Phospholipase Dδ assists to cortical microtubule recovery after salt stress
The dynamic microtubule cytoskeleton plays fundamental roles in the growth and development of plants including regulation of their responses to environmental stress. Plants exposed to hyper-osmotic stress commonly acclimate, acquiring tolerance to variable stress levels. The underlying cellular mechanisms are largely unknown. Here, we show, for the first time, by in vivo imaging approach that linear patterns of phospholipase Dδ match the localization of microtubules in various biological systems, validating previously predicted connection between phospholipase Dδ and microtubules. Both the microtubule and linear phospholipase Dδ structures were disintegrated in a few minutes after treatment with oryzalin or salt. Moreover, by using immunofluorescence confocal microscopy of the cells in the root elongation zone of Arabidopsis, we have shown that the cortical microtubules rapidly depolymerized within 30 min of treatment with 150 or 200 mM NaCl. Within 5 h of treatment, the density of microtubule arrays was partially restored. A T-DNA insertional mutant lacking phospholipase Dδ showed poor recovery of microtubule arrays following salt exposition. The restoration of microtubules was significantly retarded as well as the rate of root growth, but roots of overexpressor GFP-PLDδ prepared in our lab, have grown slightly better compared to wild-type plants. Our results indicate that phospholipase Dδ is involved in salt stress tolerance, possibly by direct anchoring and stabilization of de novo emerging microtubules to the plasma membrane, providing novel insight into common molecular mechanism during various stress events.
KeywordsPhospholipase Dδ Microtubule dynamics Salt stress Arabidopsis roots BY-2
Salinity stress is a major environmental issue worldwide, affecting most irrigated agricultural land. Excess cytoplasmic Na+ is toxic to cells because it interferes with K+ uptake and nutrition. To alleviate the effects of salt stress, plants use a variety signaling mechanisms that lead to ionic and osmotic homeostasis, regulation of cell division and expansion, and detoxification signaling for minimizing and repairing the damage (Golldack et al. 2014).
Besides biochemical processes, plants can also exhibit morphological changes to their roots. Growth rates of primary root are reduced in response to mild salt stress. Changes in cell growth rate are preceded by an alteration of microtubule dynamics, among other factors (Wang et al. 2007; West et al. 2004). Interestingly, Arabidopsis seedlings withstand salt stress better when microtubules (MTs) were depolymerized and actin filaments more polymerized during the early stages of salt stress, whereas the converse cytoskeletal situation resulted in a higher rate of seedling death (Wang et al. 2007; Wang et al. 2010).
Phospholipid signaling is one of the molecular mechanisms involved in salt stress tolerance. The important player of this signaling besides phospholipase A2 and phospholipase C is phospholipase D (PLD) which hydrolyzes membrane phospholipids to produce a free head-group and phosphatidic acid (PA), an important second messenger in plants. From the 12 subgroups of Arabidopsis PLDs, PLD α1, α3, and δ are involved in salt stress signaling (Hong et al. 2016). From those, PLDδ can directly interact with cytoskeletal components. In a tubulin pull-down assay, a membrane-associated protein was purified from tobacco as a potential microtubule-plasma membrane linking protein (Marc et al. 1996) and an antibody raised against this protein reacted with PLDδ in an Arabidopsis cDNA library screen (Gardiner et al. 2001). Arabidopsis PLDδ used directly in a pull-down assay bound to β-tubulin, actin 7, flotillin homolog, and other proteins (Ho et al. 2009).
It was shown that PA can directly interact with MAP65-1 (microtubule-associated protein 65-1) and thus enhance polymerization and bundling of microtubules under salt stress conditions (Zhang et al. 2012). Microtubule stability is probably also regulated by degradation of certain MAPs. For example, SPR1 protein, known to stabilize microtubules, is rapidly degraded under salt stress, and this in turn allows fast reorganization of cortical microtubules (Wang et al. 2011). Interestingly, many environmental factors including salt, wounding, cold, fungal infection, heat stress, and osmotic stress that affect PLD activity also affect microtubule organization (Hong et al. 2016, Zhang et al. 2017). The enhanced fitness of stressed Arabidopsis plants depends on both increased expression (activation) or inhibition of PLDs and it differs in this regard according to the type of stress situation. Interestingly, also subsequent microtubule response differs substantially depending on the stress conditions (Hong et al. 2016, Zhang et al. 2017).
We hypothesized that the ability of plants to tolerate salt stress relies on the alignment of cortical microtubule arrays and their association with the plasma membrane by links with microtubule-associated proteins (MAPs) such as PLDδ (Dixit and Cyr 2004).
In the present study, we show for the first time that PLDδ decorates microtubule structures in vivo. Further, we explored the effects of NaCl stress during the very early stages of microtubule organization. The results show that high salt concentrations induce transient microtubule depolymerization, followed by restoration of cortical microtubules and maintenance of root growth under high salt stress. Exposure of the null Atpldδ mutant to salt stress showed that the lack of PLDδ hindered the restoration of the cortical microtubule array as well as root growth, suggesting that PLDδ, rather than its product PA, is involved in salt stress tolerance.
Materials and methods
Plant material and treatment
Live imaging experiments employed 3-day-old suspension cultures of tobacco BY-2 cells (Nicotiana tabaccum L cv. Bright-Yellow 2) (Nagata et al. 1992). One tenth of the final volume of BY-2 cells (inoculum) was weekly subcultured into the fresh, sterile liquid medium (4,3 g/L Murashige and Skoog salts, 30 g/L sucrose, 100 mg/L inositol, 1 mg/L thiamin, 0,2 mg/L 2,4-dichlorophenoxyacetic acid, and 200 mg/L KH2PO4), pH = 5.8. Cells were cultivated in darkness at 26 °C on an orbital incubator (120 rpm). Co-localization experiments were performed on 9-day-old Arabidopsis thaliana (ecotype Columbia-0) plantlets transiently expressing microtubular marker RFP-MBD on the background of plants stably expressing GFP-AtPLDδ. In the experiments showing PLDδ patterns, localization, root growth after salt stress, and latrunculin treatment 5- to 12-day-old seedlings of Arabidopsis thaliana (ecotype Columbia-0) stably transformed with 35S::GFP-PLDδ construct and Arabidopsis thaliana (ecotype Columbia-0) expressing GFP-fABD2 (Voight et al. 2005) were used.
Five-day-old seedlings of wild-type (WT) Arabidopsis thaliana (ecotype Columbia-0) and the T-DNA insertional mutant Atpldδ (SALK_023247) were used in experiments investigating microtubule recovery after salt stress and root growth. Homozygous lines of T-DNA mutant Atpldδ were checked by PCR genotyping, specific primers were used according to Pinosa, who also verified lack of functional transcript in this line (Pinosa et al. 2013).
Seeds of all used lines of Arabidopsis thaliana were sterilized with 50% ethanol for 1–5 min, followed by 2.5% sodium hypochlorite with 0.1% Triton X-100 for 10 min, and washed four times in sterile Milli-Q water. Seeds were then sown under sterile conditions in Petri dishes containing nutrient medium [4.4 g/L basal salts (Murashige and Skoog 1962) supplemented with 1% (w/v) sucrose, pH to 5.7] solidified with 1.2% (w/v) agar. Seeds were imbibed at 4 °C for 48 h to ensure synchronous germination. Seedlings were grown on vertically oriented plates in a controlled growth chamber at constant light (80–100 μmol m−2 s−1) and air temperature (25 ± 1 °C).
Specified concentrations of NaCl were added to the nutrient medium before autoclaving. Arabidopsis seedlings in the experiment studying root growth were grown for 5 days on normal agar medium and then transferred to the medium containing salt using watchmaker’s forceps under sterile conditions. All chemicals were obtained from Sigma-Aldrich, USA. Stock solution of oryzalin [4-(dipropylamino)-3,5-dinitrobenzenesulfonamide] (Sigma-Aldrich, USA) was diluted in growth medium to 20 μM concentration, cytoskeletal drug latrunculin B (Sigma-Aldrich, USA), and membrane specific dye FM 4-64 (ThermoFisher, Scientific) were solved in DMSO to prepare 1000× concentrated stock solution which was then diluted by growth medium to final concentrations and applied immediately on slides with the BY-2 cells or Arabidopsis thaliana seedlings.
Preparation of plant material expressing GFP-PLDδ and RFP-MBD
Tobacco BY-2 cells were transiently transformed according to the protocol recommended and supplied with Particle Delivery System, Helios 1000/He (Bio-Rad, USA). Gold particles 1.6 μm diameter (Bio-Rad laboratories, Hercules, CA, USA) were coated with Gateway vector p2GWF7 (35S::AtPLDδ-GFP construct), p2FGW7 (35S::GFP-AtPLDδ construct), or binary vector pK7WGF2 containing cDNA of (35S::GFP-AtPLDδ construct) (Andreeva et al. 2009). After 20–24-h transformation, the expression of all constructs was observed.
Stable transformation of Arabidopsis thaliana, Col-0, wild type with binary vector pK7WGF2 (Ho et al. 2009) containing 35S::GFP-PLDδ construct were performed according to Clough and Bent (1998) using Agrobacterium tumefaciens strain C58C1.
For co-localization experiments, Arabidopsis plants stably expressing AtGFP-PLDδ were transiently transformed with A. tumefaciens strain C58C1 containing RFP tagged microtubular marker - RFP-MBD (microtubule-binding domain) inserted into the binary vector pGreen0029 (Hellens et al. 2000). Transformation procedure was performed according to Li et al. (2009) with the following modifications. Optical densities of A. tumefaciens according to OD600 reading were 0.1, 0.06, and 0.04. The expression of RFP-MBD was monitored in 24-h intervals up to 96 h. cDNA of HsMBD (Marc et al. 1998) was taken from p2RGW7 Gateway vector and later transferred into pGreen0029 binary vector with RFP reporter gene, downstream from the 35S promotor. All used Gateway vectors were provided by Prof. J. Marc, Sydney (personal communication).
Arabidopsis roots were processed for immunofluorescence microscopy using a modified method of Sugimoto et al. (2000). Whole seedlings were fixed for 40 min with 1.5% (v/v) paraformaldehyde and 0.5% glutaraldehyde in PMET buffer [50 mM PIPES, 2 mM MgSO4, 5 mM EGTA, 0.05% (v/v) triton X-100, pH 7.0] and rinsed in PMET buffer three times for 10 min. The roots of the seedlings were then digested for 15 min with 0.01% pectolyase and 0.1% pectinase in phosphate-buffered saline (PBS) [10 mM Na2HPO4, 2 mM KH2PO4, 137 mM NaCl, 2.7 mM KCl, 0.02% (w/v) NaN3, pH 7.2]. Seedlings were subsequently rinsed three times for 5 min with PME buffer [50 mM PIPES, 2 mM MgSO4, 5 mM EGTA, pH 7.0] containing 10% glycerol, extracted with − 20 °C methanol for 10 min, and rehydrated in PBS for 10 min. Seedlings were then incubated overnight at 37 °C with a mouse monoclonal antibody (class IgM) against α-tubulin at 1:1000 dilution, then washed three times for 5 min in PBS, and incubated with a sheep-anti-mouse IgG antibody conjugated with fluorescein isothiocyanate (FITC) at 1:30 dilution. After rinsing three times in PBS, the seedlings were mounted on multi-well slides (ICN Biomedicals, Cleveland OH, USA) in PBS with 20% glycerol (Agar Scientific, Stansted, Essex, UK). Cover-slips were sealed with clear nail polish.
Confocal laser scanning microscopy
Images for immunolocalization experiments and live imaging of BY-2 cells were acquired by confocal laser scanning microscope Zeiss LSM 5 DUO (Jena, Germany). Excitation at 488 nm with an argon ion laser was used for FITC fluorochrome. Optical sections of the epidermal layer in the root elongation zone were collected with a 63x Zeiss LCI Plan-Neofluar/Imm Korr/DIC (NA 1.30). Image projections for immunolocalization experiments were processed in Zeiss LSM Image Browser from at least 15 optical sections, 0.4 μm thick in the Z-axis for each Z-stack.
All of the final images were processed with Adobe Photoshop 7.0 (Adobe Systems, Mountain View, CA, USA) and Inkscape (Free Software Foundation, Inc., Boston, USA). In the microtubule density experiment the images of 6–8 roots were examined for each treatment.
Live images of Arabidopsis seedlings were obtained by confocal laser scanning microscope Zeiss 880 (Jena, Germany). Optical sections of cotyledon pavement cells were acquired with 63x Plan-Apochromat, Oil DIC M27 (NA 1.40) objective. Laser excitation 488 and 561 nm, band pass emission filter 499–540 and 579–633 nm respectively, were used for GFP and RFP visualization. Images were processed with ZEN Lite software (Zeiss, Jena, Germany) and Inkscape (Free Software Foundation, Inc., Boston, USA).
To determine the density of cortical microtubules per unit cell length, the number of microtubules crossing 20-μm longitudinal transects were counted in each of five cells in each of six replicate roots, for each time interval and two root regions (between 300 and 420 and 540–660 μm from the root apex). Number of microtubules per length unit was determined by manual counting in particular area of all photos in each Z-stack. Data obtained were analyzed for statistical significance using ANOVA.
Root growth was quantified by marking the position of the root tip on the back of vertical Petri dishes each day during the experiment. The marked Petri dishes were then scanned, and the distances from the hypocotyl to the root tip were measured with ImageJ 1.30x software (Wayne Rasband, National Institutes of Health, USA; http://rsb.info.nih.gov/ij). Fifteen roots were measured in each treatment for each time interval in two independent experiments. Time dependence of square root transformed root length were fitted by linear mixed-effects using “lmer” function from package “lme4” in R (RCoreTeam 2013), factors “plant identity” and “replication” were used as source of random variability, and factors “genotype variant” and “treatment type” as fixed effects. Significances of tested effects and their interactions were determined using “anova” function from package “lmerTest.”
Fluorescently labeled PLDδ decorates microtubules
The localization of GFP-PLDδ was also observed in response to the application of microtubule-disrupting agent, oryzalin. Ten minutes after treatment with 20 μM oryzalin, linear PLDδ structures were partially disintegrated, and after 15 min, completely disappeared (Fig. 1e, f).
Disruption of microtubules after salt stress is well documented (Zhang et al. 2012), so it was reasonable to investigate whether GFP-PLDδ linear structures would react in a similar way. Therefore, we treated plantlets expressing co-localized PLDδ and MT with 200 mM NaCl. Both linear GFP-PLDδ- and RFP-tagged microtubules were partially disintegrated after 7 min (Fig. 3d–f) and almost completely undetectable at 9 min (Fig. 3g–i). Disorganization of GFP-PLDδ structures after microtubule depolymerization is here in an agreement with oryzalin effect as described above and indicates a tight association of PLDδ with microtubules. We had therefore applied actin polymerization inhibitor latrunculin B to 12-day-old GFP-PLDδ expressing Arabidopsis seedlings, to verify specificity of this effect. Treatment with 200 nM latrunculin B had no observable effect on the stability of PLDδ linear patterns monitored over 30 min in hypocotyl cells (Fig. S2a-c).
Arabidopsis phospholipase Dδ mutant is impaired in the ability to restore microtubule density under salt stress
Whereas the recovery over the 2–5-h interval in 150 mM NaCl-treated WT reached 0.61 microtubules μm−1 on average, the recovery in the mutant was hindered and reached only 0.46 microtubules μm−1 over the same period (Fig. 4a). Exposure to 200 mM NaCl led to a slower recovery compared to 150 mM NaCl, although the recovery in the mutant was again significantly hindered (Fig. 4b). Thus, whereas the recovery in the WT over the 2–5-h interval reached on average 0.42 microtubules μm−1, the retarded recovery in the mutant reached only 0.22 microtubules μm−1 over the same period.
Taken together, we have observed an important and a previously undescribed role for PLDδ in MT restoration upon salt stress, which could have an important impact on downstream physiological responses in the cell.
Inhibition of root elongation is salt concentration-dependent and is more pronounced in Arabidopsis pldδ insertional mutant
Indications of a link between PLDδ and parts of microtubular cytoskeleton have been previously described (Andreeva et al. 2009; Ho et al. 2009), but evidence of a direct physical link between PLDδ and microtubules is lacking.
Using live imaging and confocal microscopy, we have observed co-localization between microtubules and PLDδ linear structures. Linear pattern of GFP-PLDδ alone or emerging from dotted GFP-PLDδ background was observed in several independent plant models and different cell types. In Arabidopsis seedlings, we have observed the PLDδ linear structures in leaf epidermal cells, cotyledons, hypocotyl, and root cells (Fig. 2). To assess whether our results are applicable to different, independent plant models, we examined tobacco BY-2 cells transiently expressing GFP-AtPLDδ. We observed similar structures to those in Arabidopsis. Importantly, we confirmed that those GFP-PLDδ linear structures co-localized with RFP-tagged microtubular marker - RFP-MBD in cotyledon pavement cells co-expressing both proteins.
Linear, microtubule-like localization of PLDδ was observed using N-terminally tagged constructs (GFP-PLDδ). A similar construct tagged in the C-terminal displayed cytosolic localization, similar to untagged GFP, suggesting that tagging of PLDδ at the C-terminal interferes with its insertion into the plasma membrane, a localization that is supported by various studies, including biochemical fractionation (Andreeva et al. 2009; Wang and Wang 2001). The observed localization of PLDδ was also influenced by developmental factors. The linear pattern of PLDδ was observed more often in young plants. Cells undergoing elongation, namely root and hypocotyl cells, also showed linear pattern more frequently. However, we have to point out transient character of PLDδ line patterns. PLDδ pattern is comprised of many dynamic dots which assemble into lines on transient basis. Pattern of stable lines is relatively less common phenomenon.
Proper microtubule spatial organization, specifically transversal arrangement, facilitate cell elongation, or anisotropic growth (Granger and Cyr 2001). This suggests that perhaps PLDδ may assist in microtubular reorganization during the early developmental stages of cells.
Interestingly, linear GFP-PLDδ patterns were disrupted within a few minutes upon application of oryzalin, a microtubule-disrupting drug (Fig. 1d–f), yet is unchanged upon application of latrunculin B, an actin disrupting drug, which provides an evidence of a close spatial relationship between PLDδ and microtubules. If PLDδ operates putatively like a scaffold, stabilizing microtubules to the plasma membrane (Gardiner et al. 2001), then PLDδ domains remain aligned alongside microtubules. When microtubules suddenly disappear, such as during depolymerization, PLDδ proteins aligned to microtubules would also be released, and thereby move randomly, causing a disruption of the observed linear structures of PLDδ. A similar situation would be expected for both PLDδ linear structures and microtubules in response to salt treatment.
Confocal images demonstrated tight co-localization of GFP-PLDδ line structures with microtubules as well as degradation of linear PLDδ structures after treatment, with high concentrations of salt. Involvement of either PLDδ or microtubules during salt stress response has been demonstrated separately by numerous studies (Zhang et al. 2012). We have observed the simultaneous degradation of both linear PLDδ structures and microtubules in response to treatment with high concentration of salt. Hence, we conclude there is a close physical and possibly direct interaction of microtubules and PLDδ, as hypothesized by Marc and Gardiner (Gardiner et al. 2001; Marc et al. 1996). The coordinated timing and relocalization of PLDδ and MT we observed in response to treatment with either oryzalin or salt is consistent with such a relationship.
Next, we considered and tested functional relevance for the proposed link between PLDδ and microtubules under the assumption of the aforementioned premise. Attachment of microtubules to the plasma membrane is important for microtubule stabilization and spatial organization. The AtPLDδ mutant is apparently impaired in the ability to restore microtubule density under salt stress. This behavior is consistent with the finding that PLDδ is markedly upregulated in response to dehydration and salt stress (Katagiri et al. 2001), and is therefore consistent with the proposed plasma membrane—PLDδ—microtubule interconnection (Gardiner et al. 2001) as well as with the known mechanism of microtubule stabilization by structural MAPs (Cai 2010; Dixit and Cyr 2004; Wang et al. 2011). MAP65-1 increases microtubule polymerization and bundling during salt stress after phosphatidic acid binding. These processes are initiated by enhanced activity of PLDα1 and δ and lead to the promotion of salt tolerance in Arabidopsis plants, but PLDα1 itself has no direct connection to microtubules (Zhang et al. 2012). In addition, a double mutant of Arabidopsis pldα1 and pldδ is more sensitive to salt concerning reduced root growth, compared to single mutants (Bargmann et al. 2009).
Along with decreased recovery of microtubules under salt stress, we also consistently observed a statistically significant deceleration of root growth in Atpldδ mutants in comparison to wild type (Figs. 5 and S3). This effect has been previously described (Bargmann et al. 2009), and in our study, we additionally showed the time and concentration dependency of this effect and also slightly better root growth of plants overexpressing GFP-PLDδ after treatment with 150 mM NaCl. Retarded growth rate in the mutant within 6 days following 200 mM NaCl treatment implies a significant role for PLDδ during salt stress tolerance. Although there is some redundancy between PLD isoforms concerning PA production (Johansson et al. 2014), such a long-term inability of mutant plants to overcome salt stress in the same extent as wild-type plants supports the idea of some unique feature of PLDδ, which is not completely complemented by other PLD isoforms, despite their increased activity following salt treatment (Bargmann et al. 2009). It suggests that the mechanism that involves PLDδ is independent of PA production or any following downstream responses upon salt stress (such as MAP65-1-related mechanism). Instead, this predicted mechanism may be related to PLDδ physical or scaffolding function in linking microtubules to the plasma membrane (Dhonukshe et al. 2003).
Assuming that PLDδ is implicated in the stabilization of microtubules under stress situations, it is of note that co-occurrence of both PLDδ and accumulation of microtubules under the same conditions has been reported in other processes, and the same underlying mechanism may be operating in these cases. For example, Blumeria graminis infection of Arabidopsis epidermal cells was reported to be accompanied by a higher accumulation of specifically the δ isoform of PLD, whereas microscopic analysis revealed increased accumulation of PLDδ beneath the B. graminis penetration site (Pinosa et al. 2013). The targeted accumulation of microtubules under the same conditions is also well documented (Huesmann et al. 2012). It may imply the importance of close localization of PLDδ for targeted de novo microtubule reorganization as a general molecular mechanism in response to different stress signals.
From all PLD isoforms, only PLDδ is predicted to bind to microtubules so far (Gardiner et al. 2001). We therefore propose stabilization of microtubules through direct connection of PLDδ to microtubules as being possible accessory mechanism for microtubule stabilization and improved salt tolerance and possibly more generally, also in other stress processes, where microtubular spatial organization is affected.
We thank Jan Petrášek for microscopy and further guidance and to Jiří Kubásek for the help with data processing.
This work was supported by the Czech Science Foundation Grant No. 14-09685S and by the Ministry of Education, Youth and Sports of the Czech Republic, projects NPU I, LO1417, and LM2015062. Microscope Zeiss 880 and LSM5DUO: IEB Imaging Facility is supported by OPPK—Operational Program Prague Competitiveness CZ.2.16/3.1.00/21519.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
- Granger CL, Cyr RJ (2001) Spatiotemporal relationships between growth and microtubule orientation as revealed in living root cells of Arabidopsis thaliana transformed with green-fluorescent-protein gene construct GFP-MBD. Protoplasma 216(3-4):201–214. https://doi.org/10.1007/BF02673872 CrossRefPubMedGoogle Scholar
- Johansson ON, Fahlberg P, Karimi E, Nilsson AK, Ellerstrom M, Andersson MX (2014) Redundancy among phospholipase D isoforms in resistance triggered by recognition of the Pseudomonas syringae effector AvrRpm1 in Arabidopsis thaliana. Front Plant Sci 5(639). https://doi.org/10.3389/fpls.2014.00639
- Li JF, Park E, von Arnim AG, and Nebenführ A (2009) The FAST technique: a simplified Agrobacteriumbased transformation method for transient gene expression analysis in seedlings of Arabidopsis and other plant species. Plant Methods 5:6–21Google Scholar
- Murashige T, Skoog F (1962) A revised medium for rapid growth and bio assays with tobacco tissue cultures. Physiol Plant 15(3):473–497. https://doi.org/10.1111/j.1399-3054.1962.tb08052.x CrossRefGoogle Scholar
- RCoreTeam (2013) A language and environment for statistical computing. R Foundation for Statistical Computing, ViennaGoogle Scholar
- Voight B, Timmers ACJ, Šamaj J, Müller J, Baluška F, Menzel D (2005) GFP-FABD2 fusion construct allows in vivo visualization of the dynamic actin cytoskeleton. Eur J Cell Biol 84:494–507Google Scholar