Transcriptional control of hydrogen peroxide homeostasis regulates ground tissue patterning in the Arabidopsis root

In multicellular organisms, including higher plants, asymmetric cell divisions (ACDs) play a crucial role in generating distinct cell types. The Arabidopsis root ground tissue initially has two layers: endodermis (inside) and cortex (outside). In the mature root, the endodermis undergoes additional ACDs to produce the endodermis itself and the middle cortex (MC), located between the endodermis and the pre-existing cortex. In the Arabidopsis root, gibberellic acid (GA) deficiency and hydrogen peroxide (H2O2) precociously induced more frequent ACDs in the endodermis for MC formation. Thus, these findings suggest that GA and H2O2 play roles in regulating the timing and extent of MC formation. However, details of the molecular interaction between GA signaling and H2O2 homeostasis remain elusive. In this study, we identified the PEROXIDASE 34 (PRX34) gene, which encodes a class III peroxidase, as a molecular link to elucidate the interconnected regulatory network involved in H2O2- and GA-mediated MC formation. Under normal conditions, prx34 showed a reduced frequency of MC formation, whereas the occurrence of MC in prx34 was restored to nearly WT levels in the presence of H2O2. Our results suggest that PRX34 plays a role in H2O2-mediated MC production. Furthermore, we provide evidence that SCARECROW-LIKE 3 (SCL3) regulates H2O2 homeostasis by controlling transcription of PRX34 during root ground tissue maturation. Taken together, our findings provide new insights into how H2O2 homeostasis is achieved by SCL3 to ensure correct radial tissue patterning in the Arabidopsis root.


Introduction
Plants dynamically integrate environmental signals into their genetic programs to provide flexibility for growth and development.Plant hormones have been shown to play key roles in such signaling pathways (Achard et al., 2006;Wolters and Jürgens, 2009;Verma et al., 2016).Despite the complexity of the pathways, signals are ultimately relayed to transcription factors (TFs) to spatiotemporally regulate cellular behaviors and responses (Srivastava et al., 2010;Moreno-Risueno et al., 2015;Barah et al., 2016).Identifying and characterizing transcriptional regulatory networks are important for understanding the cellular processes involved in plant growth and development (Moreno-Risueno et al., 2015;Chaiwanon et al., 2016;Dhar et al., 2022).
In this study, to better understand H 2 O 2 -mediated endodermal ACDs for MC formation, we identified and characterized a transcriptional regulatory network in ground tissue maturation.Our results have confirmed that H 2 O 2 generation facilitates MC production in the Arabidopsis root.Importantly, we provide convincing evidence that SCL3 plays a role in H 2 O 2 homeostasis through transcriptional regulation of the PRX34 gene in H 2 O 2mediated MC formation.

MC formation analysis
For phenotypic analysis of MC formation, approximately 36-40 hours post-germination (hpg) seeds were transferred to new 1/2 MS agar plates supplemented with different chemicals, including H 2 O 2 (100 mM; Sigma-Aldrich, USA), KI (1 mM; Duchefa Biochemie, Netherlands), or PAC (1 µM; Duchefa Biochemie, Netherlands).The same batches of the seeds were transferred to new 1/2 MS agar plates with no supplemented chemicals as controls.More than 300 roots from three replicate experiments (each experiment, n > 100 roots per treatment or genotype) were observed using an Axio Imager.A1 microscope (Carl Zeiss, Germany), and the frequency of endodermal ACDs for MC formation was measured, as previously described (Heo et al., 2011;Lee et al., 2016).Simultaneously, seedlings (n > 30 roots per treatment or genotype) grown on 1/2 MS agar plates with different supplements were observed using a Zeiss LSM 800 confocal laser scanning microscope (Carl Zeiss, Germany), as previously described (Lee et al., 2016;Yoon et al., 2016;Dhar et al., 2022).Student's t-test was performed using Microsoft Excel (Microsoft, USA) and the data presented herein are means values ± standard error (SEM) as described previously (Heo et al., 2011;Lee et al., 2016;Yoon et al., 2016;Dhar et al., 2022).

H 2 O 2 assays
Qualitative and quantitative assays were conducted to analyze H 2 O 2 concentrations in seedling roots.DAB staining (3,3′diaminobenzidine; Sigma-Aldrich, USA) was used to qualitatively assess H 2 O 2 concentration, as previously described (Thordal-Christensen et al., 1997;Vanacker et al., 2000;Bindschedler et al., 2006;Daudi et al., 2012) with minor modifications.DAB solution (final concentration: 1 mg/mL) was prepared in 0.05% Triton X-100 (v/v) and 10 mM sodium phosphate buffer.The seedlings were vacuum-infiltrated in the DAB solution for 10 min and subsequently incubated at 30°C for 30 min in the dark.After termination of the staining reaction, samples were fixed in the bleaching solution [ethanol:acetic acid:glycerol, 3: for 15 min.Seedling roots were observed using an Axio Imager.A1 microscope equipped with an AxioCam MRc5 digital camera (Carl Zeiss, Germany).The intensity of DAB staining in the seedling roots was quantified by NIH Image J software (http:// rsb.info.nih.gov/ij;Schneider et al., 2012), as described previously (Li et al., 2020).To quantitatively measure H 2 O 2 concentration, we used the Amplex ® Red Hydrogen Peroxide Assay Kit (cat.#A22188, Invitrogen), according to the manufacturer's instructions.The 7 dpg seedling roots grown on 1/2 MS agar plates with or without supplemented chemicals were pulverized in liquid nitrogen.Subsequently, five volumes of 50 mM sodium phosphate buffer (pH 7.4) were added, mixed thoroughly, and incubated on ice for 10 min.The samples were centrifuged at 12,000 rpm for 20 min at 4°C, and supernatants were used for the measurement of H 2 O 2 concentration as previously described (Cui et al., 2014).Each experiment was independently repeated at least three times and data were analyzed using Microsoft Excel (Microsoft, USA).

Induction of H 2 O 2 by GA deficiency facilitates MC formation
Previously, it has been shown that H 2 O 2 promotes periclinal ACDs in the endodermis for MC production (Cui et al., 2014;Li et al., 2020).These findings suggest that H 2 O 2 homeostasis plays a role in modulating the timing and extent of MC formation in the root ground tissue.We thus verified that the frequency of MC occurrence in Columbia wild-type (hereafter referred to as WT) roots was elevated by H 2 O 2 .Under normal growth conditions, cells in the endodermis of 4 dpg WT seedlings barely divided to generate the MC layer, whereas endodermal ACDs were frequently observed under H 2 O 2 (100 mM) treatment (~3.7% vs. ~10%) (Supplementary Figures 1A-C).In the presence of KI (1 mM), an efficient scavenger of H 2 O 2 (Dunand et al., 2007), MC formation in 7 dpg WT seedlings was attenuated compared to that in untreated roots (~10.7% vs. ~22%) (Supplementary Figures 1D-F).Interestingly, the application of H 2 O 2 in the presence of KI almost restored the frequency of MC formation to that in untreated WT roots (Supplementary Figure 1F).Therefore, consistent with previous studies (Cui et al., 2014;Li et al., 2020), our findings strongly support the notion that H 2 O 2 induces ACDs in the endodermis for MC production in the root ground tissue.
GA-deficient conditions caused by the loss-of-function mutant ga1-3 or by the GA biosynthesis inhibitor PAC facilitate endodermal ACDs for the MC layers (Paquette and Benfey, 2005;Cui and Benfey, 2009a;Heo et al., 2011;Koizumi et al., 2012a;Koizumi et al., 2012b;Gong et al., 2016;Lee et al., 2016).Thus, we hypothesized that frequent MC formation might be due to elevated H 2 O 2 levels induced by GA deficiency.To test this, we first analyzed H 2 O 2 concentration in roots by DAB staining in the absence or presence of PAC.DAB staining was more intense in PAC-treated WT roots than untreated controls (Figures 1A-C).In addition, we quantitatively assessed H 2 O 2 levels in PAC-treated roots relative to those in untreated controls.Consistent with the DAB staining results, H 2 O 2 accumulation was higher in PACtreated roots than untreated controls (Figure 1D).Likewise, H 2 O 2 was more accumulated in ga1-3 than WT (Figures 1A, E).In the presence of KI, ga1-3 roots showed reduced H 2 O 2 accumulation (Figures 1F-H).These findings indicate that GA-deficient conditions promoted H 2 O 2 generation.Next, we assessed the frequency of MC formation in ga1-3 with or without KI.As expected, the occurrence of the MC layers was substantially reduced in KI-treated ga1-3 compared to that in untreated ga1-3 (Figures 1J-L).Intriguingly, the KI-treated ga1-3 phenotype was almost indistinguishable from that of the WT control (Figures 1I,  K, L; Supplementary Figures 1D, F).This observation indicates that KI treatment reduced H 2 O 2 accumulation in ga1-3, which resulted in decreased MC formation.
Taken together, our results strongly suggest that the induction of H 2 O 2 under GA-deficient conditions causes more frequent MC production in the root ground tissue.In an attempt to identify the molecular component(s) underlying the H 2 O 2 -mediated regulation of MC production, we encountered an interesting report that the expression levels of class III peroxidase (PRX) genes were significantly changed in the-lossof-function spindly (spy) mutants (2-fold enrichment with P < 0.05; Cui et al., 2014), which have been shown to play a role in GA signaling (Jacobsen and Olszewski, 1993;Jacobsen et al., 1996;Sun and Gubler, 2004).Of the differentially expressed class III peroxidase (PRX) genes (Cui et al., 2014), we focused on PRX34 because this gene fulfilled our criteria: i) its expression levels were induced by GA-deficient conditions (PAC and ga1-3) and H 2 O 2 (Supplementary Figures 2A-C), ii) its expression was enriched in roots (Supplementary Figure 2D), and iii) its T-DNA insertion mutant was publicly available (Passardi et al., 2006;Daudi et al., 2012;O'Brien et al., 2012).Therefore, in this study, we aimed to elucidate the role of PRX34 in H 2 O 2 -mediated MC formation.
Because PRX34 plays a role in H 2 O 2 homeostasis (Daudi et al., 2012;O'Brien et al., 2012), we first investigated whether H 2 O 2 levels were altered in the-loss-of-function prx34 mutant.Indeed, accumulation of H 2 O 2 decreased in prx34 compared to that in WT roots (Figures 2A-D).Next, we analyzed the frequency of endodermal ACDs in WT and prx34 roots.Under normal conditions, prx34 showed a reduced frequency of MC formation compared to WT (Figures 2E, F, H).In the presence of H 2 O 2 , the occurrence of endodermal ACDs for MC generation in prx34 was almost restored to that in WT (Figures 2E, G, H).
Taken together, our findings indicate that PRX34 likely plays a role in H 2 O 2 -mediated MC production in the root ground tissue.

Involvement of PRX34 in the GA-mediated regulation of MC formation
Considering the results that i) GA deficiency induced H 2 O 2 accumulation; ii) PRX34 expression was promoted by GA deficiency and H 2 O 2 ; and iii) both H 2 O 2 level and MC occurrence were substantially attenuated in prx34, we hypothesized that PRX34 might be involved in GA-mediated MC formation in the root ground tissue.To test this, we assessed MC production in prx34 under GA-deficient conditions.In the presence of PAC, prx34 exhibited a reduced occurrence of the MC layers compared to the WT (Figures 3A-C).Next, we performed a genetic analysis using prx34 ga1-3 double mutants.Compared with ga1-3 single mutants, prx34 ga1-3 displayed attenuated MC formation, which was similar to WT (Figures 3D-G).To investigate whether the less frequent MC layers in prx34 under GA-deficient conditions (PAC or ga1-3) was due to lower H 2 O 2 production, we measured the level of H 2 O 2 in prx34 ga1-3 roots.The accumulation of H 2 O 2 in prx34 ga1-3 was indeed lower than that in ga1-3 (Figures 3H-L).These findings are consistent with our phenotypic analyses in prx34 roots under GA-deficient conditions.
Taken together, our results suggest that PRX34, via H 2 O 2 production, is involved in the GA-mediated regulation of MC formation.Previously, it has been demonstrated that SCL3 plays a role in GA-mediated MC production and thus, the loss-of-function scl3-1 results in more frequent MC formation than in WT roots (Heo et al., 2011;Zhang et al., 2011;Lee et al., 2016).Therefore, we hypothesized that the frequent MC generation phenotype in scl3-1 roots might be due to the elevated H 2 O 2 level.To test this, we assessed H 2 O 2 concentration in scl3-1 in comparison with WT roots and found that scl3-1 accumulated more H 2 O 2 than the WT (Figures 4A-D).We then analyzed the occurrence of MC in the presence of KI.The frequency of MC formation was substantially reduced in KI-treated scl3-1 roots compared to untreated controls (Figures 4E-H).These findings indicate that increased H 2 O 2 levels are likely a causative factor in the frequent generation of MC in scl3-1 roots.

SCL3 acts upstream of PRX34 in H 2 O 2mediated MC formation
Because SCL3 is likely involved in H 2 O 2 -mediated MC production, we investigated the relationship between SCL3 and PRX34.First, we performed a genetic analysis by generating prx34 scl3-1 double mutants.Under normal conditions, prx34 and scl3-1 single mutants showed opposite phenotypes in MC generation: a decreased occurrence in prx34 and an increase in scl3-1 (Figures 5A-C, E).Unexpectedly, the MC phenotype of prx34 scl3-1 double mutants resembled that of prx34 single mutants; in that the frequency of MC formation was significantly attenuated (Figures 5B, D, E).Therefore, our genetic analysis indicates that prx34 is epistatic to scl3.Next, we assessed H 2 O 2 levels in prx34 scl3-1 and demonstrated that H 2 O 2 accumulation in prx34 scl3-1 was reduced to a level similar to prx34 (Figures 5F-K).Our results strongly support the notion that the reduced production of MC in prx34 scl3-1 is due to the attenuated levels of H 2 O 2 in the double mutant roots.Furthermore, because prx34 is epistatic to scl3, PRX34 is likely to act downstream of SCL3 in the H 2 O 2 -mediated pathway for MC formation in the root ground tissue.To test this, we analyzed the expression levels of PRX34 in the loss (scl3-1) and gain (SCL3-OX) of SCL3 function plants.Interestingly, the abundance of PRX34 mRNA was elevated in scl3-1, whereas its level was reduced in SCL3-OX (Figure 6).This finding indicates that SCL3 likely acts as a negative regulator to modulate the PRX34 expression, thereby maintaining H 2 O 2 homeostasis.

Discussion
Previously, it has been reported that ROS, particularly H 2 O 2 , induce periclinal ACDs in the endodermis for MC formation in the Arabidopsis root (Cui et al., 2014;Li et al., 2020).In agreement with previous reports, we also found that H 2 O 2 facilitated the occurrence of MC.In addition, the plant hormone GA has been shown to regulate the timing and extent of MC formation (Paquette and Benfey, 2005;Cui and Benfey, 2009a;Cui and Benfey, 2009b;Heo et al., 2011;Pauluzzi et al., 2012;Koizumi et al., 2012a;Koizumi et al., 2012b;Cui et al., 2014;Choi and Lim, 2016;Cui, 2016;Gong et al., 2016;Lee et al., 2016;Bertolotti et al., 2021).Here, we demonstrated that H 2 O 2 accumulated more in ga1-3 roots than in WT, resulting in excessive endodermal ACDs for MC production.However, the molecular link between the GA pathway and H 2 O 2 homeostasis during Arabidopsis ground tissue maturation remains elusive.
To better understand the molecular events in H 2 O 2 -and GAmediated MC formation, we attempted to identify a candidate molecular link using three criteria: i) expression in both GA deficiency and H 2 O 2 , ii) enrichment of expression in the root, and iii) availability of a T-DNA insertion mutant.Previously, it was demonstrated that SPY regulated expression of some class III PRX genes (Cui et al., 2014).In particular, the expression levels of PRX33 and PRX34, which are closely related and tandemly located in the Arabidopsis genome, were substantially reduced in the spy root.Both PRX33 and PRX34 were shown to generate H 2 O 2 , which conferred resistance to pathogens during the Arabidopsis defense response (Daudi et al., 2012;O'Brien et al., 2012).In addition, both PRXs were reported to be involved in root elongation, possibly modifying cell walls (Passardi et al., 2006).However, the roles of these PRXs in H 2 O 2 -mediated MC formation remain unknown.In this study, we focused our efforts on PRX34, which fulfilled our criteria; further investigation of other PRX genes is the subject of another study, which is not discussed here.Interestingly, prx34 roots exhibited a reduction of H 2 O 2 accumulation, resulting in less frequent endodermal ACDs for MC formation than in WT roots.When applied with H 2 O 2 , the frequency of MC production in prx34 was restored to WT levels.Furthermore, prx34 showed reduced MC production in GA-deficient conditions caused by ga1-3 or PAC, compared to the mutant under normal conditions.These findings strongly support the idea that GA deficiency induces H 2 O 2 generation via PRX34, and, in turn, H 2 O 2 accumulation promotes endodermal ACDs for MC formation during Arabidopsis ground tissue maturation.
In the GA signaling pathway, SCL3 antagonizes the function of DELLA proteins, which are the major negative regulators (Silverstone et al., 1997;Silverstone et al., 1998;Sun and Gubler, 2004;Zentella et al., 2007;Heo et al., 2011;Zhang et al., 2011;Yoshida and Ueguchi-Tanaka, 2014;Weng et al., 2020;Ito and Fukazawa, 2021).The scl3-1 mutant in the ga1-3 background showed enhanced GA-deficient phenotypes in growth and development, whereas loss-of-function mutations in the DELLA genes restored the ga1-3 phenotypes (Silverstone et al., 1997;Silverstone et al., 1998;Dill and Sun, 2001;Heo et al., 2011;Zhang et al., 2011;Yoshida and Ueguchi-Tanaka, 2014;Weng et al., 2020;Ito and Fukazawa, 2021).In particular, GA-deficient conditions (ga1-3 or PAC) exacerbate the MC phenotype of scl3-1, resulting in excessive endodermal ACDs for MC formation (Heo et al., 2011).When overexpressed, SCL3-OX reduced the occurrence of MC, even under GA-deficient conditions (Heo et al., 2011).Thus, SCL3 plays a crucial role in GA-mediated MC formation (Heo et al., 2011).In this study, we demonstrated that H 2 O 2 accumulation was higher in scl3-1 than WT.In the presence of KI, the frequency of MC formation in scl3-1 was attenuated.Furthermore, the prx34 scl3-1 double mutant showed lower H 2 O 2 accumulation than scl3-1.Consistent with the low level of H 2 O 2 in prx34 scl3-1, the double mutant displayed less frequent MC production than scl3-1.Taken together, the prx34 scl3-1 double mutant was indistinguishable from the prx34 single mutant in both H 2 O 2 accumulation and MC phenotype.Thus, our genetic analysis led us to the conclusion that prx34 is epistatic to scl3.In addition, the abundance of PRX34 transcripts was higher in scl3-1 and lower in SCL3-OX than that in WT.These results strongly suggest that SCL3 serves as a negative regulator of PRX34 expression to maintain H 2 O 2 homeostasis during root ground tissue maturation.
SCL3, acting downstream of the SHR/SCR regulatory module, is uniquely positioned in controlling the timing and extent of MC formation (Heo et al., 2011;Choi and Lim, 2016;Gong et al., 2016;Lee et al., 2016).Recently, it has been reported that the SHR/SCR module physically interacted with NAC1 to restrict excessive periclinal ACDs in the endodermis during root ground tissue maturation (Xie et al., 2023).NAC1 directly inhibited the transcription of CYCD6;1 with the transcriptional co-repressor TOPLESS (TPL), resulting in reduced MC generation (Xie et al., 2023).Thus, it is tempting to speculate that the SHR/SCR regulatory module, together with NAC1, is also involved in H 2 O 2 -and GAmediated MC formation, impinging on the transcriptional regulation of SCL3.
Taken together, in a simplified model (Figure 7), the plant hormone GA negatively regulates the expression of SCL3.Hence, SCL3 acts as a convergent point for the interaction between GA and ROS.SCL3 plays a role in H 2 O 2 homeostasis via the transcriptional regulation of PRX34 in H 2 O 2 -mediated MC formation.Therefore, we have not only provided new insights into the crosstalk between GA and ROS but also unveiled a novel role for PRX34 in the maintenance of H 2 O 2 homeostasis during root ground tissue maturation.Thus, it is tempting to speculate that diverse transcriptional inputs from hormonal (e.g., GA) and developmental (e.g., SHR/SCR) pathways impinge on the tissuespecific integrator SCL3 to modulate the timing and extent of MC formation during root ground tissue maturation.performed plant work including genotyping.S-HK contributed new analytical tools and reagents and provided critical comments and suggestions on the experiments.JO, JWC, SJ, and JL wrote the manuscript with contributions from all the authors.All authors contributed to the article and approved the submitted version.

FIGURE 1 H
FIGURE 1 H 2 O 2 generation by GA deficiency facilitates MC formation in the Arabidopsis root.(A, B) DAB staining of WT roots in the absence (A) or presence (B) of PAC.(C) Quantification of DAB staining of the WT roots with or without PAC.(D) Measurement of H 2 O 2 in PAC-treated or -untreated WT roots.(E, F) DAB staining of ga1-3 roots in the absence (E) or presence (F) of KI. (G) Quantification of DAB staining of ga1-3 roots with or without KI.(H) Measurement of H 2 O 2 in KI-treated or -untreated ga1-3 roots.(I-K) Confocal images of WT (I) and ga1-3 roots in the absence (J) or presence (K) of KI.The inset in (J) shows endodermal ACDs for MC formation.The endodermis (EN), middle cortex (MC), and cortex (CO) layers are indicated with white arrowheads.(L) Proportion of WT and ga1-3 plants with MC in the absence or presence of KI.Significance of difference was statistically determined by Student's t-test (*P < 0.05; **P < 0.01; ***P < 0.001).

FIGURE 2 prx34
FIGURE 2 prx34 roots exhibit reductions of H 2 O 2 accumulation and MC formation.(A, B) DAB staining of WT (A) and prx34 (B) roots.(C) DAB staining quantification of WT and prx34 roots.(D) Measurement of H 2 O 2 in WT and prx34.(E-G) Confocal images of WT (E) and prx34 in the absence (F) or presence (G) of H 2 O 2 .The insets in (E) and (G) illustrate periclinal ACDs in the endodermis for MC formation.The endodermis (EN), middle cortex (MC), and cortex (CO) layers are indicated with white arrowheads.(H) Proportion of WT and prx34 plants with MC in the absence or presence of H 2 O 2 .Significance of difference was determined by Student's t-test (**P < 0.01; ***P < 0.001; ns: statistically not significant).

FIGURE 4 scl3
FIGURE 4 scl3 plays a role in the regulation of H 2 O 2 -mediated MC formation.(A, B) DAB staining of WT (A) and scl3-1 (B) roots.(C) Quantification of DAB staining in WT and scl3-1.(D) Measurement of H 2 O 2 in WT and scl3-1.(E-G) Confocal images of WT (E) and scl3-1 roots in the absence (F) or presence (G) of KI.The inset in (F) illustrates periclinal ACDs in the endodermis for MC formation.The endodermis (EN), middle cortex (MC), and cortex (CO) layers are indicated with white arrowheads.(H) Proportion of WT and scl3-1 plants with MC in the absence or presence of KI.Significance of difference was determined by Student's t-test (**P < 0.01; ***P < 0.001).

FIGURE 6 SCL3
FIGURE 6SCL3 negatively regulates the expression levels of PRX34 in roots.The PRX34 expression is upregulated in scl3-1, whereas its expression is attenuated in SCL3-OX.Error bars indicate ± SEM from three independent replicates.Significance of difference was determined by Student's t-test (***P < 0.001).