Underground gibberellin activity: differential gibberellin response in tomato shoots and roots

Abstract

Plant organ growth is governed and modified by developmental programs and environmental cues. In most cases, these changes are mediated by the activity of phytohormones (Bradford & Trewavas, 1994; Verma et al., 2016). Gibberellins (GAs) are growth promoting hormones that regulate many developmental processes, including organ growth and elongation (Davière & Achard, 2013). GA affects elongation by promoting cell division and expansion (Ubeda-Thomas et al., 2009). The nuclear DELLA proteins inhibit all GA-elongating responses (Locascio et al., 2013) and GA binding to the GIBBERELLIN-INSENSITIVE DWARF1 (GID1) receptor leads to DELLA degradation and activation of growth (Ueguchi-Tanaka et al., 2005, 2007). While GA plays a central role in stem elongation (Sun & Gubler, 2004), its general significance for root elongation is less clear, with numerous conflicting reports (Torrey, 1976; Feldman, 1984; Phinney, 1985; Tanimoto, 2005; Tanimoto & Hirano, 2013). It is well established that Arabidopsis (Arabidopsis thaliana) root elongation is regulated by GA. Several studies demonstrate the central role of GA in Arabidopsis primary root elongation (Ubeda-Tomás et al., 2008, 2009; Achard et al., 2009; Rizza et al., 2017). The Arabidopsis GA deficient mutant ga1-3 exhibits shorter primary root, which is rescued by GA application or loss of DELLA activity (Fu & Harberd, 2003). Similarly, the loss of all three GA receptors strongly inhibit primary root elongation (Griffiths et al., 2006). Rizza et al. (2017) showed that endogenous bioactive GA levels correlate with cell length in Arabidopsis roots. Ubeda-Tomás et al. (2008) showed that inhibiting GA signaling specifically in the endodermis of Arabidopsis roots is sufficient to disrupt root elongation, indicating that the endodermis is the key site for GA action in the regulation of root elongation. This was supported by Shani et al. (2013) who found the accumulation of exogenous bioactive tagged-GAs in the endodermis of the elongation zone. Tanimoto & Hirano (2013) suggest that roots are very sensitive to GA and therefore respond to extremely low GA concentrations. For instance, while root elongation of the ga1-3 Arabidopsis mutant was strongly induced by low concentration of GA4 (10−10 M), this treatment had no effect on leaf expansion (Arizumii et al., 2008). In other plant species however, the role of GA in root elongation is unclear; while Whaley & Kephart (1957) show that GA application to maize (Zea mays) promotes root elongation, Svensson (1972) reported that the hormone has no effect on maize root growth. In Medicago (Medicago truncatula), GA treatment reduced, and GA-biosynthesis inhibitor paclobutrazol (PAC) increased, primary root length (Fonouni-Farde et al., 2019), suggesting that GA inhibits root elongation, at least at higher concentrations. In tomato (Solanum lycopersicum), Butcher & Street (1960) show that elevating GA concentrations progressively promote root elongation, whereas Tognoni et al. (1967), showed that application of GA inhibits root elongation in a concentration-dependent manner. Barlow et al. (1991) demonstrate that roots of the GA-deficient mutant gib-1, grown in vitro, are shorter than wild-type (WT), but this phenotype was not rescued by GA3 application. Thus, the role of GA in root elongation in species other than Arabidopsis, remains ambiguous. To elucidate the role of GA in tomato primary-root elongation, we took advantage of the novel genetic resources, developed in our previous study (Illouz-Eliaz et al., 2019). We first examined primary root elongation in the GA deficient mutant gib-2 (Koornneef et al., 1990). Wild-type plants and gib-2 mutants (both on cv Moneymaker background) were grown in vermiculite in a growth room set to a photoperiod of 12 h : 12 h, night : day with light intensity (cool-white bulbs) of ~250 μmol m−2 s−1 and 25°C. After 2 wk, plants were repeatedly treated with 10−5 M GA3 and 2 wk later we measured stem and root length. Untreated gib-2 stems were a third of the length of WT. Gibberellin application induced strong stem elongation in both genetic backgrounds, and their final length was similar (Fig. 1a). Nontreated gib-2 primary roots, however, were not significantly shorter than those of the WT and GA treatment had only a very mild effect on root length (Fig. 1a). This observation suggests that either GA has no effect on tomato root elongation, or that root elongation is highly sensitive to GA and reaches saturation at very low levels of GA. Since gib-2 exhibits residual GA activity (Illouz-Eliaz et al., 2019), this may be sufficient to allow normal root growth. We therefore tested primary root elongation in the gid1 triple (gid1TRI) mutant that lacks any GA activity due to the loss of all three GA receptors, GID1a, GID1b1 and GID1b2 (Illouz-Eliaz et al., 2019). We first measured the elongation rate of primary roots and hypocotyls following germination on MS plates. Hypocotyl elongation rate of gid1TRI was 10-times lower than that of WT (both on cv M82 background, Fig. 1b), but root elongation rate of the mutant was only three-times lower (Fig. 1c). After 10 d, primary roots of gid1TRI were c. half of the length of the WT, whereas the mutant hypocotyls were a tenth of the length of the WT (Fig. 1c). Root-to-shoot ratio was c. four times higher in gid1TRI (Fig. 1d). Since the gid1TRI exhibits very slow growth, we also examined mature plants of the same physiological age (similar number of leaves in gid1TRI and WT) grown in a glasshouse under natural day-length conditions with light intensity of 700 to 1000 µmol m−2 s−1 and 18–29°C. Primary root length of gid1TRI were half of the WT, whereas gid1TRI stems were a tenth of the WT (Fig. 1e). Thus, the lack of GA activity, strongly affects shoot development, but only partially affects primary root elongation. We cannot exclude the possibility that the inhibition of root elongation in gid1TRI was a result of limited assimilate supply by the extremely small canopy. We further tested the effect of exogenous GA on root and stem elongation. To this end, 14-d-old WT (cv M82) seedlings grown in vermiculite as described earlier, were treated with 2 mg l−1 PAC followed by the application of elevating GA3 concentrations (from 10−7 M to 10−3 M). After 15 d, we measured stem and primary root length. WT stems exhibited a strong and increased elongation response to rising GA3 concentrations, reaching saturation at very high concentrations (above 10−5 M, Fig. 1f). By contrast, primary root length was hardly affected and exhibited a bell-curve response (Fig. 1g). To examine how GA affects cell length, we treated WT seedlings with PAC or PAC with GA3 (10−6, 10−4 and 10−3 M), and after 10 d we measured epicotyl and primary root epidermal cell length using confocal microscopy. Stem epidermal cell elongation strongly responded to GA treatments and cell length of stems treated with PAC and 10−3 M GA3 were four times longer than those treated only with PAC (Fig. 1h). Primary root epidermal cells exhibited a very mild response to PAC and GA3 (Fig. 1i) and the effect of the hormone was saturated already at 10−6 M. GA-treated cells were only 1.2 times longer than the PAC treated cells. It is worth noting that higher GA concentrations did not inhibit root cell length, suggesting that reduced cell number may be the cause for inhibition of root length in high GA concentration. It was previously suggested that roots are more sensitive to GA than stems (Tanimoto & Hirano, 2013). To further test root sensitivity to GA, we analyzed the response of various known GA-regulated genes, including GA biosynthesis (GA 20-OXIDASEs and GA 3-OXIDASE) and signaling (GID1s) genes that are downregulated by the hormone (Middleton et al., 2012; Illouz-Eliaz et al., 2019) and the GA-induced gene GIBBERELLIC ACID STIMULATED TRANSCRIPT1 (GAST1; Shi et al., 1992). Seedlings, grown on vermiculite as described earlier, were treated with 2 mg l−1 PAC for 10 d and then with several GA3 concentration (10−8 to 10−5 M), and 3 h later, gene expression in elongating stems and roots was analyzed by quantitative reverse transcription polymerase chain reaction (qRT-PCR) as described by Illouz-Eliaz et al. (2019). We found a clear and significant response to 10−7 M, but not to 10−8 M GA3, for all genes, in both elongating stems and roots (Fig. 2a–f), suggesting similar sensitivity to GA. Moreover, the intensity of the molecular response in roots was weaker and saturated at lower concentrations than in stems. These results are consistent with our cell elongation results (Fig. 1h,i). We previously showed that GID1a is the dominant GA receptor in tomato stems, due to its high affinity to DELLA and the fact that it is not inhibited by the feedback response to GA (Illouz-Eliaz et al., 2019). Its presence alone, in the absence of GID1b1 and GID1b2 activity, can induce the strong stem-elongation response to exogenous GA that is saturated only at very high concentrations. GID1b1 and GID1b2 exhibit a weak GA response that is saturated at low concentrations. To examine if the differential response to GA in stems and roots results from differential activity of the different GID1s, we tested the effect of GA application on the three tomato gid1 double mutants. The GID1 family in tomato is composed of three members, therefore each double mutant contains one active GID1. In stems, only gid1b1 gid1b2 with active GID1a exhibited strong elongation response, similar to WT (Fig. 2g; Illouz-Eliaz et al., 2019). By contrast, roots of this double mutant did not show elongation response (Fig. 2h), suggesting that GID1a does not promote strong GA response in roots as it does in stems. Previously we showed that GID1a and GID1b1 are highly expressed in elongating stems, while GID1b2 exhibits relatively low expression (Illouz-Eliaz et al., 2019). We analyzed available public data of tomato root transcriptome (Koenig et al., 2013; Zouine et al., 2017; Góra-Sochacka et al., 2019; Gray et al., 2020) and found very low expression of GID1a compared to GID1b1 and GID1b2 in all datasets (Fig. 2i presents the data analyzed from Góra-Sochacka et al., 2019). This raises the possibility that the minor effect of GA on root elongation is caused by low GID1a expression in roots. Further research regarding the activity of GID1a as a disjunctive component of GA signaling in aboveground and belowground organs, is thus highly warranted. An example for such studies could be exploring the effect of highly expressed GID1a under a root-specific promoter. To conclude, our results in tomato suggest that while GA plays a central role in the regulation of shoot elongation, it has only a minor effect on root elongation. Although previous studies suggest that roots are more sensitive to GA than shoots, this is probably not the case in tomato. We found however that very low GA concentrations are sufficient to saturate root elongation, but not stem elongation. Plants adjust root-to-shoot ratio to adapt to changes in the environment. Under drought conditions the root-to-shoot ratio increases to reduce transpiration and increase water uptake (Xu et al., 2015). GA accumulation is inhibited under osmotic stresses, such as drought and salinity (Achard et al., 2006; Colebrook et al., 2014). Thus, the reduced GA levels strongly affects shoot but not root growth. It is possible that this mechanism evolved as a strategy to modify the root-to-shoot ratio under stress conditions. This research was supported by a research grants from the Israel Ministry of Agriculture and Rural Development (12-01-0014), the Israel Ministry of Agriculture and Rural Development (Eugene Kandel Knowledge Center) as part of the ‘Root of the Matter’ – the root zone knowledge center for leveraging modern agriculture, and by the Israel Science Foundation (grant no. 617/20) to DW. The authors thank Prof. Naomi Ori, Dr Idan Efroni and Prof. Eilon Shani for helpful discussion.

Topics & Concepts