Introduction

The UDP-glucuronosyltransferases (UGTs) are members of a superfamily of glycosyltransferases, and catalyze the transfer of a glucuronosyl group (GlcA) from uridine 5'-diphosphoglucuronic acid to a range of acceptor substrates, which could be aglycones, such as xylan, anthocyanin and pectin. In higher plants, glucuronic acid is an important constituent of hemicellulose in cell walls. Cell walls can be divided into two types: primary walls (I) and secondary walls (II). Primary walls are formed during cell growth, and are both mechanically stable and completely extensible for cell expansion to avoid the cracking of cells under their turgor pressure. Secondary cell walls, following the cessation of expansion and division, are synthesized within the bounds of the primary walls, and confer higher mechanical stability on specialized cell types such as xylem cells. There is evidence that the glucuronic acid units of glucuronoxylans in type I walls are added to every sixth xylosyl unit of the backbone 1. Furthermore, the α-D-glucuronic acid units can be attached to the O-2 position of the xylan backbone as side groups 2. Additionally, Sawada et al. 3 demonstrated that BpUGAT (Bellis perennis glucuronosyltransferase) catalyzes the regiospecific transfer of GlcA to the 2-hydroxyl group of the 3-glucosyl moiety of the substrate anthocyanidin. These results suggest that the UGTs may be important in regulation of differentiation and secondary metabolism 4, 5, 6.

Plant UGTs belong to a multigene family 7 and have complicated expression patterns. Waldron 8 reported that a UGT activity was mainly present in the secondary wall and highest expression was detected during the elongation of the pea epicotyls. Further analysis showed that the glucuronic acid was added by the UGT during the elongation of xylan 9. Zhong et al. 10 also found that a UGT (FRA8) was specifically expressed during secondary wall thickening in fibers and vessels. Woo et al. 11 reported that a Pisum sativum UGT catalyzes conjugation of UDP-glucuronic acid to an unknown compound, and its expression is correlated with mitosis and is strongly induced in dividing cells. Inhibition of the UGT expression by constitutive expression of its antisense mRNA markedly retarded the growth and development of transgenic alfalfa. Iwai et al. 12 studied a UGT mutant called nolac-H18 in tobacco (Nicotiana plumbaginifolia) generated by T-DNA insertion. The mutation was generated in the NpGUT1 gene, which functions during pectin biosynthesis. The nolac-H18 mutant had reduced intercellular attachment. In addition, our recent work showed that a cotton UGT gene, named GhGlcAT1, was expressed in fiber and stem; and its transcript was most abundant at 15 days post-anthesis in the fiber, but was not detected in ovule, flower, seed, root and leaf, indicating that the expression of GhGlcAT1 is regulated developmentally 13. These findings have provided important insights into functions of the UGTs in plant metabolism, growth and development. However, very little work has been carried out on the structural organization and regulatory elements of their promoters that are important in controlling the expression of these genes.

The economic importance of cotton (Gossypium hirsutum) is due to extraordinary elongation and thickening of the secondary wall of ovular epithelial cells; and the fiber provides an excellent system to study cell wall morphogenesis and its polysaccharide biosynthesis 14. Our previous report suggested that the GhGlcAT1 gene is expressed in a developmentally regulated pattern and may play an important role in the synthesis of cross-linkage polysaccharides 13. In order to uncover potentially important regulatory elements in the 5' upstream region of GhGlcAT1, we isolated a 1 647 bp fragment upstream of the GhGlcAT1 coding sequence using an improved PCR-based genomic walking method and characterized its ability to direct temporal, spatial and inducible expression of a reporter gene in transgenic tobaccos. Our results demonstrate that the 1 647 bp fragment of the cotton GhGlcAT1 promoter directs specific transcriptional regulation patterns in a model plant, and provide an important insight into the regulatory mechanisms underlying the temporal- and spatial-specific expression and stress responses of the cotton GhGlcAT1 gene.

Materials and Methods

Plant material and bacterial strains

Cotton (G. hirsutum cv. CRI 12) plants were grown in a field. Tobacco (N. abacum cv. shanxi) plants were grown in a greenhouse under a 16 h light and 8 h dark, 25 °C. Escherichia coli strain DH5a was cultivated in LB medium for vector constructs and DNA manipulation. Agrobacterium tumefaciens strain LBA4404 was cultivated in YEB medium for tobacco transformation.

PCR cloning of the GhGlcAT1 promoter region

Total genomic DNA was extracted from leaf tissue as described by Paterson et al. 15, and was digested with suitable restriction enzymes. The promoter region was cloned with an improved PCR-based genomic walking method described previously 16. Eight adaptor-ligated genomic restriction fragment libraries were generated after digestion with eight restriction enzymes. The corresponding library was subjected to a first round of PCR amplification with the outer adaptor primer (AP1) (5′-GTA ATA CGA CTC ACT ATA GGG C-3′) and an outer gene-specific primer (GSP1) (5′-ATT GTG TGT AGC AAC AGC AGG GC-3′), while the inner adaptor primer (AP2) (5′-ACT ATA GGG CAC GCG TGG TC-3′) and inner gene-specific primer (GSP2) (5′-TTC CTT CTT TGT TCT CTC AGC AGA CC-3′) were used for the second round of PCR. Major bands were isolated from using the Qiagen Gel extraction III kit (Qiagen, Germany), and the isolated fragments were cloned into a pGEM-T Easy vector (Promega, USA). Recombinant plasmid DNA used for sequencing was prepared using the QIAprep Spin Mini Prop Kit (Qiagen, Germany), and the inserts were sequenced using a BigDyeTM Terminator Cycle Sequencing Ready Reaction Kit (Perkin Elmer, USA) on an ABI PRISMTM 377 DNA Sequencer.

Construction of vectors

With the cloned genomic region as a template, a 1 647 bp fragment (pGhGlcAT1) upstream of the translational start codon of GhGlcAT1 gene and its 5' deletion derivatives were generated by PCR with seven forward primers (pF1: 5'-CAAGCTTCA GAC CTG AG-T CAT TC-3′; pF2: 5′-CAAGCTTCA GTC TAA CGG GTT AGT TG-3′; pF3: 5′-CAAGC-TTAA AAC TTA AAC CCC AAT TC-3′; pF4: 5′-CAAGCTTTC TCT ATA TGA TAG TTG-GC-3′; pF5: 5′-CAAGCTTTA AAT GAC AGC ATA GAC TCC-3′; pF6: 5′-CAAGCTTA-C AAA TTA ATA GTT ACC AT-3′; pF7: 5′-CAAGCTTCC AAG CCA ACT CCT GTT AT-3′; the introduced HindIII sites are underlined) and one reverse primer (pR1: 5′-CTGGATC CC TTA TGA GTA AAA TGG AAT T-3′, the introduced BamHI site is underlined). The amplified fragments were respectively inserted into the plasmid pBI121 (Clontech) as a HindIII-BamHI fragment at the corresponding restriction sites in place of the cauliflower mosaic virus (CaMV) 35S promoter region, resulting in a series of pBI-pGhGlcAT1::GUS vectors. The seven expression vectors were named P1617 (−1617/+30, as the full-length promoter construct in this study), P1355 (−1355/+30), P1049 (−1049/+30), P813 (−813/+30), P635 (−635/+30), P457 (−457/+30) and P281 (−281/+30), respectively. Ligation reactions were carried out using T4 DNA ligase (Promega, USA), and the inserts were verified by sequencing analysis. The confirmed constructs were used to test promoter activities in transgenic tobaccos. Additionally, a pBI121 vector (Clontech, USA) with the GUS gene controlled by the CaMV 35S promoter was used as a positive control, while the plasmid pBI101 (Clontech, USA) with the promoterless GUS gene was used as a negative control.

Tobacco transformation

Individual binary vectors, including pBI101, pBI121 and a series of pBI-pGhGlcAT1::GUS, were introduced into A. tumefaciens LBA4404 by the freeze-thaw method 17. Leaf discs of tobacco were transformed as described by Horsch et al. 18. Transformed plants were selected on MS medium 19 supplemented with 500 mg/l carbenicillin and 200 mg/l kanamycin. The rooted plantlets were verified by PCR using the forward primer 5′-CTC ATT ACG GCA AAG TGT GGG-3′ and reverse primer 5′-GTG CAC CAT CAG CAC GTT ATC G-3′. Positive plantlets were transferred to soil and allowed to blossom and bear seeds. All plants were grown in the greenhouse at 25 °C with a 16 h light/8 h dark cycle. The collected tissues were frozen immediately in liquid nitrogen and stored at −70 °C.

Fluorometric quantification of GUS activity and histochemical staining

Plant tissues were ground into a fine powder using liquid nitrogen with a mortar and pestle, and suspended in GUS extraction buffer (50 mM sodium phosphate, pH 7.0; 0.1% Triton X-100; 10 mM 2-mercaptoethanol; 10 mM 1,2-diaminocyclohexane-N,N,N,N-tetra-acetic acid and 0.1% sodium lauryl sarcosine). The supernatant was collected after centrifugation at 12 000× g for 10 min at 4 °C. Fluorometric quantification of GUS activity was performed using 4-methylumbelliferyl-β-D-glucuronide substrate 20. The content of total proteins was determined using the Bradford method 21. The GUS activity is expressed as pmol of 4-methylumbelliferone per mg protein per min.

Histochemical localization of GUS activity was performed as follows: samples were fixed with 0.5% paraformaldehyde in 0.1 M sodium phosphate (pH 7.0) for 30 min, then various tissues of transgenic tobacco samples were incubated in 5-bromo-4-chloro-3-indolyl-D-glucuronic acid (X-gluc) solution at 37 °C from 3 h to overnight until the blue staining had reached sufficient intensity 20. Photosynthetic tissues were cleared of chlorophyll by passaging through a 70-100% ethanol series. Photography was performed with an anatomy microscope (Leica, Germany) or a camera (Nikon 8700, Japan).

Stress treatments

Three-week-old transgenic tobaccos (T2) were used in stress treatments. For chemical treatments, whole plants were floated on MS medium solution containing 200 mM sucrose, 200 mM glucose, 200 mM fructose, 200 mM mannitol, 200 mM sorbitol, 50 μM naphthylacetic acid, 10 μM gibberellic acid (GA3), 20% PEG8000, 50 μM abscisic acid (ABA), 250 mM NaCl, 50 μM methyl jasmonate (MeJA) and 1 mM ethylene, respectively, in Petri dishes. MS, without additions, was used as control. The GUS activity was measured after samples had been kept in a growth chamber at 25 °C for 24 h in the dark. All data represent mean value ± SE of at least five repeated experiments from different transgenic lines.

For further analysis of sugar induction, transgenic plants of each construct were treated with 0.2 M sucrose in MS solution. The GUS activity was measured after samples had been kept in a growth chamber at 25 °C in the dark for 24 h.

Results and Discussion

Sequence characterization of the GhGlcAT1 promoter region

A GhGlcAT1 promoter region was isolated as described in Materials and Methods, and after the second round PCR, two major bands were amplified from the Dra I- and Hind III-restriction fragment libraries, respectively, but no bands could be produced from other restriction fragment libraries (data not shown). Sequencing analysis showed that the larger band from the Hind III-restriction fragment library overlapped with the smaller band from the Dra I-restriction fragment library. The resulting 1 647 bp fragment, upstream of the GhGlcAT1 gene, was sequenced and designated as pGhGlcAT1, and the fragment also contains 30 nucleotides of the 5' end of the GhGlcAT1 cDNA. The first base of the cDNA was designated +1 as the putative transcription start site (Figure 1).

Figure 1
figure 1

Nucleotide sequence of the GhGlcAT1 promoter region. The putative transcription start site is designated as +1 and indicated by an arrow. The putative TATA box and CAAT box are shown in bold. The seven forward primers are underlined. Six Myb elements are in bold and shaded; eight W-boxes (TGACT) are in bold and boxed. A reverse AuxRE (GAGACA) element is in italics and underlined.

Full size image

A promoter motif search was carried out to define putative cis-elements in the pGhGlcAT1 sequence using the software programs PLACE 22 and PlantCARE 23. A number of potential regulatory motifs corresponding to known cis-elements of eukaryotic genes were found (Figure 1). A TATA box was found at the positions −39 to −34 upstream of the putative transcription start site, whereas a CAAT box was observed at positions −74 to −70 (Figure 1); these boxes function as basal promoter elements for transcription. Furthermore, regulatory elements with homology to those identified in other seed/endosperm-specific genes were found in the pGhGlcAT1 region such as an amylase-box, E-box, CARE, GARE, GCN4-motif, ACGT motif and AACA motif (Table 1). The amylase-box was required for high expression of the gene in the seeds 24, while the E-box was shown to be bound by bHLH (basic helix-loop-helix) transcription factors and to regulate the seed-specific expression of seed storage protein genes and the expression of genes involved in anthocyanin biosynthesis 25, 26, 27. The CARE and GARE boxes were required during seed germination in Arabidopsis and rice 28, 29. In addition, a combination of GCN4, ACGT and AACA motifs was sufficient to confer a detectable level of endosperm expression in rice 30, 31. Several motifs including AGAAA, AAATGA and GTGA, and six MYB response elements are also present in the promoter region (Table 1). The AGAAA, AAATGA and GTGA motifs were reported to relate to pollen grain development 32, 33, 34, while MYB response elements were related to trichome development 35, 36. Additionally, the promoter region also contains several potential motifs with homology to inducible regulatory elements such as the W-box and AuxRE (Table 1). The W-box was reported to be involved in sugar stress response in barley 37, while a reverse AuxRE (GAGACA) was found to bind to auxin response factors 38. Taken together, sequence analysis of pGhGlcAT1 promoter region suggests that it contains a number of elements that are potentially important in controlling gene expression and that the GhGlcAT1 gene may be subjected to complex regulations including tissue-specific expression and stress induction.

Table 1 Regulatory motifs found in the GhGlcAT1 promoter region
Full size table

Histochemical and quantitative analysis of GUS activity

To investigate the temporal and spatial regulation directed by the pGhGlcAT1 promoter region, transgenic tobaccos carrying the GUS reporter gene fused to the 1647 bp fragment were grown in the greenhouse, and T1 plants were histochemically assayed for GUS staining. GUS staining was observed deeply in the whole 1-day-old germinated seeds (Figure 2A), and was also strong in the coats of the 2-day-old germinated seeds (Figure 2B). In 2-day-old seedlings, GUS staining was observed in the root cap meristem zones and cotyledons, but less GUS activity was detected in the root hairs (Figure 2C). In the 3-week-old seedlings, GUS activity appeared in the root tip (Figure 2D and 2E) and the shoots (Figure 2F). GUS staining in aerial tissues was found mainly in the vascular tissues and trichomes of the upper part of the shoots (Figure 2F–2G), but not in the leaves, mature vascular bundles and trichomes of basal shoots (data not shown). Strong GUS staining in the vascular tissues of stems was also consistent with our previous report of GhGlcAT1 expression by RNA blotting 13. To monitor low levels of tissue-specific GUS activity, quantitative GUS assays were performed with 4-week-old transgenic T1 plants. As shown in Figure 3, GUS activity was mainly detected in the basal roots and the upper regions of stems, but very low activity was detected in the fourth leaves. The GUS activity in roots (111.94 pmol/mg protein/min) was 5.19-fold higher than that in the leaves. These results were in agreement with analysis of the GUS staining, indicating that the GhGlcAT1 promoter directed reporter expression in the meristematic and vascular tissues.

Figure 2
figure 2

Histochemical analysis of GUS activity in tobaccos expressing the GhGlcAT1 promoter-GUS chimeric gene at different developmental stages. (A) GUS is expressed in the 1-day-old germinated seedling. (B) Blue staining is present in the seed coat. (C) Blue staining is focused on the 2-day-old germinated seedling. (D) Blue staining can be observed in the root tip of 3-week-old plant. (E) GUS activity is present in the root meristem tissue. (F) The longitudinal section of the upper part shoot; GUS expression is in the shoot vascular tissues. (G) GUS staining is concentrated in the trichomes. GUS staining is located in pollen grains (H) and the anther (I). GUS is expressed in the seeds of longitudinal section of a 10-DAF fruit (J), cross-section of 10 DAF fruit (K) and seeds (L). (M-P) GUS activity in trichomes derived from a series of pGhGlcAT1 transgenic plants and pBI101 control plants. GUS staining is found in the trichomes of P1617 (M), P457 (N) and P281 (O) transgenic tobaccos while no staining was found in the trichomes of pBI101 control (P). sc, seed coat; rh, root hair; me, meristem; rt, root; rc, root cap; va, vascular bundle; tr, trichome; po, pollen grain; an, anther; pl, placenta; se, seed.

Full size image
Figure 3
figure 3

GUS activity assay in transgenic tobaccos. Roots (rt), stems (st) and leaves (lf) are indicated in the transgenic tobacco plants (A) and were assayed for GUS activity (B). Anthers (ant), stigmas (sti), styles (sty), sepals (sep), petals (pet), ovaries (ova), receptacles (rec) and pedicels (ped) from mature plants (C) were sampled and assayed for GUS activity (D). Mean values are expressed in pmol MU per mg protein per min with data from five independent lines. Error bars indicate SE values.

Full size image

In the reproductive growth stage, strong GUS activity was observed in anthers and pollen grains of flower organs (Figure 2H–2I). As they are T1 transgenic plants, only about half of pollen grains (1:1) showed strong GUS staining (Figure 2H), indicating that only one copy of extrinsic DNA was inserted into the tobacco genome. In 10-day-old fruits, GUS staining could be seen in the seeds, but with less activity in the placenta (Figure 2J–2K). Approximately 3/4 of the seed showed GUS staining (Figure 2L). GUS activity was hardly detected in the corresponding tissues of the negative control plants (data not shown). Consistent with the GUS activity levels, the promoter region contains several important elements, such as an amylase-box, CARE, GARE, GCN4-motif, ACGT motif and AACA motif, which are involved in seed-specific and normal embryo-specific expression 24, 28, 29, 30, 31, 39. Quantitative GUS assays showed that the highest expression was in the anthers, with relatively high expression in the stigmas and styles. However, GUS activity was detected at lower levels in the sepals, petals, pedicels, receptacles and ovaries (Figure 3). The average GUS activity in the anther was 203.12±5.42 pmol/mg protein/min, at least 6.1-fold higher than that in pedicels, and two-fold of that in styles (Figure 3). High GUS activity in anther was also consistent with GUS staining in anther and pollen grains. Several kinds of pollen development-related motifs (AGAAA, AAATGA and GTGA) present in the pGhGlcAT1 promoter region are likely related to strong GUS expression in anther and pollen grains.

A PsUGT1 gene encoding a Pisum sativum UGT was reported to be expressed in root cap meristem 11. Inhibiting PsUGT1 expression by the constitutive expression of its antisense mRNA (under control of the 35S promoter) markedly retarded the growth and development of transgenic alfalfa, especially the floral development. The floral buds from antisense transgenic plants were short and thick, with stamens producing less pollens. These results indicate that PsUGT1 plays a role during the development of pollen or other flower organs. Furthermore, the PsUGT1 promoter::GUS fusion gene was found to have a high expression in active cell division regions 40. In addition, an NpGUT1 (tobacco glucuronyltransferase 1) was also expressed predominantly in shoot and root apical meristems 12, while the expression of antisense NpGUT1 RNA produced shoots with reduced cell adhesion. Analysis of the nolac-H18 mutant tissues showed that NpGUT1 is involved in pectin biosynthesis.

In the present work, the GUS gene controlled by the GhGlcAT1 promoter region was also highly expressed in root and shoot apex meristem tissues, pollen grains as well as pistils. These patterns were similar with PsUGT1 and NpGUT1 expressions in other plant species, suggesting that GhGlcAT1 might also function during differentiation of meristem tissues, development of pollens and cell division biosynthesis.

Effects of abiotic stress on GhGlcAT1 promoter in transgenic tobacco

The 3-week-old transgenic tobaccos from the T2 homozygous seed were used for various abiotic treatments, and then their GUS activity was examined. As shown in Figure 4, glucose and sucrose treatment could obviously raise the GUS activity, which was about 88.1% and 79.5% higher than that of non-treated control plants, respectively. However, GUS activity was only slightly raised (by about 13.8%) in fructose-treated plants, and no increase in GUS activity was detected in mannitol- and Sorbitol-treated plants (Figure 4). Thus the treatments of glucose and sucrose had a greater effect on GUS gene expression than those of fructose, mannitol and sorbitol. One possible explanation may be that glucose and sucrose might contribute to the pool of GlcA, one of the substrates for UGTs. It is possible that pGhGlcAT1 might be regulated by different types of sugar. In addition, GA and NAA (which are associated with cell division) treatment also enhanced GUS activity, which was about 31.7% and 18.6% higher than that in the non-treated control plants, respectively. However, treatments with Eth, NaCl and MeJA did not cause any significant changes in the GUS activity compared with the control plants (Figure 4), while treatments with ABA and PEG8000 decreased the GUS activity (by about 14.3% and 25.3%, respectively). The possible reasons might be that ABA and PEG8000 stress could quickly repress cell meristems and growth in these tissues.

Figure 4
figure 4

GUS activity in 3-week-old pBI-pGhGlcAT1-1647::GUS transgenic tobaccos under abiotic stress, including sorbitol (Sor), mannitol (Man), glucose (Glu), sucrose (Suc), fructose (Fru), ABA, PEG8000 (PEG), ethylene (Eth), naphthylacetic acid (NAA), NaCl, MeJA, GA, and MS medium (CK), respectively. MS treatment was used as a control and taken as 100%. Each value represents the average of at least five transgenic individuals. Error bars indicate SE values.

Full size image

On the other hand, the GUS gene was expressed in root cap meristematic zones and upper part of the shoot (Figure 2D–2F), and the GUS activity increased in response to GA and NAA treatments (Figure 4). It is well known that GA can promote shoot elongation and cell wall expansion, while the shoot apical meristem is the main site of auxin biosynthesis. Therefore, GhGlcAT1 expression in those locations might be controlled by the fine regulation of plant hormones in those tissues. Additionally, meristematic zones are also the main sites of primary cell wall biosynthesis, while vascular bundles are mainly associated with the synthesis of the secondary cell wall. The high expression in those tissues indicates that the GhGlcAT1 gene might participate in primary cell wall and initial secondary cell wall synthesis. Therefore, the results presented in this paper, together with the data from Northern blot analysis 13, strongly suggest that GhGlcAT1 might be related to sugar and hormone responses in fiber development.

Deletion analysis of the promoter in transgenic tobacco

In order to localize promoter regions important for the transcriptional control of the GhGlcAT1 gene, the full-length promoter (P1617) and a series of its 5′-truncated fragments (P1355, P1049, P813, P635, P457 and P281) were fused to the GUS reporter gene (Figure 5). The up-ground shoots of five individual 3-week-old transgenic T1 tobacco plants were analyzed. The results showed that GUS activities were slightly higher in plants containing P1049 (−1l049/+30) compared to P1617 (−1l617/+30). This suggests that there might be repressor elements in the region upstream of −1l049. GUS activities were decreased in the transgenic plants harboring the P813 (−813/+30) and P635 (−635/+30) constructs, while there were less changes comparing the plants harboring P635 and P281 (−281/+30) constructs (Figure 5). It was remarkable that a 311 bp (−281/+30) promoter fragment could still drive the expression of high GUS activity. Furthermore, GUS staining could still be observed in the trichomes of transgenic tobacco apical shoots containing P457 (−457/+30) and P281 constructs as well as in the full-length construct (Figure 2M–2O), while no GUS staining could be observed in the trichomes of pBI101 transgenic tobaccos (Figure 2P). Therefore, the elements needed for trichomes-specific expression might be located in this 311 bp fragment (−281/+30). Further sequencing analysis found that there was an MYB recognition site and an E-box in this region. Fiber-specific promoter should contain critical element(s) conferring specific expression in trichomes of model plants, as cotton fibers are unicellular trichomes that are derived from the extraordinary elongation of epidermal cells of cotton ovules 41, while the regulation of trichome development in plant is dependent on a number of MYB transcription factors 42, 43, 44. It is believed that these two motifs might be related to trichome-specific expression. In addition, trichome development was also regulated by GL3, which is related to the MYC transcription factor. GL3 contains a conserved bHLH domain in its C terminus, which could bind to the E-box 45. GL3 regulates trichome development in Arabidopsis via interaction with GL1 and TTG1. The GhGlcAT1 gene is highly expressed in fiber cells at the elongation stage 13, and its promoter sequence contains six MYB recognition sites and four E-boxes (Table 1), which might be responsible for directing GUS reporter expression in tobacco trichomes. To avoid the difficulties in the regeneration of cotton plants, several fiber-specific promoters have been studied in model plants such as tobacco 46, 47, 48 and Arabidopsis 36, 41. The results showed that these promoters contain critical element(s) conferring preferential expression in trichomes of model plants, as cotton fibers are unicellular trichomes that are derived from the extraordinary elongation of epidermal cells of cotton ovules 41. Based on our results, the promoter of GhGlcAT1 may also function in a similar way as the GhGlcAT1 gene is preferentially expressed in cotton fiber 13.

Figure 5
figure 5

Structure of GUS fusion constructs and determination of GUS activity in transgenic tobaccos. The bars represent various deletion fragments of the GhGlcAT1 promoter, numbering started with the transcription start site (+1), and is indicated by a line. GUS activities presented above were the mean values (pmol MU per mg protein per minute) of (n) individual plants including SD (±), and the ratio of GUS activities between sugar-induced and uninduced plants were calculated. “+” indicates inducible and “−” uninducible.

Full size image

As pGhGlcAT1 was induced by sucrose and glucose stresses, an attempt was also made to uncover the cis-regulatory elements in response to sucrose induction. GUS activities could be strongly induced in transgenic plants harboring five of the promoter constructs when plants were placed on MS medium supplemented with sucrose (Figure 5). The highest GUS induction was found in transgenic plants containing P635 (2.14-fold). However, the GUS activity was only slightly increased (1.36-fold) in transgenic plants containing P457, while the induction was completely lost in the plants harboring P281 (Figure 5). These results indicated that the cis-element(s) required for sucrose induction might be mainly located between −635 bp and −281 bp. Sequencing analysis revealed that there are eight W-boxes in the GhGlcAT1 promoter region, and W-boxes were reported to be associated with sugar responses 37. Two W-boxes and one reverse W-box are located in the region between −635 and −457 bp, and −457 and −281 bp, respectively; and no W-boxes are present in the region between −281 bp and the translation start site (Figure 1). Our results suggested that W-boxes in the GhGlcAT1 promoter might play an important role in sugar responses.

In conclusion, the present work using GUS expression assay indicated that a 1 647 bp fragment upstream of GhGlcAT1 coding sequence confers a high level of GUS expression in trichomes, vascular tissues, pollen grains, seeds and meristematic zones of roots, and in response to sugar (glucose and sucrose) stresses in transgenic tobaccos. The results demonstrated the spatial and temporal regulation of a cotton promoter in a model plant, and provided a useful foundation to understand the biological roles of this glucuronosyltransferase in cotton fiber development. To demonstrate the contribution of the various identified elements in the promoter to tissue-specific and inducible gene expression, further study by loss of function analysis is needed including deletions analysis, linker scanning and/or point mutations.