Plant Breeding 116, 390—392 (1997) © 1997 Blackwell Wissenschafts-Verlag, Berlin ISSN 0179-9541 Short Communication PCR primers for the estimation of contamination by seeds with normal cytoplasm in rapeseed lots hearing male-sterility-inducing Ogu-INRA cytoplasm C. TiNCHANT*, M.-C. DEFRANCE' and F. 'Station de Genetique et d'Amelioration des Plantes, INRA, Centre de Versailles, Route de Saint-Cyr, F-78026 Versailles Cedex, France; ^Corresponding author With 2 figures Received November 20, 1996/Accepted March 29, 1997 Communicated by B. Schweisguth Abstract A polymerase chain reaction (PCR)-based test was developed to detect seeds bearing the 'so-called' normal rapeseed cytoplasm in seed lots with an OGU-INRA type cytoplasm. The test is based on the ampli- fication of the orfB region of male fertile rapeseed mitochondrial DNA (mtDNA). The amplification reaction uses total nucleic acids of young seedlings, extracted in bulk. After the sequencing of the orfB gene region in the normal Brassica mtDNA, primers were designed for its amplification by PCR. Although the specificity of amplification for the male fertile (mf) rapeseed cytoplasm is partly impaired by the presence of tiny amounts of this fragment in the mtDNA of male sterile (ms) plants, this test proved to be applicable for the estimation of the level of contamination in seed lots in reconstituted mixes as well as in real lots. Key words: Brassica napus — male-sterility — mitochondrial T-^xT A Ogu-INRA cytoplasm — orfB — PCR DNA The expanding use of cytoplasmic male sterility for the pro- duction of hybrid Brassica seeds has confronted breeders with the problem of seed lot impurities. The presence of 1 % male fertile (mf) plants within the female seed parent in a hybrid seed production field can reduce the level of hybridity of the harvested seeds below the level acceptable for commer- cialization. These contaminating mf plants can result from vol- unteer seeds left from a preceding culture, but can also result from pollution of the female seed lots during harvest by seeds from the maintainer line one year earlier. This latter event seems to occur quite often in field-scale seed productions. An early estimation of the level of impurity in female seed lots would be very valuable to avoid sowing contaminated lots and/or predicting whether the removal of male-fertile plants in the field at the flowering stage will be necessary. Whatever the cultivar, the distinction between female ms plants and their male fertile maintainers resides in their mito- chondrial (mt) genomes. The Ogu-INRA male sterility system is at the moment the most widely-used cytoplasmic male-sterile system for rapeseed hybrid seed production in Europe and North America. The mitochondrial genome of this male-ster- ihty-inducing cytoplasm has been engineered by recombination through protoplast fusion between Ogura radish (Ogura 1968) and normal rapeseed mitochondrial genomes (Pelletier et al. 1983). It harbours the sterility-inducing gene orfl38 in a frag- ment, the Nco2.5 fragment, originating from Ogura mtDNA (Bonhomme et al. 1991, 1992). The mt genome of the female Ogu-INRA plant is a patchwork of regions from both parental genomes (i.e. Ogura radish and normal rapeseed mt genomes), with a certain amount of redundancy in genetic information (Vedel et al. 1986, 1987). Moreover, the primary sequences of the Ogura radish and normal rapeseed mt genomes are very similar, although extensively different in organization owing to the pecuhar mode of evolution of plant mt genomes (Palmer and Herbon 1986, Nugent and Palmer 1988, Palmer 1990). However, the orfB gene, which is assumed to be essential for plant mt genomes (Gualberto et al. 1991), is present only in one copy in the mt genome of Ogu-INRA ms plants, as detected in mtDNA Southern experiments (Bonhomme et al. 1991). In these plants, the orfB gene is downstream and cotranscribed with the male-sterility gene orfl38 (Bonhomme et al. 1992) (see Fig. 1). It was decided to take advantage of sequences upstream of the orfB gene in male-fertile rapeseed mtDNA (which are different from those in male-sterile plants) in order to develop a PCR test specific for fertile Brassica mtDNA. Cloning of the orfB region from fertile Brassica mtDNA: A cosmid library (pWEl 5) of male-fertile Brassica mtDNA (504 clones with mean insert size of 30-35 kb) was screened with an or/B-specific probe, and 10 positive clones were isolated. After a rapid restriction analysis of these clones, one was chosen for the subcloning of the 2kb Xbal-BamHl fragment {XB2) containing the orfB gene and its upstream region. This fragment has been entirely sequenced using exonuclease Ill- generated ordered deletion clones (Hoheisel and Pohl 1986) and an Applied Biosystems 373A (Perkin Elmer, Foster City, CA) automated DNA sequencer. The organization of the orfB region from the male- fertile Brassica mt genome and comparison to the Ogura Nco2.5 region are shown in Fig. 1. Development of the test: The sequence was analysed (Wisconsin pack- age, version 8, Genetics Computer Group, Madison, WI) for homo- logies with known sequences in the databases and then Ohgo 4.0 soft- ware (National Biosciences Inc., Plymouth, MN) was used to design specific primers for PCR, with the following requirements: (1) the lower primer had to be anchored in the orfB coding sequence in order to avoid any interference with a possible repeat of the upstream non- coding region in the mt genome of either fertile or sterile plants; (2) the upper primer had to be in the sequence specific for the XB2 fragment compared with the Nco2.5 fragment; (3) the upper primer should not lie in the regions showing homologies with 5S and 16S rRNA and atpA genes and located in the first 350 base pairs after the Xbal site. As a U. S. Copyright Clearance Center Code Statement: 01 79-9541 /97/1 604-^390 $ 1 4.00/0