A Molecular Description of Mutations Affecting the Pollen Component of the Nicotiana alata S locus

Corresponding author: A. E. Clarke, Plant Cell Biology Research Center, School of Botany, University of Melbourne, Parkville, Victoria 3052, Australia., a.clarke{at}botany.unimelb.edu.au (E-mail)

Communicating editor: J. CHORY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Mutations affecting the self-incompatibility response of Nicotiana alata were generated by irradiation. Mutants in the M₁ generation were selected on the basis of pollen tube growth through an otherwise incompatible pistil. Twelve of the 18 M₁ plants obtained from the mutagenesis screen were self-compatible. Eleven self-compatible plants had mutations affecting only the pollen function of the S locus (pollen-part mutants). The remaining self-compatible plant had a mutation affecting only the style function of the S locus (style-part mutant). Cytological examination of the pollen-part mutant plants revealed that 8 had an extra chromosome (2n + 1) and 3 did not. The pollen-part mutation in 7 M₁ plants was followed in a series of crosses. DNA blot analysis using probes for S-RNase genes (encoding the style function of the S locus) indicated that the pollen-part mutation was associated with an extra S allele in 4 M₁ plants. In 3 of these plants, the extra S allele was located on the additional chromosome. There was no evidence of an extra S allele in the 3 remaining M₁ plants. The breakdown of self-incompatibility in plants with an extra S allele is discussed with reference to current models of the molecular basis of self-incompatibility.

FERTILIZATION in flowering plants begins when a pollen grain bearing the male gametes lands on a female stigma. Several mechanisms enable the stigma and style to discriminate between the different types of pollen it may receive, the best studied being self-incompatibility. If a pollen grain from a self-incompatible plant lands on its own stigma, or on the stigma of a genetically related plant, the pollen either will fail to germinate or will germinate to produce a pollen tube that grows poorly in the style and does not reach the ovary (DE NETTANCOURT 1977 ). In many cases, this process is controlled by a single, multiallelic locus called the S locus. In solanaceous plants such as Nicotiana alata (ornamental tobacco), the S locus acts gametophytically and a haploid pollen grain is rejected by a diploid style when the same S allele is present in both. The only known product of the solanaceous S locus is an extracellular ribonuclease produced by the style (the S-RNase; MCCLURE et al. 1989 ). S-RNases control the stylar phenotype of self-incompatible plants but do not control the pollen phenotype (LEE et al. 1994 ; DODDS et al. 1999 ). This suggests the S locus is bipartite, with different genes encoding the pollen component (pollen-S) and the style component (S-RNase) of the S locus. The product of the pollen-S gene is not known. As part of a strategy to define the nature of this product, we have generated a series of pollen-part mutations of the S locus (pollen-part mutant, PPM).

In previous studies, mutations affecting the pollen component of the S locus have been generated in N. alata (PANDEY 1965 , PANDEY 1967 ; VAN GASTEL and DE NETTANCOURT 1975 ), Petunia inflata (BREWBAKER and NATARAJAN 1960 ), and Solanum tuberosum (OLSDER and HERMSEN 1976 ; HERMSEN 1978 ). Styles of PPM plants retain the ability to reject incompatible pollen.

Different types of lesions can cause mutations affecting the self-incompatibility response of pollen. The majority of pollen-part mutations in solanaceous plants are associated with duplications of an S allele (BREWBAKER and NATARAJAN 1960 ; PANDEY 1965 , PANDEY 1967 ; VAN GASTEL and DE NETTANCOURT 1975 ). The duplicated S allele is frequently on a short, additional chromosome known as a centric fragment, which segregates independently of the S locus. The self-incompatibility phenotype of pollen from these plants is only altered when the duplicated S allele and the allele present at the S locus are different (BREWBAKER and NATARAJAN 1960 ). This phenomenon is called competitive interaction and requires two different S alleles to be present in the plant. In addition, competitive interaction results in progeny of backcross and selfed families having two different S alleles (PANDEY 1967 ; VAN GASTEL and DE NETTANCOURT 1975 ; see DE NETTANCOURT 1977 ).

A lesion in the pollen-S gene causes the other type of mutation affecting the self-incompatibility response of pollen. These are "true" pollen-part mutations and can be distinguished from plants carrying duplicated S alleles because they may be homozygous for S alleles and can produce homozygous progeny following backcross or self-pollinations.

Because none of the N. alata PPM plants generated in previous studies were available, we generated PPM plants using the same strategy applied in earlier studies (PANDEY 1967 ; VAN GASTEL and DE NETTANCOURT 1975 ). Following irradiation of S₃S₆ N. alata plants, 18 M₁ individuals were isolated and characterized by pollination, cytological examination of root tip cells, DNA blot analysis with S-RNase gene probes, and protein blot analysis with S-RNase-specific antibodies. Eleven plants had mutations affecting the pollen component of the S locus and inheritance of the pollen-part mutation in seven of these PPM plants was followed through a series of crosses. The nature of the mutation in these plants is discussed with reference to current models of self-incompatibility.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Screen for pollen-part mutants:
Mature N. alata plants (genotype S₃S₆) received a total dose of either 8 or 10 Gy from a ⁶⁰Co source (1.4 Gy/min) housed at the CSIRO Division of Plant Industry, Canberra, Australia. The target tissue was floral buds containing pollen mother cells (see DODDS et al. 1993 ). Of 149 buds irradiated, 108 received 8 Gy and 41 received 10 Gy. Following irradiation, the floral buds were labeled and the pollen was collected at anthesis. This was either used directly in pollinations or stored at -70° until needed. Whenever practicable, each pollen sample was used to pollinate two flowers from an unmutated S₃S₆ plant (i.e., an incompatible pollination). Following each pollination, 1% indole-3-acetic acid in lanolin was applied to the base of the flower. At maturity, the capsules were opened and the seeds collected. Before germination, seeds were surface sterilized for 1 hr with a hypochlorite solution (1% HClO in 0.1% Tween) and rinsed thoroughly in sterile water. Seeds were then placed on sterile MS agar containing 3% sucrose and incubated at 22°. When sufficiently strong, each seedling was transplanted into soil and grown as described by ANDERSON et al. 1989 .

Pollination analysis:
Plants were self-pollinated by spreading pollen from a dehiscent anther over the stigmas of four or more flowers. A pollination was compatible if a large capsule developed and incompatible if the flower abscised in the week following pollination. To determine the stylar self-incompatibility phenotype of a plant, immature floral buds were emasculated and pollinated with pollen from a plant of known S genotype soon after petal opening. Four such pollinations were usually done for each plant. Similar crosses were used to determine the self-incompatibility phenotype of pollen from the plant.

DNA blot analysis:
Genomic DNA was extracted from the leaves of N. alata plants as described by BERNATZKY and TANKSLEY 1986 . Leaf DNA (5 µg) was digested to completion with HindIII or BamHI (Promega, Madison, WI). The DNA was fractionated on a 0.8% agarose gel run in 1x TBE and transferred to a nylon membrane (Amersham, Buckinghamshire, UK) as described by SAMBROOK et al. 1989 . S-RNase cDNA fragments were radiolabeled with random primers (Primagene, Promega). Hybridization of the radiolabeled S-RNase cDNAs to the DNA blots was done in 50% formamide, 5x SSPE, 5x Denhardt's solution, 0.5% SDS, 50 µg/ml denatured herring sperm DNA at 42° for 12 hr. After hybridization, the membranes were washed twice (30 min each time) at 42° in 0.2x SSPE, 0.2% SDS and exposed to film.

Western blot analysis:
Styles were collected and stored at -70°. Proteins were extracted from plant tissue in an extraction buffer (100 mM Tris-HCl, pH 8, 50 mM EDTA, 0.1% polyvinylpyrrolidone, 28 mM ß-mercaptoethanol) to give a 25% solution. Protein concentrations were determined using a colorimetric assay (BRADFORD 1976 ) with BSA as a standard. Stylar proteins (15 µg) were fractionated on a 15% SDS-polyacrylamide gel according to the method of LAEMMLI 1970 . Proteins were transferred to a nitrocellulose membrane (Schleicher and Schuell, Dassel, Germany) in transfer buffer (48 mM Tris-HCl; 39 mM glycine; 0.0374% SDS; 20% methanol) using a semidry electrophoretic transfer cell apparatus (Transblot, Bio-Rad, Richmond, CA). A replicate gel with 5 µg stylar proteins in each lane was stained with silver (Bio-Rad). The membrane was incubated with a protein G-purified sheep antiserum for the S₆-RNase (DODDS et al. 1993 ) as described by HARLOW and LANE 1988 . Bound antibodies were detected using biotinylated anti-sheep immunoglobulin and streptavidin-horseradish peroxidase (Amersham) according to the manufacturer's instructions.

Cytology:
Root tips were collected from hydroponically grown plants. After harvest, the root tips were placed in a saturated solution of -bromonaphthalene and incubated for 2 hr at room temperature with occasional agitation. Root tips were then washed with water and fixed in ethanol:acetic acid (3:1) for 12 hr at 4°. After fixation, the root tips were placed in a 70% ethanol solution and stored at 4° for up to 1 month before analysis. For cytology, fixed root tips were treated with 0.2 N HCl solution for 10 min at 55°. After acid hydrolysis, the root tips were washed with water and placed in a staining solution [2% synthetic orcein (Gurr) in 45% acetic acid] for 40 min. Root tips were destained in a solution of 45% acetic acid. Macerated root tips were spread and examined under phase contrast optics using a Zeiss Universal microscope. Images were captured with a Zeiss MC63 photographic unit using Tmax100 film (Kodak, Rochester, NY).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Production of the M₁ generation:
Developing N. alata flower buds (genotype S₃S₆) were irradiated with either 8 or 10 Gy of -rays from a ⁶⁰Co source. Pollen was subsequently collected at anthesis and used to pollinate pistils of unmutated S₃S₆ plants. From 300 pollinations, only three capsules that contained seeds were recovered. Two capsules, containing 7 viable seeds, were produced following pollination with pollen irradiated with 8 Gy and one capsule, containing 11 seeds, was from pollen irradiated with 10 Gy. The seeds were germinated and the seedlings were grown to maturity. Plants were numbered M1-1 to M1-18. M1-1 to M1-4 came from one capsule, and M1-5 to M1-7 from the other capsule formed by pollen irradiated with 8 Gy. The remaining plants came from the capsule formed by pollen irradiated with 10 Gy. The plants grew normally and were not visibly different from nonmutagenized N. alata plants.

Twelve of the 18 M₁ plants formed large capsules after self-pollination and the remaining 6 plants were self-sterile (see Table 1; data are incomplete for M1-3 and M1-4, which were self-sterile but were not characterized further). Crosses to N. alata plants of known S genotypes were used to characterize the pollen and pistil selfincompatibility phenotype of each M₁ plant (Table 1). Ten plants had pistils that rejected pollen from S₃S₃ and S₆S₆ plants, which indicated their pistil phenotype was S₃S₆; five plants had pistils that rejected S₃ pollen but accepted S₆ pollen and therefore had S₃S₃ as their pistil phenotype; and one plant rejected S₆ pollen but accepted S₃ pollen, which indicated its pistil phenotype was S₆S₆.

Capsules formed following the pollination of S₃S₆ pistils with pollen from 11 of the 12 self-fertile M₁ plants. This showed these plants carried mutations affecting the self-incompatibility phenotype of their pollen. Capsules did not form following similar pollinations using pollen from the self-fertile plant M1-18, indicating this plant carried a mutation affecting the self-incompatibility phenotype of its styles (a style-part mutant, SPM). M1-18 and 3 of the 4 self-sterile M₁ plants (M1-8, M1-14, and M1-17), produced viable pollen (capsules formed after pollination of a compatible S₂S₂ pistil). Interestingly, these plants did not have mutations affecting the self-incompatibility phenotype of their pollen, even though they were all grown from seed formed after an incompatible pollination. Presumably the normal self-incompatibility response of M1-8, M1-14, M1-17, and M1-18 pollen arose because the mutation that had allowed the pollen tubes to grow through an incompatible style "reverted" to an unmutated state after fertilization. The self-incompatible M₁ plants are therefore described as revertants (REVs). M1-13 did not produce viable pollen as no capsules formed after pollination of an S₂S₂ plant.

DNA blot analysis of M₁ plants:
The S genotype of 16 M₁ plants was determined by DNA blot analysis using the S₃- and S₆-RNase cDNAs as probes (Figure 1). The S₃-RNase gene was present in all plants except M1-5 (S₆S₆). Similarly, the S₆-RNase gene was present in all plants except M1-8 and M1-17 (both S₃S₃). Thus, with the exception of M1-6 and M1-7 (both S₃S₃) and M1-18 (S₃S₃ SPM), the S genotype of the M₁ plants determined by DNA blot analysis matched the pistil phenotype.

View larger version (57K): In this window In a new window Download PPT slide   Figure 1. DNA blot analysis of the M1 plants. Each lane contains genomic DNA (5 µg) from the indicated M1 plant or an unmutated S3S6 plant (WT). DNA was digested with BamHI (top) or HindIII (bottom) and probed with the S3-RNase cDNA and the S6-RNase cDNA separately (top) or both the S3- and S6-RNase cDNAs (bottom). Molecular weight standards (in kilobases) are shown on the left of the figure and the identity of the S-RNase hybridizing bands is indicated at the right of the figure.

Detection of S-RNases in the pistils of M₁ plants:
To understand the discrepancy between the S phenotype and S genotype in plants M1-6, M1-7, and M1-18, the accumulation of S-RNases by the styles of these plants was assessed by SDS-PAGE and Western blot analysis. Buffer-soluble proteins were extracted from the styles (including stigmas) of unmutated plants (S₃S₃, S₆S₆, and S₃S₆), and from the indicated M₁ plants (Figure 2). Stylar protein was separated by SDS-PAGE and either stained with silver (Figure 2A) or transferred to a nitrocellulose membrane and incubated with an antiserum specific for the S₆-RNase (Figure 2B).

View larger version (104K): In this window In a new window Download PPT slide   Figure 2. S-RNase accumulation by the pistils of unmutated N. alata plants and selected M1 plants. Buffer-soluble protein was extracted from pistils of the indicated M1 plants and pistils of unmutated S3S3, S6S6, and S3S6 plants. M1 plants that have an S6-RNase gene but fail to reject S6 pollen are indicated (SPM). Proteins were separated by SDS-PAGE and either stained with silver (A) or blotted onto nitrocellulose and incubated with an antiserum specific to the S6-RNase (B). Protein bands corresponding to the S3- and S6-RNases (A) or the S6-RNase (B) are indicated. Molecular weight standards (in kilodaltons) are shown at the left of the figure.

In Figure 2A, the S₃-RNase appeared as a band of M_r 35 kD and the S₆-RNase as a band of M_r 33 kD. In Figure 2B, the S₆-RNase appeared as a major band of 33 kD identical in size to the protein seen in Figure 2A. A second band of M_r 31.5 kD and a third, less abundant band of M_r 30 kD were also seen. The amount of S₃- and S₆-RNase extracted from styles of M1-11 and the other nine S₃S₆ M₁ plants (data not shown) was similar to that in the style of an unmutated S₃S₆ plant. Likewise, the amount of S₃-RNase in the style of M1-8 and M1-17 (both S₃S₃) and the amount of S₆-RNase in M1-5 styles (S₆S₆) was similar to that in unmutated S₃S₃ and S₆S₆ styles, respectively (Figure 2A and data not shown). However, although M1-6, M1-7, and M1-18 had the S₆-RNase gene, little or no S₆-RNase could be detected in their styles. It is likely that the level of S₆-RNase in these plants was less than that required to reject S₆ pollen (Table 1). All three plants are therefore SPMs with a lesion affecting the style-part of the S₆ allele. This allele will be referred to as S^spm₆. Plant M1-18 had this mutation alone; plants M1-6 and M1-7 had pollen-part mutations as well. The results of protein and Western analyses of all M₁ plants are summarized in Table 2.

Cytology:
Mitotic chromosomes in the root tips of an unmutated S₃S₆ plant and all M₁ plants were stained with orcein and examined by phase-contrast microscopy. Typically, four root tips were examined from each plant and the number of chromosomes in at least four cells from each root tip was counted. Figure 3 shows representative examples of these cells. The results for each M₁ plant are summarized in Table 2. The cells of an unmutated N. alata plant and four M₁ plants contained 18 chromosomes (Figure 3A and Figure B), which is the expected number of chromosomes in this species. The morphology of individual chromosomes also matched an earlier description of N. alata chromosomes made by CARLUCCIO et al. 1974 .

View larger version (102K): In this window In a new window Download PPT slide   Figure 3. Micrographs of metaphase chromosomes in the root tip cells of an unmutated S3S6 plant (A) and representative M1 plants (B–F). The unmutated S3S6 plant (A) and plant M1-5(B) both contain 18 chromosomes. Nineteen chromosomes were seen in M1-1 (C), M1-6 (D), and M1-14 (E). The additional chromosome in M1-1 and M1-6 (C and D) was smaller than the rest (arrow). All chromosomes in M1-14 (E) were similar in size. Cells from M1-13 contained 28 chromosomes (F). Bar in A, 5 µm.

Eleven of the M₁ plants contained 19 chromosomes (Table 2) and one M₁ plant (M1-13) contained 28 chromosomes (Figure 3F). In plants with 19 chromosomes, the additional chromosome was generally smaller than the other chromosomes and varied in length from 1 µm (for example, M1-1; Figure 3C) to 1.7 µm (for example, M1-6; Figure 3D). In plant M1-14, none of the chromosomes was noticeably shorter than the others, making it difficult to say which was additional (Figure 3E). When identifiable, the additional chromosome had a constriction indicative of a centromere. In keeping with the nomenclature used by earlier researchers, the additional chromosome will be referred to as a centric fragment.

Breeding analysis of four PPM M₁ plants that had a centric fragment:
One hypothesis to account for the pollen-part mutation in the four PPM M₁ plants with a centric fragment is that the centric fragment in these plants carries a duplicated S allele. This hypothesis was tested by correlating the presence of a centric fragment with the S phenotype and S genotype of plants produced by backcrossing an M₁ plant to an unmutated S₃S₆ plant or outcrossing it to an unmutated S₂S₂ plant. In one case, a family produced by self-pollinating an M₁ plant was used instead of a backcrossed family. S genotypes were determined by DNA blot analysis using S-RNase cDNAs as probes. Table 3 summarizes the results of this analysis for plants M1-1, M1-2, M1-6, and M1-11.

All plants in the backcross family of M1-1 had S₃S₆ as their pistil phenotype and, with one exception, all were PPMs. The S genotype of five PPM plants from the backcross family was determined by DNA gel blot analysis and, as expected, both the S₃- and S₆-RNase genes were present. Cytological analysis of root tip cells from two PPM plants showed both plants contained 19 chromosomes, with the additional chromosome apparently identical to the centric fragment in plant M1-1 (Figure 3C).

Two types of plants were present in the outcross family of M1-1; self-incompatible S₂S₆ plants and PPM plants with an S₂S₃S₆ (triallelic) pistil phenotype. DNA blot analysis of five plants from each class confirmed the presence of the S₂- and S₆-RNase genes in the self-incompatible plants and the S₂-, S₃-, and S₆-RNase genes in the triallelic PPM plants (Figure 4A). Cytological analysis found a centric fragment in the root tip cells of two triallelic PPM plants but not in two self-incompatible S₂S₆ plants. Similar classes of plants were also found in families produced by outcrossing M1-1 to S₁S₁ and S₇S₇ plants (data not shown).

View larger version (59K): In this window In a new window Download PPT slide   Figure 4. DNA blot analysis of PPM plants with a centric fragment and representative progeny. Genomic DNA was isolated from the indicated plant, digested with HindIII and fractionated on an agarose gel. After transfer onto a nylon membrane, the blots were probed with 32P-labeled S-RNase cDNA probes. (A) M1-1 and representatives of the phenotypic classes (self-incompatible S2S6 and S2S3S6 PPMs) identified in the S2S2 outcross family (see Table 3). DNA from unmutated S2S2 and S3S6 plants is also shown. (B) M1-2 and representatives of the phenotypic classes identified in the S3S6 backcross family (S3S6 PPM and S6S6 PPM plants). (C) M1-2 and representatives of the phenotypic classes identified in the S2S2 outcross family (self-incompatible S2S3 and S2S6 plants and a S2S6 PPM plant). (D) M1-6 and representatives of the phenotypic classes identified in the S3S6 backcross family (S3S3 SPM/PPM and S3S6 PPM plants). (E) M1-11 and representatives of the phenotypic classes identified in the S3S6 backcross family (S3S3 SPM/PPM and S3S6 PPM plants). Molecular weight standards (in kb) are shown on the left of the figure and the S-RNase hybridizing bands are indicated at the right of the figure.

PPM plants with either S₃S₆ or S₆S₆ as their pistil phenotype were found in the backcross family of M1-2 (Table 3). DNA gel blot analysis detected both the S₃- and S₆-RNase genes in representative S₃S₆ plants but only the S₆-RNase gene in representative S₆S₆ plants (Figure 4B). Both types of PPM plant inherited the centric fragment present in plant M1-2.

Three types of plants were found in the outcross family of M1-2: self-incompatible plants, either S₂S₃ or S₂S₆, and an S₂S₆ PPM plant. The expected S-RNase genes were detected by DNA blot analysis in representatives of each type of plant (Figure 4C). The S₂S₆ PPM plant and at least one of the S₂S₃ self-incompatible plants inherited the centric fragment from M1-2 (Table 3). Other self-incompatible plants lacked centric fragments. Similar types of plants were also found in a family produced by crossing M1-2 to an S₁S₁ plant (data not shown).

S₃S₃ or S₃S₆ PPM plants were found in the backcrossed family of M1-6 and the selfed family of M1-11 (Table 3). The S₃- and S₆-RNase genes were detected in representative S₃S₆ and S₃S₃ plants from both families (Figure 4D and Figure E). This indicated some plants inherited an S^spm₆ allele. As mentioned above, plant M1-6 had a mutation affecting expression of the S₆-RNase gene and the S₃S₃ PPM plants in the backcross family presumably inherited this mutation. SDS-PAGE and Western analyses of plants from the selfed family of M1-11 found trace levels of S₆-RNase in the pistils of S₃S₃ PPM plants (data not shown). This showed M1-11 had the S^spm₆ allele and was both a pollen and a style-part mutant (SPM/PPM). Both S₃S₃ SPM/PPM and S₃S₆ PPM plants inherited centric fragments from M1-6 and M1-11 (Table 3).

Three types of plants were found in the outcrossed family of plant M1-6: self-incompatible S₂S₃ plants, S₂S₂ SPM plants, and a S₂S₃ PPM plant (Table 3). DNA blot analysis found the S₂- and S₆-RNase genes in the S₂S₂ SPM plants and S₂-, S₃-, and S₆-RNase genes in the S₂S₃ PPM plant. Both types of plants had presumably inherited the S^spm₆ allele from M1-6. Cytological examination found a centric fragment in the PPM plant.

A similar range of pollination phenotypes was also present in the outcross family of plant M1-11, except that this family included self-incompatible S₂S₆ plants and S₂S₃S₆ PPM plants (Table 3). DNA blot analysis found the S₂- and S₆-RNase genes in the S₂S₆ self-incompatible and the S₂S₂ SPM plants and the S₂-, S₃-, and S₆-RNase genes in the S₂S₃ and S₂S₃S₆ PPM plants. The S^spm₆ allele from plant M1-11 had therefore been inherited by the S₂S₂ SPM and S₂S₃ SPM/PPM plants. Cytological analysis found a centric fragment in the root tip cells of four PPM plants, but not in either the self-incompatible or SPM plants.

Breeding analysis of three PPM M₁ plants that lacked a centric fragment:
One way to account for the pollen-part mutation in the three PPM plants that lack a centric fragment is to assume the mutation is caused by a lesion in the pollen-S gene. This hypothesis was tested by examining the S phenotype and S genotype of plants produced by backcrossing an M₁ plant to an unmutated S₃S₆ plant or outcrossing it to an unmutated S₂S₂ plant. S genotype was determined by DNA blot analysis using S-RNase cDNAs as probes. Table 4 summarizes breeding data for plants M1-5, M1-7, and M1-10.

PPM plants with S₃S₃, S₃S₆, and S₆S₆ pistil phenotypes were found in the backcross family of M1-5 (Table 4). DNA gel blot analysis detected the S₃- and S₆-RNase genes in S₃S₃ and S₃S₆ PPM plants, but only the S₆-RNase gene was present in S₆S₆ PPM plants (Figure 5A). SDS-PAGE and Western analyses found only trace amounts of the S₆-RNase in the pistils of S₃S₃ PPM plants, showing these plants had inherited an S^spm₆ allele previously undetected in M1-5 (data not shown).

View larger version (33K): In this window In a new window Download PPT slide   Figure 5. DNA blot analysis of M1-5 and representative progeny. Genomic DNA was isolated from the indicated plant, digested with HindIII, and fractionated on an agarose gel. After transfer onto a nylon membrane, the blots were probed with 32P-labeled S-RNase cDNA probes. (A) M1-5 and representatives of the phenotypic classes identified in the S3S6 backcross family (S3S3 SPM/PPM, S3S6 PPM, and S6S6 PPM plants). (B) M1-5 and representatives of the phenotypic classes identified in the S2S2 outcross family (self-incompatible S2S6 plants, S2S2 SPM/PPM, and S2S6 PPM plants). Molecular weight standards (in kb) are shown on the left of the figure and the S-RNase hybridizing bands are indicated to the right of the figure.

Self-incompatible S₂S₆ plants and PPM plants with either S₂S₂ or S₂S₆ as their pistil phenotype were found in the outcross family of M1-5 (Table 4). The S₂- and S₆-RNase genes were found in all plants tested (Figure 5B). The S₂S₂ PPM plant was presumably also an SPM and had inherited the S^spm₆ allele from M1-5.

PPM plants with either S₃S₃ or S₃S₆ as their pistil phenotype were found in the backcross family of plant M1-7. As plants from both classes contained the S₃- and S₆-RNase genes, the S₃S₃ PPM plants presumably inherited the S^spm₆ allele present in plant M1-7 (see above). Two types of plants were found in the outcrossed family of plant M1-7: self-incompatible S₂S₃ plants and S₂S₂ SPM plants (Table 3). No PPM plants were among the 21 plants examined. The S₂S₂ SPM plants in this family presumably inherited the S^spm₆ allele and, consistent with this, DNA blot analysis found the S₂- and S₆-RNase genes in these plants.

S₃S₆ and S₆S₆ PPM plants were found in the backcross family of M1-10 (Table 4). DNA gel blot analysis detected the S₃- and S₆-RNase genes in the S₃S₆ PPM plants, but only the S₆-RNase gene in the S₆S₆ PPM plants (data not shown). Self-incompatible S₂S₃ and S₂S₆ plants and PPM plants with an S₂S₆ pistil phenotype were found in the outcross family of M1-10 (Table 4).

Breeding analysis of two revertant M₁ plants with a centric fragment:
The revertant plants, M1-8 and M1-17, were crossed as the male parent to self-incompatible S₂S₂ plants. All progeny of these crosses were self-incompatible with an S₂S₃ pistil phenotype (Table 5). DNA gel blot analysis detected the S₂- and S₃-RNase genes in all the plants. The centric fragment was found in two of four plants examined from each family.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Our study of pollen-part mutations is part of a broader study aimed at identifying the pollen-S gene. Earlier examinations of pollen-part mutations relied on pollination behavior to identify S alleles in individual plants. A significant advance of this study was the availability of cDNA probes to identify a particular S allele and specific antibodies to identify S-RNase products of the S locus. This gives us an independent means of identifying S alleles and an opportunity to get a more precise description of the nature of the mutations.

Ionizing radiation causes chromosomal alterations such as inversions and deletions and has frequently been used to induce pollen-part mutations (DE NETTANCOURT 1977 ). The classes of plants produced by mutagenesis in this study were similar to those described earlier (PANDEY 1967 ; VAN GASTEL and DE NETTANCOURT 1975 ). Self-compatibility was mainly due to mutations affecting pollen phenotype (PPMs). The majority of PPMs expressed two S alleles (S₃ and S₆) in their pistil but a few expressed a single S allele (either S₃ or S₆). The proportion of homozygous to heterozygous PPMs seen in this study was also similar to that reported earlier. As observed in earlier studies, many plants had a centric fragment.

Style and pollen-part mutations are independent:
We identified style-part mutations (SPMs) in three M₁ plants (M1-6, M1-7, and M1-18). Each plant had an S^spm₆ allele and two had a pollen-part mutation as well (Table 2). Molecular analyses of plants in the M₂ generation also identified S^spm₆ alleles in plants M1-5 and M1-11. Functional S₆ alleles masked the style-part mutations in these M₁ plants, which indicates the S^spm₆ allele is recessive. In plants with an S^spm₆ allele and no functional S₆ allele, low levels of stylar S₆-RNase were detected by Western blotting. The lesion in the S^spm₆ allele is unknown but appears to act in cis and affects the level of expression of the S₆-RNase gene. However, our analysis of the M₂ families indicates the S^spm₆ allele may not be completely penetrant and can in some instances become a functional S₆ allele. Breeding experiments showed that the pollen function of the S^spm₆ allele is normal and the mutation is independent of the pollen-part mutation. For example, 10 of the outcross progeny of M1-11 inherited the S^spm₆ allele but not the pollen-part mutation (Table 3). Style-part mutations affecting the expression of the S₆ allele have also been noted in N. alata plants recovered from tissue culture (H. DU, A. E. CLARKE and T. BACIC, personal communication). The S^spm₆ allele in these plants arose in the absence of irradiation. It may well be that the S^spm₆ allele described here also arose spontaneously in our stock lines and was not produced by irradiation. This mutation is not discussed further.

Evidence of a duplicated S allele in four M₁ plants:
The inheritance of the pollen-part mutation in seven M₁ plants was studied. Four plants (M1-1, M1-6, M1-7, and M1-11) had a duplicated S₃ allele as judged by DNA blot analysis with S-RNase probes (Table 6) and the three remaining plants (M1-2, M1-5, and M1-10) did not (Table 7). We will discuss the pollen-part mutations in these two groups separately.

The outcross families of M1-1, M1-6, and M1-11 contained triallelic progeny, which indicates some of the pollen produced by these plants contained both the S₃ and S₆ alleles. For M1-1 and M1-11, the lack of S₂S₃ progeny in the outcross families showed that the S₃ allele was not at the S locus. The duplicated S₃ allele (dS₃) is therefore probably on the centric fragment as centric fragments were found in all triallelic progeny. On this basis, the S genotypes of M1-1 and M1-11 are S₆S₆ duplicated S₃ (S₆S₆dS₃) and S₆S^spm₆dS₃, respectively. The outcross progeny of M1-6 indicate that this plant has both the S₃ and S^spm₆ alleles at its S locus. The duplicated allele in M1-6 is therefore either S₃ or S^spm₆ and is associated with a centric fragment. The relative intensities of the S₃- and S₆-RNase hybridizing bands on blots of M1-6 DNA indicate S₃S^spm₆dS₃ as the most likely S genotype of M1-6 (Figure 4D).

The lack of PPM plants in the outcross family of M1-7 limited interpretation of the nature of the mutation in this plant. Like M1-6, DNA blot analysis indicated that all backcross progeny had both S₃ and S₆ alleles. Lack of segregation in the backcross family is indicative of a duplicated S allele (see below). Presumably the duplicated S allele is poorly transmitted through pollen unless selection is applied for the pollen-part mutation. The S genotype of M1-7 is the same as that of M1-6 as judged by S-RNase band intensities (data not shown). As M1-7 does not have a centric fragment, the duplicated S₃ allele must have been translocated to another chromosome.

In four M₁ plants pollen-part mutations arise through competitive interaction of S alleles:
According to their S genotypes, M1-1 and M1-11 can produce either true haploid pollen containing the S₆ allele or S₆dS₃-containing pollen. An S₃S₆ pistil will reject S₆ pollen. If competitive interaction occurs, S₆dS₃-containing pollen will be compatible on an S₃S₆ pistil and all the progeny of a backcross or self-pollination will be heterozygous. This was found experimentally.

M1-6 and M1-7 can produce true haploid pollen containing either S₃ or S^spm₆ or pollen containing two S alleles, either S₃dS₃ or S^spm₆dS₃. An S₃S₆ pistil will reject all haploid pollen from M1-6 and M1-7. If competitive interaction occurs, only S^spm₆dS₃ pollen will be compatible on an S₃S₆ pistil and all the backcross progeny will have both S₃- and S₆-RNase genes. As this was observed, dS₃ must be able to interact with S^spm₆ and not S₃ to produce a PPM phenotype. It is formally possible, as suggested by PANDEY 1967 , that the S^spm₆ allele has a true pollen-part mutation and a second mutation linked to the S^spm₆ allele that is lethal in pollen. According to this model, the role of the dS₃ allele is to complement the lethal mutation. However the S₂S^spm₆ progeny in the outcross family of M1-6 and M1-7 makes the presence of a lethal mutation near the S^spm₆ allele unlikely.

Self-incompatibility models and competitive interaction:
As our analysis of four M₁ plants led us to conclude that self-incompatibility breaks down in pollen grains containing dS₃ and either S₆ or S^spm₆, we sought to explain competitive interaction using current models of the molecular basis of self-incompatibility in the Solanaceae.

The two current models of self-incompatibility are the receptor model and the inhibitor model (Figure 6; MCCLURE et al. 1989 ; THOMPSON and KIRCH 1992 ). The receptor model proposes that the pollen product of the S locus (pollen-S) is a receptor that allows extracellular S-RNases to enter the pollen tube in an allele-specific manner (Figure 6A). Specific uptake of an active ribonuclease by the pollen tube causes an increased rate of RNA breakdown and an inability to synthesize protein. This leads to the dramatically slowed growth rate characteristic of incompatible pollen tubes (LUSH and CLARKE 1997 ).

View larger version (21K): In this window In a new window Download PPT slide   Figure 6. Two models of events involved in inhibiting the growth of an S1 pollen tube in an S1S2 style. (A) The receptor model; (B) the inhibitor model. See text for details.

The behavior of pollen-part mutations resulting from a duplicated S allele can be accommodated by the receptor model if it is assumed that the self-incompatibility response of pollen is critically dependent on the number of functional receptors. A pollen tube with two different S alleles will have fewer functional receptors than a pollen tube with a single S allele if the receptor (pollen-S) is a multimer and only homomeric forms of pollen-S are functional. However, heteromeric forms of pollen-S will occur only if the monomers encoded by different S alleles assort randomly.

The second model, the inhibitor model, proposes that S-RNases enter pollen tubes nonspecifically (Figure 6B). Once inside, S-RNases encounter pollen-S, which is an inhibitor that can inactivate any S-RNases except those encoded by a matching S allele. According to this model, the inability of a pollen tube to detoxify matching S-RNases leads to increased rates of RNA degradation and consequently slowed growth. To explain why a pollen tube expressing two different S alleles can grow through an incompatible pistil, it is necessary to assume that the presence of two types of pollen-S inhibitor can inactivate all S-RNases, regardless of their allelic origin.

Although either model can explain competitive interaction, they make different predictions about the mutability of the pollen-S gene. According to the receptor model, pollen-part mutations could arise from deletions as well as duplications of the pollen-S gene. A pollen tube lacking the pollen-S gene would be unable to allow S-RNases to enter and thus would be able to grow through an incompatible style.

The inhibitor model, on the other hand, predicts that PPMs can arise only by duplication of an S allele as a pollen tube lacking the pollen-S gene is unable to detoxify any S-RNase. This makes mutations of the pollen-S gene lethal, as pollen tubes carrying these mutations are rejected by styles expressing S-RNases.

Using PPM plants to test the models of self-incompatibility:
Identifying PPM plants that appear to lack a duplicated S allele is one way to test the two self-incompatibility models. In our study, three plants fall into this category: M1-2, M1-5, and M1-10. M1-5 is homozygous at the S locus, and the backcross families of M1-2 and M1-10 include both homozygous and heterozygous PPMs. There are no triallelic progeny in the outcross families of M1-2, M1-5, and M1-10. We might conclude that these plants have true pollen-part mutations and therefore support the receptor model. Certainly the presence of S₃S₆ and S₆S₆ plants in the backcross family of M1-2 suggests a true pollen-part mutation in the S₆ allele. However, there are some uncertainties, in particular the association between a centric fragment and the PPM phenotype.

M1-5 and M1-10 have no additional chromosome or other evidence of a duplication. The inheritance of the pollen-part mutation can be explained by assuming the mutation in these plants is linked to an S₆ allele. It is possible M1-5 and M1-10 carry true mutations in the pollen-part of the S₆ allele.

To date, the only attempt to understand PPMs at a molecular level has been by THOMPSON et al. 1991 . Working with self-compatible dihaploid lines of S. tuberosum, bearing what appeared to be a translocation of an S allele, THOMPSON et al. 1991 used S-RNase gene probes and gels of stylar proteins to search for evidence of a duplicated S locus, but found none. On the basis of this lack of evidence, THOMPSON et al. 1991 concluded that the pollen-part mutation was caused either by the duplication of only a part of the S locus, or by a mutation in an allele-specific modifier locus unlinked to the S locus. Accordingly, it is possible that a duplicated pollen-part of the S₃ allele (dS^p₃), but not the style-part of the S₃ allele (the S₃-RNase gene), can account for the pollen-part mutations in M1-2, M1-5, and M1-10 (see Table 7). In M1-2, dS^p₃ is presumably on the centric fragment. In M1-5 and M1-10, dS^p₃ may be linked to an S₆ allele. Whether the lesion in M1-2, M1-5, and M1-10 is caused by a true mutation in the pollen-part of the S₆ allele or a duplication of the pollen-part of the S₃ allele is currently being investigated by DNA blot analysis using molecular markers linked to the S locus.

	FOOTNOTES

¹ Present address: Institute of Cell and Molecular Biology, Rutherford Bldg., Kings Bldgs., Mayfield Rd., Edinburgh, EH9 3JR, United Kingdom.

	ACKNOWLEDGMENTS

We thank Bruce McGinness for his assistance in the glasshouse and Drs. Peter Chandler and Jim Peacock from the CSIRO Division of Plant Industry, Canberra, Australia, for help with the irradiation experiment and access to the ⁶⁰Co source. We thank Dr. Marilyn Anderson of LaTrobe University, Melbourne, Australia for her advice at various stages of this project. J.F.G. was supported by an Overseas Postgraduate Research Scholarship from the Australian Government. Research at the Plant Cell Biology Center is funded by a Special Research Center grant from the Australian Research Council.

Manuscript received December 15, 1998; Accepted for publication March 22, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ANDERSON, M. A., G. I. MCFADDEN, R. BERNATZKY, A. ATKINSON, and T. ORPIN et al., 1989  Sequence variability of three alleles of the self-incompatibility gene of Nicotiana alata.. Plant Cell 1:483-491[Abstract/Free Full Text].

BERNATZKY, R. and S. D. TANKSLEY, 1986  Genetics of actin-related sequences in tomato. Theor. Appl. Genet. 72:314-321.

BRADFORD, M. M., 1976  A rapid sensitive method for the quantification of microgram quantities of protein using the principle of protein-dye binding. Anal. Biochem. 72:248-254[Medline].

BREWBAKER, J. L. and A. T. NATARAJAN, 1960  Centric fragments and pollen part mutations of incompatibility alleles in Petunia.. Genetics 45:699-704[Free Full Text].

CARLUCCIO, F., D. DE NETTANCOURT and A. J. G. VAN GASTEL, 1974 On a possible involvement of chromosome 3 in the formation of self-compatibility mutations in Nicotiana alata Link and Otto, pp. 41–50 in Polyploidy and Induced Mutations in Plant Breeding. IAEA, Vienna.

DE NETTANCOURT, D., 1977 Incompatibility in Angiosperms. Springer-Verlag, Berlin.

DODDS, P. N., I. BONIG, H. DU, J. RODIN, and M. A. ANDERSON et al., 1993  S-RNase gene of Nicotiana alata is expressed in developing pollen. Plant Cell 5:1771-1782[Abstract].

DODDS, P. N., C. FERGUSON, A. E. CLARKE, and E. NEWBIGIN, 1999  Pollen-expressed S-RNases are not involved in self-incompatibility in Lycopersicon peruvianum.. Sex. Plant Reprod. in press.

HARLOW, E., and D. LANE, 1988 Antibodies: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

HERMSEN, J. G. T., 1978  Genetics of self-compatibility in dihaploids of Solanum tuberosum L. 2. Detection and identification of all possible incompatibility and compatibility genotypes in six F₁'s from interdihaploid crosses. Euphytica 27:1-11.

LAEMMLI, U. K., 1970  Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685[Medline].

LEE, H. S., S. HUANG, and T. H. KAO, 1994  S proteins control rejection of incompatible pollen in Petunia inflata.. Nature 367:560-563[Medline].

LUSH, W. M. and A. E. CLARKE, 1997  Observations of pollen tube growth in Nicotiana alata and their implications for the mechanism of self-incompatibility. Sex. Plant Reprod. 10:27-35.

MCCLURE, B. A., V. HARING, P. R. EBERT, M. A. ANDERSON, and R. J. SIMPSON et al., 1989  Style self-incompatibility gene products of Nicotiana alata are ribonucleases. Nature 342:955-957[Medline].

OLSDER, J. and J. G. T. HERMSEN, 1976  Genetics of self-compatibility in dihaploids of Solanum tuberosum L. 1. Breeding behavior of two self-compatible dihaploids. Euphytica 25:597-607.

PANDEY, K. K., 1965  Centric chromosomes fragments and pollen-part mutations of the incompatibility gene in Nicotiana alata.. Nature 206:792-795.

PANDEY, K. K., 1967  Elements of the S-gene complex. II. Mutations and complementation at the S_I locus in Nicotiana alata. Heredity 22:255-284.

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

THOMPSON, R. D. and H. H. KIRCH, 1992  The S locus of flowering plants: when self-rejection is self-interest. Trends Genet. 8:381-387[Medline].

THOMPSON, R. D., H. UHRIG, J. G. T. HERMSEN, F. SALAMINI, and H. KAUFMANN, 1991  Investigation of a self-compatible mutation in Solanum tuberosum clones inhibiting S-allele activity in pollen differentially. Mol. Gen. Genet. 226:283-288[Medline].

VAN GASTEL, A. J. G. and D. DE NETTANCOURT, 1975  The effects of different mutagens on self-incompatibility in Nicotiana alata Link and Otto II. Acute irradiation with X-rays and fast neutrons. Heredity 34:381-392.

This article has been cited by other articles:

				Z.-H. Hua, A. Fields, and T.-h. Kao Biochemical Models for S-RNase-Based Self-Incompatibility Mol Plant, July 1, 2008; 1(4): 575 - 585. [Abstract] [Full Text] [PDF]

				D. Wheeler and E. Newbigin Expression of 10 S-Class SLF-like Genes in Nicotiana alata Pollen and Its Implications for Understanding the Pollen Factor of the S Locus Genetics, December 1, 2007; 177(4): 2171 - 2180. [Abstract] [Full Text] [PDF]

				Z. Hua and T.-h. Kao Identification and Characterization of Components of a Putative Petunia S-Locus F-Box-Containing E3 Ligase Complex Involved in S-RNase-Based Self-Incompatibility PLANT CELL, October 1, 2006; 18(10): 2531 - 2553. [Abstract] [Full Text] [PDF]

				S. Vilanova, M. L. Badenes, L. Burgos, J. Martinez-Calvo, G. Llacer, and C. Romero Self-Compatibility of Two Apricot Selections Is Associated with Two Pollen-Part Mutations of Different Nature Plant Physiology, October 1, 2006; 142(2): 629 - 641. [Abstract] [Full Text] [PDF]

				X. Qin, B. Liu, J. Soulard, D. Morse, and M. Cappadocia Style-by-style analysis of two sporadic self-compatible Solanum chacoense lines supports a primary role for S-RNases in determining pollen rejection thresholds J. Exp. Bot., June 1, 2006; 57(9): 2001 - 2013. [Abstract] [Full Text] [PDF]

				J. L. Stone, M. A. Sasuclark, and C. P. Blomberg Variation in the self-incompatibility response within and among populations of the tropical shrub Witheringia solanacea (Solanaceae) Am. J. Botany, April 1, 2006; 93(4): 592 - 598. [Abstract] [Full Text] [PDF]

				N. R. Hauck, H. Yamane, R. Tao, and A. F. Iezzoni Accumulation of Nonfunctional S-Haplotypes Results in the Breakdown of Gametophytic Self-Incompatibility in Tetraploid Prunus Genetics, February 1, 2006; 172(2): 1191 - 1198. [Abstract] [Full Text] [PDF]

				T. Sonneveld, K. R. Tobutt, S. P. Vaughan, and T. P. Robbins Loss of Pollen-S Function in Two Self-Compatible Selections of Prunus avium Is Associated with Deletion/Mutation of an S Haplotype-Specific F-Box Gene PLANT CELL, January 1, 2005; 17(1): 37 - 51. [Abstract] [Full Text] [PDF]

				B. McClure S-RNase and SLF Determine S-Haplotype-Specific Pollen Recognition and Rejection PLANT CELL, November 1, 2004; 16(11): 2840 - 2847. [Full Text] [PDF]

				H. Qiao, F. Wang, L. Zhao, J. Zhou, Z. Lai, Y. Zhang, T. P. Robbins, and Y. Xue The F-Box Protein AhSLF-S2 Controls the Pollen Function of S-RNase-Based Self-Incompatibility PLANT CELL, September 1, 2004; 16(9): 2307 - 2322. [Abstract] [Full Text] [PDF]

				T.-h. Kao and T. Tsukamoto The Molecular and Genetic Bases of S-RNase-Based Self-Incompatibility PLANT CELL, June 1, 2004; 16(suppl_1): S72 - S83. [Full Text] [PDF]

H. Qiao, H. Wang, L. Zhao, J. Zhou, J. Huang, Y. Zhang, and Y. Xue 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, March 1, 2004; 16(3): 582 - 595. [Abstract]