INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Transcriptional activation in eukaryotes is achieved by regulatory proteins recruiting the transcriptional machinery and chromatin remodeling factors to the promoter. Transcriptional activators have been shown to present a modular organization consisting of distinct domains for sequence-specific DNA binding and transcriptional activation through interaction with other proteins (for a review, see reference 34). DNA binding domains of many transactivators fall into readily identifiable classes. Instead, domains responsible for transcriptional activity may not possess such a well-defined organization. While no obvious sequence similarity has been detected among the activation domains of diverse activator proteins, they nevertheless often carry a distinctive amino acid composition. For example, activation domains particularly rich either in prolines, glutamines, acidic amino acids, or hydroxylated amino acids have been described (for reviews, see references 4, 21, 27, and 51). Despite the identification of targets for numerous activation domains and elucidation of the functional importance of particular residues, the structural basis for the capacity of an activation domain to stimulate transcription remains poorly understood. However, recent structural studies suggest that the -helix may be a common structural motif in the binding of transactivation factors to transcriptional activation factors (22, 52).

Transcriptional activators were also described to stimulate transcription of RNA polymerases II (Pol II) and III (Pol III) small nuclear RNA (snRNA) genes. In vertebrates, these genes contain an essential proximal sequence element (PSE) located at approximately the position 59 upstream of the transcription start site. The Pol III-dependent promoters possess additionally a TATA box located at position 30, which acts as a major determinant for RNA Pol III specificity (24, 26; for a review, see reference 15). A number of other short transcription units, such as the 7SK, Y, MRP, selenocysteine tRNA (tRNA^Sec), and H1 RNA genes, have similar basal promoter elements and can be classified as snRNA-type genes (3, 30; for a review, see reference 14). These will be here referred to as snRNA-type promoters. SNAPc, also called PTF, is a stable complex containing four protein subunits. It binds specifically to the PSE (13, 31, 37, 54, 55). Other studies have also implicated the TATA binding protein (TBP), TFIIIC1, and TFIIIB in transcription of Pol III snRNA genes (25, 41, 48, 55).

Associated with the basal promoter, the snRNA and snRNA-type genes contain a distal sequence element (DSE) around position 220, which plays a major role in transcription efficiency. The DSE contains an octamer motif which binds the transcriptional activator Oct-1 (42; for reviews, see references 14 and 16). In addition to Oct-1, Sp1 has been shown in some instances to be involved in mediating the activation properties of the DSE (1, 20, 23, 47). Recently, we have demonstrated that the zinc finger protein Staf, originally identified as the transcriptional activator of the tRNA^Sec gene (39), is also involved in transcriptional activation of snRNA and snRNA-type genes transcribed by RNA polymerases II and III (38). In the course of previous investigations, we observed that Staf also possesses the ability to stimulate CAT expression from a Pol II promoter (32, 39), therefore showing that it can activate a whole diversity of Pol II and Pol III promoters. At first sight, this result might be surprising in view of the fact that, although transcription of mRNA and Pol II snRNA genes is catalyzed by the same enzyme, the architectures of mRNA and snRNA transcription complexes have some dissimilarities (for reviews, see references 14 and 15). So far, the mechanism by which Staf can interact with these different Pol II and Pol III transcription complexes has remained unknown. In order to gain insight into it, the mapping and functional dissection of the Staf activation domain was undertaken. We report here that Staf contains in fact two distinct snRNA and mRNA promoter-specific activation domains, explaining why a single transcriptional activator can activate a variety of different promoters. This finding lays the foundation for further elucidation of the detailed, promoter-specific activation pathways.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Effector constructs. Staf-Krox-20 473 to 600 (SK473-600), the Staf-Krox-20 3'-end-truncated derivative, was obtained by double digestion of pBRN3-Staf-Krox-20 (39) with EcoRV (cleavage at position 1443 of the Staf cDNA sequence) and NotI (cleavage in the polylinker of pBRN3) and recircularization of the plasmid after filling-in of the NotI 5' extremity. Staf-Krox-20 6-248 (SK6-248), the Staf-Krox-20 5'-end-truncated derivative, was obtained by digestion of pBRN3-Staf-Krox-20 with NcoI (cleaved at positions 138 and 771 of the Staf cDNA sequence) and recircularization of the filled-in ends of the shortened plasmid. Internal deletions 6-77, 6-102, 6-130, 6-206, 6-220, 78-220, 157-220, 6-206+225-242, 84-176, 244-262, and 212-220 in Staf-Krox-20 473-600, truncation 212-220 in Staf-Krox-20, and mutants PDTIS207-211AAAAA, SIEG221-224AAAA, E214A+Q215A+K219A, L213A+Y216A, S211A, L213A, E214A, Q215A, Y216A, and K219A in Staf-Krox-20 473-600 were obtained by site-directed mutagenesis of the parental vector. S84-176K contains the two internal deletions, 6-78 and 177-262, obtained by site-directed mutagenesis of the parental vector.

Reporter constructs. p6E.tkCAT, AE.tkCAT, 3E.tRNA^Sec, and 3E.U1 are described in references 53, 32, 39, and 38, respectively.

Oocyte microinjections, nuclear localization, and DNA binding assays. Capped mRNAs were synthesized in vitro by T3 RNA polymerase as described in Schuster et al. (39) and were injected (20 nl, 1 ng) into the cytoplasm of Xenopus laevis oocytes, 20 h before nuclear injection of 20 nl containing the reporter. The concentrations of the reporters were 300 µg/ml for 6E.tkCAT and 50 µg/ml for 3E.tRNA^Sec and 3E.U1. The 3E.tRNA^Sec and 3E.U1 reporters were injected in the presence of [-³²P]GTP (800 Ci/mmol, 0.2 µCi/oocyte) and 5S RNA maxigene (5 µg/ml) as an internal control for nuclear microinjection and RNA recovery. The 6E.tkCAT reporter was injected in the presence of the pCH110 vector (300 µg/ml) as an internal control for nuclear injection. Incubation was for 16 h (3E.U1 and 6E.tkCAT reporters) or 3 h (3E.tRNA^Sec reporter). Transcription of the reporter genes was analyzed as described previously (32, 39). The transcription efficiencies of the tRNA^Sec and U1 RNA genes, relative to the 5S RNA maxigene expression, were quantitated with a Fuji Bioimage Analyzer BAS2000. For monitoring synthesis and nuclear localization of the effector proteins, oocytes microinjected with mRNAs were manually enucleated. Oocyte nuclei were homogenized in 8 µl of the extraction buffer [50 mM Tris-HCl (pH 8.0), 50 mM KCl, 0.1 mM EDTA, 5 mM MgCl₂, 25% glycerol, 1 µl of bestatin per ml, 1 µl of pepstatin A per ml, 1 µl of trans-epoxysuccinyl-L-leucylamido-(4-guanidino)-butane per ml, 1 µl of 4-(2 aminoethyl)-benzenesulfonyl fluoride per ml] per nucleus. Homogenates were clarified by centrifugation at 13,000 rpm for 5 min. ³²P-labeled double-stranded oligonucleotides containing the E element were used for Staf-Krox-20 and Krox-20-DBD bandshift assays as described previously (6).

Transfections, CAT assays, and Western analysis. Schneider line 2 (SL2) Drosophila cells were transfected by the DOTAP liposomal transfection procedure (Boehringer). The cells received 500 ng of expression plasmid, 500 ng of reporter plasmid, and 200 ng of pACH110 (49). -Galactosidase activity, normalization, and chloramphenicol acetyl-transferase (CAT) assays were performed as described previously (39). Results were quantitated with a Fuji Bioimage Analyzer BAS2000. The data represent the averages of results from three to four transfections, each performed in duplicate. For Western blot analysis, extracts from Drosophila-transfected cells were prepared essentially as described previously (2). Proteins were transferred to nitrocellulose and visualized with a polyclonal serum raised against a C-terminal Staf peptide with the use of ECL reagents (Amersham).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Mapping of the Staf activation domain. The Staf amino acid sequence can be divided into three regions (Fig. 1A): an N-terminal domain (amino acids 1 to 263), which we refer to as domain A; a central DNA binding domain B containing seven zinc fingers of the C₂-H₂ type (amino acids 264 to 468); and the C-terminal domain C (amino acids 469 to 600). Truncated Staf proteins were constructed, in order to define the Staf domains which are needed for the activation of snRNA, snRNA-type, and TATA box-containing mRNA promoters. The transactivation capabilities provided by the wild-type and truncated proteins (Fig. 1A) was assayed on various promoters by microinjection into X. laevis oocytes. To eliminate the background provided by the endogenous Staf in this assay, we designed chimeric Staf-Krox-20 proteins with altered DNA binding specificities. The Staf zinc finger domain was replaced by the corresponding sequences of the Krox-20 zinc finger domain (K-DBD), as illustrated in Fig. 1A (5, 39). This results in chimeric Staf-Krox-20 molecules (Fig. 1A) with different DNA binding specificities but overall structures that are likely to be relatively unchanged from that of the wild-type protein. The abilities of the different Staf-Krox-20 chimeric proteins to activate snRNA and snRNA-type promoters were assayed with Xenopus U1b1 snRNA (Pol II) and tRNA^Sec (Pol III) promoters. They contained, in place of the residing activator element, a multimerized version of the Krox-20 binding site E element (3E.U1 and 3E.tRNA^Sec) (Fig. 1A). The reporter 6E.tkCAT (Fig. 1A), containing six E elements upstream of the tk promoter, was used for monitoring the transcriptional activation on TATA box-containing mRNA promoters (53).

View larger version (18K): [in this window] [in a new window]   View larger version (21K): [in this window] [in a new window]   FIG. 1.   The Staf-Krox-20 chimera requires the N terminus to activate transcription from tk.CAT, U1 snRNA, and tRNASec promoters. (A) Schematic diagrams depicting the effector mRNAs synthesized in vitro and the 6E.tkCAT, 3E.U1 and 3E.tRNASec reporter genes used in the Xenopus microinjection assay. The first lane shows the structure of Staf. The A, B, and C domains are indicated with their localizations. PSE and TATA correspond to the basal promoter elements of the U1 and tRNASec genes. 3E and 6E indicate the number of multimerized E binding sites of the Krox-20 protein. SK, chimeric Staf-Krox-20 molecules. (B) CAT assays showing transcriptional activation from the 6E.tkCAT promoter by the K-DBD, SK, SK473-600 and SK6-248 effectors using extracts from microinjected Xenopus oocytes. Lane 1, no effector was expressed. (C) Transcriptional activation from the Pol II 3E.U1 (lanes 1 to 4) and Pol III 3E.tRNASec (lanes 5 to 8) promoters by the SK, SK473-600 and SK6-248 effectors using microinjected Xenopus oocytes. Lanes 1 and 5, no effector was expressed. Positions of the U1, tRNASec, and 5S maxi used as an internal standard are indicated. (D) Effector expression assayed by gel retardation with nuclear extracts from microinjected Xenopus oocytes. The identity of the expressed effector is indicated above each lane. Lanes 1 and 2, no effector was expressed. Reactions in lanes 1, 3, 5, 7, and 9 () contained no competitor; lanes 2, 4, 6, 8, and 10 (+) contained a 100-fold molar excess of unlabeled E competitor DNA.

The minimal activation domain for snRNA-type promoters is 18 amino acids long. To define the minimal region of Staf sufficient for activation of snRNA and snRNA-type promoters by RNA polymerases II and III, the 263-amino-acid activation domain A was subdivided by creating a series of deletion mutants on Staf-Krox-20 473-600 (Fig. 2A). These fusion proteins were again tested for their ability to activate transcription of the 3E.tRNA^Sec and 3E.U1 reporters in a microinjection assay. Internal deletions between amino acid 5 and amino acids 78 (6-77), 103 (6-102), 131 (6-130), and 207 (6-206) did not impair activation of the Pol III tRNA^Sec and Pol II U1 promoters (Fig. 2B, compare lane 1 and lanes 2 to 5 for the tRNA^Sec promoter; and lanes 2 and 5 for the U1 promoter in Fig. 2C). In contrast, when the internal deletions included amino acids 207 to 220 in the 6-220, 78-220, and 157-220 mutants, activation of transcription was lost (Fig. 2B, lanes 6 to 8; Fig. 2C, compare lane 2 with lanes 3 and 4). The 6-206 truncation, associated with a deletion between amino acids 224 and 243 (6-206+225-242), as well as the deletion between amino acids 243 and 263 (244-262), led to a protein as active as the parental Staf-Krox-20 (Fig. 2B, compare lane 11 with lanes 12 and 13; Fig. 2C, compare lane 2 with lanes 6 and 7). This truncation analysis demonstrates that the region from 207 to 224 is critical for activation of the tRNA^Sec and U1 promoters.

View larger version (25K): [in this window] [in a new window]   View larger version (26K): [in this window] [in a new window]   View larger version (39K): [in this window] [in a new window]   FIG. 2.   Characterization of the minimal activation domain for Pol II and Pol III snRNA promoters. (A) Schematic structures of the various truncated and fused Staf-Krox-20 proteins. The Krox-20 DBD is hatched. (B) Xenopus microinjection assay of transcriptional activation from the 3E.tRNASec promoter by the effectors depicted in panel A. Identities of the effector proteins are indicated above each lane. Lanes 9, 10, and 19, assays without effector (). Positions of the 5S maxigene and tRNASec RNAs are indicated. Autoradiographs corresponding to lanes 1 to 9, 10 to 15, 16 to 18, and 19 to 20 were from separate experiments. (C) Xenopus microinjection assay of transcriptional activation from the 3E.U1 promoter by the effectors depicted in panel A. Positions of the U1 and 5S maxi RNAs are indicated. Lane 1, assay without effector ().

Leu-213 is critical for transactivation activity on snRNA-type promoters. The region 207 to 224 does not have a preponderance of amino acids, such as glutamines, isoleucines, prolines, or serines-threonines that typify other well-characterized activation domains. To determine residues responsible for activation, mutants were created by alanine substitution mutagenesis in the context of Staf-Krox-20 473-600. Alanine substitutions of amino acids 207 to 211, 221 to 224, and serine 211 (Fig. 3A, PDTIS207-211AAAAA, SIEG221-224AAAA, and S211A, respectively) had a very minor or no effect at all on the ability of the protein to activate the 3E.tRNA^Sec and 3E.U1 reporters in the Xenopus microinjection assay (Fig. 3B). Replacement of the three residues at positions 214, 215, and 219 (Fig. 3A, E214A+Q215A+K219A) produced a twofold drop in the activity of the protein. In stark contrast, the simultaneous replacement of the two residues at positions 213 and 216 by an alanine L213A+Y216A (Fig. 3A) completely abolished the transactivating ability of the protein (Fig. 3B). To assess the contributions of Leu-213, Glu-214, Gln-215, Tyr-216, and Lys-219, they were individually mutated to alanines. Whereas changing Glu-214, Gln-215, Tyr-216, and Lys-219 had only a moderate effect, the L213A mutant by itself was severely defective compared to Staf-Krox-20 473-600 (Fig. 3B).

View larger version (24K): [in this window] [in a new window]   FIG. 3.   Identification of amino acids essential for transactivation of snRNA and snRNA-type promoters. (A) The amino acid sequence of the Staf domain 207 to 224 is shown with the changes introduced in the mutants. Dots in the mutant sequences indicate an unchanged residue. (B) Transcriptional activation assays from the Pol II 3E.U1 and Pol III 3E.tRNASec promoters by the various effectors depicted in panel A. Activity is shown as a percentage of the activity observed with Staf-Krox-20 473-600 construct containing the wild-type snRNA activation domain. Activities in microinjections assays with the 3E.U1 and 3E.tRNASec reporters are shown by solid and shaded bars, respectively. Results are the averages of two independent determinations. Identities of the effector proteins are indicated below each histogram. , assay without effector.

Staf contains distinct activation domains for mRNA and snRNA-type promoters. To determine if the snRNA activation domain of Staf can also activate transcription from an mRNA promoter, we tested the ability of truncated Staf-Krox-20 proteins to enhance the transcription level of the 6E.tkCAT reporter in a microinjection assay. Surprisingly, deletion of residues 212 to 220 in Staf-Krox-20 (212-220) and Staf-Krox-20 473-600 (212-220+473-600), corresponding to sequence ALEQYAAKV, had no effect on the transcription level of 6E.tkCAT, while the same mutant effectors failed to activate snRNA-type promoters (Fig. 4, compare lane 2 with lanes 3 and 5). In sharp contrast, deletion of residues 84 to 176 (84-176+473-600; Fig. 2A) in Staf-Krox-20 473-600 completely abolished activation from the thymidine kinase (tk) promoter of 6E.tkCAT (Fig. 4, compare lanes 2 and 4) but is absolutely innocuous to activation of transcription from the tRNA^Sec and U1 promoters (Fig. 2B, lanes 11 and 14; Fig. 2C, lanes 8 and 11). This analysis indicated that a region containing residues 84 to 176 is the major contributor to activation of transcription from an mRNA promoter. A fusion of the region containing residues 84 to 176 to the Krox-20 DBD, to yield S84-176K (Fig. 2A), led to efficient activation of transcription from 6E.tkCAT (Fig. 4, compare lanes 2 and 6). However, this same chimeric protein was unable to activate snRNA and snRNA-type promoters (Fig. 2B, lanes 16 and 17; Fig. 2C, lanes 8 and 12). Therefore, this 93-amino-acid region appears to possess an autonomous transactivation capability in the absence of the rest of Staf.

View larger version (41K): [in this window] [in a new window]   FIG. 4.   The mRNA activation domain is distinct from the snRNA counterpart. Representative CAT assays of transcriptional activation from the 6E.tkCAT promoter using extracts from microinjected Xenopus oocytes. Identities of the effectors are indicated above the lanes. The effector structures are depicted in Fig. 2A. Expression of the proteins was confirmed by gel mobility shift assays using nuclear extracts from microinjected Xenopus oocytes (data not shown).

Four repeat units for an mRNA activation domain. So far, our investigations were carried out with the use of chimeric Stafs bearing the Krox-20 DBD. To obviate the need for this heterologous DBD, which was required to eliminate the endogenous Staf background of the Xenopus oocytes, another strategy was taken. Drosophila cells, which were shown not to possess a Staf activity, were employed (39). In these cell lines, we could assay only the mRNA activation domain since the literature provides experimental evidence that transcriptional activation of snRNA promoters in Drosophila does not proceed similarly to that in vertebrates (8). Full-length Staf or various deletion mutant versions were expressed under the control of the constitutively active actin 5C promoter (Fig. 5A). The reporter AE.tkCAT contains three Staf binding site AE upstream of the tk promoter fused to the CAT reporter gene. Cotransfection of wild-type Staf and reporter plasmid resulted in an efficient transcriptional activation of the CAT gene (Fig. 5B, compare lanes 1 and 2). Deletions of the C-terminal domain C and of amino acids 212 to 220 (Staf 473-600 and Staf 212-220, respectively) (Fig. 5A) did not reduce the ability of Staf to activate the tk promoter (Fig. 5B, compare lane 2 with lanes 4 and 5). In contrast, however, deletion of the N-terminal domain A or an internal deletion between amino acids 83 and 177 (Staf6-248 and Staf84-176, respectively) (Fig. 5A) produced molecules which were unable to stimulate the promoter of the CAT reporter gene (Fig. 5B, compare lane 2 with lanes 3 and 6). The transcriptional activity could be fully recovered with Staf 84-176 (Fig. 5A), a fusion protein consisting of only amino acids 84 to 176 fused to the Staf DBD (Fig. 5B, compare lanes 2 and 7). These results, obtained with Drosophila cells, corroborate those already obtained with the Xenopus oocyte system with chimeric Staf-Krox-20 proteins.

View larger version (25K): [in this window] [in a new window]   FIG. 5.   Transactivation abilities of Staf deletion mutants from an mRNA promoter, assayed in Drosophila cells. (A) Schematic structures of the various Staf deletion mutants. The A, B, and C domains of Staf are indicated. The Staf DNA binding domain is represented in black. (B) Representative CAT assays using extracts from Drosophila cells transfected with the effectors indicated above each lane.

View larger version (28K): [in this window] [in a new window]   FIG. 6.   Maximal transcriptional activation of an mRNA promoter by Staf requires the presence of repeats R1 to R4. (A) Sequence of the Staf mRNA activation domain (amino acids 84 to 176) with the location of repeats R1 to R4. (B) Schematic representation of the seven different effector constructs. The Staf DNA binding domain and the repeats are indicated in the black and gray areas, respectively. (C) Transcriptional stimulation of the AE.tkCAT promoter containing three AE Staf binding sites, assayed in Drosophila cells. The autoradiograph from a representative experiment is shown. The names of the effector constructs and the relative transcriptional activities are indicated above the lanes. The values represent the ratio of the CAT activity to that in the absence of effector.

View larger version (21K): [in this window] [in a new window]   FIG. 7.   Alanine substitution mutagenesis of repeat R4 in the Staf mRNA activation domain. (A) Sequence alignments of repeats R1 to R4. (B) Staf 77-156 containing repeat R4 with the indicated mutations were tested in Drosophila cells for transcriptional activation of a cotransfected AE.tkCAT promoter containing three AE Staf binding sites. The activity of the Staf 77-156 construct containing wild-type R4 was defined as 100. Results are the averages (+/ values indicate standard deviations) of three independent determinations. wt, wild type.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

We have shown previously that Staf plays an important role, not only in the transcriptional activation of snRNA and snRNA-type genes transcribed by RNA polymerases II and III but also to stimulate expression from an mRNA promoter (32, 38, 39). In this study we have addressed the molecular basis of these pleiotropic effects on transcriptional activity.

Our results summarized in Fig. 8 show that domain A, corresponding to the N-terminal region of Staf, contains two physically and functionally distinct activation domains acting specifically on snRNA-type and mRNA promoters. To our knowledge, this is the first report describing such a feature in a transactivator protein. Based on our work, we propose that the two separate Staf activation domains are involved in two different transactivation mechanisms: one for the specific activation of mRNA promoters and the other for that of snRNA and snRNA-type promoters. Forsberg et al. (11) showed that Sp1 is able to stimulate transcription from mRNA and snRNA promoters. However, it must be pointed out that in this case overlapping domains are involved, while we identified two distinct, specialized domains in Staf.

View larger version (22K): [in this window] [in a new window]   FIG. 8.   Summary of Staf subdomains. (A) Schematic representation of Staf shown with delineation of the various subdomains, based on the present and earlier studies. (B) mRNA and snRNA activation domains. Left panel, positions and sequence alignments of the four substructures R1 to R4; a consensus sequence was derived at positions for which at least three of the four repeats contain identical residues. Right panel, sequence comparisons of the snRNA activation domain with regions of ZNF76 and ZNF143 (35, 50). (C) Helical wheel representation of Staf residues 212 to 220. (D) Sequence comparison of Staf residues 212 to 220 with the -helix-1 from the Oct-1 POU domain. The 1 helix is numbered according to Herr and Cleary (17).

A number of activators, enhancing transcription from mRNA promoters only, have been shown to contain multipartite activation domains. This is the situation found, for example, in TEF-1, Oct-2, and NGF1-A (19, 36, 46). Interestingly, we have shown in this work that Staf contains one single domain, necessary and sufficient for optimal transactivation of an mRNA promoter containing a TATA box. This 93-amino-acid activation domain (Staf residues 84 to 176) has a net negative charge and contains 11 and 17% of glutamine and hydroxylated residues, respectively, in a manner analogous to what is found in many other transactivation domains (21, 27, 51). A prominent feature of the Staf mRNA activation domain is its 15-amino-acid motif, repeated four times (Fig. 8B). A related situation was described for the Oct-2 activator, which contains a twofold reiterated 18-amino-acid motif (44). The Staf mRNA activation domain can be divided into subregions containing one, two, three, or four substructures that still have the potential to enhance transcription. This fact suggests that each substructure constitutes a functional unit, even though all four motifs are required to confer optimal transactivation potential. In line with the occurrence of repeated substructures in Staf, previous studies revealed that artificial reiteration of an activation domain can also lead to a significant increase in transcriptional activity (9, 18, 40, 43, 45).

From a mechanistic point of view, involvement of a reiterated motif in the activation process can be apprehended in two ways. In the first one, the affinity of the activator for the basal transcription complex is higher because the number of contacts with the target increases with the number of repeats. Alternatively, redundance of the interaction surfaces could raise the affinity of the activator by increasing the probability of binding to the target. Structural studies on the VP16 and p53 activation domains suggested that the FXX motif (X stands for any amino acid and  for hydrophobic amino acids) is a general recognition element of the acidic activation domain of TAF_II31 (22, 52). The Staf mRNA activation domain, though, does not conform to this motif and presents no obvious sequence similarity with the other known activation domains, suggesting that it may constitute the representative of a novel category.

Oct-1 has been shown to activate snRNA promoters with multipartite snRNA activation domains (28, 29, 46). Instead, our study established that Staf performs this function with one single 18-amino-acid region (residues 207 to 224) bearing no sequence similarity to other, previously identified activation domains. The mutational analysis of this domain revealed the prime importance of the 9-amino-acid subdomain of sequence ALEQYAAKV (residues 212 to 220). This subdomain is conserved in the human proteins ZNF76 and ZNF143 (35, 50), which possess significant sequence similarities to Staf (33, 39) (Fig. 8B). This, in addition to the finding that ZNF76 is able to activate transcription of snRNA genes (33), suggests that this new short motif plays a central role in the activation process by constituting an interaction surface within the snRNA transcription complex. Based on protein modeling algorithms, the subdomain composed of residues 212 to 220 is predicted to form an -helical region (7, 12). Displaying residues 212 to 220 as a putative -helix reveals its amphipathic character, with hydrophilic residues E214, Q215, and K219 distributed on one face (Fig. 8C) and hydrophobic residues L213 and Y216, on the other. From this observation, we propose that hydrophobic protein-protein interactions involving Leu-213 and Tyr-216 are critical in the snRNA gene transcriptional activation by Staf.

A follow-up in the study of domain delineation in a transcriptional activator resides in the identification of the targets that it must contact to activate transcription. Our results show clearly that a short domain of Staf specifically acts as a potent activator of Pol II and Pol III snRNA and snRNA-type gene transcription. The data additionally suggest that in the activation process the molecular target of the snRNA activation domain is a component of the basal Pol II and Pol III snRNA transcription complex. The multisubunit SNAPc complex, also called the PSE transcription factor PTF, which recognizes specifically the PSE, is the only factor known to be specific and common to Pol II and Pol III snRNA transcription complexes (13, 31, 37, 55). In this respect, it is tempting to speculate that Staf interacts directly or indirectly with SNAPc. The Oct-1 POU protein has been implicated in transcriptional activation of Pol II and Pol III snRNA promoters, as well. In this case, the POU DNA binding domain of Oct-1 not only serves to target the protein to the octamer sequence on the DNA but also is intimately involved in the transcriptional activation process by stabilizing the binding of the basal transcription complex SNAPc (10, 17, 28, 29, 31). It has been proposed that the POU domain does so by recruiting SNAPc to the PSE by direct protein-protein contacts involving the N-terminal part of the POU-specific domain. It is interesting to note that the central part of the Staf snRNA activation domain bears sequence similarity to the -helix-1 constituting the N terminus of the POU-specific domain (Fig. 8D). From this, we propose that Staf and Oct-1 recognize the same target in the SNAPc complex.

In conclusion, we have presented experimental evidence for the presence in Staf of two functionally distinct transactivation domains with specific snRNA-type and mRNA gene targets. This represents a novel feature in a transactivator protein and adds an important new dimension to structure-function studies that are being performed with transactivation domains.