The Effects of Rf4 and the Genetic Mechanism Behind Fertility Restoration of Wild Abortive Cytoplasmic Male Sterility (WA-CMS) in Japonica Rice (Oryza sativa ssp. Japonica)

Wild abortive-type cytoplasmic male sterility (WA-type CMS) has been exclusively used in hybrid seed production in indica rice cultivars, and fertility restoration in WA-type CMS is controlled by two major restorer genes, Rf3 and Rf4, through a sporophytic mechanism. However, the genetic mechanism underlying fertility restoration in WA-type CMS in japonica cultivars is poorly understood. In the present study, C418, a leading Chinsurah Boro II- (BT)-type japonica restorer line, showed partial restoration ability in WA-type japonica CMS lines. The 1:1 segregation ratio of partially fertile to sterile plants in a three-cross F1 population indicated that fertility restoration is controlled by one dominant gene. Gene mapping and sequencing results revealed that the target gene should be Rf4. The Rf4 gene restores fertility through a sporophytic mechanism, but the Rf4 pollen grains show a preferential fertilization in the testcross F1 plants. Furthermore, Rf4 was confirmed to have only a minor effect on fertility restoration in WA-type japonica CMS lines, and Rf gene dosage effects influenced the fertility restoration of WA-type CMS in japonica rice. The results of our study not only provide valuable insights into the complex genetic mechanisms underlying fertility restoration of WA-type CMS but will also facilitate the efficient utilization of WA-type CMS in japonica rice lines. Supplementary Information The online version contains supplementary material available at 10.1186/s12284-022-00605-0.

Rice is a staple food for more than half of the world's population, and the development of new and effective agricultural technologies to increase rice production is essential to meet global food demands and to ensure food security.In the past 50 years, three-line hybrid breeding in rice has made a significant contribution to the increase in yield (Chen and Liu 2014;Huang et al. 2014;Li et al. 2007;Kim and Zhang 2018).In rice, over 20 types of CMS systems have been discovered, among which three representative CMS types, including the wild abortive (WA)-, Honglian (HL)-, and Chinsurah Boro II (BT)-types have been widely used for breeding (Chen and Liu 2014;Huang et al. 2014;Li et al. 2007).At present, WA-type and HL-type CMS have been used in three-line indica hybrid seed production, and BTtype CMS has used in the breeding of japonica hybrids for cultivation (Chen and Liu 2014;Huang et al. 2014;Li et al. 2007;Yuan 1994).Several Rf loci used for the breeding of restorer lines have been identified and mapped.Rf3 and Rf4, two major fertility restorer genes for WAtype CMS, have been mapped to chromosomes 1 and 10, respectively (Ahmadikhah and Karlov 2006;Tang et al. 2014;Zhang et al. 1997).Rf5 and Rf6, two major fertility restorer genes for HL-type CMS, have been mapped to chromosomes 10 and 8, respectively (Hu et al. 2012;Huang et al. 2000Huang et al. , 2012;;Huang et al. 2015a, b).The Rf1a/Rf1b genes at the Rf1 locus of BT-type CMS have been mapped to chromosome 10 ( Akagi et al. 1996;Komori et al. 2004;Wang et al. 2006).With the exception of Rf3, all of these Rf genes have been cloned.Male fertility in rice plants carrying WA-type cytoplasm is restored sporophytically, while male fertility of rice plants carrying the BT-and HL-type cytoplasm is restored gametophytically.Furthermore, Rf5 and Rf1a were found to be the same gene, and the Rf1a(Rf5)/Rf1b locus is closely linked to Rf4 on chromosome 10.
Currently, three-line indica hybrids are widely used, but the shortcomings of BT-type CMS (unstable sterility of some BT-type CMS lines and the threat of genetic vulnerability when using a single cytoplasm source) have restricted its use, and thus the cultivation area of threeline japonica hybrids relative to the entire japonica rice planting area in China is only ~ 5% (Deng et al. 2006).Most rice breeders are of the opinion that accelerating the development and application of japonica hybrids may be an effective way to increase the total rice grain yield in China.Considering the successful use of WAtype CMS and the stable sterility of WA-type CMS lines in indica-type cultivars, breeders have attempted to use the WA-type CMS in japonica hybrids; however, they encountered several problems such as difficulty in obtaining restorer lines and deteriorating flowering habits of WA-type japonica CMS lines (Tang et al. 2008;Yang 1994).Thus, WA-type CMS has no practical use in japonica rice production, and there are few studies that have examined the genetic basis of fertility restoration in WA-type japonica CMS lines.In breeding, the japonica restorer lines used in China have been only bred for BT-type japonica CMS, and most restorers usually carry the Rf1 locus (Akagi et al. 1996;Chen and Liu 2014;Huang et al. 2014;Komori et al. 2004;Wang et al. 2006;Yang et al. 2016).Rf1 used for breeding of BT-type CMS restorer lines in China is first transferred from 'IR8' , which is a donor of Rf genes to 'IR24' , a known indica restorer for WA-type indica CMS lines carrying Rf4, and this Rf1 resource has been used in most BT-type japonica restorer lines (Yang et al. 2016).It seems reasonable to assume that most BT-type japonica restorer lines should carry Rf4.However, this is confusing, because BT-type japonica restorers usually exhibit a poor ability to restore male fertility in WA-type japonica CMS lines (Zhu et al. 2010;Zhang et al. 2018).Therefore, it is important that we understand the genetic basis of fertility restoration in WA-type japonica CMS lines, which will be helpful for breeding high-yielding WA-type japonica rice hybrids.
In breeding practice, C418 is an elite BT-type restorer line that has been widely used for breeding japonica hybrids in China.Here, we report that C418 is able to restore fertility in WA-type japonica CMS lines, and demonstrate that Rf4 is present in C418 and is responsible for the fertility restoration.Furthermore, we analyzed the functional model and effects of Rf4 on fertility restoration in WA-type japonica CMS lines.The results of our study will assist breeders in analyzing the differences in fertility restoration in WA-type CMS in indica/japonica genetic backgrounds and will strengthen the breeding of japonica restorer lines for the development of WA-type japonica hybrids.

Plant Materials
Two WA-type japonica CMS lines, 'WA-LiuqianxinA' (WA-LqxA) and 'WA-NipponbareA' (WA-NipA), and the japonica restorer line C418 were used in this study.To evaluate the fertility restoration ability of C418 in WAtype japonica CMS lines, two testcross F 1 hybrids were generated by separately crossing C418 with WA-LqxA and WA-NipA.To identify the Rf locus involved in the fertility restoration of WA-type CMS, the WA-NipA// Nip/C418, WA-NipA/C418 F 2 , and WA-LqxA/C418 F 2 populations were constructed for the genetic analysis of fertility restoration and gene mapping.To analyze the vigor of pollen grains produced by the testcross F 1 plants, the WA-NipA//WA-NipA /C418 population was generated.We crossed plants harboring the heterozygous alleles of Rf4 in the Nip//Nip/C418 population, followed by four backcrosses to Nip, and the near-isogenic line for Rf4 (NIL Rf4 ) was developed from the BC 5 F 3 population via molecular marker-assisted selection.For evaluation of the Rf4 gene's ability to restore fertility to WA-type japonica CMS lines, we made two testcross F 1 hybrids, WA-NipA/NIL Rf4 and WA-LqxA/NIL Rf4 , and generated the WA-NipA/NIL Rf4 //NIL Rf4 population.All of these materials were planted in the experimental field at Yangzhou University in Yangzhou, Jiangsu Province, China.

Fertility Scoring and Genetic Analysis
The pollen grain fertility and natural spikelet fertility of five plants from the CMS lines, 10 plants from the testcross F 1 lines, and each plant in the WA-NipA//Nip/ C418, WA-NipA/C418 F 2 , WA-LqxA/C418 F 2 , WA-NipA//WA-NipA/C418, and WA-NipA/NIL Rf4 //NIL Rf4 populations were observed.For pollen grain fertility, three mature anthers were harvested and the pollen grains were stained with 1% I 2 -KI solution.Thus, the number of dark-blue, clear (unstained), and typical abortive pollen grains in each individual were counted under a microscope, and the pollen grain fertility was estimated using the percentages of stained pollen grains.Natural spikelet fertility levels were measured as the average seed-setting rates by counting the filled and unfilled grains on the main panicle of one plant harvested 20 days after flowering.
For the genetic analysis, the stained pollen grain rate was used as the main criterion for the evaluation of sterile and partially fertile plants.Plants with a typical aborted pollen grain rate of > 95% were categorized as sterile, and those with a > 90% stained pollen grain rate were considered to be partially fertile.A chi-squared test was used to test the goodness-of-fit of the hypothesis.

DNA Extraction, PCR Amplification, and DNA Sequencing
Total genomic DNA was extracted from fresh leaves using the cetyltrimethylammonium bromide (CTAB) method with some modifications (Rogers and Bendich 1985).STS10-27 and STS10-16, two insertion/deletion markers that map to loci that are closely linked to the Rf4 gene, were used to genotype the individuals in the fertility segregation populations.The molecular marker amplifications were performed in 20 µL reactions containing 1 × supplied PCR buffer, 0.1 mmol/L of each dNTP, 1.0 U Taq polymerase, 0.2 µmol/L primer, and 20 ng template DNA.Amplification was performed using the following procedure: samples were denatured at 94 °C for 4 min, followed by 32 cycles of 94 °C for 45 s, 55 °C for 45 s, and 72 °C for 50 s, with a final extension step at 72 °C for 5 min.The amplification products were separated by electrophoresis in 3% agarose gels containing ethidium bromide, and the bands were visualized with a GEL DOC 1000 system (Bio-Rad Company, Hercules, CA, USA).DNA fragments covering Rf4 from 'C418' were amplified using high-fidelity GXL-Taq (Well-Offer Biotechnology Company, Nanjing, China).The analyzed sequences contained the 206-bp 5′-upstream region and the 164-bp 3′-downstream region of the Rf4 gene.PCR products were recycled using a recycling kit (Vazyme, Nanjing, China), and then ligated into the pEASY-Blunt Zero vector.Five plasmids were sequenced by GENEWIZ (Suzhou, China), and a clone with the correct DNA sequence was selected.A sequence alignment was performed with the BLAST algorithm-based network services at the National Center for Biotechnology Information (NCBI).The names and DNA sequences of the primers used in this study are given in Additional file 1: Table S1.
The genetic background of NIL Rf4 was determined with the 40 k rice SNP-array, a whole-genome single nucleotide polymorphism (SNP) array with 40,000 SNP and InDel markers (Greenfafa, Wuhan).

Data Analysis
The analysis of variance (ANOVA) package in SPSS15.0 was used for statistical analysis of the fertility of the lines and populations used in this study.

Evaluation of the Fertility Restoration Capability of C418
To examine the ability of C418 to restore fertility in WAtype japonica CMS lines, we determined the pollen grain and spikelet fertility levels of individual plants of two WA-CMS lines and the testcross F 1 hybrids obtained from the crosses between two WA-type CMS lines and C418.WA-NipA and WA-LqxA exhibited amorphous aborted pollen grains, and the natural spikelet levels of these two CMS lines were 0, indicating that the WAtype japonica CMS lines were completely sterile (Fig. 1a,  b, e, f ).The pollen stainability of the testcross F 1 plants was > 95%, and the natural spikelet fertility levels of these testcross F 1 plants ranged from 10.86 to 64.63%, indicating that plants carrying the WA-type CMS cytoplasm produced morphologically normal pollen, although partially stained pollen grains may lack the ability to germinate (Fig. 1).Also, the average natural spikelet fertilities of the WA-LqxA/C418 F 1 and WA-NipA/C418 F 1 plants were statistically different (24.63% ± 7.70% vs. 54.06% ± 8.47%, P = 4.24E−6).These results indicate that the restorer line C418 is able to partially restore fertility to the WA-type japonica CMS lines tested in this study, and that fertility restoration was influenced by the subtypes of the CMS line nuclear backgrounds.

Genetic Analysis of Fertility Restoration and Gene Mapping
In order to explore the mechanism underlying fertility restoration, the pollen grains and natural spikelet fertility levels of 160 plants in the WA-NipA//Nip/ C418 population were investigated.Based on the pollen grains, the plants in the WA-NipA//Nip/C418 population could be divided into two categories: (1) the plants resembled the CMS lines with degenerated anthers and shrunken pollen grains and were sterile, and (2) the plants were similar to the testcross F 1 hybrids with normal anthers and dark-stained pollen grains and were partially fertile.The segregation ratio of sterile plants to partially fertile plants was 1:1 (χ 2 = 0.21, which was < χ 2 0.05 = 3.84), indicating that fertility restoration is conditioned by one dominant restorer gene (Table 1).The natural spikelet fertilities of the sterile plants were < 3.88%, and the natural spikelet fertilities of the partially fertile plants ranged from 0 to 61.92% and showed a continuous variation.These results indicate that there might be one or more minor-effect Rf genes in C418 that affect fertility restoration in WA-type CMS lines.
Based on the breeding pedigrees of BT-type japonica hybrids in China, most BT-type restorer lines should  theoretically carry the major restorer gene Rf4 (Yang et al. 2016;Kazama and Toriyama 2014).In order to map the target Rf gene in C418, we first needed to determine whether Rf4 is present in C418.We sequenced the Rf4 allele from C418, and the nucleotide sequence of Rf4 in the C418 was identical to that from IR24, demonstrating that it carried the functional Rf4 allele.Subsequently, we tested whether Rf4 is related to the fertility restoration of WA-type CMS by genetic linkage analysis.A previous study identified two InDel marker loci, STS10-27 and STS10-16, that flank Rf1/Rf4 on Chromosome10, and these markers show polymorphisms between C418 and Nip.We used these two markers to genotype all plants in the WA-NipA//Nip/C418 population.All 88 partially fertile plants showed heterozygous genotypes at these two loci, while all 72 sterile plants were homozygous for the NIP alleles at the two marker loci.Meanwhile, three recombinants were identified.Based on the genotype and phenotype of recombinants, Rf gene was mapped in the physical region of ~ 387 kb co-segregated with molecular marker STS10-46, where Rf4 was contained (Fig. 2).
These results indicate that the Rf gene in C418 that is responsible for the fertility restoration of WA-type CMS is most probably Rf4.
The Rf4 Pollen Grains are Favored in Fertilization in the Testcross F 1 Plants A total of 350 and 254 plants in the WA-NipA/C418 F 2 and WA-LqxA/C418 F 2 populations, respectively, were grown to maturity, and the pollen grains and natural spikelet fertilities of all F 2 plants were quantified.Based on the fertility criteria (Materials and methods), there were 343 and 251 partially fertile plants, and seven and three sterile plants in the WA-NipA/C418 F 2 and WA-LqxA/C418 F 2 populations, respectively (Table 1).These results indicate that the fertility restoration in WA-type CMS japonica rice is sporophytic, but the segregation ratio of partially-fertile plants to sterile plants did not fit a one-gene (3:1) sporophytic model.To explore the reasons behind the aberrant segregation of fertility restoration, we used the markers STS10-27 and STS10-16 to genotype all of the F 2 plants.In the WA-NipA/C418 F 2 population, there were 192 plants with the Rf4Rf4 genotype and 151 plants with the Rf4rf4 genotype, all of which were partially fertile, while the seven sterile plants had the rf4rf4 genotype.Similarly, 251 partially fertile plants in the WA-LqxA/C418 F 2 population were Rf4-containing plants, including 137 plants with the Rf4rf4 genotype and 114 plants with the Rf4Rf4 genotype, and three sterile plants carried the rf4rf4 genotype.Obviously, the genotypes at the Rf4 locus deviated from the expected 1:2:1 segregation ratio in the two F 2 populations for a common single gene model.However, plant phenotypes were completely consistent with their genotypes, indicating that Rf4 is indeed correlated with the fertility restoration of WA-type CMS.
In considering the aberrant fertility and genotypic segregation ratios, we hypothesized that both Rf4 and rf4 pollen grains in the testcross F 1 plants are able to restore fertility in WA-type japonica CMS lines, but that the rf4 pollen grains are less competitive than the grains carrying Rf4.To test this hypothesis, we further constructed the WA-NipA//WA-NipA/C418 population that consisted of 37 plants, and the pollen grains and genotypes of these plants were investigated.We found that 35 plants had > 90% stained pollen grains and the Rf4rf4 genotype, while two plants showed > 95% shrunken pollens and had the rf4rf4 genotype.This distorted segregation indicated that the majority of male gametes in the WA-NipA/C418 F 1 plants involved in pollination should carry the Rf4 gene.Taken together, these results demonstrate that the pollen grains carrying Rf4 in the testcross F 1 plants have different viability compared to grains that do not carry Rf4, and that Rf4 pollen grains are preferentially selected during fertilization.

Development of NIL Rf4 in the 'Nipponbare' Japonica Genetic Background
In an attempt to identify the effects of Rf4 on the fertility restoration of WA-type CMS in japonica lines, we developed a near-isogenic line homozygous for Rf4 in the Nip genetic background using marker-assisted selection (Fig. 3a).In 2017, a BC 5 F 3 line was obtained that was very similar to Nip in terms of its main biological characteristics (Fig. 3b).In addition, the genetic background of the BC 5 F 3 line was analyzed using the 40 K rice SNP-array, which showed that the recurrent parent genome recovery was 98.45% in this line, and the C418 chromosome segment containing Rf4 had been successfully transferred into Nip.Thus, we successfully developed an Rf4 NIL, which we named NIL Rf4 , and it was suitable for further study.

The Effects of Rf4 on the Fertility Restoration of WA-type Japonica CMS Lines
We next examined the pollen grains following I 2 -KI staining and determined the natural spikelet fertility levels of WA-NipA/NIL Rf4 F 1 and WA-LqxA/NIL Rf4 F 1 plants.All F 1 plants had dark-staining pollen grains, and there were no differences between the different crosses (data not shown).The natural spikelet fertility levels of the testcross F 1 hybrids from WA-NipA and NIL Rf4 , and WA-LQXA and NIL Rf4 were all < 1%.These results indicate that Rf4 has minor effects on the fertility restoration of WA-type japonica CMS lines.In the WA-NipA/NIL Rf4 // NIL Rf4 population, there were 14 plants with the Rf4rf4 genotype and 19 plants with the Rf4Rf4 genotype, and the average natural spikelet fertility levels of the plants were 4.40% and 31.32%,respectively (Fig. 4).These results imply that Rf4 exerts a dosage effect on the fertility restoration of WA-type japonica CMS lines.

Discussion
Wild-abortive (WA) CMS is an ideal type of sporophytic CMS that has been used for the large-scale commercial production of hybrid indica rice since the 1970's in China (Barclay 2010;Li et al. 2007).However, insufficient numbers of japonica restorer lines have been a major limiting factor in the application of WA-type CMS to hybrid production of japonica rice seed.The breeding of WAtype japonica restorer lines has been a crucial issue in the development of WA-type japonica hybrids.In the japonica subspecies, only BT cytoplasm has been used for seed production, and the restorer lines have been bred for BT-type CMS lines.There are two ways to generate WAtype japonica restorer lines; the first is to systematically search for WA-type japonica restorer lines in the BTtype japonica restorer lines by testcrossing, and the second is to breed new WA-type japonica restorer lines.In our previous study, most BT-type restorer lines were able to restore fertility in WA-type japonica CMS lines, but most of the testcross F 1 plants had low spikelet fertility rates (Zhu et al. 2010;Zhang et al. 2018).It is difficult to search for WA-type japonica restorer lines in the existing BT-type japonica restorer lines, but is a reasonable strategy to breed new WA-type japonica restorer lines based on BT-type japonica restorer lines.Thus, understanding the genetic basis of fertility restoration in WA-type japonica CMS lines and the Rf genes that are present in BT-type restorer lines will assist in breeding of WA-type japonica restorer lines.Based on the breeding process of three-line japonica hybrids, C418, a leading BT-type japonica restorer line with a partial indica phenotype, has been used widely for breeding three-line hybrids in China (Yang et al. 2016).Thus, we considered C418 to be a suitable parent to use in a study of the genetic basis of fertility restoration in WA-type japonica CMS lines and to identify the Rf genes that are important for fertility restoration of WA-type CMS in BT-type japonica restorer lines.In the present study, we found that C418 can restore the fertility of WA-type CMS in testcross F 1 plants, which is consistent with the results of our previous study (Zhang et al. 2018).Many studies have shown that fertility restoration is influenced by genetic background, environmental factors (such as temperature and relative humidity), and the interaction with the different types of CMS cytoplasm.We found variable spikelet fertility levels in the testcross F 1 hybrids derived from C418.Considering that the testcross F 1 hybrids exhibited similar heading dates (data not shown), we suggest that the differences in the ability of C418 to restore fertility in the CMS lines could be related to the existence of unknown genes that participate in the fertility restoration of WAtype CMS in the WA-type japonica CMS lines.
It is well known that fertility restoration of WA-type CMS in indica lines is mainly controlled by two major Rf genes, Rf3 and Rf4.However, pollen and spikelet fertility levels usually show a continuous distribution in a given genetic population, and different genetic hypotheses have been suggested for fertility restoration in WAtype CMS in previous studies (Govinda Raj and Virmani 1988;Bharaj et al. 1991).In the present study, a dominant gene in C418 was found to restore normal anthers and stainable pollen grains to WA-type japonica CMS lines.Through DNA sequencing and linkage analysis, the Rf4 gene in C418 was identified as being responsible for fertility restoration of WA-type CMS.However, continuous variation in spikelet fertility levels of the partially fertile plants was observed in the three-cross population, indicating that there might be minor-effect Rf genes in C418 which can affect the fertility restoration of WAtype CMS.Also, the testcross F 1 plants derived from WA-LqxA and C418 in this study had stainable pollen grains and natural spikelet fertility levels of 20%, but the testcross F 1 plants derived from WA-LqxA and C9083, a BT-type japonica restorer carrying Rf4, exhibited typical abortive pollens and no natural spikelet fertility in our previous study, which provides additional evidence that there are other Rf genes in addition to Rf4 in C418 (Zhang et al. 2018).Also, the natural spikelet fertilities of WA-NipA/NIL Rf4 F 1 plants were significantly lower than in the WA-NipA/C418 F 1 plants, providing yet another piece of evidence to show that C418 carries other Rf genes as well as Rf4.To identify the unknown Rf genes in C418, we are constructing chromosome segment substitution lines from the cross of C418 (the recipient) and Nip (the donor), which will lay a solid foundation to discover the genetic basis of fertility restoration of WA-type CMS in japonica rice lines.
Fertility restoration of WA-type CMS has been documented as a sporophytic mechanism because indica rice plants with the Rfrf genotype and CMS produce all fertile pollen grains.The pollen grains produced in the testcross F 1 hybrids are thought to transmit the Rf and rf alleles with the same frequency through the paternal parent.As a result, sterile plants with the rfrf genotype can be observed in F 2 populations, and the percentage of sterile plants is approximately 25% in the single-locus model.In the present study, the appearance of malesterile individuals in the F 2 populations derived from WA-type CMS japonica lines and C418 indicates that Rf4 sporophytically restores male fertility, but the percentage of sterile plants was extremely low, deviating drastically from the expected 1:4 ratio in normal sporophytic fertility restoration.Furthermore, the ratios of plants with the Rf4Rf4, Rf4rf4, and rf4rf4 genotypes in these F 2 populations deviated from the expected 1:2:1 ratio in the sporophytic model, and the ratios of partially fertile plants with the Rf4Rf4 and Rf4rf4 genotypes were similar to a gametophytic one-gene 1:1 segregation pattern.Based on the genotypes and phenotypes of plants in the WA-NipA//WA-NipA/C418 population, we were able to confirm that the rf4 pollen grains are much less competitive than the Rf4 pollen grains in the testcross F 1 plants, and most of the rf4 grains may fail in fertilization due to lack of viability or reduced competitiveness relative to the Rf4 pollen grains.Thus, the results of our study are the first to show that pollen grains carrying Rf4 are functionally favored in fertilization over pollen grains that carry rf4 in the presence of WA-type cytoplasm in the japonica genetic background, which differs from the functional model of Rf4 acts to restore male fertility in the indica genetic background.Further studies will be required to identify factors that affect the viability of rf4 pollen grains in the japonica genetic background, which will bring new insights on the genetic basis of fertility restoration in WA-type CMS.
In WA-type indica hybrids, the two major genes Rf3 and Rf4 have been required for full fertility restoration capabilities of WA-type indica CMS lines, and these two genes exert an additive effect on fertility restoration (Ahmadikhah and Karlov 2006;Tang et al. 2014;Zhang et al. 1997Zhang et al. , 2002)).Also, the ability of Rf4 to restore fertility is greater than that of Rf3, and Rf4 is able to restore a high level of spikelet fertility to WA-type CMS lines (Cai et al. 2013;Zhang et al. 2002;Yao et al. 1997).In the present study, the natural spikelet fertilities of WA-LqxA/ NIL Rf4 F 1 and WA-NipA/NIL Rf4 F 1 plants were very low (< 1%).Thus, our results confirm that the capability of Rf4 to restore fertility in WA-type japonica CMS lines is poor, which is consistent with our previous results and also provides a good explanation of why most BT-type japonica restorer lines are not good at restoring the fertility of WA-type japonica CMS lines (Tada 2007;Zhang et al. 2018).However, the different fertility restoration capability of Rf4 in the indica/japonica genetic backgrounds indicates that minor-effect Rf genes are widely present in WA-type indica CMS lines, and that these genes can modify the effects of Rf4 on the fertility restoration of WA-type CMS.In the WA-NipA/NIL Rf4 //NIL Rf4 population, the average fertility levels of plants carrying the Rf4Rf4 genotype were greater than those of plants carrying the heterozygous Rf4rf4 genotype, indicating that the dosage of the Rf gene influences the fertility restoration of WA-type CMS japonica lines.Additionally, the Rf4Rf4 plants had natural spikelet fertilities of only 31.32%, indicating that it is not feasible to breed the WAtype japonica restorer lines only based on Rf3 and Rf4.Thus, exploring other Rf genes for WA-type CMS is crucial to develop WA-type japonica restorer lines.

Conclusion
In summary, we found that the Rf4 gene is present in C418 and is associated with the fertility restoration of WA-type japonica CMS lines.The Rf4 pollen grains produced by testcross F 1 plants have priority over the rf4 pollens in fertilization, although Rf4 restores male fertility in a sporophytic manner.Rf4 had minor effects on fertility restoration of WA-type CMS in japonica line, and there was a dosage effect of Rf on the fertility restoration of WA-type CMS lines.These results will be valuable in future breeding of WA-type japonica restorer lines.

Fig. 2
Fig. 2 Genetic mapping of the Rf gene.a Rf gene was mapped into ~ 387 kb physical region flanked by marker STS10-27 and STS10-16 on chromosome 10 that overlapped by AC068950, AC092489, AC068923 and AC079888 (b)

Fig. 3
Fig. 3 Breeding scheme and gross plant morphology of the parental line and the newly-developed NIL (near-isogenic line) in the 'Nipponbare' genetic background.a Breeding scheme showing the development of NIL Rf4 .b Mature plants of 'Nipponbare' and NIL Rf4 .Scale bars = 20 cm.c Genomic constitution of NIL Rf4 .The red and blue boxes on the chromosomes indicate the introgressed chromosomal segments derived from 'C418'; the red box represents a region that is homozygous for the C418 genotype, and the blue boxes represent heterozygous regions.The substituted regions on the distal end of Chr. 10 cover the region in which Rf4 is located

Fig. 4
Fig. 4 Natural fertility levels of plants with different Rf genotypes in the WA-NipA/NIL Rf4 //NIL Rf4 population.a Mature plants with the Rf4rf4 genotype.b Mature plants with the Rf4Rf4 genotype.c Natural spikelet fertility levels of the WA-NipA/NIL Rf4 //NIL Rf4 plants.Scale bars = 20 cm in a and b

Table 1
Fertility segregation in the F 2 populations and three-cross F 1 progeny plants