Functions of COP1/SPA E3 Ubiquitin Ligase Mediated by MpCRY in the Liverwort Marchantia polymorpha under Blue Light

Abstract

COP1/SPA1 complex in Arabidopsis inhibits photomorphogenesis through the ubiquitination of multiple photo-responsive transcription factors in darkness, but such inhibiting function of COP1/SPA1 complex would be suppressed by cryptochromes in blue light. Extensive studies have been conducted on these mechanisms in Arabidopsis whereas little attention has been focused on whether another branch of land plants bryophyte utilizes this blue-light regulatory pathway. To study this problem, we conducted a study in the liverwort Marchantia polymorpha and obtained a MpSPA knock-out mutant, in which Mpspa exhibits the phenotype of an increased percentage of individuals with asymmetrical thallus growth, similar to MpCRY knock-out mutant. We also verified interactions of MpSPA with MpCRY (in a blue light-independent way) and with MpCOP1. Concomitantly, both MpSPA and MpCOP1 could interact with MpHY5, and MpSPA can promote MpCOP1 to ubiquitinate MpHY5 but MpCRY does not regulate the ubiquitination of MpHY5 by MpCOP1/MpSPA complex. These data suggest that COP1/SPA ubiquitinating HY5 is conserved in Marchantia polymorpha, but dissimilar to CRY in Arabidopsis, MpCRY is not an inhibitor of this process under blue light.

1. Introduction

Light is an important energy source and exogenous signal that regulates plant growth and development [1]. To sense and respond to light signals of different wavelengths, plants have evolved specific photoreceptors, including blue-light receptor cryptochromes and red-/far-red light receptor phytochromes [2,3]. Under activation of light, these photoreceptors would immediately inhibit CONSTITUTIVELY PHOTOMORPHOGENIC1/SUPPRESSOR OF PHYA-105 (COP1/SPA) complex [4,5], a light signal repressor, and thus activate the downstream pathway of light signal. In Arabidopsis, COP1/SPA complex mediates the ubiquitination and degradation of various light responsive transcription factors to inhibit the light signal regulation in the dark, which include ELONGATED HYPOCOTYL5 (HY5) regulating the de-etiolation of plants [6], CONSTANS (CO) regulating the photoperiodic flowering [7], and PRODUCTION OF ANTHOCYANIN PIGMENT (PAP) regulating the anthocyanin production [8]. COP1 also exists in humans to inhibit tumorigenesis by stabilizing specific transcription factors such as P53 and cJun in a light-independent manner [9,10]. However, SPA proteins are specific to plants and regulate the activity of COP1 in a light-dependent manner [11].

Under blue light (450–500 nm of the visible spectrum), the activity of COP1/SPA is inhibited by photoreceptors including cryptochromes, FKF1 of ZEITLUPE family, and Phytochrome A [5], and such regulation is conducted in five different approaches: (1) blue light induces the nuclear exclusion of COP1, thereby separating COP1 from its target protein [12]; (2) SPA proteins degrade under blue light [13]; (3) blue light activates photoreceptors to bind to SPA1, causing the dissociation of SPA1 and COP1 [14]; (4) COP1 homodimerization would be disrupted by blue light, and thus inhibiting COP1 activity [15]; and (5) blue light activates photoreceptors to inhibit COP1 binding to target proteins through VP domain-mediated competition [16]. Cryptochromes, as typical blue-light receptors, play an important role in blue light inhibition of COP1/SPA activity. In Arabidopsis, AtCRY1 is triggered under blue light to induce the nuclear exclusion of AtCOP1, relying on AtSPA1 [17,18]. The activated AtCRY1 would interact with AtSPA1 to inhibit the interaction of AtSPA1 and AtCOP1, leading to the reduced ubiquitination activity of AtCOP1 [14]; while the interaction of AtCRY2 and AtSPA1 would not reduce that of AtSPA1 and AtCOP1 but would strengthen AtCRY2 interacting with AtCOP1 [19]. Previous studies have demonstrated that the VP domain of AtCRY2 binds to the WD40 domain of AtCOP1, thereby inhibiting AtCOP1 binding to target proteins [3,16]. It has been speculated in a recent study that in Physcomitrella patens, a kind of non-vascular PpCOP1 ubiquitinates PpHY5 independent of PpSPAs under dark conditions [20], but how blue light regulates COP1/SPA in non-vascular plants has not been studied.

The bryophyte Marchantia polymorpha is drawing attention as a new model system in the study of light signals, the genome of which contains typical photoreceptor genes and most of the downstream light signal genes [21]. Previous studies have shown the functions of the single-copy phytochrome ortholog MpPHY, the ultraviolet-B photoreceptor MpUVR8, and the blue light receptors MpPHOT and MpFKF in M. polymorpha [22,23,24,25,26]. Unfortunately, few researchers have addressed the functional differences and commonalities of COP1/SPA complex between vascular plant Arabidopsis thaliana and the bryophyte M. polymorpha. In addition, it is also an important question worth studying whether blue light regulates COP1/SPA activity in M. polymorpha.

The M. polymorpha genome contains only one CRY gene (MpCRY, Mp2g17590), one SPA gene (MpSPA, Mp3g25460), and one COP1 gene (MpCOP1 Mg5g12010). In comparison, there are four paralogs of SPAs in Arabidopsis genome and nine paralogs of COP1 in Physcomitrella. Therefore, the functional study on COP1/SPA of liverworts is comparably easy to implement. In this study, we obtained Mpspa1 knockout mutants, and elaborated the function of MpCOP1/MpSPA complex and the effect of blue light receptor MpCRY on its function through the phenotype under blue light and specific biochemical evidence.

2. Results

2.1. MpSPA Is Associated with MpCRY to Regulate the Thallus Symmetry of M. polymorpha under Blue Light

The M. polymorpha genome contains a SPA ortholog MpSPA (Figure S1a), owning three domains as AtSPA1 [27], which are N-terminal kinase-like domain, coiled-coil (CC) domain, and C-terminal WD40 domain (Figure S1a). To analyze the function of MpSPA in M. polymorpha under blue light, we used CRISPR/Cas9 technology to target the first exon of MpSPA and obtained two gene knockout mutant lines Mpspa-8 and Mpspa-11, among which 5 bases were deleted in Mpspa-8 and 11 bases deleted in Mpspa-11 (Figure 1a,c). We also obtained Mpcry-7 and Mpcry-8, two knock-out mutant lines of MpCRY. Mpcry-7 and Mpcry-8 have a 26-bp deletion and a 1-bp insertion in the 4th exon of MpCRY, respectively (Figure 1b,d). The gemmalings of WT, Mpspa mutant, and Mpcry mutant were exposed to both blue and red light for 14 days. With continuous illumination of red light, there were no significant growth differences among WT, Mpspa and Mpcry. Interestingly, under blue light, 57% individuals in Mpcry-7, 58% in Mpcry-8, 60% in Mpspa-8, and 59% in Mpspa-11 showed asymmetric growth of thallus (the growth from only one side of the gemma, hereafter referred to as ‘asymmetric growth’) under blue light, while less than 5% of individuals in WT showed asymmetric growth (Figure 1f,g).

To verify the regulation of MpCRY to MpSPA under blue light, we used CRISPR/Cas9 technology to obtain two double-knockout mutant lines Mpcryspa-6 and Mpcryspa-13 (Figure 1e). We defined the symmetric percentage as the number of the symmetrically growing thallus in 100 thalli. We found that after 14 days of continuous blue light irradiation, the symmetric percentage of Mpcryspa1-6 and Mpcryspa1-13 was 47.3% and 48.7%, which is not significantly different from that of Mpcry and Mpspa single mutant (Figure 1f,g). These results suggested that MpSPA, as MpCRY, could promote the symmetric growth of thallus under blue light, regulating the symmetric growth of thallus in with the identical signaling pathway. However, CRYs inhibit SPA1 in Arabidopsis under blue light, while the thallus of Mpcry and Mpspa1 mutants in M. polymorpha shared the identical asymmetry ratio under blue light. Therefore, we speculated that MpCRY did not inhibit the activity of MpSPA under blue light.

AtCRY2 protein is rapidly degraded by the 26S proteasome system in response to blue light [28,29,30], whereas AtCRY1 is stable. In addition, AtSPA1 and AtSPA2 also degrade in blue light [13]. In our study, MpCRY degraded in blue light and phosphorylated like Arabidopsis CRY2 (Figure S2a), while MpSPA accumulated under the same conditions (Figure S2b), which is different from AtSPA1 and AtSPA2 in Arabidopsis thaliana. Based on the above results, we speculated that MpCRY possesses partial characteristics of AtCRY2, but the stability of MpSPA protein is affected by blue light, which is different from AtSPA1.

2.2. MpSPA Interacts with MpCRY in the Nucleus

In Arabidopsis, CRY1 and CRY2 interacts with SPA1 in a blue-light-dependent way through C-terminal extension (CCE) domain [14] and N-terminal photolyases homology region (PHR) domain [19], respectively. Similarly, MpCRY possesses N-terminal PHR domain and C-terminal CCE domain as well (Figure 2a). To identify the interaction of MpCRY and MpSPA as well as the interdependent domains achieving such interaction, we first examined whether MpSPA interacts with MpCRY in a blue-light-dependent manner in yeast cells. Yeast two hybrid assay showed that the full-length MpSPA and 607 residues’ lack of the WD40 domain in MpSPA N-terminal fragment could interact with the full-length MpCRY and CCE domain of MpCRY, but the kinase domain, helix-loop-helix domain, and WD40 domain of MpSPA failed to interact with MpCRY (Figure 2b), suggesting that MpCRY interacts with the kinase domain and helix-loop-helix domain of MpSPA, but not necessarily with WD40 domain of MpSPA.

The full-length MpSPA did not interact with the PHR domain of MpCRY, although the N-terminal 607 residues of MpSPA without the WD40 domain interacted with the C-terminal PHR domain of MpCRY in yeast (Figure 2c). Based on these results, we suggested that the interaction-dependent domain of MpCRY and MpSPA1 is the same as that of AtCRY1 and AtSPA1. Moreover, MpCRY and MpSPA generated constitutive interaction in yeast under blue light and dark conditions, which is interestingly different from the blue light-specific interaction of AtCRYs and AtSPA1 in yeast two hybrid assays [14,19]. Consistent with this result, we also found that MpCRY interacted with MpSPA independently of blue light, using immunoprecipitation in HEK293T cells co-expressing Myc-MpCRY and Flag-MpSPA (Figure 2e,f) [31].

In addition, the interaction of MpCRY and MpSPA were detected through Fluorescence Complementation (BiFC) within a transient cell in Arabidopsis, the result of which showed that MpCRY and MpSPA interacted in the nucleus (Figure 2d). We then obtained the M. polymorpha plants individually expressing and co-expressing _pro35S::Citrine-MpSPA and _pro35S::MpCRY-Tdtomatod to investigate the subcelluar localization of MpSPA and MpCRY in liverworts. It was shown that MpSPA was localized exclusively to the nucleus and formed speckles under blue and dark conditions (Figure S3a). MpCRY-Tdtomato is localized to the nucleus and cytoplasm (Figure S3b) [32]. Similar to Arabidopsis AtCRY1 but different from Arabidopsis AtCRY2 which forms photobodies under blue light, individually overexpressed MpCRY-Tdtomato was not able to form photobodies under blue light. In this study, the co-expressed MpCRY and MpSPA were found to locate in the nucleus, in which MpSPA did not constitute speckles in the dark, while MpSPA and MpCRY jointly form photobodies in blue light (Figure 2g,h), showing a high affinity with AtCRY1 which cannot form photobodies when expressed alone (Figure S4), but do generate photobodies when co-expressed with AtSPA1 [33]. Based on the above results, we speculated that MpCRY possessed some properties of AtCRY1 and may form photobodies under blue light to regulate signals when working with MpSPA.

2.3. MpSPA Interacts with MpCOP1 in the Nucleus

SPAs interact with COP1 to form a tetramer to enhance the activity of COP1 to ubiquitinate target proteins [34]. MpCOP1 (Mg5g12010) is the ortholog of AtCOP1 in the genome of M. polymorpha that also contains a RING-finger domain in the N-terminal binding to ubiquitin conjugating enzyme E2, a coiled-coil (CC) domain in the middle to constitute homodimerization or to form heterodimerization with SPA1, and a WD40 domain in the C-terminal to bind the target proteins and photoreceptors (Figure 3a) [11]. We then used the yeast two-hybrid method to detect the interaction between MpSPA and MpCOP1, finding that full-length MpSPA interacted with full-length MpCOP1 (Figure 3b). Moreover, the coiled-coil domains of MpCOP1 and MpSPA interacted with each other, indicating that, just like AtCOP1-AtSPA1, the interaction of MpCOP1 and MpSPA is performed through their respective coiled-coil domain (Figure 3b). Therefore, the 607 residues at the N-terminal of MpSPA with the deletion of WD40 domain can also interact with the 278 residues at the N-terminal of MpCOP1 without WD40 domain (Figure 3b).

In addition, the yeast co-expressing the kinase domain of MpSPA coupled with the DNA binding domain (BD) of GAL4 and the isolated transcription activation domain (AD) of GAL4 grew on His- medium, indicating the transcriptional activation activity of MpSPA kinase domain (Figure 3b). Using immunoprecipitation assay to co-express Flag-MpSPA and GFP-MpCOP1 in HEK293T cells, we again demonstrated the interaction between MpSPA and MpCOP1 (Figure 3c), which is consistent with our findings that MpSPA and MpCOP1 interacted in the nuclei of the transiently transfected Arabidopsis protoplasts with the formation of speckles (Figure 3d,e). Taken together, our results indicate that MpSPA interacts with MpCOP1 in the nucleus, and it is speculated that MpSPA and MpCOP1 in M. polymorpha form the E3 ligase complex to ubiquitinate the target protein, as the AtCOP1/AtSPA1 complex. We also tried to generate MpCOP1 knock-out lines by CRISPR/Cas9. As a previous study has shown [34], these lines were lethal (data not shown).

2.4. MpCOP1/MpSPA Complex Interacts with MpHY5

HY5 is an important target protein of COP1/SPA ubiquitin ligase in Arabidopsis thaliana, which could interact with AtSPA1 and AtCOP1. In this study we focused on MpHY5 (Mp1g16800), the ortholog of AtHY5 in M. polymorpha, and aimed to figure out its interaction with MpSPA and MpCOP1. Yeast two-hybrid results showed that the coiled-coil domain of MpSPA interacted directly with the C-terminal of MpHY5, different from that of AtSPA1 and AtHY5 in Arabidopsis which requires both the WD40 domain and the coiled-coil domain (Figure 4a) [6]. Moreover, we used immunoprecipitation of HEK293T cells co-expressing GFP-MpHY5 and Flag-MpSPA and found that GFP-MpHY5 and Flag-MpSPA interacted in HEK293T cells in which the interaction of MpHY5 and MpCOP1 similarly appeared (Figure 4c–e). To investigate the ability of MpHY5 to regulate symmetric growth of M. polymorpha thallus, we utilized CRISPR/Cas9 knockout mutant Mphy5-6 and transgenic M. polymorpha (proEF1::Mphy5-Citrine/WT) overexpressing cDNA of MpHY5-Citrine in wild type plants. These plants were exposed to blue light for 14 days and did not exhibit the similar phenotype of Mpcry and Mpspa that the percentage of thallus with asymmetric growth increases (Figure 4f,g), suggesting that MpHY5 do not regulate the symmetric growth of thallus.

2.5. MpSPA Promotes That MpCOP1 Ubiquitinates MpHY5

To verify the participation of MpSPA to MpHY5 degradation, we carried out an in vitro assay [35] to detect the effect of MpSPA on the stability of MpHY5 proteins by incubating purified recombinant His-GFP-MpHY5 proteins with tissue extracts from the Mpspa mutant and WT plants. Not surprisingly, the recombinant MpHY5 proteins were much more sensitive to the seedling extract of the WT plants, compared to the seedling extract of the Mpspa, and degraded much more quickly (Figure 5a,b). It showed that MpSPA is required for the ubiquitination and degradation of MpHY5 under dark.

SPA1 could promote COP1 activity in ubiquitination of the target proteins in Arabidopsis [36]. To verify whether MpSPA possess such ability, we used HEK293T cells to analyze the regulation of MpSPA and MpCOP1 on MpHY5 ubiquitination. The immunoprecipitation-enriched GFP-MpHY5 showed weak ubiquitination when co-expressed with MpCOP1, but had a noticeably strong ubiquitination when co-expressed with MpSPA and MpCOP1 in HEK293T cells (Figure 5c,d). Thus, we wondered whether MpSPA enhancing the ubiquitination of MpHY5 originated from MpSPA promoting the interaction between MpCOP1 and MpHY5. To this end, we carried out co-immunoprecipitation using HEK293T cells co-expressing GFP-MpHY5, Flag-MpSPA, and Myc-MpCOP1. The results showed that MpSPA did not promote the interaction between MpCOP1 and MpHY5 (Figure 4d). In summary, our results indicated that MpCOP1 could ubiquitinate MpHY5, and MpSPA would promote such ubiquitination.

2.6. Blue Light Increases MpHY5 Abundance Independently of MpCRY

To investigate whether blue light affects the stability of MpHY5, we prepared the transgenic plants expressing the MpHY5 coding region fused with Citrine open reading frame, driven by the M. polymorpha ELONGATION FACTOR1a (Mp-EF1) promoter [37]. We transferred _proEF1::MpHY5-Citrine/WT transgenic gemmalings, pre-treated in the dark for 2 days, to blue light for detecting the levels of MpHY5 proteins. As observed in Arabidopsis for the AtHY5 protein [14], blue light increases the abundance of MpHY5 proteins (Figure 6a,b). In Arabidopsis, blue-light activated AtCRY1 suppresses the AtSPA1–AtCOP1 interaction and AtCOP1-dependent degradation of the transcription factor AtHY5 [14]. To investigate whether MpCRY mediates the accumulation of MpHY5 proteins, we generated the transgenic plants _proEF1::MpHY5-Citrine/Mpcry. We found that in Mpcry mutants, MpHY5 can also accumulate as in the WT plants (Figure 6a,b). This result suggested that the accumulation of MpHY5 is not dependent on MpCRY under blue light and some other blue light receptors may control this process. In Arabidopsis, AtCRY1 interacts with AtSPA1 to suppress AtSPA1-AtCOP1 interaction. To test this function of MpCRY, we performed a Co-IP assay through co-expressing GFP-MpSPA, Myc-MpCRY, and Flag-MpCOP in HEK293T cells. Different from AtCRY1, MpCRY1 failed to inhibit the interaction of MpSPA and MpCOP1 (Figure 6c,d). Because both MpCRY and MpHY5 interact with the coiled-coil domain of MpSPA, we speculated that MpCRY can act as the competitive inhibitor of the MpSPA-MpHY5 interaction. To prove this speculation, Flag-MpSPA, Myc-MpCRY, and GFP-MpHY5 were co-expressed in HEK293T cells. The result of the Co-IP assay showed that Myc-MpCRY weakly inhibited the interaction between Flag-MpSPA and GFP-MpHY5 (Figure 6e,f). To further investigate whether MpCRY suppresses the activity of MpCOP1/MpSPA complex in ubiquitination of MpHY5, we conducted an IP assay using HEK293Tcells co-expressing GFP-MpHY5, Flag-MpSPA, Flag-MpCOP1, and Myc-MpCRY and found that MpCRY failed to inhibit the activity of MpCOP1/MpSPA complex in ubiquitination of MpHY5 (Figure 5c,d). In summary, blue light increases MpHY5 abundance independently of MpCRY.

3. Discussion

Cryptochromes are conserved flavoprotein receptors identified throughout the biological kingdom with diversified roles in organism development and entrainment of the circadian clock [38,39,40]. Although a myriad of studies have reported a variety of regulatory mechanisms for blue light to regulate plant growth and development, cryptochromes are mainly found to regulate the photoperiod flowering time and hypocotyl elongation in plant lineage [39,41,42]. We have reported a classic regulatory signaling pathway explaining the mechanism of cryptochromes regulating the hypocotyl elongation in Arabidopsis, in which cryptochromes inhibit the activity of COP1/SPA1 E3 ubiquitin ligase to regulate HY5 ubiquitination and hypocotyl elongation [14]. Subsequently, the signaling pathway of cryptochrome regulating COP1 complex has been found in various plants [43,44]. Since cryptochromes and their SPA/COP1/HY5 pathway are so important and widely conserved in plant kingdom, in this study, we raised an intriguing question of whether MpCRY and its signaling pathway are conserved or diversified in the bryophyte plant without the flower and the typical hypocotyl, which is different from higher plant species.

Compared to the model plant Arabidopsis which possesses homologous genes, e.g., AtSPA1, AtSPA2, AtSPA3, and AtSPA4 [45], the M. polymorpha genome contains single MpCRY, MpSPA, MpCOP1, and MpHY5. Without the typical hypocotyl, the Mpcry mutant exhibited a novel phenotype differing from the Atcry1cry2 mutant, which decreased the percentage of asymmetric thalli under blue light. Similar to the Mpcry mutant, Mpspa mutant and Mpcryspa double mutant showed the reduction of asymmetric thalli (Figure 1f,g), suggesting that MpSPA and MpCRY are in the same signaling pathway in regulating the symmetric growth of thallus under blue light. Interestingly, although MpCRY and MpSPA are in the same signaling pathway, which is similar to AtCRY1 and AtSPA1, the molecular functions are different from those of Arabidopsis. In Arabidopsis, AtCRY1 blue light specifically inhibits the function of AtSPA1, thus under blue light, the hypocotyl of the Atcry1 mutant becomes longer [14]. However, in M. polymorpha, the interaction between MpSPA and MpCRY was independent of blue light (Figure 2b,c,e,f), which implied that MpCRY transduces its blue light signal via additional proteins or other manners except for protein–protein interaction in M. polymorpha. As we reported previously, AtCRY2 forms blue light specific photobodies after blue light irradiation and the light-activated AtCRY2 has also been shown to co-localize in nuclear bodies with AtCOP1 and AtSPA1 [19]. We thus investigate whether MpCRY could form the photobody and even transduce the signal to proteins co-localizing in the same photobody. Notably, MpCRY could not form photobodies under blue light as AtCRY2, but when co-expressing with MpSPA, MpCRY can form photobodies with MpSPA under blue light (Figure 2f). We speculated that under blue light, MpCRY and MpSPA could regulate the symmetrical growth of thalli by forming blue light specific photobodies rather than that AtCRYs blue light specifically interact with SPA1 to regulate the hypocotyl elongation in Arabidopsis.

In addition, as we reported previously, AtCRYs mediated blue light to inhibit the function of the AtCOP1/AtSPA1 complex and the ubiquitination of AtHY5 [14]. However, in M. polymorpha, MpCRY did not inhibit the interaction of MpCOP1 and MpSPA. Although MpCRY weakly inhibited the interaction between MpSPA and MpHY5 (Figure 6c), it did not affect the ubiquitination of MpCOP1/MpSPA on MpHY5, which suggested that blue light does not inhibit the activity of MpCOP1/MpSPA through MpCRY in M. polymorpha, unlike in Arabidopsis. This study exhibited the novel phenotype of plant cryptochrome and the mechanism of cryptochrome-related signaling pathway in bryophytes. It further shed a new light on the study for the biological diversity of cryptochrome signaling transduction and even enlightened the mechanism investigation of cryptochrome which may still be obscure in higher plants.

4. Materials and Methods

4.1. Plant Materials and Growth Conditions

Male and female M. polymorpha accessions, Takaragaike-1 (Tak-1) and Takara-gaike 2 (Tak-2) [46], were cultured aseptically on half-strength Gamborg’s B5 medium [47] under continuous white light of 50 to 60 μmol photon m⁻² s⁻¹ at 22 °C. F1 spores were obtained by crossing Tak-2 and Tak-1. For the observation of the phenotypes of gemmalings growing under blue or red light, we used a blue (450 nm) LED illuminator or a red (657 nm) LED illuminator to give the indicated light conditions. Light intensity was measured by LI-250A Light Meter (LI-COR).

4.2. HEK-293T Cell Culture and Transfected

HEK-293T cells (ATCC,ATCC^®CRL-11268TM) were cultured in DMEM (Biological Industries, 06-1055-57-1ACS) supplemented with 10% (v/v) FBS (Biological Industries, 04-001-1ACS), 100 U/mL penicillin, and 100 mg/mL streptomycin (Hyclone, SV30010) in humidified 5% (v/v) CO₂ in air, at 37 ℃. HEK-293T cells were seeded at a density of 3 × 10⁵ cells per well in a six-well plate and transfected using the PEI-max method as previously described [48].

Myc, Flag, and GFP were inserted into pCI (neo) (Promega, E1841, Madison, WI, USA) using the EcoRI restriction site for expressing detectable proteins in HEK-293T cells. The coding sequences for MpCRY, MpSPA, MpCOP1, and MpHY5 were then cloned into pCI (neo) Myc, pCI (neo) Flag, or pCI (neo) GFP using XbaI and XmaI restriction sites. The sequences of all primers for the constructions of these plasmids are listed in Table S1.

4.3. Generation of Transgenic Lines

To generate the Mpcry and Mpspa knockout mutant, we used the CRISPR/Cas9 genome editing system [49]. One guide RNA (gRNA) target sequence was designed in the 4th exon of the MpCRY gene and another gRNA target sequence was designed in the first exon of the MpSPA. The annealed oligonucleotides of the gRNA sequence were cloned into the BsaI site of pMpGE_En03 [50]. Using the Gateway LR reaction (Thermo Fisher Scientific), the gRNA expression cassettes were transferred to the pMpGE011 vector [50] to generate pMpGE011_MpCRY and pMpGE011_MpSPA. As described previously [26], Agrobacterium-mediated transformation of F1 spores was performed to generate the Mpcry, Mpspa mutant lines. Mpcryspa double mutants were generated by co-transformation with two kinds of agrobacterium strains respectively containing pMpGE011_MpCRY and pMpGE011_MpSPA. The Mphy5 mutant lines were provide by Nakanisi at Kyoto University.

To obtain the overexpression lines of MpCRY-Tdtomato, the coding sequence (CDS) of MpCRY without the stop codon was amplified and cloned into pENTR/D-TOPO vector (cat # K240020, Thermo Fisher Scientific, Waltham, MA, USA). Then, the cloned sequence was transferred to the destination vector pMpGWB130 [51] to generate _pro35S::MpCRY-Tdtomato binary vector, which was transformed into WT (Tak-1) thalli. To obtain the overexpression lines expressing Citrine-MpSPA, the CDS of MpSPA was amplified and cloned into pENTR-1A vector (cat # A10462, Thermo Fisher Scientific). The cloned sequence was then transferred to the pMpGWB305 [51] destination vector to generate _pro35S::Citrine-MpSPA binary vector, which was transformed into WT (Tak-1) thalli. To examine the co-location of MpCRY and MpSPA proteins, we generated transgenic lines co-expressing _pro35S::MpCRY-Tdtomato and _pro35S::Citrine-MpSPA by co-transformation with two kinds of agrobacterium strains respectively containing _pro35S::MpCRY-Tdtomato and _pro35S::Citrine-MpSPA binary vectors. To generate the overexpression lines of MpHY5-Citrine, the CDS of MpHY5 without the stop codon was amplified and cloned into pENTR/D-TOPO vector. Using the Gateway LR reaction, the cloned sequence was transferred to the destination vector pMpGWB106 [51] to generate _pro35S::MpHY5-Citrine binary vector and then was transformed into WT thalli and Mpcry mutant lines [52]. To generate the overexpression lines of MpCRY-Flag, the CDS of MpCRY without the stop codon was amplified and was transferred to the destination vector pMpGWB310 [51] to generate _proEF::MpCRY-Flag binary vector and then was transformed into WT thalli. The sequences of all primers used in transgenic plant generation are listed in Table S1.

4.4. Colocalization and Localization

Gemmae were plated on Gamborg’s B5 medium and imbibed in the dark for 3 days. Then, the gemmalings were incubated under blue light (30 μmol m⁻² s⁻¹) or kept in the dark for 6 h. Leica TCS SP8X confocal microscope was used to observe the gemmalings. A 514 nm laser was used for excitation of the fluorescent protein Citrine and a 560 nm laser was used for excitation of the furescent protein Tdtomato.

4.5. Phylogenetic Tree Analysis

For the alignment of amino acid sequences, we used the MUSCLE program in Geneious software (Biomatters, Auckland, New Zealand, version 8.1.3) [53]. After sequence alignment, sequence gap and sequences at both ends were manually removed, and conserved region was retained to calculate evolutionary distance. Phylogenetic tree was built by the online program PhyML 3.0 (www.atgc-montpellier.fr/phyml/ (accessed on 21 March 2020)) [54] with JTT model and four substitution rate categories. Its tree searching was started from aBioNJ tree, and the tree was optimized with Subtree Pruning and Regrafting (SPR) topological moves. For statistical analysis of the constructed phylogenetic tree, bootstrapping was carried out by resampling trees 1000 times. The beautification of the resulting tree was done in Geneious software.

4.6. Co-Immunoprecipitation in HEK293T Cells

Next, 36–48 h after transfection, the cells were exposed by blue light or kept in the dark for 3 h, washed twice with PBS buffer (cat # SH30256.01, HyClone, Logan, UT, USA), and then digested with TrypLE™ Express (1×) (cat # 12605-028, Gibco, Waltham, MA, USA) at 37 ℃ for 5 min. The cells were harvested for co-immunoprecipitation.

Cells transfected with different plasmid DNA were lysed by in 500 μL Pierce IP Lysis Buffer (cat # 87787, Pierce, Waltham, MA, USA) with 1× EDTA-free Protease Inhibitor Cocktail Tablets (cat # 4693159001, Roche, Waltham, MA, USA) and incubated on ice for 15 min. The mixtures were then centrifuged at 14,000× g for 10 min at 4 ℃ to remove cell debris. The supernatant was mixed with 20 μL GFP trap beads or Anti-DDDDk-tag mAb-Magnetic Agarose beads (cat # M185-10, MBL, Beijing, China), incubated with vertical blending at 4 ℃ for 2 h. The beads-protein complex was washed 4 times with washing buffer (20 mM HEPES (pH 7.5), 40 mM KCl, 1 mM EDTA) and denatured by mixing thoroughly with 30 μL 4× Loading buffer and heating at 100 ℃ for 10 min.

4.7. Yeast Two-Hybrid Assay

Experiments using the yeast two-hybrid system are according to the manufacturer’s instructions (Matchmaker user’s manual, Clontech, CA, USA). The coding sequences of MpCRY, MpSPA, and different domains of MpCRY and MpSPA were fused in-frame with the GAL4 DNA-binding domain (BD) of the bait vector pGBKT7 using EcoRI (cat # 630489, Clontech, CA, USA). The coding sequences of MpSPA, MpCOP1, MpHY5, and different domains of these proteins were fused in-frame with the GAL4 transcription-activation domain (AD) of the prey vector pGADT7 (cat # 630442, Clontech, CA, USA). The bait plasmid and the prey plasmid were co-transformed into the yeast strain Y2HGold. All sequence of primers for the constructions of these bait plasmids and prey plasmids were listed in Table S1.

The protein–protein interaction was analyzed using the histidine auxotrophy assay. Yeast colonies selected on plates (SD-LW) were patched in duplicate onto His− and His+ plates. Then, one duplicate was grown under blue light (30 μmol m⁻² s⁻¹) at 30 ℃ for 2–3 days. The second duplicate was kept in dark in the same condition.

4.8. Bimolecular Fluorescence Complementation (BiFC) Assay

Isolation and purification methods for Arabidopsis protoplasts are based on “Tape-Arabidopsis-Sandwich” [55] with some modifications. Leaves of 3-week-old Arabidopsis plants in LD (16 h light, 8 h dark) were peeled away from the lower epidermal surface using breathable tape (cat # 1530C-0, 3 M Micropore™, St. Paul, MN, USA) and colorful tape (VBWINTAPE). The remaining leaves were transferred to the enzyme solution (20 mM MES pH 5.7, 1.5% (w/v) cellulase R10, 0.4% (w/v) macerozyme R10, 0.4 M mannitol, 20 mM KCl and 10 mM CaCl). After 2 h digestion at room temperature, the solution was filtered through a 75-micron nylon mesh (Cat # 475855-1R, Calbiotecm^®, Billerica, MA, USA) and centrifugated at 100 g, 4 ℃ for 2 min to pellet the protoplasts. Then the protoplasts were washed with W5 buffer (2 mM MES pH 5.7, 154 mM NaCl, 125 mM CaCl₂, 5 mM KCl, 5 mM D-glucose) twice and incubated on ice for 30 min. For PEG-Ca²⁺ protoplast transformation, protoplasts were washed with ice-cold MMG buffer (4 mM MES pH 5.7, 0.4 M mannitol, 15 mM MgCl₂) twice and resuspended with 100 μL MMG buffer per sample. The indicate volumes of plasmids were added and mixed with protoplasts. The mixture was then mixed with 110 μL PEG-Ca²⁺ buffer (40% (w/v) PEG4000, 0.2 M mannitol, 100 mM CaCl₂) and kept at room temperature for 5 min, washed twice with W5 buffer at room temperature. The protoplasts were finally resuspended with 1 mL W5 buffer and transferred into a 6-well plate and incubated for at least 12 h in dark. For MpCOP1-MpSPA BIFC assay, the fluorescence images were captured using an Olympus FV1000 confocal laser scanning microscope. Image analyses were performed using the FV10-ASW 3.0 Software and processed with Adobe Photoshop. For the other BIFC assays, all pictures were captured by a fluorescence microscope (Axio Observer A1, Zeiss, Oberkochen, Baden-Wuberg, Germany). Image analyses were performed using the Zen software (Zeiss) and processed with Adobe Photoshop. All plasmids used in the BIFC assays were provide by Kohchi group at Kyoto University.

4.9. Ubiquitination Assays in HEK293T Cells

Ubiquitination assays in HEK293T cells were performed as previously described [56]. HEK293T cells were transiently transfected with 5 μg of the combinations of constructs for 36 h. Ubiquitinated GFP-MpHY5 was detected by IP with GFP trap beads at 4 ℃ for 2 h. The pellets were washed 4 times with washing buffer (20 mM HEPES (pH 7.5), 40 mM KCl, 1 mM EDTA) and denatured by mixing thoroughly with 30 μL 4× Loading buffer and heating at 100 ℃ for 10 min.

4.10. Antibodies Used for Immunoblotting

The following primary antibodies used in this study are commercially available: anti-Myc (cat # M192-3, MBL), anti-GFP (cat # 598, MBL), anti-Flag (cat # M185-3L, MBL), anti-HSP (cat # AbM51099-31-PU, Beijing Protein Innovation, Beijing, China), anti-Ubiquitin (cat # 14-6078-82, Invitrogen, Waltham, MA, USA), and anti-Citrine (cat # 632381, Takara, Tokyo, Japan).

4.11. Statistical Analysis

All data were analyzed using GraphPad Prism (version 8.0.2) and statistically ana lyzed by two-tailed Student’s t-test. Signals on Western blots were quantified using ImageJ software. The partition ratio was calculated based on the previous research [57]. The images of indicated time point were collected and analyzed by Image J. The intensity of particles and the intensity of background were calculated, and the final ratio was fit to the formula:

$$Partition ratio=\frac{{\sum }_{i=0}^{n}C\left(i\right)}{n·N }$$

(1)

where the C(i) represents the intensity of each condensate; n represents the condensate number of each nucleus, and N represents the mean intensity of nucleus.