Skip to main content
NCBI home page
The Plant Cell logoLink to The Plant Cell
. 2023 Nov 27;36(4):1056–1071. doi: 10.1093/plcell/koad294

A divergent tumor overexpressed gene domain and oligomerization contribute to SPIRAL2 function in stabilizing microtubule minus ends

Yuanwei Fan 1, Natasha Bilkey 2, Derek L Bolhuis 3, Kevin C Slep 4,c,✉,d, Ram Dixit 5,✉,c
PMCID: PMC10980349  PMID: 38011314

Abstract

The acentrosomal cortical microtubules (MTs) of higher plants dynamically assemble into specific array patterns that determine the axis of cell expansion. Recently, the Arabidopsis (Arabidopsis thaliana) SPIRAL2 (SPR2) protein was shown to regulate cortical MT length and light-induced array reorientation by stabilizing MT minus ends. SPR2 autonomously localizes to both the MT lattice and MT minus ends, where it decreases the minus end depolymerization rate. However, the structural determinants that contribute to the ability of SPR2 to target and stabilize MT minus ends remain unknown. Here, we present the crystal structure of the SPR2 N-terminal domain, which reveals a unique tumor overexpressed gene (TOG) domain architecture with 7 HEAT repeats. We demonstrate that a coiled-coil domain mediates the multimerization of SPR2, which provides avidity for MT binding, and is essential to bind soluble tubulin. In addition, we found that an SPR2 construct spanning the TOG domain, basic region, and coiled-coil domain targets and stabilizes MT minus ends similar to full-length SPR2 in plants. These results reveal how a TOG domain, which is typically found in microtubule plus-end regulators, has been appropriated in plants to regulate MT minus ends.

The Arabidopsis SPIRAL2 protein has a unique tumor overexpressed gene domain architecture, which, along with a basic region and coiled-coil domain, contributes to targeting and stabilizing microtubule minus ends.

Introduction

Plant cortical microtubules (MTs) play an essential role in cell morphogenesis by guiding the deposition of cellulose (Paredez et al. 2006) and matrix polysaccharides (Kong et al. 2015; Zhu et al. 2015) into the cell wall. This function relies on the creation and maintenance of distinctive cortical MT arrays in different cell types and their dynamic rearrangement in response to various internal and external stimuli (Oda 2015). Because the bulk of cortical MTs have free plus and minus ends, modulating the dynamics of both ends is important for array organization and remodeling.

The dynamics of both MT ends are controlled by a group of specialized proteins that recognize specific tubulin conformations at the MT plus end and minus end, respectively (Akhmanova and Steinmetz 2015; Nogales and Zhang 2016). The MT plus-end tracking proteins (+TIPs) specifically decorate growing plus ends and serve to stabilize plus ends and link them to other cellular structures (Galjart 2010). Likewise, the MT minus-end targeting proteins (−TIPs) stabilize free MT minus ends and attach them to other cellular structures (Akhmanova and Steinmetz 2019).

In animals, the calmodulin-regulated spectrin-associated protein (CAMSAP)/Patronin/Nezha family of −TIPs has been studied extensively. Different members of this protein family either track growing MT minus ends (Jiang et al. 2014), progressively accumulate on the MT lattice formed by the growth at the minus end (Jiang et al. 2014), or bind to and inhibit the growth of minus ends (Hendershott and Vale 2014). This protein family is distinguished by the presence of a C-terminal CAMSAP1, KIAA1078, and KIAA1543 (CKK) domain, which confers MT minus-end targeting (Atherton et al. 2017, 2019).

Plants lack the CAMSAP/Patronin/Nezha protein family. Recently, the plant-specific TORTIFOLIA1/SPIRAL2 (SPR2) protein was found to autonomously localize to and stabilize MT minus ends (Fan et al. 2018; Leong et al. 2018; Nakamura et al. 2018). The loss of SPR2 greatly enhances the frequency and rate of MT minus-end depolymerization and impairs the ability of the cortical MT array to dynamically reorient in response to blue light perception (Fan et al. 2018; Leong et al. 2018; Nakamura et al. 2018). Notably, the SPR2 protein differs from the CAMSAP/Patronin/Nezha protein family in both structure and dynamics. SPR2 lacks the CKK domain that is necessary for the CAMSAP/Patronin/Nezha proteins to recognize MT minus ends. In addition, unlike the CAMSAP/Patronin/Nezha proteins, SPR2 tracks depolymerizing MT minus ends both in vivo and in vitro and localizes to the MT lattice and, to a lesser extent, to the growing MT plus ends (Fan et al. 2018; Nakamura et al. 2018).

The Arabidopsis (Arabidopsis thaliana) SPR2 protein contains 5 predicted Huntington, elongation factor 3, phosphatase 2A, target of rapamycin 1 (HEAT) repeats at the amino terminus followed by a basic region and coiled-coil domain (Fan et al. 2018). A HEAT repeat (HR) domain followed by a basic region is reminiscent of the tumor overexpressed gene (TOG) domain–containing MT regulatory proteins XMAP215 and CLASP, which localize primarily to MT plus ends in both plants and animals (Mimori-Kiyosue et al. 2005; Ambrose et al. 2007; Brouhard et al. 2008; Kawamura and Wasteneys 2008). A TOG domain is characterized by the presence of 6 HRs that form a paddle-shaped structure, and in the case of the yeast XMAP215 member, Stu2, it binds to a tubulin dimer with high affinity (Al-Bassam et al. 2006, 2007; Slep and Vale 2007; Ayaz et al. 2012). While TOG domains are best known for their ability to bind to free tubulin subunits (Ayaz et al. 2012, 2014), the structural differences in some of the TOG domains of XMAP215 and CLASP proteins confer binding to MT lattice-incorporated tubulin (Fox et al. 2014; Byrnes and Slep 2017; Leano and Slep 2019). The proteins that contain more than 2 TOG domains (e.g. human and plant XMAP215 homologs) function as monomers (Currie et al. 2011; Widlund et al. 2011; Lechner et al. 2012), whereas those that contain only 1 or 2 TOG domains (e.g. yeast XMAP215 homologs and Cep104) function either as dimers or oligomers (Gard et al. 2004; Al-Bassam et al. 2006; Rezabkova et al. 2016), with the possible exception for the mitogen-activated protein kinase activator MEKK1 (Filipčík et al. 2020). Hence, the involvement of multiple TOG domains is an important characteristic of proteins that regulate MT dynamics.

Here, we report structural, biochemical, and cell biological data that together identify structural features that are important for the function of SPR2. We found that the N terminus of SPR2 contains a single tubulin-binding TOG domain followed by a basic region that confers binding to the MT lattice. X-ray crystallography revealed that the TOG domain of SPR2 is structurally unique with 7 HRs. In addition, we demonstrate that the multimerization of SPR2 by the coiled-coil domain increases its MT-binding affinity and is essential for SPR2 to bind and recruit soluble tubulin dimers to an MT. An SPR2 construct consisting of the TOG domain, basic region, and coiled-coil domain was able to stably localize to the minus ends of cortical MTs and inhibit their depolymerization rate when expressed in the A. thaliana spr2-2 mutant. Together, our results establish the SPR2 as a distinct TOG domain-containing protein, which has evolved to regulate the dynamics of MT minus ends.

Results

The N-terminal region of SPR2 mediates MT lattice binding and minus-end localization

To identify the domains that enable SPR2 to bind to and stabilize MT minus ends, we generated truncated versions of the SPR2 protein, which were tagged with GFP at their C terminus (Fig. 1A). The full-length and truncated GFP-labeled SPR2 proteins were recombinantly expressed and purified (Fig. 1B) and were used to assess MT binding in functional reconstitution experiments. In these in vitro experiments, we used GMPCPP-stabilized MT fragments (called “seeds”) to initiate dynamic MT polymers. The MT seeds contained a high proportion of rhodamine-tubulin to make them distinguishable from the dynamic polymer that contained less rhodamine-tubulin. We found that the N-terminal fragment of SPR2 that lacks the basic region (1 to 276 aa) does not bind to MTs (Fig. 1C). The inclusion of a portion of the basic region (1 to 340 aa) conferred weak binding to the GMPCPP-stabilized MT lattice and dynamic GDP-state MT lattice but no punctate minus end localization. The 1 to 400 aa fragment that contains the entire basic region tended to show slightly higher signal along both the GMPCPP-MT lattice and GDP-MT lattice compared with the 1 to 340 aa fragment (Fig. 1, C and G). Notably, a larger N-terminal fragment, which includes the complete basic region and an unstructured 100 aa region preceding the coiled-coil domain (1 to 500 aa), showed punctate localization along the GMPCPP-MT lattice and at the minus end of the GDP-MT lattice (Fig. 1C). While this pattern of MT localization was similar to full-length SPR2 (Fig. 1F), the extent of binding to the GMPCPP-MT lattice and signal intensity of puncta at the GDP-MT minus end were lower for the 1 to 500 aa fragment compared with full-length SPR2 (Fig. 1, G and H).

Figure 1.

Figure 1.

Open in a new tab

MT binding and minus-end localization of SPR2 fragments. A) Schematic representation of the domain architecture of SPR2 protein and the various truncations used in this study. B) Coomassie Blue–stained SDS–PAGE gels of recombinantly expressed and purified GFP-tagged SPR2 protein fragments. Dots indicate the expected size of each protein. C to F) Representative images of the localization of 500 nM GFP-tagged N-terminal fragments C), N terminus and coiled-coil domain containing fragments D), C-terminal fragments E), and full-length SPR2 F) along rhodamine-labeled MTs (Rho). Icons on top of each image indicate the position of the brightly labeled GMPCPP-stabilized seed (dark red) and dimly labeled dynamic polymer (light red), and the orientation of the plus and minus end. Scale bar = 2 μm. Arrowheads indicate SPR2-GFP signal at the minus end of dynamic GDP-state MT lattice. G) GFP signal intensity on GMPCPP-MT seeds. H) Signal intensity of GFP puncta at minus ends of GDP-MT lattice. Letters in G) and H) indicate statistically distinguishable groups determined by Kruskal–Wallis test with Dunn's multiple comparison test (P < 0.05).

The inclusion of the coiled-coil domain dramatically increased MT binding of the N-terminal fragment of SPR2 (1 to 570 aa). The 1 to 570 aa fragment occasionally showed punctate GMPCPP-MT lattice binding but more commonly showed bright continuous signal with a preference for the GMPCPP-stabilized MT lattice over the GDP-state dynamic lattice (Fig. 1, D and G). The deletion of the Ser/Thr-rich region at the extreme N terminus (38 to 570 aa) does not perturb GMPCPP-MT lattice binding and punctate localization at the minus end of GDP-MT lattice (Fig. 1, D, G, and H). However, a larger deletion that encompasses the HR region (281 to 570 aa) was unable to bind to MTs (Fig. 1D).

The C-terminal region of the SPR2 protein, with or without the coiled-coil domain, does not bind to MTs (Fig. 1E). Comparing the MT binding of the various SPR2 fragments with the full-length SPR2 protein indicated that the 38 to 570 aa fragment most closely mimics the pattern of GMPCPP-MT lattice binding and punctate GDP-MT minus end localization of full-length SPR2 (Fig. 1, G and H).

Multimerization of SPR2 enhances MT binding

Given the striking difference in the extent of MT decoration between the 1 to 500 and 1 to 570 aa fragments of SPR2, we investigated whether the coiled-coil domain increases the MT binding affinity of SPR2. Using MT cosedimentation assays, we found that the 1 to 500 aa fragment shows nonsaturating MT binding with an apparent KD of 17.4 μM (Fig. 2, A and C). By contrast, the 1 to 570 aa fragment bound to MTs with an apparent KD of 1.3 μM (Fig. 2, B and D). Therefore, the coiled-coil domain increases the MT binding affinity of SPR2 by at least one order of magnitude.

Figure 2.

Figure 2.

Open in a new tab

Multimerization state of SPR2. A and B) Coomassie Blue-stained SDS–PAGE gels of MT cosedimentation assays. Two micromolars of (1 to 500)-GFP A) or (1 to 570)-GFP B) were coincubated with increasing concentrations of taxol-stabilized MTs, 0.1 mg/mL BSA, and 20 µM taxol for 20 min. The arrowheads identify the different protein bands. C and D) MT-binding curves for (1 to 500)-GFP C) or (1 to 570)-GFP D), respectively. Each data point represents the mean ± Sd from 3 independent experiments. The data were fit to a model assuming single-site-specific binding yielding apparent KD values of 17.4 and 1.3 μM for (1 to 500)-GFP and (1 to 570)-GFP, respectively. E to G) SECMALS analysis of SPR2 33 to 333 E), SPR2 503 to 566 F), and full-length SPR2 G). Plots show the elution profile of each construct from a size exclusion column as measured using the refractive index (y axis at right) over time. The experimentally determined mass is plotted in kDa (MW, y axis at left) over time for each main peak. The average mass (±Sd) is indicated. For each construct, dashed lines indicate the monomeric formula weight. For the traces of the SPR2 503 to 566 and the SPR2 1 to 864 constructs, additional dashed lines indicate the MW of the theoretical dimer, trimer, and tetramer species. The plots shown are representative of experiments run in duplicate. S, supernatant; P, pellet.

Yeast 2-hybrid and bimolecular fluorescence complementation experiments have demonstrated that SPR2 interacts with itself (Haikonen et al. 2013; Fan et al. 2018). In addition, photobleaching analysis indicates that SPR2 exists primarily as multimers (Fan et al. 2018). To determine whether the coiled-coil domain is responsible for the multimerization of SPR2, we purified the SPR2 N-terminal domain (33 to 333 aa), central domain (503 to 566 aa), and full-length protein and analyzed each construct using size exclusion chromatography multiangle light scattering (SECMALS). The SPR2 N-terminal domain (Fig. 2E) eluted as one main peak with an experimentally determined mass (33 ± 3 kDa) similar to the monomeric formula weight (32.8 kDa). By contrast, the SPR2 central region containing the coiled-coil domain eluted as a bimodal peak (Fig. 2F). Analysis of the experimental mass across the first portion of the bimodal peak yielded a mass of 32 ± 7 kDa, while the later portion of the bimodal peak yielded a mass of 24 ± 2 kDa. The predicted masses of dimeric (19.2 kDa), trimeric (28.9 kDa), and tetrameric (38.5 kDa) species indicate that the SPR2 central domain forms a higher order oligomer, potentially a tetramer formed as a dimer of dimers. The full-length SPR2 eluted as a single main peak with an experimental mass of 381 ± 2 kDa, aligned with a predicted tetramer mass of 376 kDa (Fig. 2G). Prior SECMALS analysis of the SPR2 C-terminal domain from Physcomitrium patens and A. thaliana (649 to 864 aa) revealed a monomeric domain (Bolhuis et al. 2022; Ohno et al. 2023). These results indicate that the central coiled-coil domain mediates tetramerization and that regions flanking residues 503 to 566 contribute to tetramer stability.

The 1 to 500, 1 to 570, and 38 to 570 aa fragments of SPR2 inhibit the polymerization dynamics of MT minus ends

To determine whether the 1 to 500, 1 to 570, and 38 to 570 aa fragments of SPR2, which showed punctate localization at MT minus ends, affected the dynamics of MT plus and minus ends, we conducted kymograph analysis (Fig. 3A). We included the 1 to 400 aa fragment in this analysis to assess whether its uniform localization along the lattice influences MT dynamics.

Figure 3.

Figure 3.

Open in a new tab

Effect of SPR2 fragments on MT dynamics in vitro. A) Kymographs of MTs in the presence of 500 nM indicated SPR2 proteins. The arrowheads in the GFP channel kymographs indicate SPR2 signal at dynamic minus ends. Horizontal scale bar = 5 µm, vertical scale bar = 100 s. B) Polymerization dynamics of MT minus ends in vitro. C) Polymerization dynamics of MT plus ends in vitro. In B) and C), values are means ± Sd (n, number of events for growth and shortening rates; n, total number of kymographs for percent growth and shortening times). The letters indicate statistically distinguishable groups as determined by the Student's t-test (P < 0.05).

Under our experimental conditions, the minus ends of MTs show growth-biased polymerization dynamics in vitro (Fig. 3B), in contrast to the shortening-biased polymerization dynamics of cortical MT minus ends in plants (Fan et al. 2018; Nakamura et al. 2018). As reported previously (Fan et al. 2018), the full-length SPR2 protein greatly reduced the growth time of MT minus ends in vitro and modestly decreased the shortening time and rate compared with tubulin-alone control (Fig. 3B). Similarly, the 1 to 500, 1 to 570, and 38 to 570 aa fragments significantly decreased the growth time, shortening time, and shortening rate of minus ends when they localized to minus ends as puncta (Fig. 3, A and B; Supplemental Fig. S1). Unexpectedly, the 1 to 570 aa fragment, which showed the highest MT binding, inhibited the MT minus-end dynamics to a lesser degree than the 1 to 500 and 38 to 570 aa fragments (Fig. 3B).

The 1 to 400 aa fragment impacted the MT minus-end growth and shortening dynamics significantly less than the 1 to 500, 1 to 570, and 38 to 570 aa SPR2 fragments (Fig. 3B). All the SPR2 proteins used in this study altered the dynamics of MT plus ends to a small degree (Fig. 3C), in agreement with a previous study, which showed a small effect of full-length SPR2 on the MT plus-end dynamics in vitro (Yao et al. 2008). Given the lack of plus-end localization of the SPR2 protein in vitro (Fan et al. 2018 and this study), we propose that the modest changes in plus-end dynamics are likely an indirect effect of these proteins binding to the MT lattice.

SPR2 binds to free tubulin dimers

The Arabidopsis SPR2 protein is predicted to contain 5 HRs near the N terminus (Fig. 1A), flanked by predicted, conserved pairs of helices. We wondered whether SPR2 has an additional, cryptic 6th HR, which would collectively constitute a TOG domain. Since TOG domains are best known for their ability to bind to free tubulin dimers (Ayaz et al. 2012, 2014), we conducted pulldown experiments to test this possibility with full-length SPR2-GFP. For the experiments in this section, we used freshly solubilized tubulin below the critical concentration and without adding GTP to prevent the formation of short MTs in the solution.

We found that SPR2-GFP bound directly to soluble GDP-state tubulin (Fig. 4A). To determine the structural elements required for SPR2 to bind to free tubulin and to investigate whether SPR2 can simultaneously bind to free tubulin and MT lattice, we used a total internal reflection fluorescence microscopy-based assay. In this assay, unlabeled GMPCPP-MTs were coincubated with GFP-labeled SPR2 protein and GDP-state rhodamine-labeled tubulin (Fig. 4B). In the absence of SPR2 protein, rhodamine-labeled tubulin did not show filamentous localization, indicating that it does not incorporate into or bind to the GMPCPP lattice (Fig. 4C). However, in the presence of full-length SPR2-GFP protein, we observed a punctate rhodamine-tubulin signal that coincided with SPR2-GFP puncta along the MT lattice (Fig. 4, C to E). Similarly, rhodamine-tubulin puncta consistently colocalized with 1 to 570 and 38 to 570 aa SPR2 puncta on the MT lattice (Fig. 4, C to E). However, 1 to 500 and 1 to 400 aa SPR2 fragments were unable to recruit rhodamine-tubulin to the MT lattice (Fig. 4, C and D). Therefore, we conclude that multimerization is required for SPR2 to bind and recruit tubulin dimers to the MT lattice.

Figure 4.

Figure 4.

Open in a new tab

SPR2 binds soluble tubulin dimers. A) Coomassie Blue-stained gel of an in vitro pulldown experiment with GFP-tagged full-length SPR2 incubated with soluble tubulin. B) Schematic diagram of the in vitro assay to determine whether SPR2 and various fragments can bind simultaneously to MT lattice and tubulin subunits. C) Micrographs of 100 nM of the indicated GFP-tagged SPR2 proteins incubated with unlabeled GMPCPP-stabilized MTs and 2.5 µM 1:9 rhodamine-labeled porcine tubulin. Scale bars = 5 μm. D) Pearson's correlation coefficient between the SPR2-GFP and rhodamine-tubulin particles. E) Mander's coefficient for the extent of rhodamine signal colocalized with GFP and the extent of GFP signal colocalized with rhodamine. Letters in D) and E) indicate statistically distinguishable groups determined by Kruskal–Wallis test with Dunn's multiple comparison test (P < 0.05).

The SPR2 N-terminal domain is an architecturally distinct TOG domain

To gain insight into the architecture of the SPR2 N-terminal domain, we determined the crystal structure of a construct spanning residues 33 to 333 aa to 2.5 Å resolution (for crystallographic statistics, see Supplemental Table S1). The structure was solved by molecular replacement using search model coordinates obtained from AlphaFold (Jumper et al. 2021; Varadi et al. 2022). The final, refined model (R = 0.214, Rfree = 0.259, see final 2Fo−Fc electron density in (Supplemental Fig. S2A) encompasses residues 33 to 327 and consists of 7 HRs, designated as HRs A to G. Together, these HRs structurally resemble a TOG domain (Fig. 5A), although its α-solenoid includes an additional HR relative to the 6 found in TOG structures determined to date. The refined model aligns well with the AlphaFold prediction model (Cα RMSD = 1.1 Å), with differences localized to the HRs C and D inter-HEAT loop and the relative positioning of HR A and HR G (Supplemental Fig. S2, B and C). Like tubulin-binding TOG domains, the SPR2 N-terminal domain has an extended paddle-like conformation due to an offset in the α-solenoid structure between HR C and HR D, which limits the twist common to HR domains. Amino acid conservation maps primarily to the face of the domain formed by intra-HEAT loops, which has an overall basic charge that would complement the negatively charged MT exterior (Fig. 5B; Supplemental Figs. S3 and S4).

Figure 5.

Figure 5.

Open in a new tab

SPR2 contains a bona fide TOG domain with a unique architecture. A) Cartoon diagram of the SPR2 TOG domain, containing HRs A to G. B) SPR2 TOG domain as in A) after a 90° rotation about the x axis in the cartoon (above) and surface (middle: mapping conservation; below: electrostatics) representation. C) SPR2 (above: oriented as in A); below: after a 90° rotation) aligned with Stu2 TOG2 (Ayaz et al. 2014) over HRs B to D. Key structural differences are indicated with arrows. D) Comparison of SPR2 (above, residues colored according to conservation as in B)) and Stu2 TOG2 (below, showing key residues involved in tubulin binding) intra-HR residues. E to I) Comparison of Stu2 TOG2 bound to tubulin E) (Ayaz et al. 2014), and SPR2 TOG modeled on tubulin after alignment to Stu2 TOG2 HR B to D without F) and with Stu2 TOG2 shown G to I).

We next compared the SPR2 N-terminal domain with the Stu2 TOG2 structure from the Stu2 TOG2:αβ-tubulin complex, since the tubulin-binding determinants observed in this complex are well defined (Ayaz et al. 2014). The SPR2 N-terminal domain aligned best with Stu2 TOG2 when the alignment was focused on HRs B to D (Dali Z score 15.3; Cα RMSD = 2.1 Å, 113 of 124 residues aligned). When superimposed, the tubulin-binding intra-HEAT loops of Stu2 TOG2 align well with those of SPR2, and the Stu2 residues involved in tubulin-binding have homology to many of the SPR2 residues at equivalent positions (Fig. 5, C and D).

Significant structural differences are noted in the relative positioning of the flanking HRs (Fig. 5C). SPR2 HR A is rotated around the α-solenoid axis relative to Stu2 TOG2 HR A, while SPR2 HRs E to G are shifted along the α-solenoid axis relative to Stu2 TOG2 HRs E and F. The SPR2 HR E is shorter than the Stu2 TOG2 HR E, and its compaction into the α-solenoid effectively positions SPR2 HR F to structurally mimic Stu2 HR E, though positioned further along the α-solenoid axis (Fig. 5C). Key Stu2 TOG2 HR E tubulin-binding residues have position-equivalent homologs in the SPR2 HR F intra-HEAT loop, suggesting that SPR2 may use its elongated TOG architecture to bind a structural state of tubulin that is distinct from the curved tubulin conformation engaged by Stu2 TOG2 (Fig. 5D; Ayaz et al. 2014). Given the structural homology of Stu2 TOG2 and the structural variation observed in TOG structures determined to date (Al-Bassam et al. 2007; Maki et al. 2015; Byrnes and Slep 2017; Leano and Slep 2019), we designate the SPR2 N-terminal domain a bona fide TOG domain with distinct structural attributes.

Structural modeling suggests that the SPR2 TOG domain uses unique structural elements to bind to αβ-tubulin

Our in vitro data demonstrate that in the absence of the central coiled-coil domain, the SPR2 TOG domain is not sufficient for robust tubulin binding (Fig. 4C), consistent with several other TOG domains that require a multidomain architecture to give tubulin binding (Slep and Vale 2007). To gain insight into how individual TOG domains in an SPR2 oligomer might interact with a tubulin dimer, we used the structural alignment of the SPR2 TOG domain on the Stu2 TOG2:αβ-tubulin structure discussed above (Fig. 5C) and examined the relative positioning of HRs and the intra-HEAT loops (in which Stu2 are used to engage tubulin) for complementary contacts with the tubulin heterodimer.

The alignment of SPR2 TOG with the Stu2 TOG2: αβ-tubulin structure highlighted the unique features of the SPR2 TOG domain that may be used to engage tubulin (Fig. 5, E to I). The SPR2 TOG HRs A to D complement the surface of β-tubulin but are distinct from canonical TOG domains; the SPR2 HR A intra-HEAT loop is uniquely angled toward the β-tubulin H11′ to H12 segment (Fig. 5H). This explains why the SPR2 HR A intra-HEAT loop lacks a conserved tryptophan, found in canonical TOG domains, that engages determinants on β-tubulin H3. Instead, the SPR2 HR A loops position a conserved arginine (R56) within the hydrogen bonding distance of determinants on β-tubulin H11′. The SPR2 HR E is recessed into the TOG structure and is not predicted to contact β-tubulin or α-tubulin. However, the SPR2 HR F structurally mimics the Stu2 TOG2 HR E and is similarly positioned to contact α-tubulin H11′ (distance from H11 as modeled: 2 Å), although the SPR2 HR F intra-HEAT loop is positioned more toward the minus end of the tubulin subunit than the Stu2 TOG2 HR E (Fig. 5I). The additional SPR2 HR, HR G, is uniquely positioned to engage α-tubulin H12 (Fig. 5I). Collectively, the SPR2 TOG domain features distinct architectural determinants in HR A and HRs F and G that likely engage the H11′ to H12 segments of β-tubulin and α-tubulin, respectively. The tubulin H11′ to H12 segment is known to undergo a conformational shift in response to GTP hydrolysis in the MT lattice (Alushin et al. 2014; Zhang et al. 2015; LaFrance et al. 2022), suggesting that the SPR2 TOG domain may structurally sense the lattice nucleotide state, which may underlie the ability of SPR2 to preferentially bind to the MT lattice that is in the GMPCPP-bound state.

The 1 to 570 and 38 to 570 aa fragments of SPR2 rescue helical growth of the spr2-2 mutant

To determine whether the SPR2 fragments that bound and stabilized MT minus ends in vitro are functional in vivo, we used genetic complementation of the spr2-2 knockout mutant, which has a strong right-handed helical growth phenotype (Shoji et al. 2004).

The coding sequences of full-length SPR2, 1 to 500, 1 to 570, and 38 to 570 aa fragments fused to mRuby were expressed under the control of the SPR2 promoter in the spr2-2 mutant containing a GFP-TUB6 MT marker. Consistent with previous work (Fan et al. 2018), the expression of full-length SPR2-mRuby in the spr2-2 mutant rescued helical growth of cotyledons, true leaves, and petals (Fig. 6, A to D). For the 1 to 500 aa fragment, we analyzed 6 independent single-insertion homozygous lines. All these plants exhibited right-handed helical growth of cotyledons, true leaves, and petals, similar to the spr2-2 mutant (Fig. 6, A to D). In addition, the spr2-2 plants expressing 1 to 500 aa had upside-down cotyledons and leaves similar to the spr2-2 mutant (Fig. 6E). For the 1 to 570 aa fragment, we analyzed 20 independent single-insertion homozygous lines. Of these, 8 lines appeared like wild-type plants (e.g. Line #2-4), indicating full complementation (Fig. 6, A to E). Eleven (1 to 570)-mRuby expressing lines showed partial complementation (e.g. Line #23-3), such that cotyledons showed right-handed helical growth (Fig. 6, A and D), but true leaves and petals appeared like wild type (Fig. 6, B, C, and E). The remaining (1 to 570)-mRuby line lost the GFP-TUB6 marker.

Figure 6.

Figure 6.

Open in a new tab

Phenotypes of spr2-2 mutant expressing SPR2 fragments. Images of Col-0, spr2-2 mutant, and spr2-2 mutant complemented with SPR2-mRuby, (1 to 500)-mRuby, (1 to 570)-mRuby, or (38 to 570)-mRuby fusion proteins. A) Seedlings 10 days after germination. B) Twenty-day-old plants grown in the soil. C) Flowers from adult plants. Open arrowheads in A) point to leaves facing upside down. Closed arrowheads in A) point to twisted cotyledons in the partially complemented (1 to 570)-mRuby #23-3 line. Scale bar = 0.5 cm for seedlings and plants; scale bar = 1 mm for flowers. D) Angle between cotyledon midrib and petiole. An angle of close to 180° indicates straight cotyledons, while smaller angles indicate twisted cotyledons. Letters indicate statistically distinguishable groups determined by Kruskal–Wallis test with Dunn's multiple comparison test (P < 0.05). E) Percentage of upside-down cotyledons and leaves. Letters indicate statistically distinguishable groups within cotyledon and leaf, respectively, determined by Fisher's exact test (P < 0.05).

For the 38 to 570 aa fragment, we analyzed 5 independent single-insertion homozygous lines. Three of these lines showed full complementation (e.g. Lines #5-1 and #42-1), with cotyledons, true leaves, and petals showing wild-type–like morphology (Fig. 6, A to E). The remaining 2 lines did not show any mRuby signal, probably because of gene silencing, and therefore were not analyzed further.

The 1 to 570 and 38 to 570 aa fragments of SPR2 stabilize MT minus ends in vivo

We next performed live imaging of hypocotyl epidermal cells to determine whether the complementation results of the different SPR2 fragments can be explained by their localization pattern and effect on cortical MT dynamics.

Full-length SPR2-mRuby localized continuously at free cortical MT minus ends (Fan et al. 2018; Fig. 7, A and B), with an average dwell time of 106 s (n = 130). In striking contrast, the (1 to 500)-mRuby signal appeared only sporadically at cortical MT minus ends (Fig. 7C), with an average dwell time of 7 s (n = 102). The cortical MT minus ends in these plants depolymerized rapidly (Fig. 7D), like in the spr2-2 mutant (Fan et al. 2018; Nakamura et al. 2018).

Figure 7.

Figure 7.

Open in a new tab

Cortical MT dynamics and minus-end localization of SPR2 fragments in hypocotyl epidermal cells. Montage and kymograph showing the localization and cortical MT dynamics of full-length SPR2-mRuby A and B), (1 to 500)-mRuby C and D), (1 to 570)-mRuby Line #2-4 that shows full complementation E and F), (1 to 570)-mRuby Line #23-3 that shows partial complementation G and H), and (38 to 570)-mRuby line I and J). Numbers above each montage indicate time in seconds. The arrows and dots label the MT plus and minus ends, respectively. A purple dot indicates when the mRuby signal is present at the minus end. A white dot indicates when the mRuby signal is absent at the minus end. Arrowheads in J) point to (38 to 570)-mRuby signal at growing plus ends. Scale bar = 2 µm.

In the 1 to 570 #2-4 line that showed full complementation, we observed continuous (1 to 570)-mRuby signal at minus ends with an average dwell time of 147 s (n = 71; Fig. 7E). In addition, the cortical MT minus ends in this plant depolymerized slowly like in wild-type plants (Fig. 7F). By contrast, in the 1 to 570 #23-3 line in which cotyledons showed right-handed twisting, the (1 to 570)-mRuby signal at cortical MT minus ends was sporadic (Fig. 7G), with an average dwell time of 65 s (n = 81). The cortical MT minus ends in this plant depolymerized rapidly (Fig. 7H), like in the spr2-2 mutant. Last, in the 38 to 570 complementation lines, the (38 to 570)-mRuby signal was present continuously at cortical MT minus ends (Fig. 7I), with an average dwell time of 105 s (n = 110), and the minus ends of these cortical MTs showed wild-type–like dynamics (Fig. 7J).

We used these data to quantify the growth and shortening rates as well as the time spent growing, shortening, and pausing for both the plus and minus ends of cortical MTs. The 1 to 500, 1 to 570, and 38 to 570 aa fragments partially rescued the growth rate of the plus ends, except for the 1 to 570 aa Line #23-3 in which the plus-end growth rate was statistically indistinguishable from control plants (Table 1). The MT plus-end shortening rate was statistically the same between all the lines (Table 1). For the time spent by the plus ends in the various phases, the 1 to 500 aa lines failed to restore these parameters to control values; in fact, the deviation from control values was even more severe in these plants compared with the spr2-2 mutant. By contrast, the 1 to 570 and 38 to 570 aa lines showed at least partial rescue of these parameters, with 38 to 570 aa Line #5-1 being the most like control plants (Table 1).

Table 1.

Dynamics of cortical MT plus and minus ends

Growth rate (μm/min) Shortening rate (μm/min) Growth time (%) Shortening time (%) Pause time (%)
Plus end SPR2 4.8 ± 1.8c (148) 20.1 ± 22.5a (91) 88.7 ± 10.0b (75) 11.3 ± 10.0a (75) 0 (75)
spr2-2 8.1 ± 5.1a (117) 20.6 ± 12.4a (11) 98.1 ± 6.0a (53) 1.9 ± 6.0bc (53) 0 (53)
1 to 500 #2-2 6.2 ± 2.3ab (187) 11.4 ± 5.3a (2) 99.9 ± 1.2a (61) 0.2 ± 1.2c (61) 0 (61)
1 to 570 #2-4 5.7 ± 1.7b (86) 22.0 ± 13.6a (27) 93.0 ± 11.9b (51) 7.0 ± 11.9b (51) 0 (51)
1 to 570 #23-3 4.6 ± 1.5c (121) 19.7 ± 15.9a (32) 95.9 ± 7.4ab (45) 4.1 ± 7.4bc (45) 0 (45)
38 to 570 #5-1 6.3 ± 3.4b (161) 20.5 ± 26.2a (77) 89.3 ± 9.8b (65) 10.7 ± 9.8a (65) 0 (65)
38 to 570 #42-1 6.0 ± 1.7b (75) 25.8 ± 16.4a (29) 94.3 ± 6.8b (42) 5.7 ± 6.8b (42) 0 (42)
Minus end SPR2 NA 1.5 ± 2.7cd (288) NA 61.0 ± 38.0b (74) 39.0 ± 38.0a (74)
spr2-2 NA 7.8 ± 6.2a (884) NA 89.3 ± 20.0ab (52) 10.7 ± 20.0bc (52)
1 to 500 #2-2 NA 8.6 ± 6.7a (704) NA 98.0 ± 7.3a (61) 2.0 ± 7.2c (61)
1 to 570 #2-4 NA 2.8 ± 6.3c (248) NA 77.8 ± 33.3b (59) 22.2 ± 33.2b (59)
1 to 570 #23-3 NA 4.8 ± 6.3b (371) NA 80.4 ± 27.5b (51) 19.7 ± 27.5b (51)
38 to 570 #5-1 NA 1.2 ± 2.4d (235) NA 79.4 ± 28.4b (68) 20.6 ± 28.4b (68)
38 to 570 #42-1 NA 1.8 ± 2.2c (183) NA 84.2 ± 22.4b (38) 15.8 ± 22.4b (38)
Open in a new tab

Values are means ± Sd (n, number of events). NA, not applicable. The letters (a, b, c) indicate statistically distinguishable groups as determined by the Kruskal–Wallis test with Dunn's multiple comparison test, P < 0.05.

With respect to the cortical MT minus ends, both the shortening rate and time spent shortening were more severely altered in the 1 to 500 aa lines compared with the spr2-2 mutant (Table 1). However, in the 1 to 570 and 38 to 570 aa lines, these parameters were at least partially restored to control values (Table 1).

Discussion

The acentrosomal cortical MT cytoskeleton of higher plants consists of treadmilling polymers with abundant free minus ends (Shaw et al. 2003). The SPR2 protein has emerged as a plant-specific MT minus-end targeting protein, which localizes to and tracks free minus ends and greatly reduces their depolymerization rate (Fan et al. 2018; Leong et al. 2018; Nakamura et al. 2018). However, the structural elements of SPR2 that confer minus end localization and protection against rapid depolymerization are unknown. In this study, we show that a unique TOG domain followed by a basic region and unstructured linker, and oligomerization via a central coiled-coil domain, are important for SPR2 to target and stabilize free MT minus ends.

Our SECMALS data indicate that SPR2 exists as a tetramer. In addition, we found that oligomerization is required for SPR2 to cobind soluble tubulin and the MT polymer. However, more work is needed to determine whether an SPR2 tetramer directly binds to soluble tubulin and to elucidate the stoichiometry of such a complex. While the oligomerization of SPR2 gives rise to puncta on both the MT lattice and at the minus end, the puncta on the MT lattice show diffusive mobility (Yao et al. 2008; Wightman et al. 2013; Fan et al. 2018; Nakamura et al. 2018), while those at the minus end are stable and track depolymerizing minus ends (Fan et al. 2018; Nakamura et al. 2018). We hypothesize that a single stably bound tetramer of SPR2 is likely sufficient to protect a minus end from rapid depolymerization because 4 SPR2 molecules are typically found at static minus ends in vitro (Fan et al. 2018). However, whether all the 4 TOG domains in an SPR2 tetramer are involved in binding soluble tubulin and whether the SPR2's ability to bring free tubulin to an MT is required for minus end protection remain to be determined.

Our crystallographic structure revealed that the N-terminal region of SPR2 contains a bona fide TOG domain. We demonstrate that the TOG domain of SPR2 has a distinctive architecture consisting of 7 HRs, making it unique among TOG domain-containing proteins. Alignment of SPR2 TOG to the Stu2 TOG2: αβ-tubulin structure indicates that the unique features in HR A and HRs F and G of the SPR2 TOG domain likely engage the H11′ to H12 segments of β-tubulin and α-tubulin, respectively. Future work will test the importance of specific amino acids in the SPR2 TOG domain for tubulin binding, their contribution to the minus-end stabilizing function of SPR2, and the role of binding the H11′ to H12 segments of β-tubulin and α-tubulin.

We found that a positively charged region following the TOG domain facilitates MT lattice binding by SPR2, probably by binding to the negatively charged MT surface. The oligomerization of SPR2 greatly enhances MT binding, likely through an avidity-based mechanism in which an SPR2 tetramer uses its basic regions to engage multiple sites on the MT lattice. In the MT plus-end polymerase, XMAP215, a lattice-binding basic region is thought to be important for function at physiological concentrations by promoting the accumulation of XMAP215 at the plus end (Widlund et al. 2011). The basic region of SPR2 might play a similar role. SPR2 molecules diffuse along the MT lattice both in vitro (Fan et al. 2018) and in vivo (Yao et al. 2008; Wightman et al. 2013). Upon arrival at a minus end, diffusive SPR2 molecules become immobilized and remain persistently at the minus end (Fan et al. 2018). Thus, the lattice-binding basic region likely enhances SPR2's localization to minus ends, especially at low concentrations. In addition, lattice binding might enable SPR2 to track depolymerizing minus ends.

In our in vitro experiments, the 1 to 500 aa fragment was the smallest version that showed punctate minus-end localization. This fragment contains the TOG domain, basic region, and an ∼100 aa segment that is predicted to be disordered. Future work will examine whether the 100 aa disordered region mediates the recognition of a specific tubulin face or conformation uniquely found at the minus end.

When expressed in the Arabidopsis spr2-2 mutant, the 1 to 500 aa fragment of SPR2 is unable to complement the helical growth phenotype. While this fragment localizes to cortical MT minus ends, it does not prevent their rapid depolymerization. This inability correlates with a dramatically lower dwell time at cortical MT minus ends compared with the full-length SPR2. Hence, the stable association of SPR2 with minus ends is required for its function. The low MT binding affinity of the 1 to 500 aa fragment provides one possible explanation for its sporadic localization to cortical MT minus ends. Alternatively, this fragment might be nonfunctional due to its inability to recruit tubulin subunits to MTs.

The Arabidopsis spr2-2 mutant expressing the 1 to 570 aa fragment of SPR2 showed either complete complementation or partial complementation wherein embryonic organs remained twisted but postembryonic organs are wild-type like. Full complementation was associated with the stable localization of the 1 to 570 aa fragment to cortical MT minus ends and slow minus-end depolymerization. By contrast, partial complementation was associated with sporadic localization at cortical MT minus ends and their rapid depolymerization in hypocotyl epidermal cells. Whether the difference in complementation strength between these plants is due to differences in the expression level of this fragment remains to be determined.

The 38 to 570 aa fragment of SPR2 showed complete complementation when expressed in the Arabidopsis spr2-2 mutant. This fragment best recapitulates the MT binding and minus-end stabilizing activity of full-length SPR2 in vitro. Further, it shows continuous tracking of depolymerizing cortical MT minus ends in vivo, similar to full-length SPR2. We intentionally generated this fragment to determine whether the Ser/Thr-rich region regulates the localization and function of SPR2, potentially via phosphorylation. However, our data demonstrate that the N-terminal 37 aa are dispensable for the normal activity of SPR2.

We found that the C-terminal domain of SPR2 is not required for targeting and stabilizing MT minus ends. In addition, it does not contribute to the occasional localization of SPR2 to growing cortical MT plus ends (Fan et al. 2018), because the 38 to 570 aa fragment was able to do so. Recently, the SPR2 C-terminal domain was found to have structural homology to the C-terminal domain of the p80 katanin regulatory subunit (Ohno et al. 2023; Bolhuis et al. 2022). While this raises the possibility that SPR2 might interact with katanin to regulate MT dynamics, it remains to be experimentally determined.

How might SPR2 reduce the depolymerization rate of the MT minus end? Based on our data, one possible mechanism is that SPR2 brings new tubulin subunits to minus ends, which would slow their depolymerization rate. Alternatively, SPR2 might stabilize the MT lattice conformation at the minus end to inhibit depolymerization. A third possibility is that a tetramer of SPR2 binds to multiple lattice sites, perhaps on different protofilaments, thus preventing the outward curling of protofilaments and consequently limiting minus-end depolymerization. We note that these distinct mechanisms are not mutually exclusive. Regardless of the exact mechanism, our data establish SPR2 as a TOG domain-containing protein involved in regulating the dynamics of MT minus ends.

Materials and methods

Plant growth

A. thaliana (L.) Heynh. Columbia-0 (Col-0) plants were used for all experiments. For growth on plates, seeds were sterilized with 25% (v/v) bleach for 10 min, rinsed 5 times with sterile water, suspended in 0.1% (w/v) sterile agarose solution, and planted on plates containing 2.2 g/L MS salts with Gamborg's vitamins (Caisson Laboratories), 3 g/L sucrose, and pH 5.7. Seeds were stratified in the dark at 4 °C for 2 to 3 d and then grown at 23 °C under 16 h of 120 to 140 µmol light with Philips F96T8/TL841/HO/PLUS 86-W bulbs for 4 d. When transferred to the soil, seedlings were grown under continuous light at 120 to 140 µmol intensity with GE Ecolux with Starcoat F40T8SPX41 40-W 4100K bulbs, 70% humidity, at 22 °C.

Constructs

For live imaging, the pSPR2:SPR2-mRuby construct (Fan et al. 2018) was modified by replacing the full-length SPR2 cDNA with the appropriate truncated SPR2 cDNAs. These constructs were introduced into the spr2-2 mutant expressing a GFP-TUB6 MT marker (Fan et al. 2018) using Agrobacterium-mediated floral dip transformation. Transgenic plants were selected using 100 mg/L gentamicin with 50 mg/L kanamycin, and homozygous lines containing a single copy of the transgene were used for further study.

For protein expression, the pGEX SPR2-GFP construct (Fan et al. 2018) was modified by inserting a PaeI restriction site between SPR2 and GFP, and the PreScission protease site was replaced by a TEV protease site. SPR2 fragments were generated using PCR, restriction digested by BamHI and PaeI enzymes, and used to replace the full-length SPR2 in the modified pGEX SPR2-GFP construct. To generate His-tagged expression constructs, the SPR2 cDNA encoding residues 33 to 333 and 503 to 566 aa were obtained using PCR and individually subcloned into pET28 (MilliporeSigma) using engineered NdeI and EcoRI restriction endonuclease sites to generate thrombin-cleavable His-tagged proteins. The primers used in this study are listed in Supplemental Table S2.

Protein purification

GST-tagged SPR2 proteins were expressed in Escherichia coli Rosetta DE3 cells and purified using affinity chromatography with Glutathione Sepharose beads (GE Healthcare) in a cold room. Purified proteins were treated with His-tagged TEV protease (1.7 mg/mL) in a cleavage buffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol, 1% Triton X-100, and pH 7.0) for 20 h to remove the GST tag, and free SPR2 proteins were separated from the beads by centrifugation. The TEV protease was removed from the digested proteins by incubation with Ni2+-NTA beads for 2.5 h in a cold room. Isolated proteins were desalted using PD-10 columns (GE Healthcare) and exchanged into BRB80 buffer (80 mM piperazine-1,4-bis[2-ethanesulfonic acid], 1 mM MgCl2, 1 mM EGTA, and pH 6.8) supplemented with 50 mM NaCl to prevent protein aggregation. Proteins were either flash-frozen in liquid nitrogen and kept at −80 °C for long-term storage or kept on ice and used within 1 wk.

The His-tagged SPR2 proteins were expressed in E. coli BL21 DE3 pLysS cells and purified using Ni2+-NTA column (Qiagen). The N-terminal His6 tag was cleaved with bovine α-thrombin (Haematologic Technologies). Cleaved proteins were filtered over a benzamidine-Sepharose (GE Healthcare) column to remove thrombin, and a subsequent Ni2+-NTA column was used to remove uncleaved His6-tagged protein. Cleaved protein constructs were buffer exchanged into storage buffer (25 mM Tris pH 8.5, 500 mM NaCl, and 0.1% β-mercaptoethanol), concentrated using 3 kDa Amicon Ultra Spin Concentrators (MilliporeSigma), flash-frozen in liquid nitrogen, and stored at −80 °C.

MT cosedimentation

MTs were prepared using unlabeled porcine brain tubulin (Cytoskeleton Inc.) reconstituted in ice-cold BRB80 buffer, polymerized in the presence of 1 mM Mg-GTP for 1 h at 37 °C, and assembled MTs stabilized with 20 µM paclitaxel (Cytoskeleton Inc.). Different concentrations of taxol-stabilized MTs were coincubated with either 2 μM SPR2(1 to 500)-GFP or SPR2(1 to 570)-GFP, 0.1 mg/mL BSA, and 20 µM paclitaxel for 20 min. MTs and bound proteins were sedimented by centrifugation at 100,000 × g at 25 °C for 20 min. Supernatant and pellet fractions were resuspended in equal volumes of loading buffer and analyzed using SDS–PAGE. The amount of SPR2 protein in the supernatant and pellet fractions was quantified using densitometry.

In vitro MT dynamics reconstitution experiments

In vitro assays with dynamic MTs were conducted as described in Fan et al. (2018). Briefly, 300 nM GMPCPP-stabilized MT seeds containing 1:12 biotin-labeled and 1:10 rhodamine-labeled porcine tubulins (Cytoskeleton Inc.) bound to coverslips were used to initiate MT polymerization by flowing into 20 mM 1:25 rhodamine-labeled porcine tubulin in BRB80 buffer containing 1% methyl cellulose (4,000 cP, MilliporeSigma), 50 mM DTT, 2 mM GTP, an oxygen-scavenging system, and 500 nM of either full-length or truncated SPR2-GFP proteins. Images were collected by a 100× (NA 1.45) objective and a back-illuminated electron-multiplying CCD camera (ImageEM; Hamamatsu) at 2 s intervals.

Live imaging of Arabidopsis seedlings

Variable-angle epifluorescence microscopy images were collected from hypocotyl epidermal cells of 4-d-old, light-grown Arabidopsis seedlings as described in Fan et al. (2018).

Soluble tubulin pulldown

Anti-GFP beads (MBL Life Science) were equilibrated in binding buffer (BRB80 buffer supplemented with 50 mM NaCl and 0.05% Tween-20) and incubated with 1 µM full-length SPR2-GFP protein on a rotary shaker for 2.5 h at room temperature. Beads were then blocked with 3 µM BSA for 1 h on a shaker in a cold room, and 2 µM unlabeled porcine tubulin (Cytoskeleton Inc.) prepared in binding buffer containing 10 mM DTT and 0.2 mM PMSF was added and incubated overnight. Beads were centrifuged at 1,000 × g for 1 min, washed 3 times with binding buffer, transferred to a fresh tube, and washed again for 5 times with binding buffer. Bound proteins were then analyzed with SDS–PAGE. As controls, 1 µM of full-length SPR2-GFP and 2 µM unlabeled porcine tubulin were incubated separately with the anti-GFP beads.

Imaging soluble tubulin binding to MT-bound SPR2-GFP proteins

The GMPCPP-stabilized MTs were assembled using 50 µM unlabeled porcine tubulin containing 1:12 biotin-labeled porcine tubulin (Cytoskeleton Inc.) and polymerized in the presence of 1 mM GMPCPP (Jena Bioscience) at 37 °C for 30 min. Approximately 300 nM GMPCPP-stabilized MTs were introduced into a flow chamber and allowed to bind to a 20% antibiotin antibody (clone BN-34, MilliporeSigma) for 10 min. Then, 100 nM of the specified SPR2-GFP proteins along with 2.5 µM 1:9 rhodamine-labeled porcine tubulin were introduced into the flow chamber in BRB80 buffer containing 1% methyl cellulose (4,000 cP, MilliporeSigma), 50 mM DTT, and an oxygen-scavenging system (Fan et al. 2018). GFP and rhodamine were excited using 5 mW 488 and 561 nm diode-pumped solid-state lasers, and the images were collected by a 100× (NA 1.45) objective and a back-illuminated electron-multiplying CCD camera (Image EM; Hamamatsu) at 2 s intervals. The extent of colocalization between SPR2-GFP and soluble rhodamine-tubulin was analyzed using the JACoP plugin (Bolte and Cordelieres 2006).

Size exclusion chromatography multiangle light scattering

Full-length and truncated SPR2 constructs were individually injected onto a Superdex 200 10/300 GL size exclusion column (GE Healthcare) preequilibrated and passed through a storage buffer supplemented with 0.2 g/L sodium azide. Samples were subsequently passed consecutively through a Wyatt DAWN HELEOS II light scattering instrument and a Wyatt Optilab rEX refractometer. The light scattering and refractive index values were used to calculate the weight-averaged molar mass (MW) and the mass fraction in each peak using the Wyatt Astra V software program (Wyatt Technology Corp.).

Crystallization, data collection, and structure determination

The SPR2 N-terminal domain construct (residues 33 to 333) was crystallized using the hanging drop method: 2 μL of 15.7 mg/mL protein plus 2 μL of a 1 mL well solution containing 27.5% (w/v) PEG 3350, 150 mM ammonium phosphate, and 280 mM sodium iodide at 18 °C. An SPR2 crystal was cryo-frozen in 27.5% (w/v) PEG 3350, 150 mM ammonium phosphate, and 30% (v/v) glycerol. Two native diffraction data sets were collected on a single crystal (after a translation of the crystal along the axis of the goniometer) at the Advanced Photon Source 22-ID beamline at 100 K (each collection: 360 frames, 0.5° oscillations, 12,398.420 eV; run 1 start = −119.00°, run 2 start = −28.00°). Data were processed and merged using HKL2000 (Otwinowski and Minor 1997). AlphaFold was used to generate a model of the SPR2 N-terminal domain (residues 33 to 333; Jumper et al. 2021; Varadi et al. 2022), from which a truncated model (residues 43 to 323) was used in a molecular replacement search to phase the structure. The molecular replacement search yielded a refined log-likelihood gain of 212 for one protomer in the asymmetric unit (57% solvent). Initial models were built using AutoBuild (PHENIX) followed by reiterative buildings in Coot (Emsley et al. 2010) and refinement runs using phenix.refine (PHENIX) (Adams et al. 2010). Refinement runs used real space and reciprocal space refinement protocols, including simulated annealing with torsion angle molecular dynamics (temperatures: 5,000 K start, 300 K final, 50 steps), initial secondary structure restraints, individual B-factor refinement, and atomic displacement parameters using a maximum-likelihood target. The final refinement run yielded an Rfree value of 25.9%. The final model includes SPR2 residues 33 to 327, an N-terminal Gly-Ser-His-Met cloning artifact, a single iodine atom, and 7 water molecules. Data collection and refinement statistics are summarized in Supplemental Table S1. Structure images were generated using the PyMOL Molecular Graphics System, version 2.4.0 (Schrödinger, LLC). The structure alignment of SPR2 and Stu2 TOG2 (pdb 4U3J) (Ayaz et al. 2014) used the Dali server (Holm 2020). Sequence similarity was based on the following amino acid grouping criteria: [N,Q,H], [T,S,C], [D,E], [R,K], [F,Y,W], [L,V,I,M], [A,G,S], and [P].

Quantification and statistical analysis

All experiments were repeated at least 3 times using independent plants and protein preparations. The statistical significance of data was calculated using Student's t-test, Kruskal–Wallis test with Dunn's multiple comparison test, or Fisher's exact test, as appropriate. To quantify the dynamics of MTs in vivo and in vitro, kymographs of individual MTs were generated using the Dynamic Kymograph plugin (Zhou et al. 2020) and analyzed using the MT Kymograph Analysis plugin (Zhou et al. 2020). Graphs and curve fitting were done using GraphPad Prism. The statistical data are provided in Supplemental Data Set 1.

Accession numbers

Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the following accession number: SPR2 (At4g27060).

Supplementary Material

koad294_Supplementary_Data
koad294_supplementary_data.zip (2.1MB, zip)

Contributor Information

Yuanwei Fan, Department of Biology and Center for Engineering Mechanobiology, Washington University in St. Louis, St. Louis, MO 63130, USA.

Natasha Bilkey, Department of Biology and Center for Engineering Mechanobiology, Washington University in St. Louis, St. Louis, MO 63130, USA.

Derek L Bolhuis, Program in Molecular and Cellular Biophysics, University of North Carolina, Chapel Hill, NC 27514, USA.

Kevin C Slep, Department of Biology, University of North Carolina, Chapel Hill, NC 27599, USA.

Ram Dixit, Department of Biology and Center for Engineering Mechanobiology, Washington University in St. Louis, St. Louis, MO 63130, USA.

Author contributions

Y.F., K.C.S., and R.D. designed the research. Y.F., N.B., D.L.B., and K.C.S. performed the research and analyzed the data. Y.F., K.C.S., and R.D. wrote the article. All authors read and approved the article.

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure S1. SPR2 tracks MT minus ends.

Supplemental Figure S2. The SPR2 TOG domain structural core aligns well with the AlphaFold prediction model.

Supplemental Figure S3. The SPR2 TOG domain is highly conserved across plants.

Supplemental Figure S4. The SPR2 TOG domain face composed of intra-HEAT loops is highly conserved and basic.

Supplemental Table S1. Crystallographic data processing and refinement statistics.

Supplemental Table S2. Oligonucleotides used in this study.

Supplemental Data Set 1. Statistical analysis.

Funding

This work was supported by the National Institute of General Medical Sciences of the National Institutes of Health under award number R35GM139552 (to R.D.) and University of North Carolina Ralph W. Mosley Fund (to K.C.S.). N.B. was supported by a William H. Danforth fellowship in plant sciences, and D.L.B. was supported by the National Institutes of Health Grant T32GM008570 to the Program in Molecular and Cellular Biophysics, University of North Carolina at Chapel Hill.

Data availability

The coordinates for the A. thaliana SPIRAL2 TOG domain structure have been deposited in the Protein Data Bank under the following accession code: 8D00.

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

References

  1. Adams PD, Afonine PV, Bunkóczi G, Chen VB, Davis IW, Echols N, Headd JJ, Hung LW, Kapral GJ, Grosse-Kunstleve RW, et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D Biol Crystallogr. 2010:66(2):213–221. 10.1107/S0907444909052925 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Akhmanova A, Steinmetz MO. Control of microtubule organization and dynamics: two ends in the limelight. Nat Rev Mol Cell Biol. 2015:16(12):711–726. 10.1038/nrm4084 [DOI] [PubMed] [Google Scholar]
  3. Akhmanova A, Steinmetz MO. Microtubule minus-end regulation at a glance. J Cell Sci. 2019:132(11):jcs227850. 10.1242/jcs.227850 [DOI] [PubMed] [Google Scholar]
  4. Al-Bassam J, Larsen NA, Hyman AA, Harrison SC. Crystal structure of a TOG domain: conserved features of XMAP215/Dis1-family TOG domains and implications for tubulin binding. Structure. 2007:15(3):355–362. 10.1016/j.str.2007.01.012 [DOI] [PubMed] [Google Scholar]
  5. Al-Bassam J, van Breugel M, Harrison SC, Hyman A. Stu2p binds tubulin and undergoes an open-to-closed conformational change. J Cell Biol. 2006:172(7):1009–1022. 10.1083/jcb.200511010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Alushin GM, Lander GC, Kellogg EH, Zhang R, Baker D, Nogales E. High-resolution microtubule structures reveal the structural transitions in αβ-tubulin upon GTP hydrolysis. Cell. 2014:157(5):1117–1129. 10.1016/j.cell.2014.03.053 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Ambrose JC, Shoji T, Kotzer AM, Pighin JA, Wasteneys GO. The Arabidopsis CLASP gene encodes a microtubule-associated protein involved in cell expansion and division. Plant Cell. 2007:19(9):2763–2775. 10.1105/tpc.107.053777 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Atherton J, Jiang K, Stangier MM, Luo Y, Hua S, Houben K, van Hooff JJE, Joseph AP, Scarabelli G, Grant BJ, et al. A structural model for microtubule minus-end recognition and protection by CAMSAP proteins. Nat Struct Mol Biol. 2017:24(11):931–943. 10.1038/nsmb.3483 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Atherton J, Luo Y, Xiang S, Yang C, Rai A, Jiang K, Stangier M, Vemu A, Cook AD, Wang S, et al. Structural determinants of microtubule minus end preference in CAMSAP CKK domains. Nat Commun. 2019:10(1):5236. 10.1038/s41467-019-13247-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ayaz P, Munyoki S, Geyer EA, Piedra F-A, Vu ES, Bromberg R, Otwinowski Z, Grishin NV, Brautigam CA, Rice LM. A tethered delivery mechanism explains the catalytic action of a microtubule polymerase. Elife. 2014:3:e03069. 10.7554/eLife.03069 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Ayaz P, Ye X, Huddleston P, Brautigam CA, Rice LM. A TOG:αβ-tubulin complex structure reveals conformation-based mechanisms for a microtubule polymerase. Science. 2012:337(6096):857–860. 10.1126/science.1221698 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Bolhuis DL, Dixit R, Slep KC. Crystal structure of the Arabidopsis SPIRAL2 C-terminal domain reveals a p80-Katanin-like domain. bioRxiv 522140. 10.1101/2022.12.28.522140, 29 December 2022, preprint: not peer reviewed. [DOI] [PMC free article] [PubMed]
  13. Bolte S, Cordelieres FP. A guided tour into subcellular colocalization analysis in light microscopy. J Microsc. 2006:224(3):213–232. 10.1111/j.1365-2818.2006.01706.x [DOI] [PubMed] [Google Scholar]
  14. Brouhard GJ, Stear JH, Noetzel TL, Al-Bassam J, Kinoshita K, Harrison SC, Howard J, Hyman AA. XMAP215 is a processive microtubule polymerase. Cell. 2008:132(1):79–88. 10.1016/j.cell.2007.11.043 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Byrnes AE, Slep KC. TOG–tubulin binding specificity promotes microtubule dynamics and mitotic spindle formation. J Cell Biol. 2017:216(6):1641–1657. 10.1083/jcb.201610090 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Currie JD, Stewman S, Schimizzi G, Slep KC, Ma A, Rogers SL. The microtubule lattice and plus-end association of Drosophila Mini spindles is spatially regulated to fine-tune microtubule dynamics. Mol Biol Cell. 2011:22(22):4343–4361. 10.1091/mbc.e11-06-0520 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Emsley P, Lohkamp B, Scott WG, Cowtan K. Features and development of Coot. Acta Crystallogr D Biol Crystallogr. 2010:66(4):486–501. 10.1107/S0907444910007493 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Fan Y, Burkart GM, Dixit R. The Arabidopsis SPIRAL2 protein targets and stabilizes microtubule minus ends. Curr Biol. 2018:28(6):987–994.e3. 10.1016/j.cub.2018.02.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Filipčík P, Latham SL, Cadell AL, Day CL, Croucher DR, Mace PD. A cryptic tubulin-binding domain links MEKK1 to curved tubulin protomers. Proc Natl Acad Sci U S A. 2020:117(35):21308–21318. 10.1073/pnas.2006429117 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Fox JC, Howard AE, Currie JD, Rogers SL, Slep KC. The XMAP215 family drives microtubule polymerization using a structurally diverse TOG array. Mol Biol Cell. 2014:25(16):2375–2392. 10.1091/mbc.e13-08-0501 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Galjart N. Plus-end-tracking proteins and their interactions at microtubule ends. Curr Biol. 2010:20(12):R528–R537. 10.1016/j.cub.2010.05.022 [DOI] [PubMed] [Google Scholar]
  22. Gard DL, Becker BE, Josh Romney S. MAPping the eukaryotic tree of life: structure, function, and evolution of the MAP215/Dis1 family of microtubule-associated proteins. Int Rev Cytol. 2004:239:179–272. 10.1016/S0074-7696(04)39004-2 [DOI] [PubMed] [Google Scholar]
  23. Haikonen T, Rajamäki M-L, Valkonen JPT. Interaction of the microtubule-associated host protein HIP2 with viral helper component proteinase is important in infection with potato virus A. Mol Plant Microbe Interact. 2013:26(7):734–744. 10.1094/MPMI-01-13-0023-R [DOI] [PubMed] [Google Scholar]
  24. Hendershott MC, Vale RD. Regulation of microtubule minus-end dynamics by CAMSAPs and Patronin. Proc Natl Acad Sci U S A. 2014:111(16):5860–5865. 10.1073/pnas.1404133111 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Holm L. Using Dali for protein structure comparison. Methods Mol Biol. 2020:2112:29–42. 10.1007/978-1-0716-0270-6_3 [DOI] [PubMed] [Google Scholar]
  26. Jiang K, Hua S, Mohan R, Grigoriev I, Yau KW, Liu Q, Katrukha EA, Altelaar AFM, Heck AJR, Hoogenraad CC, et al. Microtubule minus-end stabilization by polymerization-driven CAMSAP deposition. Dev Cell. 2014:28(3):295–309. 10.1016/j.devcel.2014.01.001 [DOI] [PubMed] [Google Scholar]
  27. Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, Tunyasuvunakool K, Bates R, Žídek A, Potapenko A, et al. Highly accurate protein structure prediction with AlphaFold. Nature. 2021:596(7873):583–589. 10.1038/s41586-021-03819-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kawamura E, Wasteneys GO. MOR1, the Arabidopsis thaliana homologue of Xenopus MAP215, promotes rapid growth and shrinkage, and suppresses the pausing of microtubules in vivo. J Cell Sci. 2008:121(24):4114–4123. 10.1242/jcs.039065 [DOI] [PubMed] [Google Scholar]
  29. Kong Z, Ioki M, Braybrook S, Li S, Ye Z-H, Julie Lee Y-R, Hotta T, Chang A, Tian J, Wang G, et al. Kinesin-4 functions in vesicular transport on cortical microtubules and regulates cell wall mechanics during cell elongation in plants. Mol Plant. 2015:8(7):1011–1023. 10.1016/j.molp.2015.01.004 [DOI] [PubMed] [Google Scholar]
  30. LaFrance BJ, Roostalu J, Henkin G, Greber BJ, Zhang R, Normanno D, McCollum CO, Surrey T, Nogales E. Structural transitions in the GTP cap visualized by cryo-electron microscopy of catalytically inactive microtubules. Proc Natl Acad Sci U S A. 2022:119(2):e2114994119. 10.1073/pnas.2114994119 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Leano JB, Slep KC. Structures of TOG1 and TOG2 from the human microtubule dynamics regulator CLASP1. PLoS One. 2019:14(7):e0219823. 10.1371/journal.pone.0219823 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Lechner B, Rashbrooke MC, Collings DA, Eng RC, Kawamura E, Whittington AT, Wasteneys GO. The N-terminal TOG domain of Arabidopsis MOR1 modulates affinity for microtubule polymers. J Cell Sci. 2012:125(Pt 20):4812–4821. 10.1242/jcs.107045 [DOI] [PubMed] [Google Scholar]
  33. Leong SY, Yamada M, Yanagisawa N, Goshima G. SPIRAL2 stabilises endoplasmic microtubule minus ends in the moss Physcomitrella patens. Cell Struct Funct. 2018:43(1):53–60. 10.1247/csf.18001 [DOI] [PubMed] [Google Scholar]
  34. Maki T, Grimaldi AD, Fuchigami S, Kaverina I, Hayashi I. CLASP2 has two distinct TOG domains that contribute differently to microtubule dynamics. J Mol Biol. 2015:427(14):2379–2395. 10.1016/j.jmb.2015.05.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Mimori-Kiyosue Y, Grigoriev I, Lansbergen G, Sasaki H, Matsui C, Severin F, Galjart N, Grosveld F, Vorobjev I, Tsukita S, et al. CLASP1 and CLASP2 bind to EB1 and regulate microtubule plus-end dynamics at the cell cortex. J Cell Biol. 2005:168(1):141–153. 10.1083/jcb.200405094 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Nakamura M, Lindeboom JJ, Saltini M, Mulder BM, Ehrhardt DW. SPR2 protects minus ends to promote severing and reorientation of plant cortical microtubule arrays. J Cell Biol. 2018:217(3):915–927. 10.1083/jcb.201708130 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Nogales E, Zhang R. Visualizing microtubule structural transitions and interactions with associated proteins. Curr Opin Struct Biol. 2016:37:90–96. 10.1016/j.sbi.2015.12.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Oda Y. Cortical microtubule rearrangements and cell wall patterning. Front Plant Sci. 2015:6:236. 10.3389/fpls.2015.00236 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Ohno M, Higuchi Y, Hayashi I. Crystal structure of the C-terminal domain of the plant-specific microtubule-associated protein Spiral2. Acta Crystallogr F Struct Biol Commun. 2023:79(1):17–22. 10.1107/S2053230X22011815 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Otwinowski Z, Minor W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997:276:307–326. 10.1016/S0076-6879(97)76066-X [DOI] [PubMed] [Google Scholar]
  41. Paredez AR, Somerville CR, Ehrhardt DW. Visualization of cellulose synthase demonstrates functional association with microtubules. Science. 2006:312(5779):1491–1495. 10.1126/science.1126551 [DOI] [PubMed] [Google Scholar]
  42. Rezabkova L, Kraatz SHW, Akhmanova A, Steinmetz MO, Kammerer RA. Biophysical and structural characterization of the centriolar protein Cep104 interaction network. J Biol Chem. 2016:291(35):18496–18504. 10.1074/jbc.M116.739771 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Shaw SL, Kamyar R, Ehrhardt DW. Sustained microtubule treadmilling in Arabidopsis cortical arrays. Science. 2003:300(5626):1715–1718. 10.1126/science.1083529 [DOI] [PubMed] [Google Scholar]
  44. Shoji T, Narita NN, Hayashi K, Hayashi K, Asada J, Hamada T, Sonobe S, Nakajima K, Hashimoto T. Plant-specific microtubule-associated protein SPIRAL2 is required for anisotropic growth in Arabidopsis. Plant Physiol. 2004:136(4):3933–3944. 10.1104/pp.104.051748 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Slep KC, Vale RD. Structural basis of microtubule plus end tracking by XMAP215, CLIP-170, and EB1. Mol Cell. 2007:27(6):976–991. 10.1016/j.molcel.2007.07.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Varadi M, Anyango S, Deshpande M, Nair S, Natassia C, Yordanova G, Yuan D, Stroe O, Wood G, Laydon A, et al. AlphaFold protein structure database: massively expanding the structural coverage of protein-sequence space with high-accuracy models. Nucleic Acids Res. 2022:50(D1):D439–D444. 10.1093/nar/gkab1061 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Widlund PO, Stear JH, Pozniakovsky A, Zanic M, Reber S, Brouhard GJ, Hyman AA, Howard J. XMAP215 polymerase activity is built by combining multiple tubulin-binding TOG domains and a basic lattice-binding region. Proc Natl Acad Sci U S A. 2011:108(7):2741–2746. 10.1073/pnas.1016498108 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Wightman R, Chomicki G, Kumar M, Carr P, Turner SR. SPIRAL2 determines plant microtubule organization by modulating microtubule severing. Curr Biol. 2013:23(19):1902–1907. 10.1016/j.cub.2013.07.061 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Yao M, Wakamatsu Y, Itoh TJ, Shoji T, Hashimoto T. Arabidopsis SPIRAL2 promotes uninterrupted microtubule growth by suppressing the pause state of microtubule dynamics. J Cell Sci. 2008:121(14):2372–2381. 10.1242/jcs.030221 [DOI] [PubMed] [Google Scholar]
  50. Zhang R, Alushin GM, Brown A, Nogales E. Mechanistic origin of microtubule dynamic instability and its modulation by EB proteins. Cell. 2015:162(4):849–859. 10.1016/j.cell.2015.07.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Zhou R, Liu H, Ju T, Dixit R. Quantifying the polymerization dynamics of plant cortical microtubules using kymograph analysis. Methods Cell Biol. 2020:160:281–293. 10.1016/bs.mcb.2020.04.006 [DOI] [PubMed] [Google Scholar]
  52. Zhu C, Ganguly A, Baskin TI, McClosky DD, Anderson CT, Foster C, Meunier KA, Okamoto R, Berg H, Dixit R. The fragile Fiber1 kinesin contributes to cortical microtubule-mediated trafficking of cell wall components. Plant Physiol. 2015:167(3):780–792. 10.1104/pp.114.251462 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

koad294_Supplementary_Data
koad294_supplementary_data.zip (2.1MB, zip)

Data Availability Statement

The coordinates for the A. thaliana SPIRAL2 TOG domain structure have been deposited in the Protein Data Bank under the following accession code: 8D00.

Articles from The Plant Cell are provided here courtesy of Oxford University Press

RESOURCES