Structural features important for the U12 snRNA binding and minor spliceosome assembly of Arabidopsis U11/U12-small nuclear ribonucleoproteins

ABSTRACT

Although seven proteins unique to U12 intron-specific minor spliceosomes, denoted as U11/U12-65K, -59K, -48K, -35K, -31K, -25K, and -20K, have been identified in humans and the roles of some of them have been demonstrated, the functional role of most of these proteins in plants is not understood. A recent study demonstrated that Arabidopsis U11/U12-65K is essential for U12 intron splicing and normal plant development. However, the structural features and sequence motifs important for 65 K binding to U12 snRNA and other spliceosomal proteins remain unclear. Here, we demonstrated by domain-deletion analysis that the C-terminal region of the 65 K protein bound specifically to the stem-loop III of U12 snRNA, whereas the N-terminal region of the 65 K protein was responsible for interacting with the 59 K protein. Analysis of the interactions between each snRNP protein using yeast two-hybrid analysis and in planta bimolecular fluorescence complementation and luciferase complementation imaging assays demonstrated that the core interactions among the 65 K, 59 K, and 48 K proteins were conserved between plants and animals, and multiple interactions were observed among the U11/U12-snRNP proteins. Taken together, these results reveal that U11/U12-65K is an indispensible component of the minor spliceosome complex by binding to both U11/U12-59K and U12 snRNA, and that multiple interactions among the U11/U12-snRNP proteins are necessary for minor spliceosome assembly.

Introduction

Eukaryotic precursor (pre)-mRNA splicing is an important RNA metabolism necessary for regulated gene expression, in which introns (non-coding sequences) are removed from the pre-mRNAs by two distinct spliceosome complexes. Majority of introns are U2-type, and splicing of the U2-type introns is mediated by the major spliceosome.^1,2 U12-type introns are relatively rare comprising <1% of total introns, and splicing of the U12-type introns is mediated by the minor spliceosome.^3-9 Most U12-type introns co-exist with U2-type introns in a gene,⁴ and splicing of the U12-type introns can be a rate-determining step in the regulation of gene expression.^10-12 U12 introns contain highly conserved sequences at the 5′-splice site, 3′-splice site, and branch-point site.^13,14 The spliceosome assembly process during U2 and U12 intron splicing is highly similar and requires formation of RNA/RNA, RNA/protein, and protein/protein complexes. The U12-dependent spliceosome contains four unique small nuclear RNAs (snRNAs), denoted as U11, U12, U4atac, and U6atac.^15,16 In the earliest step of U12-type spliceosome assembly, U11 and U12 snRNAs recognize the 5splice site and branch-point site on U12 intron, respectively.^17,18

In addition to these snRNAs, more than 300 proteins have been identified in small nuclear ribonucleoprotein (snRNP) complexes. Among them, seven snRNP proteins found in the minor spliceosome, denoted as U11/U12-65K, -59K, -48K, -35K, -31K, -25K, and -20K, are conserved in animals and plants,^19,20 and their roles in minor spliceosome assembly and U12 intron splicing in humans have been previously determined. It has been demonstrated that CHHC-type zinc-finger domain-containing U11/U12-48K is involved in recognizing the 5′-splice site and interacts with U11/U12-59K during spliceosome assembly.^21,22 U11/U12-59K plays a role as a bridge between U11/U12-48K and U11/U12-65K in the minor spliceosome complex.²³ The N-terminal half of U11/U12-65K interacts with U11/U12-59K and the C-terminal half of 65 K binds to U12 snRNA at the same time.²³ Recent studies have demonstrated that U11/U12-65K in zebrafish plays an essential role in development of the digestive organs by affecting multiple genes involved in various steps of mRNA processing, including transcription, splicing, and nuclear export,²⁴ and that U11/U12-65K in humans is related to familial growth hormone deficiency by influencing U11/U12 di-snRNP formation and splicing of multiple U12-type introns.²⁵ Although these previous studies using human cell lines and animal systems clearly demonstrated the prominent roles of the minor spliceosome-specific proteins in spliceosome assembly and disease and development, the functional role of plant minor spliceosomal proteins is far less understood. It has been demonstrated that Arabidopsis (Arabidopsis thaliana) and rice (Oryza sativa) U11/U12-31K harboring RNA chaperone activity is necessary for U12 intron splicing and normal growth and development of plants.^26,27 In recent, Arabidopsis U11/U12-65K and U11-48K was determined to be essential for splicing of most of the U12 introns, which is crucial for normal plant growth and development.^28,29

In this study, we further explored the interactions between U11/U12-65K and U12 snRNA as well as other snRNP proteins unique to the minor spliceosome. We conducted domain-deletion analysis to determine structural features important for the interactions of U11/U12-65K with U12 snRNA and U11/U12-59K protein. We also employed yeast two-hybrid (Y2H) analysis and in planta bimolecular fluorescence complementation (BiFC) and luciferase complementation imaging (LCI) assays to determine the interactions between U11/U12-65K, -59K, -48K, -35K, -31K, and -25K during minor spliceosome assembly. We provide evidence that U11/U12-65K is an indispensible component of the minor spliceosome complex by interacting with U11/U12-59K and U12 snRNA in plants.

Results

U11/U12-65K binds specifically to U12 snRNA

Fig. 1A shows the domain structures of the U11/U12-snRNP proteins found in plant and animal minor spliceosome complexes. U11/U12-65K harbored two RNA-recognition motifs (RRMs) at the N-terminal and C-terminal half of the protein and a conserved proline-rich region between the two RRMs. A previous study demonstrated that human U11/U12-65K binds directly to U12 snRNA of the minor spliceosome complex (23). Analysis of the nucleotide sequences of the U12 snRNAs revealed that Arabidopsis U12 snRNA shared high sequence similarity with human U12 snRNA (Fig. 1B). The human U12 snRNA formed four stem-loop (SL) structures and has been demonstrated to bind to the branch-point site of U12-introns.^17,23 The structure of Arabidopsis U12 snRNA was predicted using an RNA fold program with the default parameters (http://mfold.rna.albany.edu,³⁰), and the results showed that Arabidopsis U12 snRNA also adopted a four SL structure, similar to the secondary structure of human U12 snRNA (Fig. S1). The sequence of the SLI that is responsible for base pairings with the branch-point site of U12 introns and the SLIII, which is the target of U11/U12-65K binding, were highly conserved between Arabidopsis and humans (Fig. 1 and Fig. S1).

Figure 1.

Schematic presentation of the domain structures of U11/U12 minor spliceosomal proteins and nucleotide sequences of U12 snRNA in A. thaliana and H. sapiens. (A) RNA-recognition motif (RRM), CHHC-type or CCHC-type zinc finger (ZF) region, and aspartate (D), glutamate (E), arginine (R), serine (S), and proline (Pro)-rich motifs are shown. (B) The sequences of U12 snRNAs in Arabidopsis (At1g61275) and human (NR_029422) were aligned by ClustalW program. The conserved sequence in the first stem-loop which is responsible for base pairing with the branch-point site of U12 introns is indicated by a thick line, and the conserved sequence in the stem-loop III which is responsible for U11/U12-65K binding is indicated by stars.

In recent, it was demonstrated that Arabidopsis U11/U12-65K binds to U12 snRNA in vitro.²⁸ To confirm whether U11/U12-65K binds to U12 snRNA in planta, we employed an RIP assay to evaluate the binding ability of 65K to U12 snRNA. Transgenic Arabidopsis expressing the U11/U12-65K-FLAG fusion protein was generated, in which the FLAG tag was fused to the C-terminal end of U11/U12-65K, and 2-week-old seedlings grown on MS medium were utilized for the RIP assay. Expression of the 65K-FLAG fusion protein in the transgenic plant was confirmed by Western blot analysis using the α-FLAG antibody (Fig. 2A). Total RNAs co-immunoprecipitated with U11/U12-65K were purified and analyzed by RT-PCR using three different primer sets designed to amplify the full-size U12 snRNA (175 nt), the 5′-half U12 snRNA (92 nt), or the 3′-half U12 snRNA (83 nt) (Fig. 2B). The results showed that each RT-PCR with the three different primer sets resulted in amplified PCR products of the expected size of U12 snRNA (Fig. 2C). The identity of each band was further verified by sequencing. In contrast, no PCR products were amplified using the primers specific to U2 snRNA gene used as a negative control (Fig. 2C). Taken together, these results confirm that U11/U12-65K binds to U12 snRNA in planta.

Figure 2.

RNA immunoprecipitation assay to determine binding of U11/U12-65K to U12 snRNA in planta. (A) Expression of the 65K-FLAG protein in the transgenic plant was confirmed by Western blot analysis using the α-FLAG antibody. The molecular size of the 65K (48.7 k_D) and FLAG-tag (2.8 k_D) fusion protein is indicated. (B) Schematic predicted secondary structure of U12 snRNA with the four stem-loop (SL) structures. Position and direction of the primers (a, b, c, and d) used in RT-PCR analysis is also indicated. (C) RT-PCR was conducted using the total RNA fraction (Input), the RNAs precipitated with FLAG antibody (IP+), and the RNAs without precipitation (IP-). For the detection of U12 snRNA, the primers a and d amplifying the full-length U12snRNA, the primers a and b amplifying the 5′-half U12 snRNA, and the primers c and d amplifying the 3′-half U12 snRNA were used. U2 snRNA was used as a negative control. The PCR products were separated on a 1% agarose gel and were visualized under UV. bp, base pair; M, molecular weight size marker.

The C-terminal RRM region of U11/U12-65K is responsible for binding to U12 snRNA

We next determined whether the N-terminal or the C-terminal RRM of U11/U12-65K is responsible for specific binding to U12 snRNA. Constructs expressing the full-length (65K-F), N-terminal region (65K-N), N-terminal region RRM (65K-N-RRM), C-terminal region (65K-C), and C-terminal RRM region (65K-C-RRM) were designed (Fig. 3A), and the recombinant GST-65K fusion proteins were purified in E. coli (Fig. S2) and utilized for an electrophoretic mobility shift assay. The full-length 65K bound to U12 snRNA (Fig. 3B) as shown previously.²⁸ The N-terminal fragment (252 amino acid residues) containing the first RRM and the conserved proline-rich region as well as the N-terminal RRM only (148 amino acid residues) did not bind to U12 snRNA, whereas the C-terminal fragment (190 amino acid residues) harboring the second RRM bound to U12 snRNA (Fig. 3C). Two shifted bands, one major and one minor bands, were observed on the gels containing the full-length of the 65K protein, possibly due to the structural changes induced by the binding of the full-length 65K protein to U12 snRNA. Next, the recombinant C-terminal fragment (303 amino acid residues) containing the conserved proline-rich region as well as the second RRM (65K-C) was purified, and its binding capability to U12 snRNA was analyzed. Contrary to the strong binding of the 65K-F and 65K-C-RRM proteins to U12 snRNA, the 65K-C protein lost almost completely the ability to bind U12 snRNA (Fig. 3C), possible due to the poor or incorrect folding of the truncated 65K-C protein. These results suggest that the C-terminal RRM region of U11/U12-65K is responsible for the specific binding to U12 snRNA.

Figure 3.

The C-terminal domain of U11/U12-65K is essential for binding to U12 snRNA. (A) Schematic domain structures of the full-length, N-terminal region, and C-terminal region of U11/U12-65K. The proline-rich conserved region between the two RRMs is indicated by a gray box. (B, C) Electrophoretic mobility shift assay to determine binding of U11/U12-65K to U12 snRNA. The ³³P-labeled U12 snRNAs were incubated with increasing amounts of (B) the full-length GST-U11/U12-65K fusion protein and (C) the N-terminal region (65K-N), the N-terminal RRM region (65K-N-RRM), the C-terminal RRM region (65K-C-RRM), or the C-terminal region (65K-C) of the 65 K protein. The protein-RNA complexes were separated on native 6% PAGE and detected with a phosphorimager. Arrows and stars indicate the RNA substrates and shifted bands, respectively.

U11/U12-65K recognizes and binds specifically to the stem-loop III of U12 snRNA

Given that human U11/U12-65K bound to the 3′ half of human U12 snRNA,²³ we investigated whether Arabidopsis U11/U12-65K recognizes and binds to the specific sequences of Arabidopsis U12 snRNA. U12 snRNAs spanning the SLI to SLIIb (designated 5′-half) or the SLIII (designated 3′-half) (Fig. 4A) were synthesized separately and used for the binding assay. The full-length or C-terminal RRM region of the 65K protein bound strongly to the 3′-half of U12 snRNA (Fig. 4B). In contrast, extremely weak shifted bands were observed on the gel containing the full-length or C-terminal RRM region of 65K protein and the 5′-half of U12 snRNA (Fig. 4C). These results indicate that U11/U12-65K binds strongly to the SLIII of U12 snRNA in plants as well as in animals.

Figure 4.

U11/U12-65K binds specifically to the third stem-loop of U12 snRNA. (A) Schematic predicted secondary structures of the full-length, 5′-half, and 3′-half of U12 snRNA. The ³³P-labeled 3′-half (B) or 5′-half (C) of the U12 snRNAs were incubated with increasing amounts of the full-length (65K-F), N-terminal region (65K-N), and C-terminal RRM region (65K-C-RRM) of the 65K protein. The protein-RNA complexes were separated on native 6% PAGE and detected with a phosphorimager. Arrows and stars indicate the RNA substrates and shifted bands, respectively.

The nucleotide sequences in the SLIII of U12 snRNA are highly conserved between human and Arabidopsis (Fig. 1B). To determine which specific nucleotides in the SLIII of U12 snRNA are important for binding to U11/U12-65K, we generated three different U12 snRNA variants in which the SLIII-M1 mutant had 5′ UA 3′ to 5′ CG 3′ changes, the SLIII-M2 mutant had 5′ UUU 3′ to 5′ AAA 3′ changes in the loop region, and the SLIII-M3 mutant had changes in the base pairs in the SLIII stem region (Fig. 5). The binding assay between U11/U12-65K and the mutant U12 snRNAs showed that binding of the 65K protein to SLIII-M1 was strongly inhibited and that binding of the 65K protein to SLIII-M2 was completely abolished. In contrast, binding of the 65K protein to SLIII-M3 was not affected (Fig. 5). These results show that the sequences in the loop region but not the base pairs in the stem region of SLIII are important for 65K protein binding. Importantly, the full-length U11/U12-65K did not bind to U11 snRNA (Fig. 5). Taken together, these results clearly indicate that U11/U12-65K recognizes and binds specifically to the SLIII of U12 snRNA.

Figure 5.

U11/U12-65K binds to the terminal hairpin of the third stem-loop of U12 snRNA. The ³³P-labeled wild type (SLIII) and mutant U12 snRNAs (SLIII-M1 to M3), in which the circled nucleotides in the third stem-loop (SL) region of the U12 snRNAs were substituted, were incubated with increasing amounts of U11/U12-65K (0, 5, and 10 pmole), and the protein-RNA complexes were separated on native 6% PAGE and detected with a phosphorimager. Arrow and stars indicate the RNA substrates and shifted bands, respectively.

Interactions between U11/U12-snRNP proteins in the minor spliceosome complex

Previous studies on human minor spliceosome components have demonstrated that U11/U12-65K interacts with U11/U12-59K and that U11/U12-59K interacts with U11/U12-48K.^21,23 To investigate whether these interactions also occur in plant minor spliceosomes and to further determine possible interactions between other U11/U12-snRNPs, the interactions between the six U11/U12-snRNPs (25 K, 31 K, 35 K, 48 K, 59 K, and 65 K proteins) were examined using the Y2H analysis and in planta BiFC and LCI assays. We first confirmed that each U11/U12-snRNP protein fused to either activation domain or binding domain vector did not activate transcription of the reporter gene in the Y2H analysis (Fig. S3). The results of Y2H analysis showed that the 65K protein interacted with the 59K protein, whereas the 59K protein interacted with the 25K, 48K, and 65K proteins. U11/U12-35K interacted with 25K protein and with 35K protein itself to form homodimer. By contrast, no interactions were observed between U11/U12-31K and other U11/U12-snRNPs (Fig. 6A). Moreover, the N-terminal region of the 65K protein as well as the full-length 65K protein interacted with the 59K protein, but the C-terminal region of the 65K protein did not interact with the 59K protein (Fig. 6A).

Figure 6.

Interactions between U11/U12-snRNP proteins in the minor spliceosome complex. (A) Interactions between each pair of U11/U12-snRNP proteins determined by yeast two-hybrid (Y2H) analysis. The cDNAs encoding each snRNP protein were cloned into the pACT2 activation domain (AD) and the pBD-GAL4 vectors. The yeast Y190 cells co-transformed with both vectors were grown on SD/- Leu/-Trp/-His medium containing 5-25 mM 3-aminotriazole for the X-gal (5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside) filter-lift assay. Interactions between the 59K protein and the full-length (65K-F), N-terminal region (65K-N), or C-terminal region (65K-C) of U11/U12-65K were analyzed as described above. (B) Interactions between each pair of U11/U12-snRNP proteins in planta determined by bimolecular fluorescence complementation (BiFC) assay. The YFP^N and YFP^C constructs expressing each fusion protein were co-infiltrated into tobacco leaves, and the emergence of yellow fluorescence signals was observed using a fluorescence microscope. (C) Interactions between each pair of U11/U12-snRNP proteins in planta determined by luciferase (Luc) complementation imaging (LCI) assay. The N-Luc and C-Luc constructs expressing each fusion protein were co-infiltrated into tobacco leaves, and the LUC image was captured using an imaging system. Unit (p/s/cm²/sr), photons/second/square centimeter/steradian. (D) Summary of interactions between U11/U12-snRNP proteins. −, no interaction; the number of + sign in BiFC and LCI assays is correlated to the relative strength of the interactions.

To further confirm whether the interactions between U11/U12-snRNP proteins revealed by Y2H analysis are really observed in plants, the BiFC and LCI assays were employed to determine the in planta interactions between the six U11/U12-snRNPs. We first confirmed that co-infiltration of the truncated N-terminal YFP (YFP^N) and the C-terminal YFP (YFP^C) as a negative control in tobacco leaves did not generate any yellow fluorescence (Fig. S4). The results of BiFC assay between different pairs of U11/U12-snRNP proteins are shown in Fig. S5 and are summarized in Fig. 6B, and the results of LCI assay are summarized in Fig. 6C. Clearly, in addition to the interactions observed in Y2H analysis, more interactions between the U11/U12-snRNPs were detected by BiFC and LCI assays. The presence and absence of interactions between the U11/U12-snRNPs and the relative strength of the interactions are summarized in Fig. 6D. Notably, the 25K protein interacted with all five U11/U12-snRNPs, whereas the 31K protein showed the least interactions with the 65K, 35K, and 25K proteins. In addition to these interactions between the different U11/U12-snRNP proteins, the 65K, 35K, and 25K proteins interacted themselves (Fig. 6D). However, it should be noted that some of the interactions were observed only in one direction between the N-terminal and C-terminal fragments of the U11/U12-snRNPs, which is commonly observed in Y2H, BiFC, and LCI assays. Based on all of these information regarding protein-RNA and protein-protein interactions, we proposed a model for minor spliceosome complexes (Fig. 7). U11/U12-65K binds to the SLIII of U12 snRNA via its C-terminal RRM and the N-terminal half of U11/U12-65K interacts with U11/U12-59K that interacts with both the U11/U12-48K and U11/U12-25K proteins. Notably, the core interactions between the U11/U12-65K, -59K and -48K proteins are conserved in plants and animals, and U11/U12-25K occupies the position that interacts with the five U11/U12-snRNP proteins.

Figure 7.

A model of interactions between U11/U12-snRNP proteins and snRNAs in the minor spliceosome complex. U11/U12-65K binds to the stem-loop III of U12 snRNA via its C-terminal region, and the N-terminal region of the 65K protein interacts with the 59K protein, which interacts with both the 48K and 25K proteins. U11/U12-25K occupies in the position that interacts with the 65K, 59K, 48K, 35K, and 31K proteins. Interactions of U11/U12-31K with the 65K, 48K, 35K, and 25K proteins allow to place the 31K protein in a position to contact with diverse RNA substrates as an RNA chaperone, which is needed to maintain the U12 introns into splicing-competent conformations for an efficient splicing of U12 introns in cells. U11 and U12 snRNAs are indicated by orange color, and U12 intron is indicated by blue color.

Discussion

Spliceosome-associated RNA-binding proteins (RBPs) harboring RRM or SR domains are required for spliceosome assembly, splicing reactions, or subsequent spliceosome disassembly.^1,31 Our current analysis demonstrated some similarities and differences in the binding of U11/U12-65K to U12 snRNA between animals and plants. Similar to the interactions found in humans,²³ Arabidopsis U11/U12-65K protein bound to U12 snRNA via its C-terminal RRM (Fig. 3), and the C-terminal RRM region of U11/U12-65K specifically recognized and bound to the SLIII of U12 snRNA (Fig. 4). Moreover, the nucleotide sequences in the SLIII of U12 snRNA were highly conserved between humans and Arabidopsis (17 and Fig. 1), and we showed via point mutation analysis that the sequences in the terminal hairpin of the SLIII of U12 snRNA are important for the interaction with U11/U12-65K protein (Fig. 5), as observed in humans. One noticeable difference in the binding of U11/U12-65K to U12 snRNA between animals and plants was also found. In contrast to the binding of the human 65K N-terminal and C-terminal halves containing each RRM and the conserved proline-rich region to U12 snRNA,²³ the Arabidopsis U11/U12-65K N-terminal and C-terminal fragments containing the conserved proline-rich region and each RRM lost the ability to bind U12 snRNA (Fig. 3). Although it is unclear at present how the truncated fragment containing the conserved proline-rich region prevents the binding of the plant 65K protein to U12 snRNA, it is possible that the truncated proteins undergo poor or incorrect folding, which results in reduced affinity to U12 snRNA. We speculate that the conserved proline-rich region of the 65K protein influences the maintenance of the protein structure necessary for its binding to U12 snRNA.

Previous studies of human minor spliceosome components have demonstrated that the 48K protein interacts with the 59K protein and then the 59K protein interacts with the 65K protein to form a core bridge for minor spliceosome assembly.^21,23 Our present Y2H, BiFC, and LCI analyses confirmed that these core interactions are also found in plant minor spliceosomes (Fig. 6). Notably, more interactions between the U11/U12-snRNPs were detected by BiFC and LCI assays than the interactions observed in Y2H analysis (Fig. 6). These interactions detected only by BiFC and LCI assays either reflect the real in planta interactions between the U11/U12-snRNPs or may result from the indirect interactions of the U11/U12-snRNPs due to the close proximity of the proteins incorporated into the di-snRNP complexes. Because some of the interactions were observed only in one direction between the N-terminal and C-terminal fragments of the U11/U12-snRNPs (Fig. 6), we proposed a model for minor spliceosome complexes (Fig. 7) by considering the relative strength of interactions observed by the three different assays (Fig. 6D). Clearly, the core interactions between the U11/U12-65K, -59K and -48K proteins are conserved in minor spliceosome complex in plants and humans. Moreover, the N-terminal region of the 65K protein as well as the full-length 65K protein interacted with the 59K protein, but the C-terminal region of the 65K protein did not interact with the 59K protein (Fig. 6A). In addition to these core interactions, we found that other U11/U12-snRNP proteins interact extensively together to form minor spliceosome complexes (Figs. 6 and 7). It appears that multiple interactions among the U11/U12-snRNP proteins are necessary to form sTable Spliceosomal complexes, which is essential for minor spliceosome assembly.

Efficient intron splicing requires maintenance of RNA substrates into splicing-competent conformations, and specific types of RBPs that can assist RNA folding are involved in intron splicing. RNA chaperones are nonspecific RBPs that bind to diverse RNA substrates with a low sequence specificity and aid the RNA folding process by preventing RNA misfolding or by resolving misfolded RNA species.^32-34 We demonstrated previously that Arabidopsis and rice U11/U12-31K proteins possess RNA chaperone activities and bind to substrate RNAs with a low sequence specificity, which are important for proper splicing of most of the U12 introns.^26,27 Contrary to U11/U12-31K, no RNA chaperone activity was observed for U11/U12-65K (Fig. S6). Because RNA chaperones generally bind sequence-nonspecifically to diverse RNA substrates,^35,36 sequence-specific binding of U11/U12-65K to U12 snRNA also supports the notion that U11/U12-65K does not act as an RNA chaperone. According to the proposed model of interactions between U11/U12-snRNP proteins and U12 intron-containing RNA transcripts (Fig. 7), U11/U12-31K interacts with the 65K, 35K, and 25K proteins, which places the 31K protein in a position that allows to contact with most of the U12 introns. It appears that this spatial organization of U11/U12-snRNP proteins is important for the function of U11/U12-31K as an RNA chaperone, which is needed to maintain the diverse U12 introns into splicing-competent conformations for an efficient splicing of U12 introns in cells.

In conclusion, our findings demonstrate that U11/U12-65K plays an essential role in U12 minor spliceosome assembly by dual binding to U11/U12-59K and U12 snRNA, and multiple interactions among the U11/U12-snRNP proteins are necessary for minor spliceosome assembly. As the nature of protein-RNA and protein-protein interactions in minor spliceosome complex is emerging, it would be of interest to determine the importance of these interactions during splicing of minor introns in plants and animals. More studies are needed to determine the functional significance of the protein-RNA and protein-protein interactions during minor spliceosome assembly and the consequence of these interactions during intron splicing and growth and development of plants.

Materials and methods

RNA immunoprecipitation assay in planta

The U11/U12-65K cDNA was fused to the FLAG-tag DNA in the pBI121 vector, which expressed the 65K-FLAG-tag fusion protein under control of the cauliflower mosaic virus 35S promoter. The vector was transformed into A. thaliana Columbia-0 ecotype by vacuum infiltration ³⁷ using Agrobacterium tumefaciens GV3101. The T₃ homozygous lines were selected and used for RNA immunoprecipitation (RIP) assay, which was conducted essentially as described.³⁸ Briefly, the plants were grown at 23°C under long-day conditions (16-h-light/8-h-dark cycle). Seedlings were harvested and fixed in formaldehyde, and nuclei were extracted from the seedlings, essentially as described.^38,39 The nuclear extract was immunoprecipitated using an anti-FLAG antibody according to the manufacturer's instructions (Sigma-Aldrich, F7425). RNAs were recovered and analyzed by reverse transcription-polymerase chain reaction (RT-PCR) using the primers listed in Supplemental Table S1. The identity of RNAs was verified by sequencing.

Expression and purification of recombinant proteins in E. coli

The U11/U12-65K coding region was cloned into the pGEX-5X-2 vector (GE Healthcare Life Sci., 28-9545-54) for expression and purification of recombinant GST-65K fusion proteins in E. coli. The recombinant GST-65K fusion proteins were expressed in BL21 DE3 competent cells (Promega, L1191) and purified using glutathione Sepharose 4B resin as described^28,35

In vitro RNA-binding assay

The cDNAs encoding the full-length, 5′-half, 3′-half, and mutant U12 snRNAs were cloned into the pET-22b(+) vector. The vector was digested with the SalI restriction enzyme, and the ³³P-UTP-labeled RNAs were synthesized by in vitro transcription using T7 RNA polymerase (Promega, P207E). The synthetic RNA substrates were incubated with the recombinant GST-65K fusion proteins in binding buffer (10 mM Tris-HCl, pH 7.5, 50 mM NaCl, 1 mM EDTA, and 7.4% glycerol) on ice for 30 min. The reaction mixture was loaded on a native 6% polyacrylamide gel, and RNA bands were visualized by a FLA7000 Phosphorimager (GE Healthcare).

Yeast two-hybrid assay

The cDNAs encoding seven spliceosomal proteins (25K, 31K, 35K, 48K, 59K, and 65K proteins) were cloned into the pACT2 vector, and the yeast Y190 cells were transformed with both the pBD-GAL4 and pACT2 vectors. The vectors and yeast strain were purchased from Clontech (Mountain View, 638822, PT3024-1). Transformants were selected on synthetically defined (SD)/-Leu/-Trp/-His medium after a 30°C incubation for 3–5 d. Surviving yeast cells were grown on SD/- Leu/-Trp/-His medium containing 5–25 mM 3-aminotriazole for the X-gal (5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside) filter-lift assay. Co-transformed yeast cells retained on the filter layer were lysed by freezing and thawing in liquid nitrogen. The filter was soaked in buffer (60 mM Na₂HPO₄, 40 mM NaH₂PO₄, 10 mM KCl, 1 mM MgSO₄, and β-mercaptoethanol) containing 20 mg/mL X-gal.

Bimolecular fluorescence complementation (BiFC) assay

For the BiFC assay, the cDNAs encoding the 25K, 31K, 35K, 48K, 59K, and 65K proteins were cloned into the N-terminal fragment of yellow fluorescence protein (YFP^N) vector pSPYNE_35S/pUC-SPYNE and the C-terminal fragment of yellow fluorescence protein (YFP^C) vector pSPYCE_35S/pUC-SPYCE as previously described.⁴⁰ The Agrobacterium strain GV3101 containing each construct and the silencing suppressor P19 strain ⁴¹ were syringe‐infiltrated in 5-week-old tobacco (Nicotiana benthamiana) leaves. After further incubation in the growth room at 24°C for 3 d under long-day conditions (16 h light / 8 h dark cycle), the YFP fluorescence in 25 mm² leaf pieces was detected with a TCS SP5 AOBS® spectral confocal and multiphoton microscope system (Leica Microsystems). The YFP images of each construct were examined from three leaves obtained from three independent plants.

Firefly luciferase complementation imaging (LCI) assay

For the LCI assay, the cDNAs encoding the 25K, 31K, 35K, 48K, 59K, and 65K proteins were cloned into the pCAMBIA1300.nLUC or pCAMBIA1300.cLUC vector as previously described.⁴² The Agrobacterium strain GV3101 containing each construct and the silencing suppressor P19 strain were syringe‐infiltrated in 5-week-old tobacco leaves. After further incubation in the growth room at 24°C for 2 days, 1 mM luciferin was sprayed onto the leaves, and the leaves were kept in dark for 2 min to quench the fluorescence. The LUC images were observed using a G:BOX Chemi XL imaging system (Syngene, ELE7606) at 5 min intervals.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Drs. M. Inouye and S. Phadtare for the BX04 mutant cells and the pINIII vector and Dr. R. Landick for the E. coli RL211 cells.

Funding

This work was supported by grants from the Mid-career Researcher Program through a National Research Foundation of Korea grant funded by the Ministry of Education, Science, and Technology (2011-0017357) and from the Next-Generation BioGreen21 Program (PJ01103601), Rural Development Administration, Republic of Korea.