E3 Ubiquitin Ligases: Ubiquitous Actors in Plant Development and Abiotic Stress Responses

Abstract

Understanding the precise regulatory mechanisms of plant development and stress responses at the post-translational level is currently a topic of intensive research. Protein ubiquitination, including the sequential performances of ubiquitin-activating (E1), ubiquitin-conjugating (E2) and ubiquitin ligase (E3) enzymes, is a refined post-translational modification ubiquitous in all eukaryotes. Plants are an integral part of our ecosystem and, as sessile organisms, the ability to perceive internal and external signals and to adapt well to various environmental challenges is crucial for their survival. Over recent decades, extensive studies have demonstrated that protein ubiquitination plays key roles in multiple plant developmental stages (e.g. seed dormancy and germination, root growth, flowering time control, self-incompatibility and chloroplast development) and several abiotic stress responses (e.g. drought and high salinity), by regulating the abundance, activities or subcellular localizations of a variety of regulatory polypeptides and enzymes. Importantly, diverse E3 ligases are involved in these regulatory pathways by mediating phytohormone and light signaling or other pathways. In this updated review, we mainly summarize recent advances in our understanding of the regulatory roles of protein ubiquitination in plant development and plant–environment interactions, and primarily focus on different types of E3 ligases because they play critical roles in determining substrate specificity.

Introduction

As sessile organisms, plants adapt to changing environments to survive under stressful conditions. Plants perceive and transmit internal or external signals, and during these processes, post-translational protein modification approaches are frequently employed (Callis 2014). Post-translational modifications, including protein acetylation, methylation, phosphorylation and ubiquitination, possess key but distinct roles during different plant developmental stages and plant–environment interactions (Wilson et al. 2016). These modifications also interact with each other and form complicated cross-talk networks.

Ubiquitin, ubiquitin-activating enzyme (E1), ubiquitin-conjugating enzyme (E2), ubiquitin ligase (E3) and the intact 26S proteasome are essential for completing the protein ubiquitination degradation process. Inactivated ubiquitin is first activated by E1, in an ATP-dependent manner, through a thioester bond between the C-terminus of ubiquitin and a cysteine residue of E1, and then the thioester-linked ubiquitin is transferred onto a cysteine residue of E2 (Fig. 1) (Haas et al. 1982). The next step is the transfer of ubiquitin from E2 to a lysine residue of a substrate protein via E3, directly or indirectly (Fig. 1). Additionally, the residues including cysteine, serine and threonine can also be modified by ubiquitin (Ishikura et al. 2010, Shimizu et al. 2010, Chen et al. 2016). Importantly, the specificity of protein ubiquitination is mainly determined by E3, and thus E3 provides recognition and binding specificity to the substrate in a temporally and spatially regulated manner (Vierstra 2009). Subsequently, the targeted protein with polyubiquitin chains (>4) is usually degraded by the 26S proteasome and the released ubiquitin is recycled (Fig. 1) (Vierstra 2009, Callis 2014).

Fig. 1.

The ubiquitin-mediated 26S proteasome degradation pathway. First, inactivated ubiquitin (gray color) is activated by E1 in the presence of ATP, through a thioester bond between the C-terminus of ubiquitin and a cysteine residue of E1; subsequently, the activated ubiquitin (black color) is transferred onto E2. Secondly, the activated ubiquitin is transferred to a lysine residue (red color) of substrate protein by E3s, directly (HECT-type E3s) or indirectly (RING, U-box and CULLIN-based types E3s). Finally, the targeted protein with a polyubiquitin chain (>4) modification is degraded by the 26S proteasome and the released ubiquitins (inactivated form, gray in color) are recycled.

Protein ubiquitination was documented several decades ago. Ubiquitin is a highly conserved protein consisting of 76 amino acids. It was first purified in 1975 and was dubbed ‘ubiquitous polypeptide’, because it exists in almost all organisms—animal, plant, bacteria and yeast cells (Goldstein et al. 1975). After a few years, one previously unknown component, APF-1 (ATP-dependent proteolysis factor 1), was found to attach covalently to the target protein, causing the target to degrade in an ATP- and magnesium ion-dependent manner (Ciechanover et al. 1980). Soon after, Wilkinson and colleagues further demonstrated that APF-1 and ubiquitin were the same compound (Wilkinson et al. 1980). Further studies revealed that ubiquitin homeostasis is tightly regulated by de-ubiquitinating enzymes (Kimura and Tanaka 2010).

The number of E1s varies in different species, with at least three in wheat and two in Arabidopsis thaliana (Hatfield et al. 1997). At least 37 E2s and eight E2-like (ubiquitin enzyme variants or UEVs) proteins were detected in the A. thaliana proteome. The more detailed functions of plant E2s and UEVs remain largely unknown. The E3 ubiquitin ligases can be classified into single- and multisubunit groups. The single-subunit group includes several subfamilies based upon their mechanisms of action and the presence of specific domains: HECT (Homology to E6-AP C Terminus), RING (Really Interesting New Gene) and U-box type E3s. The multisubunit group, Cullin-RING box1-Ligases (CRLs), are further divided into four subfamilies: SCF (S phase kinase-associated protein 1–Cullin 1–F-box), BTB (Bric-a-brac–Tramtrack–Broad complex), DDB (DNA Damage-Binding domain-containing) and APC (anaphase-promoting complex) (Vierstra 2009). There are >1,600 loci responsible for the ubiquitin–26S proteasome system in A. thaliana and these express nearly 6% of its proteome (Vierstra 2009, Vierstra 2012), among which >1,400 loci encode E3s.

Why are there such a large number of E3s in plant genomes and proteomes? It is believed that the substrate specificity is mainly determined by E3s (Vierstra 2012). In the past several decades, great attention has been paid to understanding the role of E3 ligases, and numerous E3 ligases have been characterized in detail. Although many elegant reviews of protein ubiquitination have been published in the past several years (Dreher and Callis 2007, Vierstra 2009, Hua and Vierstra 2011, Vierstra 2012, Chen and Hellmann 2013), the rapid progress in this research field means that many new breakthroughs continue to emerge. In this updated review, we mainly present the recent progress in understanding the regulatory roles of protein ubiquitination modification in signaling pathways involved in plant development and abiotic stress responses, with particular emphasis on the molecular and genetic mechanisms of diverse E3 genes, mostly from studies in model plants Arabidopsis (including A. thaliana and A. lyrata) and rice (Oryza sativa).

RING-Type E3 Ligases

The RING-type E3s are characterized by the presence of a RING domain, which is a cysteine-rich domain that co-ordinates two zinc atoms. There are >470 proteins that contain a RING domain in A. thaliana (Vierstra 2009). Increasing numbers of studies have documented the key roles of the RING-type E3s in distinct plant developmental processes, including seed germination, post-germination growth, root development, light signal transduction, gametogenesis, organ size decisions and responses to abiotic stress (Table 1).

Table 1.

General functional descriptions of recently reported E3 ligases with roles in plant development and abiotic stress responses

Type	E3 ligase	Target(s)	Species	General functions description	References
RING	OsGW2	Unknown	Oryza sativa	Negatively regulates rice grain width and weight.	Song et al. (2007); Li et al. (2010); Xia et al. (2013)
	BB/EOD1	Unknown	Arabidopsis	Negatively regulate floral organ size.	Disch et al. (2006); Xia et al. (2013)
	XBAT32	ACS4, ACS7	Arabidopsis	Negatively regulates lateral root initiation.	Prasad et al. (2010)
	SINAT5	NAC1, FLC, LHY	Arabidopsis	Negatively regulates lateral root production, while positively regulating flowering time.	Xie et al. (2002); Seo et al. (2007); Park et al. (2010a)
	OsHAF1	HD1	Oryza sativa	Positively regulates rice heading date.	Yang et al. (2015)
	OsEL5	Unknown	Oryza sativa	Positively regulatse rice root development.	Koiwai et al. (2007)
	SIS3	Unknown	Arabidopsis	Positively regulates sugar signaling.	Huang et al. (2010)
	COP1	HY5, HFR1, LAF1, phyB	Arabidopsis	Negatively regulates photomorphogenesis.	Review in Lau and Deng (2012); Jang et al. (2010)
	COP1	Unknown	Arabidopsis	Negatively regulates stomatal development.	Kang et al. (2009
	COP1	SCAR	Arabidopsis	Negatively regulates root elongation.	Dyachok et al. (2011)
	COP1	ABI4	Arabidopsis	Regulates plant seedling de-etiolation.	Xu et al. (2016)
	COP1	GI	Arabidopsis	COP1-mediated GI degradation delays flowering time.	Jang et al. (2015)
	COP1	HYL1	Arabidopsis	Positively regulates miRNA biogenesis.	Cho et al. (2014)
	ZmGW2-CHR4, ZmGW2-CHR5	Unknown	Zea mays	Negatively regulate kernel size and weight in maize.	Li et al. (2010)
	RSL1	PYL4/PYR1	Arabidopsis	Negatively regulates ABA signaling.	Bueso et al. (2014)
	AIP2	ABI3	Arabidopsis	Negatively regulates ABA signaling.	Zhang et al. (2005)
	OsDSG1	OsABI3	Oryza sativa	Negatively regulates ABA signaling.	Park et al. (2010b)
	KEG	ABI5, KEG	Arabidopsis	Negatively regulate ABA signaling.	Stone et al. (2006); Liu and Stone (2010)
	KEG	ABF1/3	Arabidopsis	Negatively regulate ABA signaling.	Chen et al. (2013)
	OsHTAS	Unknown	Oryza sativa	Positively regulates plant abiotic stress responses.	Liu et al. (2016)
	RHA2b	Unknown	Arabidopsis	Positively regulates drought response, ABA dependent.	Li et al. (2011)
	RHA2a	Unknown	Arabidopsis	Positively regulates salt and osmotic responses, also involved in seed germination.	Bu et al. (2009)
	SDIR1	SDIRIP1	Arabidopsis	Positively regulates drought and salinity response	Zhang et al. (2007, 2015)
	OsSDIR1	Unknown	Oryza sativa	Positively regulates drought and salinity response	Gao et al. (2011)
	AtAIRP1	Unknown	Arabidopsis	Positively regulates drought response, ABA dependent.	Ryu et al. (2010)
	RGLG1/2	AtERF53	Arabidopsis	Positively regulate drought response, ABA dependent.	Cheng et al. (2011)
	XERICO	Unknown	Arabidopsis	Positively regulates drought response, ABA-dependent.	Ko et al. (2006)
	OsDIS1	OsNek6	Oryza sativa	Negatively regulates rice drought stress response.	Ning et al. (2011)
	RZFP34/CHYR1	Unknown	Arabidopsis	Promote plant drought tolerance.	Ding et al. (2015)
	STRF1	Unknown	Arabidopsis	Negatively regulates salt and osmotic stresses responses.	Tian et al. (2015)
U-box	SAUL1/PUB44	AAO3	Arabidopsis	Prevent premature senescence, increase ABA content.	Raab et al. (2009)
	AtPUB9	Unknown	Arabidopsis	Negatively regulates ABA signaling.	Samuel et al. (2008)
	BnARC1	Exo70A1	Brassica napus	Involved in self-incompatibility.	Samuel et al. (2009); Indriolo et al. (2012, 2014)
	SPL11	SPIN1	Oryza sativa	Positively regulates flowering, fungal stress response.	Vega-Sanchez et al. (2008)
	PUB13	Unknown	Arabidopsis	Negatively regulates flowering, perhaps through mediating FLC–SOC1 pathway.	Li et al. (2012); Liu et al. (2012); Zhou et al. (2015)
	PUB12/PUB13	ABI1	Arabidopsis	PUB12 and PUB13 ubiquitinate ABI1 and promote ABI1 degradation.	Kong et al. (2015)
	AtCHIP	FtsH1, FtsH2, ClpP3/4/5	Arabidopsis	Involved in chloroplast development and abiotic stress response.	Shen et al. (2007a, b); Wei et al. (2015)
	AtCHIP	PP2A	Arabidopsis	Negatively regulates low temperature response.	Luo et al. (2006)
	AtPUB19	Unknown	Arabidopsis	Negatively regulates plant drought response.	Liu et al. (2011); Seo et al. (2012)
	AtPUB18	Exo70B1	Arabidopsis	Negatively regulates plant drought response.	Liu et al. (2011); Seo et al. (2012); Seo et al. (2016)
	OsPUB15	Unknown	Oryza sativa	Positively regulates salt response.	Park et al. (2011)
	PUB22/23	RPN12a	Arabidopsis	Negatively regulate salt and drought responses.	Cho et al. (2008); Stegmann et al. (2012); Chen et al. (2014)
	PUB22	Exo70B2	Arabidopsis	Negatively regulates drought response.	Seo et al. (2016)
	CaPUB1	RPN6	Capsicum annuum	Negatively regulates plant dehydration and high-salinity tolerance.	Cho et al. (2006)
	MtPUB1	Unknown	Medicago truncatula	Negatively regulates infection, nodulation and nitrogen fixation.	Mbengue et al. (2010); Vernie et al. (2016)
HECT	UPL1, UPL2	Unknown	Arabidopsis	No biological phenotype description.	Bates and Vierstra (1999)
	UPL3, UPL4	Unknown	Arabidopsis	Essential for trichome development.	Downes et al. (2003)
	UPL5	WRKY53	Arabidopsis	Regulates leaf senescence.	Miao and Zentgraf (2010)
SCF	TIR1/AFBs	Aux/IAA	Arabidopsis	Promote auxin signaling.	Reviewed in Strader and Zhao (2016); Wang and Estelle (2014)
	COI1	JAZs	Arabidopsis	Promotes JA signaling.	Reviewed in Song et al. (2014)
	D3	D53	Oryza sativa	Promotes SL signaling.	Jiang et al. (2013); Khosla and Nelson (2016); Zhou et al. (2013)
	SLY1/SNE1	DELLAs	Arabidopsis	Promote gibberellin signaling.	Colebrook et al. (2014)
	RIFP1	RCAR3	Arabidopsis	Negatively regulates ABA response.	Li et al. (2016)
	SLFs	SLFs	Arabidopsis	Involved in self-incompatibility processes.	Sun et al. (2015)
	FBL17	KRP6/7	Arabidopsis	Positively regulates cell cycle during pollen development.	Kim et al. (2008)
	UCL1	CLF	Arabidopsis	Positively regulates flowering time.	Jeong et al. (2011)
	MAX2	SMAX1	Arabidopsis	Negatively regulates plant drought stress through mediating ABA signaling.	Shen et al. (2012); Waters et al. (2014)
	EDL3	Unknown	Arabidopsis	Positive regulator in seed germination and root growth.	Koops et al. (2011)
	PP2-B11	Unknown	Arabidopsis	Positive effect in plant salt stress response.	Jia et al. (2015)
BTB	ETO1/EOL1/2	ACS5	Arabidopsis	Negatively regulate ethylene biosynthesis, and involved in freezing tolerance.	Christians et al. (2009); Catala and Salinas (2015)
	BPM	ATHB6	Arabidopsis	Affects stomatal behavior and responses to ABA, regulates fatty acid metabolism.	Lechner et al. (2011); Hu et al. (2014)
	BPM	MYB56	Arabidopsis	Positively controls flowering time.	Chen et al. (2015)
	LRB1/LRB2	FRI	Arabidopsis	Positively control flowering time.	Choi et al. (2011); Hu et al. (2014)
	LRB1/LRB2	phyB/phyD	Arabidopsis	Involved in light signaling pathway.	Christians et al. (2012); Ni et al. (2014)
DDB	ABD1	ABI5	Arabidopsis	Negatively regulates ABA signaling.	Seo et al. (2014)
	ASG2	Unknown	Arabidopsis	Negatively regulates ABA signaling.	Dutilleul et al. (2016)
	PRL1	OsAKIN10	Oryza sativa	Regulates phytohormone biogenesis, root stem cell activity and stress responses.	Lee et al. (2008); Flores-Perez et al. (2010); Ji et al. (2015)
	DWA1/2	ABI5	Arabidopsis	Negatively regulate ABA signaling.	Lee et al. (2010)
APC	CCS52A	Unknown	Arabidopsis	Controls meristem maintenance perhaps through repression of mitotic activity in the quiescent center.	Vanstraelen et al. (2009)
	TAD1	MOC1	Oryza sativa	Negatively regulates rice tillering.	Xu et al. (2012)
	TE	MOC1	Oryza sativa	Negatively regulates rice tillering, and ABA/gibberellin antagonism.	Lin et al. (2012, 2015)

RING-type E3s involved in seed biology and root elongation

Seed grain size is one of the most important agronomic traits. Several regulatory pathways that determine seed size have been identified, among which RING-type E3 ligases are involved, primarily through regulating gametogenesis and cell cycle processes. RING-type E3 DA2 negatively regulates seed size by decreasing cell proliferation in developing seeds and by functioning synergistically with the ubiquitin receptor DA1, which is also a key regulator in seed size control (Xia et al. 2013, Li and Li 2016). An earlier investigation showed that the rice RING-type E3 OsGW2 (Grain Width and Weight 2), the homolog of DA2 in rice, negatively affected grain size and final yield through mediating cell division (Song et al. 2007). Furthermore, OsGW2 can delay seed growth in A. thaliana (Xia et al. 2013). Similarly, the orthologs of OsGW2 in maize (Zea mays), ZmGW2-CHR4 and ZmGW2-CHR5, also function in controlling the variation in kernel size and weight (Li et al. 2010). DA1 functions synergistically with another RING-type E3 ligase EOD1 (ENHANCER OF DA1) to restrict organ size (Xia et al. 2013). Recently, Li’s group further demonstrated that DA1 interacts physically with and mediates UBP15 (UBIQUITIN-SPECIFIC PROTEASE15) stability, which maternally regulates organ size by promoting cell proliferation in developing seeds (Du et al. 2014). Overall, these studies suggested that orthologs DA2, OsGW2, ZmGW2-CHR4 and ZmGW2-CHR5 have conserved roles in cereal crops. The substrates of these E3 ligases remain elusive, but genetic analyses revealed that DA2 and EOD1/BB might share common downstream targets to determine plant organ size synergistically.

Ubiquitin-mediated proteolysis also has a pivotal role in root development, flowering time control and hypocotyl elongation. A RING-type E3 ligase SINAT5 (SINA of Arabidopsis thaliana 5) targets the NAC1 transcription factor for degradation to attenuate the auxin signal and negatively regulate the quantity of lateral roots (Xie et al. 2002). Later, SINAT5 was also shown to ubiquitinate FLC (Flowering Locus C) (Seo et al. 2007) and a component of the circadian oscillator, LHY (Late Elongated Hypocotyl) (Park et al. 2010a), to regulate the transition from the vegetative to the reproductive phase. These studies indicate that SINAT5 is a versatile regulator of plant development, including lateral root initiation and flowering time control, through modifying different substrates. Recently, the rice RING-type E3 ubiquitin ligase OsHAF1 (Heading date Associated Factor 1) was shown to be the key regulator during rice heading date control through regulating HD1 (Heading Date 1) degradation, which determines rice regional adaptability by modulating the photoperiodic response (Yang et al. 2015). The haf1 mutants show the later heading date phenotype under both short-day (SD) and long-day (LD) conditions, while the haf1 hd1 double mutant headed similarly to hd1 under SD conditions but mimicked haf1 under LD conditions, indicating that HAF1 may determine rice flowering time mainly through HD1 under SD conditions (Yang et al. 2015). Another RING-type E3, HOS1 (high expression of osmotically responsive genes 1) delays flowering through promoting CO (CONSTANS) degradation and thus inhibiting FLOWERING LOCUS T (FT) transcription (Lazaro et al. 2012). In addition to flowering time control, HOS1 is also involved in root development in rice. OsHOS1-RNAi (RNAi interfernce) plants show straight roots in contrast to wild-type plants that exhibit root curling. Further investigation demonstrated that OsHOS1 physically interacts with OsEREBP1 (ETHYLENE-RESPONSIVE ELEMENT BINDING PROTEIN1) and OsEREBP2, known to regulate OsRMC gene expression, which is involved in the root curling response (Lourenco et al. 2015). A recent study also showed that HOS1 regulates the AGO1 (Argonaute1) level by maintaining the appropriate transcription level of MIR168b (Wang et al. 2015). Further, as an E3, HOS1 negatively regulates hypocotyl elongation through inhibiting the transcriptional activation activity of PIF4 (PHYTOCHROME INTERACTING FACTOR 4) (Kim et al. 2017). Thus, it is believed that the E3 ligase HOS1 may also play non-proteolytic roles in the gene expression regulation process (Jung et al. 2014).

In contrast to the negative effects of SINAT5 on lateral root initiation, the positive effects of other E3 ligases on this process have also been examined, including RING-type E3 XBAT32 (XB3 ortholog 2 in Arabidopsis thaliana 32) and EL5 (Elicitor 5). The lateral root production arrest phenotype of xbat32 was fully restored similar to wild-type levels when treated with an antagonist of the ethylene receptor, silver nitrate, or an inhibitor of aminocyclopropane-1-carboxylic acid synthase (ACS) (Prasad et al. 2010). It was further revealed that XBAT32 interacts with the ethylene biosynthesis enzymes ACS4 and ACS7, catalyzes the attachment of ubiquitins to ACS4/7 and promotes the degradation of both proteins. Thus, the loss of function of XBAT32 may stabilize ACS4/ACS7, leading to increased ethylene synthesis and the subsequent suppression of lateral root formation (Prasad et al. 2010). Furthermore, EL5 was shown to act as a membrane-anchored RING-type E3, and plays a major role in maintaining cell viability in the root apical meristem in rice (Koiwai et al. 2007). The overexpression mutant EL5 plants, which also possessed an inactive E3 ligase, displayed a rootless phenotype, suggesting that EL5 ubiquitin ligase activity functions in rice root growth (Koiwai et al. 2007). Those studies indicated that E3 ubiquitin ligases are involved in lateral root development in monocots and dicots through regulating plant phytohormone biosynthesis, transport and signaling pathways or cell cycle progression.

RING-type E3s involved in flowering time control and light response

Rapid initial responses to the dark–light transition are essential for the early development of seedlings. In this process, a notable RING-type E3 protein COP1 (Constitutive Photomorphogenesis 1) down-regulates the abundance of skotomorphogenesis-positive regulators HY5, LAF1 and HFR1 to balance skotomorphogenesis and photomorphogenesis further (Lau and Deng 2012). Furthermore, COP1 targets phyB for degradation by the 26S proteasome under red light (Jang et al. 2010), and this process can be enhanced by PIF (PHYTOCHROME INTERACTING FACTOR) proteins; consequently, nuclear phyB accumulates in pif single (pif3, pif4 or pif5) or double mutants (pif3 pif4 and pif4 pif5). PIFs negatively regulate the phyB protein level through accelerating the ubiquitin–26S proteasomal process (Jang et al. 2010). These investigations indicated that there is a feedback loop between COP1 and phyB to fine-tune the plant light response. It is noteworthy that COP1 also promotes the degradation of SCAR1 under darkness, which results in the disorganization of F-actin, and subsequently negatively regulates root elongation (Dyachok et al. 2011).

Recently, CSU2 (COP1 SUPPRESSOR 2) was dissected, which represses the E3 ubiquitin ligase activity of COP1 through interaction of its coiled-coil domains (Xu et al. 2015). CSU2 also plays key roles in primary root initiation (Xu et al. 2015). Furthermore, ABI4 (Abscisic acid-Insensitive 4), a key positive regulator in the ABA signaling pathway, which is involved in seed dormancy, germination and flowering time control (Shu et al. 2013, Shu et al. 2016a, Shu et al. 2016b), is also targeted by COP1 for degradation (Xu et al. 2016). COP1 targets ABI4 for degradation to optimize plant early development, especially for seedling de-etiolation processes (Xu et al. 2016). Unsurprisingly, COP1 also integrates the cross-talk between photoperiod and ambient temperature signaling pathways through promoting GI (GIGANTEA) degradation, which directly binds to the FT promoter. Finally, COP1-mediated GI degradation delays flowering time (Jang et al. 2015). Unexpectedly, COP1 is involved in plant global microRNA (miRNA) biogenesis processes (Cho et al. 2014). A decreased level of miRNAs in the cop1 mutant was detected, and further analyses showed that the cause was the destabilization of HYL1 (HYPONASTIC LEAVES 1), an RNA-binding protein required for precise miRNA processing (Cho et al. 2014). In addition, PIF1 boosted the E3 ligase activity of COP1, and finally antagonized photomorphogenesis processes in darkness (X. Xu et al. 2014). Altogether, these studies demonstrated that the RING-type E3 ligase COP1 is indeed a versatile factor acting in numerous plant developmental periods, primarily through mediating light signaling.

RING-type E3s and ABA signaling transduction

Many studies have shown that RING-type E3s are involved in plant abiotic stress hormone (e.g. ABA) signaling transduction as well as in regulating plant responses to different abiotic stresses, including drought, salinity and extreme temperature (Table 1), and diverse regulators including ABI3, ABI4 and ABI5 are also involved (Shu et al. 2016c).

Plant stress responses are controlled by diverse pathways, including ABA-dependent and -independent pathways; and many RING-type E3s are implicated in ABA signaling networks. A RING-type E3 ligase RSL1 (RING FINGER OF SEED LONGEVITY 1) targets turnover of ABA receptors PYL4 and PYR1 in the plasma membrane, and further fine-tunes earlier ABA signaling (Bueso et al. 2014). Correspondingly, overexpressing transgenic RSL1 plants show a reduced ABA sensitivity, and rsl1 knockdown lines show hypersensitivity to ABA, because of altered ABA receptor protein levels (Bueso et al. 2014). AIP2 (ABI3-Interacting Protein), a RING-type E3 ligase, serves as a negative regulator of ABA signaling by targeting ABI3 (Abscisic acid-Insensitive 3) for degradation (Zhang et al. 2005). Similar to dicots, the ortholog of AIP2 in rice, OsDSG1 (Delayed Seed Germination 1) also possesses E3 ubiquitin ligase activity, physically interacts with OsABI3 and negatively regulates tolerance to salt and drought stresses within an ABA-dependent pathway (Park et al. 2010b). Another ABA signaling component, the bZIP transcription factor ABI5, is also targeted by ubiquitin-dependent proteolysis by a RING-type E3 ligase KEG (KEEP ON GOING) (Stone et al. 2006, Liu and Stone 2010, Liu and Stone, 2013). There is a feedback loop among ABA signaling, KEG and ABI5—in which KEG promotes ABI5 degradation to attenuate ABA signaling, whereas ABA accelerates ubiquitination and degradation of KEG and results in ABI5 accumulation and the promotion of ABA response. These studies revealed that ubiquitination is involved in ABA signaling control, and these pathways are conserved in dicots and monocots (Fig. 2). Interestingly, these studies also hinted that the abundance of AIP2 might be negatively regulated by ABA signals, similar to KEG.

Fig. 2.

Known E3 ligases involved in the regulation of the abundance of ABI3, ABI4 and ABI5, three important components in the ABA signaling cascade. RING-type E3 ligase KEG, DDB-subtype E3 ligases complexes DDB^DWA1/DWA2 and DDB^ABD1 concurrently target ABI5 for degradation in Arabidopsis. RING-type E3 ligase AIP2 in Arabidopsis and its ortholog OsDSG1 in rice ubiquitinate ABI3 and OsABI3 for degradation, while Arabidopsis RING-type E3 COP1 targets ABI4 to regulate its protein level. The E3 ligase OsDSG1 and its substrate OsABI3 are highlighted with yellow background in rice, while proteins from Arabidopsis are highlighted with blue background.

In addition to ABI5, KEG also targets two other bZIP transcription factors for degradation: ABF1 and ABF3 (Chen et al. 2013). ABF1 and ABF3 are also involved in seed germination and stress responses (Chen et al. 2013). Other ABFs, including ABF3 and ABF4, may also be involved in the KEG-mediated degradation pathway. As well as being involved in the ABA signaling transduction pathway, KEG acts in the JA (jasmonate) signaling cascade (Pauwels et al. 2015). KEG can target JAZ12 (JASMONATE ZIM-DOMAIN 12) but maintain its stability, although most JAZ proteins are substrates of SCF^COI1–E3 ligase. In addition, exogenous ABA treatment promotes JAZ12 degradation, and KEG knockdown RNAi transgenic lines showed a decrease of JAZ12, while KEG overexpression inhibits the COI1-mediated JAZ12 degradation processes (Pauwels et al. 2015). This ABA–JA cross-talk mediated by E3 ligases requires further investigation.

RING-type E3s involved in plant salt and drought stress responses

SDIR1 (SALT- AND DROUGHT-INDUCED RING FINGER 1), a RING-type E3 ligase, functions as a positive regulator of ABA signaling by promoting ABI3 and ABI5 transcription, and SDIR1-overexpressing transgenic plants consistently show a drought-tolerant phenotype (Zhang et al. 2007), although the detailed mechanisms underlying the promoting effect of SDIR1 on AIB3 and ABI5 transcription remain unknown. Furthermore, rice OsSDIR1 is also a functional E3 ligase and OsSDIR1-overexpressing rice showed markedly increased drought tolerance (Gao et al. 2011). Recently, the substrate of SDIR1, SDIRIP1 (SDIR1-INTERACTING PROTEIN 1), was characterized (Zhang et al. 2015). SDIR1 ubiquitinates SDIRIP1 and promotes its turnover through the 26S proteasome system, and finally modulates ABA-mediated responses of plants to salt stress (Zhang et al. 2015). These studies indicated that SDIR1-mediated pathways in drought and salt responses may be conserved in both dicots and monocots, and are all ABA dependent (Fig. 3). Finally, the versatile E3 ligase SDIR1 may mediate plant abiotic stress responses by affecting distinct substrate(s).

Fig. 3.

E3 ubiquitin ligase-mediated pathways involved in plant abiotic stress responses in ABA-dependent or -independent manners. This network primarily focuses on plant abiotic stress response (drought stress) pathways in which diverse E3 ligases are involved. Generally, drought stress responses are classified as ABA-dependent (SDIR1, OsSDIR1, AIRP, RHA2b, OsHTAS, PUB19 and XERICO) and ABA-independent (PUB1, PUB22/23, OsDIS1, RGLG1/2, CHYR1 and STRF1) pathways. In rice, the E3 ligase OsDIS1 and its substrates OsNek6 and OsSKIPa, and E3 ligases OsSDIR1 and OsHATS are highlighted with a yellow background, while the proteins from Arabidopsis are highlighted with a blue background. NCED3 and the bZIP transcription factors regulated at the transcriptional level are highlighted with a red background.

A RING-type E3 ligase AIRP1 (ABA-insensitive RING protein 1) was also characterized as a positive regulator in the ABA-dependent drought responses. AIRP1-overexpressing plants were significantly tolerant to severe drought stress, while the atairp1 mutant was highly susceptible to water deficiencies (Ryu et al. 2010). Another RING-type E3, RHA2b, was shown to play a positive role in ABA signaling. Overexpression of RHA2b leads to ABA-sensitive phenotypes with reduced water loss and subsequently increased drought tolerance (Li et al. 2011). In addition to ABA signaling, ABA content is also a key factor in plant drought responses. The gene XERICO, encoding the RING-type E3 enzyme, is induced by salt and osmotic stresses and confers plant drought tolerance. Detailed analyses showed that XERICO indirectly activates expression of a key ABA biosynthesis gene, NCED3, and finally increases the ABA level. In addition to being involved in drought and salt stress responses, RING-type E3 ligases were recently shown to have biological functions in plant heat responses. Rice RING-type E3 ligase OsHTAS (HEAT TOLERANCE AT SEEDLING STAGE) had a positive effect on heat tolerance at the early seedling stage, by enhancing hydrogen peroxide (H₂O₂)-induced stomatal closure (Liu et al. 2016). In detail, OsHTAS promoted H₂O₂ accumulation in shoots, altered the stomatal aperture status of rice leaves and increased ABA biosynthesis. The detailed mechanisms underlying the effect of OsHTAS on heat responses need further investigation by determining its substrates (Liu et al. 2016). Overall, SDIR1, AIRP1, RHA2b and OsHTAS positively regulate plant drought/salt/heat stress responses and are ABA signaling dependent, while XERICO improves drought tolerance by enhancing ABA biosynthesis (Fig. 3).

Another series of RING-type E3 ligases, RGLG1 (RING DOMAIN LIGASE 1), RGLG2 and OsDIS1, negatively regulate plant drought tolerance through ABA-independent pathways. RGLG1 and RGLG2 are E3 ligase homologs, and mediate AtERF53 ubiquitination for proteasome degradation and negatively regulate the drought responses (Cheng et al. 2012). Furthermore, the RING-type E3 gene OsDIS1 (Oryza sativa drought-induced SINA protein 1) also has a negative role in the rice drought stress responses (Ning et al. 2011). Interestingly, OsDIS1 has high sequence identity (82%) with the previously described RING-type E3 gene SINAT5 in Arabidopsis, and OsDIS1 negatively regulates drought responses through the transcriptional regulation of diverse stress-related genes and possibly through the post-translational regulation of OsNek6, a tubulin complex-related serine/threonine protein kinase (Ning et al. 2011). Also, considering the role of SINAT5 in lateral root production and flowering time control (Xie et al. 2002), SINAT5/OsDIS1 show diverse functions in Arabidopsis and rice.

Recently, the relationships between plant drought responses, protein ubiquitination and phosphorylation have been uncovered (Ding et al. 2015). The genetic and biochemical evidence demonstrated that RING-type E3 RZFP34/CHYR1 (CHY ZINC-FINGER AND RING PROTEIN 1) was phosphorylated by kinase SnRK2.6 on its Thr178 residue. Phenotypic analyses showed that CHYR1 promotes plant drought tolerance through mediating stomatal closure and reactive oxygen species (ROS) biosynthesis, and this effect depended on the phosphorylation status of CHYR1 mediated by SnRK2.6 (Ding et al. 2015). Another RING-type E3 STRF1 (Salt Tolerance RING Finger 1) localized on the plasma membrane and precisely regulated intracellular membrane trafficking and ROS production, and finally negatively regulated salt, ionic and osmotic stress responses (Tian et al. 2015). It is noteworthy that the RING-type E3s described above (i.e. RGLG1, RGLG2, OsDIS1, CHYR1 and STRF1) all affect drought and salt stress responses through ABA-independent pathways (Fig. 3).

U-Box-Type E3 Ligases

U-box-type E3 ligases are characterized by a conserved U-box motif of about 70 amino acids. Secondary structure prediction revealed that the U-box is a modified RING-finger domain lacking the scaffold and zinc-chelating cysteine and histidine residues that are conserved in the typical RING domain. There are about 64 predicted PUB (Plant U-box protein) genes in Arabidopsis (Vierstra 2009). Studies on PUBs from several different plant species showed that these types of E3s perform a range of functions in plant development including plant self-incompatibility, seed germination, flowering time control, chloroplast development and numerous abiotic stress responses (Table 1).

U-box-type E3s involved in plant self-incompatibility responses, seed germination, flowering control and nitrogen fixation

BnARC1 (Brassica napus Arm Repeat-Containing 1), the first reported U-box E3 in plants, was characterized as a positive regulator of the Brassica self-incompatibility response (Stone et al. 2003). BnARC1 may promote the degradation of unknown compatibility factors in the pistil, resulting in pollen rejection. Six years later, the substrate of BnARC1, Exo70A1, was characterized—this is a compatibility factor in the stigma and a negative regulator during the Brassica self-incompatibility response (Samuel et al. 2009). Further analyses revealed that ARC1 is required for self-incompatibility in A. lyrata, and ARC1 is usually absent in Brassicaceae species in which the self-incompatibility trait is lost, indicating conserved roles of ARC1 in the self-pollen rejection response within Brassicaceae (Indriolo et al. 2012). Detailed investigation demonstrated that ARC1 promotes two distinct self-pollen avoidance traits within different genetic backgrounds, including A. thaliana Col-0 and Sha ecotypes (Indriolo et al. 2014). Finally, these investigations highlighted the conserved role in self-pollen rejection in Brassica and Arabidopsis (A. thaliana and A. lyrata) species (Indriolo and Goring 2014). Consequently, ARC1 is the best-studied U-box type E3 ligase involved in the plant self-incompatibility response.

Similar to the RING-type E3s, U-box type E3 ligases have also been demonstrated to act as regulators in seed germination and leaf senescence processes by mediating ABA signaling or biosynthesis. The loss-of-function Atpub9 mutant is hypersensitive to ABA during seed germination, while the abi3 pub9 double mutant is insensitive to ABA, suggesting that the U-box E3 ligase PUB9 may act upstream of ABI3 (Samuel et al. 2008). Furthermore, the same research group documented another U-box E3 ligase, SAUL1/AtPUB44, which prevents leaf senescence through targeting ABA synthase AAO3 and promoting its degradation, and accordingly decreasing the ABA level (Raab et al. 2009). Because ABA is sufficient to trigger leaf senescence, it is consistent that the saul1 pub44 double mutant enhances ABA biosynthesis and leads to premature senescence under low light intensity conditions (Raab et al. 2009).

In rice, the U-box type E3 OsSPL11 (Spotted leaf 11) functions in flowering time control (Vega-Sanchez et al. 2008). OsSPIN1 (SPL11-Interacting protein 1) represses flowering by down-regulating the expression of the flowering promoter gene Hd3a (Heading date3a), while OsSPIN1 interacts with OsSPL11 and is monoubiquitinated by the latter; however, the result of ubiquitination does not appear to target it for degradation (Vega-Sanchez et al. 2008). In addition, the ortholog of OsSPL11 in A. thaliana, PUB13, negatively regulates flowering, possibly through mediating the FLC–SOC1 pathway (Li et al. 2012, Liu et al. 2012), and the ARM domain in PUB13 shows a predominantly negative effect (Zhou et al. 2015). Furthermore, a recent study demonstrated that PUB12 and PUB13 ubiquitinate the ABA co-receptor ABI1 and promote its degradation (Kong et al. 2015); thus, ABI1 is the common substrate for PUB12 and PUB13. However, the targets of PUB13/OsSPL11 in the flowering time control pathway need further elucidation.

U-box type E3 ligases also act in nitrogen-fixing processes, which have a key role in leguminous plant development. U-box E3 Medicago truncatula PUB1 is phosphorylated by the symbiotic receptor kinase LYK3 and negatively regulates rhizobial infection and nodulation initiation (Mbengue et al. 2010). The E3 ubiquitin ligase activity of MtPUB1 is U-box dependent, and MtPUB1 transcription is specifically induced by symbiotic conditions and Nod factors (Mbengue et al. 2010). Recently, MtPUB1 was found also to be phosphorylated by DMI2 (DOES NOT MAKE INFECTIONS 2), a key symbiotic receptor kinase of the symbiosis signaling pathway, and is required for both rhizobial and arbuscular mycorrhizal (AM) endosymbioses (Vernie et al. 2016). Furthermore, the ubiquitination activity of MtPUB1 is essential to modulate negatively successive stages of infection and the development of rhizobial and AM symbioses (Vernie et al. 2016). These studies highlighted the important regulatory roles of phosphorylation modification on MtPUB1, but dissection of the substrate(s) of MtPUB1 is still required. This will increase our understanding of the precise mechanisms of action of MtPUB1 in nitrogen fixation.

U-box-type E3s involved in drought, salinity and extreme temperature responses

Similar to RING-type E3s, U-box-type E3s also act in diverse abiotic stress responses including extreme temperature, drought and high-salinity conditions—and most of these regulatory pathways recruit phytohormone signaling (Table 1).

The U-box type E3 ligase AtCHIP (Carboxyl terminus of Hsc70-Interacting Protein) is involved in plant–environment interactions. Extreme temperature up-regulates AtCHIP expression, and AtCHIP-overexpressing transgenic lines were sensitive to both low and high temperature treatments, due to interruption of membrane function under these conditions (Yan et al. 2003). A following study demonstrated that AtCHIP interacts with and ubiquitylates an A subunit of PP2A (Protein Phosphatase 2A), which results in increased PP2A activity in AtCHIP-overexpressing lines (Luo et al. 2006). Further investigation revealed that AtCHIP ubiquitylates and promotes the degradation of a chloroplast proteolytic subunit, ClpP4, under certain stress conditions (Shen et al. 2007b). ClpP4 is a core subunit of the Clp (Caseinolytic protease) complex that is involved in protein quality control in the chloroplast (Shen et al. 2007b). Additionally, two other subunits of the chloroplast FtsH protease complex, FtsH1 and FtsH2, were demonstrated to interact with AtCHIP and were degraded by AtCHIP (Shen et al. 2007a). A recent publication revealed that AtCHIP also interacts with and ubiquitinates ClpP3 and ClpP5, and the regulatory roles of AtCHIP on Clp are conserved between A. thaliana and tobacco (Wei et al. 2015). These studies indicated that AtCHIP has complex functions involved in both plant extreme temperature responses and protein quality control during chloroplast development.

In pepper (Capsicum anuum), CaPUB1 (Putative U-box Protein 1) was identified as a U-box E3 ligase, which negatively regulates plant dehydration and high-salinity tolerance (Cho et al. 2006). CaPUB1 ubiquitinates and promotes the degradation of a subunit of the 26S proteasome complex, RPN6. A typical drought stress-induced gene RD29a was markedly decreased in CaPUB1-overexpressing lines (Cho et al. 2006). Furthermore, the orthologs of CaPUB1 in A. thaliana, AtPUB22 and AtPUB23, were also shown to interact with a subunit of the 26S proteasome complex, AtRPN12a (Cho et al. 2008). Loss-of-function pub22 or pub23 mutants are much more tolerant to drought and salt stresses, and the pub22 pub23 double mutant displayed an additive effect (Cho et al. 2008). Further research showed that PUB22 and PUB23 were also involved in biotic stress responses (Stegmann et al. 2012, Chen et al. 2014). These studies indicate that the orthologs in A. thaliana and pepper (i.e. CaPUB1, PUB22 and PUB23) have conserved functions in plant drought and salt stress responses pathways (Fig. 3).

The rice U-box E3 ligase OsPUB15 is a positive regulator of plant tolerance to salinity and drought stress (Park et al. 2011). OsPUB15 is induced by H₂O₂, salt and drought stresses, and its overexpressing plants grow better than controls under these stress conditions. In contrast to the positive effect of OsPUB15 on drought and salt responses, AtPUB19, another U-box E3, has negative regulatory roles in plant responses to ABA and dehydration (Liu et al. 2011). Loss-of-function AtPUB19 mutants show hypersensitivity to ABA, enhanced ABA-induced stomatal closing and enhanced drought tolerance, while AtPUB19 overexpression results in opposite phenotypes (Liu et al. 2011). PUB18/PUB19 and PUB22/PUB23 are negative regulators in plant drought and salt stress responses, and, interestingly, the former pair mediate drought responses in an ABA-dependent manner, while the latter pair regulate this response in an ABA-independent pathway (Seo et al. 2012). Further investigation demonstrated that PUB18/PUB19 possesses a UND (U-box N-terminal domain) but PUB22/PUB23 does not, and this UND motif is required for the negative regulation of ABA-dependent stomatal movement of PUB18. In detail, PUB18 targets the degradation of Exo70B1, a subunit of the exocyst complex, while Exo70B2 is a substrate of PUB22, and this difference depends on the UND motif (Seo et al. 2016).

HECT-Type E3s, Plant Development and Abiotic Stress Responses

HECT-type E3 ligases are single-subunit E3s defined by a signature HECT domain with a conserved catalytic cysteine. There are only seven HECT proteins in A. thaliana: UPL1–UPL7 (Downes et al. 2003). Unlike other types of E3 ligases, HECT-type E3s act as receptors of ubiquitin from E2 enzymes, and then transfer ubiquitin to a specific lysine residue in the substrate (Vierstra 2009).

The first identified HECT-type E3 ligases in A. thaliana were UPL1 (Ubiquitin-Protein Ligase 1) and UPL2. Both have been characterized at the biochemical level, but their biological functions remain largely unknown (Bates and Vierstra 1999). UPL3 plays an important role in trichome development (Downes et al. 2003), and the upl3 mutant displayes aberrant trichomes, which contain five or more branches, while the wild type possesses three. Interestingly, although UPL4 and UPL3 have structural similarities (Downes et al. 2003), no obvious trichome phenotype in upl4 mutants was detected, suggesting that UPL3 and UPL4 have distinct functions. Additionally, UPL5 was characterized as a negative regulator during leaf senescence—it ubiquitinates and promotes the degradation of transcription factor WRKY53, an important mediator regulating leaf senescence (Miao and Zentgraf 2010). No studies of UPL6 and UPL7 have yet been reported (Table 1). The precise molecular mechanisms of this type of E3s during plant development and abiotic stress responses need further investigation.

Multisubunit E3 Ligases

The Cullin–RING box1 ligases (CRLs) are multisubunit E3 ligases containing a Cullin protein, an RBX1 (RING-Box 1) protein, a variable module which determines the substrate specificity and possibly several other subunits (Vierstra 2009). The CRLs are further divided into four subtypes, SCF, BTB, DDB and APC E3s, and these E3 complexes possess similar structural composition. The most important structural feature of multisubunit E3s is that there is a substrate receptor, which mediates the real substrates (Vierstra 2009, Hua and Vierstra 2011, Vierstra 2012). SCF-subtype E3 ligases are composed of SKP1 (S phase Kinase-associated Protein 1), CUL1 (Cullin 1), RBX1 (RING-Box 1) and FBX (F-Box) proteins, with the FBX proteins determining substrate specificity (Vierstra 2009). There are >700 F-box genes in A. thaliana and some are employed in phytohormone signaling pathways, plant development and diverse abiotic stress responses (Table 1).

SCF-subtype E3s and diverse phytohormone signaling

Notably, SCF-subtype E3 ligases are employed in plant development stages and stress responses through mediating early signaling transduction of diverse phytohormones, including auxin, JA, SL (strigolactone) and gibberellin.

The F-box protein TIR1 (Transport Inhibitor Response 1) and its relatives, AFB1–3 (Auxin signaling F Box protein 1–3) and AFB5, are core components of SCF^TIR1–AFBs, E3 ligases that interact with and promote degradation of Aux/IAA repressors, and then activate ARFs (Auxin Response Factors) (Wang and Estelle 2014, Strader and Zhao 2016). Using a similar strategy, JA also employs an SCF E3 ligase complex as a receptor, in which the F-box protein is COI1 (Coronatine Insensitive 1) (Yuan and Zhang 2015). The JA signaling negative regulators JAZs are the substrates of SCF^COI1 E3 ligase complexes, which transmit the JA signal to modulate plant defense responses against bacterial and biotrophic pathogens (Song et al. 2014).

The SLs are a family of phytohormones that control diverse aspects of plant growth stages, especially plant architecture. Interaction between SL and its receptor D14 (DWARF 14) activates the SCF^D3 E3 ligase complex, and the latter promotes degradation of the SL repressor D53 (DWARF 53) (Jiang et al. 2013, Zhou et al. 2013, Khosla and Nelson 2016). The early signaling cascade of gibberellin also resembles characteristics of the TIR1/AFBs–Aux/IAA and COI1–JAZs branches, but it is noteworthy that the gibberellin receptor GID1 (Gibberellin Insensitive Dwarf 1) is not an F-box protein (H. Xu et al. 2014). SCF-type E3 ligase complexes SCF^SLY1 and SCF^SNE1 in A. thaliana, and SCF^GID2 in rice, target the DELLA proteins for 26S proteasome-mediated degradation, and the GID1–gibberellin interaction promotes this degradation (Colebrook et al. 2014). Together, DELLA proteins have a repressive function in gibberellin signaling, similar to the signaling cascades of Aux/IAA and auxin, JAZs and JA, and D53 and SL; while the corresponding SCF-subtype E3 ligase complexes mediate the degradation of these repressors.

SCF-subtype E3s involved in numerous plant developmental stages

The SCF-subtype E3 ligases regulate plant development including seed germination, self-incompatibility control, plant male gametogenesis, pollen development and flowering time control through mediating degradation of distinct substrates or phytohormone signaling transduction pathways. The A. thaliana F-box protein RIFP1 (RCAR3 INTERACTING F-BOX PROTEIN 1) negatively regulates the ABA responses by facilitating degradation of ABA receptor RCAR3 (Li et al. 2016). Correspondingly, the rifp1 mutant showed ABA hypersensitivity during seed germination and enhanced tolerance to water deficit stress (Li et al. 2016). Similar to the U-box type E3 ligase involved in plant self-incompatibility recognition (Indriolo et al. 2012, Indriolo et al. 2014), the F-box subtype of E3 is also employed in this process. Petunia F-box protein SLFs (S-Locus F-box) promote themselves through degradation using the 26S–proteasome approach, and the stability of these SLFs remarkably affects their functions in self-incompatibility processes (Sun et al. 2015). Male gametogenesis and pollen development are also regulated by F-box E3 protein through cell cycle processes. The SCF^FBL17 E3 complex targets the cyclin-dependent kinase inhibitors KRP6 and KRP7 for proteasome-dependent degradation, thus FBL17 loss-of-function leads to stabilization of the two KRPs and inhibition of germ cell cycle progression (Kim et al. 2008). Further investigation showed that this degradation was also essential for pollen development through promoting the second mitosis (Gusti et al. 2009). The F-box E3 ligase complex is also involved in plant flowering time control. Overexpression of the F-box protein-encoding gene UCL1 (UPWARD CURLY LEAF 1) reduces abundance of CLF (CURLY LEAF) protein, which is the key transcriptional repressor in flowering. Accordingly, UCL1-overexpressing lines display similar phenotypes to clf mutants, including early flowering (Jeong et al. 2011).

The F-box protein MAX2 (MORE AXILLARY GROWTH 2) is involved in karrikin and SL signaling pathways to regulate plant architecture, photomorphogenesis and leaf senescence (Shen et al. 2012, Waters et al. 2014). A recent report demonstrated that MAX2 negatively regulates plant drought stress through mediating the ABA signaling cascade (Bu et al. 2014). The F-box E3 ligase gene EDL3 (EID1-Like protein 3) is induced by osmotic stress, salt and exogenous ABA treatment (Koops et al. 2011). EDL3 functions as a positive regulator in ABA-dependent phenotypes including in seed germination, root growth and flowering (Koops et al. 2011). Another SCF-subtype E3 F-box protein PP2-B11 showed a positive effect on plant salt stress responses (Jia et al. 2015). iTRAQ (isobaric tags for relative and absolute quantitation) analyses revealed that PP2-B11 affects the level of numerous proteins, and finally influences the expression of sodium ion homeostasis genes and ROS production (Jia et al. 2015). However, the substrate(s) of PP2-B11 and EDL3 need further elucidation.

BTB-subtype E3s in plant development and abiotic stress responses

BTB-subtype E3 ligases are composed of CUL3 (Cullin 3), RBX1 and BTB (BTB domain-containing) protein, in which the BTB protein confers substrate specificity (Vierstra 2009). There are two redundant CUL3 genes which are essential for embryo development (Figueroa et al. 2005), and about 80 BTB-containing proteins in A. thaliana (Vierstra 2009).

Recent studies demonstrated that BTB-subtype E3 ubiquitin ligases play important roles in hypocotyl elongation, root development, flowering time control, phytohormone signaling transduction and abiotic stress responses. The BTB-subtype E3 ubiquitin ligases are involved in plant development through regulating ethylene signaling and biosynthesis pathways (Christians et al. 2009). BTB proteins ETO1 (ETHYLENE OVERPRODUCER 1), EOL1 (ETO1-like) or EOL2 regulate ethylene biosynthesis through promoting degradation of the type-2 ACSs (Christians et al. 2009). Consequently, the eto1, eol1 and eol2 mutants possess elevated levels of ACS5, overproduce ethylene and thus display constitutive ethylene response phenotypes including shorter and thicker hypocotyls (Christians et al. 2009). Recent investigation showed that the eto1-3 mutant possesses enhanced freezing tolerance due to increased CBF1/2/3 transcription levels (Catala and Salinas 2015).

The CUL3^{BPM(BTB/POZ–MATH)} E3 complex has been found to target a negative regulator of the ABA responses, ATHB6, to reduce plant growth and fertility, and affects stomatal behavior and responses to ABA (Lechner et al. 2011). Further investigation revealed CUL3^BPM as the substrate adaptor of this E3 complex in regulation of the fatty acid metabolism pathway, and the bpm mutant showed abnormal development and altered fatty acid accumulation in seeds (Hu et al. 2014). Recently, MYB56 was identified as the substrate of the CUL3^BPM E3 complex (Chen et al. 2015). The stability of MYB56 is controlled by the CUL3^BPM E3 complex, and MYB56 delays flowering, while BPM positively affects this process through mediating FT transcription (Chen et al. 2015).

The BTB-subtype E3 ligase complex is also involved in plant responses to vernalization, which initiates flowering through FRI (FRIGIDA), a scaffold protein involved in recruiting chromatin modifiers that regulate the transcription of flowering time control-related genes (Choi et al. 2011). A recent study showed that CUL3^{LRB1(LIGHT-RESPONSE BTB 1)} and CUL3^LRB2 target FRI for proteasomal degradation during vernalization, and the degradation of FRI further results in decreases of H3K4me3 (Histone H3Lys4 trimethylation) in FLC chromatin, and finally promotes flowering (Hu et al. 2014). This study uncovered the importance of FRI degradation via CUL3^LRB1 and CUL3^LRB2 E3 ligase complexes in the regulation of FLC transcription in response to vernalization.

In addition to the vernalization response, CUL3^LRB1 and CUL3^LRB2 E3 ligase complexes are also involved in the light signaling pathway. It was demonstrated that these E3 ligases promoted the proteasome-mediated degradation of phyB and phyD proteins in the light (Christians et al. 2012). Further study demonstrated that CUL3^LRB-mediated phyB degradation depends on the phosphorylation of PIF3 triggered by light and, interestingly, the CUL3^LRBs also promote both polyubiquitination and degradation of PIF3 (Ni et al. 2014). This excellent investigation uncovered a linked signal transmission and attenuation mechanism through mutually assured degradation of receptors and their signaling partners.

DDB-subtype E3s in plant development and abiotic stress responses

DDB-subtype E3 ligases are composed of CUL4 (Cullin 4), RBX1, DDB1 (DNA damage-binding 1) and DWD (WD40 domain-containing protein), in which the DWD protein is responsible for specific recognition of substrates (Vierstra 2009). There are about 82 DWD proteins in the A. thaliana proteome (Vierstra 2009). The ability of DDB1 to interact with different DWDs is essential for the diverse functions of these E3 complexes (Vierstra 2009, Biedermann and Hellmann 2011). Consequently, the potential DWD substrate receptors are a research hotspot, and many groups have demonstrated that DDB-subtype E3 ligases are involved in various developmental processes throughout the plant life cycle (Table 1).

A recent report showed that the DDB^{ABD1(ABA–hypersensitive DCAF1)} E3 ligase complex negatively regulates ABA signaling through promoting ABI5 degradation (Seo et al. 2014). Correspondingly, the abd1 mutant possesses a higher level of ABI5, and showed hypersensitivity to ABA during seed germination and seedling growth, enhanced stomatal closure, reduced water loss and finally increased drought tolerance (Seo et al. 2014). Recently, another DWD protein, ASG2 (ALTERED SEED GERMINATION 2), was also identified as an ABA signaling regulator through controlling the ERA1 (ENHANCED RESPONSE TO ABSCISIC ACID 1)-mediated pathway (Dutilleul et al. 2016). These studies revealed that DDB-subtype E3 ligases are involved in plant development and the water deficiency stress response through mediation of phytohormone signaling transduction.

DWD protein PRL1 (PLEIOTROPIC REGULATORY LOCUS 1) integrates sugar signaling, phytohormone biogenesis, isoprenoid metabolism and stress responses (Flores-Perez et al. 2010). Recently, PRL1 was also shown to be a modulator for root stem cell activity and meristem size through positively regulating WOX5 (WUSCHEL-related homeobox 5) and PLT (PLETHORA) transcription in the quiescent center (QC) (Ji et al. 2015). These findings revealed that PRL1 is a versatile factor that acts in diverse plant development periods and stress tolerances, possibly through affecting distinct substrates.

The abundance of ABI5 is regulated by a RING-type E3 ligase KEG (Stone et al. 2006). Further study showed that a DDB-subtype E3 ligase is also involved in controlling ABI5 stability (Lee et al. 2010). In A. thaliana, mutations in DWA1 (DWD hypersensitive to ABA 1) and DWA2 all display ABA- and salt-hypersensitive phenotypes. Biochemical analysis revealed that DWA1 and DWA2 interact with ABI5, and ABI5 accumulates to higher levels in dwa1 and dwa2 mutants (Lee et al. 2010). Actually, ABI5 is the best-studied key molecule in the ABA signaling pathway with respect to ubiquitination and two other post-translation modifications: phosphorylation and sumoylation (Yu et al. 2015). Overall, these studies revealed that, as a key component in the ABA signaling cascade, ABI5 has different regulators (Fig. 2).

APC-subtype E3s in plant development and abiotic stress responses

APC-subtype (anaphase-promoting complex) E3s regulate cell cycle progression from the metaphase to the S phase by targeted degradation of numerous cell cycle-related proteins (Peters 2006). They contain 11 or more subunits, including APC2 and APC11 (relative of Cullins), along with several interchangeable recognition subunits including CDC20 (cell division cycle protein 20), CDH1 (CDC20 homology 1) and APC10 (Vierstra 2009). The APC-subtype E3s were reported to control primarily the transitions during mitotic progression in yeast and animals (Huang and Bonni 2016); however, there is limited information regarding this subtype in plants (Table 1).

The A. thaliana APC1 subunit is critical for both female gametogenesis and embryogenesis (Wang et al. 2013). APC2 is encoded by a single-copy gene and interacts with another two subunits: APC11 and APC8. The apc2 mutant accumulates Cyclin B1 protein and shows megagametogenesis arrest (Capron et al. 2003). Two other subunits, CDC27A and CDC27B, show redundant but distinguishable functions during the gametogenesis stages. The cdc27a cdc27b double mutant gametes are non-viable (Perez-Perez et al. 2008). In addition, subunit APC4 plays critical roles in female gametogenesis and embryogenesis (Wang et al. 2012), and APC10 is essential for cell proliferation during leaf development (Eloy et al. 2011). Altogether, these studies indicated that any subunit of the APC E3 ligase complex is essential for normal plant development, such as gametogenesis, and it functions primarily through regulating cell cycle procession.

Although the biological functions of single subunits in the APC E3 ligase complex have been elucidated in recent years, identification of the substrates is a key step for critical understanding of the roles of the ligase. The APC^CCS52A E3 complexes possibly control meristem maintenance through repression of mitotic activity in the QC cells (Vanstraelen et al. 2009); however, their substrates remain elusive. APC^{TAD1 (Tillering and Dwarf 1)} promotes degradation of MOC1 (MONOCULM 1), a key positive regulator in the control of rice tiller number, in a cell cycle-dependent manner (Xu et al. 2012). Furthermore, MOC1 is also the substrate of another APC E3 ligase complex, APC^{TE (Tiller Enhancer)} (Lin et al. 2012), and TE also mediates the ABA–gibberellin antagonistic effect (Lin et al. 2015). These studies dissected the novel mechanism underlying the relationship between rice architecture and cell cycle regulation. However, compared with progress with other types of E3 ligases, the APC subtype need further investigation in the plant kingdom.

Perspectives for Future Research

It is well known that the ubiquitin–proteasome system (UPS) plays important roles in various plant growth stages and numerous plant abiotic stress responses. Dramatic advances in our understanding of the UPS were achieved especially in the first decade of the new millennium (Hua and Vierstra 2011). However, a large number of E3 ligases remain largely unknown, thus their characterization is an important and central task for the coming decades. At present, studies in this field remain at the ‘philatelic’ stage—any novel E3 ligase is a ‘stamp’, and investigators worldwide are focusing on their biological functions. Based on these studies, the deciphering of substrate(s) of single E3 ligases will help us to understand the more detailed and precise mechanisms of ubiquitination.

Unfortunately, the lack of efficient and high-throughput strategies to screen the substrates of E3 ligases is a key bottleneck. As described in this review, the substrate(s) of most E3 ligases remains elusive. Consequently, development of the corresponding protocols to uncover precisely the bona fide targets is a great future challenge. After studying individual E3 ligases, including analyzing their biological functions and substrates, constructing the regulatory networks that are joined by different signaling pathways is crucial to deepen our understanding of the cryptic roles of ubiquitination in plant development and stress responses.

Why does the UPS degradation pathway occur? The origin and evolution of the UPS including the different E1, E2 and E3 genes is another attractive research field. In this regard, the conservation and/or diversification of E3 ligases in different species are potential topics. For example, ortholog pairs such as PUB22/23 and CaPUB1, SDIR1 and OsSDIR1, AIP2 and OsDSG1, and OsGW2 and ZmGW2-CHR4/ ZmGW2-CHR5 possess conserved functions during evolution (Fig. 3); while there is functional divergence between SINAT5 and OsDIS1 ortholog pairs. Then, how does the other orthologs regarding the functions divergence? In contrast, one E3 ligase can possess different biological functions, possibly through regulating distinct substrates. For example, SINAT5 controls lateral root production and flowering time via controlling different target proteins; while SPL11 is involved in both the stress responses and transition from vegetative to reproductive phases (Table 1). It was also noted that three distinct E3 ligases are responsible for ABI5 turnover (Fig. 2)—how and when do plants choose a particular E3 ligase to regulate ABI5 degradation?

Ubiquitin mmolecules perform their functions circularly in UPS processes (Fig. 1)—how and when are inactivated ubiquitins degraded in vivo? Another interesting topic is the number of ubiquitin molecules in polyubiquitin chains, which is a key factor determining the fate of the targets (Petroski and Deshaies 2003, Vierstra 2009, Vierstra 2012). Why do degradation pathways use the number of ubiquitin molecules as a signature? Why do the chains extend beyond four ubiquitins? The energetics and topology of polyubiquitin chains will be hotspots in further research, especially in structural biology. Finally, the non-proteolytic roles of E3 ligases, such as HOS1, during plant developmental stages and stress responses will be an interesting topic in this field.

Funding

This work was supported by the China Postdoctoral Science Foundation [2014M552377 and 2016T90868]; the Natural Science Foundation of China [31071373]; and the Department of Education Sichuan Province [16ZB0040].

Glossary

Abbreviations

APC

anaphase-promoting complex

BTB

Bric-a-brac–Tramtrack–Broad complex

CRL

Cullin-RING box1-Ligase

CUL1

cullin 1

DDB

DNA Damage-Binding domain-containing

E1

ubiquitin-activating enzyme

E2

ubiquitin-conjugating enzyme

E3

ubiquitin ligase

FT

FLOWERING LOCUS T

HECT

Homology to E6-AP C Terminus

iTRAQ

isobaric tags for relative and absolute quantitation

JA

jasmonate

LD

long day

PUB

Plant U-box protein

RBX1

RING-Box 1

RING

Really Interesting New Gene

ROS

reactive oxygen species

SCF

S phase kinase-associated protein 1–Cullin 1–F-box

SD

short day

SKP1

S phase Kinase-associated Protein 1

SL

strigolactone

UEV

ubiquitin enzyme variant

UPS

ubiquitin–proteasome system

Disclosures

The authors have no conflicts of interest to declare.