Repressive C2H2 zinc finger ZAT proteins promote programmed cell death in the Arabidopsis columella root cap

Developmental programmed cell death (dPCD) controls a plethora of functions in plant growth and reproduction. In the root cap of Arabidopsis thaliana, dPCD functions to control organ size in balance with the continuous stem cell activity in the root meristem. Key regulators of root cap dPCD including SOMBRERO/ANAC033 (SMB) belong to the NAC family of transcription factors. Here we identify the C2H2 zinc finger protein ZAT14 as part of the gene regulatory network of root cap dPCD acting downstream of SMB. Similar to SMB, ZAT14 inducible misexpression leads to extensive ectopic cell death. Both thecanonical EARmotif and a conserved L-box motif of ZAT14 act as transcriptional repression motifs and are required to trigger cell death.While a single zat14 mutant does not show a cell death-related phenotype, a quintuple mutant knocking out five related ZAT paralogs shows a delayed onset of dPCD execution in the columella and the adjacent lateral root cap. While ZAT14 is co-expressed with established dPCD-associated genes, it does not activate their expression. Our results suggest that ZAT14 acts as a novel transcriptional repressor controlling a so far uncharacterized sub-section of the dPCD gene regulatory network active in specific root cap tissues.


Introduction
Developmental programmed cell death (dPCD) is a genetically controlled biological process essential for plant development (Van Durme and Nowack, 2016;Locato and De Gara, 2018).Different dPCD processes occur in a multitude of developmental contexts, including tracheary element differentiation, anther and pollen formation, seed development, and root cap turnover (Daneva et al., 2016;Locato and De Gara, 2018).The root cap is a specialized external organ at the root tip that protects the root apical meristem and acts as a sensory organ to optimize root growth and root system architecture (Kumpf and Nowack, 2015;Xuan et al., 2016;Di Mambro et al., 2018).The root cap of Arabidopsis thaliana (Arabidopsis) originates from two distinct stem cell populations: The columella initialsform a plate below the quiescent center whereas the epidermal/lateral root cap (EPI/LRC) initials are arranged in a ring around the columella initials.Iterative formative divisions of columella initials generate new layers of columella cells, while the EPI/LRC initials divide to form LRC and epidermis layers.While columella cells differentiate and expand directly, the LRC cells of the same root cap layer first undergo a number of cell divisions before differentiation to cover the root meristem until the start of the elongation zone (Kumpf and Nowack, 2015;Ganesh et al., 2022).At the start of the elongation zone, the most distal LRC cells of each cell file will undergo a precisely timed dPCD process to ensure that the extend of the LRC is restricted to the meristematic region, which is important for optimal root growth (Fendrych et al., 2014).In contrast, the columella cellsare shed in intervals as packages of living border-likecells (Vicre et al., 2005), and only complete dPCD during and after their release into the rhizosphere (Huysmans et al., 2018).
The root cap-specific TF SMB transcriptionally controls the preparation of LRC dPCD, the ultimate stage of their differentiation.Several transcriptionally regulated PCD-associated genes are positioned downstream of SMB since in the smb mutant or inducible SMB overexpression (OE) lines, their expression is decreased or increased, respectively (Fendrych et al., 2014;Huysmans et al., 2018).Accordingly, mutating of SMB leads to a delay or absence of root cap cell death (Fendrych et al., 2014), while its inducible systemic misexpression causes plant-wide ectopic cell death leading to growth arrest and eventually death of the plant (Bennett et al., 2010;Fendrych et al., 2014;Huysmans et al., 2018).

Europe PMC Funders Author Manuscripts Europe PMC Funders Author Manuscripts
We hypothesized that the PCD gene regulatory network contains additional transcriptional regulators, including transcriptional repressors, to ensure a tight and failsafe control of root cap maturation, dPCD preparation and execution, as well as post-mortem corpse clearance.Here we identify ZINC-FINGER OF ARABIDOPSIS THALIANA14 (ZAT14) (Ciftci-Yilmaz and Mittler, 2008) and related ZAT TFs as part of the SMB-controlled gene regulatory network of root cap cell death.The C2H2-type zinc finger protein family includes 176 members in Arabidopsis, being one of the largest families of putative transcriptional regulators in plants (Englbrecht et al., 2004;Ciftci-Yilmaz and Mittler, 2008;Xie et al., 2019).ZAT14 and several related ZAT proteins belong to the C2H2 C1-2i subclass containing 20 members, while other ZAT proteins fall into the related C1-3i subclass containing 8 members (Ciftci-Yilmaz and Mittler, 2008).Several ZATshave been shown to play important roles in plant stress responses by transcriptional regulation (Xie et al., 2019), but most of the ZAT proteins remain functionally uncharacterized.
Inducible ZAT14 misexpression mimics the ectopiccell death phenotype of SMB misexpressors, and SMB promotes ZAT14 expression.Single-cell RNA-sequencing (scRNA-seq) data suggest that ZAT14 is coexpressed with established dPCD-associated genes in the root cap and root xylem, which we confirmed by transcriptional and translational reporter lines.Protein domain analyses show that ZAT14 acts as a transcriptional repressor and that mutating repressive motifs interferes with ZAT14's potential to induce cell death.Transcriptome analyses show that SMB and ZAT14 misexpression cause downregulation of a common set of genes which have not been implicated in PCD so far.Though a zat14single mutant does not show any PCD-related phenotypes, higher-order mutants in related ZATgenes show a delay in columella cell death placing repressive ZAT TFs on a hitherto unrecognized branch of the root cap PCD gene regulatory network downstream of SMB.

TRANSPLANTA screening to identify additional regulators of dPCD
In order to find additional TFs of the dPCD gene regulatory network, we screened a subset of root-cap expressed TFs of the TRANSPLANTA (TPT) seed collection.The TPT collection consists of homozygous Arabidopsis lines each expressing one TF under the control of a systemic β-estradiol-inducible promoter (Coego et al., 2014).We screened for TFs which upon induction by estradiol treatment mimicked the SMB misexpression phenotype, which is characterized by ectopic cell death followed by growth arrest and plant death (Bennett et al., 2010;Fendrych et al., 2014;Huysmans et al., 2018).When proRPS5A:XVE>>SMB-GFP seeds are germinated on medium containing estradiol, their roots emerge but rapidly stop growing due to ectopic cell death and the cotyledons do not emerge from the seed coat (Figure 1A).
We did not screen the entire collection but filtered for root-cap expressed and PCDassociated TFs using a publicly available scRNA-seq dataset (Ferrari et al., 2022).Of the 650 TFs included in the collection, 320 TFs qualified as expressed in the lateral root cap (LRC) and columella clusters.When further filtering for TFs expressed in the dPCD population (clusters in whichPASPA3, BFN1 and RNS3 are highly expressed), we identified 57 TFs, expressed by a total of 150independent TPT lines.These lines were screened for a SMB-like phenotype as described above (Supplemental dataset 1).We used proRPS5A:XVE>>SMB-GFP seeds as a positive control and Col-0 seeds as a negative control.We only identified two TF genesshowing an SMB-like misexpression phenotype: one line misexpressing BEARSKIN2 (BRN2, AT4G10350) and three lines misexpressing ZINC-FINGER OF ARABIDOPSIS THALIANA 14 (ZAT14, AT5G03510) (Figure1A).As BRN2 has already been shown to induce ectopic cell death (Bennett et al., 2010), we focused on ZAT14 which has been implicated in phloem differentiation but not in dPCD (Roszak et al., 2021).

ZAT14 overexpression is sufficient to induce ectopic cell death
To independently test the effect of ZAT14 misexpression we first used transient transfection of Nicotiana benthamiana.As expected for a TF, a pro35S:ZAT14-GFP construct conveyed ZAT14-GFP localization in the nucleus 2 days after infiltration (DAI, Figure 1B).At 7 DAI ZAT14-GFPexpression induced a macroscopically visible tissue degeneration phenotype.
This phenotype was characterized by chlorosis and extensive leaf tissue death (Figure 1B), and was comparable to the phenotype generated by the expression of an established dPCD-promoting gene, pro35S:KIR1-GFP(Figure 1B) (Gao et al., 2018).Conversely, the expression of TARGET OF MONOPTEROS 5 (TMO5/BHLH32), a TF not associated with cell death, did not lead to chlorosis.In this experimental setup, TMO5 serves as a negative control (Figure 1B).
To confirm the TPT results in Arabidopsis, we generated vectors conveying estradiolinducible (Siligato et al., 2016) ZAT14 expression controlled by the ubiquitous promoter proHTR5 (Ingouff et al., 2017)(Figure 1C-E, Supplemental movie 1).In 35 independent linesexpressing ZAT14 either with or without C-terminal GFP tag for protein visualization, the overexpression of ZAT14 first triggered root growth arrest24 hours after transfer to estradiol-containing medium(Figure 1E, Supplemental figure 1B), and later led to the death of the entire seedling (Figure 1C).To visualize cell death in roots, we used the membrane-impermeable dye propidium iodide (PI).PI entry into the cell indicates plasma membrane permeabilization as a central cell death hallmark (Truernit and Haseloff, 2008;Fendrych et al., 2014).24 hours after induction (HAI), root cells in the transition and elongation zone are already intensely stained by PI indicating widespread cell death in this part of the root (Figure 1D), explaining the macroscopically visible root growth arrest.This ectopic cell death phenotypeis reminiscent of inducible misexpression of SMB, NAC046, ANAC087, and KIR1 (Gao et al., 2018;Huysmans et al., 2018).Expression analysis using RT-qPCR confirmed that ZAT14 was already strongly up-regulated at 8HAI (Figure 1F).Microscopic time-lapse imaging of proHTR5:XVE>>ZAT14-GFProots, and proHTR5:XVE>>NLS-GFPcontrol roots, revealed GFP expression 6HAI (Supplemental movies1 and 2), and first cases of ectopic cell death caused by ZAT14-GFP as early as 10 HAI (Supplemental movie 1).Our results showthat inducible misexpression of ZAT14 is sufficient to trigger ectopic cell death outside the root cap context.

L-box and EAR domain of ZAT14 are required to cause ectopic cell death
ZAT14 belongs to theC1-2i subclass of C2H2-type zinc finger proteins which contains a total of 20 members (Ciftci-Yilmaz and Mittler, 2008).There are several conserved regions present in the majority of C1-2i members: A short motif including a consensus sequence corresponding to a B-box (KXKRSKRXR) is located near the N-terminus and might act as a nuclear localization signal (Sakamoto et al., 2000).The second motif consists of acidic residues followed by hydrophobic leucine rich residues, with a consensus of "EXEXXAXCLXXL" (L-box), which is located between the B-box and the first zinc finger domain.Furthermore, there are two C2H2 zinc finger domains mediating interaction with DNA, and at the C-terminus, there is an ethylene-responsive element binding factor (ERF)-associated amphiphilic repression (EAR) motif(Supplemental figure 1A) (Sakamoto et al., 2000;Englbrecht et al., 2004;Ciftci-Yilmaz and Mittler, 2008;Xie et al., 2019).Thetranscriptional repression function of several C2H2 TFs has been attributed to regions containing an EAR motif (Ohta et al., 2001;Hiratsu et al., 2002).Together with the EAR-motif, the L-box may play roles in protein-protein interactions important for gene repression (Sakamoto et al., 2000), but experimental evidence to support this is lacking so far.
To study the importance of the differentconservedZAT14 domains contribute to the misexpression phenotype, we made different mutated versions of ZAT14.First, we deleted the 41 C-terminal amino acids containing the EAR-motif (ZAT14 ΔC ), or mutated the EARmotif modifying L 251 DLNL 255 to L 251 AAAL 255 (ZAT14 mEAR ) (Ciftci-Yilmaz et al., 2007).Next, we deleted the L-box (ZAT14 ΔL ) and ultimately, we deleted (ZAT14 ΔLΔC ) or mutated (ZAT14 mLmEAR ) both domains simultaneously.In the ZAT14 mLmEAR version, the L-box was mutated modifying C 76 LILLS 81 to C 76 AAAAS 81 and the EAR-motif was mutated as described above (Figure 2A).We introduced vectors for estradiol-inducible overexpression of all different ZAT14 versions C-terminally fused to either GFP or mTFP1 (Pasin et al., 2014)into Arabidopsis, establishing at least 20 independent lines for each construct for phenotypic analysis.In summary, we found that only joint mutation or deletion of both L-box and EAR-motif compromised the pro-PCD function of ZAT14 (Supplemental figure 1B).To maximize comparability between different lines we performed qPCR and picked two lines per construct with about 100-fold and 300-fold upregulation in comparison to the wild type (Figure 2C).We found the ZAT14 wild-type protein, ZAT14 mEAR -mTFP1, ZAT14 ΔC -mTFP1, and ZAT14 ΔL -GFP caused root growth arrest (Figure 2B, D) and rapid cell death (Figure 2E) upon induction.However, overexpression of ZAT14 ΔLΔC -GFP or ZAT14 mLmEAR -GFP did not result in root growth arrest (Figure 2B,D) or ectopic cell death (Figure 2E).To test whetherZAT14 acts via gene repression, we added the repressive SRDX motif (Hiratsu et al., 2003) to the C-terminus of ZAT14 mLmEAR .Two independent lines were selected based on the expression level via qPCR, and both showed a reconstitution of ZAT14's pro-PCD function to the wild-type level (Figure 2 B, D and E, Supplemental figure 1B).Taken together, these results suggest that ZAT14 promotes ectopic cell death by the mechanism of gene repression, and that both the EAR-motif and the L-box are important for this repressive function.

ZAT14 is highly upregulated prior to dPCD processes in the xylem and the root cap
Available scRNAseq dataset of Arabidopsis roottips (Ferrari et al., 2022) show that ZAT14 is not only upregulated prior to root cap dPCD, but also during protoxylem maturation and thus co-expressed with canonical dPCD associated genes (Olvera-Carrillo et al., 2015) (Supplemental figure 2A).
To visualize the gene expression pattern of ZAT14, as well as protein localization and dynamics, we generated transcriptional and translational reporter lines.In translationalreporterlines (proZAT14:ZAT14-GFP), 5-day-old seedlings showed strong nuclear-localized GFP signals in the LRC prior to PCD (Figure 3A, 3D, 3E), but alsoexpression in the root epidermis, endodermis and vasculature (Figure 3B, 3C).However, the expression of ZAT14 mRNA is highest in sc-RNAseq clusters of xylem and root cap tissues preparing for dPCD (Supplemental figure 2A).By crossing with the PCDreporter line proPASPA3:NLS-tdTomato (Xuan et al., 2016),proZAT14 was confirmed as co-expressingwith proPASPA3 in the LRC and the xylem cells prior to PCD (Figure 3F), consistent with the scRNA-seq data.During the dynamic turnover of the root cap, PCDprogresses towards and into the columella (Fendrych et al., 2014;Huysmans et al., 2018), and also here we observed ZAT14-GFP expression (Figure 3G-H).The transcriptional reporter lines (proZAT14:NLS-GFP-GUS) generallyshowed a similar as the translational reporter line (Supplemental figure 2B).However, transcriptional reporters did not show expression in columella cells, suggesting the ZAT14 fusion protein might either be stabilized in, or transported to, the columella.GUS histochemical staining showed GUS signal in the root and leaf vasculature, at the junction of the anther filament as well as in the floral organ abscission zone (Supplemental figure 2C-G).Interestingly, GUS signals were also detected in the phloem of the inflorescence stem (Supplemental figure 2H), suggesting ZAT14 might fulfil functions in this cell type as well.These data suggest that ZAT14 is expressed in tissues undergoing PCD, but not restricted to these tissues.

ZAT14 is downstream of the key root cap PCD regulatorSMB
Considering that the ZAT14 expression pattern is associated with cells undergoing dPCD in the root capand that ZAT14 OE phenocopied the overexpression of SMB, we investigated therelationship between SMB and ZAT14 in the dPCD regulatory network.
First, we generated an RNA-seq dataset of a dexamethasone-inducible SMB-GR line (Bennett et al., 2010)6 hours after dexamethasone treatment (Supplemental dataset 2).Our analyses revealed that ZAT14 and seven other members of the C1-2i subclass (ZAT5, ZAT15, ZAT13, AT4G16610, ZAT14L, AZF3, AZF1) showed at least 2-fold upregulation upon inducible SMB expression.Among these TFs, ZAT5 and ZAT14were highly expressed in the root cap and upregulated during dPCD.They both show an increased expression close to 16-fold change, upon SMB overexpression (Supplemental table 1).These results suggestthat the expression of ZAT14 and related members is positively regulated by SMB.
Second, when a SMB-TagBFP fusion protein was inducibly expressed under the control of the ubiquitousHTR5 promoter and transformed into the transcriptional reporter line of ZAT14, we observed strong ectopic activation of ZAT14expression in cells that misexpressed SMB (Figure 4A).
Finally, we crossed the smb-3 mutant (Willemsen et al., 2008)(hereafter called smb) with the transcriptional reporter line of ZAT14.We observed that reporterexpression in the root Europe PMC Funders Author Manuscripts Europe PMC Funders Author Manuscripts capwas weakened and occurred in less cells in smb mutant compared to wildtype with the reporter line (Figure 4B, 4C).This suggests that SMB contributes to ZAT14 expression in the root cap cells preparing for PCD, though ZAT14 expression does not seem to depend entirely on SMB activity.
Additionally, we investigated if a functional ZAT14-GFP fusion protein, expressed under the SMB promoter in the smb mutant, could complement the smbphenotype.The smbphenotype shows a longer root cap due to delayed dPCD, and an absence of corpse clearance after cell death (Fendrych et al., 2014) (Figure 4E, left panels).At 10 HAI in thesmb background proSMB:XVE>>ZAT14-GFPwas expressed throughout the root cap.At 24HAI, we observed increased cell death in the root cap, though this cell death was not restricted to the edge of the root cap as in the wild type (Figure 4E, left panels), but occurred in a more widespread fashion (Figure 4E, right panels).These results indicate that ZAT14-induced cell death can occur even in absence of SMB (Figure 4D, 4E).Altogether these results suggest that ZAT14 is downstream of SMB and does not require additional SMB-dependent genes to trigger cell death in the root cap.

ZAT14 shares partially common PCD pathway with SMB
Upon systemic inducible misexpression of SMB, ANAC087, and ANAC046, dPCDassociated genes are upregulated, indicative of an ectopically activated dPCD program (Fendrych et al., 2014;Huysmans et al., 2018).To investigate if ZAT14 indirectly promotes the expression of dPCD-associated genes as part of a core dPCD program, we tested whether the dPCD-associated genes BFN1, EXI1, PASPA3 and RNS3are upregulated upon inducible misexpression of ZAT14.The qPCR analysis showed that none of the interrogated genes were upregulated (Supplemental figure 3A), suggesting that ZAT14 does not act upstream of the known dPCD-associated genes.In line with this hypothesis, ZAT14-conferred cell death was not followed by post-mortem corpse clearance, while SMB-triggered cell death was (Figure 5A-B).This suggests that cell clearance enzymes including BFN1 are not activated downstream of ZAT14 (Supplemental figure 3A).
To investigate the function of ZAT14, we performed a transcriptome analysis of an inducible ZAT14 OE line from the TPT collection at 8h after estradiol treatment versus a mocktreated control (Supplemental dataset 3).795 genes were upregulated by ZAT14 OE, while 483 genes were downregulated compared to the mock-treated control (Figure 5C, Supplemental dataset 3).Confirming our qPCR results, PCD-associated genes including PASPA3,BFN1, RNS3, MC9 and DMP4were not upregulated in ZAT14 OE (Supplemental dataset 3).To investigate a potential gene regulatory overlap between SMB and ZAT14, we compared up-and downregulated genesupon overexpression of ZAT14 and SMB.
151 genes were commonly upregulated and 172 genes were commonly downregulated following ZAT14 and SMB OE.Both gene groups, but especially downregulated genes are much more abundant than expected by chance, as indicated by a representation factor of 3.4(Figure 5C).By contrast, not more genes than expected by chance were oppositely regulated by SMB OE and ZAT14 OE (Supplemental figure 3B,C).These results suggest that there is a common group of downregulated genes downstream of SMB and ZAT14.Commonly downregulated genes were enriched in GO terms including indole glucosinolate Europe PMC Funders Author Manuscripts Europe PMC Funders Author Manuscripts metabolic process, secondary metabolic process, cell wall modification and organization (Figure 5D).While no PCD-related GO-terms appeared enriched in commonly upregulated genes, several upregulated genes have been associated with cell death and senescence events, includingCYSTEINE ENDOPEPTIDASE 2 (CEP2, AT3G48340),TATD RELATED DNASE (TATD, AT3G03500) andRESPONSIVE TO DEHYDRATION 21B(RD21B, AT5G43060)(Supplemental dataset 4) (Shindo et al., 2012;Lampl et al., 2013;Howing et al., 2018;Buono et al., 2019).Our results indicate that ZAT14 might control a subset of the gene regulatory network downstream of SMB.As GO enrichment analysis did not identify putative mechanisms activated by ZAT14 misexpression, it remains a future challenge to understand ZAT14 function in the context of dPCD.

ZAT14 and its homologs control the onset of PCD in columella cells
To investigate the role of ZAT14 in root cap PCD, we identified a mutant carrying a T-DNA insertion in the single exon of ZAT14 (SALK_114288C; zat14-1) that was unable to generate a full-length transcript (Figure 6A).However, we did not find any obvious root cap PCD phenotype using a proPASPA3>>H2A-GFP dPCD reporter construct (Olvera-Carrillo et al., 2015)and fluorescein diacetate (FDA) combined with PI as a live-death stain (Huysmans et al., 2018) (Figure 6E, Supplemental figure 5C).To investigate whether genetic redundancy was responsible for the lack of zat14-1 mutant phenotype, we generated higherorder mutants by multiplex CRISPR-Cas9 (Figure 6B).First, we targeted C1-2i subfamily members that are most highly expressed in the root cap prior to root cap dPCD: ZAT14, AZF2, ZAT5, ZAT10 and ZAT12 (Supplemental table 2).In a proPASPA3>>H2A-GFP reporter background, we isolated a zatquintuple mutant, confirming homozygous frameshift mutations leading to early stop codons in each gene in the T3 generation (Figure 6B and Supplemental figure 4).For most genes the stop codon was located upstream of the zinc finger motifs responsible for DNA binding, suggesting the alleles generate null mutants.An exception was AZF2 in which the stop codon occurred only after the first zinc finger motif.However, a truncated AZF2 Δ165aa-173aa construct was unable to cause cell death after transient expression in N. benthamiana, while the full-length AZF2 produced cell death (Supplemental figure 5A-B), suggesting that the isolated azf2 allele is a null mutant.
In the LRC of 5-day-old quintuple mutant seedlings, we could not identify an aberrant proPASPA3>>H2A-GFP expression pattern, nor any obvious differences with the wild-type as visualized by PI staining (Supplemental figure 5C-D).However, when we cultivated seedlings for 14 days on vertical agar plates to allow the production of additional root cap layers (Feng et al., 2022), we observed a clear delay of dPCD in the columella root cap in the quintuple mutant compared to the wild-type and to the zat14-1 mutant.In the quintuple mutant living cells were still detected in the sixth layer, while in the wild type and the zat14-1 mutant these cells had already undergone PCD (Figure 6E, Supplemental figure 5E).The number of root cap layers and the columella shedding process were not different in the quintuple mutant (Figure 6F, G), suggesting that loss of ZAT proteins specifically compromised dPCD execution and not root cap development or maturation.We observed a partial reversion of this phenotype by expressing a proZAT14:ZAT14-tdTomato complementation construct (Figure 6E, G), demonstrating that ZAT14 plays a prominent role among redundantly acting ZAT proteins.

Europe PMC Funders Author Manuscripts
Europe PMC Funders Author Manuscripts However, even in the quintuple mutant, we did not observe a dPCD phenotype in the distal LRC.To remove further redundancy, we generated an octuple and an undecuple ZAT mutant by targeting ZAT8, ZAT6, ZAT14L, ZAT11, ZAT18 and AZF1 in the zat quintuple mutant background (Figure 6C-E, Supplemental figure 5C-E).However, no distal LRC dPCD phenotype was observed in these mutants (Supplemental figure 5C-D), and the delayed columella PCD process was not exacerbated (Figure 6E).Possibly, there is more extensive redundancy within the C1-2i subfamily, or even beyond among repressive C2H2-type zinc finger proteins.

Discussion
We have identified ZAT14 as a novel transcriptional regulator in the root cap dPCD gene regulatory network downstream of the established transcription factor SMB.In contrast to SMB and other NAC transcription factors involved in cell death control, ZAT14 contains a repressive EAR-motif (Englbrecht et al., 2004;Ciftci-Yilmaz and Mittler, 2008;Xie et al., 2019)that is conserved in the members of the C1-2i subfamily.We showed that deletion or mutation of the EAR-motif does not interfere with ZAT14's capacity to cause cell death upon inducible misexpression.This finding stands in contrast to results showing that removal of the C-terminal EAR-motifs were sufficient to disturb the functions of ZAT10 and ZAT11 upon transient misexpression (Ohta et al., 2001).However, other approaches showed that growth-inhibition caused by ZAT7 overexpression did not depend on the presence of the EAR-motif (Ciftci-Yilmaz et al., 2007).Likewise, the osmotic stress phenotype of zat10mutants could be complemented by a ZAT10 protein with a mutated EAR motif (Nguyen et al., 2016).
Our results showed thatsimultaneous mutations in both L-box and EAR-motif were necessary to impair ZAT14 function, suggesting that these domains have redundant or complementary functions for ZAT14's cell death regulatory capacity.The L-box in ZAT14 resembles the LXLXL motif in ZAT1 that functions as a transcription repression motif (Kagale and Rozwadowski, 2011;Song et al., 2020).The fact that we can reconstitute ZAT14 function by fusing a C-terminal SRDX domain to the ZAT14 mLmEAR version lacking both L-box and EAR motif suggests that ZAT14 indeed acts as a transcriptional repressor and that it needs both L-box and EAR-motif to exert this function.Together, these data suggest that repressive C1-2i ZAT proteins act as transcription repressors via the canonical EAR motif and other less well-defined repressive domains.
We identified ZAT14 as a dPCD-associatedgene that is co-expressed with established dPCDassociated genes in the root cap and protoxylem (Olvera-Carrillo et al., 2015).Like other dPCD-associated transcription factors, for instance,ANAC046 and ANAC087 (Huysmans et al., 2018), or VND6 and VND7 (Yamaguchi et al., 2010), inducible ZAT14 misexpression causes a rapid and widespread cell death, ultimately leading to the death of the entire plant.Interestingly, ZAT14-induced cell death occurs later in the root cap than in the root elongation zone.This might hint at the existence of root-cap expressed factors that specifically attenuate ZAT14 activity in this tissue.

Europe PMC Funders Author Manuscripts Europe PMC Funders Author Manuscripts
We demonstrate that ZAT14 is downstream of SMB, the key regulator of root cap maturation and the LRC dPCD (Fendrych et al. 2014).The three NAC TFs involved in the LRC dPCD, SMB, ANAC046 and ANAC087 partially share PCD-associated genes as their targets (Huysmans et al., 2018).But these genes are not the targets of ZAT14 as shown by our RNA-seq analysis.This is not surprising given the fact that the established NAC TFs are considered to act as transcriptional activators, while ZAT14 is a transcriptional repressor.As a repressor, ZAT14 might suppress genes that inhibit the execution of the dPCD or post-mortem corpse clearance.Alternatively, ZAT14 function might suppress other pathways that attenuate cellular viability, and therefore indirectly contribute to dPCD promotion in the the columella root cap and the xylem.Further investigation of the downstream targets of ZAT14 are necessary to shed light on the involvement of ZAT14 in PCD promotion.
However, despite ZAT14 expression in both distal and proximal root cap dPCD zones we only found a loss of function phenotype of dPCD delay in the columella and adjacent proximal LRC cells in higher-order mutants.Possibly, there are additional ZAT-related transcription factorswithin or without the C1-2i subclass that act redundantly to ZAT14.Recently, ZAT1, ZAT4, and ZAT9 have been implicated in the maturation of root cap cells as positive regulators of the dPCD-associated gene PASPA3 (Song et al., 2020).These genes belong to the C1-3i subclass of C2H2 zinc finger proteins.Even though according to a scRNA-seq dataset (Ferrari et al., 2022)ZAT1, ZAT4 and ZAT9 do not appear to be expressed in LRC cells prior to dPCD (Supplemental dataset 1), it is still possible that either these or other related genes are upregulated in the absence of highly expressed ZAT proteins in a compensatory fashion.This effect has been described for other gene families, though the underlying mechanism remains to be elucidated(El-Brolosy and Stainier, 2017).To further analyze the function of ZAT proteins in dPCD it might be necessary to generate higher-order mutants that knock out additional or alternative ZAT genes.However, such a task would be a risky endeavour given the number of potential additional genes to mutate and putative compensatory upregulation of lowly expressed paralogs.
An alternative possibility is that ZAT protein activity is dispensable for dPCD in the distal LRC, despite ZAT14 upregulation in this tissue.This situation is analogous to the fact that autophagy is activated in both the dying columella and distal LRC cells, but only required for timely PCD execution in the columella (Feng et al., 2022).Possibly, there are additional cell death mechanisms that operate in the distal LRC, but not in the columella, that can compensate for loss of ZAT14 function.
In conclusion, we have identified ZAT14 as a repressive transcription factor that operates in Arabidopsis root cap cell death downstream of the key regulator SMB.While ZAT14 misexpression is sufficient to cause ectopic cell death, knock out of ZAT14 and up to 10 related root cap expressed paralogs show a delay of cell death execution in the columella cell type of the root cap.Future studies will have to reveal how the transcriptional repression of ZAT14 target genes contributes to preparing for and executing root cap PCD.

Accession numbers
Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the following accession numbers: AZF1 (AT5G67450),

Pharmacological treatment
The stock solution of propidium iodide (Sigma-Aldrich, P4864-10 mL, 1 mg/mL in water) was diluted and added to 1/2 MS medium at the final concentration of 10 µg/mL.
The stock solution of Fluorescein diacetate (Sigma-Aldrich, F7378-5 g) was prepared using acetone as the solvent at the concentration of 2 g/mL.The stock solution was diluted and added to 1/2 MS medium at the final concentration of 2 µg/mL.
The stock solution of β-Estradiol (Sigma-Aldrich, E8875-1 g) anddexamethasone (Sigma-Aldrich, D4902-1 g) were prepared using DMSO as the solvent at the concentration of 50 mM/L.The β-Estradiol stock solution was diluted and added to 1/2 MS medium at the final concentration of 50 µM/L for spraying and 10 µM/L for plates.The dexamethasone stock solution was diluted and added to 1/2 MS medium at the final concentration of 10 µM/L.DMSO was equally diluted as the controls.

TRANSPLANTA screening
To perform the screening of the TRANSPLANTA collection (Coego et al., 2014), seeds were germinated directly on 1/2 MS medium complemented with 10 µM β-Estradiol in 8.5cm diameter round plates.After sowing around 200 seeds and two days of vernalization at 4°C, the round plates were placed horizontally in a continuous lightroom (intensity 120 µmol m 2 s −1 ) at 21°C.Pictures were taken 10 days after germination.

Europe PMC Funders Author Manuscripts
Europe PMC Funders Author Manuscripts Cloning All fragments were cloned into Gateway-compatible entry clones (Invitrogen).To create the entry clones, DNA fragments were amplified using specific primers and high-fidelityDNA polymerase (BioLabs) in a standard PCR.All cloning primers are listed in Supplemental table 3.

CRISPR vectors and mutants
For the quintuple mutant, four guide RNAs (gRNAs) were designed for AZF2, ZAT5, ZAT10, ZAT12, and ZAT14, respectively.Cloning of gRNA vectors was performed as previously described (Decaestecker et al., 2019).Briefly, these gRNAs were annealed and ligated into their respective GG entry plasmid (pGG-A-AtU6-ccdB-B, pGG-B-AtU6-ccdB-C, pGG-C-AtU6-ccdB-D, pGG-D-AtU6-ccdB-E, pGG-E-AtU6-ccdB-F).Each entry clone targets one specific ZAT TF by one of the four designed gRNAs, but each position is the pool of four gRNAs of each gene.The gRNA entry plasmids were cloned into the pFASTGK_AtCas9_AG destination vector (VIB-UGent plasmid repository (https:// gatewayvectors.vib.be)) using Golden Gate assembly.The obtained expression clones Europe PMC Funders Author Manuscripts Europe PMC Funders Author Manuscripts were transformed into proPASPA3>>H2A-GFP reporter line using the Agrobacterium tumefaciens floral dip method (Clough and Bent, 1998).24 T1 plants were genotyped using Singer sequencing and one of these 24 T1 plants showed mutation for all 5 targeted genes, AZF2, ZAT5, ZAT10, ZAT12, and ZAT14.FAST negative T2 seeds (where the Cas9 gene is absent) derived from this line were selected for genotyping to obtain homozygous lines for each gene.The quintuple mutant was obtained and confirmed in T3.Two gRNAs were designed for ZAT6, AZF1, ZAT8, ZAT11, ZAT18 and ZAT14L, respectively.And these two gRNAs per each gene were ligated into their respective GG entry plasmid (pGG-A-AtU6-ccdB-B, pGG-B-AtU6-ccdB-C, pGG-C-AtU6-ccdB-D, pGG-D-AtU6-ccdB-E, pGG-E-AtU6-ccdB-F, pGG-F-AtU6-ccdB-G).Each entry clone targets one specific ZAT TF by two designed gRNAs.The gRNA entry plasmids were cloned into the pFASTGK_AtCas9_AG destination vector using Golden Gate assembly.The obtained expression clone was transformed into quintuple mutant to obtain octuple and undecuple mutants.All primers for cloning and genotyping were listed in Supplemental table 3.

RT-PCR, RT-qPCR
Total RNA was isolated from 5 days after gemination seedlings of zat14-1 mutant and wild type.Primers used in RT-PCR were P1/P2 for ZAT14, and P4/P5 for ACTIN2.For RT-qPCR of inducible overexpression lines, about 15 FAST positive seeds were selected from T2 seeds and sowed on 1/2 MS plates.5 days old seedlings were treated with estradiol or mock for 8 h and then they were harvested.For each line, three biological replicates (independently repeated treatments) were tested.After RNA extraction using the Spectrum Plant Total RNA Kit (Sigma-Aldrich), 1 µg of RNA was used for DNA synthesis with the qScript cDNA SuperMix (Quantabio).The RT-qPCR was performed with the LightCycler 480 (Roche) using SYBR green for detection of double-stranded DNA.Analysis of the RT-qPCR data was done using https://intra.psb.ugent.be/qPCR/index.plwith PEX4 (AT5G25760) and UBL5 (AT5G42300) as housekeeping genes, and later the data were normalized against wild-type or mock samples.All primers are listed in Supplemental table 3.

ZAT14 OE and SMB-GR RNA sequencing
TRANSPLANTA ZAT14 OE line and Col-0 whole seedlings, were treated either with DMSOor estradiol for 8 h.Then, they were harvestedas three biological replicates for each condition.The whole seedlings of SMB-GR line were treated with DMSO or DEX for 6 h before they were collected as three biological replicates.Each replicate contained 25 seedlings.The RNA was extracted as described above.The quantity and quality of the RNA was checked using a NanoDrop 2000 (Nanodrop Technologies Wilmington), and RNA integrity was confirmed on an Agilent 2100 Bioanalyzer (Agilent Technologies).The RNA-sequencing was performed by VIB Nucleomics core as Illumina NEXT-seq with 75 bp paired-end sequencing.Sequencing quality as well as read mapping and summarization were performed with a software pipeline on an in-house Galaxy server.Briefly, quality of raw data was verified with FastQC (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/).Next, quality filtering was performed using Trimmomatic as described (Bolger et al., 2014).Reads were subsequently mapped to Arabidopsis genome (Araport11) (Cheng et al., 2017).

Europe PMC Funders Author Manuscripts
Europe PMC Funders Author Manuscripts DEGs were identified withthe Edge-R software package in R (Robinson et al., 2010).Genes were considered as differentially expressed if their expression levels had an absolute Log 2 (FC) > 1, p < 0.05 and FDR < 0.05.For upregulated genes upon overexpression of ZAT14 (Supplemental dataset 3), firstly, we compared ZAT14 overexpressor line treated with estradiol (ZAT14_OE_Estr) with the same line after mock treatment (ZAT14_OE_mock), 1126 upregulated genes were obtained, while by comparing ZAT14_OE_Estr with estradiol-treated wild type (WT_Estr)there are 1055 upregulated genes.To remove the genes regulated by estradiol treatment or transgenic insertion, these two upregulated gene lists were combined and 795 genes were obtained and regarded as upregulated upon the overexpression of ZAT14.The same method was performed to get the downregulated genes, by comparing ZAT14_OE_Estr with ZAT14_OE_mock or ZAT14_OE_Estr with WT_Estr, 663 genes and 815 genes were obtained respectively.483 downregulated genes were obtained after combing these two downregulated gene lists.
GO enrichment analyses were carried out with the PLAZA 5.0 website (https:// bioinformatics.psb.ugent.be/plaza/versions/plaza_v5_dicots).GO pathway enrichment bubble plot was plotted by https://www.bioinformatics.com.cn/en a free online platform for data analysis and visualization.
For representation factor calculation, the online software: http://nemates.org/MA/progs/overlap_stats.html was used."The number of genes expressed in the RNAseq" was considered as "Number of genes in the genome".

Confocal imaging
Confocal images were acquired using LSM710 (Zeiss) and SP8X (Leica) microscopes.GFP and FDAwere excited by the 488nm line of the argon laser andwere detected between 500 and 550 nm, whereas PI was excited by the 561nm line of the laser and was detected between 580 and 740 nm.TagBFP was excited by the 405nm UV diode laser and detected between 425nm and 475nm.mTFP1 was excited by the 458nm line of the argon laser and detected between 495 and 550 nm.For time course, the seedlings of the proHTR5:XVE>>ZAT14-GFP and proHTR5:XVE>>NLS-GFP were grown for 5 d on 1/2 MS medium and then transferred to a Nunc Lab-Tek Chamber (Thermo Fisher) covered by an agar slab (1/2 MS containing 10 µM/L estradiol and 10 µg/mL PI).After transfer, the seedlings were returned to the growth room for 6 h and then imaged every 30 min for 20 h.Images were processed and analyzed using Fiji (https://fiji.sc/).

GUS histochemistry
5-day-old seedlings expressing proZAT14:NLS-GFP-GUSwereincubated in X-Gluc staining buffer at 37°C in the dark overnight.Chlorophyll was removed by two sequential incubations in 70% ethanol for several hours at each step.After rehydration, samples were transferred to 50% glycerol for mounting on glass slides.Samples were photographed using an Olympus BX51 (Olympus)microscope with a 10× objective (numerical aperture 0.25) and LEICA IC90 E (Leica) stereo microscope.
(F) The quantification of the root cap layer where shedding starts.Results are means ± SD (more than 9 roots were quantified for each genotype).
(G) The quantification of the living root cap layers for each genotype.Results are means ± SD (more than 9 roots were quantified for each genotype).Statistical analysis was performed by t test; ** p < 0.01, ****P < 0.0001, ns: not significant.
See also Supplemental figures 4 and 5.

Feng
Physiol.Author manuscript; available in PMC 2023 November 30.

Figure 1 .
Figure 1.Overexpression of ZAT14 in N. benthamiana leaves and in Arabidopsis seedlings.
GFP was used as a negative control.GFP signal was imaged in infiltrated N. benthamiana leaves at 2 DAI.The representative pictures of at least three biological replicates are shown here.Bars = 50 µm.(C) Macroscopic images of WT seedlings and representative estradiol-inducible proHTR5:XVE>>ZAT14 and proHTR5:XVE>>ZAT14-GFP lines at 0 and 7 days after transfer to estradiol-containing medium.Black arrows point at root tips.Bars = 1 cm.(D) Confocal laser scanning micrograph (CLSM) of roots inducibly overexpressing ZAT14 and ZAT14-GFP 24 hours after estradiol induction were stained with the propidium iodide (PI).PI is shown in magenta and GFP is in green.Each panel consists of two stitched images.White dotted lines mark the root profile.Bars = 50 µm.(E) Quantification of the Arabidopsis root growth of two independent lines of ZAT14 and ZAT14-GFP overexpressing lines.At least 10 roots of each line were quantified.Results shown are means ± SD. (F) As indicated by RT-qPCR analysis, 8 hours after the induction by estradiol (estr) leads to strong expression of ZAT14 in two independent lines of XVE-ZAT14 and XVE-ZAT14-GFP compared to the mock.Results shown are means ± SD (three independent treatments and three technical repeats each).Statistical analysis was performed by t test; * p < 0.05, ** p < 0.01, *** p < 0.001.See also Supplemental movies 1 and 2.

Figure 2 .
Figure 2. L-box and EAR domain of ZAT14 are required to cause ectopic cell death.(A)Graphical representation of ZAT14 modified versions.Full length ZAT14 has a B-box, an L-box, two zinc finger domains, and an EAR-motif.ZAT14 ΔC has 41 deleted amino acids in the C-terminal in which the EAR-motif domain is included.ZAT14 mEAR has a mutated EAR-motif.In ZAT14 ΔL , the L-box (69aa-80aa) has been deleted.In ZAT14 ΔLΔC , the L-box domain as well as the C-terminal were deleted.In the ZAT14 mLmEAR , the L-box domain and the EAR-motif domain were mutated.In the ZAT14 mLmEAR SRDX, the SRDX motif was added at the C terminal of ZAT14 mLmEAR .

( B )
Macroscopic appearance of WT seedlings and representative estradiol-inducible ZAT14 modified versions at 0 and 6 days after transfer to estradiol-containing medium.Overexpression of the ZAT14 wild type, ZAT14 ΔC , ZAT14 mEAR , ZAT14 ΔL and ZAT14 mLmEAR SRDX resulted in root growth arrest.Similar to the wild type seedlings, the overexpression of ZAT14 ΔLΔC and ZAT14 mLmEAR did not result in root growth arrest after 6 days of estradiol induction.All these constructs were driven by HTR5 promoter.Black arrows point at root tips.Bars = 1 cm.(C) As indicated by RT-qPCR analysis, 8 hours induction by estradiol caused strong expression of ZAT14 in two independent lines of each construct compared to wild type.Results shown are means ± SD (three independent treatments and three technical repeats each).Statistical analysis was performed by t test; ** p < 0.01, *** p < 0.001, **** p < 0.0001.(D) Quantification of the Arabidopsis root growth of two independent lines of each modified ZAT14 overexpressing lines.At least 10 roots of each line were quantified.Results shown are means ± SD. (E) Overexpression of the ZAT14 wild type, ZAT14 ΔC , ZAT14 mEAR , ZAT14 ΔL and ZAT14 mLmEAR SRDX resulted in ectopic cell death in the root of Arabidopsis, stained with PI (in magenta), after 24h and 48h of estradiol treatment.The overexpression of ZAT14 ΔLΔC or ZAT14 mLmEAR did not result in aberrant cell death.mTFP1 and GFP fused to ZAT14, in green.The fluorescent signal is shown as a Z-projection, whereas the brightfield channel is shown as a single stack.Each panel consists of two stitched images.Bars = 50 μm.

Figure 3 .
Figure 3. Expression patterns of ZAT14.(A-E) 5 days old root expressing proZAT14:ZAT14-GFP, stained with PI (in magenta).ZAT14-GFP was expressed in LRC prior to PCD and also showed signal in distal root (A).Cross-section (B) and longitudinal section (C) of distal root.Cross-section (D) and longitudinal section (E) of proximal root.Endodermis cells are marked by white dotted lines in(B)  and (D).(A) consists of three stitched images.White dotted lines in (A) point at the imaging area for (B) and (C).(F) 5 days old root expressing proZAT14:ZAT14-GFP and proPASPA3:NLS-tdTomato.Asterisks indicate xylem cells expressing ZAT14 and PASPA3.Distal root is shown as a single middle stack, while the proximal root is shown as a Z-projection.(G-H) 14 days old root expressing proZAT14:ZAT14-GFP, stained with PI.Projection section (G) and longitudinal section (H).The white arrowhead indicates distal and proximal LRC cells, and white arrows indicate columella cells.Bars = 50 µm for A, F,G, 20 µm for B to E, and H. See also Supplemental figure 2.

Figure 4 .
Figure 4. ZAT14 is downstream of SMB TF. (A) proHTR5:XVE>>SMB-tagBFP (in blue) was transformed in the reporter line proZAT14:NLS-GFP-GUS (in green) and stained with PI (in magenta), which have been treated with estradiol or DMSO (as mock) for 16 h or 24 h.Each panel consists of two stitched images.White asterisks indicate LRC PCD.White arrowheads indicate ectopic cell death.White dotted lines indicate size of LRC expressing proZAT14:NLS-GFP-GUS. Bars = 50 µm.

Figure 5 .
Figure 5. ZAT14 shares partially common PCD pathway with SMB.(A) Kymograph showed the corpse clearance in the epidermis in the inducibly overexpressing ZAT14-GFP (in green; upper panel) or SMB-TagBFP (in green; lower panel), stained with PI (in magenta).Cells were imaged every 5 min.Bars = 20 µm.(B) Quantitative analysis of corpse clearance.Percentage of cells showing cell corpses persisted without any signs of rapid autolysis is indicated in grey columns, whereas white columns indicate the percentage of cells showing corpse clearance within 2 hours.In total, 28 cells were analyzed for SMB-TagBFP and 18 cells for ZAT14-GFP.(C)Venn diagram of the upregulated genes and downregulated genes upon overexpression of ZAT14 and SMB.Differentially expressed genes with at least a 2-fold increased expression in the 6 hours dexamethasone-induced pro35S:SMB-GR compared to the mock (log 2 FC > 1, P < 0.05 and FDR < 0.05) -in total 1984 genes-were overlapped with differentially expressed genes that had at least a 2-fold increased expression in the 8

Figure 6 .
Figure 6.ZAT14 and its homologs control the onset of PCD in columella cells.(A) Schematic illustration of T-DNA insertion within the genomic region of ZAT14 and transcript analysis of ZAT14 in zat14-1 mutant.Arrowhead points at the T-DNA insertion site.Arrows indicate the binding sites for RT-PCR primers.ACTIN2 (ACT2) was used as internal control for the RT-PCR.(B-D) Schematic illustration of genomic regions and CRISPR mutation sites of 11 C1-2i encoding genes, AZF2, ZAT5, ZAT10, ZAT12, ZAT14, ZAT6, AZF1, ZAT8, ZAT18, ZAT11 and ZAT14L.

Feng
Physiol.Author manuscript; available in PMC 2023 November 30.