Characterization of Oct2 from the Channel Catfish: Functional Preference for a Variant Octamer Motif

This information is current as David A. Ross, Bradley G. Magor, Darlene L. Middleton, Melanie of October 1, 2021. R. Wilson, Norman W. Miller, L. William Clem and Gregory W. Warr J Immunol 1998; 160:3874-3882; ; http://www.jimmunol.org/content/160/8/3874 Downloaded from

References This article cites 41 articles, 17 of which you can access for free at: http://www.jimmunol.org/content/160/8/3874.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on October 1, 2021

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1998 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Characterization of Oct2 from the Channel Catfish: Functional Preference for a Variant Octamer Motif1

David A. Ross,* Bradley G. Magor,2* Darlene L. Middleton,* Melanie R. Wilson,† Norman W. Miller,† L. William Clem,† and Gregory W. Warr3*

The Ig heavy chain enhancer of the channel catfish (Ictalurus punctatus) has an unusual position and structure, being found in the 3؅ region of the ␮ gene and containing eight functional octamer motifs of consensus (ATGCAAAT) and variant sequences. The presence of multiple octamer motifs suggests that an Oct2 homologue may play an important role in driving expression of the Ig heavy chain locus in a teleost fish. To test this hypothesis, two catfish Oct2 cDNAs (␣ and ␤) were cloned by screening a catfish B cell cDNA library. Catfish Oct2 ␣ and ␤ isoforms are derived by alternative RNA splicing; as determined by Southern analysis, Oct2 is a single copy gene. In comparisons with mammalian Oct2, the catfish Oct2 isoforms show high sequence conservation in their N-terminal regions and POU domains, but extensive divergence in their C-terminal regions. Catfish Oct2 ␣ and ␤ are tissue Downloaded from restricted, bind both consensus and variant octamer motifs, and activate transcription in both catfish and murine cells. In contrast, mouse Oct2 activated transcription in mouse but not catfish cells. Catfish Oct2 ␤ is a more potent transcriptional activator than Oct2 ␣. In transient expression assays, catfish Oct2 ␤ showed a marked preference for the octamer variant, ATGtAAAT, which occurs twice in the catfish enhancer. Mouse Oct2 also showed increased activity with the variant octamer when tested in mouse B cells. Gel-shift analysis competition assays indicated that catfish Oct2 binds the consensus octamer motif with an apparently higher affinity than it does the variant motif. The Journal of Immunology, 1998, 160: 3874–3882. http://www.jimmunol.org/

ranscriptional control of the Ig heavy chain (IgH)4 genes although a homologue of the ␦ gene has recently been described in mammals by the intronic enhancer (E␮) appears to in- (4), there is no evidence that the fish IgH locus can undergo typical T volve at least three major mechanisms that drive B cell- class switching via chromosomal rearrangement. In previous stud- specific expression. These include 1) E-box motifs (␮E1-5), which ies of catfish IgH gene expression (5), it was observed that the are found in both promoters and enhancers, and which associate catfish IgH enhancer (E␮3Ј) was not only present in a unique po- with transcription factors such as E47 and TFE3; 2) ␮A(␲) and sition (3Ј of the ␮ gene), but it also had an unusual diffuse struc- ␮B sites in the enhancer, which are bound by Ets-family transcrip- ture, lacking a defined core region and extending approximately tion factors (such as ets-1 and PU.1); and 3) octamer motifs, which 1.8 kb. The catfish appears to lack functional ␮A and ␮B elements, by guest on October 1, 2021 are found in both the promoter and enhancer regions and are bound considered essential to the E␮ of mice (5, 6). Although ␮E3 and by the POU-domain transcription factors (for reviews, see Refs. 1 ␮E5 motifs are present in E␮3Ј, no studies have yet defined their and 2). Thus, the mechanisms by which mammalian B cells control role. E␮3Ј contains eight octamer motifs of six different sequences, expression of IgH genes show extensive diversity, and the manner including the consensus (ATGCAAAT), capable of driving ex- whereby such a complex system evolved is unclear. pression in B cells (6). Octamer motifs are found in the regulatory Teleosts, such as the channel catfish (Ictalurus punctatus), pos- regions of many ubiquitously expressed genes (reviewed in Ref. sess an IgH locus of a similar organizational pattern to that found 7), as well as in elements regulating genes (including Igs) ex- in mammals (reviewed in Ref. 3). While fish VH, D, and JH seg- pressed specifically in B cells (8–13). Despite the structural dif- ments are similar in number and arrangement to those seen in ferences from its mammalian counterparts, the catfish IgH en- mammals, one major difference is in the number of the constant hancer shows a B cell-specific function even when tested in mouse region genes in fish. Catfish express primarily IgM-like Abs, and cells (5). Thus, the catfish IgH enhancer may rely primarily on the octamer motifs, albeit functioning in a different fashion than ob- served in mammals, particularly in the absence of ␮A and ␮B * Department of Biochemistry and Molecular Biology, Medical University of South Carolina, Charleston, SC 29425; and †Department of Microbiology, University of motifs. Mississippi Medical Center, Jackson, MS 39216 The Oct1 (14) and Oct2 (15) transcription factors, of which there Received for publication September 16, 1997. Accepted for publication December are six alternatively spliced Oct2 isoforms (16), play major roles in 10, 1997. murine IgH transcription through interactions with the octamer motif The costs of publication of this article were defrayed in part by the payment of page and coactivators. Although mouse Oct1 and Oct2 possess a very sim- charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ilar DNA binding (POU) domain, composed of a POU-specific do- 1 This research was supported by awards from the National Science Foundation main and a POU homeodomain connected by a linker region (re- (MCB-9406249) and the National Institutes of Health (R37-A1-19530). viewed in Ref. 17), differences in the activation domains confer 2 Current address: Dr. Bradley G. Magor, Hopkins Marine Station, Stanford Univer- different functions on Oct1 and Oct2. Glutamine-rich activation do- sity, Pacific Grove, CA 93950 USA. mains capable of activating transcription from a promoter, but not 3 Address correspondence and reprint requests to Dr. Gregory W. Warr, Department from a remote enhancer, have been identified in the N termini of both of Biochemistry and Molecular Biology, Medical University of South Carolina, Charleston, SC 29425. E-mail address: [email protected] Oct1 and Oct2 (18, 19). In addition, Oct2, but not Oct1, has a proline-, 4 Abbreviations used in this paper: IgH, Ig heavy chain; E␮, intronic enhancer; serine-, threonine-rich activation domain in the C-terminal region EMSA, electromobility shift assays; E␮3Ј, catfish IgH enhancer. (20). This C-terminal activation domain can activate transcription

Copyright © 1998 by The American Association of Immunologists 0022-1767/98/$02.00 The Journal of Immunology 3875

from a promoter and, with B cell specificity, from a remote enhancer (Pharmacia, Uppsala, Sweden), 2 ␮g poly(dI-dC).poly(dI-dC) (Pharmacia), 105 cpm of template (1 to 10 fmol), 3.5 to 5 ␮g nuclear extract (containing (21, 19). A B cell-specific coactivator, variously called Bob-1, OBF-1, ϩ ϩ or OCA-B (22–26), has been shown to associate with the POU-spe- 400 mM NaCl), KCl to a final Na plus K concentration of approxi- mately 120 mM, and unlabeled competitor template where indicated. Be- cific domain of Oct1 (27). Although Bob-1 is highly B cell specific, fore addition of radiolabeled template, the nuclear extract was allowed to it appears not to be the coactivator capable of mediating octamer- prebind the poly(dI-dC).poly(dI-dC) for 10 min at room temperature. The dependent activation from a distant enhancer. Instead, it is believed to binding reaction was allowed to proceed for an additional 20 min at room act through Oct1 and/or Oct2 bound at the promoter (22–24). temperature after addition of radiolabeled template and competitor (as nec- essary). Band supershifting/abrogation was tested by preincubating antisera An important role of the octamer motifs in the catfish IgH en- with nuclear extract for 30 min on ice, before addition of the DNA binding hancer can be inferred from the observation that artificial promot- mix. Samples that did not receive rabbit serum contained 0.25 mg/ml (final ers containing octamer motifs drive strong expression in catfish B concentration) of BSA (New England Biolabs, Beverly, MA). The binding cells (6). To date, Oct2 has been reported only in mammals, while reaction was resolved on a 6% (80:1 acrylamide:bis-acrylamide; Bio-Rad, Oct1 has also been described in birds and amphibians (28, 29). The Hercules, CA) PAGE gel at 10 V/cm and 4°C, with recirculating running buffer. studies reported here were undertaken to test the hypothesis that For immunization of rabbits, N-terminal (ONT, amino acids 2-108) or catfish possess a homologue of Oct2 that can act through multiple, C-terminal (O␤, amino acids 365-480) domains of the catfish ␤ isoform variant octamer motifs to drive B cell-specific transcription of the were expressed in Escherichia coli using pQE-30 vectors (Qiagen, Chats- catfish IgH gene. worth, CA) and purified. Rabbit anti-mouse Oct2 specific for the N-termi- nus (35) was a kind gift of Dr. Philippe Douville (University of Zu¨rich, Switzerland), and control rabbit sera were either anti-bovine lactoperoxi- Materials and Methods dase or preimmunization serum.

Isolation and analysis of Oct2 clones The competition EMSA were performed as specified above, with the Downloaded from addition of unlabeled competitor at the same time as labeled probe. Quan- cDNA libraries from the catfish B lymphoblastoid cell line 1B10 (30) and titation of the percent signal inhibition was by phosphoimaging, as de- the catfish monocyte-like cell line 42TA (31) were constructed in lambda scribed above for the S1 nuclease protection assays. ZAP II (Stratagene, La Jolla, CA). Approximately 2.5 ϫ 105 plaque-form- ing units of the 1B10 cDNA library were screened at low stringency with Plasmids 32P-labeled probes for mouse Oct2 cDNAs (kind gifts of Dr. T. Wirth, University of Wu¨rzburg). For the screening of the 42TA cDNA library, the Construction of the octamer-dependent reporter plasmids has been de- ␣ ␤ same number of plaque-forming units was screened at high stringency with scribed (6). Full-length coding regions of catfish Oct2 and were di- http://www.jimmunol.org/ a 32P-labeled probe for the catfish Oct2 POU domain. Sequences of catfish rectionally cloned (HindIII and NotI) into the expression vector pRc/CMV Oct2 ␣ and ␤ (Y12651 and Y12652), mouse Oct2.1 (X57936), human Oct2 (Invitrogen, San Diego, CA). The murine Oct 2.2 expression vector was a (X13810), human Oct1 (X13403), pig Oct1 (L38524), mouse Oct1 generous gift of Dr. T. Wirth. Plasmids for transfection were purified by (X68363), chicken Oct1 (M29972), frog Oct1 (X57165), and cesium chloride centrifugation and dialyzed extensively against 1ϫ TE (10 dOct2 (M93149) were aligned using the Clustal V program (DNAStar, mM Tris-Cl, 1 mm EDTA, pH 8.0). Madison WI) with PAM 250 residue weight table, gap penalty of 10, and Cell lines and DNA transfection gap-length penalty of 10. The alignment was used to generate most-parsi- monious phylogenetic trees (branch swapping, tree bisection reconnection, The catfish cell lines 1B10, 3B11, F13L-3.1, and 42TA were maintained as 100 bootstrap replicates) in the PAUP program (32). previously described (5, 6). The murine myeloma cell line J558L (Ref. 36; a kind gift of Dr. S. Morrison (UCLA, Los Angeles, CA)) was maintained by guest on October 1, 2021 Southern blots as described previously (5). The S-194 murine myeloma cell line (origi- Genomic DNA was isolated from the erythrocytes of two individual fish. nally described in Ref. 37; a kind gift of Dr. L. Pilstro¨m, Uppsala Univer- Approximately 10 ␮g of DNA was digested to completion with HindIII, sity, Uppsala, Sweden) was kept in RPMI 1640 supplemented with 5% EcoRI, or BamHI and subjected to electrophoresis on a 0.8% agarose gel. bovine calf serum and 7.5% FCS (Life Technologies). Southern blot analysis was performed with a 32P-labeled Oct2 probe Transfections into cell lines were performed as previously described (5), ϫ 6 (AJ003122), as previously described (30). Blots were then exposed to a except for S-194 cells, which were harvested at 4 10 cells/ml and ␮ storage phosphor screen (Molecular Dynamics, Sunnyvale, CA) for 20 h. electroporated at 168V, 1100 F, and 48 ohms. Six micrograms of oc- tamer-dependent reporter plasmid was used, except in F13L-3.1, which S1 nuclease protection required 12 ␮g. Equimolar concentrations (2 to 6 ␮g) of empty or Oct2 expression plasmids were cotransfected in each experiment. As controls for Ј The sequence of the catfish Oct2 probe was 5 - GTGCTGGCTCCACG transfection efficiency, 0.5 to 3 ␮g of luciferase reporter construct, driven TCACCGTTCTTGTGACGCCACGAACCTGTGTTCAGATTTAATGTGT by either an RSV promoter or CMV promoter (used in F13L-3.1, which Ј GAGTGGCCGGAACAGGGCTGGCGTCACTGCGAGGAGACGG-3 and does not express from the RSV promoter) were transfected. Assays for Ј of the catfish actin probe 5 -GGGTCACACCATCACCAGAGTCCATCAC expression of the reporter genes were performed as previously Ј Ј GATACCAGTGGGCATCAACTC-3 . The actin probe was 5 end-labeled described (5). with [␥-32P]ATP (New England Nuclear, Boston, MA), while the Oct2 probe was 3Ј end-labeled using terminal deoxynucleotidyl-transferase (Life Tech- Results nologies, Gaithersburg, MD) and [␣-32P]cordycepin (New England Nuclear). Actin and Oct2 probes were hybridized with total RNA at 43°C for 16 to 18 h. Isolation of an Oct2 homologue in catfish The S1 nuclease digestion (S1-Assay Kit; Ambion, Austin, TX) was per- A cDNA library from the catfish B lymphoblastoid cell line 1B10 formed according to the manufacturer’s instructions, with the exception that 250 U of S1 nuclease was used and digestions were conducted at 46°C for 1 h. was screened with probes containing the coding regions of mouse The protected fragments were resolved on a 12% denaturing polyacrylamide Oct 2.2 and 2.5. Of six positive cDNA clones isolated, three had gel, which was then exposed to a storage phosphor screen (Molecular Dynam- high sequence similarity to mammalian Oct transcription factors ics) and quantified using the Imagequant program (Molecular Dynamics). and were selected for further study. Two of the clones were iden- Electromobility shift assays (EMSA) and antisera production tical and contained open reading frames encoding the sequence termed Oct2 ␣. The third clone contained a partial sequence, re- Template DNA for EMSAs consisted of synthesized oligonucleotides con- lated to that of Oct2 ␣, and reverse transcriptase-PCR permitted taining one copy of the consensus (ATGCAAAT), variant (ATGtAAAT), ␤ or mutant (ACAAAATA) octamer motif (octamer motif underlined): 5Ј- the complete sequence (termed Oct2 ) to be derived. The nucle- CAATATGAATATGCAAATTACCT-3Ј and 5Ј-CATAGGTAATTTG otide sequences of Oct2 ␣ and Oct2 ␤ were identical, except in CATATTCATA-3Ј. Templates were radiolabeled by fill-in with Klenow their 3Ј regions; this was interpreted as most likely resulting from 32 fragment using [␣- P]dATP (New England Nuclear; Ref. 33) to a sp. act. alternative RNA processing events. These two sequences were ini- of 107 to 109 cpm/␮g. Nuclear from each cell line were prepared as described by tially identified as isoforms of catfish Oct2 (rather than Oct1 or Riggs et al. (34). EMSA -DNA binding reactions contained 20 mM other POU-domain transcription factors) because alignments with

HEPES, 1 mM MgCl2, 0.5 mM DTT (Sigma, St. Louis, MO), 4% Ficoll mouse Oct2 revealed clear overall sequence similarities (Fig. 1A). 3876 Oct2 OF THE CATFISH

FIGURE 1. Analysis of catfish Oct2 sequences. A, Comparison of catfish Oct2 ␣ and ␤ sequences with murine Oct 2.1. The amino acid sequences of murine Oct 2.1 (16) were aligned with catfish Oct2 ␣ (Y12651) and Oct2 ␤ (Y12652), as described in Materials and Methods. The POU-specific (POUsp) and POU ho- meodomains (POUhd) are boxed, while a putative glutamine activation domain (as defined in Seipel, Georgiev, and

Schaffner (18), residues 99-161 of mouse Downloaded from Oct 2.1) is overlined. Residues of exact identity are indicated by “.” and gaps in- troduced in the alignment are shown as “-”. B, Unrooted phylogenetic tree show- ing the relationship between various ver- tebrate Oct1, Oct2, and the Drosophila dOct2 sequences. Bootstrap resampling http://www.jimmunol.org/ values (out of 100) are shown at each node. by guest on October 1, 2021

Parsimony-based phylogenetic analysis of several available Oct1 Oct2 ␣ and ␤ are alternatively spliced from a single RNA and Oct2 sequences (Fig. 1B) confirmed that the catfish Oct2 was transcript homologous to mammalian Oct2. In an unrooted tree, the Oct1 and Upon the screening of a catfish monocyte cDNA library, four other Oct2 sequences from cluster on distinct branches. The Oct2 clones were isolated, indicating that catfish Oct2 is not ex- dOct2 of Drosophila served as an outgroup in this analysis since it clusive to B cells. One of these was a 5Ј truncated clone (Fig. 2A) is related to mammalian Oct2 only through its class II POU do- that contained an unprocessed transcript for the 3Ј region of Oct2 main (38). The Oct2 branch, which contains the catfish, ␤. Sequencing of this clone revealed that a single RNA transcript mouse, and human sequences, is strongly supported by a bootstrap can encode both Oct2 ␣ and Oct2 ␤ mRNAs. The C-terminal cod- resampling value of 99 (Fig. 1B). ing region specific for the ␤ isoform is created when 2 ␤-specific Comparisons of the inferred amino acid sequence(s) of catfish exons are spliced into a site immediately 5Ј of the stop codon for Oct2 and those of murine Oct2 (Fig. 1A) show substantial simi- the ␣ isoform. A probe generated from this clone, approximately larity in the N-terminal region, i.e., up to 48% identity depending 2.5 kb (Fig. 2A), was used to detect the Oct2 gene in Southern blot on the isoform of mouse Oct2 chosen for comparison. Catfish Oct2 analysis (Fig. 2B). A single hybridizing band was obtained with exhibits up to 66% identity with the region identified as a glu- BamHI, and two bands with EcoRI and HindIII. The two bands tamine activation domain in mouse Oct2. The POU domains of observed with EcoRI digests were predicted by the presence of an catfish and murine Oct2 are virtually identical, with only a single internal EcoRI site. These results are consistent with the presence amino acid difference in the POU-specific domain and six differ- of Oct2 as a single copy gene: the two bands observed with HindIII ences in the POU homeodomain. However, neither catfish Oct2 ␣ digestion would then represent allelic polymorphism in these sites. nor ␤ shows significant similarity to mammalian Oct2 in the C- terminal regions; in mammalian Oct2, this region contains a pro- Expression of Oct2 isoforms in catfish tissues line, serine, and threonine-rich activation domain (20). While the S1 nuclease protection assays (Fig. 3A) were used to identify the C-terminal regions of catfish Oct2 ␣ and ␤ do not show significant expression patterns of the Oct2 isoforms in catfish cell lines and sequence similarity to mammalian Oct2, they do possess an abun- tissues. The results showed that Oct2 ␣ and ␤ transcripts are both dance of proline, serine, and threonine residues, raising the possi- expressed at relatively high levels in B cells (lanes 3 and 4, Fig. bility that they may also serve as activation domains. 3B). Oct2 ␤ is the predominant isoform in B cells, comprising 71% The Journal of Immunology 3877

of total Oct2 mRNA transcripts. In contrast, both isoforms are expressed at much lower levels in T cells, about one-fifth that observed in B cells (lane 5, Fig. 3B). Oct2 is expressed in the spleen, head kidney (a major lymphohemopoietic organ of catfish), brain, kidney (a resident site for lymphoid cells in catfish), and testes (lanes 8-11 and 13, Fig. 3B). However, the gut (lane 12), skin (lane 14), liver, and muscle (data not shown) are negative for Oct2 mRNA. Thymus was not tested due to the involution of this organ in adult fish. Although the ratio of the ␣ to ␤ isoforms varies between the tissues expressing Oct2, the ␤ isoform predominates in most catfish tissues (Fig. 3B). The extent to which Oct2 expres- sion in catfish nonlymphoid tissues is associated with B cells present in the tissue is unknown.

Octamer binding proteins of catfish B cells The abilities of nuclear extracts from catfish B cells (1B10) and macrophages (42TA) to bind the consensus octamer motif were compared with that of mouse B lineage cells (J558L, a plasmacy- Downloaded from toma) in EMSA (Fig. 4). Comparing the patterns of shifted bands FIGURE 2. Catfish Oct2 is a single copy gene. A, Schematic (to scale) seen with the mouse and catfish B lineage cells, it is apparent that of the unprocessed Oct2 cDNA (AJ003122), containing coding sequence they are very similar, i.e., one distinct slow moving band and a ␣ ␤ Ј ␣ ␤ common to and , the unique 3 regions of and , and the two introns faster migrating band, or envelope of bands in the case of the 1B10 within the 3Ј region of Oct2. The stop codons and poly(A)/cleavage sites cells (cf lanes 1 and 5 in Fig. 4). These bands were interpreted as for the Oct2 ␣ and ␤ messages are indicated. PCR primers (F ϩ R) used being attributable to Oct1 and Oct2, respectively, in the case of

to generate the probe for catfish genomic Southern analysis are indicated http://www.jimmunol.org/ below the appropriate regions. B, Genomic Southern blot analysis of catfish J558L. The nature of the Oct2-shifted bands for the J558L extracts Oct2. DNA from two individual was digested with HindIII (H3), was confirmed by supershifting with anti-Oct2 Ab (cf lanes 3 and EcoRI (R1), and BamHI (Bam). DNA size standards (bp) are indicated to 4 in Fig. 4). In comparing the gel shifts obtained with catfish B cell the right of the figure. (1B10) and macrophage (42TA) extracts, it is clear that the faster by guest on October 1, 2021

FIGURE 3. Oct2 mRNA expression in catfish. A, Schematic showing the frag- ments protected from S1 nuclease diges- tion by Oct2 ␣ and ␤ mRNA. The coding regions, beginning at the POU domain, of catfish Oct2 ␣ and ␤ are shown, with the corresponding protected fragments (sizes in parentheses) of the full-length probe (101 bases), Oct2 ␣ (64 bases), and Oct2 ␤ (91 bases). B, Catfish Oct2 mRNA ex- pression in various tissues as analyzed by S1 nuclease protection. The amount of to- tal RNA loaded in each lane was normal- ized for actin message. Cell lines and tis- sues used are indicated above each lane. To the left is an indication of the ob- served bands, F O2; full-length Oct2 probe, O2 ␤; fragment protected by Oct2 ␤ mRNA, O2 ␣; fragment protected by Oct2 ␣ mRNA, F A; full-length actin probe, A; fragment protected by actin mRNA. Below each lane containing cat- fish RNA is the relative proportion of Oct2 ␣ and ␤ message, as quantified by phosphoimaging. 3878 Oct2 OF THE CATFISH Downloaded from

FIGURE 4. EMSA with consensus octamer motif. These were per- formed using mouse B cell (J558L), catfish B cell (1B10), and catfish monocyte (42TA) nuclear extracts. J558L is shown in lanes 1 to 4, 1B10 in lanes 5 to 8, and 42TA in lanes 9 and 10. Extracts alone (lanes 1, 5, and 9) and 100ϫ molar excess unlabeled competitor (lanes 2, 6, and 10) are http://www.jimmunol.org/ shown for J558L, 1B10, and 42TA. Control antiserum to bovine lactoper- oxidase (lanes 3 and 7) and murine Oct2 antiserum (lanes 4 and 8) are shown only for J558L and 1B10. The Oct2-specific bands, Oct2-super- shifted bands (lane 4), and the band predicted to be produced by the ho- mologue of Oct1 are indicated. migrating bands were more prominent in the B cells than in the macrophages (cf lanes 5 and 9 in Fig. 4). This result is interpreted by guest on October 1, 2021 as indicating that the faster migrating envelope of bands is pro- duced by the catfish Oct2, a situation tested directly by using rabbit antisera to recombinant N-terminal and C-terminal regions of the catfish Oct2 molecules in an EMSA supershift experiment. As shown in Figure 5A, antisera to both the N-terminal (ONT) and ␤ C-terminal (O ) regions produced specific supershifts only of the FIGURE 5. EMSA with consensus and variant octamer motifs. EMSA bands identified as Oct2. Rabbit antisera to murine Oct2 showed was performed using catfish B (1B10) cell nuclear extracts and the con- no cross-reactivity with catfish Oct2 and vice versa (lanes 4 and 8 sensus octamer (A) or the octamer variant, ATGtAAAT (B). For both A of Fig. 4, and data not shown). The variant octamer motif (ATG and B: lane 1, extract alone; 2, 100ϫ molar excess unlabeled competitor; tAAAT) found in the catfish enhancer, which drives strong ex- 3, prebleed to ONT (N-terminus) antiserum; 4, ONT antiserum; 5, prebleed pression in catfish B cells (6), was also tested in EMSA with cat- to O␤ (C-terminus) antiserum; 6,O␤ antiserum. The Oct2-specific bands, fish B cell extracts (Fig. 5B). The pattern of shifted bands seen Oct2-supershifted bands (lanes 4 and 6), and the band predicted to be with the variant octamer was very similar to that seen with the shifted by the homologue to Oct1 are indicated in A and B. An * indicates the envelope of bands specifically recognized by the O␤ antiserum, but of consensus motif, although EMSA with the variant motif produced unidentified nature (B, lane 6). distinct bands clearly visible within the Oct2 envelope, and the additional group of fast migrating bands (marked with an asterisk (*) in Fig. 5B) was more prominent at equivalent exposures. The putative Oct1 band and one of the bands within the Oct2 envelope induction of an octamer-dependent reporter plasmid cotransfected failed to show supershifts with antisera to catfish Oct2 N or C with vectors expressing murine Oct2, catfish Oct2 ␣, catfish Oct2 termini (Fig. 5). However, the fastest migrating group of bands (* ␤, or the empty parental vector (expressing nothing). Reporter in Fig. 5B, lane 6) was shifted by antisera to catfish Oct2 C ter- plasmids containing either the consensus or variant octamer were mini, suggesting considerable diversity in the octamer binding pro- initially compared in catfish B cells (1B10, Fig. 6A). When induc- teins of catfish. tion was set relative to 1 for the reporter construct and empty expression vector (to normalize for activation due to endogenous Catfish Oct2 has a functional preference for a variant octamer Oct2), none of the transcription factors induced high levels of ex- Previously, it has been shown that reporter constructs containing a pression from reporter plasmids containing the consensus octamer trimer of octamer motifs are highly active in catfish relative to (Fig. 6A, black bars). Catfish Oct2 ␤ showed approximately 3.5- those lacking octamer motifs; in particular, the octamer variant, fold activation over that of the empty expression vector (no dif- ATGtAAAT, had far greater activity than did the consensus (6). ferences were observed between cotransfected pRc/CMV or pBS, Transcriptional activation by exogenous Oct2 was measured by data not shown); Oct2 ␣ was somewhat less active, with murine The Journal of Immunology 3879

Oct2 showing the least activity. However, the results obtained us- ing the reporter plasmid with the variant octamer (ATGtAAAT) were in contrast to those with the consensus octamer; i.e., Oct2 ␤ showed the strongest activation (about 9-fold), whereas Oct2 ␣ was much weaker and mouse Oct2 was essentially inactive. These results indicate that, when measured in catfish B cells, Oct2 ␤ was the most active of the transcription factors tested. Furthermore, it showed a preference for the variant over the consensus octamer motif. These reporter constructs are B cell specific, as judged by their extremely low activity when transfected into T cells without additional vectors expressing Oct2 transcription factors (data not shown). When the activities of the transcription factors were tested with octamer-dependent reporter constructs in a catfish T cell line, the relative strength of Oct2 ␤ was again very clear (Fig. 6B). First, Oct2 ␤ showed a higher activation of expression than Oct2 ␣ or mouse Oct2 from reporter constructs containing either con- sensus or variant octamer motifs. Second, Oct2 ␤ showed a 28-fold higher activation from the reporter containing the variant octamer

than from the reporter with the consensus octamer. To test the Downloaded from influence of the species of host cell, consensus and variant reporter constructs along with vectors expressing Oct2 ␣ and ␤ and mouse Oct2 were transfected into a mouse plasmacytoma cell line, S-194. The results (Fig. 6C) showed that each of the three tested tran- scription factors was able to induce expression of an octamer-

dependent reporter gene. Unexpectedly, each of the transcription http://www.jimmunol.org/ factors, including the mouse Oct2, showed a greater degree of enhancement when the variant octamer motif (as opposed to the consensus) was used in the reporter construct.

Catfish Oct2 does not have a higher affinity for the variant octamer The relative affinity of catfish Oct2 for consensus or variant oc- tamer motifs was determined by competitive EMSA. Nuclear ex- tracts from the 1B10 cell line were allowed to bind either labeled by guest on October 1, 2021 consensus or variant octamer motifs, with increasing amounts of unlabeled consensus or variant octamer as competitor. Radioactiv- ity in the Oct2-specific band was then quantitated (Fig. 7). For Oct2 bound to the consensus (Fig. 7A) or variant (Fig. 7B) probes, the consensus octamer motif was a better competitor than was the variant motif, indicating that catfish Oct2 does not have a higher relative affinity for the variant octamer. The mutant octamer motif was unable to compete for Oct2 binding (Mu, Fig. 7), even at 32-fold excess, at which both consensus and variant motifs showed 80 to 90% signal inhibition of probe.

Discussion Previous studies indicated that Oct1 transcription factors are de- tectable in the chicken and , in both cases showing high similarity to mammalian Oct1 (28, 29). However, the present study represents the first report of Oct2 in a nonmammalian species, confirming that the Oct family of transcription factors has been

FIGURE 6. Induction of an octamer-driven reporter gene by Oct2 in catfish B cells (A), catfish T cells (B), and mouse B cells (C). Reporter constructs containing a minimal promoter with a trimer of either consensus (ATGCAAAT, black) or the variant (ATGtAAAT, cross-hatched) octamer motifs were cotransfected with vectors expressing murine Oct2, catfish Oct2 ␣, or catfish Oct2 ␤ into (A) the catfish B lymphoblastoid cell line 1B10, (B) the catfish F13L-3.1 T cell line, and (C) the mouse S-194 B cell line. Induction of the reporter construct is shown relative to expression seen with a cotransfected empty parental expression vector, which was arbi- trarily set at one. Values given as mean Ϯ SEM are the results of at least five replicate transfections. 3880 Oct2 OF THE CATFISH

FIGURE 7. Oct2 binding of octamer motifs: com- petition between consensus and variant sequences. A quantitative EMSA was conducted using consensus (A) or variant (B) probes in competition with consen- sus and variant octamer motifs. A, Nuclear extracts from the catfish B cell line 1B10 were incubated with a labeled consensus octamer (ATGCAAAT) probe and Downloaded from competed with increasing amounts (shown as molar ratios) of variant, consensus, or mutant (Mu) octamer motifs. Only the Oct2-specific bands are shown (as identified in Fig. 5). The relative % signal inhibition is plotted against the molar ratio of the unlabeled variant or consensus competitor. B, Experiment as in A, but http://www.jimmunol.org/ with labeled variant octamer (ATGtAAAT) probe. Each data point is shown as the mean of two indepen- dent experiments. by guest on October 1, 2021

highly conserved in the vertebrates. The POU domain of catfish that additional (yet to be defined) isoforms of catfish Oct2 may Oct2 is virtually identical to that of mammalian Oct2 proteins. exist comes both from EMSA (Fig. 5B), wherein specific bands The high degree of similarity with mammalian Oct2 also ex- can be supershifted with Abs to Oct2, and from reverse tran- tends to other regions, including the N-terminal region, the pu- scriptase-PCR analyses of catfish B cell mRNAs (unpublished tative glutamine-rich activation domain, and the linker region of observations). However, the ␣ and ␤ isoforms described in the the POU domain. The identifiable homology between mamma- present study appear to be the major Oct2 isoforms expressed, lian and catfish Oct2 breaks down in the C-terminal region, and as judged by their relative abundance in the cDNA library the sequence in catfish becomes unique: the C-terminal region screen and the presence of only the predicted protected bands in common to catfish Oct2 ␣ and ␤ isoforms is rich in proline, S1 nuclease assays (Fig. 3). serine, and threonine residues, suggesting that it may function A major finding of the present study is the preference of Oct2 ␤ as an activation domain. The observation that the ␤ isoform for the variant octamer ATGtAAAT. Previous studies have shown (which differs from the ␣ isoform by only a 34-amino acid that the variant octamer is preferred over the consensus when C-terminal extension) has substantially greater transcriptional transfected into catfish B cells in the context of an artificial pro- activity than the ␣ isoform supports the notion that the C-ter- moter (6); the present results show that this finding is most likely minal region of catfish Oct2 contains an activation domain. attributable to the preference of catfish Oct2 for this motif. This However, formal proof for such a functional domain in the cat- functional preference of Oct2 ␤ for the variant is apparent from fish Oct2 C-terminus requires further study. studies using B cells of both mouse and catfish, but is particularly Conservation between mammalian and catfish Oct2 extends striking (28-fold) when assayed in catfish T cells. Presumably, the to the pattern of RNA expression, and the existence of multiple low levels of endogenous Oct2 expressed in F13L-3.1 T cells per- RNA isoforms, generated from a single gene (39). Six Oct2 mit measurement of the full effects of the ectopically expressed isoforms have been described in the mouse (16), whereas only . It is surprising that mouse Oct2 also has a two were described for catfish in the present study. Evidence slight preference for the variant octamer (ATGtAAAT), albeit this The Journal of Immunology 3881 preference by mouse Oct2 is most likely not physiologically - References evant because the octamer motifs found in mammalian Ig promot- 1. Calame, K., and S. Ghosh. 1995. Regulation of immunoglobulin gene transcrip- ers and enhancers are typically of the consensus sequence. The tion. In Immunoglobulin Genes. T. Honjo and F. W. Alt, eds. Academic Press, only octamer variants of the sequence ATGtAAAT found in San Diego, p. 397. mouse and human Ig genes are in germ-line ␬ promoters (40, 41). 2. Ernst, P., and S. T. Smale. 1995. Combinatorial regulation of transcription II: the immunoglobulin ␮ heavy chain gene. Immunity 2:427. However, these promoters function independently of their octamer 3. Warr, G. W. 1995. The immunoglobulin genes of fish. Dev. Comp. Immunol. motifs (42). Conversely, the catfish IgH locus has been shown to 19:1. have this motif twice in the enhancer, and the deletion of one of 4. Wilson, M., E. Bengte´n, N. W. Miller, L. W. Clem, L. Du Pasquier, and G. W. Warr. 1997. A novel chimeric immunoglobulin heavy chain from a teleost these motifs from an enhancer fragment results in a loss of 66% of fish shares similarities to IgD. Proc. Natl. Acad. Sci. USA 94:4593. enhancer activity (5). In addition, Tanaka (43) has shown that 5. Magor, B. G., M. R. Wilson, N. W. Miller, L. W. Clem, D. L. Middleton, and G. W. Warr. 1994. An Ig heavy chain enhancer of the channel catfish Ictalurus binding of a transcription factor to its target is cooperative, i.e., punctatus: evolutionary conservation of function but not structure. J. Immunol. dependent not only on the interaction of the DNA binding domain 153:5556. with its motif, but also on the number of binding sites and the 6. Magor, B. G., D. A. Ross, D. L. Middleton, and G. W. Warr. 1997. Functional motifs in the IgH enhancer of the Channel Catfish. Immunogenetics 46:192. transcriptional strength of the activation domain. Consequently, 7. Verrijzer, C. P., and P. C. Van der Vliet. 1993. POU domain transcription factors. the unusual structure of catfish E␮3Ј (with multiple octamer mo- Biochim. Biophys. Acta 1173:1. tifs) could be an ideal configuration in which to achieve high levels 8. Currie, R. A., and R. G. Roeder. 1989. Identification of an octamer-binding site in the mouse kappa light-chain immunoglobulin enhancer. Mol. Cell. Biol. of transcriptional activity from a single species of transcription 9:4239. factor (i.e., Oct2) binding to multiple octamer motifs. 9. Kistler, B., P. Pfisterer, and T. Wirth. 1995. Lymphoid- and myeloid-specific Several explanations are possible for the functional preference activity of the PU.1 promoter is determined by the combitorial action of octamer Downloaded from ␤ and ets transcription factors. Oncogene 11:1095. of catfish Oct2 for the variant octamer. First, this may relate to 10. Lieberson, R., S. L. Giannini, B. K. Birshtein, and L. A. Eckhardt. 1991. An its ability to interact either with components of the basal transcrip- enhancer at the 3Ј end of the mouse immunoglobulin heavy chain locus. Nucleic tional complex or with coactivators when bound to a particular Acids Res. 19:933. 11. Matthias, P., and D. Baltimore. 1993. The immunoglobulin heavy chain locus target sequence. Viral proteins can associate with the POU do- contains another B-cell-specific 3Ј enhancer close to the ␣ constant region. Mol. mains of Oct1 in a manner that is dependent on the sequence of the Cell. Biol. 13:1547. motif to which Oct1 is bound, presumably reflecting variations in 12. Michaelson, J. S., S. L. Giannini, and B. K. Birshtein. 1995. Identification of http://www.jimmunol.org/ 3Ј␣-hs4, a novel Ig heavy chain enhancer element regulated at multiple stages of conformation of the POU domain (44). Of direct interest are the B cell differentiation. Nucleic Acids Res. 23:975. observations of Gstaiger et al. (27) and Cepek, Chasman, and 13. Parslow, T. G., D. L. Blair, W. J. Murphy, and D. K. Granner. 1984. Structure of the 5Ј ends of immunoglobulin genes: a novel conserved sequence. Proc. Natl. Sharp (45), wherein the stability of the DNA/Oct/Bob-1 ternary Acad. Sci. USA 81:2650. complex is dependent on both the octamer sequence and the con- 14. Fletcher, C., N. Heintz, and R. G. Roeder. 1987. Purification and characterization text in which the sequence occurs. Second, the POU domain of of OTF-1, a transcription factor regulating cell cycle expression of a human histone H2b gene. Cell 51:773. catfish Oct2 may have an intrinsically higher affinity for the variant 15. Scheidereit, C., A. Heguy, and R. G. Roeder. 1987. Identification and purification motif than for the consensus octamer. This altered affinity would of a human lymphoid-specific octamer-binding protein (OTF-2) that activates likely result from effects other than differences in the residues di- transcription of an immunoglobulin promoter in vitro. Cell 51:773. 16. Wirth, T., A. Priess, A. Annweiler, S. Zwilling, and B. Oeler. 1991. Multiple by guest on October 1, 2021 rectly contacting DNA, since these residues are known for human Oct2 isoforms are generated by alternative splicing. Nucleic Acids Res. 19:43. Oct1 (46) and are invariant between mammalian Oct1 and Oct2 17. Herr, W., and M. A. Cleary. 1995. The POU domain: versatility in transcrip- and catfish Oct2. While the POU domains are very flexible, and tional regulation by a flexible two-in-one DNA-binding domain. Genes Dev. 9:1679. hence can bind to many variants of the octamer motif (reviewed in 18. Seipel, K., O. Georgiev, and W. Schaffner. 1992. Different activation domains Ref. 17), the effects on DNA binding caused by sequence changes stimulate transcription from remote (’enhancer’) and proximal (’promoter’) po- at positions distant from the DNA-contacting residues are possible. sitions. EMBO J. 11:4961. 19. Friedl, E. M., and P. Matthias. 1995. Transcriptional activation and repression, For example, the linker region of the POU domains has been re- two properties of the lymphoid-specific transcription factor Oct-2a. Eur. J. Bio- ported to influence DNA binding specificity of POU domain fac- chem. 234:308. tors (47). The sequence of the linker region of the catfish differs by 20. Tanaka, M., and W. Herr. 1990. Differential transcriptional activation by Oct-1 and Oct-2: interdependent activation domains induce Oct-2 phosphorylation. Cell 34% from that of the mouse, raising the possibility that it may 60:375. affect the POU domain of catfish Oct2 such that the variant oc- 21. Annweiler, A., S. Zwilling, and T. Wirth. 1994. Functional difference between the Oct2 transactivation domains determine the transactivation potential of indi- tamer is favored. In light of the fact that catfish Oct2 does not have vidual isoforms. Nucleic Acids Res. 22:4250. a demonstrably higher affinity for the variant motif, the first ex- 22. Strubin, M., J. W. Newell, and P. Matthias. 1995. OBF-1, a novel B cell-specific planation for the octamer preference is favored, i.e., that catfish coactivator that stimulates immunoglobulin promoter activity through association with octamer-binding proteins. Cell 80:497. Oct2 binding to the ATGtAAAT variation of the octamer motif 23. Gstaiger, M., L. Knoepfel, O. Georgiev, W. Schaffner, and C. M. Hovens. 1995. leads to differentially favorable interactions with other components A B-cell coactivator of octamer-binding transcription factors. Nature 373:360. of the transcriptional complex. 24. Luo, Y., and R. G. Roeder. 1995. Cloning, functional characterization, and mech- anism of action of the B-cell-specific transcriptional coactivator OCA-B. Mol. It can be concluded that throughout the approximately 350 mil- Cell. Biol. 15:4115. lion years of evolution separating teleosts from mammals, Oct2 25. Pfisterer, P., S. Zwilling, and T. Wirth. 1995. Functional characterization of the transcription factors have remained an integral part of the regula- murine homolog of the B cell-specific coactivator BOB.1/OBF.1. J. Biol. Chem. 270:29870. tion of the IgH locus. While catfish Oct2 bridges this extensive 26. Schubart, D. B., P. Sauter, S. Massa, E. M. Friedl, H. Schwarzenbach, and evolutionary distance, functioning well in both catfish and mouse P. Matthias. 1996. Gene structure and characterization of the murine homologue B cells, murine Oct2 does not drive octamer-dependent transcrip- of the B cell-specific transcriptional coactivator OBF-1. Nucleic Acids Res. 24: 1913. tion in catfish B cells. Explanations for these observations may lie 27. Gstaiger, M., O. Georgiev, H. van Leeuwen, P. van der Vliet, and W. Schaffner. in the interactions of Oct2 with other transcription factors and/or 1996. The B cell coactivator Bob1 shows DNA sequence-dependent complex formation with Oct-1/Oct-2 factors, leading to differential promoter activation. coactivators present in the B cells of these species. Thus, although EMBO J. 15:2781. the structural relationship between mouse and catfish Oct2 is ev- 28. Petryniak, B., L. M. Staudt, C. E. Postema, W. T. McCormack, and ident, the manner in which Oct2 functions has apparently changed C. B. Thompson. 1990. Characterization of chicken octamer-binding proteins demonstrates that POU domain-containing transcription factors have as the teleost and tetrapod lineages diverged and is not yet been highly conserved during vertebrate evolution. Proc. Natl. Acad. Sci. USA understood. 87:1099. 3882 Oct2 OF THE CATFISH

29. Hinkley, C., D. Leibham, and M. Perry. 1991. Regulated binding specificity of 39. Hatzopoulos, A. K., A. S. Stoykova, J. R. Erselius, M. Goulding, T. Neuman, and Xenopus OCT-1, a maternal transcription factor. GenBank/EMBL Accession no. P. Gruss. 1990. Structure and expression of the mouse Oct2a and Oct2b, two X57165. differentially spliced products of the same gene. Development 109:349. 30. Miller, N. W., M. A. Rycyzyn, M. R. Wilson, G. W. Warr, J. P. Naftel, and 40. Van Ness, B. G., M. Weigert, C. Coleclough, E. L. Mather, D. E. Kelly, and L. W. Clem. 1994. Development and characterization of channel catfish long term R. P. Perry. 1981. Transcription of the unrearranged mouse C␬ locus: sequence B cell lines. J. Immunol. 152:2180. of the initiation region and comparison of activity with a rearranged V␬-C␬ gene. 31. Vallejo, A. N., C. F. Ellsaesser, N. W. Miller, and L. W. Clem. 1991. Sponta- Cell 27:593. neous development of functionally active long-term monocyte-like cell lines from 41. Thompson, A., E. Timmers, M. J. H. Kenter, M. E. M. Kraakman, R. W. Hendriks, channel catfish. In Vitro Cell. Dev. Biol. 27A:279. and R. K. B. Schuurman. 1992. Immunoglobulin ␬ light chain germ-line transcripts 32. Swofford, D. L. 1993. PAUP: phylogenetic analysis using parsimony, version in human precursor B lymphocytes. Eur. J. Immunol. 22:3167. 3.1.1. Illinois Natural History Survey, Champagne, IL. 33. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular Cloning: A Lab- 42. Prabhu, A., D. P. O’Brien, G. L. Weisner, R. Fulton, and B. Van Ness. 1996. oratory Manual. Cold Spring Harbor Laboratories Press, Plainview, NY. Octamer independent activation of transcription from the kappa immunoglobulin 34. Riggs, K. J., S. Saleque, K. K. Wong, K. T. Merrell, J. S. Lee, Y. Shi, and germline promoter. Nucleic Acids Res. 24:4805. K. Calame. 1993. Yin-yang 1 activates the c- promoter. Mol. Cell. Biol. 43. Tanaka, M. 1996. Modulation of promoter occupancy by cooperative DNA bind- 13:7487. ing and activation-domain function is a major determinant of transcriptional reg- 35. Chen, U., R. H. Scheuermann, T. Wirth, T. Gerster, R. G. Roeder, K. Harshman, ulation by activators in vivo. Proc. Natl. Acad. Sci. USA 93:4311. and C. Berger. 1991. Anti-IgM antibodies down modulate mu-enhancer activity 44. Walker, S., S. Hayes, and P. O’Hare. 1994. Site-specific conformational alter- and OTF2 levels in LPS-stimulated mouse splenic B-cells. Nucleic Acids Res. ation of the Oct-1 POU domain-DNA complex as the basis for differential rec- 19:5981. ognition by Vmw65 (VP16). Cell 79:841. 36. Oi, V. T., S. L. Morrison, L. A. Herzenberg, and P. Berg. 1983. Immunoglobulin 45. Cepek, K. L., D. I. Chasman, and P. A. Sharp. 1996. Sequence-specific DNA gene expression in transformed lymphoid cells. Proc. Natl. Acad. Sci. USA 80: binding of the B-cell-specific coactivator OCA-B. Genes Dev. 10:2079. 825. 37. Horibata, K., and A. W. Harris. 1970. Mouse myelomas and lymphomas in cul- 46. Klemm, J. D., M. A. Rould, R. Aurora, W. Herr, and C. O. Pabo. 1994. Crystal ture. Exp. Cell Res. 60:61. structure of the Oct-1 POU domain bound to an octamer site: DNA recognition

38. Prakash, K., X. Fang, D. Engelberg, A. Behal, and C. S. Parker. 1992. dOct2, a with tethered DNA-binding modules. Cell 77:21. Downloaded from Drosophila Oct transcription factor that functions in yeast. Proc. Natl. Acad. Sci. 47. Aurora, R., and W. Herr. 1992. Segments of the POU domain influence one USA 89:7080. another’s DNA-binding specificity. Mol. Cell. Biol. 12:455. http://www.jimmunol.org/ by guest on October 1, 2021