International Immunology, Vol. 3, No. 2, pp.


© 1991 Oxford University Press 0953-8178/91 $3.00

Nuclear protein binding to octamer motifs in the immunoglobulin 7I switch region Cynthia L. Schultz1, Laura A. Elenich, and Wesley A. Dunnick Department of Microbiology and Immunology, University of Michigan Medical School, Ann Arbor, Ml 48109-0620, USA 'Present address: DNAX Research Institute, Palo Alto, CA 94304, USA Key words. DNase-hypersensitive sites, sterile transcripts, chromatin accessibility

Initiation of the Immunoglobulin heavy chain switch DNA rearrangement event Is thought to involve conversion of the target switch region DNA to an accessible state. Accessibility is likely to be mediated by the binding of regulatory proteins to sequences In or near switch regions. A DNase hypersensitlvity assay was used to recognize possible regions of protein binding in the 7I switch region of the B cell hybridoma 470.25. A strong DNase hypersensitive site was identified 5' of the tandemly repeated ST1 sequences. Data from other laboratories suggest that this hypersensitive site Is associated with switch recombination to 7L However, the 470.25 cell does not express the 7I germllne transcript. A second group of DNase hypersensitive sites within the repetitive portion of the 7I switch region was also identified. A gel retardation assay for protein - DNA interaction revealed a sequence present in several copies in the 7I switch region that specifically binds nuclear proteins. This binding sequence, SG1BS, contains the octanucleotide sequence ATGCAAAA, a 7/8 match to the transcrlptlonal enhancer octamer motif found in Immunoglobulin promoters and the heavy chain enhancer. Binding competition studies of SG1BS demonstrate that both the octamer and flanking sequences are critical for binding. By size- and tissue-distribution, the factors that bind SG1BS are not distinguishable from the previously identified octamer-binding factors OTF-1 and OTF-2. The ability of proteins to bind the ST1 octamer motif is increased 2.3-fold upon IL-4 induction of llpopolysaccharlde-stlmulated B cells.

Introduction Immunoglobulin heavy chain switching is mediated by a deletion event that replaces the expressed heavy chain constant region gene segment with one from downstream (1). Switch regions, tandemly repeated sequences located upstream of each constant region gene segment, include the beginning and end points of switch deletions. Several lines of evidence suggest that switching to a particular constant region gene segment is directed at the molecular level and that induction of accessibility of a switch region precedes the deletion event (2-11). The induction of switch region accessibility is likely to involve frans-acting factors capaoie of recognizing one heavy chain switch region or flanking sequences and causing a conformational change, so that the target switch region is now accessible. We attempted to identify such trans-acting proteins in the nuclei of 470.25 cells. This hybridoma has undergone a successive p7 3 - 7 I switch on the non-expressed chromosome and has a rearranged 71 switch (ST1) region downstream of the expressed Correspondence to: W. A. Dunnick Transmitting editor: P. W. Tucker

73 gene (8,12). Therefore, we hypothesized that 470.25 was immortalized as a hybridoma during a successive switch from 73 to 71. Even though 470.25 may not express all the protein factors required for switch recombination to 71, it may express some of them. To detect such protein factors, we first examined the ST1 chromatin for DNase hypersensitive sites. We assumed that such hypersensitive sites would indicate the location of nonhistone protein binding to the switch region (13). We prepared DNA fragments that spanned the DNase hypersensitive sites and tested these fragments for protein binding by a gel retardation assay. We have identified a strong DNase hypersensitive site upstream of the ST1 region and a group of hypersensitive sites within the S71 region. In addition, proteins binding in vitro to consensus octamer motifs in ST1 were detected. Switch recombination to 71 and the amount of protein binding to the Sr1 octamer motifs are correlated. Relative to total octamer

Received 26 July 1990, accepted 16 October 1990

Downloaded from at Karolinska Institutet on July 19, 2015



Binding to an octamer motif in the y1 switch region

binding proteins, those that bind to the S71 octamer motifs are induced - 2-fold upon treatment of lipopolysaccharide (LPS)stimulated splenic B cells with interteukin 4 (IL-4). Methods DNase hypersensitivity assays 470.25 and EL-4 cells were propogated in Dulbecco's modified Eagle's medium for hybridomas supplemented with 1 0 - 2 0 % FCS. Samples of DNA prepared from DNase I-treated nuclei (14) were incubated with restriction endonucleases obtained from New England Biolabs or Bethesda Research Laboratories. Gel electrophoresis and Southern blotting were done as described (12). The probe used was a Sg/ll fragment (5' of the S71 region), isolated from D7I/EHIO.O by elution from low-melting-point agarose and radiolabeled by nick translation (12).

Nuclear protein extract (8 - 1 2 HQ\ 15) were incubated with 20,000 c.p.m. of radiolabeled oligonucleotide probe in a buffer containing 2.5 HQ poly d l - d C , 5 mM Tris-HCI (pH 7.6), 25 mM KCI, 2.5 mM MgCI2, 0.5 mM EDTA.Na2, 0.5 mM dithiothreitol, 12% glycerol, and 0.05% Triton X-100 at room temperature for 20 - 40 min. The reactions were loaded on a 4% polyacrylamide gel and electrophoresis was carried out at 15 mA for 5 h at room temperature in a running buffer containing 3.3 mM Na-acetate, 6.7 mM Tris-HCI (pH 7.6), and 1 mM EDTA.Na2. The gels were then baked onto silanized glass plates and autoradicgraphed. Oligonucleotides VHoct, M1, and M2 were generous gifts of F. Collins (Department of Human Genetics, University of Michigan Medical School). Nuclear extracts (15) were prepared from T cell-depleted BALB/c spleen cells treated for 4 days with 4 /ig/ml LPS or LPS and 20 ng/ml IL-4 as described (16). DNA footprinting To prepare the 71 switch region DNA binding sequence (SG1BS) probe, one strand of a 49 bp oligonucleotide probe containing SG1 BS was labeled with T4 polynucleotide kinase and annealed to the complementary strand, and unincorporated nucleotides were removed. MVHoct was constructed by inserting the oligonucleotide VHoct into M13mp18. The insert containing VHoct was then recovered by restricting in the pcdyiinker, labeled with Klenow fragment, and gel purified. 1,10-Ortho phenanthroline-copper ion (OP-copper) DNA footprinting was done as described (17). Essentially, a 5 x gel retardation assay was done and the gel was autoradiographed overnight at 4°C. The bands corresponding to the free and the bound DNA were cut out and treated with the chemical reagents. The DNA samples were recovered from the gel slice and run on a sequencing gel. To locate the sequences of interest, an A + G sequence reaction was run on the same gel. Southern blotting Genomic DNA from different strains of mouse was prepared as described (18). The DNA was digested with the indicated restriction enzyme, fractionated, transferred, and hybridized at 50°C in 6xSSC (12). The oligonucleotide probes were endlabeled using the Klenow fragment of DNA polymerase I. The nitrocellulose filter was washed in 6 x SSC, 0.5% SDS at 45°C.

To detect sterile transcripts of the 7I gene, theM13mp8subclone MM58 was used. The insert in MM58 extends from residue 1614 to residue 1790, where the Sg/ll site in the 7I sterile transcript first exon (19) begins at residue 1674. Single-stranded copies of MM58, uniformly labeled with [32P]dATP and pPJdGTP, were prepared using the universal M13 primer and the large fragment of DNA polymerase. After cutting the resulting doublestranded DNAs with EcoRI, the mixture was denatured in formamide at 100°C and fractioned on a 6% acrylamide, 7 M urea gel. The 225 bp probe was eluted from the gel at 55°C for 5 h and recovered by precipitation in ethanol. Aliquots of the probe were mixed with RNA samples, reprecipitated, recovered, dissolved in 20 pi of 0.4 M NaCI, 0.04 M PIPES (pH 6.4), 1 mM EDTA, and 50% formamide, and incubated at 100°C for 2 min and then at 42°C for 24 h The samples were diluted to 0.5 ml with 600 units S1 nuclease/ml, 0.28 M NaCI, 0.03 M Na acetate (pH 4.5), and 3 mM ZnCI2- After incubation at 37°C for 90 min, the samples were extracted with phenol and the aqueous layer was precipitated with carrier tRNA and ethanol. After 3 - 2 0 h, the samples were recovered, dissolved in 10 y\ 50% formamide, incubated in a boiling water bath for 5 min, and fractionated on a 6% acrylamide, 7 M urea gel. Size markers were the Hin\\ fragments of pBR322. The gel was exposed to X-ray film for 16 h.

Results In order to map potential sites of protein binding in the S71 region in 470.25 chromatin, we first used a DNase hypersensrtivity assay. To detect these hypersensitive sites, we used a probe that lies 5' of the ST1 region (Fig. 1). 470.25 DNA includes a single copy of sequences homologous to this probe, those sequences lying downstream of the expressed y3 gene. To locate the hypersensitive sites, a molecular clone of the ST1 region, P7I/EHIO.O (18), was completely digested with H/ndlll and partially digested with BamYW. The resulting products were detected by the same 5' probe (Fig. 1, 'EH10' lanes). Each band in the 'EH10' lanes corresponds to a DNA fragment that begins at the /-//ndlll site 5' of ST1 and ends at one of the BamYW sites in or near the ST1 region. Since the molecular clone and 470.25 DNA have exactly the same arrangement of SamHI sites 5' of the tandemly repeated sequences in S71 (12), the bands in these lanes can be used to accurately map hypersensitive sites in 470.25 chromatin 5' of S71 and to distinguish such sites from those within the tandemly repeated sequences in S 7 1. Several sub-bands are observed in the absence of added DNase I, and when 470.25 nuclei are treated with DNase I. All the sub-bands are probably due to endogenous DNase in the cells active during the harvesting of cells and preparation of nuclei (20). Lyphoid cells are known to include large amounts of DNase, and we took no special precautions to inactivate them. However, there is much evidence to indicate that endogenous DNase and DNase I act at similar, if not identical, hypersensitive sites (20). One very strong hypersensitive site maps upstream of the S71 region. This site is also found in EL-4 chromatin, but with less intense hybridization. It is located very near to the 5' end of the y1 sterile transcript (19), and a very similar DNase I hypersensitive site is induced in normal B cells by LPS and IL-4 (21,22). We used an S1 nuclease protection assay to determine if 470.25 cells

Downloaded from at Karolinska Institutet on July 19, 2015

Gel retardation assays

S1 nuclease analysis of y1 transcripts

Binding to an octamer motif in the y1 switch region EL4

470 EHIO lo 01 I

5 20 |o O.I

I 5 2d|

Bcell IL4





A. 1 2 3 4 5

B. 3 -221


+ TI


-Gyi V2



produced the 7I germline transcript Even though we could detect the transcript in RNA from normal B cells treated with LPS and IL-4 (Fig. 2, lane 5), three different preparations of 470.25 RNA did not contain significant amounts of the transcript. In addition to the strong, upstream hypersensitive site, a group of hypersensitive sites are observed in the repetitive portion of the ST1 region of 470.25; these sites are not detected in EL-4 chromatin. There are at least two possible explanations for this. It may be that EL-4 lacks these sites because proteins that create these sites are specific to 470.25. Alternatively, since EL-4 is derived from a different strain of mouse (C57BL/6 instead of BALB/c), the sequences in the S71 region may differ such that the protein binding sites responsible for the hypersensitivity are lacking (see below). After identifying the sites of DNase hypersensitivity in the repetitive portion of the ST1 region, we tested DNA fragments in this region for possible protein binding in vitro. The S71 region is a 10 kb stretch of tandemly repeated 49 bp sequence elements (18). In general, these 49mers are highly homologous to the consensus 49mer repeat unit. However, at least five 49mers contain a cluster of changes (8/16); these unique interspersed repeat elements are identical to each other (18). Restriction fragments from the central, highly repetitive portion of the ST1 region were used as probes in a gel retardation assay with nuclear extracts from 470.25. Only those fragments that contained one of the unique interspersed 49mers resulted in

Bgl.ll ' 177 bp


Fig. 2. S1 nudease analysis of 7I germline transcripts. (A) S1 nuclease analysis was performed as described in Methods. Hybridization of the single-stranded probe is schematically illustrated at the bottom of the Figure. RNA samples tested were 10 /ig Escherthia coli tRNA (lane 1), 42 /ig total RNA from 470.25 cells treated with LPS and IL-4 (lane 2), 13 fig poly(A)+ RNA from untreated 470.25 cells (lanes 3 and 4), and 4 5 ^g total RNA from B cells treated with LPS and IL-4 (lane 5). The protected fragments smaller than the full insert (177 bp) are due to 'breathing' of the ends of the probe hybridized to the mRNA and subsequent S1 digestion (B) Northern blot analysis of 40 fig of the same 470.25 RNA sample as shown in lane 3 of part (A) reveals hybridization to a Sy3 probe. Both the membrane and secreted forms of 73 mRNA are detected, demonstrating that the RNA sample is intact.

specific complex formation (data not shown). We then made a 23 bp ohgonucleotide probe, SG1BS, that spanned the cluster of changes in these unusual 49mers. We also made a probe, SGIcon, from the analogous position in the 49mer, but with the ST1 consensus sequence (Fig. 3B). Upon incubation of SG1BS with nuclear extracts from the B cell hybridoma, 470.25, two specific DNA-protein complexes were observed, G1B1 and G1B2 (Fig. 3A). Using nuclear extract from the T cell thymoma, EL4, we detected only G1B1. The probe containing the consensus 49mer sequence, SGIcon, did not show formation of any specific complexes with either extract. These results indicate that SG1BS binds specifically to nuclear factors present in both B and T lymphocytes. SG1BS includes the sequence ATGCAAAA, which is a 7/8 match to the octamer concensus sequence (ATGCAAAT) found in the promoter and enhancer of immunoglobulin, as well as other, genes (23-31). In addition, the cell distribution and size of the complexes were similar to that of two well-characterized octamer binding proteins (32-34). In a gel retardation assay, the G1B1 and G1B2 complexes observed with the 23 bp SG1 BS migrated with a mobility identical to that of the ubiquitous octamer binding factor (OTF)-1 and -2 complexes (32 - 34) observed with the 23 bp OCT probe (data not shown; the sequence of the OCT

Downloaded from at Karolinska Institutet on July 19, 2015

Fig. 1. DNase hypersensitivity assay of 470.25 and EL-4 DNA. Nuclei from 470 25 or EL-4 were treated with the indicated concentrations of DNase I, and genomic DNA was isolated and subjected to Southern analysis The BglU fragment derived 5' of the S?1 region that was radiolabeled and used as the probe is indicated by the dashed line. As size markers, a clone of the ST1 region, D7I/EHIO.O, was completely digested with Hind\\\ and partially digested with Sa/nHI and run on the same get The resulting BamHI fragments and their letter names are shown. This ladder can be used to place the hypersensitive sites within or 5' of the repetitive portion of the switch region, indicated by the bold line. 470.25 and EL-4 differ from p-yi/EHIO.O primarily in the number of 49 bp repeat units located 3' of the SamHI fragment VI. Therefore, any sites that map 5' of VI are upstream of the repetitive portion of the switch region, and any sites that map 3' of VI map in the repetitive portion of the switch region The hypersensitive sites found in the repetitive portion of the switch region of 470.25 and not in EL-4 are indicated by an arrow.



Binding to an octamer motif in the y1 switch region competitor

probe: extract:

SGIcon 470







M4 SGIcon

SG1BS 470






VHoct Ml M2




m SGIcon


Fig. 4. Competition for binding to SG1 BS. Nuclear extract from 470.25 was preincubated with 50 ng of unlabeted competitor DNA before addition of radiolabeled SG1BS. The competitors used are indicated. The sequence of the competitors is shown below, aligned with SG1 BS. Dashes represent sequence identity with SG1BS. The octamer sequence is bracketed.



Fig. 3. (A) Gel retardation assays using 1 ng of radiolabeled probe and 8 - 12 /ig of 470 25 or EL-4 nuclear extract, as indicated. The resulting complexes formed using the 23 bpSGIBS probe, G1B1 andG1B2, are identified. (B) Sequence of oligonucleotide probes used in band shift analyses. Identity with SG1 BS is represented by dashes. The 7/8 match to the consensus octamer motif is underlined. The sequence of the OCT probe is that found in the promoter of the expressed VH gene in the Bcl-I cell line.

probe is shown in Fig. 3B). Therefore, we concluded that the proteins involved in the formation of G1B1 and G1B2 were likely to be identical to or related to the OTF-1 and B cell-specrfic OTF-2 factors, respectively. Apparently, the change from T to A at the eighth position of the octamer allows recognition by these octamer binding proteins. To determine which sequences in SG1 BS contribute to binding, we used several mutant digonudeotides as competitors in a gel

retardation assay (Fig. 4). When 50-fold excess of SG1BS was added to a binding reaction containing radidabeled SG1 BS and nudear extract from 470.25, complete competition of binding was observed. An digonudeotide probe containing the octamer from an immunoglobulin heavy chain variable region probe, VHoct, exhibited roughly 50% competition. Mutations in the octamer of VHoct that destroy octamer binding (M1 and M2) also destroy competition with SG1 BS. These results indicate that the octamer is critical for binding to SG1 BS. V ^ t lacks the tenth conserved nucleotide in the decamer ATGCAAATNA, shown to contribute to the strength of binding to the octamer (27). This may explain the modest competition by VHoct, since this conserved nucleotide is present in SG1BS. Octamer-containing digonudeotides that include this A and other conserved sequence motifs flanking the octamer sequences compete more effectively for binding to SG1BS (see below). Mutations in sequences flanking the octamer have been previously shown to affect octamer binding (35,36). Consistent with this, the mutant M3, which contains the octamer from SG1 BS but the flanking sequences from SGIcon, is a very poor competitor for binding to SG1BS. This result suggests that the flanking sequences in SG1BS play a critical role in binding to

Downloaded from at Karolinska Institutet on July 19, 2015


Binding to an octamer motif in the y1 switch region EL4



rH CQ r-4

(N P3 rH










M « Ui

w rH U


U 2 (u

SG1BS. M4 has the flanking sequences of SG1BS in the absence of the octamer and does not compete, consistent with the results implicating the octamer in binding to SG1 BS. From these studies we conclude that, in addition to the octamer motif, the flanking sequences are crucial for binding to SG1 BS. To further test this, we did DNA footprinting of SG1 BS (Fig. 5). The footprint of both complexes formed with 470.25 nuclear extract as well as the footprint of the EL-4 complex are shown in Fig. 5. In all three complexes, the area of protection extends from the penultimate nudeotide of the octamer to 5 bp upstream of the octamer, ending in the hypersensitive nudeotide. We conclude from this experiment that the flanking sequences shown to be important for binding in competition studies are directly involved in binding to the protein factors.

For comparison, the OP-copper DNA footprint of VHoct is shown. The footprint of both complexes formed using this probe (designated arbitrarily as C1 and C2 in Fig. 5) show a different pattern of protection, beginning at the last nudeotide of the octamer and extending 8 bp upstream of the octamer, and there are no hypersensitive sites. To investigate a possible correlation between SG1 BS binding and switch recombination to 71, we determined the amounts of SG1BS binding in the nuclei of T cell-depleted, LPS-treated spleen cells and in the nudei of the same cells additionally treated with IL-4. B cells stimulated with LPS and IL-4 produce germline 71 transcripts (5,19,37) and subsequently switch from ^ to y\ production (38). If SG1BS plays a role in switch recombination to 7 I , one might expect a change in binding to it upon IL-4

Downloaded from at Karolinska Institutet on July 19, 2015

Fig. 5. DNA footprinting of SG1BS and VHoct The probes were synthesized by asymmetric end-labeling of SG1BS-long or MVi^oct (see Methods). A 5 x gel retardation assay was done and gel slices containing either free or bound complexes were recovered. The DNA was treated with the OP-copper ion chemical reagents to deave at exposed residues while the DNA was in the gel slice The DNA was then recovered from the acryfamide and run on a sequencing gel with an A + G sequencing ladder of the probe to determine the sequence The complexes analyzed here include G181 and G1B2 from incubation of 470.25 nudear extract with SG1BS-long, G1B1 from incubation of EL-4 extract wrth SG1BS-long, and C1 and C2 from incubation of 470.25 nudear extract wrth MVi^oct. The octamer is indicated by a line alongside the sequence and the hypersensitive nudeotide in the SG1BS complexes is indicated by a bold C


Binding to an octamer motif in the y1 switch region

1 2

probe: OCT IL4= " +











treatment. In LPS-stimulated cells, using the SG1BS probe, we observed the G1B1 and G1B2 complexes, which are the same size as the OTF-1 and OTF-2 complexes observed with the OCT probe (Fig. 6, lanes 1, 3, 5, and 7). A third complex, which migrates faster than the G1B2 (OTF-2) complex, was not observed in two other sets of B cell nuclear extracts. The third complex may be the result of limited proteolytic degradation of the nuclear extracts. To investigate changes in SG1 BS binding with IL-4 treatment, we related the amount of SG1BS binding in a given nuclear extract to the total octamer binding ability of the extract, as determined by binding to the OCT probe. In extracts from cells treated with LPS alone, there is little binding to the SG1BS probe (Fig. 6, lanes 3 and 7). Apparently, a small portion of the total octamer binding proteins in these cells bind SG1 BS well. Relative to the binding in cells treated with LPS alone, SG1 BS binding increases - 4-fold upon IL-4 induction (comparing lanes 7 and 8). When extracts are normalized for differences in total OCT binding, the portion of octamer binding proteins that bind SG1BS well is increased — 2-fold upon IL-4 induction. This small increase was observed in six experiments using three different sets of nuclear extracts (average of 2.3-fold, see Fig. 6 legend). We used SG1 BS as a probe in a Southern Wot analysis of DNA from different strains of mouse to determine whether this sequence was conserved (Fig. 7). All but C57BL/6 showed hybridization of an Ssti fragment of the size expected for the ST1 region (18). The presence of ST1 sequences was verified by washing the filter and rehybridizing it with SGicon (bottom of Fig. 7A). We also hybridized the SG1BS probe to a molecular clone of the intact C57BU6 ST1 region, and again obtained no hybridization (data not shown). However, we detected hybridization to this molecular clone and to pyl/EHIO.O (the BALB/c ST1 region) with control oligonucleotide probes that

CSS OS7Bt S a b




Fig. 7. SG1 BS sequences in various Igh haplotypes. (A) 20 jig of liver DNA from the indicated mouse strains (with Igh haplotypes indicated underneath) were digested with Ssfl and analyzed by Southern hybridization. 4 ^g of C58 liver DNA was analyzed. The probe used was end-labeled SG1BS. The same filter was stripped of hybridization and re-probed with end-labeled SGIcon (inset at bottom of Figure). (B) 20 HQ of DNA from hybndoma 224.4B3 (a 72-expressor with all S71 regions deleted) or BALB/c liver was digested with Ssfl and analyzed by Southern hybridization, using SG1BS as a probe.

matched the consensus 49mer sequence at only 18 of 21 base pairs (e.g. M3). This sensitive test implies that the C57BL/6 ST1 region lacks sequences that are at least 80% similar to SG1 BS.

Downloaded from at Karolinska Institutet on July 19, 2015

Fig. 6. Changes in SG1BS binding with IL-4 treatment. Gel retardation analysis was performed with the indicated radiolabeled probes and aliquots of nuclear extracts from T cell-depleted spleen treated with LPS alone or LPS and IL-4. Two representative experiments are shown (lanes 1 - 4 and 5 - 8 ) . To illustrate the analysis of such data, the amount of total octamer binding was determined by densitometry of lanes 5 (18 75 units), 6 (36.5 units), 7 (10 25 units), and 8 (40.0 units). Hence, the amount of SG1BS binding increases 3.9-fold upon IL-4 induction (lane 8 compared to lane 7). When corrected for the 1.95-fold increase in OCT binding (comparing lane 2 to lane 1), the overall increase was 2.0-fold. In five other experiments, using three different sets of nuclear extracts, the increases in SG1BS binding (corrected for changes in OCT binding) were 1.5-, 2.6-, 1.4-, 2.0-, and 6 1-fold (average 2.3-fold).

Binding to an octamer motif in the y1 switch region Consistent with this conclusion, we have sequenced approximately one-third of the S,1 region from C57BL/6 and have obtained no evidence for octamer-like sequences (data not shown).


We looked for protein binding to the tandemly repeated elements of the ST1 region. Octamer-containing oligonucleotides bound to nuclear proteins in vitro in the gel retardation test (Fig. 3). The octamer motifs, as well as flanking sequences, are required for this binding (Figs 4 and 5). The SG1 BS protein binding is not distinguishable from that of the well-characterized OTF-1 and OTF-2 proteins (33,34,39). Accessibility, as detected by transcription, of an S region is correlated with switching (3-6). It seems possible that the presence of an octamer motif, a strong protein binding site which regulates transcription in immunoglobuhn genes as well as other genes (23-31), would influence accessibility (and therefore DNA rearrangement) of the ST1 region. It is striking that the unusual SG1 BS sequence is accurately repeated in several copies in the S71 region of all but one of the Igh haplotypes investigated (19; Fig. 7). This genomic Southern blot also suggests that SG1BS is found only in the restriction fragment which contains the ST1 region and not elsewhere in the murine genome. SG1 BS does not hybridize to molecular clones of the other murine switch regions, nor does a GenBank search reveal homology to SG1 BS in the switch regions associated with other isotypes (data not shown). Thus, SG1BS is unique to S T 1. Arguing against a role for SG1BS in switch recombination to 71 is the apparent lack of such sequences in ST1 region of C57BL/6 {Igh^. IL4-induced switch recombination to an S 7 1 b region is almost as efficient as IL4-induced switch recombination to an S T 1 a region within the same cell (7,11). This result does not support a central role for SG1 BS in IL-4-driven switch recombination to 7 I . If SG1 BS were involved in regulation of switch recombination, then perhaps in the lghb haplotype switch recombination to 71 is regulated in a different way. Alternatively, SG1 BS may play a dispensable role in switching, e.g. in maintenance of sterile transcription. C57BL/6 mice do produce less lgG1 protein than do BALB/c mice, both before and after immunization (40), which might suggest that the SG1BS motifs result in more efficient 71 expression.

When SG1BS binding is normalized for total octamer binding proteins (as measured by binding to the OCT oligonucleotide), we observed a 2.3-fold increase in SG1 BS binding ability in cells treated with LPS and IL-4 relative to cells treated with LPS alone. The factor of 2 may underestimate changes in SG1 BS binding. If only part of the B cells in the culture were induced to switch to 7I by the LPS and IL-4, then the increase in amount of SG1 BS binding within individual cells that were induced to switch may be > 2-fold. Even though there is no evidence to implicate SG1 BS directly in switch recombination, the ability of octamer binding proteins to recognize it is apparently improved upon IL-4 treatment of T cell-depleted spleen cells (Fig. 6). Binding to SG1 BS involves sequences flanking the octamer (Figs 4 and 5). Its unique flanking sequences may allow proteins to recognize the ST1 octamer motif specifically and alter expression of the 7I gene. Acknowledgements We thank Dr Robert Coffman for his support in the latter part of this study, and Drs P. Rothman and F. Alt for their generosity in providing nuclear extracts and advice on S1 analysts We are grateful to Drs J. L. Claflin, M Impenale, and F. Collins for helpful comments and discussion of the manuscript Also, we would like to thank Dr C. Thompson, Dr D Gumucio, and T Gray for technical advice and materials. This work was supported by Public Health Service grant CA-39068 awarded by the National Cancer Institute

Abbreviations IL-4 LPS OP OTF-1 OTF-2 ST1 region SG1BS

mterleukin 4 lipopolysaccharide 1,10-orthophenanthrotine ubiquitous octamer binding factor B cell-specific octamer binding factor switch region associated with the 71 constant region gene DNA binding sequence in the 7I switch region

References 1 Lutzker, S G and Alt, F. W. 1989. Immunoglobulin heavy-chain class switching. In D. E. Berg and M. M. Howe, eds, Mobile DNA, p 693. ASM Press, Washington, DC. 2 Yancopoulos, G and Alt, F. W. 1985. DevetopmentaTly controlled and tissue specific expression of unrearranged VH gene segments. Cell 40:271. 3 Yancopoulos, G. D., DePinho, R. A., Zimmerman, K. A , Lutzker, S G., Rosenberg, N., and Alt, F. W. 1986. Secondary genomic rearrangement events in pre-B cells: VHDJH replacement by a LINE-1 sequence and directed class switching. EMBO J. 5:3259. 4 Stavnezer-Nordgren, J and Sirlin, S. 1986. Specificity of Ig heavy chain switch correlates with activity of germ-line heavy chain genes prior to switching. EMBO J 5:95. 5 Stavnezer, J., Radcliffe, G , Lin, Y.-C., Nietupski, J., Berggren, L., Sitia, R., and Severinson, E. 1988. Immunoglobulin heavy-chain switching may be directed by prior induction of transcripts from constant-region genes. Proc. Nat! Acad. Sd. USA 85:7704. 6 Rothman, P , Lutzker, S., Cook, W., Coffman, R., and Alt, F. W. 1988. Mitogen plus interieukin 4 induction of Ct transcripts in B lymphoid cells. J. Exp. Mod. 168:2385. 7 Radbruch, A., Muller, W., and Rajewsky, K. 1986. Class switch recombination is lgG1 specific on the active and inactive IgH loci of lgG1-secreting B cell blasts. Proc. Natl Acad. Sd. USA 83:3954. 8 Hummel, M., Berry, J., and Dunnick, W. 1987. Switch region content of hybridomas: the two spleen IgH loci tend to rearrange to the same isotype. J. Immunol. 138:3539.

Downloaded from at Karolinska Institutet on July 19, 2015

The strong DNase hypersensitive site 5' of the ST1 region in 470.25 chromatin maps very closely to the start site for the sterile transcripts of the ST1 region (19); the sterile transcripts are correlated with switch recombination to ST1 (5,19,37,38). Schmitz and Radbruch (21) and Berton and Vitetta (22) have shown that a DNase l-hypersensitive site, mapping at the same location as the 5' hypersensitive site in 470.25 chromatin, is induced in splenic B cells following treatment with IL-4 and LPS (21,22). If the hypersensitive site in 470.25 chromatin is the same as that induced by LPS and IL-4, this implies that 470.25 expresses some of the factors associated with switch recombination to 7 I . Nevertheless, 470.25 cells do not express the 7I germline transcript (Fig. 2), suggesting that the hypersensitive site may be correlated with expression of the germline transcript but can be dissociated from it. In concordance with this suggestion, we see the same (albeit weaker) hypersensitive site in EL4 chromatin.



Binding to an octamer motif in the y1 switch region 25 Bergman, Y., Rice, D , Grosschedl, R., and Baltimore, D 1984. Two regulatory elements for immunogtobulin x fight chain gene expression. Proc. Natl Acad Sci USA 81:7041. 26 Mason, J. O., Williams, G. T , and Neuberger, M. S. 1985. Transcnption cell type specificity is conferred by an Ig VH gene promoter that includes a functional consensus sequence. Cell 41:479. 27 Wirth, T., Staudt, L, and Baltimore, D. 1987. An octamer oligonucleotide upstream of a TATA motif is sufficient for lymphoidspecific promoter activity. Nature 329-174. 28 Mattaj, I. W., Lienhard, S., Jiricny, J., and DeRobertis, E. M. 1985. An enhancer-like sequence within the Xenopus U2 gene promoter facilitates the formation of stable transcription complexes. Nature 316:163. 29 LaBeila, F , Sive, H. L, Roeder, R G., and Heintz, N. 1988. Cell-cycle regulation of a human histone H2b gene is mediated by the H2b subtype-specific consensus element. Genes Dev 2.32. 30 O'Hare, P. and Goding C. R. 1988. Herpes simplex virus regulatory elements and the Ig octamer domain bind to a common factor and are both targets for virion transactivation. Cell 52.435. 31 Rosales, R., Vigneron, M., Macchi, M., Davidson, I, Xiao, J. H., and Chambon, P. 1987. In vitro binding of cell-specific and ubiquitous nuclear proteins to the octamer motif of the SV40 enhancer and related motifs in other promoters and enhancers. EMBO J. 6:3015. 32 Scheideret, C, Heguy, A., and Roeder, R. G. 1987. Identification and purification of a human lymphoid-specific octamer binding protein (OTF-2) that activates transcnption of an Ig promoter in vitro Ceil 51.783. 33 Staudt, L, Singh, H , Sen, R., Wirth, T , Sharp, P. A., and Baltimore, D. 1986. A lymphoid-specific protein binding to the octamer motrf of Ig genes. Nature 323.640 34 Fletcher, C , Heintz, N , and Roeder, R. G 1987. Purification and characterization of OTF-1, a transcription factor regulating cell cycle expression of a human histone H2b gene. Cell 51 773. 35 Ballard, D. W. and Bothwell, A. 1986 Mutational analysis of the immunoglobulin heavy chain promoter region. Proc. Natl Acad. Sci. USA 839626. 36 Baumraker, T., Sturm, R , and Herr, W 1988 OBP100 binds remarkably degenerate octamer motifs through specrfic interactions with flanking sequences. Genes Dev 2:1400. 37 Esser, C and Radbruch, A. 1989. Rapid induction of transcription of unrearranged S 1 switch regions in activated murine B cells by interleukin 4 EMBO J. 8.483. 38 Layton, J. E., Vitetta, E. S., Uhr, J. W., and Krammer, P. H. 1984. Clonal analysis of B cells induced to secrete IgG by T cell derived lymphokines. J. Exp Med. 160-1850 39 Landolfi, N F., Yin, X.-M., Capra, J D , and Tucker, P. W 1988. A conserved heptamer upstream of the IgH promoter region octamer can be the site of a coordinate protein - DNA interaction. Nucleic Acids Res. 16.5503. 40 Barth, W. F., Mclaughlin, C. L, and Fahey, J. L. 1965. The immunoglobulins of mice. VI. Response to immunization. J Immunol. 95781

Downloaded from at Karolinska Institutet on July 19, 2015

9 Winter, E., Krawinke), U., and Radbruch, A. 1987. Directed Ig class switch recombination in activated murine B cells EMBO J. 6:1663. 10 Katzenberg, D. R. and Birshtein, B. K. 1988. Sites of switch recombination in lgG2b- and lgG2a-producing hybridomas. J Immunol 1403219. 11 Kepron, M. R., Chen, Y.-W., Uhr, J. W., and Vrtetta, E S. 1989. IL-4 induces the specrfic rearrangement of 71 genes on the expressed and unexpressed chromosomes of lipopolysaccharide-activated normal murine B cells. J. Immunol. 143334 12 Petrim, J., Shell, B., Hummel, M., and Dunnick, W. 1987. The immunoglobulin heavy chain switch: structural features of 71 recombinant switch regions. J. Immunol. 138:1940. 13 Eissenberg, J. C , Cartwright, I. L., Thomas, G. H., and Elgin, S. C. R. 1985. Selected topics in chromatin structure. Annu Rev. Genet. 19:485. 14 Gottesdiener, K. M., Karpinski, B. A., Lindsten, T., Strominger, J. L., Jones, N. H., Thompson, C B , and Leiden, J M 1988. Isolation and structural characterization of the human 4F2 heavy-chain gene, an inducibte gene involved in T-lymphocyte activation. Mol. Cell. Bid. 83809 15 Dignam, J. D., Lebovrtz, R. M., and Roeder, R. G. 1983. Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acids Res. 11:1475. 16 Coffman, R. L. and Carty, J. 1986. A T cell activity that enhances pdyctonal IgE production and its inhibition by interferon-y J Immunol. 136.949. 17 Kuwabara, M. and Sigman, D. S. 1987. Footpnnting D N A - protein complexes in situ following gel retardation assays using 1,10-phenanthroline-copper ion. E coli polymerase - lac-promoter complexes. Biochemistry 26.7234. 18 Mowatt, M and Dunnick, W 1986 DNA sequence of the murine 7I switch segment reveals novel structural elements. J Immunol. 136.2674 19 Berton, M. T., Uhr, J. W., and Vrtetta, E. S. 1989. Synthesis of germhrte 71 immunoglobulin heavy-chain transcripts in resting B cells: induction by interleukin 4 and inhibition by interferon 7 Proc. Natl Acad. Sci. USA 86.2829. 20 Scott, W. A. and Wigmore, D. J. 1978. Sites in Simian virus 40 chromatin which are preferentially-cleaved by endonucleases. Cell 15:1511. 21 Schmrtz, J. and Radbruch, A. 1989. An interleukin 4-induable DNase I hypersensitive site indicates opening of the 7I switch region prior to switch recombination. Int Immunol V570 22 Berton, M. T. and Vrtetta, E. S. 1990. Interleukin 4 induces changes in the chromatin structure of the 7I switch region in resting B cells before switch recombination. J. Exp. Med. 172375. 23 Parslow, T. G , Blair, D. L, Murphy, W. J , and Granner, D K. 1984 Structure of 5' ends of immunoglobulin genes: a novel conserved sequence. Proc. Natl Acad. Sci. USA 81:2650. 24 Falkner, F. G. and Zachua, H. G. 1984. Correct transcription of an immunoglobulin * gene requires an upstream fragment containing conserved sequence elements Nature 310:71

Nuclear protein binding to octamer motifs in the immunoglobulin gamma 1 switch region.

Initiation of the immunoglobulin heavy chain switch DNA rearrangement event is thought to involve conversion of the target switch region DNA to an acc...
6MB Sizes 0 Downloads 0 Views