Plant Molecular Biology 19: 433-441, 1992. © 1992 Kluwer Academic Publishers. Printed in Belgium.
433
Cloning and expression of an embryo-specific mRNA up-regulated in hydrated dormant seeds Peter J. Goldmark 1' 2,, Jeanne Curry2, Craig F. Morris 2 and M.K. Walker-Simmons 2 I DJR Research, Star Route 69, Okanogan, WA 98840, USA; 2Dept. of Crop and Soil Sciences, Washington State University, U.S. Department of Agriculture/Agricultural Research Service, Wheat Genetics, Quality, Physiology and Disease Research, Johnson Hall 209, Washington State University, Pullman, WA 99164-6420, USA (* author for correspondence) Received 27 November 1991; accepted in revised form 14 February 1992
Key words: abscisic acid, Bromus, dehydration, dormancy, embryos, transcript levels Abstract
Dormant seeds do not germinate when imbibed in water even when conditions are favorable for germination. These hydrated seeds remain viable, but growth-arrested for weeks due to unknown restrictions within the embryo. As a model system for the study of the molecular processes occurring in dormant seeds, we have chosen to examine gene expression in Bromus secalinas, a grass species that produces seeds with high levels of embryonic dormancy. Using differential screening for mRNAs present in hydrated dormant embryos, we have identified a cDNA clone, pBS128, that encodes a mRNA transcript found in the embryos of hydrated seeds of B. secalinus as well as in embryos from mature dry seeds. Striking differences in pBS 128 transcript levels appear upon hydration of dormant and nondormant seeds. Upon imbibition pBS128 transcript levels increase over four-fold in dormant seeds, but rapidly decline and disappear in nondormant seeds, which subsequently germinate. The pBS 128 transcript appears to be embryo-specific since the transcript is not detectable in either non-stressed or dehydrated seedling tissue. Application of 50 #M ABA to nondormant seeds arrests germination and enhances pBS128 transcript levels. The nucleotide sequence of the nearly full-length pBS128 cDNA shows no homology to other reported genes, and the putative protein sequence does not exhibit the hydrophilic characteristics of the ABA-responsive LEA (late embryogenesis abundant) proteins.
Introduction
Seed dormancy is an important developmental state of plant growth whose molecular and bio-
chemical processes remain virtually unknown. Research in this field has focused on the physiological differences between dormant seeds and nondormant seeds. Viable seeds are dormant if
The nucleotide sequence data reported will appear in the EMBL, GenBank and DDBJ Nucleotide Sequence Databases under the accession number X63202. Contribution from the U.S. Department of Agriculture, Agriculture Research Service, and the College of Agriculture and Home Economics Research Center, Washington State University, Scientific Paper No. 9101-53. Mention of a specific product name by the U.S. Department of Agriculture is not an endorsement and does not imply a recommendation over other suitable products.
434 they do not germinate when hydrated, while seeds that germinate promptly are termed nondormant. Studies have shown that placing dormant and nondormant seeds in water results in similar rates of water uptake and new protein synthesis during the first few hours of imbibition [ 1]. Yet, undiscovered biochemical and molecular restrictions prevent cell expansion and germination in dormant seeds. Proposed mechanisms for the maintenance of dormancy in mature seeds have included structural restraints preventing water uptake and gas exchange or endogenous germination inhibitors found in the seed coat and embryo (reviewed in [ 1, 16]). Genetic evidence indicates that the plant hormone abscisic acid (ABA) is required for the induction of dormancy during seed development. Plant mutants deficient in ABA or nonresponsive to ABA produce nondormant seeds [12]. The role of ABA in maintaining dormancy in mature seeds is not yet clear, in part because mature dormant and nondormant seeds have similar ABA levels [ 19]. We are interested in identifying the molecular processes that prevent dormant seeds from germinating when moistened. We have chosen Bromus secalinas as a model system for dormancy because seeds of this wild grass species are uniformly dormant upon maturity and can be converted to nondormant seeds by a short period of postharvest storage at room temperature. Dormant B. secalinas seeds possess true embryonic dormancy meaning that embryos isolated from dormant seeds retain dormancy. The capability of wild grass seeds such as B. secalinas seeds to remain hydrated and dormant in the soil causes major agricultural problems. After extended periods the dormant seeds can sporadically germinate and successfully compete with cultivated crops for light and nutrients. In previous work we determined that wheat seeds in the hydrated dormant state exhibit prolonged expression of ABA-responsive genes [ 13 ]. Embryos isolated from dormant wheat seeds do not retain the dormancy characteristics of intact seeds and germinate within 1 to 2 days after imbibition [ 13]. Dormancy can be restored in iso-
lated wheat embryos by application of ABA. We found that application of low concentrations of ABA to isolated embryos resulted in the enhanced expression of a set of ABA-responsive genes in dormant seed embryos compared to nondormant. For seeds such as wheat with low levels of embryo dormancy we proposed that the enhanced response of dormant grain embryos to ABA could be a dormancy mechanism [ 13, 20]. In this report, we have examined gene expression in hydrated seeds of B. secalinas. Embryos from seeds of this species retain the dormancy characteristics of intact seeds and require a period ofpostharvest storage to dissipate dormancy. We have identified a gene that is up-regulated in expression and whose m R N A accumulates when dormant seeds are imbibed and remain growtharrested. This embryo-specific transcript appears to be regulated by ABA and has a possible role in dormancy maintenance.
Materials and methods Plant materials Bromus secalinas (cheat) seeds were collected from fields in northcentral Washington and propagated under controlled greenhouse conditions with 29 °C day/18 °C night temperatures. Mature seeds were separated from the plant and dried at room temperature to ensure a uniform moisture content of 12~o. A portion of the seeds was frozen at -20 ° C to preserve dormancy. Dormancy in the remaining seeds was dissipated by postharvest storage at room temperature in the dark. After postharvest storage the seeds were fully gerruinable and hydration of seeds resulted in 96100~o germination within 2 days. All germination assays were performed with embryos or whole seeds with the seed coating removed. Embryos were hand-dissected from intact seeds. Embryos or whole seeds were placed in Petri dishes on water-saturated blotters and incubated at 20 °C in the dark. Embryos or seeds were considered germinated when the coleorhiza or radicle was greater than 1 mm in length.
435
B. wctorum (downy brome) and Secale cereale (common rye) seeds were collected from fields in northcentral Washington; Avenafatua (wild oat) and Aegilops cylindrica (jointed goatgrass) were collected from the Spillman Conservation Farm near Pullman, WA. These wild grass species were propagated and stored as described above. The storage time required to dissipate dormancy is species-dependent and can be as short as 4 weeks for B. secalinas and as long as 1 year for A. fatua.
RNA extraction and northern blot analysis Total RNA was purified by pulverization of isolated embryos in liquid nitrogen followed by proteinase K digestion [2] and two phenol/ chloroform extractions. Total RNA was then recovered by precipitation with 2 M lithium chloride. RNA (15#g/lane) was denatured and electrophoresed in 1 ~o (w/v) agarose formaldehyde gels [14]. Gels were rinsed briefly in water and 10 x SSC, and then blotted to Zeta-probe (Bio-Rad, Richmond, CA) using 10 x SSC as the transfer medium. Blots were rinsed for 10 min in 2 x SSC, air-dried, and vacuum-baked for 2 h at 80 ° C. Preparation of 32p-labeled probe, hybridizations and autoradiography were accomplished by standard techniques [14]. Blots were quantified using a radioanalytic imaging system (Ambis, San Diego, CA).
cDNA library construction and screening Total RNA was purified from 800 dormant B. secalinas embryos as described above. Poly(A) + RNA was isolated from total RNA by the oligo(dT) cellulose batch method [13] and directionally cloned into 2Zap II (Stratagene, La Jolla, CA) using a Not I adapter/primer by standard procedures [10]. The primary library contained 2.1 x 106 recombinant phage. 32p-labeled cDNA synthesized from poly(A) + RNA according to the method of Morris et al. [13] was used in differential screening of dormant and nondormant embryos imbibed for 10 h in water at 20 ° C.
Conversion of the clones to plasmid form in pBluescript (Stratagene) and purification of the plasmid D N A was as previously described [13].
DNA sequencing Both strands of the pBS128 c D N A were sequenced by the dideoxy chain termination method [ 15 ] using Sequenase (United States Biochemical Corporation, Cleveland, OH) and a modified version of the supplier's protocol. A 3-4 #g portion of dsDNA were denatured in 0.125 M NaOH, 85 °C for 5 min, then neutralized with 1/5 volume 5 M ammonium acetate pH 7.5, and ethanolprecipitated at -20 ° C. Labeling and termination reactions were as described in the supplier's instructions. To prevent artifact banding patterns associated with sequencing dsDNA, terminal deoxynucleotidyl transferase was added after the termination reaction [9]. The sequence of the DNA was read visually and the data analyzed using the University of Wisconsin Genetics Computer Group Sequence Analysis software package [7] with the assistance of the VADMS Computing Center at Washington State University.
ABA analysis ABA levels were measured in whole seeds as well as isolated embryos from dormant and nondormant seeds imbibed in water from 0 to 24 h. ABA was extracted and assayed with a monoclonal antibody for (S)-ABA as previously described [19,21].
Results
Differential screening for dormancy and selection of the pBS128 cDNA clone Dormant and nondormant seeds of B. secalinas were produced under controlled conditions as described in Materials and methods. When placed in water the nondormant seeds germinated com-
436 pletely in one day, as did isolated embryos from those seeds. Hydrated dormant seeds showed no signs of germination for several weeks (Fig. 1). Embryos isolated from the dormant seeds retained most of the dormancy associated with intact seeds and did not germinate for 10 days (Fig. 1) and therefore can be considered to possess embryo dormancy. The differences in dormancy between seeds and isolated embryos (Fig. 1) are probably due to a contributing effect of the endosperm to dormancy of the intact seed. To investigate the pattern of gene expression specific for dormant seeds, poly(A) + RNA was isolated from dormant B. secalinas seed embryos imbibed in water for 10 h at 20 °C, converted to cDNA form and directionally cloned into 2Zap II. This library was differentially screened using 32p-labeled cDNA derived from poly(A) + RNA isolated from imbibed dormant and nondormant B. secalinas embryos. Plaques hybridizing with cDNA derived from dormant embryos, but not with cDNA derived from nondormant embryos were purified, converted to plasmid form in pBluescript, and characterized by northern analysis. Initial northern analysis of the clones identified one clone (pBS 128) with a hybridizing transcript that accumulated in imbibed dormant embryos. This clone was selected for further analysis.
1 O0 ~,-~=--~...... ~ w - = , - , ~ 0
lu d~e
=n
E t_
50
(P
(9 ~e
i,e
0 ~-~
0
10
20
30
Time (days) Fig. 1. Germination of B. secalinus nondormant embryos
( A - A ) , nondormant seeds (O-C), dormant embryos ( i - A ) and dormant seeds ( 0 - 0 ) imbibed in water in the dark at 20 °C.
Temporal and hormonal regulation of the pBS128 transcript Levels of mRNA hybridizing to the pBS128 cDNA probe were compared in dormant and nondormant embryos imbibed in water for 0 to 48 h (Fig. 2, upper row). Similar levels of the pB S 128 transcript were present in embryos from both dormant and nondormant dry seeds (lanes 1 and 2). The size of the mRNA species that hybridizes with the pBS128 cDNA corresponds to a length of approximately 1 kb (Fig. 2, upper rOW).
When dormant seed embryos were placed in water pB S 128 transcript levels increased by 10 h and remained elevated through 48 h (Fig. 2, upper row, lanes 3, 7 and 11). Quantification of the bands with radioanalytic imaging of the intact blot showed that pBS128 transcript increased four-fold over initial levels by 48 h after imbibition (data not shown). These dormant embryos remained growth-arrested for the course of the experiment. Upon imbibition of nondormant embryos, the pBS128 transcript was not detected after 6, 10 or 48 h (Fig. 2, upper, lanes 5, 9 and 13). Visible signs of germination were first apparent after 10 h. To determine if embryonic germination and pB S 128 expression are affected by ABA, embryos from dormant and nondormant B. secalinas seeds were imbibed in 50 #M ABA for 6, 10 and 48 h. Application of ABA blocked germination of the nondormant embryos (data not shown). ABA application to nondormant embryos also maintained low levels of the pBS128 transcript at 6 and 10 h after imbibition (Fig. 2, lanes 6 and 10) and resulted in the accumulation of higher transcript levels at 48 h. Imbibition in ABA for 48 h resulted in enhanced levels of the transcript (lane 14) compared to the water control (lane 13). Imbibition of dormant embryos in 50 #M ABA (lanes 4, 8 and 12) increased the steady-state level of pBS128 transcript present at 6, 10 and 48 h. As a control to ensure that equivalent RNA samples were present in all the lanes shown in Fig. 2 (upper row), the same blot was hybridized with the B. secalinus clone pBSll09 and the re-
437
Fig. 2. Upper row: hormonal and temporal regulation of expression of pBS128 transcript in hydrated dormant and nondormant B. secalinus embryos. Dormant (D) and nondormant (ND) embryos were imbibed in water (-) or 50 # M ABA ( + ) for the indicated times. Total RNA (15 #g/lane) was probed with the pBS128 cDNA. The arrow corresponds to a transcript size of 1 kb. Lower row: as a control to check for equivalent RNA sample loading, the same blot was hybridized with B. secalinus clone pB S 1109, a transcript that is not affected by embryo dormancy level, imbibition or ABA.
sults are shown in the lower panel of Fig. 2. As shown transcript levels of pBS1109 are not affected by embryonic dormancy levels, time of imbibition or ABA application. Since the pB S 128 transcript accumulates in the presence of ABA, differences in pBS128 transcript levels in hydrated dormant and nondormant seed embryos could be due to differences in endogenous ABA levels. Endogenous ABA concentrations were measured in dormant and nondormant embryos at 6, 12 and 24 h after imbibition in water. No significant differences in ABA concentrations were found at any time-point. For example, at 12 h after imbibition the ABA concentration (pg ABA per mg fresh weight was 63 _+ 12 (SD) for dormant seed embryos and 57 + 13 for nondormant seed embryos. Gibberellic acid (GA) is often an antagonist of ABA action and GA can overcome seed dormancy in B. secalinas (P. Goldmark, unpublished data). The effect of this hormone on pBS 128 transcript levels was determined. Dormant embryos were imbibed in water or 25 #M G A 3 and pB S 128 transcript levels were measured at 2, 4 and 6 days after imbibition (Fig. 3). Compared to water controls, steady state levels of the pBS 128 transcript
declined in the GA-treated dormant embryos to low levels by day 4 (lane 5). Visible germination of the GA-treated embryos only occurred at day 6, when pBS 128 transcript levels were very low (lane 7). Thus, GA-induced germination of dormant B. secalinas embryos appears to be preceded by a dramatic drop in steady-state levels of the pBS128 mRNA. Tissue specificity Since many ABA-responsive genes are induced when seedling tissue is dehydrated and ABA levels increase, the level of pBS128 transcript was
Fig. 3. Northern analysis ofpBS128 transcript in B. secalinus dormant embryos imbibed in water or 25 # M G A 3. Embryos were imbibed in water (-) or 25 # M G A 3 ( + ) for the indicated times. Total RNA was extracted from the embryos and probed with the pBS128 cDNA.
438 measured in dehydrated B. secalinus seedlings. pB S 128 transcript was not detected in leaf or root tissue from control or dehydration-stressed seedlings (Fig. 4, lanes 2-3), but was present in mature embryos (lane 1). As a positive control, this same blot was reprobed with two wheat cDNAs for ABA-responsive genes induced in wheat seedlings by dehydration stress [6, 22]. These include cDNA clones, pMA80 [13] encoding a dhn (tab) m R N A [5, 17] and pMA2005 encoding a group 3 lea m R N A [6, 8]. Both the dhn (rab) and group 3 lea transcripts were present in B. secalinus embryos (lane 1). The two transcripts were not detected in non-stressed seedling tissue (lane 2), but high levels of both dhn (tab) and the lea transcripts were induced in the dehydrated seedling tissue (lane 3). Thus, dehydration of B. secalinus tissue results in the induction of two ABA-responsive genes, but not the pBS 128 transcript.
Species specificity Representation of the pB S 128 transcript in other grass species was determined. Northern analysis
Fig. 4. Tissue specificityof pBS 128 transcript in B. secalinus. Northern analysis of total RNA from B. secalinus dry seed embryos (lane 1); 10-day-old seedling leaf and root tissue (lane 2); and dehydrated 10-day-old seedling leaf and root tissue (lane 3). Seedlings were slowly dehydrated for 2 days. Northern analyses were conducted with B. secalinus pB S 128, wheat Group 3 LEA (pMA2005), and wheat dhn (rab) (pMA80) cDNA probes as indicated.
with the pBS128 c D N A was performed using total RNA isolated from dissected embryos of mature dry seeds of B. secalinas, B. tectorum, A. cylindrica, S. cereale, and A.fatua (data not shown). All of the species tested contain transcripts that hybridize to pBS 128.
Sequence analysis of pBS128 cDNA Figure 5 shows the nucleotide sequence and the predicted amino acid sequence for the pBS128 cDNA. The sequenced partial c D N A has a length of 816 bp, which includes a truncated open reading frame (frame 1, nucleotides 1 through 606) encoding a putative polypeptide stretch of 202 amino acids. Downstream of the open reading frame is the probable poly(A) signal sequence, AATATA, plus a poly(A) tail of 15 residues. Another possible truncated open reading frame (frame 3, nucleotides 3-630) would encode a polypeptide stretch of 209 amino acids. In order to determine which of the two potential open reading frames would most likely be translated by B. secalinus, we first compiled a monocot codon usage frequency table (not shown) from the codon usage tables for wheat, barley, maize and rice (available from EMBL Network File Server, Heidelberg, Germany). Those species were chosen because they are all members of the Poaceae family as is B. secalinus. The pBS128 sequence was analyzed using the G C G program 'CodonPreference' [7] with the compiled frequency table. The results indicated that frame 1 is more likely to be translated. Although similarity to monocot codon usage is comparable for both frames in question, the bias toward GC in the third position (99~o for frame 1; 13 ~ for frame 3) strongly supports the translation of frame 1 [4]. Additional support comes from the fact that there are no rare codons used in frame 1, whereas frame 3 requires the use of 18 rare codons. Analysis of the deduced partial protein structure (Fig. 5) indicates a basic polypeptide having alternating hydrophilic and hydrophobic regions with a molecular mass of 22000 Da. This clone appears unique since a search of the databases revealed no significant homologies.
439 -1 c 1
TCC ACC CAT GGC AAG ATC CGC ATC CAC GAC TAC GTC GCC AAC GGC TAC GTC ATC CTC TTC Set Thr His GLy Lys I [ e Arg I r e His Asp Tyr Vat A[a Asn G[y Tyr Vat I [ e Leu Phe
61
TCG CAC CCC GGC GAT TTC ACC CCG GTG TGC ACG ACG GAG CTG GCG GCG ATG GCC AAC TAC Set His Pro G[y Asp Phe Thr Pro Va[ Cys Thr Thr G[u Leu A [ a A[a Net A [ a Asn Tyr
121 GCC /tAG GAG TTC GAG AAG AGG GGC GTC AAG CTG CTG GGC ATC TCC TGC GAC GAC GTG CAG A [ a Lys GLu Phe G[u Lys Arg G[y Va[ Lys Leu Leu G[y l i e Ser Cys Asp Asp Va[ Gin 181 TCC CAC AAG GAG TGG ACC AAG GAC ATC GAG GCC TAC AAG CCT GGG AGC /tAG GTG ACG TAC Ser His Lys Gtu Trp Thr Lys Asp [ [ e G[u A[a Tyr Lys Pro G[y Set Lys Va[ Thr Tyr 2/,1 CCG ATC ATG GCG GAC CCG GAC CGG TCG GCC ATC AAG CAG CTC AAC ATG GTG GAT CCG GAC Pro I [ e Net ALa Asp Pro Asp Arg Set A[a I [ e Lys Gin Leu Asn Net Vat Asp Pro Asp 301 GAG AAG GAC GCG GAG GGG CAG CTG CCG TCG CGC ACG CTG CAC ATC GTG GGG CCG GAC AAG G[u Lys Asp A[a G[u G[y Gin Leu Pro Set Arg Thr Leu His I [ e Vat G[y Pro Asp Lys 361 AAG GTG AAG CTG AGC TTC CTG TAC CCG TCG TGC ACG GGG AGG AAC ATG GAC GAG GTT GTG Lys Vat Lys Leu Set Phe Leu Tyr Pro Set Cys Thr G[y Arg Asn Net Asp G[u Vat Vat 421 CGC GCG GTG GAC TCG CTG CTG ACG GCG GCC AAG CAC /tAG GTG GCC ACC CCG GCC AAC TGG Arg A[a Va[ Asp Set Leu Leu Thr A[a ALa Lys His Lys Va[ A[a Thr Pro A[a Asn TrP 481 AAG CCC GGG GAG TGC GTG GTG ATC GCG CCG GGC GTG TCC GAC GAG GAG GCC AAG AAG TTG Lys Pro G[y G[u Cys Va[ Va[ I [ e A[a Pro G[y Va[ Ser Asp G[u G[u A [ a Lys Lys Leu 541 TTC CCG CAG GGG TTC GAG ACC AAG GAC CTG CCC TCC AAG AAG GGA TAC CTC CGC TTC ACC Phe Pro Gin G[y Phe G[u Thr Lys Asp Leu Pro Set Lys Lys Gty Tyr Leu Arg Phe Thr 601 AAG GTC TAGGCGTGCCCGTCGCCGTGGGCTAGCTCGTCGGCGCTTTAAGAGTTCACCTGTGGTACAAGTATTGTGTC Lys Vat 678 GTCGTTGTGGTACTTTTGGGGTACGTAGTAGCATGGCGTTTGTTTCTGTTGTAATCCAGTACTCGACTGCTGGTGGCGT 757 CGAGCTAATATATGCGTGTACTGTGTGGACTGTTGGCCCCTTTC~A~qAAAAA*t~aJ~ 815
Fig. 5. Nucleotide and deduced amino acid sequences of pB S 128. Dideoxy sequencing data of a nearly full-length (816 bp) cDNA clone from B. secalinas are shown along with the deduced, partial polypeptide. A putative polyadenylation signal, AATATA, is
underlined.
Discussion
We report the identification of a m R N A transcript that is up-regulated in hydrated dormant seeds. Transcript levels increase and are maintained at elevated levels for at least 6 days in hydrated dormant seeds (Figs. 2 and 3), but decline rapidly in germinating seeds. Additionally, treatment of dormant seeds with GA3, which overcomes dormancy in these seeds, also causes a marked decline in pB S 128 transcript levels. The results imply that dormancy is actively maintained after seed imbibition and that pBS 128 transcript expression has a role in maintaining seed dormancy. The pBS128 transcript has been identified in the grass B. secalinas, a plant that produces seeds with embryonic dormancy. The transcript ap-
pears in a wide variety of wild grass species that utilize seed dormancy as a survival mechanism. All these species produce seeds with high levels of dormancy and the conserved expression of this transcript in embryos from all these wild grasses provides evolutionary evidence for the importance of the gene. Application of ABA to nondormant embryos blocks germination and appears to positively regulate pB S 128 transcript levels (Fig. 2). This result may be due to ABA induction of transcript synthesis or stabilization of existing transcript. When dormant embryos are imbibed in ABA, pBS 128 transcript levels increase modestly compared to the water control suggesting that there is some ABA induction of pBS128 transcript in this tissue. The ABA-responsive genes that have been pre-
440
viously characterized in mature seed embryos include genes encoding storage proteins and LEA proteins. Sequence analysis of the pBS 128 cDNA clone revealed a deduced amino acid sequence unlike other ABA-responsive storage or LEA proteins. Most of the ABA-responsive gene products, particularly the LEAs, are very hydrophilic with few hydrophobic regions. The protein products are often basic and have repeating amino acid sequence motifs [5, 6, 8, 17]. In contrast the pBS128 deduced amino acid sequence lacks repeating sequences and contains alternating hydrophobic and hydrophilic regions. Many ABA-responsive genes expressed in embryos are also induced in dehydrated seedling tissue as endogenous ABA levels increase (reviewed in [3, 17]). The pBS128 transcript was not detected in control or dehydrated seedling tissue, even under the water stress conditions that result in the induction of two other ABA-responsive genes dhn (rab) and a group 3 lea (Fig. 4). In embryonic tissue it appears that increases in ABA stimulate pBS 128 transcript levels, but in dehydrated seedling tissue the ABA increases associated with tissue dehydration are not sufficient to induce pBS128 transcript levels. We conclude that pB S 128 transcript expression is restricted by tissue specificity. The fact that pBS128 transcript has been detected only in embryonic tissue suggests a role for the transcript exclusive to embryos. Other embryo-specific genes that are ABA-inducible have been identified such as carrot lea DC-8 [ 11] and maize L3 [18], but the correlation of those transcripts with seed dormancy has not been reported. Whether pBS 128 transcript levels are critical in maintaining seed dormancy in Bromus secalinus and other species or merely correlate with the dormant developmental state awaits experimental determination. References 1. Bewley JD, Black M: Seeds: Physiology of Development and Germination,pp. 190-192. Plenum Press, New York (1985). 2. Bradley JE, Bishop GA, St. John T, Frelinger JA: Asim-
ple rapid method for the purification of Poly A + RNA. Biotechniques 6:114-116 (1988). 3. Bray EA: Regulation of gene expression by endogenous ABA during drought stress. In: WJ Davies, HG Jones (eds), Abscisic Acid, Physiology and Biochemistry, pp. 81-98. Bios, Oxford, UK (1991). 4. Campbell WH, Gowri G: Codon usage in higher plants, green algae, and cyanobacteria. Plant Physiol 92:1-11 (1990). 5. Close TJ, Kortt AA, Chandler PM: A cDNA-based comparison of dehydration-induced proteins (dehydrins) in barley mad corn. Plant Mol Biol 13:95-108 (1989). 6. Curry J, Morris CF, Walker-Simmons MK: Sequence analysis of a cDNA encoding a Group 3 LEA mRNA inducible by ABA or dehydration stress in wheat. Plant Mol Biol 16:1073-1076 (1991). 7. Devereux J, Haeberli P, Smithies O: A comprehensive set of sequence analysis programs for the VAX. Nucl Acids Res 12:387-395 (1984). 8. Dure L III, Crouch M, Harada J, Ho TD, Mundy J, Quatrano R, Thomas T, Sung ZR: Common amino acid sequence domains among the LEA proteins of higher plants. Plant Mol Biol 12:475-486 (1989). 9. Fawcett,TW, Bartlett SG: An effective method for eliminating 'artifact banding' when sequencing doublestranded DNA templates. Biotechniques 9:46-48 (1990). 10. Gubler U, Hoffman BJ: A simple and very efficient method for generating cDNA libraries. Gene 25:263-269 (1983). 11. Hatzopoulos P, Fong F, Sung ZR: Abscisic acid regulation of DC 8, a carrot embryonic gene. Plant Physiol 94: 690-695 (1990). 12. Karssen CM, Brinkhorst-van der Swan D, Breekland AE, Koornneef M: Induction of dormancy during seed development by endogenous abscisic acid: studies on abscisic acid deficient genotypes ofArabidopsis thaliana (L.) Heynh. Planta 157:158-165 (1983). 13. Morris CF, Anderberg RJ, Goldmark PJ, WalkerSimmons MK: Molecular cloning and expression of abscisic acid-responsive genes in embryos of dormant wheat seeds. Plant Physiol 95:814-821 (1991). 14. Sambrook J, Fritsch EF, Maniatis T: Molecular Cloning: A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY (1989). 15. Sanger F, Nicklen S, Coulson AR: DNA sequencing with chain termination inhibitors. Proc Natl Acad Sci USA 74:5463-5467 (1977). 16. Simpson GM: Seed Dormancy in Grasses. Cambridge University Press, Cambridge, UK (1990). 17. Skiver K, Mundy J: Gene expression in response to abscisic acid and osmotic stress. Plant Cell 2:503-512 (1990). 18. Vance VB, Huang AHC: Expression of lipid body protein gene during maize seed development. J Biol Chem 263: 1476-1481 (1988). 19. Walker-Simmons M: ABA levels and sensitivity in de-
441 veloping wheat embryos of sprouting resistant and susceptible cultivars. Plant Physiol 84:61-66 (1987). 20. Walker-Simmons M: Dormancy in cereals - levels of and response to abscisic acid. In: Pharis RP, Rood SB (eds) Plant Growth Substances 1988, pp. 400-406. Springer-Verlag, New York (1990). 21. Walker-Simmons MK, Abrams SR: Uses of ABA Immunoassay. In: Davies WJ, Jones HG (eds) Abscisic Acid, Physiology and Biochemistry, pp. 53-61. Bios, Oxford, UK (1991).
22. Walker-Simmons M, Chandler PM, Sesing J, Ried JL, Morris CF: Comparison of responses to ABA in embryos and dehydrating seedlings of sprouting-resistant and susceptible wheat cultivars. In: Ringlund K, Mosleth E, Mares DJ (eds) Fifth International Symposium on PreHarvest Sprouting in Cereals, pp. 119-129. Westview Press, Boulder, CO (1990).
433
Cloning and expression of an embryo-specific mRNA up-regulated in hydrated dormant seeds Peter J. Goldmark 1' 2,, Jeanne Curry2, Craig F. Morris 2 and M.K. Walker-Simmons 2 I DJR Research, Star Route 69, Okanogan, WA 98840, USA; 2Dept. of Crop and Soil Sciences, Washington State University, U.S. Department of Agriculture/Agricultural Research Service, Wheat Genetics, Quality, Physiology and Disease Research, Johnson Hall 209, Washington State University, Pullman, WA 99164-6420, USA (* author for correspondence) Received 27 November 1991; accepted in revised form 14 February 1992
Key words: abscisic acid, Bromus, dehydration, dormancy, embryos, transcript levels Abstract
Dormant seeds do not germinate when imbibed in water even when conditions are favorable for germination. These hydrated seeds remain viable, but growth-arrested for weeks due to unknown restrictions within the embryo. As a model system for the study of the molecular processes occurring in dormant seeds, we have chosen to examine gene expression in Bromus secalinas, a grass species that produces seeds with high levels of embryonic dormancy. Using differential screening for mRNAs present in hydrated dormant embryos, we have identified a cDNA clone, pBS128, that encodes a mRNA transcript found in the embryos of hydrated seeds of B. secalinus as well as in embryos from mature dry seeds. Striking differences in pBS 128 transcript levels appear upon hydration of dormant and nondormant seeds. Upon imbibition pBS128 transcript levels increase over four-fold in dormant seeds, but rapidly decline and disappear in nondormant seeds, which subsequently germinate. The pBS 128 transcript appears to be embryo-specific since the transcript is not detectable in either non-stressed or dehydrated seedling tissue. Application of 50 #M ABA to nondormant seeds arrests germination and enhances pBS128 transcript levels. The nucleotide sequence of the nearly full-length pBS128 cDNA shows no homology to other reported genes, and the putative protein sequence does not exhibit the hydrophilic characteristics of the ABA-responsive LEA (late embryogenesis abundant) proteins.
Introduction
Seed dormancy is an important developmental state of plant growth whose molecular and bio-
chemical processes remain virtually unknown. Research in this field has focused on the physiological differences between dormant seeds and nondormant seeds. Viable seeds are dormant if
The nucleotide sequence data reported will appear in the EMBL, GenBank and DDBJ Nucleotide Sequence Databases under the accession number X63202. Contribution from the U.S. Department of Agriculture, Agriculture Research Service, and the College of Agriculture and Home Economics Research Center, Washington State University, Scientific Paper No. 9101-53. Mention of a specific product name by the U.S. Department of Agriculture is not an endorsement and does not imply a recommendation over other suitable products.
434 they do not germinate when hydrated, while seeds that germinate promptly are termed nondormant. Studies have shown that placing dormant and nondormant seeds in water results in similar rates of water uptake and new protein synthesis during the first few hours of imbibition [ 1]. Yet, undiscovered biochemical and molecular restrictions prevent cell expansion and germination in dormant seeds. Proposed mechanisms for the maintenance of dormancy in mature seeds have included structural restraints preventing water uptake and gas exchange or endogenous germination inhibitors found in the seed coat and embryo (reviewed in [ 1, 16]). Genetic evidence indicates that the plant hormone abscisic acid (ABA) is required for the induction of dormancy during seed development. Plant mutants deficient in ABA or nonresponsive to ABA produce nondormant seeds [12]. The role of ABA in maintaining dormancy in mature seeds is not yet clear, in part because mature dormant and nondormant seeds have similar ABA levels [ 19]. We are interested in identifying the molecular processes that prevent dormant seeds from germinating when moistened. We have chosen Bromus secalinas as a model system for dormancy because seeds of this wild grass species are uniformly dormant upon maturity and can be converted to nondormant seeds by a short period of postharvest storage at room temperature. Dormant B. secalinas seeds possess true embryonic dormancy meaning that embryos isolated from dormant seeds retain dormancy. The capability of wild grass seeds such as B. secalinas seeds to remain hydrated and dormant in the soil causes major agricultural problems. After extended periods the dormant seeds can sporadically germinate and successfully compete with cultivated crops for light and nutrients. In previous work we determined that wheat seeds in the hydrated dormant state exhibit prolonged expression of ABA-responsive genes [ 13 ]. Embryos isolated from dormant wheat seeds do not retain the dormancy characteristics of intact seeds and germinate within 1 to 2 days after imbibition [ 13]. Dormancy can be restored in iso-
lated wheat embryos by application of ABA. We found that application of low concentrations of ABA to isolated embryos resulted in the enhanced expression of a set of ABA-responsive genes in dormant seed embryos compared to nondormant. For seeds such as wheat with low levels of embryo dormancy we proposed that the enhanced response of dormant grain embryos to ABA could be a dormancy mechanism [ 13, 20]. In this report, we have examined gene expression in hydrated seeds of B. secalinas. Embryos from seeds of this species retain the dormancy characteristics of intact seeds and require a period ofpostharvest storage to dissipate dormancy. We have identified a gene that is up-regulated in expression and whose m R N A accumulates when dormant seeds are imbibed and remain growtharrested. This embryo-specific transcript appears to be regulated by ABA and has a possible role in dormancy maintenance.
Materials and methods Plant materials Bromus secalinas (cheat) seeds were collected from fields in northcentral Washington and propagated under controlled greenhouse conditions with 29 °C day/18 °C night temperatures. Mature seeds were separated from the plant and dried at room temperature to ensure a uniform moisture content of 12~o. A portion of the seeds was frozen at -20 ° C to preserve dormancy. Dormancy in the remaining seeds was dissipated by postharvest storage at room temperature in the dark. After postharvest storage the seeds were fully gerruinable and hydration of seeds resulted in 96100~o germination within 2 days. All germination assays were performed with embryos or whole seeds with the seed coating removed. Embryos were hand-dissected from intact seeds. Embryos or whole seeds were placed in Petri dishes on water-saturated blotters and incubated at 20 °C in the dark. Embryos or seeds were considered germinated when the coleorhiza or radicle was greater than 1 mm in length.
435
B. wctorum (downy brome) and Secale cereale (common rye) seeds were collected from fields in northcentral Washington; Avenafatua (wild oat) and Aegilops cylindrica (jointed goatgrass) were collected from the Spillman Conservation Farm near Pullman, WA. These wild grass species were propagated and stored as described above. The storage time required to dissipate dormancy is species-dependent and can be as short as 4 weeks for B. secalinas and as long as 1 year for A. fatua.
RNA extraction and northern blot analysis Total RNA was purified by pulverization of isolated embryos in liquid nitrogen followed by proteinase K digestion [2] and two phenol/ chloroform extractions. Total RNA was then recovered by precipitation with 2 M lithium chloride. RNA (15#g/lane) was denatured and electrophoresed in 1 ~o (w/v) agarose formaldehyde gels [14]. Gels were rinsed briefly in water and 10 x SSC, and then blotted to Zeta-probe (Bio-Rad, Richmond, CA) using 10 x SSC as the transfer medium. Blots were rinsed for 10 min in 2 x SSC, air-dried, and vacuum-baked for 2 h at 80 ° C. Preparation of 32p-labeled probe, hybridizations and autoradiography were accomplished by standard techniques [14]. Blots were quantified using a radioanalytic imaging system (Ambis, San Diego, CA).
cDNA library construction and screening Total RNA was purified from 800 dormant B. secalinas embryos as described above. Poly(A) + RNA was isolated from total RNA by the oligo(dT) cellulose batch method [13] and directionally cloned into 2Zap II (Stratagene, La Jolla, CA) using a Not I adapter/primer by standard procedures [10]. The primary library contained 2.1 x 106 recombinant phage. 32p-labeled cDNA synthesized from poly(A) + RNA according to the method of Morris et al. [13] was used in differential screening of dormant and nondormant embryos imbibed for 10 h in water at 20 ° C.
Conversion of the clones to plasmid form in pBluescript (Stratagene) and purification of the plasmid D N A was as previously described [13].
DNA sequencing Both strands of the pBS128 c D N A were sequenced by the dideoxy chain termination method [ 15 ] using Sequenase (United States Biochemical Corporation, Cleveland, OH) and a modified version of the supplier's protocol. A 3-4 #g portion of dsDNA were denatured in 0.125 M NaOH, 85 °C for 5 min, then neutralized with 1/5 volume 5 M ammonium acetate pH 7.5, and ethanolprecipitated at -20 ° C. Labeling and termination reactions were as described in the supplier's instructions. To prevent artifact banding patterns associated with sequencing dsDNA, terminal deoxynucleotidyl transferase was added after the termination reaction [9]. The sequence of the DNA was read visually and the data analyzed using the University of Wisconsin Genetics Computer Group Sequence Analysis software package [7] with the assistance of the VADMS Computing Center at Washington State University.
ABA analysis ABA levels were measured in whole seeds as well as isolated embryos from dormant and nondormant seeds imbibed in water from 0 to 24 h. ABA was extracted and assayed with a monoclonal antibody for (S)-ABA as previously described [19,21].
Results
Differential screening for dormancy and selection of the pBS128 cDNA clone Dormant and nondormant seeds of B. secalinas were produced under controlled conditions as described in Materials and methods. When placed in water the nondormant seeds germinated com-
436 pletely in one day, as did isolated embryos from those seeds. Hydrated dormant seeds showed no signs of germination for several weeks (Fig. 1). Embryos isolated from the dormant seeds retained most of the dormancy associated with intact seeds and did not germinate for 10 days (Fig. 1) and therefore can be considered to possess embryo dormancy. The differences in dormancy between seeds and isolated embryos (Fig. 1) are probably due to a contributing effect of the endosperm to dormancy of the intact seed. To investigate the pattern of gene expression specific for dormant seeds, poly(A) + RNA was isolated from dormant B. secalinas seed embryos imbibed in water for 10 h at 20 °C, converted to cDNA form and directionally cloned into 2Zap II. This library was differentially screened using 32p-labeled cDNA derived from poly(A) + RNA isolated from imbibed dormant and nondormant B. secalinas embryos. Plaques hybridizing with cDNA derived from dormant embryos, but not with cDNA derived from nondormant embryos were purified, converted to plasmid form in pBluescript, and characterized by northern analysis. Initial northern analysis of the clones identified one clone (pBS 128) with a hybridizing transcript that accumulated in imbibed dormant embryos. This clone was selected for further analysis.
1 O0 ~,-~=--~...... ~ w - = , - , ~ 0
lu d~e
=n
E t_
50
(P
(9 ~e
i,e
0 ~-~
0
10
20
30
Time (days) Fig. 1. Germination of B. secalinus nondormant embryos
( A - A ) , nondormant seeds (O-C), dormant embryos ( i - A ) and dormant seeds ( 0 - 0 ) imbibed in water in the dark at 20 °C.
Temporal and hormonal regulation of the pBS128 transcript Levels of mRNA hybridizing to the pBS128 cDNA probe were compared in dormant and nondormant embryos imbibed in water for 0 to 48 h (Fig. 2, upper row). Similar levels of the pB S 128 transcript were present in embryos from both dormant and nondormant dry seeds (lanes 1 and 2). The size of the mRNA species that hybridizes with the pBS128 cDNA corresponds to a length of approximately 1 kb (Fig. 2, upper rOW).
When dormant seed embryos were placed in water pB S 128 transcript levels increased by 10 h and remained elevated through 48 h (Fig. 2, upper row, lanes 3, 7 and 11). Quantification of the bands with radioanalytic imaging of the intact blot showed that pBS128 transcript increased four-fold over initial levels by 48 h after imbibition (data not shown). These dormant embryos remained growth-arrested for the course of the experiment. Upon imbibition of nondormant embryos, the pBS128 transcript was not detected after 6, 10 or 48 h (Fig. 2, upper, lanes 5, 9 and 13). Visible signs of germination were first apparent after 10 h. To determine if embryonic germination and pB S 128 expression are affected by ABA, embryos from dormant and nondormant B. secalinas seeds were imbibed in 50 #M ABA for 6, 10 and 48 h. Application of ABA blocked germination of the nondormant embryos (data not shown). ABA application to nondormant embryos also maintained low levels of the pBS128 transcript at 6 and 10 h after imbibition (Fig. 2, lanes 6 and 10) and resulted in the accumulation of higher transcript levels at 48 h. Imbibition in ABA for 48 h resulted in enhanced levels of the transcript (lane 14) compared to the water control (lane 13). Imbibition of dormant embryos in 50 #M ABA (lanes 4, 8 and 12) increased the steady-state level of pBS128 transcript present at 6, 10 and 48 h. As a control to ensure that equivalent RNA samples were present in all the lanes shown in Fig. 2 (upper row), the same blot was hybridized with the B. secalinus clone pBSll09 and the re-
437
Fig. 2. Upper row: hormonal and temporal regulation of expression of pBS128 transcript in hydrated dormant and nondormant B. secalinus embryos. Dormant (D) and nondormant (ND) embryos were imbibed in water (-) or 50 # M ABA ( + ) for the indicated times. Total RNA (15 #g/lane) was probed with the pBS128 cDNA. The arrow corresponds to a transcript size of 1 kb. Lower row: as a control to check for equivalent RNA sample loading, the same blot was hybridized with B. secalinus clone pB S 1109, a transcript that is not affected by embryo dormancy level, imbibition or ABA.
sults are shown in the lower panel of Fig. 2. As shown transcript levels of pBS1109 are not affected by embryonic dormancy levels, time of imbibition or ABA application. Since the pB S 128 transcript accumulates in the presence of ABA, differences in pBS128 transcript levels in hydrated dormant and nondormant seed embryos could be due to differences in endogenous ABA levels. Endogenous ABA concentrations were measured in dormant and nondormant embryos at 6, 12 and 24 h after imbibition in water. No significant differences in ABA concentrations were found at any time-point. For example, at 12 h after imbibition the ABA concentration (pg ABA per mg fresh weight was 63 _+ 12 (SD) for dormant seed embryos and 57 + 13 for nondormant seed embryos. Gibberellic acid (GA) is often an antagonist of ABA action and GA can overcome seed dormancy in B. secalinas (P. Goldmark, unpublished data). The effect of this hormone on pBS 128 transcript levels was determined. Dormant embryos were imbibed in water or 25 #M G A 3 and pB S 128 transcript levels were measured at 2, 4 and 6 days after imbibition (Fig. 3). Compared to water controls, steady state levels of the pBS 128 transcript
declined in the GA-treated dormant embryos to low levels by day 4 (lane 5). Visible germination of the GA-treated embryos only occurred at day 6, when pBS 128 transcript levels were very low (lane 7). Thus, GA-induced germination of dormant B. secalinas embryos appears to be preceded by a dramatic drop in steady-state levels of the pBS128 mRNA. Tissue specificity Since many ABA-responsive genes are induced when seedling tissue is dehydrated and ABA levels increase, the level of pBS128 transcript was
Fig. 3. Northern analysis ofpBS128 transcript in B. secalinus dormant embryos imbibed in water or 25 # M G A 3. Embryos were imbibed in water (-) or 25 # M G A 3 ( + ) for the indicated times. Total RNA was extracted from the embryos and probed with the pBS128 cDNA.
438 measured in dehydrated B. secalinus seedlings. pB S 128 transcript was not detected in leaf or root tissue from control or dehydration-stressed seedlings (Fig. 4, lanes 2-3), but was present in mature embryos (lane 1). As a positive control, this same blot was reprobed with two wheat cDNAs for ABA-responsive genes induced in wheat seedlings by dehydration stress [6, 22]. These include cDNA clones, pMA80 [13] encoding a dhn (tab) m R N A [5, 17] and pMA2005 encoding a group 3 lea m R N A [6, 8]. Both the dhn (rab) and group 3 lea transcripts were present in B. secalinus embryos (lane 1). The two transcripts were not detected in non-stressed seedling tissue (lane 2), but high levels of both dhn (tab) and the lea transcripts were induced in the dehydrated seedling tissue (lane 3). Thus, dehydration of B. secalinus tissue results in the induction of two ABA-responsive genes, but not the pBS 128 transcript.
Species specificity Representation of the pB S 128 transcript in other grass species was determined. Northern analysis
Fig. 4. Tissue specificityof pBS 128 transcript in B. secalinus. Northern analysis of total RNA from B. secalinus dry seed embryos (lane 1); 10-day-old seedling leaf and root tissue (lane 2); and dehydrated 10-day-old seedling leaf and root tissue (lane 3). Seedlings were slowly dehydrated for 2 days. Northern analyses were conducted with B. secalinus pB S 128, wheat Group 3 LEA (pMA2005), and wheat dhn (rab) (pMA80) cDNA probes as indicated.
with the pBS128 c D N A was performed using total RNA isolated from dissected embryos of mature dry seeds of B. secalinas, B. tectorum, A. cylindrica, S. cereale, and A.fatua (data not shown). All of the species tested contain transcripts that hybridize to pBS 128.
Sequence analysis of pBS128 cDNA Figure 5 shows the nucleotide sequence and the predicted amino acid sequence for the pBS128 cDNA. The sequenced partial c D N A has a length of 816 bp, which includes a truncated open reading frame (frame 1, nucleotides 1 through 606) encoding a putative polypeptide stretch of 202 amino acids. Downstream of the open reading frame is the probable poly(A) signal sequence, AATATA, plus a poly(A) tail of 15 residues. Another possible truncated open reading frame (frame 3, nucleotides 3-630) would encode a polypeptide stretch of 209 amino acids. In order to determine which of the two potential open reading frames would most likely be translated by B. secalinus, we first compiled a monocot codon usage frequency table (not shown) from the codon usage tables for wheat, barley, maize and rice (available from EMBL Network File Server, Heidelberg, Germany). Those species were chosen because they are all members of the Poaceae family as is B. secalinus. The pBS128 sequence was analyzed using the G C G program 'CodonPreference' [7] with the compiled frequency table. The results indicated that frame 1 is more likely to be translated. Although similarity to monocot codon usage is comparable for both frames in question, the bias toward GC in the third position (99~o for frame 1; 13 ~ for frame 3) strongly supports the translation of frame 1 [4]. Additional support comes from the fact that there are no rare codons used in frame 1, whereas frame 3 requires the use of 18 rare codons. Analysis of the deduced partial protein structure (Fig. 5) indicates a basic polypeptide having alternating hydrophilic and hydrophobic regions with a molecular mass of 22000 Da. This clone appears unique since a search of the databases revealed no significant homologies.
439 -1 c 1
TCC ACC CAT GGC AAG ATC CGC ATC CAC GAC TAC GTC GCC AAC GGC TAC GTC ATC CTC TTC Set Thr His GLy Lys I [ e Arg I r e His Asp Tyr Vat A[a Asn G[y Tyr Vat I [ e Leu Phe
61
TCG CAC CCC GGC GAT TTC ACC CCG GTG TGC ACG ACG GAG CTG GCG GCG ATG GCC AAC TAC Set His Pro G[y Asp Phe Thr Pro Va[ Cys Thr Thr G[u Leu A [ a A[a Net A [ a Asn Tyr
121 GCC /tAG GAG TTC GAG AAG AGG GGC GTC AAG CTG CTG GGC ATC TCC TGC GAC GAC GTG CAG A [ a Lys GLu Phe G[u Lys Arg G[y Va[ Lys Leu Leu G[y l i e Ser Cys Asp Asp Va[ Gin 181 TCC CAC AAG GAG TGG ACC AAG GAC ATC GAG GCC TAC AAG CCT GGG AGC /tAG GTG ACG TAC Ser His Lys Gtu Trp Thr Lys Asp [ [ e G[u A[a Tyr Lys Pro G[y Set Lys Va[ Thr Tyr 2/,1 CCG ATC ATG GCG GAC CCG GAC CGG TCG GCC ATC AAG CAG CTC AAC ATG GTG GAT CCG GAC Pro I [ e Net ALa Asp Pro Asp Arg Set A[a I [ e Lys Gin Leu Asn Net Vat Asp Pro Asp 301 GAG AAG GAC GCG GAG GGG CAG CTG CCG TCG CGC ACG CTG CAC ATC GTG GGG CCG GAC AAG G[u Lys Asp A[a G[u G[y Gin Leu Pro Set Arg Thr Leu His I [ e Vat G[y Pro Asp Lys 361 AAG GTG AAG CTG AGC TTC CTG TAC CCG TCG TGC ACG GGG AGG AAC ATG GAC GAG GTT GTG Lys Vat Lys Leu Set Phe Leu Tyr Pro Set Cys Thr G[y Arg Asn Net Asp G[u Vat Vat 421 CGC GCG GTG GAC TCG CTG CTG ACG GCG GCC AAG CAC /tAG GTG GCC ACC CCG GCC AAC TGG Arg A[a Va[ Asp Set Leu Leu Thr A[a ALa Lys His Lys Va[ A[a Thr Pro A[a Asn TrP 481 AAG CCC GGG GAG TGC GTG GTG ATC GCG CCG GGC GTG TCC GAC GAG GAG GCC AAG AAG TTG Lys Pro G[y G[u Cys Va[ Va[ I [ e A[a Pro G[y Va[ Ser Asp G[u G[u A [ a Lys Lys Leu 541 TTC CCG CAG GGG TTC GAG ACC AAG GAC CTG CCC TCC AAG AAG GGA TAC CTC CGC TTC ACC Phe Pro Gin G[y Phe G[u Thr Lys Asp Leu Pro Set Lys Lys Gty Tyr Leu Arg Phe Thr 601 AAG GTC TAGGCGTGCCCGTCGCCGTGGGCTAGCTCGTCGGCGCTTTAAGAGTTCACCTGTGGTACAAGTATTGTGTC Lys Vat 678 GTCGTTGTGGTACTTTTGGGGTACGTAGTAGCATGGCGTTTGTTTCTGTTGTAATCCAGTACTCGACTGCTGGTGGCGT 757 CGAGCTAATATATGCGTGTACTGTGTGGACTGTTGGCCCCTTTC~A~qAAAAA*t~aJ~ 815
Fig. 5. Nucleotide and deduced amino acid sequences of pB S 128. Dideoxy sequencing data of a nearly full-length (816 bp) cDNA clone from B. secalinas are shown along with the deduced, partial polypeptide. A putative polyadenylation signal, AATATA, is
underlined.
Discussion
We report the identification of a m R N A transcript that is up-regulated in hydrated dormant seeds. Transcript levels increase and are maintained at elevated levels for at least 6 days in hydrated dormant seeds (Figs. 2 and 3), but decline rapidly in germinating seeds. Additionally, treatment of dormant seeds with GA3, which overcomes dormancy in these seeds, also causes a marked decline in pB S 128 transcript levels. The results imply that dormancy is actively maintained after seed imbibition and that pBS 128 transcript expression has a role in maintaining seed dormancy. The pBS128 transcript has been identified in the grass B. secalinas, a plant that produces seeds with embryonic dormancy. The transcript ap-
pears in a wide variety of wild grass species that utilize seed dormancy as a survival mechanism. All these species produce seeds with high levels of dormancy and the conserved expression of this transcript in embryos from all these wild grasses provides evolutionary evidence for the importance of the gene. Application of ABA to nondormant embryos blocks germination and appears to positively regulate pB S 128 transcript levels (Fig. 2). This result may be due to ABA induction of transcript synthesis or stabilization of existing transcript. When dormant embryos are imbibed in ABA, pBS 128 transcript levels increase modestly compared to the water control suggesting that there is some ABA induction of pBS128 transcript in this tissue. The ABA-responsive genes that have been pre-
440
viously characterized in mature seed embryos include genes encoding storage proteins and LEA proteins. Sequence analysis of the pBS 128 cDNA clone revealed a deduced amino acid sequence unlike other ABA-responsive storage or LEA proteins. Most of the ABA-responsive gene products, particularly the LEAs, are very hydrophilic with few hydrophobic regions. The protein products are often basic and have repeating amino acid sequence motifs [5, 6, 8, 17]. In contrast the pBS128 deduced amino acid sequence lacks repeating sequences and contains alternating hydrophobic and hydrophilic regions. Many ABA-responsive genes expressed in embryos are also induced in dehydrated seedling tissue as endogenous ABA levels increase (reviewed in [3, 17]). The pBS128 transcript was not detected in control or dehydrated seedling tissue, even under the water stress conditions that result in the induction of two other ABA-responsive genes dhn (rab) and a group 3 lea (Fig. 4). In embryonic tissue it appears that increases in ABA stimulate pBS 128 transcript levels, but in dehydrated seedling tissue the ABA increases associated with tissue dehydration are not sufficient to induce pBS128 transcript levels. We conclude that pB S 128 transcript expression is restricted by tissue specificity. The fact that pBS128 transcript has been detected only in embryonic tissue suggests a role for the transcript exclusive to embryos. Other embryo-specific genes that are ABA-inducible have been identified such as carrot lea DC-8 [ 11] and maize L3 [18], but the correlation of those transcripts with seed dormancy has not been reported. Whether pBS 128 transcript levels are critical in maintaining seed dormancy in Bromus secalinus and other species or merely correlate with the dormant developmental state awaits experimental determination. References 1. Bewley JD, Black M: Seeds: Physiology of Development and Germination,pp. 190-192. Plenum Press, New York (1985). 2. Bradley JE, Bishop GA, St. John T, Frelinger JA: Asim-
ple rapid method for the purification of Poly A + RNA. Biotechniques 6:114-116 (1988). 3. Bray EA: Regulation of gene expression by endogenous ABA during drought stress. In: WJ Davies, HG Jones (eds), Abscisic Acid, Physiology and Biochemistry, pp. 81-98. Bios, Oxford, UK (1991). 4. Campbell WH, Gowri G: Codon usage in higher plants, green algae, and cyanobacteria. Plant Physiol 92:1-11 (1990). 5. Close TJ, Kortt AA, Chandler PM: A cDNA-based comparison of dehydration-induced proteins (dehydrins) in barley mad corn. Plant Mol Biol 13:95-108 (1989). 6. Curry J, Morris CF, Walker-Simmons MK: Sequence analysis of a cDNA encoding a Group 3 LEA mRNA inducible by ABA or dehydration stress in wheat. Plant Mol Biol 16:1073-1076 (1991). 7. Devereux J, Haeberli P, Smithies O: A comprehensive set of sequence analysis programs for the VAX. Nucl Acids Res 12:387-395 (1984). 8. Dure L III, Crouch M, Harada J, Ho TD, Mundy J, Quatrano R, Thomas T, Sung ZR: Common amino acid sequence domains among the LEA proteins of higher plants. Plant Mol Biol 12:475-486 (1989). 9. Fawcett,TW, Bartlett SG: An effective method for eliminating 'artifact banding' when sequencing doublestranded DNA templates. Biotechniques 9:46-48 (1990). 10. Gubler U, Hoffman BJ: A simple and very efficient method for generating cDNA libraries. Gene 25:263-269 (1983). 11. Hatzopoulos P, Fong F, Sung ZR: Abscisic acid regulation of DC 8, a carrot embryonic gene. Plant Physiol 94: 690-695 (1990). 12. Karssen CM, Brinkhorst-van der Swan D, Breekland AE, Koornneef M: Induction of dormancy during seed development by endogenous abscisic acid: studies on abscisic acid deficient genotypes ofArabidopsis thaliana (L.) Heynh. Planta 157:158-165 (1983). 13. Morris CF, Anderberg RJ, Goldmark PJ, WalkerSimmons MK: Molecular cloning and expression of abscisic acid-responsive genes in embryos of dormant wheat seeds. Plant Physiol 95:814-821 (1991). 14. Sambrook J, Fritsch EF, Maniatis T: Molecular Cloning: A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY (1989). 15. Sanger F, Nicklen S, Coulson AR: DNA sequencing with chain termination inhibitors. Proc Natl Acad Sci USA 74:5463-5467 (1977). 16. Simpson GM: Seed Dormancy in Grasses. Cambridge University Press, Cambridge, UK (1990). 17. Skiver K, Mundy J: Gene expression in response to abscisic acid and osmotic stress. Plant Cell 2:503-512 (1990). 18. Vance VB, Huang AHC: Expression of lipid body protein gene during maize seed development. J Biol Chem 263: 1476-1481 (1988). 19. Walker-Simmons M: ABA levels and sensitivity in de-
441 veloping wheat embryos of sprouting resistant and susceptible cultivars. Plant Physiol 84:61-66 (1987). 20. Walker-Simmons M: Dormancy in cereals - levels of and response to abscisic acid. In: Pharis RP, Rood SB (eds) Plant Growth Substances 1988, pp. 400-406. Springer-Verlag, New York (1990). 21. Walker-Simmons MK, Abrams SR: Uses of ABA Immunoassay. In: Davies WJ, Jones HG (eds) Abscisic Acid, Physiology and Biochemistry, pp. 53-61. Bios, Oxford, UK (1991).
22. Walker-Simmons M, Chandler PM, Sesing J, Ried JL, Morris CF: Comparison of responses to ABA in embryos and dehydrating seedlings of sprouting-resistant and susceptible wheat cultivars. In: Ringlund K, Mosleth E, Mares DJ (eds) Fifth International Symposium on PreHarvest Sprouting in Cereals, pp. 119-129. Westview Press, Boulder, CO (1990).
