The ectopic expression of Arabidopsis glucosyltransferase UGT74D1 affects leaf positioning through modulating indole-3-acetic acid homeostasis

Leaf angle is an important agronomic trait affecting photosynthesis efficiency and crop yield. Although the mechanisms involved in the leaf angle control are intensively studied in monocots, factors contribute to the leaf angle in dicots are largely unknown. In this article, we explored the physiological roles of an Arabidopsis glucosyltransferase, UGT74D1, which have been proved to be indole-3-acetic acid (IAA) glucosyltransferase in vitro. We found that UGT74D1 possessed the enzymatic activity toward IAA glucosylation in vivo and its expression was induced by auxins. The ectopically expressed UGT74D1 obviously reduced the leaf angle with an altered IAA level, auxin distribution and cell size in leaf tissues. The expression of several key genes involved in the leaf shaping and leaf positioning, including PHYTOCHROME KINASE SUBSTRATE (PKS) genes and TEOSINTE BRANCHED1, CYCLOIDEA, and PCF (TCP) genes, were dramatically changed by ectopic expression of UGT74D1. In addition, clear transcription changes of YUCCA genes and other auxin related genes can be observed in overexpression lines. Taken together, our data indicate that glucosyltransferase UGT74D1 could affect leaf positioning through modulating auxin homeostasis and regulating transcription of PKS and TCP genes, suggesting a potential new role of UGT74D1 in regulation of leaf angle in dicot Arabidopsis.

www.nature.com/scientificreports/ The leaf positioning is an important agronomic trait which is closely related to the agricultural yields and architecture. Most researches on leaf positioning were carried out in monocots (especially rice) and indicated that leaf positioning was controlled by the phytohormone brassinosteroid (BR) related signaling pathway 17-21 . Recently, Zhao and colleagues reported a mutant of IAA-amido synthetase (lc1-D) from rice 22 . This mutant showed exaggerated leaf angles because the cells at the lamina joint were stimulated to elongate. Further study showed that lc1-D mutant was particularly sensitive to exogenous BR and had significantly reduced the expression of BR biosynthetic genes, suggesting that LC1 may regulate rice leaf positioning through the interaction between auxin and BR 22 . For dicots, several researches showed that the leaf positioning was mostly related to light signaling and auxin signaling [23][24][25][26] . For example, the Arabidopsis PHYTOCHROME KINASE SUBSTRATE (PKS) family and TEOSINTE BRANCHED1/CYCLOIDEA/PCF (TCP) family were demonstrated to be required for the regulation of leaf development and leaf positioning [26][27][28] . Auxin biosynthesis and polar transport were also demonstrated to be involved in the leaf positioning control [23][24][25] . As mentioned above, auxin glucosylation plays an important role in the regulation of plant growth and development. However, whether the auxin glucosylation plays a role in the case of leaf positioning is largely unknown.
Previously, we identified the glycostransferase UGT74D1, which was found to have the glucose conjugating activity toward both IAA and its precursor IBA, with a little preference toward the latter 29 . In this study, we characterized the growth responses of ectopically expressed UGT74D1 transgenic plants and mutants. It was found that the increased UGT74D1 activity substantially altered auxin distribution in leaf primordial and resulted in accumulation of free IAA in leaves, which then dramatically stimulated cell elongation and led to BR independent change of leaf positioning, possibly by a feedback transcription regulation of PKS and TCP factors. Our work provides evidences for the link between auxin glucosylation, auxin homeostasis and leaf positioning in dicots, highlighting a distinct role of UGT74D1 from other auxin glycosyltransferases identified so far.

Results
Expression of UGT74D1 was developmentally regulated and induced by auxins. To explore the physiological role of UGT74D1, we first investigated whether the expression of UGT74D1 gene is responsive to auxins, since UGT74D1 recombinant protein can catalyze the glucose conjugation of auxins as indicated by our previous work 29 . Two types of natural auxins, IAA and IBA, were used in this experiment. The results of qRT-PCR analysis indicated that both IAA and IBA could induce the expression of UGT74D1 within different treatment duration from 1 to 24 h (Fig. 1), suggesting that UGT74D1 might function in the auxin homeostasis.
In addition, the localization of auxin activity and UGT74D1 expression in wild type plants was also investigated by pDR5::GUS and UGT74D1 promoter::GUS constructs. Interestingly, our analyses indicated that the localization of pDR5::GUS activity and UGT74D1 promoter::GUS activity was nearly overlapping at cotyledon, hypocotyl, root, root apex, leaf and leaf edge (Fig. 2). Thus, the coincidence among the zone of auxin activity and the mostly expressed site of UGT74D1 gene was particularly meaningful for role of UGT74D1 exerted in planta, suggesting a relevance of the UGT74D1 expression in mediating auxin homeostasis.
Histochemical specific staining of GUS activity in the UGT74D1 promoter::GUS transgenic lines showed that UGT74D1 was strongly expressed in cotyledons during early germination ( Fig. 3A-D). In the subsequent vegetative growth stage, UGT74D1 was expressed mainly in young leaves, leaf veins ( Fig. 3E-H). Importantly, UGT74D1 was strongly expressed in young leaf petioles (Fig. 3I), which implies the biological function of UGT74D1 in leaf petiole development. During the reproductive stage, UGT74D1 was highly expressed in flowers, young siliques and veins of cauline leaves ( Fig. 3J-M). These results indicate that the expression pattern of UGT74D1 is spatio-temporally regulated. UGT74D1 was localized in nucleus and cytoplasm. To investigate the subcellular localization of UGT74D1 protein, the plasmid 35S::74D1-GFP was transformed into Arabidopsis. The roots of the 5-day-old transgenic seedlings were detected for getting the fluorescent images. The distribution of green fluorescence signals of the UGT74D1-GFP fusion protein was similar as that of the control GFP protein, indicating that UGT74D1 protein was localized in both nucleus and cytoplasm (Fig. 4). Similar subcellular localization patterns of other UGTs in plants have also been observed in nucleus and cytoplasm, including PpUGT85A2 glycosylates linalool 30 , UGT85A1 glycosylates zeatin 31 , UGT87A2 involved in flower development regulation 32 , and UGT73C6 glycosylates brassinosteroids 33 . UGTs may also play a role in the nucleus to control the stability of nuclear receptor ligands and protect nuclear components from toxins 34,35 . UGT74D1 possessed activity toward IAA in planta. To investigate whether UGT74D1 glucosyltransferase has activity toward IAA in planta, at least four overexpression lines (74D1OE-11, -23, -24, -26) and two independent T-DNA insertion mutants (74d1ko-1, Salk_004870; 74d1ko-2, Salk_011286) were used in this study with the same line codes used previously 29 .
Using these overexpression lines and mutants, crude protein was extracted and glucosyltransferase activity toward IAA was analyzed using UDP-glucose as sugar donor. The data showed that overexpression lines with higher steady-state level of transcripts displayed higher enzyme activity toward IAA compared to wild type (Fig. 5A,B). However, the enzyme activity of mutants was not detected (data = 0). These results suggested that UGT74D1 activity toward IAA-glucose conjugation has been maintained in planta. UGT74D1 activity affected leaf positioning. Several previous studies reported that auxin glycosyltransferases could affect plant growth with curly leaves, compressed rosette, and shorter stature 14,15 . Here, UGT74D1 exhibited a distinct physiological relevance for leaves growth from other reported auxin glycosyltransferases. Although UGT74D1 knock-out mutants did not show obvious phenotype possibly because of func- www.nature.com/scientificreports/ tion redundancy, its overexpression lines displayed a clear change in leaf positioning (Fig. 6). The angle between the horizontal and the petiole of first true leaf was used as an index of leaf positioning and was measured. Our data showed that the petiole angle of UGT74D1 overexpressor plants was much smaller compared with the wildtype (Fig. 6A). To test whether the leaf positioning can be influenced by the light intensity or not, transgenic plants were grown in 30 μmol m −2 s −1 , 50 μmol m −2 s −1 and 100 μmol m −2 s −1 , respectively. It was found that the leaf petiole angle of UGT74D1 transgenic plants was always significantly smaller when compared to wild type and mutants in different light intensity (Fig. 6B). Because UGT74D1 was involved in glycosylation of IAA, this observation suggested that UGT74D1 might exert important influence on the petiole angle through changing auxin homeostasis.

UGT74D1 altered auxin level and auxin distribution of leaf petioles.
In order to know the possible influence of ectopically expressing UGT74D1 on the in vivo auxin level, the measurement of free auxins in leave petioles was performed. The most important native auxin, IAA, was detected in this research. As shown in Fig. 7A, it was found that the free form IAA was significantly increased compared to wild type, which suggests that auxin homeostasis was involved in leaf positioning and the IAA accumulation in leave petioles led to smaller petiole angle. To test this hypothesis, we used the polar auxin transport inhibitor NPA to chemically block auxin transport and disturb the auxin role. As shown in Fig. 7B, after treated with 10 μM NPA, both WT and UGT-74D1OE exhibited increased leaf inclination compared with control (mock-treated). However, the petiole angle of overexpressors was still smaller than that of wild type, indicating a relative insensitivity of overexpressors to NPA because of increased IAA level.
To investigate the possible mechanism that resulted in the change of leaf petiole angle, we used the artificial auxin-responsive DR5 promoter to monitor auxin signaling in leaf primordia 36,37 . Sections through shoot apices of Col-0/pDR5::GFP and 74D1OE/pDR5::GFP plants were imaged by confocal microscopy. Arabidopsis wild type plants harboring a pDR5::GFP construct revealed low GFP signals in leaf primordia, however, strong To test this hypothesis, the adaxial surface of leaf petiole was taken by scanning electron microscopy. As shown in Fig. 7D, cells on the adaxial side of UGT74D1OE leaf petiole were obvious longer compared to that of the WT, associated with decreased cell width. These observations suggested that the change of leaf petiole angle of UGT74D1OE was likely caused by the cell elongation on the adaxial surface of leaf petiole.
Previous studies have shown that one important aspect of the BR biological function is the regulation of leaf angle 18,19 . Accordingly, the sensitivities of UGT74D1OE lines to BR were examined. After UGT74D1OE lines were treated with 1 μM brassinolide (BL), one kind of active brassinosteroids, no obvious difference was found in leaf angle ( Supplementary Fig. S1A). In addition, we analyzed whether UGT74D1 transcription was induced by BR using the BR response factor DWF4 as control. Our results indicated that UGT74D1 transcription was not affected by BR (Supplementary Fig. S1B). Several BR related genes were also investigated for their expression  TCPs (TCP3, TCP4,  TCP10, TCP17, TCP24) 38,39 . It was found that the PKS2 transcript level was dramatically down-regulated in UGT74D1 overexpression plants (Fig. 8A), which is in good agreement with the leaf angle phenotype. Moreover, the TCP3, TCP10, TCP17 and TCP24 were modulated to a significantly up-regulated expression level in UGT74D1 overexpression lines (Fig. 8B).
In addition, to know the reason causing auxin accumulation in UGT74D1 overexpression lines, we tested the expression level of genes involved in auxin metabolic pathways. Although ugt74d1 mutants did not show a clear change in expression level of auxin related genes, UGT74D1 overexpression lines changed a lot in expression of several auxin related genes. The most important YUC genes that control IAA biosynthesis, including YUCCA2,  www.nature.com/scientificreports/ YUCCA6 and YUCCA10, were substantially up-regulated in overexpression lines, which was consistent with the accumulation of IAA in these transgenic plants (Fig. 9A). An auxin influx carrier, AUX1, was down-regulated in UGT74D1 overexpression lines compared with wild-type, which might be a result of IAA accumulation (Fig. 9B). UGT84B1 and UGT74E2 were two different glucosyltransferases toward auxins identified previously [13][14][15] . We also investigated the expression of these two UGT genes. It was found that both UGT84B1 and UGT74E2 were down-regulated in UGT74D1 overexpression lines, suggesting a functional redundancy of these auxin glucosyltransferases (Fig. 9B). IAMT1 was found to be a gene encoding methyltransferase which converts IAA to methyl-IAA (MeIAA) and its overexpression in Arabidopsis leads to a curly leaf phenotype and perturbed auxin homeostasis 40 . Our analysis of UGT74D1 transgenic plants indicated that IAMT1 was significantly upregulated in mutants but down-regulated in overexpression lines compared to wild type (Fig. 9B), implicating a link between glycosylation and methylation of auxins. Our data presented here suggested that the expression change of UGT74D1 gene caused the expression change of many genes in auxin metabolic pathways and thus perturbed the auxin homeostasis.

Discussion
Glycosyltransferase UGT74D1 catalyze the transfer of UDP-Glucose to IAA forming IAA-glucose. The free energy change of this reaction is positive and the energy of the acyl alkyl acetal bond between IAA and the aldehydic oxygen of glucose is above that of the phosphatogucose bond of UDPG 10,41 . This suggests that for the reaction to proceed, the level of UDPG must be significantly higher than that of IAA-glucose. Moreover, the accumulation of limited levels of IAA-glucose must be a second step of a transesterification. The energetics of these reactions makes it a candidate for regulatory control of IAA and IAA-glucose levels which can be looked at as a 'buffer' reaction. For example, the levels of IAA and 1-O-IAGlc track in parallel with each other in WT, low, medium and high UGT84B1 over-expression lines 14 . Similarly, a similar relationship between IAA and IAAglucose levels was showed in ectopic maize IAGLU gene expression in Arabidopsis 42 . A similar approach using antisense showed a parallel decrease in free IAA and IAA-glucose in transgenic tomato 43 . In our research, IAA www.nature.com/scientificreports/ and IAA-glucose levels also showed a parallel increase in UGT74D1 over-expression Arabidopsis lines, which was consistent with these prior studies. Hormone conjugation has been proposed to be significant contributors to hormone homeostasis. In plants, it is important to maintain appropriate hormone level in specific tissues and growth responses. In this research, an auxin-glucose conjugating enzyme, UGT74D1, was employed to explore the auxin homeostasis and the corresponding physiological responses in planta. We found that ectopically expressed UGT74D1 caused obvious homeostasis alteration of auxins and changed leaf petiole positioning. However, ugt74d1 mutants did not show clear changes in leaf growth and development. Why did ugt74d1 mutants not display obvious phenotypic or physiological change? We supposed that other auxin glucosyltransferases compensated the loss of function in ugt74d1 mutants. Besides UGT74D1, UGT84B1 and UGT74E2 were also identified as auxin glucosyltransferases of Arabidopsis [13][14][15] . Their physiological role might be partially overlapping, although ectopically expressed UGT84B1 or UGT74E2 gave a distinct phenotype 14,15 . In UGT74D1 overexpression lines, we observed that the expression level of UGT84B1 and UGT74E2 were substantially down-regulated, which also suggested at least partial overlapping physiological roles between these three auxin glucosyltransferases. However, UGT74D1 appears to have a distinct expression pattern. We observed that UGT74D1 was expressed in whole leaves and petioles in developing young leaves, while it subsequently expressed in leaf margin as leaves grew and matured, suggesting that UGT74D1 may have different physiological effects from other auxin UGTs.
Recently, OxIAA was reported to be another substrate of UGT74D1 which converts OxIAA to OxIAA-Glc 44 . OxIAA is a primary IAA catabolite formed by IAA oxidation in Arabidopsis. It is inactive in bioassays and in auxin signaling 45 . Our data showed that UGT74D1 overexpression lines had substantially increased free IAA. www.nature.com/scientificreports/ Even though UGT74D1 can catalyze the glucosylation of both IAA and OxIAA, the glucosylation of active auxin (IAA or its precursor IBA), rather than its inactive metabolite (OxIAA), might play direct and important role in modulating auxin homeostasis and leaf angle. Our experimental results indicated that the transcriptional regulation of three Arabidopsis YUCCA genes, YUCCA2, YUCCA6 and YUCCA10, was significantly enhanced in UGT74D1 overexpression lines, which may be the reason leading to IAA accumulation to very high level in leaf petiole. However, the mechanism how glucosyltransferase UGT74D1 triggers the expression of YUCCAs is unclear. Several changes were also observed in other metabolic pathways and gene expressions. For example, overexpression of UGT74D1 led to the downregulated transcription of IAA influx carrier AUX1, IAA glucosyltransferases UGT84B1 and UGT74E2, and IAA methyltransferase IAMT1. All these alterations may be a consequence of an increased IAA level in leaf petioles of UGT74D1 overexpression plants.
Leaf positioning (petiole phototropism) is one of the important agronomic traits affecting plant architecture and yield. Also, leaf positioning is one of the adaptive processes in response to environmental light signals. Increasing evidences demonstrated the importance of PKS protein family in regulating leaf positioning. PKS2 is found to be highly expressed in leaves 46,47 . The leaf position of the mutant pks2 is found to have less erect petioles than wild-type. Moreover, auxin transport assays in mesophyll protoplast indicate that PKS2 may regulate light responses by regulating auxin homeostasis 26 . However, the link between PKS protein and auxin has not been firmly established during leaf development. In this study, we found that ectopically expressed UGT74D1 resulted in clear leaf angle change. When monitoring the IAA distribution, we found that IAA was concentrated on the adaxial side of leaf primordia in UGT74D1OE lines. Consistently with this observation, cells in the adaxial side of www.nature.com/scientificreports/ UGT74D1OE leaf petiole were obvious longer than wild type and the IAA level in leave petioles of UGT74D1OE lines was also higher than wild type. These observations might suggest that UGT74D1 could modulate auxin homeostasis and asymmetric distribution in leaves, thus altering leaf angle. Moreover, our data showed that the expression of PKS2 was dramatically down-regulated in UGT74D1 overexpression lines. These findings suggested the possibility that auxin homeostasis modulated by glucosyltransferase UGT74D1 could provide feedback to PKS2 expression and then influence the leaf positioning in Arabidopsis. TCP family transcription factors are among the best-characterized regulators of leaf development 27,28 . In Arabidopsis, there were 13 class I TCPs and 11 class II TCPs 38 . TCPs play an essential role in the determination of leaf size and shape by regulating cell proliferation and differentiation. Besides, it is suggested that TCPs control leaf shape by promoting leaf maturation in a threshold activity manner 39 . Recently, a novel transcriptional repressor EAR motif protein1 (TIE1) which contain a TCP interactor was demonstrated to be a major modulator of TCP activities during leaf development 48 . It was supposed that the interaction of TIE1 and TCPs regulates the expression of auxin related genes and controls cell differentiation and leaf development 48 . In this research, we found that the ectopic expression of UGT74D1 led to a significant up-regulation for the transcription of TCP3, TCP10, TCP17 and TCP24. We supposed that a feedback circle between auxin pathway modulated by UGT74D1 glucosyltransferase and TCP pathway may exist. Considering the involvement of both PKS2 and several TCPs in the leaf positioning, our findings suggested that UGT74D1 represents a potentially unique paradigm in the regulation of leaf angle in Arabidopsis. However, the possible link between PKS protein and TCP protein remains to be answered in the case of leaf positioning.  Crude protein extraction and glucosyltransferase assay. Crude protein was extracted from 2-weekold transgenic seedlings according to Jackson et al. 14 . To investigate the glycosyltransferase activity of the crude protein extracts prepared from plant tissues, 50 mL crude protein extracts were incubated at 37 °C for 1 h according to Jin et al. 29 . The reaction mix was analyzed subsequently using reverse-phase HPLC following the method described by Jin et al. 29 .
Analysis of free IAA in leaf petioles. Leaf petioles of the 10-day-old seedlings of wild type Arabidopsis thaliana Col-0 and two UGT74D1 overexpressing lines (74D1OE-23, 74D1OE-24) were used for analysis of free IAA level. 2 cm proximal end of leaf petioles were collected in five replicates, weighed, immediately frozen in liquid nitrogen and stored at − 80 °C until extraction. Then frozen samples were ground in liquid nitrogen with mortar and pestle. IAA quantification was determined on ultra high performance liquid chromatographytriple quadrupole mass spectrometry (UPLC-MS/MS) with negative electrospray ionization mode and 100 pmol isotope-labeled 2 H 2 -IAA served as the internal standard as described by Fu et al. 51 .

Total RNA extraction and quantitative RT-PCR (qRT-PCR).
To study the expression level of leaf development related genes, 2-week-old seedlings were harvested for RNA extraction. For investigating whether UGT74D1 gene was induced by IAA and IBA, 2-week-old seedlings was first soaked with 10 μM IAA and IBA, respectively, for 0-24 h, then they were harvested for RNA extraction. Total RNA was extracted using Trizol reagent and was used as template for cDNA synthesis. The relative transcript level was normalized with Actin 2 gene according to the 2 −ΔΔCT method 52 .
Microscopy imaging. For the fluorescence images, LSM 700 confocal laser scanning microscope was used.
Images were obtained and processed using the ZEN 2009 software. For subcellular localization analysis, the UGT74D1 open reading frame without stop codon was amplified, and then inserted into p326-SGFP vector to generate the 74D1-GFP fusion gene driven by CaMV35S promoter 53,54 . 74D1-GFP fusion plasmid was transformed into Arabidopsis to get transgenic plants. The roots of the 5-day-old transgenic seedlings were detected using a confocal laser-scanning microscope at excitation wavelengths of 488 and 647 nm, respectively. Counterstaining of cell walls was achieved by mounting seedling roots in 10 μM propidium iodide. For auxin signaling analysis of shoot apices, the Col-0 seedlings harboring pDR5::GFP construct driven by CaMV35S promoter was cross-fertilized with wild-type and UGT74D1OE, respectively. The F1 generation was harvested to generate the heterozygous plants. Then Arabidopsis vegetative shoot apices were separated by removing older leaves and fluorescence signals were monitored directly using a confocal laser-scanning microscope.
Environmental scanning electron microscopy was performed for the cellular observation of leaf petioles. The basis of leaf petiole (1 cm long) was excised from first pair of true leaves (14-day-old plants) after second pair of true leaves emerged. Tissue was then transferred to peltier cooling stage (temperature setting = 5 °C). Precooled metal stubs with the samples were transferred to the cooling stage and images were recorded.