structure and expression of the rice glutelin multigene family* · 2001-07-21 · 12574 structural...

THE JOURNAL OF BIOLOGICAL CHEMISTRY Q 1989 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 264, No. 21, Issue of July 25, pp. 12573-12581, 1989 Printed in U. S. A.

Structure and Expression of the Rice Glutelin Multigene Family* (Received for publication, February 22, 1989)

Thomas W. Okitasq, Young So0 HwangS, James Hnilog, Woo Taek Kim§, Arun P. Aryan§, Raymond Larsonll , and Hari B. Krishnan** From the Institute of Biological Chemistry, Washington State University, Pullman, Washington 99164-6340

A near full-length cDNA and three genomic clones for rice (Oryza sativa L.) glutelin were isolated and studied. Based on nucleic acid sequence and Southern blot analyses, the three isolated glutelin genomic clones were representative members of three gene subfamilies each containing five to eight copies. A comparison of DNA sequences displayed by relevant regions of these genomic clones showed that two subfamilies, repre- sented by clones Gtl and Gt2, were closely related and evolved by more recent gene duplication events. The 5“flanking and coding sequences of Gtl and Gt2 displayed at least 87% homology. In contrast, Gt3 showed little or no homology in the 5”flanking sequences upstream of the putative CAAT boxes and exhibited significant divergence in all other portions of the gene. Conserved sequences in the 5“flanking regions of these genes were identified and discussed in light of their potential regulatory role. The derived primary sequences of all three glutelin genomic clones showed significant homology to the legume 11 S storage proteins indicating a common gene origin. A comparison of the derived glutelin primary sequences showed that mutations were clustered in three peptide regions. One peptide region corresponded to the highly mutable hy- pervariable region of legume 11 S storage proteins, a potential target area for protein modification. Expres- sion studies indicated that glutelin mRNA transcripts are differentially accumulated during endosperm development. Promoters of Gt2 and Gt3 were functional as they direct transient expression of chloramphenicol acetyltransferase in cultured plant cells.

Plants store a significant amount of their nitrogen, sulfur, and carbon reserves as storage proteins in seed tissue, which are utilized during the post-germinative periods of development. Based on their solubility properties, these storage pro-

* This work was supported in part by the Rockefeller Foundation and Project 0590 of the Agricultural Research Center, College of Agriculture and Home Economics Research Center, Washington State University. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked ‘‘advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Present address: Agricultural Research Institute, R. D. A., Su- weon 170, Korea.

3 Present address: Institute of Biological Chemistry, Washington State University, Pullman, WA 99164.

7 To whom correspondence should be addressed. 11 Present address: Dept. of Veterinary Microbiology and Pathol-

** Present address: Dept. of Plant Pathology, University of Mis- ogy, Washington State University, Pullman, WA 99164.

souri, Columbia, MO 65211.

teins consist of three classes: globulins, prolamines, and glutelins. Globulins, characterized by their solubility in saline solutions, serve as the major nutrient reserves in the embry- onic tissues of both dicot and monocot seeds, whereas the alcohol-soluble prolamines and relatively insoluble glutelins serve in this capacity in the endosperm tissue of monocots (reviewed in Shotwell and Larkins, 1989). Two major globulin classes, designated 7 S and 11 S based on their sedimentation properties, are present in different proportions among dicot plants.

Unlike most cereals, which utilize the alcohol-soluble prolamines as a reserve, the major proteins present in rice seeds are the glutelins. These proteins, which may constitute up to 80% of the total endosperm protein, are synthesized and accumulated during the mid-stages of endosperm development (Yamagata et al., 1982). Rice seeds also store prolamines, but this fraction consists of only 5-10% of the total endosperm protein. The synthesis of glutelins and prolamines is not coordinate. Glutelins are initially synthesized at 4-6 days post-anthesis, whereas prolamine accumulation is first detected several days later. These proteins are deposited exclu- sively into two morphologically distinct protein bodies which are formed by different cellular processes (Tanaka et al., 1980; Krishnan et al., 1986).

Despite their solubility properties, evidence has been gath- ered that the rice glutelin is structurally similar to the 11 S globulins of legumes (Zhao et al., 1983; Wen and Luthe, 1985). Both proteins are initially synthesized as a preproprotein, whereupon, processing and transport events lead to deposition of the proprotein in modified vacuoles (Yamagata et al., 1982; Krishnan et al., 1986). Subsequent proteolysis results in the formation of covalently linked acidic and basic subunits (Wen and Luthe, 1985). Recent molecular analysis of glutelin cDNA clones confirmed the homology of the glutelin primary sequence to legume ll S globulins (Takaiwa et al., 1987b, Higuchi and Fukazawa, 1987).

Efforts to engineer the glutelin proteins for improved nutritional properties necessitates a detailed understanding of the structure and expression of these genes. Here we report that the glutelins are encoded by a complex gene group consisting of at least three subfamilies each consisting of five to eight copies. Southern blot and DNA sequence analysis show that these families arose by gene duplication events. Consistent with the divergent nature of the 5”flanking sequences of the glutelin genes, analysis of the mRNA transcript levels indicate that the glutelin genes are differentially expressed during endosperm development. We also show that although glutelin structural sequences reveal strong conservation, several peptide regions seem prone to mutation not only among the glutelins but within other 11 S storage proteins as well.

12573

12574 Structural and Expression Studies of a Rice Storage Protein Multigene Family

EXPERIMENTAL PROCEDURES AND RESULTS’

Isolation of cDNA and Genes-Monospecific antibodies raised against the purified acidic subunit of the glutelin protein (Krishnan and Okita, 1986) were used to identify cDNA clones constructed in a Xgt 11 library. Several clones were identified, and preliminary analysis showed that all were truncated and lacked various portions of the coding sequence. A second cDNA library was screened resulting in the isolation of pG22, which contained a cDNA insert of about 1.8 kb? estimated to be near full length. This cDNA was then labeled with 32P and utilized as a probe to screen several rice genomic libraries constructed in various X cloning vectors (see Mini- print). Analyses of rice genomic DNA libraries containing size-selected rice DNA fragments yielded three different genomic glutelin (Gt) clones Gtl, Gt2, and Gt3. Relevant DNA fragments from all three genes were subcloned and analyzed as discussed below.

The Subfamilies of the Glutelin Genes-Physical restriction maps of all three glutelin genes are depicted in Fig. 1. Other than a common SpeI site, all three glutelin genes exhibited unique restriction maps although Gtl and Gt2 shared more sites. Conservation of several restriction enzyme sites, SpeI StuI, and SphI, were evident between Gtl and Gt2, but many others were not shared. Cross-hybridization (results not shown) and DNA sequencing (see below) analysis corrobo- rated this close relationship between Gtl and Gt2. Southern blot analysis of rice DNA was then conducted using subclones of these three glutelin genes as probes (Fig. 2). In almost all instances, unique hybridizable bands were evident when probed with the three different Gt subclones. Exceptions were the faint bands exhibited by EcoRI-digested rice DNA when probed with Gt2 subclone. The faint bands at 4.2 and 2.7 kb were due to the lower cross-hybridization of the Gt2 probe to fragments containing the Gtl and Gt3 genes, whereas the 3.4- kb band was found to be a non-glutelin DNA sequence that hybridized weakly to all glutelin probes (see Miniprint section). The absence of a major band under these conditions suggested that Gt2 resided on chromosomal DNA which lacked EcoRI sites. Other than this single exception, all of the enzyme-generated hybridizable bands exhibited by all three Gt probes were accountable by the presence of three glutelin gene subfamilies. The intensity of the bands exhibited by all three Gt probes was about the same. A comparison of the signal strengths of these bands to internal gene copy number standards indicated that each of the three subfamilies contained about five to eight copies (Fig. 2). Together, these observations showed that these three Gt genes represent the major classes, if not all, of the glutelin genes present on rice chromosomes.

DNA Sequence Analysis of Glutelin Genes-The nucleotide sequence alignments of the three glutelin genomic clones as well as a single cDNA are shown in Fig. 3. The aligned sequences begin about 900 bp upstream of the translational initiation codon and extend through the structural gene and 3’-flanking regions with one exception. The genomic clone, Gt3, was truncated and lacked genetic information for 165 bp of the carboxyl-terminal region of the basic glutelin subunit and corresponding 3‘-flanking segment. The near full-length cDNA, pG22, however, was 98% homologous to Gt3, strongly suggesting that it is a Gt3 class gene. Therefore, the close

’ Portions of this paper (including “Experimental Procedures,” part of “Results,” Fig. 8, and Tables I11 and IV) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press.

* The abbreviations used are: kb, kilobase pair(s); bp, base pair(s).

GI3

ridc*h.* lac-

FIG. 1. Physical maps of restriction endonuclease sites in subclones of glutelin genes. A 4.2-kb EcoRI fragment from a X genomic clone containing Gtl and a 2.7-kb genomic DNA EcoRI fragment containing Gt3 were subcloned into the Bluescript M13- vector. The map of Gt2 was obtained by combining individual maps obtained from analyzing subclones containing the 3.4- and 1.6-kb HindIII fragments. The rectangular boxes indicate the approximate position of the coding sequences. Dotted boxes indicate the signal peptide sequences, whereas shaded boxes depict the three introns observed in these genes.

B E H B E H B E H

18.0 kb

8.5

4.2 4.0

3.4

2.1

? - - r., I 0

A B C D FIG. 2. Reconstruction Southern blot analysis of glutelin

genes. About 5 pg of rice DNA was digested with either BamHI ( B ) , EcoRI ( E ) , or HindIII (H) and resolved on a 0.5% agarose gel. The gel was transferred to a Zeta probe filter and hybridized to 32P-labeled 4.2-kb EcoRI fragment of Gtl (A) , 3.4-kb HindIII fragment of Gt2 ( B ) , or 2.7-kb EcoRI fragment of Gt3 (C). D depicts the copy number standards representing 1,2,5,10, and 20 gene copies/haploid genome.

sequence relationship of pG22 to Gt3 not only permitted the identification of exon-intron boundaries within the coding segments but also allowed a possible indication of the relat- edness of the 3‘ region of Gt3 class genes to the other Gt genes by comparison to pG22 sequences (Fig. 3). Table I shows the degree of divergence among the three Gt type genes in segments separated by natural boundaries (introns, exons, etc.) and four arbitrary regions of the 5’-flanking sequences. The sequences of Gtl and Gt2 exhibited close similarity, whereas Gt3 sequences showed significant divergence. In all comparisons, however, the sequence identity of the exons of all three genes were much more conserved (4-6% divergence for Gtl/Gt2 and 18-23% for Gtl/Gt3 comparisons) than flanking and intron regions (8-20% divergence for Gtl/Gt2 and 18-71% for Gtl/Gt3 comparisons). Therefore, despite the

Structural and Expression Studies of a Rice Storage Protein Multigene Family 12575

FIG. 3. The DNA sequences of glutelin genomic clones Gtl, Gt2, and Gt3 and a cDNA clone, pG22. The complete 5’- and 3’- Ranking and coding sequences are shown for Gtl and Gt2, whereas Gt3 is missing a portion of the coding sequence and accompanying 3’-flanking region. Numbering is based on the translational initiation codon for Gtl which is shaded. Nucleotide replacements are depicted, whereas an asterisk denotes conservation. Gaps have been introduced to achieve the best alignment of the glutelin genes and its cDNA. The transcriptional initiation site is indicated by an arrow beginning with nucleotide -39. Conserved sequences around the putative TATA and CAAT boxes are shaded. The bold single-headed arrows represent repeat units that are shared in the 5’-flanking sequences of Gtl and Gt2. Small double-headed arrows depict the RY repeat elements

presumed lack of any enzymatic or cell structural function of these proteins, mutational constraints are, nevertheless, act- ing on their coding sequences. Presumably, the primary sequences of the glutelins have been maintained for correct folding and post-translational processing of the polypeptide and aggregation into discrete organelles in seed tissue.

The 5’- and 3’-Flanking Sequences-The 5’-flanking region immediately proximal to the translation initiation codon exhibited considerable homology among all Gt genes. The close similarity was readily apparent in the flanking regions -125 to -1 bp from the initiation codon (Table I). The translation initiation sequences of all glutelin genes agreed well with the plant consensus motif (Liitcke et al., 1987). In addition, com- parative analysis indicated that three short conserved stretches of sequences upstream of the translational start sites were shared by all Gt genes. Sl nuclease mapping studies revealed that one conserved region, -41 to -32 bp, denoted the transcriptional start sites (Fig. 4). A predominant Sl nuclease-protected band at -39 bp was readily evident together with several minor bands when probed with a DNA fragment from Gt3. An almost identical hybridization pattern was obtained when a corresponding segment of Gt2 was used in the Sl protection assay. These results differed from those obtained by Takaiwa et al. (1987a) who suggested that transcriptional initiation occurred 4 bp downstream at position -35 bp. The presence of multiple bands from the Sl nuclease protection assay may indicate multiple transcription initiation sites, although it can also be accounted for by the slight sequence differences displayed by individual polymorphic components of this multigene family. A conserved TATA box, homologous to the eucaryotic consensus, is located 28 bp from the major transcriptional initiation site. Further upstream is a third conserved region shared by all Gt genes which contains several putative CCAAT boxes. None of them, however, showed significant homology to the eucaryotic consensus. The absence of well defined CCAAT motifs seems to be a frequent occurrence for many cereal seed protein genes (Reeves and Okita, 1987).

Consistent with the strong homology in the exons exhibited by Gtl and Gt2, the promoter sequences (-900 to -125 bp) distal to the translational start were highly conserved. Signif- icant divergence between Gtl and Gt2 was mainly due to the presence of insertions/deletions of DNA fragments. One DNA insertion of Gt2, located between -266 and -310 bp, contained an inverted repeat, AATC, located near or at the borders (Fig. 3). This repeat was also evident at one of the borders of the next upstream DNA insertion. The presence of short inverted repeats are reminiscent of transposable elements, suggesting their involvement of the evolution of the rice glutelin genes. The significant divergence observed elsewhere for Gtl/GtS comparison was also evident in this region of the Gt promoter. Beginning with -186 bp, no significant homology is evident in the upstream promoter regions between Gt3 and the other two glutelin genes. Since many transcriptional regulatory elements of plant promoters reside in this region (Messing et al., 1983), the dissimilarity in nucleotide sequence exhibited by Gt3 indicated that it may not be subject to the same transcriptional controls as Gtl and

present in the promoter regions of all three genes. Small single-headed arrows indicate the small border repeats evident in the DNA insertions in the Gt2 promoter region. The three introns and their respective borders are labeled below the DNA sequences. The termination codon is located at nucleotide 1798. Consensus polyadenylation signal sequences are indicated with ouerhead lines. The polyadenylation addition sites for Gtl and Gt2 are indicated with arrowheads. GT- rich regions immediately downstream of the polyadenylation site are shown with wavy lines.


TABLE I Divergence of glutelin genes of 0. sativa

Assignments of the relevant regions of the glutelin genes and nucleotide positions are from Fig. 3. Gaps, irrespective of length, were considered as a single substitution. As Gt3 was missing the 3’ end of the gene, exon 4, and the 5’-untranslated regions of pG22, a Gt3 class cDNA, were compared with Gtl and Gt2.

Structural region

5”Flanking 5”Flanking 5”Flanking 5”Flanking 5”Flanking Exon 1 Intron 1 Exon 2 Intron 2 Exon 3 Intron 3 Exon 4 3”Untranslated

Gt l us Gt2 G t l us Gt3 Nucleotide positions No. of

gam Matches Divergence 2’f Matches Divergence

-900 to -651 -650 to -400 -267 to -399 -266 to -126 -125 to -1

1 to 331 332 to 420 421 to 698 699 to 812 813 to 1298

1299 to 1383 1384 to 1797 1798 to 1860

7 5 3 3 0 0 0 0 3 0 3 0 2

2011228 2161251 26/41

1271140 1091124 3191331

2591275 841105

451/480

3911411 58/63

78/89

75/84

11 %

14 37 9

12 4

12 6

20 6

11 5 8

2 3 1 2 0 5 2 4 2 0

No homology No homology No homology

791141 961125

2761331 65/89

2241275 621114

3761483

3411414 17/63

49/85

i

44 23 17 18 19 46 22 42 18 73

Gt2 us Gt3

No. Of Matches Divergence gaps

%

5 3 1 2 2 3 2 4 2 2

781125 901125

276/331 65/89

2271275 661114

3731483 47/85

3421414 18/63

38 28 17 18 17 42 23 45 17 71

Gt2. Evidence for differential control of the glutelin genes during seed development is presented later in this section.

Potential transcriptional regulatory sequences have been implicated by comparing seed protein genes from maize, bar- ley, and wheat. Kreis et al. (1985) suggested that a -300 bp element, similar in sequence to the SV40 enhancer core, may play some role in transcription of cereal seed protein genes due to its ubiquitous presence in a11 genes so far analyzed. Such a -300 bp element, with the consensus sequence TG(T/ A/C)AAA(G/A)(G/T) (Colot et al., 1987), was present at -694, -486, and -253 bp in the promoter regions of Gtl and Gt2 but only once in Gt3 (Table 11). This -300 bp element forms part of a larger DNA segment which had been duplicated intact or partially in Gtl and Gt2 promoter regions. The dyad sequence, CATGCATG (RY repeat), is a second motif prevalent in many seed tissue genes but not in genes expressed in other plant tissues (Dickinson et al., 1988). A perfect RY repeat is present at -285 bp in one of the larger DNA insertions of Gt2. Variant forms of this motif were observed in two of the larger DNA insertions present in the promoter region of Gt2 and elsewhere in the 5”flanking sequences of Gtl and Gt2 (Fig. 3). The possible role of these putative regulatory sequences in the temporal and tissue- specific expression of the Gt genes is presently being evaluated.

Other than the presence of polyadenylation signals, the 3‘- untranslated region of pG22, a Gt3 class cDNA, differed extensively from that observed for Gtl and Gt2. Multiple polyadenylation signals, AATAAA or AATAAC (Birnsteil et al., 1985), were present in all three genes. These signals were clustered in Gtl and Gt2 or separated by 8 bp in Gt3. As observed for the 5”flanking sequences of these genes, a single insertion/deletion event resulted in an addition or subtraction of a AATAAA sequence in Gtl and GT2, respectively. Such simple (dA-dT) sequences are highly prone to mutational events via recombination of homologous sequences present elsewhere in the genome (Slightom et al., 1980) or base pair slippage during DNA synthesis (Jones et al., 1979). Corre- sponding cDNA sequences for Gtl and Gt2 have been characterized by Takaiwa et d . (1987b) which enabled us to pinpoint the site of poly(A) tailing of their respective transcripts (Fig. 3). Immediately downstream from the poly(A) addition site are G/T clusters which have been suggested to serve as termination signals for transcription (Birnsteil et al.,

1985). The only recognizable polyadenylation signals, however, were more than 80-100 bp upstream from the polyadenylation site, a distance far greater than the usual 10-30 bp evident in most eucaryotic genes including the Gt3 class pG22 (Fig. 3). I t seems plausible that due to the conserved distance between the polyadenylation signal and the poly(A) addition site in most genes, polyadenylation signals, as yet unrecog- nized, are operational in Gtl and Gt2 (Joshi, 1987). Indeed, the polyadenylation signal seems to be flexible as many vari- ations of the AATAAA motif have been identified (Birnsteil et al., 1985). Perhaps the GATAAA motif, 17 bp from the poly(A) addition site, serves this role in Gt2 (Fig. 3).

Exon-Zntrons-The exon-intron borders were defined based on the direct comparison of the cDNA and genomic sequences (Fig. 3). Three introns interrupted the glutelin genes at identical positions of the coding sequence. The positions of these introns are identical to those observed for the pea legumin 11 S storage protein gene (Lycett et al., 1984), supporting the hypothesis that these plant genes were derived from common ancestral DNA segments (Borrotto and Dure, 1987). All of the intron borders obeyed the AG/GT rule and displayed additional properties to the plant donor-acceptor consensus (Brown, 1986). The introns of the glutelin gene were relatively A-T rich and all were relatively small, ranging in size from 75 to 111 bp. The presence of short introns, or their total absence, seems to be a prevailing feature of plant storage protein genes (Messing et al., 1983).

Protein Structure-The primary sequences of the glutelin preproproteins derived from both the genomic and cDNA sequences were shown in Fig. 5. All of the derived primary sequences showed significant homology (about 30-35%) to legumin and glycinin, storage proteins of pea and soybean, respectively (results not shown but see Higuchi and Fuka- zawa, 1987). The derived sequence at the amino terminus displayed features typical of a signal peptide: a basic residue near the amino terminus and a leucine-rich hydrophobic core. Although the exact cleavage site has not been determined due to a blocked amino terminus of the mature acidic subunit, it most likely occurs between residues 24 (Ala) and 25 (Gln). Takaiwa et al. (1986) suggested that the leader peptide was 37 residues. The putative cleavage site assigned here, however, conforms best to von Heijne’s rule (1986) for signal peptide cleavage specificities. Cleavage at this site may also result in the generation of a cyclized glutamine residue, known to be


/

a b

" T" ir

5'

c d

I 3'

FIG. 4. Mung bean nuclease analysis of the glutelin transcriptional initiation site. A 1.1-kb EcoRI-SpeI fragment, ?'P- labeled at the SpeI site by polynucleotide kinase, was denatured and hybridized with rice seed poly(A')-RNA as described in the text. After treatment with mung bean nuclease, the labeled fragments were heat-denatured and analyzed on a 6% polyacrylamide gel under denaturing conditions. Lanes a and b depict the mobility of the mung bean nuclease-resistant fragments, h e b containing three times as much sample as lane a. Lanes c and d contain products of the 32P- labeled EcoRI-SpeI fragment which have been subjected to the purine and pyrimidine reactions of Maxam and Gilbert (1980). The arrow- head indicates the predominant sized fragment protected by nuclease treatment. The nucleotide sequence around the transcriptional initiation site is depicted on the right.

intractable to Edman degradation. Moreover, alignment of the glutelin and soybean glycinin preproprotein sequences corroborates this cleavage site (Moreira et al., 1981; Scallon et al., 1987). An interesting feature displayed by the signal peptide is the presence of a Cys-Leu-(Leu/Phe)-Leu-Leu-Cys peptide residing in the central core which is presumably capable of forming an intrapeptide disulfide linkage during protein synthesis. A similar peptide motif is also present in the signal peptides of several 11 S legume storage proteins. The conservation of this peptide motif displayed by these seed proteins is strongly suggestive that it may reflect some functional significance in the events leading to the binding of the signal peptide-ribosome complex to the rough endoplasmic reticulum and subsequent transport into the endoplasmic reticulum lumen.

The site of post-translational cleavage, which results in the formation of the acidic and basic subunits of the glutelin protein, was strongly conserved among the four derived pri-


The glutelin acidic and basic subunits are believed to be linked by a disulfide bond (Wen and Luthe, 1985). Four of the 7 cysteine residues present in the glutelin subunits are conserved in both pea legumin and soybean glycinin. Two of these residues, a t positions 122 and 315, are present a t homologous positions in soybean glycinin. Since it has been established that these cysteine residues participate in disulfide linkage between the glycinin acidic and basic subunits (Stas- wick et al., 1984), the comparable residues in glutelin are likely to play the same role. Likewise, the strong conservation of 2 cysteine residues in both glutelin and legume 11 S proteins

Gtl G t 3

5 10 15 20 25 5 10 15 20 25 D A F

FIG. 6. Temporal accumulation patterns of Gtl and Gt3 mRNA transcripts during seed development. Poly(A’)-RNA samples were obtained from developing rice seeds at 5,10,15,20, and 25 days postanthesis, resolved on a formaldehyde-agarose gel, and blotted onto nitrocellulose. The blot was then probed with 3ZP-labeled Gt l or Gt3 DNA under stringent hybridization conditions, washed, and subjected to autoradiography. DAF, days after flowering.

“-a “ ~ 0 . 0

a a b b c c FIG. 7. Transient expression of glutelin promoter-chlor-

amphenicol acetyltransferase (CAT) fusion in tobacco cells. Two translational fusions to chloramphenicol acetyltransferase were constructed containing the 5’-flanking region of Gt3 and either 1332 ( a ) or 72 ( b ) bp of the coding sequence of Gt3. Physical restriction enzyme maps are depicted. The lightly shaded boxes indicate the signal peptide, whereas the darkly shaded boxes depict the three introns. Coding sequences of glutelin and chloramphenicol acetyltransferase are presented as open and dark boxes, respectively. The arrow shows the post-translational cleavage site resulting in the production of acidic and basic subunits. These plasmid constructs were transferred into NT-1 tobacco protoplasts by electroporation as described in the text and analyzed for chloramphenicol acetyltransferase activity after 24 h. Lanes a and bare duplicate chloramphenicol acetyltransferase activity reactions of construct a and b, respectively. Lane c is the result obtained for extracts obtained from protoplasts subject to electroporation conditions without any DNA.

supports the notion that they are involved in intrachain linkage.

Expression Studies of Glutelin Genes-Total RNA samples were isolated from rice seeds a t different stages of development and analyzed by the Northern blot technique using specific probes for Gtl and Gt3. These studies were conducted under stringent hybridization-wash conditions which allow only negligible cross-reaction of Gtl and Gt3 sequences (see “Experimental Procedures”). In both instances, glutelin mRNA transcripts were detected by 5 days post-anthesis, but the accumulation patterns for Gtl and Gt3 transcripts differed substantially during further seed development (Fig. 6). G t l transcripts appeared to accumulate throughout seed development, whereas those for Gt3 were differentially expressed. Maximal accumulation of Gt3 transcripts took place between 5 and 10 days and then steadily declined during seed maturation. Accumulation patterns for Gt2 appeared similar to that obtained for G t l (results not shown). These results indicate that the expression of the glutelin multigene family is not coordinate and is differentially regulated during seed development.

The DNA sequence and structural data are consistent with the notion that the isolated genomic clones contain functional glutelin genes, i.e. capable of expression. To directly test the proficiency of these genes, plasmids containing the 5“flanking and various portions of the coding segment of Gt3 were fused in-frame to a large glutelin DNA segment containing the 5”flanking sequence and 80% of the coding sequence was fused in-frame to chloramphenicol acetyltransferase (Fig. 7a), whereas a second plasmid contained a shorter translational fusion consisting of the glutelin signal peptide (Fig. 7b). These constructs were then introduced into tobacco or rice protoplasts by electroporation and evaluated for transient expression of chloramphenicol acetyltransferase activity. Chloram- phenicol acetyltransferase activity is clearly detected in tobacco cells transfected with these plasmid constructs, but not in cells subjected to mock electroporation (Fig. 7). Similar results have also been obtained with plasmids constructed with Gt2 sequences and when rice protoplasts were utilized as the recipient host cells (results not shown). These results indicate that Gt2 and Gt3 glutelin genes, isolated in this study, contain functional promoters effective in directing transcription in plant cells. In view of the conserved nature of the Gtl and Gt2 promoters, it is highly likely that Gtl is also a functional gene.

DISCUSSION

Our studies have demonstrated that the rice glutelin storage proteins are encoded by three subfamilies of genes. DNA sequence analysis has shown that components of the Gtl and Gt2 subfamilies are relatively conserved in all relevant gene regions including a t least 900 bp of the 5“flanking region, suggesting that they are products of relatively recent gene duplication events. A third gene, Gt3, shows significant divergence from the other glutelin genes, especially in its noncoding segments. All three Gt genes, however, share a highly conserved promoter region proximal (-125 to -1 bp) to the translational start and 3 small introns located at identical positions in the coding segments as well as encoded proteins of similar but not identical structures. The polymorphism of the derived amino acid sequences coded by these genes readily accounts for the size and net charge heterogeneity patterns exhibited by the glutelin fraction on two-dimensional polyacrylamide gels (Wen and Luthe, 1985).

Previous studies have demonstrated significant homology of the rice glutelin primary sequence to the legume 11 S


storage proteins and oat globulin (Takaiwa et al., 1987b; Higuchi and Fukazawa, 1987; Shotwell et al., 1988). This homologous condition at the amino acid level between legume and rice proteins, however, is not extendable to the noncoding regions particularly when the promoter regions of these genes are evaluated. Although both genes are expressed during seed development, the legume 11 S storage proteins are made and stored in embryogenic tissues, whereas the rice glutelins are synthesized and accumulated in the endosperm (Shotwell and Larkins, 1989). Except for the ubiquitous TATA box and a single RY repeat in Gt2, the promoter sequences of these genes are nonidentical, suggesting that regulatory elements conferring temporal expression have not been conserved.

It is noteworthy that nonconserved regions evident between the glutelin primary sequences are also present when comparisons are conducted among rice and legume 11 S proteins. These nonconserved regions seem prone to evolutionary change, particularly in the accumulation of insertions and deletions when 11 S proteins from different species are compared, suggesting that they are likely targets for genetic modification for improved nutritional qualities.

The similarity of the rice glutelin to legume 11 S proteins and oat globulin is also reflected at the genetic level. The legume 11 S proteins are encoded by a small multigene family, six genes for soybean glycinin and 8 genes for pea legumin (reviewed in Shotwell and Larkins, 1989), and the oat globulins are encoded by 7 to 10 genes (Shotwell et al., 1988). Interestingly, the rice glutelin gene subfamilies share this feature in containing the relatively small number of five to eight gene copies. The relative constancy of the number of gene copies per glutelin subfamilies would be consistent with the hypothesis that genes encoding the glutelin proteins al- ready existed as a multigene family which was then duplicated intact. Divergence by point and segmental mutations and subsequent homogenization by gene conversion would account for the diversity of Gtl/Gt2 and Gt3 gene subfamilies. More recent rounds of duplication and divergence of GtllGt2 would result in the generation of Gtl and Gt2 classes sharing closer sequence similarity.

The relatively small size of the glutelin gene family differs considerably from that displayed by the prolamine genes which encode the predominant protein fraction of most of the major cereals. The prolamine genes of maize (Shotwell and Larkins, 1988), wheat (Reeves and Okita, 1987), and rice (Kim and Okita, 1988) are encoded by a much larger gene family consisting 100 copies or more based on Southern blot analysis. The size difference of the glutelin and prolamine multigene families, however, is not consistent with the relative abundances of their gene products; glutelin is accumulated at 8-15-fold greater levels than prolamine, depending on the stage of seed development. Northern and dot blot analyses3 indicate that glutelin and prolamine mRNAs are present at the same abundance levels during endosperm development, but glutelin transcripts are more efficiently recruited into membrane bound polysomes. These observations suggest that both transcriptional and post-transcriptional events control storage protein biosynthesis in rice endosperm.

It is interesting to note that the primary sequences of glutelin and prolamine signal peptides are structurally distinct. The glutelin signal peptide contains 2 cysteine residues which are potentially capable of forming an intrachain linkage (Fig. 5), whereas that for prolamine (Kim and Okita, 1988) is one typically observed for secretory-like proteins (von Heijne, 1986). These possible functional differences in signal peptide sequences are consistent with our findings in an earlier study

W. T. Kim and T. W. Okita, manuscript in preparation.

(Krishnan and Okita, 1986) to process in vitro preprolamine but not preproglutelin.

In animal cells, the sorting of proteins to specific intracel- lular compartments is dictated by short peptide signals. For instance, a conserved peptide located on the carboxyl-terminus results in the retention of proteins within the endoplasmic reticulum (Munro and Pelham, 1987). As the rice glutelin is transported to protein bodies via the Golgi complex (Krishnan et al., 1987), we compared the glutelin sequence with other cereal endosperm proteins packaged by a similar mechanism. Such analyses revealed no common sequences, 4 residues or longer, except in one instance. A short peptide, Arg-Gln-Leu- Gln-Cys, located at residues 250-254 (Fig. 5) is shared by the wheat high molecular weight glutenin and is located near the mature amino terminus (Thompson et al., 1985). Although coincidental presence cannot be ruled out, the presence of this conserved peptide between structurally dissimilar proteins suggests that it may serve some functional role in targeting these proteins to protein bodies (Okita et al., 1988).

We present evidence here that the expression of the glutelin genes is not coordinate during endosperm development. Tran- scripts of the Gt3 subfamily are accumulated early during development, and their levels decline after 10 days post- anthesis, whereas maximum accumulation of Gtl and Gt2 transcripts occurs around 15 days, and these high levels are maintained a t subsequent stages. These differential patterns in mRNA accumulation are consistent with the marked differences in the promoter sequences of these genes. The Gt3 promoter diverges significantly a t -186 bp and beyond from the translational start when compared to the homologous G t l and Gt2 genes. Most conspicuously present in the promoter regions of the glutelin genes, especially in Gtl and Gt2, is the -300 bp element observed in other cereal genes (Kreis et al., 1985; Colot et al., 1987). Our demonstration that the glutelin promoters are transiently operational in tobacco protoplasts indicates that tobacco may be a useful system in the identification of sequences of these rice genes responsible for tissue specific expression.

REFERENCES

Amasino, R. (1986) Anal. Biochern. 152,304-307 Argos, P., Narayana, S. V. L., and Nielsen, N. C. (1985) EMBO J. 4,

Birnstiel, M. L., Busslinger, M., and Strub, K. (1985) Cell 41, 349-

Borroto, K., and Dure, L., I11 (1987) Plant Mol. Biol. 8, 113-131 Brown, J. W. S. (1986) Nucleic Acids Res. 7, 1513-1523 Colot, V., Robert, L. S., Kavanagh, T. A,, Bevan, M. W., and Thomp-

Dickinson, C. D., Evans, R. P., and Nielsen, N. C. (1988) Nucleic

Feinberg, A. P., and Vogelstein, B. (1983) Anal. Biochem. 132, 6-13 Fromm, M., Taylor, L. P., and Wolbot, V. (1985) Proc. Natl. Acad.

Gormon, C. M., Moffat, L. F., and Howard, B. H. (1982) Mol. Cell.

Heidecker, G., and Messing, J. (1983) Nucleic Acids Res. 11, 4891-

Henikoff, S. (1984) Gene (Amst.) 28, 351-359 Higuchi, W., and Fukazawa, C. (1987) Gene (Arnst.) 55, 245-253 Huynh, T. V., Young, R. A., and Davis, R. W. (1984) in Cloning

Technique: A Practical Approach (Glover, D., ed) pp. 49-78, IRL Press, Oxford

Jones, C. W., Rosenthal, N., Rodakis G. C., and Kafatos, F. C. (1979) Cell 18, 1317-1332

Joshi, C. P. (1987) Nucleic Acids Res. 15, 9627-9640 Kim, W. T., and Okita, T. W. (1988) Plant Physiol. 88, 649-655 Kreis, M., Shewry, P. R., Forde, B. G., Forde, J., and Miflin, B. J .

(1985) in Oxford Surueys of Plant Molecular and Cell Biology (Miflin, B. J., eds) Vol. 2, pp. 253-317, Oxford University Press, Oxford

1111-1117

359

son, R. D. (1987) EMBO J. 6, 3559-3564

Acids Res. 16, 371

Sci. U. S. A. 82, 5824-5828

Biol. 2, 1044-1051

4906

12580 Structural and Expression Studies of a Rice Storage Protein Multigene Family Krishnan, H. B., Franceschi, V. R., and Okita, T. W. (1986) Planta

Krishnan, H. B., and Okita, T. W. (1986) Plant Physiol. 81,748-753 Lutcke, H. A,, Chow, K. C., Mickel, F. S., Moss, K. A,, Kern, H. F.,

and Scheele, G. A. (1987) EMBO J . 6, 43-48 Lycett, G. W., Croy, R. R. D., Shirsat, A. H., and Boulter, D. (1984)

Nucleic Acids Res. 13, 6733-6743 Maniatis, T., Fritsch, E. F., and Sambrook, J. (1983) in Molecular

Cloning: A Laboratory Manual, Cold Spring Harbor Laboratories, Cold Spring Harbor, NY

Maxam, A. M., and Gilbert, W. (1980) Methods Enzymol. 65, 499- 560

Messing, J., Geraghty, D., Heidecker, G., Hu, N.-T., Kridl, J., and Rubenstein, I. (1983) in Genetic Engineering of Plants (Kosuge, T., Meridith, C., and Hollaender, A., eds) pp. 221-227, Plenum Pub- lishing Corp., NY

Moreira, M. A,, Hermodson, M. A., Larkins, B. A,, and Nielsen, N. C. (1981) Arch. Biochem. Biophys. 210,633-642

Munro, S., and Pelham, H. R. B. (1987) Cell 48,899-907 Murashige, T., and Skoog, F. (1962) Physiol. Plant. 15, 485-497 Murray, M. G. (1986) Anal. Biochem. 158, 165-170 Okita, T. W., Krishnan, H. B., and Kim, W. T. (1988) Plant Sci. 57,

Reeves, C. D., and Okita, T. W. (1987) Gene (Amst.) 52, 257-266 Reeves, C. D., Krishnan, H. B., and Okita, T. W. (1986) Plant Physiol.

Sanger, F., Nicklen, S. and Coulson, A. R. (1977) Proc. Nutl. Acad.

169,471-480

103-111

82,34-40

Sci. U. S. A. 74, 5463-5467

Scallon, B. J., Dickinson, C. D., and Nielsen, N. C. (1987) Mol. & Gen. Genet. 208, 107-113

Shotwell, M. A., Afonso, C., Davies, E., Chesnut, R. R., and Larkins, B. A. (1988) Plant Physiol. 87, 698-704

Shotwell, M. A., and Larkins, B. (1989) in The Biochemistry of Plants: A Comprehensive Treatise (Marcus, A., ed), Vol. 15, pp. 296-345, Academic Press, Orlando, FL

Slightom, J. L., Blechl, A. E., and Smithies, 0. (1980) Cell 21, 627- 638

%aswick, P. E., Hermodson, M. A., and Nielsen, N. C. (1984) J. Bid. Chem. 259, 13431-13435

Takaiwa, F., Kikuchi, S., and Oono, K. (1986) FEBS Lett. 206, 33- 35

Takaiwa, F., Ebinuma, H., Kikuchi, S., and Oono, K. (1987a) FEBS Lett. 221,43-47

Takaiwa, F., Kikuchi, S., and Oono, K. (1987b) Mol. & Gen. Genet.

Tanaka, K., Sugimoto, T., Ogawa, M., and Kasai, Z. (1980) Agric.

Thompson, R. D., Bartels, D., and Harberd, N. P. (1985) Nucleic

von Heijne, G. (1986) Nucleic Acids Res. 14, 4863-4690 Wen, T.-N., and Luthe, D. S. (1985) Plant Physiol. 78, 172-177 Yamagata, H., Sugimoto, T., Tanaka, K., and Kasai, Z. (1982) Plant

Zhao, W.-M., Gatehouse, J. A., and Boulter, D. (1983) FEBS Lett.

208, 15-22

Biol. Chem. 44, 1633-1639

Acids Res. 13,6833-6846

Physiol. 70, 1094-1100

162,96-102


5.2 6 . 3 8 . 2

18 4 1 3 2 . 3 4 . 3 2 0 8 3 0 . 3 5 . 2

1 . 9 4 . 3

0 7 3 9

a 4 a 5

5 . 8

2 0 . 2 3 0 2 6 2 9 2 . 3 7 . 7 0 8 4 7

8 1 5 . 5

0 6 4 1

1 . 9 1 . 9

1 3 . 6 8 1 5 . 8 J . l 6 . 3 4 1 1 9 1 . 1 3 . 1 4.1 6 . 3 1.1 0 5

8 8 7 8

6 . 3 1.3

1 2 0 14 5 10 3 10.3

2 5 7 . 4 7 . 1

2.6 4 3 5 . 0

8 . 1 4 . 1 3 . 0

8.5 1 7 1 . 0 4.1 0 . 4 5 . 4 4 . 2 6 . 6 4 9

7 3

1 3 O h 4 0

8.0

10 0 8 7

1 4 . 6 8 . 4 2 . 6 4 2 4 . 4

3 1 1.8

5 . 5 1 . 3 0 - 5

6 8

BamHl 6amHl EcoRl EcaRl EcoRl

4 I t 6 3

1 3 1 3 1 3 3 5

:7 5 2 13 1 2 5 3 11 10 1 1 11

1 6 18 5 5 9 9 1 5 9 5

I n

3 14 7

1 2 14 1 3 12 2 6

1 0 1 3 8 1

1 6 12 4 11 1 1 13 3 1 1 1

20 1 4 3 6 9 8 0 6 8 6

4 14

TGG

12 TGA TGT

5 13

TGC TAG

9 TAA 14 TAT 8 T R C 5 10 1 5 10 0 TCG I8 TCA 8 TCT 5 10

TCC CGG

12 CGR 1 2 CGT 7 CGC 9 CAG 6 1 6

CRR

1 3 CAT CAC

5 CTG 9 CTA 7 CTT 9 CTC

6 C C R 9 CCT 5 ccc

TTG TTR TTT TTC

n CCG

I 0 5 5 1 1 10 9 10 2

1 7

6 7 4 I 3 8 5 18 3 4 6 4 2 9 7 8 4 6 8 4

n

-.

3

I 6 1 0

1 3 3 9 2 6 17 0 6 9 4 3 1

1 0 4 1 6 3 8 I 5 3 8 6

0 9

12 7 3

n

~

3 0 4 6 0 1 10 6

1 2 5

14 8 I 3 4 b 1 1 3 3 8

42 7 6 1 3

1 3 8 0 8

1 0 3

7 v r 3 . 6 3 . 4 3 . 8 '$a I

4 1 3 . 6 J 5

. .

5 6 a 9 7 0 3 . 2

8 . 1 6 2

structure and expression of the rice glutelin multigene family* · 2001-07-21 · 12574 structural...

Documents