Spatial regulation of SpMTA metallothionein gene expression in sea urchin embryos by a regulatory cassette in intron 1

Mechanisms of Development 50 (1995) 131-137

Spatial regulation of ?$GK’I4 metallothionein gene expression in sea urchin embryos by a regulatory cassette in intron 1 Martin Nemer*, Elizabeth W. Stuebing,

Guang

Bail,

Henry

R. Parker2

The Institute for Cancer Research, Fox Chase Cancer Center, Philadelphia, PA 19111, IJSA

Received 10 September 1994: accepted I November 1994

Abstract The SpMTA metallothionein (MT) gene of the sea urchin Strongylocentrotuspurpuratus is restricted in its expression to the aboral ectoderm in gastrulae and pluteus larvae. The proximal 1.6 kb of the 5 ‘-flanking region together with the 1. 12-kb first intron of the SpMTA gene are sufficient for its correct cell-type specific expression in transgenic embryos. This restricted spatial expression is largely eliminated by deletion of an interior 405bp region in the intron. Within this region is a 295bp, genomically repetitive, transposon-like segment (Nemer et al., 1993), containing several sequence motifs highly homologous to posited regulatory elements in the promoters of other genes (Thiebaud et al., 1990). The P3A and P5 sites in this apparent regulatory cassette were shown

through competition to bind with relatively high affinities the same nuclear factors, bound by their counterpart sites in the CyIIIa actin promoter. Keywords:

Metallothionein;

Spatial regulation;

Sea urchin embryo

1. Introduction

The SpMTA gene of S. purpuratus (Harlow et al., 1989) is unusual for a metallothionein (MT) gene in its cell-type specificity (Nemer et al., 1991). Its restricted expression to the embryonic aboral ectoderm (Nemer et al. 1991) distinguishes SpMTA from five other MT genes in this species, including a SpMTB class, encoding a different isoform. It is similarly distinguished from all of the MT genes in another sea urchin, Lytechinus pictus, which, apparently, are ubiquitously expressed (Cserjesi et al., 1992). SpMTA, then, belongs to a cell-type specific gene set which includes the calcium-binding Spec proteins, Spec I (Bruskin et al., 1981; Lynn et al., 1983), Spec 2a, 2c, and 2d (Hardin et al., 1988), CyZZZa and CyZZZbactin (Cox et al., 1986) and arylsulfatase (Yang et al., 1989; Akasaka et al., 1991). Critical promo* Correspondingauthor. ’ Present address: Gerontology Research Center, National Institute on Aging, 4940 Eastern Ave., Baltimore, MD 21224. * Present address: Cross Cancer Institute, Edmonton, Canada. 0925-4773/95/$09.50 0 1995 Elsevier Science Ireland SSDI 0925-4773(94)-00330-P

Alberta.

ter regions in the CyZZZaactin and Spec genes have been identified as being required for quantitative (Franks et al., 1990; Gan et al., 1990a) and spatial (Davidson, 1989; Gan et al., 1990b) regulation. The SpMTA gene has been demonstrated to be functionally bipartite, insofar as its 5 ‘-flank promotes metal-regulated transcription, while its first intron serves to enhance this activity (Bai et al., 1993). The role of intron 1 in the quantitative regulation of SpMTA gene expression (Bai et al., 1993) has led us to test whether it also may be influential in spatial regulations. Previously described sequences of the SpMTA and SpMTB, genes (Harlow et al., 1989; Nemer et al., 1991) revealed that intron 1 of SpMTA contains a 295-bp insert (Fig. 1) not present in SpMTB, (Nemer et al., 1991). Moreover, this insert is transposon-like by virtue of being highly repetitive in the genome and bounded by 25-bp repeats (Nemer et al., 1993). As a posited mobile DNA element, it would seem to be unusual for its content of potential regulatory elements: It is a cassette of sequence motifs, homologous to protein-binding sites, PS, P7ZZ and P3A, in S. purpuratus promoters, and indicated in the CyZZZaactin gene promoter to be regula-

Ltd. All rights reserved

hf. Nemer et al. /Mechanisms of Development SO (1995) 131-137

132

e

f

\/ 1

I

200bp

P5(+) P711

PW+)

P3A

’ P5(-)

P5(-)

Fig. I. A schematic of a portion of the SpMTA gene including I .6 kb of 5 ‘-flanking region with its T’ATA box and intron I between exons (ex.) I and 2. Also shown are metal response elements (MREs) a-f in the promoter (Harlow et al., 1989; Bai et al., 1993), a single MRE (i) in intron I and the AvrII sites. The cassette is shown as a bold line and, also below, with locations of sequences homologous to the putative regulatory elements P7II and P3A, as well as oriented P5 sites, including those in the bracketing 2%bp inverted repeats (arrows). The P3A and all four PS sites match the consensus of potentially important sequences, stipulated by Thiebaud et al. (lb90). whereas P71ideparts by one residue from their 9-bp consensus.

tory elements (Thiebaud et al., 1990). However, the same species does contain other, similarly short, interspersed, repetitious DNA elements with demonstrable gene regulatory or protein-binding sites; e.g., the RSR element among the Spec genes (Gan et al., 1990a) and the PlO repeat in the CyZZZaactin promoter (Anderson et al., 1994). Since the SpMTA cassette may thus belong to a class of potentially influential, repetitious DNA elements bearing regulatory sites, the factor-binding capacities of these sites were analyzed. The P3A site has been shown to be bound at a YXYGCGCW core sequence by two different, cloned proteins (Calzone et al., 1991; Hoog et al., 1991). One of these proteins, P3A2, may be of broad significance, insofar as highly identical counterparts have been shown to be encoded by the developmentally significant, Drosophila erect wing (DeSimone et al., 1993) and the human NW-1 (Virbasius et al., 1993) genes. From an extended comparison among several genes, we derived the consensus TTCATKSTNKYNTNW sequence (Nemer et al., 1993) for the P.5 site, in which TTCATI’G had previously been designated as potentially important (Thiebaud et al., 1990). The SpMTA cassette contains one P3A site and four apparent P5 sites, two of which are embedded in the 25bp inverted repeats at its termini. We have found the P3A and the terminal P5 sequences to be nuclear factor-binding sites, thereby supporting the proposal that the cassette has a regulatory function. 2. Results 2.1. Regulation of the transgenically expressed SpMTA gene Expression of the endogenous SpMTA gene is localized in the aboral ectoderm in both metal-induced and control embryos, indicating that spatial control of this

gene is not overridden by metal induction and, indeed, could be linked to it (Nemer et al., 1991). Therefore, the spatial expression of chimeric constructs was tested in Cd@)-induced embryos. Zygotes were injected with the chimeric plasmid, 1.6-kb promoter-intron-1-LucZ, consisting of 1.6 kb of the 5 ‘-flanking region as well as intron 1 of SpMTA and the reporter LacZ gene, which has a nuclear targeting signal from SV40 T-antigen (Gan et al., 1990b). &galactosidase expression was detected by X-gal staining in the nuclei of whole mount gastrulae and pluteus larvae. Fig. 2 shows a pluteus (panel a) and late gastrula (panel c), in which the focal plane is at the surface, to display signal in the nuclei of ectodermal cells, and an interior plane to bring the gut into focus (panel b). Expression of this construct was largely restricted to the aboral ectoderm in the transgenic embryos displaying signal. Enrichment of expression in this tissue was calculated to be 87%, based on a correction (below) for random distribution (Table 1). For other genes, expressed specifically in the aboral ectoderm, correctly restricted expression was obtained solely with 5’flanking regions, 2.3 kb of the CyZZZu actin gene (Hot&-Evans et al., 1990) and 1.5 kb of the Spec2a gene (Gan et al., 1990b). Bai et al. (1993) observed previously that the level of expression of transgenes driven by the SpMTA gene fragment used here was onetenth to one-fifth the levels attained by promoters of CyIIIa actin and Spec transgenes. This markedly weaker strength of the SpMTA promoter-intron combination has apparently resulted here in a proportionally reduced fraction of late-stage embryos displaying signal, relative to the lo-fold greater fraction observed for the stronger CyIIIa actin and Spec2a promoters. Transgenic embryos expressing the construct 1.6-kb promoter-intron- 1A,,,,,,&cZ, in which a 405-bp A vrII segment, containing the 295-bp cassette, was deleted

M. Nemer et al. /Mechanisms of Development 50 (1995)

131-137

133

Fig. 2. Spatial expression of the LucZ gene driven by the 1.6kb SpMTA promoter and its first intron intact (left panels) or deleted by AvrII of the cassette region (right panels). Whole mount embryos, expressing the @-galactosidase reporter in nuclei, have been photographed with focus on the epithelial surface (a, c and d) or interiorly on the embryonic gut (b, e and f). Each spoke in the cross bar is 20 m, except for the higher magnification of panel f where the spoke length is 10 m.

134

M. Nemer et al. /Mechanisms of Development SO (1995) 131-137

Table 1 Spatial expression regulated by the SpMTA promoter-introit-1 fragment, and the effect of deleting the cassette from intron 1’ SpMTA gene fragment fused to lacZ

1.6-kbpromoter-intron-I 1.6-kb promoter-intron-l&,,,,

Site of expression (number of embryos) Exclusively in aboral ectoderm

Including remainder of embryo

Enrichment in aboral ectoderm

57 36

6 43

87% 28%

sConstructs were microinjected into zygotes and assayed by X-gal staining at 72 h of development. Deletion (&,,,,) was by removal of a 405-bp AvrII fragment from the 1120-bp intron I. The 142 embryos displaying signal were from about 5000 microinjected and surviving, late-stage embryos, 70% as late gastrula to prism stages and 30% as pluteus larvae. Enrichment in aboral ectoderm is the percentage beyond that expected as a result of random distribution, as described in text.

from the SpMTA intron 1 (Fig. 1) are shown in the right-hand panels of Fig. 2. The pluteus, in panel d, displays signal in the oral ectoderm, while the prism, shown in panel e and by enlargement in panel f, displays signal in the gut wall nuclei. If expression were unrestricted, the pattern would depend on the timing of the random and unequal integration of transgenes in founder cells of the different lineages (Hough-Evans et al., 1988). Considering that the founder cells of the aboral ectodermal lineage constitute 25% of the early cleaving embryo (Cameron et al., 1987), we have estimated that 28% of the embryos expressing the deleted construct have signal restricted to the aboral ectoderm, beyond simple random distribution. This degree of ectopic expression, entailing a comparably reduced percentage of aboral ectodermal expression, has been observed for other transgenes constructed from aboral ectoderm-specific genes: e.g., where the 5.6 kb, 5’4ank of the Spec 1 gene was used (Gan et al., 1990b) and where the CyIIIa actin promoter was subjected to competition with fragments containing suspected regulatory sites (Ho@-Evans, 1990). We can conclude that the spatial expression of the SpMTA gene has been altered sufftciently to implicate the deleted cassette in its regulation. 2.2. Sequence motifs in the cassette are sites bound by the same nuclear factors that bind homologous regulatory elements in the CyrIIa actin promoter The nuclear factor-binding properties of the P3A and P5 elements in the SpMTA intronic cassette were compared with their respective, homologous counterparts in the CyMcr promoter. Synthetic oligodeoxynucleotides were generated (Experimental procedures) and used in EMSA as 26-33-bp double-stranded probes, representing the regulatory elements and their adjacent regions. The P3A probes, divergent in the two genes outside of their core regions, displayed the same band shifts and competed with each other (Fig. 3). A predominant band was accompanied by two smaller, minor bands in each case. Radioactivity in the major band was measured to quantify the relative binding affinities of these probes.

P3AMTA

P3AcyIII

Fig. 3. EMSA of P3A DNA fragments representing sites in the SpMTA cassette and the CyifIa promoter. Biastula-stage nuclear ex-

tracts (5 cg protein) incubated with 0.02 pmol probe and other constituents 15 min at 15°C. Left panels: SpMTA P3A 27-bp probe only with nuclear extract (lane I), and with probe plus unlabeled 33-bp CyIIIa P3A fragment at 6, 38 and 112 pmol (lanes 2-4) or unlabeled 27-bp SpMTA P3A fragment at 38 and 112 pmol (lanes 5 and 6). Right panels: CyIIIa P3A 33-bp probe only with nuclear extract (lane I) or with probe plus 38 and 112 pmol unlabeled 33-bp CylIla P3A fragment (lanes 2 and 3) or 27-bp SpMTA P3A fragment (lanes 4 and 5). Radioactivity in the major band was measured directly, to give binding affinities of the SpMTA probe relative to the CyIIla probe (F values) of 1.82,1.19, 1.45and 1.43 in separate gels.

M. Nemer et al. /Mechanisms of Development SO (I99S)

X31-137

135

of deletion on quantitative (Bai et al., 1993) and spatial (above) expression, suggest further that the specified region in the SpMTA intron-1 region constitutes a regufatory cassette. 3. Discussion 3.1. Spatial regulation of the SpMTA gene by a region in its first intron

P5cyIII

P5MTA

Fig. 4. EMSA of PS DNA fragments representing sites in the SpMTA casette and the CyIIIa promoter. Conditions were the same as Fig. 3. Left panel: Cyllla P5 28-bp probe in lanes 1 and 2 are, respectively, without and with nuclear extract; lanes 3-5 have additions of 30, 100 and 300 pmol of unlabeled CyflIa P5 and lanes 6-8 have similar amounts of unlabeled SpMTA P5’. Right panels: SpMTA P5 26.bp probe with no additions in lane 1 and with IO, 30, 100 and 300 pmol. respectively, of SpMTA P5 (lanes 2-5) and CyIlfa P5 (lanes 6-9). Radioactivity in the major (lower) band was measured directly to give binding affinities of the SpMTA probe relative to the Cyiiia probes (F values) of 1.28, 0.90 and 1.33 in separate gels.

The relative affinities of homologous sites in other genes have been calculated with the CyZZZusites as reference and termed the F value (Thiebaud et al., 1990). An F value of 1.47 for the SpMTA P3A site, calculated by their method of ratios and based on four determinations (Table l), is indicative of a relatively high binding affinity among the several previously tested: Of 12 sites examined in the Spec 1, SM50, CyZa, CyZb, CyZc (Thiebaud et al., 1990) and SpMTA genes, this SpMTA site and two others have affmities higher than that of the CyZZZasite. The probe representing the P5 elements within the terminal inverted repeats of the SpMTA cassette shifted to give a predominant lower and a minor upper band (Fig. 4, right hand panel). The counterpart CyZZZoP5 probe shifted to the same position as the lower of these bands but displayed only a trace of the upper band. The upper band was observed to be variable for both P5 probes and tended to be eliminated by nonspecific competition. Radioactivity in the lower bands was measured in three trials and an F value of 1.17 was calculated for the SpMTA relative to the CyZZZuP 5 element: Of 11 sites examined in several genes (above), the SpMTA P5 site and two others have afftnities higher than the CyZIra site. These binding properties, together with the effects

Besides the strictly quantitative function of regulatory elements in introns, involving intronic sites for enhancers, transcriptional arrest and response to physiological signals (see Bai et al. 1993), intronic elements have also been identified for spatial and temporal control of gene expression. These include spatialtemporal control elements in the engrailed gene (Kassis, 1990), as well as histospecitic control elements in several genes, operating through the enhancement of promoter activity (Gillies et al., 1983; Hoshijima et al., 1991; Rippe et al., 1989) or through alternative splicing (Wieczorek et al., 1988; Helfman et al., 1990; Forry-Schaudies and Hughes, 1991). We noted previously that SpMTA transgene expression in blastulae was regulated by a combination of intron-1 elements and promoter MREs (Bai et al. 1993). Since the aboral ectoderm-specific genes may not be spatially restricted in their expression until the gastrula stage (Angerer et al., 1987; Gagnon et al., 1992), we cannot determine whether quantitative regulation by intron-1 elements includes otherwise spatial-controlling elements. However, the activities of SpMTA transgenes with and without a critical intron-1 region clearly indicate that late-stage embryos utilize intron-l-localized spatial regulatory elements. The localization of such elements in the SpMTA intron contrasts with their presence in the 5 ‘-flanking region of CyZZZuactin (Hough-Evans et al., 1987) and with the localization of cis determinants in the 5’-flank for the spatial expression of both CyZZZuactin (Hough-Evans et al., 1990) and Spec 2a (Gan et al., 1990b). DNA fragments, containing nuclear protein-binding sites in the CyZZZaactin promoter, were co-administered with promoter constructs to determine whether the correct aboral ectodermal expression could be altered. The expression was altered in two ways: the level of activity increased (Franks et al., 1990) and the activity was no longer spatially restricted (Hough-Evans et al., 1990). From these results, it was deduced that two sites, P3A and P711, operate negatively to restrict spatial expression. The critical, deleted intronic region of SpMTA contains both of these sites together with sequence motifs homologous to the putatively positive P5 site (Franks et al., 1990). Although the CyZZZaactin and SpMTA genes may thereby seem to be similarly regulated, it cannot be excluded that the intronic localization may entail a different mode of operation for SpMTA. Especially important would be to distinguish between

136

h4. Nemer et al. /Mechanisms

enhancer and transcriptional arrest activities. These functional sites need to be further delineated, perhaps through their mutation in both of these genes, and the mechanisms involved need to be explored. Also, the present results, involving some residual degree of restricted expression by the deleted construct, do not exclude the involvement of other parts of the SpMTA gene in spatial regulation. In spite of these caveats, the cassette has, indeed, been shown through the consequences of its deletion and by virtue of its protein binding sites to have regulatory properties. We may usefully conclude from the present data and previous results (Bai et al., 1993), that the SpMTA gene is composed of two functionally distinct components: (i) a promoter whose heavy metal ion-dependent activity is sustained through the cooperative interaction of a series of MRE sites and (ii) an intron with regulatory elements that enhance as well as restrict expression to a single set of embryonic cell lineages. Moreover, SpMTA is presently unique among metallothionein genes and among echinoderm genes in its intronic regulation. 3.2. The cassette of regulatory elements The implication of the 295-bp cassette as the determinant for spatial regulation rests upon the assumption that it operates without the involvement of the adjacent 105-bp region, also in the AvrII-deleted fragment. We tested by EMSA for the presence of factor-binding sites in that region by the use of a restriction fragment containing 90 of its base pairs (data not shown). There was no clear band shift of this fragment as probe while, in contrast, an intronic 80-bp restriction fragment bearing the P3A site displayed the same band shift as the P3A oligonucleotide probe. The cassette as the sole locus of binding sites is, therefore, consistent with it being the sole regulatory component in the deleted fragment. We have reported that this SpMTA cassette is actually a truncated version of highly identical and repetitious 375 bp elements in the S. purpurutus genome (Nemer et al., 1993). The additional 80 bp in these larger cassettes do not have identifiable regulatory motifs. The high degree of identity (94”/) among several such cassette sequences from distinct genomic loci extends to the intervening as well as the implicated factor binding sites. Our demonstration of strong binding sites in the SpMTA cassette indicates that the other 375-bp cassettes with the same sites are also capable of binding several nuclear factors. Further inquiry might be directed toward determining whether any of these multiple cassettes have regulatory functions and whether such functions are necessarily similar to those exhibited by the SpMTA cassette. 4. Experimental procedures 4.1. Chimeric constructs

The S/-flanking and first intronic regions, and their

of Development 50 (1995) 131-137

derivatives, of the SpMTA gene (Bai et al., 1993) were linked to the reporter LucZ gene containing a nuclear localization signal (Gan et al., 1990b). Into the HindIIISaLI sites of thepNL plasmid (Gan et al., 1990b) was inserted a 1.9-kb HindIII-XhoI fragment, consisting of the proximal 0.7 kb of the 5’-flank and the 1.12 kb intron 1 of SpMTA (Bai et al., 1993), to obtain the intermediate construct 0.7-kb-promoter-intron-LucZ. The 0.7-kb HindIII-XhoI SpMTA region was replaced with a 1.6-kb HindIII-XhoI fragment containing a further upstream region to give 1.6-kb promoter-intron-I-LacZ. The SpMTA portion of this construct is shown in Fig. 1. A 405-bp AvrII segment, containing the 295-bp cassette, was removed from the intron-1 region of this construct to obtain the derivative, 1.6-kb promoter-intron1Acasse,,e_LucZ. Plasmid DNA was prepared and purified through the use of the anion exchange resin procedure of Qiagen (Dusseldorf). 4.2. Transgenic embryos We employed previously described procedures for microinjecting constructs into sea urchin zygotes (Bai et al., 1993). For injection of linearized plasmid constructs with LacZ reporter gene into S. purpuratus zygotes, the procedures of McMahon et al. (1985) were modified as described by Bai et al. (1993). After injection, the zygotes were cultured to the 72-h pluteus larval stage in sterile synthetic sea water, SSW (Nemer et al., 1984), containing penicillin at 50 mg/l and streptomycin at 50 U/l. After a further 7-h incubation in 500 mM cadmium acetate, the embryos were treated according to Gan et al. (1990b) and assayed for 8-galactosidase expression. Embryos were visualized under Hoffman optics for microphotography. 4.3. Electrophoretic mobility shifr sssays {EMSA) Nuclear extracts were prepared from 24-h mesenthyme blastulae according to the procedures of Calzone et al. (1988). Embryos stored at -70°C in TEESSD [lo mM Tris, pH 7.4, 1 mM EGTA, 1 mM EDTA, 1 mM spermidine-Tri-HCl, 1 mM dithiothreitol (DTT), 0.36 M sucrose] were lysed for the isolation of nuclei. Pelleted nuclei, suspended in HEESD (10 mM HEPES, pH 7.9, 1 mM EGTA, 1 mM EDTA, 1 mM spermidine, 1 mM DTT), were brought to 0.36 M ammonium sulfate and residual chromatin removed by centrifugation. Protein was precipitated from the supernantant fluid with ammonium sulfate at 0.3 g/ml. The protein precipitate was dissolved in, then dialyzed against, buffer C (20 mM HEPES, pH 7.9, 40 mM KCl, 0.1 mM EDTA, 1 mM DTT, 20% glycerol). This protein suspension in buffer C was stored at -70°C and used for binding studies. Binding reactions of 10 ~1 each contained 20 mM HEPES (pH 7.9), 0.5 mM DTT, 75 mM KCl, 5 mM MgC12, a one-fifth dilution of buffer C, 5 c(g of noncompetitor DNA (below) and nuclear extract containing

M. Nemer et at. /Mechanisms of Development 50 (1995)

5 pg protein. After incubation at 15°C for 15 min, 2 ~1 of stop buffer (15% Ficoll400, 0.25% bromphenol blue, 0.25% xylene cyanol) was added and samples were electrophoresed in 5% acrylamide gel (30% acrylamide, 0.8% his-acrylamide) in 1 x TBE (50 mM Tris-borate, pH 8.3, 1 mM EDTA) at room temperature for 2-4 h at 200 V. After electrophoresis, the gels were dried and submitted to autoradiography. Noncompetitor DNA was sheared salmon sperm DNA (Sigma) for the P3A probes and poly[dA * dT] (Pharmacia) for the P5 probes. Double-stranded probes, labeled by Klenow fill-in reaction with [32P]dCTP (3000 Ci/mmol, New England Nuclear), were made from pairs of oligodeoxyribonucleotides: (i) The 27-bp SpMTA P3A was comprised of 5 ‘TCGCAAACCTGTGTGCGCATGCT3 ’ and 3 ‘TTTGGACACACGCGTACGATGCCS ’ ; (ii) the 26-bp SpMTA P.5 was 5’GCGTTCATGCTTGATTITGAAGT3’ and 3’GAAGTACGAACTAAAACTTC AGAGS ‘; (iii) the 33-bp CyZZZuP3A was 5 ‘CGATGCATGAAGCGCAAACAAACTATTA3’ and 3’CGTAC’ITCGCGTTT’GTTTGAAATAATTCG5 ’ ;(iv) the 28-bpCyZZZaP.5 was comprised of oligos 5 ‘GATGT’TCATTGTCGCGACATACTT3 ’ and 3 ‘AAGTAACAGCGCTGTATGAACATCS ‘; the CyZZZu sites were taken from Theze et al. (1990). Acknowledgments

We thank Margaret Campbell for excellent technical assistance. This investigation was supported by Public Health Service grants HD-04367 (M.N.), CA-61333 (M.N.), CA-06927 and RR-05539 (I.C.R.) from the National Institutes of Health and by an appropriation from the Commonwealth of Pennsylvania. References Akasaka, K., Ueda, T., Higashinagawa, T., Yamada, K. and Shimada, H. (1991) Develop. Growth & Differ. 32, 9-13. Anderson, R., B&ten, R.J. and Davidson, E.H. (1994) Dev. Biol. 163, 11-18. Angerer, L.M., Kawczynski, G., Wilkinson, D.G., Nemer, M. and Angerer, R.C. (1987) Dev. Biol. 116, 543-558. Bai, G., Stuebing, E.W., Parker, H.R., Harlow, P. and Nemer, M. (1993) Mol. Cell. Biol. 13. 993-1001. Bruskin, A.M., Tyner, A.L., Wells, D.E., Showman, R.M. and Klein, W.H. (1981) Dev. Biol. 87, 308-318. Calzone, F. J., Theze, N., Thiebaud, P., Hill, R.L, B&ten, R.J. and Davidson, E.H. (1988) Genes Dev. 2, 1074-1088. Calzone, F.J., Hoog, C., Teplow, D.B., Cutting, A.E., Zeller, R.W., Britten, R.J. and Davidson, E.H. (1991) Development 112, 335-350.

131-137

137

Cameron, R.A.. Hough-Evans, B.R., Britten, R.J. and Davidson, E.H. (1987) Genes Dev. I, 75-84. Cox, K.H., Angerer, L.M., Lee, J.J., Davidson, E.H. and Angerer, R.C. (1986) J. Mol. Biol. 188, 159-172. Cserjesi, P., Fairley, P. and Brandhorst, B.P. (1992) B&hem. Cell Biol. 70, 1142-I 150. Davidson, E.H. (1989) Development 105, 421-445. DeSimone, SM., &White, K. (1993) Mol. Cell. Biol. 13, 3641-3649. Forty-Schaudies. S. and Hughes, S.H. (1991) J. Biol. Chem. 266, 13821-13827. Franks, R.R., Anderson, R., Moore, J.G., Hough-Evans, B.R., Britten, R.J. and Davidson, E.H. (1990) Development 110, 31-40. Gagnon, M.L., Angerer L.M. and Angerer, R.C. (1992) Development 114, 451-467. Gan, L., Zhang, W. and Klein W.H. (1990a) Dev. Biol. 139, 186- 196. Gan, L., Wessel, G.M., Klein, W.H. (199Ob) Dev. Biol. 142, 346-359. Gillies, D.S., Morrison, S.L., Oi, V.T. and Tonegawa, S. (1983) Cell 33, 717-728. Hardin, P.E., Angerer, L.M., Hardin, S.H., Angerer, R.C. and Klein, W.H. (1988) J. Mol. Biol. 202, 417-431. Harlow, P., Watkins, E., Thornton, R.D. and Nemer. M. (1989) Mol. Cell. Biol. 9, 5445-5455. Helfman, D.M., Roscigno, D.M., Mulligan, G.J., Finn, L.A. and Weber, K.S. (1990) Genes Dev. 4, 98- 110. Hoog, C., Calzone, F.J., Cutting, A.E., Britten, R.J., Davidson, E.H. (1991) Development 112, 351-364. Hoshijima, K., Inoue, K., Higuchi, I., Sakamoto, H. and Shimura, Y. (1991) Science 252, 833-836. Hough-Evans, B.R., Franks, R.R., Cameron, R.A., Britten, R.J. and Davidson, E.H. (1987) Dev. Bioi. 121, 576-579. Hough-Evans, B.R., Britten, R.J. and Davidson, E.H. (1988) Dev. Biol. 129, 198-208. Hough-Evans, B.R., Franks, R.R., Zeller, R.W., Britten, R.J. and Davidson, E.H. (1990) Development 110, 41-50. Kassis, J.A. (1990) Genes Dev. 4, 433-443. Lynn, D.A., Angerer, L.M., Bruskin, A.M., Klein, W.H. and Angerer, R.C. (1983) Proc. Natl. Acad. Sci., U.S.A. 80, 2656-2660. McMahon, A.P., Flytzanis, C.N., Hot&-Evans, B.R., Katula, K.S., Britten, R.J. and Davidson, E.H. (1985) Dev. Biol. 108, 420-430. Nemer, M., Travaglini, E.C., Rondinelh, E. and D’alonzo, J. (1984) Dev. Biol. 102, 471-482. Nemer, M., Thornton, R.D., Stuebing, E.W. and Harlow, P. (1991) J. Biol. Chem. 266, 6586-6593. Nemer, M., Bai, G. and Stuebing, E.W. (1993) Proc. Natl. Acad. Sci. USA 90, 10851-10855. Rippe, R.A., Lorenzen, S.-I. Brenner, D.A. and Breindl, M. (1989) Mol. Cell. Biol 9, 2224-2227. Theze, N., Calzone, F.J., Thiebaud, P., Hill, R.L., Britten, R.J. and Davidson, E.H. (1990) Mol. Reprod. Dev. 25, 110-122. Thiebaud, P., Goodstein, M., Calzone, F.J., Theze, N., Britten, R.J. and Davidson, E.H. (1990) Genes Dev. 4, 1999-2010. Virbasius, C.A., Virbasius, J.V. and Scarpulla, R.C. (1993) Genes Dev. 7, 2431-2445. Wieczorek, D.F, Smith, C.W. and Nadal-Ginard, B. (1988) Mol. Cell. Biol. 8, 679-694. Yang, Q., Angerer, L.M. and Angerer, R.C. (1989) Dev. Biol. 135, 53-65.