BBRC

Biochemical and Biophysical Research Communications 295 (2002) 519525 www.academicpress.com

Identication of the enhancer binding protein MBF-1 of the sea urchin modulator a-H2A histone geneq
Claudia Alessandro,a,1 Paola Di Simone,a,1 Alessia Buscaino,a Letizia Anello,c Franco Palla,c and Giovanni Spinellia,b,*
a

Dipartimento di Biologia Cellulare e dello Sviluppo (Alberto Monroy), Viale delle Scienze Parco dOrleans II, Palermo 90128, Italy b Centro di Oncobiologia Sperimentale, Viale delle Scienze Parco dOrleans II, Palermo 90128, Italy c Istituto di Biologia dello Sviluppo del Consiglio Nazionale delle Ricerche, via Ugo La Malfa 153, Palermo 90146, Italy Received 14 June 2002

Abstract The modulator of the sea urchin a-H2A histone gene promoter is the only enhancer identied in the a-histone gene cluster. Binding of a single factor, denoted MBF-1, has previously detected in nuclear extracts from morula and gastrula embryos. Here, we describe the cloning of MBF-1 by screening a cDNA expression library with a tandem array of modulator binding sites. MBF-1 presents no similarity with other DNA binding proteins and contains nine Krppel like Zn ngers. In vitro translated proteins and a u factor from nuclear extracts interact with the modulator with identical specicity. In addition, MBF-1 expressed in human cells transactivates a reporter gene driven by an array of modulator sites. The DNA binding domain consists of the Zn ngers plus an adjacent basic region, while sequences in the N-terminal region mediates the transactivation function. MBF-1 is expressed in the unfertilized egg and in early and late developmental stages thus conrming that it is not a stage specic enhancer binding factor and that silencing of the a-H2A gene after hatching is not due to the lack of the transactivator. 2002 Elsevier Science (USA). All rights reserved.
Keywords: Histone genes; Sea urchin; Enhancer; Transcription factor; DNA binding domain

The sea urchin early or a-histone genes have been widely used as model system for studying the mechanisms underlying transcriptional regulation throughout embryogenesis. They are organized as tandem arrays of a repeating unit each containing one copy of each of the ve histone genes [1]. Transcription occurs during meiotic maturation of oocytes and transiently during early cleavage, reaching a maximum at around the fth division. By the gastrula stage the early transcripts are undetectable and the late set of histone genes becomes the most active form [2,3]. Silencing of the a-histone gene set at late stage of development is evidenced also by the

q The sequence reported in this paper has been deposited in the GenBank database. Accession No.: AJ430050. * Corresponding author. Fax: +3991-420-897. E-mail address: spinelli@unipa.it (G. Spinelli). 1 The rst two authors contributed equally to this work.

remodeling of the chromatin structure, such as regaining of the nucleosomal organization and modication of the nuclease hypersensitive pattern [46]. The regulative regions controlling the expression of the a-histone genes during development have been extensively investigated. Each of the ve genes, within a tandem repeat, seems to constitute an independent transcription unit regulated by both ubiquitous and gene specic cis-acting elements [79]. Previously, we reported the cis-acting regulative sequences involved in the timing of transcription of the aH2A gene during embryogenesis of the sea urchin Paracentrotus lividus. We identied near the 30 end a DNA sequence that confers temporal regulation to a transgene driven by the H2A promoter [10]. Interestingly, this region includes an insulator element [11,12], and recent dissection experiments, have shown that it is essential also for the specic temporal gene silencing that occurs at gastrula stage (unpublished observations). In

0006-291X/02/$ - see front matter 2002 Elsevier Science (USA). All rights reserved. PII: S 0 0 0 6 - 2 9 1 X ( 0 2 ) 0 0 7 0 8 - 8

520

C. Alessandro et al. / Biochemical and Biophysical Research Communications 295 (2002) 519525 shaking for 1 h, lters were washed with the probe binding solution (binding buer supplemented with 10 lg/ml of salmon sperm DNA and 0.25% nonfat dry milk). Binding reaction was carried out by exposing lters to 1 106 dpm of ligated M30 oligonucleotides in probe binding solution. Filters after overnight incubation at 4 C with gentle shaking were washed three times with a binding solution supplemented with 0.25 M guanidine hydrochloride and 0.1% Triton X-100 and exposed to Kodak X-Omat AR lms overnight. Positive plaques were picked up for plaque purication and the screening procedure repeated two more times. Construction of plasmids. M30-EGFP vector expressing the enhanced green uorescent protein was constructed by replacing the CMV enhancer/promoter cassette with an array of M30 oligonucleotide and a minimal tk promoter in the pEGFP (Clontech) vector. To express MBF-1 cDNA or its deletion mutants under the control of the CMV enhancer, full-length cDNA or its deletion mutants replaced the green uorescent protein coding region of the pGREEN-lantern vector (Life Technologies). Construct A was generated by cloning EcoRI andXbaI digested MBF-1 full-length cDNA in the pGREEN vector. Plasmids B and C were obtained by cloning, respectively, the DNA fragments produced by the SpeI and EcoRI plus HindIII digestions of the full-length cDNA. Constructs D and E were obtained by PCR by using Mg20/Mg21 and Mg20/Mg22 oligonucleotides, respectively. The Mg20 primer contains the Kozac and translation initiation site of EGFP. To generate the pSG424-MBF plasmid used in the transactivation assay the N-terminal MBF-1 coding region, obtained by PCR with Mg25/Mg26 oligonucleotides, was fused in frame with the binding domain of the yeast GAL4 transcription factor [19]. To generate GAL4-CAT plasmid containing an array of GAL4 binding sites upstream the thymidine kinase minimal promoter driving the expression of the CAT reporter. In vitro translation and electrophoretic mobility shift assays. Plasmids were linearized with HpaI, and transcribed in vitro (Stratagene) to generate capped mRNAs. One lg of RNAs was translated in a reticulocyte cell free system (Ambion). A TNT (Promega) coupled transcription and translation was alternatively used in some experiments. Labeled translation products carried in the presence of [35 S]methionine were run in a SDSPAGE 10% polyacrylamide gel. The 50 32 P-end labeled wild type and mutant M30 double stranded oligonucleotides were used in the electrophoresis mobility shift assays (EMSA). One ng of end labeled probes was incubated for 30 min on ice, with 5 lg of nuclear extracts from sea urchin embryos at gastrula stage [20] or 5 ll of the translation reaction. Five or one lg of poly(dAdT)(dAdT) was, respectively, added to the two binding reactions in 20 ll of [10 mM Hepes (pH 7.9), 60 mM KCl, 1 mM DTT, 1 mM EDTA, 4% Ficoll]. In competition experiments 100 ng of unlabeled homologous or heterologous probes were added to the preincubation mixture prior to the addition of the extracts or the translated products. Binding reactions were run on nondenaturating 5% acrylamide gels in [50 mM Tris, 50 mM H3 BO3 , and 2 mM EDTA (pH 8.3)]. Transfection of human cells and CAT assay. Human U2-OS cells were cultured (at 37 C in 5% CO2 ) in DMEM supplemented with 10% FBS (Euroclone). M30-EGFP DNA transfection and selection of clones were as described [21]. Clones were isolated, expanded, and transfected with the wild type and deletion mutants MBF-1 expressing plasmid. EGFP expression was evaluated by microscopic observations. GAL4-CAT and the pSG424-MBF, or the pSG424 were cotransfected into U2-OS cells. Two days after transfection, cells were harvested and CAT assay on cell extracts was performed as described [11]. The eciency of transfection was determined by dot blot hybridization on total cell extracts using a 32 P-labeled CAT DNA as probe. Temporal expression. RNA samples from sea urchin eggs and embryos at dierent developmental stages were analyzed by North-

the 50 region, a cis-acting element, termed modulator, was rst described by Grosschedi and Birnstiel [13,14]. By combining in vitro nuclear protein binding and transgene expression analysis in sea urchin embryos, we extended the mutation studies of Grosschedl et al. in frog oocytes [15] and found that a 30 bp sequence of the modulator is all what is required for the transcriptional activation of the early H2A promoter [10]. The H2A modulator is the only cis-acting element of the early histone repeating unit for which an enhancer activity has unambiguously demonstrated [11,16, and unpublished observations]. Interestingly, the DNA binding and enhancer activities of the modulator persist following the repression of the early histone genes at gastrula stage [10,11], suggesting that silencing of the H2A gene is not due to the inhibition of binding of the modulator specic factor. To better clarify the role of the H2A enhancer in the regulation of early histone gene expression, we cloned the gene encoding for the modulator binding factor that we denoted as MBF-1. In this paper, we describe the identication of MBF-1 and the characterization of its DNA binding and activation domains.

Materials and methods

List of oligonucleotides. Only the sense strand is reported. M30 wt: GATCAATCGCCAACAGAGGGAGCTATTCCC M30 Mut: GATCATAATAACCAACATAGGGAGCTATTCCCTA Mg25: GGATGAAAACACAGACCAGCCT Mg26: CTGGTAGACGATGTTATCCCC Mg20: ACCGGTCGCCACCATGGATAACATCGTCTACCAGA Mg21: GGAACCATCTGCTTCCTGTA Mg22: CGTGCCATCACCAGCACC Screening of a cDNA library with labeled DNA. To clone the MBF-1 transcription factor we slightly modied published procedures [17,18]. Double stranded M30 oligonucleotides, corresponding to the MBF-1 binding site were end labeled with [32 P]ATP and polynucleotide kinase and ligated at 19 C for 14 h with T4 ligase. The ligation products were run on polyacrylamide gel and a band containing seven tandem copies of the modulator was eluted and used as a probe. Escherichia coli SURE cells were infected with 2:4 105 pfu of an unamplied cDNA library from 64-cell stage embryos, constructed in k ZAP vector (Stratagene), and plated in four 25 25 cm squared NYZ agar plates. Plates were incubated at 39 C, till tiny plaques were visible (45 h). Culture plates were overlayed with 0.45 lm nitrocellulose lters, previously soaked for 30 min in 10 mM IPTG and air dried, and incubated overnight at 39 C. After cooling at 4 C, lters were removed from culture plates, air dried, and immersed, protein side up, in 50 ml of 6 M guanidine hydrochloride in binding buer containing [25 mM NaCl, 5 mM MgCl2 , 25 mM Hepes (pH 7.9), 0.5 mM dithiothreitol, 12% glycerol]. Filters were incubated at 4 C with gentle agitation. Proteins were renatured by soaking lters in binding buer containing 3 M guanidine hydrochloride at 4 C for 50 with agitation. This step was repeated fourfold, reducing at each step by half the concentration of the guanidine hydrochloride. Two additional incubations were then carried with binding buer. A blocking step was performed by immersing lters in 50 ml of binding buer supplemented with 10 lg/ml of salmon sperm DNA and 0.5% nonfat dry milk. After gentle

C. Alessandro et al. / Biochemical and Biophysical Research Communications 295 (2002) 519525 ern blot hybridization and RNase protection as previously described [22].

521

Results Cloning of MBF-1 cDNA by screening of expression library with multiple modulator target sites EMSA in nuclear extracts previously reported, strongly suggested that MBF-1 does not require formation of heterodimers for the sequence specic DNA binding [10]. Hence, as cloning strategy we screened a nonamplied k ZAPcDNA expression library from early blastula embryos, with labeled modulator array containing seven MBF-1 binding sites. This strategy of screening depends on the expression in E. coli of high level of the DNA binding domain, as b-galactosidase fusion protein, and a strong interaction between this domain and its recognition site. Several positive clones were obtained; four were isolated and characterized. The nucleotide sequence of such clones resulted identical. Interestingly, the leader region of all four cDNAs contains stop codons in the same frame of the lacZ, suggesting that initiation of protein synthesis in E. coli occurred from the eukaryotic ATG (not shown). The longest open reading frame predicts a protein of 939 amino acids (Fig. 1A). Database searching with the deduced protein sequence revealed a putative DNA

binding domain containing nine Krppel like Zn nger u motives (Fig. 1B). We observed that similarity with other DNA binding proteins is restricted to the Zn nger domain only, being the highest (around 51%) with the human CTCF protein (not shown). All nine Zn ngers plus an adjacent region are needed for the specic binding of MBF-1 to the modulator site To ascertain whether or not the protein encoded by the cloned cDNA corresponds to the endogenous MBF1 factor, we assessed the specicity of binding to the H2A modulator sequences. To this purpose, we transcribed in vitro full length and ends deleted cDNAs. The resulting RNAs were translated in a rabbit reticulocytes lysate system. Analysis of the labeled translation products by sodium dodecyl sulfate gel electrophoresis (SDSPAGE) revealed a predominant band and several less intense bands representing most probably translational arrested forms and/or proteins translated from internal in frame methionine (Fig. 2A). Fig. 2B shows the EMSA analysis carried out with the modulator binding site. As indicated by arrowheads a strong endogenous binding activity is present in the reticulocyte lysate. In fact, a DNAprotein complex formed also with the truncated proteins lacking the DNA-binding domain (lanes 13 and 14) and the lysate incubated without adding exogenous RNA (lanes 9 and 10). Two lines of evidence suggest that the in vitro translated protein corresponds to the MBF-1 factor. The in vitro translated proteins (lane 1) and nuclear extracts from gastrula embryos (lane 4) generated a predominant DNAprotein complex with the labeled wild type modulator oligonucleotide. In contrast, a mutant modulator oligonucleotide failed to interact with the proteins from both sources (lanes 2 and 5). Identical result was obtained when the translation products were challenged with a protein binding site from the H3 promoter (lane 3). Likewise, preincubation of the in vitro translated proteins with 100-fold excess of unlabeled wild type probe strongly interfered with the assembly of the labeled complex (lanes 6 and 7). As expected, 100-fold excess of the mutant modulator had no eect on DNA protein interaction (lane 8). Based on these results we conclude that the MBF-1 protein encoded by the cloned cDNA and a factor from nuclear extracts bind to the modulator with similar anity. The sequence analysis showed that MBF-1 contains nine Zn ngers comprised between residues 421 and 735 (see sequence of Fig. 1). To assess whether this protein region corresponds to a DNA binding domain, we performed EMSA analysis with the in vitro translated full length and truncated protein forms (Fig. 2A). Proteins from constructs A, B, and E having an intact Zn nger region generated, as expected, proteinDNA complex with the modulator binding site (lanes 11, 12,

Fig. 1. (A) Deduced aminoacid sequence of MBF-1. The Zn ngers are underlined. (B) Alignment of the nine Zn ngers.

522

C. Alessandro et al. / Biochemical and Biophysical Research Communications 295 (2002) 519525

(lane 16). This result strongly suggests that the DNA binding domain of MBF-1 extends beyond the Zn ngers. The additional amino acids are probably comprised between residues 725 and 793, because residues 1256 and 793939 are dispensable for binding (lanes 12, 15, 19) and mutant D lacks also these two regions. Expression of MBF-1 in human cells transactivate a reporter gene driven by an array of modulator target sequences Next, we addressed the capability of MBF-1 to transactivate the expression of a transgene driven by the modulator sequences. We performed the assay in human cells because the function of the modulatorenhancer is highly conserved among the sea urchin species. U2-OS cells were transfected with a reporter construct containing the EGFP gene, encoding for the enhanced version of the green uorescent protein, driven by a modulator enhancer cassette and a tk basal promoter. The modulator enhancer cassette, constituted by a tandem array of four intact and two deleted copies of the modulator sequence arranged in divergent orientation, has been widely used to express transgene in sea urchin embryos [11]. The EGFP reporter vector contains also the neomycin expression cassette as a selectable marker. Cells were grown under G418 for two weeks and observed under an epiuorescence microscope to monitor EGFP expression. A very low green uorescence was detected (not shown), probably indicating basal expression of the EGFP transgene. One cell clone containing stable integrated EGFP was transfected with the full-length MBF-1 and truncated forms of the MBF-1 cDNA (depicted in Fig. 2A) placed under the control of the CMV enhancer-promoter region. Expression of EGFP reporter occurred after transfection of the transgenic cell line with the constructs expressing the full-length MBF-1 (Fig. 3A) and the C-terminal deleted mutant construct B which contains the DNA binding domain (Fig. 3B). Removal of almost all the DNA binding domain resulted, as predicted, in the absence of transgene expression (Fig. 3C). Fluorescence was also not detected in cells transfected with mutants D and E (data not shown). The two proteins present the same deletion of 251 Nterminal amino acids but dier for the C-terminus. The protein from construct D did not form complex with the modulator in vitro (Fig. 2B) because it lacks the Cterminal amino acids (57 residues) of the DNA binding region. The second mutant has an intact DNA binding domain but was also trimmed from the N-terminal portion of MBF-1. To conclude, these data conrmed the identication of MBF-1 as the modulator/enhancer binding protein. They also validated the results of the EMSA analysis, which suggested that the aminoacid sequences downstream the Zn ngers are absolutely

Fig. 2. Specicity of binding to the modulator sequence of the in vitro synthesized MBF-1 and its deletion mutants. (A) Schematic representation of the proteins synthesized in a cell free system by the in vitro transcribed RNA from wild type and deletion mutants. Number of residues deduced for each protein is shown. The Zn ngers (Zn) and the basic (B) region that constitute the DNA binding domain and the activation domain (AD) are indicated. The electrophoretic migration in SDSPAGE of the [35 S]methionine labeled translation products is shown to the left. (B) Five ll of the unlabeled translation reaction mixtures (IVT) or 3 lg of nuclear extracts (NucExt) from gastrula embryos was used in EMSA with a 32 P-labeled 30 bp wild type (WT) and mutant (MUT) modulator sequences. The sequences of the two oligonucleotides are presented below the uorograph. The two probes were obtained by annealing complementary oligonucleotides. In competition experiments, nuclear extracts or translation products were pre-incubated with 100 ng of unlabeled WT or MUT modulator oligos prior to the addition of 1 ng of the labeled probe. Only the gel regions containing the nucleoprotein complexes are shown for lanes 1017. Letters above lanes refer to the mutant constructs drawn in (A). Mock indicates the complexes obtained with the cell free extracts incubated with the luciferase mRNA or without the addition of exogenous RNAs. Arrowhead points to the unspecic complex. The specic complex is indicated by an arrow. The double arrow refers to the migration of the free probe.

15, and 17). On the contrary, construct C in which all, but one, Zn ngers were removed failed to bind (lanes 13 and 14). It is of much interest to mention the results obtained with the constructs D that helped to identify the protein region involved in DNA binding. The truncated protein that was deleted from the C-terminal region up to the ninth nger and from the N-terminus up to residue 251, did not interact with the modulator

C. Alessandro et al. / Biochemical and Biophysical Research Communications 295 (2002) 519525

523

Fig. 3. MBF-1 transactivates a promoter through the modulator array in human cells. A stable human U2-OS cell line containing an EGFP integrant driven by the modulator enhancer cassette was transfected with plasmids expressing, from a CMV promoter, full-length (A) or Cterminal truncated forms (B and C) of MBF-1. The MBF-1 proteins are schematically drawn below the photographs. AD and BD stand for the activation and DNA binding domains, respectively.

Fig. 4. Transactivating function of the N-terminal region. Human U2OS cells were co-transfected with the reporter plasmid GAL4-CAT and the GAL4-MBF expression plasmid pSG424-MBF, or the pSG424 as control.

crucial for the specic interaction to the modulator site. In addition, they revealed that the protein region responsible for the transactivation of a transgene driven by the modulator sequences in human cells must be located upstream the DNA binding region. The N-terminal sequences of MBF-1 contain a transactivator activity To obtain direct evidence for the presence of a transactivation function within the N-terminal portion of MBF-1 we made an in frame fusion between the rst 256 amino acids of MBF-1 and the DNA binding domain (BD) of the yeast transcription factor GAL4. The GAL4-MBF-1 chimeric gene was placed under the control of the CMV enhancer -promoter and co-transfected in U2-OS or H1299 human cells with a CAT reporter gene driven by multiple copies of GAL4 binding sites. As control, the CAT reporter gene was also cotransfected with either bluescript plasmid or vector expressing only the GAL4-BD. As observed in a number of transient assays we obtained 15 to 20% transacetilation of chloranphenicol due, most probably, to the basal activity of the tk promoter in these cell types. Nonetheless, co-trasfection of the GAL4-MBF-1 expression plasmid always resulted in more than 95% conversion of the labeled chloranphenicol in the acetylated forms (Fig. 4). These results clearly demonstrate the capability of Nterminal region of MBF-1 fused to a heterologous DNA binding domain to enhance the expression of a reporter gene in human cells.

The MBF-1 gene is expressed in oocytes and in embryos at early and late developmental stages The enhancer and DNA binding activities in nuclear extracts of the modulator persist also in embryos at gastrula stage, i.e., in the absence of expression of the aH2A gene [10,11]. Because of these evidences we anticipated the expression of the MBF-1 gene to occur throughout embryogenesis. Indeed, a northern blot hybridization (Fig. 5A) produces a single RNA band of about 4.5 kb, compatible with the length of the cDNA clones, that appeared in unfertilized egg and in all developmental stages analyzed. The more quantitative RNase protection assay (Fig. 5B) conrmed this expression pattern and suggested that the abundance of the MBF-1 transcripts does not change a great deal, going from the unfertilized eggs up to embryos at pluteus stage (40 h post-fertilization). We conclude, that expression of MBF-1 gene is not dependent of the transcriptional state of the histone genes.

Discussion In this study we have utilized a multimerized cis-acting element for screening an expression cDNA library to clone a transcriptional activator. This screening strategy, derived from that developed for the isolation of genes by using antibody probes, it is well suited for cloning genes encoding for DNA binding proteins that do not hetero-

524

C. Alessandro et al. / Biochemical and Biophysical Research Communications 295 (2002) 519525

Fig. 5. Expression prole of MBF-1 during embryogenesis. (A) Northern blot of total RNA from eggs and dierent embryonic stages hybridized with a 32 P-labeled MBF-1 cDNA probe. The agarose gel stained with ethidium bromide is shown below the uorograph. (B) RNAse protection assay was carried out with 5 lg of total RNA, from the indicated developmental stages, or yeast tRNA as a negative control. The 32 P-labeled antisense RNA was transcribed in vitro from a MBF-1 subclone. RNAse digestion products were resolved by denaturing polyacrylamide gel together with end 32 P-labeled HpaII digested pBluescript DNA (M). Arrow indicates the expected 125 nucleotides RNase resistant fragment.

dimerize for their specic interaction with the target sequences. The modulator binding factor, MBF-1, of the 30 bp modulator/enhancer of the sea urchin a-H2A histone gene seems to belong to this category of regulators. MBF-1 contains nine Zn ngers of the Krppel class u of transcription factors and because data bank search failed to reveal any similarity with known transcription factors, it appears to be a novel enhancer binding protein. As indicated by the results of the EMSA analysis (Fig. 2B) the DNA binding domain extends beyond the Zn ngers. Although we cannot exclude that the additional amino acids include residues located upstream to the rst Zn nger, our evidence suggest that they are most probably located within the 60 amino acids adjacent to the ninth nger. Which of these are necessary for binding remains to be established. Because in the N-terminal one third of this region contains nine basic residues (lysine and arginine), it is reasonable to hypothesize that the DNA binding domain of MBF-1 consists of nine Zn ngers plus an adjacent basic region. The EMSA analysis reported in this paper with both the in vitro translation products and nuclear extracts, conrms and extends the observations of Birnstiel and coworkers on the eect of mutations on modulator function in Xenopus oocytes [15]. We showed in fact, that three nucleotide substitutions of the modulator sequence abolished the binding of MBF-1. Of particular signicance is the G to T transversion in the AACAGA sequence, in that this sequence motif is conserved in the H2A modulator of the closely related sea urchin Psammechinus miliaris [15]. In addition, as previously shown, a subset of the modulator sequence, dened as

USE 1, containing the AACAGA motif maintains the capability to bind the MBF-1 protein both in vitro and in vivo [16]. Altogether these evidence would imply the AACAGA sequence as the core sequence of the MBF-1 transcription factor binding site. We took advantage of the observation that, in contrast to Xenopus oocyte [15,16] a MBF-1 related factor is not present in U2-OS cells, to demonstrate that expression of the sea urchin MBF-1 in human cells activated transcription of a transgene driven by an array of modulator sequences (Fig. 3). Because the reporter gene was stable integrated into host chromatin, this evidence would suggest that the sea urchin MBF-1 protein has the capability to interact with the human co-activator and chromatin remodeling complexes. The N-terminal 256 amino acids mediate the activating function. This region is enriched in glutamine, glutamic acid, serine, and threonine residues that are also found in the activation domain of several transcription factors. MBF-1 does not seem to be a stage specic transactivator. In fact, when the a-H2A gene is silenced as the P. lividus embryo enters gastrulation, expression of a transgene driven by the modulator/enhancer and persistent MBF-1 binding with the key modulator sequence can be demonstrated [10,11,16]. Consistently, as shown here, MBF-1 expression occurs at almost constant level throughout embryogenesis. Hence, the general belief that enhancers are responsible for regulation of genes during development does not seem extendable to the aH2A gene of the sea urchin P. lividus. It would be interesting to know which genes are transactivated by MBF-1, beside the a-H2A histone gene, and whether they are regulated by a modulator like sequence element containing the AACAGA motif described above.

Acknowledgments
This work was supported by grants from the University of Palermo (ex 60%), MIUR (Programmi di Ricerca Scientica di Interesse Nazionale), and AIRC (Associazione Italiana Ricerca sul Cancro). Many thanks are due to A. Di Leonardo for the use of the tissue culture facilities.

References
[1] C.C. Hentschel, M.L. Birnstiel, The organization and expression of histone gene families, Cell 25 (1981) 301313. [2] G. Spinelli, F. Gianguzza, C. Casano, P. Acierno, J. Burckhardt, Evidence of two dierent sets of histone genes active during embryogenesis of the sea urchin Paracentrotus lividus, Nucl. Acids Res. 6 (1979) 545560. [3] G. Childs, R. Maxson, L.H. Kedes, Histone gene expression during sea urchin embryogenesis: isolation and characterization of early and late messenger RNAs of Strongylocentrotus purpuratus by gene-specic hybridization and template activity, Dev. Biol. 73 (1979) 153173.

C. Alessandro et al. / Biochemical and Biophysical Research Communications 295 (2002) 519525 [4] G. Spinelli, I. Albanese, L. Anello, M. Ciaccio, I. Di Liegro, Chromatin structure of histone genes in sea urchin sperms and embryos, Nucl. Acids Res. 10 (1982) 79777991. [5] P.N. Bryan, J. Olah, M.L. Birnstiel, Major changes in the 50 and 30 chromatin structure of sea urchin histone genes accompany their activation and inactivation in development, Cell 33 (1983) 843848. [6] L. Anello, I. Albanese, C. Casano, F. Palla, F. Gianguzza, M.G. Di Bernardo, R. Di Marzo, G. Spinelli, Dierent micrococcal nuclease cleavage patterns characterize transcriptionally active and inactive sea-urchin histone genes, Eur. J. Biochem. 156 (1986) 367374. [7] M. DiLiberto, Z.C. Lai, H. Fei, G. Childs, Developmental control of promoter-specic factors responsible for the embryonic activation and inactivation of the sea urchin early histone H3 gene, Genes Dev. 3 (1989) 973985. [8] J. Bell, B.R. Char, R. Maxson, An octamer element is required for the expression of the alpha H2B histone gene during the early development of the sea urchin, Dev. Biol. 150 (1992) 363371. [9] H. Fei, G. Childs, Temporal embryonic expression of the sea urchin early H1 gene is controlled by sequences immediately upstream and downstream of the TATA element, Dev. Biol. 155 (1993) 383395. [10] F. Palla, R. Mel, L. Di Gaetano, C. Bonura, L. Anello, C. Alessandro, G. Spinelli, Regulation of the sea urchin early H2A histone gene expression depends on the modulator element and on sequences located near the 30 end, Biol. Chem. 380 (1999) 159165. [11] F. Palla, R. Mel, L. Anello, M. Di Bernardo, G. Spinelli, Enhancer blocking activity located near the 30 end of the sea urchin early H2A histone gene, Proc. Natl. Acad. Sci. USA 94 (1997) 22722277. [12] R. Mel, F. Palla, P. Di Simone, C. Alessandro, L. Cali, L. Anello, G. Spinelli, Functional characterization of the enhancer blocking element of the sea urchin early histone gene cluster reveals insulator properties and three essential cis-acting sequences, J. Mol. Biol. 304 (2000) 753763. [13] R. Grosschedl, M.L. Birnstiel, Identication of regulatory sequences in the prelude sequences of an H2A histone gene by the

525

[14]

[15]

[16]

[17]

[18]

[19] [20]

[21]

[22]

study of specic deletion mutants in vivo, Proc. Natl. Acad. Sci. U S A 77 (1980a) 14321436. R. Grosschedl, M.L. Birnstiel, Spacer DNA sequences upstream of the TATAAATA sequence are essential for promotion of H2A histone gene transcription in vivo, Proc. Natl. Acad. Sci. USA 77 (1980b) 71027106. R. Grosschedl, M. Machler, U. Rohrer, M.L. Birnstiel, A functional component of the sea urchin H2A gene modulator contains an extended sequence homology to a viral enhancer, Nucl. Acids Res. 11 (1983) 81238136. F. Palla, C. Bonura, L. Anello, L. Di Gaetano, G. Spinelli, Modulator factor-binding sequence of the sea urchin early histone H2A promoter acts as an enhancer element, Proc. Natl. Acad. Sci. USA 91 (1994) 1232212326. H. Singh, J.H. LeBowitz, A.S. Baldwin Jr., P.A. Sharp, Molecular cloning of an enhancer binding protein: isolation by screening of an expression library with a recognition site DNA, Cell 52 (1988) 415423. C.R. Vinson, K.L. LaMarco, F.P. Johnson, W.H. Landschulz, S.L. McKnight, In situ detection of sequence-specic DNA binding activity specied by a recombinant bacteriophage, Genes Dev. 2 (1988) 801806. I. Sodwski, M. Ptshane, A vector for expressing GAL 4 (1147) fusions in mammalian cells, Nucl. Acids Res. 17 (1989) 7539. F. Palla, C. Bonura, L. Anello, C. Casano, M. Ciaccio, G. Spinelli, Sea urchin early histone H2A modulator binding factor 1 is a positive transcription factor also for the early histone H3 gene, Proc. Natl. Acad. Sci. USA 90 (1993) 68546858. P. Di Simone, A. Di Leonardo, G. Costanzo, R. Mel, G. Spinelli, The sea urchin sns insulator element blocks CMV enhancer following integration in human cells, Biochem. Biophys. Res. Commun. 284 (2001) 987992. M. Di Bernardo, R. Russo, P. Oliveri, R. Mel, G. Spinelli, Homeobox-containing gene transiently expressed in a spatially restricted pattern in the early sea urchin embryo, Proc. Natl. Acad. Sci. USA 92 (1995) 81808184.