The Columbia ecotype of Arabidopsis thaliana was used as a wild-type (WT) strain. Plants were grown on 0.8% agar-solidified half concentration Murashige and Skoog (MS) medium at 23°C in growth chambers with continuous light illumination (40–50 μmol photons m^–2 s^–1). After 14 days, plants were transferred to the soil (vermiculite:promix ratio of 1:1) and grown under continuous light illumination of 40–50 μmol photons m^–2 s^–1. Water containing 0.03 mM KNO₃ and different amount of estrogen (0, 0.1 or 1.0 μM) was applied to the soil. β-estradiol (estrogen) (Sigma-Aldrich) was dissolved in dimethyl sulfoxide (DMSO) at a concentration of 1.0 and 10 mM, and diluted by 10,000-times with water for 0.1 or 1.0 μM estrogen conditions, respectively. As for 0 μM estrogen conditions, water containing 0.01% DMSO was used.

Transgenic Arabidopsis lines expressing chloroplast- and/or mitochondrion-targeted YjbM were constructed as follows. First, DNA fragments encoding the chloroplast transit signal of recA (At1g79050) and the mitochondrion transit signal of γ-ATPase (At2g33040) were separately amplified by PCR using the following primer pairs: 5’-GAAGCTAGTCGACTCTAGCCATGGATTCACAGCTAGTCTTG-3’ and 5’-TCCCACTGTTTGTCATCCATGTCGCGATCGAATTCAGAACTG-3’ for the chloroplast signal, and 5’-GAAGCTAGTCGACTCTAGCCATGGCAATGGCTGTTTTCCGT-3’ and 5’-TCCCACTGTTTGTCATCCATGTCGCGGTTCTTAACACTCTTCATGCG-3’ for the mitochondrion signal. Genomic DNA from WT Arabidopsis was used as a template. Previously, we constructed an estrogen-inducible YjbM-expressing plasmid construct (Ihara and Masuda, 2016) using the pER8 plasmid (Zuo et al., 2000). An inverse PCR was performed with the pER8-yjbM plasmid (Ihara and Masuda, 2016) as a template to amplify the yjbM-coding region along with the plasmid backbone of pER8 using the primer pair 5’-ATGGATGACAAACAGTGGGAA-3’ and 5’-GGCTAGAGTCGACTAGCTTCA-3’. The DNA fragments encoding the chloroplast and mitochondrial transit peptides were then separately mixed with the inverse PCR fragment and connected using In-Fusion cloning reagent (Clontech). The resulting pER8-based plasmids, encoding yjbM fused with the chloroplast and mitochondrial target signals, were named pER8-cpYjbM and pER8-mtYjbM, respectively. After verifying the correct sequences of the inserted DNA, pER8-cpYjbM and pER8-mtYjbM were separately introduced into the WT Arabidopsis using standard Agrobacterium-dependent transformation methods. Homozygous lines were isolated in the second or third generations and used for further analysis.

Protein samples from each plant were electrophoresed through an SDS-PAGE gel (12% w/v acrylamide) and electroblotted onto a polyvinylidene fluoride membrane (GE Healthcare). Protein concentrations were determined by Lowry assay kit (BioRad), and the same amount of the protein was loaded in each lane. Membrane-bound proteins were immuno-probed with antibodies against anti-FLAG M2 (Thermo Fisher 31430), and detected using ECL Plus Western Blotting Detection kit reagents (GE Healthcare). ppGpp quantification was done by a modified method described previously (Ihara et al., 2015; Ito et al., 2020).

All experiments were designed and performed with at least three independent replicates. The collected data were calculated and the graphs and charts constructed using Microsoft Excel and R software. The statistical differences between the transgenic lines and WT were confirmed using the Dunnett test or t-test. Other details such as sample sizes are shown in each figure legend.

The ppGpp synthase activity of most ppGpp synthases, such as E. coli RelA and SpoT, is regulated at posttranslational levels (Potrykus and Cashel, 2008). However, the Bacillus subtilis ppGpp synthase YjbM, which consists solely of a ppGpp synthase domain without any activity control domains, is known to continuously synthesize ppGpp upon expression (Nanamiya et al., 2008). In this study, YjbM was modified with an additional chloroplast or mitochondrial transit peptide (cTP or mTP, respectively) and the FLAG tag at its N-terminus and the C-terminus, respectively, and the recombinant genes were separately placed downstream of an estrogen-inducible promoter ( Figure 1 ). We used the cTP from Arabidopsis RecA and the mTP from γ-ATPase, which were previously employed for specific targeting of a fluorescent protein to chloroplasts (Köhler et al., 1997) and mitochondria (Masuda et al., 2003), respectively. These constructs were separately introduced into WT Arabidopsis via Agrobacterium-mediated transformation. Given that all RSHs in Arabidopsis target and function within chloroplasts, but not mitochondria (Maekawa et al., 2015; Masuda et al., 2008; Mizusawa et al., 2008), ppGpp-dependent metabolic control may not naturally occur in mitochondria of plant cells. However, both chloroplasts and mitochondria are believed to have originated from the endosymbiosis of an ancient cyanobacterium and α-proteobacterium, respectively (Gross and Bhattacharya, 2011; Stadnichuk and Kusnetsov, 2021). Therefore, if YjbM is artificially targeted to mitochondria, it could potentially synthesize ppGpp and subsequently influence mitochondrial metabolism. As a result, several recombinant WT plants expressing cTP-YjbM or mTP-YjbM were successfully obtained.

To confirm whether the recombinant YjbM is expressed and correctly targeted to chloroplasts or mitochondria in each mutant, western blotting was performed using an anti-FLAG antibody against total protein in each plant line after estrogen treatment. For the analysis, plants were grown on agar-solidified plates in the absence of estrogen for 14 days, and then transferred to new plates containing 1.0 μM estrogen. Shoots were harvested one or two days after estrogen treatment and subjected to western blotting. The molecular weights of FLAG-tagged cTP-YjbM, mTP-YjbM, and transit peptide-cleaved YjbM, as deduced from their amino acid sequences, are 33, 32, and 27 kDa, respectively. Seven cTP-YjbM expressing lines, designated Chl-#1, Chl-#3, Chl-#4, Chl-#5, Chl-#6, Chl-#7, and Chl-#9, exhibited a 27 kDa band after one or two days of estrogen treatment ( Figure 2A ), indicating that cTP-YjbM is successfully targeted to chloroplasts in these lines, although the levels of the targeted protein varied among them. Chl-#1, Chl-#6, and Chl-#7 also showed a 33 kDa band after estrogen treatment ( Figure 2A ), suggesting that a portion of cTP-YjbM remains unprocessed and accumulates in the cytosol in these lines. In contrast, two mTP-YjbM expressing lines, designated Mi-#4 and Mi-#5, displayed a 27 kDa band after one or two days of estrogen treatment ( Figure 2B ), indicating that mTP-YjbM is targeted to mitochondria in these lines. The targeted YjbM levels were higher in Mi-#4 than in Mi-#5. Additionally, Mi-#5 and Mi-#6 exhibited a 32 kDa band after estrogen treatment ( Figure 2B ), suggesting that some mTP-YjbM remains unprocessed and accumulates in the cytosol in these lines.

To determine the amount of ppGpp accumulated in each line, ppGpp quantification was performed. For the analysis, plants were grown on agar-solidified plates in the absence of estrogen for 14 days, then transferred to new plates containing different concentrations of estrogen (0, 0.1, 1.0, or 10 μM). We quantified ppGpp levels in three cTP-YjbM-expressing lines, Chl-#1, Chl-#3, and Chl-#9, as well as in three mTP-YjbM-expressing lines, Mi-#4, Mi-#5, and Mi-#6, which exhibited different levels of targeted YjbM ( Figure 2 ). The results showed that ppGpp accumulation increased in an estrogen concentration-dependent manner, and the amount of ppGpp varied across each line ( Figure 3 ). Specifically, Chl-#1 and Chl-#3 exhibited small ppGpp accumulation in the presence of 1.0 μM and 10 μM estrogen. In contrast, Chl-#9 showed higher ppGpp levels than WT in the presence of >0.1 μM estrogen. Given that the expression level of cTP-cleaved YjbM in Chl-#9 was higher than in Chl-#1 and Chl-#3 ( Figure 2A ), ppGpp levels appear to be correlated with the amount of YjbM targeted to chloroplasts.

Among the mTP-YjbM-expressing lines, Mi-#4 accumulated the highest levels of ppGpp, with over 20 times more ppGpp than WT at estrogen concentrations >1.0 μM. Mi-#5 and Mi-#6 displayed higher ppGpp levels than WT at 1.0 μM estrogen, although Mi-#6 showed no difference in ppGpp levels compared to WT at 10 μM estrogen. Similar to the cTP-YjbM expression lines, ppGpp levels were well correlated with the amount of YjbM targeted to mitochondria; the expression level of transit peptide-cleaved YjbM increased in the order of Mi-#4, Mi-#5, and Mi-#6 ( Figure 2B ). It should be noted that the levels of non-cleaved mTP-YjbM remaining in the cytosol did not correlate with the amount of ppGpp. Specifically, the levels of non-cleaved YjbM decreased in the order of Mi-#4, Mi-#5, and Mi-#6 ( Figure 2B ), suggesting that YjbM with the transit peptide was not functional in the cytosol.

To examine the effect of ppGpp accumulation on plant growth, we grew transgenic lines of Arabidopsis on normal plates for 14 days, then transferred them to plates containing different concentrations of estrogen for 10 days. Some mutants exhibited smaller individual sizes compared to WT in the presence of estrogen ( Figure 4 ), and fresh weight of all transgenic lines tested was significantly reduced in the presence of estrogen compared to WT ( Figure 5 ). Specifically, among the transgenic lines expressing chloroplast-targeted YjbM, Chl-#9, which accumulated high amounts of ppGpp ( Figure 3 ), showed a significant reduction in fresh weight to about half that of WT at an estrogen concentration of 0.1 μM. At concentrations above 1 μM, fresh weight of Chl-#9 was reduced to one-fifth that of WT ( Figure 5 ). Phenotypically, new leaves showed white discoloration in the veins, and anthocyanin accumulation was observed on the underside of the leaves of Chl-#9 under 1.0 μM estrogen conditions ( Figure 4 ). The Chl-#9 line could not germinate on plates containing >1.0 µM estrogen ( Supplementary Figure 1 ). Additional phenotypes of the line included short roots. Mi-#4, which also accumulated high levels of ppGpp ( Figure 3 ), had less than half the fresh weight of WT at an estrogen concentration of 1.0 μM ( Figure 5 ).

To obtain more information contributing to applied research aimed at increasing plant biomass production, we next examined the effect of estrogen-induced ppGpp accumulation on plant growth in soil. Plants were first grown on plates without estrogen for 14 days, then transferred to soil and watered with solutions containing 0, 0.1, and 1.0 μM estrogen. We used low nitrogen-containing soil (0.03 mM KNO₃) to observe the impact of ppGpp accumulation on plant biomass (Goto et al., 2022, 2024). No significant effect of estrogen treatment on visible phenotype was observed in WT, Chl-#1 and Chl-#3 ( Figure 6 ). On the other hand, shoot size of the transgenic lines expressing mitochondrion-targeted YjbM (Mi-#4, Mi-#5 and Mi-#6) seem to be larger than WT when plants were watered with estrogen ( Figure 6 ). Statistical analysis indicated that Mi-#4 and Mi-#6 exhibited significantly greater fresh weight than WT under 1.0 μM estrogen conditions, with Mi-#6 also showing significantly greater fresh weight than WT under 0.1 μM estrogen conditions ( Figure 7 ). Additionally, Chl-#9 showed significantly higher fresh weight and leaf area compared to WT under 1.0 μM estrogen conditions. Notably, these phenotypes were not observed under mock (DMSO) conditions ( Figure 7 ).

Finally, we quantified ppGpp levels in plants grown in soil. For the analysis, plants were first grown on plates without estrogen for 14 days, then transferred to soil and grown for 10 days without estrogen. Plants were then watered with solutions containing 1.0 μM estrogen, grown for 2 days, and then harvested for ppGpp quantification. We first compared ppGpp levels of WT and the transgenic lines expressing chloroplast-targeted YjbM, which showed that there is no significant difference between WT and the transgenic lines in the ppGpp levels. Specifically, the measured values (pmol g^-1 fresh weight) are as follows (sample size 4): 64 ± 20 (WT), 125 ± 92 (Chl-#1), 148 ± 38 (Chl-#3) and 94 ± 14 (Chl-#9) (P > 0.05, t-test). We also found that there is no significant difference in ppGpp levels between WT and the transgenic lines expressing mitochondrion-targeted YjbM; the measured values (pmol g^-1 fresh weight) are as follows (sample size 4): 56 ± 26 (WT), 89 ± 28 (Mi-#4), 75 ± 31 (Mi-#5) and 96 ± 31 (Mi-#6) (P > 0.05, t-test). It should be noted that ppGpp is accumulated in the transgenic lines after 2 days from estrogen treatment on plates ( Figure 3 ), suggesting that the effect of estrogen on ppGpp accumulation in the transgenic lines does not last long in soil.