Loss of function of CDPK16 reduces the rate of actin turnover in pollen

To understand the regulation of actin turnover in pollen, we performed forward chemical genetic screening to uncover mutations that alter the sensitivity of pollen germination to latrunculin B (LatB). Our attention was attracted by 1 T-DNA insertion knockout mutant allele of CDPK16, designated as cdpk16-1 (S1A and S1B Fig). Germination of cdpk16-1 pollen is resistant to LatB (see below). To demonstrate that the LatB-resistant pollen germination phenotype is indeed caused by the mutation in CDPK16, we also generated another cdpk16 mutant allele by the CRISPR/cas9 approach [33], designated as cdpk16-2 (S1C Fig). We found that cdpk16-1 and cdpk16-2 mutant pollen germinates better than wild-type (WT) pollen in the presence of LatB (S1D and S1E Fig), which suggests that the germination of cdpk16 mutant pollen is resistant to LatB. Accordingly, we found that the relative growth rate of pollen tubes was increased significantly in cdpk16 mutants compared to WT in the presence of LatB (S1F and S1G Fig), which suggests that loss of function of CDPK16 renders pollen tube growth resistant to LatB. We next found that there is no overt difference in terms of the overall brightness and organization of actin filaments in cdpk16 mutant pollen grains compared to WT in the absence of LatB (S2A–S2C Fig). We noticed that actin filaments became fragmented in both WT and cdpk16 mutant pollen grains after treatment with 150 nM LatB, but the extent of fragmentation is less obvious in cdpk16 mutant pollen grains than in WT (S2A Fig). In addition, we found that LatB-triggered actin depolymerization is inhibited in cdpk16 mutant pollen grains compared to WT, as evidenced by the significantly higher relative amount of actin filaments in cdpk16 mutant pollen grains compared to WT (S2B and S2C Fig). In line with the above observations, we found that overexpression of CDPK16 (S3A Fig) renders pollen germination sensitive to LatB (S3B–S3D Fig). Thus, these data suggest that CDPK16 promotes actin turnover in pollen.

Loss of function of CDPK16 promotes pollen germination and inhibits pollen tube growth

We next determined the role of CDPK16 in regulating pollen germination and pollen tube growth. We initially compared the time course of pollen germination in WT and the 2 cdpk16 mutants, and found that pollen germination is accelerated early on in the cdpk16 mutants, but the overall pollen germination rate in cdpk16 mutants does not differ from that in WT (S4A Fig). This suggests that loss of function of CDPK16 promotes pollen germination. In addition, we found that the rate of pollen tube growth is significantly reduced in cdpk16 mutants compared to WT (S4B and S4C Fig), which suggests that CDPK16 promotes normal pollen tube growth. These data together suggest that CDPK16 maintains the normal rate of pollen germination and promotes pollen tube growth.

Loss of function of CDPK16 induces obvious accumulation of actin filaments at pollen tube tips

Next, we performed real-time visualization of the dynamics of actin filaments decorated with Lifeact-eGFP and found that, consistent with previous observations [5], actin filaments continuously polymerized from the plasma membrane (PM) within the apical region of pollen tubes (Fig 1A and S1 Movie). We found that PM-originated actin filaments are obviously brighter within the apical region of cdpk16 pollen tubes than in WT pollen tubes (Fig 1A and S2 Movie). Kymograph analysis showed that the region occupied by membrane-originated actin filaments was enlarged in cdpk16-1 pollen tubes compared to WT (Fig 1A, right panel). This is supported by physical measurements showing that the distance from the base of the apical actin structure to the pollen tube tip (mean ± SE) increased significantly in cdpk16-1 pollen tubes (6.20 ± 0.047 μm) compared to WT pollen tubes (4.27 ± 0.037 μm) (Fig 1B). Through monitoring the dynamics of individual PM-originated actin filaments (Fig 1C and S3 and S4 Movies), we found that the average severing frequency of actin filaments was significantly reduced in cdpk16-1 pollen tubes compared to WT (Fig 1D). Accordingly, the maximal filament length and lifetime increased significantly in cdpk16-1 pollen tubes compared to WT (Fig 1D). However, we did not notice obvious differences in the elongation and depolymerization rates of PM-originated apical actin filaments in cdpk16-1 pollen tubes compared to WT (Fig 1D). Thus, our results suggest that the amount of apical actin filaments is increased and their dynamics are reduced in cdpk16-1 pollen tubes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Actin dynamics is reduced in cdpk16 pollen tubes.

(A) Time-lapse images of actin filaments decorated with Lifeact-eGFP in growing WT and cdpk16-1 pollen tubes. Yellow brackets indicate the region occupied by membrane-originated actin filaments. The right panels are Kymograph analyses of the growing WT and cdpk16-1 pollen tubes shown in the left panel. The region occupied by the membrane-originated actin filaments is marked by 2 red lines. Bar = 10 μm. (B) Quantification of the width of the region occupied by membrane-originated actin filaments in WT and cdpk16-1 pollen tubes. Data are presented as mean ± SE, **P < 0.01 (Student’s t test). More than 400 time points from 10 pollen tubes were measured. Numerical data underlying this panel are available in S1 Data. (C) Time-lapse images of actin filaments at the apical region in WT and cdpk16-1 pollen tubes. Apical actin filaments are indicated by different colored dots. The severing events of actin filaments are indicated by red arrows. Bar = 5 μm. (D) Dynamic parameters of apical actin filaments in WT and cdpk16-1 pollen tubes. Data are presented as mean ± SE, with the number of filaments in parentheses. NS, no significant difference, **P < 0.01, (Student’s t test). WT, wild type.

https://doi.org/10.1371/journal.pbio.3002073.g001

CDPK16 interacts with and phosphorylates ADF7 both in vitro and in vivo and it enhances the actin-depolymerizing and severing activities of ADF7 in vitro

To determine whether CDPK16 regulates actin turnover via interaction with certain actin-binding proteins (ABPs) in vivo, we performed pull-down experiments followed by mass spectrometry to search for candidate interacting proteins of CDPK16. Interestingly, we found that ADF7 is enriched in the CDPK16 pull-down fraction (Fig 2A–2C). The direct interaction between CDPK16 and ADF7 was confirmed by the luciferase complementation imaging (LCI) assay (Fig 2D) and further validated by showing that CDPK16 can phosphorylate ADF7 in vitro (Fig 2E). To determine whether CDPK16 can phosphorylate ADF7 in vivo, we decided to treat total pollen proteins with phosphatase in the presence and absence of CDPK16, followed by 2D electrophoresis and antibody detection of ADF7. However, the currently available anti-ADF7 antibody cross-reacts with the highly similar ADF10, so we initially analyzed adf7 and adf10 mutants to distinguish ADF7 from ADF10 after electrophoresis (Fig 2F). Comparing the results from WT, adf7 and adf10, it appears that ADF7 might be subject to posttranslational modification, as there are several protein spots corresponding to ADF7 (Fig 2F). Protein spot (a) is the presumed phosphorylated form of ADF7, based on its mobility after 2D electrophoresis (Fig 2F). In support of this speculation, we found that treatment with λ-phosphatase reduced the intensity of spot (a) in total proteins extracted from WT pollen (Fig 2G). Furthermore, we found that the intensity of protein spot (a) is reduced in cdpk16 mutant pollen total extract whereas is increased in pollen total extract from CDPK16 overexpressors compared to WT (Fig 2G), which suggests that CDPK16 can phosphorylate ADF7 in pollen.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. CDPK16 interacts with and phosphorylates ADF7.

(A) Identification of proteins in CDPK16-6×His pull-down fractions by LC-MS/MS. ADF7 is one of the proteins that appears in the pull-down fraction. (B, C) Two representative peptides derived from ADF7 are presented. A full list of the peptides is presented in S2 Data. (D) CDPK16 interacts with ADF7. The interaction between ADF7 and CDPK16 was determined by qualitative analysis of luciferase (LUC) activity using the LCI assay. (E) CDPK16 phosphorylates ADF7 in vitro. The left panel shows the CDPK16-6×His and ADF7 input proteins, which were detected by Coomassie Brilliant blue R 250 staining. The right panel shows phosphorylated ADF7, which was detected by [γ-³²P] ATP autoradiography. The original pictures are available in S1 Raw Images. (F) Detection of ADF7 and ADF10 protein spots by 2D gel-electrophoresis. Total proteins from WT, adf7 and adf10 pollen were separated by 2D gel-electrophoresis. Protein spots were revealed by western blot analysis probed with anti-ADF7 antibody, which also detects ADF10. The original pictures are available in S1 Raw Images. (G) Detection of ADF7 and ADF10 in total proteins extracted from WT and cdpk16 pollen. Total proteins from WT pollen in the presence and absence of λpp or from pollen of cdpk16 mutants and CDPK16 overexpressors were separated by 2D gel-electrophoresis and subjected to western blot analysis probed with anti-ADF7 antibody. (a), (b), and (c) in (F) and (G) represent phosphorylated ADF7, ADF7, and ADF10, respectively. The original pictures are available in S1 Raw Images. ADF, actin-depolymerizing factor; LCI, luciferase complementation imaging; LC-MS/MS, liquid chromatography–mass spectrometry/mass spectrometry; WT, wild type.

https://doi.org/10.1371/journal.pbio.3002073.g002

Next, using a high-speed F-actin co-sedimentation assay, we found that CDPK16 increased the activity of ADF7 in depolymerizing actin filaments in vitro (Fig 3A and 3B), and this enhancement by CDPK16 was reduced in the absence of Ca²⁺ (Fig 3C and 3D). These results suggest that CDPK16 promotes the actin-depolymerizing activity of ADF7 in a Ca²⁺-dependent manner. We previously showed that the actin disassembly activity of ADF4 is regulated by phosphorylation in stomata [30]. Interestingly, we found that CDPK16 only weakly, albeit significantly, enhanced the actin-depolymerizing activity of ADF4 (S5 Fig), which suggests that CDPK16 promotes the actin-depolymerizing activity of ADF7 in a somewhat specific manner. In addition, we found that CDPK16 enhanced the activity of ADF7 in shortening actin filaments in vitro (Fig 3E and 3F). As ADF/cofilin also contributes to actin depolymerization via fragmentation of actin filaments [34] and ADF7 can sever actin filaments [10], we determined whether CDPK16 can promote the severing activity of ADF7. We used total internal reflection fluorescence microscopy (TIRFM) to show that CDPK16 enhanced the activity of ADF7 in severing actin filaments (Fig 3G and 3H and S5–S8 Movies). Thus, our study suggests that CDPK16 enhances ADF7-mediated actin depolymerization and severing in vitro.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. CDPK16 enhances the activity of ADF7 in a Ca²⁺-dependent manner in vitro.

(A) SDS-PAGE analysis of the protein samples from a high-speed F-actin co-sedimentation experiment in the presence of Ca²⁺. F-actin, 3 μM; ADF7, 20 μM; CDPK16 (+), 1.0 μM; CDPK16 (++), 2.5 μM; CDPK16 (+++), 5.0 μM. The supernatant fractions (S) and pellets (P) were separated on SDS-PAGE gels, and proteins were detected by Coomassie Brilliant blue R 250 staining. The original pictures are available in S1 Raw Images. (B) Quantification of the amount of actin in the supernatant fractions shown in (A). Data are presented as mean ± SE, n = 3, *P < 0.05 and **P < 0.01 by Student’s t test. Numerical data underlying this panel are available in S3 Data. (C) SDS-PAGE analysis of the protein samples from a high-speed F-actin co-sedimentation experiment in the absence of Ca²⁺. The conditions were exactly the same as in (A) except for the presence of 0.5 mM EGTA instead of 0.5 mM CaCl₂ in the kinase reaction buffer. F-actin, 3 μM; ADF7, 20 μM; CDPK16 (+), 1.0 μM; CDPK16 (++), 2.5 μM; CDPK16 (+++), 5.0 μM. The original pictures are available in S1 Raw Images. (D) Quantification of the amount of actin in the supernatant fractions shown in (C). Data are presented as mean ± SE, n = 3, ns, no significant difference by Student’s t test. Numerical data underlying this panel are available in S3 Data. (E) Images of actin filaments stained with equimolar Rhodamine-Phalloidin. The concentration of actin filaments is 2 μM. Bar = 10 μm. (F) Quantification of the length of actin filaments. Data are presented as mean ± SE, n = 3, ns, no significant difference, **P < 0.01 by Student’s t test. Numerical data underlying this panel are available in S3 Data. (G) Time-lapse images of actin filaments. F-actin, 150 nM (50% Oregon green-labeled); ADF7, 500 nM; CDPK16, 125 nM. The red arrows indicate actin filament severing events. Bar = 10 μm. (H) Quantification of the average severing frequency of actin filaments. Data are presented as mean ± SE, ns, no significant difference, *P < 0.05 by Student’s t test. Numerical data underlying this panel are available in S3 Data. ADF, actin-depolymerizing factor; CDPK, calcium-dependent protein kinase.

https://doi.org/10.1371/journal.pbio.3002073.g003

Overexpression of ADF7 alleviates the LatB-resistant pollen germination phenotype in cdpk16 and loss of function of CDPK16 enhances the LatB-resistant pollen germination phenotype in adf10

To determine whether CDPK16 promotes actin turnover through activating ADF7, we tested whether gain of function of ADF7 can alleviate the actin turnover defects in cdpk16 mutant pollen. Indeed, we found that overexpression of ADF7 suppressed the actin turnover phenotype in cdpk16 mutant pollen (S6A and S6B Fig). ADF7 and ADF10 act redundantly to control actin turnover in pollen [11], and our results above suggest that ADF7 might be the more relevant substrate of CDPK16 in vivo (Fig 2F and 2G). Therefore, we wondered whether loss of function of CDPK16 will cause an additive effect on actin turnover induced by loss of function of ADF10 in pollen. As expected, we found that the LatB-resistant pollen germination phenotype is more severe in adf10 cdpk16-1 double mutants adf10 single mutants (S6C and S6D Fig). These data together suggest that CDPK16 promotes actin turnover at least partly through up-regulating ADF7 activity in pollen.

CDPK16 enhances the actin-depolymerizing activity of ADF7 by phosphorylating its Ser128

To uncover the phosphorylation site(s) of ADF7 in vivo, we performed mass spectrometry analysis on 8His-ADF7 pulled down from total proteins extracted from pollen. An ADF7 phospho-peptide was repeatedly identified with the phosphate group conjugated to Ser128 (Fig 4A). Strikingly, we found that CDPK16 adds phosphate to the same serine of ADF7 in vitro (Fig 4B). This suggests that Ser128 in ADF7 might be the major site of phosphorylation by CDPK16. In support of this notion, we found that the extent of CDPK16-mediated ADF7 phosphorylation in vitro was reduced significantly after Ser128 was replaced with aspartic acid (ADF7^S128D) (Fig 4C and 4D). Furthermore, we found that the charge behavior of ADF7^S128D and non-phosphorylatable ADF7^S128A (Ser128 replaced with Alanine) is similar to that of the predicted phosphorylated ADF7 and non-phosphorylated ADF7 (Fig 4E), respectively. Ser128 is conserved in all the class II ADFs in Arabidopsis, i.e., ADF7, ADF8, ADF10, and ADF11 (S7 Fig) and in class II ADFs from other plant species (S8 Fig). We next generated a poly-clonal antibody against this phospho-peptide, designated as anti-phospho-ADF7(Ser128), and found that it specifically recognizes CDPK16-phosphorylated ADF7 (S9A Fig). Interestingly, we found that CDPK16 phosphorylates ADF7 in a Ca²⁺-dependent manner (S9B and S9C Fig). Importantly, we found that it recognizes 8His-ADF7 pulled down from total pollen extract, and treatment with λ-phosphatase reduced the signal (S9D Fig), which further suggests that phosphorylation of ADF7 at Ser128 does occur in vivo.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. CDPK16 can phosphorylate Ser128 in ADF7, and ADF7^S128D has enhanced actin-depolymerizing and severing activity in vitro.

(A) Identification of a phosphorylated peptide with the phosphate group conjugated to Ser128. 8His-ADF7 protein was isolated from pollen grains derived from proADF7::8His-gADF7; adf7 and subjected to LC-MS analysis. The original LC-MS data underlying this panel are available in S4 Data. The original pictures are available in S1 Raw Images. (B) CDPK16 can phosphorylate Ser128 in ADF7 in vitro. After incubation of 20 μM ADF7 with 5 μM CDPK16 in kinase buffer for 30 min, the sample was separated by SDS-PAGE. The ADF7 band was cut out and subjected to mass spectrometry analysis. A phosphorylated ADF peptide with the phosphate group conjugated to Ser128 was identified. The original LC-MS data underlying this panel are available in S4 Data. The original pictures are available in S1 Raw Images. (C) CDPK16 fails to phosphorylate ADF7^S128D-6×His. Purified ADF7-6×His and ADF7^S128D-6×His were incubated with CDPK16 in a kinase reaction buffer and the phosphorylation signals were detected by [γ-³²P] ATP autoradiography. The original pictures are available in S1 Raw Images. (D) Quantification of the relative phosphorylation level of ADF7 shown in (C). The data are presented as mean ± SE, n = 3, **P < 0.01 (Student’s t test). Numerical data underlying this panel are available in S4 Data. (E) 2D electrophoresis assay. Total proteins from mature pollen of WT, proADF7::gADF7^S128A; adf7 and proADF7::gADF7^S128D; adf7 were subjected to 2D electrophoresis analysis. Immunoblotting was performed and probed with anti-ADF7 antibody. (a) Indicates the presumed phosphorylated ADF7 or ADF7^S128D and (b) indicates unphosphorylated ADF7 or ADF7^S128A. The original pictures are available in S1 Raw Images. (F) SDS-PAGE analysis of protein samples from a high-speed F-actin co-sedimentation experiment. F-actin, 3 μM; ADF7, 20 μM; ADF7^S128A, 20 μM; ADF7^S128D, 20 μM. The original pictures are available in S1 Raw Images. (G) Quantification of the amount of actin in the supernatant fractions in (F). Data are presented as mean ± SE, n = 3, ns, no significant difference, *P < 0.05 by Student’s t test. Numerical data underlying this panel are available in S4 Data. (H) Kinetic actin filament depolymerization assay. Actin filaments were depolymerized more rapidly by ADF7^S128D than by ADF7 and ADF7^S128A. Briefly, 5 μM pre-clarified ADF7, ADF7^S128A, or ADF7^S128D were incubated with 5 μM preassembled actin filaments (50% NBD-labeled) for 2 min. The mixtures were subsequently diluted 25-fold into buffer G and actin depolymerization was monitored by measuring the changes in fluorescence. Numerical data underlying this panel are available in S4 Data. (I) Images of actin filaments stained with Rhodamine-Phalloidin. The concentration of preassembled actin filaments is 2 μm. Bar = 10 μm. (J) Quantification of the average length of actin filaments shown in (I). The average length of actin filaments was plotted, and the data are presented as mean ± SE, n = 3, ns, no significant difference, **P < 0.01 (Student’s t test). Numerical data underlying this panel are available in S4 Data. ADF, actin-depolymerizing factor; CDPK, calcium-dependent protein kinase; LC-MS, liquid chromatography–mass spectrometry; WT, wild type.

https://doi.org/10.1371/journal.pbio.3002073.g004

We next examined the actin-depolymerizing activity of Ser128 mutants of ADF7. Compared to WT ADF7, ADF7^S128D had enhanced actin-depolymerizing activity and ADF7^S128A had roughly similar activity, as determined by the high-speed F-actin co-sedimentation assay (Fig 4F and 4G) and the kinetic actin-depolymerizing assay (Fig 4H). This was further confirmed by direct visualization of actin filaments (Fig 4I and 4J). Importantly, we found that CDPK16 failed to enhance the actin-depolymerizing activity of ADF7^S128A in vitro (S10 Fig), which suggests that Ser128 of ADF7 is the major residue targeted by CDPK16. These data together suggest that Ser128 in ADF7 is phosphorylated by CDPK16, and phosphorylation of this residue increases ADF7 activity.

ADF7^S128A and ADF7^S128D fail to fully replace the role of ADF7 in promoting actin turnover in vivo

Next, we determined the biological significance of this phospho-regulation mechanism by introducing the non-phosphorylatable ADF7^S128A and phospho-mimetic ADF7^S128D into pollen. Considering that ADF7 and ADF10 redundantly regulate actin turnover in pollen [11], we assumed that the functional difference between ADF7^S128A and ADF7 or ADF7^S128D and ADF7 might be amplified in the adf10 mutant background. We selected transgenic lines containing comparable amounts of ADF7^S128A, ADF7^S128D, and ADF7 both transcriptionally (S11A Fig) and translationally (S11B and S11C Fig) for subsequent analyses. We initially found that ADF7^S128A functions almost the same as ADF7 in supporting pollen tube growth, whereas ADF7^S128D has slightly but significantly higher activity than ADF7 and ADF7^S128A in this regard (S11D and S11E Fig). However, in the presence of LatB, we found that pollen tubes harboring both ADF7^S128A and ADF7^S128D grew significantly faster than pollen tubes harboring ADF7 (S11D and S11F Fig), which suggests that ADF7^S128A and ADF7^S128D have reduced activity in promoting actin turnover in pollen tubes compared to ADF7. It is quite puzzling that ADF7^S128D has reduced capability in promoting actin turnover in pollen compared to ADF7, as ADF7^S128D has higher activity in depolymerizing and severing actin filaments than ADF7 in vitro (Fig 4F–4J). The reasonable explanation here is that ADF7^S128D cannot fully mimic the function of phosphorylated ADF7 in pollen. Strikingly, we found that the germination of pollen harboring ADF7^S128A is resistant to LatB compared to pollen harboring WT ADF7 when CDPK16 is overexpressed (S12 Fig). This suggests that phosphorylation of ADF7 at its Ser128 mainly accounts for the role of CDPK16 in promoting actin turnover in pollen.

To directly visualize the effect of phosphorylation of ADF7 at its Ser128 on the actin cytoskeleton in pollen tubes, we directly visualized the actin cytoskeleton in pollen tubes after staining with Alexa-488 phalloidin. We found that actin filaments were overall brighter in proADF7::gADF7^S128A; adf7 adf10 and proADF7::gADF7^S128D; adf7 adf10 pollen tubes than in adf10 and proADF7::gADF7; adf7 adf10 pollen tubes (Fig 5A), and the increase in the amount of actin filaments is much more obvious within the apical and subapical regions of pollen tubes (Fig 5A and 5B). Accordingly, the average fluorescence intensity of phalloidin staining within the apical and subapical regions of proADF7::gADF7^S128A; adf7 adf10 and proADF7::gADF7^S128D; adf7 adf10 pollen tubes is significantly higher than in adf10 and proADF7::gADF7; adf7 adf10 pollen tubes (Fig 5C–5E), whereas no significant difference was detected in the shank region (Fig 5C and 5F). The actin cytoskeleton exhibits severe turnover defects in the apical and subapical regions but not in the shank region of pollen tubes harboring ADF7^S128A and ADF7^S128D, which suggests that phosphorylation of ADF7 at Ser128 has biologically meaningful consequence within apical and subapical regions that have high concentration of cytosolic Ca²⁺. This, on the other hand, suggests that CDPK16-mediated regulation of ADF7 is biological significant in pollen tubes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. ADF7^S128A has less activity than ADF7 in restoring the actin turnover defects caused by loss of ADF7 in pollen tubes.

(A) Images of actin filaments in pollen tubes. Pollen tubes derived from adf10, proADF7::gADF7; adf7 adf10, proADF7::gADF7^S128A; adf7 adf10 and proADF7::gADF7^S128D; adf7 adf10 were subjected to actin staining with Alexa-488 phalloidin. The projection images and the associated optical sections are presented. Several transverse sections derived from the same pollen tube are shown in the right panels; the distance of these sections from the pollen tube tip is indicated in the images. The dashed red lines indicate the base of the apical region (3 μm from tube tip), and the dashed yellow lines indicate the base of the subapical region (10 μm from tube tip). Bars = 5 μm. (B) The 3D distribution of fluorescence intensity of actin filaments within the apical region (0–3 μm from tube tip), which was generated by ImageJ software with a 3D interactive “Surface Plot” function. Warm and cold colors indicate higher and lower fluorescence, respectively. Numerical data underlying this panel are available in S5 Data. (C) Quantification of the fluorescence intensity of actin filaments stained with Alexa-488 phalloidin in pollen tubes. The fluorescence intensity (mean ± SEM) was plotted from the tip to the base along the growth axis of pollen tubes. The dashed red lines indicate the base of the apical region (3 μm from tube tip), and the dashed yellow lines indicate the base of the subapical region (10 μm from tube tip). More than 40 pollen tubes were measured. Data are presented as mean ± SEM. Numerical data underlying this panel are available in S5 Data. (D–F) Quantification of the fluorescence intensity of Alexa-488 phalloidin within 3 different regions of pollen tubes. Data are presented as mean ± SEM, n = 3, ns, no significant difference, *P < 0.05, **P < 0.01 by Student’s t test. Numerical data underlying this panel are available in S5 Data. ADF, actin-depolymerizing factor.

https://doi.org/10.1371/journal.pbio.3002073.g005

CDPK16 mainly localizes to the plasma membrane in pollen tubes

To determine the intracellular localization of CDPK16 in pollen tubes, we generated a CDPK16-eGFP fusion construct with its expression under the control of the CDPK16 native promoter. CDPK16-eGFP can rescue the LatB-resistant pollen germination phenotype of cdpk16 mutants (S13 Fig), which suggests that CDPK16-eGFP is functional. We found that CDPK16 is mainly localized in an inner circle close to the border of ungerminated pollen grains (Fig 6A). During the onset of pollen germination, the CDPK16-eGFP signal is reduced at the germination aperture and is subsequently enriched in the region near the PM of the emerging pollen tube (Fig 6A and S9 Movie). We also examined the intracellular localization of CDPK16-eGFP in late-stage pollen tubes and found that it mainly localized to the PM, with the strongest signal in the subapical region (Fig 6B and S10 Movie). CDPK16-eGFP also forms small dots within the cytoplasm of pollen tubes (Fig 6B). We do not currently know what those structures are. The PM localization of CDPK16-eGFP was also confirmed by covisualization of the fluorescent lipophilic dye FM4-64. CDPK16-eGFP colocalized with FM4-64, and the eGFP signal was obviously stronger at the subapical region of the pollen tube (Fig 6C). We also found that CDPK16-eGFP is localized to the nucleus (Fig 6B and 6C). Our intracellular localization data are actually consistent with a previous finding that CDPK localizes to membranes via myristoylation and palmitoylation [35].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Subcellular localization of CDPK16.

(A) Intracellular localization of CDPK16-eGFP in pollen grains during germination. Yellow triangles indicate the germination aperture, and white circles indicate the growth direction of the germinating pollen tube. Bar = 10 μm. (B) Intracellular localization of CDPK16-eGFP in pollen tubes. The upper panel is the time-lapse images showing the intracellular localization of CDPK16-eGFP in growing pollen tubes. Bar = 10 μm. The lower panel shows transverse sections at 240 s. Bar = 5 μm. Asterisks indicate the growth direction of the pollen tube, and red brackets indicate the apical region with less CDPK16-eGFP signals. The dashed yellow lines indicate the base of the subapical region. The red triangles indicate the membrane of the vegetative nucleus and the yellow triangles indicate the PM. The distance of transverse sections from the tip is indicated above the images. (C) Covisualization of CDPK16-eGFP with FM4-64 in the pollen tube. FM4-64 labels the PM but not internal membranes. The right panel shows transverse sections with their distance from the tip indicated in the images. White triangles indicate the PM and blue triangles indicate the membrane of the vegetative nucleus. Bars = 10 μm. CDPK, calcium-dependent protein kinase; PM, plasma membrane.

https://doi.org/10.1371/journal.pbio.3002073.g006

Plant materials and growth conditions

The T-DNA insertion mutant of CDPK16, Salk_052257, was obtained from the Arabidopsis Biological Resource Center (ABRC) and was designated as cdpk16-1. Another mutant of CDPK16 designated as cdpk16-2 was generated by the CRISPR/Cas9 system (see below). Information on the adf7, adf10 and adf7 adf10 mutants has been described previously [10,11]. ADF7 and CDPK16 overexpressors were generated as described below. The adf10 cdpk16-1 double mutants were generated by crossing adf10 with cdpk16-1. To examine the effect of ADF7 gain-of-function on cdpk16, cdpk16-1 was crossed with ADF7 overexpressors. To observe actin dynamics in cdpk16 pollen tubes, the probe Lifeact-eGFP was introduced into cdpk16-1 via crossing with WT Arabidopsis plants expressing Lat52::Lifeact-eGFP [5]. ADF7 genomic DNA was amplified using the primer pair PG3/PG4 (S1 Table). This PCR product, with 8His fused to the N-terminus of ADF7, was moved into pCAMBIA1301 to generate pCAMBIA1301-pADF7-8His-gADF7. The pCAMBIA1301-pADF7-8His-gADF7 plasmid was subsequently transformed into adf7 using the Agrobacterium-mediated floral dip method [53]. The transgenic plants were designated as proADF7::8His-gADF7; adf7 and were used at T3. Arabidopsis Columbia-0 ecotype (Col-0) was used as WT. Arabidopsis plants were grown in media or soil at 22°C under a 16-h-light/8-h-dark photoperiod.

Generation of the CDPK16 knockout mutant using the CRISPR/Cas9 system in Arabidopsis

The egg cell-specific promoter-controlled (EPC) CRISPR/Cas9 system was employed to generate CDPK16 loss-of-function mutants as described previously [33]. Briefly, the plasmid pHEE401-CDPK16 was constructed using primer pairs PG3/PG4 and PG5/PG6 (S1 Table) and transformed into WT Arabidopsis plants to generate CDPK16 loss-of-function mutants. U6_26p-CDPK16_sgRNA1-gRNA_Sc-U6_26t-U6_29p-CDPK16_sgRNA2-gRNA_Sc-U6_26t was included in the plasmid pHEE401-CDPK16. Total DNA was extracted from the leaves of T1 transgenic plants and the target sequence was then amplified by primer pair PG7/PG8 (S1 Table). Sequencing was performed to verify the mutation. The CDPK16 loss-of-function mutant generated by CRISPR/Cas9 was designated as cdpk16-2.

Creation of ADF7 and CDPK16 overexpressors

To generate ADF7 overexpression transgenic plants, the coding region sequence (CDS) of ADF7 was amplified with primer pair O1/O2 (S1 Table) using pET28a-ADF7 [10] as the temperate. The ADF7 CDS was moved into pCAMBIA1301-Lat52 to generate pCAMBIA1301-Lat52-ADF7. To generate CDPK16 overexpressors, the CDPK16 CDS was amplified with primer pair O3/O4 (S1 Table) and subsequently moved into pCAMBIA1301-Lat52 to generate pCAMBIA1301-Lat52-CDPK16. The plasmid pCAMBIA1301-Lat52-ADF7 or pCAMBIA1301-Lat52-CDPK16 was transformed into Arabidopsis using the Agrobacterium-mediated floral dip method [53]. To generate CDPK16 overexpressors in the proADF7::gADF7; adf7 adf10 and proADF7::gADF7^S128A; adf7 adf10 backgrounds, the fragment of CDPK16 CDS fused with the Lat52 promoter at its N-terminus was amplified from pCAMBIA1301-Lat52-CDPK16 plasmid with primer pair O5/O6 (S1 Table) and subsequently moved into pK7FWG2 to generate pK7FWG2-Lat52-CDPK16. The plasmid was then transformed into proADF7::gADF7; adf7 adf10 and proADF7::gADF7^S128A; adf7 adf10 transgenic plants. T3 homozygous transgenic plants were used for subsequent analysis.

Complementation of cdpk16 mutants and visualization of the subcellular localization of CDPK16 in pollen tubes

To complement cdpk16 mutants, the genomic DNA sequence of CDPK16 containing a 3.4-kb predicted promoter sequence was amplified with primer pair PG1/PG2 (S1 Table). The DNA fragment was cloned into pEasy-Blunt to generate pEasy-Blunt-pgCDPK16 and was subsequently moved into pCAMBIA1301 digested with PstI/KpnI to generate the complementary plasmid pCAMBIA1301-pgCDPK16. Subsequently, eGFP was moved into pCAMBIA1301-pgCDPK16 digested with SacI/EcoRI to generate the plasmid pCAMBIA1301-pgCDPK16-eGFP. The plasmid pCAMBIA1301-pgCDPK16-eGFP was transformed into cdpk16-1 and cdpk16-2 to generate transgenic plants designated as proCDPK16::gCDPK16-eGFP; cdpk16-1 and proCDPK16::gCDPK16-eGFP; cdpk16-2, respectively. To determine the subcellular localization of CDPK16, proCDPK16::gCDPK16-eGFP; cdpk16-2 pollen tubes were observed under an Olympus FluoView FV1200 confocal microscope equipped with a 100× oil immersion objective (1.4 numerical aperture). Samples were excited under a 488-nm argon laser with the emission wavelengths set at 505 to 545 nm. The z-series images were collected with the z-step size set at 0.5 μm.

Determination of the effect of amino acid substitution at serine-128 of ADF7 in pollen tubes

To assay the effect of single amino acid substitution at Serine-128 of ADF7, the plasmid pEasy-Blunt-pgADF7 was used as the template to perform PCR amplification with the primer pairs M1/M2 and M3/M4 (S1 Table) to generate pEasy-Blunt-pgADF7 ^S128A and pEasy-Blunt-pgADF7^S128D, respectively. The WT and mutant inserts were subsequently moved into pFGC5941 to generate the plasmids pFGC5941-pgADF7, pFGC5941-pgADF7^S128A, and pFGC5941-pgADF7^S128D. The plasmids pFGC5941-pgADF7 and pFGC5941-pgADF7^S128A and pFGC5941-pgADF7^S128D were transformed into adf7-/-adf10+/- using the Agrobacterium-mediated floral dip method [53]. After self-segregation, adf7 adf10 plants containing the homozygous pFGC5941-pgADF7, pFGC5941-pgADF7^S128A, and pFGC5941-pgADF7^S128D were obtained, and they were designated as proADF7::gADF7; adf7 adf10 and proADF7::gADF7^S128A; adf7 adf10, respectively.

RNA extraction and qRT-PCR analysis

Total RNA was extracted from Arabidopsis mature pollen with a Total RNA Extraction Kit (Promega, LS1040). Subsequently, total RNA together with Primers Oligo (dT)₁₈ and M-MLV reverse transcriptase (Promega, M1075) were used for reverse transcription to synthesize cDNA. Quantitative real-time PCR (qRT-PCR) was conducted using 2× RealStar Green Power Mixture with ROX II (GenStar, A314-10). To determine the transcript levels of CDPK16 in WT, CDPK16 overexpressors and the cdpk16-1 mutant, a partial coding region sequence (CDS) of CDPK16 was amplified with primer pair R3/R4 (S1 Table) by qRT-PCR. eIF4A was amplified as an internal control with primer pair R1/R2 (S1 Table). To confirm the transcript levels of ADF7 in proADF7::gADF7; adf7 adf10, proADF7::gADF7^S128A; adf7 adf10 and gADF7^S128D;adf7 adf10 plants, a partial CDS of ADF7 was amplified with primer pair R5/R6 (S1 Table) by qRT-PCR. The relative amounts of ADF7 and CDPK16 transcripts were quantified by the 2^-ΔΔCt method [54].

Western blot analysis and quantification of the relative amount of ADF7 and ADF7^S128A protein in pollen

To determine the amount of ADF7 and its variant ADF7^S128A or ADF7^S128D in pollen, total proteins were isolated from Arabidopsis pollen grains derived from proADF7::gADF7; adf7 adf10, proADF7::gADF7^S128A; adf7 adf10 and proADF7::gADF7^S128D; adf7 adf10 plants according to previously published methods [55,56]. ADF7 and its variant ADF7^S128A were detected in pollen by probing with anti-ADF antibody (ADF7 polyclonal antibody) as described previously [24], and the relative amount of ADF7 was normalized to the amount of UGPase probed with anti-UGPase antibody (Agrisera, AS05086).

Two-dimensional electrophoresis assay

Two-dimensional (2D) electrophoresis assay was performed roughly according to a previously published method [30]. Briefly, total proteins were isolated from mature pollen grains derived from WT, adf7 (Salk_024576) [10], adf10 [11], proADF7::gADF7^S128A; adf7, proADF7::gADF7^S128D; adf7 cdpk16 and CDPK16 overexpressor plants, as described previously [55,56]. Briefly, mature pollen grains were finely ground in liquid nitrogen and the powder was mixed with protein extraction buffer [40 mM Tris-HCl (pH 7.5), 60 mM DTT, 2% SDS, cOmplete, EDTA-free Protease Inhibitor Cocktail (Sigma-Roche, 4693159001), and 1% Phosphatase Inhibitor Cocktail 2 (Sigma-Aldrich, P5726)]. For phosphatase treatment, total proteins were extracted in protein extraction buffer without cOmplete, EDTA-free Protease Inhibitor Cocktail 2. The proteins were then incubated with Lambda Protein Phosphatase (λpp, New England Biolabs, P0753S) at 30°C for 30 min. Then, the mixtures were incubated at 95 to 100°C for 5 min and centrifuged at 12,000 rpm for 10 min. The total pollen proteins were in the supernatant. The proteins were further purified using the 2-D Clean-Up kit and protein pellets were obtained (GE Healthcare, 80-6484-51). The pellets were re-dissolved with 2-DE buffer (8 M Urea, 2.5 M Thiourea, 4% CHAPS, 65 mM DTT, and 1% IPG buffer) and centrifuged at 13,000 g for 10 min. Then, 20 μl supernatant was separated and mixed with 115 μl Hydration buffer (8 M Urea, 2% CHAPS, 65 mM DTT, 1% IPG buffer, and 0.001% Bromophenol blue). Subsequently, isoelectric focusing and 15% SDS-PAGE were performed in a Protein I12 system chamber (Bio-Rad). After electrophoresis, immunoblot analysis was performed with anti-ADF7 antibody [24].

Production of anti-phospho-ADF7(Ser128) antibody and detection of phosphorylated ADF7 in pollen

To generate the poly-clonal antibody that specifically recognizes the ADF7 phosphorylated at Ser128, a phosphorylated peptide (ELDGIQVELQATDPSEM(P)SFDIIK) was synthesized and used to immunize rabbit to generate the antibody designated as anti-phospho-ADF7(Ser128). The specificity of the antibody was examined by performing western blot analysis after incubation of ADF7 with CDPK16 in the presence or absence of calcium. To detect the phosphorylated ADF7 in pollen, total proteins were isolated from pollen derived from proADF7::8His-gADF7; adf7 plants. To perform phosphatase treatment, Lambda Protein Phosphatase (λpp, New England Biolabs, P0753S) was added into the extraction buffer. The total protein extract was subsequently incubated with Ni-NTA agarose that was washed extensively with protein extraction buffer after centrifugation. The protein bound to the Ni-NTA agarose was separated by 15% SDS-PAGE and subjected to western blot analysis probed with Anti-phospho-ADF7(Ser128). The relative amount of loaded protein was determined by performing western blot analysis using anti-ADF7 antibody [24]. The total pollen extract without λpp treatment was used as the control.

In vitro Arabidopsis pollen germination and pollen tube growth

In vitro Arabidopsis pollen germination and pollen tube growth assays were performed as described previously [57]. Briefly, pollen grains were cultured under moist conditions at 28°C on pollen germination medium [PGM: 1 mM MgSO₄, 1 mM CaCl₂, 1 mM Ca(NO₃)₂, 0.01% (w/v) H₃BO₃, and 18% (w/v) Suc, and 0.8% (w/v) agarose (pH 6.9 to 7.0)]. To determine the effect of Latrunculin B (LatB, Sigma-Aldrich, L5288) treatment on pollen germination, pollen grains were cultured on GM in the absence or presence of 1.5 nM or 3 nM LatB for 3 h and observed under an Olympus CX21 microscope equipped with a 10× objective. More than 500 pollen grains were counted in each experiment and the experiments were conducted at least 3 times. To determine the effect of LatB on the growth of pollen tubes, liquid PGM in the absence or presence of 3 nM LatB was added onto the surface of solid GM. To determine the velocity of pollen tube growth, the length of pollen tubes was measured at 2 different time points and the extension was divided by the time interval to yield the average pollen tube growth rate. More than 30 pollen tubes were measured in each experiment and the experiments were repeated at least 3 times.

Screening for Arabidopsis T-DNA insertion mutants with a LatB-resistant pollen germination phenotype

Confirmed Arabidopsis homozygous T-DNA insertion lines were obtained from the Nottingham Arabidopsis Stock Centre (NASC). To identify genes involved in the regulation of actin turnover in pollen, the T-DNA insertions lines were screened for a LatB-resistant pollen germination phenotype. Lines were selected in which the T-DNA insertion was located in genes that showed a reasonable level of expression in pollen, based on data from www.genevestigator.com. To identify the LatB-resistant pollen germination phenotype of the T-DNA insertion lines, 1.5 nM LatB was included in solid PGM. The detailed procedure for Arabidopsis pollen germination is described above.

F-actin staining in fixed Arabidopsis pollen grains and pollen tubes

Arabidopsis pollen grains and pollen tubes were subjected to actin staining with Alexa-488 phalloidin (Thermo Fisher Scientific, A12379) as previously reported [58,59]. To determine the effect of LatB on the actin cytoskeleton, pollen grains or pollen tubes were treated with 150 nM LatB in liquid GM for 30 min before fixation with 300 μM 3-maleimidobenzoic acid N-hydroxysuccinimide ester crystalline (Sigma-Aldrich, M2786) and subsequent staining with Alexa-488 phalloidin. The samples were visualized under an Olympus FluoView FV1200 confocal microscope equipped with a 100× oil immersion objective (1.4 numerical aperture) and were excited under a 488-nm argon laser with the emission wavelength set at 505 to 545 nm. The z-series images were collected with the step size set at 0.5 μm. The amount of F-actin in pollen grains and pollen tubes was quantified by measuring the intensity of the fluorescent pixels with ImageJ software (http://rsbweb.nih.gov/ij/; version 1.48g) as reported previously [10].

Visualization and quantification of the dynamics of actin filaments in pollen tubes

To observe the dynamics of actin filaments in pollen tubes, Lifeact-eGFP was introduced into pollen tubes as described previously [5]. Briefly, pollen tubes harboring Lifeact-eGFP were observed under the Olympus IX83 microscope equipped with a 60× oil immersion objective (1.35 numerical aperture). Time-lapse z-series images were acquired with an Andor Revolution XDh spinning disk confocal system using MetaMorph software (Molecular Devices) at time intervals of 3 s and the z-step size set at 0.7 μm. To quantify the growth rate of pollen tubes and the fluorescence intensity of apical actin filaments simultaneously, a kymograph was produced using ImageJ software as previously reported [60]. To quantify the dynamics of individual actin filaments in pollen tubes, the parameters associated with them (actin filament elongation rate, actin filament depolymerization rate, maximum filament length, maximum filament lifetime, and severing frequency) were measured as described previously using ImageJ software [5,10].

FM dye staining of living pollen tubes

Pollen tubes were stained with the Lipophilic Dye FM4-64 (Thermo Fisher Scientific, T13320). Staining of pollen tubes was achieved by adding FM4-64 dye (2.5 μM in liquid pollen germination medium) onto the surface of solid pollen germination medium. After incubation for 5 min, z-series images were collected by Olympus IX83 spinning disc confocal microscopy with the z-step size set at 0.7 μm. FM4-64 dye was excited with an argon laser at 561 nm, and the emission wavelength was set in a range of 600 to 650 nm.

Luciferase complementation imaging (LCI)

To determine the interaction between CDPK16 and ADF7, the LCI assay was performed as described previously [61,62]. The coding region sequences of CDPK16 and ADF7 were fused with the N-terminus of LUC (nLUC) and the C-terminus of Luc (cLUC), respectively. The plasmids pCAMBIA1300-CDPK16-nLUC and pCAMBIA1300-cLUC-ADF7 were transformed into Agrobacterium tumefaciens strain GV3101. Bacterial cells were centrifuged and re-suspended in infiltration buffer (10 mM MES-KOH (pH 5.7), 10 mM MgCl₂, 150 μM acetosyringone). After adjusting the OD600 of the bacterial suspensions to 0.6, the bacterial cells were subsequently co-infiltrated into Nicotiana. benthamiana leaves. After incubation at room temperature for 48 to 60 h, the luciferase (LUC) activity was measured with a cooled CCD imaging apparatus (Andor iXon, Andor Technology, Belfast, United Kingdom). The empty vector was used as a control.

Protein production

The coding region sequence of CDPK16 was amplified with primer pair P3/P4 (S1 Table) and moved into pET28a to generate pET28a-CDPK16. To generate ADF7^S128A, ADF7^S128D, ADF10^S128A, and ADF10^S128D proteins, we initially generated plasmids pGEX-KG-ADF7^S128A, pGEX-KG-ADF7^S128D, pGEX-KG-ADF10^S128A, and pGEX-KG-ADF10^S128D using pGEX-KG-ADF7 and pGEX-KG-ADF10 plasmids [10,11] as the temperate with the primer pairs M1/M2, M3/M4, M5/M6 and M7/M8 (S1 Table), respectively. The plasmids pGEX-KG-ADF10, pGEX-KG-ADF10^S128A, and pGEX-KG-ADF10^S128D were directly transformed into the Escherichia coli BL21 (DE3) strain. The sequences of ADF7, ADF7^S128A, and ADF7^S128D were then amplified with primer pair P1/P2 (S1 Table) using pGEX-KG-ADF7, pGEX-KG-ADF7^S128A, and pGEX-KG-ADF7^S128D as the templates and were moved into pET28a to generate pET28a-ADF7, pET28a-ADF7^S128A, and pET28a-ADF7^S128D, respectively. The resulting plasmids were transformed into the E. coli BL21 (DE3) strain. Generation of pET28a-ADF4 and pET28a-CKL2 plasmids was performed as described previously [30]. The recombinant proteins with His-tag or GST-tag were purified with Ni-NTA Resin (Merck Millipore, 70666–4) or Glutathione Sepharose 4B (GE Healthcare, 17-0756-05) according to the manufacturer’s instructions. Actin was purified from rabbit skeletal muscle as described previously [63,64] and labeling of actin with Oregon Green 488 Iodoacetamide (Thermo Fisher Scientific, O6010) or 7-chloro-4-nitrobenzo-2-oxa-1,3-diazole (NBD, Invitrogen, C20260) was performed as described previously [65,66].

Kinase activity assay

In vitro kinase activity assays were performed to determine whether ADF is phosphorylated by CDPK16. Recombinant proteins CDPK16-6×His, ADF7-6×His, and ADF7^S128D-6×His were purified with Ni-NTA agarose. CDPK16-6×His (1 μg) was incubated with 1 μg ADF7-6×His or ADF7^S128D-6×His in a kinase reaction buffer (20 mM Tris-HCl (pH 8.0), 5 mM MgCl₂, 10 μM ATP (Jena Bioscience, NU-1103L), 1 mM DTT, and 2 μCi [γ-³²P] ATP) at 30°C for 30 min. The same volume of 2 × SDS loading buffer was added to terminate the phosphorylation reaction. The samples were analyzed by 15% SDS-PAGE and stained with Coomassie Brilliant blue R 250 (Sigma-Aldrich, C.I. 42660). The SDS-PAGE gel was exposed to a phosphor screen and the phosphorylation signals were detected the next day using a Typhoon 9410 phosphor imager (Amersham Biosciences). The phosphorylation signals were determined by densitometry using ImageJ software.

Mass spectrometry

To identify the CDPK16-interacting proteins in pollen, Ni-NTA agarose bound with CDPK16-6×His was incubated with the total protein extract isolated from WT pollen. After 3 washes with protein extraction buffer, the proteins bound to the beads were separated by 15% SDS-PAGE and stained with Coomassie Brilliant blue R 250. To perform the mass spectrometry analysis, the gel bands containing the proteins of interest were first cut into pieces, and the remaining steps were performed according to the previously reported methods [67]. Briefly, mass spectrometry samples were prepared by decolorization, drying, reduction, alkylation, drying, enzymatic hydrolysis, termination and separation, extraction, reconstitution, etc. Finally, the samples were dissolved in 20 μl 0.1% formic acid. Mass spectrometry detection was performed using the Orbitrap Fusion Tribrid mass spectrometer (Thermo Fisher Scientific). To identify the phosphorylated site(s) of ADF7 in vivo, 8His-ADF7 was isolated from pollen grains derived from the transgenic line proADF7::8His-gADF7; adf7 by Ni-NTA agarose. To identify the CDPK16-phosphorylated-site(s) of ADF7 in vitro, 20 μM ADF7 was incubated with 5 μM recombinant CDPK16-6×His in the kinase reaction buffer I (20 mM Tris-HCl (pH 8.0), 0.5 mM CaCl₂, 1 mM MgCl₂, 0.5 mM ATP, and 1 mM DTT) at 30°C for 30 min. After separation by 15% SDS-PAGE, the 8His-ADF7 band was cut out of the gel and analyzed by liquid chromatography–mass spectrometry/mass spectrometry (LC-MS/MS), which was performed at the Center of Biomedical Analysis, Tsinghua University.

High-speed F-actin co-sedimentation assay

The high-speed F-actin co-sedimentation assay was adapted from previously published methods [68]. All proteins were pre-clarified at 200,000 g for 30 min at 4°C. Briefly, preassembled actin filaments (3 μM) were incubated with 20 μM ADF7, ADF7^S128A, or ADF7^S128D. After incubation for 1 h at 30°C, the reaction mixtures were subjected to centrifugation at 100,000 g for 30 min at 4°C. To determine the effect of CDPK16 and CKL2 on the actin-depolymerizing activity of ADF7, ADF7^S128A, and ADF4, varying concentrations of CDPK16 or CKL2 were incubated with their corresponding substrates ADF7 (20 μM) in kinase reaction buffer I (20 mM Tris-HCl (pH 8.0), 0.5 mM CaCl₂, 1 mM MgCl₂, 0.5 mM ATP, and 1 mM DTT) or kinase reaction buffer II (20 mM Tris-HCl (pH 8.0), 0.5 mM EGTA, 1 mM MgCl₂, 0.5 mM ATP, and 1 mM DTT) for 30 min at 30°C before incubation with 3 μM preassembled actin filaments at 30°C for 1 h. The reaction mixtures were subjected to centrifugation at 100,000 g for 30 min at 4°C. The protein samples in the supernatant and pellet fractions were separated by SDS-PAGE and revealed by staining with Coomassie Brilliant blue R 250. The amount of actin in the supernatant was quantified by densitometry using ImageJ software.

Direct visualization of individual actin filaments by fluorescence light microscopy

The effect of ADF7 and its variants on shortening actin filaments was determined by fluorescence light microscopy as described previously [25]. Briefly, pre-polymerized actin filaments (2 μM) were incubated with 2.5 μM ADF7, 2.5 μM ADF7^S128A, or 2.5 μM ADF7^S128D for 3 min at room temperature, and 5 μM Rhodamine-Phalloidin (Thermo Fisher Scientific, R415) was subsequently added to terminate the reaction and label actin filaments. Actin filaments were diluted to a final concentration of 100 nM with buffer F (15 μg/ml glucose, 0.5% methylcellulose, 100 μg/ml glucose oxidase, 20 μg/ml catalase, 50 mM KCl, 1 mM MgCl₂, 10 mM imidazole (pH 7.0), and 100 mM DTT) and were observed under an Olympus BX53 microscope equipped with a 1.42 NA ×60 oil immersion lens. The images were collected with an Olympus DP80 camera controlled by Cell Sens Standard 1.12 software. To determine the effect of CDPK16 on the activity of ADF7, various concentrations of CDPK16 were first incubated with ADF7 in kinase reaction buffer I at 30°C for 30 min. The remaining steps were the same as for ADF7 alone.

Dilution-mediated actin depolymerization assay

To determine the ability of ADF7 and its variants on depolymerizing actin filaments, a dilution-mediated actin depolymerization assay was performed as described previously [69]. Briefly, 5 μM pre-centrifuged proteins (ADF7, ADF7^S128A, or ADF7^S128D) were incubated with 5 μM preassembled actin filaments (50% NBD-labeled) for 2 min at room temperature. The mixtures were subsequently diluted 25-fold into buffer G (5 mM Tris-HCl (pH 8.0), 0.2 mM CaCl₂, 0.02% NaN₃, 0.2 mM ATP, and 0.2 mM DTT). Actin depolymerization was traced by monitoring the changes in NBD fluorescence by the QuantaMaster Luminescence QM 3 PH Fluorometer (Photon Technology International) with the excitation and emission wavelengths set at 475 nm and at 530 nm, respectively.

Direct visualization of actin filament dynamics by total internal reflection fluorescence microscopy (TIRFM) in vitro

The dynamics of individual actin filaments were visualized by TIRFM, which was roughly adapted from the previously published method [70]. The proteins used in this assay were pre-clarified at 200,000 g for 30 min at 4°C. The flow cell was firstly incubated with 100 nM N-ethylmaleimide-myosin for 3 min in the dark, followed by washing with 1% BSA for 2 min. The flow cell was then washed with 1×TIRFM buffer (10 mM imidazole (pH 7.0), 50 mM KCl, 1 mM MgCl₂, 1 mM EGTA, 50 mM DTT, 0.2 mM ATP, 50 μM CaCl₂, 15 mM glucose, 20 μg/ml catalase, 100 μg/ml glucose oxidase, and 0.5% methylcellulose). Actin filaments at 150 nM [50% Oregon Green 488-labeled] were injected into the chamber and incubated for 5 min at room temperature in the dark. The unbound actin filaments were removed by washing with 1×TIRFM buffer. ADF7, ADF7^S128A, ADF7^S128D, or ADF7 after incubation with CDPK16 in phosphorylation reaction buffer for 30 min at room temperature was injected into the chamber. Actin filaments were observed immediately by TIRF illumination with an Olympus IX-71 microscope equipped with a 100× oil objective (1.45 numerical aperture). The time-lapse images were collected for 5 min with a Hamamatsu ORCA-EM-CCD camera (model C9100-12) driven by Micro-Manager software (www.micro-manager.org; v1.4; MMStudio) at time intervals of 3 s. More than 25 actin filaments (>10 μm) were chosen for quantification of the severing frequency (breaks/μm/s) of actin filaments with ImageJ software according to the previously published method [34].

Sequence alignment

The sequences of ADF proteins from different plant species were obtained from NCBI referring to the previous phylogenetic analyses [71–73]. The sequence alignment was performed using the software MAGA7 (https://www.megasoftware.net) and the program ClustalW, and the resulting data were exported in the FASTA format. The FASTA format file was submitted to the online server ESPript3 (http://espript.ibcp.fr/ESPript/cgi-bin/ESPript.cgi) to predict the secondary structures using AtADF1 as the template and calculate the sequence similarities based on the %Equivalent method with the global score set to 0.7. The seqlogo diagram, which reflects the amino acid prevalence at each position in the peptide sequence encompassing Ser128, was generated by using the online software program Weblogo 3 (http://weblogo.threeplusone.com/create.cgi).