Differential Expression of the Arabidopsis Cytochrome c Genes Cytc-1 and Cytc-2. Evidence for the Involvement of TCP-Domain Protein-Binding Elements in Anther- and Meristem-Specific Expression of the Cytc-1 Gene

The Cytc-1 Promoter Directs Tissue-Specific β-Glucuronidase Expression

To characterize the expression patterns conferred by the promoter region of the Arabidopsis Cytc-1 gene, we cloned a 1,236-bp genomic fragment covering nucleotides −1,182 to +54 with respect to the ATG start codon in frame with the β-glucuronidase (gus) gene-coding region contained in the binary vector pBI101.3. This clone was used to obtain Arabidopsis transgenic plants by Agrobacterium-mediated transformation. Several independent transgenic lines were analyzed by histochemical staining of GUS activity using 5-bromo-4-chloro-3-indolyl-β-D-GlcUA (X-gluc) as substrate. Three-day-old seedlings grown on petri dishes on Murashige and Skoog medium showed staining only in the root meristem (Fig. 1A). Upon progression of development, GUS activity was detected in the shoot apical meristem, at the tip of cotyledons, and in nascent leaf primordia (Fig. 1, B–E). Specific expression was also observed at the root tips, in nascent secondary roots, and at the junction of secondary and primary roots (Fig. 1, G, H, and L). At the reproductive stage, expression was observed only in anthers, especially in pollen grains, and in stigmatic papillae (Fig. 1, J and K). No activity could be detected in hypocotyls, stems, leaves, siliques, or other flower organs (Fig. 1, A, F, I, and K), suggesting that the Cytc-1 gene promoter directs preferential expression in specific cell types. Mascarenhas and Hamilton (1992) reported the existence of artifactual staining of pollen due to the diffusion of dye produced in other parts of the anther. This produces variable staining of pollen grains even in homozygous plants. The uniform staining pattern observed in plants transformed with the Cytc-1 promoter construct suggests that this is not an artifact of the type mentioned above.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Histochemical localization of GUS activity in Arabidopsis plants transformed with either the Cytc-1 (A–L) or Cytc-2 (M–X) promoters fused to the gus reporter gene. A, Three-day-old seedling showing staining in the root tip (RT). B, Upper portion of a 7-d-old seedling showing staining in cotyledon tips (CT) and apical meristem (AM). C and D, Upper portion of 12- and 15-d-old seedlings showing staining in nascent leaves (NL) and apical meristem (AM). E and F, Cotyledon and leaf of a 20-d-old plant. G and H, Primary root showing staining at the tip (RT) and nascent secondary root (NSR). I, Silique. J, Anther. K, Flower. L, Lateral root showing staining at the base (LRB). M to P, Seedlings at different developmental stages. Q and R, Cotyledon and leaf of a 20-d-old plant. S and T, Primary root and nascent secondary root showing staining in the vascular cylinder (VC). U and V, Siliques showing staining in the septum (S) and the funiculus (F). W, Flower. X, Lateral root.

Expression of the Cytc-2 Gene Differs from That of the Cytc-1 Gene

The promoter region covering nucleotides −967 to +54 of the Arabidopsis Cytc-2 gene was also analyzed using gus fusions and plant transformation. Analysis of transformed plants indicated that the Cytc-2 promoter directs a much broader expression pattern. GUS activity was observed in vascular tissues of roots, hypocotyls, cotyledons, and leaves (Fig. 1, M–R). No activity was detected in meristems or nascent leaves or secondary roots (Fig. 1, O, S, and T). In reproductive organs, expression was evident in developing anthers and pollen, in petal veins, and at the junction of flowers and siliques with pedicels (Fig. 1W). Expression was also detected within siliques, specifically in the septum and the funiculus (Fig. 1, U and V). Cytc-2 promoter activity differs, then, from that observed with the Cytc-1 promoter, showing overlap only in anthers.

The amount of GUS activity present in protein extracts from different organs of plants transformed with either the Cytc-1 or the Cytc-2 promoter fusions was quantitatively assessed by a fluorometric GUS assay (Fig. 2). Specific GUS activity in seedlings (data not shown) and rosette leaves was not significantly different from basal levels (i.e. those obtained with plants transformed with the promoterless gus gene present in pBI101.3) in plants transformed with Cytc-1. For this construct, activity was highest in flowers and was also detected in extracts from roots and siliques (Fig. 2). For Cytc-2, highest expression was observed in leaves, and activities in flowers and siliques were also higher than those observed with the Cytc-1 promoter (Fig. 2). These results agree with those obtained using histochemical detection of GUS activity.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Analysis of GUS activity in protein extracts from different organs of plants transformed with either the Cytc-1 or the Cytc-2 promoter-gus fusions. Specific GUS activity was determined using protein extracts prepared from different organs (as indicated) of plants carrying either the Cytc-1 or the Cytc-2 promoters fused to gus. Activity was also measured in extracts from plants transformed with a promoterless gus gene (pBI101). Bars indicate the mean activity values obtained with organs from three independent transformants for each construct. Error bars represent se. Similar results were obtained in other experiments in which different lines were used.

Deletion Analysis of the Cytc-1 Promoter

To define the minimal promoter regions required for correct Cytc-1 gene expression, we performed a series of nested deletions from the upstream portion of the 1,234-bp fragment fused to gus (Fig. 3A). The different constructs were introduced into Arabidopsis and GUS activity was analyzed in transformed plants. Histochemical assays indicated that a fragment located between −147 and −218 of the start codon is absolutely required for Cytc-1 transcription, since no GUS activity could be detected in any tissue or cell type of plants carrying the proximal 200-bp fragment (−146 to +54) fused to gus (data not shown). Fluorometric assays using extracts from different organs were in agreement with histochemical assays, since GUS activity values decreased to those observed with the promoterless gus gene (Fig. 3B). A smaller deletion, down to nucleotide −218, produced plants with expression in flowers reduced to about 50%, suggesting that a positive element is present between −219 and −369 (Fig. 3B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Analysis of GUS activity driven by truncated forms of the Cytc-1 promoter. A, Constructs used for the analysis of the Cytc-1 promoter. Numbers indicate the upstream end of the promoter fragment present in each construct, with respect to the translation start site (TS); the downstream end was at +54 for all constructs. Construct Δ (147–218) carries a deletion comprising the indicated nucleotides within the context of the −369 promoter fragment. Plants transformed with the promoterless gus gene (pBI101) were also used. B, GUS expression levels in organs of plants carrying the different constructs. The results indicate the mean (± se) of three independent lines. Numbers below the bars indicate the respective construct. Similar results were obtained in other experiments in which different lines were used.

Deletion of the segment located between −147 and −218 within the context of the −369 promoter fragment completely abolished expression in all organs, indicating that upstream sequences cannot functionally replace this region (Fig. 3B). The inclusion of the essential fragment in front of a −90 cauliflower mosaic virus 35S minimal promoter was unable to confer specific GUS expression to transformed plants (data not shown). Accordingly, additional sequences located downstream of −147 also seem to be required for expression. Otherwise, since the Cytc-1 gene contains a TATA-less promoter, it may be that this segment does not function correctly within the context of the cauliflower mosaic virus 35S promoter that contains a TATA box.

Mutagenesis of the Promoter Segment Required for Cytc-1 Gene Expression

A set of 10-bp scanning mutations along the 72-bp (−147/−218) fragment required for expression of the Cytc-1 gene was used to map the presence of discrete regulatory elements (Fig. 4A). Fluorometric assays using extracts from different organs of plants transformed with the mutagenized 272-bp proximal promoter fragments (−218 to +54) fused to gus revealed that mutation of positions −147 to −156 produced very low GUS activity levels, similar to those observed with the promoterless gus gene (Fig. 4B). In addition, mutation of the adjacent 10-bp segment (−157 to −166) produced a moderate decrease in expression (Fig. 4B). Plants transformed with the mutagenized −147 to −156 fragment did not show GUS histochemical staining in any tissue analyzed (Fig. 5, D, J, P, and V). All other mutants showed GUS activity in anthers and meristems with a similar location to that observed in plants transformed with the nonmutagenized fragment (Fig. 5, F, L, R, and X), but in plants carrying mutations from −157 to −166 the intensity of staining was reduced with respect to plants containing the wild-type promoter fragment (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Mutagenic analysis of the Cytc-1 promoter region required for expression. A, Sequence of the Cytc-1 promoter regions that were analyzed by mutagenesis. Introduced mutations are specified below the wild-type sequence; asterisks indicate identical nucleotides. Numbers indicate the position relative to the translation start site. Site II elements and the telo-box are underlined. B, Specific GUS activity was determined using protein extracts prepared from different organs (as indicated) of plants carrying mutagenized promoter fragments fused to gus. All constructs contain the region-spanning nucleotides −218 to +54 of the Cytc-1 gene. Activity was also measured in extracts from plants transformed with the −146 to +54 fragment (−146) and with the promoterless gus gene (pBI101). The bars represent mean (± se) values for three independent lines carrying each construct. Similar results were obtained in other experiments in which different lines were used.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Histochemical analysis of GUS activity in Arabidopsis plants transformed with mutagenized promoters fused to the gus reporter gene. Seedlings (A–F), cotyledons and nascent leaves (G–L), flowers (M–R), or roots (S–X) from plants transformed with either the wild-type (−218) promoter or the same promoter with mutations (as described in Fig. 4) in the telo-box, the site II elements, the pollenQ motif, or the regions from −147 to −156 or from −187 to −196 were analyzed by histochemical staining.

The 10-bp promoter segment required for expression contains a copy of the so-called site II element (Fig. 4A), which has been implicated in the expression of genes in proliferating cells (Kosugi et al., 1995; Kosugi and Ohashi, 1997; Trémousaygue et al., 2003). This element has the consensus sequence TGGGCC/T and is usually present in more than one copy located proximal to each other and within −120 to −180 of the start codon (Trémousaygue et al., 2003). Indeed, a second site II element is present in the Cytc-1 gene promoter between −166 and −171 (Fig. 4A). To analyze the function of site II elements in the expression of the Cytc-1 gene, we have constructed a mutant promoter in which both elements were mutagenized by changing the two central nucleotides (GG) to either A or T (Fig. 4A). A similar sequence located between the site II elements (TGGGTT) was also mutagenized to TGGATT in this construct. Histochemical staining of plants transformed with the mutagenized promoter indicated that the introduced mutations caused a complete loss of expression in all organs and developmental stages (Fig. 5, C, I, O, and U). Fluorometric assays revealed the presence of basal levels of GUS activity similar to those found in plants transformed with the promoterless gus gene (Fig. 4B). The results suggest that the integrity of site II elements is essential for Cytc-1 gene expression.

Site II elements are usually present together with downstream internal telomeric repeats or telo-boxes (AAACCCTAA; Trémousaygue et al., 2003). A sequence perfectly matching the telo-box is located between −86 and −94 in the Cytc-1 gene (Fig. 4A). Plants bearing a construct in which the telo-box was modified displayed very low GUS activity in anthers when analyzed by the histochemical method (Fig. 5N), but no staining could be observed in other parts of the plant (Fig. 5, B, H, and T). Activity in anthers was so low that it could not be differentiated from the activity present in plants transformed with the promoterless gus gene when extracts from whole flowers were analyzed by the fluorometric method (Fig. 4B). It is then likely that this mutation causes a general decrease in expression, so that GUS activity becomes detectable only in anthers, where expression levels driven by the wild-type promoter are higher. These results agree with the fact that mutation of the telo-box in other genes does not cause a complete loss of expression, suggesting that this element acts as a stimulatory element coupled to site II or other motifs (Trémousaygue et al., 1999, 2003; Manevski et al., 2000).

Finally, a region containing the sequence AGGTCA, described as a quantitative element required for high-level expression in pollen of the ZM13 maize pollen-specific gene (Hamilton et al., 1998), was identified between −109 and −114 (Fig. 4A, pollenQ). Considering that the Cytc-1 gene shows expression in pollen, the relevance of this element for expression was also studied by mutagenesis. Plants with the mutagenized Cytc-1 promoter fused to gus showed lower specific activity values (Fig. 4B), suggesting that the presence of the AGGTCA element may cause a moderate increase in Cytc-1 gene expression levels. Histochemical staining of these plants did not reveal any difference, however, with plants bearing the wild-type promoter fragment (Fig. 5, E, K, Q, and W).

Induction of the Cytc-1 Promoter by Sucrose and the Cytokinin 6-Benzylaminopurine

We have previously shown that the Cytc-1 gene is transcriptionally regulated by carbohydrates (Welchen et al., 2002). GUS activity measurements in extracts from seedlings carrying different promoter deletions indicated that all constructs that produced GUS expression also showed induction by carbohydrates at similar levels than those observed with the larger construct (data not shown). This was also true for plants carrying mutagenized promoter fragments, except for the ones with changes between nucleotides −157 to −166, in which very low induction levels were observed (Fig. 6). Since these plants also show lower GUS activity in other organs (e.g. flowers), it can be speculated that the putative elements involved in induction by carbohydrates may be the same as those required for increasing the expression in organs like flowers.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Induction of the Cytc-1 gene by Suc and cytokinins. GUS activity was measured using the fluorogenic substrate MUG and protein extracts prepared from 7-d-old seedlings grown in Murashige and Skoog (MS) medium alone or supplemented with either 3% Suc or 50 μm BAP, as indicated. The plants used for the analysis were a mix of four independent lines for each construct, as indicated. Activity was also measured in extracts from nontransformed plants (wt). Similar results were obtained in other experiments in which different lines were used.

The fact that the Cytc-1 gene shows preferential expression in proliferating tissues led us to investigate the effect of the inclusion of cytokinins in the culture medium on GUS activity levels driven by different Cytc-1 promoter fragments. The results obtained indicated that Cytc-1 expression in seedlings is induced by treatment with the cytokinin 6-benzylaminopurine (BAP; Fig. 6). The response of the different constructs closely matched the one observed for Suc treatments, but activity values were higher in the presence of the hormone. These observations reinforce the notion that expression of the Cytc-1 gene is governed by elements that respond to cell proliferation.

Proteins Present in Cauliflower Nuclear Extracts Specifically Bind to the Cytc-1 Promoter

The presence of nuclear proteins that bind to the Cytc-1 promoter region required for expression was investigated using extracts prepared from cauliflower (Brassica oleracea) inflorescences. Cauliflower was chosen because it is a readily available source of proteins expressed in meristematic tissues and because of its proximity to Arabidopsis. Figure 7A (lanes 1 to 3) shows a mobility shift assay using the inflorescence nuclear extract and fragments comprising either nucleotides −21 to −218, −21 to −146, or −126 to −218, respectively. It can be observed that both fragments that contain the region required for expression produce several shifted complexes in a similar pattern (Fig. 7A, lanes 1 and 3), while only a faint shifted band was obtained when the region between −147 and −218 was removed (Fig. 7A, lane 2). This result suggests that the cauliflower nuclear extract contains proteins that specifically bind to this region of the Cytc-1 promoter. Indeed, when a labeled fragment comprising nucleotides −126 to −218 mutagenized in its site II elements was used, no binding was observed, indicating that the nuclear proteins bind to the same region that is essential for Cytc-1 expression (Fig. 7A, lane 4). A similar conclusion could be drawn from an experiment in which excess unlabeled wild-type and mutant fragments were used to compete binding to labeled DNA (Fig. 7B). While either a 10- or a 50-fold molar excess of the wild-type fragment produced a strong decrease in binding, almost no competition was observed when similar amounts of the mutant fragment were used (Fig. 7B, lanes 1 to 5). The small amount of binding observed with the mutant fragment was also more effectively competed by DNA containing nonmutated site II elements (Fig. 7B, lanes 6 to 10). We speculate that proteins with the same binding specificity present in Arabidopsis act as key regulators of Cytc-1 gene expression.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Nuclear proteins from cauliflower inflorescences and recombinant proteins AtTCP20 and AtPuralpha specifically bind to the Cytc-1 promoter. A, Nuclear extracts (3 μg) from cauliflower inflorescences were analyzed by an electrophoretic mobility shift assay for the presence of proteins that bind to labeled DNA spanning nucleotides −21 to −218 (lane 1), −21 to −146 (lane 2), or −126 to −218 (lane 3). In lane 4, a fragment spanning nucleotides −126 to −218 with mutations in its site II elements (as described in Fig. 4) was used. B, The binding of nuclear proteins to labeled fragment −126 to −218 was analyzed in the presence of only unspecific competitor (lane 1) or with the addition of a 10- or a 50-fold molar excess of unlabeled wild-type (lanes 2 and 3) or mutated forms (lanes 4 and 5) of the same fragment. In lanes 6 to 10, a similar experiment was carried out using the mutated form of the fragment as labeled probe. C, Binding of recombinant protein AtTCP20 to the Cytc-1 promoter. An aliquot (200 ng) of recombinant protein AtTCP20 was analyzed by an electrophoretic mobility shift assay for binding to a Cytc-1 promoter fragment spanning nucleotides −126 to −218 (lane 1) or to the same fragment with mutations introduced in the regions −187/196 (lane 2), −177/−186 (lane 3), or in the site II elements (lane 4). D, Binding of recombinant protein AtPuralpha (200 ng) to Cytc-1 promoter fragments spanning nucleotides −21 to −218 (lane 1) or −21 to −146 (lane 2), or to a similar fragment with mutations within the telo-box (lane 3) as described in Figure 4.

A TCP-Domain Protein Binds to the Cytc-1 Promoter Region Required for Expression

It has been proposed that site II elements are binding sites for a group of transcription factors that contain the so-called TCP domain (Kosugi and Ohashi, 1997; Trémousaygue et al., 2003). The ability of the relevant region of the Cytc-1 promoter to function as a TCP-domain protein-binding site was assayed in vitro, using recombinant protein AtTCP20 (Trémousaygue et al., 2003) in electrophoretic mobility shift assays (Fig. 7C). Specific retarded bands were observed when this protein was incubated with DNA representing promoter regions from −126 to −218, even in the presence of an excess of unspecific competitor (Fig. 7C, lane 1). Mutations spanning nucleotides −187 to −196 or −177 to −186 did not affect binding (Fig. 7C, lanes 2 and 3). Conversely, when a fragment with mutagenized site II elements was used, no binding was observed (Fig. 7C, lane 4). These results suggest that members of the TCP family are likely candidates to act as regulators of Cytc-1 gene expression.

In a similar way, we also tested the ability of the region of the Cytc-1 promoter containing the telo-box to bind recombinant protein AtPuralpha, which has been shown to interact with these elements in vitro (Trémousaygue et al., 1999). Electrophoretic mobility shift assays with wild-type fragments covering nucleotides −21 to −218 or −21 to −146 showed a specific retarded band that was not observed when a promoter fragment with mutations within the telo-box was used (Fig. 7D). These results indicate the existence of specific binding of AtPuralpha to the telo-box present in the Cytc-1 promoter.

In support of a functional role of site II elements, we have also noted the presence of similar motifs in cytochrome c genes from other plants. Searches in data banks revealed the existence of four sequenced genomic regions located upstream of cytochrome c-encoding sequences: two from rice (Oryza sativa) and one each from Brassica rapa and Medicago truncatula. The sequence from B. rapa (accession no. AC155340), more closely related to Arabidopsis, contains four site II motifs located at −359, −250, −160, and −133 from the translation start site. In Medicago (accession no. AC142095), three motifs are present at −925, −908, and −136, respectively. From the two rice genes, the one located in chromosome 5 (AC137623), which has been shown to be expressed at high levels and in several parts of the plant (Jang et al., 2002), contains site II motifs at −295 and −81, while the gene from chromosome 1 (AP003379) has site II motifs at −641, −211, −93, and −84. It is noteworthy that this gene also contains a downstream telo-box at −27, thus presenting an arrangement very similar to the one observed in the Arabidopsis Cytc-1 gene. A telo-box is also present in the Brassica gene, but located upstream of the site II motifs in this case (−381).

Several Nuclear Genes Involved in Cytochrome c-Dependent Respiration Contain Site II Motifs in Their Promoters

The relevance of site II elements for Cytc-1 gene expression led us to investigate the existence of similar motifs in genes encoding other components of cytochrome c-dependent respiration, assuming that the expression of these components must be somehow coordinated. We have analyzed 17 Arabidopsis nuclear genes encoding either cytochrome c oxidase subunits or proteins putatively involved in the biogenesis of this complex. Two or more neighboring site II motifs, located around −200 of the translation start site, could be identified in 11 of them, representing 65% of examined promoters (Table I). A similar arrangement was observed in 9 out of 15 genes (60%) encoding complex III subunits. This is considerably more frequent than expected by chance. Indeed, when a similar search was conducted using the −1 to −500 region of 93 randomly selected Arabidopsis genes, only in 17% of them could neighboring site II motifs, separated by 25 bp or less, be detected. These results strongly suggest that site II motifs may be targets of factors that regulate the coordinated expression of nuclear genes encoding components of this portion of the plant respiratory chain. The relevance, if any, of these elements must be assessed by mutagenic analysis of the respective promoter fragments.

Plant Material and Growth Conditions

Arabidopsis (Arabidopsis thaliana L. Heyhn.) ecotype Columbia (Col-0) was purchased from Lehle Seeds. Histochemical and fluorometric GUS assays were performed on plants grown on soil in a growth chamber at 22°C to 24°C under long-day photoperiods (16 h of illumination by a mixture of cool-white and GroLux fluorescent lamps) at an intensity of approximately 200 μE m⁻² s⁻¹. Seedlings were grown in petri dishes containing Murashige and Skoog medium and 0.8% agar under similar conditions. The effect of Suc and cytokinin on expression was analyzed in dark-grown seedlings. For this purpose, either 3% Suc or 50 μm BAP, as indicated, were added to the culture medium. The dishes were kept at 4°C for 4 d and then transferred to growth chamber conditions and kept under complete darkness for 14 d.

Reporter Gene Construct and Plant Transformation

A 1.23-kb NheI/BglII fragment from the Cytc-1 gene (At1g22840), comprising nucleotides −1,182 to +54 with respect to the ATG initiation codon, was cloned in frame in the XbaI and BamHI sites of plasmid pBI101.3. Constructs containing upstream deletions were obtained by PCR amplification with oligonucleotide CATP16 and either CATP15, CATP13, CATP19, CATP17, or CATP11 (Table II) followed by cloning in the HindIII and BamHI sites of pBI101.3. A deletion comprising nucleotides −147 to −218 was constructed using partially complementary primers that hybridized at both sides of the deleted region. Primers delA1 and delA2 were used with CATP13 and CATP16, respectively, to amplify hybridizing upstream and downstream sequences. The resulting products were mixed in buffer containing 50 mm Tris-HCl (pH 7.2), 10 mm MgSO₄, and 0.1 mm dithiothreitol (DTT), incubated at 95°C during 5 min, and annealed by allowing the solution to cool to 24°C in approximately 1 h. After this, 0.5 mm of each dNTP and 5 units of the Klenow fragment of Escherichia coli DNA polymerase I were added, and incubation was followed for 1 h at 37°C. A portion of this reaction was directly used to amplify the chimeric fragments using primers CATP13 and CATP16. In a similar way, complementary primers were used to generate scanning mutations along the −147 to −218 promoter region and to mutate specific putative regulatory elements. The sequences of the oligonucleotides used are shown in Table II. All constructs were checked by DNA sequencing.

A fragment spanning nucleotides −967 to +54 of the Cytc-2 gene (At4g10040) promoter was obtained by PCR amplification of Arabidopsis genomic DNA using primers CATP22 and CATP26. The resulting fragment, containing HindIII and BglII sites, was cloned in frame in the HindIII and BamHI sites of pBI101.3.

The respective constructs were introduced into Agrobacterium tumefaciens strain GV2260 and transformed bacteria were used to obtain transgenic Arabidopsis plants by the floral-dip procedure (Clough and Bent, 1998). Transformed plants were selected on the basis of kanamycin resistance and positive PCR carried out on genomic DNA with gene-specific primers and the gus primer 5′-TTGGGGTTTCTACAGGAC-3′. Primary transformants for each construct were initially analyzed for GUS expression by histochemistry. From these, 10 independent lines with single insertions (as deduced by kanamycin resistance segregation) and with representative expression patterns (those common to a vast majority of transformants) were further reproduced and homozygous T3 and T4 plants were used to analyze gus expression. Plants transformed with pBI101.3 or pBI121 were obtained in a similar way.

GUS Assays

GUS activity of transgenic plants was analyzed by histochemical staining using the chromogenic substrate X-gluc as described by Hull and Devic (1995). Whole plants or separated organs were immersed in a 1 mm X-gluc solution in 100 mm sodium phosphate, pH 7.0, and 0.1% Triton X-100, and, after applying vacuum for 5 min, they were incubated at 37°C until satisfactory staining was observed. Tissues were cleared by immersing them in 70% ethanol.

Specific GUS activity in protein extracts was measured using the fluorogenic substrate 4-methylumbelliferyl β-d-glucuronide (MUG) essentially as described by Jefferson et al. (1987). Total protein extracts were prepared by grinding the tissues in extraction buffer (50 mm sodium phosphate, pH 7.0, 10 mm EDTA, 10 mm β-mercaptoethanol) containing 0.1% (w/v) SDS and 1% Triton X-100, followed by centrifugation at 13,000g for 10 min. GUS activity in supernatants was measured in extraction buffer containing 1 mm MUG and 20% methanol. Reactions were stopped with 0.2 m Na₂CO₃ and the amount of 4-methylumbelliferone was calculated by relating relative fluorescence units with those of a standard of known concentration. The protein concentration of extracts was determined as described by Sedmak and Grossberg (1977).

DNA-Binding Assays

Nuclear extracts were prepared from cauliflower buds (obtained from a local market) as described by Manzara and Gruissem (1995). Clones expressing recombinant proteins AtPuralpha and AtTCP20 fused to the E. coli maltose-binding protein were kind gifts of Dr. Bernard Lescure and Dr. Dominique Trémousaygue (CNRS-INRA, Castanet-Tolosan, France). Expression and purification of the recombinant proteins were carried out as indicated by the manufacturers of the pMAL-c2 system (New England Biolabs). For electrophoretic mobility shift assays, aliquots of extracts or purified proteins were incubated with double-stranded DNA (25,000 counts per minute) obtained by amplification of the corresponding fragments with primer pairs CATP16/CATP17, CATP16/CATP11, or CATP12/CATP17, followed by restriction enzyme cleavage and labeling with [α-³²P]dATP by filling in the 3′-ends using the Klenow fragment of DNA polymerase. For nuclear extracts, binding reactions (20 μL) contained, in addition to labeled DNA, 20 mm HEPES (pH 7.5), 50 mm KCl, 2 mm MgCl₂, 0.5 mm EDTA, 1.0 mm DTT, 0.5% Triton X-100, 10% glycerol, 1.5 μg poly(dI-dC), and 2 μg pBI221 (as nonspecific competitors). Reactions were incubated for 20 min on ice, supplemented with 2.5% Ficoll, and immediately loaded onto a running gel (5% acrylamide, 0.08% bis-acrylamide in 0.5× TBE plus 2.5% glycerol; 1× TBE is 90 mm Tris-borate, pH 8.3, 2 mm EDTA). The gel was run in 0.5× TBE at 30 mA and 4°C for 1.5 h and dried prior to autoradiography. For recombinant proteins, binding reactions contained 10 mm Tris-HCl (pH 7.5), 1.0 mm DTT, 10% glycerol, 22 ng/μL bovine serum albumin, and 1 μg poly(dI-dC), and incubation and electrophoresis were carried out at room temperature.

Distribution of Materials

Upon request, all novel materials described in this publication will be made available in a timely manner for noncommercial research purposes, subject to the requisite permission from any third-party owners of all or part of the material. Obtaining any permissions will be the responsibility of the requester.