Novel plant SUN-KASH bridges are involved in RanGAP anchoring and nuclear shape determination

Abstract

Inner nuclear membrane Sad1/UNC-84 (SUN) proteins interact with outer nuclear membrane (ONM) Klarsicht/ANC-1/Syne homology (KASH) proteins, forming linkers of nucleoskeleton to cytoskeleton conserved from yeast to human and involved in positioning of nuclei and chromosomes. Defects in SUN-KASH bridges are linked to muscular dystrophy, progeria, and cancer. SUN proteins were recently identified in plants, but their ONM KASH partners are unknown. Arabidopsis WPP domain-interacting proteins (AtWIPs) are plant-specific ONM proteins that redundantly anchor Arabidopsis RanGTPase-activating protein 1 (AtRanGAP1) to the nuclear envelope (NE). In this paper, we report that AtWIPs are plant-specific KASH proteins interacting with Arabidopsis SUN proteins (AtSUNs). The interaction is required for both AtWIP1 and AtRanGAP1 NE localization. AtWIPs and AtSUNs are necessary for maintaining the elongated nuclear shape of Arabidopsis epidermal cells. Together, our data identify the first KASH members in the plant kingdom and provide a novel function of SUN-KASH complexes, suggesting that a functionally diverged SUN-KASH bridge is conserved beyond the opisthokonts.

Figure 1.

Structural and sequence similarity between KASH domains and the PNS tail of AtWIP1. C termini of animal and fungal KASH proteins are aligned with the C terminus of AtWIP1. Extension of the TMD and the PNS tail are indicated below the alignment. ClustalX color is assigned to the alignment for convenient comparison. Ce, Caenorhabditis elegans; Dd, Dictyostelium discoideum; Dm, Drosophila melanogaster; Hs, Homo sapiens; Mm, Mus musculus; Sp, Schizosaccharomyces pombe.

Figure 2.

Characterization of AtSUN–AtWIP interactions. (A) Domain organization of AtWIP1 and mutant derivatives. AtWIP1 has an N-terminal domain with an unknown function (cyan), an NLS (blue), a CCD-binding AtRanGAP1 (red), a predicted TMD (yellow), and a PNS tail (shown in residues). (B) Domain organization of AtSUN2 and deletion constructs. AtSUN2 has an N-terminal domain with an unknown function (cyan), an NLS (blue), a TMD (yellow), an unknown domain (white), a CCD (red), and a SUN domain (here split to an N-terminal part [green] and a C-terminal part [orange]) (A and B) Figures are drawn to scale. (C) AtWIP1 interacts with AtSUN2 through its PNS tail. (D) AtWIP1, AtWIP2, and AtWIP3 interact with AtSUN2. (E) AtWIP1 interacts with AtSUN1 through its PNS tail. (F) AtWIP1, AtWIP2, and AtWIP3 interact with AtSUN1. (G) AtSUN2 interacts with AtWIP1 through its SUN domain. (C–G) GFP- or CFP-tagged proteins were immunoprecipitated and detected by anti-GFP antibody. RFP-Myc– or Myc-Flag–tagged proteins were detected by anti-Myc antibody, and RFP-Flag–tagged proteins were detected by anti-Flag antibody. The input/IP ratio is 1:10. Numbers on the left indicate molecular mass in kilodaltons.

Figure 3.

FRAP analysis of the interaction between AtWIP1 and AtSUN1. (A) Recovery curves of GFP-AtWIP1, GFP-AtWIP1ΔVVPT, and GFP-TDF^AtWIP1. (B) Recovery curves of GFP-AtWIP1 coexpressed with RFP-Flag-AtSUN1 or RFP-Flag-AtSUN1ΔNSUN. (C) Recovery curves of GFP-AtWIP1 coexpressed with RFP-Myc-AtSUN2 or RFP-Myc-AtSUN2ΔNSUN. (D) Recovery curves of GFP-AtWIP1ΔVVPT coexpressed with RFP-Flag-AtSUN1 or RFP-Myc-AtSUN2. (E) Recovery curves of GFP-TDF^AtWIP1 coexpressed with RFP-Flag-AtSUN1 or RFP-Myc-AtSUN2. (A–E) Error bars represent SEM (n = 60 for GFP-AtWIP1 in C; n = 30 for all others). Asterisks at the end of each curve indicate significant statistical difference of the maximum recovery compared with the green curve in each figure (P < 0.01, using a t test). Otherwise, no statistical difference has been observed (P > 0.05, using a t test).

Figure 4.

AtSUNs are required for targeting AtWIP1 and AtRanGAP1 to the NE. (A) GFP-AtWIP1 or GFP-AtWIP1ΔVVPT signal in undifferentiated root cells (top row) and corresponding intensity profiles along the magenta arrows (bottom row). C1 and C2, cytoplasmic intensity 1 and 2, respectively; N1 and N2, nuclear intensity 1 and 2, respectively. Bars, 5 µm. (B) NLI ([N₁ + N₂]/[C₁ + C₂]) calculated using the intensities measured as shown in A. Asterisks indicate significant statistical difference between compared lines (P < 0.01, using a t test; n = 50). Error bars represent SEM. (C) AtRanGAP1-GFP signal in undifferentiated root cells (top row) and corresponding intensity profiles along the magenta arrows (bottom row). Bars, 5 µm. (D) NLI calculated as described in B, using intensities measured as in C. Asterisks indicate significant statistical difference between compared lines (P < 0.01, using a t test; n = 55). Error bars represent SEM. (E) AtSUN2, AtWIP1, and AtRanGAP1 are in the same complex. AtRanGAP1-GFP was immunoprecipitated and detected by anti-GFP antibody. AtWIP1 and RFP-Myc-AtSUN2 were detected with anti-AtWIP1 antibody and anti-Myc antibody, respectively. Numbers on the left indicate molecular mass in kilodaltons.

Figure 5.

Nuclear shape change in epidermal cells of sun1-KO sun2-KD and wip1-1 wip2-1 wip3-1 plants. (A) Comparison of nuclear shapes in trichomes, leaf epidermal cells, and mature root hair cells of wild type, sun1-KO sun2-KD, and wip1-1 wip2-1 wip3-1. Nuclei were DAPI stained, and mature root hair nuclei images in the bottom row are confocal maximum intensity projections using GFP-NLS-GFP as a nuclear marker. (B) Quantitative comparison of nuclear shape changes shown in A. Asterisks indicate significant statistical difference (P < 0.01, using a t test; n = 60 for leaf epidermal cells; n = 20 for trichomes; n = 55 for root hairs) compared with wild type. Error bars represent SEM. (C) A confocal maximum intensity projection showing a super-elongated nucleus in a wild-type mature root hair using WPP-GFP as an NE marker. Bars, 10 µm.