| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Biology, Whitman College, Walla Walla, Washington 99362 USA
Received for publication March 30, 2000. Accepted for publication June 20, 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Arabidopsis • cotyledon • embryogenesis • morphogenesis • pattern formation • shoot apical meristem • suspensor • twinning
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Schwartz, Vernon, and Meinke, 1997
).
Suspensor formation involves processes of central importance to embryogenesis, including the establishment of embryo polarity, early cell differentiation, and programmed cell death; it is therefore interesting from a developmental biology perspective. In most species, the suspensor develops from the basal cell produced after the first zygotic division, and its specification likely reflects morphogenetic apical–basal polarity present in the zygote itself (Schwartz, Vernon, and Meinke, 1997
). The suspensor rapidly completes development to carry out its physiological functions and is the first specialized structure to differentiate in most plant embryos. Later in development, suspensor degeneration provides the earliest example of developmentally programmed cell death in plant development.
Arabidopsis thaliana has become an important system for genetic studies of plant development, including many aspects of plant embryogenesis (Goldberg, de Pavia, and Yadegari, 1994
; Meinke, 1995
). In Arabidopsis the suspensor is specified in the two-cell proembryo, after the zygote divides asymmetrically to form an elongate basal cell and a smaller apical terminal cell (Mansfield and Briarty, 1991
; West and Harada, 1993
). Soon after its formation, the basal cell undergoes a rapid series of divisions to produce a suspensor consisting of a single column of six to eight cells. The suspensor is fully formed by the early globular stage of embryogenesis and maintains its structure until it begins degenerating during the linear stage (West and Harada, 1993
; Yeung and Meinke, 1993
).
Characterization of Arabidopsis developmental mutants has established that interaction between the suspensor and EP is of central importance for suspensor development: many genetic defects that disrupt the embryo-proper trigger abnormal development in the suspensor. Three classes of Arabidopsis mutants with abnormal suspensors have been described in detail. In one class, consisting of the sus and rsp mutants, aberrant EP development is followed by suspensor cell proliferation and at least partial transformation, resulting in development of inviable cell masses that resemble the mutant embryo proper (Schwartz, Yeung, and Meinke, 1994
; Yadegari et al., 1994
). The second class of suspensor mutants, the twin (twn) mutants, can produce viable secondary embryos via embryogenic transformation of the suspensor and thus give rise to an abnormally high percentage of seeds containing twin embryos (Vernon and Meinke, 1994
; Zhang and Somerville, 1997
). One such mutant, twn2, has been characterized at the molecular level: the phenotype is caused by a unique regulatory mutation that eliminates expression of an essential valyl-tRNA synthase gene in the EP, but not in the suspensor. As a result, the twn2 EP degenerates early in development, but suspensor cells survive, enter into embryogenic development, and form one or more embryos (Zhang and Somerville, 1997
). It is clear from the sus, rsp, and twn2 phenotypes that cells of the suspensor have embryogenic potential, and that this potential is normally suppressed by interaction with the EP. This interpretation is consistent with classical studies of EP–suspensor interaction in a variety of species (Haccius, 1955, 1963
; Yeung and Meinke, 1993
). A third class of embryo-defective mutants, the late embryo mutants, exhibit belated suspensor degeneration relative to wild type, due to delayed or impaired development in the embryo proper (Vernon and Meinke, 1995
). Thus, many aspects of suspensor development, including cell proliferation, embryogenic potential, and even the onset of cell death, appear to be influenced by the embryo proper.
Despite the importance of EP–suspensor interactions, little is known about the nature of cell communication between these two parts of the plant embryo. One mutant that may provide clues is twn1. In 
8% of homozygous mutant seeds, twn1 produces twin embryos by embryogenic transformation of suspensor cells, while exhibiting only occasional and variable defects in the early EP (Vernon and Meinke, 1994
). In contrast to twn2, suspensor transformation in twn1 is not triggered by death of the embryo proper. Rather, twinning occurs in the presence of a viable, and often morphologically normal, EP. Thus, in twn1, suspensor transformation does not appear to be an indirect consequence of arrested development in the EP, as is the case with other suspensor mutants. Based on these observations, Schwartz, Yeung, and Meinke (1994)
proposed that TWN1 functions directly in EP–suspensor cell communication.
Three general models of TWN1 activity have been proposed to explain twinning and the occasional morphological defects observed in twn1 embryos (Vernon and Meinke, 1994
). In all three models, suspensor transformation occurs in the mutant because EP–suspensor communication is compromised and suspensor development is not properly inhibited. In the first model, TWN1 is active in the suspensor and is required for proper response to growth-inhibiting signal(s) from the embryo proper. In the second model, TWN1 is proposed to act in the embryo proper to suppress (directly or indirectly) embryogenic development in the neighboring suspensor. The third model proposes that TWN1 may have more than one developmental function and act in different embryo regions, affecting suspensor and EP development separately. One prediction of this third model is that the twn1 EP should show developmental defects that cannot be attributed to defects in the suspensor and that occur separately of twinning in twn1 homozygotes.
To better define TWN1 gene function and distinguish between the models described above, we have further characterized the twn1 mutant phenotype, with a focus on late embryo and post-embryonic morphology. We report here that the twn1 mutation alters cotyledon patterning, thus disrupting the establishment of bilateral symmetry and subsequent morphogenesis in the Arabidopsis embryo proper. The observed cotyledon defects suggest that TWN1 helps specify the location, size, and boundaries of cotyledon-forming fields in the wild-type embryonic shoot apex. Comparison of the twn1 phenotype to those of other cotyledon mutants indicates that twn1 is a unique Arabidopsis cotyledon mutant, with defects restricted to embryonic and primary leaf development. In addition to characterizing twn1, we determined that another polycotyledonous mutant, amp1, also produces twins by suspensor transformation, confirming that there is overlap between the genetic mechanisms underlying suspensor and cotyledon development. Our results suggest that TWN1 acts in pathways that impart or maintain positional information in both the suspensor and shoot apex of the developing Arabidopsis embryo.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Seeds for amp1 (amp1-1 allele), pin1 (pin1–1 allele), and pid1 (pid1–2 allele) were obtained from the Arabidopsis stock center at Ohio State University. Seeds for cuc1 and cuc2 were kindly provided by Dr. M. Tasaka (Kyoto University). Plants grown for genetic crosses, embryo microscopy, or characterization of vegetative development were seeded in soil, chilled at 4°C for 2–3 d, and grown to maturity at 22°C under 16 h light/8 h dark cycles in a controlled climate growth chamber. Seedlings grown for phenotype analyses were sterilized and germinated in culture in petri dishes containing germination medium as previously described (Vernon and Meinke, 1995
). For comparison of the dark response in twn1, amp1, and wild type seedlings, plates were exposed to light for 12–16 h to stimulate germination, then wrapped in foil and set vertically in the dark. Seedlings germinated in culture that were chosen for further use in genetic crosses, heritability studies, or seed collection were transplanted to soil after 7–10 d in culture and grown to maturity under the conditions described above.
Phenotypic analysis and microscopy
Phenotypes of seedlings germinated in culture were scored and characterized by observation under a Wild Heerbrugg M8 dissection microscope. For characterization of cotyledon veination patterns, seedlings or excised cotyledons were cleared in ethanol prior to viewing. Dark response of wild type, amp1, and twn1 seedlings was gauged by manual measurement under a dissection microscope of hypocotyls and roots of seedlings following 5 d growth on plates in the dark. To normalize for the generally reduced size of amp1 seedlings, hypocotyl : root length ratios, rather than simply hypocotyl length, were measured.
Scanning electron microscopy was carried out essentially as described by McConnell and Barton (1995)
. Seedlings were selected from plates and fixed on ice in 2% glutaraldehyde in 0.1 mol/L sodium phosphate buffer (pH 7.4; PBS) for 6 h. Seedlings were washed with cold PBS and dehydrated by 10-min incubations in increasing amounts of ethanol in PBS, starting with 5% ethanol and increasing in 10% increments to 95%. Final dehydration was accomplished by an additional 30-min incubation in 95% ethanol followed by 5 30-min incubations in 100% ethanol. Seedlings were subjected to critical point drying and sputter coating immediately following ethanol dehydration and viewed with a Jeol JSM-T300 scanning electron microscope.
For observation of developing embryos, seeds at various stages of development were excised from siliques and cleared in Hoyer's solution as previously described (Vernon and Meinke, 1994
). Cleared seeds were viewed as whole mount preparations on an Olympus BX-60 compound microscope equipped with Nomarski differential interference contrast optics.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). The polyembryonic phenotype showed 8% penetrance; that is, most mutant seeds germinated to produce a single seedling. In addition to twinning, variable defects in the EP were occasionally observed, and one model of TWN1 activity proposed that the gene may have other developmental functions affecting the EP, as well as the suspensor. The identification of twn1 EP defects that are independent of suspensor transformation would support this model and rule out alternative models in which TWN1 acts solely in the early EP or suspensor to maintain suspensor identity (Schwartz, Yeung, and Meinke, 1994
; Vernon and Meinke, 1994
). To further investigate TWN1 gene function and determine if TWN1 has developmental functions specific to the embryo proper, we characterized late embryo and seedling morphology in twn1 homozygotes.
Diverse cotyledon defects in twn1 mutant seedlings
Mutant seeds produced by selfed twn1 homozygotes were germinated in culture and examined by dissection microscope. Consistent with previous studies (Vernon and Meinke, 1994
), 8% of seeds yielded two seedlings and most seedlings resembled wild type. However, 10% of mutant seedlings (N = 1277) displayed dramatic defects in cotyledon number, arrangement, and/or morphology. Similar cotyledon defects were not observed on wild-type seedlings (N > 300). Cotyledon defects were observed on seedlings germinated from monoembryonic as well as polyembryonic seeds, indicating that the defects were not a consequence of twinning. A wild-type seedling and representative examples of twn1 seedlings with abnormal cotyledons are shown in Figs. 1–6. Defects in cotyledon number included apparent single cotyledons (Fig. 2) as well as extra cotyledons (Figs. 4 and 5). Fused cotyledons were also observed. Fusion phenotypes ranged from almost complete fusion to subtle connections at the base of clearly distinct cotyledons (Figs. 3 and 4). In all cases of fusion, cotyledons were connected from the base, indicating that fusion occurred during cotyledon formation and not postgerminatively. Defects in cotyledon number often occurred in combination with fusions of varying severity, making strict categorization of cotyledon abnormalities difficult. A small percentage of twn1 homozygotes (<0.5%) had cotyledons with trichomes (Fig. 6). In Arabidopsis, trichomes are a characteristic of vegetative organs and are not found on wild-type cotyledons. When present on twn1 cotyledons, trichomes were far less numerous and irregularly distributed than on leaves. Nevertheless, their occasional presence on cotyledons suggests that postembryonic developmental programs were precociously deployed to some extent during twn1 cotyledon development.
|
|
recently determined that supernumerary cotyledons on Arabidopsis xtc mutants are actually precocious leaves that initiate prematurely from the embryonic shoot apical meristem. However, all cotyledons observed on twn1 tricots were true cotyledons, based on the following criteria: (1) all arose from the first node; (2) they consistently had the simple veination pattern characteristic of cotyledons; and (3) precocious embryonic leaf initiation and abnormally large shoot meristems have not been observed on developing twn1 embryos, as is the case with mutants that produce cotyledon-like precocious leaves (Vernon and Meinke, 1994
; Conway and Poethig, 1997
; and see below).
Both cotyledon patterning and organogenesis are disrupted in twn1
The different cotyledon defects in twn1 seedlings could possibly be due to initiation of cotyledons at abnormal positions, or to improper recruitment of apex cells during cotyledon initiation at proper locations. The occurrence of truly tricotyledonous seedlings indicated that twn1 at least in some cases affects the position (patterning) of cotyledon initiation. Organ fusion phenotypes, however, could conceivably arise through two different developmental mechanisms: (1) cotyledons could initiate normally on opposing sides of the apex, but cells between initiation sites could be improperly recruited into cotyledon primordia, creating a single organ with a wide base; or (2) cotyledons could incorrectly initiate adjacent to each other on one side of the embryonic axis, and fusion could somehow result as a consequence of the abnormal proximity of developing organs. To determine how cotyledon fusions were occurring, we used scanning electron microscopy to view the apices of twn1 seedlings with cotyledon fusions of different severity. Figure 11 shows the apex of a representative "monocotyledonous" seedling similar to that shown at lower magnification in Fig. 2. A "collar" of cotyledon tissue encircled much of the shoot apex on such seedlings, spanning the region between normal cotyledon initiation sites on opposite sides of the apex. Thus, at least in some cases, the twn1 mutation appears to extend the field of cells recruited for cotyledon formation, as suggested in the first scenario described above. However, the more subtle organ fusions observed on some seedlings may arise through the second mechanism. Such a fusion is visible in Fig. 12, which shows the apex of a representative tricot seedling similar to the one shown at lower magnification in Fig. 4. Three distinct cotyledons are positioned around the axis. The bases do not appear to extend abnormally around the apex, but two of the organs are slightly fused near the base. Thus, cotyledon fusions on different seedlings may arise by either of the mechanisms described above, and the twn1 mutation can disrupt both the pattern of organ initiation (as in tricots), or the recruitment of cells into cotyledon primordia during organogenesis (as in seedlings with single large organs or fusions).
|
; Conway and Poethig, 1997
). Thus, twn1 cotyledon defects do not result from suspensor defects, obvious defects in embryo morphology, or premature vegetative meristem activity. Rather, the twn1 mutation appears able to specifically affect cotyledon patterning and morphology.
|
10% penetrance. All twn1 homozygotes, regardless of their cotyledon morphology, consistently produced progeny with the full range of cotyledon defects at a frequency of 10% following self-fertilization. In addition, cotyledon defects and polyembryony appeared to be separate consequences of the twn1 mutation that could occur separately in populations of twn1 seedlings: all twn1 homozygotes, whether from polyembryonic or monoembryonic seeds, produced progeny with the full range of cotyledon phenotypes, as well as twins. We carried out additional genetic analyses to confirm these aspects of phenotype heritability.
To establish the heritability and incomplete penetrance of the diverse cotyledon defects in twn1 homozygotes, we tested two specific predictions: (1) 90% of the progeny of selfed twn1 homozygotes with dramatic cotyledon defects should have cotyledons that resemble wild type; and (2) the full range of cotyledon defects should be observed among progeny of such homozygotes. To test these predictions, we collected seeds produced by five tricotyledonous twn1 seedlings, germinated them in culture, and scored progeny phenotypes. As shown in Table 1, 90% of progeny had cotyledons resembling wild type, and the total frequency of cotyledon defects in progeny from each tricot was 10%. The progeny populations from each parent exhibited the full range of cotyledon defects, including single, fused, and extra cotyledons, and combinations of these. Thus, cotyledon defects are co-inherited and consistently expressed with 10% penetrance in progeny of selfed twn1 homozygotes, regardless of parent plant morphology.
|
10%. All F2 families also produced twins at the predicted frequency. Thus, cotyledon defects exhibited a consistent pattern of inheritence following reintroduction of the twn1 mutation into a wild type background, regardless of the phenotype of the mutant parent used in the cross.
|
Taken together, these genetic studies demonstrate that cotyledon defects in twn1 follow a pattern of inheritance characteristic of a single recessive mutation that is phenotypically expressed with a penetrance of 10%. Indeed, it is difficult to conceive of an equally straightforward alternative model that can account for the observed inheritance of cotyledon defects in this mutant. Along with the demonstration of twinning in another Arabidopsis cotyledon mutant (see below), these results also support the view that cotyledon defects and polyembryony are indeed pleiotropic consequences of the twn1 mutation. Absolute proof of the single-gene basis for these disparate developmental defects will ultimately require molecular complementation with a cloned copy of the TWN1 gene.
The twn1 mutant is not allelic to other Arabidopsis cotyledon mutants
Several other Arabidopsis mutants have been identified that exhibit at least some of the cotyledon defects observed in twn1: pin1, pid1, amp1, cuc1, and cuc2 (Okada et al., 1991
; Chaudhury et al., 1993
; Bennett et al., 1995
; Aida et al., 1997
). Except for cuc2, all of these mutations have been localized to different linkage groups than twn1, which has been assigned to the lower arm of Arabidopsis chromosome five (Vernon and Meinke, 1994
). Because cuc2 can cause fused cotyledons and has also been mapped to chromosome 5, we tested for allelism between it and twn1. As a single mutation, cuc2 causes limited cotyledon fusion (with a penetrance of only 0.5%), whereas cuc1,cuc2 double mutants have a unique, fully penetrant cup-shaped cotyledon phenotype (Aida et al., 1997
). Therefore, rather than carrying out standard complementation crosses between the twn1 and cuc2, we crossed twn1 with cuc1 and scored F2 seedlings for the fully penetrant cuc1:cuc2 double mutant phenotype. If twn1 were allelic to cuc2, 6.25% of the F2 progeny from such a cross would be predicted to have a single "cup-shaped" cotyledon. Phenotypes of F2 progeny clearly indicated that twn1 was not allelic to cuc2 (Table 3).
|
0.2%, N = 1500) of twin seedlings germinated from homozygous amp1 seeds in culture (Fig. 17). This polyembryony rate is far below that of twn1, but is still approximately ten times the twinning frequency observed in wild-type Arabidopsis (Akhundova, Schevchenko, and Grinikh, 1979
; Vernon and Meinke, 1994
). Nomarski analysis of developing amp1 embryos revealed that amp1 twins develop from suspensor cells via a mechanism resembling that previously described for twn1 (Fig. 18). Thus, two Arabidopsis developmental mutants independently reveal that common mechanisms influence suspensor and cotyledon development.
|
). Because of the phenotypic similarities between these mutants, we further compared their development. In addition to its embryonic defects, amp1 also displays fully penetrant, dramatic defects in postembryonic development, including cytokinin accumulation, rapid and disorganized leaf production, abnormally high leaf number, small size, multiple bolts, and photomorphogenic development in the dark (Chaudhury et al., 1993
; Chin-Atkins et al., 1996
). To determine if twn1 exhibited any of the vegetative defects observed in amp1, we further characterized postembryonic development in twn1. Wild-type Arabidopsis have two primary leaves that arise at opposite positions around the seedling axis. The twn1 seedlings with abnormal cotyledon numbers occasionally had one or three primary leaves (e.g., see Figs. 2 and 5). Also, leaf fusion was observed between twn1 primary leaves at a very low frequency (<0.5%). These phenotypes demonstrated that the twn1 mutation can have a limited effect on development after embryogenesis. However, in contrast to amp1, discernable developmental defects did not persist beyond the first leaf pair in twn1 homozygotes: even on seedlings with obvious defects in cotyledon or primary leaf pattern, the leaf number, rate, and arrangement beyond the primary leaf(s) consistently resembled wild type. Results of a representative experiment quantifying rosette leaf rate and number are shown in Fig. 19. Furthermore, no obvious alterations in bolt number, timing, or flower morphology were observed on twn1 under our growth conditions; twn1 also differed from amp1 in its response to growth in the dark, which resembled that of wild type (Fig. 19b). Thus, although twn1 and amp1 share some cotyledon defects and can produce suspensor-derived twins, twn1 does not appear to have long-term influence on the vegetative apical meristem or apical organ production. Rather, TWN1 appears to act in an embryonic pathway that impacts only the earliest phase of apical organ patterning in Arabidopsis. This observation is consistent with the view that genetic control of cotyledon formation differs at least in part from that of postembryonic leaves (Barton, 1998
; Lenhard and Laux, 1999
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Kaplan and Cooke, 1997
; Kerstetter and Hake, 1997
). For these reasons, cotyledons have served as markers for assessing genetic control of various aspects of embryogenesis. For example, Arabidopsis mutants with severe defects in cotyledon formation, such as gurke, have helped identify putative "patterning genes" essential for proper development of the apical region of the embryo (Mayer et al., 1991
; Torrez-Ruiz, Lohner, and Jurgens, 1996
). Mutations resulting in polycotyly, such as xtc and amp1, have led to insights about the control of shoot meristem activity and timing of organ initiation (Chaudhury et al., 1993
; Conway and Poethig, 1997
). Another group of Arabidopsis cotyledon mutants, the leafy cotyledon (lec) mutants, have allowed identification of key regulatory genes that control embryo-specific programs during seed development (Meinke, 1992
; Meinke et al., 1994
; West et al., 1994
; Lotan et al., 1998
).
Other Arabidopsis mutations have been identified that cause cotyledon defects resembling those we report here for twn1. Interestingly, like twn1, these mutations are pleiotropic, cause variable defects and have incomplete penetrance with respect to their cotyledon phenotypes. The amp1 mutation results in embryos with deformed or extra cotyledons, as well as occasional precocious leaves that resemble extra cotyledons (Chaudhury et al., 1993
; Conway and Poethig, 1997
). The pin1 and pid mutations, like twn1, cause variable phenotypes, causing cotyledon fusion, deformities, or formation of a third cotyledon on some mutant individuals (Okada et al., 1991
; Bennett et al., 1995
). Another group, the cuc mutations, result in a low frequency of fused cotyledons when present as single mutations, although cuc1, cuc2 double homozygotes consistently produce a single, cup-shaped cotyledon that encircles the embryonic apex (Aida et al., 1997
). The incomplete penetrance of each of these mutations with respect to cotyledons likely reflects redundancy in the genetic control of cotyledon formation (Taylor, 1997
; Barton, 1998
). Such redundancy could be due to the existence of genes with overlapping function and/or to the presence of multiple, distinct physiological mechanisms that act in parallel on cotyledon patterning.
Although the amp1, pin, pid, and cuc mutations can affect cotyledon development similarly to twn1, all also trigger dramatic developmental defects not observed in twn1. Many of these defects are in postembryonic development. Amp1 overaccumulates cytokinins and features a greatly enlarged shoot apical meristem, bushy, disorganized vegetative growth, and abnormal seedling response to growth in the dark (Chaudhury et al., 1993
; Chin-Atkins et al., 1996
). Pin1 and pid1 are defective in auxin-related processes, and they have severe defects on the inflorescence, including spike-like bolts with few or no cauline leaves or flowers (Okada et al., 1991
; Bennett et al., 1995
). Pin1 can also have fully fused primary leaves that encircle the apex. The cuc1 and 2 mutations also affect postembryonic development as single mutations, causing a low percentage of fused floral organs (Aida et al., 1997
). Thus, twn1 is a unique Arabidopsis cotyledon mutant that exhibits a much narrower range of defects in shoot apical development than other mutants in its class.
TWN1's role in the embryonic apex
The cotyledon defects observed in twn1 embryos expand the known role for TWN1, indicating that the gene is active in the EP during later stages of embryogenesis and that it influences embryo (and seedling) symmetry through its effects on cotyledon patterning. The diversity of twn1 cotyledon defects, including fusions and deformities as well as extra cotyledons, indicates that TWN1 does not strictly affect cotyledon number. Rather, the gene appears to have a broader impact on cell fate at the embryo apex. We propose that TWN1 helps define both the location and size of cotyledon-forming fields, such that organs initiate at the proper positions around the embryonic axis, and apex cells are subsequently recruited correctly into cotyledon primordia. This interpretation is supported by microscopy of mutant seedlings, which revealed defects in both the spatial arrangement of cotyledons (patterning) and the incorporation of apex cells into cotyledon tissue. All of the cotyledon deformities we observed, including abnormal number, asymmetry, fusions, and complex venation patterns, could result from mislocalization of cotyledon-forming fields, and/or misallocation of apical cells into those fields.
We suggest two general models for how TWN1 could influence cell fate in the embryonic apex. These models take into account various aspects of the twn1 phenotype and those of other cotyledon mutants; they are summarized below and in Fig. 20. The first possibility is that TWN1 could influence cotyledon development as part of a hormone-mediated mechanism. Hormones are prime candidates for mediating cell interactions in the plant embryo, especially between the EP and suspensor and within the shoot apex (Goldberg, de Pavia, and Yadegari, 1994
). Both auxins and cytokinins are known to affect embryo symmetry and cotyledon pattern (Schiavone and Cooke, 1987
; Cooke, Racusen, and Cohen, 1993
; Lui, Xu, and Chua, 1993
; Faure, Jullien, and Caboche, 1994
). Also, the other Arabidopsis cotyldeon mutants amp1, pin1, and pid are defective in growth regulator physiology: amp1 overaccumulates cytokinin (Chaudhury et al., 1993
), pin1 is defective in auxin transport, and pid in auxin signaling (Okada, 1991
; Galweiler et al., 1998
; Christensen et al., 2000
). However, it is unlikely that TWN1 has a broad role in growth-regulator accumulation or transport like AMP1 and PIN1, because twn1 lacks most of the other severe developmental defects seen in these other cotyledon mutants (e.g., the enlarged meristem, short hypocotyl, and disorganized vegetative growth of amp1). TWN1 could instead encode a downstream component of hormone-regulated developmental pathways, with localized roles restricted to the embryonic apex and suspensor (Fig. 20). Such a scenario would explain how twn1 exhibits some of the cotyledon defects observed in amp1, pin1, and pid1, without many of the other defects caused by widespread imbalances in hormone accumulation, sensitivity, or transport in these mutants.
|
) and thus can explain the incomplete penetrance of the Arabidopsis cotyledon mutations. Within the context of both these models, TWN1 is proposed to act on or within the embryonic apex to affect the size, number, and location of cotyledon-forming fields. Thus, TWN1 might be predicted to act upstream of genes that directly define or separate cotyledon primordia, such as CUC1 and CUC2. We do not propose that TWN1 influences vegetative shoot meristem establishment, architecture, or maintenance, because twn1 individuals do not exhibit persistent defects in postembryonic organ formation. Therefore, genes that function throughout development in fundamental meristem maintenance and organization, such as STM, PINHEAD/ZLL, and CLV (Barton and Poethig, 1993
; McConnell and Barton, 1995
; Clark, 1997
; Moussian et al., 1998
; Lynn, Fernandez, and Barton, 1999
), are predicted to be unaffected by TWN1 activity. To discern between the models proposed in Fig. 20, molecular characterization of twn1 will be necessary. Physiological measurements of hormone status in twn1 are difficult because of the embryo-specificity and incomplete penetrance of the mutation, and genetic strategies (such as double mutant studies with hormone mutants) cannot be interpreted meaningfully without knowledge of the molecular nature of the twn1 mutant allele.
In addition to influencing the size and location of cotyledon primordia, TWN1 appears to occasionally affect the timing of cotyledon formation. This conclusion is based on our observation of delayed cotyledon formation in some twn1 embryos (e.g., Fig. 14), as well as the infrequent presence of trichomes on twn1 cotyledons. In wild-type Arabidopsis, trichomes form on leaves but not cotyledons, and they can be considered an indicator of postembryonic development (Meinke, 1992
). Their presence on twn1 cotyledons indicates that mutant cotyledons are partially transformed into leaves in some individuals; this transformation is probably due to delayed development. Cotyledons and leaves are homologous organs: cotyledons are modified leaves that have evolved specialized embryonic functions (Kaplan and Cooke, 1997
). Consistent with this, there is ample genetic evidence that disrupted regulation of specialized embryonic developmental programs in cotyledons can cause cotyledons to resemble leaves, and vice versa. Dramatic conversion of cotyledons to leaf-like organs with trichomes has been observed in the Arabidopsis lec mutants, in which regulators of embryo-specific programs have been disrupted (Meinke, 1992
; Meinke et al., 1994
; West et al., 1994
; Lotan et al., 1998
). A converse scenario is observed in the xtc mutant, in which leaves initiate prematurely from the embryonic shoot meristem during seed development. As a result of this early initiation, embryo-specific programs are active in early xtc leaves, and the leaves exhibit cotyledon-like features such as late-embryo storage reserves and an absence of trichomes (Conway and Poethig, 1997
). The xtc phenotype indicates that the timing of apical organ development can affect organ identity. Analogously, the presence of trichomes on twn1 cotyledons is likely an infrequent and indirect consequence of delayed cotyledon initiation in some twn1 embryos, such that cotyledons occasionally complete their development in a context in which there is decreased expression of embryo-specific genetic programs.
Distinct developmental functions for TWN1 in embryogenesis
Pleiotropic developmental mutations can be difficult to interpret, but they can also be informative, for they can reveal common genetic components of seemingly unrelated developmental processes and can indicate multiple roles for a gene in development. Given the pleiotropic effects of the twn1 mutation on cotyledon and suspensor development, overall models of TWN1 gene function must be revised to account for the role of TWN1 in the embryo apex as well as its previously defined role in suspensor cell maintenance. Previous models suggested that TWN1 was essential for proper communication between the embryo proper and suspensor, and three general scenarios had been proposed (Schwartz, Yeung, and Meinke, 1994
; Vernon and Meinke, 1994
; summarized in the introduction of this manuscript). Models in which TWN1 is active solely in EP–suspensor communication can now be ruled out, as they fail to account for the cotyledon and primary leaf defects we observed. Our results are most consistent with an alternative model in which TWN1 has separate developmental functions affecting the suspensor and embryo proper (Vernon and Meinke, 1994
). Within the context of this general model, the most straightforward scenario is one in which TWN1 functions in a cell communication mechanism that affects both these embryonic regions. Cell interaction is known to be important throughout plant development to establish positional information, especially in the shoot apex (Barton, 1998
; Lenhard and Laux, 1999
). During embryogenesis it is thought to be of particular importance both for control of suspensor fate and cotyledon patterning (Yeung and Meinke, 1993
; Schwartz, Vernon, and Meinke, 1997
; Fernandez, 1997
; Harada, 1999
). Therefore, a role in cell communication is consistent with TWN1's previously proposed role in EP–suspensor interaction and with its newfound role in cotyledon patterning.
Our finding that a second cotyledon mutant, amp1, also produces twins via suspensor transformation demonstrates that the pleiotropic link between suspensor and cotyledon development is not restricted to twn1, and it further supports the interpretation that common genetic mechanisms influence development in these different embryo regions. How can individual genes such as TWN1 or AMP1 influence such different processes as suspensor maintenance and apical organ pattern? Actually, such a scenario jibes well with our proposed role for TWN1 in cell communication. In animal systems as diverse as Drosophila and vertebrates, specific cellular signals or signal perception pathways can trigger distinct cellular responses at different times and in different regions of the developing embryo. This allows a single pathway (and its component gene products) to function in multiple, separate developmental events (Gilbert, 1997
). As complex multicellular eukaryotes, plants, like animals, have only a limited molecular "toolkit" with which to carry out the numerous cellular processes required for development, such as imparting or perceiving positional information. Therefore, it perhaps is not surprising that there is overlap between the mechanisms operating during suspensor and cotyledon development in the Arabidopsis embryo.
|
FOOTNOTES |
|---|
2 Author for correspondence (e-mail: vernondm{at}whitman.edu
; fax: 509-527-5904).
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Akhundova, G. G., V. V. Shevchenko, and L. I. Grinikh. 1979 Twin plants of Arabidopsis thaliana induced by gamma irradiation. Soviet Genetics 15: 428–432
Barton, M. K. 1998 Cell type specification and self renewal in the vegetative shoot apical meristem. Current Opinion in Plant Biology 1: 37–42[CrossRef][ISI][Medline]
———, and R. S. Poethig. 1993 Formation of the shoot apical meristem in Arabidopsis thaliana: an analysis of development in the wild type and in the shoot meristemless mutant. Development 119: 823–831[Abstract]
Bennett, S. R. M., J. Alvarez, G. Bossinger, and D. R. Smyth. 1995 Morphogenesis in pinoid mutants of Arabidopsis thaliana. Plant Journal 8: 505–520
Chaudhury, A. M., S. Letham, S. Craig, and E. S. Dennis. 1993 Amp1— a mutant with high cytokinin levels and altered embryonic pattern, faster vegetative growth, constitutive photomorphogenesis and precocious flowering. Plant Journal 4: 907–916[CrossRef][ISI]
Chin-Atkins, A. N., S. Craig, C. H. Hocart, E. S. Dennis, and A. M. Chaudhury. 1996 Increased endogenous cytokinin in the Arabidopsis amp1 mutant corresponds with de-etiolation responses. Planta 198: 549–556[ISI]
Christensen, S. K., N. Dagenais, J. Chory, and D. Weigel. 2000 Regulation of auxin response by the protein kinase PINOID. Cell 100: 469–478[CrossRef][ISI][Medline]
Clark, S. E. 1997 Organ formation at the vegetative shoot meristem. Plant Cell 9: 1067–1076[CrossRef][ISI][Medline]
Conway, L. J., and R. S. Poethig. 1997 Mutations of Arabidopsis thaliana that transform leaves into cotyledons. Proceedings of the National Academy of Science, USA 94: 10209–10214
Cooke, T. J., R. H. Racusen, and J. D. Cohen. 1993 The role of auxin in plant embryogenesis. Plant Cell 5: 1494–1495[CrossRef][ISI][Medline]
Faure, J.-D., M. Jullien, and M. Caboche. 1994 Zea3: a pleiotropic mutation affecting cotyledon development, cytokinin resistance and carbon–nitrogen metabolism. Plant Journal 5: 481–491[ISI][Medline]
Fernandez, D. E. 1997 Developmental basis of homeosis in precociously germinating Brassica napus embryos: phase change at the shoot apex. Development 124: 1149–1157[Abstract]
Galweiler, L., C. Guan, A. Muller, E. Wisman, K. Mendgen, A. Yephremov, and K. Palme. 1998 Regulation of polar auxin transport by AtPIN1 in Arabidopsis vascular tissue. Science 282: 2226–2230
Gilbert, S. F. 1997 Developmental biology, 655–662. Sinauer Associates, Sunderland, Massachusetts, USA
Goldberg, R. B., G. de Pavia, and R. Yadegari. 1994 Plant embryogenesis: zygote to seed. Science 266: 605–614
Haccius, B. 1955 Experimentally-induced twinning in plants. Nature 176: 355–356[CrossRef][ISI]
———. 1963 Restitution in acidity-damaged plant embryos: regeneration or regulation? Phytomorphology 13: 107–115
Harada, J. J. 1999 Signaling in plant embryogenesis. Current Opinion in Plant Biology 2: 23–27[CrossRef][ISI][Medline]
Kaplan, D. R., and T. J. Cooke. 1997 Fundamental concepts in the embryogenesis of dicotyledons: a morphological interpretation of embryo mutants. Plant Cell 9: 1903–1919[CrossRef][ISI][Medline]
Kerstetter, R. A., and S. Hake. 1997 Shoot meristem formation in vegetative development. Plant Cell 9: 1001–1010[CrossRef][ISI][Medline]
Lenhard, M., and T. Laux. 1999 Shoot meristem formation and maintenance. Current Opinion in Plant Biology 2: 44–50[CrossRef][ISI][Medline]
Lotan, T., M.-A. Ohto, K. M. Yee, M. A. L. West, R. Lo, R. W. Kwong, K. Yamagishi, R. L. Fischer, R. B. Goldberg, and J. J. Harada. 1998 Arabidopsis LEAFY COTYLEDON1 is sufficient to induce embryo development in vegetative cells. Cell 93: 1195–1205[CrossRef][ISI][Medline]
Lui, C.-M., Z.-H. Xu, and N.-H. Chua. 1993 Auxin polar transport is essential for the establishment of bilateral symmetry during plant embryogenesis. Plant Cell 5: 621–630
Lynn, K., A. Fernandez, and M. K. Barton. 1999 The PINHEAD/Zwille gene acts pleiotropically in Arabidopsis development and has overlapping functions with the ARGONAUTE1 gene. Development 126: 469–479[Abstract]
Mansfield, S. G., and L. G. Briarty. 1991 Early embryogenesis in Arabidopsis thaliana. II. The developing embryo. Canadian Journal of Botany 69: 461–476
Mayer, U., R. A. Torres-Ruiz, T. Berleth, S. Misera, and G. Jurgens. 1991 Mutations affecting body organization in the Arabidopsis embryo. Nature 353: 402–407[CrossRef]
McConnell, J., and M. K. Barton. 1995 Affect of mutations in the PINHEAD gene of Arabidopsis on the formation of shoot apical meristems. Developmental Genetics 16: 358–366[CrossRef][ISI]
Meinke, D. W. 1992 A homeotic mutant of Arabidopsis with leafy cotyledons. Science 258: 1647–1650
———. 1995 Molecular genetics of plant embryogenesis. Annual Review Plant Physiology and Plant Molecular Biology 46: 369–394[CrossRef][ISI]
———, L. H. Franzmann, T. C. Nickle, and E. C. Yeung. 1994 Leafy cotyledon mutants of Arabidopsis. Plant Cell 6: 1049–1064[Abstract]
Moussian, B., H. Schoof, A. Haecker, G. Jurgens, and T. Laux. 1998 Role of the ZWILLE gene in the regulation of central shoot meristem cell fate during Arabidopsis embryogenesis. EMBO Journal 17: 1799–1809[CrossRef][ISI][Medline]
Okada, K., J. Ueda, M. K. Komaki, C. J. Bell, and Y. Shimura. 1991 Requirement of the auxin polar transport system in early stages of Arabidopsis floral bud formation. Plant Cell 3: 667–684
Schiavone, F. M., and T. J. Cooke. 1987 Unusual patterns of somatic embryogenesis in domesticated carrot: developmental effects of exogenous auxins and auxin transport inhibitors. Cell Differentiation 21: 53–62[CrossRef][ISI][Medline]
Schwartz, B. W., D. M. Vernon, and D. W. Meinke. 1997 Development of the suspensor: differentiation, communication, and programmed cell death during plant embryogenesis. In B. A. Larkins and I. K. Vasil [eds.], Cellular and molecular biology of plant seed development, vol. 2, 53–72. Kluwer Academic Press, Dordrecht, The Netherlands
———, E. C. Yeung, and D. W. Meinke. 1994 Disruption of morphogenesis and transformation of the suspensor in abnormal suspensor mutants of Arabidopsis. Development 120: 3235–3245[Abstract]
Taylor, C. B. 1997 It takes two: cuc mutations and shoot organ separation. Plant Cell 9: 837–839[CrossRef][ISI]
Torres-Ruiz, R. A., A. Lohner, and G. Jurgens. 1996 The GURKE gene is required for normal organization of the apical region in the Arabidopsis embryo. Plant Journal 10: 1005–1016[CrossRef][ISI][Medline]
Vernon, D. M., and D. W. Meinke. 1994 Embryogenic transformation of the suspensor in twin, a polyembryonic mutant of Arabidopsis. Developmental Biology 165: 566–573[CrossRef][ISI][Medline]
———, and ———. 1995 Late-embryo mutants of Arabidopsis. Developmental Genetics 16: 311–320[CrossRef][ISI]
West, M. A. L, and J. J. Harada. 1993 Embryogenesis in higher plants: an overview. Plant Cell 5: 1361–1369
