routine technique in plants (Fraley et al, 1986). The development of a crop specific protocol requires the determination of the best explant source, optimal.
Plant Cell Reports
Plant Cell Reports (1988) 7:100-103
© Springer-Verlag1988
Celery transformation by A grobacterium tumefaciens: cytological and genetic analysis of transgenic plants D. Catlin 1, O. Ochoa 1, S. McCormick 2, and C.F. Quiros 1 1 Department of Vegetable Crops, University of California, Davis, CA 95616, USA 2 Plant Gene Expression Center, USDA-ARS, Albany, CA 94710, USA Received October 19, 1987 / Revised version received December 22, 1987 - Communicated by L. K. Grill
ABBOT
Transgenic celery plants were obtained followlng co-cultivation of petiole explants with ~grobacterlum tumefaciens containing pMON200, a cointegrate vector carrying genes for kanamycln resistance and nopaline synthase. Transformants were selected by ability of callus to grow in the presence of 50mg/l kanamycin. Transformation was confirmed either by the presence of nopaline or by Southern blots. Cytological analysis of 14 transformed plants revealed chromosomal aberrations, both In structure and number. Only 20% of the regenerated plants had the normal karyotype. Kanamycln resistance behaved as a monogenlc, dominant trait, segregating in a 3:1 ratio in three families derived from plants with normal karyotypes.
/~IBREVIATIONS
Explant Source and Kanamycin Sensitivity We tested two explant sources, leaf and petiole. Leaf discs from the second or third true leaves (5 mm In diameter) or young petiole explants (cross sections 2-3mm thick) were surface sterilized (70% ethanol for I minute, 10% Chlorox for 5 minutes), and rinsed three times In sterile water. For the kanamyein sensitivity tests, explants were cultured on medium (medium K) composed of MS salts (Murashlge and Skoog, 1962), B5 vitamins (Gamborg et al, 1968), 3% sucrose, 100 mg/l serlne, 0.8% agar, 0.5 mg/l 2,4-D and 0.6 mg/l klnetin (Williams and Collin, 1976). Kanamycin sensitivity of these explants was determined by supplementing the medium with kanamycTn at concentrations of 0,25,50,75,100,200,300,400 and 500 mg/l. The cultures were maintained at 27°C in the dark. The sensitivity of celery seed germination to kanamycin was determined in a replicated trial on the same medium wlth the same range of kanamycln concentrations.
KB: Kilobases, 2-4D: 2,4-dlphenoxyacetic acid
Transformation procedure I NTRODL~TION Genetlc transformation by Agrobacterlum is becoming a routine technique in plants (Fraley et al, 1986). The development of a crop specific protocol requires the determination of the best explant source, optimal conditions for co-cultivatlon and regeneration and a usable selectable marker. Celery, Aplum 9raveolens L., Is a good candidate for transformetlon because of its abillty to regenerate In culture from tissue explants. In thls paper, we report successful transformation by co-culIivation of celery petiole sections wlth Agrobacterium tumefaclens carrying the Monsanto pMON200 vector. The resulting plants were analyzed cytologically and genetically In order to determine the inheritance of kanamycln resistance. NATERIALS AND NETHODS
PLANT NATERIAL Celery seedlings of an annual strain from Thailand (P1257228) were grown in the greenhouse under optimal conditions. This strain does not require vernalization for flowering and therefore has a shorter llfe cycle than commercial celery varieties, which are biennial. Thls annual strain can be crossed with the commercial varieties and its use accelerates genetics analysis.
Offprint requests to: C. F. Quiros
Petiole explants were co-cultivated with Agrobacterlum tumefaclens carrying the kanamycln resistance vector pMON200, following the techniques of Horsch et al (1985) and McCormick et al (1986). Co-cultivation with the pMON120 vector (lacking the kanamycin resistance marker), and unlnoculated celery explants were used as controls. Explants were dipped In a diluted (1:30 In sterile water) overnight culture of Aarobacterium for two minutes, blotted dry and transferred to callusing medium (medium C, containing either 50 or 100 mg/l kanamycin and 500 mg/i oarbeniclllin). Because Horsch et al (1985) and McCormick et al (1986) stated that feeder cell layers are beneficial for transformation, we tested four feeder protocols (Table I) during the two day co-cultivation period. Explants were transferred to new selection media every four weeks. The plates were cultured in the dark at 27°C. As calll developed, I-2 mm portions were transferred to shoot/root regeneration medium (R) and cultured under 16 hr light at 27°C. Medium R Is the same as medium C, but with 0.04 mg/l kinetin. Regenerated plants were transferred to GA7 containers (Magenta Corp., Chicago, IL, USA) containing R medium for further growth, transferred to vermiculite and then to the greenhouse. Transformation was Indicated by continued growth in the presence of kanamycln and by the presence of nopallne In callI and in the leaves of regenerated plants.
101 Table I.
Feeder layer treatments.
Treatment
Solid media
1 2 3 4
AI C2 C2 C2
Cell suspension
Filte~ paper
tobacc~ 3 celery none none
yes yes yes none
in death of the tissue. On the other hand, petiole sections grown in the absence of kanamycin produced large calli. The leaf disk controls were slower in callus productlon, or failed to grow in several instances. Based on the results of these experiments, we decided to use petiole explants for the transformation experiments, and selection medium containing 50 mg/l and 100 mg/l of kanamycln to identify putative transformed caIli.
Comparison of media salts
medium (Marton and Mallga, ~B5RMNOsalts, B5 vitamins, 100 mg/l
1975). L-serine, and hormone concentrations according to Williams and 3Collin (1976):0.5 mg/l 2,4-D and 0.6 mg/l klnetln. 45 ml cell suspension grown in A medium. 55 ml cell suspension grown in C medium. Sterile 8.5 cm Whatman filter paper.
Previous reports (summarized In Browers and Orton, 1986) had used MS salts-based media for celery tissue culture. Our Tnitial transformation experiments followed such recipes, with little success (data not shown). From experiments on anther culture of celery (Ochoa and Quiros, unpublished) we found that a B5 salt-based medium was significantly better than a MS-based medium. Therefore we used a B5 salts based medium for the transformation experiments.
DNA Analysis
Frequency of kanamycln resistant callus
DNA was isolated as in Bernatzky and Tanksley (1986), and Southern hybridization analyses were performed as In Maniatis et el. (1984), using BamH1 cut pMON200 as an oligolabeled probe (Feinberg and Vogelstein, 1985). Tomato DNA (gift of J. Yoder) from a plant transformed with pMON200 was used as a positive control. Celery DNA from untreated celery plants (PI 257228) was used as negative control.
Only the treatments involving co-cultlvation of petiole explants with Agrobacterium carrying the pMON200 vector yielded calli in the selection medium. None of the petioles inoculated with Agrobacterium carrying the pMON120 vector or non-inoculated petioles survTved in this medium. Kanamycin is therefore a usable selectable marker for celery transformation. The concentration of kanamycln in the medium affected the yield of resistant calli. The 50 mg/l concentration showed a two-fold increase in the frequency of resistant calli formation over the 100 mg/l concentration, and about half the number of days to first calli.
Nopallne assay Nopaline was assayed In c a l l i or leaves of regenerated plants f o l l o w i n g the technique of Otten and Schilperoorts (1978).
C~rtologlcal Chromosome counts of the plants regenerated from the kanamycin resistant calll were done in pollen mother cells. Flower buds were fixed overnight in Carnoy's fluid (Haskell and Willis, 1968), rinsed and stored in 70% ethanol at 5°C. Anther were dissected from the flowers and squashed In a drop of 2% acetocarmine. Chromosome counts and chromosomal associations were determined in diakinesls and metaphase I. Pollen fertility was calculated from the percentage of 100 pollen grains staining with 2% acetocarmine.
Inheritance of kenmycln resistance Selfed progenies were obtalned from p u t a t i v e transformed plants. Leaf sections from 30 seedlings per progeny were surface s t e r i l i z e d , plated on medium C supplemented with 50 mg/I of kanamycln and cultured at 27°C, 16 hrs. l i g h t . This t e s t was r e p l i c a t e d three tlmes. Five leaf sections from non-transformed plants were included in the plates as c o n t r o l s . The sectlons forming c a l l i were considered kanamycin r e s l s t a n t . Callus formation was scored a f t e r four weeks In c u l t u r e .
RESULTS AND DISCUSSION
The feeder layer treatments used affected the yield of kanamycin resistant callus. Interestingly, the treatments without suspension cells in the feeder layer resulted in higher frequency and a shorter time period to calli formation (Table 2). The Nicotiana feeder layer gave the lowest frequency of callus formation and longest period to first callus. These results suggest that the tobacco cells or RMNO medium interacted negatively with the explants. It is possible that the celery suspension cells (treatment 2) competed for nutrients. Our experience suggests that the most efficient transformation procedure for celery is the co-cultivatlon of Aarobacterium inoculated petiole explants on C medium with no feeder layer, followed by transfer to selection media (C) containing 50 mg/l of kanamycin and 500 mg/l carbenlcll!in. Calli growing in this media were checked for the presence of nopaline. Nopallne positive calli were transferred to medium R without kanamycin or carbenicillin for plant regeneration. Table 2.
Recovery of kanamycln resistant calli, scored after 3 months in culture.
Co-cultivation Treatment
Ken (mg/L)
% Resist. Calli
Days to Ist callus
22.4 10.8 16.4 36.1 43.6 17.2
39 44 46 14 20 51
Tolerance of celery to kana~/cln Concentrations above 50 mg/l of kanamycin were detrimental to celery seedling growth. Although the seed germinated normally on all kanamcyin concentratlons, 75% or more of the seedlings in the treatments with 50 mg/l kanamycin showed chlorophyll deficiency 3 weeks after germination. Leaf disks and petiole sections started to show signs of kanamycin toxicity above 25 mg/l. Doses above 100 mg/l yielded very little callus formation, which stopped growing after 45 days in culture and turned yellow, resulting
I 12 2 33 4 4
i
50 100 100 50 50 100
See Table I Data from 50 kan lost due to contamination Data from 100 kan lost due to contamination
102 Plant regeneration Celery regeneration proceeds via somatic embryogenesis. Plant regeneration from all the kenamycin resistant calll was not attempted. A total of 20 plants were regenerated, 17 derived from feeder layer treatment I with 50 mg/L kenamycin In the selection media, one from treatment I but with 100 mg/L kanemycln, end two from treatment 3 with 50 mg/L kanamycln in the selection medium. We chose these callI for regeneration because they were available earlier than those generated by the higher calli yielding treatments performed later. Although the petiole sections formed callus readily, they took from 4 to 6 months to regenerate Into plants. Morphological abnormalities, such as chlorophyll deficiency and leaf shape distortion, were commonly observed in most the regenerated plants which survived transplanting to soil. We have recently optimized celery regeneration by maintaining callus cultures in the light and transferring explants to fresh media every 2 weeks, as soon as somatic embryos are visible. With these conditions we believe that celery transformation could be accomplished within 3 months.
DNA Analysis
The regenerated plants were Identified as transgenic by the presence of nopaline in their leaves (Table 3). To confirm that these plants were transgenic, Southern blots were performed. Four putative transgenic plants tested positive for the presence of the 3.7 kb BamHI Internal fragment (Fig. I). These plants showed two major border fragments and the expected 3.7 kb BBmHI internal fragment indicating that a single copy of T-DNA was inserted. One of these 4 plants (87B339) was negative for the nopaline marker, as has been occasionally reported for other species (Fraley et el., 1986, Chyi et el., 1986).
Table 3.
Cytological and transgenic characteristics of regenerated plans.
Plant #
2n
T-DNA I
Nopallne
Fertility
(%)
86A317 86A318
22 23
NA NA
+ +
30 6
86A338 86A339" 86A340 86A342. 86A343 86A345" 86A352 86A363 86A364 87A001 87A002
21 22 22 22 22 22 24 NA 22 44 22
87A003
22
NA + + NA NA NA NA + NA + NA NA
+ + + + + + + + + + NA
8 89 60 87 76 98 11 38 54 5 0 53
~tert NA not assayed t. = tertiary
F e r t i l l t y and chromosomal constitution of transganic plants Most of the transgenic plants had low fertility, as measured both by pollen staining and seed setting. Cytological analysis revealed abnormalities In chromosome structure and number (Table 3). Although about 64% of these plans were diploid, only one third were normal, the rest displayed a variable number of chromosomal translocations (Fig. 2a). Aneuploidy was seen In about 21% of the transgenlc plants, represented by either loss (Figs. 2b, 2c) or galn of chromosomes (Fig. 2d). Only one t e t r a p l o i d was observed among the transgenic plants (Fig. 2e). These aberrations resulted In a high frequency of micronuclei at the end of meiosis (Fig. 2f). Salted progenies were obtained from the normal diploids, allowing us to study the inheritance of kenamycln resistance. High degree of fertility was observed In these salted progenies. The high frequency of chromosome variants observed in this experiment is consistent with the findings of Orton (1985), who studied the karyotypes of celery cells after 6 and 12 months in culture. After 12 months in culture, he observed in the cells almost complete divergence from the original karyotype, Thus It is likely that the chromosomal changes observed In our experiment are due to the lengthy period spent In culture before regeneration, and not to the transformation procedure.
1
2
3
4
kb
Aberration
3.7 Transloc(1) Transloc(1) tert t. Monosomlc Normal Transloc(1) Normal Transloc(1) Normal Tetrasomlc NA Transloc(2) Tetraploid NA Transloc(1)
trisomlc, transloc = translocatlon, number of translocations in parenthesis *Progenies grown for kanamycin resistance Inheritance study
Fig. I . Southern showing the presence of T-DNA BamH1 Insert In transgenic celery plant (lane 4 from the l e f t ) . Controls: pMON200 (lane I ) , transgenlc tomato (lane 2), untreated celery (lane 3).
Inheritance of kanamycin reslstance A leaf callus assay was used t o determine the Inheritance of kanamycln resistance. Table 4 shows the Inheritance data.
103 Table 4.
SUIMNARY
Segregation for kanamycin resistance in progenies from three transgenic plants.
Origin
Resistant
Susceptible
X 2
P
86A339 87A342 87A345
26 24 24
4 6 6
2.17 0.40 0.40
.12 .50 .50
All three families fit the expected ratio for a monogenic trait segregating In a Mendelian fashion, confirming the presence of a single copy of the T-DNA insert detected in the Southerns. Kanamycln resistance behaved as a dominant trait which Is consistent with the findings reported for other plants (McCormick et el. 1986, Chyi et ai.1987). The leaf explants from kanamycin resistant plants started to form callus after approximately two weeks in culture, while the sensitive ones, including controls, dld not show any growth.
Thls paper demonstrates that celery Is amenable to transformation. The occurrence of chromosomal aberrations is of concern, but might be avoided or reduced by emphasis on optimized and rapid regeneration/transformation techniques. Transformatiod will be an attractive tool for the breeder to use in transferring useful traits to celery (Fischhoff et el., 1987). We plan to use the single copy T-DNA Inserts as additional markers in establishing a linkage map for celery.
^CKNOWI_EDGEI~NTS The authors are Indebted to Robert Fraley for supplying the Monsanto vectors; to John Yoder for the transgenic tomato DNA; to Cerole Meredith for the tobacco suspension cells; to Janet Stites and VInce D'Antonio for technical assistance; and to Jane Johnson for typing the manuscript. This work was supported by grants from the California Advisory Celery Board (QUI-7-86) and BARD (I-483-82) to CFQ.
REFERENCES
Brower MA and Orton TJ (1986) In: Bajaj PS (ed) BIotechnology In Agriculture and Forestry. Vol. 21 Crops I. Sprlnger-Verlag, Berlin, pp 405-420 ChyI YS, Jorgensen AJ, Goldsteln D, Tanksley SD, Loai za-FIgueroa F (1986) Mol. Gen. Genet. 204:64-69 Felnberg AP, Vogelsteln B (1983) An. Biochem. 13216-13. Fischoff DA, Bowdlsh KS, Perlak FJ, Marrone PG, McCormick SM, NIedermeyerJG, Dean DA, Kusano-Kretzmer K, Mayer EJ, Rochester DE, Rogers SG, Fraley RT. (1987) BIo/technology. 5(8):807-813 FraJey RT, Rogers SG, Horsch RB, Sanders PR, Flick JS, Adams SP, BIttner ML, Brand LA, Fink CL, Fry JS, GalluppI GR, Goldbert SB, Hoffmann NL, Woo SC (1983) PNAS 8014803-4807 Fraley TF, Rogers SG, Horsch RE} (1986) In: CRC Critical Reviews In Plant Sciences Vol 4 issue I pp 1-46 Gemborg OL, Miller RA and OJlma K (1986) Expt. Cell Res. 501151-158. Haskell G, Wills AB (1986) Primer of Chromosome Practice, O l i v e r & Boyd, London Horsch RB, Fry JE, Hoffmann NL, E i c h o l t z D, Rogers SG, Fraley RT (1985) Science 227:1229-1231 Marton and Mallga, P (1975) Plant Sci. L e t t . 5:577-581 McCormick S, Niedermeyer J, Fry J, Barnason A, Horsch R, Fraley R (1986) Plant Cell Tissue Organ Culture 4:159-169 Otten LA, Schilperoort RA (1978) Biochem Biophys Acta 527:497-500 Williams L and Collln HA (1976) Ann. Bot. 40:32-35.
Fig. 2. Pollen mother c e l l s of transgenlc plants. 2a: Plant 86A318 d i s p l a y i n g reciprocal t r a n s l o c a t i o n a t d i a k i n e s l s manifested by the presence of a quadruple. 2b: Monosomic plant 86A338. The chromosome associates s i n g l e wlth a normal p a i r forming a t r i v a l e n t . 2c1 Same p l a n t , t r i v a l e n t lagging a t metaphase I I . 2d: Tetrasomic p l a n t 86A352 at telophase I1 with e x t r a chromosomes as laggards. 2e: T e t r a p l o i d plant 87A001 a t d l a k l n e s l s . 2f: Micronuclei commonly observed In aberrant transgenic plants at the end of meiosis.
