Comparison of RF3 B. juncea to RF3 B. napus

RF3 Brassica napus L. (InVigor TM Canola) has been in commercial cultivation since 1997. RF3 B. napus was produced by Agrobacterium-mediated transformation to provide glufosniate ammonium herbicide tolerance to the crop. Canola quality (CQ) varieties of Brassica juncea L. (Czern & Coss), a close relative of B. napus, have been available in Canada for more than 20 years. Currently, BASF Agricultural Solutions is developing CQ RF3 B. juncea via conventional breeding with RF3 B. napus to provide similar herbicide tolerance in the B. juncea species. The insertion event RF3 contains the bar gene (origin Streptomyces hygroscopicus) coding for phosphinothricin acetyl trans-ferase (PAT/bar) protein which confers tolerance to glufosinate-ammonium herbicides. RF3 also contains the barstar gene (origin Bacillus amyloliquefa-ciens), coding for the Barstar protein, which is an inhibitor of the Barnase protein. In the absence of Barnase, there is no impact of Barstar. Many safety studies have been conducted during the development of the canola quality lines of RF3 B. napus and RF3 B. juncea. This report seeks to compare the two species based on the scientific data generated during the safety assess-ments of RF3 B. napus and RF3 B. juncea and discusses the utility of using studies with one species to support the safety of the other species both containing the same insertion event.


Introduction
Oilseed rape (OSR), rapeseed or canola, is a bright-yellow flowering member of the family Brassicaceae, cultivated mainly for its oil-rich seed. Production and trade statistics do not distinguish between the different sources of oilseeds (i.e. Brassica species) and its products. Currently rapeseed is the second most impor-Food and Nutrition Sciences tant oilseed crop in the world, following soybean (Table 1) [1]. This paper 1) presents a comparison of the biology and use of canola quality (CQ) B. napus and B. juncea, 2) includes the definitions of canola used by various agencies around the world; 3) provides a history of the safe use of these Brassica species, 4) describes the development of RF3 B. juncea from RF3 B. napus and 5) presents the safety studies conducted as part of the risk assessment of these products.

Biology of Brassica Species
The Triangle of U (Figure 1) demonstrates the theory developed by Nagaharu [2] about the evolution and relationships among members of the plant genus Brassica. The genomes of three ancestral species of Brassica combined to create three of the common contemporary vegetable and oilseed crop species. This theory has since been confirmed by molecular studies. It shows how three of the Brassica species were derived from three ancestral genomes, denoted by the letters  Likewise in the Korean Food Standard Codex (No. 2020-3) Section 7.1.4.3, rapeseed oil (canola oil) refers to crude oil obtained from rape and treated to be fit for human consumption. Rape contains Brassica campestris L., Brassica campestris var. chinensis T. Ito, Brassica napus L. and Brassica juncea L.

Why Canola Quality B. juncea?
B. juncea has a number of advantages over B. napus including more vigorous seedling growth and quicker ground covering ability. B. juncea is better adapted to hot, dry areas than B. napus so it can tolerate heat and drought stress [5]. B. juncea can tolerate high temperatures during flowering, which has become even more important recently for areas impacted by climate change. It effectively expands potential oilseed acreage to semi-arid zones, including the prairies of Western Canada [6]. B. juncea can be grown in more marginal and low rainfall areas and has potential as an oilseed crop in the semi-arid regions of Australia [7]. Additionally, B. juncea can naturally be grown in spring canola markets in the USA and Russia.
B. juncea is more resistant to pod shattering than B. napus and matures more uniformly, allowing for direct combining which reduces the number of passes through fields and results in less seed loss at harvest [8] [9].
Another advantage of B. juncea is its inherent blackleg resistance. Brassica species containing the B genome, such as B. juncea, have high levels of resistance to blackleg disease, a fungal disease caused by Leptosphaeria maculans, which is a major concern of canola producers worldwide [10].
From a grain processing perspective, the key differentiator between these two oilseeds is the seed color, with B. napus appearing black, while B juncea can be light brown to bright yellow. The yellow seed of B. juncea also provides a low fiber, high quality meal for feed applications for mono-gastric animals. Benefits of B. juncea meal for animal feed include:  Heavier seeds with a thinner coat resulting in a higher percentage of oil and protein  Lower fiber content resulting in better digestibility of the meal for livestock  Yellow seed coat blends better with other feedstuffs so the appearance of the final feed is not altered [11] In addition, the low seed chlorophyll content is an advantage for oil refining, making the oil from yellow-seeded varieties lighter in color.

Weed Management in Canola Production
In the early days of canola production (approx. 1965), canola was seeded on land relatively free of weeds. Land was often fallow the year before canola production.
Weed control in commercial production relied on a mixture of numerous selective herbicides targeting monocot weeds, dicot weeds, and volunteer cereals. Tillage was often combined with herbicide applications.
Since the mid-1990s, herbicide tolerant canola from B. napus has been widely adopted by farmers, greatly reducing the number of herbicide applications needed to control weeds. Currently, >99% of canola (B. napus) cultivated in Canada is herbicide tolerant: 34% is Roundup Ready canola with tolerance to glyphosate, 62% is LibertyLink ® canola with tolerance to glufosinate-ammonium, and the remaining 4% is Clearfield canola with tolerance to imazethapyr/imazamox, along with some conventional canola.
Adoption of herbicide tolerant traits has resulted in the application of less herbicide active ingredient. In 2013, the use of genetically modified herbicide tolerant canola resulted in a 2.1 million kg reduction in the amount of herbicide active ingredient used (−17.1%) [12].
The use of herbicide tolerant canola has reduced fuel usage and altered tillage practices, resulting in soil conservation and related environmental benefits such as carbon sequestering. It is estimated that in Canada in 2013 the use of genetically modified canola resulted in a fuel saving of 69 million litres and reduced carbon dioxide emissions by 185 million kg [12]. Farmers use less tillage and more direct-seeding. The reduction of tillage reduces soil erosion, contributes to less air pollution from dust, improves soil moisture retention, and reduces soil compaction.
Because of these advantages, herbicide-tolerant varieties have quickly grown in popularity since their introduction in 1995.
The first herbicide-tolerant cultivars of B. juncea were developed by a process which included the transfer of a resistance gene from B. napus [13] [14]. The objective of producing RF3 B. juncea was to bring the LibertyLink ® glufosinate-ammonium herbicide tolerance trait to CQ B. juncea lines by conventional breeding. This would capture all the aforementioned advantages of CQ B. juncea with an herbicide tolerance trait. The incorporation of the bar gene for glufosinate-ammonium tolerance provides broad spectrum weed control in B. juncea and also provide growers with an additional tool for their weed resistance management strategy.

History of Safe Use
Humans have been using Brassica oilseeds for cooking and illumination since 2000-1500 B.C. [13] [15]. Canola oil is used in the manufacture of margarine and shortening, salad and cooking oil, mayonnaise, sandwich spreads, creamers and coffee whiteners, confectionery products, low-fat foods, pharmaceuticals, and nutritional supplements.
With the recent development of CQ B. juncea varieties, seeds are used and the refined oil from CQ B. juncea lines are consumed in the same way as other canola quality rapeseed oils. The fatty acid profiles of the oil from CQ B. juncea are within the normal range of other canola quality rapeseed oils. The development of CQ B. juncea has not resulted in any change in the consumption pattern for these products. CQ B. juncea meal is used as a high protein animal feed for cattle, swine, and poultry. The animal feed can contain up to 35% of this processed commodity.
For B. juncea, Bra j 1 is the major food allergen present in Indian or Oriental mustard. It belongs to the storage protein 2S albumin family, which are abundant proteins in seeds with homologues in B. napus and B. rapa. However, oil obtained from Canola Quality grain is not considered to be an allergenic food, because protein is absent in the refined oil. Food allergy to canola quality oil has not been reported in the scientific literature [16] [17] [18] [19]. Within the Brassicaceae, related species have often been used to transfer desired traits which are not available within the primary gene pool of the species [20]. B. juncea has acquired traits through interspecific crossing and has been used as a source of traits such as yellow seed color [21] and blackleg resistance [22] in breeding B. napus.
Canola Quality characteristics, primarily low erucic acid content in the oil and low glucosinolate content in the meal, were first developed in the rapeseed species B. napus and B. rapa (campestris). These two species were used to develop CQ B. juncea. Love et al. [23] reported on the successful transfer of low glucosinolate content from B. rapa to B. juncea. Raney et al. [24] further reduced the glucosinolate content of B. juncea lines through an interspecific cross with B. napus. Malode et al. [25] utilized B. napus as a source of both low erucic acid and low glucosinolate content. Brassica napus was used to further improve the fatty acid profile of B. juncea by raising the oleic fatty acid and lowering the polyunsaturated fatty acid content [24] [26] [27].
The success of these efforts to change the quality of B. juncea by interspecific crossing led to the commercial release of CQ B. juncea cultivars in Canada [28] and Australia [29].
B. napus has been used as a source of other traits in B. juncea, such as resistance to white rust [30] [31]. The Barnase and Barstar hybridization system was also transferred to B. juncea from B. napus [32].
The first herbicide-tolerant cultivars of B. juncea were developed by a process which included the transfer of a resistance gene from B. napus [13] [14].
(origin Streptomyces hygroscopicus) coding for phosphinothricin acetyl transferase (PAT/bar) protein which confers tolerance to glufosinate-ammonium [33]. The bar gene is driven by the PssuAt plant promoter that is active in all green tissues of the plant [34]. RF3 also contains the barstar gene (origin Bacillus amyloliquefaciens), coding for the Barstar protein, which is an inhibitor of the Barnase protein [35]. The barstar gene is driven by the Pta29 promoter that restricts gene expression to the tapetal cells during anther development [36]. In the absence of Barnase, there is no impact of Barstar. The OECD identifier of RF3 B. napus is ACS-BNØØ3-6. RF3 B. juncea was developed by conventional breeding to cross an elite CQ B. juncea line with RF3 B. napus following the path shown in Figure 2.
The inserted T-DNA sequences, together with the 5' and 3' flanking genomic regions, were transferred into B. juncea through backcross breeding, the applied procedure to generate RF3 B. juncea did not involve any novel transformation. The underlying mechanism through which the RF3 B. napus was transferred into B. juncea was by a homologous recombination during meiosis. Homologous recombination is a type of genetic recombination in which nucleotide sequences are exchanged between two very similar or identical DNA molecules. Through this mechanism, the previously defined sequence of the RF3 B. napus transgenic locus (i.e. T-DNA and the defined 5' and 3' flanking genomic sequence), together with large stretches of B. napus sequence originating further downstream and upstream of the RF3 transgenic locus, were transferred from the B. napus sister chromatid to the B. juncea sister chromatid. A simplified breeding tree is provided in Figure 2:

Molecular Characterization
To confirm the RF3 B. juncea insert organization as determined in RF3 B. napus, genomic DNA from ten individual RF3 B. juncea plants was digested with restriction enzymes NsiI and EcoRV, hybridized with five different vector backbone probes and subsequently with the T-DNA probe. Only the expected fragments, based on the insert organization as determined in RF3 B. napus [37] were obtained. The membranes also contained NsiI and EcoRV digestions of RF3 B. napus DNA, which resulted in the same restriction pattern as the RF3 B. juncea plants. Therefore, the Southern blot analyses confirmed the RF3 B. juncea insert organization as determined in RF3 B. napus and the absence of vector backbone sequences.
To assess the structural stability of RF3 B. juncea, three individual plants from five generations were digested with the NsiI restriction enzyme and hybridized with the T-DNA probe. Only the expected fragments, based on the insert organization as determined in RF3 B. napus, were obtained for all individual RF3 B. juncea plants from the different generations, demonstrating the structural stability of the RF3 B. juncea transgenic locus. The membrane also contained a NsiI digestion of RF3 B. napus DNA, which resulted in the same restriction pattern as the RF3 B. juncea plants [37].
The RF3 B. juncea transgenic locus sequence, including 1 kb of each flanking genomic sequence was determined and compared with the RF3 B. napus transgenic locus sequence. Both loci were found to consist of identical sequences.
In conclusion, the molecular characterization data obtained for both RF3 species convincingly show that the RF3 B. juncea variety is fully comparable to RF3 B. napus. This is to be considered as an expected result, since RF3 B. juncea originated from RF3 B. napus through conventional breeding practices and not out of a novel transformation event. All molecular analyses of RF3 B. juncea demonstrate the B. napus nature of the RF3 T-DNA region and its flanking genomic sequences.

Protein Studies
Since it has been determined that the insertion sequences remain the same in RF3 B. juncea as in RF3 B. napus, existing PAT/bar and Barstar studies may be used to demonstrate the safety of the proteins. Studies were conducted using bacterially-produced proteins. The safety assessment for the PAT protein has been described in depth by Hérouet et al. [38]. This assessment included ex- Similar studies have been conducted for Barstar. Barstar protein was incubated for 30 minutes at 4˚C, 25˚C, 37˚C, 55˚C, 75˚C, and 95˚C followed by assessment by SDS-PAGE and western blot analyses, ELISA analyses and activity assays. The samples treated at 4˚C were used for comparison of the other temperature treated samples. These studies demonstrated the stability of the Barstar protein up to 55˚C, with some loss of stability at 75˚C, and complete loss at 95˚C. Barstar degraded within 30 seconds in presence of pepsin at pH 1.2. Greater than 90% of Barstar protein was degraded within 10 minutes in the presence of pancreatin at pH 7.5. No homology was found to known toxins or allergens.
Neither PAT nor Barstar protein when dosed at 2000 mg/kg body weight via the oral route produced any signs of systemic toxicity in male and female C57BL/ 6J mice.
A protein equivalency study comparing the structural and functional characteristics of PAT/bar protein purified from RF3 B. juncea with bacterially-produced PAT/bar protein was conducted. The identity of the RF3 B. juncea plant-purified PAT/bar protein was confirmed by intact molecular mass determination, peptide mapping and N-terminal sequence determination. The N-terminal sequence was shown to be acetylated which is a common post-translational modification in plant proteins. The apparent molecular mass of PAT/bar protein was confirmed and the immuno-reactivity of the plant-purified PAT/bar protein to an anti-PAT/bar antibody was demonstrated. The plant-purified PAT/bar protein was not glycosylated. Finally, the activity of the plant-purified PAT/bar protein was demonstrated. Therefore, the plant-purified and the bacterially-produced PAT/bar proteins were considered to be structurally and functionally equivalent.
Likewise, a protein equivalency study comparing the structural and functional characteristics of Barstar protein purified from RF3 B. juncea with bacterially-produced Barstar protein was conducted. The identity of the RF3 B. juncea plant-purified Barstar protein was confirmed by N-terminal sequence determination. The apparent molecular mass of Barstar protein was confirmed and the immuno-reactivity of the plant-purified Barstar protein to an anti-Barstar antibody was demonstrated. The plant-purified Barstar protein was not glycosylated.
Finally, specific activity of the plant-purified Barstar protein was demonstrated.
Therefore, the plant-purified and the bacterially-produced Barstar proteins were considered to be structurally and functionally equivalent.

Composition
Composition data between RF3 B. napus and RF3 B. juncea were reviewed to assess similarities or major differences in nutrient content between the two species. Composition data between the non-GM conventional B. napus and non-GM conventional B. juncea were similarly assessed. RF3 B. napus and the non-GM conventional variety were grown in 2008 in five field trials in the B. napus growing regions of Saskatchewan and Alberta, Canada. RF3 B. juncea and the non-GM conventional variety were grown in 2017 in eight field trials in the B. juncea growing regions of Canada and the USA. Entries were replicated four times in a randomized complete block design (RCBD) in both studies. RF3 B. napus and B. juncea entries received one application of Liberty ® herbicide (glufosinate-ammonium) at 500 g ai/ha. One canola grain sample was harvested from each plot for compositional analysis.
Composition analyses were conducted to determine levels of nutrients and anti-nutrients in the canola grain. Each sample was analyzed for proximates and fiber (  [3], to provide ranges for reference.    An assessment of the mineral data demonstrated that the ranges for RF3 B. juncea for calcium, phosphorus, potassium, magnesium, sodium, iron, manganese, and zinc overlapped with the ranges determined for RF3 B. napus. The range for copper in RF3 B. juncea was higher than the range determined for RF3 B. napus. A similar trend was observed between the non-GM conventionals of B. napus and B. juncea in that copper was consistently higher in B. juncea, although the ranges in the non-GM conventional did overlap slightly. The ranges for all minerals for RF3 B. juncea overlapped with the ranges established within the ILSI Crop Composition Database for non-GM B. napus. The mean values for RF3 B. juncea for all minerals also fell within the ILSI Crop Composition Database range. Therefore, it can be stated that the mineral content for RF3 B. juncea is essentially comparable to that of RF3 B. napus.
An assessment of the tocopherol data demonstrated that the ranges for RF3 B. juncea for alpha-, beta-and delta-tocopherols overlapped with the ranges determined for RF3 B. napus. The ranges for RF3 B. juncea for gamma-and total tocopherols did not overlap with the ranges determined for RF3 B. napus. A similar trend was observed between the non-GM conventional counterparts of B. napus and B. juncea. Therefore, the observation that gamma-and total tocopherol did not overlap with RF3 B. napus may be attributed to differences among the two species, since a similar trend was apparent in the non-GM conventional comparison. The ranges for RF3 B. juncea for all tocopherols did overlap with the ranges established within the ILSI Crop Composition Database for non-GM B. napus. The mean values for RF3 B. juncea for each tocopherol also fell within the ILSI Crop Composition Database range. Therefore, it can be stated that the tocopherol content for RF3 B. juncea is essentially comparable to that of RF3 B. napus.
An assessment of the anti-nutrient data demonstrated that the ranges for RF3 B. juncea for phytic acid and total glucosinolates overlapped with the ranges determined for RF3 B. napus and overlapped with the ranges established within the ILSI Crop Composition Database for non-GM B. napus. The mean values for RF3 B. juncea for each anti-nutrient also fell within the ILSI Crop Composition Database range. Therefore, it can be stated that anti-nutrient data for RF3 B. juncea is essentially comparable to that of RF3 B. napus.

Protein Expression Data
PAT and Barstar protein expression data obtained from RF3 B. napus and RF3 B. juncea samples collected from different field trials were visually inspected and compared to assess similarities or major differences between the two species.
Tissue  Table 8 and Table 9, respectively.  herbicide. This and the fact that the maximum level of PAT/bar in untreated B. napus (0.92 µg/DW) was very close to the minimum level of PAT/bar in untreated B. juncea (1.07 µg/DW) indicate that this difference is insignificant and not biologically relevant. In addition, protein expression has been shown to be highly variable due to many biological, technical, and environmental factors; therefore, interpretation of comparative results should also consider natural variation in protein expression levels [40]. Based on the observed overlapping ranges of measured analyte concentrations, it appears that PAT/bar and Barstar are similarly expressed when comparing RF3 B. juncea and RF3 B. napus.

Animal Feeding/Wholesomeness Studies
A 90-day rodent feeding study with RF3 B. napus found no adverse effects on the growth or health of Sprague Dawley rats fed a diet containing 15% RF3 B. napus meal. A 42-day broiler feeding study with RF3 B. napus found the growth and health of chickens fed a diet containing 10% RF3 B. napus meal were comparable to chickens on a nutritionally equivalent diet containing no RF3 B. napus. In addition, a 56-day channel catfish feeding study with RF3 B. napus concluded that catfish consuming a diet containing 30% RF3 B. napus meal demonstrated health and growth characteristics comparable to catfish consuming the conventional control variety.
After consideration of the factors listed above, new feeding studies were not scientifically justified and were not conducted with RF3 B. juncea because of its similar nature to RF3 B. napus. The RF3 B. napus data could be used as surrogate data for RF3 B. juncea since it is the same molecular insert, is similar in composition, and had similar PAT and Barstar expression levels as RF3 B. napus.

Conclusions
B. napus and B. juncea are closely related species within the same genus with a history of safe use. Both meet the definition of "canola" (<2% erucic acid in the oil and <30 µmol/g glucosinolates in the meal