Double low restorer lines of winter rapeseed for CMS ogura system

Tom XXIV Rośliny Oleiste 2003 Iwona Bartkowiak-Broda, Wiesława Popławska, Anna Fürguth, Katarzyna Mikołajczyk Plant Breeding and Acclimatization Institute, Department of Oilseed Crops in Poznań Instytut Hodowli i Aklimatyzacji Roślin, Zakład Roślin Oleistych w Poznaniu Double low restorer lines of winter rapeseed for CMS ogura system Podwójnie ulepszone linie restorery rzepaku ozimego dla systemu CMS ogura Key words: Słowa kluczowe: winter oilseed rape, Brassica napus L., CMS ogura, restorer gene, glucosinolates, molecular markers rzepak ozimy, Brassica napus L., CMS ogura, gen restorer, glukozynolany, markery molekularne Breeding of oilseed rape hybrid varieties in Poland is based on CMS ogura hybridization system. In the beginning, the major problem was the breeding of double low restorer lines containing glucosinolates below 15 µm/g of seeds. At present double low restorer lines have been bred and used to produce double low restored F 1 hybrids. Selection of restorer lines with desirable quality traits may be more effective when it is assisted by molecular markers. However, isoenzymatic codominant PGI-2 marker linked with restorer gene is not effective in case of selection of genotypes with alleles of restorer gene and with low glucosinolates content. In some restorer lines, the introgression of the Rfo gene from radish is modified and the allele Pgi-2 is lost. Therefore investigations on the possibility of the use of molecular markers of RAPD type have been conducted. The obtained results show that in genomes of restorer lines with extremely low glucosinolate content whose linkage between Pgi-2 allel and restorer gene has been broken PCR RAPD molecular marker OPC02 1150 is not lost with the fragment of DNA responsible for high glucosinolate content. Full linkage of this marker with restorer gene has been observed Hodowla odmian mieszańcowych rzepaku w Polsce prowadzona jest w oparciu o system CMS ogura. Początkowo głównym problemem tego systemu była hodowla linii restorerów o zawartości glukozynolanów poniżej 15 µm/g nasion. Obecnie wyselekcjonowane podwójnie ulepszone linie restorera wykorzystuje się do otrzymywania podwójnie ulepszonych mieszańców pokolenia F 1. Selekcja linii restorerów o pożądanych cechach jakościowych może być bardziej efektywna gdy jest wspomagana przez markery molekularne. Stosowany dotąd izoenzymatyczny kodominujący marker genu restorera PGI-2 jest nieefektywny w przypadku selekcji linii restorera o niskiej zawartości glukozynolanów. Niektóre linie restorery w wyniku modyfikacji zachodzących w obszarze introgresji genu Rfo pochodzącego z genomu rzodkwi utraciły allel Pgi-2. W związku z tym przeprowadzono badania nad możliwością korzystania w tym celu z markerów molekularnych typu RAPD. Uzyskane wyniki wskazują, że genomy linii restorerów o ekstremalnie niskiej zawartości glukozynolanów, które utraciły marker PGI-2, zachowały marker molekularny OPC02 1150 typu PCR RAPD mimo utraty fragmentu DNA determinującego wysoką zawartość glukozyno-

344 Iwona Bartkowiak-Broda... also in genotypes with extremely low glucosinolate content. Double low restorer lines without DNA fragment containing Pgi-2 alleles from radish genotype were selected as new recombinants for further investigations. lanów. Jak dotychczas stwierdzono pełne sprzężenie tego markera z genem restorerem także w genotypach o ekstremalnie niskiej zawartości glukozynolanów. Niskoglukozynolanowe linie restorery bez alleli Pgi-2, pochodzących z genotypów rzepaku zostały wyselekcjonowane do dalszych badań jako nowe rekombinanty,. Introduction Oilseed rape (Brassica napus L.) is known as an important source of edible oil and fodder protein as well as a good agronomic head of rotation. In Europe, especially in countries with moderate climate conditions, such as Poland, it is the most important oilseed crop. Considering the international competition oilseed rape has to increase its competitiveness (especially with soyabean) in respect to seed yield as well as seed quality. The level of seed yield may be increased by the use of heterosis effect occurring in rapeseed hybrids. Considerable heterosis effect for seed yield in F 1 hybrids of oilseed rape has been reported by various authors (Schuster 1969; Grabiec, Krzymanski 1984; Grant, Beversdorf 1985; Lefort-Buson, Dattée 1983, 1985a, 1985b; Krzymanski et al. 1993, 1994;). The results of investigations on heterosis effect encouraged the development of genetic pollination control systems. Rapeseed is the allogamouse plant, so the natural system of its propagation is composed of self- and cross fertilization in different proportions, depending on variety genotype. At present several hybridization systems using male sterility or self incompatibility are in competition in the world. In Europe first hybrid varieties were registered in 1994 in France: composite hybrid of winter rapeseed Synergy created using CMS ogura and spring variety Hybridol restored hybrid, developed using CMS polima system. These varieties were followed by a restored hybrid of winter rapeseed Pronto and Joker registered in 1996 in Germany and created using MSL NPZ Lembke system, and the first CMS ogura restored hybrid Elite registered in Denmark. At present the breeding of hybrid varieties has been developing intensively. In Europe this type of varieties covers about 40% of rapeseed acreage. In European countries majority of rapeseed breeding programmes use two hybridization systems: cytoplasmic male sterility CMS ogura elaborated in INRA in France or genic male sterility MSL (Male Sterility Lembke), private system owned by NPZ Lembke. Polish rapeseed breeding programme is based on CMS ogura system. The development of hybrid varieties with the use of CMS ogura has been one of the most important objectives in recent years for Polish breeders.

Double low restorer lines of winter rapeseed for CMS ogura system 345 CMS ogura is characterized by stable expression of male sterility in different genetic backgrounds and under different environmental conditions. The lack of restorer lines with appropriate qualitative and agronomical traits was the factor limiting the utilization of CMS ogura system in breeding of restored hybrid varieties of double low winter oilseed rape. Because of this problem the first Polish hybrid varieties registered in Official Catalogue and cultivated are composite hybrids Mazur, Kaszub, Lubusz and Pomorzanin. These varieties are composed of 70% seeds of F 1 non-restored hybrid and 30% of seeds of varieties Kana and Marita which are the source of pollen for hybrid plants. The best solution, however, seems to be the development of efficient programme of restored hybrid breeding. Restorer gene Rfo has been introduced to rapeseed genome from radish genonome (Raphanus sativus L.) (Heyn 1976) through intergeneric hybridization. The obtained recombinants retained too much genetic information from radish genotype and therefore they posses some undesirable traits: low seed set and close linkage of restorer alleles with the genes determining high glucosinolate content (Pellan-Delourme 1986). Investigations conducted by Delourme et al. (1995) revealed that the improvement of these traits and elimination of radish genetic information are possible by backcrosses with double low lines. It was confirmed also in the investigations of Poplawska (2000) and Bartkowiak-Broda, Poplawska (2001). These investigations and breeding resulted in restorer lines with low glucosinolate content reaching Polish standard for double low varieties (less than 15 µm/g of seeds), with stabilized meiotic behaviour (Poplawska 2000) and improved yielding ability. However, because of the origin of restorer gene breeding of restoring lines with stable expression of restoring ability and with desirable quality traits and also good yielding ability is time consuming and requires selection made in large population. Speeding-up this selection requires the inclusion of molecular markers to breeding programmes. Study concerning the restorer gene markers revealed that this gene is tightly linked with isozyme marker phosphoglucoisomerase PGI-2 (Delourme and Eber 1992) and four RAPD markers (Delourme et al. 1994). In the beginning selection of recombinants with low glucosinolate content and restorer gene alleles was carried out on the phenotypic expression and with the use of isozyme marker PGI-2. However, the selection of genotypes with homozygous alleles of the restorer gene and low glucosinolate content using molecular markers is still a problem. In some restorer lines, the introgression of the Rfo gene from radish is modified and the allele Pgi-2 is lost (Delourme et al. 1999). Especially in populations of restorer lines with extremely low glucosinolate content, the selection by the use of isozyme PGI-2 marker is not effective. The undertaken investigations aimed at determining whether the marker PCR RAPD OPC02 which reveals high linkage with restorer gene in the population of restorer lines studied by Delourme et al. (1994) will exhibit the same linkage with

346 Iwona Bartkowiak-Broda... restore gene in the population of restorer lines characterized by low glucosinolate content and possessing genetic background of double low varieties originating from Polish breeding. Besides, new recombinants of double low restorer lines were expected. Materials and methods Investigations were carried out on the population of 102 restorer lines. The lines were selected from F 6 generation originating from crossings between low glucosinolate male sterile lines CMS ogura (4.1 11.8 µm/g of seeds) and starting restorer line R with glucosinolates content of about 60 µm/g of seeds. Selection of genotypes with restorer gene alleles was carried out on the phenotypic expression of this trait and with the use of isozyme marker PGI-2 (Delourme and Eber 1992). Analyses were carried out according to the method described by Shilds et al. (1983) and Vallejos (1983). Restorer lines are bred on sterile cytoplasm ogura type. It allows to verify the purity of selected restorer lines using specific mtdna marker of sterile cytoplasm ogura type according to the method of Krishnasamy and Makaroff (1993) modified by Mikolajczyk et al. (1998). Extraction of DNA from young leaves for PCR RAPD analysis was performed using modified method of Doyle and Doyle (1990). The presence of molecular marker of restorer gene RAPD OPC02 1150 (Delourme et al. 1994, 1998) was investigated in lines characterized by low glucosinolate content with PGI-2 marker and with lost PGI-2 marker. The analyses of glucosinolates were performed with the method of gas chromatography of silyl derivatives of desulfoglucosinolates (Michalski et al. 1995). Results Selected plants in F 6 generation were characterized by low glucosinolate content in range 1.70 19.30 µm/g of seeds (Table 1). About 87% of 102 restorer lines selected from F 6 progeny were characterized by glucosinolate content below 15 µm/g of seeds (Polish standard for sowing material). After examination with the use of PGI-2 marker it was observed that 59.9% of restorer lines characterized by glucosinolate content in range 3.80 18.80 µm/g of seeds lost radish Pgi-2 alleles (Fig. 1), the linkage between alleles of restorer gene and Pgi-2 allele originating from radish has been broken. The remaining 40.1% of investigated lines maintaining PGI-2 marker originating from radish were characterized by glucosinolate content in range 1.70 19.30 µm/g of seeds (Fig. 1). However, within the class of plants with very low glucosinolate content below 10 µm/g of seeds the majority of plants lost PGI-2 marker from their genotype.

Double low restorer lines of winter rapeseed for CMS ogura system 347 Table 1 Glucosinolate content in restorer lines of F 6 progeny with PGI-2 marker and with lost PGI-2 marker Zawartość glukozynolanów w pokoleniu F 6 linii restorerów z markerem i bez markera PGI-2 Statistic parameters Parametry statystyczne Sum of glucosinolates in restorer lines of F 6 progeny Suma glukozynolanów w liniach restorerach pokolenia F 6 [µm/g of seeds nasion] plants with marker PGI-2 rośliny z markerem PGI-2 41 (40.1%) plants with lost marker PGI-2 rośliny bez markera PGI-2 61 (59.9%) Mean Średnia 12.21 9.86 Standard error of the mean Błąd standardowy średniej 0.65 0.49 Range Zakres 17.60 15.00 Maximum 19.30 18.80 Minimum 1.7 3.80 Standard deviation Odchylenie standardowe 4.19 3.82 Kurtosis Kurtoza 0.027 0.855 Skewness Skośność 0.57 0.30 Coefficient of variability Współczynnik zmienności 34.33 38.76 Number of plants Liczba roślin 30 25 20 15 10 5 0 1,7 5,0 5,1 10,0 10,1 15,0 15,1 19,3 Glucosinolate content [µm/g of seeds] Zawartość glukozynolanów [µm/g nasion] Restorer lines with lost marker PGI-2 Linie restorery bez markera PGI-2 Restorer lines with lost marker PGI-2 Linie restorery z markerem PGI-2 Fig. 1. Glucosinolate content in restorer lines of F 6 progeny with PGI-2 marker and with lost PGI-2 marker Zawartość glukozynolanów w pokoleniu F 6 linii restorerów z markerem i bez markera PGI-2

348 Iwona Bartkowiak-Broda... Distribution of data for glucosinolate content in seeds of plants with PGI-2 is characterized by asymmetry of distribution in the direction of low glucosinolate content with high concentration of genotypes with glucosinolate content in the range 10 to 15 µm/g of seeds. Distribution of data for glucosinolate content in seeds of plants with lost PGI-2 marker is characterized by smaller asymmetry but in the direction of higher glucosinolate content with almost normal distribution, but with the concentration of genotypes containing glucosinolates in range 5 15 µm/g of seeds. RAPD OPC02 1150 marker carried by radish introgression was tested on all restorer lines. The presence of OPC02 1150 marker was observed in all low glucosinolate genotypes, independently of the presence or absence of Pgi-2 allele of radish (Fig. 2). bp M R+ R- 1 2 3 4 5 6 7 8 9 10 11 12 13 M 2027 1904 1584 1375 bp 1150 OPC02 947 831 564 R+ model plant with restorer gene roślina wzorcowa z genem restorerem R- model plant without restorer gene roślina wzorcowa bez genu restorera 1 13 plants tested for the presence of restorer gene marker (arrow indicate polymorphic band) rośliny testowane na obecność genu restorera (strzałka pokazuje polimorficzny prążek) 1 7 plants with PGI-2 marker rośliny z markerem PGI 2 8 13 plants without PGI-2 marker rośliny pozbawione markera PGI-2 M molecular size marker (in base pairs) standard wielkości (w parach zasad) Fig. 2. 1.8% agarose gel electrophoresis of PCR RAPD products obtained with the use of primer OPC02 Elektroforetyczny rozdział produktów reakcji PCR na 1.8% żelu agarozowym z zastosowaniem startera typu RAPD OPC02

Double low restorer lines of winter rapeseed for CMS ogura system 349 Discussion On the basis of the conducted study it can be stated that after introduction of double low winter rapeseed genome to genome of restorer lines there may emerge recombinants without DNA fragments determining high glucosinolate content: with and without PGI-2 marker. The break of linkage between the restorer gene and radish isozyme PGI-2 marker occurred in the majority of investigated restorer lines. This result evidences the possibility of further modifications within the introgression of radish genome. The modifications would concern losses of radish Pgi-2 alleles together with closely linked DNA fragment, which is responsible for high glucosinolate content. However, the obtained results show that it is not necessary to eliminate radish Pgi-2 alleles to get low glucosinolate restorer lines. Restorer lines with extremely low glucosinolate content but with PGI-2 marker (Fig. 1) were obtained. Also Delourme et al. (1999) in their characteristic of radish genome introgression for various restorer lines observed the same phenomenon. Low glucosinolate genotypes with alleles of restorer gene and without PGI-2 marker have already appeared in the former generations, which we investigated, but with considerably lower frequency. Restorer lines with high and medium glucosinolate content have been selected in F 2 progeny of hybrid between low glucosinolate CMS ogura line and initial high glucosinolate restorer line and the linkage of this marker with restorer gene was very high and appeared in average in about 98.8% individuals (Poplawska 2000; Bartkowiak-Broda, Popławska 2001). In F4 progeny, frequency of linkage was smaller and equal 92.2% of examined population (Bartkowiak-Broda, Poplawska 2001). Linkage occurred only in 40.1% of examined genotypes in the presented investigations. Selection pressure is directed towards lowering of glucosinolates content. The lowering of glucosinolate content was accompanied more often with the elimination of Pgi-2 alleles originating from radish. It indicates that the use of PGI-2 marker in selection for low glucosinolate restorer lines in very low glucosinolate populations is not effective. At the same time it indicated that a part of unnecessary genetic information from radish has been eliminated and such genotypes without Pgi-2 alleles may appear valuable for further restorer lines breeding. However, further research on restorer lines requires effective molecular marker which allows to do selection of plants at every developmental stage. The obtained result turns out to be very positive because RAPD marker OPC02 1150 described by Delourme et al. (1999) occurs also in low glucosinolate restorer lines of winter rapeseed bred in Poland. RAPD marker OPC02 1150 is closely linked to the restorer gene and was not lost in extremely low glucosinolate recombinants together with the fragment of DNA responsible for high glucosinolate content. This marker can be used effectively in breeding programmes for low glucosinolate restorer lines for CMS ogura system.

350 Iwona Bartkowiak-Broda... Conclusions The break of linkage between the restorer gene and radish isozyme marker PGI-2 occurred in the majority of investigated low glucosinolate restorer lines. Because of that the PGI-2 marker is not useful for selection of restorer lines especially in very low glucosinolate populations. The tight linkage between molecular marker PCR RAPD OPC02 1150 and the restorer gene existing also in population of low glucosinolate restorer lines with genotype of winter rapeseed originating from Polish breeding was confirmed. The high use of marker OPC02 1150 in selection of recombinants with restorer gene and extremely low glucosinolate content which lost radish Pgi-2 alleles together with the fragment of DNA responsible for high glucosinolate content was observed. Low glucosinolate restorer lines without DNA fragments from radish genotype containing Pgi-2 alleles were selected. Wnioski W większości badanych niskoglukozynolanowych linii restorerów nastąpiło przerwanie sprzężenia pomiędzy genem restorerem i markerem PGI-2 pochodzącym od rzodkwi, dlatego marker ten nie jest przydatny w selekcji linii restorerów w populacjach roślin o bardzo niskiej zawartości glukozynolanów. Potwierdzono silne sprzężenie markera molekularnego PCR RAPD OPC02 1150 i genu restorera także w populacjach linii restorerów rzepaku ozimego pochodzących z polskiej hodowli. Stwierdzono dużą przydatność markera OPC02 1150 w selekcji rekombinantów o ekstremalnie niskiej zawartości glukozynolanów, podwojonych alleli Pgi-2 wraz z fragmentem DNA determinującym zawartość glukozynolanów. Wyselekcjonowano linie o bardzo niskiej zwartości glukozynolanów pozbawionych, pochodzącego od genotypu rzodkwi, fragmentu DNA zawierającego allele Pgi-2. References Bartkowiak-Broda I., Popławska W. 1999. Characteristics of double low winter rapeseed lines with introduced restorer gene for CMS ogura. Proc. 10th Intern. Rapeseed Congress, Canberra, Australia, CD ROM.

Double low restorer lines of winter rapeseed for CMS ogura system 351 Bartkowiak-Broda I., Popławska W. 2001. Introgression of Brassica napus genome into restorer lines for CMS-ogura estimated with the use of molecular markers. Biotechnologia, 1 (52): 112-118. Delourme R., Bouchereau A., Hubert N., Renard M., Landry B.S. 1994. Identification of RAPD markers linked to a fertility restorer gene for the ogura radish cytoplasmic male sterility of rapeseed (Brassica napus L.). Theor. Appl. Genet., 88: 741-748. Delourme R., Eber F. 1992. Linkage between an isozyme marker and a restorer gene in radish cytoplasmic male sterility of rapeseed (Brassica napus L.). Theor. Appl. Genet., 85: 222-228. Delourme R., Eber F., Renard M. 1995. Breeding double low restorer lines in radish cytoplasmic male sterility of rapeseed (Brassica napus L.). Proc. 9th Intern. Rapeseed Congress, Cambridge, UK, 4-7.07.1995, vol. 1, 6-8. Delourme R., Foisset N., Horvais R., Barret P., Champagne G., Cheung W.Y., Landry B.S., Renard M. 1998. Characterisation of the radish introgression carrying the Rfo restorer gene for the Ogu INRA cytoplasmic male sterility in rapeseed (Brassica napus L.). Theor. Appl. Genet., 97/1-2: 129-134. Delourme R., Horvais R., Vallé P., Renard M. 1999. Double low restored F 1 hybrids can be produced with the Ogu INRA CMS in rapeseed. Proc. 10th Intern. Rapeseed Congress, Canberra, Australia, 26 29.09.1999, CD ROM. Doyle J.J., Doyle J.L. 1990. Isolation of plant DNA from fresh tissue. Focus 12: 13-15. Heyn F.W. 1976. Transfer of restorer genes from Raphanus to cytoplasmic male sterile Brassica napus. Cruciferae Newsletter, Eucarpia, 1: 15-16. Grabiec B., Krzymański J. 1984. Badania nad wykorzystaniem zjawiska heterozji w hodowli rzepaku ozimego w Polsce. Wyniki badań nad rzepakiem ozimym: 65-73 Grant I., Beversdorf W.D. 1985. Heterosis and combining ability estimates in spring planted oilseed rape (Brassica napus L.). Can. J. Genet. Cytol., 27: 472-478. Krishnasamy S., Makaroff C. 1993. Characterization of the radish mitochondrial orfb locus: possible relationship with male sterility in Ogura radish. Curr. Genet., 24: 156-163. Krzymański J., Piętka T., Krótka K. 1993. Zdolność kombinacyjna i heterozja mieszańców diallelicznych rzepaku ozimego podwójnie ulepszonego. I. Pokolenie F 1. Postępy Nauk Rolniczych, 5: 41-52. Krzymański J., Piętka T., Krótka K. 1994. Zdolność kombinacyjna i heterozja mieszańców diallelicznych rzepaku ozimego podwójnie ulepszonego. II. Pokolenie F 2. Rośliny Oleiste Oilseed Crops, XV (1): 21-32. Lefort-Buson M., Dattée Y. 1983. L hétérosis chez le colza oleagineux (Brassica napus L.). Proc. 6th Int. Rapeseed Conference, 17-19 May, Paris, France, Vol. 1: 558-564. Lefort-Buson M., Dattée Y. 1985a. Etude de l hétérosis chez le colza oleagineux d hiver (Brassica napus L.) I. Comparaison de deux populations, l une homozygote et l autre hétérozygote. Agronomie, 5/2: 101-110. Lefort-Buson M., Dattée Y. 1985b. Etude de l hétérosis chez le colza oleagineux d hiver (Brassica napus L.) I. Structure génétique d une population de lignées. Agronomie, 5/3: 201-208. Michalski K., Kołodziej K., Krzymański J. 1995. Quantitative analysis of glucosinolates in seeds of oilseed rape. Effect of sample preparation on analytical results. Proc. 9th Intern. Rapeseed Congress, Cambridge, UK, 4-7.07.1995, vol. 1: 6-8. Mikołajczyk K., Matuszczak M., Piętka T., Bartkowiak-Broda I., Krzymański J. 1998. Zastosowanie markerów DNA do badań składników mieszańców. Rośliny Oleiste Oilseed Crops, XIX (2): 463-471.

352 Iwona Bartkowiak-Broda... Pellan-Delourme R. 1986. Etude de deux systémés de stérilité male génocytoplasmique introduits chez le colza (Brassica napus L.) par croisements intergeneriques avec Raphanus and Diplotaxis. These INRA-Rennes: 88 Pellan-Delourme R., Renard M. 1988. Cytoplasmic male sterility in rapeseed (Brassica napus L.): female fertility of restored rapeseed with ogura in cybrids cytoplasms. Genome, 30: 34-238. Popławska W. 2000. Badania nad formami restorującymi genowo-cytoplazmatyczną męską niepłodność typu Polima i Ogura u rzepaku ozimego (Brassica napus L. var. oleifera). Praca doktorska, Instytut Hodowli i Aklimatyzacji Roślin, ZRO Poznań. Schuster W., Michael J. 1976. Untersuchungen über Inzuchtdepressionen und Heterosiseffekte bei Raps (Brassica napus oleifera). Z. Planzenzüchtung 77: 56-66. Shields C.R., Orton C.J., Stuber C.W. 1983. An outline of general resource needs and procedures for the electrophoretic separation of active enzymes from plant tissues. In: Tanksley S.D. and Orton T.J. (Eds), Isozymes in plants genetics and breeding, Part A, Elsevier Sciences Publishers, B.V., Amsterdam: 443-458. Vallejos C.E. 1983. Enzyme activity staining. In: Tanksley S.D. and Orton T.J. (Eds), Isozymes in plants genetics and breeding, Part A, Elsevier Sciences Publishers, B.V., Amsterdam: 469-516.