Supplemental Figure 1. Predicted Domain Organization of the AFH14 Protein. (A) Schematic representation of the predicted domain organization of AFH14. The PTEN (phosphatase tensin) domain is shown in dark gray, the FH1 domain in black, and the FH2 domain in light gray. (B) The full-length amino acid (aa) sequence of AFH14. Structural analysis of this sequence indicated the presence of an amino-terminal PTEN-related domain (double line), and two other common structural features, the FH1 domain (dotted line), which is rich in proline (bold), and the highly-conserved FH2 1
domain (underlined). The antigen sequence (340-502 aa) used to generate monoclonal antibodies for in vivo localization of AFH14 is shown in the gray box. 2
Supplemental Figure 2. Differences between the Amplified and Predicted AFH14 Coding Sequences (CDS). Comparison of the amplified AFH14 CDS and the predicted CDS from TAIR was performed using DNAMAN software. A 591 bp sequence in the FH1 domain predicted by TAIR was absent in the CDS of amplified AFH14. The figure shows the distinct region from sequence alignment, and the numbering is based on the predicted sequence from TAIR. 3
Supplemental Figure 3. FH1FH2 and AFH14 are Specifically Recognized by Polyclonal and Monoclonal AFH14 Antibody, Respectively. (A) Purification of recombinant FH1FH2 and characterization of the anti-afh14 polyclonal antibody. Lane 1 contains purified FH1FH2 protein, which corresponds to the major band of approximately 80 kda. The band was visualized by Coomassie Blue staining. Lane 2 shows recombinant FH1FH2 stained with polyclonal anti-afh14 antibodies (1:1000). (B) The specificity of the AFH14 monoclonal antibody was assessed by immunoblotting. Lane 1 contains protein extract from a 10-day-old wildtype seedling. Lane 2 contains protein extract from a 10-day-old afh14-1 seedling. The arrow indicates a band corresponding to the AFH14 protein in Lane 1 that is absent in Lane 2, demonstrating the specificity of the monoclonal antibody. 4
Supplemental Figure 4. Oryzalin Depolymerizes Microtubules in Uninduced Control BY-2. The spindle (A) and phragmoplast (B) microtubules were depolymerized in uninduced control BY-2 cells treated with oryzalin. Scale bar = 5 μm. 5
Supplemental Figure 5. AFH14-GFP Localization is Irregular or Dispersed in BY-2 Cells Treated with Oryzalin. AFH14-GFP-overexpressing BY-2 cells were treated with oryzalin, causing depolymerization of spindle and phragmoplast microtubules. Confocal microscopy revealed that the AFH14-GFP signal became irregular (A) or dispersed (B) as a result of oryzalin treatment. Scale bar = 5 μm. 6
Supplemental Figure 6. Overexpression of AFH14-GFP Increases the Stability of Mitotic Microfilaments. AFH14-GFP-overexpressing BY-2 cells (OX) and control mock-induced cells were treated with 200 nm LatB for 50 min. (A) Microfilaments remained partially intact in OX cells. (B) Microfilaments were deploymerized in control cells. Scale bar = 5 μm. (C) The number of mitotic cells with depolymerized microfilaments (MF) was analyzed in OX and control cells. OX cells exhibited a significantly higher number of cells containing polymerized microfilaments compared to control cells (t-test, P < 0.05). Error bars indicate SE (n > 300). Quantification includes results from three independent experiments. 7
Supplemental Figure 7. Phragmoplast Microfilaments Co-distribute with Microtubules in uninduced cells. Microfilaments were clearly present in the region of the phragmoplast microtubules in uninduced control Arabidopsis suspension culture cells. Scale bar = 5 μm. 8
Supplemental Figure 8. Analysis of the AFH14 Expression Pattern in Arabidopsis Tissues and Organs. The AFH14 expression pattern in Arabidopsis was analyzed by assessing GUS activity in the tissues and organs of transgenic Arabidopsis harboring an AFH14 promoter:gus gene fusion construct. AFH14 was highly expressed in the (A) shoot apex, (B) leaf primordial and guard cells, (C) trichome, (D) primary root, lateral root, and root hair, with the exception of the root tip, (E) whole floral buds, and (F) meiotic-stage floral buds. (G) Wildtype floral buds are shown as a control. (A-D) Scale bar = 20 μm; (E-G) Scale bar = 1 mm. 9
Supplemental Figure 9. Schematic Showing the AFH14 T-DNA Insertion Locus. (A) T-DNA insertion site in the AFH14 gene. The AFH14 gene upstream region is represented by a gray line, and exons and introns are represented by black boxes and lines, respectively. The 5 - and 3 -UTR regions are boxed in gray. (B) Mutants were identified by PCR using appropriate primers (LP, RP). Lanes 1 and 3 correspond to the afh14-1 and afh14-2 mutants, and lanes 2 and 4 correspond to WT plants. (C) RT-PCR using gene-specific primers indicated that no wildtype AFH14 transcripts were detectable in the afh14-1 (Lane 1) and afh14-2 (Lane 3) lines. Lanes 2 and 4 show results from WT plants. 10
Supplemental Figure 10. Microspore Formation is Defective in afh14-2 Arabidopsis Plants. (A) Normal tetrad from a WT anther. Arrow indicates the swollen domain. (B-D) A representative dyad (B), triad (C) and polyad (D) from the afh14-2 line. (E) Irregular tetrahedral arrangement of afh14-2 microspores. Scale bar = 5 μm. 11
Supplemental Figure 11. Microtubule Array Arrangement is Defective in afh14-2 Plants. Tubulin immunolocalization was analyzed in afh14-2 male meiocytes. (A) In metaphase I, the spindle was abnormal in afh14-2 male meiocytes. (B) In telophase I, the RMSs display a skewed configuration in afh14-2 plants. (C) In metaphase II, the spindles in afh14-2 plants were oriented in a parallel arrangement. (D) In advanced cytokinesis, the phragmoplast microtubules were not clearly present in the afh14-2 mutant. Scale bar = 5 μm. 12
Supplemental Figure 12. The FH1FH2 Recombinant Protein Nucleates Microfilaments in vitro. Globular actin (G-actin) was incubated with varying concentrations of FH1FH2 for 5 min, followed by centrifugation at 20,0000 g for 60 min. The amount of actin present in the supernatant (S) and pellet (P) was analyzed by SDS-PAGE. The presence of actin in the pellet in FH1FH2-treated samples indicated that FH1FH2 can nucleate actin polymerization to form microfilaments. Results are representative of three independent experiments. The gel was stained with Coomassie Blue. 13
Supplemental Figure 13. Microtubule and Microfilament Organization in the Presence of FH1FH2. Microtubules (A-F, red) or microfilaments (G-L, green) were incubated with active FH1FH2 or heat-denatured FH1FH2. White lines and numbers indicated the diameter of 14
the filaments or filament bundles. (A) Individual taxol-stabilized microtubules were incubated with heat-denatured FH1FH2. (B) Addition of active FH1FH2 resulted in bundle formation. (C) Staining using anti-afh14 antibodies revealed the presence of dot-like structures along microtubule bundles. (D, E) Transmission electron microscopy (TEM) analysis of the reactions described in (A) and (B), respectively, confirmed results from immunofluorescence microscopy. (F) Addition of 200 mm NaCl to (B) resulted in dissociation of microtubule bundles. (G) Individual microfilaments were incubated with heat-denatured FH1FH2. (H) Addition of active FH1FH2 resulted in bundle formation. (I) Staining using anti-afh14 antibodies revealed the presence of dot-like structures along microfilament bundles. (J, K) TEM analysis of the reactions described in (G) and (H), respectively, confirmed results from immunofluorescence microscopy. (L) Addition of 200 mm NaCl to (H) resulted in dissociation of microfilament bundles. Scale bar = 5 μm for fluorescence microscopy images, and scale bar = 50 nm for TEM images. 15
Supplemental Figure 14. FH1FH2 Inhibits Cold-induced Disassembly of Microtubules in vitro. solutions containing microtubules and FH1FH2 were incubated at 4ºC, and microtubule polymerization was analyzed by confocal microscopy. (A, B) In the presence of heat-denatured FH1FH2, individual microtubules present prior to cold treatment (A) were fully disassembled after incubation at 4ºC (B). (C, D) In the presence of active FH1FH2, the majority of the microtubule bundles present prior to cold treatment (C) persisted after incubation at 4ºC (D) for 30 min or 2 h. Scale bar = 5 μm. 16
Supplemental Figure 15. The Thickness of Filaments and Filament Bundles. The thickness of the filaments and filament bundles corresponding to the result of negative stainning electron microscopy were measured. Microtubule (MT), microfilament (MF), microtubule bundles (MTB), microfilament bundles (MFB) and the mixed microtubule and microfilament bundles (MT-MF) after heat-denatured FH1FH2 (dfh1fh2) or active FH1FH2 (afh1fh2) incubation. The active FH1FH2 bundled the filaments, however, the heat-denatured FH1FH2 did not bundled them. Error bars indicate SE (n > 20). 17
Supplemental Figure 16. Microtubule and Microfilament Organization in the Presence of the FH2 Recombinant Protein. Microtubules (red) or microfilaments (green) were incubated with (1 μm) active FH2 or heat-denatured FH2. (A) Individual taxol-stabilized microtubules were incubated with heat-denatured FH2. (B) Addition of active FH2 resulted in bundle formation. (C) Staining using anti-afh14 antibodies revealed the presence of dot-like structures along microtubule bundles. (D) Addition of 400 mm NaCl to (B) resulted in dissociation of microtubule bundles. (E) Individual microfilaments were incubated with heat-denatured FH1FH2. (F) Addition of active FH1FH2 resulted in bundle formation. (G) Staining using anti-afh14 antibodies revealed the presence of dot-like structures along microfilament bundles. (H) Addition of 400 mm NaCl to (F) resulted in dissociation of microfilament bundles. (I) FH2 was added to reaction mixtures of microtubules and microfilaments, the two types of bundles co-localized. Scale bar = 5 μm. 18