HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Müller, J. R. Articles by Marcu, K. B. Search for Related Content PubMed PubMed Citation Articles by Müller, J. R. Articles by Marcu, K. B. Pubmed/NCBI databases UniGene Compound via MeSH Substance via MeSH

Generation of Switch Hybrid DNA Between Ig Heavy Chain-µ and Downstream Switch Regions in B Lymphocytes¹

Jürgen R. Müller^*,, Thomas Giese^*, Diane L. Henry, J. Frederic Mushinski^* and Kenneth B. Marcu^2,

^* Laboratory of Genetics, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892; and Department of Biochemistry and Cell Biology, State University of New York at Stony Brook, Stony Brook, NY 11794

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Ig heavy chain isotype switching in B lymphocytes is known to be preceded by transcription of a portion of the particular heavy chain gene segment that is targeted for recombination. Here, we describe an active role for these transcripts in the switch recombination process. Using an in vitro assay that exposes an artificial switch-µ (Sµ) minisubstrate to switch region transcripts in the presence of nuclear extracts from switching cells, we demonstrate that free 3' ends of the Sµ sequence are extended onto switch region transcripts by reverse transcription. The activity was induced in splenic B lymphocytes upon activation with LPS or CD40 ligand. This in vitro process is thought to be relevant to in vivo class switching for two reasons: 1) although only one-third of the Sµ minisubstrate actually contains Sµ sequence, all crossovers between switch regions occurred in the Sµ portion; and 2) treatment of B lymphocytes with IL-4, which enriches for switching to S1, increases the ratio of Sµ-S1 to Sµ-S3 hybrids by 16% after LPS treatment and by 37% after CD40 ligand activation, implicating this Sµ-primed reverse transcription of switch region transcripts as a novel mechanism of regulating the specificity of isotype switching. Further evidence for an active role of switch region transcripts was obtained by expressing S RNA in trans in the Bcl₁B₁ B lymphoma line. Endogenous Sµ-S switch circles were detected in Bcl₁B₁ cells expressing exogenous S RNA but not in mock-transfected cells.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
The production of Ig isotypes with identical Ag specificity but with different effector functions by B lymphocytes is normally accomplished by a process of DNA breakage and recombination. This process relocates a particular heavy chain variable region from the Igµ heavy chain constant region to a different downstream Ig constant region, accompanied by deletion of the intervening sequences (reviewed in Refs. 1 and 2). Recombination breakpoints in the mouse are localized to regions of 1 to 10 kbp upstream of each of the seven Ig constant regions (except Ig) that are termed switch regions. All switch regions exhibit a characteristic pattern of repetitiveness, and each switch region has a unique sequence but an overall similarity to all other switch regions (3).

Only certain cells can be induced to switch their Ig isotype in vitro: certain B cell lines and primary mature B lymphocytes that have been treated with LPS or CD40 ligand (CD40L).³ Studies of these cells revealed that switching is preceded by an unusual simultaneous transcription of both the Igµ region, to which the variable region has been productively joined, and the downstream switch region, which is targeted for recombination (4). Transcription is initiated 5' of a small exon (I exon) upstream of each switch region and is extended through the coding constant region (5). Primary transcripts are processed so that the I exon is joined directly to the exons of the constant region, by splicing out the intronic switch sequences. There is strong evidence that this pre-switch germline transcription is critical to the switching process, since knockout mice that lack the I exon from only one Ig constant gene segment fail to undergo isotype switching to that region, while switching normally to the remaining constant regions (6-8). Germline transcription of particular Ig constant gene segments can be specifically induced by certain cytokines (e.g., IL-4, IL-10, TGF-ß, IFN-), and the gene segments that respond to the cytokine by induction of germline transcription have been found to be preferentially involved in the switching (for review, see 1 . Some cytokines, however, affect the specificity of isotype switching without regulating germline transcription, implying a second mechanism of directing isotype switch (1).

The molecular mechanisms that mediate isotype switch recombination are largely unknown. The initial event is believed to be a double-strand break in a switch region (9, 10). Furthermore, an error-prone DNA synthesis is thought to be implicated in isotype switching, since switch recombination breakpoints revealed DNA strand-specific point mutations but only on one side of the recombinations (11, 12).

We recently detected the induction of an exonuclease activity in splenic B cells upon activation with either LPS or CD40L (13). Interestingly, this activity could be inhibited by the presence of switch germline transcripts. We speculated that this enzyme processes DNA strand breaks in switch regions after their generation by switch sequence-specific endonucleases.

We and others noticed the increased expression of endogenous retroviral sequences in B lymphocytes following LPS or CD40L treatment (Ref. 14; J. R. Müller, unpublished observation). It is known that some of these sequences have open reading frames that encode reverse transcriptases, and several well-studied recombinations are mediated in part by reverse transcriptases. Of particular interest appears to be the retrohoming of group II introns in yeast (reviewed in 15 . This recombination uses free 3' ends of the cleaved target DNA to prime reverse transcription of intron RNA that is inserted.

The detection of the exonuclease activity and of the expression of genes encoding putative reverse transcriptases led us to formulate a new hypothetical model of switch recombination (Fig. 1). The process is initiated by a switch sequence-specific endonuclease, and the newly generated free ends are processed by exonucleases (Fig. 1, A and B). Switch portions of germline transcripts use short homologies to anneal to the template strand of Sµ. Reverse transcriptases start DNA polymerization at a free 3' end of Sµ and generate the chimeric switch DNA (Fig. 1C). The synthesized cDNA strand undergoes homologous recombination with the respective downstream switch locus (Fig. 1, D–F). Here, we describe experiments supporting this hypothesis.

View larger version (19K): [in this window] [in a new window]   FIGURE 1. Hypothetical model of switch isotype recombination. A, The recombination process is initiated by a Sµ-specific endonuclease. B, Strand breaks are processed by an exonuclease. This activity is stopped by the presence of switch region RNA (13). C, The same switch RNA anneals to homologous sequence in Sµ. Free ends of Sµ prime a reverse transcription of the switch RNA resulting in chimeric switch DNA. D, The 3' end of the generated cDNA invades the homologous endogenous locus. Strand exchange proceeds until the end of the homology at the beginning of the Sµ portion of the chimeric DNA. E, Homologous recombination generates the switch circle. F, The remaining free ends in Sµ and in the downstream switch region are repaired using double-strand break repair processes.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Spleen cell suspensions were prepared from 5- to 12-week-old mice (DBA2/N, BALB/cAn, BALB/cJ). Erythrocytes were lysed in 0.15 M NH₄Cl, 1 mM KHCO₃, 0.1 mM Na₂-EDTA (pH 7.2), and B lymphocytes were enriched using anti-B220 mAb (Miltenyi Biotec, Bergisch Gladbach, Germany) and magnet-activated cell sorting. Cells were cultured at 3 x 10⁵ cells/ml in RPMI 1640 supplemented with 2 mM glutamine, 50 µM 2-ME, penicillin, streptomycin, and 10% FCS. Total spleen cell suspensions were activated by 50 µg/ml LPS (Sigma, St. Louis, MO). CD40L, as the membrane fraction of baculovirus-infected Sf21 cells (kindly provided by Dr. Marilyn Kehry, Boehringer Ingelheim (Ridgbury, CT) (16)) was added to sorted B lymphocytes together with 3 ng/ml anti- dextran (kindly provided by Dr. Clifford Snapper, Uniformed Services University of the Health Sciences, Bethesda, MD). If indicated, cells were stimulated with 3,000 U/ml murine IL-4 (harvested from baculovirus-infected cells or from transfected LT-1 cells (17), both of which gave virtually identical results). LPS-activated cells were lysed after 44 h, and CD40L-treated cells after 72 h. Routine FACScan analysis to determine surface Ig expression (IgG1, IgG3) was performed after 72 and 120 h, respectively. Nuclear extracts were also prepared from freshly isolated B lymphocytes, liver cells, a B lymphocyte line (CH12.LX), and a fibroblast line (NIH3T3).

Nuclear extracts were prepared by a standard method as follows (18). We washed the cells three times in PBS, and 2 x 10⁷ cells were resuspended in 400 µl 10 mM HEPES (pH 7.9), 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, then incubated on ice for 15 min. After adding Triton X-100 (final concentration 0.6%), cells were vortexed (10 s) and centrifuged twice (500 x g for 3 min). The pellet was resuspended in 50 µl of 20 mM HEPES (pH 7.9), 400 mM NaCl, 1 mM EDTA, 1 mM EGTA, and 1 mM DTT supplemented with the protease inhibitors set (Boehringer, Indianapolis, IN), gently rocked for 15 min at 4°C, and centrifuged (10,000 x g, 5 min). The resulting supernatants were adjusted to a protein concentration of 2 µg/µl, aliquoted, and stored at -70°C.

Plasmids and in vitro transcription

To generate the Sµ minisubstrate (Fig. 2A), we PCR-amplified an 87-bp fragment of mouse c-myc intron 1 (GenBank MUSCMYC1 #1625-1711) and cloned it into the HindIII/EcoRI sites of pCR2.1 (Invitrogen, Carlsbad, CA), generating plasmid pCRM. The inserted fragment harbored a previously tested high quality PCR primer-annealing site close to the HindIII end, but its sequence was otherwise considered to be nonspecific. This plasmid was cut with XhoI and EcoRV, and two annealed Sµ oligonucleotides (nontemplate strand: GAGCTGAGCTGGGGTGAGCTGAGCTGAGCTGGGGTGAGCTGA) were ligated into this site, generating pCRMµ. The resulting chimeric insert was a 169-bp fragment that contained 14 bp of upstream vector sequence, 42-bp of Sµ, 87 bp of c-myc, including the primer annealing site, and 26 bp of downstream vector sequence (Fig. 2). This insert was released by HindIII/XbaI digestion and isolated by electrophoresis on low melting point agarose gels. The 5' overhangs were filled in by Klenow polymerase with dGTP, biotinylated dATP, and dideoxy-CTP (ddCTP; BRL, Gaithersburg, MD). The ddCTP terminated the fill-in of the XbaI overhang with the first base, while the HindIII site was filled in with dGTP, dATP-biotin, and ddCTP. Thus, only the strand that contained the Sµ nontemplate sequence was biotinylated.

View larger version (48K): [in this window] [in a new window]   FIGURE 2. Generation of DNA hybrids between Sµ and non-Sµ switch sequences. A, Design of the experiments: the 169-bp Sµ minisubstrate that contains 14 bp of vector sequence, 42 bp of Sµ sequence (thick line, nontemplate strand), 87 bp of myc sequence used for PCR primer annealing, and 26 bp of downstream vector sequence was biotinylated at the HindIII site. This DNA was incubated with RNAs at 37°C for 1 h in the presence of nuclear extracts prepared from activated B cells. We postulated a limited exonuclease digestion of the Sµ DNA minisubstrate would be followed by the generation of switch hybrid DNAs by reverse transcription of switch RNAs (in the direction of the left-facing arrow and RT) using the free 3' ends of the Sµ template strand as primers. The DNA and any attached products were isolated using avidin-coated beads, and any Sµ template strand extension products were harvested by NaOH denaturation and ethanol precipitation as explained in Materials and Methods. The reaction products were analyzed by PCR (primers are indicated by arrowheads). B, The resulting PCR products are shown. In the first experiment (upper panel), hybrid sequences containing Sµ and S or S1 were generated only if both DNA and RNA were present, and no products were detected if the initial reaction contained proteinase K. The second experiment was nearly identical, with the exception that all reactions contained both RNA and DNA, and proteinase K was present in lanes 2, 4, 6, and 8. The PCR-generated products were cloned and sequenced (see Table I).

To generate switch region RNAs, switch portions of several S regions (S1, GenBank entry MUSIGHANB #3664-4983; S3, MUSIGHANA #54-2561; S, MUSIALPHA #1861-4693) were PCR amplified and cloned into the bacterial expression vector pCR2.1. Clones were sequenced to identify those with inserts that were in appropriate orientation to the plasmid T7 promoter site. RNAs corresponding to the sense orientation of downstream coding regions were generated with T7 RNA polymerase (Promega, Madison, WI) using 9 µg of linearized plasmid DNA as a template. After polymerization (4 h, 37°C), the plasmid DNA was destroyed by DNase digestion (Promega). Unincorporated nucleotides were removed by passage through Sephadex G-50 columns (5Prime3 Prime, Boulder, CO), and proteins were removed by phenol/chloroform extraction and isopropanol precipitation. The concentrations of in vitro-generated RNAs were estimated on ethidium bromide-stained agarose gels. As a negative control, we also generated antisense S RNA by this same procedure.

Generation of hybrid switch DNA

Approximately 20 ng of biotinylated Sµ minisubstrate was mixed in a 50-µl reaction with 2 µg of switch RNA (S, S1, or S3) and 2 µg of total protein from nuclear lysates in 10 mM Tris (pH 8.3), 50 mM KCl, 10 mM MgCl₂, 1 mM DTT, and 500 ng of nonspecific DNA (EcoRI-linearized pCR2.1), dATP, dGTP, dCTP, and dTTP at 20 µM each. After incubation at 37°C for 60 min, the reaction was terminated by adding 40 µg of proteinase K. To isolate the processed Sµ DNA, we added 50 µl of avidin-coated polystyrene beads (5% w/v; Baxter, Mundelein, IL) that had been preincubated with 50 µg of sheared herring sperm DNA. This suspension was gently rocked for 30 min at room temperature, followed by two washes in 0.5 ml of 300 mM NaCl, 30 mM sodium citrate, 0.1% SDS. To denature the dsDNA and release the Sµ template strand DNA (including the hypothesized reverse transcriptase-generated products, see Fig. 2A) from the biotin-tagged nontemplate strand of the Sµ minisubstrate, we resuspended the suspension in 0.2 M NaOH, incubated it for 30 min at room temperature, and removed the Sµ nontemplate strand attached to the beads by centrifugation. The avidin/biotin binding is effectively irreversible and should not be destroyed by the NaOH treatment. The denatured DNA in the supernatant was ethanol precipitated in 0.5 M Tris (pH 7.4) in the presence of 40 µg glycogen, pelleted by centrifugation, and rehydrated in 20 µl water. PCR was done with 2 µl of this preparation as a template in 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 2.5 mM MgCl₂, gelatin at 0.1 mg/ml, each primer at 0.5 µM, and each dNTP at 200 µM in a volume of 50 µl. Cycling conditions involved a 5-min initial denaturation at 95°C with the reaction subsequently held at 80°C for the addition of 1.25 U of Taq polymerase. This was followed by 40 cycles of denaturation (15 s at 95°C), annealing (15 s at 62°C), and extension (between 30 s and 2 min at 72°C). The primers used to test for hybrid generation always included the same oligonucleotide (Sµ1, 5'-AGGGATACCCGCGGATCCCAAGTAGGAATGTGAGG-3'), which annealed to the Sµ minisubstrate 69 bp downstream of the Sµ portion (Fig. 2A) and one of three primers designated to hybridize to the 3' end of each cDNA that would have been generated by reverse transcription of the switch RNAs: S1, 5'-GGGAAAGGGAGGGCATCAAGGAG-3'; S3, 5'-CAAGTTGAGCAGCCACAGGAGAGC-3'; S1, 5'-GACTGGTCTGAGGCGGGCTAATCTG-3'. Products of the PCR reactions were size fractionated by electrophoresis on 1.5% agarose gels and stained with ethidium bromide. Visible products were excised and sequenced either directly (Promega) or after cloning into pCR2.1.

To investigate whether Sµ nontemplate strand extension products were generated in the presence of S antisense RNA, we incubated nonbiotinylated MD3 substrate with S antisense RNA and nuclear extracts. DNA was isolated directly by phenol/chloroform extraction, ethanol precipitation, and rehydration in 20 µl water. PCR was performed with primers that anneal to the upstream end of the Sµ sequence and to the downstream end of the S sequence: Sµ2, 5'-TGGAGTAGCTGAGATGGGGTGAGAT-3; and S2, 5'-ATTGTAACCAGCCAAGCCAAGTTTC-3'. These PCR reactions were size fractionated by electrophoresis on 5% polyacrylamide gels followed by blotting onto nylon membranes (Hybond-N, Amersham, Arlington Heights, IL). The membranes were hybridized in Hybrisol (Oncor, Gaithersburg, MD) with a probe (corresponding to the 610-bp downstream end of the S RNA) that we generated by nick end translating S DNA with [-³²P]dCTP, washed, and exposed to x-ray film.

Quantitation of chimeric DNA generation

To quantify the rate of DNA hybrid generation, we replaced unlabeled dCTP in the above-detailed reaction with 50 µCi [-³²P]dCTP and removed unincorporated nucleotides with Sephadex G-50 columns before adding the avidin-coated beads. After washing, denaturation, precipitation, and rehydration (see above), part of the reaction (20 µl) was measured in a liquid scintillation counter. The remainder of the reaction mix was precipitated with ethanol and analyzed by electrophoresis an 8% polyacrylamide/8 M urea gels and exposed to x-ray film.

Amplification of endogenous switch circle recombinations

For comparison, we also PCR-amplified physiologic Sµ-S switch recombination sequences. This was done with DNA prepared from Peyer’s patches, since these tissues harbor a high frequency of IgA⁺ cells. Primers were used in a nested PCR that amplify Ig sequences of the circular byproducts of Sµ-S switch recombinations (switch circles): S3, 5'-TGGTCTGTACTGGGCTGGGCTAACT-3', and Sµ3, 5'-GCCAGCCCAGTTGAGTCCAGATG-3' for round one; S1 and Sµ4, 5'-CCAGCTCAGCCCCGTTCATTCAA-3' for the second round.

Expression of S RNA in Bcl₁B₁ cells

We PCR amplified the entire S region (GenBank MUSIALPHA No. 1861-4693) using DBA liver DNA as template and cloned the product into the HindIII site of the mammalian expression vector pMV12 downstream of its murine Moloney long terminal repeat. We transfected the B cell line Bcl₁B₁ with this construct using the cationic lipid compound DMRIE-C (BRL). Cells were selected for the coexpressed hygromycin resistance gene (400 µg/ml hygromycin) for 7 days. Dead cells were removed, and genomic DNA was prepared from these cells after a further 2 to 6 days in culture using standard SDS lysis, proteinase K digestion, phenol/chloroform extraction, and ethanol precipitation.

To find evidence for ongoing isotype switching, we PCR amplified switch circle recombinations. Nested PCR was performed with primers that anneal to just downstream of Sµ and just upstream of endogenous S: Sµ3 and S4, 5'-AGATAGGCTTAGTTTGGCTGATTGA-3', for round one; Sµ4 and S5, 5'-TGGGATGGGCTTGCTGAAATGGTT-3', in the second round. To avoid the detection of recombinations between transfected and endogenous switch loci, we chose pairs of PCR primers that annealed only to the endogenous and not the transfected sequences. Products of the PCR reactions were size fractionated by electrophoresis on 1.5% agarose gels and stained with ethidium bromide. Visible products were excised and directly sequenced.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Generation of chimeric switch region DNA by reverse transcription

To test whether isotype switching might be mediated by the generation of an Sµ-primed reverse transcript, we constructed an Sµ minisubstrate and biotinylated its nontemplate strand (Fig. 2A). Several switch regions (S, S1, S3) were cloned into riboprobe vectors, and RNAs were generated by in vitro transcription. Nuclear extracts were prepared from murine splenic B lymphocytes in which isotype switching had been induced by treatment with either LPS or CD40L. We coincubated the biotinylated Sµ DNA minisubstrate with different in vitro-transcribed switch RNAs in the presence of nuclear extracts, and the resulting products were isolated with avidin-coated beads followed by NaOH denaturation under conditions that dissociated complementary strands of DNA and hydrolyzed the RNA, but did not uncouple the biotinylated strand from the avidin-coated beads. These beads were removed by centrifugation, and the supernatant was analyzed. Using PCR, we detected a number of different hybrids between Sµ and sequences of the respective RNA (Fig. 2B). Omitting either the Sµ minisubstrate or switch RNA, or preincubating the reactions with proteinase K or RNase A, prevented the detection of these products.

Sequence analysis of chimeric switch DNA

Twelve PCR-generated products were subsequently isolated by electrophoresis on low melting point agarose and sequenced (Table I). The resulting sequences were compared with the known S, S1, and S3 sequences as well as with the Sµ minisubstrate. All of the detected products represented chimeric sequences that were generated by joining part of the Sµ DNA in the minisubstrate to cDNA sequences of the RNA that had been used in the in vitro reactions. None of the products, however, resulted from an extension of the entire DNA substrate, but crossover points always occurred within the Sµ portion of the DNA substrate, implying that there had been a processing of the Sµ DNA minisubstrate before extension. PCR products could have theoretically been generated if the crossover had occurred within the 125 bp between the Sµ primer that we used for PCR and the XbaI site (Fig. 2A). However, no chimeric products were found that joined cDNA sequences of the S, S1, or S3 RNAs to sequences of the Sµ DNA minisubstrate outside of its Sµ portion.

Quantitation of Sµ-primed reverse transcription

To measure the relative amounts of Sµ-primed reverse transcription, we added [-³²P]dCTP during the extension reaction and determined the amount of label that remained hybridized to the biotinylated Sµ nontemplate strand after washing.

Reverse transcription was efficient if nuclear extracts from unstimulated B cells or CH12.LX cells were used (Fig. 3A). The activity was increased by >70% in nuclear lysates from B cells that had been activated with either LPS or CD40L. Polymerization caused by fibroblast extracts was approximately one-half the amount measured in untreated B lymphocytes, and nuclear proteins from liver cells generated <10% of Sµ-primed reverse transcription compared with B lymphocytes.

View larger version (44K): [in this window] [in a new window]   FIGURE 3. Quantitation of switch hybrid DNA generation. To compare the relative amounts of Sµ-primed reverse transcription, we added [-32P]dCTP to the reactions, isolated Sµ template strand products as described above, and size fractionated them on 6% polyacrylamide/8 M urea gels. A smear is expected because 1) non-Sµ RNAs anneal at different sequences of the Sµ DNA, and 2) all polymerized products are processed by exonucleases (see Discussion). A, Cell type specificity of generating chimeric Sµ-S3 DNA. We normalized all data against the value of untreated B cells, which was arbitrarily designated to be 1.0. The reverse transcriptase activity is increased in B lymphocytes by 71% after LPS treatment and by 76% after CD40L activation. Smaller amounts of this enzymatic activity were present in nuclear lysates prepared from fibroblasts and from liver cells. B, Effective polymerization was observed with S, S1, or S3 sense RNAs in the presence of nuclear extracts from LPS-activated spleen cells (lanes 4-6). In contrast, S antisense RNA was used much less effectively (lane 7). Substitution of nonbiotinylated DNA (lane 2) or the addition of proteinase K (lane 3) prevented the appearance of products. Lane 1, Sequencing ladders used as size markers. C, S antisense RNA is used for the generation of Sµ nontemplate strand extension products. A longer Sµ DNA substrate (MD3) was exposed to nuclear lysates in the presence of S antisense RNA. We isolated DNA directly and analyzed the preparation for the presence of chimeric DNA (Sµ-S nontemplate strand recombinations). PCR reactions were size fractionated, blotted, and probed with an S probe. A strong hybridization signal was detected between 1.0 and 2.5 kb if antisense RNA was used (lane 4). Both the use of sense RNA (lane 3) and preincubation with RNase A before adding nuclear lysates (lane 1) prevented the occurrence of PCR products.

We noted that a percentage of the generated products was shorter in length than the Sµ substrate (Fig. 3B). The most likely explanation for this is a nonspecific exonuclease activity that is present in nuclear extracts from B lymphocytes. In an analogous manner, omitting nonspecific DNA during the extension reaction also led to the appearance of shorter products (not shown).

S antisense RNA was not effectively used for extending the Sµ template strand (Fig. 3B). To study whether 3' ends of the Sµ nontemplate strand prime reverse transcription of S antisense RNA, we used a longer, 692-bp Sµ substrate (MD3) that was exposed to nuclear extracts from switching cells in the presence of S antisense RNA. We isolated DNA directly and analyzed it by PCR. Figure 3C shows that these products can be detected in the presence of S antisense (lane 4) but not S sense RNA (lane 3). Preincubating the reaction mixture with RNase A before the addition of nuclear extracts prevented the detection of any products (lane 1).

Correlation between isotype switch specificity and the generation of switch hybrid DNA

The specificity of isotype switching both in LPS- and in CD40L-activated B lymphocytes can be affected by IL-4: the presence of this cytokine normally directs switching to Ig1 and suppresses the appearance of sIgG3⁺ cells (20). To determine whether this specificity correlates with a preferential usage of the corresponding switch RNA as template for Sµ-primed reverse transcription, we prepared nuclear proteins from IL-4-stimulated B lymphocytes and compared the rate of the generation of Sµ-S1 chimeric DNAs to that for Sµ-S3 chimeras.

Nuclear extracts from LPS-treated total spleen cells (Fig. 4) used S1 and S3 RNA with approximately the same efficiency, i.e., the ratio of Sµ extension onto S1 and S3 RNA equals about one. The presence of IL-4 during LPS activation increases this ratio by an average of 16.6% (from 2.3 to 39.6%). In cell populations that were enriched for B lymphocytes and then activated by CD40L, S1 was preferred over S3 by 37.8% (6.5-58.5%) if IL-4 was present. The presence of IL-4 mainly affected the generation of Sµ-S3 chimeric DNA (Fig. 4). IL-4 decreased the appearance of these chimeras in LPS-treated cultures by 12% and in CD40L-activated cells by 34%. In contrast, IL-4 had only a minor effect on the yield of Sµ-S1 chimeric sequences (not shown).

View larger version (17K): [in this window] [in a new window]   FIGURE 4. Stimulation of B lymphocytes with IL-4 affects the specificity of the switch region hybrids that are generated. Sµ DNA and S1 or S3 RNA were exposed to nuclear extracts prepared from four differently stimulated B cell populations, and the rate of generating Sµ-S1 or Sµ-S3 hybrids was measured in parallel (n = 15 using 10 different protein preparations for LPS; n = 6 with four extracts for CD40L activation). We normalized all data against the value of Sµ-S1 hybrids generated by cells in the absence of IL-4, which was arbitrarily designated to be 1.0. The measurements for the generation of Sµ-S3 hybrids are shown. The presence of IL-4 decreases the generation of Sµ-S3 chimeric DNA by 12% in LPS-activated total spleen cells and by 34% in CD40L-treated B lymphocytes. The ratio of Sµ-S1 to Sµ-S3 chimeras detected in each experiment are compared. IL-4 increases this ratio by an average of 16.6% for LPS treatment and by 37.8% if CD40L is used. In an additional five experiments, the rate of generating Sµ-S chimeric DNA was compared with the generation of Sµ-S1 hybrids. S RNA was used 40% more efficiently than S1 RNA for Sµ-primed cDNA synthesis. We found no effect of IL-4 treatment on the generation of these chimeric DNAs. Results are shown as average plus the SE of the mean. The p values comparing IL-4-stimulated and -unstimulated groups were calculated with the Mann-Whitney test for S1/S3 ratios.

Induction of switch recombination by switch sequences expressed in trans

The results described above suggest an active role for switch region RNAs. To obtain independent evidence for such a function, we transfected the B cell line Bcl₁B₁ with a construct in which S is constitutively transcribed from a long terminal repeat (Fig. 5A). Bcl₁B₁ cells do not normally switch to IgA; and no Sµ-S switch circle recombination products, as evidence for active switching, were detected by PCR in mock-transfected (empty pMV12 expression vector) cells (Fig. 5B). In contrast, a number of these recombinations were found in cells between days 11 and 15 after transfecting S sequences.

View larger version (72K): [in this window] [in a new window]   FIGURE 5. S RNA induces switch recombination if expressed in trans. A, The upper panel shows the endogenous Igµ and Ig loci. Restriction sites: X, xbaI; R, EcoRI; H, HindIII. The entire DBA S locus (thick line above S) was PCR-amplified and cloned into the mammalian expression vector pMV12 (lower panel). LTRpr, long terminal repeat promoter; Tkpr, HSV thymidine kinase promoter; pA, polyadenylation signal. We transfected Bcl1B1 cells with the resulting construct and selected for the co-expressed hygromycin resistance gene. DNAs were prepared between 11 and 15 days after transfection. We analyzed these samples for the presence of endogenous Sµ-S circle recombinations by PCR (primer annealing sites are depicted in the upper panel by horizontal arrowheads). B, The resulting PCR products are shown. No products were recovered if DNAs from mock-transfected cells were used (lanes 1–3). In contrast, S-transfected cells harbored multiple switch circle recombinations (lanes 4–9). C, PCR products were generated and gel purified from the above switch circles and the Bcl1B1 and DBA S regions. Portions of the DNAs, wherein allelic differences reside (vertical arrow in A), were directly sequenced. The upper sequence represents transfected S and reads AGACT-ATGCT from the bottom. The seventh base is replaced in endogenous Bcl1B1 sequence by a G (arrow). The three middle panels were derived from switch circle recombination products. All have the G at the specific position, indicating that the S sequence is of endogenous origin.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
The mechanisms that mediate isotype switch recombination in B lymphocytes are largely unknown. Many results show a crucial role for sterile switch region transcripts in that process (1). However, experiments using targeted gene disruption in mice showed that transcription of switch regions itself is insufficient to induce switch recombination (21, 22). We recently showed an inhibiting effect of germline transcripts on an exonuclease activity in activated B lymphocytes (13). We also detected the transcriptional up-regulation of sequences encoding putative reverse transcriptases (J. R. Müller, unpublished observation). On the basis of these independent observations, we propose a new, hypothetical model of switch recombination (Fig. 1) and provide herein experiments that support this idea.

A crucial step in our working model is the generation of a chimeric switch DNA intermediate by an Sµ-primed reverse transcription (Fig. 1C). We developed an in vitro assay to mimic in vivo conditions by incubating an Sµ DNA substrate with nuclear proteins prepared from switching cells and in vitro-generated switch region RNAs. Using this assay, followed by PCR and sequencing, we showed that these chimeric DNAs are indeed generated. No chimeric products were detected if the reactions were preincubated with either proteinase K or RNase A, indicating that both proteins and RNA are required to generate these hybrid DNAs. We cannot formally rule out one potential artifact: the initial reaction could generate switch cDNA by a reverse transcription of switch RNA without using Sµ DNA as primer. In the initial rounds of PCR, the Sµ DNA primes DNA amplification within the newly generated switch cDNA. Although these artifacts are known to occur, extensive homology is required at the crossover points (J. R. Müller, unpublished observation). Sequencing chimeric switch DNA products, we found that some of them have only limited or no homology at all at the crossover points (Table I). We believe that this should virtually exclude their artifactual generation. Therefore, the DNA sequences of the products generated by nuclear lysates of normal, switching B cells would be best explained by a mechanism in which germline switch RNAs anneal at short homologies in Sµ, and free 3' ends in Sµ are used to prime the DNA synthesis.

Most results that are presented in this paper are based on an in vitro assay. However, two results point to the significance of the mechanism described for intact cells. First, we found that Sµ sequences are necessary to prime reverse transcription of non-Sµ switch RNAs as shown by the fact that all sequenced chimeric products (n = 12) were generated by priming in the Sµ portion of the substrate that we used. None of the products resulted from priming within the nonspecific part of the DNA substrate that represented more than two thirds of the sequence. Priming from within the DNA substrate would require a prior processing, possibly by an exonuclease activity. Such exonucleases have been detected in nuclear extracts from B cells (13, 23). Second, the specificity of generating chimeric switch regions correlates with the specificity of isotype recombination as shown by the stimulation of B lymphocytes with IL-4. This cytokine normally directs the isotype switch to Ig1 (20), and nuclear proteins from IL-4-stimulated B lymphocytes preferentially use S1 RNA to generate switch region hybrids. This specificity may be explained by the induction of different reverse transcriptases that use distinct switch sequences as their templates. Alternatively, IL-4 could regulate the expression of proteins that specifically stabilize the interaction between Sµ DNA and S1 or S3 RNA. Along this line, we recently detected RNA-binding proteins that distinguish between the S1 and S3 sequence in nuclear extracts from stimulated B cells (unpublished observation).

The preferential use of specific switch RNAs as templates by cytokine-stimulated B cells implies a second level of regulation of switch isotype specificity in addition to the well-described effects of specific cytokines on the induction of germline transcripts. IL-4 is known to up-regulate germline S1 transcription. The effect of IL-4 on Ig3 germline transcription is less clear. Some results indicate that the cytokine induces a down-regulation of switching to Ig3 without affecting Ig3 germline transcription (24, 25), although transcriptional down-regulation was observed if T cell-depleted spleen cells were used (26). Similarly, both IL-10 and IFN- were found to reduce switching to IgG3 without affecting levels of germline 3 transcripts (25, 27). Our results suggest that IL-4 also contributes to the specificity of isotype switching by inhibiting the generation of hybrids that join Sµ to S3. Although the effect of IL-4 on the generation of the Sµ-S3 chimeric sequences that we detected under in vitro conditions were modest (12 and 34%, respectively), the effect may be more pronounced under in vivo conditions.

Using episomal constructs, Daniels and Lieber were able to show that isotype switching occurs only if switch sequences are transcribed in sense orientation (28). This could be a consequence of the fact that antisense transcription generates an RNA that is poorly used for the Sµ template strand-primed generation of switch region hybrids (Fig. 3). However, we observed effective reverse transcription if Sµ nontemplate strand extension products were analyzed.

The reverse transcriptase activity that generates chimeric switch DNA was not exclusively detected in switching B lymphocytes. Although this process is induced in B cells upon activation with LPS or CD40L, substantial activities were also found in untreated B cells and in fibroblasts. This suggests that the generation of these DNA hybrids is not rate limiting for the generation of switch recombinations. More likely candidates to control this process appear to be 1) a switch sequence-specific endonuclease (10) and 2) the presence of switch region germline transcripts (4). The lack of both the endonuclease and the switch germline transcripts in untreated B cells and in fibroblasts should prevent the generation of switch recombinations. Our experiments employed linear Sµ DNA substrate (i.e., no endonucleolytic step was necessary to generate chimeric switch DNAs). No generation of chimeric switch DNA was observed in the absence of in vitro-transcribed RNA. Germline transcript RNA that was likely recovered as a contaminant in our nuclear preparations could have been used, theoretically, for the generation of chimeric DNA in our assay. However, the concentration of this RNA was likely to be very low, since most RNA should have been hydrolyzed during the nuclear protein preparation. In addition, germline transcripts account for only a minor species of all nuclear RNAs.

The nature of the reverse transcriptase that generates the described chimeric DNAs is unclear. It has been reported that CD40L treatment of spleen cells leads to an expression of retroviral sequences (so-called intracisternal A particles) that encode reverse transcriptases and endonucleases (29-31). Both CD40L and LPS activation up-regulates early retrotransposon (Etn (32)) expression with a peak around day 3 and day 2, respectively (J. R. Müller, unpublished observation), coinciding with the first appearance of switched cells (1). We cloned a cDNA expressed from an Etn long terminal repeat that contained an open reading frame for a predicted protein with homology to known reverse transcriptases (GenBank accession No. AF016710). However, there is no known function for this putative protein. As a theoretical alternative to retrotransposon-related enzymes, the reverse transcriptase step could be mediated by a telomerase. Both switch regions and telomers have high G contents in their sequences. Furthermore, evolutionary conservation has preserved the location of the heavy chain locus at the ends of their respective chromosomes in man, mouse, and rat.

We hypothesize that the switch cDNAs are used for homologous recombination with the respective downstream switch loci to form switch circles (Fig. 1). This hypothesis predicts that germline transcripts could induce switching even if expressed in trans. In an attempt to find in vivo evidence for such a mechanism, we expressed S RNA in trans in a B cell line that does not normally switch to IgA expression. Sµ-S switch circle recombinations were detected only in cells that expressed S RNA. An allelic difference between the transfected and endogenous S sequences was used to demonstrate that the induced switch circles harbored endogenous S sequences adjacent to their recombinations. Several interpretations of this result are possible. Germline transcripts could possess a ribozymatic endonuclease activity to induce switching by generating strand breaks in switch regions. Alternatively, the expressed switch region transcript could be used as a template to generate chimeric switch DNAs as hypothesized in our model (Fig. 1). The 3' end of the newly generated Sµ-primed S cDNA would then invade the endogenous S locus, and strand exchange would presumably proceed until the end of the homology (i.e., the beginning of the Sµ portion of the chimeric switch DNA). Therefore, the generated cDNA would serve merely as a guide for the homologous strand exchange, but its sequence information would be lost.

Since highly sensitive PCR was used to detect these Sµ-S recombinations in BCL₁B₁ cells, their frequencies remain unknown, but this does not detract from their mechanistic significance. However, in apparent discordance with this result, mice heterozygous for I region promoter and exon knock-outs have been reported to display no switch recombination to the cis-targeted C_H allele. Furthermore, transcription of the targeted C_H allele in mice was insufficient by itself for switch recombination in the absence of appropriately regulated sterile C_H transcription and processing. Several speculative explanations can be invoked including: 1) roles for I region promoters and exons to engender S region accessibility for recombination beyond their importance for germline C_H transcription; and/or 2) the need for a "threshold concentration" of in situ processed switch region RNAs to efficiently mediate switch recombination (2, 6, 7). Nevertheless, our observation clearly demonstrates an active role for sterile switch region transcripts in switch recombination.

In summary, the results presented in this paper show evidence for some of the steps of our hypothetical switch model. However, a number of gaps remain: 1) no data are available regarding the occurrence of strand breaks in Sµ; and 2) homologous recombination of the newly generated switch cDNA with its endogenous switch locus appears to be possible, but remains a speculation. Future experiments are planned to address these issues.

	Acknowledgments

 
We thank E. B. Mushinski and D. Balzarano for expert technical assistance. J.R.M. gratefully acknowledges Dr. M. Potter for his support during the course of these studies.

	Footnotes

 
¹ This work was supported in part by U.S. Public Health Service Grant GM26939 (K.B.M.).

² Address correspondence and reprint requests to Dr. Kenneth B. Marcu, Department of Biochemistry and Cell Biology, Life Sciences Building-Room 324, State University of New York, Stony Brook, NY 11794.

³ Abbreviations used in this paper: CD40L, CD40 ligand; Sµ, switch µ region; ddCTP, dideoxy CTP.

Received for publication August 5, 1997. Accepted for publication March 26, 1998.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Stavnezer, S.. 1996. Antibody class switching. Adv. Immunol. 61:79.[Medline]
2. Snapper, C. M., K. B. Marcu, P. Zelazowski. 1997. The immunoglobulin class switch: beyond "accessibility.". Immunity 6:217.[Medline]
3. Shimizu, A, N. Takahashi, Y. Yaoita, T. Honjo. 1982. Organization of the constant-region gene family of the mouse immunoglobulin heavy chain. Cell 28:499.[Medline]
4. Stavnezer-Nordgren, J., S. Sirlin. 1986. Specificity of immunoglobulin heavy chain switch correlates with activity of germline heavy chain genes before switching. EMBO J. 5:95.[Medline]
5. Lutzker, S., F. W. Alt. 1988. Structure and expression of germ line immunoglobulin _2b transcripts. Mol. Cell. Biol. 8:1849.[Abstract/Free Full Text]
6. Bottaro, A., R. Lansford, L. Xu, J. Zhang, P. Rothman, F. W. Alt. 1994. S region transcription per se promotes basal IgE class switch recombination, but additional factors regulate the efficiency of the process. EMBO J. 13:665.[Medline]
7. Lorenz, M., S. Jung, A. Radbruch. 1995. Switch transcripts in immunoglobulin class switching. Science 267:1825.[Abstract/Free Full Text]
8. Zhang, J., A. Bottaro, S. Li, V. Stewart, F. W. Alt. 1993. A selective defect in IgG_2b switching as a result of targeted mutation of the I_2b promoter and exon. EMBO J. 12:3529.[Medline]
9. Dunnick, W., J. Stavnezer. 1990. Copy choice mechanism of immunoglobulin heavy chain switch recombination. Mol. Cell. Biol. 10:397.[Abstract/Free Full Text]
10. Wuerffel, R. A., J. Du, R. J. Thompson, A. L. Kenter. 1998. Immunoglobulin S3 DNA specific double strand breaks are induced in mitogen-activated B cells and are implicated in switch recombination. J. Immunol. 159:4139.[Abstract]
11. Dunnick, W., M. Wilson, J. Stavnezer. 1989. Mutations, duplication, and deletion of recombined switch regions suggest a role for DNA replication in the immunoglobulin heavy-chain switch. Mol. Cell. Biol. 9:1850.[Abstract/Free Full Text]
12. Li, J., G. A. Daniels, M. R. Lieber. 1996. Asymmetric mutation around the recombination break point of immunoglobulin class switch sequences on extrachromosomal substrates. Nucleic Acids Res. 11:2104.
13. Müller, J. R., K. B. Marcu. 1998. Stimulation of murine B lymphocytes induces a DNA exonuclease whose activity on switch-µ DNA is specifically inhibited by other germline switch region RNAs. J. Immunol. 16:3337.
14. Mietz, J. A., J. W. Fewell, E. L. Kuff. 1992. Selective activation of a discrete family of endogenous proviral elements in normal BALB/c lymphocytes. Mol. Cell. Biol. 12:220.[Abstract/Free Full Text]
15. Curcio, M. J., M. Belfort. 1996. Retrohoming: cDNA-mediated mobility of group II introns requires a catalytic RNA. Cell 84:9.[Medline]
16. Kehry, M. R., B. E. Castle. 1994. Regulation of CD40 ligand expression and use of recombinant CD40 ligand for studying B cell growth and differentiation. Semin. Immunol. 6:287.[Medline]
17. Yu, J.S., M. X. Wei, E. A. Chiocca, R. L. Martuza, R. I. Tepper. 1993. Treatment of glioma by engineered interleukin 4-secreting cells. Cancer Res. 53:3125.[Abstract/Free Full Text]
18. Dignam, J. D., R. M. Lebovitz, R. G. Roeder. 1983. Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acids Res. 11:1475.[Abstract/Free Full Text]
19. Dunnick, W., G. Z. Hertz, L. Scappino, C. Gritzmacher. 1993. DNA sequences at immunoglobulin switch region recombination sites. Nucleic Acids Res. 21:365.[Abstract/Free Full Text]
20. Bergstedt-Lindqvist, S., P. Sideras, H. R. Macdonald, E. Severinson. 1984. Regulation of Ig class secretion by soluble products of certain T-cell lines. Immunol. Rev. 78:25.[Medline]
21. Bottaro, A., R. Lansford, L. Xu, J. Zhang, P. Rothman, F. W. Alt. 1994. S region transcription per se promotes basal IgE class switch recombination but additional factors regulate the efficiency of the process. Embo J. 13:665.
22. Lorenz, M., S. Jung, A. Radbruch. 1995. Switch transcripts in immunoglobulin class switching. Science 267:1825.
23. Kenter, A. L., J. Tredup. 1991. High expression of a 3'5' exonuclease activity is specific to B lymphocytes. Mol. Cell. Biol. 11:4398.[Abstract/Free Full Text]
24. Esser, C., A. Radbruch. 1989. Rapid induction of transcription of unrearranged S1 switch regions in activated murine B cells by interleukin 4. EMBO J. 8:483.[Medline]
25. Severinson, E., C. Fernandez, J. Stavnezer. 1990. Induction of germ-line immunoglobulin heavy chain transcripts by mitogens and interleukins before switch recombination. Eur. J. Immunol. 20:1079.[Medline]
26. Rothman, P., Y. Y. Chen, S. Lutzker, S. C. Li, V. Stewart, R. Coffman, F. W. Alt. 1990. Structure and expression of germ line immunoglobulin heavy-chain epsilon transcripts: interleukin-4 plus lipopolysaccharide-directed switching to C epsilon. Mol. Cell. Biol. 10:1672.[Abstract/Free Full Text]
27. Shparago, N., P. Zelazowski, L. Jin, T. M. McIntyre, E. Stuber, L. M. Pecanha, M. R. Kehry, J. J. Mond, E. E. Max, C. M. Snapper. 1996. IL-10 selectively regulates murine Ig isotype switching. Int. Immunol. 8:781.[Abstract/Free Full Text]
28. Daniels, G. A., M. R. Lieber. 1995. Strand specificity in the transcriptional targeting of recombination at immunoglobulin switch sequences. Proc. Natl. Acad. Sci. USA 92:5625.[Abstract/Free Full Text]
29. Mietz, J. A., Z. Grossman, K. K. Lueders, E. L. Kuff. 1987. Nucleotide sequence of a complete mouse intracisternal A-particle genome: relationship to known aspects of particle assembly and function. J. Virol. 61:3020.[Abstract/Free Full Text]
30. Nissen-Meyer, J., T. S. Eikhom. 1981. Properties of an intracisternal A-particle-associated endonuclease activity which is stimulated by ATP. J. Virol. 40:927.[Abstract/Free Full Text]
31. Wilson, S. H., E. W. Bohn, A. Matsukage, K. K. Lueders, E. L. Kuff. 1974. Studies on the relationship between deoxyribonucleic acid polymerase activity and intracisternal A-type particles in mouse myeloma. Biochemistry 13:1087.[Medline]
32. Shell, B., P. Szurek, W. Dunnick. 1987. Interruption of two immunoglobulin heavy-chain switch regions in murine plasmacytoma P3.26Bu4 by insertion of retroviruslike element ETn. Mol. Cell. Biol. 7:1364.[Abstract/Free Full Text]

This article has been cited by other articles:

				C. Arnaudeau, L. Rozier, C. Cazaux, M. Defais, D. Jenssen, and T. Helleday RAD51 supports spontaneous non-homologous recombination in mammalian cells, but not the corresponding process induced by topoisomerase inhibitors Nucleic Acids Res., February 1, 2001; 29(3): 662 - 667. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Müller, J. R. Articles by Marcu, K. B. Search for Related Content PubMed PubMed Citation Articles by Müller, J. R. Articles by Marcu, K. B. Pubmed/NCBI databases UniGene Compound via MeSH Substance via MeSH