light-regulated and cell-specific expression of rbcs- gusa ...the gusa gene directed by the tomato...

Plant Physiol. (1993) 102: 991-1000

Light-Regulated and Cell-Specific Expression of Tomato rbcS- gusA and Rice rbcS-gusA Fusion Genes in Transgenic Rice’

Junko Kyozuka*, David McElroy’, Takahiko Hayakawa, Yong Xie3, Ray Wu, and Ko Shimamoto

Plantech Research Institute, 1 O00 Kamoshida, Midori-ku, Yokohama 227, Japan (J.K., T.H., K.S.); and Field of Botany (D.M., R.W.) and Section of Biochemistry, Molecular and Cell Biology (Y.X., R.W.), Cornell University,

Ithaca, New York 14853

~~

A previously isolated rice (Oryza sativa) rbcS gene was further characterized. This analysis revealed specific sequences in the 5‘ regulatory region of the rice rbcS gene that are conserved in rbcS genes of other monocotyledonous species. In transgenic rice plants, we examined the expression of the 8-glucuronidase (gusA) reporter gene directed by the 2.8-kbpromoter region of the rice rbcS gene. To examine differences in the regulation of monocotyledonous and dicotyledonous rbcS promoters, the activity of a tomato rbcS promoter was also investigated in transgenic rice plants. Our results indicated that both rice and tomato rbcS promoters confer mesophyll-specific expression of the gusA reporter gene in transgenic rice plants and that this expression is induced by light. However, the expression level of the rice rbcS-gusA gene was higher than that of the tomato rbcS-gusA gene, suggesting the presence of quantitative differences in the activity of these partic- ular monocotyledonous and dicotyledonous rbcS promoters in transgenic rice. Histochemical analysis of rbcS-gusA gene expression showed that the observed light induction was only found in mesophyll cells. Furthermore, i t was demonstrated that the light regulation of rice rbcS-gusA gene expression was primarily at the level of mRNA accumulation. We show that the rice rbcS gene promoter should be useful for expression of agronomically important genes for genetic engineering of monocotyledonous species.

Expression of the Rubisco small subunit genes (rbcS), which are regulated both developmentally and by light (for review,

This research was supported, in part, by grants RF86058, Allo- cation No. 60, and RF90031, Allocation No. 126, from the Rockefeller Foundation. D.M. was supported by a Science and Engineering Research Council/North Atlantic Treaty Organization studentship, a Fulbright Fellowship, scholarships from the British Universities North America Club and the St. Andrew’s Society of Washington, DC, and predoctoral fellowships from the Comell University Bio- technology Program and the Comell National Science Foundation (NSF) Plant Saence Center. The Comell NSF Plant Science Center is a unit of the United States Department of Agriculture-Department of Energy-National Science Foundation Plant Science Center Pro- gram and a unit of the Comell Biotechnology Program, which is sponsored by the New York State Science and Technology Founda- tion, a consortium of industries, and the U.S. Army Research Office.

Present address: Commonwealth Scientific and Industrial Re- search Organization, Division of Plant Industry, GPO Box 1600, Canberra, ACT 2610, Australia.

Present address: The Rockefeller University, 1230 York Avenue, New York, NY 10021.

* Corresponding author; fax 81-45-962-7492. 991

see Tobin and Silverthorne, 1985; Edwards and Coruzzi, 1990; Gilmartin et al., 1990), is one of the best models for studies of gene expression in plants (for review, see Gilmartin et al., 1990). However, compared with dicotyledonous rbcS genes, the regulation of rbcS genes in monocotyledonous species is not well understood. Part of the reason for this is the lack of a well-defined in vivo assay system. Transgenic monocots have not been available until recently, and to date monocotyledonous rbcS gene expression has been studied in transient assays of transformed mesophyll protoplasts (Schaffner and Scheen, 1991) or with particle-bombarded leaf tissues (Rolfe and Tobin, 1991; Bansal et al., 1992). However, developmental regulation of gene expression is not easily analyzed in such transient assay systems.

In other studies, transgenic tobacco has been used to examine the expression of light-regulated photosynthetic genes from monocotyledonous species (Lloyd et al., 1991; Matsuoka and Sanada, 1991). In these studies, anatomical differences between the leaf tissues of monocots and dicots have pre- vented an accurate evaluation of the observed tissue and cell specificity of gene expression. Moreover, monocotyledonous genes are not always expressed in an expected manner when introduced into transgenic dicotyledonous plants (Lamppa et al., 1985; Keith and Chua, 1986). In contrast, highly regulated expression of reporter genes directed by monocotyledonous promoters in transgenic rice (Oyza sativa) plants (Kyozuka et al., 1991; Tada et al., 1991; Zhang et al., 1991; Terada et al., 1993) has indicated that this in vivo assay system can be used for the analysis of promoters not only of rice genes but also of genes from other monocotyledonous species.

It has been suggested that the molecular mechanisms un- derlying gene expression are not the same for monocots and dicots. The promoters from rbcS genes of monocots, but not of dicots, are active in maize mesophyll protoplasts. In these protoplasts, promoter elements from dicotyledonous rbcS genes failed to replace functionally those from a monocotyledonous rbcS promoter (Schaffner and Scheen, 1991). It has also been found that important regulatory sequences in rbcS promoter regions are not conserved between monocotyledonous and dicotyledonous species. Monocotyledonous rbcS

Abbreviations: GUS, @-glucuronidase; gusA, @-glucuronidase gene; Hm’, hygromycin resistance; hph, hygromycin phosphotransferase gene; MUG, 4-methylumbelliferyl-@-~-glucuronide; nos, nopaline synthase gene; rbcS, gene for small subunit of Rubisco; X-gluc, 5- bromo-4-chloro-3-indolyl-@-glucuronic acid.

Dow

nloaded from https://academ

ic.oup.com/plphys/article/102/3/991/6067012 by guest on 01 August 2021

992 Kyozuka et al. Plant Physiol. Vol. 102, 1993

genes share a monocotyledonous consensus sequence that is not found in dicotyledonous species (Schaffner and Scheen, 1991). In this study, to begin to understand the regulation of rice

rbcS gene expression, the promoter region of a rice rbcS gene was fused to the gusA coding region and the expression of the fusion gene was examined in transgenic rice plants. Furthermore, to gain insights into the different mechanisms Óf gene regulation in dicots and monocots, the expression of the gusA gene directed by the tomato rbcS3C promoter was compared with that of the rice rbcS-gusA fusion gene.

MATERIALS AND METHODS

Sequence Characterization of the Rice rbcS Clone

A 2.8-kb PstI restriction fragment from pRRl containing the 5’ flanking region of the rice (Oyza sativa) rbcS gene (Xie et al., 1987) was cloned into pBluescriptI1-KS (Stratagene) plasmids in both orientations to produce the vectors pRR5‘53P and pRR5’35P for sequence determination.

Cenomic DNA Preparation and Analysis

Rice genomic DNA was isolated from young greenhouse- grown plants (cv IR26) as described by Dellaporta et ai. (1983) with the addition of a phenol-chloroform extraction prior to the final ethanol precipitation.

For DNA gel blot analysis, lO-pg samples of rice genomic DNA were restriction-enzyme digested, fractionated by electrophoresis using 0.9% agarose gels, and transferred to GeneScreen Plus membranes (New England Nuclear). A 0.6- kb PstI-SmaI restriction fragment spanning the rice rbcS coding region was utilized in the production of a 32P-labeled probe (Maniatis et al., 1982). Hybridization was carried out ovemight at 42OC in 5 x SSC, 50% formamide, with a final wash at 65OC in 0.1% SDS and 0.01 X SSC, following the procedure of Bematsky and Tanksley (1986). To estimate the number of integrated gusA genes in each

plant, genomic DNA was digested with EcoRI or HindIII, which have single restriction sites flanking the rbcS-gusA fusion genes in the plasmid, and the digested DNA was subjected to DNA gel blot hybridization analysis.

Analysis of the Rice rbcS RNA

Shoot tissue was collected 6 d after gennination from etiolated rice seedlings either before or after a 2-h light treatment. One milliliter of extraction buffer (100 m~ sodium acetate, 1% SDS, pH 5.0) was added per gram of tissue, the samples were extracted twice with an equal volume of pheno1:chloroform:isoamyl alcohol (25:24: I), and KCl was added to the aqueous phase to a final concentration of 0.3 M. After centrifugation, the supematant was transferred to a new tube and the RNA was precipitated with LiCl as de- saibed by Wallace (1987).

For RNA gel blot analysis 25-pg samples of rice total RNA were glyoxylated and fractionated by electrophoresis on 0.9% agarose gels and transferred to GeneScreen Plus membranes before hybridization with 0.6-kb PstI-SmaI rice rbcS coding

region probe. RNA gel blot hybridization was carried oiut for 48 h at 48OC i11 5 X SSC, 50% formamide, with a final wash at 65OC in 0.1% SDS and 0.1 X SSC, following the procedure of Broglie et al.. (1983).

Analysis of gusA RNA

RNA was isolated by a published method (Chomczynski and Sacchi, 1987) from young leaves of dark-adapted mature & transgenic plants. For the dark adaptation, plants were kept in the darlk for 4 d. Forty micrograms of total RNA were separated by lelectrophoresis on a 1.0% agarose gel and transferred to EI Hybond-N membrane (Amersham). Hybrid- ization with the gusA gene was camed out according to the manufacturer’s protocol ( Amersham) .

Rice rbcS-gusAi Chimeric Cene Construction

A 1.8-kb BamHI-SstI restriction fragment containing the Escherichia coli gusA coding region was isolated froni the plasmid pActl-F (McElroy et al., 1990) and cloned into the vector pNOS7 2, a pSP72 (Promega) derivative containing the transcription termination region of Agrobacterium tume- faciens nos, to create the gusA-nos-containing conetruct pGN72. Replacement of the 0.4-kb PstI-SnaBI restriction fragment of pCN72 with the equivalent 0.4-kb PsfI-:?naBI restriction f r a p e n t of pRAJ275 (Clontech), which coritains a eukaryotic translation initiation consensus sequence, resulted in the creation of the gusA-nos construct pGN73. Cloning of the 2.8-kb PstI restriction fragment from pRR1, containing the 5’ transcribed and non-transcribed region of the rice rbcS gene, into the PstI site of pNG73 resulted in the creation of the rbcS-gusA-nos-containing pRGN73.

Construction of the Tomato rbcS-gusA Chimeric Cene

A 1.7-kb fragpent containing the entire promoter region of the tomato rbcS3C (Sugita et al., 1987) was polymerase chain reaction-amplified with the primers 5’-CAGGPAA- CAGCTATGAC-3’, corresponding to a part of pUC plarimid, and 5’-GCTC’l’AGAAATAATTGGTTACTAAGA-3’, corresponding to the transcription start site of the rbcS3C gene. The latter primer contained an XbaI site at position +14 relative to the transcription initiation site of the tomato rbcS3C. A HindIII-XbaI fragment coqtaining 1.7 kb of 5’ transcribed and non-transcribed region isolated fromi the amplified fragment was replaced with the HindIII-XbaI fragment of pIG221 (Ohta et al., 1990), resulting in pTRIGN, consisting of the 1.7-kb tomato rbcS3C promoter, the gusA coding region carrying the first intron of castor bean catalase . 1 gene, and the nos polyadenylation region. Translation of the gusA mRNA of pTRIGN starts at the ATG derived from the castor bean catalase 1 gene.

~

Production of Transgenic Rice Plants

Protoplasts were isolated from embryogenic susperision cultures of rice (cv Nipponbare) according to Kyozuka et al. (1987). Electroporation, selection of Hm’ calli, and plant regeneration were carried out as described previously ~(Shi- mamoto et al., 1989). The rbcS-gusA chimeric genes ‘were

Dow



rbcS Expression in Transgenic Rice 993

introduced into rice protoplasts by cotransformation with a hygromycin-resistance gene, hph. Hm’ calli were selected, and plants regenerated from them were screened by staining the shoots with X-gluc.

Analysis of GUS Enzyme Activity

GUS activity was analyzed by both histochemical analysis with X-gluc and fluorometric analysis with MUG as described previously (Kyozuka et al., 1991). Incubation of tissue samples with X-gluc was canied out for 18 h at 37OC except for leaf sections of transgenic plants carrying the rice rbcS-gusA gene, which were incubated for 3 h at 37OC. Statistical analysis of the GUS specific activity results was done using the StatView SE+ program (Abacus Concepts Inc., Berkeley, CA).

Dark-grown seedlings (6- to 7-d-old) were used to examine distribution of GUS activity in the leaf tissue induced by light. The upper part of a seedling leaf was cut into sections and incubated with X-gluc. Then the rest of the seedling was grown further under light conditions (16 h light, 8 h dark). After 24,48, or 96 h of light treatment, sections were prepared from the seedling leaves and stained with X-gluc.

RESULTS

Sequence Characterization of the Rice rbcS Cene

Partia1 restriction maps and the complete nucleotide and derived amino acid sequence of the rice rbcS gene are shown in Figure 1, A and 8, respectively. Sequence analysis revealed that the rbcS sequence of pRRl showed 98.8% homology with the rice rbcS cDNA clone OSRUBPCl (Matsuoka et al., 1988). The rice rbcS genomic clone is predicted to code for a processed transcript of at least 830 nucleotides. The potential coding region of the rbcS processed transcript, if translated in vivo, would code for 175 amino aads and a precursor rbcS protein of 19.6 kD. Remova1 of a putative NH2-terminal transit peptide by cleavage between the Cys (residue 47) and Met (residue 48) would yield a mature rbcS protein of 128 amino acids with a 14.7-kD estimated molecular mass.

An inspection of the region 5’ of the putative TATA box of the rice rbcS gene identified sequences (Fig. 1C) that have previously been found to be conserved among monocotyledonous rbcS genes but that are absent from dicotyledonous rbcS promoters (Schaffner and Scheen, 1991). Analysis of the 3’ noncoding region of the rice rbcS gene revealed the pres- ente of a GT-rich sequence downstream of a putative polyadenylation region (Fig. 1B). Such GT-rich sequences have been reported to be required for effiaent gene expression in plant cells (Ingelbrecht et al., 1989). The 3’ noncoding region of the rice rbcS gene was also found to contain a 245- nucleotide sequence that represents an imperfect repeat of part of the second exon of the rbcS coding region (Fig. 1A). Alignment of the repeated sequences shows that they have 75.5% sequence similarity (data not shown).

RNA and DNA Gel Blot Analysis of the Rice rbcS Cene

Prior to canying out RNA gel blot analysis, gel blots of rice genomic DNA digested with EcoRI, BamHI, HindIII, and XhoI

were hybridized with a 32P-labeled probe from the coding region of the rice rbcS gene (Fig. 2A). Because this rice rbcS gene is known to be a member of a gene family (Matsuoka et al., 1988), relatively high hybridization stringency (42OC, 5 X SSC, 50% formamide) was used to obtain gene-specific hybridization conditions. As can be seen from Figure 2A, the rice rbcS probe hybridized to only one band from each digestion, confirming that these DNA-DNA hybridization conditions are gene specific.

The steady-state levels of transcripts from the rice rbcS gene under investigation were examined by gel blot hybridization analysis of RNA isolated from 6-d-old etiolated rice seedlings before and after a 2-h light treatment. To ensure that no cross-hybridization occurred between rice rbcS genes, hybridization was carried out at 48OC rather than 42OC to compensate for the increased stability of DNA-RNA hybrids over DNA-DNA hybrids. The abundance of the transcripts from the rice rbcS gene increased by an estimated 30-fold upon illumination of etiolated rice seedlings (Fig. 2B).

Expression of Both Rice and Tomato rbcS-gusA Chimeric Cenes 1s Mesophyll Cell Specific in Transgenic Rice Plants

We chose the promoter of the tomato rbcS3C gene (Sugita et al., 1987) to examine whether a dicotyledonous rbcS promoter could confer regulated expression of the gusA gene in transgenic rice. Expression of the tomato rbcS3C gene, which is one of five genes in the tomato rbcS gene family, is known to be strictly light dependent and is detected only in shoots and leaves (Sugita et al., 1987; Wanner and Gruissem, 1991). To construct the tomato rbcS-gusA fusion gene, we inserted the first intron of castor bean catalase 1 gene, which is known to increase the foreign gene expression in transgenic rice (Tanaka et al., 1990), into the NHz-terminal coding sequence of the gusA region (Fig. 3) .

To investigate the activity of the rice and tomato rbcS promoters in transgenic rice, the rbcS-gusA fusion genes (Fig. 3) were introduced into rice. Five transgenic plants carrying pRGN73 (rice rbcS-gusA) and three carrying pTRIGN (tomato rbcS-gusA) were obtained. The number of integrated rbcS- gusA genes in each plant was estimated by DNA gel blot hybridization (Table I). Although six of the seven plants examined were found to contain one rbcS-gusA gene, one plant transformed with pTRIGN was estimated to contain two to three copies of the tomato rbcS-gusA fusion gene.

We analyzed the tissue and cell-type expression specificity of the chimeric gusA genes by histochemical GUS analysis using X-gluc as a substrate. In transgenic rice plants, the rice rbcS promoter directed high-leve1 expression of the gusA gene in leaf blade (Fig. 4A) and sheath (Fig. 48) mesophyll cells. However, in these plants, GUS activity was not observed in leaf epidermal cells, vascular cells, roots (Fig. 4C), flower organs (Fig. 4D), or seeds (Fig. 4E). The results clearly indicate that expression of the rice rbcS-gusA gene is restricted to the mesophyll cells of transgenic rice plants.

A similar pattem of GUS expression was observed in transgenic rice plants containing the tomato rbcS3C-gusA gene. GUS activity was present in mesophyll cells of both leaf blades (Fig. 4F) and leaf sheaths (Fig. 4G), but not in roots (Fig. 4H) or flowers (data not shown). These results

Dow



994 Kyozuka et al. Plant Physiol. Vol. 102, 1993

B

A H.P.Sa.Xb. -7

p R R l

B.Sm,S.E PSaP

I -- 500 bp

l b o x Yonomat rbcS conaensua TATA box C malZe GATAAG GAACGGTG GCC ACT CCACA CTATATATGCCGT rbC-1 - 1 7 9 -174 -105 -92 -811-11 -24 -12

rbcSZm -215 -710 -133 -120 -1111-111 -11 -19

maZe GATAAG GAACGGTG GCC ACT CCACA CTATATATGCCGT

rrhM GATAAG GAACG TGAGCC ACT CCACA CTATATATACCGT rbcs -103 -98 -80 -61 -621-50 -17 -25

rice GAT AG GCACG CG GCCUTC CCACA CTATTTATA CGT -20 -90 -061-82 -1I rbcS -11s -11, -103

Figure 1. A, Restriction map of the pRRl insert. Coding region exons of the r b d gene are depicted as open boxes. The position of the second exon duplication is indicated by a closed box. The region that was subcloned for sequence determination is indicated by a double-headed arrow below the pRRl restriction map. The region that was subcloned for use as a probe in DNA and RNA gel blot hybridization is indicated by a line above the pRRl restriction map. Restriction sites are abbreviated as follows: B, BamHI; C, Bglll; E, EcoRI; R, EcoRV; H, Hindlll; P, Pstl; 5, 5stl; Sa, 5all; Sm, Smal; Xb, Xbal. B, Nucleotide and deduced amino acid sequence of rice rbc!;. An alignment between the nucleotide sequence of the rice rbc5 cDNA clone OSRUBPCl (Matsuoka et al., 1988) and that of the rbc5 genomic clone was used to identify the number and location of coding region exons within the rbc5 sequence. The end points of the rbc5 cDNA sequence are indicated by closed arrows. Differences between the rbc5 genomic and cDNA sequences are indicated above the genomic sequence. The nucleotide sequence is numbered with .tl being the presumptive transcription initiation site. The predicted translation product of the rice rbc5 transcript is indicated below the rbc5 codons. The predicted site of rbc5 transit peptide cleavage is indicated by an open arrow. Lowercase letters represent intron sequence. The exon-intron junctions within the rbc5 transcribed sequence are indicated by slashed lines. The putative TATA box,

Dow




B rice rbcS-gusA

1 2 3 4kbp10-

kktomato rbcS-guaA

1-

Figure 2. DMA gel blot analysis of rice rbcS gene. A, DMA gel blothybridization. Ten micrograms of rice genomic DMA was digestedwith: 1, EcoRI; 2, SamHI; 3, Hindlll; or 4, X/iol. The DMA was thenfractionated by electrophoresis, transferred to nylon membranes,and probed with a 0.6-kb Pstl-Smal restriction fragment spanningthe rice rbcS coding region. B, RNA gel blot hybridization analysisshowing light-inducible increase in rice rbcS transcript abundance.Lane 1, Total RNA (25 ^g) extracted from 6-d-old etiolated riceseedlings 2 h after illumination. Lane 2, Total RNA (25 ̂ g) extractedfrom 6-d-old etiolated rice seedlings before illumination.

show that both rice and tomato rbcS promoters confer me-sophyll-specific gene expression in transgenic rice plants.

To compare the rice and tomato rbcS promoter activities,GUS enzyme levels in leaves (blade and sheath) of maturetransgenic plants were examined by fluorometric analysis,using MUG as a substrate (Table I). Both the rice rbcS-gusAand the tomato rbcS-gusA genes showed no statistically sig-nificant difference between the activity in leaf blades andsheaths (P = 0.05). However, we did calculate a statisticallysignificant difference (P = 0.05) in transgenic rice leaves(blade plus sheath) between the rice and tomato rbcS-gusAfusion genes, with mean activities of 63,351 and 4,277 pmolmethylumbelliferone mg"1 min"1, respectively.

Expression of Both Rice and Tomato rbcS-gusA ChimericGenes Is Light Inducible in Mesophyll Cells of TransgenicRice Plants

The light induction of rbcS-gusA fusion gene expressionwas examined at the cellular level by histochemical GUSanalysis of RI seedlings grown in the dark for 7 d and thenilluminated for 24, 48, or 96 h (Fig. 5). GUS activity wasobserved in the mesophyll cells from young leaves of the

1kb

rice rbcS promoter1st intron ofcatalase gene

@ tomato rbcS3C promoter•• nos terminator

Figure 3. Diagrams of the rbcS-gusA fusion genes used in this study.The different regions of the rbcS-gusA fusion genes are: light stripedbox, 5' noncoding region of the rice rbcS gene; heavy striped box,5' noncoding region of the tomato rbcS3C gene; open box, codingregion of the gusA reporter gene; closed box, 3' transcriptiontermination region of nos; horizontal striped box, first intron ofcastor bean catalase 1 gene.

dark-grown seedlings carrying the rice rbcS-gusA before illu-mination (Fig. 5A), whereas the tomato rbcS-gusA gene didnot show any expression in such dark-grown shoots (Fig.5D). After 24 h of light treatment, induction of rice rbcS-gusAexpression was evident in the mesophyll cells of transgenicrice plants (Fig. 5, B and C). However, no similar inductionwas found in RI seedlings containing the tomato rbcS-gusAgene (data not shown). Induction of tomato rbcS-gusA geneexpression in the mesophyll cells of transgenic rice plantswas first visible after 48 h of illumination (Fig. 5E), andexpression continued for a further 48 h (Fig. 5F). Theseobservations establish that light induction of the gusA genedirected by both rice and tomato rbcS promoters is mesophyllcell specific in transgenic rice plants.

To determine whether the observed light induction of GUSactivity directed by the rbcS-gusA genes is due to an increasein their respective steady-state mRNA levels, total RNA wasisolated from leaves of dark-adapted primary transgenic riceplants before and after light treatment, and RNA gel blotswere hybridized with the gusA probe (Fig. 6). In this analysis,we used a rice rbcS-gusA plant with an intermediate level ofGUS specific activity (line 1) and a tomato rbcS-gusA plant(line N6-4) with the highest GUS specific activity (see TableI). A low level of gusA transcripts was detected in the dark-adapted plant containing the rice rbcS-gusA gene (Fig. 6, lane5). The level of gusA transcripts driven by the rice rbcSpromoter increased upon illumination for 6 h (Fig. 6, lane 6)and slightly decreased after 24 h of illumination (Fig. 6, lane7). We do not know the cause of this decrease in the gusAtranscripts between 6 and 24 h of illumination. In contrast,in a rice plant containing the tomato rbcS-gusA gene, the gusAtranscripts were completely absent in dark-adapted leaves(Fig. 6, lane 2) and became detectable only after 24 h of

(Continued from facing page.) polyadenylation site (PolyA) and CT consensus sequence (CT Cons) are indicated andunderlined. Restriction sites used in the subcloning and sequencing of the rbcS genomic clone are indicated andunderlined. C, Identification of putative rice rbcS regulatory elements. Sequence homologies found between maize(rbcZm! and rbcSZm3), wheat, and rice rbcS promoters are shown. Gaps have been introduced to maximize homologies.The numbers below the putative regulatory elements are relative to the presumed rbcS transcription initiation site ineach gene.

Dow



996 Kyozuka et ai. Plant Physiol. Vol. 102, 1993

Table 1. CUS specific activity of rice and tomato rbcS-gusA fusion genes in transgenic rice __

Paired t Test, I3lade vs. Sheathb

Modified Unpaired t Test, Rice vs. Tomato rbc5’

GUS Specific Activity Leaf Sheath

Fusion Gene Plant Copy No.” Leaf Blade -

Value Mean Value Mean d.f.’ t“ Poose Means d.f.h t’ Po0,’

Rice rbcS-gusA 1 1 32,939 37,274 2-2 1 34,239 24,65 1 2-3 1 31,786 15,925

3-1 1 151,750 52,090 271,626 74,611 4 -0.91 2.77 63,351 2-9 1 9,736 23,582

Tomato rbcS-gusA N 1 NT 2,969 2,965 N6-4 2-3 17,502 1,015 N6-7 1 834 7,101 378 4,277 2 1.04 4.30 4,277 9 2.22 2.26

a Estimated copy number of rbcS-gusA fusion genes from DNA gel blot hybridization analysis (data not shown). Testing the hypothesis Degrees of

Testing the hypotheris that 8 Mean

Calculated Tabulated value of t for P == 0.05.

that the mean GUS specific activities in leaf blade and leaf sheath are equal for the rice o r tomato rbcS-gusA fusion genes. freedom (n - 1 ) . the mean GUS specific activities in leaves (blade and sheath) expressing either rice or toimato rbcS-gusA fusion genes are equal. of leaf blade and leaf sheath GUS specific activity in transgenic plants expressing either rice or tomato rbcS-gusA fusion genes. degrees of freedom for a modified unpaired t test.

Calculated value of t, assuming equal variance. e Tabulated value of t for P = 0.05.

’ Calculated value of t, assuming unequal variance.

illumination (Fig. 6, lane 4). The presence of approximately equivalent amounts of intact RNA in each lane was confirmed by detecting comparable levels at various time points of hph transcripts that are driven by a constitutively expressed cauliflower mosaic virus 35s promoter (data not shown). These results indicate that, in transgenic rice plants, the expression of both the rice and tomato rbcS-gusA genes is regulated by light at the level of mRNA abundance, although the rate and length of time required for induction appear to be distinct between the two.rbcS genes.

DISCUSSION

In this work, we describe further characterization of a previously isolated rice rbcS gene (Xie et al., 1987; Xie and Wu, 1988). The 2.8-kb promoter region of the rice rbcS gene was fused to the gusA reporter gene and transferred into rice. For comparison, we also examined the promoter activity of a dicotyledonous rbcS gene, the tomato rbcS3C, in transgenic rice plants. Our analysis indicated that both rice and tomato rbcS promoters confer light-regulated and mesophyll cell- specific expression of the gusA gene in transgenic rice plants. However, our results suggest that the activity of the tomato rbcS promoter is at least 1 order of magnitude lower than that of the rice rbcS promoter.

A comparison between the rice rbcS coding region sequence presented in Figure 18 and that of the previously published sequence (Xie and Wu, 1988) revealed a number of single- nucleotide differences that result in shifts between their respective reading frames. The comparison of their predicted translation products with that of the wheat (Broglie et al., 1983) and maize (Lebrum et al., 1987) rbcS sequences clearly

indicate that a11 the differences between the two rice rbcS genomic sequences are due to errors in the previously published sequence (data not shown).

Analysis of the 3’ noncoding region of the rice rbcS gene allowed us to identify a region bearing partial homology to part of the second exon of the rice rbcS gene. We do not believe that the sequence from the 3’ end of the rbcS geinomic clone codes for any partial rbcS protein product, sinlce its potential reading frame is interrupted by numerous niucleo- tide additions and deletions. We propose that the original duplication event that gave rise to the 3’-end rbcS repeat may have resulted from an illegitimate recombination eveiit between two rice rbcS genes. However, until more rice rbcS genes are sequenced, we cannot te11 if the source of the original duplication was the rice rbcS gene described here. The full extent of the duplication could not be investigated because the pRRl genomic clone insert terminates at ihe 3’ end of the rbcS repeat.

By RNA gel blot analysis, we were able to show that the rice rbcS gene codes for a transcript whose abundance increases over 30-fold upon the illumination of etiolated rice seedlings. This is in good correspondence with the 0b.j I erva- tion that GUS activity is present at a detectable level in dark- grown seedlings of transgenic rice plants containing the rice rbcS-gusA fusion gene and increases rapidly upon light treatment. These results indicate that the light-responsive activa- tion of rice rbcS gene expression occurs largely at the level of transcription .

Regulated gusA expression from the tomato rbcS3C promoter in transgenic rice plants suggests that the molecular mechanisms essential for the qualitative expression of photosynthetic genes are conserved between rice and tomato

Figure 4 (on facing page). Histochemical localization of rice (A-E) and tomato (F-H) rbcS-gusA gene expression in transgenic rice plants. A and F, Leaf blade transverse section. Abbreviations in A: bc, bulliform cells; bs, bundle sheath cells; ep, epidermal cells; m, mesophyll cells; p, phloem; x, xylem. B and G, Leaf sheath transverse section. C and H, Root. D, A mature flower. Abbreviations: a, anther; o, ovary; s, stigma. E, A mature seed. Abbreviations: em, embryo; en, endosperm.

Dow



Figure 4. (Legend appears on facing page.)

Dow



998 Kyozuka et al. Plant Physiol. Vol. 102,1993

Figure 5. Cell specificity in the light-dependent induction of rice (A-C) and tomato (D and F) rbcS-gusA genes. A and D,Transverse section of a shoot grown in dark for 5 d. B and C, Transverse section of a dark-grown shoot after 24 h of lighttreatment. E, Transverse section of a dark-grown shoot after 48 h of light treatment. F, Transverse section of a dark-grownshoot after 96 h of light treatment.

irrespective of the low level of sequence conservation in theirrbcS promoter regions. Sequence comparison revealed thattwo elements are conserved between the rice and tomato rbcSpromoter regions; one is the I-box (GATAG, positions —115to —111, in rice rbcS, and GATAAG, positions —168 to —163,in tomato rbcS3C), and the other is, unexpectedly, a part of apreviously proposed monocotyledonous rbcS consensus se-quence (GCGGCCAAT, positions —99 to —91, in rice rbcS,and GTGGCCAT, positions -120 to -113, in tomato rbcS3C).Both the I-box and the monocotyledonous consensus se-quence have been reported to function as light-responsive

elements in the promoter of maize rbcSZml (Schaffner andScheen, 1991), tomato rbcSBC (Manzara et al., 1991), tobaccocab-E (Schindler and Cashmore, 1990), and ArabidopsisrbcSIA (Donald and Cashmore, 1990) genes. These observa-tions, along with the results described here, imply that thesetwo sequence elements may be involved in the regulation ofrice rbcS gene expression.

Despite the apparent functional conservation of rbcS pro-moters from monocots and dicots, our results suggest thatthe dicotyledonous (tomato) rbcS promoter used in our analy-sis is less competent than the monocotyledonous (rice) rbcS

Dow




1 2 3 5 6 lane

Tomato Rice rbcS promoter24 0 6 24 0 6 24 hours of light treatment

ibcS-gusA transcript

Figure 6. Light induction of gusA mRNA abundance in leaves oftransgenic rice. Gel blots containing 40 ^g of total RNA fromnontransformed rice (lane 1), transgenic rice carrying tomato rbcS-gusA gene (lanes 2-4), and transgenic rice carrying the rice rbcS-gusA gene (lanes 5-7) were hybridized with a gusA probe. RNAswere isolated from leaves of dark-adapted plants after 0 h (lanes 2and 5), 6 h (lanes 3 and 6), and 24 h (lanes 1, 4, and 7) of lighttreatment.

confers a high level of foreign gene expression with stricttissue specificity.

ACKNOWLEDGMENTS

We thank Dr. M. Sugita (Nagoya University, Nagoya) for provid-ing the tomato rbcS3C gene, and Dr. M. Matsuoka (National Instituteof Agrobiological Resources, Tsukuba) for helpful suggestions.

Received January 27, 1993; accepted March 22, 1993.Copyright Clearance Center: 0032-0889/93/102/0991/10.The EMBL accession number for the sequence reported in this article

is X07515.

promoter in transgenic rice plants. Some points need to beconsidered, however, before drawing any firm conclusion.First, there is a difference in vector constructs; an intron ispresent in the tomato rbcS-gusA gene that is absent from therice rbcS-gusA gene. We previously found that the castor beancatalase 1 intron used in this study increases expression ofthe gusA gene in transgenic rice more than 10-fold in com-bination with the cauliflower mosaic virus 35S promoter(Tanaka et al., 1990). Moreover, enhancement of foreign geneexpression by introns has been observed from other plantpromoters (Callis et al., 1987; Kyozuka et al., 1990; McElroyet al., 1990) in transformed monocot cells. These resultssuggest that it is likely that insertion of an intron in thetomato rbcS-gusA fusion gene increases its expression intransgenic rice. Second, the copy number of integrated genescould be an important factor determining expression levelsof introduced rbcS-gusA genes. Our DNA gel blot analysisestablished that six of eight transgenic plants analyzed con-tained a single copy of the introduced gene (Table I), indicat-ing that the different expression level detected in transgenicplants carrying the tomato rbcS-gusA gene and those carryingthe rice rbcS-gusA gene is not caused by the difference incopy number of the introduced gusA genes. Finally, becauserbcS is a gene family, the tomato rbcSSC gene may notrepresent the member of the gene family corresponding inits developmental and light regulation to the rice gene usedin our study. However, previous studies with the tomato rbcSgene family indicated that, among the five tomato rbcS genes,the rbcSSC gene used in our study is expressed at the highestlevel in green leaves (Sugita et al., 1987; Wanner and Gruis-sem, 1991). Taken together, our results suggest that there aredifferences in the quantitative regulation of gene expressionbetween monocotyledonous and dicotyledonous rbcS pro-moters in transgenic monocots. This notion was supportedby previous studies analyzing expression of monocot genesin transgenic tobacco plants (Lamppa et al., 1985; Elk's et al.,1986; Keith and Chua, 1986). Our results support this viewin the reciprocal sense that the activity of a dicotyledonouspromoter is reduced in a transgenic monocotyledonous plant.

Recently, virus-resistant rice plants expressing the coatprotein gene of rice stripe virus have been produced (Hay-akawa et al., 1992), suggesting that genetic engineering willbe increasingly applied to monocotyledonous crops in thenear future. The rice rbcS promoter described here should beuseful in monocot transformation, since in transgenic rice it

LITERATURE CITED

Bansal KX, Viret J-F, Haley J, Khan BM, Schantz R, Bogorad L(1992) Transient expression from cab-mi and rbcS-m3 promotersequences is different in mesophyll and bundle sheath cells inmaize leaves. Proc Natl Acad Sci USA 89: 3654-3658

Bernatsky R, Tanksley SD (1986) Genetics of actin-related se-quences in tomato. Theor Appl Genet 72: 314-321

Broglie R, Coruzzi G, Lamppa G, Keith B, Chua N-H (1983)Structural analysis of nuclear genes coding for the precursor to thesmall subunit of wheat ribulose-l,5-bisphosphate carboxylase.Bio/Technology 1: 55-61

Callis J, Fromm M, Walbot V (1987) Introns increase gene expres-sion in cultured maize cells. Genes Dev 1: 1183-1200

Chomczynski P, Sacchi N (1987) Single-step method of RNA byacid guanidium thiocyanate-phenol-chloroform extraction. AnalBiochem 162: 156-159

Dellaporta SJ, Wood J, Hicks JB (1983) A plant DNA miniprepara-tion: version II. Plant Mol Biol Rep 1: 19-21

Donald RGK, Cashmore AR (1990) Mutation of either G box or Ibox sequences profoundly affects expression from the ArabidopsisrbcS-lA promoter. EMBO J 9: 1717-1726

Edwards JW, Coruzzi GM (1990) Cell-specific gene expression inplants. Annu Rev Genet 24: 275-303

Ellis JG, Llewellyn DJ, Dennis ES, Peacock WJ (1986) Maize Adh-1 promoter sequences control anaerobic regulation: addition ofupstream promoter elements from constitutive genes is necessaryfor expression in tobacco. EMBO J 6: 11-16

Gilmartin PM, Sarokin L, Memelink J, Chua N-H (1990) Molecularlight switches for plant genes. Plant Cell 2: 369-378

Hayakawa T, Zhu Y, Itoh K, Kimura Y, Izawa T, Shimamoto K,Toriyama S (1992) Genetically engineered rice resistant to ricestripe virus, an insect-transmitted virus. Proc Natl Acad Sci USA89: 9865-9869

Ingelbrecht ILW, Herman LMF, Dekeyser RA, Van MontagueMC, Depicker AG (1989) Different 3' end regions strongly influ-ence the level of gene expression in plant cells. Plant Cell 1:671-680

Keith B, Chua N-H (1986) Monocot and dicot pre-mRNAs areprocessed with different efficiencies in transgenic tobacco. EMBOJ 5: 2419-2425

Kyozuka J, Fujimoto H, Izawa T, Shimamoto K (1991) Anaerobicinduction and tissue-specific expression of maize Adhl promoterin transgenic rice plants and their progeny. Mol Gen Genet 228:40-48

Kyozuka J, Hayashi Y, Shimamoto K (1987) High frequency plantregeneration from rice protoplasts by novel nurse culture methods.Mol Gen Genet 206: 408-413

Kyozuka J, Izawa T, Nakajima M, Shimamoto K (1990) Effect ofthe promoter and the first intron of maize Adhl on foreign geneexpression in rice. Maydica 35: 353-357

Lamppa G, Nagy F, Chua N-H (1985) Light-regulated and organ-specific expression of a wheat Cab gene in transgenic tobacco.Nature 316: 750-752

Lebrum M, Waksman G, Fressinet G (1987) Nucleotide sequence

Dow



1 O00 Kyozuka et al. Plant Physiol. Vol. 102, 1993

of a gene encoding com ribulose-1,5-bisphosphate carboxylase/ oxygenase small subunit (rbcS). Nucleic Acids Res 1 5 4360-4364

Lloyd JC, Raines CA, Jhon UP, Dyer TA (1991) The chloroplast FBPase gene of wheat: structure and expression of the promoter in photosynthetic and meristematic cells of transgenic tobacco plants. Mo1 Gen Genet 225 209-216

Maniatis T, Fritsch EF, Sambrook J (1982) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, p 543

Manzara T, Carrasco P, Gruissem W (1991) Developmental and organ-specific changes in promoter DNA-protein interactions in the tomato rbcS gene family. Plant Cel l3 1305-1316

Matsuoka M, Kano-Murakami Y, Tanaka Y, Ozeki Y, Yamamoto N (1988) Classification and nucleotide sequence of cDNA encoding the small subunit of ribulose-1,5-bisphosphate carboxylase from rice. Plant Cell Physiol29 1015-1022

Matsuoka M, Sanada Y (1991) Expression of photosynthetic genes from the C4 plant, maize, in tobacco. Mo1 Gen Genet 225:

McElroy D, Zhang W, Cao J, Wu R (1990) Isolation of an efficient actin promoter for use in rice transformation. Plant Cell 2

Ohta S, Mita S, Hattori T, Nakamura K (1990) Construction and expression in tobacco of a 8-glucuronidase (GUS) reporter gene

411-419

163-171

containing the intron within the coding sequence. Plani Cell Piys- iol31: 805-812

Rolfe S, Tobin EM (1991) Deletion analysis of a phytochrome- regulated monocot rbcS promoter in a transient assay system. Proc Natl Acad Sci USA 88: 2683-2686

Schaffner AR, Scheen J (1991) Maize rbcS promoter activity depends on sequence elements not found in dicot rbcS promoters. Plant Cel l3 997-1012

Schindler U, Cashmore AR (1990) Photoregulated gene expression may involve ubiquitous DNA binding proteins. EMBO J 9 3415-3427

Shimamoto K, Terada R, Izawa T, Fujimoto H (1989) F e d e transgenic rice plants regenerated from transformed protoplasts. Nature

Sugita M, Maiazara T, Pichersky E, Cashmore A, Gruislsem W (1987) Genornic organization, sequence analysis and expreesion of a11 five genes encoding the small subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase from tomato. Mo1 Gen Geriet 209

Tada Y, Sakamloto M, Matsuoka M, Fujimura T (1991) Expression of a monocot LHCP promoter in transgenic rice. EMBO J 1 0

Tanaka A, Mita S, Ohta S, Kyozuka J, Shimamoto K, Nalramura K (1990) Enhancement of foreign gene expression by a dicot intron in rice but not in tobacco is correlated with an increased leve1 of mRNA and an efficient splicing of the intron. Nudeic Acids Res

Terada R, Nakeyama T, Iwabuchi M, Shimamoto K (1993) A wheat histone H3 promoter confers cell division-dependent ancl -inde- pendent expression of the gusA gene in transgenic rice plants. Plant

Tobin EM, Silverthorne J (1985) Light regulation of gene expression.

Wallace DM (1987) Preparation of nucleic acids. Methods Enzymol

Wanner LA, Gnuissem W (1991) Expression dynamics of the tomato rbcS gene fandy during development. Plant Cel l3 1289-1 303

Xie Y, Peng 2, Cai Y, Wu R (1987) Cloning and analysis of histone 3 gene and Rubisco small subunit gene of rice. Sci Si Ser B 3 0

Xie Y, Wu R (1988) Nucleotide sequence of a ribulose-1,5-bisphosphate carboxylase/oxygenase small subunit gene (rbcS) in rice. Nucleic Acids Res 1 6 7749

Zhang W, McElroy D, Wu R (1991) Analysis of rice ActI 5’ region activity in transgenic rice plants. Plant Cel l3 1155-1165

338: 274-276

247-256

1803-1808

18: 6767-6770

J 3 241-252

Annu Rev Plmt Physiol36 569-593

152 41-48

706-718

Dow



light-regulated and cell-specific expression of rbcs- gusa ...the gusa gene directed by the tomato...

Documents