Research

Biochemical and Molecular Studies of S-RNase-Based Self-Incompatibility in Petunia inflata 

During the first phase of our research (1986 to 1994), we focused on the identification and characterization of the polymorphic S-RNase gene that encodes the pistil determinant of SI.  The major findings obtained include the following: (1) polymorphism of the S-RNase gene predates speciation in the Solanaceae (Ioerger et al., Proc Natl Acad Sci USA 87: 9732-9735, 1990), suggesting that S-RNase-based SI existed in the common ancestor of the solanaceous species and is thus a very ancient mechanism; (2) identification of primary structural features (including conserved and hypervariable regions) of allelic variants of S-RNase (Ioerger et al., Sex Plant Reprod 4: 81-87, 1991), which provided the basis for subsequent structure/function studies of S-RNase; (3) use of gain-of-function and loss-of-function approaches to show that S-RNase is necessary and sufficient for the pistil to recognize and reject self-pollen, thus establishing that S-RNase alone regulates pistil specificity (Lee et al., Nature 367: 560-563, 1994; cover story); (4) use of site-directed mutagenesis to show that the RNase activity of S-RNase is essential for the function of S-RNase in SI (Huang et al., Plant Cell 6: 1021-1028, 1994), suggesting that the biochemical mechanism of growth inhibition of self-pollen tubes involves degradation of pollen tube RNA; (5) use of site-directed mutagenesis to show that the glycan moiety of S-RNase is not required for its recognition function (Karunanandaa et al., Plant Cell 6: 1933-1940, 1994), suggesting that allelic specificity of S-RNase is encoded by its protein backbone; (6) use of the chimeric gene approach to show that the two hypervariable regions of S-RNase are necessary but not sufficient for allelic specificity (Kao and McCubbin, Proc Natl Acad Sci USA 93: 12059-12065, 1996).   

In the second phase of our research (1994 to 2004), we focused on the identification of the gene encoding pollen specificity.  At that time, it was thought that a single gene encodes pollen specificity, just as a single polymorphic S-RNase gene encodes pistil specificity.  Identification of the first pollen specificity gene turned out to be much more challenging than identification of the S-RNase gene.  All the attempts using the approaches designed based on the predicted properties of the pollen specificity determinant were unsuccessful (because, as it is now known, the nature of the pollen specificity determinant defies conventional wisdom!).  Our lab finally identified the first pollen specific gene of Petunia by cloning and sequencing a 328-kb region of the S-locus that contains the S2-RNase (S2-allele of S-RNase) gene (Wang et al., Plant Mol Biol 54: 724-742, 2004).  The S-locus is located in a sub-centromeric region where recombination is suppressed, so the physical distance between S-locus genes is much larger than the genetic distance would suggest.  Also, the S-locus is enriched in highly repetitive sequences, making sequencing and sequence assembly difficult, especially during the early 2000’s before the advent of next generation sequencing.  A key to the success of identifying the first pollen specificity gene was the construction of a high-quality BAC library of S2S2 genotype in our lab (McCubbin et al., Genome 43: 820-826, 2000), which was a very challenging task and took two years to complete.  The gene identified was then named PiSLF (for Petunia inflata S-Locus F-box gene), but has been renamed SLF1 after the discovery that multiple SLF genes collectively encode pollen specificity determinant (see below).  Our lab then developed an in vivo functional assay based on an intriguing finding, named competitive interaction, reported in the 1940’s, to show that SLF1 is involved in controlling pollen specificity (Sijacic et al., Nature 429: 302-305, 2004).

In the third phase of our research (2004 to 2010), we initially focused on the study of the biochemical mechanism of SI, based on the conviction then that SLF1 was the only SLF involved in pollen specificity.  The major findings obtained include the following: (1) SLF1 is the F-box protein component of an SCF complex involved in ubiquitin-mediated protein degradation of S-RNase in the pollen tube (Hua and Kao, Plant Cell 18: 2531-2553, 2006); (2) very surprisingly, non-self interactions are stronger than self-interactions between allelic variants of SLF1 and allelic variants of S-RNase (Hua and Kao, Plant Cell 18: 2531-2553, 2006); (3) there are four other SLF genes expressed in pollen, which were then collectively referred as SLF-like genes because there were thought not to be involved in pollen specificity (Hua et al., Plant Cell 19: 3593-3609, 2007), but later found to be among the multiple SLF genes involved in pollen specificity; (4) pollen specificity is controlled by at least two more paralogous genes (a joint discovery with Dr. Seiji Takayama’s lab, then at Nara Institute of Science and Technology, Japan, and now at University of Tokyo), and a model, collaborative non-self recognition, was proposed to explain the biochemical basis of compatible and incompatible pollination (Kubo et al., Science 330: 796-799, 2010).  This model predicts that for a given S-haplotype, (a) each of the multiple SLF proteins produced in pollen is capable of interacting with only a subset of its non-self S-RNases taken up by a pollen tube during its growth in the pistil; (b) at least one of the SLF proteins produced in pollen can interact with a particular non-self S-RNase, and all SLF proteins are required to collectively interact with a complete suite of their non-self S-RNases; (c) none of the SLF proteins can interact with their self S-RNase; (d) an SLF protein is a component of the SCF E3 ubiquitin ligase complex, which, in conjunction with E1 ubiquitin activating enzyme and E2 ubiquitin conjugating enzyme, catalyzes the transfer of ubiquitin chains to the S-RNase(s) recognized by the specific SLF protein in the complex; (e) all non-self S-RNases taken up into the pollen tube become ubiquitinated and degraded by the 26S proteasome, thereby resulting in cross-compatible pollination, whereas self S-RNase is not ubiquitinated and it exerts its cytotoxicity within the pollen tube, thereby resulting in self-incompatible pollination.  The findings described in (1) and (2) above played a key role in the formulation of the collaborative non-self recognition model.

In the current phase (2010-), the PI has focused on the questions uncovered from the unexpected finding that multiple SLF genes encode the pollen determinant, and that the biochemical mechanism of S-RNase-based SI is via non-self recognition, rather than the more conventional self recognition, like lock-and-key. So far, the lab has obtained the following major findings: (1) use of pollen transcriptome analysis to identify a total of 17 pollen-specific polymorphic SLF genes that collectively encode the pollen determinant in both S2-haplotype and S3-haplotype (Williams et al. Plant Cell 26: 2873-2888, 2014); (2) use of co-immunoprecipitation and mass spectrometry to show that all SLF proteins of S2 and S3 pollen are assembled into similar SCF complexes, which also contain PiSSK1 (a pollen-specific Skp1-like protein), PiRBX1 (a RING-finger protein) and PiCUL1-P (a pollen-specific Cullin1) (Li et al. Plant Reprod 27: 31-45, 2014; Plant J 87: 606-616, 2016); (3) use of the in vivo gain-of-function assay developed in the PI’s lab to establish 133 interaction relationships between 17 SLF proteins of S2-haplotype and 11 S-RNases, and 30 interaction relationships between 5 of the 17 SLF proteins of S3-haplotype and 9 of these 11 S-RNases, in order to test the predictions of the collaborative non-self recognition model (Sun and Kao Plant Cell 25: 470-485, 2013; Williams et al. Mol Plant 7: 567-569, 2014; Sun et al. Plant Cell 30: 2959-2972, 2018); (4) discovery that SLF proteins are themselves subject to ubiquitin-mediated degradation by the 26S proteasome (Sun et al. Plant J 83: 213-223, 2015), and identification of pollen proteins that may regulate the dynamic life cycle of SCFSLF complexes; (5) use of a chimeric gene approach to identify the candidate amino acids required for specific interactions between S2-SLF1 and S3-RNase (Wu et al. Plant Cell Physiol 59: 234-247, 2018); (6) Use of CRISPR/Cas9-mediated genome editing to show that PiSSK1 is essential for pollen to be compatible with pistils carrying non-self ­S-haplotypes, thus confirming that PiSSK1 is the Skp1 component of the SLF-containing SCF complex required for mediating ubiquitination and degradation of non-self S-RNases (Sun and Kao Plant Reprod 31: 129-143, 2018); (7) use of CRISPR/Cas9-mediated genome editing to show that SLF proteins are solely responsible for the SI function of pollen (Sun et al. Plant Cell 30: 2959-2972, 2018); (8) use of Bacterial Artificial Chromosome (BAC) clones collectively containing all 17 S2-SLF genes and S2-RNase for PacBio and illumina sequencing, and assembly of ~3.1 Mb sequence for comprehensive analysis of this S-locus region (Wu et al. Plant J 104: 1348-1368, 2020); (9) use of CRISPR/Cas9-mediated genome editing to show that both PiCUL1-P and PiCUL1-B can serve as the Cullin1 component of the SLF-containing SCF complexes, and that both PiCUL1-P and PiCUL1-B specifically function in SI (Sun et al. Plant Cell, doi: 10.1093/plcell/koac357).   

When our lab embarked on the study of the SI system possessed by Solanaceae more than three decades ago, very little was known about the biochemical and molecular bases for this genetically determined self/non-self recognition between pollen and pistil.  The progress made in the ensuing years by our lab and other labs worldwide has revealed that the mechanism for S-RNase-based SI is even more complex and fascinating than any researcher could envision at the beginning of the study.  As is true for any complex biological system, the more one knows about it, the more questions one uncovers.  The immense knowledge already gained will be valuable for ultimate understanding of the biochemical, molecular, and structural bases of this complex self/non-self recognition system.