The Plant Cell, Vol. 9, 703-715, May 1997 O 1997 American Society of Plant Physiologists

Downregulation of Ovule-Specific MADS Box Genes from Petunia Results in Maternally Controlled Defects in Seed Development

Lucia Colomboyall John Franken,a Alexander R. Van der Krol,b Peter E. Wit t ichyc Hans J. M. Donsya and Gerco C. Angenenta12

aDepartment of Developmental Biology, DLO-Centre for Plant Breeding and Reproduction Research (CPRO-DLO), Droevendaalsesteeg 1, P.O. Box 16, 6700 AA Wageningen, The Netherlands

The Netherlands CDepartment of Plant Cytology and Morphology, Wageningen Agricultural University, Arboretumlaan 4, 6703 BD Wageningen, The Netherlands

Department of Plant Physiology, Wageningen Agricultural University, Arboretumlaan 4, 6703 BD Wageningen,

A maternally determined seed defect has been obtained by downregulation of the petunia MADS box genes Floral Bind- ing Protein7 (FBP7) and FBP77, These genes have been previously shown to play central roles in the determination of ovule identity. Aberrant development of the seed coat and consequent degeneration of the endosperm have been ob- served in transgenic plants in which these two genes are downregulated by cosuppression. Analysis of the expression pattern of FBP 7 and FBP 7 7 and genetic analysis confirmed the matemal inheritance of the phenotype. The FBP 7 pro- moter was cloned and fused to reporter genes. One of these reporter genes was the BARNASE gene for targeted cell ablation. Our results indicate that FBP 7 promoter activity is restricted to the seed coat of developing seeds and that it is completely silent in the gametophytically derived tissues. The mutants used in this study provided a unique opportu- nity to investigate one of the poorly understood aspects of seed development: the interaction of embryo, endosperm, and maternal tissues.

.

INTRODUCTION

During floral organogenesis, five different types of organ pri- mordia emerge from the floral meristem and differentiate into the floral organs that are organized in concentric whorls. From the outer to the inner whorls, these organs are sepals, petals, stamens, carpels, and, in the center of the flower, the placenta bearing the ovules. The placenta develops as a distinct meristematic region inside the petunia carpel (Angenent et al., 1995; Colombo et al., 1995). The ovule con- tains a megasporangium surrounded and protected by one or two integuments. The megasporangium retains a single megaspore, which produces the female gametophyte (Herr, 1995). After double fertilization, the embryo develops from the zygote, and fusion between the two polar nuclei and a sperm cell nucleus leads to formation of the endosperm. During seed development, the integument(s) undergoes mor- phological changes and becomes the seed coat.

Many studies have focused on the molecular control of ei- ther ovule development or seed formation; however, it has

Current address: Dipartimento di Genetica e Biologia dei Micror- ganismi, via Celoria 26, Milan, Italy. 'To whom correspondsnce should be addressed. E-mail g.c.angenent Ocpro.dlo.nl; fax 31-31 7-418094.

been difficult to determine a relationship between the two de- velopmental programs. A number of Arabidopsis mutants dis- playing aberrant ovule development have been described. The short integuments (sinl) mutant has short integuments that fail to cover the nucellus completely (Robinson-Beers et al., 1992; reviewed in Gasser and Robinson-Beers, 1993). Su- perman (sup) ovules lack asymmetric growth of the outer in- tegument, resulting in elongated ovules of reduced fertility and suggesting a role for SUP in cellular proliferation (Gaiser et al., 1995; Sakai et al., 1995). In bell, the best-characterized ovule mutant, the outer integument is often replaced by a carpel-like structure (Robinson-Beers et al., 1992; Modrusan et al., 1994; Ray et al., 1994; Reiser et al., 1995); in the inner-no-outer (ino) mutant, the outer integument initiates but fails to develop further (Gaiser et al., 1995). Recently, the aintegumenta (ant) mutant has been described. For ant mutants, the development of the integuments is not initiated, and megasporogenesis is blocked at the tetrad stage (Elliott et al., 1996; Klucher et al., 1996).

On the other hand, severa1 Arabidopsis seed coat mutants have been described. transparent testa (tt) lacks pigmenta- tion of the seed coat because of mutations of genes involved in the flavonoid biosynthetic pathway (Koornneef, 1990); transparent testa glabra (ttg), glabrous2 (g12), and aberrant

704 The Plant Cell

fesfa shape (afs) are three other mutants in which seed coatdevelopment is affected (Koornneef, 1981; Leon-Kloosterzielet al., 1994; Meyerowitz and Somerville, 1994). In addition, inthe homeotic apeta/a2 (ap2) mutant, the seed coat is alsoaffected without changes in endosperm and embryo devel-opment (Jofuku et al., 1994). Thus, to date, evidence thatendosperm and embryo development in Arabidopsis de-pends on normal seed coat development is lacking.

One of the best-described class of plant regulatory genesis the MADS box gene family. The MADS box transcriptionfactors play essential roles in determining meristem and flo-ral organ identity, and some of them are also expressed dur-ing ovule and seed development. Recently, two petuniaMADS box genes, Floral Binding Protein? (FBP7) and FBP11,have been identified, and their role in ovule development hasbeen investigated (Angenent et al., 1995; Colombo et al.,1995). Overexpression of these genes resulted in ectopicovule formation on the two outer floral whorl organs, indicat-ing their involvement in ovule initiation and ovule identityspecification. This observation has led to the extension of theABC model (reviewed in Coen and Meyerowitz, 1991) with aD function. The D function is required to specify the identity ofthe fifth floral organ type, the ovule (Colombo et al., 1995).Several other MADS box genes are expressed during ovuleand embryo development. For example, the Arabidopsis MADSbox genes AGAMOUS-like2 (AGL2], AGL11, and AGL13 areexpressed in ovules and, at a reduced level, during seed for-mation (Flanagan and Ma, 1994; Rounsley et al., 1995), andAGL15 is specifically expressed in the embryo (Heck et al.,1995); however, their functions have not been determined yet.

In conclusion, ovule and seed development have oftenbeen studied separately, and little is known about the devel-opment of the maternally derived tissues of the seed and thegenes that control this process. In this report, we analyzethe expression of the petunia MADS box genes FBP7 andFBP11 after fertilization and their role in seed development.The activity of the FBP7 promoter was monitored by usingthe p-glucuronidase (GUS), luciferase (LUC), and BARNASEreporter gene fusions driven by the FBP7 promoter. To ana-lyze the role of FBP7 and FBP11 at later stages of ovule on-togeny, we compared seed development in wild-typepetunia plants with that in transformants in which bothMADS box genes were downregulated by cosuppression.This study provides new insights into the developmental in-teractions of the embryo, endosperm, and seed coat in a di-cotyledonous plant species.

RESULTS

Expression of FBP7 and FBP11 Genes duringFlower Development

It has been previously reported that the petunia MADS boxgenes FBP7 and FBP11 are specifically expressed in ovules

(Angenent et al., 1995). Transcripts of these genes are firstdetectable in the central meristem of the petunia flower bud,which is the progenitor tissue of the placenta and ovules.Later, these transcripts are restricted to the ovules, predom-inantly in the endothelium, which is the innermost cell layerof the integument (Angenent et al., 1995). To study the ex-pression of these MADS box genes in developing seeds,RNA gel blot analyses were performed using RNA isolatedfrom ovaries at various stages after pollination. Figure 1Ashows an RNA gel blot hybridized with probes specific forFBP7, FBP11, orpMADSS (Tsuchimoto et al., 1993). The ex-pression levels of both FBP7 and FBP11 increased immedi-ately after pollination (2 days after pollination [DAP]) anddeclined in mature seeds. In the same experiment, we fol-lowed the expression of pMADS3 during the same stages.pMADS3 is the putative petunia class C homeotic gene ho-mologous to/AG (Yanofsky et al., 1990) and PLENA (Bradleyet al., 1993), which is also expressed in ovules (Tsuchimotoet al., 1993; M. Kater, L Colombo, and G.C. Angenent, manu-script in preparation). In contrast to FBP7 and FBP11,pMADS3 expression decreased after pollination (Figure 1A).

The accumulation of FBP7 and FBP11 proteins in the de-veloping seeds was also monitored by protein gel blot anal-ysis (Figure 1B). Polyclonal antibodies were raised againstthe C-terminal part of FBP11, but they also recognizedFBP7 because of the high levels of similarity between these

B 0 1 2 3 4 7 21

Figure 1. Comparison of FBP7, FBP11, and pMADS3 Gene Expres-sion in Ovaries before and after Pollination.(A) RNA gel blot analysis of total RNA isolated from mature ovariesof the wild type (lane 0) and from ovaries 1, 2, 3, 4, 7, and 21 DAP.Each lane contains 10 ju.g of total RNA. Blots were probed with 32P-labeled gene-specific fragments for FBP7, FBP11, and pMADS3cDNAs, respectively.(B) Immunoblot analysis of FBP7 and FBP11 during seed develop-ment. Nuclear proteins were isolated from mature ovaries of the wildtype (lane 0) and from ovaries 1,2,3, 4, 7, and 21 DAP.

MADS Box Genes Involved in Seed Development 705

Figure 2. Expression of the FBP7 Gene during Ovule Development before and after Fertilization.

Longitudinal sections were hybridized with the digoxigenin-labeled antisense FBP7 transcript.(A) Ovary containing ovules at stage 12 (Angenent et al., 1995), in which the integument covers the nucellus completely.(B) Mature unfertilized ovules.(C) Fertilized ovules 3 DAP.(D) Fertilized ovules 5 DAP.cw, carpel wall; o, ovule; pi, placenta; sc, seed coat. Bars in (A) and (B) = 1.0 mm. The magnifications are the same for (A) and (C) and for (B)and (D).

two polypeptide sequences. This cross-reactivity was dem-onstrated by protein gel blots of extracts from sepals be-longing to transgenic plants overexpressing either the FBP7or the FBP11 gene (data not shown). Greater amounts ofFBP7 and FBP11 proteins were detectable in nuclear ex-tracts from ovules after pollination compared with extractsfrom ovules before the fertilization events. The smaller pro-tein band in the protein gel blot shown in Figure 1B is alwaysaccompanied by the larger FBP7/FBP11 band and probablycorresponds to a degradation product.

To determine the distribution of FBP7 mRNA in ovules be-fore and after fertilization, we performed in situ hybridiza-tions, using a 3' terminal antisense transcript of FBP7 asprobe. Figures 2A and 2B show the expression of FBP7 inovules at two stages of development. This expression pat-tern is very similar to that of FBP11, as has been reportedpreviously (Angenent et al., 1995). In ovules 3 and 5 DAP(Figures 2C and 2D), strong hybridization signals were ob-served in the tissues that form the seed coat, whereas very

V////////,pfbp? GUS

/!• //"s // *

B

Figure 3. Constructs Used to Monitor Reporter Gene Expressionunder the Control of the FBP7 Promoter.

GUS, LUC, and BARNASE were used as reporter genes.(A) Chimeric FBP7-GUS gene construct.(B) Chimeric FBP7-LUC construct.(C) Chimeric FBP7-BARNASE-BARSTAR construct.RB, right border sequence; tnos, nopaline synthase terminator; 35SpolyA, cauliflower mosaic virus terminator sequence; LB, left border.

706 The Plant Cell

Figure 4. Activity of the FBP7 Promoter Using the Reporter Genes GUS, LUC, and BARNASE.

MADS Box Genes lnvolved in Seed Development 707

low or no expression was found in other ovaty tissues. Simi- lar expression patterns were observed by using an FBP7 7- specific antisense probe (data not shown). These expression data clearly show that the petunia MADS box genes FBP7 and FBP77 are specifically expressed in ovules and that their expression persists and even increases significantly during seed development. This suggests that FBP7 and FBP7 7 are not only required for ovule development but may also be involved in postfertilization processes.

FBP7 Promoter Activity during Ovule Ontogeny

To study the regulation of FBP7 expression during seed de- velopment in more detail, we have isolated a 0.6-kb FBP7 promoter fragment by inverse polymerase chain reaction. Two chimeric gene constructs were made and consisted of the FBP 7 promoter fragment fused to either the GUS or LUC reporter gene. These chimeric constructs, schematically shown in Figure 3, were introduced into petunia line W115 by using Agrobacterium-mediated transformation (Horsch et al., 1985), and the resulting plants were tested fluorometri- cally for GUS activity and luminometrically for LUC activity. All GUS-positive plants (four of 27 transformants, namely, T51002, T51009, T51013, and T51027) showed a high leve1 of GUS activity in the ovules. GUS activity in ovules of T51013 before fertilization is shown in Figure 4A, with the activity increasing 4 DAP (compare Figures 4A and 4B). GUS activity was mainly present in the developing seed coat, excluding the outer cell layer, and in the epidermis of the placenta (Figure 4C). Although GUS is a suitable reporter gene to determine tissue specificity, LUC is a far more pre- cise reporter protein to study the temporal and relative ac- tivity of the FBP7 promoter.

To monitor LUC expression driven by the FBP7 promoter, we selected one (T59042) of 28 transformants in which pro- moter activity was high enough for a clear luminescence sig-

na1 after 30 min of measurement. To avoid accumulation effects, flower buds were treated daily with substrate before measurement, enabling the detection of de novo transcrip- tion activity. The expression at various stages of ovary de- velopment after pollination is shown in Figure 4D. Photon production due to the activity of the LUC gene driven by the FBP7 promoter was detectable in mature ovules but showed an increase after pollination, reaching a maximum at -7 DAP. Activity declined to almost zero at -14 DAP (Figure 4D). The increase in activity a few days after pollina- tion was specific for the FBP7 promoter, because it was not seen in transgenic control plants in which LUC was driven by the cauliflower mosaic virus 35s promoter (data not shown). However, the decrease in LUC activity -14 DAP is a general phenomenon due to the initial desiccation of the seed and was also seen in the control plants. Taken to- gether, our data show that FBP7 promoter activity coincides exactly with the pattern of FBP7 expression during ovule and seed development, as was determined by RNA blot analysis and in situ hybridization.

BARNASE Gene Expression Controlled by the FBP7 Promoter

The activity of the FBP7 promoter was also monitored by fusing it to the bacterial BARNASE gene (Hartley, 1988), as shown in Figure 3. The effect of expression of this cytotoxic gene was analyzed in 26 independently generated primaty transformants, namely, T52001 to T52026. The severity of the defects varied considerably among transformants. Two major classes of transformants were identified. In the first class of transformants (eight plants), the flower buds did not reach maturity and the pistil was completely missing (com- pare Figures 4E and 4F). The second class of transformants (five plants) had normal flowers with reduced pistils in which no ovules developed (compare Figures 4G and 4H). The

Figure 4. (continued).

(A) Cross-section through a mature ovary of the transgenic plant T51013 containing the FBP 7 promoter-GUS construct. Blue staining indicates GUS activity. (B) GUS activity in a cross-section through a T51 O1 3 ovary 7 DAP. (C) GUS activity in a longitudinal section through a T51013 ovary 7 DAP. The section was viewed by dark-field microscopy, with the signal color being red. (D) LUC activity driven by the FBP7 promoter in mature T59042 ovaries (O DAP) and ovaries 4, 7, 9, and 14 DAP. The carpel walls were removed before spraying. The color bar, from gray to red, indicates an increasing intensity of LUC activity. (E) Wild-type flower bud of line W115. Sepals, petals, and anthers are partly removed. (F) Detailed view of a small bud of transgenic plant T52012 expressing the BARNASE gene. The carpel is completely missing, and the develop- ment of flower organs is blocked at a young stage. The arrow indicates ablated cells (brown) in the receptacle at a position where normally the pistil develops. (G) Wild-type ovary with removed ovary wall. (H) Ovary of a T52006 flower. No ovules were observed after removing the ovary wall. cw, carpel wall; o, ovules; p, placenta; sc, seed coat.

708 The Plant Cell

Figure 5. Microscopic Analysis of Seed Development from Wild-Type and T27017 Plants.

MADS Box Genes lnvolved in Seed Development 709

other 13 plants had an intermediate phenotype, and they were not analyzed further.

One plant (T52006) belonging to the second class of transformants was used to carefully monitor FBP7 promoter activity in seed tissue of gametophytic origin. Wild-type plants were pollinated with pollen from the T52006 transfor- mant, and the developing seeds were examined from 1 DAP until they were mature. No aberrations were found during the development of the seeds, and all seeds examined ger- minated normally. However, approximately half of the prog- eny plants exhibited the same aberrant phenotype, as did the primary transformant T52006, because of BARNASE ac- tivity. Transformant T52006 contains three 1-DNA insertions that segregate as one locus. These BARNASE transgenes were present only in the progeny plants with defective pis- tils, as was confirmed by DNA gel blot analysis (data not shown). The most likely explanation for the transmission of the highly effective killer gene BARNASE to the progeny is that the activity of the fBP7 promoter in the developing seed is restricted to the seed coat and that it is not active in the gametophytically derived tissues, the embryo and en- dosperm. Therefore, a putative role for fBP7 during seed development might be restricted to the formation of the seed coat.

FBP7 and FBPll Are Required for Proper Seed Development

During the double fertilization process, one male gamete fer- tilizes the egg cell, giving rise to the zygote, whereas the second male gamete fuses with the diploid central cell, forming the triploid primary endosperm nucleus. The devel- opment of a mature embryo in petunia line W115 requires -4 weeks, with temperature and day length influencing the actual time required for embryo maturation (Sink, 1984). The seed coat, which is derived from the single integument of the petunia ovule, is thought to contribute to seed formation

by distributing essential nutrients to the developing en- dosperm and embryo (Murray, 1984, 1987).

To investigate the function of FBP7 and FBP77 during seed formation, we compared the development of wild-type petunia seeds and seeds derived from transgenic plants in which FBP7 and FBP7 7 were simultaneously downregulated by cosuppression (Angenent et al., 1995). Three transgenic plants (T27017, T27035, and T27040) were obtained show- ing the same cosuppression phenotype as was described previously (Angenent et al., 1995). One of these transgenic plants, T27017, was used for further investigation. In this transgenic plant, the expression of both fBP7 and fBP77 genes was reduced >lO-fold, and a similar reduction was observed after fertilization (data not shown). A few ovules were transformed into carpelloid structures, whereas the majority of the ovules were morphologically normal and de- veloped into seeds upon fertilization.

In contrast to the round-shaped seeds from a wild-type plant (Figure 5G), the seeds obtained by self-pollination of the transgenic plants (T27017, T27035, and T27040) were shrunken, as shown in Figure 5H. To investigate the inheri- tance of seed phenotype, we performed severa1 crosses. All seeds from the self-pollination of the T27017 plant and the seeds derived after pollination with wild-type pollen exhib- ited the shrunken phenotype. In contrast, seeds from back- crosses using pollen of transgenic origin to fertilize wild-type ovules were all phenotypically normal. The cross between wild-type and T27017 plants, in both directions, gave rise to progeny in which 19 of 35 plants for the cross W115 x T27017 and 16 of 30 plants for the cross T27017 x W115 had wild- type phenotypes. These results indicate that the shrunken seed phenotype is not lethal and is of maternal origin.

Light microscopic analysis was performed to investigate the interna1 anatomy of mature seeds from both the wild type and mutants. During development of wild-type and T27017 seed, the cellularization of the endosperm was com- parable in both types until 9 DAP (Figures 5A and 5B). How- ever, between 13 and 21 DAP, T27017 seeds showed many

'

Figure 5. (continued).

(A) Section of a wild-type seed 9 DAP; cellular endosperm with large vacuoles. (B) Section of a T27017 seed 9 DAP; endosperm similar to that of the wild type. (C) Section of a wild-type seed 19 DAP; multicellular endosperm with much cytoplasm. (D) Section of T27017 seed 19 DAP; cellularized, vacuolated endosperm. (E) Section of a mature wild-type seed; well-developed embryo and endosperm. (F) Section of a mature T27017 seed; intact embryo and degenerated endosperm. (G) Scanning electron mircroscopy of a mature wild-type seed. (H) Scanning electron microscopy of a mature T27017 seed. (I) Dissected embryo from a mature wild-type seed. (J) Dissected embryos from three mature T27017 seeds; stages ranged from globular stage to fully differentiated. e, embryo; en, endosperm; sc, seed coat. Bars in (A) and (B) = 200 pm; bars in (C) to (H) = 100 pm. The magnification for (I) and (J) is the same as for (G) and (H).

710 The Plant Cell

Figure 6. Seed Coat and Endosperm Development of the Wild Type and the T27017 Mutant of Petunia.

MADS Box Genes lnvolved in Seed Development 71 1

different forms of endosperm degeneration within one ovary. In the less severe forms, endosperm degeneration occurred at later stages of development and became apparent near maturity (compare Figures 5C and 5D). The mature seeds of the wild type showed intact endosperm and a fully devel- oped embryo (Figures 5E and 51), whereas in mature T27017 seeds, the endosperm was partly or totally degenerated (Figure 5F), and dissected embryos varied from globular and heart shaped to fully developed (Figure 5J).

To study the aberrant seed phenotype in more detail, we analyzed sections of seeds at different developmental stages. This microscopic analysis is shown in Figure 6. The petunia integument, which differentiates into the seed coat, consists of an outer epidermis, an inner epidermis called the endothelium, and approximately six layers of parenchyma cells (Figure 6A). In wild-type seed development (Figures 6A to 6E), the outer epidermis cells enlarged during develop- ment, stored large quantities of a substance that stained with toluidine blue (asterisk in Figure 6B), and formed a thick tangential cell wall against the parenchyma cells. The endo- thelium cells were papillar shaped at 5 DAP (Figure 6A) and were flattened after 9 DAP (Figure 6B). Subsequently, they also stored a substance that stained with toluidine blue, as shown in Figures 6C and 6D. After 18 DAP, the parenchyma cells and endothelium degenerated, resulting in one layer of parenchyma cells between the outer epidermis and the en- dosperm. The endosperm and embryos at 21 DAP were rich in cytoplasm and storage products (Figure 6E).

The development of the T27017 seed coat (Figures 6F to 6J) was similar to the wild type at 5 and 9 DAP (compare Figures 6A with 6F, and 6B with 6G). At 13 DAP, differences in the wild-type seed coat were observed (compare Figures 6C and 6H): in some cases, a part of the endothelium of

T27017 was degenerated (indicated with arrows), whereas adjacent endothelium cells were intact (Figure 6H). The en- dosperm cells bordering the degenerating endothelium also exhibited degeneration of cytoplasm (Figure 6H), whereas the endosperm cells bordering the intact endothelium showed less intense staining and were indistinguishable from cells found in wild-type endosperm. The endothelium cells were completely degenerated after 18 DAP (Figures 61 and 6J). The endosperm cells at 21 DAP were highly vacu- olated and partly or completely degenerated, as shown in Figure 6J. The embtyo developed more slowly in the mutant (Figure 6J) than in wild-type seeds (Figure 6E), and embryos arrested in their development were observed. Nevertheless, many T27017 seeds develop a viable embryo despite the aberration of the seed coat and endosperm.

DlSCUSSlON

Seed development is an important process during the life cycle of a plant. Here, we describe the role of two petunia MADS box genes, FBP7 and FBP77, in seed formation. Both FBP7 and FBP7 7 are essential for the determination of petunia ovule identity and development (Angenent et al., 1995; Colombo et al., 1995). By analyzing seed develop- ment in plants in which FBP7 and FBP 7 7 are simultaneously cosuppressed (Angenent et al., 1995), we have shown that the expression of both genes is also required for normal seed development. In this report, we demonstrate the role of a MADS box transcription factor in the development of the seed coat. In addition, new information is provided about the interrelationships of the three types of tissue that compose

Figure 6. (continued).

Wild-type seeds are shown in (A) to (E), and T27017 seeds are shown in (F) to (J). (A) Wild-type seed 5 DAP. Endothelium cells have a papillar shape. The endosperm consists of four cells. (6) Wild-type seed 9 DAP. Epidermal cells have storage products densely stained with toluidine blue (asterisk) and a thickening cell wall (arrow). Parenchyma cells are highly vacuolated. (C) Wild-type seed 13 DAP. Storage products of flattened endothelium cells are densely stained with toluidine blue. (D) Wild-type seed 18 DAP. The parenchyma cell layer bordering the endothelium is compressed (arrow). (E) Wild-type seed 21 DAP. The intact outer epidermis, compressed parenchyma cells, and the endothelium are shown. The endosperm con- tains large amounts of starch and surrounds a well-developed, cytoplasm-rich embryo. (F) T27017 seed 5 DAP. Endothelium cells have a papillar shape. The endosperm consists of four cells. (O) T27017 seed 9 DAP. Structurally, there is no difference between the T27017 seed and the wild-type seed. (H) T27017 seed 13 DAP. lntact endothelium cells adjacent to compressed, degenerating endothelium cells (arrows). Endosperm cells bordering the crushed endothelium cells start to degenerate. (I) T27017 seed 18 DAP. Three inner parenchyma cell layers are compressed, whereas the outer parenchyma cells are still intact. The cellular en- dosperm in this seed contains storage products. (J) T27017 seed 21 DAP. Crushed inner parenchyma cells and a degenerated endothelium are shown. The cellular endosperm in this seed does not contain storage products and is partly degenerated (arrowheads). The embryo is not as well developed when compared with the wild-type embryo. e, embryo; en, endosperm; ep, epidermis of seed coat; et, endothelium; pc, parenchyma cells. Bars in (A) and (F) = 50 km; bars in (E) and (J) =

40 km.

71 2 The Plant Cell

the dicot seed, namely, the seed coat, the endosperm, and the embryo.

Expression Pattern of FBP7 and fBP77 in Developing Seeds

ln wild-type petunia plants 2 DAP, when all of the ovules have been fertilized, the expression of both FBP7 and FBPl 1 is increased. High expression levels are maintained throughout the maturation of the seeds and decrease at later stages when the seeds start to desiccate. The expres- sion of pMADS3, another MADS box gene expressed in ovules, decreases after pollination, indicating that the induc- tion of FBP7 and FBP7 1 expression is specific for these two genes. In situ hybridization experiments using FBP7 and FBPll antisense RNAs as probes confirmed RNA gel blot data and showed that their expression is very high in seed tissue derived from the ovule integument.

The putative Arabidopsis ortholog of FBPl 1 and FBP 7, AGL l l , has been isolated and characterized (Rounsley et al., 1995). The expression pattern of this gene is very similar to the those of FBP7 1 and FBP7. AGLl7 starts to be ex- pressed in the placental tissue before the ovules arise; later in development, its expression is localized in the ovules and placental tissue. In ovules, its expression is mainly restricted to the two integuments; after fertilization, AGLl7 is pre- dominantly expressed in the chalazal region of the ovule (Rounsley et al., 1995). However, in contrast to FBP7 and FBP 11, whose expression was greatly increased after polli- nation, A G L l l expression is not detectable by RNA gel blot analysis after fertilization. This suggests either that AGL 7 1 is not the true ortholog of FBPll and FBP7 or that the function of this gene is restricted to ovule development, and it is not required for Arabidopsis seed formation.

FBP7 1s Expressed Exclusively in Maternal Tissue of the Seed

We isolated the FBP 7 promoter to investigate FBP7 expres- sion in detail by using reporter and cell ablation genes. Each reporter gene that was used provided us with specific infor- mation about FBP7 gene expression in the developing seed. Because no active LUC accumulates when the plants are regularly sprayed with luciferin (Millar et al., 1992; Wood, 1995), LUC activity reflects de novo reporter gene expres- sion. On the other hand, the GUS reporter gene was used as a tool to study tissue specificity in the seed, and high activity was observed in a major part of the seed coat.

To monitor traces of FBP7 promoter activity in embryo and endosperm tissues, we performed cell ablation, using the highly sensitive bacterial RNase BARNASE. Transgenic plants specifically expressing BARNASE in the ovules were used to study the activity of the FBP 7 promoter during seed develop- ment. Backcrosses with wild-type plants, using pollen carry-

ing the FBP7 promoter-BARNASE construct, revealed that the BARNASE phenotype can be transferred through the pol- len to the progeny. The most likely explanation for this obser- vation is that the promoter is active in maternally originated tissue of the developing seed and not in seed tissues derived from the double-fertilization process. However, it cannot be excluded that the transgene, when contributed by the male, may be silenced in endosperm and embryo.

Role of fBP7 and FBP77 in Seed Development

The induction of FBP 7 and FBPl 1 expression after fertiliza- tion supports the idea that these genes are involved in seed development. To study this proposed function, we followed the formation of seeds in the wild type and in transgenic plants in which the expression of both FBP 7 and FBPll was dra- matically reduced by cosuppression. Microscopic analysis has shown that in mature seeds from these transgenic plants, the development of the endosperm is negatively af- fected, resulting in a shrunken seed phenotype. Genetic analysis showed that the aberrant seed phenotype is mater- nally inherited. Analyses of early and intermediate develop- mental stages by light microscopy revealed that up to 9 DAP, the development of T27017 seeds is similar to wild-type seed development. However, between 9 and 13 DAP, the in- ner epidermis of the seed coat, the endothelium, starts to degenerate. At 18 DAP, the endothelium is completely de- generated, and this directly disturbs the development of the endosperm, probably because of a block in transport of nutri- ents through the seed coat to the developing endosperm and embryo. This leads to the formation of mature seeds with or without a small amount of endosperm. Genetic analysis showed that the aberrant seed phenotype is maternally inherited, indicating that the defects in endosperm develop- ment are indirect. The phenotype of the cosuppression mu- tant suggests that FBP7 and FBP 17 MADS box genes play a role in the development of the seed coat.

The interaction of endosperm, embryo, and maternal tis- sues remains one of the most complex and unresolved as- pects of seed development. It has been suggested that maternal and endosperm tissues may regulate each other’s development (reviewed in Lopes and Larkins, 1993). Cell proliferation in the endosperm and protein and starch bio- synthesis largely depend on the nutrient flow from the ma- ternal tissues. Defects in the part of the seed coat that facilitates this transport of nutrient, which is believed to be the endothelium, have negative effects on the development of the endosperm, which ultimately degenerates. In con- trast, severa1 Arabidopsis mutants have been described in which embryo and endosperm develop normally despite ab- errations in seed coat development (Jofuku et al., 1994; Léon-Kloosterziel et al., 1994). Apparently, the defects in seed coat development occur late during seed development or they do not affect nutrient flow.

MADS Box Genes lnvolved in Seed Development 713

In this report, we demonstrate the interaction between the seed coat and endosperm. Despite the degeneration of the endosperm, normal embryo development was observed, suggesting that at later stages of petunia seed development, the endosperm is not required for proper ernbryo formation. That this may be a general phenomenon for all dicot seeds is supported by the fact that in contrast to monocotyledon- ous seeds, mature dicot seeds contain only a small amount of endosperm.

METHODS

placed in a mold, and 1 mL of hardener II (Kulzer Histo-tec) was added per 15 mL of Technovit, resulting in a polymerization reaction. The Technovit blocks were sectioned by microtome. Sections were stained with a solution of 1 % toluidine blue in distilled water.

For cryoscanning, electron microscopy samples were mounted on a stub, frozen in liquid nitrogen, coated, and observed as described in Angenent et al. (1995). Ovules 7, 14, and 21 days after pollination (DAP) were histochemically stained for p-glucuronidase (GUS) activity, ac- cording to Koes et al. (1 990), before microscopic sections were made.

Bacterial Expression of FBP7 and FBPII, Immunization, and lmmunoblot Analysis

Plant Material

The Petunia hybrida variety W115 and the transgenic plants were grown under normal greenhouse conditions.

A 3’ fBP7 7 terminal cDNA fragment (0.65 kb) was cloned in the bacte- ria1 expression vector pETlId (Rosemberg et al., 1987) by using Ncol (5‘) and BamHl (3’) sites. 60th sites were generated by polymerase chain reaction (PCR), using a 5’ primer (ISOGEN; Bioscience BV, Maarssen, The Netherlands) (5’-CGCCATGGCAAATAGGCATCTG- GTGGGTGAAGG-3’; 316 bp downstream of the ATG codon) and 3’

DNA and RNA Gel Blot Analyses

Plant DNA was isolated from petunia leaves, according to Koes et al. (1990), and total RNA was isolated from ovaries, according to Verwoerd et al. (1989). Ten micrograms of DNA was digested with EcoRl or Hindlll, electrophoresed, and blotted onto Hybond N+ membranes (Amersham). For RNA gel blot analysis, 10 pg of total RNA was denatured by glyoxal (1.5 M) before electrophoresis. Equal loading of RNA in the gel slots was verified by ethidium bromide staining of the gel. Floral Binding Protein7 (fBP7) (0.6 kb), FBP77 (0.79 kb), and pMADS3 (0.6 kb) 3’ terminal fragments and the fBP7 (0.38 kb) 5‘ terminal fragment were labeled by random oligonucle- otide priming (Feinberg and Vogelstein, 1984). Blots were hybridized, as described by Angenent et al. (1992).

In Situ RNA Hybridizations

Floral buds were fixed and embedded in paraffin, and 10-pm sec- tions were prepared as described by Caiias et al. (1994). Digoxige- nin-labeled RNA probes were synthesized by in vitro transcription using the pSPT18/19 vectors (Boehringer Mannheim). For the syn- thesis of antisense RNA, we introduced 3’ terminal cDNA fragments of fBP7 (0.6 kb) into pSPT18 or pSPT19 by using the interna1 restric- tion site Hindlll. Transcripts were partially hydrolyzed by incubation at 60°C in 0.1 M Na,CO,.NaHCO, buffer, pH 10.2, for 45 min. Hybrid- ization and immunological detection were performed as described by Caias et al. (1994).

Microscopy

The material was fixed in 5% gluteraldehyde in 0.1 M sodium phos- phate buffer, pH 7.2. The material was rinsed three times for 10 min in 0.1 M sodium phosphate buffer, pH 7.2, and then dehydrated us- ing a series of alcohol solutions (30,50, 70,90, and 100%) for 30 min each. The material was embedded in hydroxyethyl methacrylate Technovit 71 00, according to the manufacturer’s protocol (Kulzer Histo-tec, Wehrheim, Germany). After embedding, the material was

primer (5’-GAGCAGATCTCTTTCTCCAGCGCCC-3’; 40 bp down- stream of the stop codon), respectively. The underlined nucleotides correspond to the cDNA. The N-terminal 106 amino acid residues were removed to avoid cross-reactions with other MADS box pro- teins. This construct was introduced into 8121, a strain of Escherichia coli containing an integrated copy of the isopropyl p-o-thiogalactopy- ranoside (IPTG) gene (Rosemberg et al., 1987). The cells were grown, hawested, and sonicated, according to Angenent et al. (1992). After centrifugation, the pellet was resuspended in loading buffer, according to Laemmli (1 970). A New Zealand rabbit was immunized with 1 O0 pg of FBPll protein purified from an SDS-polyacrylamide gel. Booster in- jections with the same amount of protein were given twice with an in- tervening period of 28 days, and the serum was collected after an additional 10 days. Mature ovules and ovules 1,2,3,4, 7, and 21 DAP were used for nuclear protein extraction, according to the method de- scribed by Angenent et al. (1 992). Protein gel was made by using 1 O0 pg of nuclear proteins for each sample. The protein detection was done according to Angenent et al. (1992).

lsolation of the FBP7 Promoter Fragment by lnverse PCR

Three misograms of genomic DNA (line W115) was digested with Hindlll in a final volume of 400 pL. The restriction reaction was per- formed for 1 O hr and followed by a phenol-chloroform extraction pro- cedure. After DNA precipitation, a ligation reaction was performed using 10 units of T4 DNA ligase in 500 pL final volume at 16°C for 16 hr. After precipitation, ~ 0 . 5 pg of the DNA was used in the PCR. In the first PCR (1 min at 94T, 1 min at 55T, and 2 min at 72°C for 30 cy- cles), a primer (5’-GTGCCmGGGAGAGGMGATAGAG-3’) con- taining the ATG codon (underlined) and a primer containing the ATG codon (underlined) but with the opposite orientation (5’-GAAGA- TCTTTCCWGGCACAATATAATATTCCTGATTAGC-3‘) were used. In the second PCR (same conditions as were given for the first one), the first primer was substituted by a nested primer (5’-CGGGATC- CCGTTCTGCAAAAGGAGAAATGG-3’; 58 bp downstream of the ATG). An amplified product of 620 bp was obtained. The promoter and fragment was cloned into pEMBL19 and sequenced. The Gen- Bank and EMBL accession number of this promoter fragment is U90137.

714 The Plant Cell

Construction of Binary Vectors and Plant Transformation

The fbp7 promoter was cloned into the binary vector pBinl9. TheGUS gene, the LUC/FERASE (LUC) gene, or the BARNASE-BARSTAR bac- teria1 operon construct (Hartley, 1988)was cloned downstream of the promoter. The recombinant pBinl9 vectors were transferred via Agro- bacterium tumefaciens (LBA4404) to petunia variety W115, using the standard leaf disc transformation method (Horsch et al., 1985). Re- generation of transformants was done as described by Van Tunen et al. (1989).

LUC Activity Measurements

Flowers of plants carrying the FBP 7-LUC construct were sprayed twice a day with a solution of luciferin (1 mM), starting 2 days before measurement. At harvest, the sepals, petals, and stamens as well as the carpel walls were carefully removed to expose the developing seeds. lmaging of photon production was done for 30 min with a two-dimensional luminometer (Hamamatzu, Hamamatzu, Japan), us- ing the Argus 50 software (Hamamatzu).

ACKNOWLEDGMENTS

We thank Drs. David Hannapel, Andrew MacCabe, and Michiel Van Lookeren-Campagne for critical reading of the manuscript, and Dr. Ruud de Maagd for his assistance with immunoblot analysis. We are grateful to Dr. Piotr Dullin for the isolation of the FBPl1 protein. Marco Busscher is acknowledged for technical assistance and Adriaan Van Aelst and Dr. Andre Van Lammeren for their help with the elec- tron micrographs. This study was supported by Human Capital and Mobility Grant No. ERBCHBICT930450 and the Dutch Organization for Scientific Reseach (NWO) to L.C.

Received January 3,1997; accepted March 6, 1997.

REFERENCES

Angenent, G.C., Busscher, M., Franken, J., MOI, J.N.M., and Van Tunen, A.J. (1992). Differential expression of two MADS box genes in wild-type and mutant petunia flowers. Plant Cell4, 983-993.

Angenent, G.C., Franken, J., Busscher, M., Van Dijken, A., Van Went, J.L., Dons, H.J.M., and Van Tunen, A.J. (1995). A nove1 class of MADS box genes is involved in ovule development in petunia. Plant Cell 7, 1569-1 582.

Bradley, D., Carpenter, R., Sommer, H., Hartley, N., and Coen E.S. (1993). Complementary floral homeotic phenotypes result from opposite orientations of a transposon at the plena locus of Antirrhinum. Cell 72, 85-95.

Caiias, L.A., Busscher, M., Angenent, G.C., Beltran, J.P., and Van Tunen, A.J. (1994). Nuclear localization of the petunia MADS box protein FBPl. Plant J. 6, 597-604.

Coen, E.S., and Meyerowitz, E.M. (1991). The war of the whorls: Genetic interactions controlling flower development. Nature 353, 31-37.

Colombo, L., Franken, J., Koetje, E., Van Went, J., Dons, H.J.M., Angenent, G.C., and Van Tunen, A.J. (1995). The petunia MADS box gene FBP7 7 determines ovule identity. Plant Cell7, 1859-1 868.

Elliot, R.C., Betzner, A.S., Huttner, E., Oakes, M.P., Tucker, W.Q.J., Gerentes, D., Perez, P., and Smyth, D.R. (1996). AIN- TEGUMENTA, an APETALAP-like gene of Arabidopsis with pleio- tropic roles in ovule development and floral organ growth. Plant Cell8, 155-168.

Feinberg, A.P., and Vogelstein, B. (1 984). A technique for radiola- beling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 137, 266-267.

Flanagan, C.A., and Ma, H. (1994). Spatially and temporally regu- lated expression of the MADS-box gene AGLP in wild-type and mutant Arabidopsis flowers. Plant MOI. Biol. 26, 581-595.

Gaiser, J.C., Robinson-Beers, K., and Gasser, C.S. (1995). The Arabidopsis SUPERMAN gene mediates asymmetric growth of the outer integument of ovules. Plant Cell7,333-345.

Gasser, C.S., and Robinson-Beers, K. (1 993). Pistil development. Plant Cell5, 1231-1239.

Hartley, R.W. (1 988). Barnase and Barstar: Expression of its cloned inhibitor permits expression of a cloned ribonuclease. J. MOI. Biol.

Heck, G.R., Perry, S.E., Nichols, K.W., and Fernandez, D.E. (1995). AGL15, a MADS domain protein expressed in developing embryos. Plant Cell 7, 1271-1282.

202,913-91 5.

Herr, J.M. (1995). The origin of the ovule. Am. J. Bot. 82,547-564.

Horsch, R.B., Fry, J.E., Hoffman, N.L., Eichholz, D., Rogers, S.G., and Fraley, R.T. (1 985). A simple and general method for trans- ferring genes into plants. Science 227, 1229-1 231.

Jofuku, K.D., den Boer, B.G.W., Van Montagu, M., and Okamuro, J.K. (1 994). Control of Arabidopsis flower and seed development by the homeotic geneAPETALA2. Plant Cell6, 1211-1225.

Klucher, K.M., Chow, H., Reiser, L., and Fischer, R.L. (1996). The AlNTEGUMENTA gene of Arabidopsis required for ovule and female gametophyte development is related to the floral homeotic gene APETALAP. Plant Cell8, 137-1 53.

Koes, R.E., Van Blokland, R., Quattrocchio, F., Van Tunen, A.J., and MOI, J.N.M. (1 990). Chalcone synthase promoters in petunia are active in pigmented and unpigmented cell types. Plant Cell 2,

Koomneef, M. (1981). The complex syndrome of ttg mutants. Arabi- dopsis Inf. Serv. 18,45-51.

Koornneef, M. (1990). Mutations affecting the testa color in Arabi- dopsis. Arabidopsis Inf. Sem. 27, 1-4.

Laemmli, U.K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685.

Léon-Kloosterziel, K.M., Keijzer, C.J., and Koornneef, M. (1 994). A seed shape mutant of Arabidopsis that is affected in integument development. Plant Cell6, 385392,

Lopes, M.A., and Larkins, B.A. (1 993). Endosperm origin, develop- ment, and function. Plant Cell5, 1383-1399.

Meyerowitz, E.M., and Somerville, C., eds (1994). Genetics of Ara- bidopsis. (Cold Spring Harbor, NY: Cold Spring Harbor Labora- tory Press).

379-392.

MADS Box Genes lnvolved in Seed Development 715

Millar, A.J., Short, S.R., Hiratsuka, K., Chua, N.-H., and Kay, S.A. (1992). Firefly luciferase as a reporter Óf regulated gene expres- sion in higher plants. Plant MOI. Biol. Rep. 10, 324-337.

Modrusan, Z., Reiser, L., Feldmann, K.A., Fischer, R.L., and Haughn, G.W. (1994). Homeotic transformation of ovules into car- pel-like structures in Arabidopsis. Plant Cell6,333-349.

Murray, D.R., ed (1984). Seed Physiology, Vol. 1. (London: Aca- demic Press).

Murray, D.R. (1987). Nutritive role of the seed coat in developing legume seeds. Am. J. Bot. 74, 1122-1 137.

Ray, A., Robinson-Beers, K., Ray, S., Baker, S.C., Lang, J.D., Preuss, D., Milligan, S.B., and Gasser, C.S. (1 994). Arabidopsis floral homeotic gene BELL (BEL7) controls ovule development through negative regulation of AGAMOUS gene (AG). Proc. Natl. Acad. Sci. USA91,5761-5765.

Reiser, L., Modrusan, Z., Margossian, L., Samach, A., Ohad, N., Haughn, G.W., and Fischer, R.L. (1995). The BELL7 gene encodes a homeodomain protein involved in pattern formation in the Arabidopsis ovule primordium. Cell83, 735-742.

Robinson-Beers, K., Pruitt, R.E., and Gasser, C.S. (1992). Ovule development in wild-type Arabidopsis and two female-sterile mutants. Plant Cell4, 1237-1 249.

Rosemberg, A.H., Lade, B.N., Chul, D.S., Lin, S.W., and Studier, F.W. (1987). Vectors for selective expression of cloned DNAs by T7 RNA polymerase. Gene 56, 125-1 35.

Rounsley, S.D., Ditta, G.S., and Yanofsky, M.F. (1995). Diverse roles for MADS box genes in Arabidopsis development. Plant Cell

Sakai, H., Medrano, L.J., and Meyerowitz, E.M. (1995). Role of SUPERMAN in maintaining Arabidopsis floral whorl boundaries. Nature 378,199-203.

Sink, K.C., ed (1984). Monographs on Theoretic and Applied Genet- ics. 9. Petunia. (Berlin: Springer-Verlag).

Tsuchimoto, S., Van der Krol, A.R., and Chua, N.-H., (1993). Ectopic expression of pMADS3 in transgenic petunia pheno- copies the petunia blind mutant. Plant Cell 5, 843-853.

Van Tunen, A.J., Mur, L.A., Brouns, G.S., Rienstra, J.D., Koes, R.E., and MOI, J.N.M. (1989). Pollen- and anther-specific chi pro- moters from petunia: Tandem prornoter regulation of the chiA gene. Plant Cell2, 393-401.

Verwoerd, T.C., Dekker, B.M.M., and Hoekema, A. (1989). A small-scale procedure for the rapid isolation of plant RNAs. Nucleic Acids Res. 17, 2362.

Wood, K.V. (1995). Marker proteins for gene expression. Curr. Opin.

Yanofsky, M.F., Ma, H., Bowman, J.L., Drews, G.N., Feldmann, K.A., and Meyerowitz, E.M. (1990). The protein encoded by the Arabidopsis homeotic gene Agamous resembles transcriptional factors. Nature 346, 35-39.

7, 1259-1 269.

Biotechnol. 6, 50-58.

DOI 10.1105/tpc.9.5.703 1997;9;703-715Plant Cell

L Colombo, J Franken, A R Van der Krol, P E Wittich, H J Dons and G C Angenentdefects in seed development.

Downregulation of ovule-specific MADS box genes from petunia results in maternally controlled

 This information is current as of May 13, 2020

 

Permissions X

https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists