Halperin and Wetherell, 1964; Borkird et al., 1988; Keller et al., 1988; Wurtele et al., 1988; Franz et al., 1989; Apuya and. Zimmerman, 1992; Wurtele and Nikolau ...
Plant Physiol. (1993) 102: 303-312
Characterization of a Gene That 1s Expressed Early in Somatic Embryogenesis of Daucus carota’ Eve Syrkin Wurtele*, Huiqing Wang, SaDly Durgerian’, Basil J.Nikolau, and Thomas H. Ulrich Department of Botany (E.S.W., H.W., S.D.), Department of Food Science and Human Nutrition (E.S.W., S.D.), and Department of Biochemistry and Biophysics (B.J.N.), lowa State University, Ames, lowa 5001 1; and Sogetal, Inc., Hayward, California 94545 (T.H.U.)
an embryo from a single cell are central to all higher life forms, little is known about the molecular events occurring during the early stages of embryo development of plants. In contrast to zygotic embryos, very young somatic embryos can be isolated in large quantities and used to characterize molecular events. In carrot (Daucus carota), cell cultures can be used to provide large amounts of somatic embryos at discrete developmental stages (e.g. Steward et al., 1958; Halperin and Wetherell, 1964; Borkird et al., 1988; Keller et al., 1988; Wurtele et al., 1988; Franz et al., 1989; Apuya and Zimmerman, 1992; Wurtele and Nikolau, 1992). Furthermore, the environment of somatic embryos can be readily altered to manipulate experimentally the factors controlling the development of embryos and the expression of particular genes during this process. In contrast to zygotic embryos, the process of desiccation does not occur in somatic embryos, thus enabling the dissociation of events that require desiccation from those that do not. In addition, comparisons between zygotic and somatic embryo development may help to elucidate the communications (Murray, 1984) that exist between the developing zygotic embryo and the maternal plant. The EMB-1 cDNA clone (Ulrich et al., 1990) represents an mRNA that is highly expressed specifically in developing embryos of carrot. Genes homologous to the emb-l gene may be expressed in all monocot and dicot embryos (Cuming, 1984; Litts et al., 1987; Baker et al., 1988; Raynal et al., 1989; Ulrich et al., 1990; Williams and Tsang, 1991; Espelund et al., 1992). Genes from cotton (Leu D19 gene) and from wheat (Em gene) that code for proteins similar to that encoded by EMB-1 are expressed primarily at the time of embryo desiccation or in response to ABA (Dure et al., 1989; Marcotte et al., 1989; Hughs and Galau, 1991). To begin to address the function and regulation of the EMB-1 protein, we isolated and characterized the emb-1 gene and investigated the spatial and temporal patterns of accumulation of EMB-1 mRNA during embryogenesis in culture and in planta.
l h e EMB-1 mRNA of carrot (Daucus carota) was isolated as an embryo abundant cDNA clone (T.H. Ulrich, E.S. Wurtele, B.J. Nikolau [1990] Nucleic Acids Res 18: 2826). Northern analyses of RNA isolated from embryos, cultured cells, and a variety of vegetative organs indicate that the EMB-1 mRNA specifically accumulates i n embryos, beginning at the early stages of embryo development. In situ hybridization with both zygotic and somatic embryos show that the EMB-1 mRNA begins to accumulate at low levels throughout globular embryos. Accumulation of EMB-1 mRNA increases and becomes more localized as embryos mature; in torpedo embryos, EMB-1 mRNA preferentially accumulates in the meristematic regions, particularly the procambium. The similarity in distribution of EMB-1 mRNA in both zygotic and somatic embryos indicates that much of the spatial pattern of expression of the emb-1 gene i s dependent on the developmental program of the carrot embryo and does not require maternal or endosperm factors. The EMB-1 protein (relative molecular weight 9910) i s a very hydrophilic protein that i s a member of a class of highly conserved proteins (typified also by the Em protein of wheat and the Lea D19 protein of cotton) that may be ubiquitous among angiosperm embryos but whose functions are as yet unknown. l h e carrot genome appears to contain one or two copies of the emb-1 gene. A 1313-base pair DNA fragment of the carrot genome containing the emb-1 gene was isolated and sequenced. l h e gene i s interrupted by a single intron of 99 base pairs. Primer extension experiments identify two EMB-1 mRNAs, differing by 6 bases at their 5’ ends that are transcribed from this gene.
The zygotic embryo of a dicotyledonous plant develops from the unicellular fertilized egg through a series of morphologically identifiable stages, beginning with globular, through heart, torpedo, and mature embryos (Raghavan, 1986). This developmental process is thought to be controlled in part by the programmed expression of specific genes in precise cell types at defined developmental times. Master regulatory genes can be considered the apex of an activating cascade of genes that results in the biochemical and morphological changes that occur during embryogenesis. Although the processes regulating the organization and formation of
MATERIALS AND METHODS Chemicals and Buffers
’ This work was supportedin part by Iowa State University, Ames,
A11 electrophoresis reagents were obtained from Bio-Rad. Magnagraph matrix and nitrocellulose (0.45-c~mpore size) were from MSI. Type 7 oligo(dT)-cellulose was from Pharmacia. Lyophilized and dialyzed Fico11 and all other biochemicals were from Sigma. [32P]dCTP(110 TBq/mmol) was from
IA. This is Journal Paper No. J.-14553 of the Iowa Agriculture and Home Economics Experiment Station, Ames, IA, Project No. 2997. Present address:Department of Zoology and Genetics, Iowa State University, Ames, I A 50011. * Corresponding author; fax 1-515-294-1337.
303
304
Wurtele et al.
NEN; 5-35S-UTP(37 TBq/mmol) and [5-3H]polyuridate(>740 MBq/mmol) were from Amersham. All nucleic acid-modifying enzymes were from BRL. Plant Crowth
Carrot (Daucus carota) plants were grown in soil in the greenhouse from seed in 12-inch pots until harvest. Seeds of wild D. carota collected in fields near Grinnell, IA, were a kind gift of Dr. Diane Robertson, Grinnell College. Roots and leaves of 3-week-old carrot plants were harvested, as were carrot flowers and seeds at various stages of development. Maintenance and Fractionation of Cell Cultures
Carrot cell lines were initiated and maintained in liquid medium containing 2,4-D according to methods described previously (Keller et al., 1988; Wurtele et al., 1988). Tissue culture lines of the domestic variety Danver (line DC2) and a wild genotype (line WC49-1) were independently established. Embryogenic cell clusters and nonembryogenic cells were fractionated by size and density centrifugation from log-phase cultures grown in a Murashige-Skoog medium containing 2,4-D [(+)2,4-D cultures] (Wurtele et al., 1988). Briefly, embryogenic cell clusters were obtained by passing the cells of the (+)2,4-D cultures through a 190-1m meshsize screen and collecting the cells from the filtrate by centrifugation. The pelleted cells were then resuspended at a density of 1 g fresh weight 20 mL-' of a 10% (w/w) Ficoll solution and centrifuged. The resulting pellet was resuspended in 15% (w/w) Ficoll at a density of 1 g fresh weight 20 mL-' and centrifuged; the cells floating on the 15% (w/w) Ficoll were collected. This fraction consisted predominantly of small embryogenic cell clusters of three to 20 cells. Nonembryogenic cells were obtained by centrifuging unfiltered (+)2,4-D cultures, resuspending the pellet in 7.5% (w/w) Ficoll at a density of 1 g fresh weight 20 mL-', and centrifuging. The cell fraction that floated was collected; this fraction consisted almost exclusively of large, vacuolated, single nonembryogenic cells. The embryogenic capacities of the fractions of embryogenic cell clusters and of the nonembryogenic cells were routinely monitored by transferring aliquots of these fractions to a medium lacking 2,4-D. The fractions of embryogenic cell clusters have a 700- to 3000-fold higher embryogenic capacity than the fractions of nonembryogenic cells (Wurtele et al., 1988). The fractions of embryogenic cell clusters and nonembryogenic cells were used as a source of RNA for the northern analysis depicted in this study. Fractions of embryos and nonembryogenic cells were obtained from cultures growing in medium lacking auxin according to procedures similar to those previously described (Keller et al., 1988). Briefly, fractions of embryos at globular plus heart, torpedo, and germinating stages of development were collected by passing the cultures consecutively through a series of screens of 1920-, 520-, and 190-pm mesh sizes, respectively. (The term "germinating" refers to the elongation of the embryonic axis in somatic embryos [Raghavan, 19861; the physiological similarities between germination in zygotic
Plant Physiol. Vol. 102, 1993
embryos and somatic embryos have not been well defined). The material passing through the 190-pm mesh screen was centrifuged in a solution containing 15% (w/w) Ficoll; the pellet was then centrifuged through a 10% (w/w) Ficoll solution. The nonembryogenic cell fraction floating on the 10% (w/w) Ficoll solution was collected. The resultant embryo and cell fractions were used as a source of RNA for this study. lsolation and Analysis of RNA and DNA
Total RNA was extracted and purified according to the procedure of Berry et al. (1985) with the following modifications. After the precipitation of RNA with 2 M LiC1, contaminating material was removed by adjusting the solution to 30% (v/v) ethanol and incubating at 22OC for 15 min. The resulting precipitate was removed by centrifugation at 10,OOOg for 10 min at 15OC. RNA was precipitated at -2OOC from the resulting supernatant by adjusting it to 0.3 M sodium acetate, 70% (v/v) ethanol. Poly(A) RNA was purified by oligo(dT)-cellulose chromatography. The relative poly(A) RNA content of each preparation was determined by solution hybridization to [5-3H]polyuridylate.An aliquot of the RNA preparation was mixed with molar excess of [5-3H]polyuridylate and allowed to anneal to the poly(A) tail of the mRNAs. Excess single-stranded [5-3H]polyuridylatewas digested with S1 nuclease, and the double-stranded 15-3H]polyuridylate annealed to the poly(A) tract of the mRNA molecules was precipitated with 20% TCA. The precipitate was collected by filtration and washed with 70% ethanol, and the radioactivity was determined by liquid scintillation counting. This procedure allows the relative quantitation of oligo(dT)-selectedmRNA preparations. Equal concentrations of polyuridylate-normalized poly(A) RNA (approximately 0.5 gg) were fractionated by electrophoresis in 2% agarose gels containing 6% formaldehyde, and the RNA was transferred to a nitrocellulose matrix (Thomas, 1983). The matrix was processed and hybridized as described previously (Balcoumbe and Key, 1980). The data presented are from poly(A) RNA isolated from var Danver line DC2 except those for the nonembryogenic cell line, for which RNA was isolated from line DC5 [a line that no longer has the capacity to form embryos when the cells are transferred to (-)2,4-D medium]. The same experiments were conducted with similar results using total RNA isolated from cells and embryos of D. carota var Danver (line DC2) and from the wild genotype (WC49-1). In these experiments, the RNA in each lane of the gel (approximately 20 Pg) was stained with ethidium bromide before transfer to the matrix. Genomic DNA from D. carota var Danver (line DC2) was isolated, digested, subjected to electrophoresis, and analyzed by Southern blot (Wurtele and Bulka, 1989). Construction of the cDNA Library and lsolation of pEMB-1 and the emb-1 Cene
Cell-suspension cultures of D. carota var Danver (line DC2) grown for 14 d in medium lacking 2,4-D (Keller et al., 1988) were used as the source of the RNA for the construction of
Gene Expression in Embryogenesis
the cDNA library. These cultures contained a heterogeneous mixture of small and large globular embryos, heart embryos, torpedo embryos, germinating embryos, and nonembryogenic cells. Double-stranded cDNA was synthesized from poly(A) RNA and EcoRI methylase treated according to methods described by Gasser et al. (1989). These cDNAs were inserted into the EcoRI site of AgtlO. A portion of this library (approximately 20,000 phages) was differentially screened (Huynh et al., 1985) with 32P-labeled cDNA, which had been reverse transcribed from poly(A) RNAs isolated from either (+)2,4-D or (—)2,4-D cultures. Inserts from recombinant X clones were subcloned into the EcoRI site of pUC19 (YanichPerron et al., 1985) for further analysis. The emb-l gene was isolated from a genomic library of carrot in the vector Charon 34 (kindly provided by Dr. Lynn Zimmerman, University of Maryland). Nucleotide and amino acid sequence analysis was done using programs from the Genetics Computer Group, University of Wisconsin. Primer Extension The accumulated mRNA transcripts, which may reflect the putative transcription initiation sites of the emb-l gene, were determined by a primer-extension experiment using as a template poly(A) RNA isolated from torpedo-stage carrot somatic embryos (similar results were also obtained using total RNA as a template). A 32P end-labeled oligonucleotide (3'-GAATGCTTCTTCTGCTGCAT-5'; 0.07 pmol) that is complementary to the EMB-1 mRNA was annealed to 500 ng of poly(A) RNA in a total volume of 10 /nL at 45°C for 30 min. After cooling to room temperature, the solution was adjusted to a total volume of 30 fiL containing 0.8 mM dGTP, 0.8 HIM dCTP, 0.8 min dATP, 0.8 mM dTTP, 6 mM DTT, 15 mM MgCl2, 40 mM Tris-HCl (pH 8.3), and 8 units of reverse transcriptase. After the solution was incubated at 42°C for 1 h, the extension products were precipitated by ethanol and analyzed by electrophoresis in a urea acrylamide-sequencing gel. The lengths of the reverse transcriptase extension products were determined by comparing them to dideoxy-terminated sequencing reaction products obtained from the genomic clone emb-l, using the identical oligonucleotide as a sequencing primer. In Situ Hybridization to RNA
In situ hybridization using paraffin-embedded sections was carried out as described by Ausubel et al. (1989), with the following modifications. 35S-RNA probes of approximately 450 bases were synthesized from pEMB-1, which contained cDNA inserts in the pBluescript (SK)+ vector. Sense and antisense probes for the EMB-1 RNA were produced by linearizing pEMB-1 with Xbal or H/ndlll and transcribing with T3 or T7 RNA polymerase. Embryos were fixed with 4% paraformaldehyde and 0.25% glutaraldehyde in 0.05 M Pipes buffer (pH 7.3). Pretreatment with 2 mg mL"1 of proteinase K was for 30 min. The hybridization solution was in a total volume of 15 to 20 /uL consisting of 50% (v/v) formamide, 0.3 M NaCl, 10 mM Tris-HCl (pH 8.0), 1 mM EDTA, IX Denhardt's solution, 500 jig mL"1 of yeast tRNA,
305
TS
o O
Q)
(»
a
II! 2 2 „
o w
• a Iu
2 s s.
c
01 Ol
o
w
01 3 Ol A 0 O
>»
a 01
a c E o
ID
z
a •B
r CO
01
X
>"
a E
Ol
01
01
c
o •o 01 a 6
CO
K
c E 01
a
01 Ol
o
01 Ol
o
J*
^
o
o
111
111
M 01
c
c
a
z
z
o
0
> 01
w
o
o
oc
•it Figure 1. Temporal accumulation of EMB-1 mRNA during embryo development in D. carota. Poly(A) RNA was fractionated by electrophoresis, transferred to a Magnagraph membrane, and hybridized with 32P-labeled cDNA clone pEMB-1. Poly(A) RNA was isolated from unfractionated cultures grown in (+)2,4-D and (-)2,4-D medium, cells and embryos of specific developmental stages, leaves and roots of 3-week-old, greenhouse-grown plants, dry seeds, flowers at anthesis, and cells of a cell line lacking embryogenic capacity. Cells lacking embryogenic capacity had been grown to the log phase of growth either in culture medium containing 0.3 mg L~' of 2,4-D or in a similar medium but without 2,4-D. A single EMB-1 mRNA is present in all stages of embryos and in seeds containing zygotic embryos.
10% dextran sulfate, 500 jug mL"1 of polyadenosine, and 10 ^L of 0.3 /ug mL"1 of 35S-labeled sense or antisense EMB-1 RNA probe. Hybridization was carried out for 12 to 16 h at 65°C in a humidified chamber. After hybridization, the tissue sections were incubated with 20 mg mL"1 of RNase A. The final, most stringent, washes were in O.lx standard sodium citrate (Ix standard sodium citrate is 0.15 M NaCl and 0.015 M trisodium citrate [pH 7.0]) and 0.1% SDS at 65°C. The stringent hybridization and wash conditions significantly reduced background hybridizations. In some instances a 32PDNA probe was used and hybridization was according to the procedures of Ausubel et al. (1989). Slides with hybridized tissue sections were coated with nuclear track emulsion (Kodak NTB 2), exposed for 12 h to 4 d, and developed. Photographs were taken with a Leitz microscope with bright-field illumination. In situ hybridizations of wild carrot were done using somatic and zygotic embryos at many stages of development, leaf and root of 3week-old plants, and fruit (including embryo, endosperm, aleurone, seed coat, and maternally derived tissues) at many stages of development from globular embryo formation through seed desiccation. The data presented depict a single embryo or seed at each stage; a minimum of four samples at
Wurtele et al.
306
K
Plant Physiol.Vol. 102, 1993
307
Gene Expression in Embryogenesis each developmental stage was examined and found to have similar patterns of EMB-1 mRNA distribution. RESULTS
lsolation of Embryo-Specific cDNAs
Somatic cells of carrot were maintained in an unorganized state by culturing them in the presence of exogenous auxin [(+)2,4-D culture]. Cells in a (+)2,4-D culture divide but do not undergo development. Cells were transferred to a culture medium lacking exogenous auxin [(-)2,4-D culture] to trigger the developmental process leading to the formation of embryos. These (+)2,4-D and (-)2,4-D cultures were fractionated into developmentally specific stages of cells and embryos using simple procedures (Keller et al., 1988; Wurtele et al., 1988). To identify gene products that accumulate specifically or differentially during embryo development, we constructed a cDNA library in the vector XgtlO using poly(A) RNA isolated from cells and embryos of a (-)2,4-D culture as the template. Poly(A) RNA (3 pg) yielded a library of 106 recombinant bacteriophage. Differential screening of 2 X 104recombinant phage with single-stranded, 32P-labeled cDNAs synthesized from poly(A) RNA isolated from (-)2,4-D and (+)2,4-D cultures identified 27 clones representing poly(A) RNA sequences that accumulated differentially between these two cultures. Twenty-one of these 27 clones represented poly(A) RNAs that were more abundant in (-)2,4-D cultures than in (+)2,4-D cultures, and the remaining six were more abundant in (+)2,4-D cultures than in (-)2,4-D cultures. To further characterize these clones, the cDNA inserts were subcloned into the EcoRI site of pUC19. The clone discussed here, pEMB-1 (Ulrich et al., 1990), was selected as representing an mRNA that accumulated to higher levels in (-)2,4-D cultures than in (+)2,4-D cultures. Embryo-Specific Accumulation of EMB-1 mRNA
The differential accumulation of the EMB-1 mRNA was confirmed by RNA gel blot analysis of poly(A) RNA isolated from (-)2,4-D and (+)2,4-D cultures. pEMB-1 hybridized to a single species of poly(A) RNA of about 560 bases in length. The abundance of this mRNA was at least 40-fold higher in
EMB-I wheat Em
1 met ala ser * * * gln gln g l u l y s l y s g l u leu asp ala arg * * * --. - - - - - - gly - - - - - - - - - arg ser rJin~ . . . . . * * * ..lYS
Em-1 wheat Em
14 ala a r g g l n g l y glu thr val val pro g l y g l y thr gly gly lys - - - - - - qlu -... _ .-. - -..-.-.._. ._. .._ .-. ..---.
EMB-1
29 t h r leu glu ala gln gln his leu ala glu g l y arg ser lYs g l y - - - .._ ... .-. gluasn ... -..... _._.. .-arg .--
wheat Em
Ser
EMH-1 wheat Em
gly gln thr arg l y s glu gln leu gly yly g l u gly tyr his glu ... .._ ._. ... arg ._. ... ",et -..gl" _....- - - ser g l n
EMB-1 wheac Em
m e t gly a r g lys gly gly leu ser asn asn asp met ser g l y g l y ... ... ._. _._ ..~ - - - - - - - - - chr ... _.glu . --- --- ---
44
59
EMH-1 w),eat Em
EMB-1 wheat Em
14 glu arg ala glu gln g l u g l y ile asp ile asp g l u ser l y s phe
asp ~ . ._. . ala arg ~ _ _.~ . . . -. ..... ... _ - - - - - - - - - - 89 arg chr lys l y s STOP lys - - - - - - s e r STOP
Figure 3. Derived amino acid sequence of the EMB-1 protein. The amino acid sequence of the EMB-1 protein (Ulrich et al., 1990) is compared with that of the Em protein of wheat (Litts et al., 1987). Dashes indicate identical amino acids; the asterisks represent an absent amino acid.
(-)2,4-D cultures than in (+)2,4-D cultures (Fig. 1). We hypothesized that the low level of EMB-1 mRNA accumulation in the (+)2,4-D cultures might be due to a small proportion of the cells that begin to divide and differentiate into globular embryos even in the presence of 0.3 mg mL-* of 2,4-D. When these globular embryos were removed from the (+)2,4-D cultures by passing through a 190-pm mesh-size screen, and the remaining cells were fractionated into populations of embryogenic cell clusters and nonembryogenic cells (Wurtele et al., 1988), no EMB-1 mRNA was detected in either fraction (Fig. 1).Thus, even the small amount of EMB1 mRNA detected in the (+)2,4-D cultures can probably be attributed to the few small globular embryos present in these cultures. To determine more precisely the stages of embryo development at which the EMB-1 mRNA accumulates, (-)2,4-D cultures were fractionated into cells and embryos at defined stages of embryogenesis, and poly(A) RNA isolated from each fraction was subjected to northem analysis. The EMB1 mRNA accumulated in globular plus heart, torpedo, and germiiiating embryos (Fig. 1).Even after very long exposures
~~~~~
Figure 2. In situ localization of EMB-1 mRNA in zygotic embryos (A-G) and somatic embryos (H-M) of carrot. Tissue was fixed, embedded in paraffin, sectioned, and hybridized in situ with 35S-labeled sense or antisense transcripts synthesized from pEMB-1 or with "P-labeled EMB-1 DNA (H-J only). The black silver grains represent hybridization to EMB-1 mRNA, indicating its tissue distribution. A, Globular zygotic embryo and seed, stained with toluidine blue. Bar, 150 pm. B, Mature torpedo zygotic embryo and seed, stained with toluidine blue. Bar, 150 pm. C, Globular zygotic embryo hybridized with EMB-1 antisense R N A showing low level of EMB-1 mRNA accumulation throughout globular embryo. Bar, 50 pm. D, Globular zygotic embryo hybridized with EMB-1 sense RNA as a negative control. Bar, 50 pm. E, Heart zygotic embryo hybridized with EMB-1 antisense RNA. Bar, 50 pm. F, Mature torpedo zygotic embryo (5-10 d after the initiation of desiccation)hybridized with EMB-1 antisense RNA. The embryo was dissected from the seed before fixation. Bar, 150 fim. G, Mature torpedo zygotic embryo (the same age as that shown in F) hybridized with EMB-1 sense RNA as a negative control. The embryo was dissected from the seed before fixation. Bar, 150 pm. H, Globular somatic embryo hybridized with EMB-1 DNA. Bar, 50 pm. I, Heart somatic embryo treated with RNase and hybridized with EMB1 DNA as a negative control. Bar, 50 pm. 1, Heart somatic embryo hybridized with EMB-1 DNA. Bar, 50 pm. K, Young torpedo somatic embryo hybridized with EMB-1 antisense RNA. Bar, 50 pm. L, Torpedo embryo hybridized with EMB-1 antisense RNA. Bar, 50 pm. M, Germinating embryo hybridized with EMB-1 antisense RNA. Bar, 150 pm.
Wurtele et al.
308
1
iu
kb 14.010.77.4 4.4 -"
3.3 -
I
1.1 ~
Figure 4. Southern blot analysis of carrot genomic DMA hybridized with pEMB-1. Total genomic DMA, digested with the restriction endonuclease H/ndlll, EcoRV, or EcoRI was fractionated by agarose gel electrophoresis, transferred to a Magnagraph membrane, and hybridized with 32P-labeled cDNA clone pEMB-1. pEMB-1 hybridizes with two or three restriction fragments.
of the RNA blots, the EMB-1 mRNA could be detected only in poly(A) RNA isolated from somatic embryos. EMB-1 mRNA was not detectable in poly(A) RNA isolated from the nonembryogenic cells from the same (—)2,4-D cultures as the embryos. The absence of detectable accumulation of EMB-1 mRNA in nonembryogenic cells purified from (—)2,4-D cultures indicates that the accumulation of this mRNA is specific for the embryogenic process and not simply a response to the transfer of the cells to a medium lacking auxin. To substantiate this conclusion further, two carrot cell lines lacking embryogenic capacity were analyzed for the accumulation of EMB-1 mRNA. These cell lines had been maintained in vitro for more than 5 years and no longer formed embryos when they were transferred to a medium lacking 2,4-D. Regardless of whether they were grown in medium containing or lacking exogenous auxin, the carrot cell lines lacking embryogenic capacity did not accumulate EMB-1 mRNA to levels detectable by RNA gel blot analyses (Fig. 1). This observation supports the conclusion that the EMB-1 mRNA accumulation is not stimulated by the transfer of the cells to medium lacking auxin per se but, rather, is specifically induced during the developmental process of embryogenesis. To examine whether EMB-1 mRNA accumulates in organs of the mature carrot sporophyte, pEMB-1 was hybridized to poly(A) RNA isolated from a variety of organs of the carrot plant, including leaves, roots, and immature inflorescences (Fig. 1). Accumulation of the EMB-1 mRNA was not detectable in leaves, roots, or inflorescences. However, dry carrot seeds containing zygotic embryos did accumulate EMB-1 mRNA (Fig. 1). These data indicate that EMB-1 mRNA accumulates specifically during embryogenesis.
Plant Physiol. Vol. 102, 1993
In Situ Localization of EMB-1 mRNA during Embryo Development
To determine whether the emb-l gene was expressed in all cells or in specific cell types, and to compare the pattern of the spatial distribution of EMB-1 mRNA within the embryo, accumulated EMB-1 mRNA was visualized by in situ hybridization to sections of zygotic and somatic embryos of carrot at different stages of development (Fig. 2). Zygotic embryos are shown in Figure 2, A to G. Wild carrot fruit are bicarpellate, composed of two mericarps; seeds partially dissected from the surrounding fruit and stained with toluidine blue are shown for orientation (Fig. 2, A and B). Carrots are endospermic (Gray et al., 1984), i.e. they retain most of the endosperm cells in the desiccated seed; the embryo grows to about one-third the length of the seed before desiccation. The endosperm of the carrot seed develops in advance of the embryo. The seed shown in Figure 2A is 20 d postanthesis and contains a globular embryo. At this stage, about 80% of the endosperm is already present (Gray et al., 1984). A group of cytoplasmically dense endosperm cells immediately surrounding the embryo are delineated by darker staining with toluidine blue (Fig. 2A); these cells will eventually degenerate and may provide a source of energy, nitrogen, phosphorus, and other substrates for the growing embryo. These cytoplasmically dense endosperm cells can be seen in higher magnification in Figure 2, C and D; in Figure 2E a heart embryo is visible, surrounded by endosperm cells that are just beginning to degenerate (as manifested by the somewhat blurred appearance of these endosperm cells). The seed shown in in Figure 2B is 55 d postanthesis; it is beginning to turn brown, is undergoing weight loss, and contains a fullsized torpedo embryo. Here, the endosperm cells immediately surrounding the embryo have completely degenerated. The density and pattern of grain distribution following hybridization with 35S-labeled antisense EMB-1 RNA as a probe reflects the regional distribution of the EMB-1 mRNA in the carrot embryos (Fig. 2, C, E, and F). Control hybridizations using 35S-labeled sense EMB-1 RNA as a probe are included for comparison (Fig. 2, D and G). EMB-1 mRNA is present at low levels in globular zygotic embryos (Fig. 2C); its accumulation appears to be uniform throughout the embryos. As pattern development is initiated with the formation of the heart-stage zygotic embryo, the accumulation of EMB1 mRNA begins to show polarity (Fig. 2E); in these embryos, EMB-1 mRNA accumulates to higher levels throughout the peripheral regions of the embryo, i.e. the location of the newly forming cotyledons, protoderm, and root. The procambial region of the late-stage heart embryo pictured has just begun to develop; it is characterized by cytoplasmic density and the absence of accumulated EMB-1 mRNA. In mature torpedo zygotic embryos EMB-1 mRNA accumulation dramatically increases (Fig. 2F; this zygotic embryo was removed from the seed before fixation). The majority of EMB-1 mRNA in torpedo zygotic embryos is in the meristematic regions, i.e. the procambium, the root and shoot meristem, and the protoderm in the region of the cotyledons, although it is present in all regions of the embryo. EMB-1 mRNA was not detected in the endosperm, aleurone layer, or maternally derived portions of the carrot fruit at any stage of development, nor
Gene Expression in Embryogenesis
309 ggaatt
-700
-651
ga:aggtaggttatgctgctata:attaegtagctaca~taa:atta:agca~ga::tc:tgacaaaaatgagtgaaattcattatgta:atatattgtt 255,
-600
aacagtatccataagatattaatcgtattgatcacggattagtcgaataatttrgataatcggatatt~aatcggttggacgaacaatcaaattaarcgg -451
-500
taattatacct:tggagaagaaaacccttgattga:tat~aa:ct~caatcatttgatctcaaacacaaatcaaataa:aaaacccttçcgttaac:gt
**
-351
-400
attctcctgcttgaattgaagctaacattgaattcaacggaagt~aaattqttttaaaatagtttgaaaactgaaaacaaatttcacatcttagcataaa -251
-300
gataiatatatggtgatataattatatatcgctttgcttggta~agaaaatgagtacg~acgttcatttcattttacatctttccaaccggcaatgatct -151
-200
ctctcgaactgcacctcaagacaagtg~ggagataatactgactt~cctagctagctgatt~=atttagtggg=cacttcttuatacut¶ttatgatatc -100
a a c a a t t g t g a b a c c c t a i ; a a c a c c a g ~ i ; a g t c a c a c g t tt c a c t i r t g a t c t g a a c t a c 4 i q c c a c t g c c tt a c t c c t c a t c a a c n c t a c a t u t ~
**
1
*50
a a t a a a c c t c g g t a a c t c c D a c ct a g t c a c t a t g c a t c t a a t g T M C C A C G A C A T T G A Z I G A G c ~ C T T A C G M G A C G A C G T A G G ~ ~ G A A T T C A 1130
191
T T T A G P A G M A G C T M G ATG GCG AGT CAA CAA GAG AAG M G GAG CTG GAT GCT AGG GCG AGG C44 GGA GAG ACC GTG GTT m e t a l a s e r g l n gln g l u ? y s l y s g l u leu asp a l a a r g a l a a r g g l n g l y g l u t h r val v a l 1169
1211
C C T GGC GGG ACT GGT GGG AAG AGT C T T GAA GCC CAG CAG CAC CT? GCT GAA Ggtacatacatataattgatcatagactccatat >
> r619
>>>>>>>> -663
~GTTACTGGAATAACTA~ACTTATAC‘rGGTTATGT~CGGGTCTATGTATGT~~ACCTACTGG~TAAATATGTGTTATATGTCTCAAGAA >>>>,>>
