- © 1998 American Society of Plant Physiologists
Abstract
An Arabidopsis transgenic line was constructed expressing β-glucuronidase (GUS) via the auxin-responsive domains (AuxRDs) A and B (BA–GUS) of the PS-IAA4/5 gene in an indoleacetic acid (IAA)–dependent fashion. GUS expression was preferentially enhanced in the root elongation zone after treatment of young seedlings with 10−7 M IAA. Expression of the BA–GUS gene in the axr1, axr4, and aux1 mutants required 10- to 100-fold higher auxin concentration than that in the wild-type background. GUS expression was nil in the axr2 and axr3 mutants. The transgene was used to isolate mutants exhibiting altered auxin-responsive gene expression (age). Two mutants, age1 and age2, were isolated and characterized. age1 showed enhanced sensitivity to IAA, with strong GUS expression localized in the root elongation zone in the presence of 10−8 M IAA. In contrast, age2 exhibited ectopic GUS expression associated with the root vascular tissue, even in the absence of exogenous IAA. Morphological and molecular analyses indicated that the age1 and age2 alleles are involved in the regulation of gene expression in response to IAA.
INTRODUCTION
The plant growth hormone auxin, typified by indoleacetic acid (IAA), has been implicated in regulating many developmental and cellular processes by altering basic patterns of gene expression (Went and Thimann, 1937; Estelle, 1992; Hobbie and Estelle, 1994; Klee and Romano, 1994; Kende and Zeevaart, 1997). The signal transduction pathway(s) leading to auxin-mediated gene activation is not known. The lack of functionally defined auxin receptors and simple phenotypic traits that are specifically associated with auxin action have hindered the elucidation of the auxin signaling apparatus responsible for gene activation using biochemical and genetic approaches.
Auxin-responsive promoter elements of early genes, that is, genes that are among the first to be transcriptionally activated by auxin, provide specific molecular probes for accessing and exploring the auxin signaling apparatus in reverse (Guilfoyle, 1986; Theologis, 1986; Key, 1989; Ballas et al., 1993, 1995; Guilfoyle et al., 1993; Takahashi et al., 1995; Abel et al., 1996; Abel and Theologis, 1996). Recently, the use of transgenic plants containing reporter genes under the control of defined promoter elements have proven useful for identifying mutations that affect specific signal transduction pathways (Susek et al., 1993; Bowling et al., 1994; Jackson et al., 1995; Li et al., 1995; Martin et al., 1997). In this study, we used the auxin-responsive domains (AuxRDs) A and B of the early PS-IAA4/5 gene (Ballas et al., 1993, 1995) to construct an Arabidopsis transgenic line expressing the β-glucuronidase (GUS) reporter in an IAA- and tissue-dependent manner. Screening a mutagenized population of plants containing this transgene led to the isolation of two mutations, age1 and age2 (for auxin-responsive gene expression), that exhibited altered patterns of IAA-induced gene expression and novel morphological phenotypes.
RESULTS
AuxRDs A and B of PS-IAA4/5 Mediate Auxin-Induced Gene Expression in Arabidopsis
The auxin-responsive region (AuxRR) of the PS-IAA4/5 promoter is a 183-bp segment extending from positions −318 to −135 (Ballas et al., 1993, 1995). This region contains two auxin-responsive domains (AuxRDs A and B) defined by linker scanning mutagenesis that mediate auxin-induced gene expression (Ballas et al., 1993, 1995). To test the ability of these domains to function in Arabidopsis, we used Agrobacterium-mediated transformation to introduce the GUS gene under the control of the AuxRR or combinations of the AuxRDs A and B, as shown in Figure 1. The activity of the constructs was tested in both the Nossen (No-0) and Columbia (Col-0) Arabidopsis ecotypes, and similar levels and patterns of GUS expression were observed (data not shown).
Promoter–GUS Constructs and GUS Activity in Transgenic Arabidopsis Plants.
(A) Schematic representation of the AuxRD–GUS fusions. The transcription start site (+1) is shown, and the lengths of the promoter elements are indicated by the numbers over the constructs. CORE, nonauxin responsive segment spanning positions −92 to +96 of PS-IAA4/5 gene; UBQ, UBQ3 promoter.
(B) The mean GUS activity observed in seedlings analyzed without auxin treatment. The degree of auxin inducibility mediated by each promoter–GUS construct is shown as the ratio of GUS activity in seedlings treated with 2 × 10−5 M IAA relative to GUS activity observed in untreated seedlings. Each point is representative of seedlings from a single line.
(C) Typical GUS staining patterns seen in No-0 plants harboring the indicated promoter–GUS fusions. The specific lines shown contain −318–GUS (line 37), CORE–GUS (line 53), A–GUS (line 5.1), B–GUS (line 20), BA–GUS (line 1.3), 4×A–GUS (line 30.4), and 4×B–GUS (line 13.2). The seedlings were treated without (−IAA) or with (+IAA) 2 × 10−5 M IAA for 6 hr followed by staining with X-gluc.
Arabidopsis lines containing four copies of the AuxRD A (4×A–GUS) or one copy of the AuxRD A plus AuxRD B (BA–GUS) showed low levels of GUS expression in the absence of IAA; however, in the presence of IAA, both lines showed a two- to 15-fold increase in the level of GUS expression (Figure 1B). Histochemical staining of 5-day-old light-grown seedlings from the BA–GUS lines showed that the majority of IAA-induced GUS expression was localized in the elongation zone of the root, although weak IAA-induced GUS staining also was seen in the petioles of the cotyledons, the hypocotyl, and the root vascular tissue (Figure 1C). In the roots, the GUS staining pattern was strongest in the atricoblast cell files, whereas the tricoblast cell file exhibited a lower level of GUS staining (Figures 2A and 2B). The enhanced levels of GUS staining in cell files that failed to form root hairs are similar to those reported for the expression of GL2, a non-auxin-regulated gene encoding a homeobox protein that appears to repress root hair differentiation (Masucci et al., 1996). Genetic analysis of Arabidopsis mutants with altered patterns of root hair formation in combination with auxin- and ethylene-related mutations suggests that the signaling pathways of these growth regulators act to promote root hair outgrowth at a relatively late stage of differentiation (Masucci et al., 1996). The significance of the expression of BA–GUS in cell files that fail to form root hairs is not known.
The auxin-induced increase in GUS activity seen in the 4×A–GUS lines appeared as variable patterns of staining in the cotyledon petioles, the hypocotyl, root elongation zone, and the root vascular tissue extending beyond the elongation zone (Figure 1C). 4×B–GUS lines showed a pattern of spatially unregulated GUS staining that is independent of IAA treatment (Figure 1C). These results suggest that specific cells are more or less competent to respond to auxin by mediating gene expression through specific auxin-responsive domains. For example, whereas the BA–GUS and 4×A–GUS lines showed similar levels of GUS expression with and without auxin treatment, the cells capable of mediating auxin-induced GUS expression from these different domains were distinct. This differential regulation could also be seen when comparing the levels and distribution of GUS expression in the 4×B–GUS lines with the BA–GUS lines. Both AuxRDs A and B functioned to mediate the amplitude of the auxin response, whereas the combination of the two domains mediated the response in a specific set of cells in the root tip (Figure 1C). It is clear from these experiments and those performed with transgenic tobacco that the roots of light-grown plants contain the highest proportion of cells that are able to sustain auxin-induced expression from the AuxRDs A and B (Ballas et al., 1993).
Plants containing single copies of the A or B domain (A–GUS and B–GUS, respectively) showed low levels of GUS activity in the absence of auxin, but the addition of IAA resulted in small increases in GUS expression. Histochemical staining of seedlings from these lines showed diffuse and variable staining patterns (Figure 1C). In contrast, plants containing the entire AuxRR (−318–GUS) showed high levels of GUS expression in the absence of auxin that was enhanced two-fold in response to IAA. Again, histochemical analysis revealed that without auxin, seedlings displayed low levels of GUS activity throughout the plant and that IAA treatment enhanced GUS expression in the root tip (Figure 1C). Finally, control lines containing the non-auxin-responsive core region (positions −92 to +96) of the PS-IAA4/5 promoter (CORE–GUS) exhibited low levels of GUS activity that do not change in response to IAA (Figure 1).
Selection of the BA3 Reporter Line
After examining the strength and consistency of GUS staining patterns of eight different BA–GUS lines, we chose one line, designated WT/BA3, for detailed characterization. The similarity in size and shape of the WT/BA3 pollen grains with those from nontransgenic plants suggests that WT/BA3 is a diploid (Altmann et al., 1994). The WT/BA3 line is homozygous for the reporter gene, as determined by GUS staining and resistance to kanamycin. Table 1 shows the effect of various growth regulators on GUS expression in the root elongation zone of the WT/BA3 line. GUS expression was strongest in the presence of IAA, 1-naphthaleneacetic acid (NAA), or 2,4-D. Indole-3-butyric acid (IBA), a less active form of auxin, mediated a weaker response. The inactive auxin analog 2,3-dichlorophenoxyacetic acid (2,3-D), the IAA transport inhibitors p-chlorophenoxyisobutyric acid (PCIB) and 2,3,5-triiodobenzoic acid (TIBA), and IAA metabolic precursors indole and tryptophan were unable to induce GUS expression after 6 hr of treatment. Fusicoccin (FC), a fungal toxin that induces proton secretion in various plant tissues, also had no effect on the levels of GUS activity. Finally, the plant hormones gibberellic acid (GA), abscisic acid (ABA), benzyladenine (BA), salicylic acid (SA), and ethylene or its precursor, 1-aminocyclopropane-1-carboxylic acid (ACC), failed to enhance GUS expression.
To confirm that the AuxRDs A and B were responsible for the observed pattern of GUS expression, we introduced a second reporter gene, that encoding the green fluorescent protein (GFP), into the WT/BA3 line under the control of one copy of the AuxRD A plus one copy of AuxRD B (BA–GFP). BA–GFP expression was similar to that of BA–GUS; that is, it was restricted in the root elongation zone and was inducible by auxin (Figure 2C). The limited fluorescence intensity, however, precluded the use of this reporter for identifying mutants displaying altered auxin-induced gene expression.
DNA gel blot analysis using GUS and neomycin phospho-transferase DNA as probes showed that WT/BA3 contains two copies of the T-DNA insert as right border/right border inverted repeats (data not shown). To define the location of the transgene in the genome, a DNA fragment was amplified by thermal asymmetric interlaced polymerase chain reaction (TAIL-PCR) and was sequenced (Liu et al., 1995). One side of the amplified fragment contained 120 bp that were identical to the pBin19 T-DNA sequence, whereas the second fragment of flanking sequence did not have identity to any entries in the databases. PCR amplification using primers within this novel region and a DNA template from a yeast artificial chromosome (YAC) library showed that this fragment is present on YAC clones yUP12G6, yUP23E10, and yUP24B8, demonstrating that the BA–GUS transgene is located near the KG17 marker in the middle of chromosome 3 (see http://cbil.humgen.upenn.edu/~atgc/physical-mapping/physmaps.html).
Details of Histochemical Localization of GUS or GFP Reporter Gene Expression.
(A) Dark-field image of GUS staining in a 5-day-old WT/BA3 root after treatment with 10 −7 M IAA for 6 hr. The GUS staining reflects the expression of GUS in the elongation zone of the root tip.
(B) Cross-section through the apical region of a WT/BA3 root treated with 10−7 M IAA for 6 hr and then stained for GUS activity. The GUS staining pattern shows that the BA promoter mediates high levels of expression in the atricoblast cell file.
(C) Fluorescent image of the root tip from a 5-day-old wild-type seedling harboring the BA–GFP transgene treated with 2 × 10−5 M IAA for 24 hr.
(D) Bright-field image of a 5-day-old axr2/BA3 root seedling stained for GUS activity without being treated with IAA.
(E) Bright-field image of a 5-day-old axr2/BA3 seedling stained for GUS activity after being treated with 10−7 M IAA for 6 hr. Note the lack of staining in the elongation zone that is seen in WT/BA3 plants.
(F) Bright-field image of 5-day-old hls3/BA3 seedlings stained for GUS activity after being treated with 10−7 M IAA for 6 hr.
(G) Bright-field image of a 2-week-old hls3/BA3 seedling stained for GUS activity after being treated with 10−7 M IAA for 6 hr. Note the GUS staining in the emerging primordia of newly forming lateral roots.
The sensitivity of the BA–GUS reporter gene to exogenous auxin was determined by examining GUS staining in seedlings treated with 0 to 10−4 M auxin (Figure 3). In the WT/BA3 seedlings, treatment with 10−7 M IAA resulted in a dramatic increase in GUS activity. Based on the intensity of staining, maximal expression of the transgene occurred at 10−4 M IAA (Figure 3), and expression was inhibited at 10−3 M IAA (data not shown). Several Arabidopsis mutants with reduced auxin sensitivity have been identified and characterized. To determine the tissue specificity of the BA–GUS transgene in these mutant backgrounds, we crossed the WT/BA3 line with axr1-12 (Lincoln et al., 1990), axr2-1 (Timpte et al., 1995), axr3-1 (Leyser et al., 1996), axr4-3 (Hobbie and Estelle, 1995), aux1-7 (Pickett et al., 1990), and hls3-12 (Lehman et al., 1996). F3 seedlings homozygous for the mutant allele and the BA–GUS transgene were treated with 0 to 10−4 M auxin, and the GUS staining pattern was determined. The results from these experiments showed that axr1/BA3, axr4/BA3, and aux1/BA3 require 10- to 100-fold higher concentration of auxin compared with the wild-type background to show GUS staining in the root elongation zone (Figure 3). These results are consistent with previous observations showing that these mutants are less sensitive to treatment with exogenous levels of auxin.
In contrast, whereas axr2/BA3 and axr3/BA3 failed to show this cell-specific GUS staining in the root elongation zone even at the highest levels of auxin tested (Figure 3), both lines showed GUS staining patterns that were significantly different from those of WT/BA3 plants. Specifically, the axr2/BA3 seedlings showed auxin-induced GUS expression in the vascular tissue of the elongation zone, suggesting that the dominant axr2 mutation may specifically decrease the ability of the cells outside of the vascular tissue to respond to exogenously applied auxin (Figures 2D and 2E). The axr3/BA3 seedlings exhibited auxin-independent expression associated with the vascular tissue throughout the root, with the highest levels observed where the root bends (Figure 3). The lack of GUS staining in the root elongation zone in the axr2/BA3 and axr3/BA3 lines (Figure 3) was not due to the suppression of the BA–GUS gene because plants resulting from backcrosses to the nontransgenic Col-0 exhibited a normal GUS expression pattern (data not shown).
Disruption of HLS3, which is an allele of ALF1, SUR1, and RTY1, results in a dramatic increase in the levels of IAA and the proliferation of lateral roots (Lehman et al., 1996). Interestingly, BA-mediated GUS expression in 5-day-old hls3/BA3 seedlings was restricted to the lower part of the hypocotyl (Figure 2F). Such intense staining was never observed in the WT/BA3 lines even after treatment with high levels of auxin (data not shown). Furthermore, when hls3/BA3 seedlings were examined after 2 weeks, during which time the number of lateral roots increased, GUS staining was observed in the emerging lateral root primordia and the leaves (Figure 2G). Because the hls3 phenotype appears to be caused by an increase in IAA (Lehman et al., 1996), the high level of GUS expression in the hypocotyl suggests that this region may generate or accumulate high quantities of auxin. Alternatively, this increase in GUS expression could be mediated by an increase in sensitivity to auxin.
Isolation of age Mutants
The strong and reproducible auxin-induced GUS expression in the WT/BA3 line suggests that it can be used for identifying mutations with altered auxin-induced gene expression patterns. Using GUS expression in the root tip of the WT/BA3 line as a marker, we initiated a screen to isolate mutants that showed altered auxin-induced gene expression (Figure 4A). To identify putative mutants, we screened M2 seedlings of an ethyl methanesulfonate–mutagenized population by using a nondestructive GUS staining assay (Figures 4B to 4D). Briefly, roots of 10- to 14-day-old seedlings were incubated with 0.5 to 1 × 10−8 M IAA for 6 hr and then stained with GUS for 15 hr (Figures 4C and 4D). Because WT/BA3 seedlings failed to show any GUS activity when treated with this level of IAA (Figure 3), seedlings showing GUS staining after this treatment were selected as candidates for further analysis. An example of this screen is shown in Figure 4D. Of the 125,000 seedlings tested, 142 mutant candidates were identified. They were transplanted to soil and allowed to set seeds. Auxin-induced GUS activity was assayed in M3 progeny of 65 of these putative mutants to determine whether the phenotype was heritable. Forty-one of the lines showed GUS activity. Two homozygous lines, 43-1c and 12-2b, showing reproducible altered auxin-regulated gene expression were chosen for further characterization. These lines were named age1 and age2, respectively.
Effect of Various Growth Regulators on GUS Expression in the Root Elongation Zone of the WT/BA3 Line
Histochemical Analysis of GUS Activity in the Root Tip of BA–GUS Plants after Treatment with 0 to 10−4 M IAA.
Homozygous seedlings containing the BA–GUS gene in wild-type (WT), axr1-12 (axr1), axr2-1 (axr2), axr3-1 (axr3), axr4-2 (axr4), or aux1-7 (aux1) backgrounds were examined.
age Mutants Have Altered Morphology
In addition to the altered expression of BA–GUS, the age1 and age2 mutations showed distinct, tightly linked morphological characteristics that include defects in root and leaf morphology. When wild-type Arabidopsis seedlings were grown on plates containing high concentrations of agar and maintained at a 45° angle, the roots exhibited a wavy growth pattern (Figure 5B) that appeared to be caused by the periodic change in the orientation of the root tip generated by the twisting of epidermal cells (Okada and Shimura, 1990). In contrast, when the age1 seedlings were grown under these conditions, there was a marked decrease in the number of times the root tip underwent reorientation, even though there were no significant differences in the length of the roots (Figure 5C). In addition, the cotyledons and leaves of age1 seedlings appeared pale when compared with the cotyledons and leaves of wild-type plants (Figures 5B to 5D). When mature, age1 plants were shorter, exhibited a bushy growth habit, had pale, epinastic rosette leaves, and showed reduced fertility (Figure 5E). In addition, the age1 plants bolted later, and the inflorescence stem was thinner than that of the wild type (data not shown).
Although mature age2 plants also showed a short, bushy growth habit, the seedlings developed normally, the leaves expanded properly, and there was no change in the wavy root growth pattern (Figure 5E). Furthermore, age2 plants flowered at the same time as did wild-type plants and had a similar inflorescence stem. Unlike previously described auxin-response mutants, neither of the age mutations disrupted the root gravitropic response, nor did they lead to increased auxin resistance in a root growth inhibition assay (data not shown).
Figure 6 shows that whereas both mutations exhibited inducible GUS expression with 10−8 M IAA (compare wild type with age1 and age2), the tissue specificity of expression was distinct. In age1 plants, GUS expression induced by 10−8 M IAA was present in the root elongation zone, whereas the age2 mutant lacked staining in this region of the seedling (Figure 6A). In addition to this localized increase in GUS activity, age1 plants showed an increase in GUS staining in the cotyledon petioles, the hypocotyl, and the upper portion of the root vascular tissue (Figure 6B). Although the age2 plants showed staining in similar parts of the seedling, GUS activity was much stronger in the vascular tissue of the upper part of the root and the lower part of the hypocotyl (Figure 6B). Furthermore, although GUS staining in the root elongation zone of age2 seedlings required the same level of IAA as wild-type plants, the staining pattern associated with the vascular tissue was independent of auxin treatment. Treatment of WT/BA3 plants with high concentrations of IAA occasionally induced a similar pattern of GUS staining associated with the vascular tissue.
Outline of the Mutant Screen.
(A) Steps in the mutant screen. EMS, ethyl methanesulfonate.
(B) Two-week-old seedlings grown vertically in a four-celled Falcon 1009 Petri plate.
(C) Two-week-old seedlings with the agar removed from cell 2 so that the roots can be treated with auxin incubation solution.
(D) Simulation of the screening. A WT/BA3 seedling treated with 2 × 10−5 M IAA for 6 hr and stained for GUS activity was placed among the roots of seedlings incubated without IAA.
Characterization of age1 and age2.
(A) Schematic representation of the map location of the age1 and age2 loci. The genetic distances, in centimorgans, to the loci indicated were determined by analyzing 190 lines for the pvv4 marker, 195 lines for the ATEAT1 marker, 114 lines for the PAI1 marker, 43 lines for the ASA1 marker, and 88 lines each for the nga106 and nga139 markers. Chr., chromosome.
(B) to (E) Five-day-old wild-type (WT) and age1 seedlings ([B] and [C], respectively) grown on 1.6% agar plates inclined at 45°. (D) shows 4-week-old wild-type (left) and age1 (right) plants. (E) shows mature plants: wild type (left), age1 (center), and age2 (right). A comparison of (B) and (C) shows the decreased number of bends in the roots of wild-type versus age1 plants. (D) shows the decreased size and pale nature of age1 plants relative to wild-type plants. (E) shows the morphological phenotype of mature wild-type, age1, and age2 plants.
Genetic Analysis of age Mutations
Backcrosses of age1 plants with unmutagenized WT/BA3 plants were used to determine the nature of this mutation. The resulting F1 plants showed reduced but detectable GUS staining in the root tip with 10−8 M IAA, suggesting that the age1 mutation may be semidominant. We observed the segregation 82:49:29 (nonstained–weakly stained–strongly stained) regarding GUS staining in F2 seedlings. In contrast, all of the F1 plants had wild-type-like morphology, and the selfed F2 progeny showed a 3:1 distribution of wild-type–to–mutant phenotype, as expected for a recessive mutation. To clarify the GUS staining pattern, we performed an additional self-cross. Analysis of F3 seedlings from 127 F2 lines showed 34 lines yielding seedlings with no staining (wild-type homozygotes), 64 lines producing a mixture of staining and nonstaining seedlings (heterozygotes), and 29 lines yielding seedlings that all showed GUS staining (mutant homozygotes). All 64 heterozygous lines showed some seedlings with weak blue staining. We observed <10 percent stained F3 seedlings in eight lines for which we quantitated GUS expression. The 29 homozygous lines exhibiting GUS staining also possessed the age1 morphological characteristics, whereas the other 98 lines showed wild-type-like morphology.
These data indicate that age1 is a single, recessive mutation. The leakiness of GUS expression in the heterozygotes may be due to the inability of the wild-type protein at one-half of its concentration to effectively suppress GUS expression. The same amount of wild-type protein is sufficient to confer the wild-type phenotype in the heterozygous plants. Mapping of the age1 mutation by analyzing simple sequence length polymorphism (SSLP) and cleaved amplified polymorphic sequence (CAPS) markers on 195 F2 lines resulting from crosses with the Landsberg erecta (Ler) ecotype shows that the mutation falls between the pvv4 and ATEAT1 markers on chromosome 1 (Figure 5A).
All F1 seedlings from multiple lines obtained by backcrossing the age2 mutation with WT/BA3 plants showed a GUS expression pattern similar to that observed in wild-type seedlings. In addition, the F2 populations showed strong GUS staining in the vascular tissue in one-quarter of the progeny. Analysis of GUS activity in 156 F3 lines exhibiting the morphological phenotype associated with the age2 mutation also showed ectopic expression of GUS, suggesting that the morphological phenotype and the pattern of GUS expression are tightly linked. Taken together, these data indicate that the age2 mutation is also a single recessive mutation. Mapping of the age2 mutation by analyzing SSLP and CAPS markers on 114 F2 lines resulting from crosses with the Ler ecotype showed that the mutation falls between nga106 and nga139 on chromosome 5 (Figure 5A). Because the BA–GUS gene is integrated in chromosome 3, this excludes the possibility that the age1 and age2 mutations are caused by mutations in the transgene.
Altered Steady State RNA Levels in age Plants
Arabidopsis has at least 25 members belonging to the Aux/IAA gene family that show complex, overlapping patterns of expression during plant development (Abel et al., 1995; Kim et al., 1997). The majority of these family members are rapidly (4 to 30 min) induced by exogenous IAA in wild-type Arabidopsis, and the expression of each of these genes is altered in the auxin-resistant mutant lines axr1, axr2, and aux1 (Abel et al., 1995). Although the pea PS-IAA4/5 promoter has been studied extensively (Ballas et al., 1993, 1995), the promoters of only two of the Arabidopsis genes, IAA-4 and IAA-5, which correspond to AtAux2-11 and AtAux2-27, respectively (Conner et al., 1990), have been sequenced. The elements responsible for their auxin-regulated gene expression have not been defined thoroughly.
To analyze the effect of the age mutations on the steady state levels of IAA mRNAs, we isolated RNA from age1 or age2 plants that were homozygous for both BA–GUS and BA–GFP. Figure 7 shows the results of the RNA gel blot analysis. In both wild-type and age1 plants, the steady state RNA levels of GUS, GFP, IAA-1, IAA-4, and IAA-5 increased after treatment with 10−7 M IAA (Figure 7A). Consistent with previous reports, neither IAA-10 nor EF-1α showed any response to exogenous auxin (Abel et al., 1995). The apparent discrepancy between the increased level of GUS or GFP activity seen in the age1 roots treated with 10−8 M IAA and the lack of response in GUS or GFP transcript levels is likely due to the limited number of cells showing increased BA-driven expression. Although the concentration of IAA required to detect an increase in auxin-induced RNA levels remained constant in both the wild type and the age1 mutant, the amplitude of the response differed in the two backgrounds. Specifically, there is a decrease in GFP, IAA-1, IAA-4, and IAA-5 mRNA levels in the age1 plants relative to the wild-type plants after treatment with 10−7 to 10−4 M IAA (Figure 7A). The failure to observe a decrease in GUS mRNA levels in the age1 mutant may be due to one of the following reasons: (1) the expression of GUS mRNA may be higher due to a positional effect on the BA3–GUS transgene; (2) the expression of the various Aux/IAA genes tested may be lower in the mutant because of their differential tissue-specific expression; or (3) the mutation may preferentially affect GUS mRNA/protein stability or GUS mRNA translatability.
In age2 seedlings that have not been treated with IAA, the levels of GUS, GFP, IAA-1, and IAA-12 RNAs were higher than in the wild type (Figure 7B). The age2 mutation had no effect on the steady state levels of IAA-4 (Figure 7B), which is known to respond to exogenous auxin (Conner et al., 1990; Abel et al., 1995). Similarly, the age2 mutation had no effect on the levels of IAA-10, EF-1α, or GST-5, a glutathione S-transferase gene that is not auxin regulated (Watahiki et al., 1995; Figure 7B).
DISCUSSION
A novel screen to identify mutants that exhibit altered auxin-regulated gene expression has been performed. The basis for this approach was the observation that the extensively characterized AuxRDs from the pea PS-IAA4/5 promoter (Ballas et al., 1993, 1995) can mediate auxin-inducible gene expression of two reporter genes, GUS and GFP, in transgenic Arabidopsis. The expression characteristics conferred by various AuxRD combinations in Arabidopsis are similar to those obtained in previous studies using transient expression in pea protoplasts (Ballas et al., 1993, 1995). Both analyses suggest that the AuxRD A contains the major auxin-responsive element(s) (AuxRE[s]) and that AuxRD A and AuxRD B act cooperatively to activate early gene expression in response to auxin (Figure 1; Ballas et al., 1995). A second feature that facilitated the screen was the cell-specific expression of BA–GUS upon auxin treatment. Finally, the non-destructive nature of the root assay allowed the facile identification of putative mutants and their subsequent rescue for setting seeds.
The pattern of BA–GUS expression observed in the axr1, aux1, and axr4 mutants (Timpte et al., 1995) is consistent with the hypothesis that there are multiple pathways mediating auxin-regulated gene expression. Although we have not examined BA–GUS expression in double mutants, it is known that AXR1 functions in a signaling pathway that is independent of either AUX1 or AXR4 (Hobbie and Estelle, 1995; Timpte et al., 1995). Because auxin-induced BA–GUS expression is altered in each of these mutants (Figure 3), multiple, distinct auxin signaling pathways affect BA-directed gene expression.
Histochemical Staining of WT/BA3, age1/BA3, and age2/BA3 Plants.
(A) GUS staining patterns in the roots of 5-day-old seedlings after a 6-hr treatment with either 10−8 or 10−7 M IAA. WT, wild type.
(B) GUS staining patterns in the aerial portion of the seedlings that had not been treated with IAA.
RNA Gel Blot Analysis of Wild Type, age1, or age2 Plants That Were Homozygous for Both BA–GUS and BA8–GFP.
(A) Total RNA was isolated from 5-day-old wild-type (WT) or age1 seedlings that had been treated with 0 to 10−4 M IAA. Twenty micrograms of RNA was separated by electrophoresis, blotted to a Nytran membrane, and probed with radiolabeled gene-specific probes for the indicated cDNAs.
(B) Twenty micrograms of RNA isolated from the roots of 5-day-old untreated age2 seedlings was analyzed as described for (A).
Interestingly, the ectopic BA–GUS expression associated with the root vascular tissue in the axr3 background is similar to that seen in seedlings expressing a SAUR–AC1 promoter—GUS fusion (Leyser et al., 1996). Because SAUR–AC1 has been identified as an early auxin response gene (Gil et al., 1994), the axr3 mutation may lie in a signaling pathway that also mediates BA–GUS expression. The recent identification of AXR3 as IAA-17, a member of the Aux/IAA gene family (Rouse et al., 1998), raises the possibility that the Aux/IAA gene products may be involved in autoregulating early auxin gene expression (Kim et al., 1997).
Mutants related to heat shock, light-regulated gene expression, circadian rhythm, systemic acquired resistance, and carbon sensing have been obtained by using screens that identify altered expression of promoter–reporter gene constructs under different selective conditions (Susek et al., 1993; Bowling et al., 1994; Jackson et al., 1995; Li et al., 1995; Martin et al., 1997). Mutants obtained by this type of screen were not previously identified by scoring morphological characteristics, suggesting that mutations affecting reporter gene expression are specific to a particular signaling pathway. Similarly, map positions and GUS expression patterns show that age1 and age2 are not allelic to previously described auxin response mutants nor do they exhibit morphological changes that mimic other auxin response mutants. The complex pattern of Aux/IAA gene expression in the age backgrounds indicates that these mutations likely control late steps in specific signaling pathways that are responsible for proper control of various auxin-regulated genes. For instance, the fact that the age2 plants contain increased levels of GUS, GFP, IAA-1, and IAA-12 RNA but do not show changes in the amount of IAA-4 RNA suggests that this mutation may be due to the loss of a specific negative regulator that normally suppresses the expression of a subset of auxin-responsive genes.
The complex phenotype exhibited by age1 plants, including the increased sensitivity of GUS expression in specific cell types, the decrease in steady state levels of auxin-inducible RNAs, and the multiple phenotypic changes, indicate that this mutation is not due to changes in IAA biosynthesis or uptake. The increased sensitivity shown in specific cells suggests that the age1 mutation may represent the loss of a negative regulator that permits expression of BA-driven GUS expression at lower auxin concentrations. In addition, if this negative regulator normally functions to suppress the expression or action of other genes that are involved in the regulation of early auxin genes, then mutations that alter its activity will lead to the aberrant expression of these genes. If one of these genes is involved in a negative feedback loop to control the steady state levels of Aux/IAA RNAs, then an increase in its expression can account for the decreased level of Aux/IAA gene expression, even though the sensitivity to auxin remains the same. The isolation and characterization of the AGE1 and AGE2 gene products will provide a better understanding of their function and their role in regulating auxin-mediated early gene expression.
METHODS
Plasmid Construction and Plant Transformation
All plasmids were constructed using standard recombinant DNA techniques; their authenticity was confirmed by DNA sequencing. The promoter–GUS fusions containing the AuxRD A, AuxRD B, AuxRR (positions −318 to +96), AuxRD A plus B, 4×AuxRD A, 4×AuxRD B, and the auxin nonresponsive CORE (−92 to +96) were constructed from previously described promoter–CAT constructs (Ballas et al., 1995). The corresponding SalI-NcoI or XhoI-NcoI fragments containing the auxin-responsive domains were joined in a three-way ligation with the NcoI-SacI fragment of pUBQncoGUS (Norris et al., 1993) and the SacI-XhoI fragment of pBI101.1. The UBQ3–GUS construct was generated by ligating the HindIII-SacI fragments from pUBQ3ncoGUS and pBI101.1.
The BA–GFP gene was constructed as follows. The BA region of the PS-IAA4/5 promoter was isolated as an XhoI-NcoI fragment from pBA–CAT, and the C-terminal region of GFP with the nos terminator was isolated as an NcoI-EcoRI fragment from pBIN35mGFP4 (Haseloff et al., 1997). The fragments were joined in a three-way ligation with EcoRI-XhoI–digested pBluescript SK+ (Stratagene, La Jolla, CA) to yield pBA–N–GFP. The N-terminal portion of GFP was introduced into this construct as a 300-bp NcoI fragment generated by polymerase chain reaction (PCR), yielding pBA–GFP containing an NcoI site at the initiating methionine. The BA–GFP–nos fragment was introduced into the XhoI-EcoRI–digested binary vector pPZP121 (Hajdukiewicz et al., 1994), which was used to transform by electroporation Agrobacterium tumefaciens LBA4404, for transforming Nossen (No-0) plants, and GV3101(pMP90), for transforming Columbia (Col-0) plants. Transformation of Arabidopsis thaliana with the Agrobacterium strains was performed according to standard procedures (Valvekens et al., 1988; Bechtold et al., 1993).
Plant Growth Conditions and Reporter Gene Assay
Surface-sterilized seeds were plated on germination media (GM) (0.5 × Murashige and Skoog salts [Gibco BRL, Gaithersburg, MD], 1% sucrose, 1 × B5 vitamins, and 0.5 g/L Mes, pH 5.8) containing 0.8 or 1.6% Bacto agar (Difco, Detroit, MI). For synchronous seedling germination, the plates were kept in the dark for 2 days at 4°C and then transferred to 23°C with continuous light. Quantitative β-glucuronidase (GUS) assays were performed as follows. Ten-day-old kanamycin-resistant seedlings grown under continuous light were treated with or without 20 μM indoleacetic acid (IAA) for 6 hr, as described previously (Theologis et al., 1985). The seedlings were washed three times with 50 mM sodium phosphate, pH 7.0, and GUS activity was determined using 4-methylumbelliferyl β-d-glucuronide as a substrate (Gallagher, 1992). For histochemical analysis of GUS expression, 5-day-old seedlings were grown vertically on GM containing 0.8% agar, treated with the concentration of chemicals listed in Table 1 for 2 hr, rinsed three times with staining buffer lacking 5-bromo-4-chloro-3-indolyl β-d-glucuronide (X-gluc) (100 mM sodium phosphate, pH 7.0, 10 mM EDTA, 0.5 mM K4Fe[CN]6, 0.5 mM K3Fe[CN]6, and 0.1% Triton X-100), and then incubated for 18 hr in staining buffer containing 1 mM X-gluc. To remove chlorophyll from the green tissues, we incubated the stained plants in 70% ethanol. GUS staining patterns were recorded using a Nikon SMZ-2T or a Zeiss Axiophot microscope. Green fluorescence protein (GFP) imaging was performed on a Zeiss Axiophot microscope (Thornwood, NY) using a standard fluorescein filter set (extension filter, BP450-490; dichroic mirror, FT 510; emission filter, LP 520).
DNA and RNA Gel Blot Analyses
Plant DNA was isolated using the cetyltrimethylammonium bromide extraction procedure described by Murray and Thompson (1980). RNA was isolated from 5-day-old seedlings treated with the IAA concentration indicated in Figure 7 by using the procedure of Theologis et al. (1985). We performed DNA and RNA hybridization analyses with Nytran membranes (Schleicher & Schuell) by using 32P-labeled probes, as described previously (Abel et al., 1995).
Thermal Asymmetric Interlaced PCR
For mapping the integration site of the BA–GUS transgene in the WT/BA3 line, thermal asymmetric interlaced (TAIL)–PCR was performed following the procedure described by Liu et al. (1995), using the following primers: arbitrary primers TAIL-2 (5′-TC[A/T]TCIG[A/C/G/T]-ACIT[GC]CTC-3′) and TAIL-3 (5′-CA[A/T]CTIC[A/C/G/T]AGIA[GC]-TCG-3′); and TL-DNA border-specific primers TL-1 (5′-CCCTATCTCGGGCTATTCTTTTGA-3′), TL-2 (5′-TATAAGGGATTTTGCCGATTTCGG-3′), and TL-3 (5′-AACCACCATCAAACAGGATTTTC-3′). TL-1, TL-2, and TL-3 were used for primary, secondary, and tertiary TAIL-PCR reactions, respectively. The amplified fragments were subcloned into pBluescript KS− vector and sequenced. From the sequence obtained, two additional PCR primers, BA3F1 (5′-TACTCGCAACATAGCTTACATCA) and BA3F2 (5′-ATGCTTAGCTCTGTCCTCTAACA), were synthesized and used to screen the yUPYAC library for yeast artificial chromosome (YAC) clones containing this region of genomic DNA. Thirty-five PCR cycles at 94°C for 1 min, 50°C for 1 min, and 72°C for 2 min were performed, and the templates yielding a product were used to identify the chromosome location by using the map generated by Ecker et al. (http://cbil.humgen.upenn.edu/~atgc/physical-mapping/physmaps.html).
Mutant Screen and Genetic Analysis
Seeds of the WT/BA3 transgenic line were mutagenized with 0.3% ethyl methanesulfonate for 17 hr at room temperature, as described by Haughn and Sommerville (1986). These M1 seeds were germinated on soil and allowed to self-pollinate. Approximately 200 surface-sterilized M2 seeds were placed in wells 1 and 4 of a four-cell Falcon 1009 Petri plate (Becton Dickinson and Co., Lincoln Park, NJ) containing 1.6% agar in GM (see Figure 4). The seeds were placed in a parallel line 0.5 cm from the divider separating cells 1 and 4 from cells 2 and 3. To facilitate coordinated seedling emergence, we placed the plates in the dark for 2 days at 4°C and then maintained them vertically at 23°C under an 18-hr-light and 6-hr-dark cycle, with cells 2 and 3 being closest to the ground. After 10 to 14 days, the agar in cells 2 and 3 was carefully removed without damaging the seedling roots and replaced with incubation buffer containing 1 × GM, 0.1% ethanol, and 10−8 or 5 × 10−9 M IAA. After a 6-hr incubation in the dark at room temperature, the auxin incubation medium was replaced by X-gluc staining solution (1 mM X-gluc, 0.25 × Murashige and Skoog salts, 0.5% sucrose, and 50 mM sodium phosphate, pH 7.0), and the plates were incubated at 37°C for 15 hr. The candidate mutant plants were isolated by using forceps, washed extensively in sterile water, and placed on GM containing 0.8% agar. After 7 to 10 days, the recovered seedlings were transferred to soil, and seeds were collected.
From the 45,000 seedlings screened using 10−8 M IAA, 100 plants were rescued, and 45 of these set seed. After a secondary screen using 10−8 M IAA, 28 lines showed GUS staining. The age1 line was isolated from this pool of seedlings. From the 80,000 seedlings screened using 5 × 10−9 M IAA, 42 plants were rescued, and 20 of these set seed. After a secondary screen using 10−8 M IAA, 13 lines showed GUS staining. The age2 line was isolated from this second pool of seedlings.
The mapping of mutations was performed using cleaved amplified polymorphic sequence (CAPS) and simple sequence length polymorphism (SSLP) markers following standard procedures (Bell and Ecker, 1994). The scoring of individuals in the mapping population was conducted for age1 by looking for the morphological characteristics of age1 and for age2 by staining with X-gluc solution for 15 hr at 37°C without auxin treatment.
The axr1-12, axr2-1, axr4-1, and aux1-7 lines were obtained from M. Estelle (Indiana University, Bloomington). The axr3-1 line was provided by O. Leyser (University of York, UK), and the hls3-12 line was from J. Ecker (University of Pennsylvania, Philadelphia). The BA–GUS gene was introduced into these backgrounds by crossing them with WT/BA3 plants.
ACKNOWLEDGMENTS
We thank Betty Fong, Terry Chew, and Samia Gargouri for their assistance with the plant work and gratefully acknowledge Drs. Jim Haseloff, Nurit Ballas, Judy Callis, Ruiz-Canton Fransisco, Kotaro T. Yamamoto, Joe Ecker, Mark Estelle, and Ottoline Leyser for providing plasmids and mutant Arabidopsis lines. Finally, the assistance of members from the Sarah Hake laboratory for microscopic analysis and David Hantz for greenhouse management is recognized. P.J.O. was supported by a National Science Foundation Postdoctoral Research Fellowship in Biosciences Related to the Envrionment (No. BIR-9627143). This work was supported by a grant from the National Institutes of Health (No. GM-35447) to A.T.
Footnotes
- Received April 28, 1998.
- Accepted August 11, 1998.
- Published October 1, 1998.