2024 Volume 74 Issue 3 Pages 193-203
Hybrid breakdown is a post-zygotic reproductive isolation that hinders genetic exchange between species or populations in both animals and plants. Two complementary recessive genes, temperature sensitive hybrid breakdown1 (thb1) and thb2, cause hybrid breakdown in rice (Oryza sativa). The present study delimited the THB1 locus to a 9.1-kb sequence, containing a single gene encoding a putative transmembrane protein with unknown functions. Haplotype analysis of THB1 in the two core collections of 119 accessions revealed that these accessions were divided into 22 haplotypes. A test cross with thb2 carrier showed that haplotype2 (H2) was assigned to thb1 and was restricted to temperate japonica. A nonsynonymous nucleotide polymorphism (SNP) specific to H2 was identified as a causal mutation in thb1. A test cross with thb1 carrier indicated that six accessions, including temperate japonica, tropical japonica, and indica, carried thb2. These results suggest that thb1 has recently evolved in temperate japonica, whereas thb2 arose in an ancient japonica and introgressed into the present three subgroups. Furthermore, we developed a derived cleaved amplified polymorphic sequence (dCAPS) marker to detect causal SNP in THB1. Our findings provide new insights into reproductive isolation and may benefit rice breeding.
Reproductive isolation hinders genetic exchange between species and populations of both plants and animals. Therefore, it is considered fundamental for speciation (Coyne and Orr 2004). Generally, reproductive isolation in plants can be classified into pre- and post-pollination, according to the developmental stage during which it occurs. Pre-pollination isolation, such as habitat divergence, temporal isolation, pollinator isolation, and mating system divergence, usually functions more effectively than post-pollination isolation (Coyne and Orr 2004). Post-pollination reproductive isolation can divide into pre-zygotic and post-zygotic isolation. Post-zygotic reproductive isolation is common across the plant and animal kingdoms (Dobzhansky 1970, Grant 1971, Stebbins 1958). Hybrids that experience postzygotic reproductive isolation are usually aborted or arrested at different stages or generations after fertilization. Based on the developmental stage at which isolation occurs and the symptoms it displays, post-zygotic reproductive isolation can be termed as hybrid inviability, hybrid weakness, hybrid necrosis, and hybrid sterility, which are observed in F1 hybrids, and hybrid breakdown in F2 and backcross generations.
Hybrid breakdown facilitates speciation by restricting gene flow between diverging taxa in both animal and plant species (Stebbins 1958) and reduces selection efficiency in crossbreeding. One of the fundamental theories on the mechanism of post-zygotic isolation is the Bateson–Dobzhansky–Muller (BDM) model (Bateson 1909, Dobzhansky 1937, Muller 1942). BDM model postulates that deleterious interactions between two or more genes derived from different species or populations cause post-zygotic isolation. Two molecular mechanisms underlying the BDM-type hybrid breakdown have been reported in plants: an elevated autoimmune response associated with the nucleotide binding site-leucine-rich repeat (NBS-LRR) genes (Alcázar et al. 2009, Bomblies et al. 2007, Yamamoto et al. 2010) and reciprocal silencing or loss of duplicated genes (Bikard et al. 2009, Kubo et al. 2022, Vlad et al. 2010). However, in many cases, the causal genes remain unknown. Elucidation of the mechanisms of hybrid breakdown is important not only to understand speciation and evolution, but also to overcome these barriers for crop breeding.
In our previous study, two complementary recessive genes, temperature sensitive hybrid breakdown1 (thb1) and thb2, that cause hybrid breakdown were serendipitously found in closely related temperate japonica cultivars of rice (Oryza sativa), ‘Yukihikari’ and ‘Kirara397’, respectively (Yoneya et al. 2021). In this study, we identified a single nonsynonymous nucleotide polymorphism (SNP) as the causal mutation in THB1 that encodes a putative transmembrane protein with unknown functions. We also demonstrated that thb1 is restricted to temperate japonica, whereas thb2 is distributed not only in temperate japonica but also in tropical japonica and indica. Furthermore, we developed a derived cleaved amplified polymorphic sequence (dCAPS) marker to determine the causal SNP genotype of THB1. Our findings provide new insights into reproductive isolation and may benefit rice breeding.
The rice plants cultivars ‘Yukihikari’ and ‘Kirara397’, and a chromosome segment substitution line YK3CSSL-6.1 (Kato and Hirayama 2021, Yoneya et al. 2021) were used as parental controls for mapping study. For the fine-scale mapping of THB1, a total of 3,754 F2 plants of a cross between ‘Kirara397’ and YK3CSSL-6.1 and self-pollinated F3 progeny were used. To characterize the distributions of thb1 and thb2, nine accessions in a Core Collection of Japanese rice landraces (JRC) of the NARO GenBank Project (Ebana et al. 2008), ‘Mansaku’ (JRC22/JP6735), ‘Shinyamadaho 2’ (JRC37/JP6962), ‘Akage’ (JP14994, JP5301), ‘Sensho’ (JRC04/JP4386), ‘Karahoushi’ (JRC44/JP10788), ‘Touboshi’ (JRC42/JP10772), and ‘Akamai’ (JRC43/JP4744, JRC21/JP9694) were used.
InDel and SNP marker analysesTwo Insertion-deletion (InDel) markers, YK3InDel06-845078_2 and YK3InDel06-646646, were used (Yoneya et al. 2021). In the present study, nine SNPs were converted into a cleaved amplified polymorphic sequence (CAPS) marker and eight dCAPS markers based on our previous research (Takano et al. 2014) and using the web-based free software program dCAPS Finder 2.0 (Neff et al. 2002) between YK3InDel06-845078_2 and YK3InDel06-646646. Appropriate PCR primer sets flanking each target SNP were designed using Primer 3.0 (version 0.4.0) software (Supplemental Table 1). Young leaves of the parents and the mapping population were collected to extract DNA and subjected to marker analyses as described by Yoneya et al. (2021).
Sanger sequencingRepeat number of T on chr06: 799,424..799,433 (IRGSP-1.0) was determined using Sanger sequencing and designated as YK3InDel-799424. Target DNA fragment was amplified using forward primer 5ʹ-TTGGGGTACCTTGAAGATGTG-3ʹ and reverse primer 5ʹ-TCCCCTCTCTAGCTCCTCTTG-3ʹ. PCR reaction were performed using a thermocycler GeneAtlas G (Astec, Fukuoka, Japan) for a total volume of 20 μL, including 2 μL 10× Ex taq buffer (Takara, Shiga, Japan), 1.6 μL dNTPs (2 mM), 0.2 μL Ex taq (5 U/μL), 0.8 μL primer each (10 μM), 2 μL DNA template (30 ng/μL), and 12.6 μL of DNase-free water. The PCR amplification program was as follows: 2 min at 98°C, 35 cycles of 10 s denaturation at 98°C, 30 s annealing at 60°C, and 30 s extension at 72°C. Purified PCR products were subjected to Sanger sequencing (Eurofins Genomics, Tokyo, Japan) to detect the number of T of the targeted bases.
Growth conditionFor genetic mapping and test cross studies, the seeds of F2 populations, F3 lines, and parental cultivars were sterilized with 0.2% (w/v) benlate solution for 1 d and germinated in reverse osmosis water for 2 d in the dark at 30°C, planted in a soil-filled cell plug tray (cell count, 8 × 16; tray size, 52 × 25 cm2; cell size, 3 × 3 × 4.4 cm3). Each cell plug tray was placed in a plastic container (53 × 34.8 × 15.6 cm3), and 3–10 cm of water was maintained depending on the plant size. To screen the recombinant F2 plants between YK3InDel06-845078_2 and YK3InDel06-646646 (Yoneya et al. 2021), F2 plants and parental control, ‘Yukihikari’, ‘Kirara397’, and YK3CSSL-6.1, were planted four times during 2020; February 14, April 3, April 24, and June 29. For progeny tests to determine THB1 genotypes, 23–56 F3 plants derived from each recombinant F2 plant were planted with parental control three times during 2020; August 17, October 16, and November 12. F2 populations for test cross study were planted on October 16, 2022. These plant materials were grown in a greenhouse at Obihiro University of Agriculture and Veterinary Medicine (OUAVM; 42°52ʹN, 143°9ʹE), with temperature maintained over 20°C from November to May. From the end of May to October, the plants were grown in a glasshouse, with the daily mean temperature maintained less than 25°C.
The F1 plants were grown under two conditions in the test cross study. First, F1 plants were grown in a greenhouse at OUAVM and grown at temperatures >25°C from December 2021 to April 2022 under natural day-length conditions. Second, F1 plants were grown at temperatures >25°C in a greenhouse from April to May and >20°C in a glasshouse at OUAVM from the end of May to October 2022 under natural day-length conditions. In both experiments, seeds were sterilized and planted as previously described. Four-week-old F1 plants were transplanted into plastic pots filled with 2L soil compost containing 1.2 g of N, P2O5, and K2O each. Spikelet fertility (percentage of filled spikelets per total number of spikelets) was assessed using five culms per plant, for five individuals from each cross.
Gene expression analysisThe plants were grown in a growth cabinet (Biotron, NK Systems, Japan) under 25°C with a 16 h-light (350 μmol/m2/s)/8 h-dark photoperiod in water. One week after germination, the shoots and roots of five seedlings were pooled as biological replicates. The experiments were performed using three biological replicates. Total RNA was extracted from each tissue sample using the RNAsuisui-S reagent (Rizo Co., Tsukuba, Japan). Reverse transcription-polymerase chain reaction (RT-PCR) was performed using a PrimeScript™ II first strand cDNA Synthesis Kit (Takara Bio, Kyoto, Japan). DNase digestion was performed using DNase I (Nippon Gene, Tokyo, Japan).
The transcriptional variants of THB1 were amplified using the following two primer sets: forward primer 5ʹ-CCATGAGATGTCCATGTACCAG-3ʹ and reverse primer 5ʹ-GAGGATTCCAATCGCCCATGA-3ʹ (thb1-1), and forward primer 5ʹ-CAAGAGCCACAGGTGGAGAG-3ʹ and reverse primer 5ʹ-TAGCACCAATGCAGCGTACA-3ʹ (thb1-2). Quantitative RT-PCR was performed using a 7300 Real-Time PCR System (Applied Biosystems, USA) and SYBR Premix Ex Taq II (Takara Bio). OsUBQ1 (Os03g0234200) was used as the reference gene (Yamamoto et al. 2010). The expression levels in each sample were calculated using three technical replicates.
Haplotype network analysisGenome-wide variation data from TASUKE+ (Kumagai et al. 2019, https://tasuke.dna.affrc.go.jp) for the rice core collection of the World Rice Core Collection (WRC) of the NARO GenBank Project (Kojima et al. 2005) and the JRC of the NARO GenBank Project (Ebana et al. 2008) were used to identify genetic polymorphisms in THB1 region. To construct a Templeton, Crandall, and Sing haplotype network (Templeton et al. 1992), Population Analysis with Reticulate Trees software (version 1.7, PopART; https://popart.maths.otago.ac.nz/download/) was used.
Phylogenetic analysisA phylogenetic tree was constructed by Molecular Evolutionary Genetics Analysis (version 11, MEGA 11; Tamura et al. 2021) using the neighbor-joining method (Saitou and Nei 1987). The bootstrap consensus tree inferred from 1000 replicates was used to represent the evolutionary history of the taxa analyzed (Felsenstein 1985). The tree was drawn to scale, keeping the units of measurement of branch lengths same as those of the evolutionary distances used to infer the phylogenetic tree. Evolutionary distances were computed using the Kimura 2-parameter method (Kimura 1980). All positions containing gaps or missing data were eliminated from the dataset (complete deletion option).
In our previous study, THB1 was mapped between two InDel markers, YK3InDel06-845078_2 and YK3InDel06-646646 (Yoneya et al. 2021). To narrow down the candidate region of THB1, 3,754 F2 plants were screened for recombination between these two InDel markers. Seventy-two recombinant F2 plants were selected and their genotypes for THB1 were evaluated based on the segregation patterns of plant growth in the F3 progeny. Weak phenotypes were discriminated based on visual observations, including leaf length, width, and color (Yoneya et al. 2021). Nine SNP markers were genotyped from these 72 individuals. Among them, three recombinants were detected between YK3SNP06-792528 and THB1. On the other hand, a single F2 plant (Plant ID: K19-24-2-3) was screened for a recombinant chromosome between YK3SNP06-801669 and THB1. Next, we determined the K19-24-2-3 genotype at YK3InDel-799424 using Sanger sequencing. The THB1 gene was co-segregated with YK3InDel-799424 and YK3SNP06-796157_2 and eventually mapped at a 9.1-kb interval between YK3SNP06-792528 and YK3SNP06-801669 on the short arm of chromosome 6 (Fig. 1A). Based on our re-sequencing data, no other polymorphisms were detected in the THB1 locus (9.1 kb) between the parent cultivars (Takano et al. 2014).
Map-based cloning of THB1. (A) High-resolution linkage map of THB1 using 72 recombinants between YK3InDel06-845078_2 and YK3InDel06-646646 from 3,754 F2 individuals in the present study. (B) Structures of alternative splicing variants of LOC_Os06g02370. The full-length protein encoded by the longest transcript variant (LOC_Os06g02370.1) is 584 amino acids in length and was taken as reference for intron/exon numbering. LOC_Os06g02370.3 was generated by miss-splicing of intron resulting in less intron and changing the position of start codon, which causes lack of 1st exon and reducing 5ʹ site of 2nd exon (332 amino acids). White boxes represent 5ʹ-UTR; white pentagons represent 3ʹ-UTR; black boxes represent exon; and horizontal lines between them represent introns. The vertical arrows indicate the polymorphisms between ‘Yukihikari’ and ‘Kirara397’ in LOC_Os06g02370. Two sets of short horizontal arrows represent the positions of primers, thb1-1 and thb1-2, for expression analyses.
YK3SNP06-792528 is located 2,803 bp downstream of LOC_Os06g02370. YK3SNP06-801669 is located 1,856 bp upstream of LOC_Os06g02370. Thus, the region delimited by fine mapping contained only LOC_Os06g02370, according to O. sativa ssp. japonica cv. ‘Nipponbare’ (annotated by Rice Genome Annotation Project, http://rice.uga.edu/) (Fig. 1B). Two splicing variants, LOC_Os06g02370.1 (584 amino acids) and LOC_Os06g02370.3 (332 amino acids), were annotated as LOC_Os06g02370. YK3InDel-799424 was assigned to the intron of LOC_Os06g02370.1 and 5ʹ-UTR of LOC_Os06g02370.3. YK3SNP06-796157_2 was located at the exon of both transcript variants and had replaced arginine in ‘Kirara397’ with histidine in ‘Yukihikari’.
Sequence analysis of LOC_Os06g02370To obtain more information about LOC_Os06g02370, bioinformatics analysis was performed using the amino acid sequence of LOC_Os06g02370.1. High similarity of protein sequences (amino acid identity >50% to LOC_Os06g02370.1) was selected for bioinformatics analysis. Twenty-three LOC_Os06g02370.1-like proteins (amino acid identity >50.4% to LOC_Os06g02370.1) were identified in plant; namely, XP_025811579.1 from Panicum hallii; XP_039842111.1 and XP_039805188.1 from P. virgatum; XP_004964339.1 from Setaria italica; NP_001144729.1 from Zea mays; XP_002437707.1 from Sorghum bicolor; XP_015694070.2 from Oryza brachyantha; XP_003557155.1 from Brachypodium distachyon; XP_051197997.1 from Lolium perenne; XP_047075120.1 from Lolium rigidum; XP_044959682.1 from Hordeum vulgare ssp. vulgare; XP_037423695.1 and XP_037458488.1 from Triticum dicoccoides; XP_044426481.1, XP_044443345.1, and XP_044426478.1 from Triticum aestivum; XP_020147246.1 from Aegilops tauschii ssp. strangulate; XP_020112363.1 from Ananas comosus; XP_008813158.2 from Phoenix dactylifera; XP_010909418.1 from Elaeis guineensis; XP_009383130.2 from Musa acuminata ssp. malaccensis; XP_060211170.1 from Lycium barbarum; and XP_059318710.1 from Lycium ferocissimum. Based on these proteins, an evolutionary tree was constructed (Fig. 2A).
Phylogenic tree and multiple sequences alignment of THB1 (LOC_Os06g02370.1) with its highly similar protein sequences. Amino acid sequences of XP_015642768.1 (THB1, LOC_Os06g02370.1) from Oryza sativa ssp. japonica group; XP_015694070.2 from Oryza brachyantha; XP_025811579.1 from Panicum hallii; XP_003557155.1 from Brachypodium distachyon; XP_039842111.1 and XP_039805188.1 from P. virgatum; XP_051197997.1 from Lolium perenne; XP_047075120.1 from L. rigidum; XP_004964339.1 from Setaria italica; XP_037423695.1 and XP_037458488.1 from Triticum dicoccoides; XP_044426481.1, XP_044443345.1, and XP_044426478.1 from T. aestivum; NP_001144729.1 from Zea mays; XP_002437707.1 from Sorghum bicolor; XP_044959682.1 from Hordeum vulgare ssp. vulgare, XP_020147246.1 from Aegilops tauschii ssp. strangulate; XP_020112363.1 from Ananas comosus; XP_008813158.2 from Phoenix dactylifera; XP_010909418.1 from Elaeis guineensis; XP_009383130.2 from Musa acuminata ssp. malaccensis; XP_060211170.1 from Lycium barbarum; and XP_059318710.1 from L. ferocissimum were compared. (A) Phylogenic tree of THB1-like proteins. (B) Amino acid sequence alignment of 17 THB1-like proteins (amino acid identity >87.1% to LOC_Os06g02370.1). The amino acid sequence of LOC_Os06g02370.3 is indicated by white box. Predicted transmembrane domains (TMD) are indicated with red boxes. Identical and similar amino acid residues are shaded in black and gray, respectively. A nonsynonymous mutation site is indicated using a triangle.
Subsequently, multiple sequence alignments were performed among LOC_Os06g02370.1 and 17 LOC_Os06g02370.1-like proteins (amino acid identity >87.1% to LOC_Os06g02370.1) (Fig. 2B). Conserved domain analysis using SOSUI 1.11 revealed that LOC_Os06g02370.1 contains eleven transmembrane domains (TMD), which is highly conserved region among monocot 17 LOC_Os06g02370.1-like proteins (Fig. 2B). The R426H mutation in LOC_Os06g02370.1 was localized in TMD 8.
Phylogenetic comparisons of these proteins showed that they are highly conserved, and exhibit relatively close genetic relationships. However, none of the selected protein sequences have been studied in detail, and their functions are unknown; therefore, we concluded that THB1 is a putative transmembrane protein with unknown function.
Phylogenic and haplotype network analyses at LOC_Os06g02370 among JRC and WRCHaplotype network and phylogenic analyses were conducted to visualize the genetic relationships between the LOC_Os06g02370 region (6503 bp, chr06: 796,310..802,812), covering all the identified exons, introns, and UTRs and the promoter region (2,000 bp), among 69 WRCs and 50 JRCs. In total, 123 polymorphisms, including 96 SNPs and 27 InDels, were detected in the 119 accessions. Nineteen haplotypes were identified in the WRC and six haplotypes were identified in the JRC. Three haplotypes (H4, H10 and H22) were shared between both collections (Fig. 3A). Twenty-two haplotype variants were identified. Two accessions, ‘Akage’ (JRC17) and ‘Fukoku’ (JRC46/JP14924) were classified to haplotype2 (H2), which was identical to ‘Yukihikari’. Because the seed of ‘Akage’ (JRC17) was not available from NARO GenBank, two accessions named ‘Akage’ (JP5301 and JP14994) were used to confirm the genotype at YK3InDel-799424 and YK3SNP06-796157_2. All results confirmed that two accessions (‘Akage’) were identical to H2. Among 22 haplotypes, only H2 had a nonsynonymous SNP at YK3SNP06-796157_2.
Phylogenic and network analysis of THB1 gene. (A) Haplotype network of the THB1 in WRC and JRC. The size of each circle is proportional to the accession numbers encompassed, and different colors indicate the subgroups. The number of hatch marks reflects the number of nucleotide differences detected between haplotypes. (B) Phylogenic tree of THB1 based on nucleotide sequences of WRC and JRC. Arrowheads indicate the JRC ID used in test-cross experiments.
To determine the genetic association of the LOC_Os06g02370 region among the classified subpopulations, we investigated the association of three major haplotypes (H4, H18, and H22) carried by more than ten accessions. H18 and H22 were closely related and specific to indica and aus (Fig. 3A). These two haplotypes were distantly associated with temperate japonica- and tropical japonica-specific H4, indicating different variations in LOC_Os06g02370 among subpopulations (Fig. 3A).
To understand the reticulated relationship between the LOC_Os06g02370 haplotypes, a phylogenetic analysis was conducted using sequence data from TASUKE+ for the WRC and JRC (Fig. 3B). One hundred and nineteen sequences were classified into three clusters: Clusters I–III. Cluster I predominantly comprised temperate japonica and tropical japonica. Clusters II and III were predominantly indica and aus. Taken together with the haplotype network and phylogenetic analyses of the LOC_Os06g02370 region, we concluded that H2 was derived from H4 via H3 in temperate japonica.
Spikelet fertilities of F1 hybrids in test cross studyTo address the distribution of thb1 and thb2, thb2 carrier ‘Kirara397’ and thb1 carrier ‘Yukihikari’ were test-crossed with nine accessions (Fig. 4, Table 1). Hybrid sterility is commonly observed in inter-subspecific crosses, leading to reduced fitness and segregation distortion in the offspring (Chen et al. 2008, Kubo et al. 2016, Li et al. 2017, Long et al. 2008, Shen et al. 2017, Yang et al. 2012). We characterized the spikelet fertility of the F1 hybrids and classified them into three groups: Group A, >61%; Group B, >21% and <60%; and Group C, <20%. F1 spikelet fertility of the crosses between testers, four temperate japonica accessions, and a tropical japonica accession was consistently maintained at >61% and classified as group A. The F1 spikelet fertility of the crosses between two testers and two indica accessions (‘Akamai’) varied from 2.1% to 65.3%. In these four cross combinations, F1 spikelet fertility in 2022 (summer) was higher than that in 2021 (winter) suggesting the F1 sterility of the crosses with these two accessions (‘Akamai’) were affected by the environmental conditions. In addition, F1 spikelet fertility of the crosses between ‘Yukihikari’ and two indica accessions (‘Akamai’) was consistently lower than that of the crosses between ‘Kirara397’ and two indica accessions (‘Akamai’), and were classified as severe sterility (group C). These results suggested that the conditional F1 sterility was diverged genetically between ‘Kirara397’ and ‘Yukihikari’.
Gene structure and sequence alignment of the THB1 region. H4 is reference haplotype being carried by 51 accessions from 119 accessions from JRC and WRC. White cells indicate the same nucleotide as that of the reference haplotype. Black cells indicate variants. a indicates physical position on chromosome 6 (Nipponbare IRGSP-1.0 reference genome). b indicates 40-bp indel polymorphism of AAGCTACACAAACACAACGGAATGGTGAAGGTATAGAGAG.
| Cross combination | F1 plants | F2 population | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ♀ | ♂ | Spikelt fertility (%) | No. of plants | Goodness of fit | |||||||||||
| Accession | SG a | HT b | 2021 | 2022 | Average | Class c | Normal | Weak | Total | χ2(15:1) | p | ||||
| Kirara397 | Akage (JP5301) | tej | H2 | 84.1 | 87.9 | 86.0 | A | 456 | 27 | 483 | 0.36 | 0.55 | |||
| Kirara397 | Akage (JP14994) | tej | H2 | 94.2 | 82.0 | 88.1 | A | 448 | 36 | 484 | 1.17 | 0.28 | |||
| Kirara397 | Mansaku (JRC22/JP6735) | tej | H3 | 62.4 | 79.6 | 71.0 | A | 463 | 2 | 465 | 26.88 | <0.001 | |||
| Kirara397 | Shinyamadaho 2 (JRC37/JP6962) | tej | H3 | 87.5 | 83.0 | 85.3 | A | 480 | 1 | 481 | 29.97 | <0.001 | |||
| Kirara397 | Sensho (JRC04/JP4386) | trj | H10 | 81.3 | 67.9 | 74.6 | A | 498 | 1 | 499 | 31.17 | <0.001 | |||
| Kirara397 | Akamai (JRC21/JP9694) | ind | H22 | 26.8 | 65.3 | 46.1 | B | 449 | 3 | 452 | 24.07 | <0.001 | |||
| Kirara397 | Akamai (JRC43/JP4744) | ind | H22 | 39.0 | 62.1 | 50.6 | B | 464 | 1 | 465 | 28.90 | <0.001 | |||
| Kirara397 | Karahoushi (JRC44/JP10788) | ind | H22 | 0.4 | 16.7 | 8.6 | C | 492 | 0 | 492 | 32.80 | <0.001 | |||
| Kirara397 | Touboshi (JRC42/JP10772) | ind | H22 | 9.2 | 24.6 | 16.9 | C | 464 | 1 | 465 | 28.90 | <0.001 | |||
| Yukihikari | Akage (JP5301) | tej | H2 | 60.6 | 78.1 | 69.4 | A | 354 | 1 | 355 | 21.58 | <0.001 | |||
| Yukihikari | Akage (JP14994) | tej | H2 | 85.1 | 88.9 | 87.0 | A | 457 | 0 | 457 | 30.47 | <0.001 | |||
| Yukihikari | Mansaku (JRC22/JP6735) | tej | H3 | 77.3 | 97.2 | 87.3 | A | 427 | 33 | 460 | 0.67 | 0.41 | |||
| Yukihikari | Shinyamadaho 2 (JRC37/JP6962) | tej | H3 | 76.9 | 73.1 | 75.0 | A | 415 | 28 | 443 | 0.00 | 0.95 | |||
| Yukihikari | Sensho (JRC04/JP4386) | trj | H10 | 81.0 | 86.9 | 84.0 | A | 378 | 24 | 402 | 0.05 | 0.82 | |||
| Yukihikari | Akamai (JRC21/JP9694) | ind | H22 | 7.5 | 27.2 | 17.4 | C | 408 | 15 | 423 | 5.28 | 0.02 | |||
| Yukihikari | Akamai (JRC43/JP4744) | ind | H22 | 2.1 | 28.6 | 15.4 | C | 415 | 15 | 430 | 5.60 | 0.02 | |||
| Yukihikari | Karahoushi (JRC44/JP10788) | ind | H22 | 2.4 | 1.2 | 1.8 | C | 162 | 1 | 163 | 8.84 | <0.001 | |||
| Yukihikari | Touboshi (JRC42/JP10772) | ind | H22 | 6.3 | 8.8 | 7.6 | C | 272 | 18 | 290 | 0.00 | 0.98 | |||
a Subgroup, tej: temperate japonica, trj: tropical japonica, ind: indica.
b Haplotype at THB1 region.
c A: more than 61%, B: more than 21% and less than 60%, C: less than 20%.
F1 spikelet fertility of the crosses between testers and two indica accessions, ‘Karahoushi’ and ‘Touboshi’, ranged from 0.4% to 24.6%, and was classified as group C. F1 fertility of the crosses between ‘Kirara397’ and the two indica accessions in 2022 (summer) was higher than that in 2021 (winter), supporting that F1 fertility of these cross combinations was affected by environmental conditions. F1 fertility of the crosses between ‘Yukihikari’ and these two indica accessions ranged from 1.2% to 8.8%. Under both conditions, F1 fertilities of ‘Yukihikari’/‘Touboshi’ were consistently higher than those of ‘Yukihikari’/‘Karahoushi’.
Identification of THB1 haplotype responsible for hybrid breakdownWe assessed the growth habits of F2 plants derived from the crosses between thb2 carrier, ‘Kirara397’, and nine accessions (Fig. 4, Table 1). Thereafter, we calculated the segregation ratio of normal versus weak plants in each F2 population to determine whether the weak plants were genetically regulated by the two complementary recessive genes. The segregations ratios of normal to weak plants were 465:27 in F2 population of ‘Kirara397’/‘Akage’ (JP5301) and 448:36 in F2 population of ‘Kirara397’/‘Akage’ (JP14994). These correspond to a ratio of 15:1 (χ2 = 0.36, p = 0.55; χ2 = 1.17, p = 0.28) for two complementary recessive genes.
Complementary interactions were identified between thb2 and thb1(H2) from landrace ‘Akage’ (JP5301, JP14994), as shown in a cross with ‘Yukihikari’ (Yoneya et al. 2021). No interactions were detected between thb2 and H3 or H10 and H22. Finally, an H2-specific nonsynonymous SNP at YK3SNP06-796157_2 was identified as a causal mutation in thb1 (Fig. 4).
Distribution of thb2We evaluated the segregation ratios of normal versus weak plants in each F2 population of the crosses between ‘Yukihikari’, the thb1 carrier, and the nine target accessions to assess whether the weak plants were genetically regulated by the complementary recessive genes. The segregation ratios of normal to weak plants were 427:33 in F2 population of ‘Yukihikari’/‘Mansaku’, 415:28 in F2 population of ‘Yukihikari’/‘Shinyamadaho 2’, 378:24 in F2 population of ‘Yukihikari’/‘Sensho’, and 272:18 in F2 population of ‘Yukihikari’/‘Touboshi’. This corresponds to a 15:1 ratio for the two complementary recessive genes. In contrast, slight segregation distortion was observed in the crosses, ‘Yukihikari’/‘Akamai’ (JRC21/JP9694) and ‘Yukihikari’/‘Akamai’ (JRC43/JP4744). The segregation ratios of normal to weak plants were 408:15 in F2 population of ‘Yukihikari’/‘Akamai’ (JRC21/JP9694) and 415:15 in F2 population of ‘Yukihikari’/‘Akamai’ (JRC43/JP4744), indicating slight deficiency of double recessive genotype in both populations.
The present results demonstrated that thb2 was distributed in not only temperate japonica (‘Mansaku’ and ‘Shinyamadaho 2’) but also tropical japonica (‘Shensho’) and indica (‘Akamai’ and ‘Touboshi’).
Gene expression analysis of LOC_Os06g02370To address the possible association between the expression level of THB1 and hybrid breakdown, the gene expression levels of LOC_Os06g02370 at the seedling stage of 11 accessions were analyzed using quantitative RT-PCR. The expression levels of LOC_Os06g02370.1 using primer thb1-1 varied among accessions in the shoots and roots (Fig. 5). In shoots, the highest expression was observed in ‘Sensho’ (JRC04/JP4386), and the lowest expression was observed in ‘Yukihikari’. In roots, the highest expression was observed in ‘Akamai’ (JRC43/JP4744), and the lowest expression was observed in ‘Kirara397’. In both tissues, significant differences were noted among japonica accessions, but not among indica accessions. In contrast, similar amounts of combined expression of LOC_Os06g02370.1 and LOC_Os06g02370.3 using primer thb1-2 were observed among all tested accessions in both tissues. Although ‘Yukihikari’ and two accessions of ‘Akage’ carry H2, the expression levels of LOC_Os06g02370.1 were independent of the haplotype. Collectively, these results exclude the possibility that the expression level of THB1 is associated with hybrid breakdown.
Relative expression level of THB1 in nine accessions in 10 days after germination. (A) Relative expression levels of the LOC_Os06g02370.1 gene using primer set of thb1-1 were detected using quantitative RT-PCR. (B) Combined relative expression level of LOC_Os06g02370.1 and LOC_Os06g02370.3 using the primer set of thb1-2 were detected using quantitative RT-PCR. All data were normalized to the expression level of OsUBQ1 (Os03g0234200) (Yamamoto et al. 2010); error bars are SD for three independent experiments. Different letters indicate significant difference at p < 0.05 (Tukey’s test).
We found that THB1 is a novel causal gene that causes hybrid breakdown. Furthermore, this study revealed that THB1-like genes are distributed as a single copy per genome in each plant, as in the case of rice. It may play an essential role in the development and survival of plants. However, to date, no evidence has revealed the function of THB1 and its ortholog in the vegetative and reproductive development of either rice or other plants. Therefore, it would be interesting to determine the functions of THB1 and its orthologs. THB1 contains multiple putative TMDs without predicted signal peptides that are conserved across orthologs. Further studies are required to clarify the subcellular localization of THB1 and elucidate its function.
The present study clarified that a single amino acid substitution of R426H in LOC_Os06g02370.1, and R172H in LOC_Os06g02370.3 in the well-conserved putative transmembrane domain is the causal mutation for hybrid breakdown. To date, two molecular mechanisms of hybrid breakdown in plants have been reported: reciprocal silencing or loss of duplicated genes (Kubo et al. 2022, Vlad et al. 2010) and the elevated autoimmune response associated with the NBS-LRR gene (Alcázar et al. 2009, Yamamoto et al. 2010). In rice, the loss of duplicated Esa-associated factor 6 (EAF6) genes on chromosomes 1 and 12 causes hybrid breakdown (Kubo et al. 2022). However, the present study showed that the rice genome does not contain duplicated THB1. In contrast, hybrid breakdown2 (hbd2) and hbd3 encode casein kinase I (CKI1) and NBS-LRR, respectively (Yamamoto et al. 2010). hbd2-CKI1 allele gains its deleterious function that causes a weak phenotype by changing one amino acid, with hybrid breakdown attributed to an elevated autoimmune response. Yamamoto et al. (2010) suggested that hbd2-CKI1 acts like Avr protein or a protein disturbed by Avr protein, with hbd3-NBS-LRR(s) recognizing hbd2-CKI1 directly or indirectly to trigger the immune response signal. As in the case of gene combinations of hwj1 and hwj2, overactivated immune responses may trigger temperature-sensitive hybrid breakdown (Soe et al. 2022). In general, ambient temperature is a key environmental factor affecting the strength of plant immune responses, and pathogen effector-triggered immunity (ETI) is preferentially activated in plants at low temperatures (Cheng et al. 2013), indicating that the genes involved in ETI are more likely to be recruited for establishing low temperature-induced hybrid breakdown. The present study demonstrated that THB1 (R426/172H) may act like Avr or a protein disturbed by Avr as in case of hbd2-hbd3 system (Yamamoto et al. 2010). Now, we are challenging the map-based cloning of THB2 to clarify the molecular mechanism whether the ETI system or a novel molecular mechanism contributes to the present low temperature-dependent hybrid breakdown.
Evolutionally history of thb1(H2) and thb2Haplotype network analysis of THB1 locus among 119 accessions in the JRC (Ebana et al. 2008) and WRC (Kojima et al. 2005) revealed that thb1(H2) was originated from the major haplotype of Thb1(H4) via Thb1(H3). Notably, both thb1(H2) and Thb1(H3) were restricted to temperate japonica of the JRC, although Thb1(H4) was distributed in temperate japonica and tropical japonica in both collections. These results suggest that thb1 has recently evolved in temperate japonica. The present molecular marker analysis using YK3SNP06-796157_2 and Sanger sequence at YK3InDel-799424 confirmed that two accessions named ‘Akage’ (JP5301 and JP14994) carry thb1(H2). ‘Akage’ is a forerunner landrace of the rice cultivars in Hokkaido, followed by its role as parents in pure-line selection and modern cross breeding. ‘Fukoku’ and ‘Yukihikari’ were developed from the progeny of ‘Akage’. Therefore, thb1(H2) may be inherited by ‘Fukoku’ and ‘Yukihikari’ from ‘Akage’. ‘Akage’ has originated from Akita Prefecture, northern Japan. It is necessary to address the geographical distribution of thb1(H2) using more accessions to clarify the evolutionary history of THB1 locus, particularly if thb1(H2) originated from Thb1(H3) in the Akita Prefecture.
The present test cross study revealed that thb2 was distributed across three subgroups: temperate japonica, tropical japonica, and indica. Previous studies have shown that the domestication of Asian cultivated rice was from a single origin (Huang et al. 2012, Molina et al. 2011), which proposes the cycles of the introgression hypothesis. Based on this theory, wild rice was first domesticated as the ancient japonica. One part of ancient japonica hybridized with local wild rice in South Asia to form the indica subgroup. Another ancient japonica group was domesticated into the modern japonica group. The japonica subgroup is widely distributed across East Asia. It is divided into the temperate japonica and tropical japonica subgroups. The results presented here suggest that thb2 arose in an ancient japonica and introgressed into the other extant subgroups. In future studies, molecular cloning of THB2 locus, haplotype analysis, and distribution of thb2 using a large collection of Asian rice will clarify the evolutionary history of thb2.
Divergence of hybrid sterility between ‘Kirara397’ and ‘Yukihikari’In the present test cross study, the spikelet fertility of F1 hybrids of temperate japonica/temperate japonica and temperate japonica/tropical japonica crosses was consistently greater than 61%, whereas those of japonica/indica crosses were varied from almost completely sterile to half sterile among the cross combinations, without hybrid weakness. Among eight temperate japonica/indica crosses, the spikelet fertility of F1 plants crossed with ‘Yukihikari’ was consistently lower than that of F1 plants crossed with ‘Kirara397’. Hybrid sterility is a complex quantitative trait controlled by multiple loci. Maekawa et al. (1998) reported that ‘Yukihikari’ carried different hybrid sterility genes from another japonica cultivar ‘T-65’ in japonica/indica hybrids. However, to date, no genetic analysis has been performed to differentiate the hybrid sterility loci in ‘Yukihikari’. More than 22 loci responsible for hybrid sterility in crosses between different subgroups have been identified using genetic mapping study (reviewed by Zhang et al. 2022). Further studies are required to confirm which hybrid sterility genes contributed to varied sterility between ‘Yukihikari’/indica and ‘Kirara397’/indica. This will clarify the possibility that the differentiation in combinations of distinct hybrid sterility loci diverged between ‘Yukihikari’ and ‘Kirara397’.
In the present study, the double recessive genotype, thb1 thb2, was slightly deficient in the F2 populations of the crosses ‘Yukihikari’/‘Akamai’ (JRC21/JP9694 and JRC43/JP4744), resulting slight segregation distortion. To date, two genetic models have been proposed to explain hybrid sterility. One model is the one-locus sporo-gametophytic interaction model (Kitamura 1962), and was followed by S5, S7, Sa, Sc, and HSA1 (Chen et al. 2008, Kubo et al. 2016, Long et al. 2008, Shen et al. 2017, Yang et al. 2012, Yu et al. 2016). The other is a duplicate gametic lethal model, interrupting two independent loci in hybrid sterility (Oka 1957, 1974), followed by DPL1/DPL2 (Mizuta et al. 2010). Among all the hybrid sterility loci, S5 is a major reproductive barrier regulator in cultivated rice (Ikehashi and Araki 1986, Song et al. 2005) and is loosely linked to THB1 on the short arm of chromosome 6 by a physical distance of almost 5 Mb. At S5 locus, the interaction of S5-i with S5-j alleles in sporophytes results in the abortion of female gametes that carry S5-j allele (Ikehashi and Araki 1986). Therefore, the selective abortion of female gametes carrying the chromosome segment of thb1–S5-j from ‘Yukihikari’ in F1 plants is suggested to lead to transmission ratio distortion of THB1–S5-i from indica. This resulted in deficient, weak F2 individuals in test crosses ‘Yukihikari’/‘Akamai’ (JRC21/JP9694) and ‘Yukihikari’/‘Akamai’ (JRC43/JP4744). Further studies are required to clarify whether S5 contributes to the segregation distortion in these cross combinations. In the test cross ‘Yukihikari’/‘Touboshi’, although severe sterility was observed in F1 hybrids, no segregation distortion of normal plants versus weak plants were observed in F2 population. We need to confirm if ‘Touboshi’ and two ‘Akamai’ accessions possess different hybrid sterility loci in F1 hybrids with ‘Yukihikari’. In addition to female gamete abortion, hybrid sterility between japonica and indica is caused by male gamete abortion and embryo abortion (Liu et al. 2004). Future studies should examine the contribution of each characteristic to the variations in spikelet sterility.
Application of dCAPS maker YK3SNP06-796157_2 in rice breedingIn the present study, we developed a codominant marker, YK3SNP06-796157_2, to determine the genotype of the causal SNP in thb1. The bulk population method has been widely employed in conventional rice breeding for decades (Matsubara 2020). In this method, the early generation population is subjected to natural and viable selection (Allard 1960, Ikehashi and Fujimaki 1980). Thus, many weak and/or sterile genotypes are likely to be eliminated from the population before the establishment of an advanced-generation population. A close linkage between favorable and deleterious alleles results in cosegregation (Lynch and Walsh 1998). Therefore, if a deleterious allele, thb1(H2) in ‘Yukihikari’, was closely linked to an allele favorable to rice breeding, the favorable allele may be eliminated in early generation. Therefore, to avoid such problems, YK3SNP06-796157_2 may be useful for selecting plants heterozygous for THB1 in early generations. This marker-assisted selection (MAS) allows the maintenance of a closely linked favorable allele until an advanced-generation population is combined with a higher probability of breaking the linkage between thb1(H2) and favorable alleles. Thus, the bulk population method with MAS using YK3SNP06-796157_2 will be a challenge for rice breeding programs in the future.
Information on the distribution of hybrid breakdown-associated alleles among cross parents is a prerequisite for rice breeding (Matsubara 2020). The SNP marker developed here, YK3SNP06-796157_2, will be useful for extensively surveying cross parents. Furthermore, amplicon sequencing of THB1 or the development of SNP arrays based on a causal SNP in THB1 will facilitate recent genomics-assisted rice breeding.
KK and TW designed the study, performed experiments, interpreted data, and wrote the manuscript. All authors have read and agreed to the final version of the manuscript.
The authors thank Drs. K. Fujino, K. Onishi, and K. Matsubara for their valuable suggestions. We thank Dr. K. Koyanagi for her advices. We also thank NARO GenBank and HRO GenBank for providing rice seeds.
