OsERF2 controls rice root growth and hormone responses through tuning expression of key genes involved in hormone signaling and sucrose metabolism

Root determines plant distribution, development progresses, stress response, as well as crop qualities and yields, which is under the tight control of genetic programs and environmental stimuli. Ethylene responsive factor proteins (ERFs) play important roles in plant growth and development. Here, the regulatory function of OsERF2 involved in root growth was investigated using the gain-function mutant of OsERF2 (nsf2857) and the artificial microRNA-mediated silenced lines of OsERF2 (Ami-OsERF2). nsf2857 showed short primary roots compared with the wild type (WT), while the primary roots of Ami-OsERF2 lines were longer than those of WT. Consistent with this phenotype, several auxin/cytokinin responsive genes involved in root growth were downregulated in nsf2857, but upregulated in Ami-OsERF2. Then, we found that nsf2857 seedlings exhibited decreased ABA accumulation and sensitivity to ABA and reduced ethylene-mediated root inhibition, while those were the opposite in Ami-ERF2 plants. Moreover, several key genes involved in ABA synthesis were downregulated in nsf2857, but unregulated in Ami-ERF2 lines. In addition, OsERF2 affected the accumulation of sucrose and UDPG by mediating expression of key genes involved in sucrose metabolism. These results indicate that OsERF2 is required for the control of root architecture and ABA- and ethylene-response by tuning expression of series genes involved in sugar metabolism and hormone signaling pathways. Electronic supplementary material The online version of this article (doi:10.1007/s11103-015-0416-9) contains supplementary material, which is available to authorized users.


Introduction
Root systems are central to the acquisition of water and nutrients, plant anchorage, seedling vigor, and responses to various stresses, which have pivotal effects on crop distribution, maximal productivity and yield stability, particularly in arid regions. Root architecture is under the tight control of genetic programs and environmental stimuli. It is well known that all phytohormones synergistically or antagonistically control root growth and development (Cuesta et al. 2013;Jung and McCouch 2013). In Arabidopsis, numerous crucial genes and intricate signaling pathways involved in root development have been well clarified, but many plant species also show their unique molecular and cellular regulatory mechanisms governing root development (Coudert et al. 2010).
Rice (Oryza sativa) is one of the most common crops worldwide. A series of mutants, genes, and genetic mechanisms governing root development have been characterized in rice, some of which are controlled by auxin and Guiqing Xiao and Hua Qin have contributed equally to this work.  (Mai et al. 2014). For examples, OsAUX1 controls auxin-mediated lateral root initiation through mediating polar auxin transport in rice (Yu et al. 2015). Mutation of OsXXT1 causes abnormal root hair development with the reduction of XyG content and the tensile strength of the cell wall . Increasing auxin-related mutants including crl1 (crownless root1), arl1 (adventitious rootless1), crl4/osgnom1, and oscand1, exhibit abnormal root formation and growth due to the disorders of auxin biosynthesis, polar transport or signaling transduction in rice (Inukai et al. 2005;Liu et al. 2005Liu et al. , 2009Wang et al. 2011). Moreover,auxin-cytokinin crosstalk plays curial roles in the control of root growth and development in rice. For examples, auxin-induced CRL5 promotes crown root initiation through repressing cytokinin signaling (Kitomi et al. 2011). OsCKX4 and EL5 are essential for crown and lateral root development by mediating crosstalk between auxin and cytokinin pathways (Gao et al. 2014;Koiwai et al. 2007;Mochizuki et al. 2014). In addition, OsMPK3/6 and OsMKK4/5 are involved in the regulation of root architecture through tuning the interplay of auxin and cytokinin (Singh et al. 2015). Sugar signals are central in determining plant growth and development by interacting with other signalling pathways (Lastdrager et al. 2014). In Arabidopsis, sucrose affects root architecture through regulating endogenous flavonols accumulation, which can suppress the effect of ABA on root growth (Nguyen et al. 2013). In rice, ABA mediates grain-filling rate by controlling the activities of key enzymes involved in sucrose-to-starch conversion in spikelet Zhu et al. 2011). Moreover, several genes involved in sucrose synthesis and metabolism or transport including OsSPS1, OsSUT1, OsSUT2, OsCYT-INV1, and OscFBP1 have been proved to be required for the control of rice growth and development (Eom et al. 2011;Hirose et al. 2010Hirose et al. , 2014Jia et al. 2008;Lee et al. 2008). For examples, mutation of OsSUT2 affected sucrose and other sugars transport and caused growth retardation including tiller number, plant height, and root dry weight (Eom et al. 2011). Disruption of OsCYT-INV1 impaired normal sucrose metabolism and root growth in rice (Jia et al. 2008).
Recent researches have proved that ethylene and its interaction with ABA play important roles in the control of rice growth and development. For examples, MHZ4/ABA4and MHZ5-mediated ABA signaling is required for ethylene-induced inhibition of root growth (Ma et al. 2014;Yin et al. 2015). Moreover, mhz6/Oseil1 and mhz7/ein2 exhibit longer root, while overexpression of MHZ6 and MHZ7 represses rice root elongation (Ma et al. 2013;Yang et al. 2015). Importantly, the inducible expression of OsERF2 (Os06g08340) by ethylene in roots was impaired in these mutants (Ma et al. 2013(Ma et al. , 2014Yang et al. 2015;Yin et al. 2015). In this paper, we found that OsERF2 was required for the regulation of primary root growth in rice. The gainof-function mutation of OsERF2 (nsf2857) shows short root, while artificial microRNA-OsERF2 (Ami-OsERF2) lines exhibit longer root, this phenotype is consentient with that of OsEIL1. At the transcript level, OsERF2 negatively regulates expression of candidate genes including ARF1, MKK4/5, MPK3/6 and PIN1b/2/9, which are involved in root growth and development through mediating auxin and cytokinin signaling. Furthermore, OsERF2 affects expression of key genes involved in ABA synthesis, which results in the enhanced accumulation of ABA in Ami-OsERF2 lines associated with root growth hypersensitivity to ABA, while those are opposite in nsf2857. Moreover, OsERF2 influence accumulation of sucrose and UDPG through mediating expression of genes related to sucrose metabolism, including INV (Os01g0332100) and OsCYT-INV1, OsSPS2, OsSUS3 and OsSUS6. Particularly, OsCYT-INV1 is closely related to the accumulation of sucrose and root growth in rice (Jia et al. 2008). Thus, our results indicate that OsERF2 is required for the control of root architecture and ABA response by regulating expression of pivotal genes involved in root development, ABA synthesis and sucrose metabolism.

OsERF2 negative regulates primary root growth
Recent studies showed that the induction of OsERF2 by ethylene was despaired in several ethylene-insensitive mutants mhz4-7 with longer primary roots . To investigate whether OsERF2 is involved in the regulation of root growth, we obtained the gain-of-function mutant of OsERF2 (nsf2857) with T-DNA insertion in its promoter from the stock seeds in SHIP (http://ship.plant signal.cn), and generated the artificial microRNA-mediated silencing lines of OsERF2 (Ami-ERF2). After checked expression of OsERF2 in different materials by qPCR ( Figure S1), three Ami-OsERF2 lines and nsf2857 were chosen for further analyses. We first compared the root growth status of these materials under normal conditions. For 7-day seedlings, the whole root system of nsf2857 is inferior to that of ZH11. Particularly, the primary root system of nsf2857 (about 6.97 cm) was significant shorter than that of ZH11 (about 10.49 cm). However, three Ami-OsERF2 lines displayed longer primary roots (about 10.20 cm) compared with Nip (about 7.04 cm) (Fig. 1a, b). These results imply that OsERF2 negatively controls primary root growth, and this phenotype is consistent with that of OsEIL1 ( Figure S2). Combined its inducible expression by key components of ethylene pathway such as OsEIN2 and OsEIL1, we speculate that OsERF2 might act as downstream regulator of ethylene signaling and is required for root growth and development.

OsERF2 regulates expression of key genes related to root architecture
To analyze the transcriptional regulation of OsERF2, we compared transcript levels of key genes involved in root growth between in nsf2857, Ami-OsERF2 and WT seedlings. As shown in Fig. 2, expression of 11 candidate genes was downregulated in nsf2857, but upregulated in Ami-OsERF2 lines. Among these genes, ARF1, MKK4/5, MPK3/6 and PIN1b/2/9 are involved in root growth through mediating auxin or cytokinin signaling. Moreover, OsKASI is required for rice root development by regulating fatty acid synthesis (Ding et al. 2015). RHL1, a homologs of Arabidopsis RHL1 (ROOT HAIRLESS 1), is required for root hair initiation (Schneider et al. 1998). These results imply that OsERF2 negatively regulates primary root growth partly due to mediating expression of genes involved in auxin/cytokinin signaling or key functional genes closely related to root development.

OsERF2 affects ABA accumulation and root growth response to ABA and ethylene
It has been documented that endogenous ABA positively regulates early root growth in rice (Chen et al. 2006), and the basal levels of endogenous ABA are required for the maintenance of normal root elongation in rice (Ma et al. 2014). In this study, we compared the content of ABA in Ami-OsERF2, nsf2857 and WT plants. Compared with WT, the content of ABA was decreased in nsf2857, but increased in Ami-OsERF2s (Fig. 3a), suggesting that OsERF2 negatively regulates ABA accumulation. In rice, Fig. 1 OsERF2 affects primary root growth. a Comparison of primary root length of 7-day-old seedlings between ZH11 and nsf2857, as well as Nip and three Ami-OsERF2 lines with similar phenotypes, bar 1 cm. b Statistical analyses (t test) were performed with graphpad prism 6 for the primary root length of ZH11 and nsf2857, as well as Nip and Ami-OsERF2. The seedlings of each sample were more than 20. Error bars represent SD. **P \ 0.01 MZH4 (ABA4) and pre-harvest sprouting genes (PHSs), including OsPDS, OsZDS, OsCRTISO and b-OsLCY, are essential for the synthesis of ABA (Fang et al. 2008;Ma et al. 2014). Here, we found that expression of these genes was downregulated in nsf2857 (Fig. 3b), but upregulated in Ami-OsERF2 lines (Fig. 3c). These transcriptional changes were consistent with the accumulation of ABA in the corresponding materials. We then analyzed root growth response to ABA and ethylene treatment in Ami-OsERF2, nsf2857 and WT. When treated with 0.5 and 1 lM ABA, the root length of ZH11 decreased from 10.4 to 8.5 and 6.9 cm, respectively, while the root length of nsf2857 was little changed in the presence of ABA (Fig. 4a, b), indicating that upregulation of OsERF2 caused root insensitivity to ABA in nsf2857. In contrast, down regulation of OsERF2 in Ami-OsERF2 plants resulted in the hypersensitivity of root growth to ABA. As shown in Fig. 3, the average primary root length of three Ami-OsERF2 lines (about 11.41 cm) was longer than that of Nip (8.33 cm) in the absence of ABA. However, the primary root length of Ami-OsERF2 lines decreased to 5.61 and 2.51 cm under the treatment with 0.5 and 1 lM ABA, which were shorter than that of Nip (about 7.63 and 3.78 cm) instead, respectively (Fig. 4c, d). The above results suggest that OsERF2 negatively regulates ABA synthesis and root growth response to ABA, which also confirm that the basal level of endogenous ABA is required for the maintenance of normal root elongation in rice.
Upon the ethylene treatment, the primary root growth of ZH11 was significantly inhibited by 64.2 % compared to that in air, but only partially inhibited by 25.7 % in nsf2857. Figure 5a, b, indicating that nsf2857 showed less ethylene-inhibited root growth. In contrast, the root growth of Ami-OsERF2 lines was inhibited by more 71 % compared to that in air, and significantly higher than that of Nip (56 %) (Fig. 5a, b), suggesting that down regulation of OsERF2 resulted in stronger ethylene-inhibited root growth.

OsERF2 transcriptionally regulates accumulation of sucrose and UDPG
Sucrose play central roles in the regulation of plant root development by generating a range of signal molecules and interacting with various hormones (Ljung et al. 2015).
Here, the contents of sucrose and UDPG (a key precursor for sucrose metabolism) were compared between nsf2857, Ami-OsERF2 and WT plants. Compared with corresponding WT, the sucrose of nsf2857 increased to 112 %, while that of Ami-OsERF2 lines decreased to about 84.7 % (Fig. 5a). In contrast to sucrose, the UDPG of nsf2857 (22.3 lg/mg) was 88 % of ZH11 (25.1 lg/mg), while that of Ami-OsERF2 (about 26.2 lg/mg) was higher than that of NiP (24.2 lg/mg) (Fig. 5b), suggesting that OsERF2 is involved in regulation of sucrose metabolism. Then, we checked transcripts of several key genes involved in sucrose metabolism including INV (Os01g0332100) and OsCYT-INV1 (encoding neutral invertase), OsSPS2 (encoding sucrose phosphate synthase), OsSUS3 and OsSUS6 (encoding sucrose synthase) between nsf2857, Ami-OsERF2 and WT plants. As shown in Fig. 5c, d, the expression of candidate genes was upregulated in Ami-OsERF2 plants, but downregulated in nsf2857. Particularly, OsCYT-INV1 has been reported to be closely related to the accumulation of sucrose and root growth in rice (Jia et al. 2008). Thus, these results suggest that OsERF2 transcriptionally regulates the accumulation of sucrose and UDPG, which partially contribute to its regulatory roles in rice root growth.

Discussion
The root system is essential for plants to absorb nutrients and water from soils, which determines plant development progress, response to drought and salinity, as well as crop qualities and yield. Rice is very susceptible to water deficit for its shallow root architecture, but the control of root system can increase its yield under drought conditions (Uga et al. 2013). Recently, many genes and mutants involved in root growth and development have been characterized in rice. Some genes are involved in auxin and cytokinin biosynthesis, transport and homeostasis, and their signal transduction (Azizi et al. 2015;Balzan et al. 2014;Coudert et al. 2010;Mai et al. 2014). For example, the mutation of CRL5, encoding an ERF transcription factor, impairs crown root initiation through repression of cytokinin signaling (Kitomi et al. 2011).
Although auxin and cytokinin govern plant root architecture, the interplay between ethylene and ABA is required for the control of root growth in rice. For examples, etiolated seedlings of Oseil1, Osein2, aba4/mhz4 and mhz5 exhibited longer primary roots, while their overexpression resulted in shorter roots, respectively. Moreover, the induction of OsERF2 by ethylene was disrupted in the roots of these mutants (Ma et al. 2013(Ma et al. , 2014Yang et al. 2015;Yin et al. 2015). In the present study, we evidenced that the gain-of-function mutant of OsERF2 (nsf2857) exhibited shorter primary root compared to WT, while down regulation of OsERF2 by artificial microRNA (Ami-OsERF2) resulted in longer primary root, suggesting that OsERF2 has a similar regulatory effect on the primary root growth with OsEIN2, OsEIL1 and MHZ5. Thus, we conjecture that OsERF2 might act as a downstream component of ethylene pathway and play key roles in the regulation of root growth (Fig. 6). ERF family factors play multiple roles in the control of plant metabolism, growth and development, and stress response (Licausi et al. 2013). In rice, many ERFs have been reported to be involved in the regulation of response to various stresses by mediating multiple signaling pathways and expression of stress-related genes (Kazan 2015). Importantly, increasing ERF members are involved in the control of development progress in rice. For examples, OsEATB influences plant height, panicle length and branching by regulating gibberellin biosynthesis (Qi et al. 2011). MFS1, SNB and OsIDS1 control inflorescence architecture and floral meristem establishment through tuning expression of spikelet meristem genes (Lee and An 2012;Ren et al. 2013). Moreover, auxin-induced CRL5 promotes crown root initiation via repressing cytokinin signaling (Kitomi et al. 2011). In this study, our findings reveal that OsERF2 negatively regulates expression of pivotal genes related to root growth. Among of these candidate genes, OsMPK3/6, OsMKK4/5 and OsPIN 1b/9 are reported to be involved in root growth by tuning auxin and cytokinin signaling (Singh et al. 2015).
ABA affects root growth and development in rice very different from how it does in Arabidopsis, and endogenous ABA positively regulates early root growth and is required for the maintenance of root elongation in rice (Chen et al. 2006;Ma et al. 2014). Here, our results suggest that several genes involved in ABA synthesis, including MZH4, OsPDS, OsZDS, OsCRTISO and b-OsLCY, are downregulated in nsf2857 seedlings, but upregulated in Ami-OsERF2 seedlings. These transcription changes are consistent with the increased ABA accumulation in Ami-OsERF2, and the decreased ABA level in nsf2857. Importantly, it has been proved that MHZ4 is responsible for ABA biosynthesis and involved in the control of ethylene response, whose mutation and overexpression causes reduced and enhanced ethylene-inhibited root growth, respectively (Ma et al. 2014). In this paper, we found that Ami-OsERF2 lines exhibited longer primary roots and enhanced sensitivity to ABA-and ethylene-treatment, while these phenotypes were the opposite in nsf2857. Moreover, our results showed that upregulation of OsERF2 caused the decreased expression of MHZ4 and ABA accumulation in nsf2857, which displayed similar phenotypes with mhz4 including the shorter roots and reduced ABA-and ethylene-responses. However, Ami-OsERF2 plants exhibited similar phenotypes with transgenic rice overexpressing MHZ4. Taken together, we speculate that OsERF2 is required for the interplay between ethylene and ABA in control of rice root growth, which might be partially related with its transcriptional control of MHZ4 expression and ABA accumulation.
As the main assimilated carbon of photosynthesis, sucrose play central roles in plant growth and development, as well as stress response by generating series of sugars as metabolites or as signaling molecules interacting the cross talks of metabolic, hormonal, and stress signals (Ljung et al. 2015;Ruan 2014). In Arabidopsis, sucrose inhibits hypocotyl elongation and promotes root growth by tuning auxin signal and cotyledon-derived long-distance signal (Kircher and Schopfer 2012;Stokes et al. 2013). In rice, mutation of OsSPS1, OsSUT1, OsSUT2, and OscFBP1 causes growth retardation and abnormal pollen development by affecting sucrose synthesis or transporter (Eom et al. 2011;Hirose et al. 2010Hirose et al. , 2014Lee et al. 2008). Similarly, UDP-glucose (UDPG) is a key metabolite for the Fig. 5 OsERF2 mediates root growth response to ethylene. a Ethylene response phenotypes in ZH11, nsf2857, Nip and two Ami-OsERF2 lines. Rice seedlings were grown in dark for 3 days in the presence or absence of 10 ppm ethylene. Bar 1 cm. b Statistical analyses (t test) of primary root length were performed with graphpad prism 6. Values represent the means of three independent biological replicates. **P \ 0.01 synthesis of sucrose, polysaccharides, glycoproteins, glycolipids and myriads of glycosylated secondary metabolites (Decker et al. 2012). UGPase plays important roles in carbohydrate metabolism through catalyzing the reversible conversion of glucose-1-phosphate to UDPG, which is essential for plant growth, pollen development, and male sterility in rice and Arabidopsis (Chen et al. 2007;Park et al. 2010;Woo et al. 2008). In rice, expression levels of genes involved in sucrose metabolism are correlated with sucrose accumulation (Maruyama et al. 2014). In this paper, we found that upregulation of OsERF2 resulted in the decreased sucrose and the increased UDPG in nsf2857 plants, while it was the opposite in Ami-OsERF2 plants. Consistent with these physiological changes, expression of key genes closely related to sucrose metabolism including INV and OsCYT-INV1, OsSPS2, OsSUS3, and OsSUS6, were downregulated in nsf2857, but upregulated in Ami-OsERF2 lines. Particularly, mutation of OsCYT-INV1 has been proved to cause the enhanced sucrose level and the reduced root growth in rice (Jia et al. 2008). Consistent with this result, we found that OsCYT-INV1 was downregulated in nsf2857, which exhibited short primary root and increased accumulation of sucrose. Thus, these findings suggest that OsERF2 transcriptionally controls the accumulation of sucrose and UDPG, which might further trigger the complex alteration of carbohydrate metabolism and multiple signaling pathways, and partially account for the effect of OsERF2 on rice root growth.

Rice materials and treatments
In this study, rice (O. sativa L. subsp. japonica cv Zhonghua 11 (ZH11) and O. sativa L. subsp. japonica cv nipponbare (Nip) were used as the wild-type (WT) for nsf2857 and amiERF2, respectively. For root growth and ABA response, more 40 geminated seeds from all samples were planted in 1/2 Murashige and Skoog (MS) medium solution Fig. 6 OsERF2 transcriptionally influences the accumulation of sucrose and UDPG. a Sucrose content was increased in nsf2857, but decreased in Ami-OsERF2 lines. b UDPG content was decreased in nsf2857, but increased in Ami-OsERF2 lines. Statistical analyses (t test) of sucrose and UDPG contents were performed with graphpad prism 6. Values represent the means of three replicates. **P \ 0.01. c, d Relative expression of key genes involved in sucrose metabolism in nsf2857 and Ami-OsERF2 lines. OsACTIN1 was used as an internal control. Data represent the means of three repeats Plant Mol Biol (2016) 90:293-302 299 supplemented with 0, 0.5 and 1 lM ABA. After growth for 7 days, root lengths were measured and photographed. For measurement of ABA, sucrose and UDPG, geminated seeds from all samples were planted in 1/2 Murashige and Skoog (MS) medium solution for 2 weeks. All data are presented as means of three times. For ethylene treatment, rice seedlings were grown on a stainless steel sieve placed in an air-tight plastic box with 10 ppm ethylene and in air as control. The seedlings were incubated at 28°C in the dark for 3 days. Artificial microRNA (amiRNA-OsERF2) construct was generated using specific primer sets (Table S1) designed through programs available on the (http://wmd3.weigel world.org/), and cloned into the binary vectors pCAM-BIA5300. amiRNA-OsERF2 plants were generated by Agrobacterium-mediated transformation and examined by qRT-PCR with specific primers (Table S1). The seeds of oseil1 mutant and OsEIL1-OX were obtained from Dr. Jinsong Zhang.

Measurements of ABA, sucrose and UDPG
ABA extraction was according to Zhu et al. (2010) measured using the ELISA kit (CSB-E09159Pl, China Agricultural University, China). 0.2 g leaves of 2-week-old seedlings were homogenized and extracted overnight at 4°C in 80 % methanol by vortex spinning. After centrifuged and concentrated to approximately 50 ll, these extracts were suspended with 1 ml TBS buffer (25 mM Tris-HCl pH 7.5, 100 mM NaCl, 1 mM MgCl2, and 3 mM NaN3) and used for ELISA according to the manufacturer's instructions. ABA concentrations were calculated as ng/g fresh weight. Each measurement was replicated three times.
Sucrose and UDPG content was measured by HPLC according to Bahaji et al. (2015). 0.1 g 2-week-old seedlings were ground in liquid nitrogen. For measurement of sucrose, the powder was suspended in 1 ml of 90 % ethanol, left at 70°C for 90 min and centrifuged at 13,0009g for 10 min. For measurement of UDPG, the powder was suspended in 1 ml of 1 M HClO 4 , left at 4°C for 2 h and centrifuged at 13,0009g for 10 min. The supernatant was neutralized with K 2 CO 3 and centrifuged at 13,0009g for 10 min. Sucrose was determined by HPLC with UPLC WATERS Acquity system by gradient separation with BEH Amide 2.1 9 100 mm 1.7 lm. UDPG was measured by HPLC on a system obtained from Waters Associates fitted with ZORBAX Carbohydrate 4.6 9 250 mm 5 lm. Standard sucrose and UDPG were purchased from Sigma.

Gene expression analysis
After geminated seeds were planted in 1/2 MS medium solution for 7 days, total RNA of roots or leaves was isolated from all samples using TRIzol reagent (Invitrogen, Beijing, China) and used to synthesize first-strand cDNA with All-in-One TM First-Strand cDNA Synthesis Kit (Transgen, China). RT-qPCR was performed using genespecific primers (Table S1) with TransStart Green qPCR SuperMix (Transgen, China). OsActin1 was used as the internal control. All experiments were repeated three times, and the average was calculated.