IDENTIFICATION AND FUNCTIONAL CHARACTERIZATION OF PISLF, PETUNIA INFLATA S-LOCUS F-BOX PROTEIN, THE MALE DETERMINANT OF S-RNASE-MEDIATED SELF-INCOMPATIBILITY
Open Access
- Author:
- Sijacic, Paja
- Graduate Program:
- Integrative Biosciences
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- February 07, 2007
- Committee Members:
- Hong Ma, Committee Member
Sarah Mary Assmann, Committee Member
Simon Gilroy, Committee Member
Teh Hui Kao, Committee Chair/Co-Chair - Keywords:
- Self-incompatibility
S-RNase
Male determinant
PiSLF - Abstract:
- Self-incompatibility (SI) is a genetic mechanism that allows pistils of flowering plants to distinguish between genetically related (self) and genetically unrelated (non-self) pollen, thereby preventing inbreeding and promoting outcrossing. The Solanaceae, Schrophulariaceae, and Rosaceae all possess S-RNase-based gametophytic self-incompatibility (GSI), which is controlled by two separate genes at the highly polymorphic S-locus: the S-RNase gene encodes the pistil determinant, and the previously unidentified pollen S-gene encodes the pollen determinant. An S-locus F-box gene of Petunia inflata, named PiSLF, was identified after the sequence analysis of a 328-kb contig that contains the S-RNase gene. PiSLF was considered the prime candidate gene for encoding the pollen determinant of SI for the following reasons: 1) it is specifically expressed in pollen and anthers; 2) it shows allelic sequence polymorphism; and 3) it co-segregates with the S-RNase gene in an S-haplotype specific manner. The major goal of this thesis research was to functionally characterize PiSLF in order to establish its potential role in controlling pollen behavior during the SI interaction. In Chapter 2, I describe the use of gain-of-function experiments to directly examine the hypothesis that PiSLF encodes the pollen determinant. The experimental design was based on the competitive interaction phenomenon, in which pollen carrying two different pollen S-alleles (heteroallelic pollen), but not pollen carrying two copies of the same pollen S-allele (homoalleleic pollen), fails to function during the SI response. I introduced the S2-allele of PiSLF into S1S1, S1S2, and S2S3 self-incompatible P. inflata plants and showed that the presence and expression of the PiSLF2 transgene specifically broke down the SI function of heteroallelic pollen but not that of homoallelic pollen. I concluded that PiSLF encodes the pollen self-incompatibility determinant. Having established the role of PiSLF in SI, I next addressed a fundamental question about the biochemical mechanism of SI: how do the pollen and pistil determinants interact to elicit S-haplotype-specific rejection of pollen tubes? Specifically, the question for the S-RNase-based mechanism is: how does a PiSLF interact with its self and non-self S-RNases differently so that only self S-RNase is allowed to exert its cytotoxic action in a pollen tube? To explain the biochemical basis of this type of SI mechanism, a “simple” inhibitor model was first proposed, which postulates that pollen S-allele products are cytotoxic inhibitors of S-RNases and that they contain an S-allele-specificity domain and an RNase inhibitor domain. A “modified” inhibitor model was subsequently proposed, which postulates that pollen S-allele products only contain the S-allele-specificity domain and that the inhibitor function resides in a general RNase inhibitor. To ascertain whether PiSLF contains both specificity and inhibitor functions, as predicted by the simple inhibitor model, I carried out loss-of-function experiments by introducing antisense- and RNAi-PiSLF2 constructs into S2S2 plants of P. inflata to suppress the expression of the endogenous PiSLF2. If absence of PiSLF2 is lethal to transgenic pollen, the results would be consistent with the simple inhibitor model. If transgenic pollen becomes compatible with pistils of any S-genotypes, the results would be consistent with the modified inhibitor model. Since the results of the antisense and RNAi experiments, described in Chapter 3, are inconclusive, it remains to be determined whether one of the inhibitor models is valid or neither is. Using in vitro binding and yeast two-hybrid assays, PiSLF was found to interact with two different proteins, S-RNases and PiSBP1 (a RING-finger protein), but, except for the N-terminal F-box domain, no other protein-protein interaction motifs could be recognized by conventional domain-prediction programs, such as SMART and Pfam. I used InterProScan and Gestalt domain detection algorithm (GDDA) programs to identify any motifs of PiSLF that could potentially be involved in interactions with S-RNases, PiSBP1 and other proteins. The results revealed the presence of an F-box-associated domain (amino acid residues 112 to 350) and a galactose oxidase central domain, which includes a kelch-like motif (amino acid residues 40 to 356) in the C-terminal region of PiSLF2. The results are included in Chapter 4. To determine whether these domains are indeed involved in interactions with S-RNases, I generated a number of truncated PiSLF2 constructs for another graduate student to test the interactions between these truncated PiSLF2 proteins and S3-RNase. The preliminary results from in vitro binding assays suggest that the region between amino acids 81 and 297 is important for the interaction with S3-RNase. These results are also reported in Chapter 4. Finally, I describe in Chapter 4 how all the results I have obtained in my thesis research have advanced our understanding and laid a foundation for future studies of the S-RNase-based SI mechanism.