Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Transcription factor Oct‐2A contains functionally redundant activating domains and works selectively from a promoter but not from a remote enhancer position in non‐lymphoid (HeLa) cells.

Transcription factor Oct‐2A contains functionally redundant activating domains and works... EMBO Journal vol.9 no. F25-1634, 1990 The pp. Transcription factor Oct-2A contains functionally redundant activating domains and works selectively a promoter but not from a remote enhancer from in cells position non-lymphoid (HeLa) is also found in promoters and enhancers of other sequence Michael M.MUller-ImmerglUck, Walter cellular and viral genes which display no cell type specificity. Schaffner and Patrick Matthias For example, it is present in the promoters of histone H2B Institut fur Molekularbiologie II der Universitiit Zurich, CH-8093 genes, in the promoters of U2 and U6 small nuclear RNA Zurich, Switzerland genes and the simian virus SV40 enhancer (Falkner et al., Communicated by W.Schaffner 1986; Harvey et al., 1982; Barberis et al., 1987; LaBella et al., 1988; Mattaj et al., 1985; Zenke et al., 1986; Rosales In non-lymphoid cells such as HeLa cells, ectopic et al., 1987; for a recent compilation of octamer sequence- expression of the lymphocyte-specific transcription factor containing genes, see Schreiber et al., 1989a). Oct-2A can reporter genes whose promoters activate Three octamer binding activities have been identified major consist of a single octamer sequence (ATTTGCAT) characterized: Oct-I (also called NF-A1, and thoroughly upstream of a TATA box. While the factor is strongly OTFI, NFIII, OBP100, oct-B3) has an apparent relative in position, it fails as an enhancer active a promoter molecular mass (Mr) of 90 000-100 000 (90-100K) and factor: an enhancer consisting of multiple copies of the has been found in all cell types investigated (Pruijn et al., of the reporter octamer sequence placed downstream 1986; Singh et al., 1986; Staudt et al., 1986; Fletcher et al., is not in cells, even at high concentration gene active HeLa 1987; Rosales et al., 1987; O'Neill and Kelly, 1988; Sturm cells, however, the same of Oct-2A. In B lymphoid et al., 1988). Oct-2A (also called NF-A2, OTF2, oct-B2) highly active. This could mean that an enhancer is has an apparent Mr Of - 60K and has been found in a few additional factor is required for enhancer activation in cell types only, notably in B lymphocytes where its presence B cells. Furthermore, we have tested the transcriptional et al., also correlates with transcription of Ig genes (Staudt activation potential of Oct-2A with a series of N-terminal Gerster al., Rosales et al., 1987; Scheidereit 1986; et 1987; and C-terminal deletions. We show that a glutamine-rich protein et al., 1987). Oct-2B is another lymphoid-specific domain near the N-terminus is required for full activity. - et al., 1988). of 75K and is related to Oct-2A (Schreiber Otherwise, large segments of the N-terminal half or the A number of other octamer binding proteins have been entire C-terminal region are dispensable in our assay, as in various tissues and cell lines: several octamer identified long as the deletions do not impinge on the conserved factors in neural tissues Oct4 -Oct0 (Scholer et al., POU domain which is sufficient for DNA binding. Vhile an embryonal carcinoma (EC) cell line-specific 1989a,b), N-terminal and C-terminal regions can functionally factor NF-A3 et 1989), another protein in (Lenardo al., for each other, a combined deletion that only compensate melanoma (independently also called NF-A3; Cox malignant retains the POU domain is a strong down mutation. We et a factor Oct-T (Barberis et al., al., 1988), testis-specific also fmd that activity depends on the promoter structure and two high mobility group proteins which 1987, 1989) of the reporter gene: the POU domain by itself shows bind both AT rich sequences and the octamer preferentially some with a promoter where the octamer activity and Birnstiel, 1989). sequence (Eckner is located very close to the TATA box, but no sequence the We and others have isolated cDNAs encoding activity with another promoter construction where the Oct-2A. Like the lymphocyte-specific transcription factor further The two octamer is located upstream. sequence RNA is almost protein it codes for, the Oct-2A expressea also respond differently to the deletion of the promoters et Muller- in lymphoid cells (Clerc al., 1988; exclusively stretch important for transcriptional glutamine-rich et Staudt Immergluck et al., 1988; Scheidereit al., 1988; it activation. From these experiments we consider likely Consistent with the et see also He et al., 1989). al., 1988; variants can that the natural octamer factor selectively is for B cell- finding that the octamer sequence important octamer- activate the different naturally occurring et et specific transcription (Mason al., 1985; Dreyfus al., containing promoters. Gerster et 1987; Wirth et al., 1987), the presence 1987; al., Key words: enhancer/Oct-2A(OTF-2)/POU/promoter/tissue- to via of Oct-2A in cells is sufficient activate, non-lymphoid specific transcription/transactivation the octamer sequence motif, lymphocyte-specific promoter Furthermore Oct-2A can et al., 1988). (Muller-Immergluck octamer-related sites also activate containing promoters et The derived from homeotic genes (Thali al., 1988). Introduction of Oct-2A et Ko deduced (Clerc al., 1988; protein sequence et Scheidereit site ATGCAAAT (or its complement et al., 1988; Muller-Immergluck al., 1988; The octamer a of to Oct-i et shows high homology is in all al., 1988) region ATTTGCAT) present immunoglobulin (Ig) as well as to two other Parslow et et tissue-specific and (Sturm al., 1988), promoters (Falkner Zachau, 1984; al., 1984) et et enhancers and Pit-l al., 1988; Ingraham al., 1988) in the chain factors, (Bodner and Ig heavy (IgH) (Falkner The of et et The same octamer and Unc-86 al., 1988). region homology Maeda (Finney Zachau, 1984; al., 1987). Press Oxford University W.Schaffner and M.M.Muller-immergluck, P.Matthias .. -, ~~~~~~~~~~~~~ -.. ~~~~~~~..I. ....~ ". ~ - ,_ f.,i _~~~~~~~~~~~~N 4Au- -- W. 5.A| Sr 1. of the driven various enhancers at the same Fig. Activity OCTA(l) promoter by downstream determined RNase position by protection analysis. Lanes 1, no lanes SV40 lanes lanes one octamer enhancer; 2, enhancer; 3, IgH enhancer; lanes six octamer in tandem. 4, sequence; 5, sequences Promoter in BJAB cells. in HeLa cells. in HeLa cells cotransfected Oct-2A (A) activity structure of the (B) (C) expressing protein. (D) reporter gene scale for the The location where the various enhancers drawn to are inserted is indicated an arrow. except enlarged promoter region. For by sequence of the see 4D. OCTA(l) promoter, Figure among these has been called the POU domain proteins and Results contains two conserved a box and subregions: POU-specific a POU-homeobox et The POU-homeobox In HeLa cells Oct-2A acts as (Herr al., 1988). preferentially a promoter is - 30% to homeoboxes found in transcription factor homologous many genes regulating Previous experiments have shown that development (Gehring, 1987). Mutagenesis multimerized octamer experiments have indicated that the entire POU domain is sites from various origins behave as a lymphoid-specific likely to represent the DNA of this class of enhancer, capable of binding region activating the fl-globin promoter from proteins (Herr et al., 1988; Sturm et The recent an upstream (Tanaka al., 1988). et al., 1988) or downstream position detection of other POU domain in mammalian (Gerster et al., 1987; Tanaka et proteins al., 1988). By cotransfection neural tissues et (He al., 1989) and Caenorhabditis experiments we have recently demonstrated that elegans cloned et (Biirglin al., 1989) could be indicative of a Oct-2A can stimulate complex specifically reporter plasmids network with regulatory overlapping specificities. containing an octamer site in the promoter (Muller- In this paper we evidence that Immergluck et present the cDNA encoded al., 1988). In order to see whether Oct-2A Oct-2A acts as a in can also account for preferentially promoter factor: a the enhancer observed activity with heterologous system, it fails to activate a remote multiple octamer sites in synthetic a downstream we position, enhancer consisting of multiple octamer sites even constructed a series of though plasmids which all contain the the same enhancer is in B highly active cells. This OCTA(1) promoter. This minimal suggests promoter consists of the that either another Oct-2 related factor is responsible for ,3-globin TATA-box and an octamer binding site (Figure 4D) enhancer activation or that and perhaps a second factor is required responds well to transactivation by Oct-2A (Muller- with Oct-2A in together order to activate the Immergluck et enhancer. By al., 1988 and Figures 4B and 5C). In testing deleted versions of Oct-2A for their transactivation the addition, plasmids contain either no enhancer or one of potential, we identified a glutamine-rich region in the N the following enhancers inserted downstream of the fl-globin terminus important for transactivation of two different test gene (Figure ID): the SV40 enhancer, the IgH enhancer, promoters. Two weaker activating domains in the C terminus a single octamer, or six octamer motifs in tandem (1 xOCTA and within the POU domain can only be monitored with one or 6 x OCTA) (Gerster et al., 1987). or the other test promoter. This suggests that a different We first tested the activity of these plasmids in B cells promoter architecture can influence the response to the same (BJA-B). The plasmids containing the SV40 enhancer, the transactivator. IgH or x enhancer, the 6 OCTA enhancer were all highly 1626 transcription factor function Oct-2A OcW-, --40 Oct-2B- - m me _9- ___ 00} Oct-2A- 0e. -- _ -_. - -%. Prnbe _ _ am Io<t; H 2 3 I 5 6 7 - I- 5 1- t) c8 "; .- Fig. 2. Bandshift experiments with nuclear extracts from Cos-7 cells expressing transfected Oct-2A cDNA deleted from the N terminus, C terminus or both ends. Autoradiograms are presented. The numbers above the lanes refer to the first and last Oct-2A amino acid present. (A) N-terminal deletions; (B) C-terminal deletions; (C) deletions from both ends. For a schematic representation of the deletions, see Figure 3. active IA lanes 2, 3 and 5). In HeLa cells, however, of and protein turnover) as well (Figure (as a consequence expression plasmid containing the SV40 enhancer showed a It will henceforth be referred to as only the as binding affinity. measurable activity which can only be seen when the 'bandshift signal'. of reference band is strongly overexposed (Figure iB, lane 2). to the functional domains Oct-2A, Secondly, identify of the mutant This weak activity depends on the octamer site in the the transcriptional activation capacity proteins cells in vivo transactivation of a promoter (Muller-Immergluck et al., 1988) and possibly was tested in HeLa by an octamer motif in the promoter reflects residual activity of endogenous Oct-I protein. As reporter plasmid containing as described can be seen in Figure IC, apart from the SV40 enhancer, and the SV40 enhancer downstream, previously cotransfection of the Oct-2A expression vector leads to (Muller-Immergliick et al., 1988). In several cases we also stimulation of every plasmid to the same extent independently such in CV-1 cells and obtained similar performed assays of the enhancer it contains. Since the basal level of transcrip- results (not shown). tion also increased for the enhancerless construct (Figure 1C, lane 1), the effect is due to weak stimulation of the Transactivation by truncated Oct-2A proteins OCTA(l) promoter by Oct-2A even in the absence of a for transactivation experiment, As control the subsequent We therefore conclude that in our functional enhancer. we first transfected the replication-competent expression functions as a experimental system, Oct-2A specifically vectors into Cos-7 cells to obtain comparably high amounts factor is not able to activate a lymphocyte- promoter and of each protein. This approach not only maps the DNA enhancer from a position far downstream. It was specific domain of Oct-2A, but assays at the same time for binding in B multiple copies of an of the mutant proteins found by others that, even cells, localization in the nucleus and stability motif were more when placed immediately octamer active under in vivo conditions. of a than far downstream (Tanaka et al., Oct-2A deleted upstream promoter As shown in Figure 2, the various proteins In the presented here, both the IgH and both ends were found in 1988). experiments from N terminus, C terminus or the x enhancers are highly active in B cells but not bound to the 6 OCTA the nuclei of Cos-7 cells and specifically in HeLa cells. Since we have not tested intermediate of the nomenclature see octamer probe (for explanation and 'far downstream' positions was observed distances between 'promoter' Materials and methods). Impaired binding only we cannot rule out quantitative, rather than qualitative, box or the POU-homeobox when part of the POU-specific in HeLa cells. Oct-2A/206-357 showed a slightly and deficiency was deleted: a reduced bandshift signal (Figure Oct-2A/207-357 strongly Deletion analysis of Oct-2A 10 and deletion of the first 15 amino 2A lanes 11). Upon For the further analysis of Oct-2A, we constructed a series of the domain (Oct-2A/210-479, acids (aa) POU-specific or the 13 C-terminal amino acids from of plasmids (Matthias et al., 1989) encoding Figure 2A lane 12) expression N from the C domain 2B lane 8) Oct-2A proteins deleted from the terminus, the POU-homeo (Oct-2A/2-344, Figure from ends. can be detected. In the case of the terminus, or both no bandshift signal domain is on the various expression Oct-I an intact First, the proteins encoded ubiquitous factor, POU-specific and were in vivo et al., for efficiency specificity (P.van plasmids expressed (Muller-Immergluck required high binding Most this is also Nuclear extracts were der communication). likely, 1989; Schreiber et al., 1989a). Vliet, personal for the reduced or absent bandshift and in a gel retardation assay with a the signal prepared analysed explanation et mutants at aa 207 or 210, octamer site probe (Muller-Immergluck al., with deletion starting 206, radiolabelled exclude a low due obtained we cannot signal The amount of retarded radioactive respectively, although 1988). probe to is affected concentration protein instability. in bandshift experiments by protein 1627 M.M.Muller-immergluck, W.Schaffner and P.Matthias Gin-rich POU-spec. POU-Homeo Leu-Z OCTA(2) w'ei 479 ++ I 11111 99: .: , 112 a 210 7777= B C 460 ++ + ) ++ ) _.!Ol. 777. 777 ++.(+) -ME= 7,.-: M., I,.!: m =.,. -1 C 161 77= 460 D, I E= 194 357 (+) 20 6 20'7 .M of and activation deleted Oct-2A The the Oct-2A considered as Fig. 3. Structure transcriptional by proteins. top drawing represents protein wild-type encoded the clone described et or this The (see Materials and methods) by previously [pOEV 1+, Muller-Immergluck al., (1988) p0/2-479, work]. to the left and to the of each to the first and last amino acid of Oct-2A in the construct. For numbers right drawing correspond present precise features of Oct-2A are the stretches and are nomenclature see Materials and methods. Some salient II III a indicated; glutamine-rich I, depicted by area in a which overall is also leucine- and somewhat also The shaded (aa 104-112, 128-135, 146-153) region proline-rich (see Figure 5A). POU domain and is solid boxes 195-269 and and the short of acidic amino bipartite (POU-specific POU-homeo) represented by (aa 297-356) patch The leucine motif is indicated vertical bars acids within the domain is indicated a area POU-specific by striped (aa 195-204). repeat by (aa 389, which with the leucine Not is a serine-threonine-rich The 396, 403, 410). particularly highlighted region overlaps repeat (aa 390-405). average was determined in at least two transfections and is indicated relative transactivation of the the various OCTA(2) promoter by proteins independent stimulation levels varied somewhat from to on the next to the the absolute transfection drawings. Although experiment experiment (depending induced the various Oct-2A mutants were structure and relative efficiency), the relative stimulation levels by very reproducible. (A) transcriptional All constructs of this series end with the identified codon at 479. structure and relative activation of N-terminal deletions. previously stop position (B) of C-terminal deletions. All clones in this series start with Oct-2A number 2 and at a codon in the vector transcriptional activation aa terminate stop et structure and relative activation of truncated from both ends. (Matthias al., 1989). (C) transcriptional proteins the deletions establish the entire An additional in brackets indicates an Taken together, presented Oct-2A/2-479. (+) and as the domain increased of the same deletion mutants POU domain (POU-specific POU-homeo) transcriptional activity of the Oct-2A factor that is stable in vivo and sufficient for which in some cases is due to increased levels of expressed DNA 2C lane see also Clerc et protein (see below). binding (Figure 6; al., 1988; Sturm and Sturm et LeBowitz et Transcriptional activity was retained upon progressive N- Herr, 1988; al., 1988; al., 1989). or C-terminal deletion unless the POU domain was affected We next examined the transactivation potential of the (Figure 3A and B). Deletions into the POU domain various deleted Oct-2A proteins. This was done by trans- (Oct-2A/206-479, result Oct-2A/207-479, Oct-2A/210479) the into HeLa cells in which cannot activate fecting expression plasmids together with proteins transcription significantly a fl-globin reporter whose over basal even the first two of plasmid promoter contains, level, though these constructs to chain retained some DNA analogous naturally occurring immunoglobulin light binding activity. (Figure 2A, lanes promoters, a single octamer site. In a first series of 10-12). Some of the truncated conferred an proteins activity we experiments used the OCTA(2) promoter which, when that exceeded the activity of For wild-type protein. example, placed together with the SV40 enhancer downstream, has a protein missing the first 99 aa of Oct-2A (Oct-2A/99-479) a particularly low basal level activity in HeLa cells (Figure confers reproducibly an -3-fold higher activity than the 4A and D; Muller-Immergluck et al., 1988). The trans- complete protein. Conversely, deletion of the second N- of the criptional activity various mutant proteins was assessed terminal glutamine-rich stretch is a down mutation (see also by measuring the amount of correctly initiated reporter gene Table I). RNA. The results are indicated in Figure 3 next to the Deletion of the C-terminal end of Oct-2A also confers drawings of the truncated proteins. The transcriptional increased activity (- 4-fold for Oct-2A/2-392 -or activation by the mutant proteins is indicated with '+' when Oct-2A/2-370) and a protein ending exactly behind the POU- it is at least 50% of the activity obtained with 'wild-type' homeobox (Oct-2A/2-357) is still fully active. As discussed 1628 Oct-2A transcription factor function C' _-O _ h -_ "m am _ ._ _m Wm O"m 4I 1 2 5 _- 4W OCTA (2) OCTAl-1) rl 14 24 -.1 OCTA(2) TATA gEigi . . . OCTA' t TATA 1 .. .. .....;~~~-+. 1-0 -. 1O99r Fig. 4. Transactivation of the OCTA(2) and OCTA(l) promoter constructs (both with an SV40 enhancer downstream) in HeLa cells by minimal Oct-2A proteins truncated from both ends determined by RNase protection analysis. (A) transactivation of the OCTA(2) promoter; (B) transactivation of the OCTA(l) promoter. Test, position of the signal from the reporter plasmid; ref, position of the signal from the reference plasmid. (C) bandshift experiments with whole cell extracts from a prepared fraction of the cotransfected cells. (D) Sequence of the OCTA(2) and OCTA(1) promoter. Both constructs have the SV40 enhancer in a downstream position (see Figure ID and Muller-Immergluck al., 1988). et below, this increase in is due to a the activity primarily greater OCTA(l) promoter, where the octamer motif is separated concentration of the truncated Oct-2A factor in the from the TATA box by 20 bases (Figure 4D). Activity was transfected cells. Note that some active deletions remove the normalized against the bandshift signals detected in whole region with the leucine repeat motif. Therefore, the putative cell extracts from the same transfected cells (Figure 4A, B leucine zipper of Oct-2A is clearly not essential for tran- and C). Indeed, deletion of the C terminus resulted in a 2-fold scriptional activation in the context of our reporter plasmid. decrease in transcription with the OCTA(1) promoter, while Surprisingly, even a protein the POU domain with the OCTA(2) promoter this decrease is not seen. This containing only is still weakly active. result suggests the presence of a weak activation domain in the C terminus of Oct-2A and that the architecture of the The POU region alone can weakly transactivate one promoter can influence the response to a given transactivator but not another promoter (see Discussion and Ruden et al., 1988). Our results show that deletions of either the or C The POU domain alone clearly activates the OCTA(2) but terminus, to of the POU do not result in up the boundaries domain, not the OCTA(1) promoter. In fact, activity is even weaker a This could be due than it in the of complete loss of transcriptional activity. appears mapping because the elevated protein to of domains in the Oct-2A level (compare lanes 4 and 5 in 4A either a redundancy activating Figure and B). We note the of a within the a short stretch of acidic amino acids within protein, presence transactivating region the N-terminal or both. To test whether the increased of the domain which POU domain itself, part POU-specific may be responsible seen with some deletion mutants reflected either a for the of this deletion mutant activity transactivating ability (see in or different Figure 3C). That particular deletion change transcription potential protein (Oct-2A/ 194/357) we transfected various deletion mutants into interfered with the of the reference since concentrations, expression gene, extracts of the HeLa cells and bandshifts with whole cell the reference was reduced. It performed intensity signal always could be that this truncated factor binds in to the transactivation highly expressed parallel experiment. octamer site is to an octamer in the enhancer. steric In the the sequence SV40 Through OCTA(2) promoter, separated to five bases and it would other factors from from the TATA box hinderance, prevent binding by only (Figure 4D), et this to the motifs so we wondered whether overlapping 'sph' (Tanaka al., 1988). promoter responded severely our studies do not address the because of its of Although directly question mutated Oct-2A proteins high degree are for the nuclear localization also Ruden et To address of what signal(s) responsible permissiveness (see al., 1988). it is that all deleted are of various mutated Oct-2A of Oct-2A, apparent this, the transactivation potential proteins present and also in nuclei of transfected cells. Since the decrease in was tested on both the the proteins OCTA(2) promoter 1629 W.Schaffner and P.Matthias M.M.Muller-Immergluck, S< _.. I. ,.-,i f i -* *4* Ew W 5. ;.-- .- W ..I:% .. ..::.. I3? :: -ML-ALA. :,&&. -11. - . ... 11'-5 := il. r M -." c-- . .-Ifll "I"N. .- Rp, of of and deletion For 5. of the of the N terminus Oct-2A. structure mutants. of salient Fig. Analysis glutamine-rich region (A) wild-type explanation to 3. The 16 and 15 leucine residues in this are underlined or indicated with a features of Oct-2A, see legend Figure glutamine (Q) (L) region dot, and RNase with the and and bandshifts of whole cell respectively. (B) (C) protection analysis OCTA(2) OCTA(l) promoter. (D) (E) quantitative extracts from a fraction of the cotransfected cells. prepared deletion into the box binding activity upon POU-specific is Two glutamine-rich stretches in the N terminus also observed with whole cell extracts (compare Figure 2A, contribute to transcription activation lane 10 and Figure 4C, lane 5), it cannot be due to In the next series of impaired experiments, we looked in more detail transport into the nucleus. at the glutamine-rich region consisting of the three glutamine- 1630 Oct-2A factor function transcription Table I. Quantification of transactivation by the deletion mutants removing the N-terminal glutamine-rich region Corresponding lanes in Figure 5 3 4 5 6 First N-terminal aa of Oct-2A 99 112 136 161 deleted glutamine-rich stretch I II III activity on OCTA(2) promoter 100% 290% 220% 80% 55% decrease in activity 1.3x 2.8x 1.5x activity on OCTA(l) promoter 100% 350% 120% 40% 35% decrease in activity 2.9x 3.0x l.lx The transcriptional stimulation has been corrected for the amount of Oct-2A protein present. For both promoters, transactivation by wild-type protein was set to 100%. The absolute increase of transcription upon cotransfection of wild-type protein was X 28x for the OCTA(2) and for the 1O OCTA(l) promoter. rich stretches Gln I, II and III that there (Figure 5A). As above, the is insufficient cotransfected Oct-2A present in experiment was with both the HeLa cells performed to out enhancer OCTA(l) and carry activation. OCTA(2) promoters. To test whether the loss of We rather speculate that either another activity octamer binding observed deletion of 12 is reproducibly upon aal -aa136 protein responsible for enhancer activation, or that (see Oct-2A Figure 3A) is due to either the deletion or itself may be altered protein target for some B cell-specific modification levels, we again quantified the it more transcriptional activation making active. Alternatively, there may be an observed with the four N-terminal deletions additional, possibly not even DNA binding, successively factor(s) present removing all three glutamine-rich stretches. The in B cells which signals would render Oct-2A mediated enhancer obtained with both the and activation OCTA(2) OCTA(1) promoter possible. Such an 'adaptor' for could, example, 5B (Figure and C) were normalized the interact with Oct-2A against bandshift via its leucine repeat. For promoter signal produced by Oct-2A in the transfected cells this factor present activity, would be dispensable since we have 5D and In (Figure E). this bandshift the excess shown that in HeLa experiment, cells, Oct-2A does not its require leucine of binding probe was sufficient to avoid between repeat. competition transiently expressed Oct-2A and Oct-I Finally, we endogenous protein. cannot rule out the existence of a As a result of this, the Oct-I signal could also be used as repression/extinction mechanism in non-lymphoid cells, an internal standard. which would affect enhancer but not promoter function. The results are summarized in Table I. On both Extinction phenomena promoters, have indeed been described for deletion of the first 99 aa leads to a octamer-containing test reproducible 3-fold genes (Junker et al., 1988, 1990; increase in - transcriptional stimulation SB Zaller et Yu et (Figure E, al., 1988; al., 1989). In this context it is lanes 2 and The of in compare 3). effect deletions the also noteworthy that transcriptional activity of another POU glutamine-rich can be region seen with the gene, the pituitary particularly specific transcription factor Pit-I with the et OCTA(i) promoter: OCTA(l) deletions (Ingraham al., 1988; Bodner et al., 1988) promoter, is cell type- of each Gln I and Gln II leads to a similar 3-fold decrease dependent: in similar cotransfection experiments in HeLa a which is about three times less leaving protein active than cells, Pit-I is able to transactivate selectively the prolactin In wild-type. the context of the and OCTA(2) promoter, only growth hormone gene promoters as well as synthetic deletion of Gln II results in a 3-fold decrease. For promoters consisting both of multiple binding sites taken from promoters, deletion of Gln Ill has a effect. either negligible Since gene (Ingraham et al., 1988). However, although Pit-I deletions in the are seen with both was still able to glutamine-rich regions transactivate a cotransfected prolactin promoters, we consider them an activation domain promoter in the important putative rat lactotroph cell line 235-1, it of Oct-2A. completely failed to activate the growth hormone gene promoter (Ingraham et al., 1988; Mangalam et al., 1989). Experiments to determine whether enhancer activation by Oct-2A could also be cell type-dependent are currently in Discussion progress. Why does Oct-2A not work from an enhancer position in HeLa cells? Activating domains of Oct-2A can compensate for Here we present evidence that Oct-2A specifically fails to each other activate x the IgH and the 6 OCTA in The data from enhancers HeLa cells, the deletion analysis of Oct-2A suggest that although the octamer containing promoter on the same Oct-2A contains several redundant transcription activating plasmid responds to cotransfected Oct-2A. The inactivity of regions on both sides of, and with possibly overlapping the the enhancer IgH might be the absence of POU domain. As explained by other long as the DNA binding domain is intact lymphocyte-specific DNA binding proteins. The and the protein is not truncated finding, from both no ends, single that even the 6 x OCTA however, enhancer was not active region seems to be essential for absolutely transcriptional upon cotransfection of Oct-2A is surprising: in HeLa cells, activation. cotransfected Oct-2A is more abundant than Two Oct-I activating domains, one in the N and one in the (Figure C 4C, lane 2). This contrasts with the situation in terminus Namalva of Oct-2A, were identified: a glutamine-rich region a B cell in cells, line which Oct-I is more abundant than in the N terminus of Oct-2A, similar to the ones in the Oct-2A. In of the 6 OCTA enhancer is factor spite this, active transcription SpI (Courey and Tjian, 1988), appears in these cells (Gerster et al., Therefore it is to be the dominant domain. An 1987). as yet undefined unlikely region 1631 M.M.Mu1ler-Immergluck, W.Schaffner and P.Matthias within the C terminus of Oct-2A can only be monitored with seems to be in some Although promoter geometry important the less permissive target promoter OCTA(l) (see below). cases et Wu and (Takahashi al., 1986; Berk, 1988), the Note that, in particular, deletions within the C terminus result sites for several binding transcription factors, including in significantly higher amounts of protein and, as a Oct-2A and can be at various distances from the Gal4, TATA consequence, elevated transcriptional activity in the box without activation impairing transcriptional (Chodosh transfected cells. et al., Wirth et Ruden et 1987; al., 1987; al., 1988). The analysis of the glutamine-rich region in the N terminus The POU region transactivates weakly by itself of Oct-2A also indicates that the OCTA(2) promoter is more than the both We also show that a truncated protein essentially consisting permissive OCTA(1) promoter. Although to deletion of the of the entire POU domain (Oct-2A/194-357) is stable in promoters respond similarly glutamine- rich stretch Gln the of the vivo and sufficient for DNA binding (for in vitro data, see II, activity OCTA(1) promoter is more affected than the deletion of Gln I LeBowitz et al., 1989). Most strikingly, this protein is still OCTA(2) by (see Table able to activate transcription from the OCTA(2) promoter, I). From our it is evident albeit with a low efficiency (Figure 4A). Thus, a third experiments that the reporter gene can be more transcription activating region of Oct-2A appears to overlap or less permissive for either full length or with DNA A cluster of six acid residues truncated Oct-2A the binding domain. factor. In this test system, the architecture at of domain between aa of the can determine the the beginning the POU-specific promoter (and enhancer) activity of a factor to a certain extent. Hence such 195-204 (see Figure 3) could be responsible for this given transcription an of DNA an with more than one transcriptional activation. Such overlap binding analysis should be done promoter domain is observed with the at a time. and activating prokaryotic phage X where the amino acids repressor (Ptashne, 1988), activating interwoven with the helix-turn-helix DNA Are different Oct-2 factors for at are binding required fine-turning et This also be the case the level? domain (Bushman al., 1989). may transcriptional for the glucocorticoid receptor (Hollenberg et al., 1987; Even our truncated Oct-2A and the though promoter Miesfeld et al., 1987). constructions do not exist as such in we that nature, suspect such a sensitive In this we note that the helix-turn-helix and differential of similar test context, region response of the POU-homeo domain of the to the same well reflect of a ubiquitous transcription promoters protein may part natural factor Oct-l contributes indirectly to transcriptional activation regulatory mechanism. Within the framework of this via a specific protein-protein contact. In elegant different variants of Oct-2 concept, splicing proteins could it has been identified as the essential have different affinities to various experiments, region octamer-containing for selective interaction with the virus transactivator promoters. This could be used to activate preferentially one herpes or the other octamer protein VP16 (or Vmw65; Stern et al., 1989). Fusions with and containing lymphocyte-specific will be Sherman et It heterologous DNA binding domains required to promoter (e.g. al., 1989). has been shown of and other that of activated dissect further the potential this putative expression promoters by the Drosophila bicoid on both the activating domains of the Oct-2A factor. morphogen critically depends concen- tration of bicoid and the of the sites affinity binding (Struhl not abolish et al., Driever et Lack of the Leu-zipper does promoter 1989; al., 1989). In the of the recent identification of activation light many additional with both the and the octamer in various tissues et The reporter plasmids OCTA(2) binding proteins (Schoeler al., contained an SV40 enhancer we would this to be OCTA(1) promoter 1989a,b), expect protein family involved downstream of the test in order to the in a network. gene augment signals. complex regulatory We show that of the the independent promoter used, enhancer effect a is still manifest with (i.e. strong signal) Materials and methods an Oct-2A mutant that lacks the leucine This indicates repeat. that, at least in the case of the SV40 enhancer, the Construction and of precise nomenclature deletion mutants cDNA deletions were introduced into the Oct-2A cDNA an (in appropriate hypothetical leucine zipper of Oct-2A cannot be involved nested deletions with either BaI3 a subclone) by creating 1, combination of in the enhancer for mediating effect by allowing, example, exonuclease and nuclease SI or natural mI by using restriction sites (Maniatis direct interactions of Oct-2A with promoter-bound proteins et The deletion of all clones was determined al., 1982). endpoint by dideoxy to the enhancer. bound SV40 Deleted cDNA molecules were inserted into vectors which sequencing. allow of deleted as an in-frame expression any N-terminally protein fusion product with the first four amino acids of the herpes simplex virus TK gene (Matthias Influence of different target promoters et Since our cDNA starts at the second amino acid of al., 1989). the extended The OCTA(2) promoter, in which the octamer site is only of Clerc et sequence al. (1988), we now use the numbering of these authors 5 from the TATA bp away box, together with an SV40 this work. For throughout simplicity we consider pO/2-479 to encode the enhancer downstream, yields the most permissive test which we call In wild-type protein Oct-2A/2-479. our nomenclature, the first number before the dash (0, 1, 2) designates the reading frame of the system. This is indicated by the fact that the POU domain vector used et expression (Matthias al., 1989) and the following two numbers itself (Oct-2A/194-357) shows some activity. However, that (which are used to define the protein in the drawing) represent the first and same POU domain is protein unable to activate the similar last amino acids of the Oct-2A sequence. octamer-TATA box promoter OCTA(l), where the octamer Using the one letter code, the translated linker sequence of the N-terminally is 20 bp away from the TATA box (compare lanes 4 and truncated clones in front of the Oct-2A sequence is as follows: pO/2479 (Oct-2A/2-479), MASWGSGTELEFR; pO/50479, MASWGSGTP; 5 in 4A and Figure D). pO/56479, MASWGSGT; p1/99479, MASWGSGVP; 12479, pIl/l Similar to truncated Oct-2A, transactivation by severely MASWGSGVP; p2/136479, MASWGSGY; p2/161479, MASWGSGY; truncated on a short distance between its Gal4 depends pl/187479; MASWGSGVP; pO/194479, MASWGSGT; p2/206479, site and the TATA box et MASWGSGYP; binding (Ruden al., MASWGSGTP. 1988). p2/207479, MASWGSGYP; pO/210479, 1632 factor function Oct-2A transcription Bodner,M., Castrillo,J.L., Theill,L.E., Deerinck,T., Ellisman,M. and The C-terminal deletions start with Oct-2A aa number 2 and were Karin,M. (1988) Cell, 55, 505-518. constructed using a similar expression vector, providing stop codons in all Bushman,F.D., Shang,C. and Ptashne,M. (1989) Cell, 58, 1163-1171. three reading frames behind the end of the Oct-2A sequence (Matthias et al., Burglin,T.R., Finney,M., Coulson,A. and Ruvkun,G. (1989) Nature, 341, 1989). The translated linker sequence of the C-terminally truncated clones 239-343. behind the Oct-2A sequence is as follows: p3S/2460, GGT; p3/2-427, PGT; Chodosh,L.A., Carthew,R.W., Morgan,J.G., Crabtree,G.R. and p3S/2-392, GYLAS; p3S/2-370, GVPS; p3S/2-357, GYLAS; p3S/2-344, Sharp,P.A. (1987) Science, 238, 684-688. VPS. Clerc,R.G., LeBowitz,J.H., Baltimore,D. and Sharp,P.A. Corcoran,L.M., Plasmids encoding Oct-2A proteins deleted at both ends were made by (1988) Genes Dev., 2, 1570-1581. recombining N-terminal and C-terminal deletion plasmids via the Sfil sites Courey,A.J. and Tjian,R. (1988) Cell, 55, 887-898. present in the Oct-2A POU domain and in the SV40 origin fragment of Cox,P.M., Temperley,S.M., Kumar,H. and Goding,C.R. (1988) Nucleic the expression vector. Acids Res., 16, 11047-11056. Dreyfus,M., Doyen,N. and Rougeon,F. (1987) EMBO J., 6, 1685-1690. Construction of clones bearing various enhancers Driever,W., Thoma,G. and Nusslein-Volhard,C. (1989) Nature, 340, The constructs (shown in Figure 1) bearing various enhancers in an EcoRI 363 -367. site downstream of the reporter gene (Westin et al., 1987) were derived Eckner,R. and Birnstiel,M.L. (1989) Nucleic Acids Res., 17, 5947 -5959. from the OCTA(2) promoter construct (containing the octamer containing Falkner,F.G. and Zachau,H.G. (1984) Nature, 310, 71 -74. oligo T.Gerster, unpublished results) with the SV40 enhancer D6, Falkner,F.G., Mocikat,R. and Zachau,H.G. (1986) Nucleic Acids Res., downstream (Figure 4D, Gerster et al., 1987, Muller-Immergluck et al., 14, 8819-8827. 1988). The SV40 enhancer containing BamHI-KpnI (Asp418) fragment Finney,M., Ruvkun,G. and Horvitz,H.R. (1988) Cell, 55, 757-769. (Gerster et al., 1987) was replaced by the corresponding fragment containing Fletcher,C., Heintz,N. and Roeder,R.G. (1987) Cell, 51, 773-781. the IgH enhancer (Gerster et al., 1987) or the enhancerless fragment from Gehring,W.J. (1987) Science, 236, 1245-1252. OVEC-1 (Westin et al., 1987). The 1 x OCTA and 6 x OCTA enhancers Gerster,T., Matthias,P., Thali,M., Jiricny,J. and Schaffner,W. (1987) (1 x, 6 x, Gerster et al., 1987) were cloned correspondingly by replacing EMBO J., 6, 1323-1330. the BamHI-XbaI fragment of the reporter plasmid. Harvey,R.P., Robins,A.J. and Wells,J.R.E. (1982) Nucleic Acids Res., 10, 7851 -7863. Preparation of nuclear mini extracts He,X., Treacy,M.N., Simmons,D.M., Ingraham,H.A., Swanson,L.W. and yg of expression plasmid DNA was transfected into Cos-7 cells by the Rosenfeld,M.G. (1989) Nature, 340, 35-42. DEAE-dextran procedure and nuclear mini-extracts were prepared after Heer,W,. Clerc,R.G., Corcoran,L.M., Baltimore,D. Sturm,R.A., 2 days, essentially as described (Muller-Immergluck et al., 1989; Schreiber Sharp,P.A., Ingraham,H.A., Rosenfeld,M.G., Finney,M. Ruvkun,G. et 1989a). al., and Horvitz,H.R. (1988) Genes Dev., 2, 1513-1516. Hollenberg,S.M., Giguere,V., Segui,P. and Evans,R.M. (1987) Cell, 49, of whole cell extracts Preparation 39-46. 10 cm plate of transfected cells was washed with 10 ml TBS and One Ingraham.H.A., Chen,R., Mangalam,H.J., Elsholtz,H.P., Flynn,S.E., from the plate with a rubber policeman in 1.2 ml TBS (+2 mM scratched Lin,C.R., Simmons,D.M., Swanson,L. and Rosenfeld,M.G. (1988) Cell, EDTA). 0.3 ml of cell suspension (25%) were pelleted for 5 s in an 519-529. centrifuge and resuspended by vortexing in 40 Il X-400 buffer 55, Eppendorf Junker,S. Nielsen,V., Matthias,P. and Picard,D. (1988) EMBO J., 7, buffer from Schreiber et al. (1988) adjusted to 400 mM KCI] [X-50 3093 -3098. complemented with 1 mM PMSF, 4 leupeptin and 10 mM DTT. Ag/ml Schreiber,E. and Matthias,P. (1990) Cell, in press. Junker,S., Pederson,S., cell extracts were prepared as described by Kumar and Chambon Whole Ko,H., Fast,P., McBride,W. and Staudt,L. (1988) Cell, 55, 135-144. four cycles of freezing and thawing. (1988) by Kumar,V. and Chambon,P. (1988) Cell, 55, 145 - 156. Sive,H.L. Roeder,R.G. and Heintz,N. (1988) Genes Dev., 2, Analytical bandshift experiments LaBella,F., 32 -39. Analytical bandshift experiments were performed as described in Schreiber LeBowitz,J.H., Clerc,R.G.. Brenowitz,M. and Sharp,P.A. (1989) Genes and Muller-Immergluck et al. (1989). et al. (1989) 3, 1625-1638. Del., Lenardo,M.J., Robbins,P., Kuang,A. Mulligan,R.C. and Staudt,L., Quantitative bandshift experiments (1989) Science, 243, 544-546. Baltimore,D. 2 of whole cell extract were incubated for 10 min at room temperature jig Maeda,H., Araki,K., Kitamura,D., Wang,J. and Watanabe,T. (1987) in 15 1il 1 x BEBBI (Muller-Immergluck et al., 1988) in the presence of Nucleic Acids Res., 15, 2851-2869. 4 ug poly(dIdC) and 18 000 c.p.m. (2.3 fmol) of labelled octamer binding Mangalam,H.J., Albert,V.R., Ingraham,H.A., Kapiloff,M., Wilson,L., site (Muller-Immergluck et al., 1988). All bandshift signals from the Oct-2 Nelson,C., Elsholtz,H. and Rosenfeld,M.G. (1989) Genes Dev., 3, proteins, normalized against the signal of Oct- 1, were quantified by cutting 946-958. out the radioactive bands and Cerenkov counting. Fritsch,E.F. and Sambrook,J. (1982) Molecular Cloning. A Maniatis,T., Laboratory Cold Spring Harbor Laboratory Press, Cold Spring Manual. Transfections for transactivation Harbor, NY. HeLa cells were transfected with 8 reporter plasmid OCTA(2) or jig Williams,G.T. and Neuberger,M.S. (1985) Cell, 41, 479-487. Mason,J.O., OCTA(1) (Muller-Immergluck et al., 1988), 4 Oct-2A expression vector, Ag Mattaj,I.W., Lienhard,S.. Jiricny,J. and De Robertis,E.M. (1985) Nature, reference plasmid and 7 sonicated herring sperm DNA as carrier /g jig 316, 163-167. et al., 1988). Two days later, cytoplasmic RNA was (Muller-Immergluck Matthias.P., Muller-Immergluck,M.M., Schrieber,E., Rusconi,S. and and analysed by RNase protection. Except for Figure 1, all levels extracted Schaffner,W. (1989) Nucleic Acids Res., 17, 6418. normalized against the signal of the reference gene, were of transcription, Miesfeld,R., Godowski,P.J., Maler,B.A. and Yamamoto,K.R. (1987) cutting out the radioactive bands and Cerenkov counting. quantified by Science, 236, 423-427. Muller-Immergluck,M.M., Ruppert,S., Schaffner,W. and Matthias,P. Acknowledgements (1988) Nature, 336, 544-551. Muller-Immergluck,M .M., Schreiber,E., Schaffner,W. and Matthias,P. We thank Edgar Schreiber for providing Figure 4C, Andreas Czank for (1989) Nucleic Acids Res., 17, 6420. excellent technical assistance. Hans-Peter Muller-Storm, Keith Harshman, O'Neill,E.A. and Kelly,T.J. (1988) J. Biol. Chem., 263, 931-937. Deborah Maguire and Markus Thali for critical reading of the manuscript Parslow,T.G., Blair,D.L., Murphy,W.J. and Granner,D.K. (1984) Proc. and valuable comments, Fritz Ochsenbein for careful preparation of the Natt. Acad. Sci. USA, 81, 2650-2654. We are indebted to T.Gerster, Iris Kemler and R.Roeder as well figures. Pruijn,G.M.J., van Driel,W. and van der Vliet,P.C. (1986) Nature, 322, as to W.Herr and M.Tanaka for communicating results prior to publica- 656-659. This work was supported by the Swiss National Science Foundation tion. Ptashne,M. (1988) Nature, 335, 683-689. and by the Kanton of Zurich. Rosales,R., Vigneron,M., Macchi,M., Davidson,I., Xiao,J.H. and Chambon,P. (1987) EMBO J., 6, 3015-3025. References Ruden,D.M., Ma,J. and Ptashne.M. (1988) Proc. Natl. Acad. Sci. USA, 85, 4262-4266. and Busslinger,M. (1987) Cell, 50, 347-359. Superti-Furga,G. Barberis,A., Heguy,A. and Roeder,R.G. (1987) Cell, 51, 783-793. Scheidereit,C., and Busslinger,M. (1989) Nucleic Acids Res, Barberis,A., Superti-Furga,G. Cromlish,J.A., Gerster,T., Kawakami,K., Balmaceda,C., Scheidereit,C., 17, 1633 M.M.Muller-Immergluck, W.Schaffner and P.Matthias Currie,R.A. and Roeder,R.G. (1988) Nature, 336, 551-557. Scholer,H.R., Hatzopoulos,A.K.. Balling,R., Suzuki,N. and Gruss,P. (1989a) EMBO J., 8, 2543-2550. Scholer,H.R., Balling,R. Hatzopoulos,A.K., Suzuki,N. and Gruss,P. (1989b) EMBO J., 8, 2551-2557. Schreiber,E., Matthias,P., Muller-Immergluck,M.M. and Schaffner,W. (1988) EMBO J., 7, 4221-4229. Schreiber,E., Matthias,P., Muller-Immergluck,M.M. and Schaffner,W. (1989a) Nucleic Acids Res., 17, 6419. Schreiber,E., Muller-Immergluck,M.M., Schaffner,W. and Matthias,P. (1989b) Renkawitz,R. (ed.), Tissue Specific Gene Expression, Verlag Chemie, Weinheim, FRG. Sherman,P.A., Basta,P.V., Heguy,A., Wloch,M.K., Roeder,R.G. and Ting,J.P.Y. (1989) Proc. Natl. Acad. Sci. USA, 86, 6739-6743. Singh,H., Sen,R., Baltimore,D. and Sharp,P.A. (1986) Nature, 319, 154-158. Staudt,L.M., Singh,H., Sen,R. Wirth,T. Sharp,P.A. and Baltimore,D. (1986) Nature, 323, 640-643. Staudt,L.M., Clerc,R.G., Singh,H., LeBowtiz,J.H., Sharp,P.A. and Baltimore,D. (1988) Science, 241, 577-580. Stern,S., Tanaka,M. and Herr,W. (1989) Nature, 624-630. 341, Struhl,G., Struhl,K. and Macdonald,P.M. 1259-1273. (1989) Cell, 57, Sturm,R.A. and Herr,W. (1988) Nature, 601-604. 336, Sturm,R.A., Das,G. and Herr,W. Genes (1988) Dev., 2, 1582-1599. Takahashi,K., Vigneron,M., Matthes,H., Wildeman,A., Zenke,M. and Chambon,P. (1986) Nature, 319, 121-126. Tanaka,M., Grossniklaus,U., Herr,W. and Hernandez,H. (1988) Genes Dev., 2, 1764-1778. Thali,M., Muller-Immergluck,M.M., DeLorenzi,M., Matthias,P. and Bienz,M. (1988) Nature, 336, 598-601. Westin,G., Gerster,T., Miuller-ImmergKick,M.M., Schaffner,G. and Schaff- ner,W. (1987) Nucleic Acids Res., 15, 6787-6798. Wirth,T., Staudt,L. and Baltimore,D. (1987) Nature, 329, 174-178. Wu,L. and Berk,A. (1988) Genes Dev., 2, 403-411. Yu,H., Porton,B., and Shen,L. Eckhardt,L.A. (1989) Cell, 58, 441-448. Zaller,D.M., and Yu,H. Eckhardt,L.A. (1988) Mol. Cell. Biol., 8, 1932-1939. Zenke,M., Grundstrom,T., Matthes,H., Wintzerith,M., Schatz,C., Wildeman,A. and Chambon,P. (1986) EMBO J., 5, 387-397. Received on December 22, 1989; revised on February 19, 1990 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals

Transcription factor Oct‐2A contains functionally redundant activating domains and works selectively from a promoter but not from a remote enhancer position in non‐lymphoid (HeLa) cells.

Download PDF
10 pages

Loading next page...
 
/lp/springer-journals/transcription-factor-oct-2a-contains-functionally-redundant-activating-T4O0I2vSfh

References (1)

Publisher
Springer Journals
Copyright
Copyright © European Molecular Biology Organization 1990
ISSN
0261-4189
eISSN
1460-2075
DOI
10.1002/j.1460-2075.1990.tb08282.x
Publisher site
See Article on Publisher Site

Abstract

EMBO Journal vol.9 no. F25-1634, 1990 The pp. Transcription factor Oct-2A contains functionally redundant activating domains and works selectively a promoter but not from a remote enhancer from in cells position non-lymphoid (HeLa) is also found in promoters and enhancers of other sequence Michael M.MUller-ImmerglUck, Walter cellular and viral genes which display no cell type specificity. Schaffner and Patrick Matthias For example, it is present in the promoters of histone H2B Institut fur Molekularbiologie II der Universitiit Zurich, CH-8093 genes, in the promoters of U2 and U6 small nuclear RNA Zurich, Switzerland genes and the simian virus SV40 enhancer (Falkner et al., Communicated by W.Schaffner 1986; Harvey et al., 1982; Barberis et al., 1987; LaBella et al., 1988; Mattaj et al., 1985; Zenke et al., 1986; Rosales In non-lymphoid cells such as HeLa cells, ectopic et al., 1987; for a recent compilation of octamer sequence- expression of the lymphocyte-specific transcription factor containing genes, see Schreiber et al., 1989a). Oct-2A can reporter genes whose promoters activate Three octamer binding activities have been identified major consist of a single octamer sequence (ATTTGCAT) characterized: Oct-I (also called NF-A1, and thoroughly upstream of a TATA box. While the factor is strongly OTFI, NFIII, OBP100, oct-B3) has an apparent relative in position, it fails as an enhancer active a promoter molecular mass (Mr) of 90 000-100 000 (90-100K) and factor: an enhancer consisting of multiple copies of the has been found in all cell types investigated (Pruijn et al., of the reporter octamer sequence placed downstream 1986; Singh et al., 1986; Staudt et al., 1986; Fletcher et al., is not in cells, even at high concentration gene active HeLa 1987; Rosales et al., 1987; O'Neill and Kelly, 1988; Sturm cells, however, the same of Oct-2A. In B lymphoid et al., 1988). Oct-2A (also called NF-A2, OTF2, oct-B2) highly active. This could mean that an enhancer is has an apparent Mr Of - 60K and has been found in a few additional factor is required for enhancer activation in cell types only, notably in B lymphocytes where its presence B cells. Furthermore, we have tested the transcriptional et al., also correlates with transcription of Ig genes (Staudt activation potential of Oct-2A with a series of N-terminal Gerster al., Rosales et al., 1987; Scheidereit 1986; et 1987; and C-terminal deletions. We show that a glutamine-rich protein et al., 1987). Oct-2B is another lymphoid-specific domain near the N-terminus is required for full activity. - et al., 1988). of 75K and is related to Oct-2A (Schreiber Otherwise, large segments of the N-terminal half or the A number of other octamer binding proteins have been entire C-terminal region are dispensable in our assay, as in various tissues and cell lines: several octamer identified long as the deletions do not impinge on the conserved factors in neural tissues Oct4 -Oct0 (Scholer et al., POU domain which is sufficient for DNA binding. Vhile an embryonal carcinoma (EC) cell line-specific 1989a,b), N-terminal and C-terminal regions can functionally factor NF-A3 et 1989), another protein in (Lenardo al., for each other, a combined deletion that only compensate melanoma (independently also called NF-A3; Cox malignant retains the POU domain is a strong down mutation. We et a factor Oct-T (Barberis et al., al., 1988), testis-specific also fmd that activity depends on the promoter structure and two high mobility group proteins which 1987, 1989) of the reporter gene: the POU domain by itself shows bind both AT rich sequences and the octamer preferentially some with a promoter where the octamer activity and Birnstiel, 1989). sequence (Eckner is located very close to the TATA box, but no sequence the We and others have isolated cDNAs encoding activity with another promoter construction where the Oct-2A. Like the lymphocyte-specific transcription factor further The two octamer is located upstream. sequence RNA is almost protein it codes for, the Oct-2A expressea also respond differently to the deletion of the promoters et Muller- in lymphoid cells (Clerc al., 1988; exclusively stretch important for transcriptional glutamine-rich et Staudt Immergluck et al., 1988; Scheidereit al., 1988; it activation. From these experiments we consider likely Consistent with the et see also He et al., 1989). al., 1988; variants can that the natural octamer factor selectively is for B cell- finding that the octamer sequence important octamer- activate the different naturally occurring et et specific transcription (Mason al., 1985; Dreyfus al., containing promoters. Gerster et 1987; Wirth et al., 1987), the presence 1987; al., Key words: enhancer/Oct-2A(OTF-2)/POU/promoter/tissue- to via of Oct-2A in cells is sufficient activate, non-lymphoid specific transcription/transactivation the octamer sequence motif, lymphocyte-specific promoter Furthermore Oct-2A can et al., 1988). (Muller-Immergluck octamer-related sites also activate containing promoters et The derived from homeotic genes (Thali al., 1988). Introduction of Oct-2A et Ko deduced (Clerc al., 1988; protein sequence et Scheidereit site ATGCAAAT (or its complement et al., 1988; Muller-Immergluck al., 1988; The octamer a of to Oct-i et shows high homology is in all al., 1988) region ATTTGCAT) present immunoglobulin (Ig) as well as to two other Parslow et et tissue-specific and (Sturm al., 1988), promoters (Falkner Zachau, 1984; al., 1984) et et enhancers and Pit-l al., 1988; Ingraham al., 1988) in the chain factors, (Bodner and Ig heavy (IgH) (Falkner The of et et The same octamer and Unc-86 al., 1988). region homology Maeda (Finney Zachau, 1984; al., 1987). Press Oxford University W.Schaffner and M.M.Muller-immergluck, P.Matthias .. -, ~~~~~~~~~~~~~ -.. ~~~~~~~..I. ....~ ". ~ - ,_ f.,i _~~~~~~~~~~~~N 4Au- -- W. 5.A| Sr 1. of the driven various enhancers at the same Fig. Activity OCTA(l) promoter by downstream determined RNase position by protection analysis. Lanes 1, no lanes SV40 lanes lanes one octamer enhancer; 2, enhancer; 3, IgH enhancer; lanes six octamer in tandem. 4, sequence; 5, sequences Promoter in BJAB cells. in HeLa cells. in HeLa cells cotransfected Oct-2A (A) activity structure of the (B) (C) expressing protein. (D) reporter gene scale for the The location where the various enhancers drawn to are inserted is indicated an arrow. except enlarged promoter region. For by sequence of the see 4D. OCTA(l) promoter, Figure among these has been called the POU domain proteins and Results contains two conserved a box and subregions: POU-specific a POU-homeobox et The POU-homeobox In HeLa cells Oct-2A acts as (Herr al., 1988). preferentially a promoter is - 30% to homeoboxes found in transcription factor homologous many genes regulating Previous experiments have shown that development (Gehring, 1987). Mutagenesis multimerized octamer experiments have indicated that the entire POU domain is sites from various origins behave as a lymphoid-specific likely to represent the DNA of this class of enhancer, capable of binding region activating the fl-globin promoter from proteins (Herr et al., 1988; Sturm et The recent an upstream (Tanaka al., 1988). et al., 1988) or downstream position detection of other POU domain in mammalian (Gerster et al., 1987; Tanaka et proteins al., 1988). By cotransfection neural tissues et (He al., 1989) and Caenorhabditis experiments we have recently demonstrated that elegans cloned et (Biirglin al., 1989) could be indicative of a Oct-2A can stimulate complex specifically reporter plasmids network with regulatory overlapping specificities. containing an octamer site in the promoter (Muller- In this paper we evidence that Immergluck et present the cDNA encoded al., 1988). In order to see whether Oct-2A Oct-2A acts as a in can also account for preferentially promoter factor: a the enhancer observed activity with heterologous system, it fails to activate a remote multiple octamer sites in synthetic a downstream we position, enhancer consisting of multiple octamer sites even constructed a series of though plasmids which all contain the the same enhancer is in B highly active cells. This OCTA(1) promoter. This minimal suggests promoter consists of the that either another Oct-2 related factor is responsible for ,3-globin TATA-box and an octamer binding site (Figure 4D) enhancer activation or that and perhaps a second factor is required responds well to transactivation by Oct-2A (Muller- with Oct-2A in together order to activate the Immergluck et enhancer. By al., 1988 and Figures 4B and 5C). In testing deleted versions of Oct-2A for their transactivation the addition, plasmids contain either no enhancer or one of potential, we identified a glutamine-rich region in the N the following enhancers inserted downstream of the fl-globin terminus important for transactivation of two different test gene (Figure ID): the SV40 enhancer, the IgH enhancer, promoters. Two weaker activating domains in the C terminus a single octamer, or six octamer motifs in tandem (1 xOCTA and within the POU domain can only be monitored with one or 6 x OCTA) (Gerster et al., 1987). or the other test promoter. This suggests that a different We first tested the activity of these plasmids in B cells promoter architecture can influence the response to the same (BJA-B). The plasmids containing the SV40 enhancer, the transactivator. IgH or x enhancer, the 6 OCTA enhancer were all highly 1626 transcription factor function Oct-2A OcW-, --40 Oct-2B- - m me _9- ___ 00} Oct-2A- 0e. -- _ -_. - -%. Prnbe _ _ am Io<t; H 2 3 I 5 6 7 - I- 5 1- t) c8 "; .- Fig. 2. Bandshift experiments with nuclear extracts from Cos-7 cells expressing transfected Oct-2A cDNA deleted from the N terminus, C terminus or both ends. Autoradiograms are presented. The numbers above the lanes refer to the first and last Oct-2A amino acid present. (A) N-terminal deletions; (B) C-terminal deletions; (C) deletions from both ends. For a schematic representation of the deletions, see Figure 3. active IA lanes 2, 3 and 5). In HeLa cells, however, of and protein turnover) as well (Figure (as a consequence expression plasmid containing the SV40 enhancer showed a It will henceforth be referred to as only the as binding affinity. measurable activity which can only be seen when the 'bandshift signal'. of reference band is strongly overexposed (Figure iB, lane 2). to the functional domains Oct-2A, Secondly, identify of the mutant This weak activity depends on the octamer site in the the transcriptional activation capacity proteins cells in vivo transactivation of a promoter (Muller-Immergluck et al., 1988) and possibly was tested in HeLa by an octamer motif in the promoter reflects residual activity of endogenous Oct-I protein. As reporter plasmid containing as described can be seen in Figure IC, apart from the SV40 enhancer, and the SV40 enhancer downstream, previously cotransfection of the Oct-2A expression vector leads to (Muller-Immergliick et al., 1988). In several cases we also stimulation of every plasmid to the same extent independently such in CV-1 cells and obtained similar performed assays of the enhancer it contains. Since the basal level of transcrip- results (not shown). tion also increased for the enhancerless construct (Figure 1C, lane 1), the effect is due to weak stimulation of the Transactivation by truncated Oct-2A proteins OCTA(l) promoter by Oct-2A even in the absence of a for transactivation experiment, As control the subsequent We therefore conclude that in our functional enhancer. we first transfected the replication-competent expression functions as a experimental system, Oct-2A specifically vectors into Cos-7 cells to obtain comparably high amounts factor is not able to activate a lymphocyte- promoter and of each protein. This approach not only maps the DNA enhancer from a position far downstream. It was specific domain of Oct-2A, but assays at the same time for binding in B multiple copies of an of the mutant proteins found by others that, even cells, localization in the nucleus and stability motif were more when placed immediately octamer active under in vivo conditions. of a than far downstream (Tanaka et al., Oct-2A deleted upstream promoter As shown in Figure 2, the various proteins In the presented here, both the IgH and both ends were found in 1988). experiments from N terminus, C terminus or the x enhancers are highly active in B cells but not bound to the 6 OCTA the nuclei of Cos-7 cells and specifically in HeLa cells. Since we have not tested intermediate of the nomenclature see octamer probe (for explanation and 'far downstream' positions was observed distances between 'promoter' Materials and methods). Impaired binding only we cannot rule out quantitative, rather than qualitative, box or the POU-homeobox when part of the POU-specific in HeLa cells. Oct-2A/206-357 showed a slightly and deficiency was deleted: a reduced bandshift signal (Figure Oct-2A/207-357 strongly Deletion analysis of Oct-2A 10 and deletion of the first 15 amino 2A lanes 11). Upon For the further analysis of Oct-2A, we constructed a series of the domain (Oct-2A/210-479, acids (aa) POU-specific or the 13 C-terminal amino acids from of plasmids (Matthias et al., 1989) encoding Figure 2A lane 12) expression N from the C domain 2B lane 8) Oct-2A proteins deleted from the terminus, the POU-homeo (Oct-2A/2-344, Figure from ends. can be detected. In the case of the terminus, or both no bandshift signal domain is on the various expression Oct-I an intact First, the proteins encoded ubiquitous factor, POU-specific and were in vivo et al., for efficiency specificity (P.van plasmids expressed (Muller-Immergluck required high binding Most this is also Nuclear extracts were der communication). likely, 1989; Schreiber et al., 1989a). Vliet, personal for the reduced or absent bandshift and in a gel retardation assay with a the signal prepared analysed explanation et mutants at aa 207 or 210, octamer site probe (Muller-Immergluck al., with deletion starting 206, radiolabelled exclude a low due obtained we cannot signal The amount of retarded radioactive respectively, although 1988). probe to is affected concentration protein instability. in bandshift experiments by protein 1627 M.M.Muller-immergluck, W.Schaffner and P.Matthias Gin-rich POU-spec. POU-Homeo Leu-Z OCTA(2) w'ei 479 ++ I 11111 99: .: , 112 a 210 7777= B C 460 ++ + ) ++ ) _.!Ol. 777. 777 ++.(+) -ME= 7,.-: M., I,.!: m =.,. -1 C 161 77= 460 D, I E= 194 357 (+) 20 6 20'7 .M of and activation deleted Oct-2A The the Oct-2A considered as Fig. 3. Structure transcriptional by proteins. top drawing represents protein wild-type encoded the clone described et or this The (see Materials and methods) by previously [pOEV 1+, Muller-Immergluck al., (1988) p0/2-479, work]. to the left and to the of each to the first and last amino acid of Oct-2A in the construct. For numbers right drawing correspond present precise features of Oct-2A are the stretches and are nomenclature see Materials and methods. Some salient II III a indicated; glutamine-rich I, depicted by area in a which overall is also leucine- and somewhat also The shaded (aa 104-112, 128-135, 146-153) region proline-rich (see Figure 5A). POU domain and is solid boxes 195-269 and and the short of acidic amino bipartite (POU-specific POU-homeo) represented by (aa 297-356) patch The leucine motif is indicated vertical bars acids within the domain is indicated a area POU-specific by striped (aa 195-204). repeat by (aa 389, which with the leucine Not is a serine-threonine-rich The 396, 403, 410). particularly highlighted region overlaps repeat (aa 390-405). average was determined in at least two transfections and is indicated relative transactivation of the the various OCTA(2) promoter by proteins independent stimulation levels varied somewhat from to on the next to the the absolute transfection drawings. Although experiment experiment (depending induced the various Oct-2A mutants were structure and relative efficiency), the relative stimulation levels by very reproducible. (A) transcriptional All constructs of this series end with the identified codon at 479. structure and relative activation of N-terminal deletions. previously stop position (B) of C-terminal deletions. All clones in this series start with Oct-2A number 2 and at a codon in the vector transcriptional activation aa terminate stop et structure and relative activation of truncated from both ends. (Matthias al., 1989). (C) transcriptional proteins the deletions establish the entire An additional in brackets indicates an Taken together, presented Oct-2A/2-479. (+) and as the domain increased of the same deletion mutants POU domain (POU-specific POU-homeo) transcriptional activity of the Oct-2A factor that is stable in vivo and sufficient for which in some cases is due to increased levels of expressed DNA 2C lane see also Clerc et protein (see below). binding (Figure 6; al., 1988; Sturm and Sturm et LeBowitz et Transcriptional activity was retained upon progressive N- Herr, 1988; al., 1988; al., 1989). or C-terminal deletion unless the POU domain was affected We next examined the transactivation potential of the (Figure 3A and B). Deletions into the POU domain various deleted Oct-2A proteins. This was done by trans- (Oct-2A/206-479, result Oct-2A/207-479, Oct-2A/210479) the into HeLa cells in which cannot activate fecting expression plasmids together with proteins transcription significantly a fl-globin reporter whose over basal even the first two of plasmid promoter contains, level, though these constructs to chain retained some DNA analogous naturally occurring immunoglobulin light binding activity. (Figure 2A, lanes promoters, a single octamer site. In a first series of 10-12). Some of the truncated conferred an proteins activity we experiments used the OCTA(2) promoter which, when that exceeded the activity of For wild-type protein. example, placed together with the SV40 enhancer downstream, has a protein missing the first 99 aa of Oct-2A (Oct-2A/99-479) a particularly low basal level activity in HeLa cells (Figure confers reproducibly an -3-fold higher activity than the 4A and D; Muller-Immergluck et al., 1988). The trans- complete protein. Conversely, deletion of the second N- of the criptional activity various mutant proteins was assessed terminal glutamine-rich stretch is a down mutation (see also by measuring the amount of correctly initiated reporter gene Table I). RNA. The results are indicated in Figure 3 next to the Deletion of the C-terminal end of Oct-2A also confers drawings of the truncated proteins. The transcriptional increased activity (- 4-fold for Oct-2A/2-392 -or activation by the mutant proteins is indicated with '+' when Oct-2A/2-370) and a protein ending exactly behind the POU- it is at least 50% of the activity obtained with 'wild-type' homeobox (Oct-2A/2-357) is still fully active. As discussed 1628 Oct-2A transcription factor function C' _-O _ h -_ "m am _ ._ _m Wm O"m 4I 1 2 5 _- 4W OCTA (2) OCTAl-1) rl 14 24 -.1 OCTA(2) TATA gEigi . . . OCTA' t TATA 1 .. .. .....;~~~-+. 1-0 -. 1O99r Fig. 4. Transactivation of the OCTA(2) and OCTA(l) promoter constructs (both with an SV40 enhancer downstream) in HeLa cells by minimal Oct-2A proteins truncated from both ends determined by RNase protection analysis. (A) transactivation of the OCTA(2) promoter; (B) transactivation of the OCTA(l) promoter. Test, position of the signal from the reporter plasmid; ref, position of the signal from the reference plasmid. (C) bandshift experiments with whole cell extracts from a prepared fraction of the cotransfected cells. (D) Sequence of the OCTA(2) and OCTA(1) promoter. Both constructs have the SV40 enhancer in a downstream position (see Figure ID and Muller-Immergluck al., 1988). et below, this increase in is due to a the activity primarily greater OCTA(l) promoter, where the octamer motif is separated concentration of the truncated Oct-2A factor in the from the TATA box by 20 bases (Figure 4D). Activity was transfected cells. Note that some active deletions remove the normalized against the bandshift signals detected in whole region with the leucine repeat motif. Therefore, the putative cell extracts from the same transfected cells (Figure 4A, B leucine zipper of Oct-2A is clearly not essential for tran- and C). Indeed, deletion of the C terminus resulted in a 2-fold scriptional activation in the context of our reporter plasmid. decrease in transcription with the OCTA(1) promoter, while Surprisingly, even a protein the POU domain with the OCTA(2) promoter this decrease is not seen. This containing only is still weakly active. result suggests the presence of a weak activation domain in the C terminus of Oct-2A and that the architecture of the The POU region alone can weakly transactivate one promoter can influence the response to a given transactivator but not another promoter (see Discussion and Ruden et al., 1988). Our results show that deletions of either the or C The POU domain alone clearly activates the OCTA(2) but terminus, to of the POU do not result in up the boundaries domain, not the OCTA(1) promoter. In fact, activity is even weaker a This could be due than it in the of complete loss of transcriptional activity. appears mapping because the elevated protein to of domains in the Oct-2A level (compare lanes 4 and 5 in 4A either a redundancy activating Figure and B). We note the of a within the a short stretch of acidic amino acids within protein, presence transactivating region the N-terminal or both. To test whether the increased of the domain which POU domain itself, part POU-specific may be responsible seen with some deletion mutants reflected either a for the of this deletion mutant activity transactivating ability (see in or different Figure 3C). That particular deletion change transcription potential protein (Oct-2A/ 194/357) we transfected various deletion mutants into interfered with the of the reference since concentrations, expression gene, extracts of the HeLa cells and bandshifts with whole cell the reference was reduced. It performed intensity signal always could be that this truncated factor binds in to the transactivation highly expressed parallel experiment. octamer site is to an octamer in the enhancer. steric In the the sequence SV40 Through OCTA(2) promoter, separated to five bases and it would other factors from from the TATA box hinderance, prevent binding by only (Figure 4D), et this to the motifs so we wondered whether overlapping 'sph' (Tanaka al., 1988). promoter responded severely our studies do not address the because of its of Although directly question mutated Oct-2A proteins high degree are for the nuclear localization also Ruden et To address of what signal(s) responsible permissiveness (see al., 1988). it is that all deleted are of various mutated Oct-2A of Oct-2A, apparent this, the transactivation potential proteins present and also in nuclei of transfected cells. Since the decrease in was tested on both the the proteins OCTA(2) promoter 1629 W.Schaffner and P.Matthias M.M.Muller-Immergluck, S< _.. I. ,.-,i f i -* *4* Ew W 5. ;.-- .- W ..I:% .. ..::.. I3? :: -ML-ALA. :,&&. -11. - . ... 11'-5 := il. r M -." c-- . .-Ifll "I"N. .- Rp, of of and deletion For 5. of the of the N terminus Oct-2A. structure mutants. of salient Fig. Analysis glutamine-rich region (A) wild-type explanation to 3. The 16 and 15 leucine residues in this are underlined or indicated with a features of Oct-2A, see legend Figure glutamine (Q) (L) region dot, and RNase with the and and bandshifts of whole cell respectively. (B) (C) protection analysis OCTA(2) OCTA(l) promoter. (D) (E) quantitative extracts from a fraction of the cotransfected cells. prepared deletion into the box binding activity upon POU-specific is Two glutamine-rich stretches in the N terminus also observed with whole cell extracts (compare Figure 2A, contribute to transcription activation lane 10 and Figure 4C, lane 5), it cannot be due to In the next series of impaired experiments, we looked in more detail transport into the nucleus. at the glutamine-rich region consisting of the three glutamine- 1630 Oct-2A factor function transcription Table I. Quantification of transactivation by the deletion mutants removing the N-terminal glutamine-rich region Corresponding lanes in Figure 5 3 4 5 6 First N-terminal aa of Oct-2A 99 112 136 161 deleted glutamine-rich stretch I II III activity on OCTA(2) promoter 100% 290% 220% 80% 55% decrease in activity 1.3x 2.8x 1.5x activity on OCTA(l) promoter 100% 350% 120% 40% 35% decrease in activity 2.9x 3.0x l.lx The transcriptional stimulation has been corrected for the amount of Oct-2A protein present. For both promoters, transactivation by wild-type protein was set to 100%. The absolute increase of transcription upon cotransfection of wild-type protein was X 28x for the OCTA(2) and for the 1O OCTA(l) promoter. rich stretches Gln I, II and III that there (Figure 5A). As above, the is insufficient cotransfected Oct-2A present in experiment was with both the HeLa cells performed to out enhancer OCTA(l) and carry activation. OCTA(2) promoters. To test whether the loss of We rather speculate that either another activity octamer binding observed deletion of 12 is reproducibly upon aal -aa136 protein responsible for enhancer activation, or that (see Oct-2A Figure 3A) is due to either the deletion or itself may be altered protein target for some B cell-specific modification levels, we again quantified the it more transcriptional activation making active. Alternatively, there may be an observed with the four N-terminal deletions additional, possibly not even DNA binding, successively factor(s) present removing all three glutamine-rich stretches. The in B cells which signals would render Oct-2A mediated enhancer obtained with both the and activation OCTA(2) OCTA(1) promoter possible. Such an 'adaptor' for could, example, 5B (Figure and C) were normalized the interact with Oct-2A against bandshift via its leucine repeat. For promoter signal produced by Oct-2A in the transfected cells this factor present activity, would be dispensable since we have 5D and In (Figure E). this bandshift the excess shown that in HeLa experiment, cells, Oct-2A does not its require leucine of binding probe was sufficient to avoid between repeat. competition transiently expressed Oct-2A and Oct-I Finally, we endogenous protein. cannot rule out the existence of a As a result of this, the Oct-I signal could also be used as repression/extinction mechanism in non-lymphoid cells, an internal standard. which would affect enhancer but not promoter function. The results are summarized in Table I. On both Extinction phenomena promoters, have indeed been described for deletion of the first 99 aa leads to a octamer-containing test reproducible 3-fold genes (Junker et al., 1988, 1990; increase in - transcriptional stimulation SB Zaller et Yu et (Figure E, al., 1988; al., 1989). In this context it is lanes 2 and The of in compare 3). effect deletions the also noteworthy that transcriptional activity of another POU glutamine-rich can be region seen with the gene, the pituitary particularly specific transcription factor Pit-I with the et OCTA(i) promoter: OCTA(l) deletions (Ingraham al., 1988; Bodner et al., 1988) promoter, is cell type- of each Gln I and Gln II leads to a similar 3-fold decrease dependent: in similar cotransfection experiments in HeLa a which is about three times less leaving protein active than cells, Pit-I is able to transactivate selectively the prolactin In wild-type. the context of the and OCTA(2) promoter, only growth hormone gene promoters as well as synthetic deletion of Gln II results in a 3-fold decrease. For promoters consisting both of multiple binding sites taken from promoters, deletion of Gln Ill has a effect. either negligible Since gene (Ingraham et al., 1988). However, although Pit-I deletions in the are seen with both was still able to glutamine-rich regions transactivate a cotransfected prolactin promoters, we consider them an activation domain promoter in the important putative rat lactotroph cell line 235-1, it of Oct-2A. completely failed to activate the growth hormone gene promoter (Ingraham et al., 1988; Mangalam et al., 1989). Experiments to determine whether enhancer activation by Oct-2A could also be cell type-dependent are currently in Discussion progress. Why does Oct-2A not work from an enhancer position in HeLa cells? Activating domains of Oct-2A can compensate for Here we present evidence that Oct-2A specifically fails to each other activate x the IgH and the 6 OCTA in The data from enhancers HeLa cells, the deletion analysis of Oct-2A suggest that although the octamer containing promoter on the same Oct-2A contains several redundant transcription activating plasmid responds to cotransfected Oct-2A. The inactivity of regions on both sides of, and with possibly overlapping the the enhancer IgH might be the absence of POU domain. As explained by other long as the DNA binding domain is intact lymphocyte-specific DNA binding proteins. The and the protein is not truncated finding, from both no ends, single that even the 6 x OCTA however, enhancer was not active region seems to be essential for absolutely transcriptional upon cotransfection of Oct-2A is surprising: in HeLa cells, activation. cotransfected Oct-2A is more abundant than Two Oct-I activating domains, one in the N and one in the (Figure C 4C, lane 2). This contrasts with the situation in terminus Namalva of Oct-2A, were identified: a glutamine-rich region a B cell in cells, line which Oct-I is more abundant than in the N terminus of Oct-2A, similar to the ones in the Oct-2A. In of the 6 OCTA enhancer is factor spite this, active transcription SpI (Courey and Tjian, 1988), appears in these cells (Gerster et al., Therefore it is to be the dominant domain. An 1987). as yet undefined unlikely region 1631 M.M.Mu1ler-Immergluck, W.Schaffner and P.Matthias within the C terminus of Oct-2A can only be monitored with seems to be in some Although promoter geometry important the less permissive target promoter OCTA(l) (see below). cases et Wu and (Takahashi al., 1986; Berk, 1988), the Note that, in particular, deletions within the C terminus result sites for several binding transcription factors, including in significantly higher amounts of protein and, as a Oct-2A and can be at various distances from the Gal4, TATA consequence, elevated transcriptional activity in the box without activation impairing transcriptional (Chodosh transfected cells. et al., Wirth et Ruden et 1987; al., 1987; al., 1988). The analysis of the glutamine-rich region in the N terminus The POU region transactivates weakly by itself of Oct-2A also indicates that the OCTA(2) promoter is more than the both We also show that a truncated protein essentially consisting permissive OCTA(1) promoter. Although to deletion of the of the entire POU domain (Oct-2A/194-357) is stable in promoters respond similarly glutamine- rich stretch Gln the of the vivo and sufficient for DNA binding (for in vitro data, see II, activity OCTA(1) promoter is more affected than the deletion of Gln I LeBowitz et al., 1989). Most strikingly, this protein is still OCTA(2) by (see Table able to activate transcription from the OCTA(2) promoter, I). From our it is evident albeit with a low efficiency (Figure 4A). Thus, a third experiments that the reporter gene can be more transcription activating region of Oct-2A appears to overlap or less permissive for either full length or with DNA A cluster of six acid residues truncated Oct-2A the binding domain. factor. In this test system, the architecture at of domain between aa of the can determine the the beginning the POU-specific promoter (and enhancer) activity of a factor to a certain extent. Hence such 195-204 (see Figure 3) could be responsible for this given transcription an of DNA an with more than one transcriptional activation. Such overlap binding analysis should be done promoter domain is observed with the at a time. and activating prokaryotic phage X where the amino acids repressor (Ptashne, 1988), activating interwoven with the helix-turn-helix DNA Are different Oct-2 factors for at are binding required fine-turning et This also be the case the level? domain (Bushman al., 1989). may transcriptional for the glucocorticoid receptor (Hollenberg et al., 1987; Even our truncated Oct-2A and the though promoter Miesfeld et al., 1987). constructions do not exist as such in we that nature, suspect such a sensitive In this we note that the helix-turn-helix and differential of similar test context, region response of the POU-homeo domain of the to the same well reflect of a ubiquitous transcription promoters protein may part natural factor Oct-l contributes indirectly to transcriptional activation regulatory mechanism. Within the framework of this via a specific protein-protein contact. In elegant different variants of Oct-2 concept, splicing proteins could it has been identified as the essential have different affinities to various experiments, region octamer-containing for selective interaction with the virus transactivator promoters. This could be used to activate preferentially one herpes or the other octamer protein VP16 (or Vmw65; Stern et al., 1989). Fusions with and containing lymphocyte-specific will be Sherman et It heterologous DNA binding domains required to promoter (e.g. al., 1989). has been shown of and other that of activated dissect further the potential this putative expression promoters by the Drosophila bicoid on both the activating domains of the Oct-2A factor. morphogen critically depends concen- tration of bicoid and the of the sites affinity binding (Struhl not abolish et al., Driever et Lack of the Leu-zipper does promoter 1989; al., 1989). In the of the recent identification of activation light many additional with both the and the octamer in various tissues et The reporter plasmids OCTA(2) binding proteins (Schoeler al., contained an SV40 enhancer we would this to be OCTA(1) promoter 1989a,b), expect protein family involved downstream of the test in order to the in a network. gene augment signals. complex regulatory We show that of the the independent promoter used, enhancer effect a is still manifest with (i.e. strong signal) Materials and methods an Oct-2A mutant that lacks the leucine This indicates repeat. that, at least in the case of the SV40 enhancer, the Construction and of precise nomenclature deletion mutants cDNA deletions were introduced into the Oct-2A cDNA an (in appropriate hypothetical leucine zipper of Oct-2A cannot be involved nested deletions with either BaI3 a subclone) by creating 1, combination of in the enhancer for mediating effect by allowing, example, exonuclease and nuclease SI or natural mI by using restriction sites (Maniatis direct interactions of Oct-2A with promoter-bound proteins et The deletion of all clones was determined al., 1982). endpoint by dideoxy to the enhancer. bound SV40 Deleted cDNA molecules were inserted into vectors which sequencing. allow of deleted as an in-frame expression any N-terminally protein fusion product with the first four amino acids of the herpes simplex virus TK gene (Matthias Influence of different target promoters et Since our cDNA starts at the second amino acid of al., 1989). the extended The OCTA(2) promoter, in which the octamer site is only of Clerc et sequence al. (1988), we now use the numbering of these authors 5 from the TATA bp away box, together with an SV40 this work. For throughout simplicity we consider pO/2-479 to encode the enhancer downstream, yields the most permissive test which we call In wild-type protein Oct-2A/2-479. our nomenclature, the first number before the dash (0, 1, 2) designates the reading frame of the system. This is indicated by the fact that the POU domain vector used et expression (Matthias al., 1989) and the following two numbers itself (Oct-2A/194-357) shows some activity. However, that (which are used to define the protein in the drawing) represent the first and same POU domain is protein unable to activate the similar last amino acids of the Oct-2A sequence. octamer-TATA box promoter OCTA(l), where the octamer Using the one letter code, the translated linker sequence of the N-terminally is 20 bp away from the TATA box (compare lanes 4 and truncated clones in front of the Oct-2A sequence is as follows: pO/2479 (Oct-2A/2-479), MASWGSGTELEFR; pO/50479, MASWGSGTP; 5 in 4A and Figure D). pO/56479, MASWGSGT; p1/99479, MASWGSGVP; 12479, pIl/l Similar to truncated Oct-2A, transactivation by severely MASWGSGVP; p2/136479, MASWGSGY; p2/161479, MASWGSGY; truncated on a short distance between its Gal4 depends pl/187479; MASWGSGVP; pO/194479, MASWGSGT; p2/206479, site and the TATA box et MASWGSGYP; binding (Ruden al., MASWGSGTP. 1988). p2/207479, MASWGSGYP; pO/210479, 1632 factor function Oct-2A transcription Bodner,M., Castrillo,J.L., Theill,L.E., Deerinck,T., Ellisman,M. and The C-terminal deletions start with Oct-2A aa number 2 and were Karin,M. (1988) Cell, 55, 505-518. constructed using a similar expression vector, providing stop codons in all Bushman,F.D., Shang,C. and Ptashne,M. (1989) Cell, 58, 1163-1171. three reading frames behind the end of the Oct-2A sequence (Matthias et al., Burglin,T.R., Finney,M., Coulson,A. and Ruvkun,G. (1989) Nature, 341, 1989). The translated linker sequence of the C-terminally truncated clones 239-343. behind the Oct-2A sequence is as follows: p3S/2460, GGT; p3/2-427, PGT; Chodosh,L.A., Carthew,R.W., Morgan,J.G., Crabtree,G.R. and p3S/2-392, GYLAS; p3S/2-370, GVPS; p3S/2-357, GYLAS; p3S/2-344, Sharp,P.A. (1987) Science, 238, 684-688. VPS. Clerc,R.G., LeBowitz,J.H., Baltimore,D. and Sharp,P.A. Corcoran,L.M., Plasmids encoding Oct-2A proteins deleted at both ends were made by (1988) Genes Dev., 2, 1570-1581. recombining N-terminal and C-terminal deletion plasmids via the Sfil sites Courey,A.J. and Tjian,R. (1988) Cell, 55, 887-898. present in the Oct-2A POU domain and in the SV40 origin fragment of Cox,P.M., Temperley,S.M., Kumar,H. and Goding,C.R. (1988) Nucleic the expression vector. Acids Res., 16, 11047-11056. Dreyfus,M., Doyen,N. and Rougeon,F. (1987) EMBO J., 6, 1685-1690. Construction of clones bearing various enhancers Driever,W., Thoma,G. and Nusslein-Volhard,C. (1989) Nature, 340, The constructs (shown in Figure 1) bearing various enhancers in an EcoRI 363 -367. site downstream of the reporter gene (Westin et al., 1987) were derived Eckner,R. and Birnstiel,M.L. (1989) Nucleic Acids Res., 17, 5947 -5959. from the OCTA(2) promoter construct (containing the octamer containing Falkner,F.G. and Zachau,H.G. (1984) Nature, 310, 71 -74. oligo T.Gerster, unpublished results) with the SV40 enhancer D6, Falkner,F.G., Mocikat,R. and Zachau,H.G. (1986) Nucleic Acids Res., downstream (Figure 4D, Gerster et al., 1987, Muller-Immergluck et al., 14, 8819-8827. 1988). The SV40 enhancer containing BamHI-KpnI (Asp418) fragment Finney,M., Ruvkun,G. and Horvitz,H.R. (1988) Cell, 55, 757-769. (Gerster et al., 1987) was replaced by the corresponding fragment containing Fletcher,C., Heintz,N. and Roeder,R.G. (1987) Cell, 51, 773-781. the IgH enhancer (Gerster et al., 1987) or the enhancerless fragment from Gehring,W.J. (1987) Science, 236, 1245-1252. OVEC-1 (Westin et al., 1987). The 1 x OCTA and 6 x OCTA enhancers Gerster,T., Matthias,P., Thali,M., Jiricny,J. and Schaffner,W. (1987) (1 x, 6 x, Gerster et al., 1987) were cloned correspondingly by replacing EMBO J., 6, 1323-1330. the BamHI-XbaI fragment of the reporter plasmid. Harvey,R.P., Robins,A.J. and Wells,J.R.E. (1982) Nucleic Acids Res., 10, 7851 -7863. Preparation of nuclear mini extracts He,X., Treacy,M.N., Simmons,D.M., Ingraham,H.A., Swanson,L.W. and yg of expression plasmid DNA was transfected into Cos-7 cells by the Rosenfeld,M.G. (1989) Nature, 340, 35-42. DEAE-dextran procedure and nuclear mini-extracts were prepared after Heer,W,. Clerc,R.G., Corcoran,L.M., Baltimore,D. Sturm,R.A., 2 days, essentially as described (Muller-Immergluck et al., 1989; Schreiber Sharp,P.A., Ingraham,H.A., Rosenfeld,M.G., Finney,M. Ruvkun,G. et 1989a). al., and Horvitz,H.R. (1988) Genes Dev., 2, 1513-1516. Hollenberg,S.M., Giguere,V., Segui,P. and Evans,R.M. (1987) Cell, 49, of whole cell extracts Preparation 39-46. 10 cm plate of transfected cells was washed with 10 ml TBS and One Ingraham.H.A., Chen,R., Mangalam,H.J., Elsholtz,H.P., Flynn,S.E., from the plate with a rubber policeman in 1.2 ml TBS (+2 mM scratched Lin,C.R., Simmons,D.M., Swanson,L. and Rosenfeld,M.G. (1988) Cell, EDTA). 0.3 ml of cell suspension (25%) were pelleted for 5 s in an 519-529. centrifuge and resuspended by vortexing in 40 Il X-400 buffer 55, Eppendorf Junker,S. Nielsen,V., Matthias,P. and Picard,D. (1988) EMBO J., 7, buffer from Schreiber et al. (1988) adjusted to 400 mM KCI] [X-50 3093 -3098. complemented with 1 mM PMSF, 4 leupeptin and 10 mM DTT. Ag/ml Schreiber,E. and Matthias,P. (1990) Cell, in press. Junker,S., Pederson,S., cell extracts were prepared as described by Kumar and Chambon Whole Ko,H., Fast,P., McBride,W. and Staudt,L. (1988) Cell, 55, 135-144. four cycles of freezing and thawing. (1988) by Kumar,V. and Chambon,P. (1988) Cell, 55, 145 - 156. Sive,H.L. Roeder,R.G. and Heintz,N. (1988) Genes Dev., 2, Analytical bandshift experiments LaBella,F., 32 -39. Analytical bandshift experiments were performed as described in Schreiber LeBowitz,J.H., Clerc,R.G.. Brenowitz,M. and Sharp,P.A. (1989) Genes and Muller-Immergluck et al. (1989). et al. (1989) 3, 1625-1638. Del., Lenardo,M.J., Robbins,P., Kuang,A. Mulligan,R.C. and Staudt,L., Quantitative bandshift experiments (1989) Science, 243, 544-546. Baltimore,D. 2 of whole cell extract were incubated for 10 min at room temperature jig Maeda,H., Araki,K., Kitamura,D., Wang,J. and Watanabe,T. (1987) in 15 1il 1 x BEBBI (Muller-Immergluck et al., 1988) in the presence of Nucleic Acids Res., 15, 2851-2869. 4 ug poly(dIdC) and 18 000 c.p.m. (2.3 fmol) of labelled octamer binding Mangalam,H.J., Albert,V.R., Ingraham,H.A., Kapiloff,M., Wilson,L., site (Muller-Immergluck et al., 1988). All bandshift signals from the Oct-2 Nelson,C., Elsholtz,H. and Rosenfeld,M.G. (1989) Genes Dev., 3, proteins, normalized against the signal of Oct- 1, were quantified by cutting 946-958. out the radioactive bands and Cerenkov counting. Fritsch,E.F. and Sambrook,J. (1982) Molecular Cloning. A Maniatis,T., Laboratory Cold Spring Harbor Laboratory Press, Cold Spring Manual. Transfections for transactivation Harbor, NY. HeLa cells were transfected with 8 reporter plasmid OCTA(2) or jig Williams,G.T. and Neuberger,M.S. (1985) Cell, 41, 479-487. Mason,J.O., OCTA(1) (Muller-Immergluck et al., 1988), 4 Oct-2A expression vector, Ag Mattaj,I.W., Lienhard,S.. Jiricny,J. and De Robertis,E.M. (1985) Nature, reference plasmid and 7 sonicated herring sperm DNA as carrier /g jig 316, 163-167. et al., 1988). Two days later, cytoplasmic RNA was (Muller-Immergluck Matthias.P., Muller-Immergluck,M.M., Schrieber,E., Rusconi,S. and and analysed by RNase protection. Except for Figure 1, all levels extracted Schaffner,W. (1989) Nucleic Acids Res., 17, 6418. normalized against the signal of the reference gene, were of transcription, Miesfeld,R., Godowski,P.J., Maler,B.A. and Yamamoto,K.R. (1987) cutting out the radioactive bands and Cerenkov counting. quantified by Science, 236, 423-427. Muller-Immergluck,M.M., Ruppert,S., Schaffner,W. and Matthias,P. Acknowledgements (1988) Nature, 336, 544-551. Muller-Immergluck,M .M., Schreiber,E., Schaffner,W. and Matthias,P. We thank Edgar Schreiber for providing Figure 4C, Andreas Czank for (1989) Nucleic Acids Res., 17, 6420. excellent technical assistance. Hans-Peter Muller-Storm, Keith Harshman, O'Neill,E.A. and Kelly,T.J. (1988) J. Biol. Chem., 263, 931-937. Deborah Maguire and Markus Thali for critical reading of the manuscript Parslow,T.G., Blair,D.L., Murphy,W.J. and Granner,D.K. (1984) Proc. and valuable comments, Fritz Ochsenbein for careful preparation of the Natt. Acad. Sci. USA, 81, 2650-2654. We are indebted to T.Gerster, Iris Kemler and R.Roeder as well figures. Pruijn,G.M.J., van Driel,W. and van der Vliet,P.C. (1986) Nature, 322, as to W.Herr and M.Tanaka for communicating results prior to publica- 656-659. This work was supported by the Swiss National Science Foundation tion. Ptashne,M. (1988) Nature, 335, 683-689. and by the Kanton of Zurich. Rosales,R., Vigneron,M., Macchi,M., Davidson,I., Xiao,J.H. and Chambon,P. (1987) EMBO J., 6, 3015-3025. References Ruden,D.M., Ma,J. and Ptashne.M. (1988) Proc. Natl. Acad. Sci. USA, 85, 4262-4266. and Busslinger,M. (1987) Cell, 50, 347-359. Superti-Furga,G. Barberis,A., Heguy,A. and Roeder,R.G. (1987) Cell, 51, 783-793. Scheidereit,C., and Busslinger,M. (1989) Nucleic Acids Res, Barberis,A., Superti-Furga,G. Cromlish,J.A., Gerster,T., Kawakami,K., Balmaceda,C., Scheidereit,C., 17, 1633 M.M.Muller-Immergluck, W.Schaffner and P.Matthias Currie,R.A. and Roeder,R.G. (1988) Nature, 336, 551-557. Scholer,H.R., Hatzopoulos,A.K.. Balling,R., Suzuki,N. and Gruss,P. (1989a) EMBO J., 8, 2543-2550. Scholer,H.R., Balling,R. Hatzopoulos,A.K., Suzuki,N. and Gruss,P. (1989b) EMBO J., 8, 2551-2557. Schreiber,E., Matthias,P., Muller-Immergluck,M.M. and Schaffner,W. (1988) EMBO J., 7, 4221-4229. Schreiber,E., Matthias,P., Muller-Immergluck,M.M. and Schaffner,W. (1989a) Nucleic Acids Res., 17, 6419. Schreiber,E., Muller-Immergluck,M.M., Schaffner,W. and Matthias,P. (1989b) Renkawitz,R. (ed.), Tissue Specific Gene Expression, Verlag Chemie, Weinheim, FRG. Sherman,P.A., Basta,P.V., Heguy,A., Wloch,M.K., Roeder,R.G. and Ting,J.P.Y. (1989) Proc. Natl. Acad. Sci. USA, 86, 6739-6743. Singh,H., Sen,R., Baltimore,D. and Sharp,P.A. (1986) Nature, 319, 154-158. Staudt,L.M., Singh,H., Sen,R. Wirth,T. Sharp,P.A. and Baltimore,D. (1986) Nature, 323, 640-643. Staudt,L.M., Clerc,R.G., Singh,H., LeBowtiz,J.H., Sharp,P.A. and Baltimore,D. (1988) Science, 241, 577-580. Stern,S., Tanaka,M. and Herr,W. (1989) Nature, 624-630. 341, Struhl,G., Struhl,K. and Macdonald,P.M. 1259-1273. (1989) Cell, 57, Sturm,R.A. and Herr,W. (1988) Nature, 601-604. 336, Sturm,R.A., Das,G. and Herr,W. Genes (1988) Dev., 2, 1582-1599. Takahashi,K., Vigneron,M., Matthes,H., Wildeman,A., Zenke,M. and Chambon,P. (1986) Nature, 319, 121-126. Tanaka,M., Grossniklaus,U., Herr,W. and Hernandez,H. (1988) Genes Dev., 2, 1764-1778. Thali,M., Muller-Immergluck,M.M., DeLorenzi,M., Matthias,P. and Bienz,M. (1988) Nature, 336, 598-601. Westin,G., Gerster,T., Miuller-ImmergKick,M.M., Schaffner,G. and Schaff- ner,W. (1987) Nucleic Acids Res., 15, 6787-6798. Wirth,T., Staudt,L. and Baltimore,D. (1987) Nature, 329, 174-178. Wu,L. and Berk,A. (1988) Genes Dev., 2, 403-411. Yu,H., Porton,B., and Shen,L. Eckhardt,L.A. (1989) Cell, 58, 441-448. Zaller,D.M., and Yu,H. Eckhardt,L.A. (1988) Mol. Cell. Biol., 8, 1932-1939. Zenke,M., Grundstrom,T., Matthes,H., Wintzerith,M., Schatz,C., Wildeman,A. and Chambon,P. (1986) EMBO J., 5, 387-397. Received on December 22, 1989; revised on February 19, 1990

Journal

The EMBO Journal – Springer Journals

Published: May 1, 1990

There are no references for this article.