Plant Cell Reports
Plant Cell Reports (1992) 1 !: 4 9 9 - 505
9 Springer-Verlag 1992
Transformation and regeneration of Brassica rapa using Agrobacterium tumefaciens Sharon E. Radke, Joann C. Turner, and Daniel Facciotti Calgene, Inc. 1920 Fifth Street Davis, California 95616, U S A Received M a y 12, 1992/Revised version received July 1, 1992 - C o m m u n i c a t e d by C.F. Quiros
Summary. Transformation and regeneration procedures for obtaining transgenic Brassica rapa ssp. oleifera plants are described. Regeneration frequencies were increasedby using silver nitrate and by adjusting the duration of exposure to 2,4D. For transformation, Agrobacterium tumefaciens strain EHA101 containing a binary plasmid with the neomycin phosphotransferase gene (NPT II) and the b-glucuronidase gene (GUS) was cocultivated with hypocotyl explants from the oilseed B. rapa cvs. Tobin and Emma. Transformed plants were obtained within three months of cocultivation. Transformation frequencies for the cultivars Tobin andEmma were 1-9%. Evidence for transformation was shown by NPT II dot blot assay, the GUS fluorometric assay, Southern analysis, and segregation of the kanamycin-resistance trait in the progeny. The transformation and regeneration procedure described here has been used routinely to transform two cultivars ofB. rapa and 18 cultivars ofB. napus.
Key words: Brassica - Oilseed rape - Gene transfer Transgenic plants - Silver nitrate
Introduction. Brassica rapa ssp. oleifera (syn. B. campestris), oilseed rape,
is an important edible and industrial oilseed crop worldwide. It is grown widely in the Indian subcontinent, Northwest China, Western Canada, and Northern Europe. B. rapa can be cultivated in northern areas with short growing seasons or in climates unsuitable for the culture of B. napus. For canola production in Canada, total acreage ofB. rapa is about equal to that of B. napus. Genetic engineering techniques have been applied to B. napus to introduce new traits such as modified oil composition (Knutzon et al. 1992), herbicide tolerance (De Block et al. 1989), and altered protein composition (Altenbach et al. 1992). It would be advantageous to extend these techniques to B. rapa in order to broaden the utility of that crop. Correspondence to: D. Facciotti
The routine improvement of genotypes by transformation or by mutagenesis of explants requires the development of efficient tissue culture methods. Inthepast, shootregeneration frequencies for B. rapa were low in comparison to B. napus (Murataand Orton 1987). More recently, higher frequencies were obtained from oilseed B. rapa ssp. oleifera using cotyledonarypetioles (Hachey etal. 1991) andhypocotyls (Facciotti et al. 1989) and from vegetable B. rapa ssp. chinensis, ssp. parachinensis and ssp. pekinensis using hypocotyls and cotyledons (Chi and Pua 1989; Chi et al. 1990). Several investigators (Facciotti et al. 1989; Chi and Pua 1989; Chi et al. 1990) reported that higher regeneration frequencies were obtained fromB, rapa hypocotyl and cotyledon explants when inhibitors of ethylene action such as silver nitrate were used. Methods havebeen reported forAgrobacteriumtumefaciensmediated transformation and regeneration from explants of the following Brassica species: B. napus (Pua et al. 1987; Fry et al. 1987; Radke et al. 1988), B. oleracea (Srivestava et al. 1988; De Blocket al. 1989),B.juncea (BarfieldandPua 1991). Previously we presented preliminary results showing transformation ofB. rapa hypocotyl explants (Facciotti et al. 1989). Here we present a more complete transformation and regeneration study of B. rapa ssp. oleifera. The transformation protocol described is also used for B. napus and may be applicable to other Brassica species. Recently, a modification of oil composition was demonstrated by the insertion of an antisense stearoyl-ACP desaturase gene into B. rapa and B. napus using this transformation method (Knutzon etal. 1992).
Materials and methods Plant Regeneration. B. rapa cultivars used in this study were Ante, Candle, Emma, Span, Sylvi, Tobin and R-500. Seeds were surface-sterilized for 20 min and planted in Magenta boxes containing modified MS medium (0.1 X) as described previously (Radke et al. 1988). Seeds were germinated and grown in a cell cultu re room maintained at 22-24 ~ with a 16 h light/8 h dark photoperiod at a light intensity of 45-55 mEm-2s-x. Hypocotyls and roots were dissected from 3 to 7-day-old seedlings and cut
500 into segments 5-10 mm in length. Cotyledonary explants were approximately 3 mm in size. Petioles and leaves were collected from 2 to 3-weekold seedlings and cut into 3-10 mm sections. The explants were first placed onto B5-1 callus induction medium which contained B5 salts (Gibco), B5 vitamins, 1 mg/l 2,4-dicblorophenoxyaeetic acid (2,4-D), 3 % sucrose, 0.6% Phytagar (Gibco), pH 5.8. After 3-16 days, the explants were transferred to B5-BZ shoot regeneration medium (B5 salts and vitamins, 3 mg/1 6beuzylaminopurine (BAP), 1 mg/1 zeatin, 1% sucrose, 0.6% Phytagar, pH 5.8) in the presence or absence of AgNO3 (5 or 10 rag/l). Two to eight weeks later regenerated shoots were transferred to B5 root induction medium (B5 salts and vitamins, 2 mg/lindole-3-butyric acid, 1% sucrose, 0.6% Phytagar, pH 5.8). In 2-4 weeks, rooted plants were transferred to soil.
standard protocols (Maniatis et al. 1982). The blot was washed and reprobed with a radiolabelled 1.9 kb GUS DNA probe. Inheritance of the NPT II gene in the progeny was determined by segregation of the kanamycin-resistance trait in the presence of the antibiotic G418, an analog of kanamycin. Seeds were surface-sterilized for 10 rain as above and then placed onfilterpaper (Whatman 3MM) in Magentaboxes containing 7 rol of 0.1 X MS salts and 75 mg/l G418 (Geneticin, Sigma). Seeds were germinated at 24 ~ under continuous fight of 110 mEm-:s-1. After one week, seedlings were scored for resistance (expanded green cotyledons, elongated axis, and presence of roots) or sensitivity (yellow embryos, axis not elongated) to the antibiotic compared to untransformed B. rapa seedlings.
Planttransformation. Agrobacterium tumefaciens strain EHA101 (Hood et al. 1986)harboring binary plasmid pCGN7001 (Fig. 1D, Comai et al. 1990) which contains an NPT II gene and a GUS gene was used to transform B. rapa. In some experiments pCGN1557-based binary plasmids (McBride and Summeffelt 1990) which contain an NPT II selectable markergene expressed from a CaMV 35S 5' region and a tml 3' region and several genes of interest were used. For an example of one of these plasmids, pCGN3242, see Knutzon et al. (1992). The hypocotyl transformation protocol developed forB. napus (Radke et al. 1988) was followed with modifications. Seeds ofB. rapa cvs. Emma and Tobin were surface-sterilized as above. Hypocotyls were excised from 6 to 7-day-old seedlings, cut into segments 2-4 mm in length, and placed onto filter paper over tobacco feeder cells (previously described) which were spread ontoMS-1 medium (MS salts (Gibco), 100 mg/1 myo-inositol, 1.3 rag/ 1thi amine-HC1, 200 mg/1KH2PO4,1 rag/12,4-D, 3 % sucrose, 0.6 % Phytagar, pH5.8). Explants were incubated for 18-24 h on feeder plates under indirect continuous light. Explants were immersed for 10 min in a suspension of lxl08 bacteria/ml, then returned to feeder plates. After two days of cocultivation with Agrobacterium, explants were transferred to B5-1 medium supplemented with 500 rag/1 carbenicillin and with or without 25 mg/1 kanamycin (Boehringer-Mannheim). Plates were sealed with Parafilm or Micropore (3M) paper tape. Explants were cultured at 24 ~ under continuous light at 75 mEm-2s "1for 3-7 days, then transferred to B5-BZ shoot regeneration medium supplemented with 0, 10 or 25 mg/1 kanamycin, 500 mg/1 carbenicillin, and 0, 5, or 10 mg/l AgNO 3. Explants were subsequently transferred every two weeks to fresh medium without AgNO v The initial plating density was 40-50 explants per plate (100 x 25 ram). It was reduced to 20-25 per plate after the second transfer to B5-BZ medium. Five to nine weeks after culture initiation, green shoots were cut from kanamycinselected calli and placed onto B5-0 shoot maturation medium (135 medium with 1% sucrose and lacking hormones) supplemented with 300 rag/1 carbenicillin and 50 mg/lkanamycin. Two weekslater, shoots were trimmed to contain 2-3 nodes and placed on B5 root induction medium supplemented with 50 mg/1 kanamycin. Roots developed on some of the shoots after two weeks. Shoots which had not rooted were re-cut at the base and placed back onto the medium for another 2-4 weeks.
Plant growth conditions. Regenerated plants were transferred to soft-less mix in six-pack containers and placed in a growth chamber at 22 ~ with a 16h light/8 h dark photoperiod. After 2-3 weeks theplants weretransplanted into 2-gaUon pots and grown in a greenhouse. Self-pollinated seed were obtained eitherby bud pollination or by placing the flowering plants in a CO2enriched (up to 4%) atmosphere for 24 h to overcome self-incompatibility (Nakanishi and Hinata 1973). Analysis of transformed plants. Leaf samples from kanamycin-selected plants were assayed for NPT II enzyme activity using a dot blot assay (Radke et al. 1988). Leaf samples were assayed for GUS enzyme activity using a fluorometric assay (Jefferson 1987). For Southern blot analysis, DNA was isolated from leaves (Dellaporta et al. 1983), and purified once by cesium chloride density gradient centrifugation. A 1 kb NPT II DNA fragment was radiolabelled by nick-translation (BRL kit) and hybridized to BamHI orEcoRI digested DNA, which had been electrophoresed on a 0.7% agarose gel and transferred to nitrocellulose using
Results and discussion
Plant regeneration
The morphogenic potential of several B. rapa tissues was determined. The most efficient shoot regeneration system was achieved through the use of hypocotyl explants, and is described here. Hypocotyl explants were cultured for 3-4 days on B5-1 callus induction medium followed by seven days on B5-BZ shoot regeneration medium containing 5-10 mg/1 silver nitrate, and then transferred to B5-BZ without silver. Under these conditions, friable callus was observed at the cut ends of explants a few days after the transfer onto medium containing silver nitrate. The callus was derived from cells of, or near, the vascular cambium. Within two weeks, compact green nodular tissues with numerous shoot primordia were produced from the calli on most of the explants. Shoots were transferred to rooting medium as early as four weeks after culture initiation. Plants were transferred to pots 6-12 weeks after the beginning of culture. Eighty percent of the transplants survived. In most of the plants derived from spring cultivars the onset of flowering occurred early, sometimes within five days after transplanting. In the winter cultivar Sylvi, the vernalization requirement for flowering was not modified by the passage through culture. The oil composition of seeds from regenerated plants was not significantly different from that of control seeds (data not shown). Effect of 2,4-D The induction of shoot primordia, as described above, required exposure of the hypocotyl explants to the auxin 2,4-D prior to culture on cytokinin-enriched shoot regeneration medium. The duration of exposure to 2,4-D (1 mg/1) was critical to achieve efficient regeneration. For the variety Emma, a minimum of two days of exposure to 2,4-D was required to obtain frequencies (percentage of explants which produced at least one shoot) of 40% (27/68). The highest frequency of 63% (94/150) was obtained after 3-4 days of exposure. Lower frequencies (below 25%) were obtained from explants following 11-14 days on 2,4-D and only 3% of the explants regenerated after a one-day exposure. This trend was also observed in the cultivar Candle, for which the shoot regeneration frequencies were 32% (48/150) after three days,
501 19% (37/200) after five days, and 17% (34/200) after seven days on 2,4-D. For both Emma and Candle 3-4 days of exposure to 2,4-D was optimal.
Silver nitrate concentrations of 5 mg/l and 10 mg/l were found to promote the formation of shoot primordia at the same frequency (76% and 78% respectively).
Effect of silver nitrate
Effect of cultivar and explant source
The use of silver nitrate to promote adventitious shoot formation by inhibition of ethylene action in tissue culture was first reported forNicotiana and wheat (Pumhauser et at. 1987). This beneficial effect was recently reported forB. rap a (Chi and Pua 1989) and other Brassica species (De Block et at. 1989; Chi et al. 1990; Sethi et at. 1990)o
The pattern of morphogenesis from hypocotyl explants was similar for six of the cultivars (Emma, Candle, Tobin, Ante, Span, and Sylvi) of B. rapa that were tested. With R-500, shoot primordia formed on most explants (80-95%) but they grew into abnormal structures and only a few explants produced shoots (4.5%). Hachey et at. (1991) also reported a reduced regeneration for R-500 compared with other culti-
Table 1. Frequencies of shoot primordia induction and regeneration from hypocotyl explants ofBrassica rapa cv. Emma.
VarS.
Silver Nitrate Exposure
Primordia: No. of explantsa Average % Range (%) Shoots: No. of explantsa Average % Range (%)
withoutb
continuousb
94/218 43 (30-77)
373/617 61 (33-92)
43/364 12 (1-29)
99/433 23 (10-62)
one weekc
606/1250 49 (26-63)
nPooled data from 3-7 experiments. Frequencies are the number of explants with primordia or shoots/total of explants. bCulture conditions not optimized for 2,4-D. CCulture conditions optimized, 3-4 days on 2,4-D.
In our experiments withB, rapa, the addition of silver nitrate to the B5-BZ medium had a dramatic effect on shoot primordium induction and elongation. The number of primordia on silver nitrate-treated explants was consistently higher, up to 40 per explant, than on untreated explants which had fewer than 10 per explant. This indicates that silver nitrate increases the number of cells from an individual explant which undergo morphogenesis. In the absence of silver nitrate, primordia reverted to unorganized callus after 3-4 weeks in culture When silver nitrate was used, primordia did not revert to calli and shoots elongated more rapidly. In a few cases up to eight shoots developed per treated explant, although formation of 1-3 shoots was typical. Either competition for nutrients or dominance among primordia presumably limited the frequency of primordia which formed shoots from the same explant. Silver increased the average frequency of explants with primordia to 61% from the 43% observed for untreated explants. An average of 23 % of the treated explants produced shoots comparedwith 12% of untreated explants (Table 1). In general, more than one-third of the explants which had primordia developed shoots. Although silver nitrate was undoubtedly effective in promoting shoot formation, continuous exposure to it caused symptoms of vitrification. Limiting the silver exposure to the first seven days on regeneration medium increased the average regeneration frequency to 49% (Table 1) and decreased symptoms of vitrification.
Several tissues in addition to hypocotyls were used as explant sources. Regeneration frequencies were up to 36% from petiole cross-sections, 15% from cotyledon and leaf explants, and 1% from root segments. Although we determined that several B. rapa tissues are capable of a morphogenic response, only hypocotyl explants were used to develop a transformation protocol. Plant transformation The A. tumefaciens hypocotyl transformation procedure de-
scribed previously (Radke et al~ 1988) and the regeneration procedure described above have been modified to improve transformation frequencies ofB. napus and to achieve routine wansformation of B. rapa. Kinetin was eliminated from the B5 callus induction medium and MS salts rather than B5 salts were used during the cocultivationperiod (results not shown). With these modifications transformedB, rapaplants have been produced routinely. In addition to the modifications described above other parameters were tested to improve the transformation frequency (percentage ofcocultivatedexplants that produced at least one transformed pianO. Kanamycin selection The antibiotic kanamycin was added to media after the twoday cocultivation period and maintained throughout the regeneration and rooting processes. When non-cocultivated control explants were selected on 25 mg/1kanamycin, yellow or white calli grew slowly and shoot regeneration was completely inhibited (Table 2). When only 10 mg/l kanamycin was used for selection, some green calli developed and pale green shoots regenerated from control explants. Because ot these escapes, the higher concentration (25 mg/1) was used during shoot regeneration from cocultivated explants even though the highest transformation frequency (9%) was obtained using the lower concentration of kanamycin (10 mg/l). Following selection of the explants on regeneration medium containing 25 mg/l kanamycin, the regenerated shoots were transferred to shoot maturation (B5-0) and rooting media containing 50 mg/l kanamycin. The higher concentration of kanamycin was used because control shoots rooted occasion-
502 Table 2. Comparison of parameters for the transformation of Brassica rapa cv. Emma hypocotyl explants. + Feeder Cells Kan. Conc. (mg/1)
2,4-D Exposure (Days)
Control pCGN 7001 pCGN 7001 pCGN7001
10 10 10 10
Control pCGN7001 pCGN 7001 pCGN7001
25 25 25 25
- Feeder Cells
KanR Callia (%)
KanR Shootsb (%)
Transformed Plants c (%)
10 6 8 10
5/12 66/82 64/87 67/86
(42) (80) (74) (77)
2/12 14/82 8/87 19/86
(17) (17) (9) (22)
0/82 3/87 8/86
10 6 8 10
0/12 8/87 78/85 74/82
(0) (93) (92) (90)
0/12 8/81 16/85 15/82
(0) (10) (19) (18)
1/81 5/85 5/82
KanR Callia
(%)
(0) (3) (9)
1/11 55/89 53/89 69/88
(9) (62) (60) (78)
(1) (6) (6)
0/1l 52/89 54/88 68/86
(0) (58) (61) (79)
KanR Shoots b (%) 1/11 7/89 7/89 6/88
(9) (8) (8) (7)
0/11 (0) 3/89 (3) 7/88 (8) 12/86 (14)
Transformed Plants c (%) 0/89 0/89 1/88
1/89 1/88
4/86
(0) (0) (1)
(1) (1) (5)
aNumber of explants with green kanamycin-resistant calli/total number of explants. bNumber of explants with green shoots that regenerated on kanamycin/total number of explants. CNumber of explants from which at least one green shoot rooted on kanamycin and tested positive for GUS or NPT II/total number of explants.
ally and the leaves remained partially green on 25 mg/1 kanamycin. Although some of the shoots which regenerated on 25 mg/1kanamycin were escapes, other shoots which were transformed were subsequently lost on the more stringent selection (50 rag/l) during the shoot maturation and rooting processes. These losses resulted from either low expression of the NPT II gene or inability to root. In the experiment shown in Table 2, the recovery of transformed Emma plants from regenerated shoots was about 33% from explants exposed to 2,4-D for 10 days and initially selected on 25 rag/1kanamycin. A more typical recovery rate for B. rapa is 17%. Among Tobin plants which were regenerated from cocultivated kanamycin-selectedexplants, 97% (182/187) tested positive for NPT II activity using the dot blot procedure (Fig. 1A). Maintaining selection throughout the regeneration and rooting process was essential to achieve the high percentage of rooted plants that were transformed. With these stringent selection conditions it would not be necessary to screen the plants for NPT II activity before continuing to analyze plants for the gene of interest. 2,4-D optimization To attain the highest transformation frequency, the duration of 2,4-D exposure was optimized as in the regeneration procedure discussed above. One of the experiments, summarizedin Table 2, shows that transformation frequencies were highest (3-9%) when hypocotyl explants were cultured on media containing 2,4-D (MS-1 followedby B5-1) for the first 8 or 10 days from culture initiation. In other experiments, no transformed plants were recovered when explants were incubated for only three days on 2,4-D and then placed onto regeneration medium (B5-BZ) directly after cocultivation. It was also observed that in general, more kanamycin-selected shoots regenerated when explants were cultured on 2,4-D for a total of 10 days as opposed to 8 days. The 10-day 2,4-D treatment for Emma and Tobin cocultivated explants is longer than the optimal 3-4 day Ireatment reported above for regeneration of
hypocotyl explants. StressfromAgrobacteriurninfectionas well as from selection on kanamycin may decrease the rate of cell division and the number of cells that can grow, resulting in the need for a longer exposure to 2,4-D to promote callus induction before transfer to cytokinin medium. OurB. napus results (unpublished) indicate that the duration of 2,4-D exposure may be optimizedfor differentgenotypes to achieve high transformation frequencies. In both cultivars tested, differences in regeneration response were observed for explants that were not selected on kanamycin after exposure to the stress of Agrobacterium. The regeneration frequency ofcocultivatedTobinexplants averaged 34% while non-cocultivated control Tobin explants averaged 15%. On the other hand, cocultivated Emma explants had frequencies of 37% while non-cocultivated controls had slightly higher frequencies of about 43%. We have also observed genotype-dependent responses to Agrobacterium infection from B. napus, in which regeneration of non-selected explants increased after cocultivation with Agrobacterium in some cultivars while it decreased or remained the same in others. Within a cultivar, regeneration frequencies could probably be increased for purposes of mutagenesis or transformation by optimization of the exposure to 2,4-D after a particular treatment. Feeder cells The use of tobacco feeder cells during the preincubation and two-day eocultivation periods increased the number of explants that produced green kanamycin-selected calli and increased the transformation frequency in many of the treatments shown in Table 2. Although feeder cells were not required to obtain transformed plants they were routinely used in our experiments. Transformation frequencies ranging from
Plant Cell Reports (1992) 1 !: 4 9 9 - 505
9 Springer-Verlag 1992
Transformation and regeneration of Brassica rapa using Agrobacterium tumefaciens Sharon E. Radke, Joann C. Turner, and Daniel Facciotti Calgene, Inc. 1920 Fifth Street Davis, California 95616, U S A Received M a y 12, 1992/Revised version received July 1, 1992 - C o m m u n i c a t e d by C.F. Quiros
Summary. Transformation and regeneration procedures for obtaining transgenic Brassica rapa ssp. oleifera plants are described. Regeneration frequencies were increasedby using silver nitrate and by adjusting the duration of exposure to 2,4D. For transformation, Agrobacterium tumefaciens strain EHA101 containing a binary plasmid with the neomycin phosphotransferase gene (NPT II) and the b-glucuronidase gene (GUS) was cocultivated with hypocotyl explants from the oilseed B. rapa cvs. Tobin and Emma. Transformed plants were obtained within three months of cocultivation. Transformation frequencies for the cultivars Tobin andEmma were 1-9%. Evidence for transformation was shown by NPT II dot blot assay, the GUS fluorometric assay, Southern analysis, and segregation of the kanamycin-resistance trait in the progeny. The transformation and regeneration procedure described here has been used routinely to transform two cultivars ofB. rapa and 18 cultivars ofB. napus.
Key words: Brassica - Oilseed rape - Gene transfer Transgenic plants - Silver nitrate
Introduction. Brassica rapa ssp. oleifera (syn. B. campestris), oilseed rape,
is an important edible and industrial oilseed crop worldwide. It is grown widely in the Indian subcontinent, Northwest China, Western Canada, and Northern Europe. B. rapa can be cultivated in northern areas with short growing seasons or in climates unsuitable for the culture of B. napus. For canola production in Canada, total acreage ofB. rapa is about equal to that of B. napus. Genetic engineering techniques have been applied to B. napus to introduce new traits such as modified oil composition (Knutzon et al. 1992), herbicide tolerance (De Block et al. 1989), and altered protein composition (Altenbach et al. 1992). It would be advantageous to extend these techniques to B. rapa in order to broaden the utility of that crop. Correspondence to: D. Facciotti
The routine improvement of genotypes by transformation or by mutagenesis of explants requires the development of efficient tissue culture methods. Inthepast, shootregeneration frequencies for B. rapa were low in comparison to B. napus (Murataand Orton 1987). More recently, higher frequencies were obtained from oilseed B. rapa ssp. oleifera using cotyledonarypetioles (Hachey etal. 1991) andhypocotyls (Facciotti et al. 1989) and from vegetable B. rapa ssp. chinensis, ssp. parachinensis and ssp. pekinensis using hypocotyls and cotyledons (Chi and Pua 1989; Chi et al. 1990). Several investigators (Facciotti et al. 1989; Chi and Pua 1989; Chi et al. 1990) reported that higher regeneration frequencies were obtained fromB, rapa hypocotyl and cotyledon explants when inhibitors of ethylene action such as silver nitrate were used. Methods havebeen reported forAgrobacteriumtumefaciensmediated transformation and regeneration from explants of the following Brassica species: B. napus (Pua et al. 1987; Fry et al. 1987; Radke et al. 1988), B. oleracea (Srivestava et al. 1988; De Blocket al. 1989),B.juncea (BarfieldandPua 1991). Previously we presented preliminary results showing transformation ofB. rapa hypocotyl explants (Facciotti et al. 1989). Here we present a more complete transformation and regeneration study of B. rapa ssp. oleifera. The transformation protocol described is also used for B. napus and may be applicable to other Brassica species. Recently, a modification of oil composition was demonstrated by the insertion of an antisense stearoyl-ACP desaturase gene into B. rapa and B. napus using this transformation method (Knutzon etal. 1992).
Materials and methods Plant Regeneration. B. rapa cultivars used in this study were Ante, Candle, Emma, Span, Sylvi, Tobin and R-500. Seeds were surface-sterilized for 20 min and planted in Magenta boxes containing modified MS medium (0.1 X) as described previously (Radke et al. 1988). Seeds were germinated and grown in a cell cultu re room maintained at 22-24 ~ with a 16 h light/8 h dark photoperiod at a light intensity of 45-55 mEm-2s-x. Hypocotyls and roots were dissected from 3 to 7-day-old seedlings and cut
500 into segments 5-10 mm in length. Cotyledonary explants were approximately 3 mm in size. Petioles and leaves were collected from 2 to 3-weekold seedlings and cut into 3-10 mm sections. The explants were first placed onto B5-1 callus induction medium which contained B5 salts (Gibco), B5 vitamins, 1 mg/l 2,4-dicblorophenoxyaeetic acid (2,4-D), 3 % sucrose, 0.6% Phytagar (Gibco), pH 5.8. After 3-16 days, the explants were transferred to B5-BZ shoot regeneration medium (B5 salts and vitamins, 3 mg/1 6beuzylaminopurine (BAP), 1 mg/1 zeatin, 1% sucrose, 0.6% Phytagar, pH 5.8) in the presence or absence of AgNO3 (5 or 10 rag/l). Two to eight weeks later regenerated shoots were transferred to B5 root induction medium (B5 salts and vitamins, 2 mg/lindole-3-butyric acid, 1% sucrose, 0.6% Phytagar, pH 5.8). In 2-4 weeks, rooted plants were transferred to soil.
standard protocols (Maniatis et al. 1982). The blot was washed and reprobed with a radiolabelled 1.9 kb GUS DNA probe. Inheritance of the NPT II gene in the progeny was determined by segregation of the kanamycin-resistance trait in the presence of the antibiotic G418, an analog of kanamycin. Seeds were surface-sterilized for 10 rain as above and then placed onfilterpaper (Whatman 3MM) in Magentaboxes containing 7 rol of 0.1 X MS salts and 75 mg/l G418 (Geneticin, Sigma). Seeds were germinated at 24 ~ under continuous fight of 110 mEm-:s-1. After one week, seedlings were scored for resistance (expanded green cotyledons, elongated axis, and presence of roots) or sensitivity (yellow embryos, axis not elongated) to the antibiotic compared to untransformed B. rapa seedlings.
Planttransformation. Agrobacterium tumefaciens strain EHA101 (Hood et al. 1986)harboring binary plasmid pCGN7001 (Fig. 1D, Comai et al. 1990) which contains an NPT II gene and a GUS gene was used to transform B. rapa. In some experiments pCGN1557-based binary plasmids (McBride and Summeffelt 1990) which contain an NPT II selectable markergene expressed from a CaMV 35S 5' region and a tml 3' region and several genes of interest were used. For an example of one of these plasmids, pCGN3242, see Knutzon et al. (1992). The hypocotyl transformation protocol developed forB. napus (Radke et al. 1988) was followed with modifications. Seeds ofB. rapa cvs. Emma and Tobin were surface-sterilized as above. Hypocotyls were excised from 6 to 7-day-old seedlings, cut into segments 2-4 mm in length, and placed onto filter paper over tobacco feeder cells (previously described) which were spread ontoMS-1 medium (MS salts (Gibco), 100 mg/1 myo-inositol, 1.3 rag/ 1thi amine-HC1, 200 mg/1KH2PO4,1 rag/12,4-D, 3 % sucrose, 0.6 % Phytagar, pH5.8). Explants were incubated for 18-24 h on feeder plates under indirect continuous light. Explants were immersed for 10 min in a suspension of lxl08 bacteria/ml, then returned to feeder plates. After two days of cocultivation with Agrobacterium, explants were transferred to B5-1 medium supplemented with 500 rag/1 carbenicillin and with or without 25 mg/1 kanamycin (Boehringer-Mannheim). Plates were sealed with Parafilm or Micropore (3M) paper tape. Explants were cultured at 24 ~ under continuous light at 75 mEm-2s "1for 3-7 days, then transferred to B5-BZ shoot regeneration medium supplemented with 0, 10 or 25 mg/1 kanamycin, 500 mg/1 carbenicillin, and 0, 5, or 10 mg/l AgNO 3. Explants were subsequently transferred every two weeks to fresh medium without AgNO v The initial plating density was 40-50 explants per plate (100 x 25 ram). It was reduced to 20-25 per plate after the second transfer to B5-BZ medium. Five to nine weeks after culture initiation, green shoots were cut from kanamycinselected calli and placed onto B5-0 shoot maturation medium (135 medium with 1% sucrose and lacking hormones) supplemented with 300 rag/1 carbenicillin and 50 mg/lkanamycin. Two weekslater, shoots were trimmed to contain 2-3 nodes and placed on B5 root induction medium supplemented with 50 mg/1 kanamycin. Roots developed on some of the shoots after two weeks. Shoots which had not rooted were re-cut at the base and placed back onto the medium for another 2-4 weeks.
Plant growth conditions. Regenerated plants were transferred to soft-less mix in six-pack containers and placed in a growth chamber at 22 ~ with a 16h light/8 h dark photoperiod. After 2-3 weeks theplants weretransplanted into 2-gaUon pots and grown in a greenhouse. Self-pollinated seed were obtained eitherby bud pollination or by placing the flowering plants in a CO2enriched (up to 4%) atmosphere for 24 h to overcome self-incompatibility (Nakanishi and Hinata 1973). Analysis of transformed plants. Leaf samples from kanamycin-selected plants were assayed for NPT II enzyme activity using a dot blot assay (Radke et al. 1988). Leaf samples were assayed for GUS enzyme activity using a fluorometric assay (Jefferson 1987). For Southern blot analysis, DNA was isolated from leaves (Dellaporta et al. 1983), and purified once by cesium chloride density gradient centrifugation. A 1 kb NPT II DNA fragment was radiolabelled by nick-translation (BRL kit) and hybridized to BamHI orEcoRI digested DNA, which had been electrophoresed on a 0.7% agarose gel and transferred to nitrocellulose using
Results and discussion
Plant regeneration
The morphogenic potential of several B. rapa tissues was determined. The most efficient shoot regeneration system was achieved through the use of hypocotyl explants, and is described here. Hypocotyl explants were cultured for 3-4 days on B5-1 callus induction medium followed by seven days on B5-BZ shoot regeneration medium containing 5-10 mg/1 silver nitrate, and then transferred to B5-BZ without silver. Under these conditions, friable callus was observed at the cut ends of explants a few days after the transfer onto medium containing silver nitrate. The callus was derived from cells of, or near, the vascular cambium. Within two weeks, compact green nodular tissues with numerous shoot primordia were produced from the calli on most of the explants. Shoots were transferred to rooting medium as early as four weeks after culture initiation. Plants were transferred to pots 6-12 weeks after the beginning of culture. Eighty percent of the transplants survived. In most of the plants derived from spring cultivars the onset of flowering occurred early, sometimes within five days after transplanting. In the winter cultivar Sylvi, the vernalization requirement for flowering was not modified by the passage through culture. The oil composition of seeds from regenerated plants was not significantly different from that of control seeds (data not shown). Effect of 2,4-D The induction of shoot primordia, as described above, required exposure of the hypocotyl explants to the auxin 2,4-D prior to culture on cytokinin-enriched shoot regeneration medium. The duration of exposure to 2,4-D (1 mg/1) was critical to achieve efficient regeneration. For the variety Emma, a minimum of two days of exposure to 2,4-D was required to obtain frequencies (percentage of explants which produced at least one shoot) of 40% (27/68). The highest frequency of 63% (94/150) was obtained after 3-4 days of exposure. Lower frequencies (below 25%) were obtained from explants following 11-14 days on 2,4-D and only 3% of the explants regenerated after a one-day exposure. This trend was also observed in the cultivar Candle, for which the shoot regeneration frequencies were 32% (48/150) after three days,
501 19% (37/200) after five days, and 17% (34/200) after seven days on 2,4-D. For both Emma and Candle 3-4 days of exposure to 2,4-D was optimal.
Silver nitrate concentrations of 5 mg/l and 10 mg/l were found to promote the formation of shoot primordia at the same frequency (76% and 78% respectively).
Effect of silver nitrate
Effect of cultivar and explant source
The use of silver nitrate to promote adventitious shoot formation by inhibition of ethylene action in tissue culture was first reported forNicotiana and wheat (Pumhauser et at. 1987). This beneficial effect was recently reported forB. rap a (Chi and Pua 1989) and other Brassica species (De Block et at. 1989; Chi et al. 1990; Sethi et at. 1990)o
The pattern of morphogenesis from hypocotyl explants was similar for six of the cultivars (Emma, Candle, Tobin, Ante, Span, and Sylvi) of B. rapa that were tested. With R-500, shoot primordia formed on most explants (80-95%) but they grew into abnormal structures and only a few explants produced shoots (4.5%). Hachey et at. (1991) also reported a reduced regeneration for R-500 compared with other culti-
Table 1. Frequencies of shoot primordia induction and regeneration from hypocotyl explants ofBrassica rapa cv. Emma.
VarS.
Silver Nitrate Exposure
Primordia: No. of explantsa Average % Range (%) Shoots: No. of explantsa Average % Range (%)
withoutb
continuousb
94/218 43 (30-77)
373/617 61 (33-92)
43/364 12 (1-29)
99/433 23 (10-62)
one weekc
606/1250 49 (26-63)
nPooled data from 3-7 experiments. Frequencies are the number of explants with primordia or shoots/total of explants. bCulture conditions not optimized for 2,4-D. CCulture conditions optimized, 3-4 days on 2,4-D.
In our experiments withB, rapa, the addition of silver nitrate to the B5-BZ medium had a dramatic effect on shoot primordium induction and elongation. The number of primordia on silver nitrate-treated explants was consistently higher, up to 40 per explant, than on untreated explants which had fewer than 10 per explant. This indicates that silver nitrate increases the number of cells from an individual explant which undergo morphogenesis. In the absence of silver nitrate, primordia reverted to unorganized callus after 3-4 weeks in culture When silver nitrate was used, primordia did not revert to calli and shoots elongated more rapidly. In a few cases up to eight shoots developed per treated explant, although formation of 1-3 shoots was typical. Either competition for nutrients or dominance among primordia presumably limited the frequency of primordia which formed shoots from the same explant. Silver increased the average frequency of explants with primordia to 61% from the 43% observed for untreated explants. An average of 23 % of the treated explants produced shoots comparedwith 12% of untreated explants (Table 1). In general, more than one-third of the explants which had primordia developed shoots. Although silver nitrate was undoubtedly effective in promoting shoot formation, continuous exposure to it caused symptoms of vitrification. Limiting the silver exposure to the first seven days on regeneration medium increased the average regeneration frequency to 49% (Table 1) and decreased symptoms of vitrification.
Several tissues in addition to hypocotyls were used as explant sources. Regeneration frequencies were up to 36% from petiole cross-sections, 15% from cotyledon and leaf explants, and 1% from root segments. Although we determined that several B. rapa tissues are capable of a morphogenic response, only hypocotyl explants were used to develop a transformation protocol. Plant transformation The A. tumefaciens hypocotyl transformation procedure de-
scribed previously (Radke et al~ 1988) and the regeneration procedure described above have been modified to improve transformation frequencies ofB. napus and to achieve routine wansformation of B. rapa. Kinetin was eliminated from the B5 callus induction medium and MS salts rather than B5 salts were used during the cocultivationperiod (results not shown). With these modifications transformedB, rapaplants have been produced routinely. In addition to the modifications described above other parameters were tested to improve the transformation frequency (percentage ofcocultivatedexplants that produced at least one transformed pianO. Kanamycin selection The antibiotic kanamycin was added to media after the twoday cocultivation period and maintained throughout the regeneration and rooting processes. When non-cocultivated control explants were selected on 25 mg/1kanamycin, yellow or white calli grew slowly and shoot regeneration was completely inhibited (Table 2). When only 10 mg/l kanamycin was used for selection, some green calli developed and pale green shoots regenerated from control explants. Because ot these escapes, the higher concentration (25 mg/1) was used during shoot regeneration from cocultivated explants even though the highest transformation frequency (9%) was obtained using the lower concentration of kanamycin (10 mg/l). Following selection of the explants on regeneration medium containing 25 mg/l kanamycin, the regenerated shoots were transferred to shoot maturation (B5-0) and rooting media containing 50 mg/l kanamycin. The higher concentration of kanamycin was used because control shoots rooted occasion-
502 Table 2. Comparison of parameters for the transformation of Brassica rapa cv. Emma hypocotyl explants. + Feeder Cells Kan. Conc. (mg/1)
2,4-D Exposure (Days)
Control pCGN 7001 pCGN 7001 pCGN7001
10 10 10 10
Control pCGN7001 pCGN 7001 pCGN7001
25 25 25 25
- Feeder Cells
KanR Callia (%)
KanR Shootsb (%)
Transformed Plants c (%)
10 6 8 10
5/12 66/82 64/87 67/86
(42) (80) (74) (77)
2/12 14/82 8/87 19/86
(17) (17) (9) (22)
0/82 3/87 8/86
10 6 8 10
0/12 8/87 78/85 74/82
(0) (93) (92) (90)
0/12 8/81 16/85 15/82
(0) (10) (19) (18)
1/81 5/85 5/82
KanR Callia
(%)
(0) (3) (9)
1/11 55/89 53/89 69/88
(9) (62) (60) (78)
(1) (6) (6)
0/1l 52/89 54/88 68/86
(0) (58) (61) (79)
KanR Shoots b (%) 1/11 7/89 7/89 6/88
(9) (8) (8) (7)
0/11 (0) 3/89 (3) 7/88 (8) 12/86 (14)
Transformed Plants c (%) 0/89 0/89 1/88
1/89 1/88
4/86
(0) (0) (1)
(1) (1) (5)
aNumber of explants with green kanamycin-resistant calli/total number of explants. bNumber of explants with green shoots that regenerated on kanamycin/total number of explants. CNumber of explants from which at least one green shoot rooted on kanamycin and tested positive for GUS or NPT II/total number of explants.
ally and the leaves remained partially green on 25 mg/1 kanamycin. Although some of the shoots which regenerated on 25 mg/1kanamycin were escapes, other shoots which were transformed were subsequently lost on the more stringent selection (50 rag/l) during the shoot maturation and rooting processes. These losses resulted from either low expression of the NPT II gene or inability to root. In the experiment shown in Table 2, the recovery of transformed Emma plants from regenerated shoots was about 33% from explants exposed to 2,4-D for 10 days and initially selected on 25 rag/1kanamycin. A more typical recovery rate for B. rapa is 17%. Among Tobin plants which were regenerated from cocultivated kanamycin-selectedexplants, 97% (182/187) tested positive for NPT II activity using the dot blot procedure (Fig. 1A). Maintaining selection throughout the regeneration and rooting process was essential to achieve the high percentage of rooted plants that were transformed. With these stringent selection conditions it would not be necessary to screen the plants for NPT II activity before continuing to analyze plants for the gene of interest. 2,4-D optimization To attain the highest transformation frequency, the duration of 2,4-D exposure was optimized as in the regeneration procedure discussed above. One of the experiments, summarizedin Table 2, shows that transformation frequencies were highest (3-9%) when hypocotyl explants were cultured on media containing 2,4-D (MS-1 followedby B5-1) for the first 8 or 10 days from culture initiation. In other experiments, no transformed plants were recovered when explants were incubated for only three days on 2,4-D and then placed onto regeneration medium (B5-BZ) directly after cocultivation. It was also observed that in general, more kanamycin-selected shoots regenerated when explants were cultured on 2,4-D for a total of 10 days as opposed to 8 days. The 10-day 2,4-D treatment for Emma and Tobin cocultivated explants is longer than the optimal 3-4 day Ireatment reported above for regeneration of
hypocotyl explants. StressfromAgrobacteriurninfectionas well as from selection on kanamycin may decrease the rate of cell division and the number of cells that can grow, resulting in the need for a longer exposure to 2,4-D to promote callus induction before transfer to cytokinin medium. OurB. napus results (unpublished) indicate that the duration of 2,4-D exposure may be optimizedfor differentgenotypes to achieve high transformation frequencies. In both cultivars tested, differences in regeneration response were observed for explants that were not selected on kanamycin after exposure to the stress of Agrobacterium. The regeneration frequency ofcocultivatedTobinexplants averaged 34% while non-cocultivated control Tobin explants averaged 15%. On the other hand, cocultivated Emma explants had frequencies of 37% while non-cocultivated controls had slightly higher frequencies of about 43%. We have also observed genotype-dependent responses to Agrobacterium infection from B. napus, in which regeneration of non-selected explants increased after cocultivation with Agrobacterium in some cultivars while it decreased or remained the same in others. Within a cultivar, regeneration frequencies could probably be increased for purposes of mutagenesis or transformation by optimization of the exposure to 2,4-D after a particular treatment. Feeder cells The use of tobacco feeder cells during the preincubation and two-day eocultivation periods increased the number of explants that produced green kanamycin-selected calli and increased the transformation frequency in many of the treatments shown in Table 2. Although feeder cells were not required to obtain transformed plants they were routinely used in our experiments. Transformation frequencies ranging from
