Background

Protein interactome analysis enables the comprehensive identification of protein complexes and protein modifications, thus hel** us to understand specific biological mechanisms, elucidate cellular functions, and decipher genotype–phenotype relationships [1, 2]. The yeast two-hybrid (Y2H) system has been widely used to analyze direct protein–protein interactions and to provide high-quality binary interactomes [3]. It has been used to develop high-throughput human binary protein–protein interactome maps, from the first-generation maps comprising 2754 pairwise interactions in 2005 [4, 5] to the third-generation maps comprising 53,000 pairwise interactions in 2020 [1].

Protein ubiquitination is required for a plethora of cellular processes in eukaryotes, including proteasome-dependent protein degradation, cell cycle progression, transcriptional regulation, and signal transduction [6] and is sequentially catalyzed by three different types of enzymes: the ubiquitin-activating enzyme E1, the ubiquitin-conjugating enzyme E2, and the ubiquitin ligase E3. E1 catalyzes the ATP-dependent activation of ubiquitin and the formation of a thioester bond with ubiquitin. E2 binds to the activated ubiquitin and transfers it to its substrate via an E3 ligase [6]. E3 ligases, the most heterogeneous of these enzymes, mediate substrate specificity. Their direct physical interactions with substrates determine the modification mode of the substrates. E3 ligases can be divided into single- and multi-subunit types. The single-subunit type includes HECT (homologous to the E6AP carboxyl terminus), RING (really interesting new gene), and U-box, while the multi-subunit type includes SCF (SKP1–CUL1–F-box), CUL3–BTB, CUL4–DDB1–DWD, and APC/C complex (anaphase-promoting complex/cyclosome), in which F-box, BTB, DWD, and APC co-activator subunits determine substrate specificity [7]. The identification of E3 ligases and their substrates is critical for understanding the protein ubiquitination mediated biological processes.

In the past decade, significant progress has been made in identifying ubiquitinated proteins in animals using a monoclonal antibody that specially recognizes the putative ubiquitination sites diglycine-modified lysines (K-Ɛ-GG) [8,9,10]. Using this antibody, over 20,000 distinct endogenous ubiquitination sites were identified in human cells [11], and 1543 putative ubiquitinated proteins were identified in rice (Oryza sativa L.) leaves with and without treatment with inducers of plant defense responses [14]. These approaches have led to the identification of a huge number of putative ubiquitinated proteins in both animals and plants; however, their cognate E3s remain largely unknown.

Plant genomes encode approximately 1500 E3s; this family has expanded by more than two-fold compared to E3s in mammals and other species [6, 15]. This observation suggests that E3 ligases may be involved in regulating many more biological processes in plants than in other species. Although Y2H screening has been widely used to identify E3–substrate pairs in plants, the efficiency of screening E3 ligase genes is low for several reasons. First, two-thirds of the clones from a conventional cDNA library are not amenable to fusion in frame with the N-terminal GAL4 activation domain required to validate interactions [16]. Second, some E3 ligase genes are expressed only in specific tissues or developmental stages or under some stress conditions [17]. Therefore, a complete E3 ubiquitin ORFeome library is essential for analyzing ubiquitination interactome in plants.

Rice is an important food crop and a model monocot plant [18]. In the current study, a ubiquitin E3 ligase gene (UbE3) library covering 98.94% of the 1515 E3 ligase genes in rice was generated. In addition to the known E3s of four substrates, several new E3s were identified by using the UbE3 library. These interactions and substrate degradation were confirmed by in vitro and in vivo assays. Furthermore, when the phenylalanine ammonia lyases OsPAL1–OsPAL9 were used as baits, only the F-box-type E3 ligase OsFBK16 interacted with OsPAL1–OsPAL7. We further verified that OsFBK16 degrades OsPAL1, OsPAL5, and OsPAL6 in vivo and demonstrated that overexpression of OsPAL1 and OsPAL6 in rice as well as loss-of-function of OsFBK16 enhanced rice blast disease resistance. Thus, our UbE3 library provides a powerful proteomic resource for the global identification of E3 ligases and analysis of ubiquitination interactome and biological networks in plants.

Results

Putative ubiquitinated proteins in rice and annotation of the ubiquitin E3 ligases

We previously identified 1543 proteins containing ubiquitinated sites in two studies [Yeast two-hybrid (Y2H) screening

Equal amounts of plasmid DNA harboring each E3 ligase gene were mixed well and transformed into yeast strain Y187. The yeast cells were spread on SD-Leu plates, cultured for 3 days, harvested, and mixed. Glycerol was added to the solution to a concentration of 15% for long-term storage at − 80 °C.

Before the screening, auto-activation of each pGBKT7-bait construct was determined by co-transforming yeast strain AH109 with the empty vector pGADT7. The lack of growth of a yeast culture on an SD-Leu-Trp-His-Ade plate indicated that no autoactivation occurred and the culture could be used for screening. We mated yeast strain AH109 carrying the pGBKT7-bait construct with strain Y187 containing plasmids with all the E3 ligase genes in the pGADT7 background at 30 °C with gentle shaking at 37 rpm for 20–24 h. The culture was checked under a microscope until a 3-lobed structure or a shape resembling “Mickey Mouse” appeared. Following centrifugation and re-suspension, the culture was spread onto SD-Leu-Trp-His-Ade plates and incubated at 30 °C for 3–10 days. The clones were picked, transferred to new SD-Leu-Trp-His-Ade plates, and incubated at 30 °C for 3 more days to further confirm the positive interaction. For PCR amplification, individual yeast clones were picked, transferred into ddH2O, and quickly lysed using liquid nitrogen. PCR was performed using lysed yeast as templates, the pGADT7 vector primers AD-F: CTATTCGATGATGAAGATACCCCACCAAACC and AD-R: GTGAACTTGCGGGGTTTTTCAGTATCTACGATT; the PCR products were purified and subjected to sequencing.

Co-immunoprecipitation (Co-IP) assay

Co-IP assays were carried out by agro-infiltration of 4-week-old N. benthamiana leaves. Agrobacterium cultures carrying plasmids harboring GFP-tagged OsSKIPa, OsNRPD1aC, NRR, rTGA2.1, OsPAL1, and their cognate E3 ligase genes fused with HA tag were mixed and co-infiltrated into N. benthamiana leaves. Samples were collected at 72 h after agroinfiltration, and total protein was extracted as previously described [48]. Immunoprecipitation was performed with anti-GFP antibody (MBL, D153-11) or anti-HA antibody and protein G agarose beads. Immunoblotting was performed using an anti-GFP antibody (MBL, 598-7) or anti-HA antibody (MBL, M18907).

Protein degradation assay in planta

Protein degradation experiments were performed via transient protein expression in N. benthamiana leaves [49, 50] and rice protoplasts. Agrobacterium cultures carrying plasmids harboring GFP-tagged OsSKIPa, OsNRPD1aC, NRR, rTGA2.1, OsPAL1, and their cognate E3 ligase genes fused with HA tag were mixed and co-infiltrated into N. benthamiana leaves. After 48 h, 50 μM MG132 (Millipore) or an equal volume of DMSO solution was infiltrated in the leaves, which were collected for protein extraction 24 h after MG132 treatment. For the protein degradation assay in rice protoplasts, plasmids harboring GFP-tagged NRR or rTGA2.1 and their cognate E3 ligase genes fused with HA tag as well as the control GUS-HA construct were co-expressed in Nipponbare protoplasts. After 16 h, 50 μM MG132 or an equal volume of DMSO solution was added, and 4 h later, the protoplasts were collected for protein extraction. Protein abundance was detected by immunoblotting using anti-HA or anti-GFP antibody. ACTIN (Abmart, M20009L) or HSP was used as an internal protein control. OsSKIPa, OsNRPD1aC, or OsPAL1 transcript levels were measured by RT-PCR, and ACTIN was used as the internal control.

E3 ligase activity and ubiquitination assay in vitro

The full-length CDS of OsRING77, OsRING113, OsRFPH2-10, P3IP1, and OsPUB46 were individually fused to the C-terminus of MBP in the pMal-C2X vector, the full-length OsRING116 was fused to the C-terminus of GST in the pGEX6p-1 vector, and the fusion proteins were expressed in Escherichia coli strain BL21. For RING-type E3 ligases, ubiquitination reaction mixtures containing E1 (wheat E1), E2 (AtUBC8), 2 μg/μL ubiquitin (U-100At, Boston Biochem, USA), and purified MBP-OsRING77, MBP-OsRING113, or MBP-OsRFPH2-10 were mixed in 1× reaction buffer (50 mM Tris-HCl, pH 7.4, 10 mM MgCl2, 5 mM ATP, and 2 mM DTT). The reactions were incubated at 30 °C for 2 h, and in vitro E3 ligase activity was determined using an anti-Ub antibody (Millipore, 05-944) and anti-MBP antibody (Abbkine, A02070-2) [51]. For U-box-type E3 ubiquitin ligase, purified MBP-P3IP1 or MBP-OsPUB46 was individually pre-incubated in total rice extracts prior to the assay [52]. For the substrate ubiquitination assay, equal amounts of purified GST-OsSKIPa or GST-OsNRPD1aC were added to the reaction mixture, and ubiquitination was measured using anti-Ub, anti-MBP, or anti-GST (BPI, AbM59001-2H5-PU) antibodies.

Rice transformation and M. oryzae inoculation

The full-length coding fragments of OsPAL1 and OsPAL6 were amplified from the rice cultivar NPB and inserted into the pRHV-cGFP vector driven by the maize ubiquitin promoter to generate the OsPAL1 and OsPAL6 overexpression constructs [53]. The generated constructs were introduced into the calli of NPB via Agrobacterium tumefaciens-mediated transformation as described previously [54]. The osfkb16 mutants were generated via CRISPR-Cas9 technology [55]. OsPAL1 and OsPAL6 overexpression lines were identified by quantitative real-time polymerase chain reaction (qRT-PCR) and the mutations in the osfkb16 mutants were analyzed by sequencing.

For punch inoculation with M. oryzae, isolate RB22 was cultivated on an oat medium in darkness for 1 week at room temperature and then moved to light for spore induction. After 7–10 days, spore suspension (5 × 105 spores/mL) of RB22 in 0.025% (v/v) Tween 20 was used for punch inoculation on the second leaf (from the top) of 6-week-old plants as previously described [56]. Disease symptoms on leaves were scored 14 days after inoculation. Relative fungal biomass was calculated by measuring the expression of the M. oryzae MoPot2 with the DNA-based quantitative PCR assay.

Accession numbers of the genes used in this study

The following are the accession numbers of the genes used in this study: OsFBX466 (LOC_Os02g38499), OsFBX55 (LOC_Os02g38589), OsFBO24 (LOC_Os12g05609), OsFBX481 (LOC_Os12g05709), OsRING62 (LOC_Os02g35347), OsRING66 (LOC_Os02g35365), OsRING202 (LOC_Os11g04280), OsRING203 (LOC_Os11g04281), OsRING77 (LOC_Os02g19140), OsRING113 (LOC_Os03g26370), OsRING171 (LOC_Os01g58400), OsRING394 (LOC_Os12g04590), OsRING199 (LOC_Os11g04680), OsFBX503 (LOC_Os07g17570), OsRING344 (LOC_Os08g42640), OsRING336 (LOC_Os05g41520), OsRING176 (LOC_Os01g58780), OsRING337 (LOC_Os09g12720), OsRING375 (LOC_Os04g22240), BTBT3 (LOC_Os11g37520), HBTB8 (LOC_Os12g08720), OsPUB28 (LOC_Os01g67500), OsPUB46 (LOC_Os04g34140), OsPUB49 (LOC_Os10g41220), OsPUB69 (LOC_Os08g13780), OsFBX68 (LOC_Os02g56810), OsFBX82 (LOC_Os03g20500), OsFBX389 (LOC_Os10g35920), BTBZ1 (LOC_Os01g66890), MBTB47 (LOC_Os10g29180), OsRING116 (LOC_Os01g38700), and OsFBK16 (LOC_Os06g39370).