Plant Molecular Biology 11:561-573 (1988) © KluwerAcademicPublishers, Dordrecht - Printed in the Netherlands

561

Constitutive and light-induced DNAseI hypersensitive sites in the rbcS genes of pea (Pisum sativum) Andrea G6rz, Willi Sch~ifer l, Eiji Hirasawa 2 and Gfinter Kahl

Plant Molecular Biology Group, Department of Biology, University of FrankfurtZMain, Federal Republic of Germany; lpresent address: Department of Plant Pathology, Cornell University, Ithaca, N Y 148535908, USA; 2present address: Osaka City University, Faculty of Science, Sumiyoshi-ku, Sugimoto-cho, Osaka 558, Japan Received 27 January 1988; accepted in revised form 10 August 1988

Key words: chromatin structure, DNAseI-hypersensitive sites, methylation, photoinduction, promoter, rbcS genes

Abstract

The chromatin structure of pea (Pisum sativum) rbcS genes in inactive (root), potentially active (dark-grown leaf), and active states (light-grown leaf) was analysed using (a) pancreatic DNAseI to detect general DNAseI sensitivity and DNAseI-hypersensitive sites, and (b) methyl-sensitive restriction endonucleases to probe for cytosine methylation within the promoter region. We showed that within the same organ individual members of the pea rbcS multigene family are differentially sensitive to DNAseI suggesting differential protection in nuclei. During light activation general DNAseI sensitivity increases in some genes, especially their 5' upstream regulatory sequences. DNAseI-hypersensitive sites are constitutively present in 5' upstream regulatory sequences around positions -335, -465, -650, and -945 (5' constitutive domain) and in the coding region around position + 340, + 450, + 530, + 640, and + 810 (3' constitutive domain). One additional hypersensitive site appears after light induction (inducible site). This region is centred around position -190 and flanked by light-responsive elements (LREs). In spite of changes in the chromatin structure of rbcS genes during their transition from an inactive to an active state, their cytosine methylation at Alu I, Fnu 4HI, Hae III, Sau 3AI and Sau 96I sites in the promoter region remains uniform.

Introduction

The small subunit of ribulose-l,5-bisphosphate carboxylase/oxygenase (rubisco) is encoded by a small multigene family of five or more members (rbcS genes) in various di- and monocotyledonous plants [4, 9]. The rbcS mRNAs constitute a major fraction of total mRNA in photosynthetically active higher plant tissues. Their translation products are transported into the chloroplast with concomitant posttranslational processing. Within the chloroplast

eight small subunits together with eight chloroplastencoded large subunits form the rubisco holoenzyme which catalyses the fixation of CO 2. The levels of rbcS m R N A are tightly regulated by light via a phytochrome- and/or blue light receptor-mediated response, and regulation is predominantly at the transcriptional level [16, 18, 41]. Transcription is either extremely low or not detectable in the dark but rapidly activated in the light. Consequently rbcS transcripts are most abundant in photosynthetic organs such as leaves and pericarps [9, 16]. Simpson

562 and co-workers [36] found rbcS expression to be restricted to cells containing plastids of a defined developmental stage. Light responsiveness and organ specificity has been appointed to sequences upstream of the transcription start site using transgenic calli and plants harbouring gene fusions of the rbcS promoter and a reporter gene [reviewed in 17 and 28]. However, the light-responsive elements (LREs) alone do not determine the rate of gene expression, as in various transgenic plants the rbcS transcript levels vary from 5to 10-fold [27], or even more [24]. This suggests that other factors besides nucleotide sequence influence the extent of rbcS transcription, as are the sites of integration of the genes or changes in their chromatin structure [17, 28, 32]. The role of chromatin structure for gene regulation has been supported by several reports [reviewed in 5, 11 and 39]. A more open configuration of regulatory sequences seems to be a prerequisite for gene activation. Such 'open windows' occur preferentially in the 5' upstream region o f a gene and can then be associated with regulatory proteins and other transcription factors. Generally, chromatin of active genes is about 10 times more sensitive towards the endonuclease DNAseI than is bulk chromatin. This general sensitivity is again 10-fold higher for distinct stretches of a gene (DNAseIhypersensitive sites). These areas, though referred to as sites, may extend over a length of 5 0 - 200 bp [12]. In numerous animal genes a positive correlation between the occurrence of these DNAseIhypersensitive sites and gene expression has been demonstrated but still little is known about plant genes. General DNAseI sensitivity of active plant genes has been observed [8, 31, 37, 38, 47]. However, DNAseI-hypersensitive sites in plant genes have only recently been reported [34, 42, 47]. Another possible determinant of gene activity is methylation of specific cytosine residues which in viral promoter sequences leads to an inactivation of the neighbouring gene [10]. This mechanism also works in some animal [reviewed in 35] and plant genes [1, 22, 23, 43]. The differential expression of the individual rbcS genes within one and the same tissue and their response towards light could arise from different packing of their chromatin. We reasoned that chromatin

of inactive rbcS genes would be tightly packed but becomes more 'open' before or upon activation. Since it is possible to detect these changes, we investigated general DNAse I sensitivity, appearance or disappearance of DNAseI-hypersensitive sites and cytosine methylation in the promoter region of pea rubisco genes in their inactive, potentially active and active states.

Materials and methods

Plant material Green leaves were harvested from pea plants (Pisum sativum, cv. Progress No 9) 3 - 4 weeks old grown under normal light conditions. For dark-grown leaf tissue plants were transferred into complete darkness 8 days prior to harvesting, which together with the initial isolation steps was performed in complete darkness omitting any safety lights to avoid possible effects of green light [29].

Gene probes Probes were derived from the genomic clone of rbcS 3.6 (pPS R6 [6]). The physical map of this clone is shown in Fig. 1. In subsequent figures this individual clone .is drawn as a continuous line. Standard procedures for recombinant DNA work were performed according to Maniatis et al. [30].

Isolation of total genomic DNA Total genomic DNA was isolated according to Bedbrook [2] except that the initial extraction buffer was the same as described for the preparation of nuclei. The resulting DNA had a length of more than 50 kb.

Preparation of nuclei About 40 g of leaf tissue and 80 g of root tissue was homogenized together with 10 volumes of extraction buffer (250 mM sucrose, 10 mM NaC1, 10 mM

563 Pea Rubisco ssu C.ene

Start mRNA

poly(A)site I

"~-~--:---[ Eco RI

~50 TATA Exon 1

Intron I

[

Intron2 Exon2

Exon3

3"UT

I

[ H1nfl

[

--po~322

HindIII

HpalI

Eco RV

Hinfl

HpoI

EcoRV

Hinf I

Fig. 1. Physical map of clone rbcS 3.6. Restriction sites of enzymes used in generating hybridizationprobes are indicated.

MES (pH 6.0), 0.5 mM EDTA, 20 mM ~mercaptoethanol (/3MCE), 0.6% Nonidet P40, 0.5 % bovine serum albumin (BSA), 0.2 mM PMSF) in a Waring blender, filtered through two layers of miracloth and centrifuged for 10 min at 2000 g and 4 °C. The pellet was washed with 0.5 vol of extraction buffer. The crude nuclear pellet was suspended in digestion buffer containing 10 mM PIPES, 0.25 M sucrose, 10 mM NaCI, 3 mM MgC12, 5 mM ~MCE. Intactness of nuclei was controlled by light microscopy after staining with carmine acetic acid and by Micrococcus nuclease digestion to give the typical nucleosomal ladder (data not shown).

contaminating RNA by RNAse A digestion (15/~g/ml) followed by phenol extraction and ethanol precipitation as described above. DNA concentration of the different samples was measured spectrophotometrically. Total genomic DNA ('naked' DNA) at a concentration of 200 tzg/ml in digestion buffer was digested with 0.5 #g DNAseI/ml for the same time intervals as described for nuclei. Purified DNA was restricted to completion with Eco RV and separated on agarose gels. Pst I fragments of wild-type k-DNA were used as molecular weight markers.

Analysis of cytosine methylation D N A s e I digestion of isolated nuclei and total genomic D N A Nuclei suspensions were adjusted to 150/~g D N A / m l as determined spectrophotometrically after lysis with 5 M urea/2 M NaC1. DNAseI digestion was performed at 4 °C with 5/zg DNAseI/ml. Aliquots were removed after 0.25, 0.5, 1, 2, 4, 8 and 16 min, and the reaction stopped by adding 0.1 vol of 200 mM EDTA, 5% SDS. Control samples (o) were mock-digested for 16 min. Proteinase K was then added to a final concentration of 0.5 mg/ml and samples incubated for 5 min at 60 °C and 3 h at 37°C. DNA was extracted with phenol/chloroform/isoamyl alcohol (25:24:1), and ethanolprecipitated. Protein-free DNA was purified from

Total genomic DNA (about 25 #g) was restricted to completion with 100 U each of Alu I, Alu I + Eco RV, Fnu 4HI, Hae III, Sau 3AI, Sau 96I, and Eco RI in a volume o f 500/A at 37°C for 4 h in the recommended buffers. Digestion was monitored by comparing aliquots of the digest with intact pea genomic DNA in ethidium bromide (EtBr) stained test gels and by the following method: 20/zl aliquots of the complete digestion mixtures including enzyme were added to 5/~1 (0.5/zg) R-DNA. These test samples were incubated exactly as described for the digestion samples and checked on 0.8 %0 agarose gels for correct X-DNA restriction patterns under this sub-optimal DNA/enzyme ratio (control of correct and efficient restriction). Totally digested DNA was

564 purified once with phenol/chloroform/isoamyl alcohol (25:24:1), concentrated by ethanol precipitation, separated electrophoretically in 2070 agarose, blotted, and hybridized to a probe spanning the first 973 bp upstream o f the transcription start.

Hybridization DNA was blotted onto nitrocellulose and hybridized for 20 h at 50°C in 50°7o formamide, 5 0 m M Na2HPO4/NaH2PO 4, 450 mM NaC1, 45 mM sodium citrate, 10 x Denhardt's solution, 100 #g/ml sssalmon sperm DNA. Cloned DNA was labelled to a specific activity o f 2 x 109 either by random priming [13] or by the synthesis o f single-stranded probes from an M13 template [7]. Blots were washed 2 x 30 min at 55 °C in 0.05 x SSC (7.5 mM NaC1, 0.75 mM sodium citrate). Under these stringent conditions the 5' probes hybridize specifically to only two rbcS gene copies.

Sensitivity of chromatin from different tissues to parthTl digestion with DNAse! Nuclei isolated from tissues of different organs (leaves and roots) grown under different environmental conditions (light- and dark-grown leaves) were digested with DNAseI for different times at otherwise identical conditions (5 #g DNAseI/150/~g DNA in 1 ml digestion volume). As shown in Fig. 2 (upper panel) comparison of the EtBr patterns after electrophoretic separation of the resulting DNA fragments reveals a much greater sensitivity of chromatin from root tissue as compared to both light- and dark-grown leaf tissue. Since mock-digested nuclei (o) yielded high molecular weight DNA similar to that of leaf tissue we conclude that this more rapid degradation does not simply result from a higher endogenous nuclease activity of root tissue during preparation. The amount of DNA in the digestion samples was adjusted according to photometric determination after lysis in urea/NaCl and proved to be highly reproducible both for leaf and root tissue.

Results

Methods for the isolation of nuclei Changing standard protocols for the isolation of intact nuclei turned out to be necessary for the study of chromatin structure using DNAseI. The extraction buffer had to be used without any polyamines such as spermine or spermidine to exclude any interaction with chromatin and to avoid erroneous digestion patterns. Moreover, extraction procedures were kept extremely short by leaving out final purification steps to minimize destruction of nuclei during isolation. As a consequence of these modifications the initial EDTA concentration had to be scaled down to 0.5 mM, so that it could be more easily washed out before adding DNAseI. DNA isolated directly from these nuclei had a molecular weight of more than 50 kb as compared to intact X-DNA. The smear in the EtBr patterns (Fig. 2, upper panel) results from overloading the gel).

Comparison of general DNAseI sensitivity of different members of the rbcS multigene family To compare general DNAseI sensitivity of the individual rbcS gene copies a gene probe spanning from - 270 to + 355 was labeled by random priming [13]. Isolated DNA from nuclear digestions was restricted withEco RV. The members of the gene family share a common conserved Eco RV site 37 bp downstream of the transcription initiation site thus generating fragments which can be assigned either to the 5' or to the 3' region of the gene. Under stringent hybridization and washing conditions as described in Material and methods the probe (Fig. 2) formed hybrids with fragments 1.1 and 1.4 kb of the 5' untranscribed region and fragments 1.2, 1.9, 2.1, 2.5, and 11.0 kb for the coding and 3' untranslated region. According to sequence homology [18, 40] promoter fragments 1.1 and 1.4 kb can be assigned to genes rbcS 3.6 and rbcS 3A, respectively, whereas for fragments downstream of the transcription start site only rbcS 3.6 (21. kb) and rbcS E9 (1.9 kb) can

565

Fig. 2. General DNAseI sensitivity of bulk chromatin and individual genes of the rbcS gene family from pea. Nuclei from light-grown

(L) and dark-grown (D) leaf tissue and roots (R) were increasingly digested with DNAseI. DNA (1/zg/lane) was separated in 0.4°7oagarose gels and subsequently stained with ethidium bromide (0.5/~g/ml; upper panel). For hybridisation with the gene-specific probe 15 #g DNA/lane were restricted with Eco RV and electrophoresed in 1070agarose. Southern blots were hybridized to the indicated probe and autoradiograms are shown in the lower panel. RV: Eco RV site identical to all rbcS genes. (RV): Eco RV site specific for individual members of the rbcS gene family. Distances from the transcription start site were calculated by comparison with Pst I-restricted ~,-DNA. Arrowheads: fragments upstream of transcription initiation. Lines: fragments downstream of transcription initiation.

clearly be identified. Identification o f the three remaining fragments is based o n the correlation between transcriptional activity [18] and D N A s e I sensitivity o f these genes (see Discussion) and are therefore put in parenthesis in Fig. 3. Classification o f the different rbcS genes is according to Fluhr et al. [18]. To quantitate the degree o f general D N A s e I sensitivity o f the individual rbcS fragments autoradiographs f r o m three independent experiments were scanned and signal intensity was expressed as a percentage o f the c o r r e s p o n d i n g control (o). D a t a o f in-

duced and n o n - i n d u c e d leaf tissues are summarized in Fig. 3. By c o m p a r i n g the general D N A s e I sensitivity o f individual rbcS copies (Fig. 2, lower panel; Fig. 3) we f o u n d that upstream region o f rbcS 3A (1.4 kb) to be more sensitive than that o f rbcS 3.6 (1.1 kb). D o w n s t r e a m we observed two fragments o f high (1 1.0 and 2.5 kb), two fragments o f intermediate (2.1 and 1.2 kb), and one fragment o f low D N A s e I sensitivity (1.9 kb). This pattern o f differential sensitivity a m o n g the rbcS copies is similar in all three tissues. The change in D N A s e I sensitivity o f the individu-

566 UPSTREAM

DOWNSTREAM Eco RV

oo

80

~

rbcS-3A

rbcS-3.6

(rbc5-8.0)

rbcS-E9

rbcS-3,6

(rbcS-3A)

(rbc5-3C)

L4kb

1.1kb

1.2kb

1.9kb

2.1kb

2.5kb

11.0kb

I

k

II h

L

[-'1

so

~ ,o

m

0 DNAseI-.I~

)DN,4sel ~

ODNAseI--I~

ODNAsel.--Ib

OON,4sel--I~

ODNAse7..-I~

ODNAseI---~

/~LIo 25.2%

/~L/o 8.7%

L~L/D 4..5°1o

/~LID l,&°/*

L~L/D 8.3°/°

/~LIDI2.5°/o

/~LID 15"/,

Fig. 3. Comparison of the general DNAseI sensitivity of active (L) and inactive (D) individual rbcS gene copies in pea leaf tissue (see Fig. 2 for further details). The individual rbcS signals were quantified by scanning autoradiograms from three independent experiments. The assignment of the various fragments to particular rbcS copies is explained in the discussion. A L / D : mean difference between light and dark as a percentage of dark.

al fragments after light induction of the corresponding gene is indicated as A L/D. Data of the respective DNAseI-digestion series (Fig. 3) was summed up and the differrence between light and dark expressed as a percentage of the dark value. For the downstream region the difference between light and dark is maximal for those fragments which are also most sensitive towards DNAseI (11.0 and 2.5 kb) and decreases gradually from fragment 2.1 over 1.9 to 1.2 kb. In the upstream region A L / D of rbcS 3A is significantly higher than for rbcS 3.6 (Fig. 3).

Mapping of DNAseI hypersensitive sites in the 5' untranscribed region of two rbcS genes Four hypersensitive regions centered around positions - 3 3 5 , - 4 6 5 , - 6 5 0 , and - 9 4 5 are present

constitutively, i.e. irrespective of the state of activity of the gene (Fig. 4A). One additional hypersensitive site appeared in light-induced leaves only (Fig. 4B) and was m a p p e d around position -190. In general, we observed that the DNAseI-hypersensitive sites become less pronounced in potentially active genes (dark-grown leaves) and did not reliably separate from each other in inactive genes (root tissue). For this reason we shall directly compare the chromatin structure of active and inactive genes in leaf tissue only. Prerequisites for the reproducible detection of these hypersensitive sites were probes labelled to high specific activity, large amounts of D N A per lane, and long exposure times of the blots. Our promoter probe, spanning from the transcription start site to - 2 7 0 , was cloned into M13mp8 and singlestranded fragments were synthesised via primer ex-

567

Fig. 4A. (top) DNAseI hypersensitive sites in the promoter re-

c:

~ \

Lm D ........

iv

zzz

zr z

\

tension and used for indirect end-labelling [45]. Chromatin from root tissue was digested with reduced amounts o f DNAseI because o f its higher sensitivity towards the enzyme. Samples which showed about the same extent of digestion as those from leaf tissues, as evaluated by EtBr patterns, were then selected for the following experiments. Hypersensitive sites generated by mild DNAseI digestion o f nuclei are not the consequence of sequence-specific cuts but rather a feature o f chromatin structure. Digestion of purified genomic D N A ('naked' DNA) did not reveal any low molecular weight fragments when probed for either the upstream or the downstream region of the gene. Only

gion of rbcS genes 3.6 and 3A (1.1 and 1.4 kb fragments, respectively). DNA (30 #g) was electrophoresed in 1.5% agarose gels, blotted, and hybridized to an Ml3-generated probe. Separate gels were run for samples from light- and dark-grown leaves, and for roots and "naked" DNA, respectively. Arrowheads mark regions o f DNAseI hypersensitivity, the open arrowhead symbolizes the

photoinducible site. For reasons of sequence comparison (see Fig. 7) the positions of DNAseI-hypersensitivesites are marked relative to the transcription start site. B 0eft) Densitometricscans of lanes 4 of both L (light) and D (dark) autoradiograms in A. The arrow indicates the position of the light-inducible site.

the autoradiogram of the promoter is shown (Fig. 4A).

Hypersensitive regions 3' to transcription initiation Five DNAseI hypersensitive sites were detected using an M13-generated probe spanning from +37 to +255 (Fig. 5). According to sequence data of the different rbcS genes [18, 40] these hypersensitive sites map to areas within exons as well as introns. All sites are present in all three different tissues although there seem to be tissue-specific differences in the in-

568

.....\

Fig. 5. DNAseI-hypersensitive sites downstream from the tran-

.30

scription start site of the pea rbcS gene family. For explanation, consult legends of Fig. 2 and 4. Major and minor sites are symbolized by the size of the arrowheads. Figures indicate the size of the fragments generated by Eco RV cleavage at +37.

DH$ B

.78

.60 .49 .ZJ

~... ~

tensity of particular signals. In active chromatin (light-grown leaves) one major and several minor sites are present and the major site maps further downstream than in the inactive chromatin from dark-grown leaves (as exemplified in Fig. 5B). In contrast, all sites are about equally intense in chromatin from root tissue (inactive rbcS genes).

Cytosine methylation of 5' regulatory sequences The state of cytosine methylation of the 5' region for

two of rbcS genes was assessed by mapping cleavage by methyl-sensitive nucleases within high molecular weight DNA, isolated from light- and dark-grown leaves and roots of pea. We hybridized against the entire upstream sequence of rbcS 3.6 (from - 1 to -973). In contrast to our experiments with DNAseI this longer probe hybridizes against rbcS 3.6 and rbcS E9 (see Eco RI digest in Fig. 6) as the homology further upstream is higher for these two gene copies than it is for rbcS 3.6 and 3A. A variety of different methyl-sensitive restriction endonucleases were tested (Alu I, Fnu 4HI, ttae III, Sau 3 AI, and Sau 96I)

569

Fig. 6. Methylation of cytosine residues within the promoter region of pea rbcS genes 3.6 and E9. Total genomic DNA from light-grown (L) and dark-grown (D) leaves and roots (R) was digested with methyl-sensitive restriction endonucleases, 20 #g DNA/lane was separated in 207oagarose gels and hybridized to the indicated probe. An Eco RI (RI) digest of genomic DNA was included to ascertain specific hybridization to the two rbcS genes. Arrowheads: fragments arising from cleavage at restriction sites in the promoter region of rbcS 3.6 as indicated in the upper panel. M" Marker, Pst I digest of wild-type h-DNA. A: Alu I; F: Fnu 4H1; S: Sau 3AI. C: + Methylated cytosine preventing nucleolytic cleavage.

[26]. A d o u b l e digest o f g e n o m i c D N A with A l u I a n d E c o RV was also i n c l u d e d to o b t a i n f r a g m e n t s o f the p r o m o t e r size only. N o difference was o b s e r v e d in the m e t h y l a t i o n p a t t e r n o f D N A f r o m the different tissues. B a n d s n o t visible o n t h e a u t o r a d i o g r a m s in Fig. 6 showed u p after l o n g e r e x p o s u r e (e.g. A l u I, lower two bands).

Discussion We have used sequences o f the rbcS 3.6 gene f r o m p e a to l o o k for c h a n g e s in its c h r o m a t i n structure c o n c o m i t a n t with gene expression. O n e i n d i c a t i o n

o f such a c h a n g e is the increased sensitivity o f a gene t o w a r d s D N A s e I after gene a c t i v a t i o n as it has b e e n f o u n d in several a n i m a l a n d few p l a n t systems [11, 25, 44]. A n o t h e r feature o f the c h r o m a t i n o f active genes is the a p p e a r a n c e o f D N A s e I - h y p e r s e n s i t i v e sites, regions o f 5 0 - 2 0 0 b p o f greatly increased D N A s e I sensitivity a n d p o t e n t i a l address sites for r e g u l a t o r y proteins. O u r results present evidence for b o t h the increase o f general D N A s e I sensitivity a n d the a p p e a r a n c e o f D N A s e I - h y p e r s e n s i t i v e sites in rbcS genes d u r i n g o r after their activation. First, there is a distinct p a t t e r n o f broad D N A s e I sensitivity a m o n g the five gene copies in one a n d the s a m e tissue. D o w n s t r e a m , fragm e n t s 2.5 a n d 11.0 kb reveal high sensitivity ( + + +

570 in Tab. 1), fragments 1.2 and 2.1 kb intermediate (+ +), and fragment 1.9 kb lowest sensitivity (+). According to sequence data and size comparison in an Eco RI digest o f D N A from DNAseI-treated nuclei (data not shown) we could assign fragments 2.1 and 1.9 kb to genes rbcS 3.6 and E9, respectively. The remaining three fragments of high and intermediate sensitivity can then be correlated to the high and intermediate transcriptional activity of rbcS 3A, 3C, and 8.0, respectively [18]. These three clones are parenthesized in Fig. 3 and in the s u m m a r y of data given in Table 1. The positive correlation between transcriptional activity and general DNAseI sensitivity is even more pronounced in the promoter region (genes rbcS 3.6 and 3A). However, since this pattern o f broad sensitivity is retained in all three tissues tested it rather seems to reflect their potential to become activated than their actual state of activity. The only possible exception might be rbcS E9 (fragment 1.9 kb), which seems to become somewhat more susceptible in root tissue. We cannot decide whether this enhanced sensitivity is a feature of the gene itself or due to tissue-specific factors since bulk chromatin o f root tissue proved to be more easily digested by DNAseI than chromatin of leaf tissue. Second, after light induction each o f the five rbcS genes becomes slightly more accessible to DNAseI, with the difference between the inactive and the active state being minimal in one of the least active gene (rbcS E9) and maximal in rbcS 3A and 3C, the most active ones. Interestingly, the most extreme changes occur in the promoter region of the most active gene (rbcS 3A). In the rbcS gene family two types of DNAseIhypersensitive sites (DHS) are present: several constitutive sites both upstream and downstream of the transcription initiation site and one inducible site

around position -190. 'Constitutive sites' are regions that are highly acccessible to DNAseI whether the adjacent sequences are transcribed or not. These constitutive regions consequently appear in chromatin of both dark- and light-grown leaves as well as in roots. Nevertheless, during transition from the inactive to the active state of rbcS genes changes in signal intensity of certain constitutive sites might reflect an influence of photoinduction (see Fig. 4 and 5). We also observed constitutive D H S to become increasingly 'blurred' and hardly to distinguish in inactive tissues, especially in roots. The hypersensitive site around position - 1 9 0 in rbcS 3A and/or 3.6 appears only after light induction and is therefore referred to as 'inducible site'. Constitutive and inducible D H S have also been found in other animal systems [19, 48] and in the maize Adh 1 gene [34]. Especially the general architecture o f proximal inducible and distal constitutive sites in the promoter region of the Adh 1 gene matches well with the situation we found for the rbcS genes. In the Adh 1 gene a constitutive region encircling 6 hypersensitive sites is located upstream ( - 160 to - 700) of two inducible sites mapping at - 38 and at - 1 0 0 including TATAA and CAAT boxes. Anaerobic stress leads to the activation of the Adh 1 gene, the occurrence of the inducible D H S and an increase in the intensity of constitutive DHS. The inducible D H S we observed in rbcS genes is located in a region that is most important for the regulation of the rbcS genes. For both rbcS 3A and E9 a region up to about - 3 5 0 has been reported to be sufficient to confer light responsiveness as well as tissue specificity to reporter genes in transgenic plants [17, 28, 32]. Several light regulatory elements (LREs) between - 1 6 8 and -110 could be identified [27] which are conserved in all rbcS genes of pea

Table 1. Correlation between general DNAsel sensitivity and transcription potential of the individual members of rbcS gene family.

Relative abundance of rbcS transcripts [18] General DNAseI sensitivity

n.d. = not determined.

rbcS

(8.0)

E9

3.6

(3A)

(3C)

downstream upstream

19°70 + + n.d.

+

++

40o7o +++

34o7o +++

n.d.

+ +

+ + +

n.d.

7O7o

571 -270 I

,

-260

-250

t

I

-240

-230

,

I

-220 I

-210

-200

I

-190

I

I

GTTTAATCCTTCTACTGTTGTTAGTTTTTTCAGTTAGCTTAATGGGCATGTTACACGTGGCATTA . . . . TCCTATTGGTGGCAAATCATA GCACCATCACACATTTAcAcTCTTcACATGAAAAGATAAGATcAGTGAGGTAATATCCACATGGCAcTGTCcTATTGGTG-6-cTTATGATA III"

II * DHS I

-180 i

-170 I

-160 I

-150

-140

I

I

-130 I

-120

-110

I

AGGTTAGGACACACAACTF~TCT~6TGTGGTTAATAT6pCTGCAAAGTTtTCATTT-CAC~ATCT AGGCTAGCACACAAAA--~TTCAAIATCTI~GTGTGGTTAATATC~GCTGCAAACTTIlATCATTTTCAClTATCT

I

II

I

3"

rbcS - 3.6 rbcS-

3A

Ill

Fig. 7. Sequences o f the rbcS 3.6 and 3A 5' region around the light-inducible DNAsel-hypersensitive site. Nucleotide sequences are derived from Kuhlemeier et al. [27] and Timko et al. [40] and are aligned for greatest sequence homology beginning at the transcription start site. I, II, III, II*, III*: Light-responsive elements according to Kuhlemeier et al. [27].

DHS I: Centre of the light-inducible DNAseI hypersensitive region.

(LRE box II even in petunia, tobacco and soybean). Boxes II and I I I are reiterated further upstream between position - 250 and - 210. The locations of the different LREs for rbcS 3A and 3.6 are indicated in Fig. 7 (I, II, III, II*, III*). The centre of the photoinducible site reported here lies in between two sets of these LREs. The nature and function of the photoinducible D H S will have to be further investigated. The observation that LREs II and III function negatively but only in concert with an additional enhancer and that additional positively acting elements are present in their neighbourhood stresses the highly complex nature of light-regulated gene expression. Initial studies concerning the role o f transacting fators [20] revealed the different LREs of the pea rbcS genes to be possible binding sites for one or more nuclear proteins. Binding of such proteins may then alter the adjacent chromatin architecture and in turn the susceptibility to DNAseI. There is still much debate as to whether rnethylation of cytosine residues in promoters or other parts of a gene has anything to do with gene control [3, 10, 14]. In higher plants about 30% o f all cytosine residues at C-X-G and C-G sequences are usually methylated as compared to only 2 - 8% of total cytosines appearing as 5-methylcytosine in animals [21]. Such a high degree of methylation could well influence transcription of plant genes. Actually demethylation of nopaline synthase gene in the flax crown gall t u m o u r cell line FT 37/1 with 5-

azacytidine in vivo leads to a drastic activation of the otherwise very faintly active gene [22]. Experiments with other systems support the hypothesis that actively transcribed genes are hypomethylated as compared to inactive genes [1, 10, 23]. There are exceptions o f this rule; for example, no methylation of cytosine residues was observed within a 900 bp promoter fragment of maize Adh 1 gene which is totally suppressed [33]. In this particular case genomic sequencing [7] has been used for the detection of cytosine methylation. Most other reports on C methylation used Msp I/Hpa II restriction patterns to detect changes in cytosine methylation in inactive vs. active genes. However, Msp I/Hpa II sites represent only a small fraction of all methylable sequences. Cytosines at strategic positions might thus have been missed in these experiments. Therefore we have used a series of other enzymes known to be methylsensitive (rbcS 3.6 does neither contain any Msp I/Hpa II nor Hha I sites). There was no difference in methylation of cytosine residues in the promoter of rbcS genes in their inactive, potentially active or activated state. All sites recognized by the restriction endonucleases used here have been cut and are therefore not methylated, even in the case of one o f the least active genes, rbcS 3.6. We conclude that methylation at these sites in the 1.0 kb promoter fragment does not play a role in rbcS gene regulation.

572

Acknowledgements This work was supported by a grant from the Fritz Thyssen Stiftung (Cologne). Andrea G6rz is a recipient of a fellowship from the FAZIT-Stiftung (Frankfurt), and Eiji Hirasawa acknolwedges a fellowship from the Alexander yon Humboldt-Stiftung (Bonn). We thank Mr S. Hall for inspiration, Mrs S. Kost for the graphics, and Professor H. Bohnert (Tucson, USA) for providing us with the clones.

References 1. Amasino RM, Powell ALT, Gordon MP: Changes in T-DNA methylation and expression are associated with phenotypic variation and plant regeneration in a crown gall tumor line. Mol Gen Genet 197:437-446 (1984). 2. Bedbrook J: A plant nuclear DNA preparation procedure. PMB Newsl. 2:24 (1981). 3. Bird AP: DNA methylation - How important in gene control? Nature 307:503-504 (1984). 4. Broglie R, Coruzzi G, Fraley RT, Rogers SG, Horsch RB, Niedermayer JG, Fink CL, Flick JS, Chua NH: Lightregulated expression of a pea ribulose-l,5-bisphosphate carboxylase small subunit gene in transformed plant cells. Science 224:838-843 (1984). 5. Cartwright IL, Elgin SCR: Chromatin of active and inactive genes. In: Kahl G (ed.) Architecture of Eurkaryotic Genes, pp. 283-300. Verlag Chemie, Weinheim (1988). 6. Cashmore AR: Nuclear genes encoding the small subunit of ribulose-l,5-bisphosphate carboxylase. In: Kosuge T, Meredith CP, Hollaender A (eds) Genetic Engineering of Plants, pp. 29-38. Plenum Press, New York/London (1983). 7. Church GM, Gilbert W: Genomic sequencing. Proc Natl Acad Sci USA 81:1991-1995 (1984). 8. Coates D, Taliercio EW, Gelvin SB: Chromatin structure of integrated T-DNA in crown gall tumors. Plant Mol Biol 8: 159-168 (1987). 9. Coruzzi G, Broglie R, Edwards C, Chua NH: Tissue specific and light-regulated expression of a pea nuclear gene encoding the small subunit of ribulose-l,5-bisphosphate carboxylase. EMBO J 3 : 1 6 7 1 - 1 6 7 9 (1984). 10. Doerfler W, Langner K-D, Knebel D, Hoeveler A, Mfiller U, Lichtenberg U, Weisshaar B, Renz D: Eukaryotic gene inactivation by sequence-specificpromoter methylation and the release of the transcription block. In: Kahl G (ed.) Architecture of Eurkaryotic Genes, pp. ???-???. Verlag Chemie, Weinheim (1988). ll. Eissenberg JC, Cartwright 1L, Thomas GH, Elgin SCR: Selected topics in chromatin structure. Ann Rev Genet 19: 485-536 0985).

12. Elgin SCR: Anatomy of hypersensitive sites. Nature 309: 213-214 (1984). 13. Feinberg AP, Vogelstein B: A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 132:6-13 (1983). 14. Felsenfeld G, McGhee J: Methylation and gene control. Nature 295:602-603 (1982). 15. Ferl R J: Modulation of chromatin structure in the regulation of the maize A d h 1 gene. Mol Gen Genet 200:207-210 (1985). 16. Fluhr R, Chua NH: Developmental regulation of two genes encoding ribulose-bisphosphatecarboxylase small subunit in pea and transgenic petunia plants: Phytochrome response and blue-light induction. Proc Natl Acad Sci USA 83: 2358-2362 (1986). 17. Fluhr R, Kuhlemeier C, Nagy F, Chua NH: Organ-specific and light-induced expression of plant genes. Science 232: 1106-1112 (1986). 18. Fluhr R, Moses P, Morelli G, Coruzzi G, Chua NH: Expression dynamics of the pea rbcS multigene family and organ distribution of the transcripts. EMBO J 5:2063 -2071 (1986). 19. Fritton HP, Igo-Kemenes T, Nowock M J, Strech-Jurk U, Theisen M, Sippel AE: Alternative sets of DNAse I hypersensitive sites characterize the various functional states of the chicken lysozyme gene. Nature 311:163-165 (1984). 20. Green PJ, Kay SA, Chua NH: Sequence-specific interations of a pea nuclear factor with light-responsive elements upstream of the rbcS-3A gene. EMBO J 6:2543-2549 (1987). 21. Gruenbaum Y, Naveh-Many T, Cedar H, Razin A: Sequence specificity of methylation in higher plant DNA. Nature 292: 860-862 (1981). 22. Hepburn AG, Clarke LE, Pearson L, White J: The role of cytosine methylation in the control of nopaline synthase gene expression in a plant tumor. J Mol Appl Genet 2:315-329 (1983). 23. Hepburn AG: DNA methylation and its effect on transformed gene expression. In: Kahl G (ed.) Architecture of Eukhryotic Genes, pp. 419-433. Verlag Chemie, Weinheim (1988). 24. Jones JDG, Dunsmuir P, Bedbrook J: High level expression of introduced chimaeric genes in regenerated transformed plants. EMBO J 4:2411-2418 (1985). 25. Kahl G, G6rz A, Weising K, Schfifer W, Hirasawa E: Chromatin structure of plant genes. In: Kahl G (ed.) Architecture of Eurkaryotic Genes, pp. 301 - 322. Verlag Chemie, Weinheim (1988). 26. Kessler C, Neumaier TS, Wolf W: Recognition sequences of restriction endonucleases and methylases - a review. Gene 33: 1-102 (1985). 27. Kuhlemeier C, Fluhr R, Green PJ, Chua NH: Sequences in the pea rbcS-3A gene have homology to constitutive mammalian enhancers but function as negative regulatory elements. Genes and Devel 1:247-255 (1987). 28. Kuhlemeier C, Green PJ, Chua NH: Regulation of gene expression in higher plants. Ann Rev Plant Physio138: 221-237 (1987).

573 29. Mandoli DF, Briggs WR: Phytochrome control of two lowirradiance responses in etiolated oat seedlings. Plant Physiol 67:733-739 (1981). 30. Maniatis T, Fritsch EF, Sambrook J: Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY (1982). 31. Murray MG, Kennard WC: Altered chromatin conformation of the higher plant gene phaseolin. Biochemistry 23: 4225-4232 (1984). 32. Nagy F, Morelli G, Fraley RT, Rogers SG, Chua NH: Photoregulated expression of pea rbcS gene in leaves of transgenic plants. EMBO J 4: 3063-3068 (1985). 33. Nick H, Bowen B, Ferl JR, Gilbert W: Detection of cytosine methylation in the maize alcohol dehydrogenase gene by genomic sequencing. Nature 319:243-246 (1986). 34. Paul AL, Vasil V, Vasil IK, Ferl RJ: Constitutive and anaerobically induced DNAseI hypersensitive sites in the 5' region of the maize Adh 1 gene. Proc Natl Acad Sci USA 84: 799- 803 (1987). 35. Reeves R: Transcriptionally active chromatin. Biochim Biophys Acta 782:343-393 (1984). 36. Simpson J, Van Montagu M, Herrera-Estrella L: Photosynthesis-associated gene families: difference in response to tissue-specific and environmental factors. Science 233:34-38 (1986). 37. Spiker S, Murray MG, Thompson WF: DNAseI sensitivity of transcriptionally active genes in intact nuclei and isolated chromatin of plants. Proc Natl Acad Sci USA 80:815-819 (1983). 38. Steinmiiller K, Batschauer A, Apel K: Tissue-specific and light-dependent changes of chromatin organization in barley (Hordeum vulgare). Eur J Biochem 158:519-525 0986). 39. Thomas GH, Siegfried E, Elgin SCR: DNAse! hypersensitive sites: a structural feature of chromatin associated with gene expression. In: Reeck G, Goodwin G, Puigdomenech P (eds)

40.

41.

42.

43.

44.

45.

46.

47.

48.

Chromosomal Proteins and Gene Expression, pp. 77-101. Plenum Press, New York (1986). Timko MP, Kausch AP, Hand JM, Cashmore AR, HerreraEstrella L, Van den Broeck G, Van Montagu M: Structure and expression of nuclear genes encoding polypeptides of the photosynthetic apparatus. In: Steinback KE, Bonitz S, Arntzen C J, Bogorad L (eds) Molecular Biology of the Photosynthetic Apparatus, pp. 381- 394. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY (1986). Tobin EM, Silverthorne J: Light regulation of gene expression in higher plants. Ann Rev Plant Physiol 36:569-593 (1985). Vayda ME, Freeling M: Insertion of the Mul transposable element into the first intron of maize Adh 1 interferes with transcription elongation but does not disrupt chromatin structure. Plant Mol Biol 6:441-454 (1986). Watson JC, Kaufmann LS, Thompson WF: Developmental regulation of cytosine methylation in the nuclear ribosomal genes of Pisum sativum. J Mol Biol 193:15-26 (1987). Weintraub H: A dominant role for DNA secondary structure in forming hypersensitive structures in chromatin. Cell 32:1191-1203 (1983). Wu C: The 5' ends of Drosophila heat shock genes in chromatin are hypersensitive to DNAsel. Nature 286: 854-860 (1980). Wu C: Two protein-binding sites in chromatin implicated in the activation of heat-shock genes. Nature 309:229-234 (1984). Wurtzel ET, Burr FA, Burr B: DNAseI hypersensitivity and expression of the Shrunken-I gene of maize. Plant Mol Biol 8:251-264 (1987). Zaret KS, Yamamoto KR: Reversible and persistent changes in chromatin structure accompany activation of a glucocorticoid-dependentenhancer element. Cell 38: 29-38 (1984).