left of sequences indicate positions in the switch sequence with the first nucleotide of SalI (GTCGAC) in the upstream S region or ClaI (ATCGAT) in the downstream S region as +1 (7). Fig. 4. View largeDownload slide Breakpoint sequences in chicken Sμ, MCS and Xenopus Sμ. Alignment of recombination junctions in SCG(gμ,2α), SCG(Xμ,2α) and SCG(MCS,2α). Each breakpoint is represented by three sequences: upstream S sequence (top), recombinant S sequence (middle) and downstream S sequence (Sα, bottom). Pairs of identical sequences are shadowed, while mutations and deletions are indicated by a gap and an asterisk respectively. Each breakpoint is represented as a transition of the' inverted' area from the top to the bottom. Numbers to the left of sequences indicate positions in the switch sequence with the first nucleotide of SalI (GTCGAC) in the upstream S region or ClaI (ATCGAT) in the downstream S region as +1 (7). Fig. 5. View largeDownload slide Palindrome-clustered sequence, MCS, but not telomere sequence can serve as a switch sequence. (a) CD8α–GFP+ cells of SCG(μ,TEL) and SCG(MCS,2α) transfectants were analyzed by FACS after cytokine stimulation. CD8α–GFP+ cells from SCG(MCS,2α) transfectants were sorted by MACS for breakpoint sequencing. Frequencies of CD8α–GFP+ cells are indicated below each plot. Stim, stimulation. (b) Preswitch or postswitch transcripts were examined by RT-PCR before (–) or after (+) cytokine stimulation. (c) Genomic PCR shows diverse postswitch products in CD8α–GFP+ clones derived from SCG(MCS,2α) #756. Arrowhead and bracket indicate expected sizes of the preswitch and postswitch products of SCG(MCS,2α) respectively. (d) Telomere sequence and MCS were comparably transcribed. RT-PCR shows the comparison of Pre-Tr1 and Pre-Tr2 among each transfectant of SCG(μ,TEL) and SCG(MCS,2α). The expression was compared by PCR cycles, at 25 or 30 cycles. The internal control, HPRT, was expressed at similar levels in all transfectants. Fig. 5. View largeDownload slide Palindrome-clustered sequence, MCS, but not telomere sequence can serve as a switch sequence. (a) CD8α–GFP+ cells of SCG(μ,TEL) and SCG(MCS,2α) transfectants were analyzed by FACS after cytokine stimulation. CD8α–GFP+ cells from SCG(MCS,2α) transfectants were sorted by MACS for breakpoint sequencing. Frequencies of CD8α–GFP+ cells are indicated below each plot. Stim, stimulation. (b) Preswitch or postswitch transcripts were examined by RT-PCR before (–) or after (+) cytokine stimulation. (c) Genomic PCR shows diverse postswitch products in CD8α–GFP+ clones derived from SCG(MCS,2α) #756. Arrowhead and bracket indicate expected sizes of the preswitch and postswitch products of SCG(MCS,2α) respectively. (d) Telomere sequence and MCS were comparably transcribed. RT-PCR shows the comparison of Pre-Tr1 and Pre-Tr2 among each transfectant of SCG(μ,TEL) and SCG(MCS,2α). The expression was compared by PCR cycles, at 25 or 30 cycles. The internal control, HPRT, was expressed at similar levels in all transfectants. Fig. 6. View largeDownload slide Breakpoint distribution on secondary structures of target sequences. (a) Locations of breakpoints on secondary structures of murine Sα, chicken Sμ, Xenopus Sμ and MCS. I–III, murine Sα; IV–VIII, chicken Sμ; IX–XI, Xenopus Sμ; XII–XIII, MCS. Breakpoints in murine Sα identified downstream of SCG(gμ,2α), SCG(Xμ,2α) and SCG(MCS,2α) and those in chicken Sμ, Xenopus Sμ and MCS in Fig. 4 were mapped. Lengths and locations of examined sequences are as follows: murine Sα (I and II, 100 bases; III, 110 bases) I, 331–530; II, 631–830; III, 901–1110; chicken Sμ (VI, 200 bases; the other, 100 bases) IV, 1042–1141; V and 1541–1640; VI, 3201–3400; VII, 3800–3899; VIII, 4800–4899; Xenopus Sμ (100 bases) IX, 381–480; X, 1601–1700; XI, 2036–2135; MCS (XII, 36 bases; XIII, 170 bases) XII, 1–36; XIII, master sequence in MCS derived from BSIIKS. (b) Locations of breakpoints and their surrounding sequences used for examination of the secondary structure. The bar shows the locations of sequences used in S sequences. In the case of I, II and XIII, breakpoints fell onto the same repeats in the switch region and the number corresponds to the location. The arrow shows the breakpoints in each sequence. In the case of chicken Sμ, G-rich and C-rich regions are indicated. (c) Scheme for counting the distance from the stem–loop junction. Numbers indicate the nucleotide distance from the stem–loop junctions indicated by open circle. For instance, `0' shows immediately adjacent nucleotide positions. Closed circles, nucleotides other than stem–loop junctions. Fig. 6. View largeDownload slide Breakpoint distribution on secondary structures of target sequences. (a) Locations of breakpoints on secondary structures of murine Sα, chicken Sμ, Xenopus Sμ and MCS. I–III, murine Sα; IV–VIII, chicken Sμ; IX–XI, Xenopus Sμ; XII–XIII, MCS. Breakpoints in murine Sα identified downstream of SCG(gμ,2α), SCG(Xμ,2α) and SCG(MCS,2α) and those in chicken Sμ, Xenopus Sμ and MCS in Fig. 4 were mapped. Lengths and locations of examined sequences are as follows: murine Sα (I and II, 100 bases; III, 110 bases) I, 331–530; II, 631–830; III, 901–1110; chicken Sμ (VI, 200 bases; the other, 100 bases) IV, 1042–1141; V and 1541–1640; VI, 3201–3400; VII, 3800–3899; VIII, 4800–4899; Xenopus Sμ (100 bases) IX, 381–480; X, 1601–1700; XI, 2036–2135; MCS (XII, 36 bases; XIII, 170 bases) XII, 1–36; XIII, master sequence in MCS derived from BSIIKS. (b) Locations of breakpoints and their surrounding sequences used for examination of the secondary structure. The bar shows the locations of sequences used in S sequences. In the case of I, II and XIII, breakpoints fell onto the same repeats in the switch region and the number corresponds to the location. The arrow shows the breakpoints in each sequence. In the case of chicken Sμ, G-rich and C-rich regions are indicated. (c) Scheme for counting the distance from the stem–loop junction. Numbers indicate the nucleotide distance from the stem–loop junctions indicated by open circle. For instance, `0' shows immediately adjacent nucleotide positions. Closed circles, nucleotides other than stem–loop junctions. Transmitting editor: T. Watanabe We wish to thank Dr H. Yagita, Targeted Genetics Co., Dr H. Kitao, Professor A. Shimizu, Dr H. Nishimura and Dr K. Kobuke for providing the HM40-3 mAb, the tgCMV/HyTK plasmid, pE13-A plasmid, pPD-1I3 plasmid and Xenopus laevis DNA. We also thank Y. Tabuchi, K. Yurimoto and T. Toyoshima for their technical assistance, T. Nishikawa for secretarial help, and M. Muramatsu, C. G. Lee, X. C. Chen, S. Yoshikawa, Il-Mi Chung, T. Muto and T. Eto for critical discussions and helpful advice. References 1 Oettinger, M. A. 1999. V(D)J recombination: on the cutting edge. Curr. Opin. Cell. Biol.  11: 325. Google Scholar 2 Wagner, S. D. and Neuberger, M. S. 1996. Somatic hypermutation of immunoglobulin genes. Annu. Rev. Immunol.  14: 441. Google Scholar 3 Snapper, C. M. and Finkelman, F. D. 1999. Immunoglobulin class switching. In Paul, W. E., ed., Fundamental Immunology, 4th edn, p. 831. Lippincott-Raven, Philadelphia, PA. Google Scholar 4 Kinoshita, K. and Honjo, T. 2000. Unique and unprecedented recombination mechanisms in class switching. Curr. Opin. Immunol.  12: 195. Google Scholar 5 Muramatsu, M., Kinoshita, K., Fagarasan, S., Yamada, S., Shinkai, Y. and Honjo, T. 2000. Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell  102: 553. Google Scholar 6 Revy, P., Muto, T., Levy, Y., Geissmann, F., Catalan, N., Forveille, M., Dufourcq-Lagelouse, R., Plebani, A., Gennery, A., Sanal, O., Tezxan, I., Ersoy, F., Kayserili, H., Brousse, N., Muramatsu, M., Notarangelo, L. D., Kinoshita, K., Honjo, T., Fischer, A. and Durandy, A. 2000. Activation-induced cytidine deaminase (AID) deficiency causes autosomal recessive form of the hyper-IgM syndrome (HIGM2). Cell  102: 565. Google Scholar 7 Kinoshita, K., Tashiro, J., Tomita, S., Lee, C. G. and Honjo, T. 1998. Target specificity of immunoglobulin class switch recombination is not determined by nucleotide sequences of S regions. Immunity  9: 849. Google Scholar 8 Kitao, H., Arakawa, H., Kuma, K., Yamagishi, H., Nakamura, N., Furusawa, S., Matsuda, H., Yasuda, M., Ekino, S. and Shimizu, A. 2000. Class switch recombination of the chicken IgH chain genes: implications for the primordial switch region repeats. Int. Immunol.  12: 959. Google Scholar 9 Mussmann, R., Courtet, M., Schwager, J. and Du Pasquier, L. 1997. Microsites for immunoglobulin switch recombination breakpoints from Xenopus to mammals. Eur. J. Immunol.  27: 2610. Google Scholar 10 Youn, H. J., Harriss, J. V. and Gottlieb, P. D. 1988. Nucleotide sequence analysis of the C.AKR Lyt-2a gene: structural polymorphism in alleles encoding the Lyt-2.1 T-cell surface alloantigen. Immunogenetics.  28: 345. Google Scholar 11 Takebe, Y., Seiki, M., Fujisawa, J., Hoy, P., Yokota, K., Arai, K., Yoshida, M. and Arai, N. 1988. SRα promoter: an efficient and versatile mammalian cDNA expression system composed of the simian virus 40 early promoter and the R-U5 segment of human T-cell leukemia virus type 1 long terminal repeat. Mol. Cell. Biol.  8: 466. Google Scholar 12 Mizushima, S. and Nagata, S. 1990. pEF-BOS, a powerful mammalian expression vector. Nucleic Acids Res.  18: 5322. Google Scholar 13 Lupton, S. D., Brunton, L. L., Kalberg, V. A. and Overell, R. W. 1991. Dominant positive and negative selection using a hygromycin phosphotransferase–thymidine kinase fusion gene. Mol. Cell. Biol.  11: 3374. Google Scholar 14 Kitao, H., Arakawa, H., Yamagishi, H. and Shimizu, A. 1996. Chicken immunoglobulin mu-chain gene: germline organization and tandem repeats characteristic of class switch recombination. Immunol. Lett.  52: 99. Google Scholar 15 Blackburn, E. H. 1984. Telomeres: do the ends justify the means? Cell.  37: 7. Google Scholar 16 Yen, C. H., Hohman, C. and Elliott, R. W. 1997. Mapping and characterization of three YAC clones containing TTAGGG arrays. Mamm. Genome  8: 775. Google Scholar 17 Nishimura, H., Agata, Y., Kawasaki, A., Sato, M., Imamura, S., Minato, N., Yagita, H., Nakano, T. and Honjo, T. 1996. Developmentally regulated expression of the PD-1 protein on the surface of double-negative (CD4–CD8–) thymocytes. Int. Immunol.  8: 773. Google Scholar 18 Nakamura, M., Kondo, S., Sugai, M., Nazarea, M., Imamura, S. and Honjo, T. 1996. High frequency class switching of an IgM+ B lymphoma clone CH12F3 to IgA+ cells. Int. Immunol.  8: 193. Google Scholar 19 Sen, D. and Gilbert, W. 1988. Formation of parallel four-stranded complexes by guanine-rich motifs in DNA and its implications for meiosis. Nature  334: 364. Google Scholar 20 Sen, D. and Gilbert, W. 1990. A sodium–potassium switch in the formation of four-stranded G4-DNA. Nature  344: 410. Google Scholar 21 Kataoka, T., Miyata, T. and Honjo, T. 1981. Repetitive sequences in class-switch recombination regions of immunoglobulin heavy chain genes. Cell  23: 357. Google Scholar 22 Taylor, L. D., Carmack, C. E., Huszar, D., Higgins, K. M., Mashayekh, R., Sequar, G., Schramm, S. R., Kuo, C. C., O'Donnell, S. L., Kay, R. M., et al. 1994. Human immunoglobulin transgenes undergo rearrangement, somatic mutation and class switching in mice that lack endogenous IgM. Int. Immunol.  6: 579. Google Scholar 23 Mendez, M. J., Green, L. L., Corvalan, J. R., Jia, X. C., Maynard-Currie, C. E., Yang, X. D., Gallo, M. L., Louie, D. M., Lee, D. V., Erickson, K. L., Luna, J., Roy, C. M., Abderrahim, H., Kirschenbaum, F., Noguchi, M., Smith, D. H., Fukushima, A., Hales, J. F., Klapholz, S., Finer, M. H., Davis, C. G., Zsebo, K. M. and Jakobovits, A. 1997. Functional transplant of megabase human immunoglobulin loci recapitulates human antibody response in mice [published erratum appears in Nat. Genet. 1997;16:410]. Nat. Genet. 15:146. Google Scholar 24 Richardson, N., Navaratnam, N. and Scott, J. 1998. Secondary structure for the apolipoprotein B mRNA editing site. Au-binding proteins interact with a stem loop. J. Biol. Chem.  273: 31707. Google Scholar 25 Maas, S. and Rich, A. 2000. Changing genetic information through RNA editing [In process citation]. BioEssays  22: 790. Google Scholar 26 Cheng, S. W., Lynch, E. C., Leason, K. R., Court, D. L., Shapiro, B. A. and Friedman, D. I. 1991. Functional importance of sequence in the stem–loop of a transcription terminator. Science  254: 1205. Google Scholar 27 Dunnick, W., Hertz, G. Z., Scappino, L. and Gritzmacher, C. 1993. DNA sequences at immunoglobulin switch region recombination sites [published erratum appears in Nucleic Acids Res. 1993;21:2285]. Nucleic Acids Res. 21:365. Google Scholar 28 Goossens, T., Klein, U. and Kuppers, R. 1998. Frequent occurrence of deletions and duplications during somatic hypermutation: implications for oncogene translocations and heavy chain disease. Proc. Natl Acad. Sci. USA  95: 2463. Google Scholar 29 Wilson, P. C., de Bouteiller, O., Liu, Y. J., Potter, K., Banchereau, J., Capra, J. D. and Pascual, V. 1998. Somatic hypermutation introduces insertions and deletions into immunoglobulin V genes. J. Exp. Med.  187: 59. Google Scholar 30 Papavasiliou, F. N. and Schatz, D. G. 2000. Cell-cycle-regulated DNA double-strand breaks in somatic hypermutation of immunoglobulin genes. Nature  408: 216. Google Scholar 31 Bross, L., Fukita, Y., McBlane, F., Démollière, C., Rajewsky, K. and Jacobs, H. 2000. DNA double-strand breaks in immunoglobulin genes undergoing somatic hypermutation. Immunity  13: 589. Google Scholar 32 Peters, A. and Storb, U. 1996. Somatic hypermutation of immunoglobulin genes is linked to transcription initiation. Immunity  4: 57. Google Scholar 33 Tumas-Brundage, K. and Manser, T. 1997. The transcriptional promoter regulates hypermutation of the antibody heavy chain locus. J. Exp. Med.  185: 239. Google Scholar 34 Fukita, Y., Jacobs, H. and Rajewsky, K. 1998. Somatic hypermutation in the heavy chain locus correlates with transcription. Immunity  9: 105. Google Scholar 35 Storb, U., Klotz, E. L., Hackett, J., Jr, Kage, K., Bozek, G. and Martin, T. E. 1998. A hypermutable insert in an immunoglobulin transgene contains hotspots of somatic mutation and sequences predicting highly stable structures in the RNA transcript. J. Exp. Med.  188: 689. Google Scholar 36 Short, J. M., Fernandez, J. M., Sorge, J. A. and Huse, W. D. 1988. Lambda ZAP: a bacteriophage lambda expression vector with in vivo excision properties. Nucleic Acids Res.  16: 7583. Google Scholar 37 Blackburn, E. H. 1984. Telomeres: do the ends justify the means? Cell  37: 7. Google Scholar 38 Yen, C. H., Hohman, C. and Elliott, R. W. 1997. Mapping and characterization of three YAC clones containing TTAGGG arrays. Mamm. Genome  8: 775. Google Scholar 39 Nikaido, T., Nakai, S. and Honjo, T. 1981. Switch region of immunoglobulin Cμ gene is composed of simple tandem repetitive sequences. Nature  292: 845. Google Scholar 40 Arakawa, H., Iwasato, T., Hayashida, H., Shimizu, A., Honjo, T. and Yamagishi, H. 1993. The complete murine immunoglobulin class switch region of the alpha heavy chain gene-hierarchic repetitive structure and recombination breakpoints. J. Biol. Chem.  268: 4651. Google Scholar 41 Storb, U., Peters, A., Klotz, E., Kim, N., Shen, H. M., Hackett, J., Rogerson, B., O'Brien, R. and Martin, T. E. 1998. Immunoglobulin transgenes as targets for somatic hypermutation. Int. J. Dev. Biol.  42: 977. Google Scholar 42 Pasqualucci, L., Migliazza, A., Fracchiolla, N., William, C., Neri, A., Baldini, L., Chaganti, R. S., Klein, U., Kuppers, R., Rajewsky, K. and Dalla-Favera, R. 1998. BCL-6 mutations in normal germinal center B cells: evidence of somatic hypermutation acting outside Ig loci. Proc. Natl Acad. Sci. USA  95: 11816. Google Scholar © 2001 Japanese Society for Immunology
TI - Palindromic but not G-rich sequences are targets of class switch recombination
JF - International Immunology
DO - 10.1093/intimm/13.4.495
DA - 2001-04-01
UR - https://www.deepdyve.com/lp/oxford-university-press/palindromic-but-not-g-rich-sequences-are-targets-of-class-switch-blSJhaDy5B
SP - 495
EP - 505
VL - 13
IS - 4
DP - DeepDyve
ER -