Introduction

The sessile nature of plants often exposes them to unfavorable environmental conditions in the forms of biotic and abiotic stresses. To resist these threats, plants have evolved specific defensive strategies to contribute to their diversification and fitness to constantly changing environments (Zhu 2016; Yang and Guo 2018; Gong et al. 2020; Karlova et al. 2021; Zhao et al. 2021; Gul et al. 2022; Kohli et al. 2022; Kumar et al. 2023; Leisner et al. 2023; Xiao and Zhou 2023; Ibeas et al. 2024). Root hairs are cellular extensions of root epidermal cells found on nearly all vascular plants, and are considered to be important sensors and defense structures in plants (Salazar-Henao et al. 2016; Arif et al. 2019; Karlova et al. 2021; Kohli et al. 2022; Liu et al. 2023; Qian et al. 2024). The presence of root hairs extends the surface area of the root system, playing crucial roles in anchorage, storage, and mineral and water acquisition (Ruiz et al. 2020; Karlova et al. 2021; Kohli et al. 2022). In addition, root hairs are also beneficial for shaping soil structure, enhancing soil health and promoting microbial diversity (Koebernick et al. 2017; Robertson-Albertyn et al. 2017). Moreover, root hair traits like density and length display high developmental plasticity in response to different environments, which affects water and nutrient uptake and in turn overall plant growth (Ruiz et al. 2020; Karlova et al. 2021; Kohli et al. 2022; Ibeas et al. 2024). Therefore, elucidating the regulatory mechanisms associated with root hair development and stress response will be beneficial for deeper understanding of defense mechanisms in plants, and furthermore, for genetically improving stress resistant crops.

Root hair development occurs in four main stages including cell fate determination, initiation, elongation, and maturation (Grierson et al. 2014). Studies in Arabidopsis have shown that the root hair cell fate is specified via interactions between protein complexes of transcriptional activators including WEREWOLF (WER), GLABRA 3 (GL3)/ENHANCER OF GL3 (EGL3), and TRANSPARENT TESTA GLABRA 1 (TTG1), and repressors including CAPRICE (CPC), TRYPTICHON (TRY), EHANCER OF TRY AND CPC1 (ETC1)/ETC2/ETC3, and GLABRA2 (GL2) in different cell files (Salazar-Henao et al. 2016). Following specification, root hair initiates and elongates through a cascade action of basic helix-loop-helix (bHLH) transcription factors (Masucci and Schiefelbein 1994; Yi et al. 2010; Datta et al. 2015).

Besides transcription factors, filamentous-actin (F-actin) also plays crucial roles in the establishment of cell polarity and the maintenance of elongation of root hairs (Bibikova et al. 1999; Stephan 2017; Bascom Jr. et al. 2018; Szymanski and Staiger 2018; Wang et al. 2020; Bi et al. 2022; Liu et al. 2023; Qian et al. 2024). In the process of root hair development, F-actin organization is in temporally and spatially dynamic changed. As root hair initiation, actin are fine, short, and form meshwork in bulges (Baluska et al. 2000; Chin et al. 2021). Accompanying the bulge elongated and transitioned to rapid tip growth, the base and shank of the cell consist of thick, longitudinal F-actin bundles along the growth direction, while the dome region of the cell still retains tip-focused dense actin network (Baluska et al. 2000; Chin et al. 2021). When root hair stops elongating, and gets mature, F-actin in tips are in thick cables instead of meshwork (Baluska et al. 2000; Chin et al. 2021). These phenotypic observations imply the precise regulation of actin organization during root hair development. As expected, people found that treatment Arabidopsis seedlings with the actin-disrupting compound latrunculin B (LatB) inhibits root hair elongation and unidirectional growth (Bibikova et al. 1999; Gibbon et al. 1999). Also, perturbing genes associated with F-actin biosynthesis or dynamics result in impaired root hairs with abnormal actin structures (Tominaga-Wada et al. 2011; Wang et al. 2020; Bi et al. 2022; Qian et al. 2024). The actin-related protein 2/3 (ARP2/3) complex is well-characterized actin nucleator in plants (Yanagisawa et al. 2013), playing crucial roles in a variety of eukaryotic cellular processes (Szymanski 2005; Yanagisawa et al. 2013; Chin et al. 2021). Functional loss of ARP2/3 complex subunits DISTORTED1 (DIS1) and CROOKED (CRK) lead to the formation of wavy root hairs (Mathur et al. 2003a, 2003b). However, activation of ARP2/3 for efficient nucleation of F-actin requires the SCAR/WAVE complex that is comprised by Sra1/PIR121/CYFIP1, Nap1/NAP125, Abi-1/Abi-2, Brick1 (BRK1)/HSPC300, and Scar/WAVE (Davidson and Insall 2013; Yanagisawa et al. 2013). Among these, BRK1/HSPC300 is critical for activation of ARP2/3 since BRK1 genetically interacts with ARP2/3 components and selectively stabilizes the ARP2/3 activators SCAR1 and SCAR2 (Djakovic et al. 2006; Le et al. 2006). Knockdown of HSPC300 in Drosophila cultured cells results in a reduction of cortical F-actin and alterations in cell morphology (Kunda et al. 2003). In plants, mutation of BRK1 gene causes the loss of pavement cell lobe formation and the production of distorted trichomes with aberrantly localized cortical F-actin organization (Frank et al. 2003; Djakovic et al. 2006). However, whether there are factors that regulate BRK1 activities remains unclear. Chin et al. (2021) reported that SPIRRIG (SPI) facilitates actin-dependent root hair development potentially by modulating BRK1 activities. SPI encodes a beige and Chediak Higashi (BEACH) domain containing protein (Saedler et al. 2009) and was initially isolated as one of a set of ‘DISTORTED’ genes, whose loss-of-function mutants show twisted and wavy trichomes (Hülskamp et al. 1994; Schwab et al. 2003). Further investigations demonstrate that spi mutants also display severer root hair developmental defects (Saedler et al. 2009; Chin et al. 2021). In particular, during the transition from initiation to tip growth in root hairs, BRK1 depletion coincides with SPI accumulation (Chin et al. 2021), implying the interaction between BRK1 and SPI in the processes of root hair development, but the genetic evidence is currently lacking.

In addition to itsr roles in fundamental developmental processes, F-actin cytoskeleton is also critical for abiotic stress response and resilience either as a direct target or a signal transducer (Wang and Mao 2019; Lv et al. 2021; Wang et al. 2021a, b; Bi et al. 2022; Kumar et al. 2023; Sun et al. 2023; Qian et al. 2024). For example, F-actin undergoes cycles of depolymerization and repolymerization under hyper-salinity condition (Wang et al. 2010). During early stages of salt stress, actin bundles are induced, while F-actin polymerization disappears after high salt stress, indicating F-actin mediated stress response and defensive mechanisms. Consistent with this, genetic analyses demonstrated that actin-related and actin-binding proteins participate in stress response by modulating the dynamics of the F-actin cytoskeleton (Zhao et al. 2013). Under hyper-salinity condition, the ARP2/3 complex promotes actin assembly and modulates cytoplasmic Ca2+ levels and ion homoeostasis, which may confer plant salinity tolerance. Whether there are any other F-actin cytoskeleton components facilitate in stress response is still interested.

In this study, we showed that SPI and BRK1 are required for salt stress induced root hair developmental plasticity in an F-actin cytoskeleton-dependent manner. Moreover, SPI modulates BRK1 stability during root hair development. Our work uncovers the key role of SPI and BRK1 on root hair developmental response to environmental stressors. Furthermore, we also provide direct evidence for the regulation of BRK1 by SPI at the post-translational level, addressing the molecular relationship between SPI and SCAR/WAVE-ARP2/3 complexes during cell morphogenesis.

Results

Isolation and analyses of the som1-1 mutant

To explore the possible mechanisms that plants use to respond to environmental stress, we performed forward genetic screen through examining root hair developmental response to salt stress by using gl2-3 EMS mutagenesis pool we established before (Shi et al. 2016). As a result, a recessive Arabidopsis mutant that exhibits hyper-sensitivity to salt stress was isolated, and was designated as salinity over-sensitive mutant 1–1 (som1-1).

Under our initial screen condition with 50 mM NaCl treatments, although both root hair initiation and elongation were inhibited in either the wild type (WT) or the som1-1 plants, som1-1 mutant showed a more dramatic response (Fig. 1A). During the treatment, the average root hair length of 4-day-old WT seedlings was decreased from ~ 362.21 ± 40.60 μm to ~ 65.11 ± 23.07 μm, with a decreased ratio of 82.82% (Fig. 1B, D). Similarly, root hair density in WT was also reduced, with a reduction ratio of 58.41% (~ 59.90 ± 4.39 before treatment and ~ 24.55 ± 4.58 after treatment, respectively) (Fig. 1C, E). In contrast, root hair initiation and elongation in som1-1 were almost inhibited under salt stress (Fig. 1A). After 50 mM NaCl treatment, the root hair length and density were ~ 0.49 ± 1.56 μm and ~ 3.50 ± 2.74 in som1-1 (Fig. 1D and E), with the decreased ratio of 98.35% and 88.64%, respectively (Fig. 1B and C). These results suggest that SOM1 gene may participate in salt stress response in the aspect of plant root hair development. To confirm the participation of SOM1 in salt stress, the som1-1 plants were also subjected to grow in NaCl-containing medium in different concentrations and root hair phenotypes were investigated. As shown in Fig. 1A, we found a dosage-dependent response of WT and som1-1 root hairs to salt stress. However, with the increase of the NaCl concentration, the repression of root hair elongation and initiation were significantly stronger in the som1-1 mutant compared to the WT (Fig. 1A, D and E). To evaluate the effects of SOM1 on the key genes that associate with root hair development, we detected the expression levels of the genes and found the transcripts levels of most of genes were reduced compared to those in the WT (Fig. S1), suggesting that SOM1 may act upstream of these genes to regulate root hair development. Taken together, these phenotypic observations indicate the important role of SOM1 in plant response to salt stress during root hair development.

Fig. 1
figure 1

Isolation of som1-1 mutant. A Root hair phenotypes of 4-day-old WT and som1-1 seedlings grown on 1/2 MS medium with or without NaCl. Scale bars = 200 μm. B and C Reduction ration of root hair length (B) and root hair density (C) of WT and som1-1 under 50 mM NaCl treatment. D and E Quantitative analyses of root hair length (D) and root hair number (E) in WT and som1-1 under NaCl treatment with different concentrations. For B and C, data are mean ± SD (n = 20) from three biological replicates. For D and E, the experiments were repeated for three times with similar results and one set of the data was represented (n = 20). ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05 (Student’s t test relative to controls)

Full size image

som1-1 is a new mutant allele of SPI

To determine the genetic identity of SOM1, a map-based cloning strategy was employed to isolate the SOM1 gene. Initial bulked segregant analyses (BSA) revealed that the SOM1 locus is close to molecular markers F12K11 and F3I6 on chromosome I (Figs. 2A; S2). Fine mapping confined SOM1 to an interval upstream of molecular marker F20D22 (Fig. 2A). Literature searches showed that the SPI (AT1G03060) gene resides in this region, and its loss-of-function mutants display similar root hair phenotypes as those of the som1-1 mutant (Hülskamp et al. 1994; Schwab et al. 2003; Saedler et al. 2009). SPI encodes a 3, 571 amino acid long BEACH domain containing protein that participates in a plethora of cellular and developmental processes (Saedler et al. 2009). To determine whether an SPI mutation is responsible for the developmental defects of som1-1, a genomic fragment of SPI from som1-1 was isolated and sequenced and a C to T single nucleotide transition was identified in the 10th exon of the SPI gene (Fig. 2A). Theoretically, this single nucleotide mutation would convert the 624th amino acid residue from glutamine to stop codon (Q624*) (Fig. 2A), resulting in a premature termination of protein translation. Due to the large size of SPI, we failed to generate a construct containing the native SPI promoter driving the full-length SPI cDNA or genomic DNA to carry out genetic complementation analyses. Alternatively, we obtained two independent SPI loss-of-function mutants spi-3 and spi-4 and carried out allelic tests. The spi-3 and spi-4 mutants carry T-DNA insertions in the 10th and 15th exons of SPI gene, respectively (Fig. 2B; Steffens et al. 2015). The transcripts accumulation of SPI in som1-1, spi-3, and spi-4 were examined by using RT-qPCR (Fig. 2C and D). Phenotypic examination revealed that both spi-3 and spi-4 exhibit similar root hair as well as trichome and epidermal pavement cell defects as those in som1-1 (Figs. 2E; S3A and B). Quantification analyses also showed similarities in decreased root hair length and root hair density, trichome branch length, and pavement cell complexity among som1-1, spi-3, and spi-4 (Table 1; Fig. S3C). The F1 plants from crosses between som1-1 and spi-3 as well as between som1-1 and spi-4 displayed root hair, trichome, and pavement cell morphologies resembling the respective single mutants (Table 1; Figs. 2E; S2A, B and C), indicating the failure of spi-3 and spi-4 to complement som1-1 in the F1 hybrids. In addition, spi mutants have also been reported to display salt hyper-sensitivity (Steffens et al. 2015), we found the growth of aerial parts and primary root of som1-1 are sensitive to NaCl treatment as well (Fig. S4A-C). Collectively, these results suggest that SOM1 represents the same genetic locus as SPI, and som1-1 represents a new mutant allele of SPI. Thus, som1-1 was renamed as spi-142 based on earlier nomenclature (Saedler et al. 2009). We also investigated the effects of salt stress on SPI expression and found SPI transcript is not apparent changed (Fig. S4D), implying the effect of salinity on SPI may be on protein or subcellular localization level.

Fig. 2
figure 2

Allelic analyses of som1-1 and spi mutants. A Map-based cloning of SOM1 locus. SOM1 gene was preliminarily linked to markers F12K11 and F3I6 on chromosome I through BSA analyses. Further fine-mapping mapping with additional molecular markers placed SOM1 upstream of F20D22. Numbers of recombinants were shown under each marker. The asterisk indicated the position of SPIRRIG (SPI) gene, AT1G03060. In the represented gene structure, boxes and solid lines represented the exons and introns, respectively. The 5´ and 3´ untranslated regions were shown as shaded boxes. The highlighted nucleotides below the gene model indicated the exact position of the mutation site. B Schematic representation of the mutation sites in som1-1, spi-3 (Salk_065311), and spi-4 (GK_420D09). Boxes and solid lines represented the exons and introns, respectively. The 5´and 3´untranslated regions were shown as shaded boxes. C and D Real-time quantitative RT-PCR (RT-qPCR) analyses of transcripts levels of SPI in som1-1 and spi alleles with primers F1 coupled with R1 in the 10th exon (C), and primers F2 coupled with R2 in the 15th exon (D). Fold changes were calculated with respect to the expression levels in the WT. Data are mean ± SD of three biological replicates. ****P < 0.0001, n.s. not significant (P > 0.05, Student’s t test). E Root hairs phenotypes of 4-day-old 1/2 MS medium grown WT, som1-1, spi-3, spi-4, som1-1 × spi-3 F1, and som1-1 × spi-4 F1 plants. Scale bars = 200 μm

Full size image
Table 1 Statistical analyses with respect to the cellular dimensions of root hairs and trichome cells in WT and spi mutants
Full size table

SPI is required for F-actin cytoskeleton associated salt stress induced root hair development

Since SPI is demonstrated to function in actin-mediated root hair development and the actin cytoskeleton is known to play vital roles in salt stress tolerance in Arabidopsis (Wang et al. 2010; Chin et al. 2021), we asked whether the hyper-sensitivity of spi-142 root hairs to salt stress is relevant to abnormal actin cytoskeleton organization dynamics. To this end, we examined the F-actin organization in WT (ABD2-GFP) and spi-142 root hairs under salt stress by generating spi-142 ABD2-GFP plants. As reported before (Baluska et al. 2000; Chin et al. 2021), prior to NaCl treatment, F-actin in WT root hair bulges were in network consisting of short, discontiguous filaments (Fig. 3A), while in elongating root hairs, the base and shank parts contained thick actin bundles along the growth axis and the apical region still remained in actin meshwork (Fig. 3A). When growth stopped, thick actin bundles extended parallel or obliquely to the growth axis and also protruded to the tip of the root hairs to replace the tip-focused actin meshwork (Fig. 3A). In contrast, under NaCl treatment, distinct tip-focused F-actin meshwork was unable to form and thick F-actin bundles directly protruded to tip region in either root hair bulges or pre-matured root hairs in WT plants, which resembled those observed in untreated spi-142 root hairs (Fig. 3A). We also examined F-actin organization in spi-142 root hairs under salt stress, comparable phenotypes to those in the untreated spi-142 were observed (Fig. 3A). Such an actin cytoskeleton structure may be responsible for salt stress induced formation of shorter and fewer root hairs since spi-142 root hairs showed similar growth phenotypes (Fig. 3A). These results also suggested that the hyper-sensitivity of spi-142 root hairs to salt stress could be due to the defects in actin cytoskeleton organization. To confirm the effects of SPI on salt stress induced F-actin cytoskeleton dynamics, we examined the F-actin structures in root epidermal cells as well. As shown in Fig. 3B, similar F-actin features were also observed in NaCl treated WT plants and untreated spi-142 plants, further suggesting a certain degree of correlation between the F-actin organization and the hyper-sensitivity of spi-142 to salt stress. As reported by Wang et al. (2010), we found prolonged salt stress resulted in F-actin depolymerization in WT plants (Fig. 3B). Consistently, measurements of F-actin density showed that WT root epidermal cells grown in 1/2 MS medium supplied with 50 mM NaCl had more F-actin arrays, which was reminiscent of those in untreated spi-142 plants (Fig. 3C). Moreover, we also observed the formation of twisted and disorganized actin bundles induced by NaCl treatment in WT plants that was dramatically different from those regularly arranged F-actin bundles in untreated plants (Fig. 3B). Quantitative analyses of these F-actin organizations were consistent with visual observations. The angles of the actin bundles were about 0°-45°with a significant peak between 0°-15° in WT plants without salt stress, while they displayed a relatively even distribution in different orientations with a peak between 45°-60° under stress condition, like those in the untreated spi-142 mutant (Fig. 3D). Consistently, investigation of the F-actin anisotropy exhibited that the mean anisotropy in WT without salt stress was about 0.31, compared to ~ 0.17 in WT under salt stress and ~ 0.14 in untreated spi-142 mutant, respectively (Fig. 3E). Taken together, these results indicate that SPI might contribution to salt stress induced developmental response through F-actin cytoskeleton related pathways.

Fig. 3
figure 3

Investigation of F-actin cytoskeleton organization in WT and spi-142 in the condition of NaCl treatment. A F-actin arrays in root hairs of 4-day-old ABD2-GFP and spi-142 ABD2-GFP seedlings grown on 1/2 MS with or without NaCl observed under spin-disc confocal microscopy. Scale bars = 10 μm. B F-actin arrays in the root epidermal cells of ABD2-GFP and spi-142 ABD2-GFP seedlings grown on medium with or without salt stress. Scale bars = 10 μm. C-E Quantitative analyses of the F-actin density (C), angles (D), and anisotropy (E) in root epidermal cells. F-actin filaments parallel to the cell’s longitudinal axis were define as 0° while those perpendicular to the cell extension direction were defined as 90°. The anisotropy score 0 was defined for nor order (purely isotropic arrays) and 1 for perfectly ordered (purely anisotropic arrays). Data are shown as mean ± SD. ****P < 0.0001, n.s. not significant (P > 0.05, Student’s t test)

Full size image

BRK1 mediates salt stress induced root hair developmental response

Given that SPI had been shown to facilitate actin-dependent root hair development probably through temporally and spatially regulating BRK1 expression (Chin et al. 2021), and BRK1 plays pivotal roles in F-actin nucleation (Djakovic et al. 2006; Le et al. 2006), we hypothesized whether SPI’s function related to F-actin in salt stress induced root hair developmental plasticity is associated with BRK1 expression. To address this issue, we evaluated the effects of BRK1 expression levels on root hair development under salt stress. First, we identified the BRK1 loss-of-function mutant brk1-1 (Fig. 4A and B; Djakovic et al. 2006), and whose root hair phenotypes were investigated. As shown in Fig. 4D, E and F, an increased root hair length and root hair density were found compared those in the WT, implying the negative regulation of BRK1 on both root hair tip-growth and initiation. In order to confirm the influence of BRK1 on root hair development, we also examined the root hair phenotypes of BRK1 overexpression plants by generating pBRK1:BRK1-GFP transgenic plants. The pBRK1:BRK1-GFP construct could restore the trichome defects of brk1-1 (Fig. S5A), indicating that BRK1-GFP is functional in planta. The overexpression of BRK1 in pBRK1:BRK1-GFP plants was verified by investigating transcripts levels and protein abundance of BRK1-GFP through RT-qPCR and western blot analyses, respectively (Fig. 4B and C). We found in 4-day-old pBRK1:BRK1-GFP plants, root hair density was comparable to that of the WT (Fig. 4D, F), but root hair length was remarkably reduced, resembling to those in spi-142 mutant (Fig. 4D and E). In WT seedlings, root hair length was ~ 288.65 ± 23.34 μm, while it was ~ 131.84 ± 17.99 μm—~ 146.14 ± 26.93 μm in pBRK1:BRK1-GFP plants (Fig. 4E). To test whether the reduced root hair length in pBRK1:BRK1-GFP is caused by reduced tip extension, we examined root hair growth rate. Like that in the spi-142 mutant, root hair growth rate was lower in pBRK1:BRK1-GFP than that in the WT (Fig. S5B and C), suggesting the function of BRK1 in root hair cell tip growth. Then, we analyzed the response of brk1-1 and pBRK1:BRK1-GFP root hairs to salt stress. As shown in Fig. 5A, root hair development in brk1-1 plants exhibited a resistance to salt stress while pBRK1:BRK1-GFP plants showed a hyper-sensitive response compared to those in the WT. After 50 mM NaCl treatment, root hair length and root hair density were ~ 55.86 ± 28.98 μm and ~ 22.75 ± 4.51 in the WT, whereas those were ~ 296.10 ± 73.31 μm and 45.45 ± 4.95 in brk1-1, and ~ 3.89 ± 4.82 μm and 5.45 ± 3.09 in pBRK1:BRK1-GFP plants, respectively (Fig. 5B, D). The reduction ratio of root hair length and root hair density was 82.01% and 56.35% in WT, 50.72% and 45.78% in brk1-1, and 97.46% and 83.99% in pBRK1:BRK1-GFP plants, respectively (Fig. 5C, E), which were consistent with the visual examinations. Taken together, these results indicate that BRK1 negatively modulates root hair developmental response to salt stress and may participate in the SPI-mediated regulation on salt stress induced root hair developmental plasticity.

Fig. 4
figure 4

Characterization of the effects of BRK1 on root hair development. A Schematic representation of the mutation site in brk1-1 (cs86554). Boxes and solid lines represented the exons and introns, respectively. The 5´and 3´untranslated regions were shown as shaded boxes. B RT-qPCR analyses of transcripts levels of BRK1 in WT, brk1-1, and independent BRK1 overexpression lines (pBRK1:BRK1-GFP). C Immuno-blotting examination of BRK1 abundance in BRK1 overexpression lines. The amounts of PBA1 were used as loading control. D Root hair phenotypes of 4-day-old 1/2 MS medium grown brk1-1 and BRK1 overexpression plants. Scale bars = 200 μm. E and F Root hair length (E) and root hair density (F) of WT, brk1-1, and BRK1 overexpression plants. For B, data are mean ± SD from three biological replicates . For E and F, the experiments were repeated for three times with similar results and one set of the data was represented (n = 20). ****P < 0.0001, ** P < 0.01, n.s. not significant (P > 0.05, Student’s t test)

Full size image
Fig. 5
figure 5

Examination of the response of BRK1 loss-of-function mutant and overexpression plants root hairs to salt stress. A Root hair phenotypes of 4-day-old WT, brk1-1, and BRK1 overexpression line (BRK1-OE #2) seedlings grown on 1/2 MS medium or the medium supplemented with different concentration of NaCl. Scale bars = 200 μm. B-E Quantification of root hair length (B) and root hair number (D) as well as the reduction ration of root hair length (C) and root hair number (E) under 50 mM NaCl treatment. For B and D, the experiments were repeated for three times with similar results and one set of the data was represented (n = 20), data are mean ± SD (n = 20). For C and E, data are mean ± SD (n = 20) from three biological replicates. ****P < 0.0001, *** P < 0.001, * P < 0.05 (Student’s t test)

Full size image

SPI modulates BRK1 protein stability during root hair development

To investigate the possible mechanism through which BRK1 participates in SPI-mediated regulation on root hair development under salt stress, we characterized whether SPI is involved in BRK1 stability control, given that SPI is required for BRK1 depletion during root hair development and the highly similar root hair phenotype of SPI loss-of-function mutant and BRK1 over-expression plants (Figs. 1 and 4; Chin et al. 2021). To this end, we first investigated the stability of BRK1 through in vitro and in vivo approaches. In in vitro cell-free degradation assay, GST-BRK1 recombinant protein was expressed and purified in E.coli and added to total proteins extracted from 2-week-old WT rosette leaves. After monitoring the GST-BRK1 abundance during the time course, we showed that GST-BRK1 is unstable. After 6 h incubation with the cell extracts, only ~ 37% of GST-BRK1 remained (Fig. 6A and B). Quantification of the degradation of kinetics showed that the half-life (t1/2) was about 3.9 h. We confirmed the instability of BRK1 through a protoplast-based in vivo assay. Through transiently expressing p35S:BRK1-GFP in Arabidopsis leaf mesophyll protoplasts, we examined BRK1-GFP fluctuations at the indicated times. As shown in Fig. 6C and D, without cytosolic translation inhibitor cycloheximide (CHX), the amounts of BRK1-GFP continuously increased as expression times prolonged (Fig. 6C and D). However, after 4 h CHX treatment, the BRK1-GPF levels apparently decreased (Fig. 6C and D), suggesting the degradation of BRK1-GFP and instability of BRK1. In addition, we also examined in planta BRK1 degradation by using pBRK1:BRK1-GFP transgenic plants. Similar to those observed in in vitro and in vivo experiments, we found BRK1-GFP levels decreased during the time course in the presence of CHX (Fig. 6F and G), suggesting that BRK1 is unstable during plant growth.

Fig. 6
figure 6

Degradation assay of BRK1. A In vitro cell-free protein degradation assay of recombinant protein GST-BRK1. GST-BRK1 was expressed and purified from E. coli and added to the cell extracts prepared from WT with our without proteasome inhibitor MG132 (200 μM). GST-BRK1 was detected with anti-GST antibody. B Quantitative analyses of GST-BRK1 levels in WT cell extracts with or without MG132. The half-life times were predicted based on the regression equations. C Protoplast-based in vivo protein degradation assay of BRK1-GFP protein. Transient expression vector p35S:BRK1-GFP was expressed in protoplast from WT leaf mesophyll and total proteins were extracted. The abundance of BRK1-GFP was detected when the cytosolic translation inhibitor CHX (50 μM) and MG132 were present or absent. D and E Quantification of the degradation kinetics of BRK1-GFP levels under CHX treatment (D) or under combined treatment with CHX and MG132 (E). The half-life times were predicted based on the regression equations. F In planta degradation assay of BRK1-GFP in pBRK1:BRK1-GFP transgenic plants. Total proteins were extracted from pBRK1:BRK1-GFP and treated with CHX, and BRK1-GFP levels were detected by western blot. G Quantitative assessments of BRK1-GFP levels in pBRK1:BRK1-GFP plants after CHX treatment. The half-life times were predicted based on the regression equations. For A, C, and F, the same volume of DMSO was used as control, and the amount of PBA1 was used as the loading control. For B, D, E, and G, data are presented as mean ± SD of three biological replicates

Full size image

Given that the ubiquitin/26S proteasome system (UPS) controls the turnover of numerous proteins in cells (Vierstra 2009), we analyzed whether BRK1 is subject to 26S proteasomal destruction. Recombinant GST-BRK1 protein was added to protein extracts from WT with or without the proteasome inhibitor carbobenzoxy-L-Leucyl-L-Leucyl-L-Leucinal (MG132). We found that in the absence of MG132, the amount of GST-BRK1 clearly decreased after the incubation (Fig. 6A and B). However, the abundance of GST-BRK1 was retained when MG132 was present (Fig. 6A and B), indicating the repression of BRK1 degradation and BRK1 degradation may be mediated by 26S proteasome. The 26S proteasome-dependent BRK1 destruction was further confirmed by protoplast-based in vivo assays. As shown in Fig. 6C, D, when CHX was added to the transfected cells, protein gel blotting showed a decrease of BRK1-GFP abundance within 4 h. In contrast, the BRK1-GFP levels were almost sustained when CHX and MG132 were presented simultaneously (Fig. 6C, E).

As most proteins that are subjected to UPS-dependent degradation are marked with ubiquitin chain that is covalently linked to lysine (Lys) residues (Vierstra 2009), we sought to identify the potential Lys residues that mediate BRK1 degradation. All four Lys residues in the BRK1 protein were mutated to Arginine (Arg), respectively (Fig. S6A), and the mutant proteins were expressed and purified from E. coli. The stability of the mutant proteins was investigated in a cell-free degradation assay. As shown in Fig. S6B and C, we found that the Lys to Arg substitution at K3 leads to apparent stabilization for GST-BRK1, suggesting that this residue may be required for ubiquitin-dependent degradation of BRK1.

Next, we investigated whether SPI mediates BRK1 turnover by examining the effect of SPI inactivation on BRK1 abundance. The pBRK1:BRK1-GFP construct was transformed into spi mutants and BRK1-GFP levels were determined using anti-GFP antibodies. Transgenic lines with equivalent BRK1 transcripts to that of pBRK1:BRK1-GFP were selected (Fig. 7A), and the BRK1-GFP levels were examined. As shown that in Fig. 7B and C, the steady-state levels of BRK1-GFP were higher in spi mutants than in the WT, suggesting the excess accumulation of BRK1 when SPI is inactive. To further examine the regulation of SPI on BRK1 protein stability, we compared BRK1 abundance in WT and in spi mutants using the in vitro and in planta degradation assays. Similar to the results shown in Fig. 7B, the amount of in vitro added GST-BRK1 decreased during the time course in the cell extracts from WT but was retained in spi-142 cell extracts (Fig. 7D and E). After 6 h incubation, the GST-BRK1 in the cell extracts from WT was almost exhausted (~ 11% remained), while in spi-142 cell extracts, ~ 35% GST-BRK1 protein could still be detected. These results suggested that SPI mutation increases the stability of BRK1. To validate the involvement of SPI in BRK1 stability control, we examined in planta degradation dynamics of BRK1 in WT and spi mutant background respectively. We found that when cytosolic translation is inhibited by CHX, the amount of BRK1-GFP decreased in WT after 1.5 h incubation (~ 29% remained; Fig. 7F and G). Conversely, BRK1-GFP degradation was significantly relieved in spi mutant background (Fig. 7F and G). After 1.5 h incubation, abundant BRK1-GFP (~ 64%) could still be detected, which was in consistent with that in in vitro experiments. Taken together, these results suggested the involvement of SPI in regulating BRK1 stability, and loss of SPI function leads to the accumulation of BRK1. In addition, we also examined the effects of NaCl treatment on BRK1 stability. As shown in Fig. S7, the amount of BRK1-GFP after treatment is higher than those before treatment, suggesting that salt stress may stabilize BRK1. Further, we found that SPI may involve in the influence of salinity on BRK1 stability since BRK1-GFP is more abundant in spi mutant than that in the WT after NaCl treatment (Fig. S7).

Fig. 7
figure 7

Investigation of the effects of SPI on BRK1 degradation. A Transcripts levels of BRK1 in pBRK1:BRK1-GFP and spi pBRK1:BRK1-GFP plants examined by RT-qPCR. Fold changes were calculated with respect to the expression levels in the WT. B Accumulation of steady-state BRK1-GFP in WT and spi mutants. The amount of PBA1 was used as the loading control. C Quantification of the abundance of BRK1-GFP in WT and spi mutant backgrounds. D In vitro cell-free protein degradation assay of GST-BRK1 in WT and in spi-142 mutant. Purified GST-BRK1 was incubated with the cell extracts prepared from WT and spi-142 respectively and the abundance of GST-BRK1 was detected with anti-GST antibody. The amount of PBA1 was used as the loading control. E Degradation plot of GST-BRK1 in WT and in spi-142 extracts. The half-life times were predicted based on the regression equations. F In planta protein degradation assay of BRK1-GFP protein in pBRK1:BRK1-GFP and spi pBRK1:BRK1-GFP plants. Total proteins were extracted from pBRK1:BRK1-GFP and spi pBRK1:BRK1-GFP plants and treated with CHX. BRK1-GFP levels were detected by western blot with anti-GFP antibody at the indicated times. The amount of PBA1 was used as the loading control. G Quantitative assessments of BRK1-GFP levels in pBRK1:BRK1-GFP and spi pBRK1:BRK1-GFP plants. The half-life times were predicted based on the regression equations. For A, C, E, and G, data are means ± SD of three biological replicates. ****P < 0.0001, *P < 0.05, n.s. not significant (P > 0.05, Student’s t test)

Full size image

Discussion

As the major communicative interface of plants and their environments, root hairs display high developmental plasticity under adverse environmental conditions and the presence of root hairs therefore serves as an effective defensive strategy for plants to cope with stress challenges (Wang and Li 2008a; Wang et al. 2020; Karlova et al. 2021; Bi et al. 2022; Kohli et al. 2022; Ibeas et al. 2024; Qian et al. 2024). Salt stress inhibits root hair initiation and tip growth in Arabidopsis (Wang et al., 2008b; Jin et al. 2023), which is considered to be an adaptive strategy of protection (Shan et al. 2005; Wang et al. 2008b; Ji et al. 2013). Root hair developmental response to salt stress was reported to be caused by ion disequilibrium and mediated by Salt Overly Sensitive (SOS) pathway (Shi and Zhu 2002, Wang et al. 2008b; Van Zelm et al. 2020). Single mutant of SOS genes exhibits apparent developmental-defective phenotypes for root hairs and differential salt response (Shi and Zhu 2002, Wang et al. 2008b). Recent studies showed that SOS2 coordinates with Rho GTPase of Plants 2 (ROP2) and Rho GTPase GDP Dissociation Inhibitor 1 (RhoGDI1) to regulate the growth of root hair under salt stress (Liu et al. 2023). Transcription factors also serve important roles in salt stress induced root hair development. Overexpression of bHLH transcription factor ROOT HAIR DEFECTIVE 6 (RHD6) results in plants sensitive to NaCl treatment (Jin et al. 2023), while bZIP proteins ABSCISIC ACID RESPONSIVE ELEMENT-BINDNG FACTOR 1, 3, and 4 (ABF1, 3, 4) physically interact with RHD6 and suppress its transcriptional activity to positive modulate the salinity induced inhibition of root hair development (Jin et al. 2023). Salt stress disturbs root hair growth also via affecting the biosynthesis and signaling of phytohormones (Lv et al. 2018; Wang et al. 2019). Salt stress influences the localization of auxin efflux carrier PIN-FORMED 2 (PIN2) and induced its endocytosis, as a consequences, altering auxin transport and positioning selection of root hairs (Wang et al. 2019). In crops and other plants, root hair developmental plasticity plays vital roles for plant growth as well under salt stress. RHD3 was cloned from the salt tolerant hybrid wheat variety, and whose transcription is down-regulated under salt stress (Shan et al. 2005). Overexpression of heat shock transcription factor OsHsfA7 exhibits less, shorter root hair and salt tolerance in rice (Liu et al. 2013). SbbHLH85 modulates resilience to salt stress by regulating root hair growth in sorghum (Song et al. 2022). These results provide theoretic evidences and candidate genes for stress molecular breeding. Limonium bicolor is considered as a model recretohalophyte, and can survive on high salinity environment (Yuan et al. 2022). Molecular studies showed that heterologous overexpression of bHLH transcription factor and importin-β protein SUPER SENSITIVE TO ABA AND DROUGHT 2 (SAD2) from L.bicolor reduces root hair number and enhances salt tolerance while expression LbTRY could induce root hair formation and leads plants to sensitive to salt stress (Wang et al. 2021a, b; Leng et al. 2021; Xu et al. 2021), presenting new insight into plant salt tolerance mechanisms. In this study, we report that BEACH domain containing protein SPI involves in salt stress induced root hair developmental plasticity in Arabidopsis. Previous studies showed that SPI’s function is important for root hair development as spi mutants exhibit shorter root hairs (Saedler et al. 2009; Chin et al. 2021). We found root hair initiation is also affected in spi mutants (Table 1; Figs. 1 and 2) and the growth of spi root hairs is hyper-sensitive to salt stress (Fig. 1), which is consistent with previous reports that SPI mutation leads to the growth of primary root and seedling sensitive to salinity (Steffens et al. 2015). Moreover, the developmental response of spi mutants under salt stress is similar to those in the sos mutants (Shi and Zhu 2002; Wang et al. 2008a,b), implying that SPI may also associate with SOS pathway to response salt environment, but the details need further investigations.

F-actin cytoskeleton dynamic plays essential roles in orchestrating root hair cell shape during development and in environmental stress response and resilience (Wang et al. 2011; Soda et al. 2016; Stephan 2017; Bascom Jr. et al. 2018; Szymanski and Staiger 2018; Wang and Mao 2019; Liu et al. 2021; Bi et al. 2022; Kumar et al. 2023; Sun et al. 2023; Liu et al. 2023; Qian et al. 2024). However, whether these two F-actin related processes work in concert to help plants to adapt and survive in unfavorable environments remains unclear. Short-term salt stress induces actin cytoskeleton assembly and bundling, but treatment with either long-term or high concentration induces actin cytoskeleton disassembly, and the SOS pathway and calcium signaling are associated with these processes (Wang et al. 2010, 2019). Actin reorganization in sos mutants is abnormal in response to salt stress (Wang et al. 2010). Consistently, disruption of the actin filaments with actin-filament-disrupting drugs increases death of sos2 seedling under salt treatment conditions (Wang et al. 2010). Salt stress induces calcium accumulation in the cytosol, and ARP2/3 complex mediates calcium elevation under salinity (Qian et al. 2019; Zhao et al. 2013). Mitochondria-dependent [Ca2+]cyt increase is enhanced in arp2 mutant, resulting in its hyper-sensitivity to salt, suggesting an interaction between actin cytoskeleton and calcium oscillation under salinity. The actindepolymerizing factor (ADF) is highly conserved among eukaryotes and plays critical roles in the various processes of plant growth and stress response via remodeling actin cytoskeleton architecture (Sun et al. 2023). It has been showed that ADF1 is regulated by MYB transcription factor MYB73 and is involved in response to salt stress by affecting actin filament organization (Wang et al. 2021a, b). adf1 mutants show significantly reduced survival rate, increased percentage of actin cable, and reduced density of actin filaments under salt stress. Systematically Genome-wide identification of ADFs under various abiotic stresses in soybean (Glycine max) revealed that 18 GmADF genes show distinct expression patterns under drought and salt stresses, which may facilitate the engineering of salt-tolerant crops (Sun et al. 2023). Villin (VLN) that is considered to be one of the most important actin-binding proteins plays essential roles in plant development and in resisting adverse environments (Lv et al. 2021). GhVLN3 and GhVLN10 are highly and preferentially expressed in elongation cotton fibers and distinctly upregulated by abiotic stresses (Lv et al. 2021). Arabidopsis ADF7 and VLN1 are positively and negatively involved in root hair formation respectively (Bi et al. 2022). ADF7 inhibits the expression of VLN1 and leads to the decline of F-actin bundling and thick bundle formation as well as the increase of F-actin depolymerization to promote root hair formation (Bi et al. 2022). Further investigation showed that GL2 directly binds to the promoter of VLN1 and positively regulates VLN1 expression and actin dynamics in root hairs (Wang et al. 2020).

SPI has been reported to facilitate root hair development through an F-actin cytoskeleton-dependent pathway (Saelder et al. 2009; Chin et al. 2021). Moreover, SPI regulates the expression of genes that are relevant to salt stress response (Steffens et al. 2015), implying that SPI could be an integrator that responses to F-actin dependent developmental and environmental signals during root hair development, but direct evidences are currently lacking. Based on our salt sensitive experiments and the investigation of actin cytoskeleton organization, we found that spi mutant root hair displays similar actin structures to those in the WT under salt stress (Fig. 3), suggesting that SPI may mediate F-actin associated root hair developmental response under salt stress. Furthermore, based on genetic and biochemical studies, we showed SPI interacts with BRK1 to modulate salt stress induced root hair developmental plasticity in F-actin cytoskeleton associated manner.

The formation of diverse actin organization is mediated by actin-binding proteins that nucleate, destabilize and bundle actin filaments (Szymanski and Staiger 2018). BRK1 that encodes a subunit of SCAR/WAVE complex is considered to be essential for the proper organization of actin cytoskeleton because the function of actin nucleators ARP2/3 complex apparently depends on the activity of BRK1 (Le et al. 2006; Djakovic et al. 2006). Accordingly, BRK1 has been showed to participate in a range of cellular and developmental events and exhibit distinct expression patterns (Djakovic et al. 2006; Le et al. 2006; Liu et al. 2024). During trichome development, BRK1 is tip localized to trichome branches (Yanagisawa et al. 2018), while in the processes of root hair development, BRK1 is only expressed in the root hair initiation stage, and disappears during root hair elongation (Chin et al. 2021). This kind of specific expression pattern suggested that the activities of BRK1 may be tightly regulated, but the pathways that associate with modulating BRK1 activities stay to be identified. In addition, whether BRK1 is involved in stress induced F-actin organization is also very attractive. Based on genetic analyses and NaCl treatment experiments, we found BRK1 is relevant to salt stress and negatively regulates salt stress induced root hair development (Figs. 4 and 5). Furthermore, we also found that BRK1 is unstable and is subjected to 26S proteasome for proteolysis (Fig. 6). More importantly, we revealed that SPI is required for BRK1 degradation (Fig. 7). SPI mutation results in the accumulation of steady-state BRK1 in vitro and in vivo, which is consistent with the cellular observation by Chin et al. (2021). Our genetic analyses also showed that BRK1 overexpression results in root hair developmental defects that resemble those in the spi mutants (Figs. 2E and 4D-F), further support the regulatory role of SPI on BRK1 stability. Considering the spatial and temporal expression differences between SPI and BRK1 during root hair development, the modulation of SPI on BRK1 stability may be indirectly. Studies by Steffens et al. (2015) showed that SPI regulates pleiotropic transcriptional changes under salt stress, so the modulation of SPI on BRK1 stability may be due to the transcriptional alternation of the factors that involve in ubiquitin-26S proteasome pathway, but the precise associated components need to be isolated. Taken together, our results uncover the functions of SPI and BRK1 and their regulatory relationship in response to root hair development and salt stress, revealing a pathway that modulates the activities of actin cytoskeleton related protein during plant cell development and under environmental stress.

Materials and methods

Plant materials and growth condition

All plant materials used in the study are in the Columbia-0 (Col-0) background except those used for map-base cloning. The wide type (WT) refers to Col-0 plants. T-DNA insertion lines for SPI (Salk_065311, spi-3; GK_420D09, spi-4) and BRK1 (CS86554, brk1-1) were obtained from the Arabidopsis Biological Resource Center (ABRC). The T-DNA insertion sites and homozygous plants were confirmed by genomic PCR and phenotypic examination. Primer pairs used for genotyping are listed in Table S1. The marker line ABD2-GFP that indicates the F-actin microfilaments was kindly provided by Prof. Zhaosheng Kong (Chinese Academy of Sciences).

For plant culture, seeds were stratified at 4 °C for 2 days and then sown on commercial soil mix (Pindstrup) for gemination and growth in growth room at 22 ± 1 °C under continuous illumination (~ 100 μmol m–2 s–1). For protoplast preparation, seeds were sowed and grown on Jiffy-7-Peat Pellets (Jiffy Group) in a growth chamber (Conviron A1000) with day/night cycle (12 h/12 h) at 22 ± 1 °C, and fully expanded rosette leaves of 30-day-old plants were collected.

Root hair, trichome, and cotyledon pavement cell phenotype characterization

Root hairs in the primary root of 4-day-old 1/2 MS-medium grown seedlings with or without NaCl were examined under the Leica M295 dissection microscope, and those in the root tips (approximately 3 mm long) were imaged using the equipped Nikon N995 digital camera. The length of the 20 longest root hairs with observable ends were measured with Image J software for each seedling (Masucci and Schiefelbein 1994). For root hair number, all visible root hairs in the root tips were counted. For measurement of the root hair growth rate, a time-lapse strategy was used as described by Szymanski and Nielsen (2009). The time interval between each measurement was 2 min, and root hair elongation was measured using Image J software.

Trichome phenotypes were examined as described by Saedler et al. (2009). Plants in different genetic backgrounds were grown in soil for 2 weeks and trichome morphologies in the 5th and 6th rosette leaves were examined with a stereoscope (Olympus, SZ61). Representative trichomes were imaged by using a tabletop scanning electron microscope (Hitachi, SEM TM3030). Trichome branch length was measured with the Image J software.

Epidermal pavement cell shapes in cotyledons of 4-day-old 1/2 MS-medium grown seedlings were examined by confocal microscopy. The outlines of pavement cells were indicated by staining with Propidium Iodide (PI) (1 mg/mL). The complexity of pavement cells was quantified using the formula: perimeter2/(4 × π x area) (Saedler et al. 2009).

For statistical purposes, at least 20 individual plants for each genotype in different conditions were used for analyses, and all experiments were repeated three times. Student’s ttest was used to assess the difference.

Map-based cloning

The som1-1 mutant was isolated from a gl2-3 (Salk_039825) ethyl methanesulfonate (EMS)-mutagenesis population we established previously (Shi et al. 2016). To identify the mutation site in som1-1, map-based cloning was employed according to Lukowitz et al. (2000). First, the som1-1 mutant was crossed with Arabidopsis ecotype Landsberg erecta (Ler) to generate an F2 mapping population. Then, bulked segregant analyses (BSA) were conducted with 25 pairs of molecular markers covering all 5 Arabidopsis chromosomes and a mixed genomic DNA pool with 94 individuals from the F2 segregation population. Next, fine mapping was carried out to narrow down the physical interval harboring the mutation site with additional molecular markers. Finally, candidate genes in the resulting chromosomal region were sequenced to identify the mutation site. Sequence information of the molecular markers used is listed in the Table S1.

Protein expression and purification

To express and purify the recombinant native GST-BRK1 and mutant GST-BRK1 proteins in vitro, pGEX4T-1-BRK1, pGEX4T-1-BRK1-M1 (K3R), pGEX4T-1-BRK1-M2 (K49R, K51R), and pGEX4T-1-BRK1-M3 (K58R) plasmids were constructed and transformed into Escherichia coli Rosetta (DE3) strain. The cells were cultured in liquid LB medium at 37 °C at 220 rpm until OD600 = 0.6. Then, GST-BRK1 recombinant proteins were induced to express with 1 mM isopropyl-ß-D-thiogalactopyranoside (IPTG). Soluble proteins were purified with Glutathione Sepharose 4B beads (17–0756-01, GE Healthcare) according to the manufacturer’s instructions. Primers used for plasmid construction are listed in Table S1.

In vitro and in vivo protein degradation assay

The experiments were performed according to Wang et al. (2009). For in vitro degradation assay, total proteins from 2-week-old plants were extracted with degradation buffer (25 mM Tris (pH = 7.5), 10 mM NaCl, 10 mM MgCl2, 4 mM PMSF, 5 mM DTT, 10 mM ATP). Protein concentration was determined by the Bio-Rad protein assay, and the concentration in different plant samples were adjusted to be equal using degradation buffer. Then, about 500 ng GST-BRK1 recombinant protein was added to the total protein extracts and incubated at 22 °C for the indicated times. At each time interval, 20 μL of the reaction products were taken and boiled in the same volume SDS sample buffer for 10 min to stop the reaction. The abundance of GST-BRK1 was determined by western blot assay with anti-GST antibody (ab18184, Abcam, England).

For protoplast-based in vivo protein degradation assay, mesophyll protoplasts were isolated from the WT and transfected with transient expression plasmid p35S:BRK1-GFP according to Yoo et al. (2007). The transfected cells were cultured in the dark for 10 h in W5 buffer (154 mM NaCl, 125 mM CaCl2, 5 mM KCl, 2 mM MES). Then, the cells were treated with 50 μM cytosolic translation inhibitor cycloheximide (CHX, 66–81-9, Merk) for the indicated times. The same volume of DMSO was used as the control. Total protein was extracted from the treated cells and BRK1-GFP levels were examined by western blot with anti-GFP antibody (ab290, Abcam, England).

To perform in planta degradation assay, the stable expression plasmid pBRK1:BRK1-GFP was constructed and transformed into WT, spi mutants, and brk1-1 mutant respectively. The transgenic plants were obtained through antibiotic screening, genomic PCR, and GFP florescence examination. In the T3 generation, total protein were extracted from the latest emerging rosette leaves of 2-week-old soil grown plants with degradation buffer, and treated with CHX for indicated times. The amount of BRK1-GFP protein was evaluated by western blot.

For all immuno-blotting assay, the amounts of PBA1 detected with anti-PBA1 antibody (ab98861, Abcam, England) were used as loading controls. Primers used for p35S:BRK1-GFP and pBRK1:BRK1-GFP plasmids construction are listed in Table S1.

RNA isolation and real-time quantitative RT-PCR (RT-qPCR)

Total RNAs were extracted from the roots of 4-day-old seedlings with TRIzol reagent (15,596–026, Invitrogen, USA). The first-strand DNAs were synthesized with oligo (dT15) primer by using Uelris RT mix with DNase (All-in-one) (R2020, US Everbright® Inc, USA) with 1 μg total RNAs. Real-time qPCRs were carried out with the FastStart Essential DNA Green Master kit (06402712001, Roche, Switzerland) and Bio-Rad CFX96 real-time PCR system. Relative expression levels of the target genes were calculated with 2−△Ct. Expressions levels of GADPH were used as the internal controls. Primers used are listed in Table S1.

Salt sensitivity assay

The experiments were carried out according to Wang et al. (2008b) and Jin et al. (2023). Seeds of the plants from different genotypes were surface sterilized and sown on 1/2 MS medium with or without different concentration of NaCl. Growth status of root hair, primary root, and germination of 4-day-old seedlings were examined. Root hair length, root hair density, primary root length, and biomass were quantitatively measured.

F-actin microfilaments imaging and quantification

For in planta F-actin observation, the ABD2-GFP marker line was crossed into the spi-142 background and unsegregated spi-142 ABD2-GFP plants were selected in subsequent generations by genomic PCR and GFP fluorescence analyses. F-actin organization in root hairs and root epidermal cells were examined with a spinning disk confocal system built on a DMi8 inverted microscope (Lecia) equipped with a CSU-W1 confocal scanner unit (Yokogawa) and an iXon Ultra 888 EMCCD camera (Andor) with a HC PL APO 63 × N.A.1.30 glycerol objective (Leica). Images were taken through Z-stacking with a 0.5 um z-step. GFP was excited at 488 nm.

Actin density measurements were performed according to Guan et al. (2021). Images acquired were opened with Fiji-Image J, and a line of fixed length (10 μm) perpendicular to the directions of most F-actin was draw and the numbers of F-actin across the line were counted. F-actin angles and anisotropy quantifications were performed according to Boudaoud et al. (2014). Images were loaded in Image J and angles of individual F-actin were measured. Actin filaments parallel to the cell extension direction were defined as 0° while those perpendicular to the cell’s longitudinal axis were defined as 90°. For F-actin anisotropy evaluation, the following convention was used: anisotropy score 0 = no order (purely isotropic arrays) and 1 = perfectly ordered (purely anisotropic arrays). For each scenario, a total of 8-10 seedling roots were analyzed, and 2-3 root epidermal cells with clear F-actin configuration in each seedling were selected for quantitative analyses. For F-actin angles analyses, 10 actin filaments or bundles in each cell were selected for measurements, therefore, about 300 (10 x 3 x 10) actin filaments were surveyed in each condition. Values of F-actin angle were divided into 6 intervals (0-15°, 15-30°, 30-45°, 45-60°, 60-75°,75-90°), respectively. The frequency of the F-actin angles in each interval was analyzed, and finally exhibited as a heatmap.

Statistical analyses

Statistical analyses were carried out using Graphpad Prism software. Data are presented as means and standard diviations of the distributions. Student’s t test were performed to determine the significance of differences between data sets. All tests were two tailed, performed at the significance level P = 0.05. For all analyses, P < 0.05 was considered statistically significant (*P < 0.05; ** P < 0.01; *** P < 0.001; **** P < 0.0001; n.s. not significant, P > 0.05).