Barley aleurone cell development: molecular cloning of ... - Springer Link

Plant Molecular Biology 12: 285-293, 1989 © 1989 Kluwer Academic Publishers. Printed in Belgium

285

Barley aleurone cell development: molecular cloning of aleurone-specific cDNAs from immature grains Kjetill Jakobsen,1 Sonja Sletner Klemsdal, 1 Reidunn B. Aalen, l Marie Bosnes,1 Danny Alexander2, 3 and Odd-Arne Olsen 1.

1Department of Biology, Division of General Genetics, University of Oslo, PO Box 1031 Blindern, 0315 Oslo 3, Norway (*author for correspondence); 2Arco Plant Cell Research Institute, 6560 Trinity Court, Dublin, CA 94568, USA; 3present address: Calgene, 1920 Fifth Street, Davis, CA 95616, USA Received 20 July 1988; accepted in revised form 8 November 1988

Key words." aleurone, cDNA library, differential screening, DNA sequence, tissue-specific expression Abstract

The cloning of 11 different homology groups of cDNAs representing genes expressed in aleurone, but not in starchy endosperm of 20-day-old barley grains is described. Among the cDNAs, four are aleurone-specific, while the remaining are also expressed in the embryo, but not in any other part of the plant. Sequence analysis of one of the aleurone-specific clones, BllE, reveals an open reading frame coding for an unidentified 10.4 kDa protein with a putative signal sequence and a possible metal-binding finger. The BllE gene has a high GC content in the 5' leader sequence (63%), as well as in the coding region (70%) compared to known cDNAs from the barley starchy endosperm. Northern analysis of BllE indicates maximum mRNA abundance around mid-phase of grain development. When isolated immature aleurone/pericarp is incubated in tissue culture medium (MS) the BllE message disappears, indicating a requirement for a diffusible factor from the intact grain for its continued presence.

Introduction

The aleurone of barley grains is a suitable model system for developmental studies for several reasons. First, it starts to differentiate in the cellular endosperm approximately 5 days post anthesis (dpa), and forms a well defined layer surrounding the starchy endosperm. Aleurone cell formation is thus well defined both in time and space. Second, a number of zygotic mutants affecting aleurone cell differentiation have been isolated [3]. In addition, the existence of maternal mutants affecting endosperm development [7] may indicate that maternal factors are involved in aleurone development. The present study is aimed at isolating aleurone-

specific cDNAs for use as tissue markers in developmental studies. In spite of the early discovery of aleurone cells [32], as well as their recent use in hormone studies, resulting in the isolation of several highly abundant cDNAs [29, 36, 37, 38], little is known about the genes or gene products participating in aleurone cell development. One reason is that developing aleurone cells have not previously been isolatfd separate from adhering starchy endosperm and maternal tissue. Aleurone cells have a conspicuous morphology; they contain several distinct organelles such as aleurone grains [12] consisting of phytin-containing globoids, protein carbohydrate bodies, and a ground

286 matrix [4, 18, 19]. Other aleurone organelles are lipid-containing spherosomes and microbodies [18, 19]. The specific functions of different organelles during development or germination have not been identified. However, the unique ultrastructure of aleurone ceils indicates that several aleurone-specific genes may be involved in aleurone cell development. In this communication we report the isolation of aleurone-specific cDNA clones from developing grains. The clones were isolated by differential screening of an aleurone cDNA library from 20 dpa grain, using cDNA complementary to starchy endosperm mRNA as an excluding probe. The stage of 20 dpa corresponds approximately to the mid-phase of barley grain development and to the peak of storage protein mRNA synthesis in the starchy endosperm [28, 34].

ter germination. Roots, ranging in length from 2 to 5 cm, were collected from germinating seeds.

Microscopy

Tissue was prepared for light microscopy and embedded in plastic according to Olsen and Krekling [31].

Isolation of nucleic acids and poly(A) + RNA selection

Poly(A) ÷ RNA was isolated as described by Mundy et ai. [27], total RNA as described by Galau et aL [9].

In vitro translation Materials and methods

Plant material and culture conditions

Barley (Hordeum vulgare L. vat. disticum cv. Bomi) plants were cultivated as described earlier [21]. Immature and mature aleurone layers and embryos were incubated in the presence of abscisic acid (ABA; 50/~M) and gibberellic acid (GA3; 10 #M) [28]. Germination of mature barley seeds and incubation of aleurone layers were according to Mozer [26].

Prepai'ation of tissue

After removal of palea, lemma and the embryo from grains harvested 20 dpa, the starchy endosperm was squeezed out and immediately frozen in liquid nitrogen. The remaining aleurone/pericarp, contaminated by some remaining starchy endosperm, was either frozen directly in liquid nitrogen or, to obtain pure aleurone, cut in two and scraped to remove adhering starchy endosperm cells. The aleurone layer was then peeled off the pericarp under a dissecting microscope and frozen. Embryos were hand-dissected from 20 and 30 dpa grains, whereas light-grown plantlets were harvested approximately one week af-

Poly(A) + RNA was in vitro translated and separated on SDS-PAGE [22]. Immunoprecipitation of 105 cpm of total translate with antibodies against protein fraction X [14], the proteins of beer, was performed as described by Mundy et al. [28]. The antibodies were kindly provided by J. Mundy.

Construction of cDNA libraries and differential screening

DNA complementary to poly(A) + RNA of pure 20 dpa aleurone was cloned into the polylinker of the pARC7 vector by the dimer-primer method of Alexander et al. [2]. Transformation of competent DH-5ot Escherichia coli cells yielded 3.2 × 10 6 colonies per pmol vector ends. Control ligation (without linker) indicated that approximately 10°70 of the colonies were non-recombinant. Average length of the inserts was 450 bp. Differential screening for aleurone-specific cDNAs was done by plating 3 × 104 cells according to Taub and Thompson [40]. The filters were first hybridized to [32p]_dCTP_labelled first-strand cDNA complementary to poly(A) + RNA from starchy endosperm. After 24 h exposure, the filters were discharged by boiling in 1 mM EDTA (pH 8.0)

287 and rehybridized to cDNA complementary to poly(A) + RNA from aleurone/pericarp. Aleuronepositive clones obtained by this procedure were then rescreened by hybridizing to cDNA complementary to poly(A) + RNA (prepared as above) from 20 dpa pericarp, 30 dpa embryos, roots and green leaves. A second cDNA library was constructed from poly(A) ÷ RNA of 20 dpa aleurone/pericarp in lambda ZAP (Stratagene) as described by Crawford [5], except that a Biogel A50 column was used to eliminate unligated linkers [17].

Hybridizations and cDNA probes Total RNA was separated by glyoxal agarose gel electrophoresis [24] or by formaldehyde agarose gel electrophoresis [10] and blotted onto nylon membranes [10, 11]. Hybridization criteria were as indicated in figure captions. Cross hybridization between aleurone-specific cDNAs was determined by hybridizing replicate blots of enzyme-digested cDNA clones to each of the purified nick-translated inserts. The following cDNA clones were used for Northern hybridizations: B3- and C-hordein (clones pB7 and pcP387, respectively) [8]. The pcP387 and pB7 clones were kindly provided by Dr M. Kreiss, and PAPI (Alf) [29] by Dr J. Rogers.

Results

Purity of aleurone preparations Plastic-embedded sections of aleurone preparation demonstrate that they were virtually free from contaminating starchy endosperm cells (Fig. 1). Only occasionally were remains of sub-aleurone cell walls and adhering starch granules seen. Likewise, the starchy endosperm preparations (not shown) were correspondingly free from aleurone cells. B- and C-hordeins are the most abundant storage proteins of the starchy endosperm and their mRNAs are thus suitable for assessing the purity of aleurone mRNA preparations. As shown in Figure 2 only weak hybridization was found for both probes in the aleurone, the amount of B- and C-hordein mRNA being approximately 150-fold higher in the starchy endosperm than in the aleurone.

Differential screening for aleurone-specific cDNAs. Screening of the aleurone cDNA library (in pARC7) resulted in 63 aleurone-positive, starchy endospermnegative clones. Further hybridization of the 63 clones to cDNA complementary to mRNA from pericarp, young green leaves and roots were all

DNA sequencing Inserts from the pARC7 cDNA library were subcloned into the vector Bluescript KS + (Stratagene) and sequenced in both directions by using the Sequenase kit from United States Biochemical Corporation. To sequence inserts from the lambda cDNA library, Bluescript SK- vector containing the inserts of interest were formed by the lambda ZAP automated excision process. Internal 17-mer oligonucleotide sequencing primers were synthesized on a Pharmacia Gene Assembler.

Fig. L Transversesection(1 /~m,stainedwithtoluidineblue) of aleurone froma pure aleuronepreparation of 20-day-oldgrains. Starchyendospermside is facingdown.Blackspheresrepresent aleurone grains.

288

Fig. 2. Demonstrationof the purityof the isolatedaleuronelayers. Northernblots of total RNA frompure aleurone(AL)and starchyendosperm(SE) from20-day-oldgrain hybridizedto Bhordein (pB7) and C-hordein(pCP387) probes. Each lane contains 5 gg total RNA. Hybridizationconditionswere 5x SSC, 50~/0formamide,5x Denhardt'ssolutionat 42°C and washing was done at 0.Sx SSC, 42°C. negative. When hybridized to cDNA complementary to mRNA from 30 dpa embryos, however, only 14 of the 63 clones were completely negative. The remaining 51 clones, belonging to at least 7 crosshybridization groups, hybridized relatively strongly to cDNA from the embryo. On the basis of cross-hybridization studies and restriction mapping, the 14 aleurone-positive pARC7 inserts were assigned to 4 groups, termed BILE, B14A, B12A and B14D, after one clone from each group. In this paper, we analyze two of the clones, pBllE and pB12A in the pARC7 vector and their full-length lambda ZAP versions zBllE and zB12A. The lambda clones were selected by plating 20000 lambda ZAP recombinants on duplicate filters, which gave 3 positives with B12A as probe and 21 clones with BllE,

Sequence analysis The sequences within the overlap of pBllE (189 bp) and its full-length lambda version zBllE (485 bp)

were found to be identical (data not shown). The pARC7 insert contained a poly(A) tail of 93 bp, compared to only 8 bp in zBllE. The sequence contains only one open reading frame, with eight potential translation start codons between nucleotides 58 and 121 (ATG, underlined, Fig. 3A). With only one known exception [16] plant RNAs initiate translation at the first ATG codon [13]. Nucleotides in the vicinity of the first ATG (an A in position - 15, a purine in position - 3 , a G in position +4 and a C at + 5) fit the consensus plant initiation sequence [13, 23]. Searching nucleic acid and protein databases, no protein homologous to BllE was found. The deduced protein sequence consists of 102 amino acids (Fig. 3A), with an estimated molecular weight of 10.4 kDa, and an isoelectric point of 6.9. The aminoterminal part of this protein has all the characteristics of a signal peptide, with charged residues preceding a hydrophobic region of 17 amino acids. Two potential cleavage sites were found (arrows, Fig. 3A). The processed protein would be 74 or 71 amino acids long with an estimated molecular weight of 7.5 or 7.3 kDa, respectively. In the 119 bp 3' untranslated region the consensus polyadenylation sequence AATAAA is located 60 nucleotides upstream of the poly(A) tail. BIlE has an unusually high G + C content. The coding sequence is 70% G+C. The 5' untranslated region is 63% G+C, while the 3' untranslated region has a lower G + C content (54%). Southern blots probed with BIlE insert indicate that the gene is present in one or only a few copies (Fig. 4), although reconstitution experiments are needed to determine the exact number. Southern hybridizations of BIlE at lower stringency (Tm28°C) did not give additional bands (data not shown). Sequence analysis of the second aleurone-specific clone, zB12A, identified the insert as PAPI, a previously cloned gene from mature aleurone [29]. We have sequenced three independent clones in this group. With the exception of three single-base insertions and two base substitutions in the nontranslated regions, our sequences were identical with that of PAPI. These minor sequence changes may reflect the fact that different cultivars have been used in the two studies.

289

A 1 CGAGAGCGAGCGTGTGAGTGTAGCCGAGTAGATCACCGTACGACGACGACGAGGGGCATG 1 MET 61 GCGATGGCGATGGGGATGGCGATGAGGAAGGAGGCAGCGGTGGCCGTGATGATGGTGATG 2 AIaMETAIaMETGIyMETAlaMETArgLysGIuAIaAIaValAIaValMETMETValMET 121 GTGGTGACGCTGGCGGCGGGTGCGGACGCGGGAGCGGGAGCGGCGTGCGAGCCGGCGCAG 22 V a l V a l T h r I ~ l ~ l a G l ~ A l ~ s p A l % G 1 y A l a G l y A l ~ l a C y s G l u P r o A 1 a G l n 181 CTGGCGGTGTGCGCGTCGGCGATCCTGGGCGGGACGAAGCCGAGCGGCGAGTGCTGCGGG 42 LeuAlaValCysAlaSerAlaIleLeuGlyGlyThrLysProSerGlyGluCysCysGly 241 AACCTGCGGGCGCAGCAGGGGTGCTTGTGCCAGTACGTCAAGGACCCCAACTACGGGCAC 62 AsnLeuArgAlaGlnGlnGlyCysLeuCysGlnTyrValLysAspProAsnTyrGlyHis 301 TACGTGAGCAGCCCACACGCGCGCGACACCCTCAACTTGTGCGGCATACCCGTACCGCAC 82 TyrValSerSerProHisAlaArgAspThrLeuAsnLeuCysGlyIleProValProHis 361 TGCTAGCCGCCTAGCCGATCGAGGGCTCCAGGCACGCATGCATGTTCCTGTTATGTGGAT 102 Cys*** 421

G'I~~TGC~GGTGATCTATGGCGGCTAGCTTGC'I~CCTGGC~rAGCAGC~GCTG

481 TAATG (A)93

B zBIIE DNBEVI

30 D A G A G A ~ p E 1021 T R K F V L

ZBIIE DNBEVI

56SGEC 1046 E S S

L

P~GQ L A ~ S

A IIL G GIT K p

R

S S

F--

D

AQQGCLCQY

Q

G I I S E G G A A

G G

A H

C zBIIE HaG5

55 P S G E ~ G 79

Q L Q Q

D~R Q E

A Q Q~Q~L~Q Q N I E

Q

Y~D E A

P N Y Q V

F R

Fig. 3. A. Complete nucleotide and amino acid sequence of the aleurone specific zBllE cDNA insert. ATG codons are underlined and the putative polyadenylation signal is boxed. The suggested cleavage sites of the signal sequence is indicated by arrows. B. Alignment of part of the BIlE amino acid sequence with the region of the herpes simplex virus protein (see in the text). C. Alignment of part of the BIlE amino acid sequence with a region of the sunflower storage protein (see text).

Expression of aleurone-specific messages The B l l E c D N A hybridizes to a single transcript in 20 and 30 d p a aleurone (Fig. 5), corresponding roughly to the size predicted f r o m the sequence in Figure 3A. N o signal was observable on N o r t h e r n

blots o f R N A f r o m 20 dpa starchy endosperms or 20 and 30 dpa embryos, confirming the results o f the initial screenings. F r o m Figure 5 it also appears that B I l E was present at its m a x i m u m level a r o u n d 20 dpa. The level in mature aleurone was roughly 100-fold lower t h a n at 20 dpa. Estimates f r o m dot

290 Fig. 4. Hybridization of [32P]-1abelled insert BllE to genomic DNA from barley. DNA (5 #g) digested with Eco RI (E) or Hind III (H) was separated on 0.8% agarose gels and transferred onto GeneScreen nylon membrane. Hybridization and washes were in 1x SSC equivalent buffer, at Tin-15 °C as described by Galau et al. [8]. Marker lanes contain pBR325 fragments.

blots of relative transcript abundances indicated that BIlE was approximately 6-fold more abundant at 20 dpa than PAPI (B12A) (data not shown). When isolated mature or' immature aleurone/pericarp layers were incubated for 20 h with ABA or GA3, no regulatory effect on transcript abundance of BIlE could be observed (Fig. 5). During the 20 h incubation of aleurone/pericarp in MS medium, however, the BIlE message declined at least 100-fold (Fig. 5). Subsequent hybridization of this Northern filter to PAPI, an ABA-inducible cDNA and a ribosomal probe (data not shown) demonstrated that RNA from incubated aleurone layers was not degraded non-specifically.

Fig. 5. Steady-state BIlE mRNA levels in aleurone, starchy endosperm and embryo. Northern blot of total RNA from 20 and 30 dpa aleurone/pericarp and mature aleurone extracted directly after dissection (to), incubated in 1/2 x MS medium [30] (M) or in MS medium containing 50 #M abscisic acid (A), 10 tzM gibberellic acid (G). The blot was hybridized to [32p]-labelled insert of BllE. Hybridizations and washes were in 1× SSC equivalent buffer at 68 °C (Tm-15 °C).

291 Detection o f a putative B I l E protein

Discussion

A possible candidate protein corresponding to the BllE message was found by immunoadsorbing proteins translated in vitro from mRNA of aleurone, embryo and starchy endosperm with antibodies against protein fraction X. With these antibodies a 10 kDA protein was precipitated from the translation products of aleurones, but not from the starchy endosperm or the embryo (Fig. 6). This protein thus fits both the pattern of expression as well as the molecular weight of BllE, since no protein processing occurs in such translation systems.

Isolation of sufficient quantities of pure aleurone cells by dissection is optimal around 20 dpa. At earlier stages the tissue is fragile and the yield is low. After 20 dpa the thickening of the sub-aleurone cell walls prevents the separation of starchy endosperm from the aleurone/pericarp. Genes expressed exclusively at development stages before or after 20 dpa will therefore escape cloning by the present methods. Identification of the function of the cloned genes in aleurone cell development must await further experiments. The grouping of the aleurone positive, starchy endosperm negative clones into embryo negative and embryo positive, is, however, interesting in the light of the similar function of the aleurone and the scutellum. During germination, both cell types produce a-amylase and/~-glucanase [25, 27]. The B11E transcript and the remaining potentially aleurone-specific clones may therefore represent functions exclusive to aleurone, whereas clones found in both the aleurone and the embryo may represent common characteristics of the two tissues. No specific function for BllE can be suggested, since no homologous protein has been characterized. The transcript shares one feature with other known barley aleurone transcripts, including oramylases and PAPI, namely G + C richness. BllE has a G + C content of 70% in the coding region. In contrast, a G + C content of approximately 50°70 has been found in genes expressed in starchy endosperm [15, 35]. A special feature of the BllE sequence, not found in any previously studied plant mRNA [13, 36], is the high G + C (630/0) content of the 5' untranslated leader. An additional feature of BIlE is that the consensus polyadenylation sequence AATAAA is 60 nucleotides upstream of the poly(A) addition site and not 25 to 30 as commonly found [13]. It remains to be seen whether the observed codon bias and the unusual location of the polyadenylation signal is of significance in the regulation of aleurone-specific gene expression. Although no protein homologous to the open reading frame of BIlE was found, two cases of similarity within limited regions of the protein sequence deserve mentioning, although they are not

Fig. 6. SDS°PAGE of in vitro translated proteins of aleu-

rone/pericarp (lane 4), starchyendosperm(lanes 1 and 3) from 20 dpa grains and 30 dpa embryos (lane2) immunoabsorbed withantibodiesagainstproteinfractionX. Arrowindicatespossible BllE protein.

292 statistically significant. The first, which obtains a score of 2.6 using the Dayhoff MDM-78 matrix (number of random runs = 100) [6], is a 30 amino acid overlap of the Herpes simplex virus DNAbinding protein [33] (Fig. 3B). This region, extending from amino acid 36 to 64, shows 40% homology. It contains three cysteines which might be a part of a metal-binding "finger", the fourth being located in position 69 (arrows, Fig. 3B). Similarity is also present for the stretch around the last two cysteines of the putative finger with the metal binding site of a group of metallothioneines [20]. The combination of a leader peptide and a metal-binding finger may seem surprising. However, this has recently been found in a nodulin gene family [39]. The nodulin genes contain two highly conserved fingers in addition to the leader sequence. The second case of similarity, which has a score of 2.2 [6], is within a part 18 amino acids long of a sunflower seed storage protein [1] (Fig. 3C). This possibility is interesting in light of the presence of storage protein in the aleurone grains [3, 16, 17] which is not present in the scutellum. If the in vitro synthesized protein immunoadsorbed with protein fraction X (beer protein) antibodies is the BllE protein (Fig. 6), its presence in mature aleurone can be inferred. Studies using antibodies directed against a BllE fusion protein are underway to confirm its existence, as well as its sub-cellular localization in vivo. The main purpose of the present work was to isolate aleurone-specific cDNAs for the use in studies of aleurone cell differentiation in wild-type and mutant endosperms with abnormal aleurone [3]. We feel that BllE has several characteristics suited for this purpose, such as strict tissue specificity, as well as an interesting pattern of expression in organ culture. This may open up for studies of the interaction between cis-regulatory elements and trans-acting factors. The disappearance of the B11E message during organ culture of aleurone, if due to an effect at the level of transcription, may allow identification of factors involved in communication between the plant tissues.

Acknowledgements We thank P. Filner, formerly at ARCO PCRI, Dub1in, California, R. Davis, Stanford University and G. Galau, University of Georgia, in whose laboratories some of the experiments were carried out, for valuable help and support. Thanks are also due to S. Engebretsen and A. Keiserud for excellent technical assistance. This work was supported by the Norwegian Agricultural Research Council (NLVF).

References 1. Allen RD, Cohen EA, Vonder Haar RA, Adams CA, Ma DP, Nessler CL, Thomas TL: Sequence and expression of a gene encoding an albumin storage protein in sunflower. Mol Gen Genet 210:211-218 (1987). 2. Alexander D, McKnight TD, Williams BG: A simplified and efficient vector-primer cDNA cloning system. Gene 31: 7 9 - 89 (1984). 3. Bosnes M, Harris E, Aigeltinger L, Olsen O-A: Morphology and ultra-structure of 11 barley shrunken endosperm mutants. TAG 74:177-187 (1987). 4. Buttrose MS: Ultrastructure of barley aleurone cells as shown by freeze-etching. Planta (Berl.) 9 6 : 1 3 - 2 6 (1971). 5. Crawford NM, Campbell WH, Davis RW: Nitrate reductase from squash: cDNA cloning and nitrate regulation. Proc Natl Acad Sci USA 83:8073-8076 (1986). 6. Dayhoff MO: In: Atlas of protein sequence and structure, vol. 5, suppl. 3, pp. 1-8. National Biomedical Foundation, Washington DC (1978). 7. Felker FC, Peterson DM, Nelson OE: Anatomy of immature grains of eight maternal effect shrunken endosperm barley mutants. Am J Bot 72:248-256 (1985). 8. Forde BG, Kreis M, Williamson MS, Fry RP, Pywell J, Shewry PR, Bunce N, Miflin BJ: Short tandem repeats shared by B- and C-hordein cDNAs suggest a common evolutionary origin for two groups of cereal storage protein genes. EMBO J 4 : 9 - 1 5 (1985). 9. Galau GA, Legocki AB, Greenway SC, Dure L III: Cotton messenger RNA sequences exist in both polyadenylated and nonpolyadenylated forms. J Biol Chem 256:2551-2560 (1981). 10. Galau GA, Hughes DW, Dure L III: Abscisic acid induction of cloned cotton late embryogenesis-abundant (Lea) mRNAs. Plant Mol Biol 7:155-170 (1986). 11. Galau GA, Bijaisoradat N, Hughes DW: Accumulation kinetics of cotton late embryogenesis abundant mRNAs and storage protein mRNAs: Coordinate regulation during embryo genesis and the role of abscisic acid. Dev Biol 123:198-292 (1987).

293 12. Guillermond A: Recherches cytologiques sur la germination des grains de Gramin6es et contribution h l'etude des grains d'aleurone. Arch Anat Micr Morph Exp 10:9-43 (1908). 13. Heidecker G, Messing J: Structural analysis of plant genes. Ann Rev Plant Physiol 37:493-466 (1986). 14. Hejgaard J, Sorensen SB: Characterization of a protein-rich beer fraction by two-dimensional immunoelectrophoretic techniques. C R Trav Lab Carlsberg 40:187-203 (1975). 15. Hejgaard J, Rasmussen SK, Brandt A, Svendsen J: Sequence homology between barley endosperm protein Z and protease inhibitors of the al-antitrypsin family. FEBS Lett 180: 89- 94 (1985). 16. Hoffman LM, Ma Y, Barker RF: Molecular cloning of Phaseolus vulgaris lectin mRNA and use of cDNA as a probe to estimate lectin transcript levels in various tissues. Nucleic Acids Res 10:7819-7828 (1982). 17. Huynh TV, Young RA, Davis RW: Constructing and screening cDNA libraries in gtl0 and gtll. In: Glover DM (ed) DNA Cloning: A Practical Approach, vol. 1 pp. 4 9 - 78. IRL Press, Oxford (1985). 18. Jacobsen JV, Knox RB, Pyliotis NA: The structure and composition of aleurone grains in the barley aieurone layer. Planta (Berl.) 101:189-209 (1971). 19. Jones RL: The fine structure of barley aleurone cells. Planta (Berl.) 85:359-375 (1969). 20. Karin M, Richards RI: Human metallothionein genes: molecular cloning and sequence analysis of the mRNA. Nucleic Acids Res 10:3165-3173 (1982). 21. Klemsdal SS, Kvaale A, Olsen O-A: Effects of the barley mutants Riso 1508 and 527 high lysine genes on the cellular development of the endosperm. Physiol Plant 67:453-459 (1986). 22. Klemsdal SS, Olsen O-A, R~rvik KA: The barley high lysine genes of mutants 1508 and 527 after hordein polypeptide composition quantitatively, but not qualitatively. Hereditas 107:107-114 (1987). 23. Kozak M: Point mutations define a sequence flanking the AUG initiator codon that modulates translation by eucaryotic ribosomes. Cell 44:283-292 (1986). 24. Maniatis T, Fritsch EF, Sambrook J: Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY (1982). 25. McFadden GI, Ahluwalia B, Clarke AE, Fincher GB: Expression sites and developmental regulation of genes encoding (13, 1-4)-/3-glucanases in germinated barley. Planta 173: 500-508 (1988). 26. Mozer TJ: Control of protein synthesis in barley aleurone layers by the plant hormones gibberellic acid and abscisic acid. Cell 20:479-485 (1980).

27. Mundy J, Brandt A, Fincher GB: Messenger RNA from the scutellum and aleurone of germinating barley encode (1-3, 1-4)-~-D-glucanase, a-amylase and carboxypeptidase. Plant Physiol 79:867-871 (1985). 28. Mundy J, Hejgaard J, Hansen A, Hallgren L, Joergensen KG, Munck L: Differential synthesis in vitro of barley aleurone and starchy endosperm proteins. Plant Physiol 81: 630-636 (1986). 29. Mundy J, Rogers JC: Selective expression of a probable amylase/protease inhibitor in barley aleurone cells: Comparison to the barley amylase/subtilisin inhibitor. Planta (Berl) 169: 51-63 (1986). 30. Murashige T, Skoog F: A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol Plant 15: 473-497 (1962). 31. Olsen O-A, Krekling T: Grain development in normal and high lysine barley. Hereditas 93:147-160 (1980). 32. Pfeffer W: Untersuchungen uber die Protein-korper und die Bedeutung des Asparagins beim Keimen der Samen. Jb Wiss Bot 8:529-571 (1872). 33. Quinn JP, McGeoch DJ: DNA sequence of the region in the genome of herpes simplex virus type 1 containing the genes for DNA polymerase and the major DNA binding protein. Nucleic Acids Res 13:8143-8163 (1985). 34. Rahman S, Kreis M, Forde BG, Shewry PR, Miflin BJ: Hordein-gene expression during development of the barley (Hordeum vulgare) endosperm. Biochem J 223:315-322 (1984). 35. Rasmussen SK, Hopp HE, Brandt A: Nucleotide sequences of cDNA clones for B1 hordein polypeptides. Carlsb Res Commun 48:187-199 (1983). 36. Rogers JC, Miltiman C: Isolation and sequence analysis of a barley (Hordeum vulgare) c~-amylasecomplementary D N A clone. J Biol Chem 258:8169-8174 (1983). 37. Rogers JC: Two barley c~-amylase gene families are regulated differently in aleurone cells. J Biol Chem 260(6): 3731- 3738 (1985). 38. Rogers JC: Aleurain: a barley thiol protease closely related to mammalian cathepsin H. Proc Natl Acad Sci USA 82: 6512-6516 (1985). 39. Sandal NN, Bojsen K, Marcker KA: A small family of nodule specific genes from soybean. Nucl Acids Res 15: 1507-1519 (1987). 40. Taub F, Thompson EB: An improved method for preparing large arrays of bacterial colonies containing plasmids for hybridization: In situ purification and stable binding of DNA on paper filters. Anal Biochem 126:222-230 (1982).