Review

Protein Language: Post-Translational

Modifications Talking to Each Other
Lam Dai Vu,1,2,3,4 Kris Gevaert,3,4,5,@ and Ive De Smet1,2,5,*,@

Post-translational modifications (PTMs) are at the heart of many cellular sig- Highlights
naling events. Apart from a single regulatory PTM, there are also PTMs that Recent studies highlight the impor-
function in orchestrated manners. Such PTM crosstalk usually serves as a fine- tance of crosstalk between different
PTMs in several plant signaling
tuning mechanism to adjust cellular responses to the slightest changes in the pathways.
environment. While PTM crosstalk has been studied in depth in various species;
Combinatory PTM codes are the result
in plants, this field is just emerging. In this review, we discuss recent studies on of distinct molecular mechanisms,
crosstalk between three of the most common protein PTMs in plant cells, being leading to different outcomes,
phosphorylation, ubiquitination, and sumoylation, and we highlight the diverse although mainly to ensure a tight reg-
ulation in response to environmental
underlying mechanisms as well as signaling outputs of such crosstalk. changes.

By its abundance, protein phosphory-

lation has a central role in protein
crosstalk, but emerging studies on ubi-
quitination and sumoylation also high-
PTM Crosstalk Adds More Complexity to Cellular Signaling light roles for these modifications in
Plants are constantly exposed to environmental changes and developmental cues that require plant protein crosstalk.

fast cellular sensing and response mechanisms. To decipher these mechanisms, the genome, Advances in mass spectrometry allow
transcriptome, and proteome have been explored in depth, each adding an additional level of the identification of PTM crosstalk,
complexity to signaling networks. In addition, PTMs of proteins are diverse, with 461 different enabling a more precise understand-
ing of plant signaling processes.
types of modified amino acid registered for eukaryotic proteins in UniProt; these PTMs affect
protein activity, stability, localization, and interactions [1]. This diversity, together with the
dynamic nature of many PTMs, is an advantage at the protein level, especially since PTMs
are often considered as switchboxes for cellular signaling.

Proteins originating from the same gene often harbor various PTMs, resulting in so-called
‘proteoforms’ (Box 1) [2], each of which might exhibit different activities or functions. In this way,
responses to even slight changes in the environment can be precisely fine-tuned, without
much, if any, evolutionary cost for the organism [3]. For example, in the model plant Arabidopsis
1
thaliana, over 1000 proteins are annotated with at least two different residues modified [1]. This Department of Plant Biotechnology
and Bioinformatics, Ghent University,
list is not exhaustive, since many common PTMs that often work interdependently, such as B-9052 Ghent, Belgium
modifications by ubiquitin and small ubiquitin-related modifier (SUMO), are not fully listed in the 2
VIB Center for Plant Systems
UniProt database. Indeed, proteomic studies in plants suggest the number of modified plant Biology, B-9052 Ghent, Belgium
3
Department of Biochemistry, Ghent
proteins is significantly higher [4–6]. University, B-9000 Ghent, Belgium
4
VIB Center for Medical
Given that phosphorylation is the most widely investigated PTM, its crosstalk with other PTM Biotechnology, B-9000 Ghent,
Belgium
has been prominently observed, especially in recent studies [7,8]. In this review, we provide 5
These authors contributed equally
insight into the most recent research into PTM crosstalk between phosphorylation and @
Twitters: @KrisGevaert_VIB,
ubiquitination or sumoylation in plant proteins. The latter involves two modifications that @IveDeSmet1978

use similar enzymatic machineries, but often serve different functions. In addition, we also
*Correspondence:
discuss a functional link between ubiquitin and SUMO modifications. Furthermore, we also ive.desmet@psb.vib-ugent.be
highlight advances in detecting PTM crosstalk. (I. De Smet).

1068 Trends in Plant Science, December 2018, Vol. 23, No. 12 https://doi.org/10.1016/j.tplants.2018.09.004
© 2018 Elsevier Ltd. All rights reserved.
 Box 1. Proteoforms as Functional and Regulatory Protein Units
Glossary
An organism expands the basic protein information encoded in its genome in various ways [86]. Each (chemical)
Autoubiquitination and
variation of a protein encoded by the same gene is termed a ‘proteoform’ [2], and proteoforms can exhibit diverse
autophosphorylation: self-
functions. In human cells, most proteoforms arising from alternative splicing share less than 50% of common protein–
modification of a modifying enzyme
protein interactions [87]. Furthermore, PTMs are widespread and a single protein can harbor multiple PTM sites. Hence,
can occur either independently of the
different combinations of PTMs on the same protein are a massive source for proteoform diversification [86]. Since
oligomerization status of the protein,
protein activity largely depends on PTMs, the combinatory PTM code has a significant influence on the function of each
where the protein can modify itself (in
proteoform. Therefore, specific proteoforms might be favored by the cell under certain cellular or environmental
cis), or in an oligomer, where each
conditions [86].
protomer can modify another
protomer (in trans).
Theoretically, the number of proteoforms of a single protein increases exponentially with the number of possible PTMs. E1-E2-E3 conjugation cascade:
For example, reversible modifications, such as phosphorylation, can result in 2n proteoforms, with n being the number of ubiquitin/SUMO is activated by an
modifiable amino acid residues [86]. However, the copy number of each proteoform can be limited by PTM crosstalk ATP-dependent formation of a
when the absence or presence of one or more PTM(s) is a prerequisite for other PTMs to occur. This may ensure that the thioester bond between a catalytic
copy number of nonfunctional proteoforms is kept low to avoid misregulation of proteins. cysteine on the E1 ligase and the
ubiquitin/SUMO protein, which is
transferred to an E2 ligase via trans-
thioesterification. Finally, the E2
ligase catalyzes the conjugation of
How Do PTMs Communicate?
ubiquitin/SUMO to the substrate via
The heart of the dynamics of many PTMs, including phosphorylation, ubiquitination, and complex formation with the substrate
sumoylation, is found in the reversibility of the modifications, carried out by ‘writer’ and ‘eraser’ and the E3 ligase core [44,78]. E1-
enzymes that specifically target a (modified) amino acid residue to add or remove a modification E2-E3-protein conjugation cascades
commonly form complex
[9]. Understandably, PTM crosstalk takes place when the catalytic machineries for different
polyubiquitin or polySUMO chains.
modifications modify each other to regulate their activities (Figures 1B–D and 2C) [10]. While both SUMO and ubiquitin have
Furthermore, crosstalk of two PTMs can occur on their common substrate (Figures 1A,E, various similar nonproteolytic
2A,B, and 3) [10]. Such PTM crosstalk is defined by the position of the modified amino acids, functions in cells, in contrast to
ubiquitin, no proteasomal
the type of modifications, and the effect of PTMs on one another [11]. Crosstalk can be
machineries have been yet shown to
orthosteric when different PTM types act at the same site or more distally via allosteric use the polySUMO chain directly as
regulation, in which a PTM alters the protein conformation, which, in its turn, will affect the a marker for degradation [60].
occurrence of a second PTM [3]. This crosstalk is either a positive or negative regulation, where E3 SUMO/ubiquitination ligases:
‘writer’ enzymes that interact with
one PTM will induce or abrogate the other, respectively [11,12]. substrates and are important for
defining substrate specificity; they
Crosstalk complexity can increase with the number of the PTMs that a protein harbors. can exhibit a high diversity in both
However, crosstalk studies are often limited to the interplay between two different PTMs. structure and function [79]. However,
only a few E3 SUMO ligases have
Nevertheless, the underlying mechanism can be diverse with various signaling outcomes. Here, been identified and characterized in
we explore these options and illustrate them with relevant examples. plants: the SAP and Miz (SIZ) family
(one gene in Arabidopsis), the HIGH
PLOIDY2/METHYL METHANE
Phosphorylation–Ubiquitination Crosstalk
SULFONATE SENSITIVITY21 (HPY2/
The large number of genes encoding proteins with predicted kinase, phosphatase (see Glossary) MMS21) family (one gene in
[13,14], or ubiquitin ligase [15–17] activity strongly hints at a frequent occurrence of phosphor- Arabidopsis), and the PROTEIN
ylation and ubiquitination during signal transduction. Furthermore, these enzymes can coregulate INHIBITOR OF STAT (PIAS) family
(two genes in Arabidopsis) [80].
the same signaling modules through crosstalk between phosphorylation and ubiquitination.
Protein kinases: ‘writer’ enzymes
catalyzing the phosphorylation of a
Phosphodegrons Facilitate Ubiquitination protein by transferring the
The so-called ‘phosphodegron’ is a clear example of crosstalk between phosphorylation and g-phosphate group of ATP to a
functional group on the side chain of
ubiquitination, where one or several phosphosites induce the ubiquitination and subsequent
an amino acid residue. In plants, the
degradation of a substrate in a cis-regulatory manner (Figure 1A) [18]. Here, a specific phos- targeted amino acids are mainly
phomotif on the substrate serves as the docking site for a ubiquitin ligase. While frequently serine and threonine, and, to a lesser
occurring in cellular signaling, only a few phosphodegrons have been described in detail for plant extent, also tyrosine and histidine
[81]. Protein kinases often contain
proteins [19,20]. For example, the Arabidopsis IRON-REGULATED TRANSPORTER 1 (IRT1) is other domains involved in ligand
controlled by the regulatory mechanism of a phosphodegron [20]. IRT1 is not only essential for binding or protein interactions, which
iron uptake into the cell, but also able to transport other heavy metal ions that become harmful ultimately can affect, or be affected
when present in high intracellular concentrations [21]. In the latter case, a short histidine-rich by, the kinase activity.

Trends in Plant Science, December 2018, Vol. 23, No. 12 1069

peptide sequence in the cytoplasmic loop of IRT1 binds to the heavy metal ions, which signals Phosphatases: ‘eraser’ enzymes
for the recruitment of the calcineurin-B like proteins (CBL)-interacting protein kinase 23 reversing phosphorylation (i.e.,
dephosphorylation) by hydrolyzing
(CIPK23) to phosphorylate several nearby serines and threonines [20]. The resulting phospho- the phosphoester bond and
degron provides a binding site for the RING E3 ubiquitin ligase IRT1-DEGRADATION removing the phosphate group as a
FACTOR 1 (IDF1). IDF1 catalyzes the polyubiquitination of two lysines within the IRT1 degron free ion [81].
and promotes IRT1 late endocytosis and degradation. As expected, the absence of CIPK23 Skp1-Cullin-F-Box (SCF) protein
scaffold: a multiprotein E3 ubiquitin
leads to the accumulation of intracellular IRT1 and hypersensitivity to heavy metal-rich ligase complex comprising a Cullin
conditions [20]. protein serving as a scaffold for
binding a small RING-box protein
(Rbx) and a S-phase kinase-
Phosphorylation can also disrupt an interaction between an E3 ligase and a substrate and, thus,
associated protein 1 (Skp1). In turn,
the term ‘phospho-inhibited degron’ was introduced (Figure 1A) [22]. Removing the phosphate the Rbx protein serves as the
group (dephosphorylation) using a phosphatase triggers (poly)ubiquitination of the substrate. docking site for the E2-ubiquitin
While no such system has been yet described in detail, including the actual modified sites, conjugate and Skp1 holds the
recruitment site for the F-box protein
dephosphorylation has been linked to ubiquitin-mediated proteasomal degradation, such as in
[82].
hormone signaling pathways like abscisic acid (ABA) [23] or gibberellic acid (GA) signaling [24]. SUMO-interacting motif (SIM):
contains a hydrophobic amino acid
Phosphorylation of E3 Ubiquitin Ligases Controls Interaction with Substrates consensus that can recognize and
directly bind SUMO noncovalently
Similar to the phosphodegron of a substrate, phosphorylation of an E3 ligase can promote its [83].
interaction with a substrate (Figure 1B). Arabidopsis 14-3-3 proteins bind to the E3 ligase Ubiquitin proteases/SUMO
ARABIDOPSIS TOXICOS LEVADURA 31 (ATL31) following recognition of a specific phos- proteases: ‘eraser’ enzymes that
phorylation pattern comprising four phosphosites on the C-terminal region of ATL31 [25]. This cleave the isopeptidyl bond between
the ubiquitin/SUMO and the lysine of
interaction is essential for ATL31-dependent ubiquitination and subsequent degradation of 14- the substrate protein to remove the
3-3 proteins [26]. At least one of the phosphosites, threonine 209 (T209), is phosphorylated by modification [84,85].
CIPK7, CIPK12, and CIPK14 under high carbon (C)/low nitrogen (N) conditions [27]. In addition,
CIPK-mediated phosphorylation of ATL31 is enhanced by Ca2+, suggesting a link between
calcium signaling and C/N-nutrient response [27].

Phosphorylation of an E3 ligase can also antagonize its interaction with its substrate (Figure 1B).
An important checkpoint for modulation of the GA response comprises the TARGET OF GNS
KINASE 2 (TAGK2), GA RECEPTOR RING E3 UBIQUITIN LIGASE (GARU), and the GA receptor
GA INSENSITIVE DWARF1 (GID1). Using the tyrosine kinase inhibitor genistein, it was shown
that tyrosine phosphorylation of GARU by TAGK2 inhibited its interaction with its substrate
GID1 [28]. Inhibition of TAGK2 activity by genistein led to increased ubiquitination of GID1, its
destabilization, and a reduced GID1-dependent degradation of DELLA proteins [28,29].

Phosphorylation of E3 Ubiquitin Ligases Controls Stability of E3 Ubiquitin Ligases

Apart from mediating E3 ubiquitin ligase–substrate interactions, phosphorylation can modulate
the E3 ubiquitin ligase activity by controlling its stability (Figure 1C) [30,31]. Indeed, autou-
biquitination is a common self-regulatory mechanism of E3 ubiquitin ligases. In Arabidopsis,
phosphorylation of PLANT U-BOX (PUB) E3 ligases can disrupt this [30]. Here, MITOGEN-
ACTIVATED PROTEIN KINASE 3 (MPK3), activated by pathogen-associated molecular pat-
terns (PAMPs), phosphorylates PUB22, a negative regulator of PAMP-induced signaling, at
T62 and T88 in its U-box domain, thereby impeding PUB22 dimerization (Figure 1C). Given that
PUB22 can self-regulate its degradation by trans-autoubiquitination via dimer formation,
phosphorylated PUB22 is stabilized. In this respect, a phosphomimetic mutant of PUB22
leads to higher PUB22 levels and increased susceptibility toward infection [30]. Hence, this
negative feedback loop controlled by a phosphorylation checkpoint on the E3 ligase is an
important modulator of the immune response mediated by MPK3. This crosstalk might work
independently from substrate recruitment because, in many U-Box type ligases, substrate
binding does not involve the U-Box domain [32].

1070 Trends in Plant Science, December 2018, Vol. 23, No. 12

(A) (B)
Phosphodegron Phosphoinhibited degron

(C) (D)

PM

StabilizaƟon Signaling
aƩenuaƟon

(E)
Dark Low light High light
Key:
PHYB
PHYB

PHYB
PHYB

Kinase

Phosphatase
PIF3
PHYB
PHYB

EBF
PIF3
PIF3 E3 ubiquiƟn ligase
EBF

UbiquiƟnaƟon/phosphorylaƟon substrate

Cullin
PHYB
PHYB

PIF3 Skp1 protein

E2 PIF3
EBF E2
ASK Rbx LRB Rbx Rbx protein
CUL1 CUL3
E2 E2 ubiquiƟn ligase
UbiquiƟn
PIF3
PIF3
Phosphogroup
PHYB
PHYB

Receptor

Ligand
Degraded protein
Photomorphogenesis Photomorphogenesis Photomorphogenesis Light signaling
aƩenuaƟon

Figure 1. Multifaceted Crosstalk between Phosphorylation and Ubiquitination. (A) Phosphorylation of the substrates affects the binding of E3 ubiquitin ligase.
(B) Phosphorylation of the E3 ubiquitin ligase affects the binding to the substrate. (C) Receptor-like kinase regulates its own turnover after activation upon ligand binding
via phosphorylation of a downstream E3 ubiquitin ligase in the signaling cascade. (D) Phosphorylation regulates autoubiquitination of E3 ubiquitin ligase. (E) Dose of
stimulus-dependent and phosphorylation-mediated ubiquitination and degradation of a PIF3 and its kinase, PHYB. Adapted, with permission, from [41] (E).

Trends in Plant Science, December 2018, Vol. 23, No. 12 1071

(A)
Key:
E3 ubiquiƟn ligase

E3 SUMO ligase

UbiquiƟnaƟon/SumoylaƟon substrate

UbiquiƟn

SUMO

Degraded protein

STUbL with SUMO-interacƟng moƟf

Lysine

(B)

(C)

SIZ1 COP1 HY5 Photomorphogenesis COP1

HY5

COP1 HY5
StabilizaƟon HY5

Dark Light

Figure 2. Multifaceted Crosstalk between Sumoylation and Ubiquitination. (A) Antagonistic crosstalk on the same lysine residue. (B) small ubiquitin-related
modifier (SUMO) as a marker for ubiquitination. (C) Mutual modifications of CONSTITUTIVE PHOTOMORPHOGENIC1 (COP1) and SIZ1, a E3 ubiquitin ligase and a E3
SUMO ligase, respectively, to regulate the stability of ELONGATED HYPOCOTYL5 (HY5), a positive regulator of photomorphogenesis.

The major ABA signaling regulator KEEP ON GOING (KEG) exhibits an interesting protein
structure that contains both an E3 ubiquitin ligase domain and a protein kinase domain [33].
Autoubiquitination mediates KEG degradation in the presence of ABA, and this process also
requires phosphorylation of KEG [34]. Interestingly, the inactive kinase domain mutant does not

1072 Trends in Plant Science, December 2018, Vol. 23, No. 12

 SnRK2.8 Cytoplasm
S55/59

NPR1 NPR1 NPR1 NPR1 NPR1

+ Salicylic acid + Salicylic acid
S589/T373 TF TF
Key: Defense gene
Kinase
Phosphatase
NPR1
E3 SUMO ligase
S11/15
E3 ubiquiƟn ligase
NegaƟve cis-regulatory element
PosiƟve cis-regulatory element
TranscripƟonal repressor
NPR1 NPR1
TranscripƟonal acƟvator
UbiquiƟn TF
Phosphogroup - +
SUMO Nucleus

Figure 3. Schematic Interpretation of Canonical and Interdependent Modifications of NONEXPRESSER OF PR GENES 1 (NPR1) upon Infection in
the Salicylic Acid Signaling Pathway. Here, NPR1 is sumoylated on multiple positions, but it is not ruled out that NPR1 is poly-sumoylated on the same residue(s).

show an altered ability of KEG to ubiquitinate its substrates, but does show a significantly
decrease in KEG degradation upon ABA treatment. This indicates that autophosphorylation
by the kinase domain affects KEG autoubiquitination activity. However, other kinases may also
have a role in this process [35].

Ubiquitination of (Receptor-Like) Kinases Attenuates Signaling Cascades

Many signaling pathways depend on input from membrane-associated receptor-like kinases
(RLKs) following binding of specific ligands. Crosstalk between phosphorylation and ubiquiti-
nation can take place as soon as RLKs are activated, leading to their internalization and
possible degradation (Figure 1D) [36]. Turnover of receptor proteins soon after their activation is
important to attenuate signaling.

Two prominent examples are the crosstalk-controlled turnover of the RLKs FLAGELLIN-
SENSITIVE2 (FLS2) and BRASSINOSTEROID INSENSITIVE 1 (BRI1). Ligand perception [fla-
gellin and brassinosteroids (BR), respectively] triggers complex formation of FLS2 or BRI1 with
the BRI1-ASSOCIATED RECEPTOR KINASE 1 1 (BAK1) to activate a cascade of downstream
kinases [37,38]. Importantly, the ligands also mediate the interaction of two E3 ubiquitin ligases,
PUB12 and PUB13, with FLS2 and BRI1, via phosphorylation of conserved residues in PUB12/
13. However, the mechanism leading to the recruitment of the E3 ligases differs for FLS2 and
BRI1. While phosphorylation of PUB12/13 solely depends on the activity of BAK1 in the case of
FLS2 [37], the BRI1 activity is essential for phosphorylation of and its interaction with PUB12/13
[38]. Thus, phosphorylation serves as a marker for RLK ubiquitination and turnover after their
activation to attenuate signaling pathways.

Trends in Plant Science, December 2018, Vol. 23, No. 12 1073

Such a signaling attenuation mechanism using phosphorylation–ubiquitination crosstalk also
occurs in cytoplasmic receptor-like or nonreceptor-like kinases [36,39]. For example, ABA
treatment promotes phosphorylation and activation of sucrose-nonfermented-related protein
kinases (SnRKs), such as SnRK2.3, and also targets SnRK2.3 to F-BOX PHLOEM PROTEIN 2-
B11 (PP2-B11) for ubiquitin-dependent proteasomal degradation to attenuate ABA signaling
and the abiotic stress response [40].

A Unique Case: Multilayered Phosphorylation-Mediated Control of Ubiquitination

Usually, substrate recruitment to an E3 ligase depends on a preceding (de)phosphorylation
event (see above). However, the phosphorylation-mediated ubiquitination and subsequent
degradation of PHYTOCHROME-INTERACTING FACTOR3 (PIF3), a negative regulator of
light signaling, exhibits a unique regulatory mechanism (Figure 1E) [41,42]. PIF3 binds
constitutively to the ETHYLENE INSENSITIVE 3 (EIN3)-BINDING F-BOX proteins (EBFs),
and this is independent of its light-induced phosphorylation by PHYTOCHROME B (PHYB)
[41]. However, the multisite phosphorylation of PIF3 induced by low light mediates the
recruitment of the substrate–F-box complex (PIF3-EBF) to the core Skp1-Cullin-F-Box
(SCF)proteinscaffold, possibly via a conformational change in the PIF3-EBF complex.
PIF3 is then ubiquitinated by the SCFEBF1/2 complex, resulting in its degradation to promote
photomorphogenesis [41]. In parallel, under high light intensity, PIF3 shows increased
affinity for the [Light-Response Bric-a-Brack/Tramtrack/Broad (LRB)]–CULLIN 3 (CUL3)-
type E3 ubiquitin ligase complex [41,42]. This catalyzes the assembly of a PIF3-PHYB
complex and the CUL3LRB E3 ligase machinery, and promotes PIF3 and PHYB ubiquiti-
nation and degradation. Hence, this second pathway, occurring in parallel with the EBF-
dependent pathway, while still allowing a light response due to the degradation of PIF3, also
acts as a negative feedback to modulate PHYB levels. While phosphorylation of PIF3 is
essential for both EBF- and LRB-mediated ubiquitination, it remains unclear how light levels
control the phosphostatus of PIF3 and whether different phosphoforms of PIF3 favor one or
both degradation pathways.

Furthermore, TOPP4, a type 1 protein phosphatase, interacts directly with PIF5, another
repressor of photomorphogenesis and a direct target of PHYB, and dephosphorylates PIF5
in an antagonistic manner to PHYB during photomorphogenesis to modulate the phosphor-
ylation and ubiquitination of PIF5 [43]. This may provide a second layer of fine-tuning control of
PHYB-dependent phosphorylation and ubiquitination of PIF proteins to attenuate light
signaling.

Phosphorylation–Sumoylation Crosstalk
In contrast to phosphorylation–ubiquitination crosstalk, coordinated functions of phosphory-
lation and sumoylation in plants are studied to a lesser extent, not only because of the less
frequent occurrence of sumoylation, but also due to the infancy of our understanding of plant
SUMO modifications [8].

Sumoylation is best characterized for nuclear proteins, especially related to abiotic and biotic
stresses [44–46]. The first phosphorylation–sumoylation crosstalk in plants was identified in the
basic helix-loop-helix (bHLH) transcription factor CESTA (CES), a component of the BR
response machinery [47,48]. Upon treatment with BR, CES is strongly sumoylated by SUMO1
and SUMO2 at lysine 72 (K72) and relocalized to nuclear bodies, an event essential for its
transcriptional activity [48]. Noticeably, CES phosphorylation at serine 55 (S75) and S77 by
CALCIUM-DEPENDENT KINASEs (CDPKs) impairs its sumoylation and nuclear body formation
[48]. However, it is unclear whether BR triggers an unknown phosphatase to antagonize

1074 Trends in Plant Science, December 2018, Vol. 23, No. 12

phosphorylated S75 and S77 or whether CDPK activity is merely required to maintain a basal
level of CES that can be sumoylated.

A more complex interplay between phosphorylation and sumoylation is found in NONEX-

PRESSER OF PR GENES 1 (NPR1), a salicylic acid receptor important for immune induction
(Figure 3) [49]. Sumoylation of NPR1 is blocked by phosphorylation of two serine residues, S55
and S59, to keep the protein quiescent under nonstressful conditions. Following exposure to
pathogens, salicylic acid mediates phosphorylation at S589 and T373 by SnrK2.8 to import
NPR1 from the cytoplasm into the nucleus [50]. Here, NPR1 is transiently sumoylated by
SUMO3, which may alter its affinity for transcriptional repressors of defense genes [49].
Furthermore, sumoylation is required for phosphorylation at two other serines, S11 and
S15, which, in turn, enhances further NPR1 sumoylation. This proteoform of NPR1 binds
to a transcriptional activator and triggers the expression of downstream pathogenesis-related
genes. Importantly, after activation, phosphorylation-dependent sumoylation eventually leads
to ubiquitin-mediated NPR1 degradation (Figure 3) [49]. This is an excellent example of a
system where crosstalk between multiple PTMs within a single protein triggers consecutive
signaling events, from activation to turnover of a protein, ensuring a dynamic but tightly
controlled regulation.

Apart from NPR1, the activity of other transcriptional regulators of plant defense is also
controlled by both sumoylation and phosphorylation [51]. Noticeably, those that are targeted
by MAP kinases upon immune activation are heavily modified by sumoylation. Examples
include WRKY DNA-BINDING PROTEINS, ETHYLENE-INSENSITIVE 3 (EIN3) and its close
homolog EIN3-LIKE1 (EIL1), as well as the histone deacetylase HDA19. While direct regulation
between sumoylation and phosphorylation is yet to be demonstrated for these proteins, these
observations appear to suggest that a complex interplay between both modifications com-
monly occurs during immune signaling [51].

Ubiquitination–Sumoylation Crosstalk
Sumoylation and ubiquitination can regulate each other in an antagonistic or a synergetic
manner [52]. However, while both modifications share similarities in their executive apparatus,
which involves a E1-E2-E3 conjugation cascade, their crosstalk exhibits distinct mecha-
nisms. SUMO and ubiquitin can compete for the same lysine for modification (Figure 2A). Upon
salt stress, accumulation of the DELLA protein REPRESSOR OF GA (RGA) is promoted by
sumoylation of the lysine 65 residue (K65) in a GA-independent manner [53]. Interestingly,
mutating K65 to arginine (R) also has a stabilizing effect on RGA. This indicates that K65 in RGA
is targeted by both sumoylation and ubiquitination, and that sumoylation shields RGA from GA-
mediated ubiquitination and degradation. In another example, SUMO and ubiquitin compete for
the same quartet of lysines in the RECEPTOR ACTIVATED C KINASE 1B (RACK1B) [54]. In fact,
sumoylation of these four lysines increases RACK1B resistance to ubiquitin-mediated degra-
dation in response to ABA. [54]. Interestingly, in nonplant organisms, crosstalk with a third
modification, such as phosphorylation, steers which modification will occur on the same lysine
residue [55].

Similar to phosphorylation, in several systems, sumoylation also serves as a marker for

subsequent ubiquitination events, albeit with distinct mechanisms (Figure 2B). This crosstalk
is best demonstrated by the activity of SUMO-interacting motif (SIM)-containing SUMO-
targeted ubiquitin E3 ligases (STUbLs) [56]. STUbLs can recognize and bind a sumoylated
substrate and mediate the ubiquitination of unoccupied lysine residues in a manner similar to a
phosphodegron. Furthermore, hybrid conjugates of SUMO and ubiquitin were detected in

Trends in Plant Science, December 2018, Vol. 23, No. 12 1075

proteome studies, confirming that SUMO can be modified by ubiquitin [4,57–59]. Indeed, upon
recognition and binding to poly-SUMO chains, many STUbLs are able to add (poly-)ubiquitins
on these poly-SUMO chains (Figure 2B) [60]. In this way, the ubiquitin-SUMO hybrid side-chain
also directs the substrate to proteasomal degradation [56]. In Arabidopsis, six RING-type
STUbL homologs were identified via yeast two-hybrid using SUMO1, 2, and 3 as baits [61]. Of
note, STUbL4 interacts with, and potentially targets, CYCLING DOF FACTOR2 (CDF2), a
transcriptional repressor of CONSTANS (CO), to control floral transition [61]. CDF2 was shown
to be modified by SUMO1 and overexpression of STUbL4 reduced CDF2 protein levels [61,62].
Overall, SUMO-dependent degradation is increasingly observed in plants [49,63,64], suggest-
ing an important role of plant STUbLs in signaling events.

Sumoylation and ubiquitination machineries can modify and regulate the activity of the other
(Figure 2C) [65]. A positive regulation integrated with a negative feedback loop between
sumoylation and ubiquitination was observed in the mutual modification of the ubiquitin E3
ligase CONSTITUTIVE PHOTOMORPHOGENIC1 (COP1), a repressor of photomorphogene-
sis, and the E3SUMOligase SIZ1. In the dark, sumoylation of COP1 by SIZ1 enhances COP1
activity, which increases the ubiquitination of its targets, such as ELONGATED HYPOCOTYL5
(HY5), to negatively regulate photomorphogenesis [65]. However, COP1 in turn induces SIZ1
ubiquitination and proteasomal degradation and, upon prolonged light exposure, SUMO-
conjugated COP1 levels are decreased. Hence, the COP1-SIZ1 interplay serves as a tight
regulation to ensure proper photomorphogenic development [65].

Triple the Fun: Confluence of Phosphorylation, Ubiquitination, and

Sumoylation
Ultimately, all three PTMs together (phosphorylation, ubiquitination, and sumoylation) can
control a single regulatory event. In nonplant organisms, a prominent example is the coordi-
nated occurrence of these three PTMs during genotoxic stress to sequentially facilitate the
nuclear import, activation, and nuclear export of the regulatory subunit NEMO of the IkB
kinase (IKK) [66]. Such PTM regulatory codes are also found in plants. As mentioned above, the
integration of phosphorylation, sumoylation, and ubiquitination affects NPR1 activity and
stability (Figure 3) [49]. In addition, the PHYB-PIF module is controlled by those three PTMs.
The phosphorylation-mediated ubiquitination of PIF5 is hampered by sumoylation occurring at
the C terminus of PHYB, which may weaken the PHYB–PIF5 interaction [67]. Sumoylation of
PHYB, while being enhanced by red and white light, may also serve as a modulator of light
signaling, because mutation of the SUMO target K996 in PHYB leads to a red light-hypersen-
sitive phenotype.

Another example is the Arabidopsis protein kinase SnRK1, which functions as a modulator of
cellular C status and links it to various stress responses [63]. Activation of SnRK1 requires
phosphorylation of T175 in its T-loop region. However, this also leads to its SIZ1-dependent
sumoylation, which eventually dampens SnRK1 signaling by targeting the protein for ubiquiti-
nation and proteasomal degradation [63]. Nitrate reductases (NRs) are not only targets of
SnRK1, but also regulated by other PTMs [68]. On the one hand, phosphorylation deactivates
NRs by promoting binding with 14-3-3 proteins [69], while on the other hand, sumoylation of
NRs by SIZ1 increases their activity [70]. Furthermore, new evidence suggests a regulation of
NRs by ubiquitin-dependent proteasomal degradation [68], which is in contrast to an earlier
observation showing that NRs were not modified by ubiquitin [71]. However, the overall
orchestration and possible crosstalk of these PTMs in response to nitrogen status remain
elusive.

1076 Trends in Plant Science, December 2018, Vol. 23, No. 12

 Box 2. Mass Spectrometry-Based Proteome-Wide Analysis of PTM Crosstalk
In-depth PTM studies require highly confident identification of modified sites, usually using mass spectrometry (MS)-based methods. Most PTMs exist in the cells in
substoichiometric quantities. Therefore, an enrichment step for modified peptides is essential before MS analysis. For phosphorylation, affinity enrichment using
metal oxides is commonly used [88], whereas immunoaffinity is commonly applied for ubiquitinated and sumoylated proteins or peptides [89–92]. Unlike ubiquitin,
SUMO does not leave a di-glycine tag on peptides following tryptic cleavage. Thus, fully functional modified SUMO proteins that leave a small tryptic remnant that is
distinct from that of ubiquitin can be conveniently detected in tandem MS analysis when expressed in transgenic plants (Figure IA) [6,58].

Furthermore, attempts to globally co-analyze different PTM types have been initiated. Large-scale PTM crosstalk analyses usually require serial enrichment (Figure IB)
[93]. Here, modified peptides are enriched for the first PTM, followed by enrichment for the second PTM, resulting in the enrichment of co-modified peptides. Swaney
et al. applied two complemental strategies for simultaneous analysis of the phosphoproteome and the ubiquitinome that includes a serial enrichment of
phosphorylated and ubiquitinated peptides [94]. By applying a proteasome inhibitor to prevent degradation of ubiquitinated proteins, several phosphosites within
predicted phosphodegron consensus sequences were observed, especially those that were found co-occurring with a ubiquitination site in close proximity. A similar
approach was applied for sumoylation and ubiquitination profiling in human cells overexpressing C-terminally modified SUMOs [59,95]. The ubiquitin-SUMO co-
modification was highly enriched in nuclear proteins, suggesting an important role for their crosstalk in gene regulation and the DNA damage response. Additionally,
an ultra-deep analysis of the human sumoylome revealed a large number of peptides co-modified by both SUMO and phosphorylation in close proximity, especially in
the disordered region of proteins [96].

Peptide inference methods, such as bottom-up proteomics, have been successfully applied to identify individual PTM types, but are not suitable for mapping
proteoforms because some share the same peptide sequences. Top-down proteomics has made steady progress in recent years and is more suitable for this
purpose because intact proteins are analyzed instead of peptide surrogates [97]. Alternatively, middle-down proteomics, which analyzes larger peptide species (50
and more amino acids), is a good compromise, in which larger proteoform fragments can be analyzed [98]. In this way, coexisting modifications that are not in close
proximity in the protein sequence may also be detected [98,99]. To facilitate PTM crosstalk research in plants, a central resource comprehensively visualizing
MS-identified PTMs on plant proteins (http://www.psb.ugent.be/PlantPTMViewer) was recently presented [100].

(A)

AtSUMO1 N-Term- -(target protein)

AtSUMO1 N-Term- 6xHis-Tag -(target protein)

(modified)
(B)
IP against
IP against (GG)K- (QTGG)K-

Ni-NTA Trypsin digesƟon

Phospho- IP against
SUMO UbiquiƟn Phosphosites 6xHis enrichment (QTGG)K-

Figure I. Strategy to Detect Peptides Co-Modified by Different PTMs. (A) Modified Arabidopsis SUMO1-H89R with a 6xHis-tag for pre-enrichment for
sumoylated proteins. Tryptic cleavage site is marked with a scissor symbol. (B) A schematic approach for the enrichment of peptides co-modified by small ubiquitin-
related modifier (SUMO) and ubiquitin or phosphorylation. Abbreviation: IP, immunoprecipitation.

Comprehensively Detecting PTM Crosstalk: A Challenge

Although individual cases have been described, comprehensive data on PTM crosstalk in
plants are still lacking. This is mostly due to difficulties in identifying PTM crosstalk. Indeed, the
coexistence of multiple proteoforms complicates the identification of the functional ones in a
specific biological context. Moreover, because PTM crosstalk is usually highly dynamic, a
refined spatial and temporal experimental resolution is required. Nevertheless, advances in
PTM detection, especially in mass spectrometry-based techniques, allow for large-scale

Trends in Plant Science, December 2018, Vol. 23, No. 12 1077

identification of phosphorylation–ubiquitination–sumoylation crosstalk (Box 2). In addition, Outstanding Questions
indirect approaches have also been applied for this purpose. Recently, profiling the phos- How distinct are plant crosstalk motifs
phoproteome of mutants devoid of SUMO E3 and E3-like ligases SIZ1 and PROTEIN compared with other organisms?

INHIBITOR OF ACTIVATED STAT-LIKE 1 (PIAL1) and PIAL2 resulted in the identification

Crosstalk studies in plants largely
of phosphopeptides that are potentially affected by sumoylation in Arabidopsis [72]. Indeed, a focus on the interplay of the ‘writer’
large portion of phosphoproteins exhibits sumoylation attachment consensuses. In addition, proteins. How many crosstalks are
using specific kinase inhibitors in combination with profiling of protein ubiquitination and/or there that are mainly controlled by
‘eraser’ proteins (phosphatases, ubiq-
sumoylation is also an interesting way to determine PTM crosstalk, as has been shown in the
uitin proteases, and SUMO
study of the phosphorylation–ubiquitination crosstalk underlying GARU activity [28]. Further- proteases)?
more, more direct approaches that are able to capture multiple PTMs simultaneously have
been successfully applied for nonplant systems, but remain to be implemented in plants How expansive is ubiquitination–phos-
(Box 2). phorylation crosstalk among plant pro-
tein kinases after their activation as
readout for signaling modulation?
Concluding Remarks
Cellular functions are not limited by genetic codes, but are vastly expanded by modifications of Since phosphorylation of SUMO E3
biomolecules, including proteins. While PTMs are increasingly investigated in plants, their ligases has been identified, is there
any mutual regulation between kinases
interplay has thus far been largely ignored. Such crosstalk occurs in a range of biological
and E3 SUMO ligases in
processes. The coordination of multiple PTMs ensures quick and tight regulation of signal phosphorylation–sumoylation cross-
transduction and, in many cases, mediates both activation and attenuation of the same talk similar to other PTM crosstalks?
pathway. In addition, crosstalk mechanisms of different PTMs can be highly complex, with
mutual modification of the PTM machineries as well as their substrates. While phosphorylation Apart from SIMs that recognize poly-
SUMO chains, do STUbLs exhibit any
of plant E3 SUMO ligases has been detected [73,74], its role is yet to be determined.
mechanism for substrate specificity?

Mass spectrometry-based determination of PTM crosstalk is promising, but the dynamic How can one resolve the temporal and
nature of PTMs is a major obstacle to overcome for systematic characterization of PTM spatial resolution to increase PTM
crosstalk in plants (see Outstanding Questions). Finally, algorithms that enable the prediction crosstalk identification?

of PTM crosstalk use validated crosstalk pairs to analyze PTM positions and coevolution of
Is large-scale crosstalk profiling and
those [75–77]. While not yet implemented in plants, these tools can be used as a proxy to prediction that has been successfully
identify well-conserved crosstalk modules in plants. However, these detection techniques and implemented in other systems also
prediction algorithms are suitable for the identification of positive crosstalk, such as phospho- adaptable for plants?
degrons, where both PTMs coexists, but not for the identification of negative crosstalk, such as
phosphoinhibited degrons. How can we tackle large-scale detec-
tion and prediction methods for nega-
tive crosstalk, such as phospho-
Acknowledgments inhibited degron?
L.D.V. is the recipient of a VIB International PhD Fellowship.

References
1. Bateman, A. et al. (2017) UniProt: the universal protein knowl-
edgebase. Nucleic Acids Res. 45, D158–D169
2. Smith, L.M. and Kelleher, N.L. (2013) Proteoform: a single term
describing protein complexity. Nat. Methods 10, 186–187
3. Nussinov, R. et al. (2012) Allosteric post-translational modifica-
tion codes. Trends Biochem. Sci. 37, 447–455
4. Walton, A. et al. (2016) It’s time for some ‘site’-seeing: novel
tools to monitor the ubiquitin landscape in Arabidopsis thaliana.
Plant Cell 28, 6–16
5. Vlastaridis, P. et al. (2017) Estimating the total number of phos-
phoproteins and phosphorylation sites in eukaryotic proteomes.
Gigascience 6, 1–11
6. Rytz, T.C. et al. (2018) SUMOylome profiling reveals a diverse
array of nuclear targets modified by the SUMO ligase SIZ1
during heat stress. Plant Cell 30, 1077–1099
7. He, Z. et al. (2017) Sumoylation, phosphorylation, and acetyla-
tion fine-tune the turnover of plant immunity components medi-
ated by ubiquitination. Front. Plant Sci. 8, 1–7
8. Tomanov, K. et al. (2018) Sumoylation and phosphorylation:
hidden and overt links. J. Exp. Bot. 69, 4583–4590
9. Prabakaran, S. et al. (2012) Post-translational modification: nature’s
escape from genetic imprisonment and the basis for dynamic infor-
mation encoding. Wiley Interdiscip. Rev. Syst. Biol. Med. 4, 565–583
10. Hunter, T. (2007) The age of crosstalk: phosphorylation, ubiq-
uitination, and beyond. Mol. Cell 28, 730–738
11. Csizmok, V. and Forman-Kay, J.D. (2018) Complex regulatory
mechanisms mediated by the interplay of multiple post-transla-
tional modifications. Curr. Opin. Struct. Biol. 48, 58–67
12. Hunter, T. (2007) The age of crosstalk: phosphorylation, ubiq-
uitination, and beyond. Mol. Cell 28, 730–738
13. Kerk, D. et al. (2002) The complement of protein phosphatase
catalytic subunits encoded in the genome of Arabidopsis. Plant
Physiol. 129, 908–925
14. Zulawski, M. et al. (2014) The Arabidopsis kinome: phylogeny
and evolutionary insights into functional diversification. BMC
Genomics 15, 548

1078 Trends in Plant Science, December 2018, Vol. 23, No. 12

15. Vierstra, R.D. (2009) The ubiquitin-26S proteasome system at
the nexus of plant biology. Nat. Rev. Mol. Cell Biol. 10, 385–397
16. Hua, Z. and Vierstra, R.D. (2011) The cullin-RING ubiquitin-
protein ligases. Annu. Rev. Plant Biol. 62, 299–334
17. Isono, E. and Nagel, M.-K. (2014) Deubiquitylating enzymes and
their emerging role in plant biology. Front. Plant Sci. 5, 56
18. Filip9
cík, P. et al. (2017) When worlds collide—mechanisms at the
interface between phosphorylation and ubiquitination. J. Mol.
Biol. 429, 1097–1113
19. Hardtke, C.S. et al. (2000) HY5 stability and activity in arabi-
dopsis is regulated by phosphorylation in its COP1 binding
domain. EMBO J. 19, 4997–5006
20. Dubeaux, G. et al. (2018) Metal sensing by the IRT1 transporter-
receptor orchestrates its own degradation and plant metal nutri-
tion. Mol. Cell 69, 953–964.e5
21. Barberon, M. et al. (2011) Monoubiquitin-dependent endocyto-
sis of the IRON-REGULATED TRANSPORTER 1 (IRT1) trans-
porter controls iron uptake in plants. Proc. Natl. Acad. Sci. 108,
E450–E458
22. Holt, L.J. (2012) Regulatory modules: coupling protein stability
to phopshoregulation during cell division. FEBS Lett. 586, 2773–
2777
23. Dai, M. et al. (2013) The PP6 phosphatase regulates ABI5
phosphorylation and abscisic acid signaling in Arabidopsis.
Plant Cell 25, 517–534
24. Qin, Q. et al. (2014) Arabidopsis DELLA protein degradation is
controlled by a type-one protein phosphatase, TOPP4. PLoS
Genet. 10, e1004464
25. Yasuda, S. et al. (2014) Phosphorylation of Arabidopsis ubiquitin
ligase ATL31 is critical for plant carbon/nitrogen nutrient balance
response and controls the stability of 14-3-3 proteins. J. Biol.
Chem. 289, 15179–15193
26. Sato, T. et al. (2011) Identification of 14-3-3 proteins as a target
of ATL31 ubiquitin ligase, a regulator of the C/N response in
Arabidopsis. Plant J. 68, 137–146
27. Yasuda, S. et al. (2017) Arabidopsis CBL-interacting protein
kinases regulate carbon/nitrogen-nutrient response by phos-
phorylating ubiquitin ligase ATL31. Mol. Plant 10, 605–618
28. Nemoto, K. et al. (2017) Tyrosine phosphorylation of the GARU
E3 ubiquitin ligase promotes gibberellin signalling by preventing
GID1 degradation. Nat. Commun. 8, 1–7
29. Hussain, A. et al. (2007) Identification of conserved tyrosine
residues important for gibberellin sensitivity of Arabidopsis
RGL2 protein. Planta 226, 475–483
30. Furlan, G. et al. (2017) Changes in PUB22 ubiquitination modes
triggered by MITOGEN-ACTIVATED PROTEIN KINASE3
dampen the immune response. Plant Cell 29, 726–745
31. Wang, J. et al. (2015) The E3 ligase OsPUB15 interacts with the
receptor-like kinase PID2 and regulates plant cell death and
innate immunity. BMC Plant Biol. 15, 49
32. Metzger, M.B. et al. (2014) RING-type E3 ligases: master manip-
ulators of E2 ubiquitin-conjugating enzymes and ubiquitination.
Biochim. Biophys. Acta 1843, 47–60
33. Stone, S.L. et al. (2006) KEEP ON GOING, a RING E3 ligase
essential for Arabidopsis growth and development, is involved in
abscisic acid signaling. Plant Cell 18, 3415–3428
34. Liu, H. and Stone, S.L. (2010) Abscisic acid increases Arabi-
dopsis ABI5 transcription factor levels by promoting KEG E3
ligase self-ubiquitination and proteasomal degradation. Plant
Cell 22, 2630–2641
35. Lyzenga, W.J. et al. (2017) The kinase activity of calcineurin B-
like interacting protein kinase 26 (CIPK26) influences its own
stability and that of the ABA-regulated ubiquitin ligase, Keep on
Going (KEG). Front. Plant Sci. 8, 502
36. Zhimin, L. and Tony, H. (2009) Degradation of activated protein
kinases by ubiquitination. Annu. Rev. Biochem. 78, 435–475
37. Lu, D. et al. (2011) Direct ubiquitination of pattern recognition
receptor FLS2 attenuates plant innate immunity. Science 332,
1439–1442
38. Zhou, J. et al. (2018) Regulation of Arabidopsis brassinosteroid
receptor BRI1 endocytosis and degradation by plant U-box
PUB12/PUB13-mediated ubiquitination. Proc. Natl. Acad.
Sci. 115, E1906–E1915
39. Monaghan, J. (2018) Conserved degradation of orthologous
RLCKs regulates immune homeostasis. Trends Plant Sci. 23,
554–557
40. Cheng, C. et al. (2017) SCFAtPP2-B11 modulates ABA signal-
ing by facilitating SnRK2.3 degradation in Arabidopsis thaliana.
PLoS Genet. 13, e1006947
41. Dong, J. et al. (2017) Light-dependent degradation of PIF3 by
SCF EBF1/2 promotes a photomorphogenic response in Ara-
bidopsis. Curr. Biol. 27, 2420–2430
42. Ni, W. et al. (2014) A mutually assured destruction mechanism
attenuates light signaling in Arabidopsis. Science 344, 1160–
1164
43. Yue, J. et al. (2016) TOPP4 regulates the stability of PHYTO-
CHROME INTERACTING FACTOR5 during photomorphogen-
esis in Arabidopsis. Plant Physiol. 170, 1381–1397
44. Saracco, S.A. et al. (2007) Genetic analysis of SUMOylation in
Arabidopsis: conjugation of SUMO1 and SUMO2 to nuclear
proteins is essential. Plant Physiol. 145, 119–134
45. Miller, M.J. et al. (2010) Proteomic analyses identify a diverse
array of nuclear processes affected by small ubiquitin-like modi-
fier conjugation in Arabidopsis. Proc. Natl. Acad. Sci. 107,
16512–16517
46. Augustine, R.C. et al. (2016) Defining the SUMO system in maize:
SUMOylation is up-regulated during endosperm development and
rapidly induced by stress. Plant Physiol. 171, 2191–2210
47. Poppenberger, B. et al. (2011) CESTA, a positive regulator of
brassinosteroid biosynthesis. EMBO J. 30, 1149–1161
48. Khan, M. et al. (2014) Interplay between phosphorylation and
SUMOylation events determines CESTA protein fate in brassi-
nosteroid signalling. Nat. Commun. 5, 4687
49. Saleh, A. et al. (2015) Posttranslational modifications of the
master transcriptional regulator NPR1 enable dynamic but
tight control of plant immune responses. Cell Host Microbe
18, 169–182
50. Lee, H.-J. et al. (2015) Systemic immunity requires
SnRK2.8-mediated nuclear import of NPR1 in Arabidopsis.
Plant Cell 27, 3425–3438
51. van den Burg, H.A. and Takken, F.L.W. (2010) SUMO-, MAPK-
and resistance protein-signaling converge at transcription com-
plexes that regulate plant innate immunity. Plant Signal. Behav.
5, 1597–1601
52. Geoffroy, M.C. and Hay, R.T. (2009) An additional role for SUMO
in ubiquitin-mediated proteolysis. Nat. Rev. Mol. Cell Biol. 10,
564–568
53. Conti, L. et al. (2014) Small ubiquitin-like modifier protein SUMO
enables plants to control growth independently of the phyto-
hormone gibberellin. Dev. Cell 28, 102–110
54. Guo, R. and Sun, W. (2017) Sumoylation stabilizes RACK1B and
enhance its interaction with RAP2.6 in the abscisic acid
response. Sci. Rep. 7, 1–13
55. Ulrich, H.D. (2005) Mutual interactions between the SUMO and
ubiquitin systems: a plea of no contest. Trends Cell Biol. 15, 525–532
56. Praefcke, G.J.K. et al. (2012) SUMO playing tag with ubiquitin.
Trends Biochem. Sci. 37, 23–31
57. Elrouby, N. and Coupland, G. (2010) Proteome-wide screens for
small ubiquitin-like modifier (SUMO) substrates identify Arabi-
dopsis proteins implicated in diverse biological processes. Proc.
Natl. Acad. Sci. 107, 17415–17420
58. Miller, M.J. et al. (2010) Proteomic analyses identify a diverse
array of nuclear processes affected by small ubiquitin-like modi-
fier conjugation in Arabidopsis. Proc. Natl. Acad. Sci. 107,
16512–16517
59. McManus, F.P. et al. (2017) Identification of cross talk between
SUMOylation and ubiquitylation using a sequential peptide
immunopurification approach. Nat. Protoc. 12, 2342–2358
Trends in Plant Science, December 2018, Vol. 23, No. 12 1079
60. Perry, J.J.P. et al. (2008) A SIM-ultaneous role for SUMO and
ubiquitin. Trends Biochem. Sci. 33, 201–208
61. Elrouby, N. et al. (2013) Identification of Arabidopsis SUMO-
interacting proteins that regulate chromatin activity and devel-
opmental transitions. Proc. Natl. Acad. Sci. U. S. A. 110,
19956–19961
62. Budhiraja, R. et al. (2009) Substrates related to chromatin and to
RNA-dependent processes are modified by Arabidopsis SUMO
isoforms that differ in a conserved residue with influence on
desumoylation. Plant Physiol. 149, 1529–1540
63. Crozet, P. et al. (2016) SUMOylation represses SnRK1 signaling
in Arabidopsis. Plant J. 85, 120–133
64. Castro, P.H. et al. (2015) SIZ1-dependent post-translational
modification by SUMO modulates sugar signaling and metabo-
lism in Arabidopsis thaliana. Plant Cell Physiol. 56, 2297–2311
65. Lin, X.L. et al. (2016) An Arabidopsis SUMO E3 ligase, SIZ1,
negatively regulates photomorphogenesis by promoting COP1
activity. PLoS Genet. 12, 1–21
66. Huang, T.T. et al. (2003) Sequential modification of NEMO/IKKg
by SUMO-1 and ubiquitin mediates NF-kB activation by geno-
toxic stress. Cell 115, 565–576
67. Sadanandom, A. et al. (2015) SUMOylation of phytochrome-B
negatively regulates light-induced signaling in Arabidopsis thali-
ana. Proc. Natl. Acad. Sci. 112, 11108–11113
68. Kim, J.Y. et al. (2018) Nitrate reductases are relocalized to the
nucleus by atSIZ1 and their levels are negatively regulated by
COP1 and ammonium. Int. J. Mol. Sci. 19, 1–15
69. Lambeck, I.C. et al. (2012) Molecular mechanism of 14-3-3
protein-mediated inhibition of plant nitrate reductase. J. Biol.
Chem. 287, 4562–4571
70. Park, B.S. et al. (2011) Arabidopsis nitrate reductase activity is
stimulated by the E3 SUMO ligase AtSIZ1. Nat. Commun. 2,
400–410
71. Weiner, H. and Kaiser, W.M. (1999) 14-3-3 proteins control
proteolysis of nitrate reductase in spinach leaves. FEBS Lett.
455, 75–78
72. Nukarinen, E. et al. (2017) Protein sumoylation and phosphor-
ylation intersect in Arabidopsis signaling. Plant J. 91, 505–517
73. Heazlewood, J.L. et al. (2008) PhosPhAt: a database of phos-
phorylation sites in Arabidopsis thaliana and a plant-specific
phosphorylation site predictor. Nucleic Acids Res. 36,
D1015–D1021
74. Durek, P. et al. (2009) PhosPhAt: The Arabidopsis thaliana
phosphorylation site database. An update. Nucleic Acids Res.
38, 828–834
75. Peng, M. et al. (2014) Identification of enriched PTM crosstalk
motifs from large-scale experimental data sets. J. Proteome
Res. 13, 249–259
76. Minguez, P. et al. (2012) PTMcode: a database of known and
predicted functional associations between post-translational
modifications in proteins. Nucleic Acids Res. 41, D306–D311
77. Huang, Y. et al. (2015) Systematic characterization and predic-
tion of post-translational modification cross-talk. Mol. Cell. Pro-
teomics 14, 761–770
78. Wang, F. and Deng, X.W. (2011) Plant ubiquitin-proteasome path-
way and its role in gibberellin signaling. Cell Res. 21, 1286–1294
79. Chen, L. and Hellmann, H. (2013) Plant E3 ligases: flexible
enzymes in a sessile world. Mol. Plant 6, 1388–1404
1080 Trends in Plant Science, December 2018, Vol. 23, No. 12
80. Novatchkova, M. et al. (2012) Update on sumoylation: defining
core components of the plant SUMO conjugation system by
phylogenetic comparison. New Phytol. 195, 23–31
81. Dissmeyer, N. and Schnittger, A. (2011) The age of protein
kinases. Methods Mol. Biol. 779, 7–52
82. Willems, A.R. et al. (2004) A hitchhiker’s guide to the cullin
ubiquitin ligases: SCF and its kin. Biochim. Biophys. Acta
1695, 133–170
83. Kerscher, O. (2007) SUMO junction—what’s your function?
New insights through SUMO-interacting motifs. EMBO Rep.
8, 550–555
84. Zhou, H. et al. (2017) UBIQUITIN-SPECIFIC PROTEASES func-
tion in plant development and stress responses. Plant Mol. Biol.
94, 565–576
85. Yates, G. et al. (2016) SUMO proteases: uncovering the roles of
deSUMOylation in plants. J. Exp. Bot. 67, 2541–2548
86. Aebersold, R. et al. (2018) How many human proteoforms are
there? Nat. Chem. Biol. 14, 206–214
87. Yang, X. et al. (2016) Widespread expansion of protein interac-
tion capabilities by alternative splicing. Cell 164, 805–817
88. Nilsson, C.L. (2012) Advances in quantitative phosphoproteo-
mics. Anal. Chem. 84, 735–746
89. Walton, A. et al. (2016) It’s time for some ‘site’-seeing: novel
tools to monitor the ubiquitin landscape in Arabidopsis thaliana.
Plant Cell 28, 6–16
90. Udeshi, N.D. et al. (2013) Large-scale identification of ubiquiti-
nation sites by mass spectrometry. Nat. Protoc. 8, 1950–1960
91. Hendriks, I.A. and Vertegaal, A.C.O. (2016) A high-yield double-
purification proteomics strategy for the identification of SUMO
sites. Nat. Protoc. 11, 1630–1649
92. Lamoliatte, F. et al. (2014) Large-scale analysis of lysine
SUMOylation by SUMO remnant immunoaffinity profiling. Nat.
Commun. 5, 5409
93. Mertins, P. et al. (2013) Integrated proteomic analysis of post-
translational modifications by serial enrichment. Nat. Methods
10, 634–637
94. Swaney, D.L. et al. (2013) Global analysis of phosphorylation
and ubiquitylation cross-talk in protein degradation. Nat. Meth-
ods 10, 676–682
95. Lamoliatte, F. et al. (2017) Uncovering the SUMOylation and
ubiquitylation crosstalk in human cells using sequential peptide
immunopurification. Nat. Commun. 8, 14109
96. Hendriks, I.A. et al. (2017) Site-specific mapping of the human
SUMO proteome reveals co-modification with phosphorylation.
Nat. Struct. Mol. Biol. 24, 325–336
97. Toby, T.K. et al. (2016) Progress in top-down proteomics and
the analysis of proteoforms. Annu. Rev. Anal. Chem. 9, 499–519
98. Schräder, C.U. et al. (2018) Simultaneous proteoform analysis of
Histones H3 and H4 with a simplified middle-down proteomics
method. Anal. Chem. 90, 3083–3090
99. Tvardovskiy, A. et al. (2015) Top-down and middle-down pro-
tein analysis reveals that intact and clipped human histones
differ in post-translational modification patterns. Mol. Cell. Pro-
teomics 14, 3142–3153
100. Willems, P. et al. (2018) The Plant PTM Viewer, a central
resource exploring plant protein modifications. From site-seeing
to protein function. bioRxiv Published online September 13,
2018. http://dx.doi.org/10.1101/415802