- © 1998 American Society of Plant Physiologists
Abstract
Cell elongation is a developmental process that is regulated by light and phytohormones and is of critical importance for plant growth. Mutants defective in their response to light and various hormones are often dwarfs. The dwarfed phenotype results because of a failure in normal cell elongation. Little is known, however, about the basis of dwarfism as a common element in these diverse signaling pathways and the nature of the cellular functions responsible for cell elongation. Here, we describe an Arabidopsis mutant, dwarf4 (dwf4), whose phenotype can be rescued with exogenously supplied brassinolide. dwf4 mutants display features of light-regulatory mutants, but the dwarfed phenotype is entirely and specifically brassinosteroid dependent; no other hormone can rescue dwf4 to a wild-type phenotype. Therefore, an intact brassinosteroid system is an absolute requirement for cell elongation.
INTRODUCTION
Plant growth is accomplished by orderly cell division and tightly regulated cell expansion. In plants, the contribution of cell expansion to growth is of much greater significance than in most other organisms; all plant organs owe their final size to a period of significant cell elongation, which usually follows active cell division. In Arabidopsis, cell elongation is largely responsible for hypocotyl growth in germinating seedlings and extension of inflorescences (bolting) at the end of vegetative growth. Coordinate control of plant growth is regulated by both external stimuli and internal mechanisms. Of the external signals, the most obvious is light (Deng, 1994). Light inhibits hypocotyl elongation and promotes cotyledon expansion and leaf development in seedlings, and photoperiod is crucial for flower initiation in a large number of species.
The internal components of plant signaling are generally mediated by chemical growth regulators (phytohormones; reviewed in Klee and Estelle, 1991). Gibberellic acid (GA) and cytokinins promote flowering; in addition, GA stimulates stem elongation, whereas cytokinins have the opposite effect, reducing apical dominance by stimulating increased axillary shoot formation. Conversely, auxins promote apical dominance and stimulate elongation by a process postulated to require acidification of the cell wall by a K+-dependent H+-pumping ATPase (Rayle and Cleland, 1977). A less well understood class of plant growth substances is the brassinosteroids (BRs; Mitchell et al., 1970; Grove et al., 1979; Mandava, 1988). Although several such compounds have been identified and are known to effect elongation of cells in various plant tissues, their biosynthesis, regulation, and mechanism of action have only recently begun to be elucidated (reviewed in Clouse, 1996; Fujioka and Sakurai, 1997).
In Arabidopsis, a number of mutations have been identified that affect either light-dependent or hormone signaling pathways and whose pleiotropic phenotypes include defects in cell elongation. The majority of these mutants also have other alterations in their phenotypes. The range of phenotypes in the deetiolated (det) and constitutive photomorphogenic (cop) light-regulatory mutants is broad. Mutations in DET1, COP1, COP8, COP9, COP10, and COP11 result in constitutive derepression of substantial portions of the photomorphogenic program (Chory et al., 1989b; Deng and Quail, 1992; Wei and Deng, 1992; Wei et al., 1994), whereas mutations in COP4 seem to affect only morphology and gene expression (Hou et al., 1993). The only invariant phenotype in this class of light-regulatory mutants is a substantial reduction in height in both light and darkness.
A number of hormone mutations also lead to reduced plant height. At least five GA mutants have been described as being reduced in stature (Koornneef and Van der Veen, 1980). GA biosynthetic mutants may also have no or defective flower development and are marked by an absence of viable pollen. Reduced levels of endogenous gibberellins are also a characteristic (Barendse et al., 1986; Talon et al., 1990), and their phenotype can be nearly restored to that of the wild type by the addition of exogenous GA. Another hormone mutation, auxin resistant2 (axr2), results in plants with a dwarf phenotype both in the light and in darkness as well as increased resistance to high levels of auxin, ethylene, and abscisic acid (Timpte et al., 1992). An interesting relationship exists between light regulation and cytokinin levels. Arabidopsis seedlings grown in the dark in the presence of cytokinins have open cotyledons, initiate chloroplast differentiation and leaf development, and activate transcription from the chlorophyll a/b binding protein gene (CAB) promoter. Importantly, they also display a cytokinin dose-dependent dwarf phenotype. Moreover, det1 and det2 have a decreased requirement for cytokinins in tissue culture and appear to be saturated for a cytokinin-dependent delay in senescence (Chory et al., 1994).
The most recently discovered class of plant growth substances, the BRs, has been to date the least studied; however, rapid progress toward understanding BR biosynthesis and regulation is now being made (Yokota, 1997). Three Arabidopsis mutants with a dwarf phenotype, dwarf1 (dwf1; Kauschmann et al., 1996; B.P. Dilkes and K.A. Feldmann, unpublished data), constitutive photomorphogenesis and dwarfism (cpd; Szekeres et al., 1996), and det2 (Li et al., 1996), have been shown to be defective in steroid biosynthesis. In addition, BRs have been shown to cause elongation in a variety of mutants with a dwarfed phenotype (Szekeres et al., 1996). Interestingly, many of these mutants were previously identified as being defective in light-mediated (cop, det, and fusca [fus; another group of mutants with some members perturbed in light-regulated growth]) or hormone-mediated (axr2) regulation. Evidently, many of the signaling pathways regulated by light and phytohormones result in defective cell elongation.
Currently, little is known about the downstream events that occur in response to these signals and thereby directly control cell size. This is because the biochemical and cell biological processes involved have thus far been difficult to address. In addition, there is little information about the integration of regulatory signals converging at the cell from different signaling pathways and the ways they are coordinately controlled. In particular, the interaction of light and hormones in the control of cell elongation is not clear. Additional mutants in each of these pathways may help to further unravel these interactions.
In this work, we describe dwf4, an Arabidopsis mutant in which the major defect of the mutation is a block in cell elongation. This mutant also shares some characteristics of light-grown plants when grown in the dark; we present evidence suggesting that this latter phenotype is a consequence of defective cell elongation. Like det2 and cpd, dwf4 is rescued to a wild-type phenotype by exogenous brassinolide (BL). Because the defect in cell elongation blocks both the response to darkness and the effects of exogenously supplied GA and auxin, our results suggest that steroid signaling is necessary for normal plant growth and development.
RESULTS
Isolation of dwf4
As formally defined, a plant with a dwarf phenotype is one that has a short, robust stem and short, dark green leaves. dwf1 is the first dwarf isolated via T-DNA mutagenesis (Feldmann and Marks, 1987) and has been described in detail (Feldmann et al., 1989; referred to as diminuto in Takahashi et al. [1995] and Szekeres et al. [1996]). In a screen of 14,000 transformed families (T3), a number of additional dwarfs were identified. Two independent lines were found that segregated for a similar phenotype: both were shorter than dwf1, but their rosette diameter was comparable to that mutant. These dwarfs were also essentially infertile. The most striking aspect of the morphology of these mutants is their similarity to det2 (Chory et al., 1991). For this reason, further analysis was conducted with these lines. After being found to be allelic to each other, both were designated as dwf4. The dwf4 mutation was subsequently shown to be inherited as a monogenic, recessive Mendelian trait that, in dwf4-1, cosegregates with the dominant kanamycin resistance marker contained in the T-DNA (data not shown), suggesting that the mutation in this line may be a disrupted, tagged allele. dwf4-2 also contains a single kanamycin resistance marker, but it failed to cosegregate with the dwarf phenotype (data not shown). Two additional alleles (dwf4-3 and dwf4-4) were identified among dwarf mutants obtained from the Nottingham Arabidopsis Resource Centre (Nottingham, UK; N365 and N374, respectively; Figure 1). Unless otherwise indicated, all experiments presented below were performed with dwf4-1.
The dwf4 Phenotype
As shown in Figure 1, dwf4 mutants are significantly smaller than the wild type and are dark green in color. They have short, rounded leaves. Again, the dwf4 phenotype is reminiscent of the light-regulatory mutant det2 (Chory et al., 1991); however, complementation analysis has shown that the two mutations are not allelic, with the dwf4 mutation mapping to the lower arm of chromosome 3 and det2 mapping to chromosome 2 (Chory et al., 1991). The results presented in Table 1 show that soil-grown dwf4 plants attained a height of <3 cm at 5 weeks, whereas wild-type plants grew to >25 cm. Moreover, individual organs, such as leaves, were invariably shorter in dwarf plants. dwf4 siliques were also markedly shorter than those of the wild type and were infertile. The loss of fertility of dwf4 was due to the reduced length of the stamen filaments relative to the gynoecium, which resulted in mature pollen deposition on the ovary wall rather than on the stigmatic surface. Hand pollination of dwf4 flowers with either mutant or wild-type pollen resulted in good seed set without significantly changing the size of the siliques.
Morphological Comparison of dwf4 Alleles at 5 Weeks after Germination.
The wild type is >25 cm tall at this age (data not shown). dwf4-1 (left) and dwf4-2 (second from left) are in the Wassilewskija background; dwf4-3 (second from right) and dwf4-4 (right) are in the Enkheim background. Bar = 1 cm.
Another feature of dwf4 plants is a reduction in apical dominance, as is evident by the threefold increase in the number of inflorescences at 5 weeks of age (Table 1). Mutants also have twice the number of rosette leaves, which may be explained by a prolonged vegetative phase in the dwf4 plants. Development of flowers on the primary inflorescence is delayed by ~4 days in dwf4, but the flowering phase is significantly longer in the mutant, with senescence of the last flower occurring at ~98 days compared with ~57 days for the wild type. One result of this delay in senescence is that dwf4 plants contain almost three times the number of siliques as does the wild type (Table 1).
The Development of Wild-Type and dwf4-1 Plants
The reduced stature observed in soil-grown dwf4 was also observed in hypocotyls of agar-grown seedlings. Measurements of hypocotyl length over time, shown in Figure 2, indicated that not only are dwf4 seedlings shorter than wild-type seedlings immediately after germination but also that the rate of growth was retarded in the mutants. In addition, dwf4 hypocotyls reached their terminal length in <5 days, whereas wild-type seedlings continued to grow.
The Growth Defect of dwf4 Is Due to a Reduction in Cell Length
Both the short stature and the reduced growth rate of dwf4 could be due to a defect in cell division or cell elongation or both. To distinguish between these possibilities, we analyzed sections from 7-day-old hypocotyls and 5-week-old inflorescence stems, by light microscopy, as described in Methods. To minimize variations due to the developmental stage of the sample, we always took the stem sections from the fourth internode. Figures 3C and 3D show that the average cell size in dwf4 is significantly smaller than in wild-type plants (Figures 3A and 3B), whereas no differences were detected in the number of cells along the length of either organ between the wild type and dwf4 (data not shown). Therefore, the short stature and reduced organ length of dwf4 are largely or exclusively due to a failure of individual cells to elongate. No differences were observed in the number of cell layers contained in the wild type and dwf4.
The small size of dwf4 cells offers a possible explanation for the dark green color of the mutant plants. Chlorophyll measurements were taken, and leaf mesophyll protoplasts were prepared, stained, and measured to visualize and count chloroplasts, as described in Methods. Although there were no significant differences in total chlorophyll content, the chlorophyll a/b ratio, or the absorption spectra between wild-type plants and mutants (data not shown), the mean plane area (the apparent two-dimensional surface area of mounted cells) of dwf4 leaf mesophyll protoplasts was 376 mm2, whereas that of wild-type protoplasts was 599 mm2. The two-dimensional comparison of plane area represents a dramatic reduction in volume for dwf4 cells. However, the number of chloroplasts per cell was only slightly lower: the mean number of chloroplasts per cell was 40 for dwf4 and 44 for the wild type. Therefore, dwf4 cells contain a greatly increased number of chloroplasts per unit cell volume. As a consequence, the chloroplasts are brought closer to each other, making the color of the leaves appear darker. Chloroplast size was the same in both lines (data not shown).
Comparison of Wild-Type and dwf4 Hypocotyl Growth Rates.
Circles indicate wild type; squares indicate dwf4. Each data point represents the average of 10 seedlings.
dwf4 Is Specifically Rescued by BRs
The reduced length of cells in dwf4 hypocotyls and inflorescence stems is indicative of a failure of these cells to elongate during development. A variety of endogenous and environmental signals is responsible for stimulating elongation in plants; therefore, a series of experiments was performed to determine whether dwf4 is affected in a specific signaling pathway or is blocked in elongation as a response to various signals.
Of the endogenous (hormonal) signals that might be deficient in dwarf plants, an obvious candidate is GA, because gibberellin-deficient mutants are shorter in stature than are the wild-type plants (Koornneef and Van der Veen, 1980). Our results, however, indicate that dwf4 is not defective in the synthesis of gibberellins. When germinated on 10−5 M GA, wild-type seedlings demonstrated an elongation response (Figure 4), whereas dwf4 seedlings responded minimally, if at all. At 10−4 M GA, wild-type seedlings elongated slightly more than at 10−5 M (data not shown), but the dwf4 seedlings were essentially saturated for elongation at 10−5 M GA. Similar results were obtained when soil-grown plants were sprayed with 1 mM GA once inflorescences first became visible: dwf4 inflorescence stems elongated by only 28% above the untreated controls, whereas those of the wild type elongated by 45% above the untreated controls (data not shown). Mutants that owe their reduced stature to decreased levels of endogenous gibberellins can be fully rescued by added hormone (Koornneef and Van der Veen, 1980; Talon et al., 1990). In addition, dwf4 seeds germinate in the absence of exogenously supplied GA. Our results therefore suggest that dwf4 is not deficient in endogenous GA. A corollary conclusion from this experiment is the demonstration that dwf4 is capable of detecting GA; that is, it is not likely to be affected in signal perception but rather is defective in the extent to which it can respond to this signal.
Longitudinal Sections of Wild-Type and dwf4 Tissues.
(A) Hypocotyl from a 7-day-old wild-type plant. Average cell length is 92.7 μm.
(B) Stem from a 5-week-old wild-type plant. Average cell length is 79.2 μm.
(C) Hypocotyl from a 7-day-old dwf4 plant. Average cell length is 32.2 μm.
(D) Stem from a 5-week-old dwf4 plant. Average cell length is 15.8 μm.
(A) and (C) are at the same magnification; (B) and (D) are at the same magnification. Bars in (A) and (B) = 100 μm.
Response to Cell Elongation Signals.
All measurements were performed as described in Methods. BL measurements were performed with dwf4-3 and the corresponding wild-type control, Enkheim. Open bars indicate the wild type; filled bars indicate dwf4. Lines above the bars represent one standard deviation. Light, light-grown control; Dark, dark-grown control; hy2, DWF4 and dwf4 plants in a hy2 background; GA, 10−5 M GA; 2,4-D, 10−8 M 2,4-D; −BR, liquid-grown controls; +BR, liquid-grown controls with 10−6 M BL.
Auxin can also stimulate cell elongation. This effect is especially visible in young seedlings (Klee and Estelle, 1991). The response of wild-type and dwf4 plants to auxin was tested by growing seedlings for 10 days on vertically oriented plates containing various concentrations of the synthetic auxin 2,4-D. At all concentrations assayed, inhibition of root growth was evident (data not shown). Figure 4 shows that at 10−8 M 2,4-D, hypocotyl elongation in wild-type and dwf4 seedlings was similar to that of the controls. Higher concentrations of auxin were inhibitory for both wild-type and dwf4 seedlings, and lower concentrations had no effect (data not shown). In view of the inhibition of root growth, it is clear that dwf4 is not auxin resistant (data not shown); rather, its elongation response is compromised.
As mentioned above, the most obvious exogenous signal for plants is light. Therefore, to investigate whether light-regulated cell elongation is altered in dwf4, wild-type and dwf4 seedlings were grown in the dark, as described in Methods. Figure 4 shows that as expected, wild-type seedlings displayed hypocotyl elongation typical of etiolated growth. By contrast, dark-grown dwf4 seedlings were only slightly longer than those grown in the light. To assess the relationship between the dwf4 phenotype and light sensing by dwf4, the mutation was crossed into a mutant defective in the HY2 gene. All hy mutants share the common phenotype of an elongated hypocotyl that mimics part of the etiolation response in the light. Specifically, hy2 is deficient in active phytochrome because chromophore biosynthesis does not take place (Chory et al., 1989a). Figure 4 shows that dwf4 hy2 double mutants displayed a dwarfed phenotype indistinguishable from that of dwf4 HY2 (light-grown control); therefore, the elongation block due to the dwf4 mutation is epistatic to a defect in phytochrome activity.
In the course of our studies, we prepared a genomic library from dwf4-1, from which we isolated a clone in which a fragment of T-DNA interrupts a gene encoding a putative cytochrome P450 steroid hydroxylase (data not shown). Because BRs have been shown to elicit elongation in Arabidopsis (Clouse et al., 1993) and because BR-deficient mutants have been recently described (Kauschmann et al., 1996; Li et al., 1996; Szekeres et al., 1996), we tested the effect of BL on Arabidopsis seedlings by germinating seeds in liquid medium containing different amounts of BL. As shown in Figure 4, the dwf4 hypocotyls were restored to wild-type height by 10−6 M BL. This, together with our identification of a disrupted gene encoding a putative BR biosynthetic enzyme, strongly suggests that the phenotype of dwf4 is specifically due to a defect in BR biosynthesis (see Choe et al., 1998).
The Elongation Defect of dwf4 Leads to a Light-Regulatory Phenotype
The BR-deficient mutant det2 was originally identified as defective in regulation by light (Chory et al., 1991). Given the similarity of det2 and dwf4 phenotypes and functions and in view of the observation that dwf4 is epistatic to hy2, one can predict that the etiolation response, which includes significant hypocotyl elongation, would not be normal in dwf4. To assess to what extent the etiolation response is affected by BR-dependent cell elongation, we grew dwf4 and wild-type plants on agar under continuous light or in complete darkness, as described in Methods. Figure 5A shows that after 7 days of growth in the light, wild-type seedlings displayed open and expanded cotyledons as well as emerging leaf buds. In contrast, the overall appearance of light-grown dwf4 seedlings (Figure 5B) is strikingly similar to that of det2 (Chory et al., 1991). dwf4 hypocotyls are very short, and the cotyledons are smaller than those of the wild type, displaying significant epinastic growth. As expected, dark-grown wild-type seedlings have a typical etiolated appearance, with a highly elongated hypocotyl and closed, unexpanded cotyledons (Figure 5C). However, dwf4 hypocotyls failed to elongate (Figure 5D). That the dwf4 mutation can abolish the elongation component of the etiolation response is in agreement with the notion that the block in cell elongation in dwf4 is specifically a BR-dependent process.
Aberrant Skotomorphogenesis.
(A) Seven-day-old light-grown wild-type plant.
(B) Seven-day-old light-grown dwf4 plant.
(C) Seven-day-old dark-grown wild-type plant.
(D) Seven-day-old dark-grown dwf4 plant.
(E) Shoot apex of 7-day-old dark-grown dwf4 plant. dwf4 develops true leaves in the dark, as shown by the presence of emerging trichomes.
Bar in (D) = 1 mm for (A) to (D); bar in (E) = 100 μm.
In addition to short hypocotyls, dark-grown dwf4 seedlings displayed partially open cotyledons and leaf primordia, with up to four leaf buds clearly visible (Figures 5D and 5E). This has not been observed with the wild type, although it occurs with certain light-regulatory mutants (Chory et al., 1989b; Deng et al., 1991; Wei and Deng, 1992). dwf4 leaf development can continue in the darkness for several weeks, resulting in significant expansion of rosette leaves (data not shown). These results indicate that dwf4 plants can initiate what is normally a photomorphogenic pathway in the absence of light. Although this is often diagnostic of a light-regulatory mutant, wild-type Arabidopsis can perform leaf development and even flowering in complete darkness when grown in liquid culture (Araki and Komeda, 1993).
The cause for this dark-flowering effect is not understood; therefore, the possibility exists that leaf development in dark-grown dwf4 is related to dark flowering and not to a light-regulatory defect. For example, perhaps the proximity of the dwf4 shoot apical meristem to the surface of the agar, due to the shortness of the hypocotyls, mimics some effect of submerged culture, such as a high water potential or a high concentration of some nutrient. To test this possibility, wild-type seedlings were grown in complete darkness for 6 weeks in vertically oriented dishes to maximize contact between the seedling and the medium. Wild-type seedlings grown in this fashion displayed open cotyledons and underwent at least partial leaf development (data not shown). In fact, all wild-type seedlings grown along the surface of the agar showed development of an inflorescence with at least one cauline leaf and a terminal flower bud. We conclude, therefore, that the appearance of leaves in dark-grown dwf4 may be due simply to its short size and the culture conditions.
A number of light-regulatory mutants have been described that undergo photomorphogenesis in the dark at the cellular level. In mutants such as cop1, cop8, cop9, cop10, and cop11, stomata undergo photomorphogenic maturation (Deng and Quail, 1992; Wei and Deng, 1992; Wei et al., 1994); of these, cop1 and cop9 as well as det1 (Chory et al., 1989b) also initiate differentiation of plastids into chloroplasts. To determine whether dwf4 plants undergo photomorphogenic cellular differentiation in the dark, we analyzed cotyledons from light- and dark-grown plants by transmission and scanning electron microscopy; the results are shown in Figure 6. Analysis of plastids in thin sections from 7-day-old dark-grown seedlings showed no difference between the wild type and dwf4. Both lines contained normal chloroplasts when grown in the light (Figures 6A and 6B), whereas dark-grown seedlings contained etioplasts, with their characteristic prolamellar body and no significant organization of thylakoids (Figures 6C and 6D). Analysis of stomatal structures on the underside of cotyledons from 7-day-old seedlings indicates that stomatal development was not completed in the dark, because the stomatal opening was occluded in both lines (Figures 6E and 6F). The majority of light-regulatory mutants analyzed to date displayed light-grown morphology in the dark without concomitant chloroplast or stomatal development. As in these mutants, therefore, the dwf4 mutation uncouples the developmental pathway of seedling morphology from that of light-regulated cellular differentiation.
Plastid and Cell Differentiation in Cotyledons of 7-Day-Old Light-Grown and Dark-Grown Wild-Type and dwf4 Seedlings.
(A) Plastids from a light-grown wild-type plant.
(B) Plastids from a light-grown dwf4 plant.
(C) Plastids from a dark-grown wild-type plant.
(D) Plastids from a dark-grown dwf4 plant.
(E) Stomatal guard cells from a dark-grown wild-type plant.
(F) Stomatal guard cells from a dark-grown dwf4 plant.
Bar in (B) = 1 μm for (A) to (D); bars in (E) and (F) = 10 μm.
An additional feature of many light-regulatory mutants is that photomorphogenesis in the dark is accompanied by expression of genes that normally are light induced (Chory et al., 1989b, 1991; Deng et al., 1991; Wei and Deng, 1992; Hou et al., 1993; Wei et al., 1994). To assess whether dwf4 is able to induce light-regulated transcripts in the dark, we compared the activity of a CAB promoter fused to the Escherichia coli gene uidA, encoding β-glucuronidase (GUS), in light- and dark-grown dwarf and wild-type plants. The CAB–uidA fusion in pOCA107-2 (Li et al., 1994) was crossed into dwf4, and F2 dwarf and wild-type plants were grown in the dark or light for 12 days, followed by determination of GUS activity by fluorometry (Gallagher, 1992). The results (data not shown) demonstrate that when grown in the light, both wild-type and dwf4 seedlings contain GUS activity, which is significantly reduced in both lines when grown in the dark. Moreover, dark-grown dwf4 seedlings display no GUS activity above the background present in dark-grown wild-type plants. The absence of light-induced gene expression in the dark is a distinguishing feature of certain cop and det mutants, such as cop2, cop3, and det3. Because we have shown that the defect in cell elongation of dwf4 is specifically rescued by BRs, even in the presence of light, we conclude that this is not a light-regulatory mutant. That its phenotype is partially deetiolated or constitutively photomorphogenic is a secondary effect of its reduced stature and the growth conditions.
DISCUSSION
We report the isolation and characterization of four alleles of dwf4. dwf4 plants are reduced in stature but robust and healthy overall (Figure 1). The phenotypic effects of the dwf4 mutation can be summarized as follows: (1) cell elongation appears to be the primary defect in the mutant; (2) growth is affected in all organs examined; and (3) a deetiolated/constitutively photomorphogenic morphology may be the result of defective cell elongation. The arrest of cell elongation in dwf4 can be rescued by exogenous application of BL.
The dwf4 Phenotype
Morphologically, the dwf4 phenotype can be described as being due to both primary and secondary effects of reduced cell elongation. The primary effect is simply a reduction in the length of individual organs exclusively along their normal growth axis; that is, organ width is not reduced (Table 1). The secondary effects of reduced cell elongation are themselves due to the reduction in organ length. The dark green color of the leaves, for example, may be due exclusively to the existence of a wild-type number of chloroplasts in a significantly smaller cell. Similarly, the sterility of mutants is a consequence of the shortness of the stamens, which fail to deposit their pollen on the stigmatic surface. In addition to the morphological alterations of dwf4, mutants display delayed development, the first sign of which occurs at flowering (Table 1). Because rosette leaves are produced continuously during vegetative development, delayed flowering results in dwf4 rosettes having almost twice the number of leaves observed in the wild type.
A Block in Cell Elongation
Experimental evidence for a block in cell elongation in dwf4 is first seen from the measurements shown in Figure 2. The rate of growth was significantly reduced in agar-grown dwf4 seedlings, which ceased to grow when their hypocotyl length was <20% of the final wild-type length. Because all of the cells in a hypocotyl before the initiation of leaf development are present in the embryo, the initial growth of seedlings is due exclusively to cell expansion, which therefore must be reduced in dwf4. A similar situation applies to soil-grown plants. Five weeks after germination, well after plants had bolted, dwf4 plants were shorter than wild-type plants (Figure 1 and Table 1). Although the mutants continued growing for several weeks more than did the wild type, they remained shorter through senescence. That cell elongation is the direct cause of this decreased growth is shown by measurements of cell length both in 7-day-old hypocotyls (Figures 3A and 3C) and in 5-week-old stems (Figures 3B and 3D). Not only is the reduction in cell length in good agreement with the reduction in organ length, but insofar as could be determined, there is no difference in the number of cells between dwf4 and wild-type plants.
Organ growth by cell elongation in plants occurs as part of normal development in response to a variety of input signals. Mutants that are defective in these signaling pathways invariably fail to elongate normally in response to the appropriate stimuli. A mutant with a block at a step that is common to several individual pathways would therefore be expected to have defective responses to all of the corresponding signals. dwf4 appears to be such a mutant. Figure 4 shows that elongation induced by the hy2 mutation is blocked in a dwf4 hy2 double mutant. Not surprisingly, in view of this result, dwf4 also failed to display hypocotyl elongation as a response to growth in complete darkness (Figure 5). In addition, dwf4 was capable of perceiving GA, but its response was severely compromised. This mutant could also respond to the inhibitory effects of auxin but was incapable of auxin-stimulated elongation. It was only exogenous BL that fully restored wild-type length to dwf4 hypocotyls (Choe et al., 1998).
Because dwf4 failed to respond to at least three independent signaling pathways but responded fully to only one, the most likely explanation for the dwarf phenotype is therefore that a fully functional BR system is required for a full response to GA, auxin, and deetiolation. From the perspective of cellular economy, it may be advantageous that the downstream elements involved in cell elongation are shared among at least some of the signaling pathways that evoke this response. The interaction of various pathways at a common step provides the plant with a potential point for the integration of signals produced by diverse independent stimuli. Our results indicate that BRs act at this downstream step.
Abnormal Skotomorphogenesis as a Consequence of the Dwarf Growth Habit
When dwf4 is grown in the light, its morphology is similar to that of various cop and det mutants, with multiple short stems, short rounded leaves, loss of fertility due to reduced stamen length, and delayed development (see Figures 1 and 4). Dark-grown dwf4 seedlings possess short hypocotyls, open cotyledons, and developing leaves (Figure 5). Therefore, it is tempting to speculate that this mutant may be defective in the control of light-regulated processes. On the other hand, because a dark-flowering phenotype has been demonstrated for liquid-grown Arabidopsis (Araki and Komeda, 1993), and given that agar medium is mostly water, it is especially significant that it is the dwarf seedlings, whose apical meristems are very close to the agar surface, that display a light-grown phenotype in the dark. Furthermore, because wild-type seedlings grown along the surface of the agar reproduce the dark-flowering phenotype, it is possible that the apparent light-regulatory defect of dwarf seedlings is a dark-flowering response. This possibility is strengthened by the observation (data not shown) that wild-type seedlings (ecotype Wassilewskija [Ws-2]) grown in the dark on horizontally oriented plates occasionally bend down and touch the agar surface, and these seedlings invariably produce leaves. In addition, of the eight DWF loci identified in this laboratory, only the shortest mutants displayed open cotyledons and leaf bud development; in the case of dwf1 (Feldmann et al., 1989), this aberrant skotomorphogenesis is confined to the most severely affected alleles (B. Dilkes and K.A. Feldmann, data not shown).
In addition to the presence of a short hypocotyl and at least partially open cotyledons in the dark, cop1 (Deng and Quail, 1992), det1 (Chory et al., 1989b), and det3 (Cabrera y Poch et al., 1993) have been shown to initiate leaf formation in the dark. In mutants such as cop1, cop8, cop9, cop10, and cop11, stomata undergo photomorphogenic maturation (Deng and Quail, 1992; Wei and Deng, 1992; Wei et al., 1994); of these, cop1 and cop9 as well as det1 (Chory et al., 1989b) also initiate differentiation of plastids into chloroplasts. dwf4 displayed, in addition to a light-grown dwarf phenotype, a dark-growth phenotype of short hypocotyls, open cotyledons, and developing leaves; however, in contrast with the light-regulatory defect seen with whole plants, the cellular differentiation phenotype was unaffected (Figure 6). In dark-grown dwarf seedlings, stomata did not complete their development, and differentiation of chloroplasts was not observed. The absence of a cellular light-regulatory phenotype in dwf4 is similar to that of a number of photomorphogenic mutants, such as det2, det3, cop2, cop3, and cop4 (Chory et al., 1991; Cabrera y Poch et al., 1993; Hou et al., 1993).
In view of the dark-flowering phenotype on agar and the absence of a light-regulatory defect in differentiating cells, we conclude that at least in the case of dwf4, aberrant skotomorphogenesis may be a consequence of a dwarf growth habit rather than dwarfism being part of a defect in the control of light-regulated processes. This effect may also explain the light-regulatory phenotype found in other mutants with severely reduced height, such as axr2 (Timpte et al., 1992; see below), and strong alleles of dwf1, both of which are also rescued by exogenous BRs (Szekeres et al., 1996).
BRs Interact with Multiple Signaling Pathways
Because a dwarflike phenotype is the only invariant trait of all known light-regulatory mutants, it follows that the primary defect in dwf4 mutants is identical to the minimal cop/det/fus phenotype. It is possible, then, that the dwarf phenotype in all of the above mutants is due to defective cell elongation, which therefore may not be exclusively related to regulation by light. GA-deficient mutants also display a variable set of traits with different degrees of severity. But as with the photomorphogenic mutants, the similarity in phenotype between dwf4 and the ga mutants is confined to reduced cell elongation. The evidence is compelling for BL involvement at or near a downstream control point where multiple signaling pathways interact. First, our results (Figure 4) show that BL is required for cell elongation as a response to darkness as well as GA and auxin. In addition, previous studies (Kauschmann et al., 1996; Li et al., 1996; Szekeres et al., 1996) and this work show that BR can compensate for the cell elongation defect of mutants as diverse as det2, cpd, dwf4, det1, cop1, and dwf1. This places BRs downstream of all the cellular functions affected in these mutants. Finally, at least one of the BR biosynthetic genes has been shown to be modulated by light, cytokinins, and the carbon source (Szekeres et al., 1996).
Mutations in axr2 result in a dwarf growth habit and a dark-grown phenotype with short hypocotyl and open cotyledons (Timpte et al., 1992). In addition, axr2 mutants are resistant to auxin, ethylene, and abscisic acid and have defective root and shoot gravitropism. The dwarf phenotype in axr2 mutants has been shown to be due to reduced cell elongation and is rescued by BL (Szekeres et al., 1996). This suggests that at least one of the multiple hormone signaling pathways affected in axr2 involves a BR-dependent step. Mutations at another locus, acaulis1, also have a significant reduction in cell elongation, but the defect is confined to inflorescence stems and leaves (Tsukaya et al., 1993). Flowers are fully fertile and mature into normal-sized siliques with normal seed set. There is no change in hypocotyl length. If BRs are directly involved in this apparently organ-specific signaling pathway, it may be due to organ-specific responsiveness to individual BR species.
With regard to the mechanism of action of BRs, at the moment one can only speculate that the target may be a component of the cell expansion machinery. Perhaps steroid signaling initiates a series of events leading to cell wall loosening. Future studies of cell wall composition together with analysis of genetic interactions between BRs and cell wall mutants (Reiter et al., 1993) may provide information about the role of these compounds in cell elongation.
METHODS
Plant Growth Conditions
The dwf4-1 and dwf4-2 mutations are in the Arabidopsis thaliana ecotype Wassilewskija (Ws-2) background; the dwf4-3 and dwf4-4 mutations are in the Enkheim (En-2) background. The conditions used for plant growth were essentially as described previously (Feldmann, 1991; Forsthoefel et al., 1992), except that agar-solidified medium contained 0.5% sucrose. Seedlings up to 2 weeks of age (6 weeks of age for dark-growth experiments) were grown on agar-solidified medium; older plants were grown in potting soil. Germination of seeds for dark growth experiments was induced by overnight exposure of the seeds to light immediately after removing the plates from incubation at 4°C.
Analytical Methods
Protoplasts were obtained by overnight incubation of sliced leaves in 0.1% cellulysin, 0.1% driselase, 0.1% macerase (Calbiochem, San Diego, CA) in 125 mM Mes, pH 5.8, 0.5 M mannitol, and 7 mM CaCl2 (Galbraith et al., 1992). Immediately before observation, chloroplasts were stained with a solution of 1.5% KI and 1% I2. Measurements were performed as described for tissue sections (see below), and plane areas were calculated according to the formula A = πr2.
Chlorophyll determinations were performed from 2-week-old soil-grown plants. Green tissue was weighed, frozen in liquid nitrogen, and extracted in dim light with 80% acetone in the presence of a mixture of equal parts sand, NaHCO3, and Na2SO4. After brief centrifugation, the supernatant was collected and the extraction was repeated twice, pooling the supernatants from each sample. Chlorophylls a and b were measured spectrophotometrically, as described in Chory et al. (1991).
Growth Signal Response Measurements
Gibberellic acid (GA) response was assayed on plants grown individually in 5.7-cm pots. Once inflorescences reached 1 to 2 mm in height, they were sprayed weekly with 1 mM GA3 (Sigma). Control plants were sprayed with water. One week after the third spraying, plants were collected, and the length of the main stem was measured between the top of the rosette and the base of the most distal pedicel; 13 to 18 plants of each line were measured per treatment. Auxin response was tested by growing seedlings for 10 days under 16 hr of light on vertically oriented agar plates containing various concentrations of 2,4-D (Gibco, Grand Island, NY). Genetic interaction with the hy2 mutation was tested by growing seedlings under continuous light for 7 days. Brassinolide (BL) response was determined in liquid culture, as described by Clouse et al. (1993), except that three or four seedlings were grown in each well of a 24-well culture plate for 7 days. Measurements were taken for 10 to 20 seedlings for each genotype and condition, under a dissection microscope fitted with an ocular micrometer.
Microscopy
Tissues were fixed in 2% glutaraldehyde and 0.05 M sodium cacodylate, pH 6.9, for 2 hr at room temperature or overnight at 4°C, followed by three washes in buffer. For light microscopy, 1% safranin was included in the first wash, and embedding was performed in Paraplast Plus (Oxford Labware, St. Louis, MO). Ten-millimeter sections from five individual plants per line were analyzed and photographed, and cell measurements were taken using a ruler on 5 × 7 inch prints. A print of a hemocytometer grid at the same final magnification was used for calibration. At least 25 cells were measured per sample, with a minimum of 150 cells per line. For electron microscopy, the tissues were treated after fixation with 1% tannic acid in buffer for 30 min, washed three times, and postfixed in 1% OsO4 in buffer for 2 hr, followed by five washes and dehydration through an ethanol series. Samples for transmission electron microscopy were embedded in Spurr's resin. Sections (90 nm) were stained with saturated uranyl acetate followed by Reynolds's lead citrate (Reynolds, 1963) and examined in a JEOL (Tokyo, Japan) 100-CX instrument. For scanning electron microscopy, samples were transferred to freon 113, critical point dried, and sputter-coated with 30 to 50 nm of gold. Analysis was performed in a microscope (ISI model DS130; Topcon, Inc., Paramus, NJ) with an accelerating voltage of 15 kV.
ACKNOWLEDGMENTS
Electron microscopy was performed at the Electron Microscope Facility, Division of Biotechnology, Arizona Research Laboratories, University of Arizona, where Dave Bentley and Beth Huey provided essential suggestions and assistance. Access to the light microscope was kindly provided by Martha Hawes. We thank Joanne Chory for det2 and pOCA107-2 seeds; Shozo Fujioka for BL; Brian Dilkes for performing the BL experiment; Steve Clouse, Thomas Altmann, and Barbara Fishel for invaluable advice; and Frans Tax and Dean DellaPenna for critical reading of the manuscript. K.A.F. acknowledges support from National Science Foundation (NSF) Grant No. 9602433; R.A. was supported by an NSF postdoctoral fellowship awarded in 1992.
Footnotes
-
↵1 Current address: Department of Biochemistry, University of Arizona, Tucson, AZ 85721.
- Received September 17, 1997.
- Accepted December 15, 1997.
- Published February 1, 1998.
REFERENCES
