Identification of a new class of pistil-specific proteins of Petunia inflata that is structurally similar to, but functionally distinct from, the self-incompatibility factor HT
Pollen–pistil interactions are thought to involve a wide variety of intercellular recognition events controlled by diverse proteins and other molecules. One of the best characterized interactions is the S-RNase-based gametophytic self-incompatibility (GSI) system found in Solanaceae, Rosaceae and Scrophulariaceae. Although the S specificity of the pistil and the pollen in these families is determined by the S locus-encoded proteins S-RNase and SLF/SFB, respectively, these proteins alone are not sufficient for operation of the GSI reaction. Other factors are also required and are classified into three groups. To date, the only known factor is the pistil-expressed small asparagine-rich protein HT-B in three solanaceous genera Nicotiana, Lycopersicon and Solanum. HT-B is a Group 2 factor that is required for pollen rejection but do not affect S-RNase expression; factors in the other groups have not yet cloned. Here, we identified a new class of HT-like proteins in the style of Petunia inflata and named it HTL. Through alternative splicing, it was found that two isolated homologous HTL cDNAs, HTL-A and HTL-B, derived from a single gene. Like HT-B, HTL showed pistil-specific accumulation as well as significant sequence similarity to HT including conserved cystein residues at the C-terminal region and a signal peptide for extracellular localization. However, unlike HT-B, HTL lacked an asparagine-rich domain. Thus, it represents a new class of HT proteins. To determine whether HTL is involved in GSI function, RNA silencing constructs for HTL-A and HTL-B were introduced into self-incompatible P. inflata. Although several transgenic lines showed no detectable levels of both HTL-A and HTL-B transcripts, they retained normal GSI function and produced large fruits upon compatible pollination. This suggests that since silencing of the HTL gene alone is not sufficient to affect reproductive physiology, the gene is functionally distinct from the GSI factor HT-B.
KeywordsPetunia Pollen–pistil interactions RNA silencing Self-incompatibility
Pollen–pistil interactions represent a critical step in the successful reproduction of plants. Through the interactions, the pistil rejects undesirable pollen while allows desirable, compatible pollen to grow for fertilization. While it is thought that various molecules in both the pistil and the pollen are involved in these cell–cell recognition mechanisms, little is known about their identities and roles.
The most characterized pollen–pistil interaction is self-incompatibility (SI), which promotes out-crossing by enabling the pistil to reject genetically related pollen. The dominant class of SI is gametophytic SI (GSI) which is, in general, controlled by a multiallelic locus S; when the S allele of the pollen matches one of the two S alleles of the pistil, then the pollen is recognized to be self and is rejected (de Nettancourt 2001; McCubbin and Kao 2000). The molecular basis of the S specificity of GSI has been extensively studied in Solanaceae, Rosaceae and Scrophulariaceae. S specificity in both the pistil and the pollen is controlled by distinct but tightly linked genes at the S locus, pistil-S and pollen-S, respectively. The pistil-S product is a ribonuclease called S-RNase, which is thought to act as a cytotoxin to the self pollen tube (McClure et al. 1989, 1990; Anderson et al. 1986; Lee et al. 1994; Murfett et al. 1994; Xue et al. 1996; Sassa et al. 1992, 1996, 1997; Broothaerts et al. 1995). The product of pollen-S was unclear until very recently when molecular characterization of the S locus regions identified two F-box proteins, SFB and SLF, as pollen-S candidates (Lai et al. 2002; Entani et al. 2003; Qiao et al. 2004; Ushijima et al. 2003; Yamane et al. 2003). Further functional analysis showed that the Petunia F-box protein PiSLF encodes the S specificity of the pollen (Sijacic et al. 2004). Furthermore, Ushijima et al. (2004) and Sonneveld et al. (2005) showed that mutant S haplotypes of Prunus that confer pollen-specific self-compatibility are defective in the SFB gene, strongly supporting that SFB is the long sought after pollen-S gene.
Although the S specificity is exclusively controlled by the S locus products, genetic analyses also have shown that the S locus alone is not sufficient to operate GSI function; other factors are also required (Ai et al. 1991; Bernatzky et al. 1995; Hosaka and Hannemen 1998a, b). McClure et al. (2000) has classified these factors into three groups according to their functions. Group 1 factors directly affect the expression of S-specificity genes. Group 2 factors interact, either genetically or biochemically, with the specificity determinants and are required for pollen rejection but have no wider role in pollination. Group 3 factors include genes that function both in pollen rejection and in other pollen–pistil interactions (McClure et al. 2000). To date, the only factor characterized at the molecular level is the pistil-expressed small protein HT found in three solanaceous genera Nicotiana, Lycopersicon and Solanum (McClure et al. 1999; Kondo et al. 2002a, b; O’Brien et al. 2002). HT, which is characterized by an asparagine-rich region at the C-terminal and shows no significant homology to other proteins and, thus represents a new protein family (McClure et al. 1999). Although its biochemical function is not clear, HT is necessary for the pistil to reject self pollen (McClure et al. 1999; O’Brien et al. 2002). In Solanum and Lycopersicon, two highly related proteins, HT-A and HT-B, are found in a single species (Kondo et al. 2002a, b; O’Brien et al. 2002). In comparison to HT-B’s amino acid sequence, HT-A has a small deletion at the terminus of the mature protein. RNA silencing experiments showed that the pistil requires expression of HT-B but not HT-A for GSI function (O’Brien et al. 2002). To date, only HT-B has been found in Nicotiana. The biological role of HT-A is not yet known. In other solanaceous genera with S-RNase information such as Petunia and Physalis, involvement of HT in GSI is not yet clarified. Since HT does not affect the expression of S-RNase, it is an example of a Group 2 factor (McClure et al. 2000). Although some S-RNase binding proteins are possible candidates for Group 3 factors in Nicotiana (McClure et al. 2000; Cruz-Garcia et al. 2005), molecular information about these proteins and their role in pollen rejection is still limited.
Characterization of the proteins/genes expressed in pistils and pollen are important for understanding the molecular networking of GSI and other pollen rejection mechanisms. Here, we describe identification of two new HT-like proteins from the pistils of self-incompatible Petunia inflata, HTL-A and HTL-B (for HT-like protein A and B, respectively). HTLs show a sequence homology to HT but lack the asparagine-rich region; thus, they represent a new class of HT proteins. RNA silencing experiments showed that loss of HTL expression is not sufficient for affecting either self-incompatible or compatible reactions.
Materials and methods
Petunia inflata seeds were obtained from Dr. J.B. Power (University of Nottingham, UK). RT-PCR and RACE (Rapid Amplification of cDNA Ends; Frohman et al. 1988) procedures were used to isolate the two S-RNase encoding cDNAs S3L and Sk1 (DDBJ accession numbers AB094599 and AB094600, respectively). The S3L-RNase sequence was identical to the reported S3-RNase cDNA of P. inflata (Ai et al. 1990) with the exception of a single nucleotide difference that leads to the amino acid change of 135th valine residue (GTG) to leucine (CTG). A young flower bud of a petunia plant was pollinated to produce a selfed population, which was used to test the inheritance of the S3L- and Sk1-RNase genes. Genomic fragments of the RNase genes (ca. 400 b) were amplified by REDExtract-N-Amp Plant PCR kit (SIGMA, St. Louis, MO, USA) with a primer pair of FSSR1 (5′-ATGATCTGGAACGCCACTGG-3′) and RSSR1 (5′-CTTCGTGCTGTCCGTTTCATC-3′), and then treated with EcoT22I to cut Sk1 but not S3L. Segregation of the RNase genotypes fitted the expected 1:2:1 ratio, indicating that they are allelic (S3LS3L:S3LSk1:Sk1Sk1 = 7:25:16, 0.100<P<0.250). A mature flower of the S3LS3L homozygous plant was pollinated by an S3LSk1 heterozygous plant, and the RNase genotypes of the 21 progenies determined. The genotypes were all S3LSk1. Since the S3L pollen was rejected by the S3LSk1 pistil, this confirms that the cloned genes are alleles of the S locus. Progenies of Sk1Sk1 × S3LSk1 were all S3LSk1 (16 plants), providing further support that the clones correspond to S-RNases.
Cloning of cDNA and genomic fragment
Total RNA was extracted from the pistils of P. inflata (S3LS3L), and reverse transcribed with a d(T)17 adapter primer to generate cDNA as described in Sassa and Hirano (1998) and Ushijima et al. (2003). A PCR primer (FHT1, 5′-GTTCTTTTGATAATATCATCAG-3′) was designed on the basis of an alignment of reported HT sequences. The primer was designed to amplify the HT-like sequence of petunia by a 3′RACE with ExTaq DNA polymerase (Takara, Otsu, Japan) using the cDNA template and the 24-mer adapter sequence at the 3′ (Sassa and Hirano 1998). Primers RHT2 (5′-ACTTTTACTGCAGTACCCATG-3′) and R2PiHT1 (5′-TTACAACAGTTACATTCTTGGC-3′) were used for 5′RACE of HTL-A and HTL-B, respectively. A 3′RACE for amplification of full-length cDNAs for HTL-A and HTL-B was conducted using both an FPHT (5′-AGAATTTATTCATCAAAATGG-3′) as a gene-specific primer, and a high fidelity DNA polymerase Pyrobest (Takara). The PCR products were cloned into a pZErO-2 vector (Invitrogen, Carlsbad, CA, USA) and sequenced. The deduced amino acid sequences of petunia proteins were aligned with those of solanaceous HT proteins by using ClustalX with default parameter settings (Thompson et al. 1997). Phylogenetic tree of HT-like proteins was generated from the alignment by neighbor-joining method (Saitou and Nei 1987). Bootstrap analysis was performed for 1,000 replicates.
Genomic fragments of HTL-A and HTL-B genes were amplified from P. inflata DNA by a Pyrobest DNA polymerase with the primer pairs FPHT and RPHTPT15 (5′- CGCGGATCCTTAGGAAGAACATTTTTTAATAC-3′), and FPHT and R2PiHT1, respectively. They were then cloned and sequenced.
Gel blot analysis of DNA and RNA
Genomic DNA of P. inflata was extracted as described in Doyle and Doyle (1990). Ten microgram of DNA were digested by restriction enzymes, separated in an agarose gel, blotted onto a nylon membrane (Biodyne Plus; Pall), and hybridized with digoxygenin (DIG) labeled probes as described in Sassa and Hirano (1998).
Pistil RNA from both wild type and transgenic lines was extracted. Ten microgram were separated in a formaldehyde gel, blotted onto a nylon membrane and hybridized with DIG-labeled probes as described in Sassa et al. (1997). The probes specific to the HTL-A and HTL-B genes were amplified from their respective 3′RACE clones using FAHTmHlf (5′-CAAATGATCAACGCCAGAACAC-3′) and a vector-derived primer M13-47 (5′-CGCCAGGGTTTTCCCAGTCACGAC-3′) and FBHTmHlf (5′- CAAATGAAGGGCTGAACACAAG-3′) and M13-47, respectively.
Expression in E. coli, production of antiserum and immunoblot analysis
The coding region of HTL-A was amplified from the cDNA clone using FPHTPT15 (5′-GAAAGTATCATATGGTTGCTGCTAGGGAAATAAC-3′) and RPHTPT15 (5′-CGCGGATCCTTAGGAAGAACATTTTTTAATAC-3′) with a Pyrobest DNA polymerase. It was then digested by NdeI and BamHI and cloned into pET15b (Novagen, Darmstadt, Germany) to generate pETPiHTLA. pETPiHTLA was transferred to an E. coli strain Rosetta-gami (DE3) (Novagen) to express a recombinant HTL-A protein. This recombinant HTL-A protein was recovered from an insoluble fraction by solubilization in 8 M urea and was purified by adsorption to a HiTrap Chelating column according to the manufacture’s instruction (AmershamBioscience Corp., Piscataway, NJ, USA). The HTL-A protein was further separated into SDS-PAGE gels, recovered by electroelution and used to immunize a rabbit to obtain antiserum as described in Sassa and Hirano (1998).
Proteins were extracted from the various floral organs and leaves of P. inflata with an extraction buffer (8 M urea, 50 mM Tris–HCl pH 6.8, 0.2% 2-mercaptoethanol) and quantified by the method described in Bradford (1976). Equal volumes of 2x SDS-sample buffer (125 mM Tris–HCl pH 6.8, 5% SDS, 10% 2-mercaptoethanol, 20% glycerol) were mixed and loaded to Tricine SDS–PAGE gel as described in Schägger and von Jagow (1987). Separated proteins were transferred to a PVDF membrane and probed by the antiserum against recombinant HTLA protein as described in Sassa and Hirano (1998) and Ushijima et al. (2001).
Self-complementary hairpin RNA-mediated gene silencing constructs were prepared by using pHANNIBAL (Wesley et al. 2001). The HTL-A coding region was amplified by a Pyrobest polymerase with FPHTHAN (5′-GCTCTAGACTCGAGTAGAATTTATTCATCAAAATGG-3′) and RPHTHAN2 (5′-GCATCGATGGTACCATCAGGTGTACAACATTGGC-3′) and was cloned into pHANNIBAL in sense and antisense orientation separated by Pdk intron (Wesley et al. 2001). The resultant silencing cassette was released by NotI digestion and moved to a binary vector pART27 (Gleave 1992) to create pARTHANHTLA. A silencing cassette for HTL-B was prepared by amplifying HTL-B cDNA with FPHTHAN and RHANPHTB (5′-GCATCGATGGTACCGAAGTTATTCAGGTCATTCC-3′) and by cloning it into pHANNIBAL as described. The HTL-B silencing cassette was excised by SacI and SpeI, and inserted into SacI–XbaI sites of pBINPLUS (van Engelen et al. 1995) to obtain pBINHANHTLB. The silencing constructs pARTHANHTLA and pBINHANHTLB were introduced to Agrobacterium tumefaciens LBA4404 to transform P. inflata (S3LS3L) by leaf disk method (Horsch et al. 1985).
Isolation of a new class of HT-like genes from the pistils of P. inflata
Alternative splicing of the HTL gene
Organ-specific expression of the HTL protein
Effect of HTL gene silencing on self-incompatible and compatible pollinations
HTL-A and HTL-B are similar to the solanaceous HT proteins in terms of overall amino acid sequence, location of cystein residues, occurrence of putative signal peptide and style-specific accumulation. In addition, it is noteworthy that the deletion of five to seven residues in the amino terminal region represents a structural characteristic for isoform A of HT (O’Brien et al. 2002), and that the corresponding region of petunia HTL is much closer to the GSI factor isoform B than the non-SI factor isoform A. However, the significant structural difference between the petunia HTLs and other HTs is the lack of the asparagine-rich region in HTL. Although the asparagine-rich region is retained in HT-A proteins which are known not to be involved in SI (O’Brien et al. 2002), its biological role is still unknown and may constitute a necessary part of the HT-B function in the operation of SI.
Genomic sequence analysis showed that two isoforms of petunia HTL are derived from alternative splicing of a single gene. Although Solanum and Lycopersicon express two isoforms of HT transcripts, the nucleotide differences between the isoforms of a single species are dispersed in the sequences (Kondo et al. 2002a, b; O’Brien et al. 2002). This and RFLP analysis of the genes by O’Brien et al. (2002) suggest that the isoforms are encoded by different genes. Generations of variants through alternative splicing may be a characteristic of HTL.
In order to determine whether HTL is involved in GSI function, RNA silencing plants for the HTL gene were produced. The hairpin RNA induced gene silencing strategy (Wesley et al. 2001) was found to be very effective in knocking down the expression of HTL gene. While some transgenic lines showed no detectable levels of accumulation of both HTL-A and HTL-B transcripts, the silenced plants showed no morphological change and retained normal self-incompatibility and compatibility behavior. One possibility is that HTL is redundantly involved in GSI with unidentified related protein(s) and that knock-down of HTL alone is not sufficient to change reproductive physiology. Identification and silencing of other HT-like gene(s) in petunia, comparison of it with the HTL silencing phenotype and, if necessary, production of simultaneous silencing plants for both HT-like genes will be needed to test this possibility.
It is also likely that HTL is required under specific conditions not tested in this study. One possibility may be the involvement of HTL in an incompatibile reaction against pollen from different species. Another possibility is that HTL is involved in the defense mechanism of the pistil against invasion of pathogens. While pistils are very rich in the nutrients required for pollen tube growth and are, thus, vulnerable to pathogen invasion, infection of the pistil is rare (Atkinson et al. 1993). This is consistent with the findings that pistils accumulate varieties of potential antimicrobial proteins such as defferent pathogenesis-related (PR) proteins (eg., Sassa and Hirano 1998). Sessa and Fluhr (1995) showed that the suppression of the pistil-specific glucanase (PR3) by antisense approach had no influence on reproductive physiology of tobacco, and postulated that the glucanase is involved in defense of the pistil. It is noteworthy that a tobacco S-like RNase (RNaseNE) inhibits hyphal elongation of plant pathogens; this represents an additional cytotoxic activity of the extracellular class of plant RNase beyond the cytotoxic effect of the S-RNase on self pollen in the GSI system (Hugot et al. 2002). Moreover, it is also interesting that the S-RNase-based cytotoxic effect against self-pollen requires HT-B and that the petunia pistil accumulates HTL and the S-like RNases, the latter being implicated in the defense mechanism of the pistil (Singh et al. 1991; Lee et al. 1992). Although HT-B is postulated to interact with pollen tubes and to facilitate uptake of S-RNase, its biochemical function in SI mechanism is not yet known (McClure et al. 1999). The elucidation of the precise role of any member of HT-like protein family will provide insight into the biochemical function of other members and help us better understand the pollen–pistil interactions.
We thank Dr. J.B. Power for P. inflata seeds, Dr. P.M. Waterhouse for pHANNIBAL, Dr. A.P. Gleave for pART27 and Dr. W.J. Stiekema for pBINPLUS. Drs. Y. Ogawa, Y. Hoshino and H. Washida are acknowledged for their technical advice. We also thank Melody Kroll for proof reading this manuscript. This work was supported in part by the Grants-in-Aid for Scientific Research (C, 13660011) and the Grants-in-Aid for Young Scientists (A, 16688001) from the Ministry of Education, Science, Sports and Culture of Japan to H.S.
- de Nettancourt D (2001) Incompatibility and incongruity in wild and cultivated plants. Springer, Berlin Heidelberg New YorkGoogle Scholar
- Doyle JJ, Doyle JL (1990) Isolation of plant DNA from fresh tissue. Focus 12:13–15Google Scholar
- O’Brien M, Kapfer C, Major G, Laurin M, Bertrand C, Kondo K, Kowyama Y, Matton DP (2002) Molecular analysis of the stylar-expressed Solanum chacoense small asparagine-rich protein family related to the HT modifier of gametophytic self-incompatibility in Nicotiana. Plant J 32:985–996CrossRefPubMedGoogle Scholar
- Olmstead RG, Sweere JA, Spangler RE, Bohs L, Palmer JD (1999) Phylogenetic and provisional classification of the Solanaceae based on chloroplast DNA. In: Nee M, Symon DE, Lester PN, Jessop JP (eds) Solanaceae IV. Royal Botanic Gardens, Kew, pp 111–137Google Scholar
- Qiao H, Wang H, Zhao L, Zhou J, Huang J, Zhang Y, Xue Y (2004) The F-box protein AhSLF-S2 physically interacts with S-RNases that may be inhibited by the ubiquitin/26S proteasome pathway of protein degradation during compatible pollination in Antirrhinum. Plant Cell 16:571–581CrossRefPubMedGoogle Scholar
- Sassa H, Hirano H, Ikehashi H (1992) Self-incompatibility-related RNases in styles of Japanese pear (Pyrus serotina Rehd.). Plant Cell Physiol 33:811–814Google Scholar