Phenylmercury stress induces root tip swelling through auxin homeostasis disruption

Abstract

We previously reported that in Arabidopsis, the phytochelatin-mediated metal-detoxification machinery is also essential for organomercurial phenylmercury (PheHg) tolerance. PheHg treatment causes severe root growth inhibition in cad1-3, an Arabidopsis phytochelatin-deficient mutant, frequently accompanied by abnormal root tip swelling. Here, we examine morphological and physiological characteristics of PheHg-induced abnormal root tip swelling in comparison to Hg(II) stress and demonstrate that auxin homeostasis disorder in the root is associated with the PheHg-induced root tip swelling. Both Hg(II) and PheHg treatments severely inhibited root growth in cad1-3 and simultaneously induced the disappearance of starch-containing plastid amyloplasts in columella cells. However, further confocal imaging of the root tip revealed distinct effects of Hg(II) and PheHg toxicity on root cell morphology. PheHg treatment suppressed most major genes involved in auxin homeostasis, whereas these expression levels were up-regulated after 24 h of Hg(II) treatment. PheHg-triggered suppression of auxin transporters PIN1, PIN2, and PIN3 as GFP-fusion proteins was observed in the root tip, accompanied by an auxin reporter DR5rev::GFP signal reduction. Supplementation of indole-3-acetic acid (IAA) drastically canceled the PheHg-induced root swelling, however, Hg(II) toxicity was not mitigated by IAA. The presented results show that the collapse of auxin homeostasis especially in root tips is a cause for the abnormal root tip swelling under PheHg stress conditions.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11103-024-01538-6.

Key Message

An organomercury phenylmercury disrupts auxin homeostasis and induces abnormal root tip swelling.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11103-024-01538-6.

Introduction

The edaphic environment exhibits a diversity of metal composition, resulting from anthropogenic and/or geological influences. A large area of arable soils is deficient in the availability of nutritional essential metals such as iron and zinc (Alloway 2009; Abadía et al. 2011), while anthropogenically influenced soils often contain practically considerable levels of toxic metals such as cadmium, lead, or mercurials (Awasthi et al. 2016; Stein et al. 2017; Uraguchi et al. 2019). As sessile organisms, plant roots encounter their surrounding soil environment and are forced to selectively absorb essential elements and cope with toxic metal stress to maintain growth. Toxic metal contamination of agricultural soils is recognized as a risk factor harming both crop production and human health (Panaullah et al. 2008; Clemens 2019; Wang et al. 2019). Understanding how plant roots respond to and cope with toxic metals in soils is therefore essential to provide a basis for mitigating these risks.

A major and general mechanism for toxic metal detoxification in plants is the phytochelatin (PC)-mediated machinery. PCs are cysteine-rich peptides with the general structure of (γ-Glu-Cys)n-Gly (n is usually between 2 and 7) (Grill et al. 1987), and are non-ribosomally synthesized by PC synthases (PCSs) from their precursor glutathione (GSH) (Clemens et al. 1999; Ha et al. 1999). PCs are synthesized in the cells challenged by a variety of toxic metal ions (Howden et al. 1995; Tennstedt et al. 2009; Uraguchi et al. 2018), and can capture cytosolic toxic metal ions (Sadi et al. 2008; Liu et al. 2010; Uraguchi et al. 2021). In addition to these initial steps, the subsequent ABC transporter-mediated vacuolar sequestration is also physiologically essential for plants to accomplish metal detoxification (Song et al. 2010; Park et al. 2012; Uraguchi et al. 2020). These results demonstrate the primary role of the PC machinery in plant metal detoxification, corroborated by the drastic collapse of metal tolerance in the loss-of-function mutants of Arabidopsis PCS1, ABCC1, and ABCC2.

Several significant and interesting questions remain regarding PC-mediated plant metal tolerance. Among is the cell-type specificity of the PC-system, especially in the roots that encounter the toxic metal ions in soils. In Arabidopsis, an early study using GFP-reporter lines suggests that AtPCS1 is cytosol-localized and expressed both in roots and shoots (Blum et al. 2010). Recently, we further examined the cell-type specificity of AtPCS1 expression in roots using the pAtPCS1-AtPCS1-GFP line and demonstrated that its expression is restricted to lateral root cap in the root apical meristem and epidermal cells in the elongation zone (Uraguchi et al. 2022b). These cell-type specific expression patterns driven by the AtPCS1-native promoter sufficiently rescued the hypersensitivity of the AtPCS1 loss-of-function mutant cad1-3 under arsenite or mercurials stress conditions, suggesting that AtPCS1-mediated PC machinery functions in these surface root cell types as a firewall protecting the inner root cells from the toxic metal ions (Uraguchi et al. 2022b).

Another question raised from our observations of Arabidopsis WT and cad1-3 roots is the existence of metal species-specific or chemical form-specific root responses (Uraguchi et al. 2017, 2018, 2020, 2022b). Typical metal-specific root responses are found for arsenite [As(III)] and cadmium [Cd(II)] stress. As(III) simply stunts the root elongation, but Cd(II) treatment leads to wavy root growth in the WT. Chemical form-specific differences of the root phenotypes are observed under the mercury stress, inorganic mercury [Hg(II)], and phenylmercury (PheHg), an organomercurial that was used as a pesticide worldwide in the middle of the 20th century (Ishiyama et al. 1965). Hg(II) stress tends to increase the slanted root elongation to the left on vertical agar plates (when seeing the plate from the front) (Uraguchi et al. 2020), on the other hand, PheHg treatment induces an abnormal root tip enlargement/swelling phenotype (Uraguchi et al. 2022b). These root responses suggest that the mode-of-actions of toxicity are different between Hg(II) and PheHg. The swollen root tip induced by PheHg stress is associated with the severe arrest of root elongation and is more frequently observed in the AtPCS1 loss-of-function mutant cad1-3. Complementation of cad1-3 with pAtPCS1-AtPCS1-GFP rescued the swollen root phenotype and growth defect under PheHg treatment, demonstrating that AtPCS1-mediated phytochelatin synthesis detoxifies PheHg as in the case of Hg(II) and other inorganic toxic metal ions. These results suggest the root responses and morphology were unique to the respective mercurial, suggesting a possibility that the mode-of-actions of toxicity are different between PheHg and Hg(II).

The objective of this study is to advance our understanding possible mechanisms underlying the metal-specific morphological changes using the merucials as a model, and to offer broader implications for improving plant resilience in metal-contaminated soils. We characterized the PheHg-induced root tip swelling in comparison to the effects of Hg(II) stress to understand the unique morphological changes of the roots responding to the toxic metals. We found that auxin homeostasis disorder is a key target related to the PheHg toxicity in roots. We conducted time-course analyses of microscopy and found detailed morphological differences between PheHg and Hg(II)-treated roots. Transcript analyses revealed contrasting effects of PheHg and Hg(II) stress and the auxin-related genes were generally down-regulated by PheHg. GFP-reporter analyses further supported the down-regulation of PIN transporters’ expression in the root tips by PheHg exposure, followed by DR5-signal depression. Supplementation of indole-3-acetic acid (IAA) drastically alleviated the PheHg toxicity, whereas it did not function for mitigating Hg(II) stress. These different responses of the root tip against PheHg and Hg(II) advance our understanding of plant root responses to toxic metal stress and provide a basis for developing strategies to mitigate the metal stress impact on agricultural systems. In addition to the wider Hg(II) pollution of soils (Li et al. 2017; Tang et al. 2018), PheHg contamination still exists in some soils (Hintelmann et al. 1995; Deonarine et al. 2015).

Materials and methods

Plant materials

Arabidopsis thaliana Col-0 as wild-type and the AtPCS1 null mutant cad1-3 (Howden et al. 1995) were used for physiological and microscopic root observations. cad1-3 was used for transcript analyses. The auxin-related GFP reporter lines, pPIN1-PIN1-GFP (Benková et al. 2003), pPIN2-PIN2-GFP (Xu and Scheres 2005), pPIN3-PIN3-GFP (Žádníková et al. 2010), and DR5rev::GFP (Friml et al. 2003) were used for confocal microscopic observations.

Growth conditions and mercury treatment

One-tenth strength modified Hoagland medium was used for plant cultivation [100 µM (NH₄)₂HPO₄, 200 µM MgSO₄, 280 µM Ca(NO₃)₂, 600 µM KNO₃, 5 µM Fe- N, N’-di-(2-hydroxybenzoyl)-ethylenediamine-N, N’-diacetic acid (HBED), 1% (w/v) sucrose, 5 mM MES, pH 5.7] (Tennstedt et al. 2009; Kühnlenz et al. 2014). Essential microelements other than Fe were not added to the medium to avoid possible interaction between mercurials and essential microelements. 1.5% purified agar (Nacalai Tesque) was added for the medium solidification. This agar reagent contains much lower levels of various mineral contents compared to the others and enables clear phenotype observations under metal(loid) stress (Uraguchi et al. 2020).

Arabidopsis seeds were surface sterilized by 5 min immersion with 70% (v/v) ethanol, followed by 1-min immersion with 99.5% ethanol. After air-drying, the sterilized seeds resuspended with 0.1% (w/v) agar were sown on control agar plates which did not contain mercurials. After stratification at 4 °C, plants were grown vertically in a growth chamber (16 h light/8 h dark, 22 °C). Plants were vertically grown for 3 or 4 d under control conditions and then transferred to HgCl₂ containing agar plates as Hg(II) treatment and C₆H₅HgOCOCH₃ containing plates as PheHg treatments, respectively.

Root growth assay

To compare the occurrence of the PheHg-induced root tip swelling between Col-0 and cad1-3, a frequency of the root tip enlarged plants was counted after exposure to 0.08 µM [6 days after transfer (DAT)] or 0.2 µM PheHg (4 DAT). Normally 30 seedlings for each genotype were examined per plate. An assay was conducted with at least three plates and repeated multiple times.

Root elongation under mercurial stress was also compared between Col-0 and cad1-3 under Hg(II) and PheHg stress conditions. Uniformly grown 4-d-old plants were transferred to 20 µM Hg(II) or 0.2 µM PheHg containing plates. Plants were photographed every day (1–4 DAT) and the primary root length of the identified seedlings was traced and measured with Image J software and the plugin Smart Root (Lobet et al. 2011).

Auxin supplementation assay

Effects of auxin supplementation on root sensitivity to mercurial toxicity were examined for Hg(II) and PheHg, respectively. 20 nM 3-indoleacetic acid (IAA, Sigma-Aldrich), 1-naphthaleneacetic acid (NAA, Sigma-Aldrich), or 2,4-dichlorophenoxyacetic acid (2,4-D, Sigma-Aldrich) was co-treated with 20 µM Hg(II) or 0.08 µM PheHg. Primary root length was quantified as the Hg(II) toxicity and a root tip swelling frequency was measured as the PheHg toxicity as described elsewhere.

Quantitative RT-PCR

4-d-old plants were transferred to 25 µM Hg(II) or 0.2 µM PheHg containing plates. After 4 h and 24 h of the treatment, roots and shoots were separately harvested and subjected to RNA extraction. For roots, the root tip (0–1 cm) was used. NucleoSpin RNA Plant (MACHEREY-NAGEL) was used for total RNA extraction. DNase treatment was applied during the RNA extraction. PrimeScript RT Master Mix (Takara Bio) was used for cDNA synthesis and quantitative RT-PCR was performed with PowerUp SYBR Green Master Mix (Thermo Fisher Scientific). For accurate normalization, geometric averaging of multiple internal control genes was applied (Vandesompele et al. 2002). In addition to the well-used housekeeping gene AtEF1a (At5g60390) (Remans et al. 2008; Kühnlenz et al. 2014; Uraguchi et al. 2022a), a SAND family gene (At2g28390), and an F-box protein gene (At5g15710) also served as internal control genes (Uraguchi et al. 2023). These two genes were suggested as more stable internal control genes, less affected by Cd and other heavy metal stress (Remans et al. 2008). A relative quantification method using standard curves was applied. The primer sequences were designed using the QuantPrime tool (Arvidsson et al. 2008) and are listed in Supplementary Table S1.

Lugol staining

To observe starch granules of the amyloplast in the columella cells, Lugol’s staining of Col-0 and cad1-3 was carried out as previously described (Stahl et al. 2009) with minor modifications. Roots treated with 20 µM Hg(II) or 0.2 µM PheHg (1–4 DAT) were mounted in a 1:5 dilution of Lugol’s solution (Sigma-Aldrich) with a clearing solution containing 70% (w/v) chloral hydrate and 10% (v/v) glycerol. After 5 min incubation, differential interference contrast (DIC) images of the root tips were obtained by a laser scanning confocal microscope (FV-3000, Olympus).

Propidium iodide staining

Counterstaining of cell walls and dead cells was achieved by propidium iodide (PI) to observe the root meristem morphology under 20 µM Hg(II) or 0.2 µM PheHg stress conditions. The roots of Col-0 and cad1-3 (1–4 DAT) were stained with 10 µg/mL propidium iodide (PI) for 2–5 min and then subjected to confocal imaging. The excitation and detection wavelengths for PI were 561 nm and 570–670 nm, respectively.

GFP expression analyses

GFP expression in the roots of the PIN-GFP reporter lines and the DR5rev::GFP line was observed by the confocal microscope after the fixation and clearing procedures using the TOMEI reagent (Hasegawa et al. 2016), following the manufacturer’s instruction for TOMEI-II method (Tokyo Chemical Industry). TOMEI-II is a clearing method suitable for fluorescent observation of inner roots because it scarcely decreases the fluorescence of reporter proteins due to its mild fixation. Briefly, plants exposed to 0.2 µM PheHg (1 DAT and 2 DAT) were fixed with 4% paraformaldehyde in PBS. The fixed plants were washed three times with PBS and stained with 1 µg/mL 4’,6-diamidino-2-phenylindole (DAPI) for 10 min, followed by another three-time rinsing with PBS. The roots were then mounted with 20% TOMEI reagent in PBS and incubated for 10 min to achieve tissue clearing. Z-stack images of roots were obtained, and the images were processed with Fluoview software (Olympus) to create maximum-intensity projection images except for pPIN2-PIN2-GFP. ROIs were set for respective reporter lines and signal intensities of GFP and DAPI were quantified using Image J software. GFP signal intensity was normalized by that of DAPI. The excitation and detection wavelengths were 405 nm and 430–470 nm for DAPI, and 488 nm and 500–530 nm for GFP.

FM4-64 internalization assay

FM4-64 (Thermo Scientific) was used as a plasma membrane/endocytic marker. pPIN2-PIN2-GFP plants treated with 0.2 µM PheHg (1 DAT) or grown on control plates were stained with 4 µM FM4-64 in one-tenth strength modified Hoagland medium solution. After 45 min incubation to induce internalization of FM4-64, epidermal cells in the root tip were observed by the confocal microscope. The excitation and detection wavelengths were 488 nm and 640–740 nm for FM4-64, and 488 nm and 500–530 nm for GFP.

Statistical analyses

The JMP Pro 17 software was used for statistical analyses. Two-tailed Exact Wilcoxon rank sum test (p < 0.05) was applied to assess statistical differences between the two groups, and one-way ANOVA, followed by the Steel-Dwass test (p < 0.05) was applied to compare statistical differences among the treatments.

Results

Root phenotypic differences under hg(II) and PheHg stress conditions

In the previous study, we macroscopically observed that PheHg treatment induced an abnormally enlarged/swollen root tip, and the swelling resulted in a shorter root length of cad1-3. However, the frequency of the root tip swelling was not examined quantitatively. We thus first quantified the root tip swelling frequency in Col-0 and cad1-3 grown under PheHg treatments (Fig. 1a). Under 0.08 µM PheHg, which was used in the previous study to differentiate Col-0 and cad1-3 root growth, the occurrence of the swollen root tip was 2-times higher in cad1-3 (approx. 60%) than in Col-0 (approx. 30%). Effects of the higher PheHg dose (0.2 µM) were next examined and the root swelling was induced under this condition in all the tested plants even in Col-0 wild-type (Fig. 1a).

Fig. 1.

Root responses to mercurial stress. (a) Frequency of root tip swelling under 0.08 µM and 0.2 µM PheHg conditions. Col-0 and cad1-3 plants were transferred to the PheHg plates and numbers of the plants showing the abnormally enlarged root tip were counted at 6 DAT for 0.08 µM and 4 DAT for 0.2 µM. The frequency was calculated per plate. Data are presented as box plots (center line at the median, upper bound at 75th percentile, lower bound at 25th percentile) with whiskers at minimum and maximum values. Each plot represents data from a single assay. Data were obtained from at least two independent experiments (n = 6–10). Asterisks indicate a significant difference between Col-0 and cad1-3 (*** P < 0.001, Two-tailed exact Wilcoxon rank sum test). n.s., not significant. (b, c) Root phenotypes of Col-0 and cad1-3 under 20 µM Hg(II) and 0.2 µM PheHg. 4-d-old plants were transferred to the mercury-containing plates and photographed at 4 DAT (b). Scale bars = 1 cm. The root length of each seedling was tracked and measured until 4 DAT using Image J and the plugin Smart Root (c). Data represent means with SD [n = 20 for control and Hg(II) and n = 10 for PheHg]

Under this PheHg condition, root response and elongation were compared to 20 µM Hg(II) after transfer to the mercury-containing medium (Fig. 1b-c). Under 0.2 µM PheHg, both Col-0 and cad1-3 exhibited the root swelling phenotype: the root tip looked thick, and the elongation was terminated (Fig. 1b-c). Under Hg(II) stress, the elongation of Col-0 roots was slanted to their right-hand side (to the left from our view) (Fig. 1b), suggesting the disruption of the gravitropism. In cad1-3, the loss of gravitropism was more remarkable (Fig. 1b), and the root elongation of cad1-3 was severely inhibited (Fig. 1c). These results demonstrate that both Hg(II) and PheHg terminate the root growth of cad1-3, however, there are differences in root morphology and responses between Hg(II) and PheHg treatments.

Time-dependent morphological changes under mercurial stress

Morphological characteristics of the root tips under Hg(II) and PheHg stresses were further examined by PI staining in Col-0 (Supplementary Figure S1) and cad1-3 (Fig. 2). In Col-0, corresponding with the control-like root growth under Hg(II) stress (Fig. 1c), little morphological changes were detected (Supplementary Figure S1). However, in cad1-3, time-dependent morphological changes were clearly observed (Fig. 2): epidermal cells of the transition zone started to expand slightly from 1 day-after-transfer (DAT) and the morphology of both the transition zone and meristematic zone was damaged at 4 DAT. Another notable symptom of Hg(II) toxicity in cad1-3 was sparsely observed dead cells that were completely stained with PI.

Fig. 2.

Time-dependent morphological changes of the root tips under mercurial stress. cad1-3 plants were transferred to 20 µM Hg(II) and 0.2 µM PheHg plates and the root tips were stained with PI and observed by a laser scanning confocal laser microscopy from 1 DAT to 4 DAT. Representative images for each time point and each condition were shown. Scale bars = 100 μm. The results of the same experiment conducted with Col-0 are shown in Supplementary Figure S1

More severe and drastic morphological changes were observed for PheHg-treated root tips of both Col-0 (Supplementary Figure S1) and cad1-3 (Fig. 2). Round-shaped epidermal cells were observed from 1 DAT around the transition zone and the distribution of cells with a similar round shape was gradually increased towards the root tip, regardless of the cell types. Consequently, radial expansion of the meristem by PheHg was much more severe than that induced by Hg(II), whereas the longitude meristem size was reduced by both mercurial. PI-positive dead cells were also sparsely observed from 1 DAT of PheHg treatment. These PheHg-induced phenotypes were similarly observed in both Col-0 and cad1-3.

Disappearance of starch granules in root caps under mercurial stress

Starch granules accumulated in the amyloplast of columella cells were visualized by Lugol staining in Col-0 (Supplementary Figure S2) and cad1-3 (Fig. 3) after being transferred to the respective mercurial-containing agar plates. Corresponding with the genotypic difference of the root growth inhibition (Fig. 1c), the Hg(II) treatment negligibly affected the starch accumulation and amyloplast distribution in Col-0 (Supplementary Figure S2). However, in cad1-3 exposed to Hg(II), starch granules gradually disappeared from 2 DAT to 4 DAT along with some radial expansion of the meristematic zone (Fig. 3). In contrast, the responses of root tips were similar between Col-0 and cad1-3 under PheHg treatment, and the PheHg-induced phenotypes were more severe than those observed under Hg(II) stress. At 2 DAT, the Lugol staining signal was weak in PheHg-exposed root tips of both Col-0 (Supplementary Figure S2) and cad1-3 (Fig. 3) and nearly completely disappeared at 3 DAT. Notably, the amyloplast disappearance was accompanied by remarkable radial expansion of the whole root meristem.

Fig. 3.

Time-dependent changes of starch granules in the root caps under mercurial stress. cad1-3 plants were transferred to 20 µM Hg(II) and 0.2 µM PheHg plates and the roots were subjected to Lugol staining and observed by a laser scanning confocal laser microscopy from 1 DAT to 4 DAT. Representative images for each time point and each condition were shown. Scale bars = 50 μm. The results of the same experiment conducted with Col-0 are shown in Supplementary Figure S2

Transcript responses of auxin-related genes under mercurial stress

We focused on a plant hormone auxin to explore possible mechanisms underlying root tip morphological changes and physiological responses induced by Hg(II) and PheHg stresses, because auxin regulates root growth and environmental responses. A set of genes involved in auxin biosynthesis, signaling, and transport was selected for shoot and root, respectively, and the transcript changes after 4 h and 24 h of Hg(II) and PheHg treatments were analyzed by quantitative RT-PCR in cad1-3 (Fig. 4). In the shoot, which is the main organ for auxin synthesis, most of the tested genes for auxin biosynthesis and signaling were upregulated at 4 h of the treatment but were on the contrary repressed at 24 h (Fig. 4a). PheHg treatment did not elicit such drastic changes to the gene expression in shoots.

Fig. 4.

Transcriptional responses of the auxin-related genes in shoot (a) and root (b) under mercurial stress. cad1-3 plants were transferred to 25 µM Hg(II) and 0.2 µM PheHg plates and harvested after 4 h and 24 h of the treatments. Transcripts of the auxin-related genes were quantified by quantitative RT-PCR. Data represent the log2 transformed fold change from each control condition, derived from four independent biological replicates

In the root, the effect of Hg(II) stress was opposite to that in the shoot. Most of the tested genes were down-regulated after 4 h of Hg(II) treatment but were up-regulated after 24 h (Fig. 4b). PheHg stress overall repressed the genes for auxin biosynthesis (TAA1 and YUCs), transport (PINs and LAXs), and signaling (IAAs) in roots even after 4 h of the exposure. The down-regulation tendency by PheHg stress was pronounced after 24 h for the most tested genes in roots. These results suggest that Hg(II) and PheHg stresses both disrupt the auxin homeostasis but with the different manners, and the effects of PheHg stress are more severe in the roots.

Expression of PIN-GFP and DR5rev::GFP under PheHg stress

Because the qPCR results suggested that PheHg treatment severely repressed the genes for auxin synthesis, transport, and signaling in roots, we further examined the expression of GFP reporter lines for PIN1, PIN2, PIN3, and DR5 (Fig. 5). Confocal imaging was conducted at 1 DAT and 2 DAT when the root tip swelling and morphological damages by PheHg were not so severe (Figs. 2 and 3, Supplementary Figures S1-S2), and confocal imaging of GFP was feasible. PIN1, PIN2, and PIN3 are major auxin transporters that mediate polar auxin transport in the root tip and the formation of auxin maxima around the stem cell niche (SCN). Expression of PIN-GFP driven by respective native promoters was decreased with a time-dependent manner after the exposure to PheHg, although the cell-type specificity of each PIN expression was not affected (Fig. 5a). Accordingly, the GFP signal driven by the DR5rev promoter was reduced at 2 DAT. In contrast, DR5rev::GFP signal was not decreased by Hg(II) treatment (Supplementary Figure S3).

Fig. 5.

Expression of the PIN-GFP and DR5rev::GFP (green) in the root tips under PheHg stress. Respective reporter plants were transferred to 0.2 µM PheHg plates and at 1 DAT and 2 DAT, the roots were observed by laser scanning confocal laser microscopy after fixation and DAPI staining (white). (a) Representative images for each time point and each condition were shown. Scale bars = 50 μm. (b) Relative GFP signal intensity. DAPI signal intensity within each ROI was used for normalization. Data are presented as box plots (center line at the median, upper bound at 75th percentile, lower bound at 25th percentile) with whiskers at minimum and maximum values. Each plot represents data from a single root. Data were obtained from at least two independent experiments (n = 4–6). Asterisks indicate a significant difference between the control and PheHg treatment (* P < 0.05, ** P < 0.01, Two-tailed exact Wilcoxon rank sum test). n.s., not significant

These observations were quantitatively assessed using the GFP signal intensity that was normalized by that of DAPI (Fig. 5b). The normalization was conducted to reduce the possible technical effect of root thickness induced by PheHg. The quantification results showed statistically significant repression of PIN-GFP at 1 DAT as well as 2 DAT, and a significant decrease of DR5rev::GFP signal at 2 DAT. Overall, the confocal imaging suggested that PIN transporters expression was repressed at the early stages of PheHg-induced root tip swelling, which resulted in a decrease of auxin maxima at SCN.

Polarity and internalization of PIN2-GFP under PheHg stress

We also examined the effects of PheHg stress on polar localization and internalization/degradation pathway for PINs using PIN2-GFP as a model (Fig. 6). PIN2 was selected for this experiment because it was specifically expressed in epidermis and cortical cells in roots, and thus the polar localization and membrane trafficking analyses of PIN2-GFP under PheHg stress conditions were more easily conducted compared to other PINs that expressed in inner and smaller cell types.

Fig. 6.

Polarity and internalization of PIN2-GFP (green) in the epidermal cells of the roots under PheHg stress. pPIN2-PIN2-GFP plants were transferred to 0.2 µM PheHg plates and at 1 DAT, the roots were observed by laser scanning confocal laser microscopy after 45 min of FM4-64 staining (magenta). Scale bars = 10 μm. White arrowheads indicate internalized GFP signal merged with FM4-64 signal

FM4-64 was used as a counterstaining for plasma membrane and as an endocytosis marker. Under the control condition, the PIN2-GFP signal was preferentially distributed to the apical side of the epidermal cells, and the internalized PIN2-GFP signal was also frequently observed (Fig. 6). These PIN2-GFP signals on the plasma membrane as well as internalized signals were well co-localized with FM4-64 signals, suggesting the substantial endocytosis activity in these cells and its involvement in PIN2 membrane trafficking. In contrast, after 1 DAT under the PheHg condition which significantly reduced the PIN2-GFP expression levels (Fig. 5b), the internalization of PIN2-GFP and FM4-64 was rarely detected, although the polar localization of PIN2-GFP to the apical side was maintained (Fig. 6). The results suggested that PheHg stress decreased endocytosis activity at least in the epidermal cells, and enhanced internalization/degradation was not a reason underlying the reduced expression of PIN2-GFP under PheHg stress conditions.

Effects of auxin supplementation on mercurial toxicity

Finally, we examined whether auxin supplementation could mitigate the toxicity of Hg(II) and PheHg (Fig. 7). IAA as a natural auxin was first tested. 20 nM IAA supplementation did not rescue the root growth inhibition by Hg(II) in both cad1-3 and Col-0 (Fig. 7a). In contrast, the same concentration of IAA supplementation almost canceled the root tip swelling, the typical PheHg toxicity in both Col-0 and cad1-3 (Fig. 7b). Then, we similarly examined the effects of the synthetic auxin analogs NAA and 2,4-D on PheHg-induced root tip swelling. The mitigation effects of these two auxin analogs were intermediate compared to that of IAA (Fig. 7b).

Fig. 7.

Effects of auxin supplementation on Hg(II) (a) and PheHg toxicity (b) to the roots. (a) Col-0 and cad1-3 plants were transferred to 20 µM Hg(II) containing plates or 20 µM Hg(II) and 20 nM IAA containing plates and the root length of each seedling was tracked and measured at 4 DAT using Image J and the plugin Smart Root. Data are presented as box plots (center line at the median, upper bound at 75th percentile, lower bound at 25th percentile) with whiskers at minimum and maximum values. Each plot represents data from a single root. Data were obtained from three independent assays (n = 21). Asterisks indicate a significant difference between Hg(II) and Hg(II) + IAA treatments (*** P < 0.001, Two-tailed exact Wilcoxon rank sum test). n.s., not significant. (b) Col-0 and cad1-3 plants were transferred to 0.2 µM PheHg containing plates or 0.2 µM PheHg and 20 nM IAA/NAA/2,4-D containing plates. Numbers of the plants showing the swollen root tip were counted at 6 DAT. Data are presented as box plots (center line at the median, upper bound at 75th percentile, lower bound at 25th percentile) with whiskers at minimum and maximum values. Each plot represents data from a single assay. Data were obtained from at least three independent experiments (n = 10–15). Data sharing the same letter are not significantly different (P < 0.05, Steel-Dwass test)

Discussion

PheHg was used as a pesticide worldwide in the middle of the 20th century, taking advantage of its strong anti-fungal activity. In Japan during the 1950–1960 s, the post-war period with high demand for stable rice production, PheHg-based pesticides were widely used in rice paddy to control rice blast (Ishiyama et al. 1965). The toxicity of PheHg was then recognized and its application was prohibited to avoid health risks, but PheHg contamination still exists in some soils (Hintelmann et al. 1995; Deonarine et al. 2015). Phytotoxicity of PheHg was also reported in the 1960–1970 s mainly focusing on its potential as a regulator/inhibitor of photosynthetic reactions. For instance, PheHg application induces stomatal closure and decreases photosynthesis (Zelitch and Waggoner 1962; Shimshi 1963). PheHg blocks electron flow from PS II, subsequent ferredoxin function at PS I, and ATP synthesis (Nozaki et al. 1961; Siegenthaler and Packer 1965; Honeycutt and Krogmann 1972). On the other hand, PheHg toxicity in roots and its detoxification mechanism have been little examined.

We recently reported that the metal-binding peptide PC machinery is essential for the detoxification of the organomercurial PheHg in Arabidopsis (Uraguchi et al. 2022b). The study demonstrates that AtPCS1-mediated PC synthesis and the subsequent vacuolar sequestration by AtABCC1 and AtABCC2 are both essential for the detoxification of PheHg and maintaining proper root growth under PheHg stress conditions. Since PC’s roles have been well established for the detoxification of Cd, As, and Hg of their inorganic forms, our finding provides a very rare example of PC-mediated detoxification of an organic form of toxic elements. There is only one more evidence demonstrating PC’s involvement in detoxification for an organic form of arsenic, monomethylarsenate (Tang et al. 2016).

Despite that the detoxification pathway is shared between PheHg and other inorganic metals including Hg(II), the root phenotypes elicited by Hg(II) and PheHg stress were different (Fig. 1b). The root tip swelling is a unique phenotype elicited by PheHg. The result suggests the existence of chemical form-specific metal toxicity in the roots. We aimed to further characterize the PheHg toxicity focusing on the root tip swelling, however, our previous PheHg stress condition (0.08 µM) was not suitable for the time-course analysis, because not all the plants (approximately 20–40% of Col-0 and 50–70% in cad1-3, Fig. 1a) exhibited the root tip swelling. We found that the higher PheHg treatment (0.2 µM) induced the root tip enlargement at the rate of 100% in both Col-0 and cad1-3 (Fig. 1a) and thus 0.2 µM was applied to the following experiments.

Under these conditions, microscopic observations detailed morphological differences between Hg(II) and PheHg toxicity (Fig. 2 and Supplementary Figure S1). Round-shaped cellular expansion of the epidermal cells at the upper meristem/transition zone was initially induced by both mercury stress, however, the morphological changes at 4 DAT were more drastic under PheHg conditions. It was difficult to conduct a standard imaging analysis of meristem cell numbers and length due to the severe tissue damage and morphological changes, but the longitude meristem size looks decreased by both mercurial stress (Fig. 2 and Supplementary Figure S1). The round-shaped expansion of the root tip cells is likely the cellular alteration responsible for the abnormal root tip swelling by PheHg.

We previously hypothesized that PheHg toxicity is induced by PheHg itself, not by degraded inorganic Hg from PheHg (Uraguchi et al. 2022b). This is supported by the presented results in this study that the macroscopic and microscopic phenotypes are distinguishable between Hg(II) and PheHg and much lower concentrations of PheHg [0.08 µM or 0.2 µM, compared to 20 µM Hg(II)] induced the root swelling. The PheHg-induced root damages are unique and different from the reported root phenotypes. For example, DNA-damaging stress often induces severe root damage in the respective related mutants but it can be characterized by enhanced root hair formation at the root tips along with tissue disorder and patchy dead cells (Sakamoto et al. 2011, 2019; Takahashi et al. 2019; Bisht et al. 2023), which are different from the PheHg-damaged roots (Fig. 2).

Disruption of the root gravitropism/abnormal root orientation was also suggested as a common mercury-induced root response, although there was again a difference between Hg(II) and PheHg (Fig. 1b). The Hg(II) stress enhanced a slanted root elongation to their right-hand side in Col-0 (leftward from our view), and the cad1-3 mutation enhanced the slanting tendency. Under 0.2 µM PheHg, a root tip was swollen, and its orientation looked ununiform and divergent. To examine mechanisms underlying the differences of the root tip responses, Lugol staining was conducted to observe starch granules in amyloplasts (Fig. 3 and Supplementary Figure S2). Amyloplasts are starch-accumulating plastids in the root cap columella cells. Sedimentation of amyloplasts is crucial for root gravity sensing and subsequent auxin flux control at the root tip, which are mediated by LAZY1-LIKE (LZY) family proteins (Taniguchi et al. 2017; Nishimura et al. 2023). Thus, the degradation/destarching of amyloplasts can lead to the loss of root gravitropism (Nakayama et al. 2012; Zheng et al. 2024), and it consequently affects auxin distribution within the root. Under the PheHg stress condition, the root tip swelling was accompanied by the complete loss of starch granules in the columella cells of both Col-0 and cad1-3 (Fig. 3 and Supplementary Figure S2), suggesting a possibility that auxin homeostasis is disrupted under PheHg stress conditions. Hg(II) stress also resulted in the absence of starch granules at 3 DAT in cad1-3 (Fig. 3), which exhibited severe root slanting (Fig. 1b). However, in Col-0, the Lugol staining signal was not changed by Hg(II) treatment (Supplementary Figure S2), although the root was moderately slanted (Fig. 1b). For this case, enhanced root twisting (torsion) (Migliaccio et al. 2009) may be one possibility in addition to or rather than gravitropism disruption. Microtubule reorientation causes root twisting in the case of halotropism (Yu et al. 2022). The possibility of root twisting should be further examined to characterize Hg(II) toxicity mechanisms.

To further explore mechanisms underlying the PheHg-induced root swelling, transcript responses of the auxin-related genes were analyzed in cad1-3. The results show the unique transcript profiles for Hg(II) and PheHg treatments, respectively (Fig. 4), highlighting the differences between Hg(II) and PheHg toxicity at the molecular level. For PheHg, auxin synthesis genes TAA1 and YUCs, and auxin signaling related genes IAAs in the roots were remarkably down-regulated after 24 h of the treatment. This is the time point when the morphological changes are not so severe (Figs. 2 and 3), demonstrating that the disruption of the auxin homeostasis at the transcript level occurs ahead of the root tip enlargement. It should be also noted that the effects of Hg(II) treatment were not time-dependent, and the responses were opposite between roots and shoots. The up-regulation in shoots (4 h) and in roots (24 h) would be compensative responses to the down-regulation after 4 h, which is likely to be elicited by the Hg(II) toxicity. These responses were unique to Hg(II), suggesting the specificity of the PheHg and Hg(II) toxicity at the transcriptional level.

Compared to TAA1, YUCs, and IAAs, the decrease of PIN transcript by the PheHg stress was not so drastic or clear (Fig. 4b). Because post-translational modification and subcellular membrane trafficking systems play key roles in PIN regulations (Zhang et al. 2010; Leitner et al. 2012), we analyzed the PIN-GFP expression levels in the PheHg treated roots. In contrast to the slight transcript changes, GFP signals of PIN1, PIN2, and PIN3 fusions were significantly decreased from 1 DAT of the PheHg treatment (Fig. 5). PINs are auxin efflux carriers and characterized by the cell type-specific expression at the tissue level, and polar localization at the subcellular level. Both are important for polar auxin transport in the roots. Our results demonstrate that PheHg stress decreases the PIN expression levels but does not affect the cell type-specificity of each PIN in the root tips. We also examined the effect of PheHg stress on the polarity and internalization of PINs using PIN2 as the model (Fig. 6). Clathrin-mediated endocytosis plays a key role in PIN’s recycling and thereby its polarity formation (Kitakura et al. 2011). However, surprisingly, our results show that PheHg stress does not affect shootward PIN2 polarity in the epidermis (Wiśniewska et al. 2006), while endocytosis is almost inhibited by PheHg stress. Mechanisms underlying this phenomenon remain unknown, but at least, our assay suggests that the reduced expression of the PINs is not attributed to enhanced endocytosis and degradation pathways.

Each PIN coordinately fine-tunes polar auxin flux within the root (Blilou et al. 2005; Wiśniewska et al. 2006) to form auxin maxima around the SCN. Following the PIN-GFP depression from 1 DAT, an auxin marker DR5-driven GFP signal was decreased by PheHg stress at 2 DAT (Fig. 5). The reduced PIN1-GFP expression and the consequent disruption of polar auxin transport from the shoot towards the root tip would be one reason for the decrease of auxin level at the SCN. Moreover, de novo auxin synthesis in the root tip is likely to be inhibited by PheHg, as suggested by the transcript analysis showing the overall repression of TAA1 and YUCs (Fig. 4b). TAA1 and YUCs are the first and second step enzymes, respectively, for the IPA pathway, a major pathway for IAA production (Mashiguchi et al. 2011). Our results suggest that PheHg stress inhibits polar auxin transport to the root tip as well as de novo auxin synthesis through the IPA pathway, and these phenomena occur at the early phase of the root tip enlargement. Interestingly, CYP79B2 was upregulated after 24 h of the PheHg treatment in roots (Fig. 4B). CYP79B2 and CYP79B3 are suggested as the first-step enzymes responsible for IAA production through the IAOx pathway, which is independent of the IPA pathway with TAA1 and YUCs (Sugawara et al. 2009; Kasahara 2016). The upregulation of CYP79B2 may be a compensative response against the diminished activity of the IPA pathway under PheHg stress, but even if that is the case, the response is likely insufficient to maintain the root tip homeostasis disrupted by PheHg.

To examine whether the auxin disruption/decline at the root tip is a cause of the PheHg-induced root tip swelling, we conducted the auxin supplementation assay (Fig. 7). The IAA supplement almost canceled the PheHg toxicity, whereas the effects of the two synthetic auxin analogs NAA and 2,4-D were intermediate (Fig. 7b). IAA is a natural auxin which is transported by AUX1 and its homologous influx carriers as well as the efflux carrier PINs, whereas NAA is a specific substrate for PINs, and 2,4-D is preferentially recognized by the influx carriers rather than PINs (Delbarre et al. 1996; Petrášek et al. 2006; Yang et al. 2006; Simon et al. 2013). The different effects of IAA, NAA, and 2,4-D on the PheHg-induced root swelling (Fig. 7) suggest that both influx and efflux auxin transport activities are diminished by PheHg stress, and the auxin imbalance/decline in the root tip is a major cause of the abnormal root tip swelling under PheHg stress. Hg(II) stress did not decrease the root tip DR5rev::GFP level (Supplementary Figure S3), and reasonably, no complementation effect was observed for IAA supplementation under the Hg(II) stress condition (Fig. 7a). These results support our hypothesis again that PheHg toxicity is physiologically different from that of Hg(II) and is not attributed to the degradation of PheHg.

In conclusion, the present study differentiates PheHg toxicity and Hg(II) toxicity, especially focusing on the root phenotypes. PC synthesis is crucial for the detoxification of both PheHg and Hg(II) as previously reported, however, the root response and morphological changes are different between the two mercurial stresses. The abnormal root tip swelling is a unique phenotype induced by PheHg stress, and it is accompanied by the auxin homeostasis disorder in the root tip (Fig. 8). We propose that polar auxin transport as well as de novo auxin synthesis through the IPA pathway are major targets of the PheHg toxicity. Further studies are needed to dissect the molecular target of the PheHg toxicity, for instance, using loss-of-function mutants of the respective enzymes. In vitro assay using TAA1 and YUC recombinant proteins would be also needed to directly evaluate a possible inhibitory effect of PheHg. The IAA supplementation at the physiological concentration (20 nM) rescues the root abnormality under PheHg stress, suggesting a potential of IAA or its analogs for conferring the PheHg stress tolerance to plants. Application of IAA or its analogs may also enhance metal tolerance of the plants against such as cadmium and arsenic which causes agricultural problems worldwide. On the other hand, Hg(II) toxicity in roots is less likely to target auxin homeostasis. Thus, its toxicity mechanism on the root should be further examined. Possible effects of Hg(II) and PheHg on root gravitropism indicated by our assay should be also investigated further.

Fig. 8.

Model for PheHg toxicity disrupting auxin homeostasis and root elongation. PheHg stress decreases PINs expression in the root and thereby inhibits polar auxin transport toward the root tip. de novo auxin synthesis is also a possible target of PheHg toxicity, decreasing the transcript of the auxin synthesis genes in roots. These two adverse effects result in the decrease of auxin level in the root tip, leading to abnormal root tip enlargement/swelling and root growth inhibition. IAA supplementation rescues the PheHg toxicity, supporting the hypothesis that the collapse of auxin homeostasis in the root tip is a trigger for the root tip swelling. PheHg toxicity is enhanced in the AtPCS1 mutant cad1-3, demonstrating that phytochelatins play a crucial role in maintaining auxin homeostasis through chelation of cytosolic PheHg for vacuolar sequestration as previously reported by Uraguchi et al. (2022b)

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

SU and MK designed the experiments. SU, MS, CH, MH, KO, MO, HS, and AO conducted the experiments. SU, MS, CH, YO, RN, YT, and MK analyzed the data. SU wrote the manuscript, and all authors approved the manuscript.

Funding

This work was supported in part by the Japan Society for the Promotion of Science (Grant nos. 18K05377, 21K05330, 24K08649 to S.U.).

Declarations

Conflict of interest

The authors declare no conflicts of interest.