Plant Physiol, January 2002, Vol. 128, pp. 17-20

SCIENTIFIC CORRESPONDENCE

Monoallelic Expression and Dominance Interactions in Anthers of Self-Incompatible Arabidopsis lyrata¹

Department of Plant Biology, Cornell University, Ithaca, New York 14853

	ARTICLE

TOP ARTICLE LITERATURE CITED

Genetic dominance and recessiveness are most commonly addressed in the context of mutated alleles that confer aberrant phenotypes but have rarely been explained for functional variants. An opportunity to gain a mechanistic understanding of interactions between naturally occurring functional allelic variants is presented by the self-incompatibility (SI) system of crucifers. This intraspecific mating barrier, which allows the epidermal cells of the stigma to recognize and reject self-related pollen, is based on the activity of a large number of haplotypes of the S-locus complex. Each haplotype encodes highly divergent allelic variants of the S-locus receptor kinase (SRK), a transmembrane protein of the stigma epidermis that determines SI specificity in the stigma (Stein et al., 1991; Takasaki et al., 2000), and the S-locus Cys-rich protein (SCR), a pollen coat-localized ligand for SRK (Kachroo et al., 2001), which determines SI specificity in pollen (Schopfer et al., 1999; Takayama et al., 2000). Self-pollination is proposed to trigger an S haplotype-specific receptor-ligand interaction between SRK and SCR, which leads to the arrest of self-related pollen at the stigma surface (Nasrallah, 2000).

In self-incompatible crucifers, including Brassica species and Arabidopsis lyrata, a wild, self-incompatible relative of Arabidopsis, pollen SI specificity is determined by the diploid genotype of the pollen-producing parent rather than by the genotype of individual haploid pollen grains (Bateman, 1954; Thompson and Taylor, 1966; Kusaba et al., 2001; Schierup et al., 2001). Consequently, genetic interactions between S haplotypes occur in the specification of SI phenotype in pollen as well as in stigmas. Allelic interactions of codominance, dominance, incomplete dominance, or mutual weakening occur, and these interactions can differ in stigma and pollen, consistent with the activity of distinct determinants of SI specificity in these two tissues.

Recessiveness in pollen confers an advantage on an S haplotype by allowing pollen of the recessive genotype to elude the S haplotype-specific stigmatic surveillance mediated by SRK. In fact, "pollen-recessive" alleles attain high frequencies in populations (Uyenoyama, 2000). Elucidation of the molecular basis of S haplotype recessiveness in pollen is important for understanding the mechanism of SI and the evolution and maintenance of S haplotypes in a population. Here, we examine the genetic interaction of two S haplotypes of A. lyrata and elucidate the molecular basis of their dominant/recessive relationship in pollen.

We recently isolated the SRK and SCR genes from two A. lyrata S haplotypes designated Sa and Sb (Kusaba et al., 2001). Reciprocal crosses of SaSb to SaSa and SbSb revealed that, in the stigma, Sa and Sb exhibit a codominant interaction with "weakening" of Sa. In pollen, Sa is recessive to Sb, and pollen grains from SaSb plants exhibit Sb specificity. These interactions imply that SRKa and SRKb are both active in heterozygotes, although the SRKa allele exhibits somewhat lower activity. In contrast, the activity of SCRa is completely masked in SaSb heterozygotes.

We found significant differences in the temporal and spatial distribution of SCRa and SCRb transcripts in Sa and Sb homozygotes, respectively. On RNA gel blots (Fig. 1A) and by reverse transcriptase-PCR, SCRa transcripts were detected in early stage anthers, which contain a tapetum (a cell layer derived from diploid cells of the sporophyte that serves as nurse tissue for the developing haploid microspores and that degenerates before anther dehiscence), but were not detectable in late-stage anthers, which lack tapetal cells (Fig. 1A). In contrast, SCRb transcripts were detected at early and late stages of anther development (Fig. 1A) as previously described (Kusaba et al., 2001). In situ hybridization of SbSb anthers demonstrated that SCRb is expressed sporophytically in the tapetal cell layer and gametophytically in microspores (Fig. 2), as described for all Brassica SCR alleles examined to date (Schopfer et al., 1999; Schopfer and Nasrallah, 2000; Takayama et al., 2000; Shiba et al., 2001). In contrast, SCRa, which is the only "pollen-recessive" allele isolated to date, exhibits strict sporophytic expression (Fig. 2). Thus, functional SCR alleles can vary dramatically in their expression pattern, and expression of SCR in the tapetal cell layer is sufficient for SI. The additional gametophytic expression exhibited by SCRb and all known Brassica SCR alleles might be redundant or serve to boost SCR levels in individual pollen grains.

View larger version (24K): [in this window] [in a new window]   Figure 1.   Differential expression of S-locus genes in A. lyrata. A, Developmental regulation of SCRa and SCRb expression in A. lyrata anthers. Total RNA (15 µg per lane) was isolated from SaSa and SbSb leaves (L), pistils (P), and anthers at three stages of development: 3 anthers (with intact tapetum) and 1 anther (after degeneration of the tapetum) were collected at 3 d and 1 d before flower opening, respectively. Mature anthers (containing mature pollen grains) were collected from open flowers (0). Blots were probed with SCRa or SCRb cDNAs, which, being only 35% similar, serve as allele-specific probes. Hybridization with actin served as a loading control. Hybridization signals were quantitated using a PhosphorImager and the ImageQuant program (Molecular Dynamics, Sunnyvale, CA). B, Expression of SRKa in A. lyrata stigmas. poly(A+) RNA (2 µg per lane) was isolated from SaSa (aa) and SaSb (ab) stigmas (1 stage) and probed with the SRKa ectodomain and with actin as a loading control. C, Expression of SCRa and SCRb in SaSa (aa), SbSb (bb), and SaSb (ab) anthers. Total RNA (15 µg per lane) was isolated from anthers (A) and microspores (M) collected from 3-stage buds. Probes are as in A.

View larger version (107K): [in this window] [in a new window]   Figure 2.   In situ localization of SCRa and SCRb transcripts. Paraffin-embedded sections were prepared from SaSa, SbSb, and SaSb 3-stage anthers and hybridized with dioxigenin-labeled RNA probes transcribed in vitro essentially according to protocols at http://www.Arabidopsis.org/cshl-course. The probes were: s-a, antisense SCRa; s-b, antisense SCRb. Negative controls: Sense SCRa (s-a) and sense SCRb (s-b) RNA probes. T, Tapetum; M, microspores. Some microspores in SaSb anthers did not hybridize with the s-b probe, in keeping with the expected segregation of Sa and Sb microspores. Magnification, 450×.

We examined the expression of the SRK and SCR alleles in SaSb heterozygotes. We found no difference in the level of SRKa transcripts beyond that expected from reduced gene dosage in heterozygous stigmas relative to homozygous stigmas (Fig. 1B). Thus, the weakening of Sa activity in heterozygous stigmas, like dominant/recessive relationships in the Brassica stigma (Hatakeyama et al., 2001), is not related to differences in SRK expression levels. It may be based on interference between receptor or ligand isoforms either in the SRK-SCR interaction or in the recruitment of downstream effectors of the SI response.

In contrast, the SCR alleles were differentially regulated in heterozygotes. SCRb transcripts were detected in SaSb anthers (Fig. 1C) and were localized to tapetum and microspores as in Sb homozygotes (Fig. 2). However, SCRa transcripts were drastically reduced in heterozygotes relative to Sa homozygotes (Figs. 1C and 2), with average reductions of approximately 80-fold and 30-fold estimated from long exposures of RNA gel blots and quantitative reverse transcriptase-PCR, respectively. In contrast, SCRb transcripts were reduced by only approximately 10% in heterozygotes relative to Sb homozygotes. Importantly, comparison of eight SaSa and eight SaSb plants generated by forced selfing of an SaSb plant in which SCRa was "silent" showed that SCRa was expressed in SaSa progeny and "silenced" in their SaSb sibs. Thus, the low-expression state of SCRa is not heritable and is probably not due to an unlinked modifier gene influencing SCRa transcription or the stability of its transcripts.

Why is expression of the SCRa allele suppressed in SaSb heterozygotes? This effect might be due to direct interference from the SCRb allele or from other sequences within the Sb haplotype, or it might result from SCRa-specific properties. Several eukaryotic genes exhibit monoallelic expression, with selection of the expressed allele occurring either stochastically, according to parental origin (genomic imprinting), or based on allele-inherent characteristics (Rothenburg et al., 2001). Furthermore, severe down-regulation or silencing has been documented for a number of eukaryotic genes in the form of transgene effects (Kooter et al., 1999) and other trans-sensing phenomena, such as paramutation in maize (Zea mays) and transvection in Drosophila melanogaster (Tartof and Henikoff, 1991). Many of these examples are associated with increased DNA methylation (Martienssen and Colot, 2001) or with RNA degradation effected by aberrant small (21-25 nt) interfering RNAs (Hamilton and Baulcombe, 1999; Mallory et al., 2001; Matzke et al., 2001a). We compared the DNA of leaves and anthers of SaSa and SaSb plants by methylation-sensitive restriction enzyme digestion and by genomic bisulfite sequencing (Clark et al., 1994). We detected no consistent differences between SaSa and SaSb plants in the methylation state of SCRa within the two exons and one intron of the gene and within approximately 500 bp of sequence 5' of the initiating codon. We also failed to detect SCRa degradation products or small SCRa-related RNA species in small RNA-enriched fractions isolated from SaSb anthers at two stages of development. Nevertheless, neither phenomenon can be categorically ruled out, because modifications restricted to tapetal cells, which constitute only a small proportion of anther cells, might not be detected by current methods.

An alternative explanation for the differential expression of SCRa and SCRb in homozygotes and heterozygotes is suggested by the approximately 65% sequence divergence of SCRa and SCRb and by the extensive structural heteromorphism that distinguishes the Sa and Sb haplotypes (Kusaba et al., 2001), two features that are likely to interfere with chromosome pairing. It is possible that expression of the SCRa allele, but not that of the SCRb allele, is dependent on homolog pairing. Such dependence has been described for some eukaryotic genes (Aramayo and Metzenberg, 1996; Goldsborough and Kornberg, 1996; Matzke et al., 2001b), with expression being affected even by transient pairing of homologous chromosomes in some cases (LaSalle and Lalande, 1996). Interestingly, chromosome pairing has been described in tapetal cells (Aragon-Alcaide et al., 1997). The possibility that SCRa is expressed only in Sa homozygotes (S-locus homozygotes can occur naturally in the case of recessive alleles) or in heterozygous combinations that allow S haplotype pairing is at least consistent with the absence of SCRa transcripts in haploid microspores. However, further analysis of SCRa expression in the presence of different S haplotypes is required to test this hypothesis.

Irrespective of the underlying mechanism(s) for gene silencing, the recessive/dominant interaction exhibited by the SCRa and SCRb alleles in pollen is explained by the severe down-regulation of the recessive SCRa allele in the tapetum of SaSb heterozygotes, which, together with the lack of SCRa expression in microspores, results effectively in monoallelic expression of the dominant SCRb allele. The observed reduction in SCR concentration of approximately 30-fold or more results in loss of the corresponding SI specificity in pollen because too few SCR molecules are delivered to the stigma surface by any individual pollen grain for SRK activation to occur. We propose that this unusual feature of allelic differences in the temporal and spatial pattern of SCR gene expression, as well as allele-specific differences in susceptibility to silencing, may explain many, if not all, cases of dominant/recessive interactions and mutual weakening of S haplotypes in the pollen of crucifers.

	ACKNOWLEDGMENTS

We thank V. Vance and A. Mallory for advice on small RNA isolation, and M. Wofner, T. Fox, and U. Grossniklaus for helpful comments.

	FOOTNOTES

Received August 28, 2001; returned for revision September 24, 2001; accepted September 28, 2001.

¹ This work was supported by the National Institutes of Health (grant no. GM5752) and the National Science Foundation (grant no. IBN-0077289).

² Present address: Institute of Radiation Breeding, National Institute of Agrobiological Science, Ohmiya-machi, Naka-gun, Ibaraki 319-2293, Japan.

^* Corresponding author; e-mail jbn2{at}cornell.edu; fax 607-255-5407.

www.plantphysiol.org/cgi/doi/10.1104/pp.010790.

	LITERATURE CITED

TOP ARTICLE LITERATURE CITED

• Aragon-Alcaide L, Reader S, Beve A, Shaw P, Miller T, Moore G (1997) Curr Biol 7: 905-908[CrossRef][Web of Science][Medline]
• Aramayo R, Metzenberg RL (1996) Cell 86: 103-113[CrossRef][Web of Science][Medline]
• Bateman AJ (1954) Heredity 8: 305-332
• Clark SJ, Harrison J, Paul CL, Frommer M (1994) Nucleic Acids Res 22: 2990-2997[Abstract/Free Full Text]
• Goldsborough AS, Kornberg TB (1996) Nature 381: 807-810[CrossRef][Medline]
• Hamilton AJ, Baulcombe DC (1999) Science 286: 950-952[Abstract/Free Full Text]
• Hatakeyama K, Takasaki T, Suzuki G, Nishio T, Watanabe M, Isogai A, Hinata K (2001) Plant J 26: 69-76[CrossRef][Web of Science][Medline]
• Kachroo A, Schopfer CR, Nasrallah ME, Nasrallah JB (2001) Science 293: 1824-1826[Abstract/Free Full Text]
• Kooter JM, Matzke MA, Meyer P (1999) Trends Plant Sci 4: 340-347[CrossRef][Web of Science][Medline]
• Kusaba M, Dwyer K, Hendershot J, Vrebalov J, Nasrallah JB, Nasrallah ME (2001) Plant Cell 13: 627-643[Abstract/Free Full Text]
• LaSalle JM, Lalande M (1996) Science 272: 725-728[Abstract]
• Mallory AC, Ely L, Smith TH, Marathe R, Anandalakshmi R, Fagard M, Vaucheret H, Pruss G, Bowman L, Vance VB (2001) Plant Cell 13: 571-583[Abstract/Free Full Text]
• Martienssen RA, Colot V (2001) Science 293: 1070-1073[Abstract/Free Full Text]
• Matzke M, Matze AJM, Kooter JM (2001a) Science 293: 1080-1083[Abstract/Free Full Text]
• Matzke M, Mette MF, Jakowitsch J, Kanno T, Moscone EA, van der Winden J, Matzke AJM (2001b) Genetics 158: 451-461[Abstract/Free Full Text]
• Nasrallah JB (2000) Curr Opin Plant Biol 3: 368-373[CrossRef][Web of Science][Medline]
• Rothenburg S, Koch-Nolte F, Thiele HG, Haag F (2001) Immunogenetics 52: 231-241[Medline]
• Schierup MH, Mable BK, Awadalla P, Charlesworth D (2001) Genetics 158: 387-399[Abstract/Free Full Text]
• Schopfer CR, Nasrallah JB (2000) Plant Physiol 124: 935-939[Free Full Text]
• Schopfer CR, Nasrallah ME, Nasrallah JB (1999) Science 286: 1697-1700[Abstract/Free Full Text]
• Shiba H, Takayama S, Iwano M, Shimosato H, Funato M, Nakagawa T, Che FS, Suzuki G, Watanabe M, Hinata K (2001) Plant Physiol 125: 2095-2103[Abstract/Free Full Text]
• Stein JC, Howlett BH, Boyes DC, Nasrallah ME, Nasrallah JB (1991) Proc Natl Acad Sci USA 88: 8816-8820[Abstract/Free Full Text]
• Takasaki T, Hatakeyama K, Suzuki G, Watanabe M, Isogai A, Hinata K (2000) Nature 403: 913-916[CrossRef][Medline]
• Takayama S, Shiba H, Iwano M, Shimosato H, Che FS, Kai N, Watanabe M, Suzuki G, Hinata K, Isogai A (2000) Proc Natl Acad Sci USA 97: 1920-1925[Abstract/Free Full Text]
• Tartof KD, Henikoff S (1991) Cell 65: 201-203[CrossRef][Web of Science][Medline]
• Thompson KF, Taylor JP (1966) Heredity 21: 345-362
• Uyenoyama MK (2000) Genetics 156: 351-359[Abstract/Free Full Text]