Pleiotropic regulatory locus 1 maintains actin cytoskeleton integrity and cellular homeostasis to enable Arabidopsis root growth

Summary

Cell functions are based on the integrity of actin filaments. The actin cytoskeleton is typically the target but also the source of signals. Arabidopsis PRL1 (Pleiotropic Regulatory Locus 1), regulates multiple cellular processes and physiological responses. However, the precise mechanisms underlying PRL1`s multiple functions are unclear. Here, we show that PRL1 maintains actin integrity and concomitant cellular homeostasis. The cortical actin cytoskeleton was de-polymerized in the prl1 mutant, causing the developmental root defect. Actin depolymerization, rather than reactive oxygen species (ROS) imbalance, constituted the fundamental cause of retarded root growth in prl1. ANAC085 upregulation by, and cooperation with, actin depolymerization triggered stele cell death in prl1 roots. Differential gene expression and alternative splicing defects resulting from actin depolymerization occurred independently in prl1. Our work establishes the cause-effect relationships between actin depolymerization and downstream stress-related signals, revealing a novel function of PRL1 and enhancing the understanding of PRL`s functional mechanisms.

Graphical abstract

Highlights

• •PRL1 maintains cortical actin integrity for root development
• •Actin depolymerization initiates hierarchical stress-related signals
• •Stress-related hierarchical signals make succedent and local stele cell death of prl1
• •Actin depolymerization is the end but also the source of signals

Biological sciences; Plant biology; Plant development

Introduction

Pleiotropic regulatory locus 1 (PRL1), which is a typical nuclear-localized WD40 domain-containing protein,^1,2 has been extensively studied for its multiple biological and cellular functions through analyzing various defects in its null mutants.^3,4,5,6 In Arabidopsis, an allelic prl1 mutant was first isolated for identifying a sugar signaling component and found to exhibit multiple defects in root development, hypocotyl elongation, and enhanced hypersensitivity to various phytohormones, as indicated by their developmental phenotypes.⁶ Subsequently, PRL1 was identified as a core component of the MAC/NTC complex responsible for plant innate immunity.^5,7 Further experiments revealed that PRL1 is involved in the activation of the spliceosome complex, microRNA processing, genome maintenance,^{3,8,9,10,11,12,13} and protein turnover.^1,14,15 However, the links between its biological functions and the molecular and biochemical functions of PRL1 remain unclear, and the mechanisms underlying its multiple biological functions remain undetermined.

The prl1 mutant is characterized by a short primary root, yet the underlying mechanism of this developmental defect remains unclear.^6,8,16 At the cellular level, root length is determined by the combination of controlled cell division in the meristem (MZ) zone and regulated anisotropic cell expansion in the elongation zone (EZ).¹⁷ Reactive oxygen species (ROS) have dual roles in regulating cellular actions. Under favorable developmental and environmental conditions, ROS maintains balance and facilitates root development by controlling the spatial transition from cell proliferation to cell differentiation, which is mediated by UPBEAT1-controlled ROS distribution.¹⁸ However, in response to encountered exogenous or endogenous stresses, ROS imbalance is generated, leading to detrimental effects such as reducing MZ size, preventing cell expansion, damaging DNA, and initiating the resultant DNA Damage Response (DDR), including cell-cycle arrest, cell death, and growth inhibition.^19,20,21 Previous studies have shown that stress-responsive genes are de-repressed in prl1 mutant, indicating the production of intracellular stress signals.^10,22 However, the cause of endogenous stress and its derivation in the prl1 mutant remain unclear.

In both plants and animals, the cytoskeleton is a fundamental component that provides structure and supports cellular behaviors such as cell division and expansion, and it is often considered a target for signals.²³ Dynamic homeostasis of actin is essential for normal cytoskeleton function, and this is regulated by two types of actin-binding proteins (ABPs): those that promote actin filament polymerization and those that facilitate actin filament depolymerization. These ABPs are regulated by upstream signals to maintain actin homeostasis.^24,25 Cells need to sense actin configuration and give feedback to its reorganization. However, when the dynamic actin cytoskeleton is disrupted, what signals do cells recognize and transmit? Recent research has shown that abnormal cortical actin structures can be recognized as damage-associated molecular patterns (DAMPs), which lead to a series of downstream events such as ROS accumulation, DNA fragmentation, and widespread changes in the transcription of defense/stress-responsive genes.^{26,27,28,29,30,31,32} Therefore, the actin cytoskeleton serves both as a cellular scaffold and as a signal.

In this study, our focus is on exploring the mechanisms underlying the developmental defect in the primary root of the prl1 mutant and investigating the general biological functions of PRL1. We conducted a series of investigations, including cellular, pharmacological, genetic, and molecular approaches, to identify the cause of the short root phenotype in prl1. Our findings revealed that the depolymerization of cortical actin was the principle and immediate cause of the short root phenotype in prl1. Moreover, we found that the imbalance of ROS was produced in response to actin de-polymerization, while ANAC085 was upregulated by and specifically cooperated with actin cytoskeleton depolymerization to implement root stele cell death. We also got the de-repression of stress/defense-related genes as a secondary signal in response to actin depolymerization and independent of alternative splicing defects in prl1. These results indicate that the fundamental function of PRL1 is to maintain actin integrity and concomitant cellular homeostasis. Our findings represent a significant contribution to the current understanding of the functions of PRL1 and may inspire further investigations on the molecular mechanisms underlying its functions.

Results

Mutation in pleiotropic regulatory locus 1 retards primary root development

To reveal general mechanisms for PRL1 functions, we specially investigated the cause of the root growth retardation resulting from mutations in PRL1. The 6-day primary root of prl1 was observed and quantified as dramatically shorter than the roots of wild type (WT) plants (Figure 1A). Time-lapse analysis confirmed the low rate of root growth (Figure 1B). Because the root meristem is the cellular basis for continuous primary root growth,³³ we compared the meristem zone size and cell status in the root tip between prl1 and WT. The root meristem zone size of the prl1 plant was obviously smaller than that of wild type and the cell numbers of the meristem zone were less than those of WT (Figures 1C and 1D). Cell death (Figure 1E; Figure S1) indicated by propidium iodide (PI) staining³⁴ and the quiescent center (QC) expansion marked by pWOX5-promoted green fluorescent protein (GFP)-encoding gene expression and GUS staining (QC25: GUS) (Figure 1E; Figure S2) were found in the root tip of prl1 but not in WT, both indicating root cells of prl1 were encountered endogenous stress.³⁵ Cell division was reduced in the root meristem zone of prl1 compared with WT, which was indicated by a decrease of the GUS intensity driven by the CycB1; 1 promoter (Figure 1F). Because root length is the combined output of cell division and cell elongation, we found that the lengths of cells in different zones of prl1 were shorter than the lengths of WT cells (Figure S3). These results showed that mutation in PRL1 retarded primary root growth by generally disturbing root cell behaviors and indicated the root cells of prl1 encountered endogenous stress.

Figure 1.

Mutation in PRL1 retards primary root development

(A) Representative root images of WT and prl1 in 6-day seedlings vertically cultured in ½ MS. Bar = 1cm.

(B) Quantification of primary root length of WT and prl1 from 1 to 15 days after germination. Value is mean ± SD (unpaired t-test, two tailed p < 0.0001, n > 50).

(C) Representative root meristem region of WT and prl1 in 6-day seedlings. The size of the root meristem region is typically defined from QC to a distinctive cortical cell with double size to its neighboring QC-closer cell. Scale bar = 100μm.

(D) Quantification of cell numbers in meristem zones of WT and prl1. Values are mean ± SD of more than 30 roots at 6 days after germination. Asterisk indicates a significant difference (unpaired t-test, two-tailed p < 0.0001, n > 50).

(E) pWOX5:: GFP expression and PI staining displayed stele cell status 6 days after germination WT and prl1. Scale bar = 20 μm.

(F) CyclinB1;1: GUS expression levels in 6-day roots of WT and prl1. Scale Bar = 100 μm.

Reactive oxygen species imbalance is sufficient to cause root growth defect but undetermined the necessary cause for pleiotropic regulatory locus 1 root retardation

Stress negatively affects cell behavior. Reactive oxygen species (ROS) imbalance is always generated in response to exogenous and endogenous stresses, which, in turn, harms organelles and molecules and arrests cell behaviors such as expansion and division.²⁰ To determine whether ROS were imbalanced in the prl1 mutant, we used diaminobenzidine (DAB) and nitroblue tetrazolium (NBT) staining to measure the accumulation of hydrogen peroxide (H₂O₂) and superoxide anion (O2^·-) respectively, we found significant H₂O₂ accumulation that had invaded the location where root meristem should be. There was also reduced O₂^·- content and loss of the O2^·- gradient in the meristem zone (Figures 2A and 2B), which indicated a ROS imbalance that would lead to premature cell differentiation in the root.¹⁸ To determine whether ROS imbalance hindered root growth, we treated seedlings with ո-propyl gallate (PG) and Diphenyleneiodonium (DPI) for 5 days to mimic ROS imbalance and found that the PG- and DPI-treated WT roots were seriously shortened similar to the roots of prl1, suggesting that ROS imbalance was sufficient to produce short roots and might be the cause for the retardation of primary root growth of prl1 (Figures 2A and 2B). To confirm that ROS imbalance is the cause of short roots in prl1, we treated plants with potassium iodide (KI) to scavenge H₂O₂ and restore ROS balance. KI treatment of WT eliminated hydrogen peroxide accumulation and the roots were longer compared with untreated WT; however, the KI treatment of prl1 did not alter the ROS accumulation and distribution pattern and did not restore root length (Figures 2A and 2B).

Figure 2.

ROS imbalance is enough to causing but undetermined the necessary cause for the root retardation of prl1

(A) Hydrogen peroxide concentration in roots of different genetic backgrounds and chemical treatments. Scale bar = 200 μm.

(B) Superoxide concentration in roots of different genetic backgrounds and chemical treatments. Scale bar = 200 μm.

(C) Seedlings with different genetic backgrounds and chemical treatments. Scale bar = 1 cm.

(D) Statistical quantification of root length in seedlings with different genetic backgrounds and chemical treatments. Values are mean ± SD (unpaired t-test, two-tailed p < 0.05, n > 50).

(E) Living status of root stele cells in 6-day seedlings of WT, atm, atr, sog1, prl1, atm prl1, atr prl1, sog1 prl1 indicated by PI staining. Scale bar = 50 μm.

We then turned to a genetic method to restore ROS balance. By analyzing RNA-seq data, we uncovered many peroxidase-encoding genes (PERs) that were downregulated in whole 6-day seedlings; reduced expression was confirmed by real-time quantitative polymerase chain reaction (RT-qPCR) and reverse-transcription polymerase chain reaction (RT-PCR) (Figures S4–S6). To avoid disturbance of native promoter activity on the expression of PER23, we chose the constitutive 35S promoter to drive the expression of PER23 and measured its expression levels in different overexpression (OE) lines (Figure S7). By crossing, we constructed a homozygous PER23OE prl1 double mutant. In comparison with WT, the PER23OE line displayed longer roots and had diminished the accumulation of H₂O₂ and upward distribution of O₂^·- along the longitudinal root axis (Figures 2A–2D). However, overexpression of PER23 did not restore ROS balance or root length in prl1 plants (Figures 2A–2D) indicating that other unidentified factors may have prevented the elimination of ROS accumulation and rescue of root growth. Hence, a ROS imbalance was not the necessary cause to root retardation in prl1. As the persistence of ROS imbalance, and considering that the precursor RNA processing 19/Nineteen complex/MOS4-associated complex (Prp19/NTC/MAC) complex is involved in genome maintenance,^{19,20,21,36,37} we surmised that ROS imbalance would damage DNA, which led to canonical ATAXIA TELANGIECTASIA MUTATED kinase (ATM)/ATM/RAD30-RELATED (ATR)-SUPPRESSOR OF GAMMA RESPONSE 1 (ATM/ATR-SOG1) signaling and stele cell death.^34,38 However, contrary to expectation, stele cell death still occurred in double homozygous atm prl1, atr prl1, and sog1 prl1 mutants, and their root lengths were similar to roots of prl1 (Figure 2E; Figure S9). Thus, stele cell death and a developmental defect in the primary root of prl1 were independent of the ATM/ATR-SOG1 signaling cascade, further suggesting that either ROS imbalance is not the cause of cell death or that other molecules signal ROS imbalance to cell death. Together, ROS imbalance is enough to cause root growth defects such as in prl1, but it was undetermined the necessary cause for prl1 root retardation.

The principal cause of root growth retardation is the de-polymerization of microfilaments, which also causes reactive oxygen species imbalance, and stele cell death in pleiotropic regulatory locus 1

In light of the general disruption of root cell behaviors in prl1, we considered the alteration of the cytoskeleton, the fundamentally cellular determinant for growth and development, as a foundation stone for developmental root defects. Previously, we determined that cortical actin filaments, not microtubules, was depolymerized in pavement cells of cotyledons of prl1, which explained the polar developmental defects.³⁹ Thus, we constructed a homozygous double GFP- fABD2 prl1 mutant and found that the cortical actin cytoskeleton in the epidermal cells of prl1 root was depolymerized (Figure 3A; Figure S10), which indicated that PRL1 functions in maintaining cortical actin integrity. We treated WT seedlings with actin polymerization inhibitor Latrunculin B (Lat B) and found that it depolymerized the actin cytoskeleton (Figure 3A) and hindered root growth (Figures 3A and 3B), which suggested that actin depolymerization was sufficient for the inhibition of root growth.

Figure 3.

Depolymerization of microfilaments is the principal cause of root growth retardation, ROS imbalance, and stele cell death in prl1

(A) Cortical actin configurations in epidermal cells of WT, prl1, WT (1 μM Lat B), and WT (0.5 mM PG). In WT, actin formed filaments. In prl1, actin was disassembled. In Lat B-treated WT, actin filaments were absent. In PG-treated WT, actin formed into a transverse arranged bundle. Scale bar = 10 μm.

(B) Roots of WT and Lat B-treated WT. One day after germination, seedlings were transferred into ½ Murashige and Skoog medium (MS) containing 0 μM or 1 μM Lat B and raised for 5 days. Scale bar = 1cm.

(C) Quantification of root length of WT and 1 μM Lat B-treated WT seedlings. Values are means ± SD of more than 30 roots 6 days after germination. Asterisk indicates a significant difference (unpaired t-test, two-tailed p < 0.0001).

(D) Roots of WT and Lat B-treated WT stained with DAB. Scale bar = 200 μm.

(E) Roots of WT and Lat B-treated WT stained with NBT. Scale bar = 200 μm.

(F) PI staining of root stele cell shows live status in pWOX5::GFP, anac085, prl1, anac085prl1, Lat B-treated 6-/14-day after germination anac085 and pWOX5::GFP, PG-treated and DPI-treated WT, the ratio presents the frequency of dead cells in observed cells. pWOX5::GFP was used to indicate QC position. Scale bar = 50μm. “※” marks dead stele cell.

As actin microfilament (MF) depolymerization or ROS imbalance alone could lead to root growth retardation, to clarify their relationship in prl1, we measured ROS status in the root of Latrunculin B (Lat B)-treated WT and found that ROS were similar to the prl1-like ROS imbalance (Figures 3C and 3D), which indicated that actin depolymerization can cause ROS imbalance. To rule out the possibility that ROS imbalance was a side effect of Lat B treatment, we observed that the configuration of the actin cytoskeleton in response to ROS imbalance generated by PG treatment was transverse-arranged actin bundles in cells of elongation zone (EZ) (Figure 3A). This configuration was distinct from that in prl1, which demonstrated that ROS imbalance was induced by actin depolymerization in prl1 and further suggested that ROS imbalance had a different mechanism for root growth inhibition from prl1. In prl1 roots, we observed stele cell death (Figure 3F). Because ROS imbalance is always the mediator of cell death in response to stress,^21,37 we determined whether ROS imbalance (pWOX5: GFP used for convenient observation of QC position; Figure 3F) would generate cell death. We did not find any cell death in roots of PG- and DPI-treated WT, which indicated that ROS imbalance was not the cause of the stele cell death in prl1; instead, both ROS imbalance and cell death were results in response to PRL1 mutation. Further, although we could not completely reproduce stele cell death in Lat B-treated WT for 5 days, we observed root stele cell death in some roots, which indicated that actin depolymerization was likely the cause of stele cell death in prl1 (Figure 3F).

By analyzing the RNA-seq data, we found that ANAC085/NAC085 was highly induced in prl1. The ANAC085/NAC085 gene is a member of the stress-activated ANAC/NAC transcription factor family⁴⁰ and is usually expressed at a relatively low level in all tissues. We used RT-qPCR to measure ANAC085 expression and found that it was upregulated compared to other members ANAC044, ANAC065, and ANAC071 of the ANAC transcriptional factor family in prl1 (Figure S11), and this result was reproduced in Lat B-treated WT, and PG-treated WT (Figure S12); thus, both actin depolymerization and ROS imbalance can induce ANAC085 expression. We presumed that ANAC085 would transduce the prl1 mutation-triggered signal to stele cell death. Indeed, in a double homozygous anac085 prl1 strain, but not in an anac044 prl1 mutant, we found stele cell death was either completely absent or the degree of cell death was greatly diminished (Figure 3F), which indicated that ANAC085 was necessary for cell death in prl1. Further, there was no cell death in the Lat B-treated anac085 null mutant (Figure 3F), which strengthened the conclusion that root stele cell death was the result of actin depolymerization and mediated by ANAC085. Although ANAC085 was induced by ROS imbalance (Figures S12 and S13), we did not find cell death in roots of PG- and DPI-treated WT (Figure 3F), suggesting that the up-regulation of ANAC085 alone was not sufficient for stele cell death. In prl1, transcription factor ANAC085 in combination with actin cytoskeleton depolymerization produced stele cell death. We proposed that actin depolymerization was the primary cause for retarded root growth, ROS imbalance, and stele cell death in prl1.

In addition to its function as an intracellular scaffold supporting cell division and expansion, the reorganization of the actin cytoskeleton or disruption of the equilibrium of actin turnover is recognized as a danger signal that triggers events such as cell death, ROS accumulation, and DNA fragmentation in pollen of self-incompatible plants, yeast, and mammalian cells.^30,31 Hence, most likely actin depolymerization in prl1 was recognized as a danger signal to initiate sequential events. In normal growth conditions, single anac085 and anac044 mutants have a wild-type phenotype to WT; however, in a genetic background of prl1, defects in ANAC085 and ANAC044 make more serious seedlings growth inhibition at 11 days after germination (Figure S14). These observations support the aforesaid conclusion that the prl1 mutation triggered stress which further inhibited seedling growth and the findings reaffirmed the function of ANAC085 and ANAC044 in making a trade-off between stress resistance and growth.⁴⁰

Depolymerization of microfilaments in pleiotropic regulatory locus 1 causes widespread changes in stress-related transcription

To confirm at the transcriptional level that the PRL1 mutation-caused actin depolymerization-initiated hierarchical stress signals, we performed RNA-seq analysis to prl1, Lat B-treated WT, and WT in three biological replicates. The significant thresholds for differentially expressed genes (DEGs) between each experimental group and WT were settled with |log2FoldChange|≥1 and padj<0.05. We identified 714 upregulated (hyper-DEGs) and 1,218 down-regulated (hypo-DEGs) genes in prl1 (Figure S15A, File Set S1). There were 1,596 hyper-DEGs and 1,431 hypo-DEGs in the Lat B-treated WT (Figure 4B (Figure S15B, File Set S2). There were 800 DEGs in common between prl1 and Lat B-treated WT and the overlap of the DEGs was significant (Fisher`s exact test, p = 0; Figure 4A), indicating that prl1 and actin depolymerization regulated the same group of genes and shared similar actin depolymerization-related signaling pathways.

Figure 4.

PRL1 maintains microfilament homeostasis and cellular homeostasis

(A) Venn diagram showing the degree of overlap between differentially expressed genes (DEGs) in prl1 and Lat B-treated WT.

(B) GO enrichment analysis of DEGs in prl1 showing that they are stress-related.

(C) GO enrichment analysis of DEGs in Lat B-treated WT showing that they are stress-related.

(D) A Venn diagram showing the degree of overlap between genes with retained introns in prl1 and Lat B-treated WT.

(E) GO enrichment analysis of overlapped intron retention genes in prl1 and Lat B-treated WT showing that they are RNA splicing-related.

(F) A model for PRL1 functions in root development. Mutation in PRL1 leads to cortical actin depolymerization which is responsible for growth and development defect of prl1 with short primary root. In parallel, actin depolymerization triggers hierarchical stress/danger signals including ROS imbalance, DEGs, splicing defects, and cell death which would make feedback on the actin cycle and plant development.

To analyze their functional correlations, we performed Gene Ontology (GO) enrichment analysis on the DEGs. The DEGs in two independent experiments were enriched in stimulus response, including response to toxic substances, response to chemicals, and response to stress (Figures 4B and 4C). Thus, the GO analysis indicated that the prl1 mutation and actin depolymerization were related to stress, further indicating that PRL1 performs its functions by maintaining the integrity of actin polymerization. Interestingly, in prl1 and Lat B-treated WT, DEGs were enriched for biological function related to oxidative stress. The ROS imbalance was the common result of null PRL1 mutation and actin depolymerization; thus, we compared DEGs between PG-treated WT and WT and identified 795 hyper-DEGs and 1,724 hypo-DEGs (Figure S15C, File Set S3). We found that 549 DEGs overlapped, and the overlap level was significant (SuperExactTest, p = 0) (Figure S15D), which indicated an extensive stress-related transcriptional alteration in response to ROS imbalance. This result demonstrated that DEGs were not only related to PRL1-regulated transcription but they were also related to responses to actin depolymerization- or ROS imbalance-related stress.

In addition to transcription regulation, PRL1 within the MAC/Prp19 complex regulates alternative splicing by the activation of the spliceosome complex.³⁶ Because alternative splicing is a strategy for plants to respond to stresses,⁴⁰ we hypothesized that an alternative splicing defect in prl1 was the result of stress originated by actin depolymerization. Intron retention is a major type of alternative splicing defect in Arabidopsis⁴¹; thus, we analyzed intron retention in prl1 and Lat B-treated WT and found that 664 cases of intron retention in prl1 and 401 cases in Lat B-treated WT. One hundred eighty-seven of these cases were shared between the two conditions (Fisher`s exact test, p = 0) (Figure 4D), which indicated one case that intron retention defects partially resulted from actin depolymerization in prl1.

We performed GO enrichment analysis of the common cases of intron retention and found that they were enriched in genes for spliceosome function and regulation of mRNA splicing (Figure 4E), including splicing regulator SR30 (AT1G09140), SR33 (AT1G55310), and SUA (AT3G54230) (File Set S4), which indicated that prl1 and actin depolymerization shared alternative splicing defects (intron retention) originated from alternative splicing defects in regulators of spliceosome. Biotic and abiotic stresses are known to change the alternative splicing of splicing regulators.⁴² Hence PRL1 mutation-initiated stress would be an alternative cause of the alternative splicing defects but not only its direct role in the activation of spliceosome by interacting with specific components of the spliceosome complex. Thus, DEGs and alternative splicing defects in prl1 were probably the result of actin depolymerization-initiated stress.

To determine whether there was a correlation between the presence of splice variants and gene expression, we analyzed the distribution of intron-retained genes and found them randomly distributed in increased, reduced, and unchanged regions. Thus, there was no correlation between intron retention defects in prl1 and Lat B-treated WT (Figures S15F and S1G) which was consistent with a report by Jia et al..¹⁰ In sum, DEGs and alternative splicing defects were independent responses to actin depolymerization in prl1.

Discussion

Microfilament de-polymerization is dominant to the effect of reactive oxygen species imbalance on retarded root development

The actin cytoskeleton is an essential determinant of various cell behaviors, including cell division, cell expansion, cell morphogenesis, and vesicle transport, which have been experimentally demonstrated.^25,43 The length of a root is determined by a combination of cell proliferation and directional cell elongation, with actin playing a crucial role in both processes.^17,44 During mitosis, the actin cytoskeleton undergoes dynamic changes to perform temporally specific cell division functions. In the elongation zone, cells expand directionally parallel to the longitudinal axis of the primary root.^44,45 Root development is influenced by both exogenous and endogenous signals, with the actin cytoskeleton being a typical target of upstream signals that modulate ABPs for the turnover of actin configuration. Therefore, actin de-polymerization is intrinsic to the immediate cause of the short primary root in the prl1 mutant.

ROS imbalance is a common response to internal or external stress signals, which can affect various processes in cells, such as physiology, signaling, membrane properties, cell wall relaxation, and eventually cell death.¹⁹ Consequently, ROS imbalance can indirectly suppress root growth. Either ROS imbalance or actin depolymerization could alone be the sufficient cause for the short root by disrupting cell division and elongation. To elucidate the exact mechanism underlying PRL1-regulated root development, it is essential to know the cause-effect relationship between ROS accumulation and actin depolymerization in prl1. Gradually, there were increasing evidence showing that actin depolymerization was recognized by cells to be damaged and triggered hierarchical damage-associated signals and cell events including ROS imbalance, DEGs, genome instability, and so on.^30,32,46 Our experiment showed that actin depolymerization could initiate ROS imbalance but ROS imbalance caused the formation of the transversely arranged actin bundle different from actin configuration in prl1 which is enough to say actin depolymerization is the cause of ROS imbalance. Cellular homeostasis is always characterized by the accumulated level of ROS.¹⁹ So, our evidence indicated that PRL1 maintains actin integrity for root development and concomitantly maintains cellular homeostasis.

Stele cell death is a locally specific cell response to actin depolymerization but is not responsible for whole developmental root defect in pleiotropic regulatory locus 1

Actin turnover homeostasis is always in the control of the exogenous and endogenous signals.²⁵ Meanwhile, the actin configurations are always monitored by the cell system to keep its constant turnover.^25,47,48 Mounting evidence indicated that the long-lasting alteration of actin configuration such as actin depolymerization would be recognized as a damage-associated molecular pattern triggering dangerous signals causing DEGs, DNA damage, and ROS accumulation^28,30,31,32 which in turn, in some conditions, could lead to development defects. But when all these events exist simultaneously, it should be thoughtful of the cause-effect relationships among them. In this study, we observed stele cell death and QC expansion in the prl1 root, then we predicted that the prl1 root encountered stress although we did not know whether stele cell death or QC expansion was responsible for its developmental root defects. So, we checked ROS status and found its imbalance in the prl1 root. As hydrogen peroxide accumulation could cause DAN damage which in turn initiates cell defense mechanisms such as DNA damage responses (DDR) including cell cycle inhibition or cell death to keep DNA integrity,^{19,48,49,50,51} we surmised stele cell death is the result of hydrogen peroxide accumulation and mediated by ATM/ATR-SOG1 signaling module.^34,38 Unexpectedly, experimental evidence showed stele cell death was not mediated by SOG1 but by another ANAC family member ANAC085 indicating an independent signaling pathway transducing to cell death. There were evidence that showed different members of the transcription factor ANAC family are responsive to different stress signals to cause their own specific cell events.^40,52,53,54 Our evidence showed that ANAC085 did not message ROS accumulation but actin depolymerization to stele cell death. Does that mean actin depolymerization-caused stele cell death or ANAC085 expression enhancement is independent of ROS imbalance or of ROS imbalance-caused DNA damage? In combination with the two facts that both actin depolymerization and ROS imbalance could activate expression of ANAC085 and artificially depolymerizing actin filament could make ROS imbalance, it should say ROS imbalance is a concomitant cellular response to actin depolymerization and ROS imbalance is needed for the activation of ANAC085 expression. Thus, it also reinforced our opinion that actin depolymerization is the principle cause of developmental root defects in prl1. The disappearance of stele cell death and persistence of short root of anac085 prl1 indicated that stele cell death was not responsible for whole root defects and further indicated that stele cell death and whole root defects were parallel results of actin depolymerization. So, those indicated that PRL1 keeps stele cells alive by maintaining actin integrity. In terms of role of PRL1 in DNA damage or genome maintenance, without the addition of DNA toxic chemicals into the medium, the root of atr or atm mutant, without damage DNA repairment-requiring ability, grows normally without cell death, meaning random DNA damage would not cause stele cell death and hinder root growth; while prl1 and atr prl1, atm prl1 and sog1 prl1 with or without damage DNA repairment-requiring ability growing in ½ MS shared stele cell death indicating PRL1 does not function in the repairment of DNA damage but in keeping endogenous cellular homeostasis. Although some articles suggested PRL1 functioned in the repairment of DNA damage, it should be known that would not be its principle function. Without ANAC085, anac085 prl1 got shorter root and kept ROS imbalance compared to prl1 indicating the regular role of ANAC family members in trade-off between growth and stress resistance and ANAC085-mediated stele cell death is a local cellular response to actin depolymerization caused by loss-of-function PRL1.

What is the role of pleiotropic regulatory locus 1 in regulating gene transcription?

Previous studies showed that PRL1 could regulate gene transcription by controlling alternative splicing as a component of the activation complex to spliceosome or by affecting the metabolism of non-coding RNA (miRNA/SiRNA),^3,9,10,11,36 but the exact mechanisms were unknown. Other studies illustrated that PRL1 negatively controls stress-related signals.^10,22 What is the function relationship between PRL1 in alternative splicing or ncRNA metabolism and stress-related differentially expressed genes? Are genes alternative splicing or ncRNA metabolism related to the roles of PRL1 in development or innate immunity? Here, we would ask whether PRL1-controlled actin integrity is controlled by those molecular functions? Jia`s work had given some answers that the AS role of PRL1 was independent of its regulatory role in differential gene expression. If PRL1 performs as a neutral regulatory component in ncRNA metabolism or precursor message RNA alternative splicing without upstream selective signals, why are DEGs related to stress signals? In most circumstances, stresses could cause defects in AS or the metabolism of ncRNA. In this work, our RNA-seq results that were consistent with Jia`s and we got further information that AS defects and DEGs were related to actin deplymerization-initiated stress signals. That is to say PRL1 is not a regulatory component in molecular machinery but an indirect regulator by maintaining intracellular endogenous homeostasis. As previous work given some evidence that physical interactions PRL1 physically interacts with some molecular components in regulatory molecular machinery, it had to be considered that PRL1 has two split roles in regulating gene transcription levels, one as a component of molecular machinery, the other as a defender for intracellular homeostasis. As the revealing molecular mechanism underlying actin depolymerization in prl1 is the definite target of our following work, this ambiguous circumstance will be solved sooner.

Using our limited data, we proposed an alternative model that mutation in PRL1 led to cortical actin depolymerization which resulted in growth and development defects. Concomitantly, actin depolymerization was recognized as a stress/danger signal that initiates signaling events such as ROS imbalance, DEGs, splicing defects, and cell death. These signaling events had further effects on the development of the plant by making feedback effects (Figure 4F).

Limitation of the study

Although the finding of actin depolymerization is the key point for developmental root defect in prl1 and the finding of PRL1`s new function in maintaining actin integrity is important, what a pity is it without revealing the molecular mechanism of how actin filament is depolymerized. It makes many connections among functions of PRL1 unestablished and ambiguous problems unsolved.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Bacterial and virus strains
E. coli: DH5α	ThermoFisher Scientific	18265017
E. coli: DB3.1	Lifescience	S0043
A.tumefaciens: GV3101	Lifescience	STR3005
Chemicals, peptides, and recombinant proteins
Latrunculin B	Sigma-Aldrich	L5288
Diaminobenzidine (DAB)	Sigma-Aldrich	D12384
Nitroblue tetrazolium (NBT)	Sigma-Aldrich	11383213001
ո-propyl gallate (PG)	Sigma-Aldrich	48710
Diphenyleneiodonim (DPI)	Sigma-Aldrich	D290
Potassim iodine (KI)	Sigma-Aldrich	30315
Deposited data
Raw data	This paper	GAS: CRA016896 https://bigd.big.ac.cn/gsa/brrowse/CRA016896
Experimental models: Organisms/strains
Arabidopsis: Col-0, pr1-12 (Salk_039427), atr (Salk_083543), atm (Salk_040423), sog1 (Salk_039420), anac085 (Salk_208662), anac044 (Salk_054551)	ABRC	N/A
Arabidopsis: prl1-10	Gao et al.39	N/A
Arabidopsis: QC25: GUS	Sabatini et al.55	N/A
Arabidopsis: pWOX5:: GFP	Haecker et al.56	N/A
Arabidopsis: CycB1;1: GUS	Colón-Carmona et al.57	N/A
Arabidopsis: GFP-fABD2	Sheahan et al.58	N/A
Arabidopsis: p35S::PER23-GFP, QC25:GUS prl1, pWOX5::GFP prl1, CycB1;1: GUS prl1, GFP-fABD2 prl1, P35S::PER23 prl1	This paper	N/A
Oligonucleotides
Primers used in this paper see Table S1	This paper	N/A
Recombinant DNA
Plasmid pGWB605	Dr. Nakagawa (Shimane University)	https://doi.org/10.1271/bbb.100184
Plasmid pDONR207	Lifescience	PVT11146
Plasmid 207-PER23	This paper	N/A
Plasmid 35S::PER23-GFP	This paper	N/A
Software and algorithms
Arabidopsis Information Resource	N/A	https://www.arabidopsis.org/
ImageJ	Schindelin et al.59	https://imagej.nih.gov/ij/
GraphPad Prism8	Graphpad	https://www.graphpad.com/scientific-software/prism/

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Dr Xiaowei Gao (gxw_cell@fafu.edu.cn).

Materials availability

Mutants generated in this study are available without any restriction from the lead contact, Xiaowei Gao (gxw_cell@fafu.edu.cn).

Data and code availability

All data reported in this paper will be shared by the lead contact upon request.

This paper does not report original code.

All additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Experimental model and study participant details

Plant materials and growth conditions

The Arabidopsis thaliana plants were in the background of the Columbia (Col-0) ecotype. For convenience, the prl1 mutants are referred to as prl1 and included a single nucleotide substitution mutant (prl1-10) and a T-DNA insertional line Salk_039427. All the T-DNA insertion mutants were obtained from the Arabidopsis Biological Resource Center and included Salk_039427(prl1-12), Salk_054551(anac044), Salk_208662(anac085), Salk_039420(sog1), Salk_040423(atm), and Salk_083543(atr). All mutations were confirmed by PCR with primers listed in Table S1. The QC25: GUS, pWOX5: GFP, and CycB1;1: GUS seeds were kindly provided by Dr. Chuanyou Li (IGDB, CAS, China). The seeds of actin marker line GFP-fABD2 were provided by Dr. Deshu Lin (HIST, FAFU, China). All double mutants, pWOX5: GFP prl1, GFP-fABD2 prl1, p35S: PER23, p35S: PER23 prl1, anac085 prl1, anac044 prl1, atm prl1, atr prl1, sog1 prl1, and cycB1;1: GUS prl1 were obtained by crossing and confirmed by PCR (primers listed in Table S1). Seedlings were germinated on ½ MS agar plates containing 1% sucrose at 22°C with 16 h light and 8 h dark periods.

Mehod details

Construction and of p35S: PER23 overexpression lines

For construction of the p35S: PER23-GFP overexpression line, the coding sequence of PER23 (AT2G38390) was cloned by PCR, inserted into pDONR207, and sequenced to ensure accuracy; the fragment was then cloned in pGWB605, which was introduced into Agrobacterium tumefaciens GV3101 using the routine floral dip method.⁶⁰ The T1 seeds were broadcast in the soil, germinated for two days, and spayed with diluent Basta several times; surviving seedlings were transplanted to new pots. Seeds of transgenic lines were produced individually, and the homozygous transgenic lines were screened in T2 and T3 generations. The selected homozygous transformants were used for experiments.

Chemical pretreatment of seedlings

To evaluate the relation between reactive oxygen species (ROS) and actin depolymerization and their effects on root development, the germinated seeds in ½ MS medium for one day were transferred to PG- (0.5 mM), DPI- (0.25 μM), KI- (1 mM) and LatB-containing (1 μM) ½ MS medium separately and raised for five days further. The 6-day seedlings were used for root length evaluation, ROS staining, actin imaging, and RNA-seq experiments.

Analysis of root length

For root length measurements, the 6-day roots of WT or homologous mutant lines were photographed while growing on agar. The root length data were collected using Image J software³¹ and statistically analyzed with GraphPad Prism 8 with three biological repeats.

Root zones and cortical cell number analysis

Root tips of seedlings were photographed with DIC optics on an Eclipse Nikon Ni-U Upright Microscope. The number (root meristem cell number defined as the number of cells in the cortex file extending from the QC to the transition zone) and length of epidermal cells were analyzed using Image J.

Root PI staining and QC analysis (cell death observation)

Stain roots with 10 μg/mL PI (Sigma) for 5-10 min before imaging. After staining, mount the roots of pWOX5: GFP and pWOX5: GFP prl1 on glass slides and add distilled water to dip them. Gather confocal images with Zeiss LSM880. Use 488 nm and 561 nm laser light to excite GFP and PI fluorescence, respectively. Collect the emitted fluorescent light between 500 and 550 nm (GFP) and 600 and 656 nm (PI). Obtain Z stacks images with 0.5-μm sections, which were averaged two times. Use Zeiss confocal software and Adobe Photoshop CC 2018 to process gathered images.

Beta-glucuronidase staining

Six-day seedlings of CycB1;1: GUS, CycB1;1: GUS prl1, QC25: GUS and QC25: GUS prl1 grown on ½ MS media in light were used for beta-glucuronidase staining. All roots were immersed in staining solutions (0.1 M NaPO₄ pH7.0, 10 mM EDTA, 0.1% (V/V) Triton X-100, 1 mM K₃Fe (CN)₆, 2 mM X-Gluc) for 2 h to overnight at 37°C, followed by washing three times with 75% ethanol until root tissue was clear (Aida et al., 2004; Bieleszová et al., 2019). The stained roots were observed with a Nikon Ni-U Upright Microscope equipped with a Digital Slight 10 camera.

Starch granules (Lugol) staining

The root of the 6 DAG seedlings were submerged in Lugol solution (Sigma) for 5min, rinsed with distilled water, cleared with clearing buffer (chloral hydrate:glycerol:water in 8:3:1 ratio),⁶¹ then observed and photographed using a DIC optic on a Nikon Ni-U upright Microscope equipped with a Nikon Digital Slight 10 Camera.

Nitroblue tetrazolium (NBT) staining

Whole seedlings were used 6 days after germination. Roots were stained for 10-15 min in 2 mM nitroblue tetrazolium in 20 μM phosphate buffer pH 6.1. The reaction was stopped by transferring the seedlings to distilled water.⁶² The roots were observed with an Eclipse Nikon Ni-U Upright Microscope equipped with a Digital Slight 10 camera. Settings were identical for all images in an experiment. Each experiment was repeated at least three times with similar results.

Diaminobenzidine (DAB) staining

Whole seedlings were used 6 days after germination. The seedlings were placed whole in freshly prepared DAB solution (10 mg 3,3`-diaminobenzidine in 10ml distilled water, pH adjusted to 3.8 with HCl) for 4 h in the dark at ambient temperature. The staining solution was removed and the seedlings were washed twice with distilled H₂O.⁶² Images were acquired with Nikon Ni-U Upright Microscope equipped with a Digital Slight 10 camera. Settings were identical for all images in an experiment. Each experiment was repeated at three times.

Confocal microscopy live imaging of cortical actin

To visualize the cortical actin configuration, laser-scanning confocal microscopy (LSM 880 with Airyscan) was used to image the epidermal cells in different zones of the primary roots of GFP-fABD2, PG- and Lat B-treated GFP-fABD2 and GFP-fABD2 prl1. Roots grown to six days were mounted on slides, distilled water was added, and the slides were covered with coverslips. Z stacks of optical sections and Z-series maximum intensity projections of GFP fluorescence were generated to determine actin configurations.

Real-time qPCR

RNA was extracted using TRIzol Reagent (Invirtrogen) from the seedlings of WT, prl1, anac085, PER23 over-expressing lines, the LatB- and PG-treated WT respectively. First-strand cDNA was synthesized using the SuperScript III cDNA Synthesis kit (Invitrogen). RT qPCR was performed using 2×Hieff qPCR SYBR Green Master Mix and PCR reactions and fluorescence detection were performed in a Mastercycler ep realplex (Eppendorf). ACTIN 2 (ACT2) was used as the internal control. Three technical replicates of the RT-qPCR were performed using three biological replicates. Primers used for qPCR are listed in Table S1.

RNA-seq analysis

RNA-seq analysis was conducted at the Personal Biotechnology Co., Ltd (Shanghai, China). Total RNA was extracted with the RNeasy Plant Mini Kit (Qiagen). The RNA-seq libraries were constructed using the TruSeq RNA Sample Prep Kit (Illumina). The mRNA was purified using oligo(dT) magnetic beads and then fragmented to about 300 nucleotides. The first-strand cDNA was synthesized by reverse transcription with random hexamer primers, followed by conversion to double-stranded cDNA. The fragments were amplified with adaptor-specific primers to enrich the fragment of library. The DNA library quality was tested with an Agilent 2100 Bioanalyzer, then the total and effective concentration of DNA was measured. The libraries were applied to the Illumina Next-generation Sequencing platform for 150-nt Paired-end sequencing. Three biological replicates were performed. For each sample, the raw read number was over 38 million. After filtration, the clean reads occupied 92% of sequenced reads. The filtration steps included removal of reads containing only the adaptor sequences and removal of reads with average quantity lower than molecular quantity 20. About 96% of the useful reads were uniquely mapped to the reference Arabidopsis thaliana TAIR10 genome using HISAT2 (http://ccb.jhu.edu/software/hisat2/index.shtml). At least 99.3% of the clean reads could be mapped to exons, and about 99.0% of clean reads could be mapped to genes. Gene annotation was performed by reference to Ensembl (http://www.ensembl.org/). Gene expression levels were normalized based on fragments per kilobase of exon model per million mapped fragments. DESeq was used for analyzing differentially expressed genes (DEGs; |log2 FoldChange|>1, P-value <0.5). We analyzed the numbers of common and specific DEGs, and constructed venn diagram of these DEGs, and used Fisher`s Exact Test to evaluate overlapping significance of DEGs between two experimental groups and SuperExactTest to evaluate overlapping significance of DEGs among three experimental groups about their functional intersection. The program topGO was used for GO enrichment analysis. The software rMATs (http://rnaseq-mats.sourceforge.net/index.html) was used for analyzing alternative splicing events. The quantification of alternative splicing events was conducted with rMATS statistic model. Δф (exon inclusion level) was used as the standard confirming production of alternative splicing between two experimental samples (Δф>5% and FDR<= 1%). We also used Fisher`s Exact Test to evaluate the significance of genes shared between prl1 and LatB-WT experimental groups and having retained introns.^10,63 The raw sequence data reported in this paper have been deposited in the Genome Sequence Archive in National Genomics Data Center, the accessing link: https://bigd.big.ac.cn/gsa/browse/CRA016896.

Quantification and statistical analysis

Details of RNA-seq data analysis were indicated in method. Root length measurement and root meristem cell number count were carried out using ImageJ. In each related experiment, number of repeats, sample size and P value were indicated in the figure legends or the results. Data were analyzed using GraphPad8 software. Significance difference were determined using unpaired t-test and indicated in legends.

Published: June 28, 2024