Mol Gen Genet (1990) 221:164-~70 © Springer-Verlag 1990

Multiple proteins bind to the P2 promoter region of the zein gene pMS1 of maize Uwe-G. Maier*, Klaus D. Grasser, Michael M. Haafl, and Giinter Feix Institute of Biology III, Schfinzlestrasse 1, D-7800 Freiburg, Federal Republic of Germany Summary. A 216 bp promoter fragment of the 19 kDa protein zein gene pMSI, containing the CCAAT and T A T A boxes, was analysed by a variety of techniques for in vitro interactions with nuclear proteins from endosperm tissue. H M G proteins were found to form stable complexes with these A/T-rich promoter sequences and several specific DNA-binding proteins appear to be involved in the formation of DNA-protein complexes with tiffs fragment. A 29 bp region spanning the two CCAAT boxes was protected from DNase I digestion in footprinting experiments. Key words: Z e a m a y s - Zein - P2 promoter - Transcription factors - H M G protein

Introduction The correct and regulated functioning of eucaryotic R N A polymerase II promoters depends on the prior formation of specific protein-DNA complexes in the region of the promoters or further upstream (e.g. Maniatis et al. 1987). The promoter-proximal sequence elements (TATA and CCAAT boxes) involved in these reactions have been found to be located at similar positions in most Pol II genes, while the more distant protein-binding sites show considerable variability in sequence and position (Wingender 1988). The composition and molecular details of some of the specific protein-DNA complexes formed have been analysed (e.g. the D r o s o p h i l a heat shock activator protein; Wu et al. 1987). Pol II genes from plants also contain TATA and CCAAT boxes (Messing et al. 1983); in some cases genes display an A G G A box instead of the CCAAT box (Heidecker and Messing 1986). Detailed study of the proteinD N A interactions occurring at these elements in plant genes, and analysis of their complexity, is still in progress. It is anticipated that mechanisms similar to those observed with genes from yeast and animals will also be found to occur in the specific activation of plant promoters, and particularly for the highly regulated zein genes from Z e a m a y s , which are the subject of this report. Genes coding for zeins, the storage proteins of maize, are expressed specifically in the endosperm tissue from approximately the 14th day after pollination (14 dap). Large * Present address: Institute for Biology II, Cell Biology, Sch~inzlestrasse 1, D-7800 Freiburg, Federal Republic of Germany Offprint requests to .' G. Feix

amounts of the protein are synthesized and stored in protein bodies. In the case of the zein genes isolated from maize variety A619 (pMS1, pMS2, pML1 and pML2) the presence of two promoters (P1 at - 1 0 0 0 and P2 at - 5 0 ) have been observed in the 5' flanking region (Langridge and Feix 1983; Langridge et al. 1985; Brown et al. 1986). Specific binding of nuclear proteins from endosperm and seedling tissue has been demonstrated recently to severn regions of the 5' flanking region of pMS1 (Maier et al. 1988). In particular, an endosperm-specific interaction occurring at a 15 bp consensus sequence (the - 3 0 0 box occurring in all zein genes and some storage protein genes from other cereals) could be shown (Maier et al. 1987). We have now extended the analysis of specific protein-DNA interactions occurring in the 5' flanking region of pMS1. We concentrated our analysis on the P2 promoter and immediately adjacent sequences. Evidence is presented for the involvement of several nuclear proteins in the specific proteinD N A complexes and, furthermore, the binding of nonhistone chromosomal high mobility group (HMG) proteins (Spiker 1988) to the A/T-rich promoter region.

Materials and methods Restriction endonucleases, alkaline phosphatase, Klenow polymerase and DNase I were purchased from Boehringer Mannheim. T4 polynucleotide kinase and radioactive nucleotides were from Amersham, DEAE membranes from Schleicher and Schuell, standard molecular weight markers, 1,10-phenanthroline and heparin agarose (type II) from Sigma and CNBr-activated Sepharose from Pharmacia. The plasmid pMS1 is described by Brown et al. 1986. All standard D N A manipulations were done as described by Maniatis et al. (1982). Sequencing was carried out as described by Maxam and Gilbert (1980). Nuclear extracts were prepared from hand-pollinated maize variety A619 as described by Maier et al. (1988). Isolation and phosphorylation of H M G proteins were carried out as described by Grasser et al. (1989). B a n d shift assay. The experiments were performed as outlined in Schmitz et al. (1989). Briefly, binding reactions containing in 30 gl up to 10 gg nuclear proteins, 5 ~tg carrier D N A (sonicated pSP65 DNA), 15% (v/v) glycerol in binding buffer (final concentration: 10 mM TRIS/HC1, pH 7.5, 50 m M NaC1, 1 mM dithioerythritol, 1 mM EDTA) were preincubated at room temperature for 10 min and after the

165 addition of 5000 cpm of the labelled fragments, were further incubated for 2 rain. After the addition of 3 gl of 0.05% (w/v) bromphenol blue and 0.05% (w/v) xylene cyanol FF, the reaction mixture was loaded on a 5% native polyacrylamide gel (30:0.8 acrylamide:bisacrylamide) and run for 3 h at 10V/cm. The electrophoresis buffer contained 50 mM TRIS, 380 m M glycine and 2 mM EDTA.

a

uo w

,

Pl

~

E

m

p2 m

®V

c

"'",....

/,//

///

~'""-.,.

Heparin-agarose chromatography. Nuclear proteins from 320 g endosperm tissue (10-12 dap) in 20 ml of column buffer (50 mM TRIS/HC1, pH 7.9, 1 mM EDTA, 12.5 ml MgC12, 20% glycerol, 1 mM dithiothreitol and 0.5 mM phenylmethylsulfonyl fluoride) were loaded on a 2 ml heparin-agarose (type II) column. After washing with column buffer, the bound protein was eluted with a linear salt gradient (8 ml, 0-1 M KC1 in column buffer). The fractions were dialysed against column buffer and stored in aliquots at - 70 ° C. DNA affinity chromatography. Five hundred micrograms of the subcloned SnaBI--BamHI fragment (see Fig. 1 a) were coupled to 1 ml of CNBr-activated Sepharose 4 B according to the manufacturer's recommendations. The active fractions from the heparin-agarose column were precipitated with 80% NH4SO4, dissolved in 1 ml and dialysed against column buffer. The chromatography was carried out as described by Kadonaga and Tjian (1986). SDS-gel electrophoresis. Nuclear proteins were separated in discontinuous 12% SDS-polyacrylamide gels according to Laemmli (1970) and silver stained according to Merril et al. (1979). Results

Basic structure of the P2 promoter of p M S l and its binding to H M G proteins The 19 kDa zein gene pMS1 is contained on a 4.4 kb EcoRI maize genomic D N A fragment isolated from the maize variety A619 (Brown et al. 1986). The region used in this investigation contains the P2 promoter and spans from positions - 8 to - 2 2 3 (relative to the protein start site at + 1) and was subcloned in pUC13. As indicated by Fig. lb, this 216 bp promoter fragment displays a TATA box at positions - 80 to - 88, two CCAAT boxes at positions - 121 to - 132 and, furthermore, from positions - 184 to - 2 0 8 the "7-11-7 element" which has been found to occur at

m // V;

100 "-,.

/////

DNase I footprint experiments. The method used was according to Galas and Schmitz (1978) with modifications. For the footprint experiments (final volume 100 pl) the binding buffer used in band shift assays additionally contained 10 mM MgClz. After completion of the binding reaction, 150 ng DNase I was added and incubated for 7.5 rain at room temperature. The reaction was stopped with 20 mM EDTA and loaded on a 5% native polyacrylamide gel. Free and protein-bound fragments were eluted from respective gel pieces in buffer containing 0.3 M NH4C1 and 20 mM TRIS/HC1, pH 7.5. After phenol:chloroform (1:1) and chloroform:isoamylalcohol (24:1) extraction the D N A was ethanol precipitated. The D N A fragment pattern obtained by electrophoresis in a 6% gradient urea sequencing gel was visualized by autoradiography.

I

.

C~

,

o~

7-1 1 -7

EC

b

o t

P2

eE

T

I

~ "-, ~

[

03

"-.

'l [

1 B0 bp -223 GTA T A A A T T A T A A CCCAACAAAG

-200 TAATCACTEA ATGTEAAAAC CAAETAGATA CCATgTCATC T E T A C E T T A T CTTACTAATA i -11313 GAETDAGATG

TETTTTTBCA

AATTCGAAAT TAATETTGCA CAAGCACAAD i

ATATTATTBA

GAEEAACTAG CAAEATAGAA AGCACAATAB TG

T D T A T A A A T A TCTETTAGAT TAGTAGCTAA TATATCGEAC

Fig. l. a Schematic representation of the zein clone pMS1. The thick line indicates the coding region, the dot 3' to the coding region a polyadenylation signal and the circles 5' to the coding region the CCAAT and TATA boxes of the promoters Pl and P2, respectively. The triangles indicate the PI and P2 promoters. c, CCAAT box; t, TATA box. b Sequence of the P2 promoter fragment. The promoter fragment was obtained by recutting the SnaBI-BamHI fragment at the RsaI site at position --8. The dotted line indicates the 7-11-7 sequence, the thick lines the CCAAT and TATA boxes and the bracket the footprint position

this location in other plant genes and is specifically expressed in the endosperm (Boronat et al. 1986). The promoter fragment was previously shown to interact specifically with nuclear proteins from endosperm and seedling tissue (Maier et al. 1987, 1988). The aim of the investigations described here was further analysis of the proteinD N A complexes formed, with emphasis on the TATA and CCAAT boxes and A/T-richness in general. Animal H M G proteins have been shown to be involved in protein-DNA complexes formed preferentially at A/Trich regions and in the stimulation of in vitro transcription from Pol I, II and III promoters (Brown and Anderson 1986; Solomon et al. 1986; Tremethick and Molloy 1988; Yang-Yen and Rothblum 1988). Thus we analysed first the binding of isolated H M G proteins to the A/T-rich promoter fragment (68% A/T). For such an analysis we could rely on the previous characterization of H M G proteins from maize endosperm and on the demonstration that these proteins are phosphorylated by a casein type II kinase activity of nuclei from maize endosperm (Grasser et al. 1989). Therefore, H M G proteins of about 20 kDa could be labelled with aZp before use in band-shift experiments with unlabelled promoter fragment. In order to demonstrate that the retarded band observed in such experiments is actually caused by a complex formed between the promoter fragment and H M G proteins, radioactive protein bound to the promoter fragment was eluted from the retarded band and then co-electrophoresed with purified H M G proteins. Figure 2a shows that the eluted protein (lane 1) and H M G proteins (lane 2) display the same electrophoretic

166 a

b

/~g carrier

0.25

0.25 .i 0.20-

--37 -27

1.0 0.9 0.8 O.7

~0.15. 0

0.10.

0.6_

0.5

-.4

0.4

-17 -14

0

t

b

l

3

-"~ II

1

2

0.3 0.2 0.1 0

0.05

1

2

3

4

l

i

5

i

J

7

~

9

l

l

l

11

l

l

l

13

15

b

5

Fig. 2. a SDS-PAGE of HMG proteins. Lane 1, autoradiography of 32p-labelled HMG proteins isolated by elution from the retarded band in electrophoretic band-shift experiment with unlabelled promoter fragment. Lane 2, electrophoretic mobility of HMG proteins (purified from maize endosperm) after Coomassie staining of the gel. The arrows indicate the running positions of the HMG proteins. The numbers at the side indicate the positions of molecular weight markers in kda. b Electrophoretic band-shift experiment of HMG protein-DNA complexes. The binding assays of 30 gl in lanes 1-5 each contained 5000 cpm of 32p-labelled promoter fragment, 5000 cpm of 32p-labelled DNA fragment of the coding region of pMS1 and, except for lane 1, 2 gg of HMG protein. In addition, the assays in lanes 3-5 contained 0.25, 1 and 5 gg of carrier DNA, respectively. The running positions of the proteinfree promoter and coding fragment are indicated by I and lI, respectively

mobility. The complex formation between H M G proteins and the P2 promoter fragment is more sensitive to the addition of carrier D N A than the complex involving specific binding proteins. As shown in Fig. 2b, the presence of 0.25 gg of carrier D N A effects the complex formation already (lane 3) whereas 1 gg carrier D N A results in a slight band retardation only (lane 4). The gradual decrease of the band-shift effect indicates the binding of several proteins with changing affinities. Figure 2b also demonstrates that a D N A fragment from the coding region of pMS1 with 53% A/T did not lead to band retardation. The results are indicative of preferential binding of H M G proteins to the A/T-rich promoter fragment.

F

5

6

7

F

5

6

7

8

9

10

9

10

11

12

C

8

11

12

Fig. 3. a Elution profile of a heparin-agarose column loaded with endosperm nuclear proteins. The heavy line represents the protein content, the thin line the salt gradient, b Band-shift experiments of the P2 promoter fragment with protein from the indicated column fractions of a or without protein (lane F). e Band-shift experiments as in b except that a fragment from the coding region of pMS1 was used

Specific binding of nuclear proteins from endosperm to the P2 promoter fragment Band-shift analysis of the labelled P2 promoter fragment was performed with a mixture of nuclear proteins enriched from a crude extract by heparin-agarose chromatography, yielding protein preparations with improved DNA-binding properties. Figure 3 shows the resulting protein elution profile, and band-shift analysis performed with the individual column fractions. In contrast to the reactions with H M G proteins, the binding reactions performed here were always executed in the presence of high amounts of carrier D N A

which inhibits the binding of H M G proteins. It can be seen that use of fractions 8 to 10 of the column protein peak led to the appearance of a strong retarded band while band-shift experiments performed with fractions eluting later displayed a different banding pattern. The specificity of the band-shift analysis is indicated by the absence of retarded bands when a D N A fragment of the coding region ofpMS1 is used (Fig. 3c). In order to test whether the mixture of D N A binding proteins also contains proteins with zinc finger characteris-

167

iiii!iii iiiiiiiiii ii !iiiiiiii T T C

i!~ii!i!iiii~ii!

1

2

3

Fig. 4. Influence of zinc chelator on band-shift experiments. Bandshift experiments were done with the promoter fragment and protein from fraction 11 of the heparin-agarose column in the absence (lane 2) or presence (lane 3) of 10 mM 1,10-phenanthroline (in ethanol). The assay in lane 1 contained no protein while the assay in lane 2 contained the same amount of ethanol as that in lane 3. The arrow points to the retarded band tics, protein binding tests were performed with aliquots of the column fractions in the presence of the zinc chelator 1,10-phenanthroline (Posorske et al. 1979). In the case of the major histocompatibility complex class II gene, E~, (Hooft van Huijsduijnen et al. 1987) the addition of this chelator agent inhibited the D N A binding of a protein with a zinc finger domain. Figure 4 shows that there is a loss of D N A binding activity under the influence of the chelator with proteins from column fraction 11 (cf. lanes 2, 3). Addition of the zinc chelator to fractions 8 to 10 did not interfere with the formation of the retarded band in the band-shift experiment, which is a further indication for the occurrence of different D N A binding proteins in the various column fractions. The results confirmed that the ethanol used as chelator solution medium was not a cause for the disappearance of the retarded band in lane 3.

Localization of a protein binding reaction on the promoter fragment by DNAse I footprinting A footprint analysis was performed by separating protein bound D N A fragments from the naked D N A in preparative gels after the DNase I nicking reaction. This method has the advantage of identifying binding sites recognized in complete protein-DNA reactions by using the protein bound D N A from band-shift experiments only. For the experiment presented in Fig. 5, 20 gg protein were incubated first with carrier D N A for 10 min, then 2 x 1 0 6 cpm of the P2 promoter fragment were added and allowed to react for 3 min. A partial DNase I reaction was subsequently conducted, followed by preparative band-shifts to separate bound protein from protein free D N A . Lane 1 of Fig. 5 shows the DNase I pattern of this reaction, demonstrating that a footprint of 29 bp overlaps with the two C C A A T boxes. Lane 2 shows the results of the DNase I reaction without prior addition of nuclear proteins.

T A A T T A A A 13 C T T A A A E 13 T T T T T C T A T A A

Fig. 5. DNase I footprint experiment of P2 promoter fragment. The autoradiogram shows electrophoretic patterns of DNase I fragments obtained by prior incubation of the promoter fragment with nuclear proteins (lane 1) or without preincubation with nuclear proteins (lane 2). The nucleotide sequence of the protected region is given at the right

Enrichment of promoter binding proteins by affinity chromatography Affinity chromatography has become a potential tool for the purification of D N A binding proteins (e.g. K a d o n a g a

168 ing specifically to the P2 promoter fragment of the affinity chromatography column.

-84 Discussion

-48 -37 -27

1 2 3 Fig. 6. SDS-PAGE of nuclear proteins. Silver-stained protein patterns from: lane 1, crude nuclear extract; lane 2, peak fraction of a heparin-agarose column; and lane 3, protein adsorbed to the promoter fragment affinity column. The positions of molecular weight markers are given in kda

1

2

Fig. 7. Band-shift experiments with the P2 promoter fragment and protein bound on the DNA-affinity column. Lane 1, migration of the protein-free P2 promoter fragment; lane 2, P2 promoter fragment after incubation with the DNA-affinity column fraction. The arrows point to band shift positions and Tjian 1986). To benefit from this technique, the active fractons of a heparin-agarose column were first concentrated with 80% ammonium sulphate, then preincubated with high concentrations of carrier D N A to avoid binding of H M G proteins, and finally applied to the P2 promoter fragment affinity column. Several proteins bind to the affinity column as identified by silver staining of eluted proteins separated on SDS gels (Fig. 6, lane 3). For comparison, lanes 1 and 2 of Fig. 6 show the proteins detected by silver staining in gel-separated crude nuclear extracts or a heparin-agarose chromatography peak fraction, respectively. Affinity enriched proteins, in a m o u n t s not detectable by the Bradford method (Bradford 1976) were used for bandshift experiments. The same two retarded bands were observed as before, when about 5 gg protein from the different heparin-agarose fractions were used. The band-shift experiment of Fig. 7 indicates a high enrichment of proteins bind-

From detailed work with a number of eucaryotic promoters it can be assumed that promoters are activated by complex interactions with nuclear proteins before the initiation of R N A synthesis takes place. The specific effect of the DNAprotein complexes formed may additionally involve proteinprotein interactions and protein modifications. The evidence presented in this paper indicates that a number of proteins appear to be involved in the complexes formed stably in vitro with the P2 promoter region of zein genes. This was suggested previously by genetic studies with the regulatory Opaquetype mutants, which indicated the action of different trans-acting proteins in the specific expression of zein genes (Di Fonzo et al. 1979). The band-shift experiments with binding reactions performed in the presence of high amounts of carrier D N A resolve several retarded bands which may represent complexes with different amounts of the same or different proteins bound to the promoter fragment. Since some of the observed differences in the band-shift patterns were dependent upon the heparin-agarose chromatography fractions used, a slight separation of different proteins that are complex associated is indicated, although concentration effects of the proteins involved should also be considered. The presence of different D N A binding proteins is supported by the inhibition of their specific binding to D N A when the zinc chelator 1,10-phenanthroline is added. Only with non-peak fractions of the heparin-agarose column could such effects be demonstrated, indicating the presence of zinc finger proteins in these column fractions. The DNase I footprint analysis resulted in the protection of a 29 bp segment, including the CCAAT boxes, and indicates the simultaneous binding of several proteins to this region, since D N A binding of protein monomers usually leads to protection of shorter D N A fragments. This finding is in accordance with the results obtained with other gene systems where several CCAAT binding proteins have been identified (e.g. Dorn et al. 1987; Chodosh et al. 1988; Santoro et al. 1988). The borders of the protein binding sites at the CCAAT boxes, however, may not be mapped precisely by this method since DNase I does not hydrolyse the A/T-rich regions of D N A very efficiently. A protein binding to the TATA box region of the P2 promoter region could not be demonstrated. This may be caused by the instability of a bound protein or by insufficient levels of binding proteins in the preparation used. The difficulty in demonstrating a specific protein binding to T A T A boxes in vitro has also been encountered in other gene systems. Further protein-DNA interactions assumed to occur additionally in the 216-bp-long P2 promoter fragment probably also eluded detection by the footprinting method. A candidate for such a protein interaction is the 7-11-7 sequence element from positions - 1 8 4 to - 2 0 8 , which could be involved in the interaction with regulatory proteins as it occurs at a similar location in comparable genes (Boronat et al. 1986). The presence of several nuclear proteins binding specifically to the P2 promoter fragment was also demonstrated by affinity chromatography experiments. The large excess of carrier D N A (2.5 gg DNA/less than I gg protein) used

169 in band-shift analysis o f the proteins enriched by the affinity column demonstrates the high specificity o f the D N A - p r o tein interactions at this stage. Because o f our present problems in scaling up the p r e p a r a t i o n o f nuclear proteins, the affinity column could only be used analytically and a correlation between individual proteins and specific D N A - b i n d ing sites has yet to be determined. The results obtained with the H M G proteins indicate that m o r e general D N A - b i n d i n g proteins m a y also be an i m p o r t a n t part o f the overall complexes connected with active zein promoters. Recent findings in animal systems showed that H M G proteins (at least in the case of H M G 1 and 2) bind A/T-rich structures Preferentially and can be competed with carrier D N A (Brown and A n d e r s o n 1986; S o l o m o n et al. 1986). These results could be analogous to our findings concerning the P2 promoter. F u r t h e r analysis of the interaction o f the non-histone c h r o m o s o m a l H M G proteins with other complex associated proteins will be very helpful for an assessment of the special chromatin structures connected to zein gene activation. The complex protein binding patterns taking place in the P2 p r o m o t e r region are thought to be insufficient for final p r o m o t e r activation which is likely to require the complementary action o f elements located further upstream. This is indicated by transient transformation experiments with deletion mutants in carrot protoplasts, which demonstrate that sequences other than the p r o m o t e r region used in the present investigation are necessary for detectable functioning of the P2 p r o m o t e r o f a related zein gene (Boston et al. 1987; Roussell et al. 1988). Furthermore, specific protein binding studies with nuclear proteins from seedling and endosperm tissue, as well as transient transformation assays in protoplasts with pMS1 and p M L I sequences o f up to 1.5 kb from the coding region also indicate the importance of these regions for a regulated expression o f the zein genes (unpublished results). F o r these upstream regulatory elements to be functional, the basic protein binding complexes at the P2 p r o m o t e r region are a necessary prerequisite. C o m b i n e d action o f proximally and distally located protein complexes in conjunction with each other is anticipated to occur in regulated expression o f the zein gene. Acknowledgements. The technical assistance of E. Brutzer and A. Schfifer is gratefully acknowledged. We thank D. Horn, Dr. T. Quayle and M.L. Schmitz for comments on the manuscript. This work was supported by grants from the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie.

References

Boronat A, Martinez MC, Reina M, Puigdonech P, Palau J (1986) Isolation and sequencing of a 28 kD glutelin-2 gene from maize. Common elements in the 5' flanking regions among zein and glutelin genes. Plant Sci 47:95-102 Boston RS, Becwar MR, Ryan RD, Goldsbrough PB, Larkins BA, Hodges TK (1987) Expression from heterologous promoters in electroporated carrot protoplasts. Plant Physiol 83 : 742-746 Bradford MM (1976) A rapid and sensitive method for the quantification of microgram quantities of proteins utilizing the principle of protein-dye binding. Anal Biochem 72:248-254 Brown JW, Anderson JA (1986) The binding of the chromosomal protein HMG-2a to DNA regions of reduced stabilities. J Biol Chem 261 : 1349 1354 Brown JWS, Wandelt C, Feix G, Neuhaus G, Schweiger H-G

(1986) The upstream regions of zein genes: sequence analysis and expression in the unicellular green alga Acetabularia. Eur J Cell Biol 42:161-170 Chodosh LA, Baldwin AS, Carthew RW, Sharp PA (1988) Human CCAAT-binding proteins have heterologous subunits. Cell 53 : 11-24 Di Fonzo N, Fornasari E, Salamini F, Reggiani A, Stave C (1979) Interactions of maize mutants floury-2 and opaque-2 in the synthesis of endosperm proteins. J Hered 71 : 397-402 Dorn A, Bollekens J, Staub A, Benoist C, Mathis D (1987) A multiplicity of CCAAT box binding proteins. Cell 50: 863-872 Galas DJ, Schmitz A (1978) DNase footprinting: a simple method for the detection of protein-DNA binding specificity. Nucleic Acids Res 5:3157-3170 Grasser KD, Maier U-G, Feix G (1989) A nuclear casein type II kinase from maize endosperm phosphorylating HMG proteins. Biochem Biophys Res Commun 162:456-463 Heidecker G, Messing J (1986) Structural analysis of plant genes. Annu Rev Plant Physiol 37 : 439-466 Hooft van Huijsduijnen RAM, Bollekens J, Dorn A, Benoist C, Mathis D (1987) Properties of a CCAAT box-binding protein. Nucleic Acids Res 15 : 7265 7282 Kadonaga JT, Tjian R (1986) Affinity purification of sequencespecific DNA binding proteins. Proc Natl Acad Sci USA 83 : 5889-5893 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature 227 : 680-685 Langridge P, Feix G (1983) A zein gene of maize is transcribed from two widely separated promoter regions. Cell 34:1015-1022 Langridge P, Brown JWS, Pintor-Toro JA, Feix G, Neuhaus G, Neuhaus-Url G, Schweiger H-G (1985) Expression ofzein genes in Aeetabularia mediterranea. Eur J Cell Biol 39:257-264 Maier U-G, Brown JWS, Toloczyki C, Feix G (1987) Binding of a nuclear factor to a consensus sequence in the 5' flanking region of zein genes from maize. EMBO J 6:17-22 Maier U-G, Brown JWS, Schmitz LM, Schwall M, Dietrich G, Feix G (1988) Mapping of tissue-dependent and independent protein binding sites to the 5' upstream region of a zein gene. Mol Gen Genet 212:241-245 Maniatis T, Fritsch EF, Sambrook J (1982) Molecular cloning: A laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York Maniatis T, Goodbourn S, Fischer JA (1987) Regulation of inducible and tissue specific gene expression. Science 236:1237-1245 Maxam AM, Gilbert W (1980) Sequencing end-labeled DNA with base-specific chemical cleavages. Methods Enzymol 65:499-560 Messing J, Gerathy D, Heidecker G, Hu N-T, Kridl J, Rubenstein I (1983) Plant gene structure. In: Hollaender A, Kosuge T, Meredith C (eds) Genetic engineering of plants. Plenum, New York, pp 211-228 Merril CR, Switzer RC, Van Keuren ML (1979) Trace polypeptides in cellular extracts and human body fluids detected by twodimensional electrophoresis and a highly sensitive silver stain. J Biol Chem 255 : 11992-11996 Posorske CH, Cohn M, Yanagisawa N, Auld DS (1979) Methionyl-tRNA synthase of Escherichia eoli, a zinc metallo protein. Biochem Biophys Acta 576:128-133 Roussell DL, Boston RS, Goldsbrough PB, Larkins BA (1988) Deletion of DNA sequences flanking an M~ 19000 zein gene reduces its transcriptional activity in heterologous plant tissues. Mol Gen Genet 211 : 20~209 Santoro C, Mermod N, Andrews PC, Tjian R (1988) A family of human CCAAT-box-binding proteins active in transcription and DNA replication: cloning and expression of multiple cDNAs. Nature 334:218-224 Schmitz ML, Maier U-G, Brown JWS, Feix G (1989) Specific binding of nuclear proteins to the promoter region of a maize nuclear rRNA gene unit. J Biol Chem 264:1467-1472 Solomon MJ, Strauss F, Varshavsky A (1986) A mammalian high

170 mobility group protein recognizes any stretch of six A/T base pairs in duplex DNA. Proc Natl Acad Sci USA 83:1276-1280 Spiker S (1988) Histone variants and high mobility group nonhistone chromosomal proteins of higher plants: Their potential for forming a chromatin structure that is either poised for transcription or transcriptionally inert. Physiol Plant 74:200-213 Tremethick D J, Molloy PL (1988) Effects of high mobility group proteins 1 and 2 on initiation and elongation of specific transcription by RNA polymerase II in vitro. Nucleic Acids Res 16:11107-11123 Wingender E (1988) Compilation of transcription regulating proteins. Nucleic Acids Res 16:1879-1902 Wu C, Wilson S, Walker B, Dawid I, Paisley T, Zimarino V, Ueda

H (1987) Purification and properties of Drosophila heat shock activator protein. Science 238 : 1247-1253 Yang-Yen HF, Rothblum LI (1988) Purification and characterization of a high-mobility-group-like DNA-binding protein that stimulates rRNA synthesis in vitro. Mol Cell Biol 8:3406-3414

C o m m u n i c a t e d by R.G. H e r r m a n n

Received October 29, 1989