Salt Stress–Induced Disassembly of Arabidopsis Cortical Microtubule Arrays Involves 26S Proteasome–Dependent Degradation of SPIRAL1^,

This study examines the relationship between salt stress tolerance and dynamic instability of microtubules, revealing an important role for proteasome-dependent regulation of SPR1 in the survival of plants challenged by high salinity.

Abstract

The dynamic instability of cortical microtubules (MTs) (i.e., their ability to rapidly alternate between phases of growth and shrinkage) plays an essential role in plant growth and development. In addition, recent studies have revealed a pivotal role for dynamic instability in the response to salt stress conditions. The salt stress response includes a rapid depolymerization of MTs followed by the formation of a new MT network that is believed to be better suited for surviving high salinity. Although this initial depolymerization response is essential for the adaptation to salt stress, the underlying molecular mechanism has remained largely unknown. Here, we show that the MT-associated protein SPIRAL1 (SPR1) plays a key role in salt stress–induced MT disassembly. SPR1, a microtubule stabilizing protein, is degraded by the 26S proteasome, and its degradation rate is accelerated in response to high salinity. We show that accelerated SPR1 degradation is required for a fast MT disassembly response to salt stress and for salt stress tolerance.

INTRODUCTION

The ubiquitin/26S proteasome system (UPS) regulates many fundamental cellular processes by controlling the degradation rates of numerous proteins (Hershko and Ciechanover, 1998; Vierstra, 2009). For the majority of UPS substrates, the concerted action of E1, E2, and E3 enzymes leads to the covalent attachment of a multiubiquitin chain to the protein destined for degradation. The polyubiquitinated target protein is then degraded by the 26S proteasome (Glickman, 2000), an evolutionarily conserved multicatalytic protease that contains an enclosed proteolytically active core particle and one or two regulatory particles (RPs). The main roles of the RPs are in substrate recognition, which is performed by RP non-ATPase subunits (RPNs), and in the unfolding and translocation of substrates to the core particle by RP triple A ATPase subunits (RPTs) (Smalle and Vierstra, 2004; Kurepa and Smalle, 2008).

In Arabidopsis thaliana, like in other eukaryotes, proteasome mutants and proteasome activity inhibitors are used to uncover the identity of UPS-regulated pathways and UPS target poteins (Kurepa and Smalle, 2008). Loss of function of RP subunits RPN1a and RPN10, for example, revealed that the 26S proteasome is essential for cell division and expansion, the modulation of responses to hormones and proteostatic drugs, and gametophyte development (Smalle et al., 2003; Kurepa et al., 2008, 2009a, 2009b, 2010; Wang et al., 2009). These studies also revealed that the stress responses of rpn1a and rpn10 mutant plants are altered: compared with the wild type, proteasome mutants are more tolerant of oxidative stress and less tolerant of protein misfolding stresses such as heat shock and salt stress (Smalle et al., 2003; Kurepa et al., 2008; Wang et al., 2009).

The most frequently used proteasome inhibitor is MG132, a reversible, cell-permeable peptidyl aldehyde that inhibits the proteasome-specific chymotrypsin-like protease β5 (Lee and Goldberg, 1998). Studies using MG132 have shown that the UPS is involved in the regulation of plant cell microtubule (MT) networks (Yanagawa et al., 2002; Oka et al., 2004; Sheng et al., 2006; S. Wang et al., 2011). MTs are polymers of α/β-tubulin heterodimers, which are incorporated into MTs directionally so that α-tubulin is exposed at the so-called lagging (−) and β-tubulin at the leading (+) ends. MTs play important roles in numerous cellular processes, including cell division and directional cell expansion (Nogales, 2001; Sedbrook and Kaloriti, 2008).

The polymerization-depolymerization dynamics of MTs are essential for their functionality in cellular processes, and they include two types of GTP hydrolysis-dependent dynamic behaviors: dynamic instability and treadmilling. Treadmilling is the net growth of MTs at their (+) ends and shortening at their (−) ends, and dynamic instability is the switching between episodes of growth (polymer assembly) and shortening (polymer disassembly) at the (+) ends. The transition from growing to shortening is a dynamic instability parameter called catastrophe, and transition from shortening to growing is known as rescue. Growth, shrinkage, catastrophe, and rescue rates depend on MT-associated proteins (MAPs) and, in particular, a subclass of MAPs called (+)-end-tracking proteins (+TIPs) (Hirokawa, 1994; Mandelkow and Mandelkow, 1995; Lloyd and Hussey, 2001; Schuyler and Pellman, 2001; Sedbrook, 2004; Hamada, 2007; Sedbrook and Kaloriti, 2008; Lyle et al., 2009a, 2009b).

A large number of +TIPs have been identified, and some of them are found in widely diverged species, suggesting that they regulate a basic, evolutionarily conserved component of the dynamic instability process (Bisgrove et al., 2004). For example, Arabidopsis has functional homologs of end binding protein 1 (EB1) and cytoplasmic linker-associated protein 1 (CLASP1), which were first identified in human cells (Chan et al., 2003; Mathur et al., 2003; Galjart, 2005; Vaughan, 2005; Ambrose et al., 2007; Kirik et al., 2007; Bisgrove et al., 2008; Komaki et al., 2010). Other +TIPs, such as SPIRAL1 (SPR1), are found only in plants (Nakajima et al., 2004, 2006; Sedbrook et al., 2004). Because plants overexpressing SPR1 have increased tolerance to MT-destabilizing drugs, and spr1 mutants have morphological defects consistent with altered MT stability, it has been concluded that this protein, similar to EB1 and CLASP1, is important for the regulation of dynamic instability (Nakajima et al., 2004, 2006; Sedbrook et al., 2004; Abe and Hashimoto, 2005; Ishida et al., 2007).

Similar to other eukaryotes, the structure of the plant MT cytoskeleton is modulated not only by various developmental cues, but also in response to environmental signals and stress conditions (Smertenko et al., 1997; Himmelspach et al., 1999; Mathur and Chua, 2000; Wang and Nick, 2001; Abdrakhamanova et al., 2003; Van Bruaene et al., 2004; C. Wang et al., 2007, 2011; Hamant et al., 2008). For example, short-term salt stress promotes the reorientation of MTs in maize (Zea mays) roots from transverse to parallel relative to the longitudinal axis and in tobacco (Nicotiana tabacum) BY-2 cells from a structured to a seemingly random organization (Blancaflor and Hasenstein, 1995; Dhonukshe et al., 2003). Long-term salt stress also affects cortical MT organization in Arabidopsis and suppresses the right-handed helical growth of spr1 mutants (Shoji et al., 2006). Furthermore, prolonged exposure to salt stress conditions was shown to trigger a biphasic response starting with a massive MT depolymerization phase followed by the formation of new MT networks that are thought to be better suited for surviving high salinity (Wang et al., 2007). Because treatments with a MT-destabilizing drug improved survival and growth under salt stress conditions and treatments with a MT-stabilizing drug caused salt stress hypersensitivity, the initial MT depolymerization response is believed to be essential for maintaining salt stress tolerance (Wang et al., 2007).

The specific mechanism that mediates the initial, salt stress–induced MT destabilization is currently not well understood. Because all tested proteasome mutants are hypersensitive to salt stress, we hypothesized that the MT reorganization needed for salt stress tolerance might be impaired in these lines due to an altered dynamic instability of MTs. Since the dynamic instability is regulated by the action of MAPs and in particular +TIPs, we tested if one of these proteins is conditionally degraded by the proteasome to allow MT restructuring required for an optimal salt stress response. Here, we show that in Arabidopsis, salt stress accelerates the 26S proteasome–dependent degradation of SPR1 and that this facilitates MT disassembly and promotes salt stress tolerance.

RESULTS

26S Proteasome Mutants Have Increased Tolerance to MT-Destabilizing Drugs

To assay the stability of cortical MTs in 26S proteasome mutants, we first tested the responses of rpn1a-4, rpt2a-2, rpn10-1, and rpn12a-1 mutants to the MT-destabilizing drugs oryzalin, which binds to α-tubulin, and propyzamide, which binds to β-tubulin (Nakamura et al., 2004; Lyons-Abbott et al., 2010). We showed previously that the total 26S proteasome activity in the four tested proteasome mutants is affected to different levels, with rpt2a-2 carrying the weakest and rpn10-1 the strongest defect in 26S proteasome function (Kurepa et al., 2008). We determined the effects of MT-destabilizing drugs by measuring their inhibition of root elongation and their promotion of root tip swelling, two plant growth responses that were shown to be caused by MT disassembly (Baskin et al., 1994).

All proteasome mutants were more tolerant to both propyzamide and oryzalin (Figures 1A and 1B). The strongest mutant, rpn10-1, was the most tolerant, whereas the weakest mutant, rpt2a-2, showed a statistically significant increase in propyzamide and oryzalin tolerance only at a single dose (6 μM for propyzamide, P < 0.05; 100 nM for oryzalin, P < 0.05). In addition, the propyzamide tolerance of the double mutant rpn1a-4 rpn10-1 was further increased compared with the respective single mutants (see Supplemental Figure 1 online).

Figure 1.

26S Proteasome Mutants Have Increased Tolerance to MT-Destabilizing Drugs.

(A) and (B) Effect of propyzamide (A) and oryzalin (B) on primary root elongation in the wild type (Col-0) and proteasome mutants rpn1a-4, rpt2a-2, rpn10-1, and rpn12a-1. Seedlings grown on MS/2 medium for 5 d were transferred to media containing the denoted doses of propyzamide or oryzalin. The increase in root length was measured after 3 d of treatment. The root length of untreated seedlings of each genotype was set at 100%, and mean values ± sd (n ≥ 25) are shown as percentages of the root length of the respective controls. The difference in root length between Col-0 and rpn1a-4, rpn10-1, or rpn12a-1 was statistically significant for all tested doses (n ≥ 25, P < 0.0001; two-way ANOVA followed by Bonferroni multiple comparisons post-test). For clarity, only the statistical significance for Col-0 versus rpt2a-2 is marked in the graphs (*P < 0.05).

(C) Effects of 4 μM propyzamide on primary root morphogenesis. The experimental conditions and timeline were the same as in (A). Bar = 250 μm.

We also analyzed the morphology of propyzamide-treated roots. Earlier reports have shown that seedlings grown on media supplemented with MT-destabilizing drugs contain swollen primary roots (Furutani et al., 2000). After a 3-d-long treatment with 4 μM propyzamide, root swelling was obvious in the Columbia-0 (Col-0) plants, whereas the diameter of primary roots of all proteasome mutants did not increase (Figure 1C; see Supplemental Figure 1C online).

Changes in rpn10-1 MT Dynamics

To analyze the stability of MTs in proteasome mutants further, we compared the dynamics of individual MTs in Col-0 and rpn10-1 backgrounds (Figure 2). Previous studies have shown that green fluorescent protein (GFP) fusions with α- (GFP-TUA6) or β-tubulin (GFP-TUB6) are incorporated into MTs and can be used to visualize MT dynamics in vivo (Ueda et al., 1999; Nakamura et al., 2004; Abe and Hashimoto, 2005). However, whereas GFP-TUB6–labeled MTs have wild-type properties, the incorporation of GFP-TUA6 changes MT dynamics and induces right-handed helical growth (Nakamura et al., 2004; Abe and Hashimoto, 2005). Thus, we chose to introduce the 35S:GFP-TUB6 transgene into the rpn10-1 mutant and analyzed MT dynamics using confocal time-lapse imaging.

Figure 2.

MT Plus-End Dynamics Are Altered in the rpn10-1 Mutant.

(A) Kymograph of a 160-s-long recording showing (+)-end dynamics of individual MTs. Four-day-old seedlings expressing 35S:GFP-TUB6 in the Col-0 and rpn10-1 backgrounds were used for time-lapse confocal imaging. Epidermal cells of upper hypocotyl regions were photographed every 4 s. Bar = 5 μm.

(B) The life history plots (length versus time) of three GFP-TUB6–labeled MTs in Col-0 and three GFP-TUB6–labeled MTs in the rpn10-1 background. MT length was measured from confocal micrographs using Image J.

[See online article for color version of this figure.]

Kymographs of individual MT (+)-ends and representative life history plots showed that in rpn10-1, the MT growth phase was longer and the shrinkage phase was shorter than in the wild type (Figure 2). Indeed, most of the parameters of MT dynamic instability were altered in the rpn10-1 background (Table 1). The average growth rate of the leading ends and the frequency of catastrophe in rpn10-1 were significantly lower than in Col-0, while the frequency of rescue events in rpn10-1 was increased. Thus, individual MTs in rpn10-1 were less dynamic and more prone to polymerization. In contrast with the (+)-end, there was no significant difference in (−)-end dynamics between rpn10-1 and the wild type (Table 1), implying that it is the process of dynamic instability and not treadmilling that was affected by the inhibition of proteasome activity. These results combined with the increased tolerance of proteasome mutants to MT-destabilizing drugs indicated that 26S proteasome–dependent proteolysis plays an important role in the regulation of MT dynamic instability.

Table 1.

Dynamic Instability Parameters in Col-0 and rpn10-1

	(+)-Ends		(−)-Ends
Dynamic Parameters	Col-0	rpn10-1	Col-0	rpn10-1
Growth rate (μm/min)	5.44 ± 4.61	4.04 ± 2.73*	1.91 ± 1.15	1.72 ± 1.03
Shrinkage rate (μm/min)	10.3 ± 8.89	8.91 ± 8.44	3.53 ± 4.16	3.82 ± 4.17
Catastrophe (events/second)	0.041	0.030*	0.212	0.196
Rescue (events/second)	0.096	0.141*	0.047	0.051
Time spent on growth	67.2%	76.6%	9.3%	10.7%
Time spent on pause	9.6%	11.2%	48.4%	48.7%
Time spent on shrinkage	20.4%	12.3%	42.4%	40.6%
Dynamicity (μm/min)	5.91 ± 1.46	4.19 ± 2.41*	1.67 ± 1.92	1.99 ± 1.28

MT dynamic instability parameters were quantified from confocal micrographs. Velocities were calculated from 38 leading and 22 lagging ends for 35S:GFP-TUB6 in Col-0 and 45 leading and 32 lagging ends for 35S:GFP-TUB6 in the rpn10-1 background. Total measurements include 1111 and 1763 velocities (4-s intervals) for GFP-TUB6 in Col-0 and rpn10-1, respectively. Dynamic parameters are expressed as mean ± sd. Statistical significance was calculated using Student’s t test comparing the mutant and wild-type values (*P < 0.05).

SPR1 Mediates the Tolerance of Proteasome Mutants to Propyzamide

Previous studies showed that overexpression of the plant-specific +TIP SPR1 leads to increased propyzamide tolerance (Nakajima et al., 2004). Since proteasome mutants are also more tolerant of MT-destabilizing drugs (Figure 1) and have altered MT dynamics, (Figure 2), we tested whether SPR1 is a 26S proteasome target and whether its stabilization in proteasome mutants can explain the observed phenotypes.

To assay SPR1 stability, we used cycloheximide (CHX) to inhibit de novo protein synthesis. After a 12-h-long CHX treatment, the SPR1 level was reduced to ~30% of the control. This decrease was blocked by the proteasome inhibitor MG132, suggesting that SPR1 is an unstable protein that is targeted for 26S proteasome–dependent proteolysis (Figure 3A). Indeed, the SPR1 level was increased in 26S proteasome mutants rpn1a-4 and rpn10-1 and was even more abundant in the rpn1a-4 rpn10-1 double mutant (Figure 3B). The SPR1 mRNA abundance remained unchanged in the proteasome mutant backgrounds, confirming that the increase in SPR1 protein level was caused by a posttranscriptional mechanism (see Supplemental Figure 2 online). Finally, CHX chase immunoblotting analyses (Yewdell et al., 2011) showed that the SPR1 degradation rate was decreased in proteasome mutant backgrounds (Figures 3C and 3D), confirming a role for the UPS in regulating SPR1 abundance.

Figure 3.

SPR1 Is a 26S Proteasome Target.

(A) Immunoblotting analyses using anti-SPR1 and anti-GS antisera. Seven-day-old seedlings were treated with 100 μM MG132 and/or 200 μM CHX for 16 h and used for the extraction of total protein. The anti-GS sera recognizes both the chloroplastic (45 kD) and cytosolic (40 kD) GS isoforms.

(B) SPR1 levels in 8-d-old Col-0, rpn1a-4, rpn10-1, and rpn1a-4 rpn10-1 seedlings. Anti-RPN1 and anti-RPN10 sera were used to confirm the genotype of proteasome mutants.

(C) Representative CHX-chase immunoblots. The stability of SPR1 was tested on total protein extract of 10-d-old wild-type and mutant seedlings treated with 200 μM CHX for the indicated time periods.

(D) Quantification of SPR1 stability in CHX-treated Col-0, rpn1a-4, and rpn10-1 plants. Immunoblots, representatives of which are shown in (C), were used to quantify signal intensities. The average signal intensity of the zero time point sample for each line was set to 100%, and the mean values ± sd (n = 3) are shown as percentages of the respective control. The asterisks represent the statistical significance of the difference between degradation rates in Col-0 and both proteasome mutants (****P < 0.0001; ANOVA followed by Bonferroni multiple comparisons post-test).

[See online article for color version of this figure.]

To test if SPR1 accumulation is a cause for the increased tolerance of proteasome mutants to MT-destabilizing drugs, we crossed the spr1-3 mutation into the rpn1a-4 and rpn10-1 mutant backgrounds (Figure 4; see Supplemental Figure 3 online). Analyses of the root tip morphology in propyzamide-treated double mutants showed that the reduced root tip swelling in proteasome mutants was suppressed by the spr1-3 mutation (Figure 4A). To quantify the effect of the spr1-3 mutation, we measured the width of the primary root in the elongation zone (Figure 4B). As expected, the propyzamide treatment did not promote any substantial root swelling in the proteasome mutants. The spr1-3 mutation caused a mild increase in root width compared with the wild type (344.5 ± 5 μm and 372.9 ± 7 μm for Col-0 and spr1-3, respectively; n = 25, P < 0.001). By contrast, the root width of both rpn1a-4 spr1-3 and rpn10-1 spr1-3 double mutants was strongly increased compared with the respective single proteasome mutants (n = 25, P < 0.001), although the increases in swelling did not reach the wild-type level (P < 0.01 for rpn1a-4 spr1-3 and P < 0.001 for rpn10-1 spr1-3 compared with Col-0). We concluded that the removal of SPR1 suppressed the propyzamide tolerance of proteasome mutants but did not fully revert this phenotype to the wild-type level, suggesting the existence of other MT-stabilizing proteins that are also degraded by the 26S proteasome.

Figure 4.

spr1-3 Suppresses the Propyzamide Tolerance of Proteasome Mutants.

(A) Effects of 4 μM propyzamide on primary root morphogenesis in Col-0, rpn1a-4, spr1-3, and the double mutant rpn1a-4 spr1-3. Five-day-old seedlings were transferred to fresh MS/2 medium or MS/2 medium with propyzamide. The seedlings were grown on vertically positioned plates for 3 d and then photographed. Bar = 1 mm.

(B) Effects of 4 μM propyzamide on root width. Seedlings were grown and treated as in (A). The width of primary roots at the elongation zone was measured from micrographs using ImageJ. The significance was analyzed using ANOVA followed by Bonferroni multiple comparisons post-test. Crosses represent the significance of the differences between the untreated Col-0 and untreated mutant lines, and asterisks mark the significance between the treated Col-0 and treated mutant lines (**P < 0.01; ††† and ***P < 0.001; and †††† and ****P < 0.0001).

(C) Root lengths of 5-d-old seedlings grown on MS/2 media with 4 μM propyzamide (****P < 0.0001; significance for Col-0 versus mutants). The statistical analyses and data presentation are as in (B).

(D) In vivo analyses of the MT-destabilizing effect of 20 μM propyzamide in the Col-0 wild type and in the rpn10-1 and rpn10-1 spr1-3 mutants. The 35S:GFP-TUA6 transgene was crossed into the rpn10-1 and rpn10-1 spr1-3 mutant backgrounds, and the GFP-TUA6–labeled cortical MTs were analyzed by confocal microscopy. Hypocotyl epidermal cells of 4-d-old seedlings treated for 1 h are shown. Bar = 20 μm.

[See online article for color version of this figure.]

The root elongation assay confirmed that SPR1 contributes to the propyzamide tolerance of proteasome mutants (Figure 4C). The rpn1a-4 spr1-3 and rpn10-1 spr1-3 double mutants displayed an increase in right-handed helical growth (see Supplemental Figure 3 online). For example, the right-handed root skewing of spr1-3 was enhanced in both the rpn1a-4 and rpn10-1 mutant backgrounds. This was the most striking in rpn1a-4 spr1-3 seedlings that had upward growing roots (see Supplemental Figure 3C online). However, low propyzamide doses fully suppressed the enhanced root skewing phenotype in both double mutants (see Supplemental Figure 4 online), allowing us to compare the propyzamide tolerance levels of all lines by the root elongation assay. Similar to the root tip swelling response, the root elongation assay showed that spr1-3 suppresses the propyzamide tolerance of both the rpn1a-4 and rpn10-1 mutants (Figure 4C).

To analyze directly the effects of spr1-3 on MT stability, we attempted to cross the 35S:GFP-TUB6 transgene into the spr1-3 and rpn10-1 spr1-3 mutants. We were unable to isolate spr1-3 or rpn10-1 spr1-3 mutants expressing GFP-TUB6 in spite of screening more than 1000 F2 seedlings, and we hypothesized that the 35S:GFP-TUB6 transgene is located in the proximity of the SPR1 locus. However, analysis of the 35S:GFP-TUB6 locus revealed that the T-DNA is not linked to SPR1 (data not shown). The reason GFP-TUB6 expression and the spr1-3 mutation are incompatible is currently unknown. As an alternative, we introduced the 35S:GFP-TUA6 transgene into the spr1-3 and rpn10-1 spr1-3 mutants. To assay the MT stability in these lines, we treated 4-d-old seedlings with a high dose of propyzamide to ensure a fast and complete MT disruption (20 μM propyzamide for 1 h) and analyzed epidermal cells of the upper hypocotyl regions by confocal microscopy (Figure 4D). In rpn10-1, the GFP-TUA6–labeled cortical MTs remained partially intact, while MTs in the wild-type and rpn10-1 spr1-3 cells were disrupted, and the GFP signal was localized in randomly dispersed aggregates. Collectively, these experiments suggested that the stabilization of SPR1 in proteasome mutants is a cause for both their increased tolerance to MT-destabilizing drugs and altered MT dynamicity.

Salt Stress Promotes Proteolysis of SPR1

Based on earlier reports of the salt stress hypersensivitiy of 26S proteasome mutants (Smalle et al., 2003; Wang et al., 2009) and the fact that salt stress tolerance requires MT disassembly (Wang et al., 2007), we hypothesized that loss of proteasome function increases salt stress sensitivity by stabilizing SPR1, which in turn leads to increased MT stability. To test this, we first determined if salt stress leads to a conditional 26S proteasome–dependent degradation of SPR1, which could facilitate the rearrangement of cortical MTs needed for salt tolerance.

Prolonged treatment (16 h) of 4-d-old Col-0 seedlings with 150 mM NaCl caused a decrease in SPR1 level, and this decrease was blocked by MG132 (Figure 5A). CHX-chase immunoblotting analyses showed that a 16-h-long NaCl treatment led to an ~80% reduction of the SPR1 level in Col-0, whereas the SPR1 level in rpn1a-4 and rpn10-1 mutants was reduced by only ~30 and ~20%, respectively (Figures 5B and 5C). These results indicated that the salt stress–induced decrease in SPR1 level was a result of proteasome-dependent degradation.

Figure 5.

Salt Stress Promotes 26S Proteasome–Dependent Proteolysis of SPR1.

(A) Immunoblotting analyses of SPR1 in Col-0 seedlings treated with NaCl and MG132. Four-day-old seedlings were treated for 16 h with 150 mM NaCl alone or 150 mM NaCl and 100 μM MG132.

(B) CHX-chase immunoblot. Ten-day-old seedlings were transferred to liquid MS/2 media containing 200 μM CHX and 200 mM NaCl. Seedlings were treated for the indicated times.

(C) Quantification of immunoblots represented by those in (B). The signals of the untreated samples were set at 100%. Data are shown as mean values ± sd (n = 3) as percentages of the respective controls. The asterisks represent the statistical significance of the difference between the degradation rates in Col-0 versus rpn10-1 and rpn1a-4 (****P < 0.0001; ANOVA followed by Bonferroni multiple comparisons post-test).

(D) and (E) Immunoblotting analyses of SPR1 in Col-0 seedlings treated with the denoted doses of NaCl and (D) propyzamide or mannitol (E). Four-day-old seedlings were transferred to water containing the test compounds, treated for 3 h, and used for the isolation of total proteins.

[See online article for color version of this figure.]

During treatments with a higher concentration of salt (200 mM), degradation of SPR1 was detectable already after 3 h of incubation (see Supplemental Figures 5A to 5C online). To test if high salt concentrations promote SPR1 degradation via a global increase in 26S proteasome–dependent proteolysis, we monitored total proteasome activity by analyzing the abundance of polyubiquitinated proteins and by measuring proteasomal chemotryptic activity (see Supplemental Figures 5D and 5E online). After 5 h of exposure to 200 mM NaCl, we observed a significant decrease in the SPR1 protein level, no change in the overall levels of ubiquitinated proteins, and a mild but significant decrease in proteasome activity. We concluded that the accelerated proteasome-dependent SPR1 degradation during salt stress is not caused by a general increase in proteasome activity but is more likely the result of a specific destabilization mechanism that is activated by salt stress. Finally, we confirmed that the SPR1 depletion in response to salt is indeed the result of a posttranscriptional mechanism by observing that the SPR1 mRNA level did not change in response to salt stress treatments (see Supplemental Figure 5F online).

Previous studies have shown that cortical MT arrays depolymerize in response to salt stress (Wang et al., 2007). Since the molecular mechanisms that govern this process were unknown, it remained possible that the salt stress–induced degradation of SPR1 reflects not a direct effect on SPR1, but an indirect effect caused by the MT disassembly. For example, the MT disassembly might increase the amount of free SPR1 that could be more susceptible to degradation than the tubulin-bound version. To distinguish between these possibilities, we monitored the SPR1 level in Col-0 plants treated with propyzamide. In Col-0 seedlings treated for 3 h with 10 μM propyzamide, the cortical MTs were depolymerized (see Supplemental Figure 6 online), but the SPR1 level remained the same as in the untreated control (Figure 5D). The propyzamide treatment also did not influence the salt stress–induced degradation of SPR1 (Figure 5D).

In addition to ionic stress, high salinity is also known to cause osmotic stress (Zhu, 2001). To test if osmotic stress influences the stability of SPR1, we treated Col-0 seedlings with high doses of the osmolyte mannitol (Figure 5E). The mannitol treatments did not affect SPR1 stability, suggesting that the signal triggering SPR1 destabilization was specifically ionic stress (Figure 5F). We concluded that salt stress, but not osmotic stress or MT disassembly per se, promotes the 26S proteasome–dependent degradation of SPR1.

SPR1 Stabilization Causes Salt Stress Hypersensitivity

To test if SPR1 stabilization causes the salt hypersensitivity of 26S proteasome mutants (Smalle et al., 2003; Wang et al., 2009), we first analyzed if the increased MT stability is responsible for the changes in salt tolerance (Figure 6A), then determined the salt tolerance levels of rpn1a-4 spr1-3 and rpn10-1 spr1-3 mutants (Figures 6B to 6D), and finally compared the effects of salt stress on the MT arrays of rpn10-1 and rpn10-1 spr1-3 cells (Figure 7).

Figure 6.

spr1-3 Suppresses the Salt Hypersensitivity of 26S Proteasome Mutants.

(A) Salt hypersensitivity of proteasome mutants was suppressed by propyzamide treatments. Five-day-old seedlings grown on MS/2 medium were transferred to fresh MS/2 plates or MS/2 plates with 100 mM NaCl and the indicated doses of propyzamide. Plates were positioned vertically, and the root lengths were measured after 3 d. Data represent relative root length (n ≥ 20) with sd. The average root length of untreated Col-0 seedlings was set at 100%.

(B) Five-day-old seedlings grown on MS/2 medium were transferred to fresh MS/2 medium or MS/2 containing 100 mM NaCl. Test plates were positioned vertically, and plants were grown for 2 weeks. The arrows highlight necrotic spots in the rpn10-1 mutant grown on 100 mM NaCl.

(C) and (D) The salt tolerance levels were quantified by comparing the fresh weight after a 2-week-long treatment (C) and root lengths after a 3-d-long treatment (D). Data are shown as mean values ± sd of three independent experiments (n ≥ 15 for each). The asterisks represent the significances of the differences between the double mutants and the respective single proteasome mutants and were calculated using ANOVA followed by Bonferroni multiple comparisons post-test (****P < 0.0001, ***P < 0.001, and **P < 0.01).

Figure 7.

SPR1 Stabilization in rpn10-1 Delays the Salt-Induced Disassembly of Cortical MT Arrays.

(A) Visualization of GFP-TUA6–labeled cortical MTs in upper hypocotyl epidermal cells from 4-d-old Col-0, rpn10-1, and rpn10-1 spr1-3 seedlings treated with 200 mM NaCl for the indicated times. Bar = 10 μm.

(B) The density of cortical MTs per unit length in upper hypocotyl cells was determined by counting the MTs crossing the longitudinal axis of a cell. For each line and treatment, a minimum of 32 hypocotyl cells from five separate seedlings was photographed and used for measurements. The asterisks represent the significance of the difference between Col-0 and the mutants treated for the same time and were calculated using ANOVA followed by Bonferroni multiple comparisons post-test (****P < 0.0001).

[See online article for color version of this figure.]

To test the relationship between MT stability and salt tolerance, we analyzed the effects of combined salt stress and propyzamide treatments on the root elongation of rpn1a-4 and rpn10-1 mutants (Figure 6A). In the absence of propyzamide, root length of wild-type plants grown on 100 mM NaCl was ~70% of the control, whereas the root length of proteasome mutants was reduced to ~50%. As expected, low doses of propyzamide counteracted the salt-induced inhibition of root elongation in all lines. However, whereas 1 μM propyzamide led to an 8% increase in root length in the wild type (statistically nonsignificant, P > 0.05, n ≥ 15; analysis of variance [ANOVA] with Bonferroni post-test), the increase in root length in the proteasome mutants was ~25% (P < 0.001, n ≥ 15). Thus, since low doses of an MT-destabilizing drug reverted the salt hypersensitivity of proteasome mutants to wild-type levels, we concluded that the increased MT stability in proteasome mutants is indeed responsible for their salt hypersensitivity.

This conclusion, together with our observation that spr1-3 suppressed the increased MT stability in rpn1a-4 and rpn10-1 mutants (Figure 4), prompted us to test if the spr1-3 mutation also suppresses the salt hypersensitivity of proteasome mutants (Figures 6B and 6D). Salt stress induces a range of morphological alternations in both the aerial parts of the plant and the roots. For example, NaCl treatments lead to reductions in root length and lateral root number, a decrease in leaf and petiole size, and a delay in leaf emergence (Burssens et al., 2000). Furthermore, prolonged or intense salt stress treatments cause leaf chlorosis, bleaching, and necrosis. All adverse effects of salt stress, including leaf necrosis and the inhibition of root and rosette growth, were apparent at lower doses in rpn10-1 compared with the wild type and were suppressed by the spr1-3 mutation (Figure 6B). The fresh weight of rpn1a-4 spr1-3 and rpn10-1 spr1-3 plants grown for 2 weeks on salt-containing media was significantly increased compared with the single rpn1a-4 and rpn10-1 mutants but did not reach the wild-type levels (Figure 6C). The effects of NaCl on root elongation were measured after 3 d of treatment (Figure 6D). On the control media, spr1-3 suppressed root elongation both in the wild type and in proteasome mutants, probably because it caused the right-hand skewing of roots. On NaCl-containing media, root elongation of rpn1a-4 spr1-3 and rpn10-1 spr1-3 plants was increased compared with the respective single 26S proteasome mutants but again did not reach the length of the wild-type roots treated with the same dose of NaCl (Figure 6D). Thus, the spr1-3 mutation did indeed suppress the salt hypersensitivity of proteasome mutants, but this suppression was partial.

Next, we performed a time-course analysis of the impact of salt stress on GFP-TUA6–labeled MT arrays in wild-type, spr1-3, rpn10-1, and rpn10-1 spr1-3 plants (Figure 7). In the wild type, MT disassembly was nearly complete after a 4-h-long exposure to 200 mM NaCl (Figure 7A). By contrast, the cortical MT network in rpn10-1 cells remained largely intact up until 6 h into the treatment. The spr1-3 mutation suppressed the delayed MT depolymerization in the rpn10-1 mutant, but similarly to the effects of spr1-3 on whole-seedling salt tolerance, it did not fully revert the MT disassembly rate back to the wild-type level. Analyses of the number of MTs per unit length confirmed both the delayed salt response in rpn10-1 and the partial suppression of this delay by the spr1-3 mutation (Figure 7B). We concluded that the SPR1 stabilization in proteasome mutants causes salt hypersensitivity by slowing down the salt stress–induced depolymerization of MTs. However, since the proteasome mutant salt stress responses did not completely revert back to the wild-type level when the spr1-3 mutation was introduced, we also concluded that SPR1 is not the only proteasome target that inhibits salt stress–induced MT depolymerization.

The fast, salt-induced MT depolymerization response in cells of the wild type and rpn10-1 spr1-3 double mutant was followed by the start of new MT formation (8-h time point; Figure 7). By contrast, after 8 h of treatment, the MT disassembly was just completed and no distinct MTs could be detected in the rpn10-1 seedlings. This confirms an earlier report stating that fast MT disassembly in response to salt stress is required for the timely formation of new MT networks potentially better suited to tolerate high salt concentrations (Wang et al., 2007).

Additional proof that increased SPR1 abundance is a cause for salt stress hypersensitivity was obtained by analyzing transgenic plants that overexpress SPR1. In theory, plants that overexpress SPR1 should respond to salt stress similarly to proteasome mutants in which the SPR1 level is increased because of a reduced degradation rate. Previous work already showed that SPR1 overexpression indeed leads to increased tolerance to propyzamide in shoot organs (Nakajima et al., 2004). Here, we show that increased tolerance to propyzamide can also be observed in the root-swelling assay and is indeed associated with salt stress hypersensitivity (see Supplemental Figure 7 online). Collectively, these results confirm the importance of SPR1 removal for maintaining salt stress tolerance.

MG132 Promotes MT Stability and Salt Stress Hypersensitivity

Finally, to independently test the role of SPR1 proteolysis in the salt stress–induced restructuring of cortical MT arrays, we analyzed the combined effects of salt, propyzamide, and MG132 on Col-0 and spr1-3 seedlings (Figure 8).

Figure 8.

MG132 Stabilizes MTs and Increases the Sensitivity to Salt Stress.

(A) and (B) Four-day-old Col-0 and spr1-3 seedlings were transferred to MS/2 medium containing 4 μM propyzamide, 40 μM MG132, or 4 μM propyzamide and 40 μM MG132. Plants were grown vertically for three days before the root tips were photographed ([A]; bar = 0.5 mm), and the root lengths were measured (B). In (B), data represent an average root length with sd (n = 26), and the asterisks represent the significances of the differences between the wild type and spr1-3 (****P < 0.0001; ANOVA followed by Bonferroni multiple comparisons post-test).

(C) Visualization of GFP-TUB6–labeled cortical MTs in upper hypocotyl epidermal cells. Four-day-old seedlings were pretreated with 100 μM MG132 or DMSO in liquid MS/2 medium for 6 h, and then 20 μM propyzamide or 200 mM NaCl was added. Propyzamide treatments were done for 1 h and the NaCl treatments for 3 h prior to microscopy. Bars = 10 μm.

(D) and (E) Four-day-old Col-0 and spr1-3 seedlings were transferred to MS/2 medium containing 100 mM NaCl, 50 μM MG132, or both. Seedlings were photographed, and the root lengths were measured after 3 d of treatment. In (E), data represent an average root length with sd (n ≥ 25), and the asterisks represent the significances of the differences between the wild type and spr1-3 (****P < 0.0001, ***P < 0.001, and **P < 0.01; ANOVA followed by Bonferroni multiple comparisons post-test).

Similar to the genetic suppression of proteasome activity, pretreatment of Col-0 seedlings with MG132 inhibited the propyzamide-induced swelling of root tips (Figure 8A). Although MG132 also inhibited the swelling of spr1-3 roots, this effect was not as strong as in the wild type (Figure 8A). MG132 also counteracted the inhibitory effect of propyzamide on root elongation (Figure 8B). Treatment with either MG132 or propyzamide inhibited root elongation in Col-0 and spr1-3 seedlings. However, while MG132 reduced the root lengths to ~85% of the respective controls, the propyzamide treatment caused a more severe growth inhibition (to ~30% of the control values). Roots of Col-0 plants treated with both MG132 and propyzamide were ~50% longer than when treated with propyzamide alone, indicating that the proteasome inhibitor counteracts the effect of the MT destabilizing drug. Furthermore, this MG132 effect was also partially attenuated in the spr1-3 mutant, implying a role for SPR1 in the MG132-induced MT stabilization (Figure 8B).

To test if MG132 also increases salt sensitivity, we analyzed the response of GFP-TUB6–labeled MT arrays and the growth of Col-0 and spr1-3 on medium supplemented with 200 mM NaCl (Figures 8C to 8E). Indeed, MG132 suppressed the salt-induced depolymerization of cortical MTs (Figure 8C) and decreased the salt tolerance of wild-type seedlings (Figures 8D and 8E). However, compared with the wild type, MG132 was less effective in suppressing leaf expansion (Figure 8D) and root elongation of salt-treated spr1-3 seedlings (Figure 8E), confirming that the salt hypersensitivity caused by proteasome inhibition requires SPR1 function. Thus, we find that the increased MT stability and salt stress hypersensitivity of proteasome mutants can be phenocopied by treating wild-type plants with the proteasome inhibitor MG132.

DISCUSSION

Posttranscriptional and, in particular, proteasome-dependent regulation of tubulin/MT system dynamics have been described in considerably greater detail in animals than in plants. MT dynamics depend on the availability of tubulin heterodimers and on the activities of MAPs, and recent studies have shown that proteasome-dependent protein degradation determines the stability of both MT dynamics determinants. For example, the role of the proteasome and tubulin-specific chaperones in the degradation of tubulins released by MT depolymerization has been documented in animals (Bhamidipati et al., 2000; Ren et al., 2003; Bartolini et al., 2005; Voloshin et al., 2010). Depolymerization of MTs also leads to the proteasome-dependent degradation of tubulins in plants, but the molecular players that prime tubulin for proteolysis and that deliver tubulin heterodimers (or monomer-chaperone complexes) to the proteasome are still unidentified (S. Wang et al., 2011).

In animals, the proteasome has also been shown to regulate MT dynamics by influencing the stability of MAPs (David et al., 2002; Petrucelli et al., 2004; Peth et al., 2007; Poruchynsky et al., 2008; Ban et al., 2009). Our study extends this observation to plants and reveals that proteasome-dependent stability control is essential for the restructuring of MT arrays and the fine-tuning of MT dynamics. We show that genetic and pharmacological inactivation of 26S proteasome activity in Arabidopsis leads to an increase in MT stability and, consequently, to an improved tolerance of MT destabilizing drugs. The increased MT stability in proteasome mutants was largely the result of stabilization of the plant-specific +TIP SPR1. Thus, the accumulation of SPR1 resulting from either overexpression of the SPR1 transgene (Nakajima et al., 2004) or a reduced degradation rate (this study) is sufficient to promote increased MT stability in plant cells.

Although the increased stability of SPR1 in 26S proteasome mutants implies that this MAP is targeted for degradation in a ubiquitin-dependent manner, we were unable to detect any candidate ubiquitinated SPR1 forms with higher molecular weights. Similar results were described for several other known proteasome targets (Dill et al., 2001; Lopez-Molina et al., 2001; Xie et al., 2002; Smalle et al., 2003; Gagne et al., 2004), suggesting that the detection of ubiquitin conjugates can be technically challenging. Proteomic studies have shown that only a very small fraction of any given 26S proteasome target protein exists in its ubiquitinated form (Kaiser and Tagwerker, 2005). Furthermore, it can be envisioned that even upon partial stabilization in proteasome mutant backgrounds, the ubiquitinated SPR1 forms remain undetectable due to the actions of ubiquitin proteases that are known to revert a substantial fraction of ubiquitinated proteins back to their unmodified forms (Smalle and Vierstra, 2004; Kaiser and Tagwerker, 2005).

While our data show that SPR1 stabilization plays a pivotal role in the altered MT dynamics of proteasome mutants, they also suggest the involvement of other MAPs. For example, whereas the increased rescue frequencies and decreased frequencies of catastrophe (Figure 2, Table 1) that reflect an overall increase in MT stability in rpn10-1 cells are in agreement with the predicted effects of increased SPR1 action, the slower growth of MTs (Table 1) cannot be explained by an increase in SPR1 activity and suggests that loss of proteasome function also affects one or more proteins that control the MT polymerization rate. Another example is presented in Figure 4: The spr1-3 mutation did not fully suppress the increased propyzamide tolerance of proteasome mutants, suggesting the stabilization of one or more additional proteins that promote MT stability. The involvement of MT-stabilizing factors other then SPR1 could also explain the observation that the right skewing of spr1-3 roots was enhanced in the rpn10-1 spr1-3 and rpn1a-4 spr1-3 double mutants (see Supplemental Figure 3 online). The spr1 mutants are unusual in that they combine a decrease in MT stability with the right skewing of roots even though this developmental phenotype is typically the result of MT stabilization (Sedbrook and Kaloriti, 2008). Because spr1-3 partially reversed the increased MT stability of proteasome mutants (Figure 4), we expected that the right skewing of spr1-3 roots would also be reversed. However, since the right skewing was enhanced in the double mutants, we propose that this phenotype reflects the function of other MT-stabilizing MAPs whose activity is enhanced by their stabilization in the proteasome mutant backgrounds.

Our observation that MT stability depends on the SPR1 degradation rate suggested that the activity of this protein is controlled posttranslationally and implied the existence of developmental or environmental cues that influence MT dynamics by regulating SPR1 stability. Indeed, we found that SPR1 proteolysis is enhanced under salt stress conditions (Figure 5) and that SPR1 stabilization leads to salt stress hypersensitivity in 26S proteasome mutants (Figure 6). Furthermore, we showed that the increased salt stress sensitivity of proteasome mutants is caused by a slower rate of MT disassembly, which is a direct result of the SPR1 stabilization. These results are in agreement with the results of an earlier study that described that suppression of MT disassembly by the MT-stabilizing drug paclitaxel caused salt stress hypersensitivity (Wang et al., 2007).

Arabidopsis 26S proteasome mutants are also characterized by a decrease in growth rate that reflects a reduced rate of mitosis (Kurepa et al., 2009b). Since MT depolymerization is also known to inhibit mitosis, an alternative explanation for the salt stress hypersensitivity of 26S proteasome mutants is that the reduced mitotic rates make them more susceptible to the growth inhibitory effects of salt stress. However, this explanation is unlikely because the spr1-3 mutation suppressed the salt hypersensitivity but not the reduced growth rate of proteasome mutants (Figure 6). Furthermore, the SPR1 overexpression lines were also salt hypersensitive but did not display any decrease in growth rate (see Supplemental Figure 7 online), thus confirming that the increased SPR1 abundance in 26S proteasome mutants is indeed the most likely cause for their salt stress hypersensitivity.

Collectively, our data reveal an important role for proteasome-dependent regulation of SPR1 in the survival of plants challenged by high salinity. Figure 9 outlines a model that summarizes the role of SPR1 in the MT disassembly response to salt stress. According to this model, SPR1 is a moderately stable protein under normal growth conditions and is localized predominantly at the growing ends of MTs where it inhibits their disassembly. Upon salt stress, 26S proteasome–dependent degradation of SPR1 is accelerated, and the MT depolymerization needed for the survival of plant cells under salt stress is facilitated. By contrast, salt stress–induced SPR1 degradation is attenuated in proteasome mutants, thus slowing down MT disassembly and causing salt stress hypersensitivity. Since the spr1-3 mutant was not more tolerant to salt stress than the wild type, we can also conclude that this destabilization mechanism is strong and fast enough to suppress SPR1 activity to a level where it does not interfere with the salt-induced MT disassembly process. Accordingly, we propose that SPR1 destabilization does not initiate the salt-induced MT disassembly, but its removal is required to allow the timely completion of this process.

Figure 9.

Model Summarizing the Role of SPR1 Proteolysis in the Salt Stress Tolerance of Wild-Type and Proteasome Mutant Cells.

SPR1, a (+)-end MAP, is degraded by the 26S proteasome (26SP). In 26S proteasome mutants, the degradation rate is reduced and the MTs are more stable due to SPR1 accumulation. Upon salt stress perception, a still unknown mechanism leads to the increased proteasome-dependent degradation of SPR1 that facilitates MT depolymerization. In proteasome mutants, salt stress–induced degradation of SPR1 is reduced, which slows down MT depolymerization and causes salt stress hypersensitivity. This schematic is simplified for the purpose of clarity: no other (+)-end MAPs but SPR1 are depicted, and all (−)-end components are omitted.

The rapid MT depolymerization response to salt stress is thought to facilitate the formation of a new MT network that allows cells to better withstand the damaging impacts of high salt concentrations. Indeed, in plant cells exposed to prolonged salt stress, the initial massive MT depolymerization was followed by the formation of new MT networks (Wang et al., 2007). It has been established that salt stress leads to a reduction and reorientation of cell expansion and that these growth alterations are important for adapting to and surviving high salinity (Munns and Tester, 2008). As major determinants of the direction and rate of cell expansion, MT networks would indeed have to be rapidly reorganized to promote and facilitate such changes in growth. The new MT network has a more random MT organization compared with cells of unstressed plants (Wang et al., 2007). This corresponds well with the need for a reduced growth rate, as rapid cell elongation tends to require a transverse orientation of MTs to the direction of growth (Chan et al., 2011; Crowell et al., 2011).

Because MT depolymerization is known to increase calcium channel activity (Thion et al., 1998), the initial MT disassembly response is also thought to be important for increasing the cytosolic calcium concentration, which is a major requirement for the adaptation to salt stress (Wang et al., 2007; Mahajan et al., 2008). The calcium burst was also shown to be essential for the formation of new MTs after prolonged salt stress exposure, suggesting that the initial MT depolymerization response not only allows the development of new MT networks but also establishes optimal conditions for MT synthesis (Wang et al., 2007).

While our study highlights the importance of proteasome-dependent proteolysis in the regulation of MT dynamics during salt stress, it also raises a number of questions. The first question relates to the stress-specific effects on MTs. We have shown that osmotic stress does not cause MT disassembly and does not promote SPR1 destabilization. Other stresses, such as cold, heat, and treatments with nanoparticles (Smertenko et al., 1997; S. Wang et al., 2011), induce changes in plant MT networks, and future studies need to address whether this is also mediated via the stability control of MAPs. Whereas SPR1 is currently the only known 26S proteasome target among the plant MAPs, the partial suppression of 26S proteasome mutant MT phenotypes by spr1-3 suggests the existence of other proteins with SPR1-like activities that are also stabilized when proteasome function is impaired. Some obvious candidates to consider are the family of SPR1-like proteins, which were shown to have functions similar to SPR1 (Nakajima et al., 2006), and the evolutionarily conserved, MT-stabilizing protein EB1, which is known to be targeted for proteasome-dependent proteolysis in human cells (Peth et al., 2007) and was reported to interact with SPR1 potentially to regulate directional plant cell expansion (Kaloriti et al., 2007).

The second question relates to the mechanism by which the perception of the salt stress signal leads to proteasome-dependent degradation of SPR1. Signal-induced site-specific phosphorylation often initiates the ubiquitin-dependent targeting of a protein to the 26S proteasome (Chen et al., 1995; Matsuzaki et al., 2003; Smalle and Vierstra, 2004). Phosphorylation is also a general regulation mechanism that reduces the binding affinity of MAPs for MTs and thus leads to MT destabilization (Drewes et al., 1998; Matenia and Mandelkow, 2009; Beck et al., 2010). On the other hand, it has been reported that a mitogen-activated protein kinase (MAPK) cascade plays a critical role in the salt stress response of Arabidopsis (Teige et al., 2004). Therefore, salt stress–induced proteasome-dependent control of MT depolymerization could involve a salt stress–activated MAPK cascade that leads to the phosphorylation of SPR1, followed by its interaction with a specific ubiquitin ligase and degradation by the 26S proteasome. This sequence of events would require that a salt stress–responsive MAPK localizes close to SPR1 or is relocated to SPR1 upon stress (i.e., constitutively or conditionally associated with MTs). Recent studies in Arabidopsis identified a number of MAPKs (MAK18 and MAK4) involved in stress signaling that are associated with MTs or are involved in the regulation of MT dynamics (Walia et al., 2009; Beck et al., 2010). On the other hand, the SPR1 protein contains nine putative phosphorylation sites (as predicted by NetPhos 2.0), and one of them (Thr-76) is a part of MAPK consensus sequence PXS/TP or S/TP, suggesting that the potential for MAPK-dependent SPR1 regulation is indeed an interesting topic for future research.

METHODS

Plant Materials and Growth Conditions

Arabidopsis thaliana plants were grown on plates containing half-strength Murashige and Skoog medium with 1% Suc (MS/2) as described previously (Smalle et al., 2003). Plants were grown in a controlled environment chamber at 22°C with continuous light (140 μmol photons m⁻² s⁻¹). The proteasome mutants rpn1a-4, rpt2a-2, rpn10-1, and rpn12a-1 (all in Col-0 background, and all carrying the kanamycin resistance gene) have been described (Kurepa et al., 2008; Wang et al., 2009). The spr1-3 mutant and GFP-TUA6 and GFP-TUB6 overexpression lines (all in Col-0 background) have also been described (Nakajima et al., 2004; Abe and Hashimoto, 2005). For the generation of double mutants, putative homozygous double mutants were selected based on their phenotypes, and their genotypes were confirmed by immunoblotting analyses.

Treatments

Oryzalin and propyzamide were purchased from Sigma-Aldrich and MG132 from Enzo Life Sciences. All drugs were made as 1000× stocks. MG132, CHX, and propyzamide were dissolved in DMSO and oryzalin in 100% ethanol. All control experiments included a 1× dose of the solvent. For all root elongation and root tip assays, plants grown on vertically positioned MS/2 plates for 5 d were transferred to drug- or mock-supplemented media. Test plates were positioned vertically, and the root length was marked daily. After 3 d of treatments, plants were photographed. Photomicrographs of representative root tips were taken with an Olympus SZX12 microscope equipped with a DP12 camera. For stress tolerance assays, 4- or 5-d-old seedlings were transferred to test plates with NaCl or mannitol and were grown vertically. For all morphometric analyses, the relevant parameter was measured from digital images using ImageJ (http://rsb.info.nih.gov/ij/). Unless specified otherwise, data are presented as mean values, and the error bars represent standard deviation. Statistical significance was determined by ANOVA tests followed by post hoc Bonferroni multiple comparison test. Post hoc statistical significance is indicated in the figures by asterisks or crosses. For all experiments, descriptive statistics, plotting, and the hypothesis testing were done using Prism 5.0d software (GraphPad Software).

Confocal Microscopy Analysis of MTs and Scoring

MT distributions were analyzed in 35S:GFP-TUA6 and 35S:GFP-TUB6 lines in Col-0, rpn10-1, and rpn10-1 spr1-3 backgrounds using an Olympus Fluoview FV1000 confocal laser scanning microscope equipped with an argon ion laser for the excitation of GFP (excitation 488 nm and barrier 500 to 550 nm) essentially as described (S. Wang et al., 2011). For all experiments, 4-d-old seedlings were used, and epidermal cells of upper hypocotyl regions were analyzed. Except for the MT dynamics measurements, seedlings were mounted, incubated, and observed in water or an aqueous solution of the tested drugs. For the measurement of individual MT dynamics, seedlings were mounted in liquid MS/2 medium, and time-lapse imaging was performed at 1% laser power and 4-s intervals for 3 to 6 min. Images were processed and analyzed using ImageJ as previously described (Buschmann and Lloyd, 2008). The dynamics and the frequency of catastrophe and rescue were calculated as described (Dhonukshe and Gadella, 2003; Abe and Hashimoto, 2005). To quantify MT network density, MTs crossing the middle long axis of upper hypocotyl cells were counted. The results were expressed as MT number per unit distance as described previously (Ishida et al., 2007).

Immunoblotting Analyses

For all analyses, plants were weighed and transferred to 1.7-mL tubes for treatment. After treatment, plants were blotted, frozen in liquid N₂, and ground in three volumes of 2× Laemmli sample buffer. Total proteins were separated by SDS-PAGE (14% acrylamide separating gel for SPR1 analyses, 8% for RPN1 analyses, and 10% for all other analyses). Proteins were transferred to nitrocellulose membranes (Hybond C-Extra; GE) and probed as previously described (S. Wang et al., 2011). Antibodies against RPN1 and SPR1 were described (Nakajima et al., 2004; Wang et al., 2009). To test if the signal intensity for SPR1 in all experiments falls within the linear dynamic range of the assay, immunoblots with a dilution series of total protein extracts were probed with anti-SPR1 (dilution 1:1000), developed using chemiluminescent substrate (Thermo Scientific Pierce ECL Plus substrate) and analyzed using the linear best fit analyses (see Supplemental Figure 8 online). The anti-RPN10 sera were obtained from Enzo Life Sciences. The anti-glutamine synthase (GS) serum was purchased from Agrisera AB. GS was assayed as a control. This is a stable protein, and its steady state levels are not affected by inhibition of proteasome activity (Kurepa et al., 2010). Horseradish peroxidase–conjugated and alkaline phosphatase–conjugated secondary antibodies were purchased from Santa Cruz Biotechnology.

Transcript Analyses

Total RNA was isolated using TRIzol reagent (Invitrogen). RNA was treated with TURBO DNase (Ambion), and the RNA concentrations were determined spectrophotometricaly (NanoDrop 2000; Thermo Scientific). One microgram of RNA per sample was used for the synthesis of the first-strand cDNA (iScript reverse transcription supermix; Bio-Rad).

For the quantification of SPR1 transcripts, real-time RT-PCR was performed using the StepOne real-time PCR system (Applied Biosystems) and the DyNAmo Flash SYBR Green qPCR kit (Finnzymes) in total reaction volume of 20 μL. The SPR1 primers used were 5′-AGCCTGCAGAGCTTAACAAG-3′ and 5′-TGAACTTTGGTCGAAGGACG-3′, and the primers for reference genes ACT2, ACT8, EF-1α, and GADPH were as described (Czechowski et al., 2005). Selection of the reference gene(s) best for the normalization in each experiment was calculated using geNorm (Vandesompele et al., 2002).

Analyses of Ubiquitin Conjugates and Proteasome Activity

Immunoblotting analysis of ubiquitin conjugates was done as described (Kurepa et al., 2008). The antipolyubiquitinated protein serum was from Enzo Life Sciences. For proteasome activity assays, samples were ground in 1.25 volumes of extraction buffer as described (Kurepa et al., 2008). Protein concentrations were measured using Bradford reagent (Bio-Rad), and total proteasome activity was measured using the Suc-LLVY-AMC assay as described (Kurepa et al., 2008).

Generation of Transgenic Plants Overexpressing SPR1

To generate the overexpression construct, SPR1 cDNA was amplified using attB PCR primers attB1SPR1 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTAATGGGTCGTGGAAACAGC-3′ and attB2SPR1 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTCTTACTTGCCACCAGTGAAGA-3′. The cDNA was introduced into pDONR221 via BP reaction and then to pEarlyGate100 (Earley et al., 2006) by LR recombinase (Invitrogen). Transgenic plants were selected on MS/2 plates containing 10 μM l-phosphinothricin (Gold Biotechology), and homozygous T3 plants were used for the analyses.

Accession Numbers

Sequence data from this article can be found in The Arabidopsis Information Resource (http://www.Arabidopsis.org/) under the following accession numbers: RPN1a (At2g20580), RPT2a (At4g29040), RPN10 (At4g38630), RPN12a (At1g64520), SPR1 (At2g03680), TUA6 (At4g14960), and TUB6 (At5g12250). Arabidopsis T-DNA insertion mutants and transgenic lines and their identification numbers are SALK_027970 (rpn1a-4), SALK_005596 (rpt2a-2), 35S:GFP-TUA6 (CS6551), and 35S:GFP-TUB6 (CS6550).

Supplemental Data

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

• Supplemental Figure 1. Visible Phenotypes and Propyzamide Responses of the Double Mutant rpn1a-4 rpn10-1.

• Supplemental Figure 2. The Steady State SPR1 Transcript Levels in Proteasome Mutants Are the Same as in the Wild Type.

• Supplemental Figure 3. Right-Handed Helical Growth Is Enhanced in rpn1a-4 spr1-3 and rpn10-1 spr1-3 Mutants.

• Supplemental Figure 4. Propyzamide Treatment Reverts Amplified Root Skewing in the rpn1a-4 spr1-3 Mutant.

• Supplemental Figure 5. Effects of NaCl Treatments on the SPR1 Transcript Level, SPR1 Protein Abundance, Accumulation of Ubiquitinated Proteins, and Proteasome Activity.

• Supplemental Figure 6. Treatment with 10 μM Propyzamide Causes Depolymerization of Cortical MTs.

• Supplemental Figure 7. Overexpression of SPR1 Increases MT Stability and Leads to Salt Stress Hypersensitivity.

• Supplemental Figure 8. Linear Relationship between SPR1 Abundance and Signal Intensity on Immunoblots Developed Using Chemiluminescent Substrate.