Protein Phosphatase 2Cs and Microtubule-Associated Stress
Protein 1 Control Microtubule Stability, Plant Growth, and
Drought Response
Govinal Badiger Bhaskara, Tuan-Nan Wen,
1
Thao Thi Nguyen,
1,2
and Paul E. Verslues
3
Institute of Plant and Microbial Biology, Academia Sinica, Taipei 11529, Taiwan
ORCID ID: 0000-0001-5340-6010 (P.E.V.)
Plant growth is c oordinated with environmental factors, including water availability during times of drought. Microtubules
influence cell expansion; however, the mechanisms by which environmental signals impinge upon microtubule organization and
whether micro tubule-r elated factors limit growth during drought remains unclear. We found that three Clade E Growth-Regulati ng
(EGR) Type 2C protein phosphatases act as negative growth regulators to restrain growth during drought. Quantitative
phosphoprote omics indicated that EGRs target cytoskeleton and plasma membrane-associat ed proteins. Of these, Microtubule-
Associated Stress Protein 1 (MASP1), an uncharacterized protein, increased in abundance during stress treatment and could bind,
bundle, and stabilize microtubules in vitro. MASP1 overexpression enhanced growth, in vivo microtubule stability, and recovery of
microtubule organization during drought acclimation. These MASP1 functions in vivo were dependent on phosphorylation of
a single serine. For all EGR and MASP1 mutants and transgenic lines examined, enhanced microtubule recovery and stability were
associated with increased growth during drought stress. The EGR-MASP1 system selectively regulates microtubule recovery and
stabilit y to adjust plant growt h and cell expansion in response to changi ng environmenta l cond iti ons. Modificati on of EGR-MASP1
signaling may be useful to circumvent negative growth regulation limiting plant productivity. EGRs are likely to regulate additional
proteins involved in microtubule stability and stress signaling.
INTRODUCTION
Drought and reduced soil water potential (c
w
) decrease plant
growth by restricting both cell expansion and cell division (Skirycz
and Inzé, 2010; Tardieu et al., 2011). Drought is thus a major
constraint on plant productivity (Boyer, 1982; Claeys and Inzé,
2013).Activedownregulationofgrowth, particularly shootgrowth,
during drought can be beneficial to plant survival by limiting the
leaf area for transpiration and thus allowing conservation of soil
water. However, such downregulation can also decrease biomass
production andyield more thanis needed under moderateseverity
drought. Understanding growth regulation in general, as well as
identifying factors that detect changes in water status and limit
growth duringdrought,has the potential to improvecrop biomass,
productivity, andyieldstability (Claeys and Inzé, 2013). Regulation
of growth under moderate drought is distinct from mechanisms
determining survival of severe water deprivation where no growth
occurs (Skirycz et al., 2011; Verslues, 2016).
A major facet of growth regulation is the control of cell expansion.
The extent of cell expansion depends on solute deposition to gen-
erate turgor, cell wall properties, mechanical cues from neighboring
cells, as well as many types of int ra- and extracellular signaling
(Hamant et al., 2008; Sampathkumar et al., 2014; Feng et al.,
2016). Cell shape and expansion are greatly influenced by events
at the plasma membrane-cell wall interface, including vesicle
trafficking and cortical microtubule (MT)-mediated control of cell
wall deposition (Landrein and Hamant, 2013; McMichael et al.,
2013; Lei et al., 2014, Endler et al., 2015; Feng et al., 2016).
Disruption of cortical MT organization by oryzalin disrupts di-
rectional (anisotropic) cell expansion and causes isotropic cell
swelling (Baskin et al., 1994; Sugimoto et al., 2003). M any mu-
tants with altered MT organization or stability are impaired in
anisotropic growth and have developmental defects including
spiral, twisting growth patterns (Thitamadee et al., 2002; Sho ji
et al., 2004; Bannigan et al., 2006; Ambrose et al., 2007; Korolev
et al., 2007; X. Wang et al., 2007; Buschmann et al., 2009;
Nakamura and Hashimoto, 2009; Hamada et al., 2013; Liu et al.,
2016). Cortical MTs may themselves act as sensors or trans-
ducers of mechanical signals that regulate growth (Landrein and
Hamant, 2013; Nick, 2013).
There is arapid(1 h or less) loss of cortical MT organization upon
exposure to salt or osmotic stress (Komis et al., 2001; Dhonukshe
et al., 2003; Ban et al., 2013; Fujita et al., 2013; Endler et al., 2015).
For osmotic stress, loss of MT organization is mediated in large
part by phosphorylation of a-tubulin, which blocks the polymer-
ization of a-b tubulin dimers (Ban et al., 2013; Fujita et al., 2013).
As minus end depolymerization and MT catastrophe events
continue even though polymerization is inhibited, the result is
a loss of MT organization. Salt stress-induced loss of MT or-
ganization also involves proteasome-dependent degradation of
the MT plus end binding protein Spiral 1 (SPR1; Wang et al.,
2011).F or salt stress, thereis evidence thatsubsequentrecovery
of MT organization is essential for acclimation and resumption of
1
These authors contributed equally to this work.
2
Current address: Department of Biochemistry , University of Wisconsin,
Madison, WI 53706.
3
Address correspondence to paulv@gate.sinica.edu.tw.
The author responsible for distribution of materials integral to the findings
presented in this article in accordance with the policy described in the
Instructions for Authors (www.plantcell.org) is: Paul E. Verslues (paulv@
gate.sinica.edu.tw).
www.plantcell.org/cgi/doi/10.1105/tpc.16.00847
The Plant Cell, Vol. 29: 169–191, January 2017, www.plantcell.org ã 2016 American Society of Plant Biologists. All r ights reserved.
Downloaded from https://academic.oup.com/plcell/article/29/1/169/6099020 by guest on 20 August 2022
growth(C. Wang etal., 2007;Wang et al.,2011), andthis recovery
involves the microtubule-associated proteins Companion of
CelluloseSnynthase1 (CC1)andCC2 (Endleret al.,2015).For low
c
w
and osmotic stress, the role of MT recovery in stress accli-
mationand growthislessclear.Phosphorylation ofa-tubulin was
observed under lethal, or near lethal, plasmolyzing osmotic
treatments, which precluded observations of longer term re-
covery and acclimation. Loss of MT organization induced by
a-tubulin phosphorylation may mainly be a protectio n mecha-
nismto prevent cellular damage(Ban etal., 2013). Treatmentwith
high molecular weight polyethylene glycol (PEG), which reduces
c
w
and mimics many aspects of soil drying while avoiding
plasmolysis, also causes an initial loss of MT organization fol-
lowed by later recovery (Mei et al., 2012). This later MT recovery
has been little investigated and the role of MT in growth and cell
expansion during drought is uncertain.
The planthormonesethylene and abscisic acid (ABA) are potent
regulators of plant growth, drought response, and microtubule
organization. Increased ethylene production or increased ethyl-
ene response inhibits anisotropic cell elongation and induces
isotropic radial swelling of the hypocotyl and roots (Guzmán and
Ecker, 1990; Pierik et al., 2007). The effect of ethylene on cell
expansion involves rapid and extensive rearrangements of MT
organization (Roberts et al., 1985; Pierik et al., 2007; Polko et al.,
2012; Ma et al., 2016). Drought stress can lead to increased
ethylene production that, in combination with ABA accumulation,
regulates root and shoot growth during drought (Spollen et al.,
2000; LeNoble et al., 2004). Exogenous ABA treatment can also
induce MT reorganization (Seung et al., 2013; Takatani et al.,
2015), although whether such MT reorganization is involved in
ABA effects on growth during drought is not as clear.
Protein phosphorylation-dephosphorylation is important for
drought and ABA signaling as well as MT organization. Phar-
macological studies found that inhibitors of Type 1 and Type 2A
(PP2A) protein phosphatases, as well as kinase inhibitors, caused
root swelling (loss of anisotropic cell expansion) and MT
disorganization (Baskin and Wilson, 1997). Also, some plant
MT-associated proteins (MAPs) are known to be phosphorylated.
For example, phosphorylation of MAP65 family proteins changes
their MT binding properties (Komis et al., 2011). An important
aspect of drought-related signaling is the activity of Type 2C
protein phosphatases (PP2Cs), especial ly th e Cl ade A PP2Cs
whoseactivityis regulatedby interactionwith PyrabactinResistant-
Like/Regulatory Component of Absc isic Acid Rec eptor (PYL /
RCAR) ABA receptors (Cutler et al., 2010; Fuchs et al., 2013). In-
teraction withPYLsisspecific toClade A PP2Cs,which accountfor
only nine of the 80 PP2Cs present in Arabidopsis (Arabidopsis
thaliana; Fuchs et al., 2013; Sugimoto et al., 2014). For most of the
other PP2Cs in Clades B to F, there is little information on their
physiologicalfunctionoronthephosphorylationsitesthey regulate.
Untargeted “shotgun” phosphoproteomics has been used to
identify phosphopeptides whose abundance is rapidly altered in
response to osmotic stress or exogenous ABA (Kline et al., 2010;
Steckeretal.,2014; Minkoffetal.,2015)andto identify targetsofthe
SnRK2 kinases (Umezawa et al., 2013; Wang et al., 2013). Quan-
titative phosphoproteomics is a promising method to identify
PP2C-regulated phosphoproteins; however, we are not aware of
any studies that have attempted such an analysis.
A search in our laboratory for genes affecting drought-regulated
traits, including proline accumulation and plant growth, led us to
characterize the function of three Clade E PP2Cs that we found to
regulate plant growth and MT organization. Phosphoproteomic
analysisidentifiedputative targets of thesePP2Cs, including a new
MT binding protein that accumulated during low c
w
stress and
promoted MT stability and growth in a phosphorylation-specific
manner. Our results uncover cellular functions of uncharacterized
Clade E PP2Cs, identify regulators of MT organization, and dem-
onstrate the importance of MT organization and stability in con-
tinued plant growth during drought stress.
RESULTS
The EGR Clade E PP2Cs Are Negative Regulators of Growth
during Drought Stress
We hypothesized that some of the nearly 70 relatively uncharac-
terized Arabidopsis PP2Cs (Schweighofer et al., 2004; Fuchs et al.,
2013) may function in drought signaling and identified three of the
13 Clade E PP2Cs as genes of particular interest (Supplemental
Figure 1A; Schweighofer et al., 2004). These genes are hereafter
referred to as Clade E-Growth-Regulating PP2C-1 (EGR1), EGR2,
and EGR3. We focused on the EGRs in part because expression
ofEGR1 iscorrelated withthedrought-regulatedgeneD
1
-pyrroline-
5-carboxylate synthetase1 in the Plant Gene Expression Database
(Horan et al., 2008) and EGR2 and EGR3 are close homologs
of EGR1. EGR3 has a lso been recently identified as a putative
leaf growth reg ulator by transcriptome analysis of stress- and
circadian-regulated genes (Dubois et al., 201 6). EGR expres-
sion was upregulated by low c
w
in bo th the w ild type and the
ABA-deficientmutantaba2-1 (Figure1A)butwas lessresponsiveto
exogenous ABA (Supplemental Figures 1B and 1C), suggesting
that EGR function may be distinct from ABA-regulated drought
responses.
We isolated T-DNA mutants for all three EGRs (Supplemental
Figure 1D). Because EGR1 and EGR2 were strongly induced by
stress and are the most closely related to each other in sequence
(Supplemental Figure 1), we also generated an egr1-1 egr2-1
double mutant. Initial experiments showed that egr mutants
had increased proline accumulation after transfer to PEG-infused
agar plates adjusted to c
w
representing conditions typical of mild
to more severe, but not lethal, drought stress (20.5 to 21.2 MPa;
Supplemental Figure 1E). We further assayed seedling growth in
unstressed control conditions (20.25 MPa) and at low c
w
(21.2
MPa). Growthof all genotypes, includingthe wild type(Supplemental
Figure 2) , was inhibited by low c
w
; however, egr mutants maintained
higher root elongation and seedling dry weight (Figures 1B and 1C).
egr1-1 egr2-1 also had significantly increased growth in the un-
stressed control condition (Figure 1B). For EGR2 and EGR3,two
T-DNA alleles had essentially identical effects on growth (Figure 1B;
egr2 and egr3 are combined data of the two alleles for each gene)
and proline accumulation (Supplemental Figure 1E). We were unable
to obtain multiple T-DNA alleles for EGR1; however , Pro35S:YFP-
EGR1 complemented egr1-1 (Fi gure 1D). The complementation and
similar phenotypes ofallegr mutants indicated that theegr1-1 T-DNA
insertion in the 59 UTR of EGR1 likely blocks protein translation even
170 The Plant Cell
Downloaded from https://academic.oup.com/plcell/article/29/1/169/6099020 by guest on 20 August 2022
Figure 1. Clade E EGR PP2Cs Are Negative Regulators of Growth and Cell Expansion.
(A) EGR expression at low c
w
(21.2 MPa) relativeto the unstressed control forCol-0 (W.T.) andABA-deficient mutantaba2-1. Dataare means 6 SE, n =6 from
two independent experiments.
(B) Seedling dry weight (D.W.)androotelongation of theEGR mutantandoverexpression lines inunstressedcontrol conditions or 10daftertransfer to low c
w
(21.2 MPa). Data are relative to the wild type (mean 6 SE, n = 6 to 8 for seedling dry weight and n = 18 to 24 for root elongation, asterisk indicates significant
difference compared with the wild type by one-sided t test [P # 0.05]). Dashed red line indicates the wild-type level of growth (100%). Three or four seedlings
were combined for each dry weigh measurement. egr2 and egr3 are combined data of two T-DNA alleles for each gene (Supplemental Figures 1A and 1D).
Growth values of Col-0 wild type used for normalization are shown in Supplemental Figure 2.
(C) Representative seedlings of Col-0 wild type and egr1-1 egr2-1 after low c
w
(21.2 MPa) treatment. Five-day-old seedlings were transferred to low c
w
and
photographs taken 10 d later. Bar = 1 cm.
Phospho-Signaling, Microtubules, and Growth 171
Downloaded from https://academic.oup.com/plcell/article/29/1/169/6099020 by guest on 20 August 2022
though some mRNA is still produced in the mutant (Supplemental
Figure 1D). Conversely, overexpression of EGR1 (intheCol-0wild-
type background) decreased growth and proline accumulation
(Figure 1D). Together, these data indicate that EGRs act as negative
regulators of several drought-related phenotypes.
To show that the egr growth phenotypes extended across
different developmental stages andmethodsof low c
w
imposition,
18-d-old wild-type and egr mutant plants were exposed to partial
soil drying followed by quantification of rosette fresh and dry
weight. The soil drying treatment, with partial rewatering to ensure
that all plants were exposed to the same level of moderate soil
drying stress, lasted 20 d andat the endof thisperiod rosettefresh
and dry weight of the stress-treated wild type was decreased by
nearly 70% compared with the well-watered control. Because the
degreeof soildrying wasmaintainedat amoderate level,there was
little wilting or decrease in leaf relative water content for any of the
genotypes assayed. Consistent with the PEG plate assays, egr
mutants had increased rosette fresh and dry weight in the soil
drying treatment. egr1-1 egr2-1 had the strongest effect and also
had increased growth in the well-watered control (Figures 1E and
1F). Scanning electron microscopy showed that egr1-1 egr2-1
plants had a larger average size of epidermal pavement cells
compared with the wild type under water stress conditions while
still maintaining normal leaf morphology (Figures 1G and 1H;
Supplemental Data Set 1). Thus, the increased growth of egr
mutants was, at least in part, due to greater cell expansion and
ability to maintain cell expansion despite water limitation. These
data further demonstrated the role of EGRs in restricting growth.
Putative EGR Targets Identified by Quantitative
Phosphoproteomics Are Enriched in Cytoskeleton- and
Plasma Membrane-Associated Proteins
Toidentifytargetso fEGR regulation, weperformedquantitative
phosphoproteome and transcriptome analyses of the wild type
and egr1-1 egr2-1 under control or low c
w
(21.2MPa,96h)
treatments. The gene expression data were mainly used as
a compa nion for the phos phoprote omics data t o compare
proteins with stress-induced increase or decrease in phos-
phopeptide abundance to genes transcriptionally re gulate d by
low c
w
. Our phosphoproteomic analysis was unique compared
with recent studies (Kline et al., 2010 ; Umezawa et al., 2013;
Wang et al., 20 13; Stecker et al., 2014; Minkoff et al., 2015) in
that we used longer term (96 h; Supplemental Figure 3A) low c
w
stress rather than short-term ABA or dehydration treatment.
This was done primarily because EGR expression, as well as
that of many ot her phosphata ses and kinases (Bhaskara et al.,
2012), was ind uced by such lo nger te rm stress tr eatment and
because egr growt h an d pro line ph enoty pes were observed i n
longer term low c
w
treatment (Figure 1; Supplemental Figure 1).
Our analysis identified more than 1500 phosphoproteins
(Supplementa l Figure 3B and Supplemental Data Sets 2 to 4),
manyof which were not previously listed in the PhosPhat database
(Supplemental Figure 3C). In the wild type, 119 phosphopeptides
were significantly more abundant at low c
w
compared with the
control condition, while 23 were significantly less abundant (Figure
2A; Supplemental Data Sets 5 and 6). Many of these proteins were
fromgeneswhoseexpressionwasunaffected bylowc
w
(Figure2A;
Supplemental Data Sets 5 and 6) and were not identified in
previous phosphoproteomic studies (Supplemental Figure 3D
and Supplemental Data Sets 4 and 7). Thus, our phosphopro-
teomic data revealed aspects of the low c
w
response that could
not be inferred from transcriptome data or previously e xisting
phosphoproteo mic data. Howeve r, it should be kept in mind
that the longer term low c
w
treatment used in our study allowed
ample time for protein abundance to change. Thus, differences
in phosphopeptide abundance could reflect a stress-indu ced
difference in phosphorylation stoichiometry or a change in protein
abundance. Both are of interest and distinct from transcriptional
regulation; however, further experiments will be needed to distin-
guish between these two possibilities.
The egr1-1 egr2-1 phosphoproteomic data identified putative
targets of EGR regulation that were consistent with EGR locali-
zation at the cell periphery and suggested a cytoskeleton-related
function for the EGRs. Phosphopeptide abundances in egr1-1
egr2-1 were compared with those of the wild type for both the
Figure 1. (continued).
(D) Pro35S:YFP-EGR1 complements the increased growth and proline accumulation of egr1-1 and suppresses growth and proline in the wild type. Proline
was measured 96 hafter transfer to21.2 MPa. Dataare combined fromtwo transgenic linesfor both theegr1-1 and wild-typebackgrounds and aremeans 6
SE (n = 4 to 12); asterisk indicates P # 0.05 compared with the wild type. Dashed red line indicates the wild type level of growth (100%).
(E) Rosette fresh weight (F.W.) and dry weight for egr mutants in well-watered control plants or in plants subjected to partial soil drying. Data are expressed
relative to the Col-0 wild type and are means 6 SE (n = 6 to 9) combined from two to three independent experiments. Two rosettes from plants grown in the
samesector of the same potwereusedforeach fresh and dry weightmeasurement.Asterisks indicate significantdifferencecompared with the wildtype(P#
0.05 by one-sided t test). Dashed red line indicates the wild-type level of growth (100%).
(F) Representative rosettes of the wild type and egr1-1 egr2-1 in control and soil drying treatments. Plants were 40 d old and were grown under short-day
conditions. The soil drying treatment started at 18 d after planting and continued for 22 d with partial rewatering to control the extent of soil drying (see
Methods for further details). Bars = 1 cm.
(G) Areas of epidermal pavement cells of leaf 6 from 38-d-old Col wild type and egr1-1 egr2-1 in control and soil drying treatments. Open boxes with green
median lines showdata for theunstressed control, andgray boxes withred median linesare stress treatment(black lines ineach box indicate the mean, while
box and whiskersindicate the 25to 75 and5 to 95percentile ranges, respectively, and black circles showoutlying data points). Data aremeans 6
SE (n =40to
80) combined from four to six plants. Lowercase letters above each box indicate significantly different groups (ANOVA on ranks, P # 0.05; Supplemental
Data Set 1).
(H) Representative scanning electron microscopy images with example epidermal pavement cells outlined in orange to illustrate the increased cell size in
egr1-1 egr2-1, while retaining normal morphology. Bars = 50 mm.
172 The Plant Cell
Downloaded from https://academic.oup.com/plcell/article/29/1/169/6099020 by guest on 20 August 2022
Figure 2. Phosphoproteomics Analysis of Wild-Type and egr1-1 egr2-1 Plants Reveals Distinct Drought Effects on the Phosphoproteome and Identifies
MASP1, an EGR-Interacting Phosphoprotein.
(A) Phosphopeptide abundance ratioversus gene expressionfor wild type(W.T.) stress versus control. Dark-red symbols indicate significant changes
in phosphopeptide abundance (P # 0.05 by one-sided t test and fo ld ch ange $1.5). All other phosphope ptide data are plotted using gray symbols.
Diagonalline indicatesidenticalchange in phosphopeptide abundance and gene expression. Data are fold changein phosphopeptide abundance(as
indicatedintheticklabels)plottedonalogarithmicscale.Fortheticklabels,ratioslessthanonewereinvertedandshownasnegativefoldchangefor
clarity of presentation.
(B) Phosphopeptide abundance versusgene expression for egr1-1 egr2-1 versus the wild type in control and stress treatments. Format ofdata presentation
is as describedfor (A).Dark red or green symbolsindicate phosphopeptides with P# 0.05 by one-sided t test and fold change $1.5(with additionof MASP1,
which has P = 0.07).
(C) BiFC interaction of MASP1 with EGRs and with itself in transient expression assays using intact Arabidopsis seedlings. Images of leaf epidermal
pavement cells of unstressed seedlings are shown. Essentially identical results were seen in stress-treated (21.2 MPa) seedlings. Bars = 20 mm.
(D) Representative image showing lack of interaction from BiFC analysis of EGRs with AT1G78320, a close homolog of MASP1, which lacks the MASP1
phosphorylation site (Supplemental Figures 3E and 3F and Supplemental Data Set 11) and SAY1 (Fig. 2B). Bars = 20 mm.
(E) Representative images showing the lack of BiFC fluorescence signal from the Clade A PP2C Highly ABA-Induced 1 (HAI1; AT5G59220) and MASP1 as
well as EGR1 and PYL10 (identical results were seen for EGR2 and EGR3). The HAI1-PYL10 interaction was used as a positive control. Bars = 20 mm.
(F) Localization of YFP-EGR1 and YFP-EGR2 in stable transgenic plants. Cells in the root maturation zone of unstressed 11-d-old seedlings are shown.
Similar localization was observed at low c
w
. Bars = 10 mm.
(G) Coimmunoprecipitation ofEGR1,EGR2,and MASP1. HAI1 andtheuninfiltratedAvr-PTO line (W.T.)wereusedasa negative controls. YFP-taggedEGR1,
EGR2, or HAI1 was transiently expressed along with FLAG-MASP1 and immunoprecipitation (I.P.) was performed with GFP-Trap resin to capture the YFP-
tagged phosphatase. Immunoblot (I.B.) of the total protein extract (input) and immunoprecipitated proteins was performed using FLAG antisera. The
experiment was repeated with similar results.
(H) Immunoblot using MASP1-specific antisera to detect endogenous MASP1 in wild-type seedlings under control or stress (21.2MPa,96h)
conditions shows inductionof MASP1 protein level at low c
w
(100 m g of total protein was loaded per lane). Replicateblots were probed with HSC70 as
aloadingcontrol.
(I) Quantitative RT-PCR analysis shows no increase of MASP1 expression in seedlings transferred to either 20.7 or 21.2 MPa low c
w
stress for 96 h
compared with unstressed plants. Data are means 6 SE (n = 6). Dashed red line indicates the level of expression in unstressed seedlings.
(J) Phos-tag gel analysis of MASP1 in stress-treated (21.2 MPa,96 h) seedlings. The identity of the phosphorylated MASP1 bandwas confirmed by analysis
of masp1-1 and by treating the wild-type sample with calf intestinal phosphatase (C.I.P.). The same samples were also run on SDS-PAGE gels to assess
Phospho-Signaling, Microtubules, and Growth 173
Downloaded from https://academic.oup.com/plcell/article/29/1/169/6099020 by guest on 20 August 2022