A complex dominance hierarchy is controlled by ... · A complex dominance hierarchy is controlled by ... 1Graduate School of Biological Sciences, Nara Institute of Science and Technology,
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
In the format provided by the authors and unedited.
A complex dominance hierarchy is controlled bypolymorphism of small RNAs and their targetsShinsuke Yasuda1‡, Yuko Wada1‡, Tomohiro Kakizaki2‡, Yoshiaki Tarutani1†, Eiko Miura-Uno1,Kohji Murase1, Sota Fujii1, Tomoya Hioki1, Taiki Shimoda1, Yoshinobu Takada3, Hiroshi Shiba1†,Takeshi Takasaki-Yasuda4, Go Suzuki5, Masao Watanabe3* and Seiji Takayama1,6*
In diploid organisms, phenotypic traits are often biased byeffects known as Mendelian dominant–recessive interactionsbetween inherited alleles. Phenotypic expression of SP11alleles, which encodes the male determinants of self-incompat-ibility in Brassica rapa, is governed by a complex dominancehierarchy1–3. Here, we show that a single polymorphic 24nucleotide small RNA, named SP11 methylation inducer 2(Smi2), controls the linear dominance hierarchy of the fourSP11 alleles (S44 > S60 > S40 > S29). In all dominant–recessiveinteractions, small RNA variants derived from the linkedregion of dominant SP11 alleles exhibited high sequence simi-larity to the promoter regions of recessive SP11 alleles andacted in trans to epigenetically silence their expression.Together with our previous study4, we propose a new model;sequence similarity between polymorphic small RNAs andtheir target regulates mono-allelic gene expression, whichexplains the entire five-phased linear dominance hierarchy ofthe SP11 phenotypic expression in Brassica.
Most Mendelian dominance relationships are explained by loss-of-function mutations in haplosufficient genes, which make thewild-type allele dominant over the mutant allele. In a few cases,however, dominance is thought to be controlled by other geneticelements called ‘dominance modifiers’5,6. In one such case, a smallRNA (sRNA) controls the dominance of pollen self-incompatibilityphenotypes in Brassica rapa4. Self-incompatibility is a geneticmechanism, controlled by multiple S-haplotypes at the S-locus,for avoiding self-fertilization by rejecting self-pollen to maintaingenetic diversity in the species7. For the Brassicaceae self-incompat-ibility, each S-haplotype encodes a male determinant, S-locusprotein 11 (SP11, also called SCR)8–10, and a female determinant,S-locus receptor kinase (SRK)11,12; the S-haplotype-specific inter-action between these determinants triggers a self-incompatibilityresponse in the stigma epidermis13,14.
SP11 is sporophytically expressed in the anther tapetum, and theself-incompatibility phenotypes in pollen are determined by domi-nance relationships between the two S-haplotypes that the plantcarries. According to the dominance of pollen self-incompatibilityphenotypes, the S-haplotypes in B. rapa have been categorized asclass-I (for example, S8, S9, S12 and S52) or class-II (for example,S44, S60, S40 and S29), with class-I S-haplotypes dominating overclass-II S-haplotypes1–3. When class-I and class-II S-haplotypesare heterozygous, a 24 nucleotide (nt) sRNA called SP11
methylation inducer (Smi) derived from an inverted repeat sequenceof class-I S-haplotype induces DNA methylation of the recessiveclass-II SP11 promoter and results in mono-allelic gene silencing4,15.Interestingly, class-II S-haplotypes exhibit linear dominancerelationships, resulting in a complicated dominance hierarchy:class-I S > (S44 > S60 > S40 > S29)
3. However, the molecular mechan-ism underlying the linear dominance hierarchy within class-IIS-haplotypes remains unknown.
To elucidate the molecular mechanism underlying the lineardominance hierarchy of class-II S-haplotypes in B. rapa, wesequenced the full S-locus regions of three class-II S-haplotypes,S44, S60 and S40, for which partial S-locus sequences were previouslypublished16,17. We also determined the partial S-locus sequence ofS29 haplotype. The sRNA gene candidates in the S44, S60 and S40S-locus regions were identified using three de novo sRNA predictionprograms. Predicted stem-loops encoding sequences homologous tothe four S-haplotypes of SP11 sequences ±1 kb were further selectedusing a BLAST search. As a result, we identified two inverted repeatsequences in all of the S-haplotypes (Supplementary Tables 1–3):the previously identified SMI3 and a novel SMI-like sequence,SP11 METHYLATION INDUCER 2 (SMI2) (SupplementaryFig. 1). In each S-haplotype, SMI2 was located 1.9–8.6 kb down-stream of SRK but was not observed in the SP11 genomic regionof the class-I S-haplotype in B. rapa8,18,19 (Fig. 1a). SMI2 was onlyobserved in class-II S-haplotypes, which formed a distinctcluster from class-I S-haplotypes in the phylogenetic analysis(Supplementary Fig. 2 and Supplementary Methods).
To determine whether SMI2 in each S-haplotype is expressed andprocessed into sRNAs at the relevant developmental stage, wesequenced sRNAs from early stage (stages 1–3) anthers10, duringwhich SP11 DNA methylation occurs15. We obtained 11–17million sRNA sequences for each class-II S-homozygote(Supplementary Table 4). The mapping of sRNA sequences tocognate SMI2 transcripts revealed that 24 nt sRNAs (designatedSmi2) with sequence similarity to class-II SP11 promoters were pro-cessed from the stem structures of S44-, S60- and S40-Smi2 precursors(Fig. 1b). Although we observed multiple distinct sRNA fragmentsfrom the precursor of the most recessive S29-haplotype, we didnot detect mature Smi2 (Supplementary Fig. 3). To accuratelymeasure the amount of processed Smi2, we performed stem-loopPCR with reverse transcription (RT–PCR)20 using sRNA extractedfrom early stage anthers from S44-, S60-, S40- and S29-homozygotes.
1Graduate School of Biological Sciences, Nara Institute of Science and Technology, Ikoma, Nara 630-0192, Japan. 2Division of Vegetable Breeding, Instituteof Vegetable and Floriculture Science, NARO, Tsu, Mie 514-2392, Japan. 3Graduate School of Life Sciences, Tohoku University, Sendai, Miyagi 980-8577,Japan. 4Graduate School of Agricultural Science, Kobe University, Kobe, Hyogo 657-8501, Japan. 5Division of Natural Science, Osaka Kyoiku University,Kashiwara, Osaka 582-8582, Japan. 6Department of Applied Biological Chemistry, The University of Tokyo, Tokyo 113-8657, Japan. †Present address:Department of Integrated Genetics, National Institute of Genetics, Mishima, Shizuoka 411-8540, Japan (Y.T.); Graduate School of Life and EnvironmentalSciences, University of Tsukuba, Tsukuba, Ibaraki 305-8577, Japan (H.S.). ‡These authors contributed equally to this work. *e-mail: [email protected];[email protected]
LETTERSPUBLISHED: XX XX 2016 | VOLUME: 3 | ARTICLE NUMBER: 16206
A complex dominance hierarchy is controlled bypolymorphism of small RNAs and their targetsShinsuke Yasuda1‡, Yuko Wada1‡, Tomohiro Kakizaki2‡, Yoshiaki Tarutani1†, Eiko Miura-Uno1,Kohji Murase1, Sota Fujii1, Tomoya Hioki1, Taiki Shimoda1, Yoshinobu Takada3, Hiroshi Shiba1†,Takeshi Takasaki-Yasuda4, Go Suzuki5, Masao Watanabe3* and Seiji Takayama1,6*
In diploid organisms, phenotypic traits are often biased byeffects known as Mendelian dominant–recessive interactionsbetween inherited alleles. Phenotypic expression of SP11alleles, which encodes the male determinants of self-incompat-ibility in Brassica rapa, is governed by a complex dominancehierarchy1–3. Here, we show that a single polymorphic 24nucleotide small RNA, named SP11 methylation inducer 2(Smi2), controls the linear dominance hierarchy of the fourSP11 alleles (S44 > S60 > S40 > S29). In all dominant–recessiveinteractions, small RNA variants derived from the linkedregion of dominant SP11 alleles exhibited high sequence simi-larity to the promoter regions of recessive SP11 alleles andacted in trans to epigenetically silence their expression.Together with our previous study4, we propose a new model;sequence similarity between polymorphic small RNAs andtheir target regulates mono-allelic gene expression, whichexplains the entire five-phased linear dominance hierarchy ofthe SP11 phenotypic expression in Brassica.
Most Mendelian dominance relationships are explained by loss-of-function mutations in haplosufficient genes, which make thewild-type allele dominant over the mutant allele. In a few cases,however, dominance is thought to be controlled by other geneticelements called ‘dominance modifiers’5,6. In one such case, a smallRNA (sRNA) controls the dominance of pollen self-incompatibilityphenotypes in Brassica rapa4. Self-incompatibility is a geneticmechanism, controlled by multiple S-haplotypes at the S-locus,for avoiding self-fertilization by rejecting self-pollen to maintaingenetic diversity in the species7. For the Brassicaceae self-incompat-ibility, each S-haplotype encodes a male determinant, S-locusprotein 11 (SP11, also called SCR)8–10, and a female determinant,S-locus receptor kinase (SRK)11,12; the S-haplotype-specific inter-action between these determinants triggers a self-incompatibilityresponse in the stigma epidermis13,14.
SP11 is sporophytically expressed in the anther tapetum, and theself-incompatibility phenotypes in pollen are determined by domi-nance relationships between the two S-haplotypes that the plantcarries. According to the dominance of pollen self-incompatibilityphenotypes, the S-haplotypes in B. rapa have been categorized asclass-I (for example, S8, S9, S12 and S52) or class-II (for example,S44, S60, S40 and S29), with class-I S-haplotypes dominating overclass-II S-haplotypes1–3. When class-I and class-II S-haplotypesare heterozygous, a 24 nucleotide (nt) sRNA called SP11
methylation inducer (Smi) derived from an inverted repeat sequenceof class-I S-haplotype induces DNA methylation of the recessiveclass-II SP11 promoter and results in mono-allelic gene silencing4,15.Interestingly, class-II S-haplotypes exhibit linear dominancerelationships, resulting in a complicated dominance hierarchy:class-I S > (S44 > S60 > S40 > S29)
3. However, the molecular mechan-ism underlying the linear dominance hierarchy within class-IIS-haplotypes remains unknown.
To elucidate the molecular mechanism underlying the lineardominance hierarchy of class-II S-haplotypes in B. rapa, wesequenced the full S-locus regions of three class-II S-haplotypes,S44, S60 and S40, for which partial S-locus sequences were previouslypublished16,17. We also determined the partial S-locus sequence ofS29 haplotype. The sRNA gene candidates in the S44, S60 and S40S-locus regions were identified using three de novo sRNA predictionprograms. Predicted stem-loops encoding sequences homologous tothe four S-haplotypes of SP11 sequences ±1 kb were further selectedusing a BLAST search. As a result, we identified two inverted repeatsequences in all of the S-haplotypes (Supplementary Tables 1–3):the previously identified SMI3 and a novel SMI-like sequence,SP11 METHYLATION INDUCER 2 (SMI2) (SupplementaryFig. 1). In each S-haplotype, SMI2 was located 1.9–8.6 kb down-stream of SRK but was not observed in the SP11 genomic regionof the class-I S-haplotype in B. rapa8,18,19 (Fig. 1a). SMI2 was onlyobserved in class-II S-haplotypes, which formed a distinctcluster from class-I S-haplotypes in the phylogenetic analysis(Supplementary Fig. 2 and Supplementary Methods).
To determine whether SMI2 in each S-haplotype is expressed andprocessed into sRNAs at the relevant developmental stage, wesequenced sRNAs from early stage (stages 1–3) anthers10, duringwhich SP11 DNA methylation occurs15. We obtained 11–17million sRNA sequences for each class-II S-homozygote(Supplementary Table 4). The mapping of sRNA sequences tocognate SMI2 transcripts revealed that 24 nt sRNAs (designatedSmi2) with sequence similarity to class-II SP11 promoters were pro-cessed from the stem structures of S44-, S60- and S40-Smi2 precursors(Fig. 1b). Although we observed multiple distinct sRNA fragmentsfrom the precursor of the most recessive S29-haplotype, we didnot detect mature Smi2 (Supplementary Fig. 3). To accuratelymeasure the amount of processed Smi2, we performed stem-loopPCR with reverse transcription (RT–PCR)20 using sRNA extractedfrom early stage anthers from S44-, S60-, S40- and S29-homozygotes.
1Graduate School of Biological Sciences, Nara Institute of Science and Technology, Ikoma, Nara 630-0192, Japan. 2Division of Vegetable Breeding, Instituteof Vegetable and Floriculture Science, NARO, Tsu, Mie 514-2392, Japan. 3Graduate School of Life Sciences, Tohoku University, Sendai, Miyagi 980-8577,Japan. 4Graduate School of Agricultural Science, Kobe University, Kobe, Hyogo 657-8501, Japan. 5Division of Natural Science, Osaka Kyoiku University,Kashiwara, Osaka 582-8582, Japan. 6Department of Applied Biological Chemistry, The University of Tokyo, Tokyo 113-8657, Japan. †Present address:Department of Integrated Genetics, National Institute of Genetics, Mishima, Shizuoka 411-8540, Japan (Y.T.); Graduate School of Life and EnvironmentalSciences, University of Tsukuba, Tsukuba, Ibaraki 305-8577, Japan (H.S.). ‡These authors contributed equally to this work. *e-mail: [email protected];[email protected]
LETTERSPUBLISHED: XX XX 2016 | VOLUME: 3 | ARTICLE NUMBER: 16206