Plant Cell Reports
Plant Cell Reports (1995) 15:102-105
© Springer-Verlag1995
Transformation of the wild tomato Lycopersicon chilense Dun. by Agrobacterium tumefaciens Zahra Agharbaoui, Ann Francine Greer, and Zohreh Tabaeizadeh Department of Biological Sciences, Universityof Quebec in Montreal, P.O. Box 8888, Station "Centre-ville", Montreal, QC, H3C 3P8, Canada Received 26 September 1994/Revised version received 13 December 1994 - Communicatedby I. K. Vasil
Summary. Leaf disc transformation-regeneration
Jones 1987).
technique was applied to the drought tolerant wild relative of cultivated tomato, Lycopersicon chilense, using a plasmid construct which contained the coding sequences of neomycin phosphotransferase (NPTII) and chloramphenicol acetyltransferase (CAT) genes. The two genotypes used, LA2747 and LA1930, showed a distinct difference in their aptitude to transformation; a higher success rate was obtained for the first genotype in every stage of the process. Shoots were formed on the regeneration medium containing 100 ~tg/ml kanamycin through direct or indirect organogenesis. Root formation became only possible when the concentration of kanamycin was reduced to 50 Ixg/ml. Expression of chloramphenicol acetyltransferase gene was observed in all of the kanamycin-screened plantS after they matured; the activity of the gene was absent or low in some of the young plants. The presence of the CAT gene in transgenic plants was further confirmed by Southern blot analysis. Although transgenic plants grew to maturity, they did not produce fruit, owing to the self incompatibility of L. chilense.
e s c u l e n t u m and L. chilense has not been successful
Key words: Agrobacterium tumefaciens - Lycopersicon chilense - Regeneration - Transformation.
Abbreviations: BAP, 6-benzylaminopurine; CAT, chloramphenicol acetyltransferase; 2,4-D, 2,4dichlorophenoxyacetic acid; IAA, indole-3-acetic acid; LB, Luria Broth; EDTA, ethylenediamine-tetraacetic acid. Introduction chilense Dun., a wild relative of the cultivated tomato, is a natural inhabitant of extremely arid regions in South America. It has been found to tolerate drought and saline conditions (Rick 1973;
Lycopersicon
Correspondence to: Z. Tabaeizadeh
Sexual hybridization between L.
owing to interspecific incompatibility (Martin 1961; Rick 1973; V. Poysa, personal communication). With the goal of understanding the molecular mechanisms of drought tolerance in L. chilense, we have isolated three groups of drought- and ABA-induced cDNAs from this species (Chen and Tabaeizadeh 1992). Two of these cDNA clones have been characterized. The first, pLC30-15 shares few common characteristics with several desiccation and ABA induced genes isolated from other higher plants. However, the encoded protein is unique due to a high content of glutamic acid and the presence of three lysine rich repeats in its primary structure (Chen et al. 1993). Phylogenetic analysis indicated that this protein, compared to other ABA and drought induced proteins, belongs to a separate category that diverged early in the course of evolution (Bellemare and Tabaeizadeh, unpublished data). Characterization of the second cDNA, pcht28, revealed that it encodes an acidic chitinase; this is the first known drought-and-ABA induced chitinase gene (Chen et al. 1994). Although these genes are induced by drought stress and their level of expression is proportional to the degree of drought tolerance (comparing different genotypes of L. chilense among each other as well as with the cultivated tomato), it is yet to be demonstrated if they play a role in drought tolerance. A logical approach to achieve this goal would be to produce and subsequently compare L. chilense plants in which the expression of these genes is increased or reduced. Therefore, the availability of an efficient genetic transformation system in this species is necessary. The cultivated tomato has been transformed in several laboratories (McCormick et al. 1986; Chyi et al. 1986; Filatti et al. 1987; van Rockel et al. 1993; Lipp Joao and Brown 1993). However, no such report exists for L. chiIense. The present study was undertaken to develop an efficient procedure for the production of transgenic plants in L. chilense which although not edible, provides a valuable genetic source of agronomically useful traits.
103 Materials and Methods P/ant materla/. Seeds of L chilense (LA1930 Peru and LA2747 Chifi), kindly provided by Dr. C.M. Rick (Tomato Genetics Stock Center, University of California, Davis, CA), were surface sterilized and sown on moist filter paper in Petri dishes and kept at 23 ° C in the dark for 4 days. After 9 days, seedlings were transferred to soil and kept in a growth chamber under a 16-h photoperiod (150-180 mE m'2s -1 ) at a temperature of 23 ° C during the day and 19° C at night with a relative humidity of 55%. To avoid the problem of genetic variability among the plants, one plant (for each genotype) was propagated by cuttings and the resulting clones were used for transformation experiments approximately 3 months after the cuttings were initiated.
Plasmid construct. The construct used contained the coding sequence of chloramphenicol acetyltransferase (CAT) gene finked to 19S promoter from cauliflower mosaic virus. This cassette had been cloned into the binary vector pBIN19 (Bevan 1984) a n d then introduced into Agrobacterium tumefaeiens strain LBA4404. The map and details on the construction of this plasmid have been presented by Delatmey eta/. (1988).
Lyeopersicon esculentum (LA157) cell suspensions established and maintained in MS (Murashige and Skoog 1962) medium supplemented with 2 mg/l 2,4-dichlorophenoxyacetic acid (Greer and Tabaeizadeh, unpublished results) were used. They were kept on a gyratory shaker (100 rpm) at 23-25°C under a 16-h photoperiod (15-16 mE m-28-1) and subcultured weekly. Feeder layer plates were prepared 3-4 hours before cocultivation experiments by pipetdng 2 ml of a suspension culture (three days after subculture) onto a plate containing solidified culture medium comprised of MS (Murashige and Skoog 1962) salts and vitamins, 3% sucrose, 0.65% agar and appropriate growth regulators (pH 5.7). Cells were evenly distributed on the plate and covered with a piece of Whatman No. 1 filter paper.
CAT assays. The reaction mixture was prepared as described by Fromm et al. (1985). Reaction products were separated by thin-layer chromatography and TLC plates were autoradiographed for 24 hours at room temperature.
Southern blot analysis. Nuclear DNA of leaf tissue from transformed and untransformed plants was isolated according to a technique developed by Dr. S. Tanksley's laboratory (personal communication). Briefly, nuclei were first isolated from leaves using a buffer containing 0.35M sorbitol, 0.1M Tris-HC1 (pH 7.5), 5mM EDTA and 20 mM sodium bisulfate. The nuclei were then treated with nuclear lysis buffer containing IM Tris-HC1 (pH 7.5), 0.25M EDTA, 5M NaC1 and 20 g/l Hexadecyl trimethyl-ammonium bromide (CTAB). After chloroform/octanol extraction the DNA was precipitated with isopropanol. The DNA pellet was washed with ethanol, dried, and dissolved in TE buffer (Maniatis et al. 1982). 20 Ixg of DNA was digested with KpnI, fractionated on a 0.8% agarose gel electrophoresis, blotted overnight onto a Hybond-N membrane and hybridized with 32p. labelled CAT fragment according to Maniatis eta/. (1982). Filters were washed twice at room temperature and once at 65 ° C in SSC buffer (Maniatis eta/. 1982). Lambda DNA digested with HindlII was used as size marker.
Preparation of feeder layer.
Transformation procedure. The leaf disk transformation procedure of Horsch eta/. (1985) was used with some modifications. A. tumefaciens containing the transformation vector was cultured overnight at 28 ° C in liquid LB medium containing 50 Ixg/ml kanamycin and 50 I.tg/ml streptomycin, pH 7.5. The bacterial suspension was then centrifuged at 1000 g for 15 minutes and the pellet was diluted to an O.D. reading of 0.5 at 560 nm, using MS liquid medium without hormones. The 2nd, 3rd and 4th leaves were excised from plants, surface sterilized in a 15% solution of sodium hypochlorite for 15 minutes and washed three times with sterilized bidistilled water. Leaves were then cut into 0.5 cm 2 segments. Explants were submerged in the diluted A. tumefaciens inoculum and carefully agitated for 2-3 minutes. They were then blotted dry on sterile filter paper and transferred to the feeder culture. Petri dishes were kept at 24-26 ° C under a 16 h photoperiod and the light intensity of 50-70 mE m-2 s-1. After 3 days on the feeder culture, leaf segments were transferred to Petri dishes containing the same medium supplemented with 100 I.tg/ml kanamycin and 500 I.tg/ml carbenicillin. As controls, uninoculated explants were cultured in the same manner on selection medium containing antibiotics and on medium without antibiotics. Regenerated shoots were transferred to the rooting medium (MS medium without hormones) containing 50-100 Ixg/ml kanamycin and 500 ~tg/ml carbenicillin. Rooted plantlets were acclimated for 2-3 weeks after being transferred to sterile soil. They were kept in a growth chamber under a 16-h photoperiod at (150-200 mEre-28-1) at 24:20 ° C (fight:dark). Chloramphenicol acetyltransferase (CAT) assay. Selected leaf tissue (200 rag) was frozen in liquid nitrogen and macerated to a fine powder. Tissue was suspended in 200 I-tl of 250 mM Tris-HC1 buffer pH 7.8 and spun in a microfuge for 5 minutes where the debris was pelleted. Supernatant was heated for 10 minutes at 65 ° C. Protein concentrations were determined and aliquots containing 200 I.tg protein were used for
Results
Genetic transformation and regeneration of transgenic plants. In a series of preliminary experiments, shoot production from leaf explants of the two genotypes was verified using various culture media. These media differed from each other with regards to the composition and concentration of supplemented growth regulators. Based on the results obtained, the medium MSZ (containing 2 mg/l zeatin) which induced a high percentage of direct organogenesis was used for the two genotypes. Moreover, media MS4 (containing 2.25 mg/l BAP and 1.75 mg/l IAA) and MSB (containing 3 rag/1 IAA and 2 mg/1 BAP) which promoted callus and shooting were also used for LA1930 and LA2747, respectively. On MSZ medium supplemented with 100 ~tg/ml kanarnycin, shoot primordia were apparent by the eighth week of transformation on inoculated explants of both genotypes. This concentration of kanamycin was chosen based on preliminary dose response experiments (data not shown). Explants from LA2747 cultured on this medium demonstrated a higher capacity for shoot formation when compared to LA1930 (Fig. 1A). Large numbers of explants also produced light yellow-green soft calli on this medium. Some of these calli regenerated shoots when dissected from explants and recultured on the same medium. On MS4 medium which was used for LA2747, 98% of the explants formed calli and 60% of them produced between 1-3 shoot primordia. On MSB medium which was used for LA1930, 88% of explants produced calli, but only 8% of them demonstrated the ability for direct organogenesis. Calli produced on both MS4 and MSB media were dissected and transferred to MSZ medium where approximately 60% of those from LA2747 produced shoots; the number of shoots produced from LA1930 was below 1%. No callus or shoot
104 primordia were obtained from the uninoculated explants on the above mentioned media (containing 100 ~tg/ml kanamycin).
was found to be more recalcitrant; a large number of plants degenerated gradually soon after being transferred to soil. Only 30% of the plantlets survived the acclimatization period. Acclimatized plants grew to maturity (Fig. 1C) and produced flowers; however, no fruits were formed because of self incompatibility. No morphological differences were observed between transformed and control (in vitro and soil grown) plants.
Biochemical and molecular characterization of transformed plants: The expression of chloramphenicol acetyltransferase gene was verified in transformed plants. Initially a few plants were subjected to CAT assays shortly after the acclimatization period (approximately 2 months after transfer to soil). Most of these plants showed low CAT activity and some of them did not express the gene (data not shown). However, when the same plants were tested again after 6 months, using identical amounts of protein as in the first experiment, they showed 2-3 fold higher CAT activity (Fig. 2A). No activity was detected in untransformed plants. Y2Y5 R2
Rs
Re
Rs I:tl0 Rll R12 Rls R20
C
3 acI ac-
c-
A
Fig. 1. Leaf disc transformation regeneration in Lycopersicon chilense LA2747. (A) Shoot and callus formation from inoculated leaf explants on MSZ medium supplemented with 100 Ixg/rnl kanamycin, 10 weeks after transformation (B) A transgenic plantlet 8 weeks after transfer to the rooting medium containing 50 Itg/ml kanamycin. (C) A transgenic plant 5 months after transfer to soil.
Regenerated shoots of approximately 1 cm were dissected from leaf explants or calli and transferred to MSZ medium where they were kept for 4-5 weeks to enhance elongation. They were then transferred to the rooting medium. Initially, the rooting medium included 100 ~tg/ml of kanamycin. However, none of the putative transformed shoots produced roots on this medium, even though they survived for a long period of time. When the concentration of kanamycin was reduced to 50 ltg/ml, approximately 56% of LA2747 and 25% of LA1930 shoots produced roots after 6-8 weeks (Fig. 1B). In the controls, shoots produced on the nonselective medium showed frequencies of rooting (on the medium without kanamycin) which were almost the same as those recorded for transformation experiments. As expected, these shoots were not able to produce roots on medium containing 50 ~g/ml kanamycin. Rooted plantlets were transferred to soil where 80% of those of LA2747 survived. At this stage once again, the genotype LA1930
C1
Ys
Y2
R12
Rs
C2
kb 0.91 0.79
B Fig. 2. Biochemical and molecular analysis of transgenic plants. (A) Expression of CAT gene in LA1930 (Y) and LA2747 (R) transgenic plants; c, chloramphenicol; lae, 1-acetate chloramphenicol; 3ac, 3acetate chloramphenicol. Lane C is the extract from untransformed plant. (B) Southern blot of transformed and untransformed L. chilense plants probed with 32p labeled CAT fragment. C1, untransformed plant; Y5 and Y2, LA1930 transformed plants; R12 and R6, LA2747 transformed plants. Lane C2 contains 39 pg of the CAT fragment.
Some of the plants which expressed CAT activity were further characterized by Southern blot analysis, using the CAT fragment as the probe. The analysis confirmed that CAT activity was correlated with the presence of the introduced CAT gene. KpnI digested genomic DNA from transgenic plants contained the diagnostic 0.79 kb CAT fragment that comigrates with the fragment used as the probe (Fig. 2B). An additional fragment in one of the transgenic plants (Y5) represents the rearranged copy of
105 the integrated gene. DNA from untransformed plants did not contain sequences that hybridize to the CAT probe. Discussion
Owing to our interest in the genetic potential of Lycopersicon chilense, we have verified the capacity for in vitro plant regeneration of this species in the past (Greer and Tabaeizadeh 1991). During the course of the present study, we found that even though L. chilense had demonstrated a higher capacity for plant regeneration, compared to those reported for the cultivated tomato, it seemed more recalcitrant during the transformation process (from inoculation of leaf explants to production of mature plants). The density of Agrobacterium used for the inoculation of the explants was found to be critical. Low inoculum concentration (O.D. reading of 0.3-0.4) resulted to low ferequency of transformation events judging from the number of calli and shoot primordia produced from the inoculated leaf discs. On the other hand, in those experiments where the concentration of the inoculum was high (O.D. reading higher than 0.5), we were not able to stop the growth of Agrobacterium after transfer of the explants to the media containing carbenicillin. In fact, all of the explants resulting from these experiments gradually degenerated and died. Shoots produced on the selection medium were not able to produce roots in the presence of 100 I.tg/ml kanamycin. The fact that the uninoculated explants did not show any sign Of calli or shoot primordia on kanamycin containing medium strongly indicate that the shoots produced under selection were transgenic. In one of the reported studies on genetic transformation of cultivated tomato (van Roekel et al. 1993), 33 ~tg/ml of kanamycin was used in the rooting medium of shoots which had been produced on 100 ~tg/ml of kanamycin. The authors, however, did not indicate whether they tested higher concentrations of kanamycin in the rooting medium. Another noticeable point of this study was that in some young transgenic plants the CAT activity was either low or absent. However, the activity increased significantly once the plants matured. It is not known whether the problem of CAT expression in young transgenic plants was at the transcriptional or translational level. The loss of gene expression in transgenic plants has also been reported for petunia (Horsch et al. 1985) and tomato (McCormick et al. 1986). As expected, the transgenic plants did not produce fruits owing to the self incompatibility of L. chilense (McGuire and Rick 1954; Taylor 1986). Cross pollination between these plants and untransformed plants used in this study, could not be effective since they all originated from one single clone. Progenies could probably be produced by pollination of the transgenic plants with a mixture of pollen, derived from different seed grown L. chilense plants. However, the existing polymorphism among
different L. chilense individuals (inside each genotype) could make the interpretation of the results very difficult, if not impossible. This problem of genetic diversity has also been discussed by Bonnema and O'Connell (1992) in a report regarding somatic hybridization of L. chilense and cultivated tomato. The analysis of progenies is an important part in genetic transformation studies. However, it should be recognized that L. chilense is a wild species. Thus, it does not present the same type of interest, compared to crop plants, with regard to genetic improvement, using transformation approaches. We believe that the obstacle in production of progeny from L. chilense transgenic plants will not affect our ongoing research for demonstrating the involvement of the genes that we have isolated, in the tolerance of this species during drought stress. Transgenic plants were propagated and maintained by cuttings. Some of the resulting cuttings (randomly chosen) were successfully micropropagated using kanamycin (100 ~tg/ml) containing medium.
Acknowledgements. This research was supported in part by a grant from the Natural Sciences and Engineering Research Council of Canada. References Bevan M (1984) Nucl. Acids Res. 12:8711-8721. Bonnema AB, O'Connell (1992) Plant Cell Rep. 10:629-632. Chen RD, Tabaeizadeh Z (1992) Biochem. Cell Biol. 70:199-206. Chert RD, Campeau N, Greer AF, Bellemare G, Tabaeizadeh Z (1993) Plant Physiol.103:301. Chen RD, Yu LX, Greer AF, Cheriti H, Tabaeizadeh Z (1994) Mol. Gen. Genet. 245:195-202. Chyi YS, Jorgenson RA, Goldstern D, Tarksley SD, Loaiza-Figueroa F (1986) Mol. Gen. Genet. 204:64-69. Delauney AJ, Tabaeizadeh Z, Verma DPS (1988) Proc. Natl. Acad. Sci. USA 85:4300-4304. Filatfi JJ, Kiser J, Rose R, Luca C (1987) Bit/Technology 5:726-730. Greer AF, Tabaeizadeh Z (1991) Can. J. Bot. 69:2257-2260. Horsch RB, Fry JE, Hoffmann NL, Eichholtz D, Rogers SG, Fraley RT (1985) Science 227:1229-1231. Fromm M, Taylor LP, Walbot V (1985) Proc. Natl. Acad. Sci. USA. 82:5824-5828. Jones RA (1987) In: Nevis DJ, Jones RA (eds) Tomato Biotechnology: proceedings of a symposium at the University of California, Davis, Alan R. Liss, Inc., New York, pp 125-137. Lipp Joao KH, Brown TA (1993) Plant Cell Rep. 12:422-425. McCormick S, Niedermeyer J, Fry J, Barnason A, Horsch R, Fraley R (1986) Plant Cell Rep. 5:81-84. McGnire DC, Rick CM (1954) Hilgardia 23:101-124. Maniatis T, Fritsch EF, Sambrook J (1982) Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Martin FW (1961) Genetics 46:1443-1454. Murashige T, Skoog, F (1962) Physiol. Plant. 15:473-497. Rick CM (1973) In: Srb AM (ed) Genes, Enzymes, and Populations, Plenum Press, New York, pp 255-269. van Roekel JSC, Damm B, Melchers LS, Hoekema A (1993) Plant Cell Rep. 12:644-647. Tylor IB (1986) In: Atherton JG, Rudich J (eds) The tomato crop: A scientific basis for improvement, Chapman and Hall, New York, pp 134.
Plant Cell Reports (1995) 15:102-105
© Springer-Verlag1995
Transformation of the wild tomato Lycopersicon chilense Dun. by Agrobacterium tumefaciens Zahra Agharbaoui, Ann Francine Greer, and Zohreh Tabaeizadeh Department of Biological Sciences, Universityof Quebec in Montreal, P.O. Box 8888, Station "Centre-ville", Montreal, QC, H3C 3P8, Canada Received 26 September 1994/Revised version received 13 December 1994 - Communicatedby I. K. Vasil
Summary. Leaf disc transformation-regeneration
Jones 1987).
technique was applied to the drought tolerant wild relative of cultivated tomato, Lycopersicon chilense, using a plasmid construct which contained the coding sequences of neomycin phosphotransferase (NPTII) and chloramphenicol acetyltransferase (CAT) genes. The two genotypes used, LA2747 and LA1930, showed a distinct difference in their aptitude to transformation; a higher success rate was obtained for the first genotype in every stage of the process. Shoots were formed on the regeneration medium containing 100 ~tg/ml kanamycin through direct or indirect organogenesis. Root formation became only possible when the concentration of kanamycin was reduced to 50 Ixg/ml. Expression of chloramphenicol acetyltransferase gene was observed in all of the kanamycin-screened plantS after they matured; the activity of the gene was absent or low in some of the young plants. The presence of the CAT gene in transgenic plants was further confirmed by Southern blot analysis. Although transgenic plants grew to maturity, they did not produce fruit, owing to the self incompatibility of L. chilense.
e s c u l e n t u m and L. chilense has not been successful
Key words: Agrobacterium tumefaciens - Lycopersicon chilense - Regeneration - Transformation.
Abbreviations: BAP, 6-benzylaminopurine; CAT, chloramphenicol acetyltransferase; 2,4-D, 2,4dichlorophenoxyacetic acid; IAA, indole-3-acetic acid; LB, Luria Broth; EDTA, ethylenediamine-tetraacetic acid. Introduction chilense Dun., a wild relative of the cultivated tomato, is a natural inhabitant of extremely arid regions in South America. It has been found to tolerate drought and saline conditions (Rick 1973;
Lycopersicon
Correspondence to: Z. Tabaeizadeh
Sexual hybridization between L.
owing to interspecific incompatibility (Martin 1961; Rick 1973; V. Poysa, personal communication). With the goal of understanding the molecular mechanisms of drought tolerance in L. chilense, we have isolated three groups of drought- and ABA-induced cDNAs from this species (Chen and Tabaeizadeh 1992). Two of these cDNA clones have been characterized. The first, pLC30-15 shares few common characteristics with several desiccation and ABA induced genes isolated from other higher plants. However, the encoded protein is unique due to a high content of glutamic acid and the presence of three lysine rich repeats in its primary structure (Chen et al. 1993). Phylogenetic analysis indicated that this protein, compared to other ABA and drought induced proteins, belongs to a separate category that diverged early in the course of evolution (Bellemare and Tabaeizadeh, unpublished data). Characterization of the second cDNA, pcht28, revealed that it encodes an acidic chitinase; this is the first known drought-and-ABA induced chitinase gene (Chen et al. 1994). Although these genes are induced by drought stress and their level of expression is proportional to the degree of drought tolerance (comparing different genotypes of L. chilense among each other as well as with the cultivated tomato), it is yet to be demonstrated if they play a role in drought tolerance. A logical approach to achieve this goal would be to produce and subsequently compare L. chilense plants in which the expression of these genes is increased or reduced. Therefore, the availability of an efficient genetic transformation system in this species is necessary. The cultivated tomato has been transformed in several laboratories (McCormick et al. 1986; Chyi et al. 1986; Filatti et al. 1987; van Rockel et al. 1993; Lipp Joao and Brown 1993). However, no such report exists for L. chiIense. The present study was undertaken to develop an efficient procedure for the production of transgenic plants in L. chilense which although not edible, provides a valuable genetic source of agronomically useful traits.
103 Materials and Methods P/ant materla/. Seeds of L chilense (LA1930 Peru and LA2747 Chifi), kindly provided by Dr. C.M. Rick (Tomato Genetics Stock Center, University of California, Davis, CA), were surface sterilized and sown on moist filter paper in Petri dishes and kept at 23 ° C in the dark for 4 days. After 9 days, seedlings were transferred to soil and kept in a growth chamber under a 16-h photoperiod (150-180 mE m'2s -1 ) at a temperature of 23 ° C during the day and 19° C at night with a relative humidity of 55%. To avoid the problem of genetic variability among the plants, one plant (for each genotype) was propagated by cuttings and the resulting clones were used for transformation experiments approximately 3 months after the cuttings were initiated.
Plasmid construct. The construct used contained the coding sequence of chloramphenicol acetyltransferase (CAT) gene finked to 19S promoter from cauliflower mosaic virus. This cassette had been cloned into the binary vector pBIN19 (Bevan 1984) a n d then introduced into Agrobacterium tumefaeiens strain LBA4404. The map and details on the construction of this plasmid have been presented by Delatmey eta/. (1988).
Lyeopersicon esculentum (LA157) cell suspensions established and maintained in MS (Murashige and Skoog 1962) medium supplemented with 2 mg/l 2,4-dichlorophenoxyacetic acid (Greer and Tabaeizadeh, unpublished results) were used. They were kept on a gyratory shaker (100 rpm) at 23-25°C under a 16-h photoperiod (15-16 mE m-28-1) and subcultured weekly. Feeder layer plates were prepared 3-4 hours before cocultivation experiments by pipetdng 2 ml of a suspension culture (three days after subculture) onto a plate containing solidified culture medium comprised of MS (Murashige and Skoog 1962) salts and vitamins, 3% sucrose, 0.65% agar and appropriate growth regulators (pH 5.7). Cells were evenly distributed on the plate and covered with a piece of Whatman No. 1 filter paper.
CAT assays. The reaction mixture was prepared as described by Fromm et al. (1985). Reaction products were separated by thin-layer chromatography and TLC plates were autoradiographed for 24 hours at room temperature.
Southern blot analysis. Nuclear DNA of leaf tissue from transformed and untransformed plants was isolated according to a technique developed by Dr. S. Tanksley's laboratory (personal communication). Briefly, nuclei were first isolated from leaves using a buffer containing 0.35M sorbitol, 0.1M Tris-HC1 (pH 7.5), 5mM EDTA and 20 mM sodium bisulfate. The nuclei were then treated with nuclear lysis buffer containing IM Tris-HC1 (pH 7.5), 0.25M EDTA, 5M NaC1 and 20 g/l Hexadecyl trimethyl-ammonium bromide (CTAB). After chloroform/octanol extraction the DNA was precipitated with isopropanol. The DNA pellet was washed with ethanol, dried, and dissolved in TE buffer (Maniatis et al. 1982). 20 Ixg of DNA was digested with KpnI, fractionated on a 0.8% agarose gel electrophoresis, blotted overnight onto a Hybond-N membrane and hybridized with 32p. labelled CAT fragment according to Maniatis eta/. (1982). Filters were washed twice at room temperature and once at 65 ° C in SSC buffer (Maniatis eta/. 1982). Lambda DNA digested with HindlII was used as size marker.
Preparation of feeder layer.
Transformation procedure. The leaf disk transformation procedure of Horsch eta/. (1985) was used with some modifications. A. tumefaciens containing the transformation vector was cultured overnight at 28 ° C in liquid LB medium containing 50 Ixg/ml kanamycin and 50 I.tg/ml streptomycin, pH 7.5. The bacterial suspension was then centrifuged at 1000 g for 15 minutes and the pellet was diluted to an O.D. reading of 0.5 at 560 nm, using MS liquid medium without hormones. The 2nd, 3rd and 4th leaves were excised from plants, surface sterilized in a 15% solution of sodium hypochlorite for 15 minutes and washed three times with sterilized bidistilled water. Leaves were then cut into 0.5 cm 2 segments. Explants were submerged in the diluted A. tumefaciens inoculum and carefully agitated for 2-3 minutes. They were then blotted dry on sterile filter paper and transferred to the feeder culture. Petri dishes were kept at 24-26 ° C under a 16 h photoperiod and the light intensity of 50-70 mE m-2 s-1. After 3 days on the feeder culture, leaf segments were transferred to Petri dishes containing the same medium supplemented with 100 I.tg/ml kanamycin and 500 I.tg/ml carbenicillin. As controls, uninoculated explants were cultured in the same manner on selection medium containing antibiotics and on medium without antibiotics. Regenerated shoots were transferred to the rooting medium (MS medium without hormones) containing 50-100 Ixg/ml kanamycin and 500 ~tg/ml carbenicillin. Rooted plantlets were acclimated for 2-3 weeks after being transferred to sterile soil. They were kept in a growth chamber under a 16-h photoperiod at (150-200 mEre-28-1) at 24:20 ° C (fight:dark). Chloramphenicol acetyltransferase (CAT) assay. Selected leaf tissue (200 rag) was frozen in liquid nitrogen and macerated to a fine powder. Tissue was suspended in 200 I-tl of 250 mM Tris-HC1 buffer pH 7.8 and spun in a microfuge for 5 minutes where the debris was pelleted. Supernatant was heated for 10 minutes at 65 ° C. Protein concentrations were determined and aliquots containing 200 I.tg protein were used for
Results
Genetic transformation and regeneration of transgenic plants. In a series of preliminary experiments, shoot production from leaf explants of the two genotypes was verified using various culture media. These media differed from each other with regards to the composition and concentration of supplemented growth regulators. Based on the results obtained, the medium MSZ (containing 2 mg/l zeatin) which induced a high percentage of direct organogenesis was used for the two genotypes. Moreover, media MS4 (containing 2.25 mg/l BAP and 1.75 mg/l IAA) and MSB (containing 3 rag/1 IAA and 2 mg/1 BAP) which promoted callus and shooting were also used for LA1930 and LA2747, respectively. On MSZ medium supplemented with 100 ~tg/ml kanarnycin, shoot primordia were apparent by the eighth week of transformation on inoculated explants of both genotypes. This concentration of kanamycin was chosen based on preliminary dose response experiments (data not shown). Explants from LA2747 cultured on this medium demonstrated a higher capacity for shoot formation when compared to LA1930 (Fig. 1A). Large numbers of explants also produced light yellow-green soft calli on this medium. Some of these calli regenerated shoots when dissected from explants and recultured on the same medium. On MS4 medium which was used for LA2747, 98% of the explants formed calli and 60% of them produced between 1-3 shoot primordia. On MSB medium which was used for LA1930, 88% of explants produced calli, but only 8% of them demonstrated the ability for direct organogenesis. Calli produced on both MS4 and MSB media were dissected and transferred to MSZ medium where approximately 60% of those from LA2747 produced shoots; the number of shoots produced from LA1930 was below 1%. No callus or shoot
104 primordia were obtained from the uninoculated explants on the above mentioned media (containing 100 ~tg/ml kanamycin).
was found to be more recalcitrant; a large number of plants degenerated gradually soon after being transferred to soil. Only 30% of the plantlets survived the acclimatization period. Acclimatized plants grew to maturity (Fig. 1C) and produced flowers; however, no fruits were formed because of self incompatibility. No morphological differences were observed between transformed and control (in vitro and soil grown) plants.
Biochemical and molecular characterization of transformed plants: The expression of chloramphenicol acetyltransferase gene was verified in transformed plants. Initially a few plants were subjected to CAT assays shortly after the acclimatization period (approximately 2 months after transfer to soil). Most of these plants showed low CAT activity and some of them did not express the gene (data not shown). However, when the same plants were tested again after 6 months, using identical amounts of protein as in the first experiment, they showed 2-3 fold higher CAT activity (Fig. 2A). No activity was detected in untransformed plants. Y2Y5 R2
Rs
Re
Rs I:tl0 Rll R12 Rls R20
C
3 acI ac-
c-
A
Fig. 1. Leaf disc transformation regeneration in Lycopersicon chilense LA2747. (A) Shoot and callus formation from inoculated leaf explants on MSZ medium supplemented with 100 Ixg/rnl kanamycin, 10 weeks after transformation (B) A transgenic plantlet 8 weeks after transfer to the rooting medium containing 50 Itg/ml kanamycin. (C) A transgenic plant 5 months after transfer to soil.
Regenerated shoots of approximately 1 cm were dissected from leaf explants or calli and transferred to MSZ medium where they were kept for 4-5 weeks to enhance elongation. They were then transferred to the rooting medium. Initially, the rooting medium included 100 ~tg/ml of kanamycin. However, none of the putative transformed shoots produced roots on this medium, even though they survived for a long period of time. When the concentration of kanamycin was reduced to 50 ltg/ml, approximately 56% of LA2747 and 25% of LA1930 shoots produced roots after 6-8 weeks (Fig. 1B). In the controls, shoots produced on the nonselective medium showed frequencies of rooting (on the medium without kanamycin) which were almost the same as those recorded for transformation experiments. As expected, these shoots were not able to produce roots on medium containing 50 ~g/ml kanamycin. Rooted plantlets were transferred to soil where 80% of those of LA2747 survived. At this stage once again, the genotype LA1930
C1
Ys
Y2
R12
Rs
C2
kb 0.91 0.79
B Fig. 2. Biochemical and molecular analysis of transgenic plants. (A) Expression of CAT gene in LA1930 (Y) and LA2747 (R) transgenic plants; c, chloramphenicol; lae, 1-acetate chloramphenicol; 3ac, 3acetate chloramphenicol. Lane C is the extract from untransformed plant. (B) Southern blot of transformed and untransformed L. chilense plants probed with 32p labeled CAT fragment. C1, untransformed plant; Y5 and Y2, LA1930 transformed plants; R12 and R6, LA2747 transformed plants. Lane C2 contains 39 pg of the CAT fragment.
Some of the plants which expressed CAT activity were further characterized by Southern blot analysis, using the CAT fragment as the probe. The analysis confirmed that CAT activity was correlated with the presence of the introduced CAT gene. KpnI digested genomic DNA from transgenic plants contained the diagnostic 0.79 kb CAT fragment that comigrates with the fragment used as the probe (Fig. 2B). An additional fragment in one of the transgenic plants (Y5) represents the rearranged copy of
105 the integrated gene. DNA from untransformed plants did not contain sequences that hybridize to the CAT probe. Discussion
Owing to our interest in the genetic potential of Lycopersicon chilense, we have verified the capacity for in vitro plant regeneration of this species in the past (Greer and Tabaeizadeh 1991). During the course of the present study, we found that even though L. chilense had demonstrated a higher capacity for plant regeneration, compared to those reported for the cultivated tomato, it seemed more recalcitrant during the transformation process (from inoculation of leaf explants to production of mature plants). The density of Agrobacterium used for the inoculation of the explants was found to be critical. Low inoculum concentration (O.D. reading of 0.3-0.4) resulted to low ferequency of transformation events judging from the number of calli and shoot primordia produced from the inoculated leaf discs. On the other hand, in those experiments where the concentration of the inoculum was high (O.D. reading higher than 0.5), we were not able to stop the growth of Agrobacterium after transfer of the explants to the media containing carbenicillin. In fact, all of the explants resulting from these experiments gradually degenerated and died. Shoots produced on the selection medium were not able to produce roots in the presence of 100 I.tg/ml kanamycin. The fact that the uninoculated explants did not show any sign Of calli or shoot primordia on kanamycin containing medium strongly indicate that the shoots produced under selection were transgenic. In one of the reported studies on genetic transformation of cultivated tomato (van Roekel et al. 1993), 33 ~tg/ml of kanamycin was used in the rooting medium of shoots which had been produced on 100 ~tg/ml of kanamycin. The authors, however, did not indicate whether they tested higher concentrations of kanamycin in the rooting medium. Another noticeable point of this study was that in some young transgenic plants the CAT activity was either low or absent. However, the activity increased significantly once the plants matured. It is not known whether the problem of CAT expression in young transgenic plants was at the transcriptional or translational level. The loss of gene expression in transgenic plants has also been reported for petunia (Horsch et al. 1985) and tomato (McCormick et al. 1986). As expected, the transgenic plants did not produce fruits owing to the self incompatibility of L. chilense (McGuire and Rick 1954; Taylor 1986). Cross pollination between these plants and untransformed plants used in this study, could not be effective since they all originated from one single clone. Progenies could probably be produced by pollination of the transgenic plants with a mixture of pollen, derived from different seed grown L. chilense plants. However, the existing polymorphism among
different L. chilense individuals (inside each genotype) could make the interpretation of the results very difficult, if not impossible. This problem of genetic diversity has also been discussed by Bonnema and O'Connell (1992) in a report regarding somatic hybridization of L. chilense and cultivated tomato. The analysis of progenies is an important part in genetic transformation studies. However, it should be recognized that L. chilense is a wild species. Thus, it does not present the same type of interest, compared to crop plants, with regard to genetic improvement, using transformation approaches. We believe that the obstacle in production of progeny from L. chilense transgenic plants will not affect our ongoing research for demonstrating the involvement of the genes that we have isolated, in the tolerance of this species during drought stress. Transgenic plants were propagated and maintained by cuttings. Some of the resulting cuttings (randomly chosen) were successfully micropropagated using kanamycin (100 ~tg/ml) containing medium.
Acknowledgements. This research was supported in part by a grant from the Natural Sciences and Engineering Research Council of Canada. References Bevan M (1984) Nucl. Acids Res. 12:8711-8721. Bonnema AB, O'Connell (1992) Plant Cell Rep. 10:629-632. Chen RD, Tabaeizadeh Z (1992) Biochem. Cell Biol. 70:199-206. Chert RD, Campeau N, Greer AF, Bellemare G, Tabaeizadeh Z (1993) Plant Physiol.103:301. Chen RD, Yu LX, Greer AF, Cheriti H, Tabaeizadeh Z (1994) Mol. Gen. Genet. 245:195-202. Chyi YS, Jorgenson RA, Goldstern D, Tarksley SD, Loaiza-Figueroa F (1986) Mol. Gen. Genet. 204:64-69. Delauney AJ, Tabaeizadeh Z, Verma DPS (1988) Proc. Natl. Acad. Sci. USA 85:4300-4304. Filatfi JJ, Kiser J, Rose R, Luca C (1987) Bit/Technology 5:726-730. Greer AF, Tabaeizadeh Z (1991) Can. J. Bot. 69:2257-2260. Horsch RB, Fry JE, Hoffmann NL, Eichholtz D, Rogers SG, Fraley RT (1985) Science 227:1229-1231. Fromm M, Taylor LP, Walbot V (1985) Proc. Natl. Acad. Sci. USA. 82:5824-5828. Jones RA (1987) In: Nevis DJ, Jones RA (eds) Tomato Biotechnology: proceedings of a symposium at the University of California, Davis, Alan R. Liss, Inc., New York, pp 125-137. Lipp Joao KH, Brown TA (1993) Plant Cell Rep. 12:422-425. McCormick S, Niedermeyer J, Fry J, Barnason A, Horsch R, Fraley R (1986) Plant Cell Rep. 5:81-84. McGnire DC, Rick CM (1954) Hilgardia 23:101-124. Maniatis T, Fritsch EF, Sambrook J (1982) Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Martin FW (1961) Genetics 46:1443-1454. Murashige T, Skoog, F (1962) Physiol. Plant. 15:473-497. Rick CM (1973) In: Srb AM (ed) Genes, Enzymes, and Populations, Plenum Press, New York, pp 255-269. van Roekel JSC, Damm B, Melchers LS, Hoekema A (1993) Plant Cell Rep. 12:644-647. Tylor IB (1986) In: Atherton JG, Rudich J (eds) The tomato crop: A scientific basis for improvement, Chapman and Hall, New York, pp 134.
