Eur. J. Biochem. 207,195-200 (1992) @) FEBS 1992
Enhanced cell-free transcription of the ribosomal protein L3 gene by the polyoma virus enhancer PEA3 DNA-binding protein Thillainathan YOGANATHAN ’, Alison COWIE’, John A. HASSELL’ and Bruce H. SELLS‘ Molecular Biology and Genetics, College of Biological Science, University of Guelph, Ontario, Canada The Institute for Molecular Biology and Biotechnology, McMaster University, Hamilton, Ontario, Canada (Received January 1 , 1992) - EJB 92 0074
The mouse-ribos~mal-protein-L~~-gene promoter contains a 12-bp sequence motif within the 5’upstream region termed the p element which shows significant similarity with the consensus sequence of the polyoma-virus-enhancer PEA3. A cloned PEA3 DNA-binding protein, expressed in Escherichiu coli and purified, activates the expression of the ribosomal-pr0tein-1~~ gene in a cell-free system. Moreover, the PEA3 protein participates in the formation of the ribosomal-protein-L32-promoterpreinitiation-transcription complex. The preinitiation complex formed with PEA3 is resistant to competition by oligonucleotides containing the p element. In addition anti-PEA3 serum interacts with a factor in mouse L1210 nuclear extract that binds to the /3 element, causing a supershift in a mobility-shift assay. Our study demonstrates for the first time that the PEA3 protein can transactivate a cellular gene in a cell-free transcription system.
The ribosome particle is composed of more than 70 different ribosomal proteins whose formation is coordinately regulated such that each protein is produced in equimolar amounts. How this coordination is achieved is not yet clear since the genes for these proteins are unclustered in the eucaryotic genome and are located on different chromosomes [l]. One approach to solving this question is to examine the promoter regions of the ribosomal-protein genes to identify the factors which interact with specific sequences and to determine whether any of these factors play a role in coordinating transcription. Transcription of eucaryotic genes is controlled by the combined action of several sequence-specific DNA-binding proteins [2, 31. Basic transcription factors and RNA polymerase bind in the vicinity of the transcription-initiation site to ensure basal levels of transcription, while elevated levels are induced by additional factors which bind to regulatory elements located some distance upstream of the start site. These latter protein factors influence the rate of initiation by interacting with the transcriptional apparatus. Several of the proteins associated with the pol-11-promoter enhancer and transactivator regions have been identified [2, 31. These transactivators are generally bipartite in structure, containing a DNA-binding domain and a transcription-activation domain [3]. The DNA-binding domains of these regulatory proteins form a variety of structures which make contact with DNA, i.e. zinc finger or helix-turn-helix motifs [2, 31. More recently a new class of DNA-binding proteins which contains none of the known DNA-binding structures has been identified. These proteins are represented by the E-twenty-six-avian-leukaemia-
virus-specific (ETS) family of proto-oncogene products which includes the PEA3 DNA-binding protein [4 - 71. In the present study we have focused our attention on the gene encoding the ribosomal-protein L32 (rpL3J. This gene has a promoter region with several distinctive features. It lacks both a TATA box and an Spl-binding site [8]. Additional studies have revealed that the upstream region of the rpL32 gene (nucleotides -80 to -69) contributes to the efficiency of its transcriptional activity [9, 101. In a recent report we have identified a 50 - 55-kDa protein, the B factor, which binds to the sequence in this (nucleotides - 80 to - 69) region and participates in the cell-free transcription of this gene [lo]. We have postulated that the p element in the r ~ L ~ ~ - g promoter ene acts as an upstream transactivating element. The p element has significant sequence similarity to the PEA3 motif of the polyoma virus enhancer 14, 81. A protein-binding activity has been identified in mouse cell nuclear extracts that recognises the PEA3 sequence [l 11. Furthermore, a cDNA that encodes PEA3-binding activity has been cloned by screening an expression library with a probe containing the PEA3 motif [7]. Based on its DNA-binding specificity, we postulated that the PEA3 protein activates m o ~ s e - r p L ~ ~ - g e n e expression through its interaction with the j? element. We have, therefore, investigated the function of the PEA3 DNAbinding protein in a mouse cell-free transcription system in an attempt to determine whether PEA3 and the p factor are interchangeable in stimulating transcription of the rpL32gene. In this study, we report that PEA3 DNA-binding protein transactivates the rpL,, gene through the p element of this promoter.
Correspondence to B. H. Sells, College of Biological Science, Molecular Biology and Genetics, University of Guelph, N1G 2W1 Guelph, Ontario, Canada Ahhreviations. rp, ribosomal protein; ETS, E-twenty-six-avianleukaemia-viros-specific; GST, glutathione-S-transferase. Enzyme. Restriction endonucleases (EC 3.1.21.4).
EXPERIMENTAL PROCEDURES Cell culture and nuclear-extract preparation Mouse L1210 cells were maintained in a rotating culture in Dulbecco’s modified Eagle’s medium supplemented with
196 5% fetal calf serum. Nuclear extracts were obtained using the procedure of Dignam et al. [12]. Nuclear extracts were fractionated on DEAE-cellulose (Whatman DE52) according to the protocol of Zahradka et al. [13].
A Poiyomavirus enhancer
5’-CAGGAAGTGACT-3’
rpL32 fl element 5’-CCGGAAGTGCTT-3’
Plasmids
m 622
3 09
Individual reaction mixtures contained 12.5 mM Hepes, pH 7.9, 5 mM MgC12, 70 mM KC1, 10% glycerol, 500 pM (each) ATP, CTP and GTP, 25 pM [32P]UTP,0.4 mM EDTA, 0.5 mM dithiothreitol, 100 U/ml RNA guard (Pharmacia), 0.25 pg plasmids DNA and 1 - 3 p1 DEAE chromatography fraction of L1210 nuclear extract, in 25 pl total volume. Samples were incubated for 60 min at 30°C and the reactions were arrested by the addition of 175 pl stop buffer [14]. The transcripts were deproteinized, precipitated with ethanol and separated by 4% urea/PAGE and radioactivity detected by autoradiography.
Synthetic oligonucleotides used in these studies were produced using a Biosearch 8600 DNA synthesizer. Each oligonucleotide was HPLC purified and the purity checked by running on a preparative 20%-polyacrylamide/urea sequencing gel. Double-stranded oligonucleotide was made by annealing the chemically synthesised oligonucleotides essentially as described in [15]. Gel-mobility-shift assay The radiolabelled fragments used for the gel-mobility-shift assays were prepared essentially as described by Wall et a1.(1988) [I 51. A specific double-stranded oligonucleotide that contained the sequence in the p-element recognition site was prepared by annealing the following chemically synthesized oligonucleotides, 32P-labelled using a kinase: 5’-CCCAGAGCCGGAAGTG-3’ and 5’-TGGGCACTTCCGGCTC-3’. Gel-mobility-shift assays were performed under transcription conditions, except that the NTP were omitted. 1.5 pg poly [d(I-C)] and 0.1 ng radiolabelled oligonucleotide fragments were incubated with nuclear extract at 30°C for 30 min and protein-DNA complexes resolved by electrophoresis on a 4% polyacrylamide gel in a Tris/glycine running buffer. Cloning and purification of PEA3 protein A partial cDNA encoding the carboxy-terminal(38 kDa) of PEA3 was expressed in Escherichia coli as a fusion protein with glutathione-S-transferase (GST; Pharmacia). The protein was purified by affinity Chromatography on glutathioneSepharose as described by the manufacturer (Pharmacia). The purified GST-PEA3 fusion protein migrated as a single band at the expected molecular mass on SDSjPAGE [7].
at-
urn
- - P O
Cell-free transcription
Synthetic oligonucleotides
0
B
Plasmid P3AR2.8, containing a fragment of the murine rpL32gene was obtained from R. P. Perry (Institute of Cancer Research, Philadelphia, PA); plasmid pAd97, containing the adenovirus major late promoter was obtained from N. Miyamoto (Ontario Cancer Institute, Toronto) and the histone-H4 plasmid from G. Stein (University of Massachuset t s) .
147
1
2
3
4
Fig. 1. The nucleotide sequences of the PEA3 enhancer and rpL,, element and enhancement of transcription by PEA3 protein in vitro. (A) Similarities between the nucleotide sequence of the PEA3 enhancer and rpL32 element. (B) Transcription enhancement by PEA3 protein in vitro. Template plasmids were digested with EcoRl. Molecularmass standards (lane 1); control (lane 2); 100 ng purified PEA3 protein (lane 3); 100 ng purified GST protein (lane 4).
Preparation of anti-PEA3 serum The antiserum used in this study was generated in rabbits using a bacterially expressed PEA3 protein. RESULTS Transcription activation by purified PEA3 protein To determine the effect of PEA3 protein on rpL3, transcription, we have conducted cell-free transcription of When the rpL32 gene was transcribed in L1210 extracts, the templates generated a 1500-bp nucleotide-run-off transcript corresponding to specific initiation at the in vivo cap site (Fig. lB, lane 2). Addition of PEA3 protein stimulated this specific transcription (Fig. lB, lane 3). Addition of GST had no effect on the transcription (Fig. lB, lane 4). These results indicate that PEA3 protein activated transcription of To further evaluate whether this transcriptional activation occur with other pol-I1 genes, we have performed transcription experiments with histone and adenovirus major-late-promoter genes. Purified-PEA3-fusion-protein transactivation specific to the rpL,, gene To determine the effect of PEA3 upon transcription of other pol-I1 genes, the following experiments were performed.
197 A rpL32
- ,PEA3 -,,
GST
< 1
2
B Histone H4
3
m
-
a
n
4
5
I-
1
m
U
< 1
2
3
C
< 1
2
3
Fig. 2. Effects of PEA3 protein on the transcriptional activities of various promoters. Nucleotide-run-off assays were performed using different genes. (A) EcoRI-digested rpLjz gene; lane 1, containing no PEA3-GST; lanes 2 and 3, containing 100 ng and 200 ng of PEA3GST fusion protein respectively; lanes 4 and 5, containing 200 ng and 200 ng of GST respectively. (B) XboI-digested histone-H4 gene; lane 1, no PEA3; lane 2, with 200 ng PEA3-GST fusion protein; lane 3, with 200 ng GST. (C) Ad MLP: AatII-digested pAd97 containing the adenovirus 2 major late promoter; lane 1, no PEA3; lane 2, with 200 ng PEA3-GST protein; lane 3, with 200 ng GST protein. Specific transcripts are indicated throughout by arrows.
In the first the PEA3-GST fusion protein, synthesized in E. roli and purified, was incubated with a DNA template containing the m o ~ s e - r p L ~ ~ - g epromoter ne and a partially purified mouse L1210 nuclear extract (see Experimental Procedures). Fig. 2A shows that addition of the purified PEA3 fusion protein (100 ng or 200 ng) to the reaction mixtures increased transcription of rpL32gene (Fig. 2A, lanes 2 and 3). Addition of the same amount of GST protein alone produced no stimulation (Fig. 2A, lanes 4 and 5). Furthermore, inclusion of the PEA3-GST fusion protein (200 ng) in the transcription-reaction mixture containing the histone-H4 promoter resulted in no significant change in transcription (Fig.2 B, lanes 1-3). Similarly, addition of the PEA3-GST fusion protein (200 ng) produced no changes in adenovirus major late promoter transcription (Fig. 2C, lanes 1- 3). These data imply that PEA3 functions as a specific transcriptional activator of the mouse rpL32gene. In a second series of experiments to assess the involvement of PEA3, nuclear extract was incubated with the oligonucleotide containing the element to deplete the reaction mixture of the p factor. Fig. 3 (lane 2 ) reveals that following incubation the resulting extract is incapable of supporting transcription of the rpL32 gene. This reduction in transcription is specific to the p-element sequence since an unrelated oligonucleotide of the same size failed to depress transcription of the rpL32 gene (Fig. 3, lane 6). To determine whether the factor(s) removed by incubation with the p-element sequence could be replaced by PEA3, depleted extracts were incubated with the PEA3-GST fusion protein. The results indicate that addition of the fusion protein restored the ability of the depleted extract to transcribe the r ~ L , ~ - g e n(Fig. e 3, lanes 4 and
2
3
4
5
6
Fig. 3. Transcriptional activation of the rpL32 gene by PEA3 is dependent upon binding to a specific upstream sequence. Oligonucleotide competition of the transcription reactions supplemented with PEA3 protein is shown. Transcription assays and reaction conditions were as described in Experimental Procedures. Specific and control oligonucleotides were incubated with 4 kl DEAE-chromatography fraction for 15 min prior to addition of template. In addition, transcription reaction mixtures were supplemented with various amounts of PEA3-GST protein or GST. Lane 1 , no oligonucleotide; lane 2. 200 ng specific oligonucleotide ;lane 3,200 ng specific oligonucleotide and 200 ng GST; lane 4,200 ng specific oligonucleotide and 100 ng PEA3-GST fusion protein; lane 5,200 ng specific oligonucleotide and 200 ng PEA3-GST protein; lane 6, 200 ng non-specific oligonucleotide alone. Oligo., oligonucleotide; un., non-specific.
5) whereas addition of GST alone failed to restore transcription( Fig. 3, lane 3). These results suggest that the PEA3 protein is involved in DNA binding and stimulation of rpL32gene transcription. PEA3 protein forms a transcription-initiation complex with the rpL3*-genepromoter
These studies have been designed to determine the stage of involvement of the PEA3 protein in transcription by employing a cell-free system. The following experiments were performed to assess whether this protein functions in the formation of the preinitiation complex and whether it remains committed to the template following complex formation. Preinitiation complexes were prepared as described by Hai et al. (1988) [16], and the reaction was arrested at this stage by incubation in the absence of NTP. Transcription was then initiated by the addition of NTP and the transcripts generated in 60 min were analyzed. Fig. 4 (lanes 1 and 2) illustrates the level of transcription with or without PEA3 protein. Addition of 200 ng jelement to the reaction mixture before assembly of the preinitiation complex substantially reduced transcriptional stimulation by the PEA3-GST fusion protein (Fig. 4, lane 3). In contrast, addition of the p element following formation of the preinitiation complex resulted in no significant effect on the activation induced by PEA3 (Fig. 4, lane 4). These observations suggest that PEA3 is required during assembly of the preinitiation complex and that this complex is a tightly bound one and resistant to competition by exogenously added jelement. Moreover, addition of the same amount of a control oligonucleotide, either before or after preinitiation, had no effect on enhancement by PEA3 (Fig. 4, lanes 5 and 6). These results infer that the PEA3 protein functions as a transcriptional activator in the rpL32-cell-free-transcriptionsystem and participates in the formation of the preinitiation complex. PEA3 protein binds to the fi element
The PEA3 binding site within the polyoma virus enhancer and the rpL32 p element have considerable nucleotide se-
198 0
0
0
0 9
.9 .9
2 0
6 6
2 5
c
Q
C
&,"s i-\
PEA3
1
2
3
4
5
- + + ++
6
Fig. 4. PEA3 participates in the stabilization of preinitiation complexes. Transcription assay and the reaction conditions are as described in Experimental Procedures. Lane 1, control transcription-reaction mixture without PEA3; lane 2, transcription-reaction mixture with 150 ng PEA3; lane 3, transcription-reaction mixture with 150 ng PEA3 and 200 ng p-element oligonucleotide (oligonucleotide added prior to the formation of the preinitiation complex); lane 4, transcription-reaction mixture with 150 ng PEA3 and 200 ng /I-element oligonucleotide (oligonucleotide added after formation of the preinitiation complex); lane 5, transcription-reaction mixture with 150 ng PEA3 and 200 ng control oligonucleotide (added prior to formation of the preinitiation complex); lane 6, transcription-reaction mixture with 150 ng PEA3 and 200 ng non-specific oligonucleotide (added after formation of the preinitiation complex).
quence similarity (Fig. 1A). This sequence similarity suggests that the PEA3-GST fusion protein should bind to the rpL3, fl element. To test this hypothesis, cloned and purified PEA3 protein was included in a gel-mobility-shift experiment. PEA3 protein was able to bind to the radiolabelled fl element to produce a radiolabelled complex migrating more slowly than the radiolabelled oligonucleotide (Fig. 5). Formation of the complex was abolished by the addition of excess fl element, whereas the same amount of a control oligonucleotide had no effect on complex formation (Fig. 5, lanes 3 - 5). Binding of PEA3 to the rpL,,-gene-promoter element suggests that the PEA3 polypeptide contains or shares DNA-binding specificity with the factor. Specificity of the nuclear protein binding to the 1 element
To assess the nature of the fi factor present in the transcriptionally active nuclear extract we have carried out gel-mobility-shift assays. A gel-mobility-shift-assay using a fl-element probe, produced two specific complexes (Fig. 6A). The formation of these complexes, designated CI and CII, was inhibited by the addition of specific oligonucleotide (Fig. 6A, lanes 2 and 3) but not by the non-specific oligonucleotide (Fig. 6A, lanes 4 and 5). These results suggest that these nuclear proteins are specific to the fl element. Antibodies directed against PEA3 recognize the 1 factor
To further understand the relationship between the fl factor and PEA3 protein, the influence of antisera raised against PEA3 was examined in a gel-mobility-shift assay. Nuclear extract was incubated with anti-PEA3 serum and tested for its ability to form a complex with the fl element. From Fig. 6B it is evident that the amount of the more-rapidly migrating complex (CI) was reduced following incubation of the nuclear extract with antibodies to PEA3. The more slowly
1 2 3 4 5
Fig. 5. Gel-mobility-shift assays showing that PEA3 protein binds to the radiolabelled j element. Lane 1, no PEA3 protein; lane 2, 50 ng PEA3 protein; lanes 3 and 4,50 ng PEA3 protein plus a 50-fold molar excess and a 25-fold molar excess of non-radiolabelled p oligonucleotide, respectively; lane 5, 50 ng PEA3 protein and a 50-fold molar excess of non-specific oligonucleotide.
migrating complex (CII) was unaffected. The preimmune serum failed to modify the migration of either of these bands (data not shown). In addition to reducing the formation of the CI complex, the presence of the anti-PEA3 serum gave rise to a higher-molecular-mass complex containing the fi element resulting in the production of a supershift (CIII). As the antibody concentration was increased, the higher-molecular-mass complex became more intense (Fig. 6, lanes 2-4). These data suggest that the more-rapidly migrating band (CI) is composed of a protein in the nuclear extract structurally related to PEA3. DISCUSSION The mouse-rpL3,-gene promoter contains a 12-bp sequence motif within the upstream region termed the fl element. This fl element displays significant similarity with the consensus sequence of the polyoma-virus enhancer PEA3 (Fig. 1A) Our current studies suggest that the protein interacting with the PEA3-binding site also binds to the B element and induces transcription of the rpL3, gene. Induction appears to be specific to the rpL3, gene since addition of PEA3 protein to a cell-free transcription system containing either the adenovirus major-late-promoter gene or the histone-H4 gene fails to activate transcription of these genes. These observation are not surprising since a DNA sequence resembling the PEA3-binding site is absent in these latter two gene promoters. Several reports have indicated that ETS protein represents a new class of DNA-binding protein which activates transcription in vivo [4 - 61. No published information currently exists demonstrating transcriptional activation by this DNA-binding protein using a cell-free transcription system.
199
A
CII
c
1
2
3
4
5
A n t i PEA3
A Clll
+
CI I
4
CI
4
+ 1
2
3
Super Shift
Free Probe
4
Fig. 6. Gel-mobility-shit assays using oligonucleotide-competitionand antibody binding. (A) Gel shift assay with oligonucleotide-competition analysis. Nuclear extract was incubated with the p probe in the absence of competitor (lane 1); in the presence of a 50-fold molar excess and a 25-fold molar excess of double-stranded oligonucleotide sequence containing the element, respectively (lanes 2 and 3); in the presence of 50-fold molar excess and 25-fold molar excess of non-specific double-stranded oligonucleotide, respectively (lanes 4 and 5). CI and CII indicate the positions of the complexes observed. Spec. oligo., specific oligonucleotide. (B) Formation of a supershift complex with the p factor using anti-PEA3 serum. DNA-binding mobility-shift mixtures containing L1210 nuclear extract were incubated with various amount of anti-PEA3 serum. Lane 1 , no antiserum; lanes 2-4, 0.25, 0.50 and 1.0 pl anti-PEA3 serum, respectively.
Recent studies [7] on the amino-acid sequence of PEA3 have shown strong sequence similarity with the carboxy-terminus of the DNA-binding domain of the ETS-protooncogene product. This domain within the ETS protein is believed to mediate its binding to a specific sequence within
DNA. Furthermore, evidence exists that the ETS-protooncogene product activates transcription of polyoma virus transcription and T-cell-receptor-a-gene expression through the PEA3 motif [4, 61. Other workers [4] have reported that the ETS protein cooperates with C-fos and C-jun to activate transcription by the polyoma-virus enhancer. The nucleotide-binding sequence of the C-jun-C-fos complex, referred to as the AP-1 site is represented by the sequence TGAGTCA. An examination of the DNA sequence of the rpL,, promoter failed to reveal the presence of an AP-1 site. In the current study, PEA3 protein interaction with the rpL,,-gene promoter presumably results from another mechanism. Our results reveal that addition of PEA3 alone is sufficient to enhance the level of transcription by the rpL3,-gene promoter. Whether this enhancement is achieved by interaction of PEA3 with endogenous C-jun and C-fos in the nuclear extract without their binding to the DNA cannot be ruled out. The isolated clone used to prepare the PEA3 protein in the current studies was a partial cDNA encoding approximately 2/3 of the full-length protein but containing the complete DNA-binding domain (ETS domain) and an additional 244 amino acids towards the amino-terminus of the protein [7]. Since this polypeptide is capable of inducing transcription of the rpL,, gene effectively, we presume that this fragment of 329 amino acids contains an activation domain in addition to the DNA-binding domain, as it is sufficient for the transactivation function. Additional studies by Hassell’s laboratory also show that this clone contains an activation domain (Cowie, A. and Hassell, J. A., unpublished results). Production of ribosomal components is coordinately regulated during growth and development, although the mechanism(s) involved in coordination is not completely understood. A report by Hariharan et al. (1989) [17] revealed that the rpLJo and rpLJ2genes contain identical fi elements. In the current study, we have demonstrated that the element of the rpL,,-gene promoter utilizes PEA3 (ETS) DNA-binding protein for transactivation. This binding protein is a possible candidate for maintaining coordinate regulation of ribosomal protein genes at the transcriptional level. Over the past few years, a variety of genes have been described whose protein products have predicted amino-acid sequences showing strong similarity to that of the mouse ETS oncogene [5, 18-21]. Many of these protein have been identified as transcription factors and their target in cellular and viral genes have been identified [4-6, 221. Among these, mouse ETS-2 appears to be broadly distributed in most cells examined and associated with cell proliferation [23]. However, ETS-1 is most abundant in T lymphocytes and not as broadly distributed as ETS-2 [24]. It is tempting to speculate, therefore, that ETS-2 might be the real target for rpL3,-gene-transcription activation. It will be important to determine whether ETS-2 participates in the regulation of rpL,,. In a gel-mobility-shift analysis, the p element interacts with two polypeptides in the nuclear extract. Currently the nature of these two polypeptides is unknown. There are reports in the literature indicating that multiple proteins can bind to the same promoter element [25]. Gel-mobility-shift experiments with a DEAE-chromatography fraction of the nuclear extract and anti-PEA3 serum revealed that, in the presence of these antibodies, the levels of the more-rapidly migrating complex was reduced with a concomitant appearance of a complex with a much slower mobility (supershift). This supershift results from the interaction of the anti-PEA3 serum and a protein (p factor) bound to the /3 element. The anti-PEA3 serum
appeared to interact specifically with a component(s) of the more-rapidly migrating band. Why the upper band was unaffected by the antibodies is unclear. We can only speculate that the upper band is structurally distinct from the lower band or that the appropriate epitope is unexposed. Further studies are needed to resolve these possibilities. In conclusion, our results suggest that the PEA3 (ETS) DNA-binding protein transactivates the rpL,, gene through the fl element. The mechanism of activation is similar to other DNA-binding transcription factors. This research was supported by the Medical Research Council (Canada), the National Science and Engineering Research Council (Canada) and National Cancer lnstitute of Canada.
REFERENCES 1. Wiedemann, L. M., D’Eustachio, P., Kelly, D. E. & Perry, R. P. (1987) Somatic Cell Mol. Genet. 13, 77-80. 2. Ptashne, M. (1988) Science 335, 683-689. 3. Mitchell, P. J. & Tjian, R. (1989) Science 245, 371 -378. 4. Wasylyk, B., Wasylyk, C., Flores, P., Begue, A,, LePrince, D. & Stethelin, D. (1990) Nuture 346, 191 -193. 5. Gunther, C. V., Nye, J. A,, Bryner, R. S. & Graves, B. J. (1990) Genes & Dev. 4, 667 -679. 6. Ho, I. C., Bhat, N . K., Gottschalk, L. R., Lindsten, T.; Thompson, C. B.. Papas, T. S. & Leiden, J. M. (1990) Science 250, 814-818. 7. Xin, J.-H., Cowie, A., Lachance, P. & Hassell, J. A. (1992) Genes d; D ~ v6,481 . -496. 8. Dudov, K. & Perry, R. P. (1984) Cell37, 457-468. 9. Atchison, M. L., Meyuhas, 0. & Perry, R. P. (1989) Mol. Cell. Biol. 9. 2067 - 2074.
10. Yoganathan, T. & Sells, B. H. (1991) FEBS Lett. 286, 163- 166. 11. Martin, M. E., Piette, J., Yaniv, M., Tang, W.-J. & Folk, W. R.
(1988) Proc. Natl Acud. Sci. USA 85, 5839-5843. 12. Dignam, J. D., Lebowitz, R. & Roeder, R. G. (1983) Nucleic Acids Res. 11, 1475-1489. 13. Zahradka, P., Larson, D. E. & Sells, B. H. (1989) Exp. Cell. Res. 185, 8 - 20. 14. Zahradka, P. & Sells, B. H. (1988) Eur. J . Biochem. 171, 37-43. 15. Wall, L., deBoer, E. & Grosveld, F. (1988) Genes & Dev. 2,10891100. 16. Hai, T., Horikoshi, M., Roeder, R. G. & Green, M. R. (1988) Cell 54, 1043- 1051. 17. Hariharan, N., Kelly, D. E. & Perry, R. P. (1989) Genes & Dev. 3,1789-1800. 18. Watson, D. K., McWilliams, M. J., Lapis, P., Lautenberger, J. A,, Schweinfest, C. W. & Papas, T. S. (1988) Proc. Nutl Acud. Sci. U S A 85, 7862 - 7866. 19. Boulukos, K. E., Pognonec, P., Begue, A., Galibert, F., Gesquiere, J. C., Stehelin, D. & Ghysdael, J. (1988) EMBO J . 7,697-705. 20 Pribyl, L. J., Watson, D. K., McWilliams, M. J., Ascione, R. & Papas, T. S. (1988) Dev. B i d . 127,45-53. 21 Bosselut, R., Duvall, J. F., Gegonne, A., Brady, J . & Ghysdeal, J. (1990) EMBO J . 9, 3137-3144. 22 Wasylyk, C., Gutman, A., Nicholson, R. & Wasylyk, B. (1991) E M 3 0 J . I0,1127-1134. 23 Bhat, N. K., Fisher, R. J., Fujiwara, S., Ascione, R. & Papas, T. S. (1987) Prnc. Nutl Acad. Sci. U S A 84, 3161 -3165. 24. Bhat, N. K., Komschlies, K. L., Fujiwara, S . , Fisher, R. J., Mathieson, B. J., Gregorio, H. A,, Young, J. W., Kasik, K., Ozato, K. & Papas, T. S. (1989) J . Zmmunol. 142,62-68. 25. Hai, T., Liu, F., Coukos, W. J. & Green, M. R. (1989) Genes & Dev. 3,2083 - 2090.
Enhanced cell-free transcription of the ribosomal protein L3 gene by the polyoma virus enhancer PEA3 DNA-binding protein Thillainathan YOGANATHAN ’, Alison COWIE’, John A. HASSELL’ and Bruce H. SELLS‘ Molecular Biology and Genetics, College of Biological Science, University of Guelph, Ontario, Canada The Institute for Molecular Biology and Biotechnology, McMaster University, Hamilton, Ontario, Canada (Received January 1 , 1992) - EJB 92 0074
The mouse-ribos~mal-protein-L~~-gene promoter contains a 12-bp sequence motif within the 5’upstream region termed the p element which shows significant similarity with the consensus sequence of the polyoma-virus-enhancer PEA3. A cloned PEA3 DNA-binding protein, expressed in Escherichiu coli and purified, activates the expression of the ribosomal-pr0tein-1~~ gene in a cell-free system. Moreover, the PEA3 protein participates in the formation of the ribosomal-protein-L32-promoterpreinitiation-transcription complex. The preinitiation complex formed with PEA3 is resistant to competition by oligonucleotides containing the p element. In addition anti-PEA3 serum interacts with a factor in mouse L1210 nuclear extract that binds to the /3 element, causing a supershift in a mobility-shift assay. Our study demonstrates for the first time that the PEA3 protein can transactivate a cellular gene in a cell-free transcription system.
The ribosome particle is composed of more than 70 different ribosomal proteins whose formation is coordinately regulated such that each protein is produced in equimolar amounts. How this coordination is achieved is not yet clear since the genes for these proteins are unclustered in the eucaryotic genome and are located on different chromosomes [l]. One approach to solving this question is to examine the promoter regions of the ribosomal-protein genes to identify the factors which interact with specific sequences and to determine whether any of these factors play a role in coordinating transcription. Transcription of eucaryotic genes is controlled by the combined action of several sequence-specific DNA-binding proteins [2, 31. Basic transcription factors and RNA polymerase bind in the vicinity of the transcription-initiation site to ensure basal levels of transcription, while elevated levels are induced by additional factors which bind to regulatory elements located some distance upstream of the start site. These latter protein factors influence the rate of initiation by interacting with the transcriptional apparatus. Several of the proteins associated with the pol-11-promoter enhancer and transactivator regions have been identified [2, 31. These transactivators are generally bipartite in structure, containing a DNA-binding domain and a transcription-activation domain [3]. The DNA-binding domains of these regulatory proteins form a variety of structures which make contact with DNA, i.e. zinc finger or helix-turn-helix motifs [2, 31. More recently a new class of DNA-binding proteins which contains none of the known DNA-binding structures has been identified. These proteins are represented by the E-twenty-six-avian-leukaemia-
virus-specific (ETS) family of proto-oncogene products which includes the PEA3 DNA-binding protein [4 - 71. In the present study we have focused our attention on the gene encoding the ribosomal-protein L32 (rpL3J. This gene has a promoter region with several distinctive features. It lacks both a TATA box and an Spl-binding site [8]. Additional studies have revealed that the upstream region of the rpL32 gene (nucleotides -80 to -69) contributes to the efficiency of its transcriptional activity [9, 101. In a recent report we have identified a 50 - 55-kDa protein, the B factor, which binds to the sequence in this (nucleotides - 80 to - 69) region and participates in the cell-free transcription of this gene [lo]. We have postulated that the p element in the r ~ L ~ ~ - g promoter ene acts as an upstream transactivating element. The p element has significant sequence similarity to the PEA3 motif of the polyoma virus enhancer 14, 81. A protein-binding activity has been identified in mouse cell nuclear extracts that recognises the PEA3 sequence [l 11. Furthermore, a cDNA that encodes PEA3-binding activity has been cloned by screening an expression library with a probe containing the PEA3 motif [7]. Based on its DNA-binding specificity, we postulated that the PEA3 protein activates m o ~ s e - r p L ~ ~ - g e n e expression through its interaction with the j? element. We have, therefore, investigated the function of the PEA3 DNAbinding protein in a mouse cell-free transcription system in an attempt to determine whether PEA3 and the p factor are interchangeable in stimulating transcription of the rpL32gene. In this study, we report that PEA3 DNA-binding protein transactivates the rpL,, gene through the p element of this promoter.
Correspondence to B. H. Sells, College of Biological Science, Molecular Biology and Genetics, University of Guelph, N1G 2W1 Guelph, Ontario, Canada Ahhreviations. rp, ribosomal protein; ETS, E-twenty-six-avianleukaemia-viros-specific; GST, glutathione-S-transferase. Enzyme. Restriction endonucleases (EC 3.1.21.4).
EXPERIMENTAL PROCEDURES Cell culture and nuclear-extract preparation Mouse L1210 cells were maintained in a rotating culture in Dulbecco’s modified Eagle’s medium supplemented with
196 5% fetal calf serum. Nuclear extracts were obtained using the procedure of Dignam et al. [12]. Nuclear extracts were fractionated on DEAE-cellulose (Whatman DE52) according to the protocol of Zahradka et al. [13].
A Poiyomavirus enhancer
5’-CAGGAAGTGACT-3’
rpL32 fl element 5’-CCGGAAGTGCTT-3’
Plasmids
m 622
3 09
Individual reaction mixtures contained 12.5 mM Hepes, pH 7.9, 5 mM MgC12, 70 mM KC1, 10% glycerol, 500 pM (each) ATP, CTP and GTP, 25 pM [32P]UTP,0.4 mM EDTA, 0.5 mM dithiothreitol, 100 U/ml RNA guard (Pharmacia), 0.25 pg plasmids DNA and 1 - 3 p1 DEAE chromatography fraction of L1210 nuclear extract, in 25 pl total volume. Samples were incubated for 60 min at 30°C and the reactions were arrested by the addition of 175 pl stop buffer [14]. The transcripts were deproteinized, precipitated with ethanol and separated by 4% urea/PAGE and radioactivity detected by autoradiography.
Synthetic oligonucleotides used in these studies were produced using a Biosearch 8600 DNA synthesizer. Each oligonucleotide was HPLC purified and the purity checked by running on a preparative 20%-polyacrylamide/urea sequencing gel. Double-stranded oligonucleotide was made by annealing the chemically synthesised oligonucleotides essentially as described in [15]. Gel-mobility-shift assay The radiolabelled fragments used for the gel-mobility-shift assays were prepared essentially as described by Wall et a1.(1988) [I 51. A specific double-stranded oligonucleotide that contained the sequence in the p-element recognition site was prepared by annealing the following chemically synthesized oligonucleotides, 32P-labelled using a kinase: 5’-CCCAGAGCCGGAAGTG-3’ and 5’-TGGGCACTTCCGGCTC-3’. Gel-mobility-shift assays were performed under transcription conditions, except that the NTP were omitted. 1.5 pg poly [d(I-C)] and 0.1 ng radiolabelled oligonucleotide fragments were incubated with nuclear extract at 30°C for 30 min and protein-DNA complexes resolved by electrophoresis on a 4% polyacrylamide gel in a Tris/glycine running buffer. Cloning and purification of PEA3 protein A partial cDNA encoding the carboxy-terminal(38 kDa) of PEA3 was expressed in Escherichia coli as a fusion protein with glutathione-S-transferase (GST; Pharmacia). The protein was purified by affinity Chromatography on glutathioneSepharose as described by the manufacturer (Pharmacia). The purified GST-PEA3 fusion protein migrated as a single band at the expected molecular mass on SDSjPAGE [7].
at-
urn
- - P O
Cell-free transcription
Synthetic oligonucleotides
0
B
Plasmid P3AR2.8, containing a fragment of the murine rpL32gene was obtained from R. P. Perry (Institute of Cancer Research, Philadelphia, PA); plasmid pAd97, containing the adenovirus major late promoter was obtained from N. Miyamoto (Ontario Cancer Institute, Toronto) and the histone-H4 plasmid from G. Stein (University of Massachuset t s) .
147
1
2
3
4
Fig. 1. The nucleotide sequences of the PEA3 enhancer and rpL,, element and enhancement of transcription by PEA3 protein in vitro. (A) Similarities between the nucleotide sequence of the PEA3 enhancer and rpL32 element. (B) Transcription enhancement by PEA3 protein in vitro. Template plasmids were digested with EcoRl. Molecularmass standards (lane 1); control (lane 2); 100 ng purified PEA3 protein (lane 3); 100 ng purified GST protein (lane 4).
Preparation of anti-PEA3 serum The antiserum used in this study was generated in rabbits using a bacterially expressed PEA3 protein. RESULTS Transcription activation by purified PEA3 protein To determine the effect of PEA3 protein on rpL3, transcription, we have conducted cell-free transcription of When the rpL32 gene was transcribed in L1210 extracts, the templates generated a 1500-bp nucleotide-run-off transcript corresponding to specific initiation at the in vivo cap site (Fig. lB, lane 2). Addition of PEA3 protein stimulated this specific transcription (Fig. lB, lane 3). Addition of GST had no effect on the transcription (Fig. lB, lane 4). These results indicate that PEA3 protein activated transcription of To further evaluate whether this transcriptional activation occur with other pol-I1 genes, we have performed transcription experiments with histone and adenovirus major-late-promoter genes. Purified-PEA3-fusion-protein transactivation specific to the rpL,, gene To determine the effect of PEA3 upon transcription of other pol-I1 genes, the following experiments were performed.
197 A rpL32
- ,PEA3 -,,
GST
< 1
2
B Histone H4
3
m
-
a
n
4
5
I-
1
m
U
< 1
2
3
C
< 1
2
3
Fig. 2. Effects of PEA3 protein on the transcriptional activities of various promoters. Nucleotide-run-off assays were performed using different genes. (A) EcoRI-digested rpLjz gene; lane 1, containing no PEA3-GST; lanes 2 and 3, containing 100 ng and 200 ng of PEA3GST fusion protein respectively; lanes 4 and 5, containing 200 ng and 200 ng of GST respectively. (B) XboI-digested histone-H4 gene; lane 1, no PEA3; lane 2, with 200 ng PEA3-GST fusion protein; lane 3, with 200 ng GST. (C) Ad MLP: AatII-digested pAd97 containing the adenovirus 2 major late promoter; lane 1, no PEA3; lane 2, with 200 ng PEA3-GST protein; lane 3, with 200 ng GST protein. Specific transcripts are indicated throughout by arrows.
In the first the PEA3-GST fusion protein, synthesized in E. roli and purified, was incubated with a DNA template containing the m o ~ s e - r p L ~ ~ - g epromoter ne and a partially purified mouse L1210 nuclear extract (see Experimental Procedures). Fig. 2A shows that addition of the purified PEA3 fusion protein (100 ng or 200 ng) to the reaction mixtures increased transcription of rpL32gene (Fig. 2A, lanes 2 and 3). Addition of the same amount of GST protein alone produced no stimulation (Fig. 2A, lanes 4 and 5). Furthermore, inclusion of the PEA3-GST fusion protein (200 ng) in the transcription-reaction mixture containing the histone-H4 promoter resulted in no significant change in transcription (Fig.2 B, lanes 1-3). Similarly, addition of the PEA3-GST fusion protein (200 ng) produced no changes in adenovirus major late promoter transcription (Fig. 2C, lanes 1- 3). These data imply that PEA3 functions as a specific transcriptional activator of the mouse rpL32gene. In a second series of experiments to assess the involvement of PEA3, nuclear extract was incubated with the oligonucleotide containing the element to deplete the reaction mixture of the p factor. Fig. 3 (lane 2 ) reveals that following incubation the resulting extract is incapable of supporting transcription of the rpL32 gene. This reduction in transcription is specific to the p-element sequence since an unrelated oligonucleotide of the same size failed to depress transcription of the rpL32 gene (Fig. 3, lane 6). To determine whether the factor(s) removed by incubation with the p-element sequence could be replaced by PEA3, depleted extracts were incubated with the PEA3-GST fusion protein. The results indicate that addition of the fusion protein restored the ability of the depleted extract to transcribe the r ~ L , ~ - g e n(Fig. e 3, lanes 4 and
2
3
4
5
6
Fig. 3. Transcriptional activation of the rpL32 gene by PEA3 is dependent upon binding to a specific upstream sequence. Oligonucleotide competition of the transcription reactions supplemented with PEA3 protein is shown. Transcription assays and reaction conditions were as described in Experimental Procedures. Specific and control oligonucleotides were incubated with 4 kl DEAE-chromatography fraction for 15 min prior to addition of template. In addition, transcription reaction mixtures were supplemented with various amounts of PEA3-GST protein or GST. Lane 1 , no oligonucleotide; lane 2. 200 ng specific oligonucleotide ;lane 3,200 ng specific oligonucleotide and 200 ng GST; lane 4,200 ng specific oligonucleotide and 100 ng PEA3-GST fusion protein; lane 5,200 ng specific oligonucleotide and 200 ng PEA3-GST protein; lane 6, 200 ng non-specific oligonucleotide alone. Oligo., oligonucleotide; un., non-specific.
5) whereas addition of GST alone failed to restore transcription( Fig. 3, lane 3). These results suggest that the PEA3 protein is involved in DNA binding and stimulation of rpL32gene transcription. PEA3 protein forms a transcription-initiation complex with the rpL3*-genepromoter
These studies have been designed to determine the stage of involvement of the PEA3 protein in transcription by employing a cell-free system. The following experiments were performed to assess whether this protein functions in the formation of the preinitiation complex and whether it remains committed to the template following complex formation. Preinitiation complexes were prepared as described by Hai et al. (1988) [16], and the reaction was arrested at this stage by incubation in the absence of NTP. Transcription was then initiated by the addition of NTP and the transcripts generated in 60 min were analyzed. Fig. 4 (lanes 1 and 2) illustrates the level of transcription with or without PEA3 protein. Addition of 200 ng jelement to the reaction mixture before assembly of the preinitiation complex substantially reduced transcriptional stimulation by the PEA3-GST fusion protein (Fig. 4, lane 3). In contrast, addition of the p element following formation of the preinitiation complex resulted in no significant effect on the activation induced by PEA3 (Fig. 4, lane 4). These observations suggest that PEA3 is required during assembly of the preinitiation complex and that this complex is a tightly bound one and resistant to competition by exogenously added jelement. Moreover, addition of the same amount of a control oligonucleotide, either before or after preinitiation, had no effect on enhancement by PEA3 (Fig. 4, lanes 5 and 6). These results infer that the PEA3 protein functions as a transcriptional activator in the rpL32-cell-free-transcriptionsystem and participates in the formation of the preinitiation complex. PEA3 protein binds to the fi element
The PEA3 binding site within the polyoma virus enhancer and the rpL32 p element have considerable nucleotide se-
198 0
0
0
0 9
.9 .9
2 0
6 6
2 5
c
Q
C
&,"s i-\
PEA3
1
2
3
4
5
- + + ++
6
Fig. 4. PEA3 participates in the stabilization of preinitiation complexes. Transcription assay and the reaction conditions are as described in Experimental Procedures. Lane 1, control transcription-reaction mixture without PEA3; lane 2, transcription-reaction mixture with 150 ng PEA3; lane 3, transcription-reaction mixture with 150 ng PEA3 and 200 ng p-element oligonucleotide (oligonucleotide added prior to the formation of the preinitiation complex); lane 4, transcription-reaction mixture with 150 ng PEA3 and 200 ng /I-element oligonucleotide (oligonucleotide added after formation of the preinitiation complex); lane 5, transcription-reaction mixture with 150 ng PEA3 and 200 ng control oligonucleotide (added prior to formation of the preinitiation complex); lane 6, transcription-reaction mixture with 150 ng PEA3 and 200 ng non-specific oligonucleotide (added after formation of the preinitiation complex).
quence similarity (Fig. 1A). This sequence similarity suggests that the PEA3-GST fusion protein should bind to the rpL3, fl element. To test this hypothesis, cloned and purified PEA3 protein was included in a gel-mobility-shift experiment. PEA3 protein was able to bind to the radiolabelled fl element to produce a radiolabelled complex migrating more slowly than the radiolabelled oligonucleotide (Fig. 5). Formation of the complex was abolished by the addition of excess fl element, whereas the same amount of a control oligonucleotide had no effect on complex formation (Fig. 5, lanes 3 - 5). Binding of PEA3 to the rpL,,-gene-promoter element suggests that the PEA3 polypeptide contains or shares DNA-binding specificity with the factor. Specificity of the nuclear protein binding to the 1 element
To assess the nature of the fi factor present in the transcriptionally active nuclear extract we have carried out gel-mobility-shift assays. A gel-mobility-shift-assay using a fl-element probe, produced two specific complexes (Fig. 6A). The formation of these complexes, designated CI and CII, was inhibited by the addition of specific oligonucleotide (Fig. 6A, lanes 2 and 3) but not by the non-specific oligonucleotide (Fig. 6A, lanes 4 and 5). These results suggest that these nuclear proteins are specific to the fl element. Antibodies directed against PEA3 recognize the 1 factor
To further understand the relationship between the fl factor and PEA3 protein, the influence of antisera raised against PEA3 was examined in a gel-mobility-shift assay. Nuclear extract was incubated with anti-PEA3 serum and tested for its ability to form a complex with the fl element. From Fig. 6B it is evident that the amount of the more-rapidly migrating complex (CI) was reduced following incubation of the nuclear extract with antibodies to PEA3. The more slowly
1 2 3 4 5
Fig. 5. Gel-mobility-shift assays showing that PEA3 protein binds to the radiolabelled j element. Lane 1, no PEA3 protein; lane 2, 50 ng PEA3 protein; lanes 3 and 4,50 ng PEA3 protein plus a 50-fold molar excess and a 25-fold molar excess of non-radiolabelled p oligonucleotide, respectively; lane 5, 50 ng PEA3 protein and a 50-fold molar excess of non-specific oligonucleotide.
migrating complex (CII) was unaffected. The preimmune serum failed to modify the migration of either of these bands (data not shown). In addition to reducing the formation of the CI complex, the presence of the anti-PEA3 serum gave rise to a higher-molecular-mass complex containing the fi element resulting in the production of a supershift (CIII). As the antibody concentration was increased, the higher-molecular-mass complex became more intense (Fig. 6, lanes 2-4). These data suggest that the more-rapidly migrating band (CI) is composed of a protein in the nuclear extract structurally related to PEA3. DISCUSSION The mouse-rpL3,-gene promoter contains a 12-bp sequence motif within the upstream region termed the fl element. This fl element displays significant similarity with the consensus sequence of the polyoma-virus enhancer PEA3 (Fig. 1A) Our current studies suggest that the protein interacting with the PEA3-binding site also binds to the B element and induces transcription of the rpL3, gene. Induction appears to be specific to the rpL3, gene since addition of PEA3 protein to a cell-free transcription system containing either the adenovirus major-late-promoter gene or the histone-H4 gene fails to activate transcription of these genes. These observation are not surprising since a DNA sequence resembling the PEA3-binding site is absent in these latter two gene promoters. Several reports have indicated that ETS protein represents a new class of DNA-binding protein which activates transcription in vivo [4 - 61. No published information currently exists demonstrating transcriptional activation by this DNA-binding protein using a cell-free transcription system.
199
A
CII
c
1
2
3
4
5
A n t i PEA3
A Clll
+
CI I
4
CI
4
+ 1
2
3
Super Shift
Free Probe
4
Fig. 6. Gel-mobility-shit assays using oligonucleotide-competitionand antibody binding. (A) Gel shift assay with oligonucleotide-competition analysis. Nuclear extract was incubated with the p probe in the absence of competitor (lane 1); in the presence of a 50-fold molar excess and a 25-fold molar excess of double-stranded oligonucleotide sequence containing the element, respectively (lanes 2 and 3); in the presence of 50-fold molar excess and 25-fold molar excess of non-specific double-stranded oligonucleotide, respectively (lanes 4 and 5). CI and CII indicate the positions of the complexes observed. Spec. oligo., specific oligonucleotide. (B) Formation of a supershift complex with the p factor using anti-PEA3 serum. DNA-binding mobility-shift mixtures containing L1210 nuclear extract were incubated with various amount of anti-PEA3 serum. Lane 1 , no antiserum; lanes 2-4, 0.25, 0.50 and 1.0 pl anti-PEA3 serum, respectively.
Recent studies [7] on the amino-acid sequence of PEA3 have shown strong sequence similarity with the carboxy-terminus of the DNA-binding domain of the ETS-protooncogene product. This domain within the ETS protein is believed to mediate its binding to a specific sequence within
DNA. Furthermore, evidence exists that the ETS-protooncogene product activates transcription of polyoma virus transcription and T-cell-receptor-a-gene expression through the PEA3 motif [4, 61. Other workers [4] have reported that the ETS protein cooperates with C-fos and C-jun to activate transcription by the polyoma-virus enhancer. The nucleotide-binding sequence of the C-jun-C-fos complex, referred to as the AP-1 site is represented by the sequence TGAGTCA. An examination of the DNA sequence of the rpL,, promoter failed to reveal the presence of an AP-1 site. In the current study, PEA3 protein interaction with the rpL,,-gene promoter presumably results from another mechanism. Our results reveal that addition of PEA3 alone is sufficient to enhance the level of transcription by the rpL3,-gene promoter. Whether this enhancement is achieved by interaction of PEA3 with endogenous C-jun and C-fos in the nuclear extract without their binding to the DNA cannot be ruled out. The isolated clone used to prepare the PEA3 protein in the current studies was a partial cDNA encoding approximately 2/3 of the full-length protein but containing the complete DNA-binding domain (ETS domain) and an additional 244 amino acids towards the amino-terminus of the protein [7]. Since this polypeptide is capable of inducing transcription of the rpL,, gene effectively, we presume that this fragment of 329 amino acids contains an activation domain in addition to the DNA-binding domain, as it is sufficient for the transactivation function. Additional studies by Hassell’s laboratory also show that this clone contains an activation domain (Cowie, A. and Hassell, J. A., unpublished results). Production of ribosomal components is coordinately regulated during growth and development, although the mechanism(s) involved in coordination is not completely understood. A report by Hariharan et al. (1989) [17] revealed that the rpLJo and rpLJ2genes contain identical fi elements. In the current study, we have demonstrated that the element of the rpL,,-gene promoter utilizes PEA3 (ETS) DNA-binding protein for transactivation. This binding protein is a possible candidate for maintaining coordinate regulation of ribosomal protein genes at the transcriptional level. Over the past few years, a variety of genes have been described whose protein products have predicted amino-acid sequences showing strong similarity to that of the mouse ETS oncogene [5, 18-21]. Many of these protein have been identified as transcription factors and their target in cellular and viral genes have been identified [4-6, 221. Among these, mouse ETS-2 appears to be broadly distributed in most cells examined and associated with cell proliferation [23]. However, ETS-1 is most abundant in T lymphocytes and not as broadly distributed as ETS-2 [24]. It is tempting to speculate, therefore, that ETS-2 might be the real target for rpL3,-gene-transcription activation. It will be important to determine whether ETS-2 participates in the regulation of rpL,,. In a gel-mobility-shift analysis, the p element interacts with two polypeptides in the nuclear extract. Currently the nature of these two polypeptides is unknown. There are reports in the literature indicating that multiple proteins can bind to the same promoter element [25]. Gel-mobility-shift experiments with a DEAE-chromatography fraction of the nuclear extract and anti-PEA3 serum revealed that, in the presence of these antibodies, the levels of the more-rapidly migrating complex was reduced with a concomitant appearance of a complex with a much slower mobility (supershift). This supershift results from the interaction of the anti-PEA3 serum and a protein (p factor) bound to the /3 element. The anti-PEA3 serum
appeared to interact specifically with a component(s) of the more-rapidly migrating band. Why the upper band was unaffected by the antibodies is unclear. We can only speculate that the upper band is structurally distinct from the lower band or that the appropriate epitope is unexposed. Further studies are needed to resolve these possibilities. In conclusion, our results suggest that the PEA3 (ETS) DNA-binding protein transactivates the rpL,, gene through the fl element. The mechanism of activation is similar to other DNA-binding transcription factors. This research was supported by the Medical Research Council (Canada), the National Science and Engineering Research Council (Canada) and National Cancer lnstitute of Canada.
REFERENCES 1. Wiedemann, L. M., D’Eustachio, P., Kelly, D. E. & Perry, R. P. (1987) Somatic Cell Mol. Genet. 13, 77-80. 2. Ptashne, M. (1988) Science 335, 683-689. 3. Mitchell, P. J. & Tjian, R. (1989) Science 245, 371 -378. 4. Wasylyk, B., Wasylyk, C., Flores, P., Begue, A,, LePrince, D. & Stethelin, D. (1990) Nuture 346, 191 -193. 5. Gunther, C. V., Nye, J. A,, Bryner, R. S. & Graves, B. J. (1990) Genes & Dev. 4, 667 -679. 6. Ho, I. C., Bhat, N . K., Gottschalk, L. R., Lindsten, T.; Thompson, C. B.. Papas, T. S. & Leiden, J. M. (1990) Science 250, 814-818. 7. Xin, J.-H., Cowie, A., Lachance, P. & Hassell, J. A. (1992) Genes d; D ~ v6,481 . -496. 8. Dudov, K. & Perry, R. P. (1984) Cell37, 457-468. 9. Atchison, M. L., Meyuhas, 0. & Perry, R. P. (1989) Mol. Cell. Biol. 9. 2067 - 2074.
10. Yoganathan, T. & Sells, B. H. (1991) FEBS Lett. 286, 163- 166. 11. Martin, M. E., Piette, J., Yaniv, M., Tang, W.-J. & Folk, W. R.
(1988) Proc. Natl Acud. Sci. USA 85, 5839-5843. 12. Dignam, J. D., Lebowitz, R. & Roeder, R. G. (1983) Nucleic Acids Res. 11, 1475-1489. 13. Zahradka, P., Larson, D. E. & Sells, B. H. (1989) Exp. Cell. Res. 185, 8 - 20. 14. Zahradka, P. & Sells, B. H. (1988) Eur. J . Biochem. 171, 37-43. 15. Wall, L., deBoer, E. & Grosveld, F. (1988) Genes & Dev. 2,10891100. 16. Hai, T., Horikoshi, M., Roeder, R. G. & Green, M. R. (1988) Cell 54, 1043- 1051. 17. Hariharan, N., Kelly, D. E. & Perry, R. P. (1989) Genes & Dev. 3,1789-1800. 18. Watson, D. K., McWilliams, M. J., Lapis, P., Lautenberger, J. A,, Schweinfest, C. W. & Papas, T. S. (1988) Proc. Nutl Acud. Sci. U S A 85, 7862 - 7866. 19. Boulukos, K. E., Pognonec, P., Begue, A., Galibert, F., Gesquiere, J. C., Stehelin, D. & Ghysdael, J. (1988) EMBO J . 7,697-705. 20 Pribyl, L. J., Watson, D. K., McWilliams, M. J., Ascione, R. & Papas, T. S. (1988) Dev. B i d . 127,45-53. 21 Bosselut, R., Duvall, J. F., Gegonne, A., Brady, J . & Ghysdeal, J. (1990) EMBO J . 9, 3137-3144. 22 Wasylyk, C., Gutman, A., Nicholson, R. & Wasylyk, B. (1991) E M 3 0 J . I0,1127-1134. 23 Bhat, N. K., Fisher, R. J., Fujiwara, S., Ascione, R. & Papas, T. S. (1987) Prnc. Nutl Acad. Sci. U S A 84, 3161 -3165. 24. Bhat, N. K., Komschlies, K. L., Fujiwara, S . , Fisher, R. J., Mathieson, B. J., Gregorio, H. A,, Young, J. W., Kasik, K., Ozato, K. & Papas, T. S. (1989) J . Zmmunol. 142,62-68. 25. Hai, T., Liu, F., Coukos, W. J. & Green, M. R. (1989) Genes & Dev. 3,2083 - 2090.
