Molecular mapping of restorer-of-fertility 2 gene identified from a sugar beet (Beta vulgaris L. ssp. vulgaris) homozygous for the non-restoring restorer-of-fertility 1 allele

Key message By genetically eliminating the major restorer - of - fertility gene ( Rf ), a weak Rf gene was unveiled. It is an allele of Z , long known as an elusive Rf gene in sugar beet. Abstract In the hybrid breeding of sugar beet, maintainer-genotype selection is a laborious process because of the dependence on test crossing, despite the very low occurrence of this genotype. Marker-assisted selection (MAS) of the maintainer genotype is highly desired by sugar beet breeders. The major restorer-of-fertility gene (Rf) was identified as Rf1, and its non-restoring allele (rf1) was discriminated at the DNA level; however, some of the rf1rf1 selections retained an as yet unidentified Rf, another target locus for MAS. The objective of this study was to identify this Rf. An rfrf1 plant was crossed to a cytoplasmic male-sterile sugar beet and then backcrossed to obtain progeny segregating the unidentified Rf. The progeny exhibited partial male-fertility restoration that was unstable in single plants. The segregation ratio of restored vs. non-restored plants suggested the involvement of a single Rf in this male-fertility restoration, designated as Rf2. We confirmed the feasibility of molecular tagging of Rf2 by identifying four shared amplified fragment length polymorphism (AFLP) fragments specific to 17 restored plants. Bulked segregant analysis also was performed to screen the Rf2-linked AFLP markers, which were subsequently converted into 17 sequence-tagged site markers. All the markers, as well two additional chromosome-IV-assigned markers, were linked to each other to form a single linkage map, on which Rf2 was located. Our data suggested that Rf2 is likely an allele of Z, long known as an elusive Rf gene in sugar beet. We also discuss the importance of Rf2 for sugar beet breeding. Electronic supplementary material The online version of this article (doi:10.1007/s00122-014-2398-4) contains supplementary material, which is available to authorized users.

conditioned by the M locus on chromosome IV (Hogaboam 1957;Roundy and Theurer 1974; 94 Schondelmaier and Jung 1997). Hjerdin-Panagopoulos et al. (2002) detected Z as two linked 95 quantitative trait loci (QTLs) for fertility restoration that were located on chromosome IV. The precise 96 map position of Z is unknown.

102
However, none of the selected rf1rf1 plants from other populations had the maintainer genotype 103 (Moritani et al. 2013). Therefore, an as yet unidentified Rf reduced the maintainer-genotype frequency 104 in the rf1rf1-selection.

105
It was not known whether this unidentified Rf was an allele of Z, whose impact on 106 maintainer-genotype selection has not been elucidated. If this unidentified Rf is an allele of Z, it will 107 become very clear that Z is not a minor Rf but an important locus for sugar beet breeding. To this end, 108 we focused our analysis on the sugar beet line that is the most recalcitrant against rf1rf1-based MAS,

140
EcoRI and MseI were selected as the restriction endonucleases. Adapter-ligated DNA was 141 pre-amplified using Takara Ex Taq (Takara Bio, Ohtsu, Japan) using pairs of primers in which one of 142 the four nucleotides at the 3' terminus as a selective nucleotide. The PCR protocol was 20 cycles of 143 94°C for 30 sec, 56°C for 1 min and 72°C for 1 min. For selective amplification, pairs of primers 144 having three selective nucleotides were used. The PCR protocol was 94°C for 5 min; 13 cycles of 145 94°C for 30 sec, 65°C (annealing temperature was decreased by 0.6°C /cycle) for 30 sec and 72°C for 146 1 min; and 13 cycles of 94°C for 30 sec, 56°C for 30 sec and 72°C for 1 min. The amplified products 147 were electrophoresed in the high efficiency genome scanning system (Kawasaki and Murakami 2000; (MoBio Laboratories, Carlsbad, CA, USA). Purified PCR products were cloned into the pBluescript 168 (SK+) vector using Ligation high ver. 2 (Toyobo, Osaka, Japan) and sequenced on an ABI3130 169 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) using a BigDye Terminator v3.  (Table S3), and we analyzed these data using the mapping software. As a result, we obtained a map of 281 34.5 cM (Fig. 2b). The arrangement of markers is fairly well preserved between the BC 1 F 1 and the 282 BC 1 F 3 (Fig. 2).

283
We conducted QTL analysis for fertility restoration in the BC 1 F 3 to map Rf2.

290
The closest markers to Rf2 appeared to be sc4, sc3, sc7, sc10 and ca2, because the presence 291 or absence of these five markers showed the best association with male-fertility restoration (107/114 292 in the BC 1 F 1 and 132/146 in the BC 1 F 3 ). Therefore, Rf2 was located in the interval between sc4 and 293 ca2 (one of four markers, see Fig. 2a

329
Rf2 is situated in the region delimited by nir and ant on chromosome IV, on which Z is the 330 only Rf known to be located (Schondelmaier and Jung 1997). Thus Rf2 is likely an allele of Z. In 331 addition, the low restoration effect of Rf2 is consistent with the postulation of Z provided by many 332 sugar beet researchers (Owen, 1945;Hogaboam, 1957;Theurer, 1971).

333
In this context, we think Rf2 may correspond to one of the two-linked QTLs for fertility        Table S3 Segregation of marker genotypes in BC 1 F 3 6 These markers were assumed to be configured in repulsion in the linkage analysis.