(C) PLOS One This story was originally published by PLOS One and is unaltered. . . . . . . . . . . Actin depolymerizing factor ADF7 inhibits actin bundling protein VILLIN1 to regulate root hair formation in response to osmotic stress in Arabidopsis [1] ['Shuangtian Bi', 'College Of Bioscience', 'Biotechnology', 'Shenyang Agricultural University', 'Shenyang', 'Mingyang Li', 'Caiyuan Liu', 'Xiaoyu Liu', 'Jianing Cheng', 'Lu Wang'] Date: 2022-11 Actin cytoskeleton is essential for root hair formation. However, the underlying molecular mechanisms of actin dynamics in root hair formation in response to abiotic stress are largely undiscovered. Here, genetic analysis showed that actin-depolymerizing protein ADF7 and actin-bundling protein VILLIN1 (VLN1) were positively and negatively involved in root hair formation of Arabidopsis respectively. Moreover, RT-qPCR, GUS staining, western blotting, and genetic analysis revealed that ADF7 played an important role in inhibiting the expression and function of VLN1 during root hair formation. Filament actin (F-actin) dynamics observation and actin pharmacological experiments indicated that ADF7-inhibited-VLN1 pathway led to the decline of F-actin bundling and thick bundle formation, as well as the increase of F-actin depolymerization and turnover to promote root hair formation. Furthermore, the F-actin dynamics mediated by ADF7-inhibited-VLN1 pathway was associated with the reactive oxygen species (ROS) accumulation in root hair formation. Finally, ADF7-inhibited-VLN1 pathway was critical for osmotic stress-induced root hair formation. Our work demonstrates that ADF7 inhibits VLN1 to regulate F-actin dynamics in root hair formation in response to osmotic stress, providing the novel evidence on the F-actin dynamics and their molecular mechanisms in root hair formation and in abiotic stress. Root hairs are required for plants to absorb nutrients and water. The dynamics of cytoskeleton such as actin filaments (F-actin) are necessary for the formation of root hairs, which is regulated by different kinds of cytoskeleton-binding proteins. At the same time, the dynamics of cytoskeleton are also involved in plant abiotic stress tolerance. However, there are few studies on the underlying molecular mechanisms of F-actin dynamics in root hair formation in response to abiotic stress. Actin depolymerization factor 7 (ADF7) and actin bunding protein Villin 1 (VLN1) are important actin-binding proteins in Arabidopsis. Here, we describe a pathway that ADF7 inhibits VLN1 to regulate F-actin dynamics in root hair formation in response to osmotic stress, providing a new evidence for the studies on the molecular mechanisms of F-actin dynamics in root hair formation and in plant abiotic stress tolerance. Funding: This research was supported by the National Key Research and Development Program of China (2022YFE0108200), National Natural Science Foundation of China (Nos. 31970183 and 31470358), Liaoning Revitalization Talents Program (XLYC2002065), Liaoning Education Foundation, Hundred Thousand Ten Thousand Talent Project of Liaoning Province, and Shenyang Agricultural University Graduate Innovation Cultivation Fund. It was also partly supported by the open funds of the State Key Laboratory of Plant Physiology and Biochemistry (SKLPPBKF1905). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Data Availability: All relevant data are within the manuscript and its Supporting Information files. The TAIR accession numbers for the sequences of the genes used in this study are as follows: ADF7 (AT4G25590) and VLN1 (AT2G29890). Copyright: © 2022 Bi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. ROS accumulation plays an important role in root hair formation and elongation [ 35 – 38 ]. In plants, NADPH oxidase catalyzes ROS production. NADPH oxidase-mediated ROS production is the best-characterized mechanism during root hair development [ 35 – 37 ]. NADPH oxidase is encoded by ROOT HAIR DEFECTIVE 2 (RHD2/RBOHC) gene [ 35 , 36 ]. rhd2 mutants display defects in root hair formation and elongation, correlated with reduced ROS levels in roots and root hairs [ 35 – 37 ]. FER, FERONIA receptor-like kinase, is required for root hair development by regulating NADPH oxidase-dependent ROS production in root and root hairs [ 37 ]. Actin dynamics are involved in the regulation of ROS level in vivo [ 26 , 39 ]. Both moderate actin polymerization and moderate actin depolymerization increase NADPH oxidase activity in microglia [ 40 ]. F-actin depolymerization elevates ROS levels in roots by regulating RHD2 expression in salt stress in Arabidopsis [ 39 ]. Actin disrupting drug Lat B resulted in a significant increase in immunogenic peptide flg22-induced ROS production [ 26 ]. Therefore, we suppose that the integration of the actin dynamics and ROS signaling might play a key role during root hair development. Here, we found that ADF7 inhibited the expression and the function of VLN1, resulting in elevating F-actin depolymerization, fine F-actin amount, F-actin turnover, and ROS accumulation in root epidermal cells and new emerged root hair cells, which plays an important role in osmotic stress-induced root hair formation, providing the first evidence on the molecular mechanisms of F-actin depolymerization and F-actin bundling in root hair formation and in osmotic stress. Actin depolymerization factors (ADFs) are responsible for de-polymerizing and severing single F-actin [ 24 – 27 ]. Arabidopsis ADF7 is one of 11 ADF family proteins and possesses the mild activities of single F-actin depolymerizing and severing, compared with the other ADFs [ 24 , 28 ]. Previous findings showed that ADF7 is required for pollen tip growth by severing actin-mediated turnover of F-actin [ 28 ]. Additionally, ADF7 highly expresses in the microspore stage using ADF7-GFP expression analysis [ 25 ]. Villins (VLNs) prominently possess F-actin bundling abilities [ 29 – 31 ]. Arabidopsis genome encodes 5 VLN isoforms (VLN1-5) [ 29 , 32 ]. VLN1 displays a simple actin bundling capacity in a calcium ion (Ca 2+ ) and Ca M in an independent manner [ 32 ]. VLN1 is highly expressed in various plant tissues including leaves, hypocotyls, roots, and root hairs [ 29 , 33 ]. VLN1 and VLN3 play a partially overlapping role in the turnover of actin bundle formation in vitro [ 34 ]. VLN1 interacts with ADFs to affect F-actin dynamics in vitro [ 32 ]. Additionally, VLN1 negatively regulates root hair elongation mediated by transcription factor GL2 in osmotic stress [ 33 ]. A few actin-binding proteins (ABPs) are identified in root hair initiation. Profilin plays a primary role in accelerating F-actin assembly; additionally, it promotes actin bundles/cables and inhibits actin nucleation [ 14 – 19 ]. Arabidopsis profilin1 mutant, prf1-1, seedlings develop higher density and longer root hairs [ 15 ], suggesting that PRF1 is involved in root hair formation and elongation. CROOKED/ARPC5 is molecularly identified as a subunit of the ARP2/3 complex that possesses the function of nucleating actin assembly [ 20 , 21 ]. Crooked seedlings grow more than one root hair from the same hair cell, suggesting that ARPC5 negatively regulates root hair formation [ 21 ]. AtFH8 participates in several cytoskeletal functions such as nucleating, capping, binding, severing F-actin, and binding to profilin [ 22 , 23 ]. Overexpression of AtFH8 leads to producing more than one root hair on one hair-forming site, indicating that FH8 is involved in bulges formation rather than root hair formation [ 23 ]. Because the actin arrays and dynamics mediated by PRF1 and ARPC5 in root hair formation are not reported, how PRF1 and ARPC5 regulate actin dynamics to affect root hair formation remained unknown. Therefore, the molecular mechanisms of actin dynamics mediated by ABPs in root hair formation are largely unknown. The Arabidopsis actin single mutant act2 and act7 seedlings show fewer root hairs, and double mutant act2 act7 seedlings display full defects in root hair formation, indicating that actin cytoskeleton is required for root hair formation [ 3 , 9 – 11 ]). Before root hair formation, F-actin dynamics show that numerous longitudinal filament actin (F-actin) (parallel growth axes) surrounds the nuclei near the end walls in root epidermal cells of the root hair emission region in the wild type (WT) [ 12 , 13 ]. In the root epidermal cells of act2 act7, the thicker and more transversely oriented F-actin bundles or rod-like structures instead of the finer and more longitudinal F-actin in that of WT seedlings [ 10 ]. This indicates that the changes of F-actin architecture including the F-actin orientation and thickness in root epidermal cells are closely related to root hair formation. Root hair is used as an ideal model to study plant cell elongation and differentiation. Root hairs occupy 77% of the root surface area as the critical water and nutrient absorption site during plant growth, development, and stress management [ 1 ]. Under normal conditions, hair cells (H cells), rather than non-hair cells (N cells), in the root epidermal cells rapidly differentiate into root hairs [ 2 – 4 ]. Underwater and nutrient deficiency conditions, plants grow excessively numerous root hairs in response to environmental stresses [ 5 – 8 ]. Results Actin-depolymerizing protein ADF7 is positively involved in root hair formation To explore the role of ADF7 in root hair formation, we first identified T-DNA insert mutants adf7-2 and constructed complementation lines (ADF7-comps) by transforming ADF7 promoter (pADF7)::ADF7 in adf7-2 plants and overexpressing lines (ADF7 OEs) by transforming 35S::ADF7 in Col-0 plants (S1 Fig)). Reverse transcription-PCR (RT-PCR) and RT-quantitative PCR (RT-qPCR) analyses showed no detectable expression in adf7-2, indicating that adf7-2 is a knock-out line, and Col-like expression in ADF7 comp #2 and #7, ~2.3-fold higher expression in ADF7 OE #13, and~3.1-fold higher in ADF7 OE #14 (S1B–S1E Fig). Then, we calculated root hair numbers in Col-0, adf7-2, ADF7 comp #2, ADF7 comp #7, ADF7 OE #13, and ADF7 OE #14, as described previously [4]. Our results showed that Col-0 seedlings grew ~40 root hairs, similar to the previous report [4]. adf7-2 displayed~28 root hairs, respectively. The defects in root hair number in adf7 was rescued in ADF7 comp #2 and #7 (Fig 1A and 1B), confirming that ADF7’s loss-of-function caused the root hair number phenotype in adf7. In contrast, both ADF7 OE #13 and #14 seedlings had ~60 root hairs (Fig 1A and 1B). PPT PowerPoint slide PNG larger image TIFF original image Download: Fig 1. Actin depolymerization protein ADF7 is positively involved in root hair formation and actin bunding protein VLN1 is negatively involved in root hair formation. (A) Images of root hairs from wild type (Col-0), adf7-2, ADF7 comp #2, ADF7 comp #7, ADF7 OE #13, and ADF7 OE #14. Scale bar, 200 μm. OE, overexpression. (B) Histogram depicting root hair number in (A). (C) Quantification of the percentage of root hairs from H and N cells and root hair formation position in H cells in (A). Values given are means ± SD. (D) Images of root hairs from Col-0, vln1-1, vln1-2, VLN1 comp #9, VLN1 comp #14, VLN1 OE #7, and VLN1 OE #8. Scale bar, 200 μm. (E) Histogram depicting root hair number in (D). (F) Quantification of the percentage of root hairs from H and N cells and root hair formation position in H cells in (D). Values given are means ± SD. Significant difference (P< 0.05) indicated by different letters among genotypes is determined for each condition by one-way ANOVA followed by Tukey’s test in (B), (C), (E), and (F). https://doi.org/10.1371/journal.pgen.1010338.g001 Next, we calculated the percentage of root hairs from H and N cells in ADF7 genotype seedlings using the previous method [41]. Col-0 seedlings grew ~99% hairs from H cells and ~1% hair from N cells. Compared with Col-0, adf7-2 showed decreased root hair numbers from H cells and no significant change from N cells, while ADF7 OE #13 and #14 seedlings showed no change in root hair number from H and the increased number from N cells (Fig 1C). These results indicated that ADF7 is positively involved in the differentiation of H and N cells in root hair initiation. To explore root hair formation position from H cells, we calculated the root hair number along a line of H cells in a fixed zone between 2.5 and 3.5 mm from the primary root tips in Col-0 and ADF7 genotype seedlings according to the previous method [42]. The results mean that the density of H cells in the root epidermal cells from various genotype seedlings. Our results showed that Col-0 seedlings grew ~6 root hairs per mm along with H cells (Figs 1C and S2 similar to the previous report [42]. Compared with Col-0, adf7 mutants showed a decreased root hair number along with H cells, whereas ADF7 OE seedlings displayed an increased number (Figs 1C and S2). The results showed that ADF7 is involved in increasing the density of H cells in the root epidermal cells. The results illustrate that ADF7 plays a positive role in root hair formation. Actin-bundling protein VLN1 is negatively involved in root hair formation To explore whether VLN1 is involved in root hair formation, vln1-1, vln1-2, VLN1 comp #9 and #14 and VLN1 OE #7 and #8 were used ([33], S3 Fig) to calculated three parameters including root hair number in roots, the percentage of root hairs from H and N cells, and root hair number along H cells, using the same methods mentioned above. The results showed that vln1 mutants (~ 60 root hairs) grew more root hair numbers than Col-0 (~ 40 root hairs) (Fig 1D and 1E), associated with vln1 displayed a higher percentage of root hairs from N cells and more root hairs along H cells (Figs 1F and S2). These root hair phenotypes of vln1 mutants were restored in VLN1 comp #9 and #14 (Fig 1D–1F). VLN1 OE #7 and #8 seedlings displayed ~ 30 root hairs, associated with a lower percentage of root hairs from H cells and fewer root hairs along H cells (Figs 1 and S2). These results indicate that VLN1 is negatively involved in root hair initiation. ADF7 inhibits the expression and function of VLN1 during root hair formation Next, we investigate whether ADF7 and VLN1 interact during root hair formation. RT-qPCR analysis showed that VLN1 expression in roots was significantly increased in adf7 mutants, but inhibited in ADF7 OE seedlings (Fig 2A). While ADF7 expression in roots wasn’t significantly changed in vln1 mutants and VLN1 OEs (Fig 2B). Additionally, we generated the GUS staining genotype seedlings by introducing the VLN1 promoter (pVLN1)::GUS into adf7-2 and ADF7 OE #13, respectively, and introducing the ADF7 promoter (pADF7)::GUS into vln1-2 and VLN OE #8 by crossing, respectively. We also introduced a pVLN1::VLN1::GFP and a pADF7::ADF7::GFP into vln1-2 and adf7-2, respectively, which results in fully rescuing the root hair phenotypes of vln1-2 and adf7-2, respectively (S4 and S5 Figs), indicating that the GFP fusion protein may be used in protein expression analysis. Analysis of GUS staining also showed that VLN1 doesn’t affect ADF7 expression in roots and ADF7 negatively regulates VLN1 expression in roots (Fig 2C–2H). Further, western blotting supported that ADF7 affects VLN1 expression in roots (Fig 2C–2H). Further, we generated ADF7 and VLN1 double gene genotype seedlings, including adf7 vln1 #1 and #2, and adf7 VLN1 OE #1 and #2 from crossing adf7-2, vln1-2, and VLN1 OE #8 respectively (S6 Fig). Compared with Col-0, adf7 vln1 seedlings displayed more root hair number, similar to single vln1 mutants (Fig 2I and 2J). In adf7 VLN1 OE seedlings had contrasting results, similar to VLN1 OE seedlings (Fig 2I and 2J). The results confirm that ADF7 is upstream of VLN1 in root hair formation, and ADF7 inhibits the expression and function of VLN1 in root hair formation. PPT PowerPoint slide PNG larger image TIFF original image Download: Fig 2. ADF7 inhibits the expression of function VLN1 in root hair formation. (A) RT-qPCR quantification of VLN1 expression level in Col-0, adf7-2, ADF7 comp #2, ADF7 comp #7, ADF7 OE #13, and ADF7 OE #14. (B) RT-qPCR quantification of ADF7 expression level in Col-0, vln1-1, vln1-2, VLN1 comp #9, VLN1 comp #14, VLN1 OE #7, and VLN1 OE #8. (C) GUS analysis of VLN1 expression in root tips from Col-0, adf7-2, and ADF7 OE seedlings. (D) GUS analysis of ADF7 expression in root tips from Col-0, vln1, and VLN1 OE seedlings. (E) Western blotting of VLN1 expression level in Col-0, adf7-2, and ADF7 OE seedlings. Rubisco as a loading control. (F) Quantification of the relative grayscale value in (E). (G) Western blotting of ADF7 expression level in Col-0, vln1, and VLN1 OE seedlings. Rubisco as a loading control. (H) Quantification of the relative grayscale value in (G). (I) Images of root hairs from ADF7 and VLN1 double gene genotypes. Scale bar, 200 μm. (J) Histogram depicting root hair number in (I). Significant difference (P< 0.05) indicated by different letters among different genotypes is determined for each condition by one-way ANOVA followed by Tukey’s test. Values are means ± SD of three independent biological replicates. *** P< 0.001, Student’s t-test compared to Col-0, in (A) and (B), and compared to VLN1::GFP and ADF7::GFP in (F) and (H). https://doi.org/10.1371/journal.pgen.1010338.g002 ADF7 inhibits VLN1-mediated thick bundling activity in epidermal cells of root apices F-actin arrays are correlated with the organization of single actin filament dynamics [47]. Previous findings have shown that ADF7 possesses actin severing and depolymerizing capacities in vitro and in pollen cells, and VLN1 displays a simple mechanism of bundling capacity in vitro and in root hairs [28,32–34]. Therefore, we put our efforts toward characterizing the actin organization capacities of ADF7 and VLN1 in root apices. The observed regions were the elongation zone of root apices (Fig 4). The results showed that ADF7 loss-of-function led to a decline in F-actin severing frequency and depolymerization rate and an increase in bundling frequency, as well as the decreased F-actin turnover based on the increased maximum filament length and maximum filament lifetime, and the decreased severing frequency and depolymerization rate (Fig 4 and S1–S6 Videos). VLN1 mutation showed contrasting results (Fig 4 and S7 Video). Moreover, adf7 vln1 double mutants displayed similar F-actin dynamics to vln1 (Fig 4 and S8 Video). These results illustrate that ADF7 is responsible for organizing single F-actin depolymerizing and severing, and VLN1 functions in single F-actin bundling. The results also indicate that ADF7 inhibits VLN1-mediated single F-actin bundling, which leads to a significant increase of F-actin turnover, in root apices. ADF7-inhibited-VLN1 pathway activates root hair formation by regulating F-actin dynamics in root tips Considering ADF7 and VLN1 can bind to actin in vitro [28,32], and the direct roles of ADF7 and VLN1 in controlling actin dynamics [28,32,48,49], therefore, we proposed that ADF7 and VLN1 might be via directly controlling F-actin dynamics to affect root hair formation. Then, we conducted actin pharmacological experiments. We firstly calculated root hair number in Col-0, adf7-2, ADF7 OE #14, vln1-2, VLN1 OE #8, adf7 vln1 #1, and adf7 VLN1 OE #1 seedlings from 3-d-old seedlings treated with the presence or absence of actin disrupting drug latrunculin-A (Lat A) for 3 h then removed in no drug media for 3-d growth. In Col-0 seedlings, the low concentrations of Lat A (0.2 and 0.4 μM) stimulated significant root hair formation (Fig 5A and 5B and S7). Moreover, Lat A treatments rescued the defects in root hair formation in adf7, VLN1 OE, and adf7 VLN1 OE seedlings (Figs 5A and 5B and S7), illustrating that actin depolymerization activates ADF7-inhibited-VLN1-regulated root hair formation. No significant difference between 0.2 and 0.4 μM Lat A treatments was found (Figs 5A and 5B and S7). PPT PowerPoint slide PNG larger image TIFF original image Download: Fig 5. The F-actin depolymerization regulated by ADF7-inhibited-VLN1 promotes root hair formation. (A) Images of root hairs from Col-0, adf7-2, ADF7 OE #14, vln1-2, VLN1 OE #8, and adf7 vln1 #1 under CK and Lat A treatments (0.2 μM). Scale bar, 200 μm. (B) Histogram depicting root hair number in (A). (C) Confocal microscopy images of epidermal cells visualized by the expression of fABD2::GFP in root apices from Col-0, adf7-2, vln1-2, and adf7 vln1 #1 seedlings under Lat A treatments (0.2 μM). Enlarged views from the red boxes are in the bottom row. Scale bar, 25 μm. (D) Violin plot showing the average and contribution of fluorescence intensity of actin cables in (C). The red line represents the average fluorescence intensity in different genotypes. (E) Histogram depicting skewness of F-actin in (C). (F) Histogram depicting the percentage of occupancy of F-actin in (C). Significant difference (P< 0.05) indicated by different letters among genotypes is determined for each condition by one-way ANOVA followed by Tukey’s test in (B), (D), (E), and (F). https://doi.org/10.1371/journal.pgen.1010338.g005 Next, we observed the actin dynamics in Col-0, adf7, vln1, and adf7 vln1 seedlings under Lat A treatments in root epidermal cells in root tips (Fig 5C). 0.2 μM Lat A treatments led to the decline of fluorescence intensity and skewness parameter and the increase of occupancy percentage, suggesting that actin disrupting drug with low concentrate increased actin depolymerization and inhibited actin thick bundles in Col-0, adf7, vln1, and adf7 vln1 seedlings (Fig 5C–5F). Furthermore, the mild actin depolymerization treatments increased F-actin turnover in roots from Col-0, adf7, vln1, and adf7 vln1 seedlings (Figs 5C–5F and 4D). These results indicate that actin disrupting drug increased F-actin depolymerization, fine F-actin and F-actin turnover in roots, consequently promoting root hair formation in all the seedlings including Col-0, ADF7, and VLN1 genotype seedlings, highlighting that ADF7-inhibited-VLN1 pathway organizes F-actin dynamics including the increase of F-actin depolymerization, fine F-actin amount and F-actin turnover in root tips to promote root hair formation. [END] --- [1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1010338 Published and (C) by PLOS One Content appears here under this condition or license: Creative Commons - Attribution BY 4.0. via Magical.Fish Gopher News Feeds: gopher://magical.fish/1/feeds/news/plosone/