Alterations in hormonal signals spatially coordinate distinct responses to DNA double-strand breaks in Arabidopsis roots

Plant hormones cytokinin and auxin orchestrate differential DNA damage responses in Arabidopsis roots.

Abstract

Plants have a high ability to cope with changing environments and grow continuously throughout life. However, the mechanisms by which plants strike a balance between stress response and organ growth remain elusive. Here, we found that DNA double-strand breaks enhance the accumulation of cytokinin hormones through the DNA damage signaling pathway in the Arabidopsis root tip. Our data showed that activation of cytokinin signaling suppresses the expression of some of the PIN-FORMED genes that encode efflux carriers of another hormone, auxin, thereby decreasing the auxin signals in the root tip and causing cell cycle arrest at G₂ phase and stem cell death. Elevated cytokinin signaling also promotes an early transition from cell division to endoreplication in the basal part of the root apex. We propose that plant hormones spatially coordinate differential DNA damage responses, thereby maintaining genome integrity and minimizing cell death to ensure continuous root growth.

INTRODUCTION

Plant roots play a crucial role in water and nutrient uptake, anchorage to soil, and sensing of the rhizosphere environment. Since these functions have a substantial impact on overall plant growth, root development is precisely controlled in response to changing underground conditions. Environmental stresses usually inhibit root growth; salinity, oxidation, or heat stress severely retards root growth. Previous studies demonstrated that roots that encounter high levels of boron or aluminum, and soil-borne pathogens, suffer from DNA damage and exhibit a delay or cessation of cell division, thereby suppressing root growth (1–3). This is an active response to DNA damage that is governed by the cell cycle checkpoint mechanism, in which cell cycle progression is arrested at a specific stage to ensure DNA repair or to provoke cell death in severe cases. As do other eukaryotes, plants have two protein kinases, ATAXIA TELANGIECTASIA MUTATED (ATM) and ATM AND RAD3-RELATED (ATR), that sense DNA damage and trigger cell cycle checkpoints (4, 5). ATM is activated by DNA double-strand breaks (DSBs), whereas ATR primarily senses single-strand DNA and replication stress caused by DNA replication fork blocking. In animals and fungi, DNA damage signals are transmitted to Checkpoint-1 (CHK1) and CHK2 kinases, and ATM, ATR, CHK1, and CHK2 phosphorylate and activate the tumor suppressor protein p53 (6). However, orthologs of CHK and p53 are missing in plants; instead, the plant-specific NAM, ATAF and CUC (NAC)–type transcription factor, named SUPPRESSOR OF GAMMA RESPONSE 1 (SOG1), plays a central role in transmitting the signal from ATM and ATR (7). SOG1 is phosphorylated and activated by ATM and ATR (8, 9) and binds to the sequence CTT[N]₇AAG to induce the expression of target genes involved in DNA repair and recombination, cell cycle control, and plant immunity (10).

In Arabidopsis roots, cells actively divide and proliferate in the meristematic zone in the tip region. After several rounds of cell division, cells stop dividing and start endoreplication, in which DNA replication is repeated without mitosis or cytokinesis. Endoreplicating cells begin to elongate and constitute the transition zone and, eventually, start more rapid cell elongation through actin reorganization in the elongation/differentiation zone (11). It was shown that DSBs arrest the cell cycle at G₂ in the meristematic zone (12) and promote an early onset of endoreplication (13), reducing the meristem size and inhibiting root growth. At the same time, selective cell death is induced in stem cells, which reside in the most apical part of the growing root and surround the quiescent center (QC) cells, in response to DSBs (14). These DNA damage responses are under the control of the ATM-SOG1 pathway, although how distinct DNA damage responses are coordinated to ensure continuous root growth remains elusive. We showed previously that two repressor-type R1R2R3-Myb transcription factors (Rep-MYBs), MYB3R3 and MYB3R5, participate in DNA damage–induced G₂ arrest by suppressing the expression of G₂-M–specific genes (12). We have recently demonstrated that protein accumulation of Rep-MYBs is regulated by the transcription factors ANAC044 and ANAC085 (15), while cyclin-dependent kinases (CDKs) also play a crucial role in Rep-MYB accumulation; namely, Rep-MYBs phosphorylated by CDKs are targeted for degradation, whereas low CDK activity caused by DNA damage stabilizes Rep-MYB proteins, leading to G₂ arrest (12). However, it remains unclear how CDK activities are inhibited by DNA stress in the root tip. Although DNA damage induces the expression of several CDK inhibitors and represses the gene encoding activator-type R1R2R3-Myb transcription factor, which induces G₂-M–specific genes, these transcriptional responses are known to be insufficient for G₂ arrest (10, 12, 15). Therefore, some unknown mechanism(s) should function in down-regulation of CDK activities under DNA stress. Moreover, how stem cell death and the early onset of endoreplication are induced also remains unexplained.

Cytokinin plant hormones participate in various developmental and physiological processes such as seed germination, flowering, and senescence (16). The initial step of cytokinin biosynthesis is catalyzed by adenosine triphosphate/adenosine diphosphate (ATP/ADP) ISOPENTENYLTRANSFERASES (IPTs), which produce ribosylated and phosphorylated precursors of N⁶-(∆²-isopentenyl) adenine (iP) (17). These products are converted to ribosylated and phosphorylated forms of trans-zeatin by the action of CYP735A, which hydroxylates the trans-end of the prenyl side chain (18). LONELY GUYs (LOGs) are essential enzymes for converting these precursors to biologically active cytokinins (19). Previous studies demonstrated that cytokinins are systemically transported in Arabidopsis: iP types are moved from shoots to the root meristem via the phloem (20), and trans-zeatin types are transported from roots to shoots via the xylem (18). CYP735A2 is predominantly expressed in developing vascular tissues of roots, thereby converting iP to trans-zeatin type and supplying trans-zeatin types to shoots (18). The ABC transporter ABCG14 contributes to trans-zeatin–type cytokinin transport by facilitating xylem loading in roots (21).

Cytokinins are known to inhibit primary root growth and lateral root formation by suppressing the expression of the auxin efflux carrier PIN-FORMED (PIN) (22, 23). We previously reported that DSBs inhibit lateral root formation and that this inhibition is partially suppressed in cytokinin biosynthesis or signaling mutants, suggesting the possibility that DNA damage regulates lateral root development through influencing cytokinin signaling (24). However, it had remained elusive whether cytokinin biosynthesis or signaling is directly targeted by DNA damage signals and how other hormones are associated with DNA damage responses. Consequently, the exact role of hormonal signaling in controlling distinct DNA damage responses in roots is still totally unknown. In this study, we found that DSBs increase the endogenous cytokinin level in Arabidopsis roots, thereby suppressing the expression of PIN1, PIN3, and PIN4 and decreasing the auxin level in the meristematic zone. We propose that reduced auxin signaling is a key to cause G₂ arrest and stem cell death, while enhanced cytokinin signaling promotes the early onset of endoreplication, representing the mechanisms underlying spatial regulation of DNA damage responses by two plant hormones.

RESULTS

DSBs activate cytokinin signaling in roots

To test whether cytokinin signaling is affected by DNA damage in Arabidopsis root tips, we first examined the promoter activity of the cytokinin-inducible type-A ARABIDOPSIS RESPONSE REGULATOR (ARR) gene ARR5 (25). Five-day-old pARR5:GUS seedlings were transferred onto a medium containing 8 μM DSB-inducing reagent zeocin (26) and subjected to β-glucuronidase (GUS) staining after 24 hours. GUS activity was increased in the vasculature around the boundary between the meristematic zone and the transition zone, vascular initial cells and their daughters, and the root cap (Fig. 1A and fig. S1). We also observed roots carrying the cytokinin response marker Two Component Signaling Sensor new (TCSn):GFP, which reflects the transcriptional activity of type-B response regulators (27). The expression pattern of this marker was similar to that of pARR5:GUS, except that green fluorescent protein (GFP) fluorescence was also detected in the epidermis (Fig. 1B). Quantification of GFP fluorescence revealed that the expression level was elevated by zeocin treatment in four epidermal cells above and below the boundary between the meristematic zone and the transition zone (Fig. 1C). A higher level of GFP fluorescence was also observed in vascular cells encompassing a 100-μm region around the boundary (Fig. 1D). The elevated expression of pARR5:GUS and TCSn:GFP was similarly observed when seedlings were treated with bleomycin (0.6 μg/ml), another DSB inducer (fig. S2) (26). Our time-course experiment showed that the TCSn:GFP signal increased after 12 hours of zeocin treatment in both the epidermis and the vasculature, while propidium iodide (PI)–stained dead cells appeared in vascular initial cells and their daughters after 24 hours (fig. S3), suggesting that activation of cytokinin signaling is an early response to DSBs. To confirm that enhanced cytokinin signaling is not a consequence of disorganized tissue structure caused by DNA damage, we observed the expression patterns of several cell type–specific markers, pAHP6:GFP (protoxylem), pAPL:GFPer (phloem), pCO2:H2B-YFP (cortex), pSCR:GFP-SCR (endodermis and QC), and pWOX5:NLS-YFP (QC) (fig. S4). Zeocin treatment induced stem cell death after 24 hours, as mentioned above, but did not change the expression pattern of any marker, indicating that our experimental conditions did not cause severe defects in root tissue organization. Increased expression of TCSn:GFP was also observed after 1.5 mM aluminum treatment for 24 hours, which is also known to cause DSBs (1), while expression patterns of the cell type–specific markers were not altered (fig. S5).

Fig. 1. DSBs activate cytokinin signaling in the root tip.

(A) Zeocin response of the ARR5 promoter activity. Five-day-old pARR5:GUS seedlings were transferred to Murashige and Skoog (MS) plates supplemented with (+ zeocin) or without (− zeocin) 8 μM zeocin and grown for 24 hours, followed by GUS staining. Arrowheads indicate the boundary between the meristematic zone and the transition zone. Scale bar, 100 μm. (B) Zeocin response of the synthetic cytokinin reporter TCSn:GFP. Five-day-old TCSn:GFP seedlings were treated with (+ zeocin) or without (− zeocin) 8 μM zeocin for 24 hours, and GFP fluorescence was observed after counterstaining with PI. Arrowheads indicate the boundary between the meristematic zone and the transition zone. Magnified images of the areas marked by white boxes are shown on the right [see also (D)]. Scale bars, 100 μm. (C) GFP fluorescence intensity in the root epidermis of TCSn:GFP seedlings. Cell position “1” indicates the first endoreplicated cell, which is preceded by the last mitotic cell before entry into the endoreplication (cell position “−1”). Data are presented as means ± SD (n > 8). Significant differences from the control without zeocin treatment were determined by Student’s t test, *P < 0.05 and ***P < 0.001. (D) GFP fluorescence intensity in the vasculature of TCSn:GFP seedlings. GFP fluorescence was measured in the areas surrounded by yellow dotted lines shown on the right in (B), which encompass a 100-μm region around the boundary between the meristematic zone and the transition zone. The value relative to that of the control without zeocin treatment is shown. Data are presented as means ± SD (n = 10). The significant difference from the control was determined by Student’s t test, ***P < 0.001.

In response to DSBs, ATM phosphorylates and activates the plant-specific transcription factor SOG1, which then induces hundreds of genes to trigger DNA damage responses (7). We found that zeocin treatment increased pARR5:GUS expression in wild type (WT), but not in the atm-2 or sog1-1 knockout mutant (fig. S1). This indicates that activation of cytokinin signaling in roots is a programmed response to DSBs that requires the ATM-SOG1 pathway.

DSBs elevate cytokinin level in the root tip

In Arabidopsis, iP and trans-zeatin are perceived by three receptors, ARABIDOPSIS HISTIDINE KINASE 2 (AHK2), AHK3, and AHK4/CRE1. The cytokinin signal activates the transcription factors, type-B ARRs, via the His-Asp phosphorelay pathway (16). Previous studies demonstrated that the type-B response regulators ARR1 and ARR2 are expressed around the transition zone and up-regulate cytokinin signaling to promote cell differentiation and restrict the meristem size (28, 29). However, our quantitative real-time polymerase chain reaction (qRT-PCR) data showed that neither ARR1 nor ARR2 was induced by zeocin in roots (fig. S6A). The marker lines pARR1:ARR1-GUS and pARR2:ARR2-GUS displayed no change in expression levels after zeocin treatment (fig. S6B). These results indicate that DSBs do not up-regulate ARR1 or ARR2 in roots.

We therefore examined whether the endogenous cytokinin level increases in response to DNA damage. Cytokinin content was separately measured in cells constituting the transition zone or the meristematic zone. We performed fluorescence-activated cell sorting (FACS) on protoplasts prepared from ROOT CLAVATA HOMOLOG 1 (RCH1) promoter:GFP or RCH2 promoter:CFP reporter lines. Since the RCH1 and RCH2 promoters are active in the meristematic zone and the transition zone, respectively, both in the presence and absence of zeocin (fig. S7) (28), GFP- or cyan fluorescent protein (CFP)–expressing protoplasts are expected to be derived from each zone. Our cytokinin measurements revealed that, in the presence of zeocin, the levels of trans-zeatin, and cytokinin glucosides trans-zeatin-9-N-glucoside (tZ9G) and isopentenyladenosine-7-N-glucoside (iP7G), increased in the transition zone, but not in the meristematic zone (Fig. 2A). The amounts of cytokinin precursors trans-zeatin riboside and isopentenyladenosine riboside and cytokinin glucosides trans-zeatin-O-glucoside and trans-zeatin-7-N-glucoside (tZ7G) were elevated in both the meristematic zone and the transition zone, although the increases were higher in the transition zone (Fig. 2A). As a result, the total amount of trans-zeatin– and iP-type cytokinins was more markedly increased in the transition zone than the meristematic zone after zeocin treatment (Fig. 2B). We could not detect iP in any sample, probably because iP is efficiently degraded by CYTOKININ OXIDASE (CKX) enzymes and converted to N-glucosides (30).

Fig. 2. DSBs elevate cytokinin level in the root tip.

(A and B) Five-day-old seedlings of the marker lines pRCH1:GFP and pRCH2:CFP were transferred to MS plates with (+ zeocin) or without (− zeocin) 8 μM zeocin and grown for 24 hours. FACS was conducted to collect GFP- or CFP-positive protoplasts, which were analyzed for their cytokinin concentration using liquid chromatography–tandem mass spectrometry. Amounts of each trans-zeatin- and iP-type cytokinin per gram of fresh weight (FW) in GFP- or CFP-positive protoplasts, which were derived from cells in the meristematic zone or the transition zone, respectively (A), and those of total trans-zeatin– or iP-type cytokinin species (B) are shown. Data are presented as means ± SD (n = 4). Significant differences from the control without zeocin treatment were determined by Student’s t test, *P < 0.05, **P < 0.01, and ***P < 0.001. <LOD, below limit of detection. tZ, trans-zeatin; tZR, tZ riboside; tZOG, tZ-O-glucoside; tZROG, tZR-O-glucoside; tZ7G, tZ-7-glucoside; tZ9G, tZ-9-glucoside; tZR5MP, tZR-5-monophosphate; iP, N⁶-(∆²-isopentenyl) adenine; iPR, iP riboside; iP7G, iP-7-glucoside; iP9G, iP-9-glucoside; iPR5MP, iPR-5-monophosphate.

DSB-dependent induction of cytokinin biosynthesis genes inhibits root growth

To identify the cause of the DNA damage–dependent increase in endogenous cytokinin content, we measured the transcript levels of 17 cytokinin biosynthesis genes: 7 ATP/ADP IPTs (IPT1 and IPT3–IPT8), 2 CYP735As (CYP735A1 and CYP735A2), and 8 LOGs (LOG1–LOG8). We did not analyze tRNA IPTs (IPT2 and IPT9) because they are engaged in the synthesis of cis-zeatin, a less physiologically active cytokinin than iP or trans-zeatin (31). qRT-PCR using RNA from whole seedlings revealed that zeocin treatment elevated the mRNA levels of IPT1, IPT3, IPT5, IPT7, CYP735A2, and LOG7, but not LOG3, LOG4, or LOG8 (Fig. 3A). We could not detect transcripts of IPT4, IPT6, IPT8, CYP735A1, LOG1, LOG2, LOG5, or LOG6 regardless of zeocin treatment. When qRT-PCR was conducted using RNA from roots, we could detect LOG1 and LOG5 transcripts; however, they were not increased by zeocin treatment (fig. S8A). Induction of IPT1, IPT3, IPT5, IPT7, CYP735A2, and LOG7 was not observed in the atm-2 or sog1-1 mutant (fig. S8B), indicating that their induction is under the control of the ATM-SOG1 pathway. Although the tZ7G and tZ9G levels were elevated after zeocin treatment (Fig. 2A), the expression of the cytokinin N-glucosyltransferase UGT76C2 that is involved in the formation of cytokinin glucosides was not induced at all (Fig. 3A), suggesting the possibility that DSBs enhance trans-zeatin production and consequently increase the tZ7G and tZ9G levels.

Fig. 3. DSBs induce cytokinin biosynthesis genes.

(A) Transcript levels of cytokinin biosynthesis genes. Five-day-old WT seedlings were transferred to MS plates with or without 8 μM zeocin and grown for 24 hours. Total RNA was extracted from whole seedlings and subjected to qRT-PCR. Transcript levels of cytokinin biosynthesis genes were normalized to that of ACTIN2 and are indicated as relative values, with that of the control without zeocin treatment set to 1. Data are presented as means ± SD calculated from three biological and technical replicates. Significant differences from the control were determined by Student’s t test, ***P < 0.001. (B to E) Zeocin response of pIPT1:GUS, pIPT3:GUS, pIPT5:GUS, pCYP735A2:GUS, and pLOG7:GUS. Five-day-old seedlings were treated with (+ zeocin) or without (− zeocin) 8 μM zeocin for 24 hours. GUS-stained samples were observed for whole seedlings (B), root tips (C), and cotyledons and shoot apices (D). A cross section around the transition zone of zeocin-treated pLOG7:GUS is shown in (E). Arrowheads in (C) indicate the boundary between the meristematic zone and the transition zone. Scale bars, 1 cm (B), 100 μm (C), 1 mm (D), and 20 μm (E). Photo credits: Masaaki Umeda, Nara Institute of Science and Technology.

We then observed the expression patterns of the DSB-induced genes in tissues using the promoter:GUS reporter lines. IPT1, which is known to be expressed in the procambium (32), displayed increased expression in zeocin-treated root tips (Fig. 3C). The promoter activities of IPT3 and IPT5 were elevated in cotyledons and shoot apices, respectively, but not detected in the root tip (Fig. 3, B to D). pIPT7:GUS showed no GUS signal regardless of zeocin treatment, probably because the promoter region used for the reporter construction lacks essential cis-element(s). Zeocin highly induced CYP735A2 and LOG7 in the vasculature and in the epidermis and cortex, respectively, around the transition zone (Fig. 3, C and E). Consistent with the TCSn:GFP expression (fig. S3), the induction of CYP735A2 and LOG7 was observed after 12-hour zeocin treatment (fig. S9). These results suggest that zeocin taken up by roots induces IPT3 and IPT5 in shoots and provides more cytokinin precursors and that the induction of CYP735A2 and LOG7 around the transition zone leads to higher accumulation of iP- and trans-zeatin–type cytokinins in the root tip, as revealed by our cytokinin measurement (Fig. 2). The induction of pLOG7:GUS was also observed under DSB-causing aluminum stress (fig. S10).

We then examined whether the induction of cytokinin biosynthesis genes is involved in root growth inhibition in response to DNA damage. Since the ipt1;3;5;7 quadruple mutant exhibits a severe growth defect (17), we used the ipt3-2;5-1;7-1 triple mutant together with cyp735a2-1 and log7-1. When 5-day-old seedlings were transferred onto a medium containing 8 μM zeocin, root growth was less severely inhibited in the three mutants than in WT (Fig. 4A). ipt3-2;5-1;7-1 roots, which showed the highest zeocin tolerance, also grew faster than WT in the absence of zeocin (Fig. 4A), supporting the previous report that cytokinin precursors and active forms were markedly reduced in the triple mutant (17). Counting the cortical cell number in the meristematic zone showed that after zeocin treatment for 24 hours, the meristem size was reduced to 50% in WT, but only to 96, 78, and 79% in ipt3-2;5-1;7-1, cyp735a2-1, and log7-1, respectively (Fig. 4, B and C). We also measured the area of dead cells in the vasculature of PI-stained root tips. As shown in Fig. 4D, cell death area was markedly reduced in the three mutants as compared to WT. These results suggest that, although enhanced production of cytokinin precursors by IPT3 and IPT5 in shoots contributes to DSB-induced cytokinin accumulation in roots, CYP735A2- and LOG7-mediated synthesis of iP and trans-zeatin around the transition zone is crucial for root growth inhibition, reduction of the meristem size, and induction of stem cell death.

Fig. 4. Roots of cytokinin biosynthesis mutants are tolerant to DSBs.

(A) Root growth of WT, ipt3-2;5-1;7-1, cyp735a2-1, and log7-1. Five-day-old seedlings were transferred to MS plates supplemented with (+ zeocin) or without (− zeocin) 8 μM zeocin, and root length was measured every 24 hours. Data are presented as means ± SD (n > 20). Significant differences from WT were determined by Student’s t test, ***P < 0.001. (B) Images of root tips. Five-day-old seedlings were treated with (+) or without (−) 8 μM zeocin for 24 hours and subjected to PI staining. Arrowheads indicate the QC (bottom) and the boundary between the meristematic zone and the transition zone (top). Scale bar, 100 μm. (C and D) Cortical cell number in the meristematic zone and cell death area in vascular stem cells and their daughters. Five-day-old seedlings were treated with (+ zeocin) or without (− zeocin) 8 μM zeocin for 24 hours. The number of cortical cells between the QC and the first elongated cell was counted (C). The area of PI-stained dead cells in roots grown in the presence of zeocin was measured using ImageJ software (D). Data are presented as means ± SD (n > 20). Significant differences from the control without zeocin treatment (C) and WT (D) were determined by Student’s t test, **P < 0.01 and ***P < 0.001.

Enhanced cytokinin signaling around the transition zone is involved in DSB-induced meristem size reduction and stem cell death

Cytokinin signaling is up-regulated around the transition zone of Arabidopsis roots due to localized expression of type-B ARRs (28). Therefore, we speculated that a reduction of the meristem size under DNA damage conditions is a consequence of enhanced cytokinin signaling around the transition zone, which is caused by higher cytokinin accumulation in the root tip. To test this possibility, we generated transgenic plants expressing CKX1, which encodes a cytokinin-degrading enzyme, in the transition zone under the RCH2 promoter (fig. S11A). When 5-day-old seedlings were transferred onto zeocin-containing medium, root growth was only mildly inhibited in the two independent transgenic lines as compared with WT (fig. S11B). In the absence of zeocin, the cortical cell number in the meristematic zone was higher in pRCH2:CKX1 than in WT (Student’s t test, P < 0.05, n > 20) (fig. S11, C and D), matching a previous report (28). The transgenic lines displayed less reduction of the meristem size after zeocin treatment, and similar lower sensitivity was also observed in zeocin-induced cell death that rapidly occurs in vascular stem cells and their daughters (fig. S11, C to E). These results suggest that the activation of cytokinin signaling around the transition zone is associated with meristem size reduction and stem cell death after DNA damage.

DSBs reduce auxin signaling in the root tip

Under normal growth conditions, one of the causes of meristem size restriction is the ARR1/12-mediated induction of SHORT HYPOCOTYL 2 (SHY2)/IAA3, which encodes a member of the Aux/IAA protein family of auxin signaling repressors, around the transition zone (23). SHY2 promoter activity was elevated around the transition zone in response to zeocin treatment (fig. S12A). Moreover, the shy2-31 loss-of-function mutant, which has a larger meristem than WT (23), exhibited higher tolerance to DNA damage: The reduction of the meristem size and the cell death induction by zeocin were partially suppressed in shy2-31 (Student’s t test, P < 0.05, n > 20) (fig. S12, B to D).

DSB-induced cell death occurs around the stem cell niche except in the SHY2 expression domain, suggesting the possibility that SHY2-dependent down-regulation of PIN expression inhibits auxin transport and causes stem cell death as well as meristem size reduction (23). To test this possibility, we then examined the response of PIN genes to DNA damage. In roots, PIN1 and PIN4 are required for downward auxin flow in the stele, and PIN2 functions in upward transport in the lateral root cap (LRC) and the epidermis (33). PIN3 and PIN7 regulate downward auxin flow in the stele and redirection in the columella for lateral transport to the LRC and the epidermis (33). Our qRT-PCR data showed that the transcript levels of PIN1, PIN3, and PIN4, but not PIN2 or PIN7, were reduced by 8 μM zeocin treatment (Fig. 5A). This result was supported by the analysis of pPIN:PIN-GFP reporter lines; in the absence of zeocin, PIN1-GFP and PIN3/4-GFP accumulate in the apical and basal parts of the vasculature, respectively (33), whereas the GFP fluorescence was markedly decreased by zeocin treatment (Fig. 5B and fig. S13). It is noteworthy that the PIN1-GFP signal diminished in the meristematic zone, although SHY2 is induced in the transition zone, but not in the meristematic zone (fig. S12A). Since PIN1 expression is up-regulated by auxin in a time- and concentration-dependent manner (34), it is likely that SHY2-mediated PIN3 and PIN4 (and PIN1) repression around the transition zone disturbs downward auxin flow, thereby reducing the auxin level and suppressing PIN1 expression in the meristematic zone. On the other hand, no change in expression patterns/levels of PIN2-GFP or PIN7-GFP was observed after zeocin treatment (Fig. 5B).

Fig. 5. DSBs inhibit expression of PIN1, PIN3, and PIN4.

(A) Transcript levels of PIN1, PIN2, PIN3, PIN4, and PIN7. Five-day-old WT seedlings were transferred to MS plates containing 8 μM zeocin and grown for 0, 12, and 24 hours. Total RNA was extracted from roots and subjected to qRT-PCR. Transcript levels of PIN1, PIN2, PIN3, PIN4, and PIN7 were normalized to that of ACTIN2 and are indicated as relative values, with that for 0 hours set to 1. Data are presented as means ± SD calculated from three biological and technical replicates. Significant differences from the 0-hour sample were determined by Student’s t test, *P < 0.05 and ***P < 0.001. (B) Zeocin response of pPIN1:PIN1-GFP, pPIN2:PIN2-GFP, pPIN3:PIN3-GFP, pPIN4:PIN4-GFP, and pPIN7:PIN7-GFP. Five-day-old seedlings were treated with (+ zeocin) or without (− zeocin) 8 μM zeocin for 24 hours. GFP fluorescence was observed after counterstaining with PI. Arrowheads indicate the boundary between the meristematic zone and the transition zone. Scale bar, 100 μm.

We next investigated the auxin response using the auxin-responsive marker DR5rev:GFP (35) and the GFP reporter driven by the IAA2 promoter (20), which is one of the primary downstream targets of auxin signaling. As shown in Fig. 6A, the DR5rev:GFP signal in the QC and columella cells, and the pIAA2:GFP signal in the vasculature, QC, and columella cells, were reduced by 24-hour zeocin treatment, indicating that DSBs decrease auxin signaling in the meristematic zone, QC, and columella. We also used the R2D2 reporter line, in which the mDII-tdTomato:DII-3xVenus ratio indicates relative auxin levels (36). Our observation revealed that zeocin treatment reduced the auxin level in the root tip (Fig. 6, B and C), corresponding with the decrease in auxin signaling.

Fig. 6. DSBs inhibit auxin transport and signaling in the root tip.

(A) Zeocin response of DR5rev:GFP and pIAA2:GFP. Five-day-old seedlings were treated with (+ zeocin) or without (− zeocin) 8 μM zeocin for 24 hours. Magnified images of the areas marked by white boxes in DR5rev:GFP are shown on the right. Arrowheads indicate the boundary between the meristematic zone and the transition zone. Scale bars, 100 μm. (B and C) Zeocin response of the R2D2. Five-day-old seedlings were treated with (+ zeocin) or without (− zeocin) 8 μM zeocin for 24 hours, and DII-3xVenus and mDII-tdTomato were observed (B). The fluorescence was measured in the meristematic zone, and the ratio of mDII-tdTomato:DII-3xVenus is shown (C). Data are presented as means ± SD (n > 8). The significant difference from the control without zeocin treatment was determined by Student’s t test, *P < 0.05. Scale bar, 100 μm. (D) Cortical cell number in the meristematic zone. Five-day-old seedlings were treated with or without 8 μM zeocin and/or 10 μM naphthylphthalamic acid (NPA) for 24 hours. The number of cortical cells between the QC and the first elongated cell was counted. Data are presented as means ± SD (n > 18). Bars with different alphabetical letters are significantly different from each other (Student’s t test, P < 0.05). (E) IAA2 transcript level in pin1;3;4. Five-day-old WT and pin1;3;4 seedlings were transferred to MS plates with or without 8 μM zeocin and grown for 24 hours. Transcript levels of IAA2 were normalized to that of UBQ10 and are indicated as relative values, with that for the WT control without zeocin treatment set to 1. Data are presented as means ± SD calculated from three biological and technical replicates. Significant differences from the control without zeocin treatment were determined by Student’s t test, ***P < 0.001.

To examine the involvement of auxin regulation in the DRR, we used naphthylphthalamic acid (NPA), which inhibits polar auxin transport (37). When WT roots were treated with 10 μM NPA, the meristem size was reduced to the same extent as with 8 μM zeocin treatment (Fig. 6D). However, cotreatment with zeocin and NPA did not generate an additive effect, suggesting that down-regulation of auxin level in the root tip is crucial for DSB-induced meristem size reduction. Since the cellular pattern in the root tip of the pin1;3;4 mutant is severely disorganized (38), we could not estimate phenotypic defects in the DNA damage responses. However, in the triple mutant that exhibits lower IAA2 expression than WT, zeocin did not further reduce the expression level, suggesting that suppression of PIN1, PIN3, and PIN4 is involved in the decrease of auxin signaling in the root tip after zeocin treatment (Fig. 6E). Notably, zeocin-induced repression of PIN1, PIN3, PIN4, and IAA2 was not observed in the atm-2 or sog1-1 mutant (fig. S14).

Activation of cytokinin biosynthesis is crucial for reducing auxin signaling in the meristem

We then asked whether reduced auxin signaling in the meristematic zone is caused by enhanced cytokinin biosynthesis. Our expression analyses showed that zeocin-triggered repression of PIN1, PIN3, PIN4, and IAA2 was suppressed in ipt3-2;5-1;7-1, cyp735a2-1, and log7-1 (Fig. 7A). Moreover, PIN1-GFP accumulation was not decreased in log7-1 after zeocin treatment (Fig. 7B). As described above, LOG7 is specifically induced around the transition zone by DSBs (Fig. 3C), suggesting that enhanced biosynthesis of active cytokinins around the transition zone plays a pivotal role in perturbing downward auxin flow and decreasing auxin signaling in the meristematic zone.

Fig. 7. LOG7 is involved in DSB-induced suppression of auxin flow and cell division.

(A) Transcript levels of PIN1, PIN3, PIN4, and IAA2 in ipt3-2 ipt5-1 ipt7-1, cyp735a2-1, and log7-1. Five-day-old seedlings were transferred to MS plates with or without 8 μM zeocin and grown for 24 hours. Total RNA was extracted from roots and subjected to qRT-PCR. Transcript levels of PIN1, PIN3, PIN4, and IAA2 were normalized to that of ACTIN2 and are indicated as relative values, with that for the control without zeocin treatment set to 1. Data are presented as means ± SD calculated from three biological and technical replicates. Significant differences from the control without zeocin treatment were determined by Student’s t test, *P < 0.05 and ***P < 0.001. (B) PIN1-GFP expression in log7. Five-day-old seedlings of WT and log7-1 harboring pPIN1:PIN1-GFP were treated with (+ zeocin) or without (− zeocin) 8 μM zeocin for 24 hours. GFP fluorescence was observed after counterstaining with PI. Arrowheads indicate the boundary between the meristematic zone and the transition zone. Scale bar, 100 μm. (C) G₂ progression in log7. Five-day-old seedlings of WT and log7-1 were transferred to MS plates with or without 8 μM zeocin and grown for 12 hours. After pulse labeling with 20 μM EdU for 15 min, seedlings were transferred back to MS plates with or without 8 μM zeocin and grown for 0, 4, and 6 hours. Cells in the meristematic zone were double-stained with ethynyl deoxyuridine (EdU) and 4′,6-diamidino-2-phenylindole (DAPI), and the percentage of EdU-labeled cells among those with mitotic figures was calculated. Data are presented as means ± SD (n > 8). Significant differences from the control without zeocin treatment were determined by Student’s t test, ***P < 0.001.

Reduced auxin signaling triggers both G₂ arrest and stem cell death

We previously reported that DSBs arrest the cell cycle preferentially at G₂ in dividing cells (16). To test whether enhanced cytokinin biosynthesis is involved in G₂ arrest in the meristematic zone, we conducted 5-ethynyl-2′-deoxyuridine (EdU) incorporation experiments to monitor cell cycle progression. EdU is incorporated into newly synthesized DNA during the S phase. After EdU-labeled cells pass through G₂, cells that enter mitosis display mitotic figures (12, 15). Five-day-old WT and log7-1 seedlings treated with or without 8 μM zeocin for 12 hours were incubated with EdU for 15 min, and the number of EdU-labeled cells with mitotic figures in the meristematic zone was counted after 4 and 6 hours. In WT, the percentage of these cells was significantly reduced in zeocin-treated roots, indicating retardation of G₂ progression (Fig. 7C). However, in log7-1, we could not find any difference between zeocin-treated and nontreated samples (Fig. 7C). These data suggest that activation of cytokinin biosynthesis is associated with DSB-induced G₂ arrest in the meristematic zone.

We next asked whether reduced auxin signaling is the cause of DSB-induced G₂ arrest. To address this issue, we treated WT roots with indole-3-acetic acid (IAA) at 5 nM, a very low concentration that did not change the meristem size at all (Fig. 8, A and B). When 5 nM IAA was applied together with 8 μM zeocin, LOG7 induction was observed similarly to the sole zeocin treatment (fig. S15), suggesting that 5 nM IAA does not affect DSB-dependent activation of cytokinin biosynthesis. However, IAA treatment partially suppressed zeocin-induced meristem size reduction, and EdU incorporation experiments showed that G₂ arrest was also suppressed in the presence of both zeocin and IAA (Fig. 8, A to C). Zeocin-induced cell death was rarely observed in IAA-treated roots (Fig. 8, A and D), suggesting that a reduction in auxin signaling causes stem cell death and G₂ arrest under DNA damage conditions.

Fig. 8. DSB-induced reduction in auxin signaling causes G₂ arrest.

(A and B) Root meristem size after treatment with zeocin and/or IAA. Five-day-old WT seedlings were transferred to MS plates supplemented with or without 8 μM zeocin and/or 5 nM IAA and grown for 24 hours. Arrowheads indicate the QC (bottom) and the boundary between the meristematic zone and the transition zone (top). Scale bar, 100 μm (A). The number of cortical cells between the QC and the first elongated cell was counted (B). Data are presented as means ± SD (n > 20). Bars with different alphabetical letters are significantly different from each other (Student’s t test, P < 0.01). (C) G₂ progression in the presence of zeocin and/or IAA. Five-day-old WT seedlings were transferred to MS plates with or without 8 μM zeocin and/or 5 nM IAA and grown for 12 hours. After pulse labeling with 20 μM EdU for 15 min, seedlings were transferred back to MS plates with or without 8 μM zeocin and/or 5 nM IAA and grown for 0, 4, and 6 hours. Cells in the meristematic zone were double-stained with EdU and DAPI, and the percentage of EdU-labeled cells among those with mitotic figures was calculated. Data are presented as means ± SD (n > 8). Significant differences from the control without zeocin or IAA were determined by Student’s t test, *P < 0.05 and ***P < 0.001. (D) Cell death area in vascular stem cells and their daughters. Five-day-old WT seedlings were treated with or without 8 μM zeocin and/or 5 nM IAA for 24 hours, and the area of PI-stained dead cells was measured. Data are presented as means ± SD (n > 20). Bars with different alphabetical letters are significantly different from each other (Student’s t test, P < 0.01).

A possible scenario is that the DSB-induced reduction of the auxin level triggers stem cell death, which consequently inhibits cell cycle progression in the meristematic zone. To examine this possibility, we activated auxin signaling in the vasculature under DNA stress. In the auxin receptor mutant tir1-1, we expressed ccvTIR1, which encodes a modified receptor that specifically recognizes the synthetic auxin cvxIAA and transmits the auxin signal (39) under the WOODENLEG promoter that is active in vascular stem cells and the stele (40). In the transgenic plants, auxin signaling is activated within the vascular tissue by cvxIAA treatment. Note that treatment with 10 nM cvxIAA alone affected neither the meristem size nor the G₂ progression in WT, tir1-1, or three independent transgenic lines (nos. 10, 25, and 34) (fig. S16, A and B). In the transgenic lines, 8 μM zeocin application reduced the meristem size and induced G₂ arrest irrespective of cvxIAA treatment, as observed in WT and tir1-1 (fig. S16, A and B). This result is reasonable because we counted the meristem cell number and EdU-labeled cells with mitotic figures in the cortex and the epidermis, respectively, but not in the vasculature where ccvTIR1 was expressed. However, zeocin-induced cell death was significantly suppressed by cvxIAA treatment in the transgenic lines (fig. S16C), indicating that up-regulation of auxin signaling in vascular stem cells suppresses zeocin-induced stem cell death, although this suppression cannot restore the defect in G₂ progression in the meristematic zone. In contrast, when ccvTIR1 was expressed using the TIR1 promoter, which is active throughout the root tip (41), cvxIAA treatment could suppress not only zeocin-induced cell death but also meristem size reduction and G₂ arrest (fig. S16, D to F). These results suggest that DSB-induced cell cycle arrest requires a reduction in auxin signaling but is not a consequence of stem cell death.

ARR2-CCS52A1 promotes an early onset of endoreplication in response to DNA damage

As mentioned above, the shy2-31 mutation suppressed zeocin-induced meristem size reduction, but only partially (fig. S12, B and C). This suggests that in addition to SHY2-mediated inhibition of auxin signaling, another mechanism functions in meristem size control, probably through regulating the onset of endoreplication. We previously reported that cytokinin-activated ARR2, which is specifically expressed around the transition zone, induces the expression of CELL CYCLE SWITCH PROTEIN 52 A1 (CCS52A1), encoding an activator of the E3 ubiquitin ligase anaphase-promoting complex/cyclosome (APC/C), and promotes degradation of mitotic cyclins, thereby enhancing the transition from cell division to endoreplication (29). Since cytokinins are elevated by DSBs as described above, it seemed likely that the ARR2-CCS52A1 pathway is associated with an early onset of endoreplication under DNA stress. To test this possibility, we first observed the promoter activity of CCS52A1 using a GFP reporter line. As shown in Fig. 9A, the GFP signal in epidermal and cortical cells around the transition zone was significantly elevated after zeocin treatment. qRT-PCR revealed that CCS52A1 transcripts increased by about fourfold in the presence of zeocin, while no such induction was observed in atm-2, sog1-1, log7-1, or arr2-4 (Fig. 9, B and C), suggesting that CCS52A1 is up-regulated through ARR2-mediated cytokinin signaling, which is activated by the higher accumulation of cytokinins. Zeocin-induced meristem size reduction was partially suppressed in the ccs52a1-1 or arr2-4 knockout mutant, and a similar level of suppression was observed for the ccs52a1-1 arr2-4 double mutant (Fig. 9D), suggesting that the ARR2-CCS52A1 pathway enhances endoreplication onset and contributes to meristem size reduction.

Fig. 9. The ARR2-CCS52A1 pathway participates in meristem size reduction caused by DSBs.

(A) CCS52A1 expression in the presence of zeocin. Five-day-old pCCS52A1:NLS-GFP seedlings were transferred to MS plates with (+ zeocin) or without (− zeocin) 8 μM zeocin and grown for 24 hours. GFP fluorescence was observed after counterstaining with PI. Arrowheads indicate the boundary between the meristematic zone and the transition zone. Magnified images of the areas marked by white boxes are shown on the right. Scale bars, 100 μm. (B and C) Transcript levels of CCS52A1 in atm, sog1, log7, and arr2. Five-day-old seedlings were transferred to MS plates with or without 8 μM zeocin and grown for 24 hours. Total RNA was extracted from roots and subjected to qRT-PCR. Transcript levels of CCS52A1 were normalized to that of ACTIN2 and are indicated as relative values, with that for the control without zeocin treatment set to 1. Data are presented as means ± SD calculated from three biological and technical replicates. Significant differences from the control without zeocin treatment were determined by Student’s t test, ***P < 0.001. (D to F) Cortical cell number in the meristematic zone. Five-day-old seedlings of WT, ccs52a1-1, arr2-4, shy2-31, ccs52a1-1 arr2-4, arr2-4 shy2-31, and ccs52a1-1 shy2-31 were transferred to MS plates with or without 8 μM zeocin and grown for 24 hours. The number of cortical cells between the QC and the first elongated cell was counted. Data are presented as means ± SD (n > 20). Significant differences from the control without zeocin treatment were determined by Student’s t test, *P < 0.05, **P < 0.01, and ***P < 0.001.

To reveal whether the SHY2-dependent signaling and the ARR2/CCS52A1-mediated pathway cooperate in DSB-induced meristem size reduction, we performed genetic experiments. After zeocin treatment, the root meristem size was reduced to 44% in WT and to 63 and 55% in arr2-4 and shy2-31, respectively, but only to 76% in the arr2-4 shy2-31 double mutant (Fig. 9E). Similarly, the ccs52a1-1 shy2-31 double mutant was less sensitive to zeocin than either single mutant (Fig. 9F). These results suggest that both ARR1/12-SHY2 and ARR2-CCS52A1 pathways function in meristem size control under DNA stress.

DISCUSSION

In this study, we revealed that DSB-stimulated ATM and SOG1 enhance cytokinin biosynthesis. As a result, cytokinin signaling is activated around the transition zone, and SHY2 is up-regulated to repress PIN1, PIN3, and PIN4, thereby perturbing downward auxin flow. Reduction of auxin causes cell cycle arrest in the meristem and stem cell death. At the same time, ARR2 induces CCS52A1 and promotes an early onset of endoreplication around the boundary between the meristematic zone and the transition zone. The combined effects of SHY2- and CCS52A1-dependent pathways reduce the root meristem size and, together with stem cell death, lead to root growth inhibition (Fig. 10). In plants, selective death of stem cells is a characteristic feature of the DNA damage response; in mammals, normal somatic cells undergo cell death in response to severe DNA damage (42). The genome integrity of stem cells needs to be highly maintained to ensure postembryonic organ development in plants. Therefore, stem cell death and the immediate supply of new stem cells from the organizing center, in which DNA repair genes are highly expressed (43), are a reasonable strategy to cope with DNA stress. On the other hand, cell death can disrupt the local tissue structure because plant cells do not migrate within tissues; thus, plants may prevent normal somatic cells from undergoing cell death and, instead, activate the cell cycle checkpoint and accelerate the transition to endoreplication, thereby reducing the number of dividing cells, which are particularly sensitive to DNA damage. The present study clearly indicates that two plant hormones, cytokinin and auxin, orchestrate the differential DNA damage responses in roots, enabling the maintenance of genome integrity and continuous organ growth.

Fig. 10. Model for root meristem size reduction and stem cell death in response to DSBs.

DSBs induce IPT3 and IPT5 in shoots and CYP735A2 and LOG7 in roots through the ATM-SOG1 pathway. Around the transition zone, conversion from iP- to trans-zeatin–type cytokinins and the final activation step of cytokinin biosynthesis are enhanced, resulting in an increase of active cytokinins. Elevated cytokinin signaling up-regulates SHY2 and suppresses PIN1, PIN3, and PIN4 expression, thereby inhibiting downward auxin flow and reducing the auxin level in the meristematic zone. This leads to G₂ arrest of dividing cells and stem cell death. Increased cytokinin signaling also up-regulates CCS52A1 around the boundary between the meristematic zone and the transition zone and promotes degradation of mitotic cyclins, triggering an early onset of endoreplication. The combinatorial effect of G₂ arrest and an early transition from cell division to endoreplication causes meristem size reduction. TZ, transition zone; MZ, meristematic zone.

It has been reported recently that cytokinin-activated ARR1 regulates root meristem size through the control of auxin quantity in the LRC by induction of GH3.17 and PIN5, which inactivate IAA and pump auxin from the cytoplasm into the endoplasmic reticulum, respectively (44). In the present study, we have not investigated the involvement of GH3.17 or PIN5 in the DNA damage response, but it is possible that enhanced cytokinin biosynthesis affects the auxin level in the LRC and reduces the meristem size. We previously reported that DSBs inhibit lateral root formation (24), although how and to what extent hormonal signals are associated with this response have remained unknown. It is probable that cytokinin signaling activated by DNA damage inhibits auxin flow, in a manner similar to that shown in this study, and disturbs the proper organization of the LR meristem. Further studies will answer the question of whether the mechanisms identified here are responsible for controlling growth of the whole root system under genotoxic stress conditions.

One of the key findings in this study is that DNA damage induces the cytokinin biosynthesis genes IPT1, IPT3, IPT5, IPT7, LOG7, and CYP735A2. Their promoters do not have the consensus sequence for SOG1 binding (10); therefore, downstream transcription factor(s) may be involved in their induction. However, we cannot exclude the possibility that some of these cytokinin biosynthesis genes are directly regulated by SOG1, because SOG1 can still bind to sequences with a few nucleotide substitutions in the consensus motif (10). The promoter:GUS lines showed that IPT1 expression increased in zeocin-treated roots, while IPT3 and IPT5 were induced in cotyledons and shoot apices, respectively (Fig. 3). Since the ipt3-2;5-1;7-1 triple mutant exhibited a zeocin-tolerant phenotype in roots (Fig. 4), it is conceivable that cytokinin precursors produced by IPT3 and IPT5 are transported from shoots to roots and promote the DNA damage response in the root tip (Fig. 10). In support of this hypothesis, the levels of cytokinin precursors were significantly elevated in the root tip after zeocin treatment (Fig. 2). However, it should be noted that zeocin in the solid medium is efficiently taken up by roots in our experimental system; therefore, how IPT3/5 induction in shoots contributes to DNA damage responses in roots under natural environmental conditions remains obscure. DSBs also induce CYP735A2 and LOG7 specifically around the transition zone (Fig. 3), indicating that conversion from iP- to trans-zeatin–type cytokinins and the final activation step of cytokinin biosynthesis are enhanced around the transition zone. Note that LOG7 has the highest specificity constant (K_cat/K_m) among the eight LOG proteins in Arabidopsis (19), suggesting that LOG7 induction facilitates efficient production of biologically active cytokinins around the transition zone, thereby enabling a rapid and strong response to DNA stress in terms of the control of meristem size and stem cell death. The log7 mutant exhibited clear suppression in meristem size reduction; G₂ arrest; stem cell death; decreased expression of PIN1, PIN3, and PIN4; and induction of CCS52A1 (Figs. 4, 7, and 9).

We previously reported that Rep-MYBs, which repress the expression of G₂-M–specific genes, play an essential role in inhibiting G₂ progression in response to DNA damage (12). Under normal growth conditions, CDKs phosphorylate Rep-MYBs and promote their degradation through the ubiquitin-proteasome pathway, leading to a release of transcriptional repression of G₂-M–specific genes. In response to DNA damage, CDK activity is reduced, and Rep-MYBs are stabilized to cause G₂ arrest (12). This model indicates that how CDK activity is suppressed is a key to trigger G₂ arrest when roots are exposed to DNA stress. The present study showed that DSBs activate cytokinin signaling around the transition zone and inhibit downward auxin flow, thereby decreasing the auxin level in the meristem, and that DSB-induced G₂ arrest occurred independently of stem cell death. Therefore, we propose that a decline in auxin level has a central role in reducing CDK activity and causing G₂ arrest. It has been shown that in Arabidopsis, auxin up-regulates the expression of CDKA;1, which encodes the functional ortholog of yeast Cdc2/Cdc28p (45). Genes for A2-type cyclins (CYCA2s), and for the CDK inhibitors KIP-RELATED PROTEIN 1 (KRP1) and KRP2, are induced and repressed, respectively, by exogenous auxin treatment (46), and the auxin antagonist α-(phenyl ethyl-2-one)-IAA (PEO-IAA) rapidly decreased the transcript levels of CDKB1;1 and CYCA2;3 (47). Moreover, the heterodimeric transcription factor E2 PROMOTER BINDING FACTOR (E2F)-DIMERIZATION PARTNER (DP) is known to mediate auxin signaling and directly induce CDKB1;1, representing a clear connection between auxin and CDK expression (48, 49). These observations suggest that a decline in auxin level leads to a reduction in overall CDK activities, thereby enabling stabilization of Rep-MYBs. Rep-MYBs are phosphorylated by all types of CDKs (A-, B1-, and B2-types) in Arabidopsis (12). Our recent study revealed that the transcription factors ANAC044 and ANAC085 are also involved in the control of Rep-MYB stability under DNA stress (15); therefore, the next important question is how reduced auxin level and the ANAC044/085-dependent pathway cooperate in DNA damage–induced cell cycle arrest, which specifically occurs at G₂ phase.

Here, we revealed that a decrease in auxin signaling is one of the causes of stem cell death and cell cycle arrest under DNA stress. Canher et al. (50) recently reported that DSB-induced cell death obstructs auxin flow and facilitates auxin accumulation around dead cells, thereby enhancing stem cell regeneration. In this model, the regeneration process starts after cell death occurs, representing a scenario in which DSBs first reduce the auxin level in the meristematic zone and the stem cell niche, followed by a local auxin accumulation that contributes to stem cell regeneration. A key question regarding the initial event triggered by DNA damage is how reduced auxin signaling induces stem cell death. A previous study showed that treatment of cultured tobacco BY-2 cells with the auxin antagonist PEO-IAA resulted in chromatin relaxation (51) and that Arabidopsis mutants defective in the histone chaperone chromatin assembly factor-1 (CAF-1), which is involved in heterochromatin formation, accumulated high levels of DNA damage and frequently exhibited cell death in the root tip (52). Therefore, it is possible that reduced auxin signaling leads to chromatin decondensation, which makes the genome hypersensitive to DNA damage and promotes cell death. This is a fascinating hypothesis, but further research is essential to reveal how auxin controls the chromatin structure and the causal relationship between chromatin accessibility and DNA damage accumulation. Another interesting question is why only stem cells undergo cell death, although the auxin level decreases all over the meristematic zone. Lozano-Elena et al. (53) demonstrated that DSB-induced stem cell death was suppressed in the Arabidopsis mutant of BRASSINOSTEROID INSENSITIVE 1 (BRI1), which encodes one of the brassinosteroid receptors, implying that another layer of regulation is necessary to give rise to stem cell–specific cell death. Further studies will deepen our understanding of how plants accomplish postembryonic organ development through maintaining genome stability under fluctuating environmental conditions.

MATERIALS AND METHODS

Plant growth conditions

Arabidopsis thaliana [ecotype Columbia-0 (Col-0)] plants were grown vertically under continuous light conditions at 22°C on Murashige and Skoog (MS) plates [0.5× MS salts, 2-(N-morpholino)ethanesulfonic acid (0.5 g/liter), 1% sucrose, and 1.2% phytoagar (pH 6.3)]. For DNA damage treatments, 5-day-old seedlings were transferred to new MS plates supplemented with or without 8 μM zeocin (Invitrogen Life Technologies) or bleomycin (0.6 μg/ml). For aluminum treatments, 5-day-old seedlings were transferred to new MS plus 5 mM succinic acid plates (pH 4.2) supplemented with or without 1.5 mM aluminum.

Plant materials and constructs

pARR5:GUS (25), TCSn:GFP (27), atm-2 (4), sog1-1 (7), pAHP6:GFP (54), pAPL:GFPer (55), pCO2:H2B-YFP (56), pWOX5:NLS-YFP (57), pSCR:GFP-SCR (56), pRCH1:GFP (28), pRCH2:CFP (28), pARR1:ARR1-GUS (11), pARR2:ARR2-GUS (11), pIPT1:GUS (32), pIPT3:GUS (32), pIPT5:GUS (32), pCYP735A2:GUS (18), pLOG7:GUS (19), ipt3-2;5-1;7-1 (17), cyp735a2-1 (18), log7-1 (19), pSHY2:GUS (58), pPIN1:PIN1-GFP (33), pPIN2:PIN2-GFP (33), pPIN3:PIN3-GFP (33), pPIN4:PIN4-GFP (33), pPIN7:PIN7-GFP (33), DR5rev:GFP (35), pIAA2:GFP (20), R2D2 (36), pin1-1;3-5;4-3 (38), pCCS52A1:NLS-GFP (29), ccs52a1-1 (29), arr2-4 (29), and shy2-31 (29) were described previously. sog1-1 (Col-0/Ler background) was backcrossed to Col-0 WT three times before use. atm-2 and sog1-1 carrying pARR5:GUS were generated by crossing the mutants and the pARR5:GUS line. Similarly, pPIN1:PIN1-GFP log7-1 was generated by crossing pPIN1:PIN1-GFP and log7-1. The full-length open reading frame of CKX1 was amplified from Arabidopsis cDNA by PCR with F_CKX1 (5′-AAAAAGCAGGCTTCATGGGATTGACCTCATCCTTAC-3′) and R_CKX1 (5′-AGAAAGCTGGGTCTTATACAGTTCTAGGTTTCGGC-3′) primers and cloned into the pDONR221 entry vector (Thermo Fisher Scientific) by a BP recombination reaction according to the manufacturer’s instructions. The RCH2 promoter was amplified from Arabidopsis genomic DNA by PCR with FP_RCH2 (5′-ATAGAAAAGTTGGGAGGTAAGAATCATGAGAGTGGAG-3′) and RP_RCH2 (5′-TTGTACAAACTTGACATTTGCCTCAAAATACGAAAAGAAG-3′) primers and cloned into the pDONRP4P1R entry vector (Thermo Fisher Scientific) by a BP recombination reaction. To generate the fusion construct of the RCH2 promoter and CKX1, the entry clones were mixed with the R4pGWB601 destination vector (59) in an LR recombination reaction according to the manufacturer’s instructions (Thermo Fisher Scientific). In addition, the TIR1 open reading frame was PCR amplified from Arabidopsis cDNA with F_TIR1 (5′-AAAAAGCAGGCTTCATGCAGAAGCGAATAGCCTTGTCG-3′) and R_TIR1 (5′-AGAAAGCTGGGTCTTATAATCCGTTAGTAGTAATGATTTGC-3′) primers and cloned into the pDONR221 entry vector by a BP recombination reaction. To make the ccvTIR1 entry clones, PCR-mediated site-directed mutagenesis was conducted using the TIR1 entry vector as a template with F_mTIR1 (5′-GAAAACCTCACGGTGCTGACTTTAATTTGGTACCTGAC-3′) and R_mTIR1 (5′-CAAATTAAAGTCAGCACCGTGAGGTTTTCCTTTAAGCTCCAC-3′) primers. The WOL and TIR1 promoter were PCR amplified from Arabidopsis genomic DNA with FP_WOL (5′-ATAGAAAAGTTGGGCAACCCAAATACGAAATACTCGTCC-3′) and RP_WOL (5′-TTGTACAAACTTGAATCTGAGCTACAACAATAGAGAAC-3′) for the WOL promoter, and FP_TIR1 (5′-ATAGAAAAGTTGGGGAGTACGAAACCCGAGACTAGGAG-3′) and RP_TIR1 (5′-TTGTACAAACTTGATGCGGCCAAATAACCTCGAGATC-3′) for the TIR1 promoter, and cloned into the pDONRP4P1R entry vector (Thermo Fisher Scientific) by a BP recombination reaction. To generate the pWOL:ccvTIR1 and pTIR1:ccvTIR1 constructs, entry clones were mixed with the R4pGWB601 destination vector (59) in an LR recombination reaction. All constructs were introduced into the Agrobacterium tumefaciens GV3101 strain harboring the plasmid pMP90. The obtained strains were used to generate stably transformed plants by the floral dip transformation method.

Quantitative RT-PCR

Total RNA was extracted from root tips or whole seedlings with the Plant Total RNA Mini Kit (Favorgen Biotech). First-strand cDNA was prepared from total RNA with ReverTra Ace (Toyobo) according to the manufacturer’s instructions. For quantitative PCR, a THUNDERBIRD SYBR qPCR mix (Toyobo) was used with 100 nM primers and first-strand cDNA. PCR reactions were run on a LightCycler 480 real-time PCR system (Roche) according to the following conditions: 95°C for 5 min, 45 cycles at 95°C for 10 s, 60°C for 10 s, and 72°C for 15 s. Transcript levels were normalized to that of ACTIN2. Three biological and technical replicates were performed for each experiment. The following primers were used: 5′-CTGGATCGGTGGTTCCATTC-3′ and 5′-CCTGGACCTGCCTCATCATAC-3′ for ACTIN2, 5′-AGAGATCACAACGAATCAGATTACGT-3′ and 5′-ATGACGCCGAGGAGATGGT-3′ for IPT1, 5′-CGGGTTCGTGTCTGAGAGAG-3′ and 5′-CTGACTTCCTCAACCATTCCA-3′ for IPT3, 5′-AGTTACAGCGATGACCACCA-3′ and 5′-GGCAGAGATCTCCGGTAGG-3′ for IPT5, 5′-ACTCCTTTGTCTCAAAACGTGTC-3′ and 5′-TGAACACTTCTCTTACTTCTTCGAGT-3′ for IPT7, 5′-CATGTTCTAGGGGTCATTCCA-3′ and 5′-CTCCGATGGTCTCACCAGTT-3′ for CYP735A2, 5′-GAACTCGGAACCGAACTGG-3′ and 5′-TCAAACCCATTAAACCAATGC-3′ for LOG1, 5′-TGATGCTTTTATTGCCTTACCA-3′ and 5′-CCACCGGCTTGTCATGTAT-3′ for LOG3, 5′-GTTTGATGGGTTTGGTTTCG-3′ and 5′-CACCGGTCAACTCTCTAGGC-3′ for LOG4, 5′-GAGGGTTTGTGTGTTCTGTGGTAGC-3′ and 5′-GAGACCAAACCCATGAGACCAATG-3′ for LOG5, 5′-CTCCGATGGTCTCACCAGTT-3′ and 5′-CATGTTCTAGGGGTCATTCCA-3′ for LOG7, 5′-ATTGCACTCCCTGGAGGTTA-3′ and 5′-CCCATCAACATTCAATAGACCA-3′ for LOG8, 5′-CTCTCCACGTACTGGTTGTCGTTAC-3′ and 5′-CGGAAGAGATCTTCACATGTTTGTG-3′ for PIN1, 5′-AAGTCACGTACATGCATGTG-3′ and 5′-AGATGCCAACGATAATGAGTG-3′ for PIN2, 5′-CTTGCTGGATGAGCTACAGCTTTGG-3′ and 5′-CAAGTCAAACAGCCATGACGCCAAG-3′ for PIN3, 5′-CGAAAGAGTAATGCTAGAGGTGGTG-3′ and 5′-AATATCAGTCGTGTCATCACACTTG-3′ for PIN4, 5′-CGAAAGAGTAATGCTAGAGGTGGTG-3′ and 5′-AATATCAGTCGTGTCATCACACTTG-3′ for PIN7, 5′-GAAGAATCTACACCTCCTACCAAAA-3′ and 5′-CACGTAGCTCACACTGTTGTTG-3′ for IAA2, and 5′-CACGCTGCAAGAGAACAAGA-3′ and 5′-ACCACTTGAGTCCGCATACC-3′ for CCS52A1.

GUS staining

Seedlings were incubated in a GUS staining solution [100 mM sodium phosphate, 5-bromo-4-chloro-3-indolyl β-d-glucuronide (1 mg/ml), 0.5 mM ferricyanide, and 0.5 mM ferrocyanide (pH 7.4)] in the dark at 37°C. The samples were cleared with a transparent solution [chloral hydrate, glycerol, and water (8 g:1 ml:1 ml)] and observed under a light microscope (Olympus).

Microscopic observation and measurement of cell death area

Five-day-old roots were stained with 10 μM PI solution for 1 min at room temperature, and root tips were observed under a confocal laser scanning microscope (Olympus, FluoView FV1000). Cell death area was measured using Fiji image analysis software (http://fiji.sc) by defining the field in which PI infiltrates the cells.

Root growth analysis

Seedlings were grown vertically in square plates, and root tips were marked every 24 hours. Plates were photographed, and root growth was calculated by measuring the distance between successive marks along the root axis with ImageJ software (http://rsb.info.nih.gov/ij/).

Quantification of cytokinin levels in isolated cell populations

Protoplasts were isolated from root tips of 5-day-old pRCH1:GFP and pRCH2:CFP seedlings, which were treated with or without 8 μM zeocin for 24 hours. GFP- or CFP-positive protoplasts were collected through FACS, and their cytokinin concentration was determined using liquid chromatography–tandem mass spectrometry. Protoplast isolation, cell sorting, and cytokinin purification and quantification were performed as described previously (60). Analysis and sorting were performed using a BD FACSAria I and BD FACSDiva software. GFP fluorescence was excited using a 488-nm laser and collected using a fluorescein isothiocyanate filter set (emission filter, 530 ± 30 nm), while CFP fluorescence was excited using a 405-nm laser and collected using an Alexa Fluor 430 filter set (emission filter, 550 ± 30 nm). The collected fluorescent protoplasts were snap-frozen in liquid nitrogen after each sorting experiment. Before cytokinin purification and subsequent quantification, the samples were independently collected from four biological replicates per genotype and treatment (60,000 to 160,000 protoplasts per replicate).

EdU pulse labeling

Five-day-old seedlings were grown on MS medium supplemented with or without 8 μM zeocin, 5 nM IAA, or 10 nM cvxIAA for 12 hours and transferred to MS medium with or without zeocin, IAA, or cvxIAA and with 20 μM EdU for 15 min. After washing with MS medium, the seedlings were transferred back to MS medium supplemented with or without zeocin, IAA, or cvxIAA. EdU staining was performed with a Click-iT Plus EdU Alexa Fluor 488 imaging kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. Roots were double-stained with EdU and 4′,6-diamidino-2-phenylindole, and epidermis cells were observed with a confocal microscope (FV1000, Olympus).

Statistical analysis

Significant differences between different treatments and genotypes were analyzed by Student’s t test. Details of statistical analyses are provided in the figure legends.

SUPPLEMENTARY MATERIALS

Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/7/25/eabg0993/DC1

View/request a protocol for this paper from Bio-protocol.