DNA marker assisted breeding in interspecific crosses to improve canola (Brassica napus L.)

Transcription

1 DNA marker assisted breeding in interspecific crosses to improve canola (Brassica napus L.) Christopher James Schelfhout BSc.Agric.(Hons) This thesis is presented for the degree of Doctor of Philosophy of The University of Western Australia Faculty of Natural and Agricultural Sciences School of Plant Biology 2007

2 Statement of candidate contribution I declare that this thesis is my own composition and the result of my own research work. All the contributions made by other individuals have been duly acknowledged. This thesis contains no material which has been accepted for the award of any other degree in this university or another institution and to the best of my knowledge and belief, contains no material previously published or written by another person, except where due reference has been made in the text. Several papers have been published from this thesis (page iv). Chris Schelfhout contributed 90% of the work for all experimental research, preliminary drafts and literature research. The co-authors contributed 10% for research guidance, editing and funding. All co-authors agree to the use of these published works as chapters in this thesis.. Christopher James Schelfhout... Associate Professor Wallace Cowling. Dr. Janet Wroth i

3 Abstract In order to expand the gene pool of canola-quality rapeseed (Brassica napus) reciprocal interspecific crosses were made between B. napus cv. Mystic and near canola-quality B. juncea breeding line JN29. F 1 progeny from these crosses were used to make backcrosses to both parents in all possible combinations and directions, and were selfed to form F 2 -derived lines. The highest frequencies of viable F 2 and BC 1 progeny were obtained when B. napus was the maternal parent of the interspecific hybrid. BC 1 and F 2 progeny (and subsequent generations) were grown under field conditions to identify agronomic improvements over the parents. Transgressive segregation was observed in F 2 and BC 1 and in subsequent generations for agronomic traits (seed yield under high or low rainfall conditions, plant biomass, harvest index, height, branching and days to anthesis) and seed quality traits (oil, protein, glucosinolates, oleic acid). The majority of progeny conformed to B. napus morphology, and a minority segregated to B. juncea morphology in subsequent generations. Some of the B. juncea morphotypes had lower glucosinolates and higher oleic acid than the parent JN29, with no detectable erucic acid, and thereby conformed to canola quality. Methods were developed for tracing B-genome in interspecific progeny. A repetitive DNA sequence pbnbh35 from B. nigra (genome BB, 2n = 16) was used to identify B-genome chromosomes and introgressions in interspecific progeny. Specific primers were designed for pbnbh35 in order to amplify the repetitive sequence by PCR. A cloned sub-fragment of 329 bp was confirmed by sequencing as part of pbnbh35. PCR and hybridisation techniques were used on an array of Brassica species to confirm that the pbnbh35 subfragment was Brassica B-genome specific. Fluorescence in situ hybridisation (FISH) in B nigra, B. juncea (AABB, 2n=36) and B. napus (AACC, 2n=38) showed that the pbnbh35 sub-fragment was present on all eight Brassica B- genome chromosomes and absent from A- and C-genome chromosomes. The pbnbh35 repeat was localised to the centromeric region of each B-genome chromosome. FISH clearly distinguished the B-genome chromosomes from the A-genome chromosomes in the amphidiploid species B. juncea. This is the first known report of a B-genome repetitive marker that is present on all Brassica B- genome chromosomes. ii

4 The pbnbh35 sub-fragment was used as a PCR-based marker and FISH probe to trace and identify Brassica B-genome chromosomes and chromosomal introgressions in the interspecific progeny B. napus x B. juncea. B-genome positive PCR products were identified in 67% of fertile F 2 progeny, which was more than double the proportion found in fertile BC 1 progeny. The majority of these progeny were indistinguishable from B. napus in morphology. Four B-genome positive F 2 -derived families and one BC 1 -derived family were fixed or segregating for B. juncea morphology. Based on FISH assays, the B. juncea morphotypes could be grouped into two types in the F 4 and BC 1 S 2 generation: (i) those with a B. juncea complement of 36 chromosomes, 16 of which gave B-genome positive signals, and (ii) those with a B. napus complement of 38 chromosomes and no (or few) B-genome signals. One of these B. juncea morphotypes with 38 chromosomes exhibited isolated and weak B-genome FISH signals on 11 chromosomes and typical A-genome FISH signals. FISH signals were not detected in several B-genome PCR positive progeny. The results suggest that novel B. napus genotypes have been generated containing introgressions of B-genome chromatin from B. juncea chromosomes. B. juncea morphology occurred in interspecific progeny with a chromosome complement similar to B. napus (2n = 38) and without the entire B- genome present. It also is highly likely that recombination has occurred between the A-genome of the two Brassica species. This research has demonstrated that the secondary gene pool of B. napus may be accessed by selfing interspecific hybrids, and without sacrificing canola quality, if the B. juncea parent is near canola-quality. Interspecific progeny may be screened to enhance the proportion with B-genome positive signals. Some progeny with B. juncea type morphology had improved seed quality over the JN29 parent. iii

5 Acknowledgements This thesis was completed under the supervision of Associate Professor Wallace Cowling and Dr. Janet Wroth. I am very grateful and fortunate to have had such committed supervisors who lead by example and strive for excellence in the work they do. Their encouragement and enthusiasm has enabled me to achieve so much during the course of this study. Wallace provided a seemingly endless supply of innovative ideas and he has shown me that persistence and commitment is to be rewarded. Janet is a lateral thinker who can always be relied upon to give an alternative perspective on things. Thanks also to Janet for all the little pieces of maternal advice you gave along the way. I thankfully acknowledge the financial support of the Australian Research Council and industry partners Council of Grain Grower Organisations Ltd and Export Grains Centre Ltd, who funded my scholarship and operating expenses. There are many others who supported my research in various ways. These include Milton Sanders and Kylie Edwards who tirelessly helped with field trials and Jeremy English and Catherine Borger who also assisted with harvest. Thanks to Michael Blair for your assistance with trials at the Shenton Park field station. Thank you to Graham Walton, Tim Trent and staff at the Merredin Research Station for their assistance with the trials at Merredin. I would also like to thank Matthew Nelson, Anouska Cousin and Michael Francki for the numerous discussions and advice given during our shared time in the molecular genetics laboratory. Thanks also to Nick Larkin for your friendship and advice during our time shared in the office. Many thanks to Dr. Rod Snowdon and colleagues at Justus Liebig Universitat, Germany. Rod s generous donation of time, advice and resources proved to be a productive partnership. It was a very rewarding and memorable experience. I sincerely thank my parents for their support and assistance along the way. Their continued encouragement has been invaluable. I must thank my three very own F 1 progeny Louis, Mary and George who all arrived during the course of this study and kept life interesting. I apologise for the late nights and missed opportunities. Finally, to my beautiful wife Anni, thank you so much for your patience and support, it has been a long time coming but now I m finally at the end and I would like to dedicate this thesis to you. iv

6 Journal publications arising from this thesis Schelfhout CJ, Snowdon R, Cowling WA, Wroth JM (2004) A PCR based B- genome-specific marker in Brassica species. Theoretical and Applied Genetics 109, Schelfhout CJ, Snowdon R, Cowling WA, Wroth JM (2006) Tracing B-genome chromatin in Brassica napus B. juncea interspecific progeny. Genome 49, Schelfhout CJ, Wroth JM, Yan G, Cowling WA (2008) Enhancement of genetic diversity in canola-quality Brassica napus and B. juncea by interspecific hybridisation. Australian Journal of Agricultural Research (in press). Conference publications arising from this thesis Schelfhout, C., Cowling, W. and Wroth, J. (2002) Tracing the B genome in B. napus x B. juncea interspecific hybrids. Proceedings of the 12 th Australasian Plant Breeding Conference, Perth Western Australia. Cowling, W., Wroth, J. and Schelfhout, C. (2002). Interspecific crossing to widen the gene pool of canola. Proceedings of the 12th Australasian Plant Breeding Conference, Perth, Western Australia. Schelfhout, C., Cowling, W. and Wroth J. (2003) B-genome detection and influence in interspecific crosses of Brassica napus and B. juncea. Proceedings of the 11th International Rapeseed Congress. Copenhagen, Denmark. v

7 Table of contents Statement of candidate contribution.... i Abstract ii Acknowledgments iv Publications arising from this thesis v Table of contents......vi Chapter 1. General Introduction Chapter 2. Literature Review...6 Chapter 3. Published paper A PCR based B-genome specific marker in Brassica species Chapter 4. Published paper Tracing B-genome chromatin in B. napus B. juncea interspecific progeny Chapter 5. In press paper Enhancement of genetic diversity in canola-quality Brassica napus and B. juncea via reciprocal gene exchange Chapter 6. General Discussion References vi

8 Chapter 1 - General Introduction Rotation with canola (Brassica napus L.) has improved weed, insect and disease control in subsequent cereal crops in Australia, and has improved profitability in Australian farming systems. The area sown to canola increased markedly during the early 1990s (Norton et al. 1999). Expansion into rainfall zones as low as 280 mm per annum was made possible with the introduction of shorter season varieties (Salisbury and Wratten 1999) however these options do not always extend to growers the same relative gain as in higher rainfall cropping regions. In drier areas canola crops are usually only sown when opportunistically favourable conditions prevail early in the cropping season. Australian canola growers are also faced with threats from blackleg (Leptosphaeria maculans) disease as a result of peak ascospore releases in June, July and August coinciding with crop sowing and establishment (Bokor 1975, Salisbury et al. 1995, Khangura and Barbetti 2004). In 30 years of oilseed rape breeding in Australia since 1970, B. napus was converted into canola quality, based on Canadian sources of low erucic acid in oil and European sources of low glucosinolates in meal, and selected for adaptation to the Australian winter-spring growing environment with high levels of blackleg resistance (Salisbury and Wratten 1999). By default, the Australian breeding programs became one large closed recurrent selection program with an effective population size of 11 and a population inbreeding coefficient of 0.21 after 5 cycles of selection over 30 years (Cowling 2007). This narrowing of the gene pool in Australia over 30 years is of concern, and new sources of genetic variation are vital to the future of Australian canola breeding. The scope for genetic improvement of B. napus from within its primary gene pool is limited due to the narrowness of the gene pool. Becker et al. (1995) discuss probable causes for such a narrow gene pool, such as its relatively recent origins from a few natural interspecific hybridizations between B. rapa and B. oleracea, and selection for adaptation to winter and spring cropping environments in Europe, followed by further selection into canola quality germplasm. New genetic diversity in B. napus will aid the adaptation of this species to new environments, as suggested by Lewis and Thurling (1994) in Australia. 1

9 The secondary gene pool of B. napus, including all the diploid and amphidiploid species in U s triangle (U 1935), has long been considered as a source of new alleles for B. napus breeding. B. juncea (genome AABB) is grown widely in drought-prone areas and has been the subject of selection for drought tolerance (Chauhan et al. 2007). This species may provide alleles to improve adaptation to low-rainfall environments in B. napus (genome AACC). Previous crosses between B. napus (canola quality) and B. juncea (mustard quality) resulted in B. napus progeny with potentially useful blackleg resistance, but lower seed quality (Barret et al. 1998, Roy 1980a, 1980b, 1984, Sacristan and Gerdemann 1986), earliness (Rao et al. 1993a) or shatter resistance (Prakash and Chopra 1988). Potentially beneficial alleles from B. juncea or other donor species are mostly lost during backcrossing to restore canola quality, unless they are major alleles that are readily tracked by phenotype or with molecular markers. A potential solution is to use near canola-quality B. juncea as the donor of beneficial alleles. B. juncea shares a common genome (AA) with B. napus and may also contribute useful B-genome alleles through introgression with the A or C genomes of B. napus. This research aimed to determine if progeny of such crosses maintained their canola quality upon selfing, with possible reciprocal benefits to both species through a single interspecific cross. With this in mind, experiments were set up with reciprocal crossing and backcrossing in all possible combinations to examine the potential for introducing valuable alleles from B. juncea into B. napus and also, in the reverse direction, useful quality and agronomic traits into B. juncea from B. napus. The first backcross was used to test the value of increasing the proportion of alleles from the recurrent parent. The progeny and parental lines were observed in the field in the F 2 and BC 1 generation, and fertile selfed selections were tested in contrasting agricultural environments (low and high rainfall) in the F 3 and BC 1 S 1. Measurement of agronomic, disease and quality traits allowed a comparison of performance of various cross combinations (reciprocals and backcross versus self-pollination). Molecular and cytogenetic techniques were developed to assist in the detection of B-genome within B. napus-type progeny. An RFLP type B-genome marker has been reported (Gupta et al. 1992). If converted to a PCR-based B- 2

10 genome specific marker, this would be useful to detect B-genome in interspecific progeny and parents. With such a test, individuals carrying B- genome could be selected in the F 2 and BC 1, and those with putative B-genome introgressions could be detected by FISH reactions in F 2 -derived and BC 1 - derived lines. If canola quality is maintained in interspecific progeny through the approach outlined here, then interspecific progeny may be selected for superior performance in target agricultural environments with the knowledge that the selections are of immediate value in commercial canola breeding programs. Further backcrossing to the parent species to improve canola quality should not be necessary. Minor alleles contributing to valuable economic traits, such as yield, quality or disease resistance, from interspecific crossing will be available for further genetic improvement in elite canola breeding programs. The main hypothesis of this thesis is that progeny from crossing between canola-quality B. napus (AACC) and near canola-quality B. juncea (AABB) will retain their canola quality and demonstrate transgressive segregation for key agronomic traits in the fertile F 2 and BC 1 progeny. Further, meiosis and recombination in the F 1 (AABC) will result in few fertile F 2 and BC 1 progeny. B- genome chromatin will be present in some of these progeny. It is expected, based on past results (Roy 1980a), that cross direction will greatly influence the fertility of progeny. It is hypothesized that both B. napus (AACC) and B. juncea (AABB) progeny will be detected in F 2 -derived and BC 1 -derived progeny, and some of these will be canola quality with superior performance in for relevant agricultural traits in target environments. A secondary aim of the thesis was to develop PCR-based molecular assays to identify individuals with B-genome introgressions from B. juncea early in the selfing and backcrossing process. It is hypothesised that many of these B-genome positive progeny will have B. napus morphology. FISH cytogenetic studies will help to localise the site of introgression of B-genome in B. napus type progeny. The association of B-genome with new agronomic variation within the progeny will be examined. It may also be possible to identify specific influences of B-genome on traits observed in B. napus type progeny, such as blackleg resistance or adaptation to low rainfall environments. Finally, the research aimed to identify unique interspecific progeny with superior agronomic, disease or quality traits. Superior lines may be of either B. 3

11 juncea or B. napus morphology. It is hypothesised that BC 1 progeny backcrossed to B. juncea will have higher proportions of B-genome and more B. juncea type progeny, than those backcrossed to the B. napus parent; likewise, progeny backcrossed to B. napus will have less B-genome than selfed progeny. On the basis of previous results (Roy 1980a), it was expected that the majority of progeny derived from selfing of the F 1 would be B. napus morphology. The use of near canola-quality B. juncea in crossing with canola quality B. napus should reduce the so-called deleterious linkage drag which is often observed in interspecific crosses, where extensive backcrossing is necessary to restore seed quality or agronomic traits. Selection of high quality progeny with agronomic improvements from B. juncea may be possible in target environments, such as the low rainfall cropping regions of Australia. By minimizing backcrossing, minor alleles for complex agronomic traits should be available for selection in interspecific progeny with canola quality. This thesis is divided into a number of chapters as listed below. Apart from the published and submitted papers all of the references are listed at the end of the thesis: Chapter 1. General introduction: provides the background to this research project. Chapter 2. Literature review: reviews the literature on; The history of B. napus as a crop species in Australia Agronomic factors that limit expansion of canola in Australian farming systems B. napus, its evolution and relatives and, Sources of new genetic variation for the improvement of B. napus Chapter 3. Published paper: Schelfhout CJ, Snowdon R, Cowling WA, Wroth JM (2004) A PCR based B- genome-specific marker in Brassica species. Theoretical and Applied Genetics 109,

12 This chapter outlines the isolation and development of a B-genome specific marker that is believed to be the first B-genome marker present on all Brassica B-genome chromosomes. Chapter 4. Published paper: Schelfhout CJ, Snowdon R, Cowling WA, Wroth JM (2006) Tracing B-genome chromatin in Brassica napus B. juncea interspecific progeny. Genome 49, This chapter outlines the further development of the B-genome specific probe outlined in Chapter 3. Here the marker is developed into a probe for fluorescence in-situ hybridisation (FISH) and used to identify putative B-genome introgressions in B. napus B. juncea interspecific progeny. Genomic relationships are discussed. Chapter 5. in press (June 2008): Schelfhout CJ, Wroth JM, Yan G, Cowling WA (2008) Enhancement of genetic diversity in canola-quality Brassica napus and B. juncea by interspecific hybridisation. Australian Journal of Agricultural Research (in press). This chapter outlines the extensive field testing of the Brassica interspecific progeny. It also discusses the transgressive segregation observed and the reciprocal nature of genetic improvement in both B. napus and B. juncea. It also discusses the effects of the interspecific cross on seed quality. Chapter 6. General discussion; This chapter critically discusses the key information gained from this research and proposes some opportunities for further investigation. 5

13 Chapter 2 - Review of Literature Brassica napus as a crop species in Australia History Brassica napus first appeared as a commercial crop (rapeseed) in Australia in 1969 during a period when cereal growers were faced with the introduction of wheat quotas (Colton and Potter 1999). Lack of resistance to the fungal disease blackleg (causal agent Leptosphaeria maculans) in these early varieties from Canadian germplasm caused widespread failure of rapeseed crops within a few years. In Western Australia blackleg disease was so severe that the area sown to rapeseed was reduced from 49,000 hectares in 1973 to 3,000 hectares in 1974 (Salisbury et al. 1995). The major focus of research and breeding during the 1970s was to improve blackleg resistance and to reduce anti-nutritional components, namely erucic acid in oil and a range of glucosinolates in the seed, as discussed in several reviews (Downey 1990, Downey and Rimmer 1993, Scarth 1995). The term double-low rapeseed was used to denote varieties of rapeseed with less than 2% erucic acid and less than 30 µmoles of total glucosinolates per gram of meal, and in 1979 the term canola was introduced in Canada to describe varieties with these improved seed quality traits (Anon. 2007, Colton and Potter 1999). The current quality requirements for Australian canola being traded in the open bulk commodity market are listed in Table 1. Canola breeding in Australia in the 1980s continued to focus on the improvement of blackleg resistance as well as the introduction of hybrid and herbicide resistant varieties. Development of Australian cultivars is based on a very limited germplasm pool of about 11 ancestral parents introduced in 1970 (Cowling 2007). There has been little introduction of new germplasm over the past 30 years and recent Australian varieties are the product of crossing within this gene pool over 5 generations of closed recurrent selection (Cowling 2007). 6

14 Table 1. Quality standards for Australian canola (modified from Australian Oilseeds Federation, Salisbury and Potter 1999). Commodity Quality trading standard Price adjustments and comments Canola (Brassica napus or Brassica rapa) Oil 40% (Western Australia 42%) Glucosinolates as specified: maximum 30 micromoles (any one or any mixture of 3-butenyl glucosinolate, 4-pentenyl glucosinolate, 2-hydroxy- 3 butenyl glucosinolate, and 2-hydroxy-4-pentenyl glucosinolate) per gram of oil-free air-dry solids Erucic acid in oil maximum 2% Impurity max. 3% (Rejected above ) Moisture max. 8% (Reject above) Broken seed max. 7% (Rejected above) Damaged seed - 3% Sprouted max. 5% Green - nil (2%) Total max. 10% (Rejected above) 1.5% premium or deduction for each 1% above or below 40% respectively Gross weight will be adjusted by 1% for each 1% up to 3% max. 2 for 1 penalty over 4% If accepted over max, 2% deduction for each 1% over the allowed level If accepted over max, 0.5% deduction for each 1% over the allowed level 0.5% deduction for each 1% over the allowed level - except for green seeds which shall incur a penalty of 1% for each 1% over zero (nil) to a maximum of 2% over which green seed level, the seed is rejectable Canola production in Australia: the current status During the 1990s, improved varieties were released and Australian canola production increased from approximately 0.1 Mt in 1990 to 1.7 Mt in 1999 (Colton and Potter 1999). While production was limited to the higher rainfall areas, by the late 1990s herbicide tolerant and shorter season varieties became available leading to expansion of the industry into rainfall zones as low as 280 mm per annum (Salisbury and Wratten 1999). Australian production peaked in 1999 and has declined in recent years due to several unfavourable seasons. Table 2 shows the area sown to canola and the corresponding production for the last nine seasons (Australian Oilseeds Federation, 2007). 7

15 Table 2. Figures for the area sown to canola in Australian and the production in years 1998/99 to 2007/08 est. Source: and_figures. Season 99/00 00/01 01/02 02/03 03/04 04/05 05/06 06/07 07/08 est. Area sown ( 000 ha) Production ( 000 t) The benefits of growing canola In addition to direct benefits in seed oil and meal, canola provides grain growers with several useful rotational benefits. Canola, a broadleaf species, permits greater control of grass weeds for subsequent cereal crops and allows chemical group rotation to avoid herbicide resistance in weeds (Oilseeds WA 2006). Canola also acts as a break crop for cereal diseases (Norton et al. 1999). Kirkegaard et al. (1994, 1997, 2000) have suggested that canola may exhibit soil fumigation properties by releasing glucosinolates that are toxic to soil borne plant pathogens. While the mechanism by which canola provides benefits to subsequent cereal crops is debated, the boost in wheat yield is significant from a survey of 226 wheat crops in 1995 across Victoria, wheat yield following canola was 3.9 t/ha where wheat following wheat was 2.8 t/ha (Norton et al. 1999). Rising oil prices and increasing interest in biofuels has bolstered farmers confidence in canola as a profitable crop. Recent drought and tight supply for oilseeds has kept global prices for canola buoyant with an outlook for possible future rises (Gorey 2007). 8

16 Agronomic factors that limit expansion of canola in Australian farming systems Climate Traditionally canola has been restricted to higher rainfall regions of southern Australia (>450 mm). With the advent of shorter season varieties canola, production has expanded into low rainfall environments with annual averages of mm (Salisbury and Wratten et al. 1999). Unlike many other canola regions of the world, canola crops in southern Australia mature under increasing temperatures and ensuing terminal drought stress. Soil moisture stress at pod fill can cause considerable yield loss through poor seed set, pod abortion and quality problems such as reduced oil (Mailer and Pratley 1990) and increased seed glucosinolate content (Champolivier and Merrien 1996). Lewis and Thurling (1994) observed higher yields in low rainfall environments from B. juncea compared with B. napus cultivars and concluded that the higher yields were achieved through higher rates of post-anthesis growth in the B. juncea lines. In tandem with decreasing soil moisture, temperatures rise during the maturity of Australian canola crops. High temperatures and low rainfall during pod fill are associated with low oil content of canola seed (Pritchard et al. 2000). There is a strong inverse relationship of seed oil and protein content under Australian conditions (Si et al. 2003). Si and Walton (2004) reported limits to oil content in short season, low rainfall environments. High temperatures at crop maturity contributed to an increase in the production of erucic acid in the seed oil (Wilmer et al. 1997). There have been suggestions that high temperatures may reduce unsaturated fatty acid composition, however experiments utilising short periods of high temperature have failed to prove this (Pritchard et al. 2000). Crop establishment Good crop establishment can be hampered by inconsistent rainfall events and patchy soil moisture. By selecting larger seed, Kant and Tomar (1995) demonstrated in B. juncea that germination, emergence seedling length and early vigour are improved. This may help establishment under low moisture conditions. 9

17 Soils and crop nutrition Some areas in South-Eastern Australia are young, fertile soils, whilst those of Western Australia are ancient, infertile soils and require the regular application of fertilisers to be productive (Hunt and Gilkes 1992). Canola crops require approximately 25% more nitrogen, phosphorous and potassium and five times more sulphur than Australian wheat crops (Hocking et al. 1999). The higher input costs of canola compared to wheat increases the financial risk of growing canola, especially in low rainfall regions. Extensive use of nitrogen fertilisers has led to the development of soil acidity over large areas of Australia s agricultural regions (Hunt and Gilkes 1992). Low soil ph values can result in aluminium toxicity that causes stunted root growth (Kochian 1995, Huang et al. 2002) and manganese toxicity that induces chlorosis and distorted leaves (Hocking et al. 1999). Application of lime is a common approach to combating low soil ph. However, introgression of aluminium tolerance from wild Brassica relatives has been proposed to improve oilseed rape production on acid soils (Huang et al. 2001), and there have been reports of aluminium tolerance in Brassica species (Huang et al. 2002). Diseases The fungal pathogen that causes blackleg, Leptosphaeria maculans (asexual form Phoma lignam) is by far the most serious disease of canola and commonly causes crop lodging during the seed fill stage (Salisbury et al. 1995). The Mediterranean climate of southern Australia is particularly favourable for the retention of the blackleg fungus on crop residues over the summer months. Hot dry summers inhibit fungal growth and development, which resumes in the cooler autumn and winter months with the release of ascospores that are usually dispersed several hours after rain (Hall 1992) or following heavy dew (McGee 1977). In Australia, peak ascospore showers occur in June, July and August (winter) and coincide with the time of crop sowing and establishment (Bokor 1975, Salisbury et al. 1995, Khangura and Barbetti 2004). Resistance to blackleg has increased in Australian breeding programs since the 1970s (Cowling 2007). Cargeeg and Thurling (1980) proposed that 10

18 adult plant resistance to blackleg was polygenic due to the range in the level of resistance among breeding populations. Early work by Roy (1978) and Sacristan and Gerdemann (1986) putatively transferred adult plant resistance to blackleg from the A genome of Brassica juncea (Genome = AABB) to the A genome of B. napus (Genome = AACC). By 1984 a low erucic acid line (Onap JR ) with complete (seedling and adult) resistance to blackleg had been found in progeny from B. napus B. juncea interspecific hybridisation (Roy 1984). Onap JR was later crossed with partially resistant B. napus lines to produce progeny adapted to cold, wet and highly diseased conditions (Roy 1984). Roy (1984) did not conduct cytological studies in an attempt to identify introgressed donor germplasm and was unable to establish homozygous or stable lines with the B. juncea type resistance. Attempts to incorporate the B genome resistance in B. napus were made by producing B. napus B. nigra somatic hybrids (Gerdemann-Knorck et al. 1995). High levels of resistance to blackleg were achieved using this method, although further breeding work is required to stabilise the resistance in individual plants. B. napus B. juncea hybrids generated by Barret et al. (1998) were claimed to have high levels of blackleg resistance (including cotyledon resistance) introgressed from B. juncea. Two theories were proposed for the introgression of this fragment. The first theory suggested that the resistance was carried on a chromosome that had substituted for an entire B. napus chromosome. The second theory suggested the loci carrying resistance were located on a fragment homologous to that in B. napus, thus permitting homologous recombination. Earlier work by Roy (1978, 1984) indicated that the B. juncea complete resistance was carried on the B genome thus making it unlikely to be transferred via homologous recombination. Results published by Barret et al. (1998) indicates possible incorporation of improved A genome type resistance from B. juncea. In the study of a cross between the susceptible B. napus cultivar Tower and a B. juncea-type resistant B. napus line, Pang and Halloran (1996) suggested that adult plant resistance to blackleg is governed by three major genes with complex interactive effects. Mayerhofer et al. (1997) mapped blackleg resistance to a single locus in the Australian cultivars Shiralee and Maluka and subsequently used these cultivars to generate Canadian cultivars with high levels of blackleg resistance. 11

19 The B. napus cultivar Surpass 400, released in 2000 in Australia, contained a powerful new resistance gene that was monogenically dominant (Li and Cowling 2003). Cultivars with this gene, introduced from a wild B. sylvestris plant (Crouch et al. 1994), had excellent resistance to blackleg until a break down of the B. sylvestris resistance was observed (Li et al. 2003). Breakdown of this resistance was subsequently confirmed by surveys conducted in 2003 on cultivars with this monogenic resistance (Canola Association of Australia 2004). In the same year canola crops on the Eyre Peninsula and several other regions of Australia suffered serve blackleg related losses. L. maculans can rapidly overcome major gene resistance to blackleg in Australia (Li et al. 2005) and in Europe (Brun et al. 2001). Pests Key invertebrate pests of canola attack crops at the establishment stage or at the flowering and pod fill phase (Miles and McDonald 1999). During crop establishment canola is attacked by a number of insect pests including the redlegged earth mite (Halotydeus destructor), blue oat mite (Penthaleus spp.), lucerne flea (Sminthurus viridis) and false wireworm (Isopteron punctatissimus, Adelium spp.). Pests that specifically attack canola crops during the flowering and pod fill phase include aphids (turnip aphid, Lipaphis erysimi, cabbage aphid, Brevicoryne brassicae, and green peach aphid, Myzus persicae), native budworm (Helicopvera puntigera) and Rutherglen bug (Nysius vinitor). Recently diamondback moth (Plutella xylostella) has emerged as a serious threat to canola crops. This pest became prominent in the Northern agricultural region of Western Australia in 2000 and the major outbreak was attributed to unseasonal weather conditions (Cook et al. 2000). Yield Yield improvement in canola remains the primary breeding objective as for most other crop species. Yield is a complex trait, the end result of many physiological and morphological traits that are quantitatively inherited as well as environmental factors that affect a crop over the entire growing season. Thurling (1974a, 1974b) found that a reduction of yield occurred in both B. napus and B. rapa with delays in sowing date and that this was most likely a 12

20 result of a reduced duration of pre-anthesis growth. Thurling and Vijendra Das (1979) suggested that higher yield is determined by pre-anthesis vegetative growth. In low rainfall zones of Southern Australia, particularly at lower latitudes, drought stress and high spring temperatures have potential to hamper crop yield and quality. One solution is the breeding of early flowering varieties that can form siliqua and start filling seed prior to the onset of heat and drought stress. However, early flowering itself may not increase yield because this trait shortens the period of vegetative growth prior to anthesis (Thurling 1983). In order to maintain viable yields whilst reducing the time to flowering it is important to increase the relative growth rate of the plant during this preanthesis period. In addition to this, Lewis and Thurling (1994) demonstrated that yield may be improved in crops facing terminal drought through improved water use efficiency during the post-anthesis period. In that study B. juncea had better post-anthesis water use efficiency than B. napus and B. rapa, but similar total water use during the growing season. The efficient post-anthesis water use of B. juncea enabled this species to generate a higher yield (Lewis and Thurling 1994). Niknam et al (2003) found that osmotic adjustment occurred in water stressed Brassica and varied between species and cultivars. It was also observed that those accessions with a high osmotic adjustment suffered less yield reduction in water limiting conditions. Weeds Canola is regularly challenged by a number of grass and broadleaf weeds including many weeds from the Brassicaceae family such as wild turnip (B. tuornefortii) and wild radish (Raphanus raphanastrum) (Sutherland 1999). Minimum or no-till cropping relies on herbicide application at both pre- and postemergent stages of crop development. Post emergent weed control is also dependent on resistance in the crop species to the herbicide being applied. Herbicide resistance is a valuable trait carried in a number of Australian canola varieties. The most common resistance is to the Group C triazine herbicides. Triazine tolerance was discovered in B. rapa (Maltais and Bouchard 1978) and was later found to be carried in cytoplasmic DNA and inherited maternally (Souza-Machado et al 1978). This trait has since been introduced in B. napus populations (Beversdorf et al 1980). This expands the control options 13

21 for related (and other) broadleaf leave weeds such as wild radish as well as annual ryegrass that are serious competitors and contaminators of canola crops (Littlewood and Garlinge 2001). A yield penalty is observed with the triazine tolerant (TT) cultivars (Röbbelen 1987, Robertson et al 2002). Despite this yield suppression, canola production is dominated by triazine tolerant canola in Australia due to major benefits from weed control in minimum till cropping systems. Recently varieties with resistance to the Group B imidazolinone herbicides have been introduced to Australia, which provide farmers with greater flexibility in weed control (Littlewood and Garlinge 2001). However, resistance to Group B herbicides develops rapidly in weed species and consequently producers must plan carefully to integrate these herbicide resistant canola varieties into their crop rotations (Sutherland 1999). Silique shatter Most Australian canola crops are swathed to regulate seed maturity and to limit pod shattering. Swathing adds $20/ha to input costs (Carmody 2001) and has a significant bearing on the gross margin of canola production, particularly in marginal regions where yields are low. While breeding has reduced the degree of pod shatter, almost all varieties will shatter to some degree and yield losses increase proportionally as swathing is delayed beyond maturity. Optimal swathing time is when half or more of the seed has lost its green colour (Carmody 2001). A number of Brassica relatives with higher levels of resistance to shatter have been identified as a potential source of shatter resistance in B. napus, such as Sinapis alba (Brown et al. 1997) and B. juncea (Prakash and Chopra 1988; Prakash and Chopra 1990). Shatter resistance has also been introduced into B. napus through the resynthesis of B. napus from its diploid progenitor species B. rapa and B. oleracea (Morgan et al. 1998), and is believed to be inherited polygenically and independently of other agronomic traits (Morgan et al. 2000). Oil content and quality Seed oil content is inversely related to seed protein content (Singh et al. 2001, Si et al. 2003) and this is particularly prominent in low rainfall, high temperature regions. The oil content of canola in these areas often falls below the current 14

22 standard (Table 1) and in unfavourable years large areas of the Australian canola crop are affected by low seed oil. Further breeding effort is needed to stabilize oil content and quality under drought and heat stress conditions. Nevertheless, Australian canola meets the global standards for oil content and is widely recognised for its low levels of moisture, admixture and chlorophyll, all of which assure a major advantage for buyers and processors of Australian canola (Australian Oilseeds Federation 2007). The frying vegetable oil market may demand a reduction in saturated fats and increase in the mono-unsaturated oleic acid content of Australian canola in the future (Mailer 1999). In 1999 large areas in Western Australia were planted to a single cultivar, Karoo (Nelson 2000). Unfortunately, this variety had a low proportion of individual plants with high levels of erucic acid that affected the overall oil quality and caused problems for the Australian export market (Nelson 2000, Cowling et al. 2001). Canola breeding is necessary to develop high quality cultivars with stable seed quality and greater adaptation to season-end stresses. Brassica napus, its evolution and relatives Over 95% of oilseed rape grown in Australia is Brassica napus var. oleifera (West et al. 2001). B. napus is a member of the Brassicaceae family of plants that have a multitude of uses and have benefited mankind for thousands of years. This family includes species that are cultivated as oilseeds, vegetables and condiments. The vegetable Brassicas (B. oleracea and B. rapa) have a range of morphotypes including leafy rosettes, heading forms, swollen stems and enlarged roots (Labana and Gupta 1993), whereas the other species B. nigra, B. napus, B. juncea and B. carinata are predominantly oleiferous forms with some morphotypes suitable for use as vegetable crops. It is believed that B. napus first appeared near Flanders around the year 1600 where it quickly replaced the winter turnip rape throughout north-western Europe (Labana and Gupta 1993). Its origins, despite being relatively recent, parallel other amphidiploid Brassica species that appeared many years earlier. It is assumed that B. napus originated in South-Eastern Europe around the 15

23 Mediterranean where populations of its progenitor species B. rapa and B. oleracea overlap (Labana and Gupta 1993). U s triangle and the Brassica genomes A Korean botanist working in Japan (U 1935) proposed the theory of a triangular relationship amongst six oilseed and vegetable Brassicas. This triangle (Fig. 2), traditionally known as U s triangle, allocates alphabetic genome descriptors and illustrates the relationship between the three main diploid species B. rapa (AA), B. nigra (BB) and B. oleracea (CC) and their amphidiploid progeny B. juncea (AABB), B. napus (AACC) and B. carinata (BBCC). B. juncea AABB 2n = 4x = 36 B. rapa AA 2n = 2x = 20 B. napus AACC 2n = 4x = 38 B. nigra BB 2n = 2x = 16 B. carinata BBCC 2n = 4x = 34 B. oleracea CC 2n = 2x = 18 Figure 2. The Triangle of U, showing genome descriptors and progenitor genomes of the amphidiploid species (from U 1935). The triangle of U has been well supported by a number of research groups using different techniques. These included studies on amphidiploid production and chromosome pairing (Prakash and Hinata 1980), nuclear DNA content (Verma and Rees 1974), DNA analysis (Erickson et al. 1983) and the use of genome specific markers (Hosaka et al. 1990). Cytogenetics and genome relationships among Brassica crops and relationships to the model plant Arabidopsis are reviewed recently by Snowdon (2007). 16

24 Studies on pachytene chromosome analysis provide evidence to suggest there are six Brassica genomes observed within the diploid Brassica species (Röbbelen 1960, Venkateswarlu and Kamala 1971) and are classified as the A, B, C, D, E, and F genomes. Other reports suggest there may have been as few as five or as many as seven (Truco et al. 1996). Chromosomes of these five to seven original types within the present three diploid genomes have lost homology due to duplication and rearrangements (Prakash and Chopra 1993, Truco et al. 1996). Cytogenetic reports indicate the three diploid species evolved from a common progenitor species (Prakash and Chopra 1993), however more recent molecular studies (Song et al. 1988, Yanagino et al. 1987, Truco et al. 1996), suggest a biphyletic origin of the diploid Brassicas where B. rapa and B. oleracea evolved from one lineage and B. nigra evolved separately along another lineage. The amphidiploid oilseed B. napus appears to have arisen spontaneously through natural hybridisation of B. oleracea and B. rapa (Chen and Heenan 1989). The progenitor species of B. napus (B. rapa and B. oleracea) have many sub-species or varieties (Prakash and Hinata 1980). Chen and Heenan (1989) deduced that the narrow range of genetic variation found within B. napus resulted from the hybridisation of very few of these parental genomes. Other investigators (Hosaka et al. 1990, Olsson 1960 and Song and Osborn 1992) suggest that B. napus arose from polyphyletic origins rather than a single hybridisation event between the two parent genomes. Despite these theories of polyphylogeny in B. napus it is accepted that the genetic diversity of oilseed rape (Brassica napus) is small. Becker et al. (1995) identifies two probable causes for this: (1) rapeseed is of relatively recent origin and extensive cultivation and breeding of this crop species did not occur until little over 50 years ago; and (2) the species has a narrow genetic base. The present breeding material of oilseed rape is derived from very few interspecific hybrid plants that occurred spontaneously some centuries ago (Becker et al. 1995). Many studies have attempted to re-create the amphidiploid Brassica species using a number of hybridisation techniques. B. carinata and B. juncea were easier to resynthesise than B. napus (Prakash and Chopra 1993), possibly as a result of the more recent evolution of B. napus compared with its amphidiploid relatives. Approaches taken to generate synthetic B. napus from 17

25 B. rapa and B. oleracea have included grafting (Hosoda et al and Namai 1971), mixed pollination (Feng 1955 and Sarashima 1964), style excision (Hososda et al. 1963), embryo culture (Nishi et al. 1961, Nishi et al and Matsuzawa 1984) and ovary culture (Inomata 1978). More recently resynthesis of B. napus has become a common practice for the introduction of desirable variation from its diploid parents (Chen and Heenan 1989, Becker et al. 1995) and has practical implications for crop improvement. The Brassica genomes have been the subject of many studies on chromosome interaction, homology and wide hybridisation. The basis of these studies has paved the way for future research into germplasm improvement of B. napus and its oilseed relatives. Introducing genetic variation for the improvement of Brassica napus Kumar et al. (1988) state that the genus Brassica, with its large number of wild species, had the potential to donate many new nuclear/organelle genes to the different cultivated species for improvement in the range of edible oils and vegetables. However, because of pre- and post fertilisation barriers, hybridisation between wild and cultivated species has not been very successful. Sources of genetic variation for Brassica napus Spontaneous mutations may often go unnoticed although occasionally they provide useful variation for crop improvement (Muangprom and Osborn 2004). Induced variation has been used extensively with Brassicas to provide genetic variation for disease resistance (Yadav et al. 2001), altered oil profile (Velasco et al. 1998), shatter resistance (Kadkol 1987) and yield (Rahman and Das 1991). Despite valuable results that have been obtained by a number of groups, the use of induced mutagenesis is limited due to diplontic selection, adverse linkages and the undesirable pleiotropic effects normal associated with mutants (Dhillon et al. 1993). Genetic transformation is another method of creating genetic variation in B. napus where gene sequences may be taken from any organism. Genetic modification by plant transformation has advanced more rapidly in Brassica 18

26 species than many others because of the rapid advance of tissue culture methods (Moloney and Holbrook 1993). By using genetic transformation, interspecies and wider genetic barriers can be by-passed. Genetic transformation is an expensive and time consuming approach but does provide a direct approach for the introduction or improvement of a desired trait. New variation has been observed for a number of agronomic traits in progeny derived from B. juncea somaclones (Katiyar 1997, Jain et al. 1989). Despite promising variation arising from somaclonal variation, direct application of these techniques to generate variation are uncommon in contemporary breeding programs where breeders are seeking specific trait improvements. In most crop species, crossing and recombination of genes and alleles is still the most widely used and basic method of introducing new genetic variation. However, its application becomes limited when the target alleles lie across species barriers. This problem is compounded by the restricted genetic diversity within the primary gene pool of B. napus oilseed rape germplasm (Becker et al. 1995). The secondary and wider gene pools Centres of origin as first described by Vavilov are regions of high genetic diversity for domesticated species (Vavilov 1951). These centres of origin may provide a source of new genes and alleles for particular traits in breeding programs. This is valuable in some crop species where local varieties have become extinct or where domestication has reduced the genetic base to a limited number of varieties (Borojevic 1990). Wild populations of B. napus have not been found (Prakash and Hinata 1980) around its proposed centre of origin, thus the only ancestral sources of variation for B. napus may lie with domesticated and wild relatives of the progenitor genomes in B. rapa and B. oleracea. Using interspecific hybridisation to introduce genetic variation in B. napus Hermsen (1992) outlined the importance of selecting new alleles for crop improvement from within the species where possible. This approach avoids the introduction of large quantities of alien DNA and the subsequent associated genetic disruption. As previously outlined, the gene pool of B. napus is limited 19

27 (Becker et al. 1995). Breeders have little choice but to search the secondary gene pool to find sources of improvement for desired traits. Two populations constitute different species when there is no gene flow between them because of genetic barriers (Lacadena 1978). With manipulation these interspecies barriers may be overcome to make use of beneficial germplasm of closely related species. Interspecific hybridisation within the Brassicas has enabled improvement of disease resistance (Roy 1984, Somers et al. 2002), herbicide resistance (Ayotte et al. 1987), pod shatter resistance (Prakash and Chopra 1990), oil profile (Getinet at al. 1994) and flowering time (Rao et al. 1993a, Rao et al. 1993b). Interspecific barriers to wide hybridisations Barriers such as spatial isolation, non-synchronous flowering and cleistogamy may be easily controlled in a breeding program (Hermsen 1992). For the complete process of interspecific hybridisation to take place, there must be no restriction imposed on the processes that take place after pollination and there must be homology in the genetic processes that take place between the pollen and the host plant. Hogenboom (1973) uses several terms when referring to interspecific crosses such as incongruity to describe non-homology in the genetic process, penetration capacity to describe mechanisms controlled by the pollen genes to circumvent barriers against hybridising alien females, and barrier capacity to note the genetic regulation of barriers against hybridisation within the female plant by male pollen. Prezygotic barriers may occur on the stigma, style or ovary of the host plant and are well documented for many interspecies crosses. Hybridisation may not occur when pollen fails to germinate due to a mismatch of stigma fluids. However, Heizmann et al. (2000) found few problems with the adhesion of pollen to stigmas in both intra- and interspecific crosses of Brassicaceae species. Wider crosses in this family were shown to exhibit reduced pollen adhesion to the stigma (Luu et al. 1998). Barriers to pollen tube growth occur in either the stigma or style of the host plant. Yadav et al. (2002) observed variation in the rate of pollen tube growth amongst several Brassica interspecific hybrids. Rate and frequency of pollen tube growth was seen to be affected by the direction of the cross, where reciprocal crosses favoured one direction. Luu 20

28 et al. (1998) suggested most interspecific pre-zygotic barriers occur within the style after the pollen grains have germinated. Post zygotic impedance of sexual hybridisation in remotely related species may result in fully sterile F 1 plants due to lack of homologous pairing of chromosomes (Hermsen 1992). Chromosome elimination, sub-lethality, or other abnormalities of plant growth, male sterility, poor flowering of F 1 plants, disharmonious karyotypes in segregating generations, all causing hybrid breakdown are typical post-zygotic barriers to hybridisation. Subsequently hybrids with these faults rarely produce viable seed and their genetic variation is lost. Overcoming barriers to interspecific hybridisation Pre-fertilisation barriers can be overcome by grafting, mixed pollination, style excision, treatment of the style with chemicals and embryo rescue (Prakash and Chopra 1993), whilst post-fertilisation barriers are overcome by in-vitro techniques such as ovary, ovule and embryo rescue. Techniques used to overcome Brassica interspecific hybridisation barriers are listed in Table 2 (modified from Prakash and Chopra 1993). Table 2. Techniques for overcoming barriers to interspecific hybridisation in Brassica species (modified from Prakash and Chopra 1993). Technique Reference Grafting Hosoda et al. (1963), Namai (1971) Mixed pollination Feng (1955), Sarashima (1964) Style excision Hosoda et al. (1963) Embryo culture Nishi et al. (1961), Ayotte et al. (1987), Chrungu et al. (1999), Kumar et al. (2001) Ovary culture Inomata (1978), Inomata (2002) Despite wild B. napus populations never being found there is an immense amount of genetic diversity within its progenitor genomes. Research has shown that interspecific barriers in the Brassica family can be overcome. 21

29 Analysis of Brassica interspecific hybrid progeny Lacadena (1977) stated that the main objective of interspecific hybridisation in plant breeding was to expand the gene pool and to introduce alien genes carried by wild species into cultivated varieties. This statement may be extended to include not only wild species but also closely related species in all stages of domestication. Location and identification of genetic sequences responsible for new variation in recipient Brassica genomes assist in the stabilisation of these traits in future generations. Interspecific progeny have been analysed both phenotypically and/or genotypically to characterise the range of new genetic variation. Phenotypic analyses The progeny of Brassica interspecific hybrids are screened for morphological and physiological traits such as yield, disease and/or quality traits that differ from the parental material. Environmental influences are known to significantly affect phenotypic analyses. Richards (1978) found maximum response to selection for drought stress in B. rapa and B. napus in the target (low rainfall) environment. Genotypic analyses The ability to detect, locate and characterise foreign DNA from a donor genome in progeny of an interspecific cross will assist selection of desirable genotypes and phenotypes (Basunanda et al. 2007). Selection may occur for alien DNA in the early generations of selfing or backcrossing, thus increasing the probability of securing useful new traits to the host species. Identification of alien chromatin in wide crosses has been achieved by using cytogenetic techniques to identify chromosome insertions (Snowdon 1997), flow cytometry to quantify chromosome content (Sabharwal and Dolezel 1993) and molecular markers to trace chromosome and DNA transfer (Quiros 1991, Basunanda et al. 2007). Cytogenetics Direct observation of chromosomes can be a useful tool in the analysis of interspecific hybrids. Chromosomal rearrangements may be observed and 22

30 correlations drawn between these and phenotypic observations. The in situ detection of nucleic acid sequences, whether of genes on chromosomes or viruses or of mrna in tissues, provides a direct visualisation of the spatial location of sequences that is crucial for elucidation of the organisation and function of genes (Wilkinson 1998). In situ hybridisation involves using a labelled probe (radioactive or fluorescent) to detect complementary sequences in target DNA. The probe may be either digested genomic DNA (genomic in situ hybridisation, or GISH) or any gene or DNA sequence used as a probe (fluorescence in situ hybridisation, or FISH). GISH has been used as tool to identify Brassica genomes in interspecific and intergenomic hybrids. By using genomic DNA probes from the three diploid species the A and B genome chromosomes could be distinguished in B. juncea, as well as the B and C genome chromosomes in B. carinata (Itoh et al. 1991, Snowdon et al. 1997). Hosaka et al. (1990) were not able to distinguish between the chromosomes of these two genomes, possibly due to the relative homology of the A and C genomes. On the basis of this report, along with the small size of Brassica chromosomes, it appears GISH may not be a suitable method for identifying C genome introgressions in the AA (B. rapa) or AABB (B. juncea) background. Using a sequence originally reported by Harrison and Heslop-Harrison (1995), Armstrong et al. (1998) used FISH with a highly repeated sequence (pbckb4) from B. campestris (B. rapa) to demonstrate the presence of C- genome in wide Brassica crosses. This probe was observed to co-localise with pericentomeric heterochromatin on the nine chromosomes of B. oleracea (CC). The probe was not totally genome specific, but may prove useful in the detection of whole C genome chromosomes in the B. juncea (AABB) background. This marker may however fail to identify translocation of distal or telomeric C genome chromosome fragments in an A or B genome background. FISH markers, based on reverse transcriptase domains in retroelements of B. oleracea, showed characteristic distributions on the C genome (Alix et al. 2005). Localisation of alien introgressions There has been much discussion on the ability of transgenes in B. napus to recombine with other species in interspecific and other wide crosses 23

31 (Jorgensen and Andersen 1994). This has prompted investigations of safe integration sites within the B. napus genome which minimise transgene recombination in wide hybrids (Mikkelsen et al. 1996, Tomiuk et al. 2000). This concept has implications for the introgression of alien DNA during the creation of Brassica interspecific hybrids. It would be virtually impossible to target specific sites in B. napus for introgression through interspecific hybridisation. However, it would be noteworthy to determine if there are certain chromosomal regions that are consistently recombining in these wide crosses. The stability of B genome additions or introgressions in the A or C genome of B. napus needs further investigation. Non-homologous chromosome pairing has been reported in the A and B genomes of Brassicas (Prakash 1973) but putative introgressions have not always been stable (Roy 1984). There is little research to date that shows preferential introgression or recombination in the progeny of Brassica interspecific hybrids. Molecular genetics Koornneef et al (1997) state that genetics becomes more powerful when combined with molecular genetics, which links DNA and phenotype. An array of molecular techniques has emerged over the past 25 years, many of which have benefited plant breeding. Perez de la Vega (1993) outlines three marker systems that supplement phenotypic analysis in plant populations. These are (1) biological compounds or non-protein low molecular weight plant compounds, (2) proteins, including enzymes and storage proteins and (3) DNA markers obtained by restriction enzymes, by PCR and base sequences. The third group mentioned here have become particularly prominent in modern plant breeding. Some examples of molecular genetic uses for a range of breeding and genome mapping exercises in the Brassica species include DNA markers based on restriction fragment length polymorphisms (RFLP) by Lagercrantz and Lydiate (1996), randomly amplified polymorphic DNA (RAPD) (Thormann et al. 1994), and PCR based microsatellite DNA sequences (Uzunova and Ecke 1999) and have been comprehensively reviewed by Snowdon and Friedt (2004). These techniques are often used with other genotypic analyses to study Brassica interspecific hybrids and have all proven to be useful tools in the advancement of Brassica germplasm. 24

32 Genome specific sequences Genome and species specific marker studies in Brassica have been reported a number of times. Chen and Heenan (1990) have used genome specific isozyme markers to study chromosome behaviour in interspecific hybrids of B. napus and B. campestris (B. rapa). Chèvre et al. (1991) used isozyme, RFLP and fatty acid markers that expressed genome specificity in the analysis of B. napus B-genome addition lines. Hosaka et al. (1990) generated genome specific markers based on RFLP probes that were useful in identifying A and C genome DNA (Table 3). The C genome specific markers were specific to the C genome when probed to genomic DNA in B. oleracea, B. napus and a synthetic B. napus line. Two probes were identified that bound exclusively to C genome DNA (pb547 and pb870) in both diploid B. oleracea and amphidiploid B. napus. Not all of the probes studied bound exclusively to species containing C genome. Hosaka et al. (1990) stated that these two probes were dispersed and provided evidence to suggest that these probes hybridise to five of the C genome chromosomes. It is not stated whether these probes hybridise to locations on all of the nine C genome chromosomes. Some of the probes that initially appeared to be C genome specific also hybridised to fragments from alternative genomes (pb845) suggesting a degree of genome homology. Iwabuchi et al. (1991) isolated and cloned an 88 base pair repetitive sequence present in the A genome of B. campestris (syn. rapa) (pcs1). This sequence showed no homology to the C-genome found in B. oleracea. This probe may also hybridise to A-genome chromosomes of B. napus. The sequence was reported to be a middle repetitive element with an approximate copy number of 1,680 (Iwabuchi et al. 1991). In situ hybridisation analysis exhibited strong signals on three of the ten A-genome chromosomes indicating that this sequence is not distributed across the entire A-genome. Itoh et al. (1991) successfully used the probe CS1 in the detection of A-genome chromosomes in somatic hybrids of B. campestris and B. oleracea. The genome specific markers mentioned here are based on digested genomic clones. 25

33 Table 3. Genome specific markers generated by Hosaka et al. (1990) Clone Presence Absence pb177 A C B pb185 C A B pb320 C A B pb370 A C B pb485 A B C pb488 A B C pb547 C A B pb845 C A B pb850 C A B pb870 C A B Quiros et al. (1991) reported the application of RAPD based genome specific markers. Sixty-five genome specific markers were identified found in both the diploid and derived amphidiploid species. Sixteen of these markers were A- genome specific, 37 were B-genome specific and 12 were C-genome specific. Out of the 37 B-genome specific markers, 11 were mapped to four independent B-genome chromosomes in B. napus-nigra addition lines. RAPD markers have also been used to identify seven of the eight B-genome chromosomes (Struss et al. 1992). To date, there have been no publications that identify A or C genome specific markers that are dispersed across the entire genome of the respective diploid species. Quiros et al. (1991) could not confirm whether their markers were dispersed across every chromosome for each of the three genomes. The difficulty involved in identifying unique A and C genome specific markers supports theories on the relative homology of these two genomes (Song et al. 1988). Snowdon et al. (2002) was able to distinguish between A and C-genome chromosomes in B. napus using cytogenetic techniques however molecular markers would be even more useful for plant breeding applications wherever wide crosses are made through interspecific hybridisation of Brassica species. Gupta et al. (1992) reported the isolation of Brassica B genome specific sequences. Two sequences were used as hybridisation probes with one proving to be highly B genome specific. This 496 base pair sequence (pbnbh35) was observed to hybridise to species carrying Brassica B-genome chromosomes. Gupta et al. (1992) identified another B genome specific repeat 26

34 sequence (pbn-4), that is believed to be highly dispersed across all B-genome chromosomes (Kapila et al. 1996). With respect to their repetitive and broad distribution, coupled with the low homology of the B genome with the A and C genomes (Song et al. 1988, Yanagino et al. 1987), these markers may prove efficient in identifying B genome introgressions into A and or C genome backgrounds subject to further investigation and development. Conclusions Rapid adoption of canola in Australia during the 1990s demonstrates the need for this crop in Australian farming systems, although recent reductions in area indicate that genetic improvements are needed. Yield, quality and disease resistance remain major breeding objectives coupled with improved agronomic practices. It is unlikely that the yield limits of this crop are fully realised. Improved cultivars will come from breeding programs that select new genetic variation and useful allelic combinations for the benefit of the Australian canola. Canola in Australia has similar restrictions in genetic diversity to canola in other countries. Given the limited genetic base of B. napus, genetic resources in the secondary gene pool become important for future genetic improvements. There is a useful array of genetic tools and analyses that can be used to assist in the improvement of crop species today. 27

35 Chapter 3 Published paper Schelfhout CJ, Snowdon R, Cowling WA, Wroth JM (2004) A PCR based B- genome-specific marker in Brassica species. Theoretical and Applied Genetics 109,

36 Theor Appl Genet (2004) 109: DOI /s x ORIGINAL PAPER C. J. Schelfhout. R. Snowdon. W. A. Cowling. J. M. Wroth A PCR based B-genome-specific marker in Brassica species Received: 19 January 2004 / Accepted: 27 April 2004 / Published online: 26 June 2004 # Springer-Verlag 2004 Abstract Previous hybridisation studies showed that the repetitive DNA sequence pbnbh35 from Brassica nigra (genome BB, 2n=16) bound specifically to the B-genome and not to the A- or C-genomes of Brassica species. We amplified a sub-fragment of pbnbh35 from B. nigra by PCR, cloned and sequenced this sub-fragment, and confirmed that it was a 329-bp sub-fragment of pbnbh35. PCR and hybridisation techniques were used to confirm that the pbnbh35 sub-fragment was Brassica B-genome-specific. Fluorescence in situ hybridisation (FISH) in B. nigra, B. juncea (AABB, 2n=36) and B. napus (AACC, 2n=38) showed that the pbnbh35 subfragment was present on all eight Brassica B-genome chromosomes and absent from the A- and C-genome chromosomes. The pbnbh35 repeat was localised to the centromeric region of each B-genome chromosome. FISH clearly distinguished the B-genome chromosomes from the A-genome chromosomes in the amphidiploid species B. juncea. This is the first known report of a B-genome repetitive marker that is present on all B-genome chromosomes. It will be a useful tool for the detection of B chromosomes in interspecific hybrids and may prove useful for phylogenetic studies in Brassica species. Communicated by H. C. Becker C. J. Schelfhout (*). W. A. Cowling. J. M. Wroth School of Plant Biology, Faculty of Natural and Agricultural Sciences, The University of Western Australia, 35 Stirling Hwy, Crawley, WA, 6009, Australia cschelfh@agric.uwa.edu.au Tel.: Fax: R. Snowdon Institut für Pflanzenbau und Pflanzenzüchtung I, Justus Liebig Universitat, Heinrich-Buff-Ring 26-32, Giessen, Germany W. A. Cowling Canola Breeders Western Australia Proprietary Limited, 15/219 Canning Highway, South Perth, WA, 6151, Australia Introduction Large regions of eukaryote genomes are characterised by repetitive DNA sequences (Flavell 1980; Jelinek and Schmid 1982) which generally appear in one of several different forms tandemly repeated sequences, retroelements and other unique classes such as telomeric sequences or rdna units (Heslop-Harrison 2000). These repetitive DNA sequences have been useful as markers in phylogenetic studies (Halldén et al. 1987) as repetitive sequence motifs tend to be highly conserved within species but vary across species (Heslop-Harrison 2000). Species-specific sequences have been identified in Brassica species, and homology often exists between A- and C-genome repetitive sequences while there are fewer similarities between B-genome and A- or C-genome sequences (Hosaka et al. 1990; Chèvre et al. 1991). This is consistent with the results of genomic studies of Brassica species that have identified close relationships between the A- and C-genomes and more distant associations with the B-genome (Song et al. 1988; Quiros et al. 1991; Truco et al. 1996). Prior to this communication no Brassica B-genome specific sequences had been shown to be present on all eight B-genome chromosomes. Gupta et al. (1992) cloned two B. nigra genome-specific sequences, one of which pbnbh35 proved to be highly B-genome specific. Southern hybridisation analysis showed that this 496-bp sequence hybridised to all species carrying the Brassica B- genome (Gupta et al. 1992). Southern hybridisation analysis suggested that pbnbh35 is a highly dispersed sequence. Kapila et al. (1996) showed that pbnbh35 was present on five monosomic addition lines of B. nigra in a B. napus background (Chèvre et al. 1991), however they were unable to identify this sequence on all eight B. nigra chromosomes, and there have been no further studies reported. The objective of the investigation reported here was to convert pbnbh35 into a PCR-based marker and confirm its specificity to all eight chromosomes of the Brassica B- genome using PCR, DNA hybridisation and in situ

37 918 hybridisation techniques. The potential use of this sequence as a B-genome-specific marker is discussed. Materials and methods Plant material The plant material used in this study is shown in Table 1. Seeds were sown in pots containing a standard potting mix and slow-release fertiliser and were propagated in an air-conditioned glasshouse at C. At the six-leaf stage, the youngest two leaves from each plant were collected, snap frozen in liquid nitrogen and stored at 80 C. Total genomic DNA extraction and quantification Sufficient leaf tissue to half-fill a 1.5-ml tube was removed from 80 C storage and macerated in 600 μl of DNA extraction buffer (10 g l 1 N-lauroyl-sarcosine, 3.2 g l 1 EDTA, 12.1 g l 1 Trizma base, 12.6 g l 1 sodium sulfite, 5.8 g l 1 sodium chloride, ph 8.5). The tubes were kept on ice for 10 min, after which 600 μl of phenol: chloroform:iso-amyl alcohol (25:24:1) was added. The contents were mixed for 1 min and centrifuged at 13,000 rpm for 10 min. The supernatant was transferred to a fresh tube along with 40 μl of3m sodium acetate and 250 μl isopropanol, and the tubes were inverted to encourage precipitation. After 10 min at room temperature the tubes were centrifuged at 13,000 rpm for 10 min. The supernatant was discarded and the DNA pellet washed with 500 μl 70% ethanol, then centrifuged at 13,000 rpm for 5 min. Excess ethanol was decanted and the DNA pellet gently vacuum-dried for 10 min. DNA was resuspended in 100 μl of R40 (2 μl RNase A in 1 ml TE buffer). Total genomic DNA concentration was estimated against a DNA mass ladder (Gibco, Gaithersburg, Md.) by electrophoresis on a 1.5% agarose gel followed by ethidium bromide staining and UV light visualisation. PCR assay Primers were designed to amplify a sub-fragment of the pbnbh35 sequence published by Gupta et al. (1992) (Fig. 1). A PCR protocol was optimised to amplify this sequence in species containing the Brassica B-genome. Template DNA (1 μl of 100 ng/μl) from each of the respective Brassica species (Table 1) was added to 24 μl of master mix [2.5 μl 10 reaction buffer (Sigma, St. Louis, Mo.), 4.0 μl 1.25 mm dntps (Promega, Madison, Wis.), 1 μl 10mM Primer 1 (Geneworks), 1 μl 10mM Primer 2 (Geneworks), 1 μl REDTaq polymerase (Sigma) and 15.3 μl MilliQ water]. The PCR analyses were run on the following program in a Hybaid thermocycler: one cycle of 4 min at 94 C; 30 cycles of 1 min at 94 C, 2 min at 60 C, 2 min at 72 C; a final hold at 4 C. Fig. 1 pbnbh35, a 496-bp B-genome-specific sequence published by Gupta et al. (1992) with primers designed (bold) to amplify a 329-bp sub-fragment Sequence analysis The size of the PCR products was determined by electrophoresis on a 1.5% agarose gel, ethidium bromide staining and visualisation under UV light. The PCR products were isolated from the gel and cleaned on PCR cleanup columns (Roche, Indianapolis, Ind.). Cleaned samples were again checked for size, then ligated into a pgem T-Easy (Promega) plasmid vector and transformed into Escherichia coli-competent cells (JB 109, Promega) before overnight cultivation in LB media at 32 C. A sample of the clone containing the pbnbh35 sub-fragment was stored at 80 C in glycerol stocks. Sequence analysis was conducted on two independently cloned PCR products isolated from B. nigra. Sequencing was performed using the BigDye terminator protocol (Applied Biosystems, Foster City, Calif.) and sequence comparisons made using Vector NTI software. Homology of the amplified subfragment of pbnbh35 to published DNA sequences was examined using a BLASTN database query (Web Angis-BioManager website: on the Genebank main and EST databases. Specificity of PCR product To check the B-genome specificity of pbnbh35, DNA was extracted from each Brassica species in U s triangle (U 1935) (Table 1) and used as template DNA in the PCR-based assay described above. The cleaned PCR products (10 μl) were screened on a 1.5% agarose gel as described above. Fragment size was estimated against a 100-bp ladder (Promega). Slot blot hybridisation assay Genomic DNA (20 ng) from B. juncea and B. napus was denatured in an equal volume of 0.4 N sodium hydroxide and blotted onto Hybond N+ nylon membrane that had been pre-soaked in 2 SSC Table 1 Brassica species used in this study Brassica species Variety name Genome description B. rapa Pak Choy (Chinese cabbage) 2n=20 (AA) B. nigra Black mustard (ATC 90745) 2n=16 (BB) B. oleracea Cabbage 2n=18 (CC) B. napus Canola cv. Mystic 2n=38 (AACC) B. juncea Indian mustard (line JN29) 2n=36 (AABB) B. carinata Ethiopian mustard 2n=34 (BBCC)

38 919 (0.3 M NaCl, 0.03 M Na Citrate) using a Bio-Rad (Hercules, Calif.) slot blot apparatus. Following blotting, the membrane was rinsed in 2 SSC, blotted again and dried at 60 C for 30 min. Membranes were pre-hybridised for 3 4 h at 65 C in the pre-hybridisation solution [per 30-ml tube: 0.3 g dextran sulphate, ml sterile water, 6 ml of 20 SETS (3 M NaCl, 20 mm EDTA, 0.6 M Tris-HCl, 11 mm tetra-sodium pyrophosphate, ph 8), 6 ml 50 Denhardt s solution and 3 ml of 10% sodium dodecyl sulphate (SDS)]. The pbnbh35 sub-fragment (100 ng) was used as a probe in the hybridisation. It was prepared by denaturation for 10 min at 100 C with 1.2 μl random primers (50 μg ml 1 ) and immediate quenching on ice. This solution was then combined with 3.2 μl sterile water, 2.0 μl oligo buffer (Promega), 0.6 μl Klenow (Promega) at 5 U μl 1 and 3 μl [ 32 P]-dCTP (20 MBq) and incubated at 37 C for 2 h. Following the incubation period 30 μl 0.8 M EDTA was added to stop the labelling reaction, and the contents of the tube were briefly centrifuged. The probe solution was then centrifuged for 10 min at 10,000 rpm in a sephadex spin column to remove unincorporated nucleotides. The pre-hybridisation solution was replaced with the hybridisation solution (per tube: 0.5 g dextran sulphate, 2.7 ml sterile water, 1 ml SETS, 1 ml 50 Denhardt s and 50 μl 10% SDS). The probe solution was added and the membrane hybridised overnight at 65 C. The following day the hybridisation solution was discarded, and the membrane was given two 15-min washes in 2 SSC. Membranes were exposed to film (Fuji medical X-ray) for approximately 4 days at 80 C and developed in an auto-developer (All-pro imaging). The slot blot matrix was crossed-checked for the presence of bands. Fluorescence in situ hybridisation (FISH) assay Binding of the amplified sub-fragment (pbnbh35-sub) to chromosomes within the Brassica complex was examined by FISH in order to confirm its B-genome specificity and distribution within B- genome chromosomes of B. juncea. Four-day-old root tips were obtained from Brassica species (Table 1) for observation of mitotic chromosomes. Seeds were germinated on canola germination media [MS salts + B5 vitamins, 3% (w/v) sucrose, 0.8% purified agar and myo-inositol (100 mg l 1 ] at 25 C for 3 days and moved to 4 C for 24 h prior to the excision of 1-cm-long root tips. To accumulate metaphases we placed the excised root tips in 2 mm 8-hydroxyquinoline for 2 h at room temperature on a shaker and then transferred them to 4 C for 2 h without shaking. The root tips were blotted dry on filter paper and transferred to Farmer s fixative (3:1, ethanol:acetic acid). After 24 h, the root tips were transferred to fresh Farmer s fixative and stored at 20 C. Root tips were transferred to 100% ethanol for long-term storage at ambient temperatures. To make chromosome preparations we removed the root tips from the ethanol and washed them briefly in sterile distilled water. Fifteen root tips were selected from each Brassica species (Table 1). Onemillimeter segments of the root tip were excised and transferred to 50 μl enzyme solution consisting of 2% (w/v) cellulase (Calbiochem) and 20% (v/v) pectinase (Sigma) dissolved in enzyme buffer (40 mm citric acid, 60 mm tri-sodium citrate, ph 4.8; filter-sterilised through a 0.2-μm filter). The material was digested for 2 h at 37 C. The enzyme solution was then replaced with 50 μl hypotonic solution (75 mm KCl), renewed once and left at room temperature for 40 min. The hypotonic solution was replaced with 50 μl 60% acetic acid, renewed once and left for 25 min at room temperature, with gentle mixing every few minutes. The second acetic acid solution was replaced with 50 μl Farmer s fixative and renewed once. The root tips were gently sheared with a pipette to separate cells, and μl of the cell suspension was loaded onto precleaned cold microscope slides. We immediately added 30 μl of Farmer s fixative, and the slides were allowed to air dry slowly. The slides were viewed by phase-contrast microscopy to identify suitable metaphase stage cells for hybridisation. For preparation of the FISH probe the pbnbh35 sub-sequence was directly labelled with the fluorochrome Cy3-11-dUTP (Amersham Life Science, UK) via PCR amplification. Unincorporated fluorochrome was removed using a PCR cleanup column (Qiagen, Valencia, Calif.), and the labelled probe was eluted at a concentration of approximately 10 ng μl 1 in a hybridisation solution containing 50% formamide, 2 SSC and 10% dextran sulphate. Approximately 15 μl of probe solution was applied to the area of the slide to be hybridised, covered with a mm cover slip and sealed with rubber cement. Denaturation was achieved by incubation at 80 C for 4 min on a heated metal plate. The slides were immediately transferred to 37 C for overnight hybridisation. The following day the rubber seal and cover slip were carefully removed and the slides washed at 42 C for 5 min in 2 SSC followed by two washes for 5 min in 0.2 SSC and a final wash of 5 min in 2 SSC. The slides were stored in 4 SSC containing 0.5% Tween prior to staining. DAPI-Antifade (20 μl) (Appligene-Oncor) was added to each slide, covered with a cover slip, left for 2 min, then squashed under a paper towel to remove excess stain. Slides were stored in the dark until viewing. They were viewed under a Leica DM-R fluorescence microscope with a single-bypass and photographed with a Cohu 4912 uncooled CCD camera. Individual images were merged using the Leica Q-FISH software. No image manipulation was necessary. Results Sequence analysis of the two independently cloned subfragments of pbnbh35, amplified by PCR from B. nigra, confirmed their homology with the sequence published by Gupta et al. (1992) (Fig. 1). When aligned using VECTOR NTI software both of the analysed sequences showed 92% homology with pbnbh35. The BLASTN programme indicated no homology greater than 10% with any other sequence in the Genebank main and EST databases. The PCR product showed B-genome specificity Fig. 2 PCR based assay of six Brassica species with different genome compositions. L 100-bp ladder, AA Brassica rapa, BB B. nigra, CCB. oleracea, ABB. juncea, ACB. napus, BCB. carinata, N negative control

39 920 when amplified from several Brassica species from U s triangle (U 1935) (Fig. 2). The pbnbh35 sub-fragment was amplified only from those samples where the template DNA contained the Brassica B-genome (B. nigra, B. juncea and B. carinata) (Fig. 2). A single band, equivalent to the expected size of 329 bp from pbnbh35 (Fig. 1), was generated in each Brassica species with the B-genome (Fig. 2). Slot blot hybridisation of the pbnbh35 sub-fragment gave positive signals with B. juncea (AABB) but not with B. napus (AACC) (data not shown). Results from in situ hybridisation with the pbnbh35- sub-fragment are shown in Figs. 3 and 4. The molecular cytogenetic results supported the results of the PCR assay and the Southern hybridisations. Strong FISH signals from the pbnbh35 sub-fragment were observed on B. nigra chromosomes, whereas no signals were seen in B. napus. Strong hybridisation signals were also observed on 16 chromosomes in B. juncea (2n=36), presumably corresponding to the 16 B-genome chromosomes. No pbnbh35 sub-fragment signals were observed in the remaining 20 chromosomes of B. juncea. The pbnbh35 repeat was localised in large blocks surrounding the centromeres of all B-genome chromosomes, with a gap often visible at the centromere (Fig. 3). Signals did not extend to the telomeric regions. Fig. 4 B. napus probed with Cy3-labelled, B-genome-specific sequence. pbnbh35: no red labels Discussion Fig. 3 Fluorescence in situ hybridisation to B. juncea mitotic metaphase chromosomes with Cy3-labelled B-genome-specific sequence pbnbh35 (red). Green signals show a FITC-labelled A- genome-specific repeat sequence (Snowdon, unpublished results) that hybridises to four A-genome chromosomes in B. juncea. Chromosomes are counterstained with the blue fluorescent dye DAPI FISH has become the method of choice for analysis of the chromosomal distribution of repetitive DNA sequence elements in plant genomes (Heslop-Harrison 2000). In the investigation described here, FISH was used to investigate the previously unknown distribution of a Brassica B- genome repeat sequence on B-genome chromosomes. We confirmed that pbnbh35 is a repetitive sequence with B- genome specificity, as indicated by Gupta et al. (1992), and we converted the sequence into a PCR product that is amplified only in species containing the Brassica B- genome. The FISH assay showed that this sequence is localised in high-copy number on either side of the centromeres of all eight Brassica B-genome chromosomes and appears with low frequency or is absent in the interstitial and telomeric regions. This is the first report confirming a B-genome-specific marker that is distributed across all B-genome chromosomes. The PCR sub-fragment from pbnbh35 resolved as a unique single band on agarose gel at the expected size of 329 bp, and two independently cloned products from this band produced identical sequences with the predicted size of 329 bp. The pbnbh35 sub-fragment has the properties of a highly conserved dispersed tandem repeat sequence, and there was no evidence for sequence homology with any known retro-element. Strong signals surrounding the centromeres of the chromosomes indicate high-copy numbers in pericentromeric heterochromatin, however there are virtually no signals towards the telomeric regions. Such a distribution, frequently observed in genomic in situ hybridisation (GISH) with Brassica

40 921 hybrids (Fahleson et al. 1997; Skarzhinskaya et al. 1998; Snowdon et al. 2000), reflects the generally low number of dispersed repeat sequences in the interstitial and telomeric chromosome regions of Brassica and related genera (Heslop-Harrison and Schwarzacher 1996). A variety of applications can be envisaged for this PCRbased, B-genome-specific marker. The marker could be used for detecting the B-genome in the progeny of wide interspecific and intergeneric crosses where this has not been possible for studies in the past (Roy 1984; Prakash and Chopra 1988; Chèvre et al. 1991; Rao et al. 1993) where it may be present in whole chromosomes (addition lines) or introgressions. This marker would complement existing randomly amplified polymorphic DNA markers where linkage to B-genome alleles have been useful in aiding the selection of beneficial traits in Brassica wide crosses (Chèvre et al. 1997). Previous work has demonstrated the utility of GISH to distinguish Brassica B- genome chromosomes from those of the A- and C- genomes (Snowdon et al. 1997). However GISH gives only cytogenetic information on anonymous genomespecific repeat sequences. In contrast, the availability of a PCR-based tandemly repeated marker enables a combination of exact cytological characterisation by FISH with molecular genetic analyses. The latter can give considerably more insight into the presence or absence of a given sequence or its near homologues and enables a much more accurate estimation of repeat copy numbers. Furthermore, tandem repeats tend to remain phylogenetically well conserved after they are amplified to high-copy numbers in genomes. Hence this B-genome-specific marker will also be useful in comparative mapping and phylogenetic studies among the Brassicaceae. For example, there has been speculation from cytological studies that B. nigra might be more closely related to Sinapis species than it is to the Brassica A- and C-genome diploids. The availability of PCR-based repeat sequence markers will allow detailed investigation of the sequences and distributions of genome-specific tandem repeats throughout different crucifer genera and give new molecular phylogenetic information on Brassica genome evolution. Acknowledgements This research was funded by an Australian Research Council Strategic Partnership Industry Research Training grant with co-funding provided by the Export Grains Centre Ltd and the Council of Grain Grower Organisations Ltd. We thank Michael Francki and Matthew Nelson for support and advice in various stages of this research. References Chèvre AM, This P, Eber F, Deschamps M, Renard M, Delseny M, Quiros CF (1991) Characterization of disomic addition lines Brassica napus Brassica nigra by isozyme, fatty acid, and RFLP markers. Theor Appl Genet 81:43 49 Chèvre AM, Barret P, Eber F, Dupuy P, Brun H, Tanguy X, Renard M (1997) Selection of stable Brassica napus B. juncea recombinant lines resistant to blackleg (Leptoshaeria maculans). 1. Identification of molecular markers, chromosomal and genomic origin of the introgression. Theor Appl Genet 95: Fahleson J, Lagercrantz U, Mouras A, Glimelius K (1997) Characterization of somatic hybrids between Brassica napus and Eruca sativa using species-specific repetitive sequences and genomic in situ hybridization. Plant Sci 123: Flavell R (1980) The molecular characterization and organization of plant chromosomal DNA sequences. Annu Rev Plant Physiol 31: Gupta V, Lakshmisita G, Shaila MS, Jagannathan V, Lakshmikumaran MS (1992) Characterization of species-specific repeated DNA sequences from B. nigra. Theor Appl Genet 84: Halldén C, Bryngelsson T, Sall T, Gustafsson M (1987) Distribution and evolution of a tandemly repeated DNA sequence in the family Brassicaceae. J Mol Evol 25: Heslop-Harrison JS (2000) Comparative genome organization in plants: from sequence and markers to chromatin and chromosomes. Plant Cell 12: Heslop-Harrison JS, Schwarzacher T (1996) Genomic southern and in situ hybridization for plant genome analysis. In: Jauhar PP (ed) Methods of genome analysis in plants. CRC Press, Boca Raton, pp Hosaka K, Kianian SF, McGrath JM, Quiros CF (1990) Development and chromosomal localization of genome-specific DNA markers of Brassica and the evolution of amphidiploids and n=9 diploid species. Genome 33: Jelinek WR, Schmid CW (1982) Repetitive sequences in eukaryotic DNA and their expression. Annu Rev Biochem 51: Kapila R, Negi MS, This P, Delseny M, Srivastava PS, Lakshmikumaran M (1996) A new family of dispersed repeats from Brassica nigra: characterization and localization. Theor Appl Genet 93: Prakash S, Chopra l (1988) Introgression of resistance to shattering in Brassica napus from Brassica juncea through non-homologous recombination. Plant Breed 101: Quiros CF, Hu J, This P, Chèvre AM, Delseny M (1991) Development and chromosomal localization of genome-specific markers by polymerase chain reaction in Brassica. Theor Appl Genet 82: Rao MVB, Babu VR, Radhika K (1993) Introgression of earliness in Brassica napus L. I. An interspecific B. juncea and B. napus cross. Int J Trop Agric 11:14 19 Roy NN (1984) Interspecific transfer of Brassica juncea-type high blackleg resistance to Brassica napus. Euphytica 33: Skarzhinskaya M, Fahleson J, Glimelius K, Mouras A (1998) Genome organisation of Brassica napus and Lesquerella fendleri and analysis of their somatic hybrids using genomic in situ hybridization. Genome 41: Snowdon RJ, Köhler W, Friedt W, Köhler A (1997) Genomic in situ hybridization in Brassica amphidiploids and interspecific hybrids. Theor Appl Genet 95: Snowdon RJ, Winter H, Diestel A, Sacristán MD (2000) Development and characterization of Brassica napus Sinapis arvensis addition lines exhibiting resistance to Leptosphaeria maculans. Theor Appl Genet 101: Song KM, Osborn TC, Williams PH (1988) Brassica taxonomy based on nuclear restriction fragment length polymorphisms (RFLPs) 1. Genome evolution of diploid and amphidiploid species. Theor Appl Genet 75: Truco MJ, Hu J, Sadowski J, Quiros CF (1996) Inter- and intragenomic homology of the Brassica genomes: implications for their origin and evolution. Theor Appl Genet 93: U N (1935) Genome-analysis in Brassica with special reference to the experimental formation of B. napus and peculiar mode of fertilisation. Jpn J Bot 7:

41 Chapter 4 Published paper Schelfhout CJ, Snowdon R, Cowling WA, Wroth JM (2006) Tracing B-genome chromatin in Brassica napus x B. juncea interspecific progeny. Genome 49,

42 1490 Tracing B-genome chromatin in Brassica napus B. juncea interspecific progeny C.J. Schelfhout, R. Snowdon, W.A. Cowling, and J.M. Wroth Abstract: We used polymerase chain reaction (PCR) and fluorescence in situ hybridization (FISH) techniques to demonstrate the presence of Brassica B-genome chromosomes and putative B-genome introgressions in B. napus B. juncea interspecific progeny. The B-genome - specific repeat sequence pbnbh35 was used to generate PCR products and FISH probes. The highest frequencies of viable progeny were obtained when B. napus was the maternal parent of the interspecific hybrid and the first backcross. B-genome - positive PCR assays were found in 34/51 fertile F 2 progeny (67%), which was more than double the proportion found in fertile BC 1 progeny. Four B-genome - positive F 2-derived families and 1 BC 1-derived family were fixed or segregating for B. juncea morphology in the F 4 and BC 1S 2, respectively, but in only 2 of these families did B. juncea-type plants exhibit B. juncea chromosome count (2n = 36) and typical B-genome FISH signals on 16 chromosomes. The remaining B. juncea-type plants had B. napus chromosome count (2n = 38) and no B- genome FISH signals, except for 1 exceptional F 4-derived line that exhibited isolated and weak B-genome FISH signals on 11 chromosomes and typical A-genome FISH signals. B. juncea morphology was associated with B-genome - positive PCR signals but not necessarily with 16 intact B-genome chromosomes as detected by FISH. B-genome chromosomes tend to be eliminated during selfing or backcrossing after crossing B. juncea with B. napus, and selection of lines containing B-genome chromatin during early generations would be promoted by use of this B-genome repetitive marker. Key words: B genome, introgression, FISH, PCR, Brassica napus, Brassica juncea, canola, oilseed rape. Résumé : Les auteurs ont employé des techniques PCR et de l hybridation in situ en fluorescence (FISH) pour démontrer la présence de chromosomes du génome B de Brassica et d introgressions putatives de tels chromosomes dans des progénitures interspécifiques B. napus x B. juncea. Laséquence répétitive pbnbh35, spécifique du génome B, a été employée pour générer des produits PCR et des sondes FISH. Les plus grandes fréquences de progénitures viables ont été obtenues lorsque le B. napus était employé comme parent maternel de l hybride interspécifique et du premier rétrocroisement. Des résultats positifs en PCR ont été obtenus pour 34/51 (67 %) de la progéniture F 2 fertile, ce qui est plus de deux fois la proportion trouvée au sein de la progéniture BC 1 fertile. Quatre familles dérivées en F 2 et une dérivée enbc 1, toutes positives pour le génome B, étaient fixées ou en ségrégation pour la morphologie de type B. juncea au sein de la F 4 et de la BC 1S 2 respectivement. Dans seulement deux de ces familles, les plantes de type B. juncea affichaient le nombre chromosomique attendu (2n = 36) et des signaux FISH typiques du génome B sur 16 chromosomes. Les autres plantes de type B. juncea présentaient un nombre chromosomique typique du B. napus (2n = 38) et aucun signal FISH du génome B, à l exception d une plante dérivée enf 4 qui montrait des signaux FISH faibles et isolés sur 11 chromosomes et des signaux FISH typiques du génome A. La morphologie B. juncea était associée àdes produits PCR spécifiques du génome B, mais pas nécessairement avec 16 chromosomes intacts du génome B tels que détectés par FISH. Les chromosomes du génome B tendent à subir une élimination suite à l autofécondation ou au rétrocroisement dans les progénitures issues de croisements entre le B. juncea et le B. napus. Ce marqueur de l ADN répété du génome B pourrait faciliter la sélection de lignées contenant de la chromatine du génome B au cours des premières générations. Mots clés :génome B, introgression, FISH, PCR, Brassica napus, Brassica juncea, colza, colza oléagineux. [Traduit par la Rédaction] Received 13 March Accepted 9 August Published on the NRC Research Press Web site at on 30 January C.J. Schelfhout. School of Plant Biology, Faculty of Natural and Agricultural Sciences, The University of Western Australia, 35 Stirling Hwy, Crawley WA, 6009, Australia. R. Snowdon. Institut für Pflanzenbau und Pflanzenzüchtung I, Justus Liebig Universitat, Heinrich-Buff-Ring 26-32, D Giessen, Germany. W.A. Cowling 1 and J.M. Wroth. School of Plant Biology, Faculty of Natural and Agricultural Sciences, The University of Western Australia, 35 Stirling Hwy, Crawley WA, 6009, Australia; Canola Breeders Western Australia Pty Ltd, Locked Bag 888, Como WA 6952, Australia. 1 Corresponding author ( wcowling@cyllene.uwa.edu.au). Genome 49: (2006) doi: /g # 2006 NRC Canada

43 Schelfhout et al Introduction Stable introgression of Brassica B genome into rapeseed (Brassica napus L., also known as canola or oilseed rape) through wide crossing has been an objective for many rapeseed breeders as a means to improve the core germplasm. The Brassica B genome is associated with many valuable traits, such as nonshattering siliqua or disease resistance, which are not found in the A or C genomes of B. napus. Brassica juncea (L.) Czern. et Coss, an amphidiploid with A and B genomes, is often used as the source of B genome in interspecific crossing with B. napus. There have been several attempts to introgress resistance to blackleg disease (caused by the fungus Leptosphaeria maculans [Desm] Ces. & De Not.) from the B genome of B. juncea into B. napus (Roy 1984; Gerdemann-Knorck et al. 1995; Chèvre et al. 1997; Dixelius and Wahlberg 1999). Siliqua shatter resistance has been transferred from B. juncea into B. napus (Prakash and Chopra 1988). Problems with cross incompatibility and hybrid sterility have limited the success of Brassica interspecific hybridization, and successful crossing outcomes vary, depending on the species, subspecies, or varieties used (Roy 1980a). Strong reciprocal effects were observed in Brassica interspecific crosses (Roy 1980a) and seed set was higher when B. juncea was used as the maternal parent in crosses with B. napus. While B. juncea was the more successful maternal parent, Roy (1980b) reported a very low frequency of progeny with B. juncea morphology from this cross. Cytogenetic and molecular studies of the Brassica genomes have supported the view that the 3 diploid species (B. rapa L., B. nigra [L.] Koch, and B. oleracea L.) evolved from a common progenitor species (Warwick and Black 1991; Prakash and Chopra 1993). However, contrasting molecular studies (Song et al. 1988; Yanagino et al. 1987; Truco et al. 1996) suggest a biphyletic origin of the diploid Brassicas, in which B. rapa (A genome) and B. oleracea (C genome) evolved from 1 lineage, and B. nigra (B genome) evolved separately along another lineage. The A and C chromosomes of B. napus pair readily with their respective A and C homologues in B. oleracea and B. rapa (Parkin and Lydiate 1997). Support for the biphyletic origin was provided by Axelsson et al. (2000) based on the behaviour of the A and B genomes in B. juncea interspecific crosses. Nevertheless, the 3 diploid Brassica spp. are united in lineage back to an Arabidopsis progenitor million years ago, and up to 90% of the B. napus genome can be traced to 21 conserved blocks from the Arabidopsis genome (Parkin et al. 2005). The Brassica B genome appears to be excluded in favour of homologous and homoeologous pairing of A and C genomes in interspecific crosses among Brassica species (Meng et al. 1998). The B genome may be closer to the C genome than the A genome, as the substitution of B-genome chromosomes by C-genome chromosomes in interspecific progeny has been observed (Banga 1988). Attia and Röbbelen (1986) observed similar trends in chromosomal behaviour among the 3 genomes but also noted that the perceived distance of the B genome may be a product of genic control. Detection of alien DNA in wide crosses has been achieved by quantification of chromosome content by flow cytometry (Sabharwal and Dolezel 1993) and tracing of chromosome and DNA transfer with molecular markers (Quiros et al. 1991; Chèvre et al. 1991). Visualization of alien chromatin in interspecific hybrids using in situ hybridization techniques (Fahleson et al. 1997; Skarzhinskaya et al. 1998; Snowdon et al. 1997, 2000) enables pinpointing of introgressions to specific chromosomes. A B-genomespecific repeat sequence, pbnbh35 (Gupta et al. 1992), was developed to trace and identify B genome using polymerase chain reaction (PCR) and fluorescence in situ hybridization (FISH) (Schelfhout et al. 2004). The aim of this study was to demonstrate the potential of pbnbh35 (Schelfhout et al. 2004) to act as a marker for B- genome chromosomes and chromosomal introgressions in the self and backcross progeny of an interspecific hybrid population between B. napus and B. juncea. The marker would help to enrich the population for B genome and counter the tendency for B chromosomes to be eliminated in early generations (Meng et al. 1998). The potential for introgression of B-genome chromatin into the A and C genomes of B. napus would be improved by selecting earlygeneration progeny with B-genome chromosomes. We used PCR amplification and FISH of pbnbh35 to detect B- genome chromatin in F 2 and BC 1 progeny and in subsequent generations. We discuss the stability and fate of B-genome chromosomes and putative introgressions in these interspecific hybrids. Materials and methods Production of interspecific hybrids Reciprocal crosses 99X022 and 99X055 were made between B. napus canola Mystic and B. juncea near-canola quality line JN29 in 1999 (Fig. 1). Twelve viable F 1 interspecific hybrid seeds from each cross were sown under glasshouse conditions in January The F 1 lines were self-pollinated by hand to generate F 2 seed or backcrossed to both parents by hand-pollination in reciprocal combinations (Fig. 1). More than 1000 F 2 and BC 1 seeds were sown in seedling trays under glasshouse conditions in July 2000, of which 725 germinated (Table 1). Leaf tissue was taken from each seedling after 3 weeks and immediately stored at 80 8C for later use of DNA in PCR amplification studies. After 4 weeks, seedlings were transplanted directly into soil in an insect-proof screenhouse. Plants were maintained with slow release fertilizer and controlled irrigation. Before flowering the primary inflorescence of every plant was covered with a pollination bag, and bags were regularly shaken to promote self-pollination. Fertile F 2 and BC 1 plants were harvested at the end of 2000 (total 358, Table 1), and those that had sufficient seed were sown in rows (20 seeds per row, with 2 replicates) in the field under normal agronomic management for canola at Shenton Park and Merredin, Western Australia, in June There were 59 F 2 -derived F 3 and 229 BC 1 S 1 lines in the field in Agronomically desirable plants were bagged at flowering to prevent crosspollination. The presence of B. napus morphology and B. juncea morphology was noted within rows. Lines were assessed for agronomic traits, blackleg disease resistance (measured as percentage # 2006 NRC Canada

44 1492 Genome Vol. 49, 2006 Fig. 1. Breeding strategy with all possible reciprocal combinations in the cross B. napus Mystic B. juncea JN29, including BC 1 families and subsequent self-fertilized generations. In the first cross, the maternal parent is listed first. In the backcrosses, (,) denotes the maternal parent and (<) denotes the paternal parent. Table 1. F 2 and BC 1 progeny from the interspecific cross Brassica napus Mystic B. juncea JN29 : number of viable seeds and fertile plants generated from the 2 reciprocal crosses and 8 backcross families. Cross pedigree Cross number No. of F 1 seeds harvested Jan F 2 or BC 1 No. of seeds harvested June 2000 No. of viable seeds germinated July 2000 No. of fertile plants with mature seed Dec Parents B. napus Mystic (Mys) B. juncea JN B. napus, primary hybrid Self (Mys/JN29) 99X022 > BC 1 (Mys//Mys/JN29) 00X BC 1 (Mys/JN29//Mys) 00X BC 1 (Mys/JN29//JN29) 00X <0.1 BC 1 (JN29//Mys/JN29) 00X Total progeny B. napus, B. juncea, primary hybrid Self (JN29/Mys) 99X055 > BC 1 (JN29/Mys//JN29) 00X BC 1 (JN29//JN29/Mys) 00X BC 1 (JN29/Mys//Mys) 00X BC 1 (Mys//JN29/Mys) 00X Total progeny B. juncea, Note: Refer to Fig. 1 for an explanation of cross numbers. Mean mass of harvested seed (g/plant) plant survival under field disease conditions), seed oil content, protein, and total glucosinolate content. Single plants and row-bulks with near-canola quality were advanced to the F 4 and BC 1 S 2 generation for further field testing at Shenton Park in Total genomic DNA extraction and quantification A rapid DNA extraction protocol was used to isolate total genomic DNA from leaf material of interspecific progeny for the PCR assays (Schelfhout et al. 2004). PCR assay Frozen leaf material from fertile F 2 and BC 1 interspecific progeny, grown in the field in 2001 as F 3 and BC 1 S 1 rows, was screened using a Brassica B-genome - specific marker following the protocol described in Schelfhout et al. (2004). DNA was isolated from 248 of the 288 fertile progeny that showed reasonable agronomic attributes in PCR products were screened on 1.5% agarose gels to test for the presence of a single band at 329 bp that was confirmed to be the B-genome - specific marker pbnbh35 (Schelfhout et al. # 2006 NRC Canada

45 Schelfhout et al ). Positive (B. juncea JN29 ) and negative (B. napus Mystic ) control reactions along with a size marker ladder were included in all electrophoresis gel assays to confirm that the PCR fragment was the B-genome - specific pbnbh35 subfragment. FISH assay Ten PCR-positive lines were selected from F 2 -derived families and 18 from BC 1 -derived families for FISH assay. The F 2 -derived families were tested as F 2 -derived F 4 bulks or F 4 -derived F 5 single plant-selections. The BC 1 -derived families were tested as BC 1 S 2 bulks or BC 1 S 2 -derived BC 1 S 3 single-plant selections. The families selected for FISH assay included those segregating for B. napus and B. juncea morphology, or those with outstanding or unusual agronomic or seed quality attributes. The B. napus and B. juncea parents were included as controls. FISH analysis followed the protocol described in Schelfhout et al. (2004). The pbnbh35 subsequence probe was directly labelled with the fluorochrome Cy3 11-dUTP (Amersham Biosciences, Piscataway, N.J.) by PCR. An A- genome - specific repeat sequence that hybridizes to between 4 and 6 A-genome chromosomes (R. Snowdon, unpublished results) was directly labelled with FITC-11-dUPT (Amersham Life Science) and included in the hybridization mix as a control FISH reaction and to confirm the presence of A-genome chromosomes. No cell debris or cytoplasm was present in the preparations, and no unlabelled blocking DNA was required. Chromosomes were counterstained with 4,6-diamidino-2-phenylindole, and fluorescence was visualized using a Leica DM-R microscope (Leica Microsystems, Wetzlar, Germany). At least 5 well-hybridized metaphases, free from nonspecific background hybridization signals, were examined for each plant investigated. Digital images were recorded using a Cohu 4912 uncooled CCD camera (Cohu Inc., San Diego, Calif.) and Leica QFISH software. Results Production of interspecific progeny The majority of F 1 seeds from the reciprocal crosses between B. napus B. juncea germinated precociously in the pods before seed maturity, but several viable F 1 seeds were harvested from each cross (Table 1). The F 1 hybrid plants were clearly distinguished from parent plants by the presence of galls at the root shoot junction and other morphological differences, including poor seed set and infertility. B. napus Mystic was the more successful maternal parent (cross 99X022) with more viable F 1 seeds and more fertile F 2 plants than B. juncea JN29 as the maternal parent (cross 99X055) (Table 1). There were more failed self-pollinations and siliqua with aborted seeds in cross 99X055. Only 1 F 3 line from 99X055 survived to the F 4 generation for further testing (99X S). BC 1 families gave similar results: B. napus Mystic and 99X022-F 1 were more successful maternal parents than B. juncea JN29 and 99X055-F 1 (Table 1). BC 1 families with lower fertility also produced seed with lower viability. Only 4/8 backcrosses produced viable BC 1 S 1 seed for field trials in 2001, and 3 of these involved B. napus or 99X022- F 1 as the maternal parent to produce the F 1 or BC 1 (Table 1). Only 3 families expressed greater than 33% fertility (00X011-BC 1, 00X012-BC 1, and 99X022-F 2 ). For those families that survived to BC 1 S 1 and F 3, the fertility rate was high from that generation onwards. Five of the 51 F 2:3 progeny of 99X022 and 7 of the 194 BC 1 S 1 progeny of 00X011 and 00X012 in the field in 2001 contained plants with B. juncea morphology. The remaining families had stable B. napus morphology. The sole surviving F 2 - derived progeny of 99X055 (B. juncea as maternal parent) contained plants with B. juncea and B. napus morphology. Four of the selected F 4 families and 1 of the selected BC 1 S 2 families segregated for B. juncea morphology in the field in 2002 (Table 2). PCR-based assay A single band at 329 bp was confirmed to be the B- genome - specific PCR fragment pbnbh35 by sequence analysis in previous work (Schelfhout et al. 2004) and by its presence in the positive control (B. juncea JN29 ) and absence in the negative (B. napus Mystic ) control reactions. The fragment was also present in other B-genome - containing Brassica species (as shown in fig. 2 of Schelfhout et al. 2004). The proportion of fertile F 2 or BC 1 progeny with B genome detected by the PCR-based assay was lower in progeny of backcrosses 00X011 (23 of 80 lines, or 29%) and 00X012 (29 of 114 lines, or 25%) than in the selfed progeny of line 99X022 (34 of 51 lines, or 67%). In all these crosses, B. napus was the maternal parent in the interspecific hybrid (Table 1). B-genome reactions were detected in each of the progeny when B. juncea was the maternal parent, in backcross 00X013 (2 lines) and F 2 progeny of 99X055 (1 line) (Table 1). FISH assay The B. juncea and B. napus parents ( JN29 and Mystic, respectively) showed FISH signals typical of the species. JN29 was characterized by 16 chromosomes with B- genome FISH signals that were strong around the centromere but weak in the interstitial regions. Mystic showed the A-genome - specific repeat sequence in 4 6 chromosomes. B-genome FISH signals were detected in plants from 2 of the 10 selected F 2 -derived F 4 bulks (99X and 99X ) but in none of the 18 selected BC 1 - derived families. All plants from the F 2 -derived F 4 bulk and F 4 -derived F 5 lines from family 99X were typical B. juncea morphotypes with B. juncea chromosome count (2n = 36), positive PCR signals, and strong B-genome FISH signals on 16 chromosomes (for example, line 99X M2-SP02 in Fig. 2a). Plants in F 2 -derived F 4 progeny of 99X segregated for B. napus and B. juncea morphology and varied in FISH signals (Table 2): one plant (99X ) showed strong B-genome signals on 16 chromosomes and had typical B. juncea morphology and chromosome number (not pictured), whereas another (99X ) showed no B-genome signals, B. napus chromosome count, typical A-genome FISH signals, and B. napus morphology (Fig. 2b). One exceptional line, F 4 -derived F 5 line 99X M2- SP02, displayed B. juncea phenotypic traits, including leaf morphology, shatter-resistant siliqua, yellow seed coats, and # 2006 NRC Canada

46 1494 Genome Vol. 49, 2006 Table 2. FISH assay and chromosome count in parents, F 2-derived, and BC 1-derived families from the cross B. napus B. napus that were fixed or segregating for B. juncea morphology and were positive for B-genome - specific markers by PCR assay. Generation Family no. or cultivar Morphotype Blackleg resistance* B-genome PCR assay B-genome FISH assay Chromosome no. (2n) B. napus parent Mystic B. napus 38 B. juncea parent JN29 B. juncea (16) 36 BC 1-derived 00X Segregating B. napus and B. juncea + Segregating 38 and 36 F 2-derived 99X Segregating B. napus and B. juncea + + +/ Segregating 38 and 36 F 2-derived 99X B. juncea (16) 36 F 2-derived 99X B. juncea + 38 F 2-derived 99X Segregating B. napus and B. juncea + Segregating 38 and 36 *Assessment of blackleg resistance occurred in the field at Shenton Park in 2001 and 2002, and is summarized by + for moderate to high resistance and for susceptible. moderate blackleg resistance, but a diploid chromosome count of 38. B-genome FISH signals were observed on 11 chromosomes with the B-genome - specific pbnbh35 probe (Fig. 2c). Six chromosomes exhibited FISH signals with the A-genome - specific repeat sequence. This pattern of FISH signals was confirmed in several cells in the preparation. The line was not as resistant to blackleg as the B. juncea parent JN29. Another F 2 -derived F 4 family with B. juncea morphology (2n = 38), 99X , showed positive B- genome PCR but was negative for B-genome FISH signals and was not resistant to blackleg (Table 2). None of the PCR-positive BC 1 interspecific progeny showed visible B-genome introgressions with FISH, although some segregated for B. juncea morphological traits, for example, family 00X (Table 2). Discussion Substantial F 1 sterility was observed in the cross B. napus B. juncea in both cross directions, but in our case B. napus was the more successful maternal parent. This indicates unbalanced chromosome behaviour among the Brassica A, B, and C genomes especially when B. juncea was used as the maternal parent (Fig. 1, Table 2). In contrast to our observation, Roy (1980a) reported that B. juncea was more successful as the maternal parent than B. napus. However, like Roy (1980b), we observed a low frequency of interspecific progeny with B. juncea morphology. The favoured direction of the initial hybridization appears to be genotype specific, but from that point on the favoured chromosome complement is B. napus, and B-genome chromosomes tend to be discarded (Meng et al. 1998). Genotype-specific infertility in the interspecific hybrid plant may be caused by pre- or post-zygotic barriers at the stigma, style, or ovary. Luu et al. (1998) suggest that most interspecific pre-zygotic barriers in Brassicaceae occur within the style after the pollen grains have germinated. There may be further genetic factors that contribute to hybrid infertility, including interactions with cytoplasmic genetic factors of the maternal parent, and these interactions may be genotype dependent. B-genome chromatin was detected by PCR at twice the frequency in selfed progeny (67%) as in progeny backcrossed to B. napus (25% 29%). More opportunities for homoeologous pairing of the B genome with A or C genome occur with selfing. We conclude that selfing of earlygeneration progeny that react positively in the B-genome PCR assay, followed by selection of progeny that retain PCR-positive signals, is a promising way to increase the frequency of B-genome introgression into B. napus. Intercrossing PCR-positive sibling lines may be even more successful than selfing to achieve this goal. The strong homology between amphidiploid Brassica genomes and their diploid progenitors indicates that very little change has occurred since the formation of the amphidiploids in agricultural times (Parkin and Lydiate 1997; Axelsson et al. 2000). In the cross B. napus B. juncea, A- genome chromosomes are expected to pair normally at meiosis and are unlikely to pair with B- or C-genome chromosomes. Many of the unmatched B- and C-genome chromosomes (especially B chromosomes) will be lost during meiosis, as observed by Meng et al. (1998). Some nonpairing chromosomes will become incorporated into the gametes and be carried on to the following generation. C-genome chromosomes appear to be retained preferentially over B- genome chromosomes in this process. The frequency of introgression of B genome into the A or C genome may be increased if the selfed population is enriched for B genome through use of the B-genome - specific marker. Whole chromosome substitutions may occur, as has been observed previously in Brassica interspecific crosses (Banga 1988). It was expected that backcrossing to B. juncea would allow homologous pairing of the B-genome chromosomes in the F 1, thus favouring the retention of the B genome. As observed in this study, B genome was present in all 3 surviving backcrosses to B. juncea (00X015), but unfortunately fertility was low, and few of these lines survived past the BC 1 generation. It is not clear why viable seeds are rarely produced from backcrossing to B. juncea or why the predominant progeny have B. napus morphology and chromosome constitution, but similar observations have been made in the past (Roy 1980b). FISH was effective at identifying B. juncea-type progeny with a normal complement of B-genome chromosomes (Fig. 2a), but no B-genome FISH reactions were observed for the majority of B. napus-type progeny with positive B- # 2006 NRC Canada

47 Schelfhout et al genome PCR signals. One F 4 -derived F 5 progeny with B. juncea morphology (99X M2-SP02) showed positive B-genome FISH signals and putative B-genome introgressions on 11 chromosomes, as well as A-genome FISH signals and a chromosome complement of 2n = 38 (Fig. 2c). None of the chromosomes carried both A-genome - Fig. 2. (a) FISH patterns on prometaphase chromosomes of 99X022-58M2-SP02, an F 4-derived F 5 line of B. napus Mystic B. juncea JN29 with 16 B-genome chromosomes (labelled red with a probe derived from the B-genome - specific repeat sequence pbnbh35) and a total chromosome count of 2n = 36. The green probe hybridized specifically to 4 6 A-genome chromosomes. Plants showing patterns of this type were morphologically B. juncea type and tested positive for B-genome chromatin by the PCR assay. (b) FISH patterns on prometaphase chromosomes of 99X , a plant from an F 2-derived F 4 bulk of B. napus Mystic B. juncea JN29 with no B-genome chromosomes apparent by FISH assay but positive signals for pbnbh35 by PCR assay and a total chromosome count of 2n = 38. Four chromosomes are labelled with the green A-genome - specific probe (circled). Plants showing patterns of this type were morphologically B. napus type. (c) FISH patterns on prometaphase chromosomes of 99X M2-SP02, an F 4-derived F 5 plant from B. napus Mystic B. juncea JN29. B-genome - specific signals from the pbnbh35 subfragment (red) are visible on 11 chromosomes (circled). Six chromosomes carry A-genome - specific signals (green). The total chromosome count is 2n = 38, indicating a B. napus type with B- genome introgressions. However, the morphology of the plant resembled B. juncea. The bar in each photomicrograph represents 10 mm. specific and B-genome - specific FISH signals, which indicates that the introgression of B genome did not occur in these particular A chromosomes. The majority of B-genome signals in this genotype were less intense, narrower, and not concentrated around the centromeres as in B. juncea control lines or other B. juncea morphotypes. Some of the stronger B-genome signals in 99X M2-SP02 may represent whole B chromosomes or A/C introgression into B chromosomes (Fig. 2c). However the small size and low intensity of the majority of signals favours the conclusion that they result from B-genome introgression into B. napus A- or C-genome chromosomes. The small size of Brassica chromosomes and the confinement of heterochromatic FISH and genomic in-situ hybridization signals to centromeric regions (Harrison and Heslop- Harrison 1995; Snowdon et al. 1997) prevent exact localization and cytological characterization of the introgressions without the use of additional molecular techniques. In line 99X M2-SP02, there was an absence of B-genome signals on the 6 A-genome chromosomes identified by an A-genome specific sequence (Fig. 2c). Nevertheless, the introgressions may involve the remaining A-genome or C- genome chromosomes. Line 99X M2-SP02 provides physical evidence that B. juncea morphology can occur in interspecific progeny without all of the B genome present and with a chromosome complement similar to that of B. napus (2n = 38). Support for this contention is found in family 99X , which has B. juncea morphology and B. napus chromosome complement (2n = 38), is positive for B-genome PCR, and negative for B-genome FISH signals (Table 2). The pbnbh35 repeat sequence was detected exclusively in pericentromeric heterochromatin flanking the centromeres of B-genome chromosomes (Schelfhout et al. 2004), where it most likely occurs in high copy numbers. Most of the lines that were positive for B genome by PCR assay were negative for B genome by FISH assay. These lines may con- # 2006 NRC Canada