Artigo Acesso aberto Revisado por pares

Cell differentiation by interaction of two HMG-box proteins: Mat1-Mc activates M cell-specific genes in S.pombe by recruiting the ubiquitous transcription factor Ste11 to weak binding sites

1997; Springer Nature; Volume: 16; Issue: 13 Linguagem: Inglês

10.1093/emboj/16.13.4021

ISSN

1460-2075

Autores

Søren Kjærulff, Dennis Dooijes, Hans Clevers, Olaf Nielsen,

Tópico(s)

Genomics and Chromatin Dynamics

Resumo

Article1 July 1997free access Cell differentiation by interaction of two HMG-box proteins: Mat1-Mc activates M cell-specific genes in S.pombe by recruiting the ubiquitous transcription factor Ste11 to weak binding sites Søren Kjærulff Søren Kjærulff Department of Genetics, Institute of Molecular Biology, University of Copenhagen, DK- 1353 Copenhagen K, Denmark Search for more papers by this author Dennis Dooijes Dennis Dooijes Department of Immunology, University Hospital, PO Box 88500, 3508 GA, Utrecht, The Netherlands Search for more papers by this author Hans Clevers Hans Clevers Department of Immunology, University Hospital, PO Box 88500, 3508 GA, Utrecht, The Netherlands Search for more papers by this author Olaf Nielsen Corresponding Author Olaf Nielsen Department of Genetics, Institute of Molecular Biology, University of Copenhagen, DK- 1353 Copenhagen K, Denmark Search for more papers by this author Søren Kjærulff Søren Kjærulff Department of Genetics, Institute of Molecular Biology, University of Copenhagen, DK- 1353 Copenhagen K, Denmark Search for more papers by this author Dennis Dooijes Dennis Dooijes Department of Immunology, University Hospital, PO Box 88500, 3508 GA, Utrecht, The Netherlands Search for more papers by this author Hans Clevers Hans Clevers Department of Immunology, University Hospital, PO Box 88500, 3508 GA, Utrecht, The Netherlands Search for more papers by this author Olaf Nielsen Corresponding Author Olaf Nielsen Department of Genetics, Institute of Molecular Biology, University of Copenhagen, DK- 1353 Copenhagen K, Denmark Search for more papers by this author Author Information Søren Kjærulff1, Dennis Dooijes2, Hans Clevers2 and Olaf Nielsen 1 1Department of Genetics, Institute of Molecular Biology, University of Copenhagen, DK- 1353 Copenhagen K, Denmark 2Department of Immunology, University Hospital, PO Box 88500, 3508 GA, Utrecht, The Netherlands *Corresponding author. E-mail: [email protected] The EMBO Journal (1997)16:4021-4033https://doi.org/10.1093/emboj/16.13.4021 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The Schizosaccharomyces pombe mfm1 gene is expressed in an M cell-specific fashion. This regulation requires two HMG-box proteins: the ubiquitous Ste11 transcription factor and the M cell-controlling protein Mat1-Mc. Here we report that the mfm1 promoter contains a single, weak Ste11-binding site (a so-called TR-box) that can confer M-specificity on a heterologous promoter when present in eight copies. In vitro, both Mat1-Mc and Ste11 can bind this box with approximately the same affinity. The Mat1-Mc protein caused a dramatic increase in the DNA-binding of Ste11 to this box, under conditions where we could not detect Mat1-Mc in the resulting protein–DNA complex. When we changed a single base in the mfm1 TR-box, such that it resembled those boxes found in ubiquitously expressed genes, Ste11 binding was enhanced, and in vivo the mfm1 gene also became expressed in P cells where Mat1-Mc is absent. These findings suggest that M-specificity results from Mat1-Mc-mediated Ste11 binding to weak TR-boxes. We have also defined a novel motif (termed M-box), adjacent to the mfm1 TR-box, to which Mat1-Mc binds strongly. A DNA fragment containing both the TR- and the M-box allowed the formation of a complex containing both Ste11 and Mat1-Mc. A single copy of this fragment was sufficient to activate a heterologous promoter in an M-specific fashion, suggesting that these two boxes act in a synergistic manner. Introduction The HMG-box is a recently discovered DNA-binding element present in several eukaryotic proteins. Sequence analysis indicates that the HMG-box is a stretch of ∼70 amino acids, with a net positive charge and an abundance of aromatic residues and prolines (Baxevanis and Landsman, 1995). Based on sequences and binding characteristics, the HMG-box protein family can be divided into two groups. One group includes proteins such as HMG1–2 and UBF that often contain multiple HMG-boxes and recognize DNA in a structure- rather than sequence-dependent fashion (reviewed by Bustin and Reeves, 1996). The second group contains proteins with a single HMG-box, and these bind DNA in a sequence-specific manner and some of them are activators of transcription. Members of this subfamily include the mammalian sex-determining factor SRY (Gubbay et al., 1990; Sinclair et al., 1990) and the related SOX gene products, the lymphoid-specific LEF-1/TCF-1 proteins (Travis et al., 1991; Van de Wetering et al., 1991; Waterman et al., 1991) and several proteins involved in mating-type determination in fungi, e.g. Mat1-Mc and Ste11 from the fission yeast Schizosaccharomyces pombe (Kelly et al., 1988; Sugimoto et al., 1991; Dooijes et al., 1993). Although the similarity of the primary sequence is modest, all HMG-box proteins probably fold up in the same tertiary structure. Thus, in nuclear magnetic resonance (NMR) studies, the HMG-boxes of HMG1, SRY and SOX4 (Weir et al., 1993; van Houte et al., 1995; Werner et al., 1995) were found to adopt highly similar structures consisting of three α-helixes, arranged in an L shape. This may provide a structural basis for the unusual DNA-binding characteristics of the HMG-box proteins, which include interaction with the minor groove of the DNA helix, binding to irregular DNA structures, such as the sharp bends present in four-way junction DNA molecules, and the ability to modulate the DNA helix by bending (Bustin and Reeves, 1996). The biological significance of this DNA bending is still obscure. However, it was reported recently that DNA bending induced by the LEF-1 protein facilitates the formation of a higher-order nucleoprotein complex in the T-cell receptor α (TCRα) enhancer (Giese et al., 1995), suggesting that HMG-box proteins may have an architectural role in assembling such complexes. Supporting this idea, a mutant SRY protein that binds DNA with almost normal affinity but bends DNA in a different angle has been found in a sex-reversed XY patient (Pontigga et al., 1994). Although data on HMG-box proteins and their interaction with DNA have accumulated during the last few years, little is known about their in vivo action, and the mechanism by which the SRY protein controls male development in mammals is still largely unknown (Schafer and Goodfellow, 1996). SRY is highly homologous to the mating-type protein Mat1-Mc from S.pombe (Kelly et al., 1988; Gubbay et al., 1990; Sinclair et al., 1990). SRY and Mat1-Mc bind to the same sequence, CTTTGTT, in vitro (Dooijes et al., 1993), and Mat1-Mc has formally a function similar to SRY in establishing sex-specific gene expression. The sexual differentiation process in S.pombe is activated under conditions of nitrogen starvation, where the cells are induced to exhibit either minus (M) or plus (P) mating behaviour, depending on which gene they express from the mat1 locus. Expression of the mat1-Mc gene generates an M cell, whereas expression of the mat1-Pc gene gives rise to a P cell (Kelly et al., 1988). The Mat1-Mc and Mat1-Pc proteins specify the mating type by activating a number of, respectively, M- or P-specific genes. The products of these are the cell type-specific components of the pheromone communication system that enables the two cell types to identify each other prior to mating (reviewed by Nielsen and Davey, 1995). The M-specific genes controlled by Mat1-Mc include three structural genes for the M-factor pheromone, mfm1-3 (Davey, 1992; Kjærulff et al., 1994), the mam1 gene encoding an M-factor transporter (Christensen et al., 1997), a gene, mam2, that encodes the receptor for the P-factor pheromone (Kitamura and Shimoda, 1991) and the sxa2 gene encoding a P-factor-degrading protease (Imai and Yamamoto, 1992; Ladds et al., 1996; Yabana and Yamamoto, 1996). The Ste11 protein, which is a key transcription factor in the sexual differentiation pathway in S.pombe, is one of the few HMG-box proteins with known target sites. Ste11 binds to the so-called TR-box, TTCTTTGTTY (Sugimoto et al., 1991), the core of which is identical to the Mat1-Mc-binding site. Ste11 is activated by nitrogen starvation (Li and McLeod, 1996), and TR-boxes are found in the promoter regions of many genes that are expressed in a Ste11-dependent manner during mating. These include the M-specific genes as well as genes that are expressed in both cell types (Sugimoto et al., 1991; Kjærulff et al., 1994; Petersen et al., 1995). In the present study, we investigate the mechanism by which M-specific genes are activated during sexual differentiation in S.pombe. We show that M-specificity is conferred on the mfm1 gene by a special version of the TR-box that binds Ste11 poorly. Both Ste11 and Mat1-Mc can bind to this box and, under conditions of limiting amounts of Ste11, the Mat1-Mc protein can recruit Ste11 to the TR-box. We propose that ubiquitously expressed genes harbour a strong TR-box, to which Ste11 can bind on its own, whereas M-specificity results from Mat1-Mc-dependent Ste11 binding to a weak TR-box. Results Mat1-Mc and Ste11 bind the TR-box of the mfm1 promoter As a representative M cell-specific gene we chose mfm1, one of three structural genes encoding M-factor pheromone (Davey, 1992; Kjærulff et al., 1994). Expression from mfm1 was monitored using a fusion of the mfm1 promoter and the Escherichia coli lacZ gene (Figure 1A). This fusion behaves like the wild-type mfm1 gene; expression is limited to M cells, is induced by nitrogen starvation and further stimulated by a pheromone signal from P cells (Figure 1B, wt; Kjærulff et al., 1994). Figure 1.Mutagenesis reveals two elements important for proper expression of the mfm1 gene. (A) Sequence of the mfm1 promoter. The TR-box and the M-boxes are in bold and underlined. Transcription start points (tsp) are indicated by an arrow. The KpnI site is the fusion point between mfm1 and lacZ. The right panel shows a primer extension analysis (P.) to determine the transcription start point of the mfm1–lacZ construct. The sequence of mfm1–lacZ was run in parallel (GATC). The arrowheads indicate the positions of the tsps. (B) Mutational analysis of the mfm1–lacZ fusion. wt is the wild-type mfm1–lacZ construct. Ste11 + pnmt-Mc is the wild-type mfm1–lacZ construct integrated in an M ste11 strain containing a plasmid overexpressing mat1-Mc. The sequence located 69–98 bp upstream of the transcription start point has been highlighted. Point mutations introduced in this region are indicated by lower case letters (TR-mut1-3 and M-box mut1-2). All constructs were integrated in the mfm1 locus, at the SmaI site. β-Galactosidase activities were measured in exponentially growing cultures (+), in cultures starved of nitrogen for 5 h (÷) and in cultures starved of nitrogen and exposed to pheromone for 5 h (÷, P). β-Galactosidase activities are expressed in Miller units and represent the mean of three separate trials. Restriction sites: Sm, SmaI; EV, EcoRV; H, HindIII; EI, EcoRI. Download figure Download PowerPoint The Ste11 transcription factor is required for induction of mfm1 (Kjærulff et al., 1994). While Ste11 may regulate mfm1 expression through its control of mat1-Mc (Sugimoto et al., 1991), it also appears to play a more direct role, since a ste11 strain harbouring a plasmid that produces functional Mat1-Mc protein from the nmt promoter still fails to transcribe the mfm1–lacZ fusion (Figure 1B, Ste11 + pnmt-Mc). Consistent with this, a TR-box is situated 79 bp upstream of the transcription start point (tsp) in mfm1, and all other known M-specific genes contain a TR-box at a similar position (Figure 1A, Table I). To determine the functional significance of this element, we altered the conserved G of the mfm1 TR-box to a T. This mutation prevented binding of Ste11 in vitro (see below) and almost completely abolished promoter function (Figure 1B, TR-mut1). Table 1. Comparison of TR-boxes found in M-cell-specific genes and in genes expressed in both cell types Gene Sequence Position (C from ATG) Orientation Reference In genes expressed in both cell types fus1a,b GTATTTCTTTGTTCTTTA −247 > 1 fusla,b AACTTTCTTTGTTCGGTT −158 < 1 map1a GTGTTTCTTTGTTACAAA −59 < 2,3 mei2a CGATTTCTTTGTTCCTAT −1890 > 4,5 mei2a AAGTTTCTTTGTTTTACA −1868 > 4,5 mei2a GAGATTCTTTGTTTACTT −1696 > 4,5 mei2a,b TAACTTCTTTGTTCTCTA −1516 > 4,5 mei2a,b TCTTTTCTTTGTTTGTTT −911 > 4,5 rep1a ATTTTTCTTTGTTTACAT −178 > 6 rep1a TACATCCTTTGTTTACAA −165 > 6 spk1 AAGTTTCTTTGTTAATGT −602 < 7 spk1 AACTTTCTTTGTTATTGT −598 > 7 ste4a TGCTTTCTTTGTTATAAA −84 > 8 ste6a GAATTTCTTTGTTTACTA −174 < 9 ste11a TTGTTTCTTTGTTGCAAT −1375 > 5 zfs1 ATTTTTCTTTGTTTGACG −493 > 10 zfs1 CAACTTCTTTGTTCGTTT −92 > 10 Consensus TTTCTTTGTT In M cell-specific genes mam1a ATGGGCCTTTGTTAGGTA −203 > 11 mam2a ATTCCTCTTTGTTTAGAA −123 < 12 mfm1a,b AGACTTCTTTGTTGGTCG −157 < 13,14 mfm2a TGATCTCTTTGTTCATTT −138 < 13 mfm3a AGACTTCTTTGTTGTTTC −154 < 15 sxa2 GGGTGTCTTTGTTGCCCA −322 > 16 Consensus TCTTTGTT a Expression has been shown to be reduced in a ste11− mutant. b TR-box has been shown to be required for expression in vivo. References: 1, Petersen et al. (1995); 2, Yabana and Yamamoto (1996); 3, Nielsen et al. (1996); 4, Watanabe et al. (1988); 5, Sugimoto et al. (1991); 6, Sugiyama et al. (1994); 7, Toda et al. (1991); 8, Okazaki et al. (1991); 9, Hughes et al. (1990); 10, Kanoh et al. (1995); 11, Christensen et al. (1997); 12, Kitamura and Shimoda (1991); 13, Davey (1991); 14, this work; 15, Kjærulff et al. (1994); 16, Imai and Yamamoto (1992). The Mat1-Mc protein was shown previously to bind the sequence CTTTGTT (Dooijes et al., 1993), which constitutes the core of the TR-box, and we therefore compared the abilities of the Ste11 and Mat1-Mc proteins to bind the TR-box of the mfm1 promoter in vitro. These experiments showed that E.coli-expressed GST–Ste11 and malE–Mat1-Mc fusion proteins bind to an oligonucleotide covering the TR-box of mfm1 with approximately equal affinity (Kd ∼10−8 M). In both cases, the retarded complex was competed efficiently by the TR-box, but not by the mutagenized TR-box (TR-mut1), nor by two oligonucleotides containing unrelated sequences (Figure 2A and B). Figure 2.The HMG-boxes of Mat1-Mc and Ste11 bind specifically to the TR-box of the mfm1 gene. EMSA performed on various labelled probes (see Materials and methods) using E.coli-expressed, purified GST–Ste11 fusion protein (A) or malE–Mat1-Mc fusion protein (B). Competitor DNA was added in 100- or 500-fold molar excess, as indicated by the triangles. FP indicates unbound probe. The mutation in TR-mut is shown in Figure 1B (TR-mut1). The TCF-box and LEF-box oligonucleotides contain binding sites for mammalian HMG-box proteins. Download figure Download PowerPoint The TR-box of mfm1 confers M-specific expression on a heterologous promoter Given the fact that Mat1-Mc binds the mfm1 TR-box, we investigated whether this element could confer M-specific expression on a heterologous promoter. Various copies of it were inserted in the Saccharomyces cerevisiae CYC1 minimal promoter, which was fused to the E.coli lacZ gene (Lowndes et al., 1992). A single mfm1 TR-box in the minimal promoter produced negligible β-galactosidase activity (Figure 3). However, when eight copies of this TR-box were present, a high level of β-galactosidase activity was induced by nitrogen starvation, and this activity was stimulated further by a pheromone signal. Most importantly, however, induction of expression was restricted to M cells. Figure 3.Eight copies of the TR-box of mfm1 confer M cell-specific expression on a minimal promoter from the S.cerevisiae cytochrome c gene fused to the E.coli lacZ gene. One or eight copies of an oligonucleotide containing the mfm1 TR-box were inserted in the vector pSPΔ178 (Lowndes et al., 1992), giving pcyc-TR (1/8). One or eight copies of an oligonucleotide covering the most downstream M-box of mfm1 (see below) were inserted in pSPΔ178, giving pcyc::M–box (1/8). The mfm1–lacZ fusion (see Figure 1) inserted in the vector pDW232 (Weilguny et al., 1991) was used as positive control (pmfm1). The vector pSPΔ178 (pcyc) was used as a negative control. The constructs were transformed into h− (M) and h+ (P) strains and assayed for β-galactosidase activities. The results are expressed in Miller units and each number is the average of three separate trials. Open boxes represent activities in vegetatively growing cultures, boxes hatched vertically are activities in nitrogen-starved cultures and boxes hatched horizontally are activities in nitrogen-starved cultures treated with pheromone. Download figure Download PowerPoint The TR-boxes of ubiquitously expressed genes and those of M-specific genes differ in sequence These observations indicated that the M-specificity of the mfm1 promoter may lie in the TR-box or sequences in its immediate vicinity. This is quite surprising, since TR-boxes are also found in genes expressed in both cell types. However, we noticed a striking difference between the TR-boxes found in M-specific genes and those found in ubiquitously expressed genes (Table I). All promoters of the latter class contain at least one copy of the 10 bp motif, TTTCTTTGTT. This 10 bp motif is not found in any of the six known M-specific genes. Here the consensus is somewhat smaller, namely the 8 bp motif TCTTTGTT. To test whether this sequence difference was responsible for M-specificity, we changed the TR-box of the mfm1–lacZ fusion into the version found in ubiquitously expressed genes by substituting the 5′ C with a T. Interestingly, this construct was now expressed in both M and P cells (Figure 1B, TR-mut2). Hence, the addition of only one T to the 5′ end of the TR-box converts mfm1 from an M-specific gene into a gene that is expressed in both cell types. Furthermore, we note that in P cells expression requires a pheromone signal. Recently, we found by site selection that the preferred binding site of Ste11 in vitro is the 13 bp motif, TTTCTTTGTTCTC (Dooijes et al., in preparation), which resembles the TR-box found in genes expressed in both cell types. Insertion of this sequence in the mfm1 promoter also renders the gene ubiquitously expressed (Figure 1B, TR-mut3). Thus, the TR-box from ubiquitously expressed genes and the optimal Ste11-binding site both confer non-cell type-specific expression on the mfm1 gene. Furthermore, methylation interference experiments have showed that the two 5′ Ts, which are missing in the M-specific TR-boxes, indeed are contacted by the Ste11 protein (Dooijes et al., in preparation). Mat1-Mc stimulates Ste11 binding to the mfm1 TR-box How can the absence of this T residue render expression dependent on the Mat1-Mc protein? We found that Ste11 binds the TR-box of the M-specific mfm1 gene more weakly than it binds the version found in ubiquitously expressed genes, whereas the Mat1-Mc protein seemed to bind the two different boxes equally well (Figure 4A). Western analysis showed that expression of Ste11 is induced by nitrogen starvation and that the level of Ste11 protein does not appear to be higher in M cells than in P cells (Figure 4B). We therefore speculated that Mat1-Mc may control the M-specific genes by assisting binding of Ste11 to their TR-boxes, and the following observations support this idea. Under conditions of limiting amounts of purified Ste11, where virtually no complex occurred with the mfm1 TR-box probe, we observed that addition of small amounts of purified Mat1-Mc caused a significant increase in appearance of shifted complex (Figure 4C). Interestingly, this complex co-migrated exactly with the binary Ste11–DNA complex. Moreover, addition of Ste11 antibodies to the induced complex gave rise to a supershift, whereas addition of malE antibodies (which recognize the malE–Mc fusion protein) had no effect. This indicates that the induced complex detected in this assay consists mainly of Ste11 and that Mat1-Mc is not a stable component of it. The enhancement of Ste11 binding seems to be mediated specifically by Mat1-Mc, since addition of purified human SRY protein (Sinclair et al., 1990), that binds the TR-box in vitro, had no stimulatory effect on Ste11 binding (data not shown). Figure 4.(A) Ste11 binds better to the version of the TR-box found in ubiquitously expressed genes than to the mfm1 TR-box, whereas mat1-Mc binds the two versions equally well. EMSA was performed with a labelled mfm1 TR-box probe and a labelled mfm1 TR-box mut2 probe using E.coli-expressed GST–Ste11 and malE–Mat1-Mc proteins. (B) The expression pattern of the Ste11 protein is identical in the two cell types. Western analysis of the expression of Ste11 in M cells (h−) and in P cells (h+). Protein extracts (100 μg) from mitotically growing cells (nitrogen +), from nitrogen-starved cells (nitrogen −) and from nitrogen-starved cells treated with pheromones (pheromone +) were Western blotted and probed with affinty-purified anti-Ste11 antibodies. (C) Mat1-Mc stimulates specific binding of Ste11 to the TR-box of mfm1. In an EMSA, the labelled TR-box was incubated with 5 nM of purified GST–Ste11 and increasing amounts of malE–Mat1-Mc (1, 10 nM). In the last two lanes, anti-malE antibodies or anti-Ste11 antibodies were added to the binding reaction after 20 min of incubation. Download figure Download PowerPoint Mat1-Mc generates a DNase I-hypersensitive site in the mfm1 promoter Taken together, the results described above strongly indicate that M-specificity is conferred on the mfm1 gene by the presence of a special version of the TR-box, to which the binding of Ste11 is mediated by Mat1-Mc. To investigate further the mechanism by which Mat1-Mc may enhance the binding activity of Ste11, we performed an in vitro DNase I footprint on the mfm1 leader (Figure 5). As expected, both Mat1-Mc and Ste11 could protect the same 12 bp region spanning the TR-box. However, Mat1-Mc created a strong hypersensitive site just 3 bp downstream of the TR-box, suggesting that binding of this protein causes a strong distortion of the mfm1 promoter. Ste11 did not give rise to this hypersensitive site, and when Ste11 was added together with Mat1-Mc it became less apparent, consistent with a mechanism where Ste11 replaces Mat1-Mc at the TR-box. In summary, these observations confirm that Ste11 and Mat1-Mc both have the ability to bind the mfm1 TR-box. However, they also reveal that, upon binding, the two proteins modulate the DNA helix differently. Mat1-Mc seems to produce a strong distortion of the DNA, which may be crucial for efficient binding of Ste11 to this TR-box. Figure 5.Mat1-Mc creates a DNase I-hypersensitive site at the TR-box and binds several regions of the mfm1 promoter. Solid phase DNase I footprinting analysis of a 222 bp region of the mfm1 promoter spanning the TR-box. ÷ indicates DNase I digestion of naked DNA. Triangles indicate increasing amounts of added protein; 0.2–1.0 μM malE–Mat1-Mc; 0.1–0.5 μM Ste11. Vertical bars show the protected regions. The arrow indicates the DNase I-hypersensitive site. The sequence of the analysed region is given at the left. Download figure Download PowerPoint Mat1-Mc binds to two different elements Unexpectedly, we found that Mat1-Mc also protected a 21 bp A-rich region starting 8 bp upstream of the TR-box (Figure 5). In fact, Mat1-Mc seems to protect this region better than the TR-box. To a lesser extent, the Ste11 protein also protected this upstream region (Figure 5). We therefore compared the sequences next to the TR-boxes in the six known M-specific genes (Table II). Five of these genes each habour two ACAAT-boxes that are located, respectively, 14–16 bp and 24–26 bp from the inverted TR-box. The mam2 gene is an exception: here we only found the somewhat diverged sequence, ACATA, located 26 bp from the TR-box. We next compared the abilities of Mat1-Mc and Ste11 to bind an oligonucleotide covering the most downstream ACAAT-box of mfm1. Purified Ste11 only forms a weak complex with this element, whereas Mat1-Mc binds strongly to the ACAAT-box (Figure 6). Actually, Mat1-Mc prefers the ACAAT-box to the TR-box. Binding of Mat1-Mc was abolished by changing the conserved C of the ACAAT-box to an A. We refer to this element as an M-box, since it preferably makes complexes with Mat1-Mc. Figure 6.Mat1-Mc binds more strongly to an M-box-containing probe than to the mfm1 TR-box. EMSA on TR-box- or M-box-containing probes using E.coli-expressed malE–Mat1-Mc and GST–Ste11 proteins. The mutations M-box mut1 and M-box mut2 are shown in Figure 1. Download figure Download PowerPoint Table 2. Comparison of sequences adjacent to the TR-box in M cell-specific genes Gene Sequence mam1a TTCATTTTGTATGGTGGACAACAATGGAGAGTACCTAACAAAGGCCCATTGTGTAC mam2a ACTTTTGAGACATAGAAGTGTTTTCTGGAAATTCTAAACAAAGAGGAATTATTGGC mfm1a TGAGTATTAACAATTGACTAGACAATGGGTCCGACCAACAAAGAAGTCTCAGTTTT mfm2a CCAGAATTAAACAATGGGTCAAACAATAGGCAAATGAACAAAGAGATCACAGTTTC mfm3b TCAGTTGTAACAATTAACTAGACAATAGGCCCAACCAACAAAGAAGTCTCAGATTT sxa2b GTCCATTGTTTACAATCAACAACAATAGAGATGGGCAACAAAGACACCCAGCGAAG a Expression has been shown to be induced by nitrogen starvation. b Efficient expression requires both nitrogen starvation and a pheromone signal. To determine the functional significance of this element, we altered the conserved C of the mfm1 M-box to an A. This mutation severely reduced the mfm1 promoter function (Figure 1B, M-box mut1), demonstrating that the M-box indeed is important for expression of M-specific genes. However, the construct still supported a relatively high level of pheromone-induced expression in M cells. Based on their expression pattern, the M-specific genes can be divided into two groups. The four genes mam1, mam2, mfm1 and mfm2 are highly induced by nitrogen starvation per se, whereas mfm3 and sxa2 in addition to starvation require a pheromone signal for efficient expression (Kitamura and Shimoda, 1991; Imai and Yamamoto, 1994; Kjærulff et al., 1994; Christensen et al., 1997). A closer examination of the M-boxes revealed that mam1, mam2, mfm1 and mfm2 all have a conserved G residue in the 3′ end of one of their M-boxes, whereas mfm3 and sxa2 lack this G (Table II). To test whether this sequence difference could explain the pheromone-induced nature of mfm3 and sxa2, we changed the G in the M-box of the mfm1 promoter to an A. This mutation also reduced binding of the Mat1-Mc protein to the M-box in vitro (Figure 6), and now expression was only moderately induced by nitrogen starvation (Figure 1B, M-box mut2). However, this construct could still be induced to a relatively high level by a pheromone signal. Hence, strong interaction of Mat1-Mc with the M-box is required for expression induced by nitrogen starvation, whereas this interaction apparently is not important for pheromone stimulation. Synergistic function of the TR-box and the M-box of mfm1 We were unable to demonstrate any activation of the S.cerevisiae CYC1 promoter by the M-box, even when eight copies of it were inserted (Figure 3). However, an oligonucleotide containing the M-box combined with its downstream TR-box conferred M-specific expression on the CYC1 promoter—even in one copy (Figure 7A). Given the fact that a single TR-box had no effect on the minimal promoter (Figure 3), this result implies that the two HMG-box-binding sites work in a synergistic fashion. This synergy could reflect stable interaction between the Mat1-Mc and Ste11 proteins and DNA. To test this idea, we used the oligonucleotide containing the M-box and the TR-box in an electrophoretic mobility shift assay (EMSA). As expected, both Mat1-Mc and Ste11 bind this oligonucleotide on their own (Figure 7B). However, when both HMG-box proteins were present simultaneously, a unique complex was formed with slower mobility. Addition of antibodies against Ste11 or malE–Mat1-Mc both caused a supershift of this slow migrating complex, showing that the complex is of ternary nature containing Ste11, Mat1-Mc and DNA. Hence, Mat1-Mc may have two roles in the mfm1 promoter; it enhances the binding of Ste11 to the TR-box (Figure 4C) and it forms a ternary complex with Ste11 when both the TR- and M-box are present (Figure 7B). Figure 7.(A) Synergistic function of the TR-box and M-box of mfm1 on the CYC1 minimal promoter. One or four copies of an oligonucleotide containing the most downstream M-box and the TR-box of mfm1 were inserted in pSPΔ178, giving pcyc-TR-M-box (1/4). The constructs were transformed into h− (M) and h+ (P) strains and assayed for β-galactosidase activities as described in Figure 3. (B) Mat1-Mc and Ste11 form a stable ternary complex with the mfm1 TR-box combined with its upstream M-box. In an EMSA, a labelled probe containing the TR-box and the most downstream M-box was incubated with purified Ste11 protein (50 nM) and increasing amounts of malE–Mat1-Mc (20–50 nM). In the last two lanes, anti-malE antibodies or anti-Ste11 antibodies were added to the binding reaction after 20 min of incubation. Download figure Download PowerPoint We next asked whether Mat1-Mc and Ste11 also interact in vivo. To address this question, we tagged the Mat1-Mc protein N-terminally with an influenza haemagglutinin (HA

Referência(s)