Novel Arabidopsis microtubule-associated proteins track growing microtubule plus ends

Abstract

Background: Microtubules (MTs) are polarized polymers with highly dynamic plus ends that stochastically switch between growth and shrinkage phases. In eukaryotic cells, a plethora of MT-associated proteins (MAPs) regulate the dynamics and higher-order organization of MTs to mediate distinct cellular functions. Plus-end tracking proteins (+TIPs) are a group of MAPs that specifically accumulate at the growing MT plus ends, where they modulate the behavior of the MT plus ends and mediate interactions with cellular targets. Although several functionally important + TIP proteins have been characterized in yeast and animals, little is known about this group of proteins in plants.

Results: We report here that two homologous MAPs from Arabidopsis thaliana, Growing Plus-end Tracking 1 (GPT1) and GPT2 (henceforth GPT1/2), contain basic MT-binding regions at their central and C-terminal regions, and bind directly to MTs in vitro. Interestingly, GPT1/2 preferentially accumulated at the growing plus ends of cortical MTs in interphase Arabidopsis cells. When the GPT1/12-decorated growing plus ends switched to rapid depolymerization, GPT1/2 dissociated from the MT plus ends. Conversely, when the depolymerizing ends were rescued and started to polymerize again, GPT1/2 were immediately recruited to the growing MT tips. This tip tracking behavior of GPT proteins does not depend on the two established plant + TIPs, End-Binding protein 1 (EB1) and SPIRAL1 (SPR1).

Conclusions: The Arabidopsis MAPs GPT1 and GPT2 bind MTs directly through their basic regions. These MAPs track the plus ends of growing MTs independently of EB1 and SPR1 and represent a novel plant-specific + TIP family.

Keywords: Arabidopsis; EB1; GPT1; GPT2; Plus-end tracking proteins; SPIRAL1.

Fig. 1

The positively charged domains of GPT1/2 bind MTs in onion epidermal cells. Full-length and truncated fragments of GPT1/2 were fused to GFP at their C-termini, and transiently expressed in onion epidermal cells. a and d Charge plots of GPT1 (a) and GPT2 (d). b and e Protein structures of GPT1 (b) and GPT2 (e). The positions of basic amino acid residues are indicated by blue lines. The N-terminal (N) regions are negatively charged, whereas the middle (M) and C-terminal (C) regions are positively charged. c and f Localization of GFP-fused GPT fragments to cortical MTs is indicated by ++ (strong localization), + (substantial localization), +/- (weak localization), and – (no localization). The numerator and denominator show the number of cells with positive localization patterns and the number of total cells that expressed GFP fusion proteins, respectively. The GPT2-GFP fusion (f) was co-expressed with tagRFP-MAP4 to visualize MTs. Left panels, GPT2-GFP; middle panels, tagRFP-MAP4; right panels, merged images. Scale bars, 10 μm

Fig. 2

Recombinant GPT1 and GPT2 bind to MTs in vitro. (a) Purified MBP-fused GPT1 and GPT2 proteins were incubated with or without taxol-stabilized MTs, pelleted by ultracentrifugation, and analyzed by SDS-PAGE and Coomassie Brilliant Blue staining. The positions of full-length MBP-GPT1 and MBP-GPT2 are indicated by arrowheads, and the MT-binding partial degradation products of MBP-GPT1 are indicated with an asterisk. Tubulin is also indicated. S, supernatant fraction; P, pellet fraction; and M, size markers. (b) Quantitative analysis of the binding of MBP-GPT2 to MTs. Various concentrations of purified MBP-GPT2 was mixed with taxol-stabilized MTs and then subjected to co-sedimentation analysis, as in A. Assuming that there is one GPT2 binding site on the tubulin dimer, the binding equation (see Methods) was fitted

Fig. 3

GPT1/2 decorate mitotic MT structures in Arabidopsis root epidermal cells. GFP fusions (green) were expressed in transgenic Arabidopsis plants harboring the mCherry-TUB6 MT marker (magenta). Epidermal cells in the cell division zone of the roots of seedlings were analyzed. Dividing cells at preprophase, metaphase, anaphase, and cytokinesis are shown. a GFP-GPT1 was expressed under the control of the cauliflower mosaic virus 35S promoter. b GPT2-GFP was expressed under the control of the endogenous regulatory elements of GPT2. Scale bars, 5 μm

Fig. 4

GPT1 tracks the growing plus ends of cortical MTs. a The subcellular localization of GFP-GPT1 (green) and mCherry-TUB6, which labels cortical MTs (magenta), was analyzed in interphase cells of the Arabidopsis root epidermis. Scale bar, 5 μm. b Average fluorescence intensity (FI) profiles of GFP-GPT1 and mCherry-TUB6 were obtained by analyzing and plotting data from 20 MT images. The data were normalized and the peak intensity of GFP-GPT1 was set to 1. The error bars indicate SEM. c Representative time-lapse sequence and corresponding kymograph of a MT that underwent catastrophe at t = 14 s. GFP-GPT1 disappeared from the tip immediately after the MT began to depolymerize. Dashed yellow arrows indicate the point at catastrophe occurred. Scale bar, 2.5 μm. d Representative time-lapse sequence and corresponding kymograph of a MT that underwent rescue at t = 12 s. GFP-GPT1 localized to the tip immediately after MT growth resumed. Dashed yellow arrows indicate the point at which rescue occurred. Scale bar, 2.5 μm

Fig. 5

GPT1 and EB1 label similar regions of growing MT ends. a Subcellular localization of GFP-GPT1 (green) and EB1-mCherry (magenta) was analyzed in interphase cells of the Arabidopsis root epidermis. Scale bar, 5 μm. b Representative time-lapse sequences. Left, mCherry signals were analyzed first, followed by GFP signals. Right, GFP signals were analyzed first, followed by mCherry signals. Scale bars, 2.5 μm. c and d Average FI profiles of GFP-GPT1 and EB1-mCherry were obtained by analyzing and plotting data from 20 MT images. The data were normalized and the peak intensities of GFP-GPT1 and EB1-mCherry were set to 1. The error bars indicate SEM. c mCherry signals were analyzed first, followed by GFP signals. d GFP signals were analyzed first, followed by mCherry signals

Fig. 6

GPT1 does not require EB1 or SPR1 to track the MT end. a and b GFP-GPT1 was expressed in the Arabidopsis eb1a eb1b eb1c triple mutant. a The subcellular localization of GFP-GPT1 was analyzed in interphase cells of the Arabidopsis root epidermis. Scale bar, 5 μm. b An average FI profile of GFP-GPT1 was obtained by analyzing and plotting data from 20 MT images. The data were normalized and the peak intensity of GFP-GPT1 was set to 1. The error bars indicate SEM. c and d GFP-GPT1 was expressed in the spr1 mutant that also expressed mCherry-TUB6 as a MT marker. c The subcellular localization of GFP-GPT1 and mCherry-TUB6 was analyzed in interphase cells of the Arabidopsis root epidermis. Scale bar, 5 μm. d Average FI profiles of GFP-GPT1 and mCherry-TUB6 were obtained by analyzing and plotting data from 20 MT images. The data were normalized and the peak intensity of GFP-GPT1 was set to 1. The error bars indicate SEM