ANGUSTIFOLIA Regulates Actin Filament Alignment for Nuclear Positioning in Leaves^,

In dark-adapted Arabidopsis cells, nuclei position toward the body center of the leaf, a centripetal nuclear positioning process regulated by ANGUSTIFOLIA through the alignment of actin filaments.

Abstract

During dark adaptation, plant nuclei move centripetally toward the midplane of the leaf blade; thus, the nuclei on both the adaxial and abaxial sides become positioned at the inner periclinal walls of cells. This centripetal nuclear positioning implies that a characteristic cell polarity exists within a leaf, but little is known about the mechanism underlying this process. Here, we show that ANGUSTIFOLIA (AN) and ACTIN7 regulate centripetal nuclear positioning in Arabidopsis (Arabidopsis thaliana) leaves. Two mutants defective in the positioning of nuclei in the dark were isolated and designated as unusual nuclear positioning1 (unp1) and unp2. In the dark, nuclei of unp1 were positioned at the anticlinal walls of adaxial and abaxial mesophyll cells and abaxial pavement cells, whereas the nuclei of unp2 were positioned at the anticlinal walls of mesophyll and pavement cells on both the adaxial and abaxial sides. unp1 was caused by a dominant-negative mutation in ACTIN7, and unp2 resulted from a recessive mutation in AN. Actin filaments in unp1 were fragmented and reduced in number, which led to pleiotropic defects in nuclear morphology, cytoplasmic streaming, and plant growth. The mutation in AN caused aberrant positioning of nuclei-associated actin filaments at the anticlinal walls. AN was detected in the cytosol, where it interacted physically with plant-specific dual-specificity tyrosine phosphorylation-regulated kinases (DYRKPs) and itself. The DYRK inhibitor (1Z)-1-(3-ethyl-5-hydroxy-2(3H)-benzothiazolylidene)-2-propanone significantly inhibited dark-induced nuclear positioning. Collectively, these results suggest that the AN-DYRKP complex regulates the alignment of actin filaments during centripetal nuclear positioning in leaf cells.

The proper spatial arrangement of nuclei is essential for various cellular activities during cell division, growth, migration, and differentiation in eukaryotes (Takagi et al., 2011; Gundersen and Worman, 2013; van Bergeijk et al., 2016). The proper nuclear positioning also is required to help plants cope with environmental stimuli, including pathogen infection, touch, temperature, and light (Takagi et al., 2011; Griffis et al., 2014). We recently reported that the nuclei of Arabidopsis (Arabidopsis thaliana) leaf cells move to the anticlinal walls of cells in response to strong blue/ultraviolet A light to reduce DNA damage and cell death caused by ultraviolet irradiation (Iwabuchi et al., 2016). Furthermore, we found that nuclei in Arabidopsis leaf cells are positioned facing toward the body center of the leaf in the dark (Iwabuchi et al., 2016). Specifically, the nuclei of pavement and mesophyll cells on both the adaxial and abaxial sides of the leaf are positioned at inner periclinal walls. These findings indicate that nuclear positioning is independent of gravity. This process is induced repeatedly in the dark (Iwabuchi et al., 2007), indicating that this nuclear positioning process is active. This finding led to the observation that there is a characteristic cell polarity in the adaxial-abaxial direction of a leaf.

Light-induced nuclear positioning depends on chloroplast movement. Namely, nuclei cannot move by themselves; instead, nucleus-attached chloroplasts carry the nuclei to anticlinal walls of cells (Higa et al., 2014). Light-induced chloroplast movement is a well-known process that occurs in a wide variety of plants whose underlying mechanism has been investigated in depth (Wada, 2013, 2016; Suetsugu et al., 2017). Chloroplast movement is regulated by the blue light-receptor phototropins, actin filaments, and several regulatory proteins (Kong and Wada, 2016). Hence, in mutants lacking these proteins, this type of nuclear positioning does not occur.

Unlike light-induced nuclear positioning, the mechanism underlying dark-induced nuclear positioning is not fully understood. The actin cytoskeleton is essential for dark-induced nuclear positioning. The plant-specific motor myosin XI-i, which is localized to the nuclear envelope, is involved in dark-induced nuclear positioning (Iwabuchi et al., 2010, 2016; Tamura et al., 2013). Some proteins that regulate chloroplast movement also are involved in this response (Higa et al., 2014; Suetsugu et al., 2015, 2017). In dark-adapted pavement cells, thick, longitudinally aligned actin filaments are associated with nuclei at inner periclinal walls (Iwabuchi et al., 2010; Iwabuchi and Takagi, 2010; Takagi et al., 2011). These actin filaments are thought to be required for the positioning of the nucleus to the inner periclinal walls of the cell. Moreover, an analysis of actin8D and phototropin2 mutants indicates that dark-induced nuclear positioning is differentially regulated between pavement cells and mesophyll cells (Iwabuchi et al., 2007, 2010, 2016).

Here, we screened for mutants defective in nuclear position in the dark to identify additional regulatory proteins involved in dark-induced nuclear positioning. We obtained two independent mutants, which we designated unusual nuclear positioning1 (unp1) and unp2. Whereas unp1 is a previously unreported dominant-negative mutant of ACTIN7, unp2 is a recessive mutant of the gene ANGUSTIFOLIA (AN). AN encodes a plant homolog of C-terminal-binding protein/brefeldin A-ADP ribosylated substrate (CtBP/BARS; Folkers et al., 2002; Kim et al., 2002). AN is involved in determining leaf and cell shapes, root formation, microtubule organization, and abiotic stress responses in Arabidopsis (Tsuge et al., 1996; Folkers et al., 2002; Kim et al., 2002; Bai et al., 2013; Gachomo et al., 2013; Bhasin and Hülskamp, 2017). Our findings reveal the relationship between AN and the actin cytoskeleton in centripetal nuclear positioning in Arabidopsis leaves.

RESULTS

Isolation of Two Arabidopsis Mutants with Defects in Nuclear Positioning in the Dark

To explore the mechanism of dark-induced nuclear positioning, we employed a forward genetics approach. We isolated the unp1 mutant by screening an ethyl methanesulfonate-mutagenized population of transgenic Arabidopsis plants expressing the nuclear marker Nup50a-GFP (Tamura et al., 2013). In dark-adapted wild-type leaves, most nuclei in palisade mesophyll and pavement cells were positioned at the inner periclinal wall of the cell. In unp1 leaves, by contrast, 52% of nuclei were aberrantly positioned at the anticlinal walls of mesophyll cells, although most nuclei in pavement cells were positioned at the inner periclinal walls, as in wild-type cells (Fig. 1). Leaf nuclei are lens shaped; thus, the projection area of the nucleus correlates negatively with the rate of nuclear positioning at the anticlinal wall (Iwabuchi et al., 2016). This was observed in unp1 mesophyll cells (Supplemental Fig. S1).

Figure 1.

The unp1 and unp2 mutants exhibit aberrant nuclear positioning in the dark. A, Distribution patterns of nuclei in palisade mesophyll cells and adaxial pavement cells of wild-type, unp1, and unp2 leaves in the dark. The left and middle columns show horizontal sections with nuclei (blue) stained with Hoechst 33342. Cells are outlined with yellow dotted lines. The right column shows cross sections, including nuclei (green) stained with Hoechst 33342, cell walls (blue) stained with Calcofluor, and chloroplasts (magenta). Bars = 20 µm. B, Percentage of nuclei positioned on the anticlinal walls of palisade mesophyll and adaxial pavement cells of wild-type, unp1, and unp2 leaves in the dark and after illumination with blue light (100 µmol m⁻² s⁻¹ for 3 h). Data represent means ± se (n = 5 leaves; **, P < 0.01 with Student’s t test). Mesophyll and pavement cells were observed in each of five leaves from different plants; the mean numbers of each cell type observed per leaf were as follows: wild-type leaves, 100 mesophyll and 67 pavement cells; unp1 leaves, 103 mesophyll and 49 pavement cells; and unp2 leaves, 135 mesophyll and 88 pavement cells.

Nuclear positioning after exposure to 100 µmol m⁻² s⁻¹ blue light for 3 h also was investigated in unp1. In pavement cells, 55% of wild-type nuclei and 35% of unp1 nuclei moved to the anticlinal walls, although in mesophyll cells, 87% of wild-type nuclei and 83% of unp1 nuclei moved to the anticlinal walls (Fig. 1B; Supplemental Fig. S2). These results indicate that the unp1 mutation affected blue light-induced nuclear positioning in pavement cells. We also observed the positions of chloroplasts in unp1 mesophyll cells and found no differences between the wild type and unp1; the chloroplasts of both lines were distributed in the lower half of the cell in the dark and at the anticlinal walls in blue light (Fig. 1A; Supplemental Fig. S2). Thus, the unp1 mutation did not appear to affect chloroplast positioning.

The leaf petioles of the unp1 mutant were bent upward (Supplemental Fig. S3A), and plant height, seed number per fruit, and fruit length were reduced significantly in unp1 compared with wild-type plants (Supplemental Fig. S3, B–D). Furthermore, the nuclei of unp1 pavement cells were spherical, while those of wild-type cells were spindle shaped (Supplemental Fig. S3E). The nuclei of unp1 and wild-type cells were almost the same size (Supplemental Fig. S3F). These data show that, in addition to changes in nuclear positioning, the unp1 mutant exhibited pleiotropic phenotypes in various organs.

We isolated the unp2 mutant from an Arabidopsis population carrying T-DNA insertions. Dark-induced nuclear positioning was impaired in pavement and mesophyll cells of unp2 leaves; 65% of mesophyll nuclei and 36% of pavement nuclei were positioned aberrantly at the anticlinal walls (Fig. 1), indicating that the unp2 mutation affected a gene required for dark-induced nuclear positioning. After irradiation with 100 µmol m⁻² s⁻¹ blue light for 3 h, 89% of mesophyll nuclei and 69% of pavement nuclei had moved to the anticlinal walls (Fig. 1B; Supplemental Fig. S2), suggesting that the unp2 mutation did not affect blue light-induced nuclear positioning. The positions of chloroplasts in unp2 under dark and blue light conditions were similar to those in the wild type (Fig. 1A; Supplemental Fig. S2), suggesting that the unp2 mutation did not affect chloroplast positioning.

We next investigated dark-induced nuclear positioning in spongy mesophyll cells and abaxial pavement cells. Dark-induced nuclear positioning was impaired in these cells in unp1 and unp2 leaves: 64% of spongy mesophyll nuclei and 28% of abaxial pavement nuclei were positioned to anticlinal walls in unp1, while 40% of spongy mesophyll nuclei and 20% of abaxial pavement nuclei were positioned at the anticlinal walls in unp2 (Fig. 2). Thus, the unp1 and unp2 mutations were involved in centripetal nuclear positioning in leaves.

Figure 2.

Dark-induced nuclear positioning on the abaxial side of wild-type, unp1, and unp2 leaves. A, Distribution patterns of nuclei in spongy mesophyll cells and abaxial pavement cells of wild-type, unp1, and unp2 leaves in the dark. Nuclei (blue) were stained with Hoechst 33342. Cells are outlined with yellow dotted lines. Bars = 20 µm. B, Percentage of nuclei positioned on the anticlinal walls of spongy mesophyll and abaxial pavement cells of wild-type, unp1, and unp2 leaves in the dark. Data represent means ± se (n = 5 leaves; *, P < 0.05 and **, P < 0.01 with Student’s t test). Mesophyll and pavement cells were observed in each of five leaves from different plants; the mean numbers of each cell type observed per leaf were as follows: wild-type leaves, 45 spongy mesophyll and 44 abaxial pavement cells; unp1 leaves, 30 spongy mesophyll and 45 abaxial pavement cells; and unp2 leaves, 43 spongy mesophyll and 57 abaxial pavement cells.

ACTIN7 and AN Are Responsible for the unp1 and unp2 Phenotypes, Respectively

Backcrossing of unp1 with wild-type plants showed that unp1 was a dominant-negative mutant. Map-based cloning revealed a point mutation in the second exon of ACTIN7 (At5g09810) in the unp1 mutant (Fig. 3A), which resulted in the substitution of Gly-38 with Ser in a region containing the DNase I-binding loop (D-loop) of actin protein subdomain 2 (Fig. 3B). The Gly residue of subdomain 2 is widely conserved in fungi, animals, and plants (Supplemental Fig. S4). To confirm that this mutation was responsible for the unp1 phenotype, we introduced ACTIN7 harboring the unp1 mutation, designated ACTIN7(G38S), into wild-type plants. When these transgenic plants were placed in the dark, 94% of mesophyll nuclei were positioned at the anticlinal walls of cells (Fig. 3C), indicating that the unp1 phenotype resulted from the mutation in ACTIN7.

Figure 3.

ACTIN7 and AN are required for nuclear positioning in the dark. A, Schematic representation of ACTIN7 (At5g09810) showing the position of the unp1 mutation. Gray boxes, Untranslated regions; black boxes, exons; solid lines, introns. B, Front view of a model of actin protein (Protein Data Bank code, 3MFP; Fujii et al., 2010). C, Adaxial pavement and palisade mesophyll cells of plants expressing ACTIN7, ACTIN2, or ACTIN8 containing unp1-type mutations in the dark. Blue, Nuclei stained with Hoechst 33342. Cells are outlined with yellow dotted lines. Values below the images indicate the percentages of nuclei positioned on the anticlinal walls. Data represent means ± se (n = 5 leaves). Mesophyll and pavement cells were observed in each of five leaves from different plants; the mean numbers of each cell type observed per leaf were as follows: ACTIN2(G38S), 102 mesophyll and 43 pavement cells; ACTIN7(G38S), 101 mesophyll and 40 pavement cells; and ACTIN8(G38S), 92 mesophyll and 35 pavement cells. Bars = 20 µm. D, Schematic representation of AN (At1g01510) showing the positions of the unp2 and an-1 mutations. Gray boxes, Untranslated regions; black boxes, exons; solid lines, introns. E, RT-PCR of AN and ACTIN2 (control) transcripts in wild-type and unp2 leaves. F, Palisade mesophyll and adaxial pavement cells of an-1 plants in the dark. Blue, Nuclei stained with Hoechst 33342. Cells are outlined with yellow dotted lines. Values below the images indicate the percentages of nuclei positioned on anticlinal walls. Data represent means ± se (n = 5 leaves). Mean numbers of 101 mesophyll cells and 50 pavement cells were observed in each of five leaves from different an-1 plants. Bars = 20 µm.

We also examined the effect of the unp1-type mutation on other actin isoforms, including ACTIN2 (At3g18780) and ACTIN8 (At1g49240); like ACTIN7, these genes are strongly expressed in vegetative tissues of Arabidopsis (Meagher et al., 1999). When either ACTIN2(G38S) or ACTIN8(G38S) was expressed under the control of the ACTIN7 promoter, the mesophyll nuclei of the transgenic plants were positioned at the anticlinal walls of cells in the dark [Fig. 3C; 62% ± 6.6% for ACTIN2(G38S) and 79% ± 4.6% for ACTIN8(G38S)]. This suggested that all three actin isoforms were involved in dark-induced nuclear positioning in mesophyll cells.

Genotyping revealed a T-DNA insertion in the first intron of AN (At1g01510) in unp2 mutants (Fig. 3D). Reverse transcription (RT)-PCR analysis did not detect intact AN mRNA in unp2 (Fig. 3E). Another allele, an-1, exhibited similar changes in nuclear positioning to unp2: the percentages of nuclei positioned at the anticlinal walls were 73% ± 3.5% for mesophyll cells and 38% ± 6.4% for pavement cells (Fig. 3F). This indicated that the unp2 phenotype resulted from the deficiency in AN caused by the T-DNA insertion.

Pavement cells of wild-type leaves had jigsaw piece-like, irregularly shaped anticlinal walls that interlocked with those of neighboring cells. The pavement cells of unp2 leaves were smoother than those of wild-type cells, as determined by the circularity index (Supplemental Fig. S5, A and B). We compared the circularity index of pavement cells having nuclei positioned on the anticlinal walls with that of cells having nuclei positioned on the inner periclinal walls. There were no differences in circularity indices between either wild-type or unp2 cells (Supplemental Fig. S5C). Therefore, AN-dependent dark-induced nuclear positioning was not related to cell shape.

The unp1 and unp2 Mutations Affect Actin Filament Organization

We previously demonstrated the actin-dependent regulation of dark-induced nuclear positioning by fluorescently labeled phalloidin staining and immunofluorescence staining (Iwabuchi et al., 2010). We examined how the unp1 mutation affected actin organization by fluorescently labeled phalloidin staining. In the pavement cells of wild-type leaves, nuclei were associated with thick, longitudinally aligned actin filaments at inner periclinal walls in the dark (Fig. 4A). Actin filaments also were well organized at the upper cell surface (Fig. 4A). The number of actin filaments was reduced drastically in unp1 pavement cells (Fig. 4, A and B), even though the levels of actin protein (Fig. 4C) and actin bundling (Fig. 4D) were comparable between the wild type and unp1.

Figure 4.

Organization of actin filaments in pavement cells of wild-type, unp1, and unp2 leaves in the dark. A, Maximum intensity projections of actin filaments in the vicinity of the nuclei (left) and at the surface (right) of adaxial pavement cells of wild-type, unp1, and unp2 leaves in the dark. Magenta (arrows), Nuclei stained with Hoechst 33342; green (arrowheads), actin filaments stained with phalloidin. Cells are outlined with sold white lines. Sequential images were taken from the top to bottom of cells along the optical z axis at 0.5-µm intervals. Round, green structures approximately 5 µm in diameter (indicated with black asterisks) are chloroplasts in the underlying mesophyll cells. Bars = 20 µm. B, Occupancy rates of actin filaments at the cell surfaces of adaxial pavement cells in wild-type, unp1, and unp2 leaves (n = 34 cells in wild-type leaves, 33 cells in unp1 leaves, and 48 cells in unp2 leaves; *, P < 0.05 and **, P < 0.01 with Student’s t test). Data represent means ± se. C, Immunoblot analysis of wild-type, unp1, and unp2 leaf proteins using anti-actin and anti-tubulin antibodies. Tubulin was used as a loading control. D, Fluorescence levels of actin filaments at the cell surface of adaxial pavement cells in wild-type, unp1, and unp2 leaves (n = 34 cells in wild-type leaves, 33 cells in unp1 leaves, and 48 cells in unp2 leaves). Data represent means ± se.

To investigate actin organization in living cells, we examined transgenic plants expressing Lifeact-Venus (Era et al., 2009). The actin filaments were severely fragmented in the pavement cells of transgenic plants (Fig. 5A), although their lateral movements were normal in unp1 (Fig. 6; Supplemental Movies S1 and S2). The unp1 mutation also affected cytoplasmic streaming: unp1 cells showed a 36% decrease in the velocity of cytoplasmic streaming compared with the wild type (Fig. 5B; Supplemental Movie S3). These results suggested that Gly-38 in the D-loop of ACTIN7 was required for the polymerization of actin filaments and cytoplasmic streaming.

Figure 5.

Fragmentation of actin filaments affects the velocities of particles in pavement cells of unp1 leaves. A, Maximum intensity projections of actin filaments at the surfaces of adaxial pavement cells in Lifeact-Venus/wild-type and Lifeact-Venus/unp1 leaves. Cells are outlined with dotted yellow lines. Sequential images were taken from the top to bottom of cells along the optical z axis at 0.5-µm intervals. Bars = 20 µm. B, Maximal velocities of particles in adaxial pavement cells in wild-type, unp1, and unp2 leaves. Particles in pavement cells located above the midveins of the leaves were imaged at 0.13-s intervals for 30 s (see Supplemental Movie S3). The distances traveled by the particles in 5 s were determined. Data represent means ± se (n = 21 cells in wild-type leaves, 17 cells in unp1 leaves, and 23 cells in unp2 leaves) from three to four leaves (**, P < 0.01 with Student’s t test).

Figure 6.

Lateral movements of actin filaments in leaf pavement cells of Lifeact-Venus/wild-type and Lifeact-Venus/unp1 plants. A, Movement of actin filaments. Actin filaments were imaged at 5-s intervals for 60 s (see Supplemental Movies S1 and S2). Images at the 0-, 5-, and 10-s time points were merged. Arrows indicate nuclei. Bars = 20 µm. B, Cross-correlation coefficients between the first image and the subsequent time-sequential images of actin filaments in the vicinity of nuclei (left) and at the cell surface (right) in Lifeact-Venus/wild-type and Lifeact-Venus/unp1 plants. Data represent means ± se (n = 4–6 leaves). A mean of four pavement cells were observed in each of four to six leaves from different plants.

We also examined the effect of the unp2 mutation on the actin cytoskeleton in dark-adapted pavement cells using phalloidin staining but not Lifeact-Venus, which caused artifactual defects in actin organization (for an explanation of our choice of protocol, see “Materials and Methods”). We observed longitudinally aligned actin filaments associated with the nuclei in the wild type and unp2 (Fig. 4A). However, the actin filaments in the wild type were aligned along inner periclinal walls, while those in unp2 were aligned aberrantly along the anticlinal walls (Fig. 4A). The number of actin filaments was slightly but significantly reduced at the cell surface in unp2 cells compared with wild-type cells (Fig. 4B). By contrast, the levels of actin protein (Fig. 4C), actin bundling (Fig. 4D), and the velocity of cytoplasmic streaming (Fig. 5B) were normal in the mutant. In addition, to eliminate the possibility that the aberrant cell shape affected actin filament alignment in unp2, we examined pavement cells positioned just above the leaf midvein. The shapes of these pavement cells in wild-type and unp2 plants were similar: the cells were long, narrow, and less wavy than pavement cells in other areas of the leaf (Fig. 7A). Even in such pavement cells of unp2, nucleus-associated actin filaments were aligned aberrantly along the anticlinal walls (Fig. 7A). As shown in Figure 1B, some pavement nuclei were positioned at the anticlinal and inner periclinal walls in wild-type and unp2 cells, respectively, despite dark adaptation. At such times, actin filaments were aligned with the nuclei along the anticlinal walls in the wild-type cells and on the inner periclinal walls in unp2 cells (Fig. 7B), producing a strong relationship between the positions of nuclei and the alignment of actin filaments. Taken together, these results indicated that AN was required for the proper alignment of nucleus-associated actin filaments.

Figure 7.

Organization of actin filaments in pavement cells of wild-type and unp2 leaves in the dark. Maximum intensity projections are shown for actin filaments in the vicinity of nuclei and at the surfaces of adaxial pavement cells of wild-type and unp2 leaves (A) and actin filaments associated with nuclei occasionally positioned at the anticlinal wall of wild-type pavement cells and at the inner periclinal wall of unp2 pavement cells (B). Magenta (arrows), Nuclei stained with Hoechst 33342; green (arrowheads), actin filaments stained with phalloidin. Cells are outlined with dotted yellow lines. Sequential images were taken from the top to bottom of cells along the optical z axis at 0.5-µm intervals. Round, green structures approximately 5 µm in diameter (indicated with black asterisks) are chloroplasts of the underlying mesophyll cells. Bars = 20 µm.

AN Constitutively Localizes to the Cytosol under Dark and Blue Light Conditions

We also investigated the intracellular localization of AN in pavement and mesophyll cells of transgenic plants expressing both AN-GFP and the nuclear membrane marker SUN2-tRFP. Under dark conditions, AN-GFP was detected predominantly in the cytosol rather than in the nucleus in both pavement and mesophyll cells (Supplemental Fig. S6). The cytosolic localization of AN-GFP was not affected by irradiation with 100 µmol m⁻² s⁻¹ blue light for 1 h, which normally causes nuclei to begin to move to the anticlinal walls of cells (Supplemental Fig. S6). These results indicated that AN was constitutively present in the cytosol under both dark and blue light conditions.

AN Associates with Dual-Specificity Tyr Phosphorylation-Regulated Kinases

To identify proteins that interacted with AN, we performed an interactome analysis of transgenic plants expressing AN-GFP. Mass spectrometry analysis of the fraction pulled down by the anti-GFP antibody identified several forms of plant-specific dual-specificity tyrosine phosphorylation-regulated kinases (DYRKPs; DYRKP-1, DYRKP-2A, DYRKP-2B, and DYRKP-3; Fig. 8, A and B; Supplemental Fig. S7).

Figure 8.

AN associates with the protein kinase DYRKP. A, Silver staining of products from transgenic plants expressing either free GFP or AN-GFP pulled down by anti-GFP antibody. AGI, Arabidopsis Genome Initiative. B, Identification of putative AN-associating protein kinases by mass spectrometry. C, Yeast two-hybrid assay showing physical interactions of AN with DYRKP-2B and AN. D, GFP signals in the adaxial pavement and palisade mesophyll cells of 35S:GFP-DYRKP-2A transgenic plants. Bars = 20 µm.

To confirm a physical interaction between AN and DYRKPs, we performed a yeast two-hybrid assay. AN interacted physically with DYRKP-2B but not DYRKP-3 (Fig. 8C). In addition, AN self-assembled in this assay (Fig. 8C). We also investigated the intracellular localization of DYRKP-2A in transgenic plants expressing GFP-DYRKP-2A. Distribution was similar to that of AN, being localized predominantly to the cytosol in both pavement and mesophyll cells (Fig. 8D). Thus, the AN-DYRKP complex appears to function in the cytosol.

Several inhibitors have been used to investigate the function of DYRK in animal cells (Pozo et al., 2013). We examined the effect of INDY [(1Z)-1-(3-ethyl-5-hydroxy-2(3H)-benzothiazolylidene)-2-propanone], a benzothiazole compound that inhibits animal Dyrk1A (Ogawa et al., 2010), on dark-induced nuclear positioning. INDY binds to the ATP pocket of Dyrk1A and is a highly selective inhibitor of its kinase activity (Ogawa et al., 2010). After exposure to 100 µmol m⁻² s⁻¹ blue light for 3 h, leaves were treated with INDY at concentrations of 10 µm, 100 µm, or 1 mm for 5 h in the dark. Concentrations of 100 µm and 1 mm INDY significantly inhibited the movement of nuclei from the anticlinal walls to the inner periclinal walls in both pavement cells and mesophyll cells (Fig. 9, A and B). By contrast, 100 µm INDY did not inhibit blue light-induced nuclear positioning (Fig. 9, C and D). Therefore, DYRKPs appear to be involved in dark-induced nuclear positioning but not in blue light-induced nuclear positioning.

Figure 9.

INDY inhibits dark-induced nuclear positioning but not blue light-induced nuclear positioning. A, Distribution patterns of nuclei in palisade mesophyll and adaxial pavement cells of leaves treated with either 100 µm INDY or dimethyl sulfoxide (DMSO; Mock) for 5 h in the dark following exposure to 100 µmol m⁻² s⁻¹ blue light for 3 h. Nuclei (blue) were stained with Hoechst 33342. Cells are outlined with yellow dotted lines. Bars = 20 µm. B, Effects of INDY on dark-induced nuclear positioning. Data represent means ± se (n = 5–6 leaves; *, P < 0.05 and **, P < 0.01 with Student’s t test). Mean numbers of 121 palisade mesophyll cells and 78 adaxial pavement cells were observed in each of five to six leaves from different plants. C, Distribution patterns of nuclei in palisade mesophyll and adaxial pavement cells of leaves exposed to 100 µmol m⁻² s⁻¹ blue light for 3 h in the presence of either 100 µm INDY or DMSO (Mock). The INDY or DMSO treatments were performed 1 h before the end of the dark treatment. Nuclei (blue) were stained with Hoechst 33342. Cells are outlined with yellow dotted lines. Bars = 20 µm. D, Percentage of nuclei positioned on the anticlinal walls of palisade mesophyll and adaxial pavement cells in blue light. Data represent means ± se (n = 5 leaves). Mean numbers of 106 mesophyll cells and 72 pavement cells were observed in each of five leaves from different plants.

We next focused on the phosphorylation of actin protein. To examine the level of actin phosphorylation, we subjected wild-type leaf extracts to immunoblotting on a SuperSep Phos-tag gel with anti-actin antibody. Pretreatment of the leaf extract with alkaline phosphatase affected the migration of actin protein (Fig. 10). Unphosphorylated actin protein was not detected in the absence of alkaline phosphatase treatment (Fig. 10), clearly showing that most actin proteins were phosphorylated in planta. The phosphorylation status of actin proteins did not differ between unp2 and wild-type plants (Fig. 10), suggesting that the AN-DYRKP complex was not involved directly in the phosphorylation of actin.

Figure 10.

Representative phosphorylation status of actin in wild-type, unp1, and unp2 leaves. Top, Immunoblot using a SuperSep Phos-tag gel showing the shift in the actin band toward a higher mass in samples not treated with alkaline phosphatase (−) compared with those treated with alkaline phosphatase (+). Bottom, Immunoblot using a standard gel as the loading reference.

DISCUSSION

A Model for Centripetal Nuclear Positioning in Plants

Our results indicated that centripetal nuclear positioning required the presence of actin filaments aligned at the inner periclinal walls in an AN-dependent manner. We propose a model for dark-induced nuclear positioning in inner periclinal walls in palisade mesophyll and pavement cells of plants (Fig. 11). In this model, the AN-DYRKP complex regulates the phosphorylation of an unknown protein X in the cytosol. Protein X regulates the alignment of actin filaments at the inner periclinal walls. The actin filaments are composed of ACTIN2, ACTIN7, and ACTIN8, all of which are phosphorylated. Further studies should focus on clarifying the detailed functions of AN and the DYRKPs and on identifying component(s) downstream from the AN-DYRKP complex. We showed previously that, in the dark, nuclei move to the centers of the inner periclinal walls via a process involving the myosin XI-i and WITs. These proteins, together with the nuclear membrane proteins WIP and SUN and the lamina-like proteins CROWDED NUCLEI and KAKU4 (Tamura et al., 2013, 2015; Goto et al., 2014), form the LINC complex. It would be useful to investigate whether centripetal nuclear positioning involves a relationship between the AN-DYRKP and LINC complexes.

Figure 11.

Model of dark-induced nuclear positioning. AN associates with DYRKPs to regulate the alignment of actin filaments at the inner periclinal walls through the phosphorylation of the putative target, X. Actin filaments are composed of ACTIN2, ACTIN7, and ACTIN8, and the actin proteins are highly phosphorylated. Actin filaments associate with the linker of nucleoskeleton and cytoskeleton (LINC) complex, which is composed of myosin XI-i and the nuclear membrane proteins WPP domain-interacting tail-anchored proteins (WITs), WPP domain-interacting protein (WIP), and Sad1/UNC-84 (SUN).

unp1 Is a Dominant-Negative Mutant of ACTIN7

The D-loop is required for the actin-actin interactions in mammal skeletal muscles (Khaitlina and Strzelecka-Gołaszewska, 2002; Oda et al., 2009; Fujii et al., 2010). Additionally, in the slime mold Dictyostelium discoideum, phosphorylation of Tyr-53 in the actin D-loop disrupts associations between actin monomers, resulting in a decrease in polymerization (Liu et al., 2006; Baek et al., 2008). In unp1, Gly-38 of the D-loop of subdomain 2 of ACTIN7 was substituted with Ser, suggesting that the D-loop has an important role in actin polymerization in plants. The D-loop also is important for actin-myosin interactions in D. discoideum (Kubota et al., 2009). In plants, the actomyosin system drives cytoplasmic streaming (Ueda et al., 2010), possibly due to the actin-myosin interactions through the actin D-loop.

As discussed previously (Nishimura et al., 2003; Kato et al., 2010), many normal actins are polymerized into actin filaments in unp1, but mutated actins (unp1-type actins) also are incorporated into the filaments with some frequency. These mutated actins inhibit the further incorporation of normal actins, as they are unable to undergo actin-actin interactions; thus, short actin filaments are formed. Therefore, short actin filaments in unp1 may be able to move laterally via myosin (Ueda et al., 2010).

There are two other dominant-negative mutants of actin: act2-2D, in which Arg-149 is substituted by Cys in subdomain 3 of ACTIN2 (Nishimura et al., 2003), and actin8D/fryzzy1, in which Glu-272 is replaced with Lys in the hydrophobic loop between subdomains 3 and 4 in ACTIN8 (Kato et al., 2010). unp1 exhibited similar pleiotropic phenotypes to these mutants (Supplemental Fig. S1), although, in contrast to the severe growth defects of act2-2D and actin8D, the growth of unp1 plants was relatively normal. This dissimilarity might result from the positions of the mutations, as there were no obvious growth differences between ACTIN2(G38S), ACTIN7(G38S), and ACTIN8(G38S) plants.

actin8D, like unp1, exhibits normal nuclear positioning in pavement cells in the dark (Iwabuchi et al., 2016), which suggests that nuclear position in pavement cells requires an actin-independent nuclear centering system. Alternatively, the seemingly normal positioning of nuclei in unp1 and actin8D may result from an absence of light-induced nuclear positioning, in which case the nuclei would always keep a position at the inner periclinal wall. Furthermore, dark-induced nuclear positioning was affected by the unp1 mutation in abaxial pavement cells but not in adaxial pavement cells, possibly due to a difference in the response to the unp1 mutation between adaxial and abaxial pavement cells.

unp1 is a new dominant-negative mutant of ACTIN7 and also is a valuable tool for manipulating actin filaments in plants. As actin polymerization can be inhibited genetically by the expression of the dominant-negative ACTIN2, ACTIN7, and ACTIN8 proteins, the role of the actin cytoskeleton in cellular processes could be investigated in tissues of interest by generating transgenic plants expressing a dominant-negative actin gene under a tissue-specific or inducible promoter. This tissue specificity cannot be obtained using actin inhibitors.

AN Regulates the Alignment of Actin Filaments for Centripetal Nuclear Positioning

AN controls the arrangement of cortical microtubules in Arabidopsis (Folkers et al., 2002; Kim et al., 2002). On the other hand, the arrangement of microtubules affects the alignment of actin filaments and vice versa (Mineyuki and Palevitz, 1990; Seagull, 1990; Eleftheriou and Palevitz, 1992; Tominaga et al., 1997; Era et al., 2013). Actin filaments in unp2 pavement cells were aligned aberrantly to anticlinal walls; thus, AN might regulate the actin filament alignment directly or indirectly. The shape of unp2 pavement cells did not affect the nuclear position in the dark; thus, it is also possible that AN regulates nuclear positioning independently of cell shape formation.

Several studies suggest diverse roles for AN. For example, deficiency of AN results in various phenotypes, including narrow leaves (Tsukaya et al., 1994; Tsuge et al., 1996), trichomes with reduced branching (Cho et al., 2005), twisted roots and seed pods (Bai et al., 2013; Kwak et al., 2015), and increased tolerance to drought and pathogen attack (Gachomo et al., 2013). The molecular roles of AN in these phenotypes remain to be clarified.

AN encodes a protein related to CtBP/BARS (Folkers et al., 2002; Kim et al., 2002). CtBP/BARS has an NAD-binding motif, the NAD(H)-dependent d-hydroxyacid dehydrogenase catalytic triad, and PXDLS-binding domains, required for functioning as a transcriptional corepressor (Chinnadurai, 2002, 2007). AN, however, has mutations in all these regions (Stern et al., 2007), has a plant-specific region at the C terminus (Kim et al., 2002; Stern et al., 2007), and functions outside the nucleus (Minamisawa et al., 2011), suggesting that its action differs from that of CtBP.

AN Associates with the Protein Kinases DYRKPs

AN-GFP was detected previously not only in the cytosol of Arabidopsis cells but also in the close vicinity of the trans-Golgi, where it formed several small particles (Folkers et al., 2002; Minamisawa et al., 2011). These small particles, however, were not observed under our experimental conditions, even though we used the same transgenic line as a previous study (Minamisawa et al., 2011); thus, we conclude that cytosolic AN is required for dark-induced nuclear positioning.

We identified DYRKP-2B as an AN-interacting protein (Fig. 8); its homolog DYRKP-1 also interacts physically with AN (Bhasin and Hülskamp, 2017). DYRKP-1 and DYRKP-2B are members of a plant-specific subgroup in the DYRK family (Kajikawa et al., 2015), which is widely conserved in animals, plants, fungi, and protists, and is involved in developmental processes, cell homeostasis, and cellular stresses (Aranda et al., 2011). DYRK3 is involved in stress granule formation in HeLa cells (Wippich et al., 2013), and AN is involved in this process in Arabidopsis leaf cells (Bhasin and Hülskamp, 2017), suggesting that DYRK3 and AN function in the same pathway. Given that DYRKPs had a high Mascot score in mass spectrometry analysis (Fig. 8), it is possible that AN and DYRKPs constitutively associate with each other to function in the cytosol. The DYRK inhibitor, INDY, also inhibited dark-induced nuclear positioning, suggesting that DYRKP, together with AN, was involved in regulating dark-induced nuclear positioning. The inhibitory effect of INDY on dark-induced nuclear positioning was not a side effect because blue light-induced nuclear positioning was statistically normal in the presence of INDY. Nevertheless, further analyses using DYRKP-deficient or kinase-dead mutants are required to clarify the molecular mechanisms through which the DYRKPs affect centripetal nuclear positioning in plants.

MATERIALS AND METHODS

Plant Materials and Growth Conditions

Arabidopsis (Arabidopsis thaliana) accession Columbia was used as the wild type. The unp1, unp2 (SALK_026489), an-1 (Tsukaya et al., 1994), and Lifeact-Venus (Era et al., 2009) plants were all in the Columbia background. The unp1 plants expressing Lifeact-Venus (Lifeact-Venus/unp1) were generated by crossing unp1 with transgenic plants expressing Lifeact-Venus. The plants were grown on compost for 2 to 5 weeks at 22°C under a 16-h-white light (30–50 µmol m⁻² s⁻¹)/8-h-dark cycle. Alternatively, the plants were grown aseptically on germination medium plates (one-half-strength Murashige and Skoog salts, 0.025% [w/v] MES-KOH, pH 5.7, and 0.5% [w/v] Gellan gum) under continuous white light.

Isolation of the unp1 Mutant and Map-Based Cloning of UNP1

Seeds of a transgenic plant expressing Nup50a-GFP were mutagenized by ethyl methanesulfonate treatment as described previously (Tamura et al., 2013). M2 seeds were collected from individual M1 plants to generate M2 lines. Each seedling was examined with a fluorescence microscope (Axioskop 2 plus; Zeiss), and a mutant line that exhibited aberrant nuclear positioning was selected. Map-based cloning was performed as described previously (Tamura et al., 2005). The unp1 mutant expressing Nup50a-GFP was crossed with a wild-type plant to remove Nup50a-GFP for nuclear positioning analysis.

Generation of Transgenic Plants Expressing ACTIN2(G38S), ACTIN7(G38S), and ACTIN8(G38S)

The Gateway cloning system (Invitrogen) was used to construct ProACTIN7:ACTIN7(G38S). The ProACTIN7:ACTIN7(G38S) fragment from unp1 DNA was amplified by PCR using the following primers: 5′-AACCAATTCAGTCGACGATATATCACGAAAACCGATC-3′ (ACTIN7 forward) and 5′-AAGCTGGGTCTAGATATCCGTTTCGCATCAGAATGGTAATACAC-3′ (ACTIN7 reverse). The amplified fragments were subcloned into the pENTR1A entry vector (Invitrogen) that had been digested with SalI and EcoRV using an In-Fusion HD Cloning Kit (Takara). The ProACTIN7:ACTIN7(G38S) fragment was introduced into the pFAST-G01 plant expression vector via the LR recombination reaction (Shimada et al., 2010).

To construct ProACTIN7:ACTIN2(G38S), the promoter region of ACTIN7 and the genomic regions of ACTIN2 and ACTIN8 were amplified from DNA from wild-type plants via PCR using the following primer sets: 5′-AACCAATTCAGTCGACGATATATCACGAAAACCGATC-3′ (ACTIN7-promoter forward) and 5′-TTTTCACTAAAAAAAAAG-3′ (ACTIN7-promoter reverse); 5′-TTTTTTTAGTGAAAAATGGCTGAGGCTGATGAT-3′ (ACTIN2-genomic forward) and 5′-TAACAACACTGGGAAAAACA-3′ (ACTIN2-G38S reverse); 5′-TTTCCCAGTGTTGTTAGTAGGCCAAGACATCATGGT-3′ (ACTIN2-G38S forward) and 5′-TAAGACAAGACACACTTAGAAACATTTTCTGTGA-3′ (ACTIN2-genomic reverse); and 5′-GTGTGTCTTGTCTTATCTGG-3′ (ACTIN7-terminator forward) and 5′-AAGCTGGGTCTAGATATCCGTTTCGCATCAGAATGGTAATACAC-3′ (ACTIN7-terminator reverse).

ProACTIN7:ACTIN8(G38S) was constructed as described for ProACTIN7:ACTIN2(G38S) using the following primer sets: 5′-AACCAATTCAGTCGACGATATATCACGAAAACCGATC-3′ (ACTIN7-promoter forward) and 5′-TTTTCACTAAAAAAAAAG-3′ (ACTIN7-promoter reverse); 5′-TTTTTTTAGTGAAAAATGGCCGATGCTGATGACATTC-3′ (ACTIN8-genomic forward) and 5′-TAACAACACTGGGGAAAACC-3′ (ACTIN8-G38S reverse); 5′-TTCCCCAGTGTTGTTAGTCGACCTAGACATCATGGT-3′ (ACTIN8-G38S forward) and 5′-TAAGACAAGACACACTTAGAAGCATTTTCTGTGGAC-3′ (ACTIN8-genomic reverse); and 5′-GTGTGTCTTGTCTTATCTGG-3′ (ACTIN7-terminator forward) and 5′-AAGCTGGGTCTAGATATCCGTTTCGCATCAGAATGGTAATACAC-3′ (ACTIN7-terminator reverse).

The fragments were fused and subcloned into the pENTR1A entry vector using an In-Fusion HD Cloning Kit (Takara). The ProACTIN7:ACTIN2(G38S) and ProACTIN7:ACTIN8(G38S) fragments were introduced into the pFAST-G01 plant expression vector. Plants were stably transformed with these vectors by Agrobacterium tumefaciens-mediated transformation.

Generation of Transgenic Plants Expressing GFP-DYRKP-2A

The DYRKP-2A cDNA sequence was amplified from cDNA of Columbia plants by PCR using KOD Plus Neo (Toyobo). The cDNA fragment was cloned into the entry vector, pENTR/D-TOPO (Life Technologies). The sequence of DYRKP-2A on the entry vector was inserted into a binary vector, pH35GW, with the LR reaction using LR clonase II (Life Technologies) to construct the binary vectors for pro35S:GFP-DYRKP-2A. The transformation was performed by the floral dip method (Clough and Bent, 1998). To select transformants, sterilized seeds were sown on Murashige and Skoog medium (Murashige and Skoog, 1962) containing Gamborg’s B5 vitamins (Gamborg et al., 1968), 3% (w/v) Suc, 0.8% (w/v) agar, and 10 mg L⁻¹ hygromycin (Wako Pure Chemical Industries). The fluorescence of GFP was detected using a confocal laser scanning microscope (FV1000; Olympus). GFP signal was excited with a 473-nm laser and detected with a 490- to 540-nm window.

Isolation of the unp2 Mutant and RT-PCR

Mature leaves of Arabidopsis T-DNA-tagged plants were examined with a fluorescence microscope (Axioskop 2 plus), and a mutant line that exhibited aberrant nuclear positioning in the dark was selected. Nuclei were visualized by Hoechst 33342 nuclear staining as described below.

Total RNA was isolated from leaves with an RNeasy Plant Mini Kit (Qiagen) according to the manufacturer’s instructions. Total RNA was subjected to first-strand cDNA synthesis using Ready-To-Go RT-PCR Beads (GE Healthcare), and the cDNA was amplified by PCR under the following conditions: 30 cycles of 95°C for 15 s, 60°C for 15 s, and 72°C for 60 s. The primer sets were as follows: 5′-GCCTCACGTCGTTACACTC-3′ (AN forward) and 5′-CCTGTTGCCTACTGGTGGAT-3′ (AN reverse); and 5′-AGAGATTCAGATGCCCAGAAGTCTTGTTCC-3′ (ACTIN2 forward) and 5′-AACGATTCCTGGACCTGCCTCATCATACTC-3′ (ACTIN2 reverse).

Dark and Light Treatments

For dark treatment, detached leaves were placed on GM plates and incubated in the dark for at least 16 h. For light treatment, detached leaves were irradiated with 100 µmol m⁻² s⁻¹ blue light (470 nm) using a light-emitting diode light source system (IS-mini; CCS). Light intensity was measured using a quantum sensor (LI-190SA; LI-COR).

Visualization of Nuclei, Cell Walls, and Chloroplasts

Samples were fixed in buffer solution (50 mm PIPES, 10 mm EGTA, and 5 mm MgSO₄, pH 7) containing 2% (w/v) formaldehyde (Wako) and 0.3% (w/v) glutaraldehyde (Wako) for 1 h. The fixed samples were stained with 5 µg mL⁻¹ Hoechst 33342 (CalBiochem) diluted in buffer solution supplemented with 0.03% (v/v) Triton X-100 for at least 1.5 h.

Measurement of Nuclear Area

The projection area of the nucleus was measured as reported previously (Iwabuchi et al., 2016).

Visualization of Actin Filaments

Actin filaments were visualized in leaf samples that had been fixed as described for the visualization of nuclei (Iwabuchi et al., 2010). Samples were stained with 200 nm Alexa Fluor 488 phalloidin (Invitrogen) and 5 µg mL⁻¹ Hoechst 33342, diluted in buffer solution supplemented with 0.03% (v/v) Triton X-100, for at least 1.5 h.

Lifeact-Venus/unp1 (F3) and Lifeact-Venus/unp2 (F3) lines were generated by crossing the Lifeact-Venus line and unp1 and unp2, respectively, and actin filaments were inspected in living cells. The Lifeact-Venus/unp2 line exhibited abnormally and significantly thicker actin bundles than the Lifeact-Venus line. There was also an overaccumulation of Lifeact-Venus protein in the Lifeact-Venus/unp2 line; therefore, the possibility that the thicker actin-bundle formation was caused by an overaccumulation of Lifeact-Venus rather than by the unp2 mutation could not be excluded. Given these facts, only phalloidin was applied for staining unp2.

To visualize actin filaments in transgenic plants expressing Lifeact-Venus, leaves were fixed in buffer solution (50 mm PIPES, 10 mm EGTA, and 5 mm MgSO₄, pH 7) containing 0.02% (w/v) paraformaldehyde (Wako) and 0.003% (w/v) glutaraldehyde (Wako) for 5 min to stop the movement of actin filaments.

Cells were imaged using a confocal scanning microscope (LSM800; Zeiss), and sequential images were taken from the outer periclinal walls to the inner periclinal walls of the cell along the optical z axis at 0.5-µm intervals. Maximum intensity projections were generated using the Fiji image-processing package.

Measuring the Movements, Occupancy, and Fluorescence Levels of Actin Filaments

To analyze lateral movements of the actin filament, pavement cells of transgenic plants expressing Lifeact-Venus were examined. Actin filaments were imaged at 5-s intervals for 60 s with a confocal laser scanning microscope (LSM800). The sequential images were aligned to correct experimental drift using the Fiji plug-in Align Slices in Stack (https://sites.google.com/site/qingzongtseng/template-matching-ij-plugin/tuto2#updates), as reported previously (Tseng et al., 2011). To evaluate the lateral movements of actin filaments, cross-correlation coefficients between the first image and the subsequent time-sequential images were determined with the Fiji plug-in Image CorrelationJ (http://www.gcsca.net/IJ/ImageCorrelationJ.html), as reported previously (Chinga and Syverud, 2007).

To evaluate actin filament occupancy in a cell, each pavement cell was outlined using Fiji. Actin filaments visualized by phalloidin staining were then extracted as skeletonized images using the Fiji plug-in LPX Filter2d (https://lpixel.net/products/lpixel-imagej-plugins/), as reported previously (Ueda et al., 2010). Actin filament occupancy was calculated as the percentage of pixel numbers of the skeletonized actin filaments versus those of cells using Fiji and Excel. Fluorescence levels of the skeletonized actin filaments were determined using Fiji.

Measuring the Velocity of Cytoplasmic Streaming

Cytoplasmic streaming was examined in pavement cells located above the midvein of the leaf. To estimate the velocity, particles present in the cytosol were imaged at 0.13-s intervals for 30 s using a fluorescence microscope (Axio Scope.A1; Zeiss) equipped with a CCD camera (VB-7010; Keyence). The greatest distance traveled by particles in a 5-s interval was determined using Fiji.

Measurement of the Circularity Index of Pavement Cells

Pavement cells were traced using Fiji to determine A and P, where A is the area and P is the perimeter of the cell. The circularity index of pavement cells was calculated as (4πA)/P², as reported previously (Iwabuchi et al., 2010). The circularity index reflects the extent to which a shape approaches a circle (a perfect circle has a circularity index of 1).

SDS-PAGE and Immunoblot Analysis

Protein extracts from mature leaves were subjected to SDS-PAGE followed by immunoblot analysis. Immunoreactive signals were detected with the ECL detection system (GE Healthcare), using an anti-actin antibody (clone 10-B3; Sigma-Aldrich) at a dilution of 1:2,000, an anti-microtubule antibody (clone B-5-1-2; Sigma-Aldrich) at a dilution of 1:2,000, and ECL anti-mouse IgG horseradish peroxidase-linked species-specific whole antibody (GE Healthcare) at a dilution of 1:1,000.

SDS-PAGE was performed using a SuperSep Phos-tag gel (Wako) or a conventional precast gel (Criterion XT; Bio-Rad). For dephosphorylation, protein extracts were treated with alkaline phosphatase (Sigma-Aldrich) at a dilution of 1:100 for 60 min at 30°C. Silver staining was performed using a Silver Staining Kit (GE Healthcare) according to the manufacturer’s instructions.

Immunoprecipitation

Immunoprecipitation was performed using a µMACS Epitope Tag Protein Isolation Kit (Miltenyi Biotec) as reported previously (Tamura et al., 2010). Seven-day-old transgenic plants expressing AN-GFP or free GFP were homogenized in buffer containing 50 mm Tris-HCl, pH 8, 150 mm NaCl, and 1% (v/v) Triton X-100. Homogenates were centrifuged at 20,400g at 4°C to remove cellular debris. The supernatants were mixed with magnetic beads conjugated to an anti-GFP antibody (Miltenyi Biotec) and incubated on ice for 15 min. The mixtures were applied to µ Columns (Miltenyi Biotec) in a magnetic field to capture the magnetic antigen-antibody complex. After extensive washing with buffer containing 50 mm Tris-HCl, pH 8, 150 mm NaCl, 1% (v/v) Igepal CA-630 (formerly Nonidet P-40), 0.5% (w/v) sodium deoxycholate, and 0.1% (w/v) SDS, immunoaffinity complexes were eluted with buffer containing 50 mm Tris-HCl, pH 6.8, 50 mm DTT, 1% (w/v) SDS, 1 mm EDTA, 0.005% (w/v) Bromophenol Blue, and 10% (v/v) glycerol.

Preparation of Peptides for Tandem Mass Spectrometry Analysis

Peptides were prepared for analysis as reported previously (Tamura et al., 2010). For in-gel digestion, the protein components of the immunoprecipitates were separated on a 2.5-cm-long SDS gel. The gel was cut into three fractions. Each excised gel fraction was treated with 100% (v/v) acetonitrile for 15 min and dried in a vacuum concentrator. The dried gel was treated with 10 mm DTT in 50 mm ammonium bicarbonate for reduction, followed by 55 mm 2-iodoacetamide in 50 mm ammonium bicarbonate for 30 min for alkylation. After washing, the gel was treated with 50% (v/v) acetonitrile in 50 mm ammonium bicarbonate and dried in a vacuum concentrator. The gel was treated with 10 µg mL⁻¹ trypsin in 50 mm ammonium bicarbonate and incubated overnight at 37°C. The digested peptides were recovered twice with 50 µL of 5% (v/v) formic acid in 50% to 70% (v/v) acetonitrile. The extracted peptides were combined and evaporated to 10 µL in a vacuum concentrator.

Mass Spectrometric Analysis and Database Searching

Liquid chromatography-tandem mass spectrometry analysis was performed using the LTQ-Orbitrap XL-HTC-PAL system and the Mascot server as described previously (Tamura et al., 2010).

Yeast Two-Hybrid Assay

Gene constructs were prepared as reported previously (Takagi et al., 2013). Briefly, the coding regions of AN, DYRKP-2B, and DYRKP-3 were amplified by PCR from wild-type cDNA using the following primer sets: 5′-ATGAGCAAGATCCGTTCGTCTG-3′ (AN forward) and 5′-ATCGATCCAACGTGTGATACC-3′ (AN reverse); 5′-ATGGCAGACCAAAGCTCTGTTG-3′ (DYRKP-2B forward) and 5′-AGCAGAGATTGGCTCGTATGGG-3′ (DYRKP-2B reverse); and 5′-ATGGCGGTTGATGTTAAATC-3′ (DYRKP-3 forward) and 5′-ATTGTAAGATGAAGAAGAAG-3′ (DYRKP-3 reverse). The amplified fragments were subcloned into the pENTR1A entry vector (Invitrogen) that had been digested with SalI and EcoRV using an In-Fusion HD Cloning Kit (Takara). The fragments were introduced into destination vector pDEST-GBKT7 or pDEST-GADT7 via the LR reaction. The Saccharomyces cerevisiae Y2H gold strain (Clontech) was transformed with these vectors. The transformed yeasts were first selected on synthetic defined (SD)/-Leu/-Trp plates. The selected yeasts were examined subsequently on either SD/-Leu/-Trp plates or SD/-Leu/-Trp/-Ade/-His/+X-α-gal/+aureobasidin A plates at 30°C. As negative controls, pGBKT7 and pGADT7 vectors were used. The yeast failed to proliferate when expressing either DYRKP-2A or DYRKP-1.

Inhibitor Treatment

Sample leaves were vacuum infiltrated using a syringe containing INDY (Sigma-Aldrich) solution. For the analysis of blue light-induced nuclear positioning, INDY was applied to the leaves for 1 h before the end of the dark treatment, and subsequent light irradiation was performed in the presence of the inhibitor. For the analysis of dark-induced nuclear positioning, sample leaves were irradiated with 100 µmol m⁻² s⁻¹ blue light for 3 h, and subsequent dark treatment was performed for 5 h in the presence of the inhibitor.

Statistical Analysis

Values are represented as means ± se, as determined using StatPlus. P values were determined using an unpaired Student’s t test.

Accession Numbers

Sequence data from this article can be found in the GenBank/EMBL data libraries under the following accession numbers: At5g09810 (ACTIN7), At3g18780 (ACTIN2), At1g49240 (ACTIN8), At1g01510 (AN), At3g17750 (DYRKP-1), At1g73460 (DYRKP-2A), At1g73450 (DYRKP-2B), and At2g40120 (DYRKP-3).

Supplemental Data

The following supplemental materials are available.

• Supplemental Figure S1. Nuclear projection areas of mesophyll and pavement cells of wild-type, unp1, and unp2 leaves in the dark.

• Supplemental Figure S2. Blue light-induced nuclear positioning in the wild type and unp1 and unp2 mutants.

• Supplemental Figure S3. Pleiotropic phenotypes of unp1 plants.

• Supplemental Figure S4. Alignment of actin proteins.

• Supplemental Figure S5. Nuclear positioning is not related to cell shape.

• Supplemental Figure S6. AN is localized to the cytosol under both dark and blue light conditions.

• Supplemental Figure S7. List of 116 proteins identified by mass spectrometry as being present in AN-GFP plants but not in free-GFP plants.

• Supplemental Movie S1. Lateral movements of actin filaments in the vicinity of nuclei in leaf pavement cells of Lifeact-Venus/wild-type and Lifeact-Venus/unp1 plants.

• Supplemental Movie S2. Lateral movements of actin filaments at the cell surface in leaf pavement cells of Lifeact-Venus/wild-type and Lifeact-Venus/unp1 plants.

• Supplemental Movie S3. Cytoplasmic streaming in leaf pavement cells of wild-type, unp1, and unp2 plants.