Plant Cell Reports (2000) 19 : 235240

Q Springer-Verlag 2000

K.V. Krishnamurthy 7 K. Suhasini 7 A.P. Sagare M. Meixner 7 A. de Kathen 7 T. Pickardt O. Schieder

Agrobacterium mediated transformation of chickpea (Cicer arietinum L.) embryo axes

Received: 30 May 1997 / Revision received: 18 September 1997 / Accepted: 22 March 1999

Abstract Embryo axes of four accessions of chickpea (Cicer arietinum L.) were treated with Agrobacterium tumefaciens strains C58C1/GV2260 carrying the plasmid p35SGUSINT and EHA101 harbouring the plasmid pIBGUS. In both vectors the GUS gene is interrupted by an intron. After inoculation shoot formation was promoted on MS medium containing 0.5 mg/l BAP under a selection pressure of 100 mg/l kanamycin or 10 mg/l phosphinothricin, depending on the construct used for transformation. Expression of the chimeric GUS gene was confirmed by histochemical localization of GUS activity in regenerated shoots. Resistant shoots were grafted onto 5-day-old darkgrown seedlings, and mature plants could be recovered. T-DNA integration was confirmed by Southern analysis by random selection of putative transformants. The analysis of 4 plantlets of the T1 progeny revealed that none of them was GUS-positive, whereas the presence of the nptII gene could be detected by polymerase chain reaction. Key words Chickpea 7 Transformation 7 Agrobacterium

Abbreviations BAP: Benzylaminopurine 7 PPT: Phosphinothricin 7 MS: Murashige and Skoog (1962) 7 GUS: b-Glucuronidase 7 NPTII: Neomycin phosphotransferase

Introduction
Chickpea (Cicer arietinum L.), an important grain legume, suffers from heavy losses due to fungal diseases and insect pests, mainly Ascochyta rabiei and chickpea pod borer (Singh et al. 1994). Although wild species of Cicer have numerous desirable traits, the cross-incompatibility between the wild and cultivated varieties has deterred improvement of the crop by conventional plant breeding techniques (van Rheenen et al. 1993). The introduction of specific genes into chickpea could be achieved by genetic engineering. The prerequisites for a successful gene transfer of desirable traits is the establishment of an efficient transformation protocol. To date, only a few reports on transformation are available in chickpea. Transformed callus was obtained using wild strains of Agrobacterium (Mohapatra and Sharma 1991; Islam et al. 1994), while hairy roots were obtained by treating chickpea plants with root-inducing strains of Agrobacterium (Siefkes-Boer et al. 1995). Transformed plants were obtained by Fontana et al. (1993) and Kar et al. (1996) by treating seed-derived embryo axes deprived of the apical meristem with Agrobacterium tumefaciens strain LBA 4404 harbouring the binary vector pBI121. The inheritance of the transgenes was not demonstrated in these studies. In the present report, we describe the combination of a simple protocol for multiple shoot formation from mature seed-derived embryos with Agrobacterium mediated gene transfer. From four cultivars of chickpea, kanamycin-as well as phosphinotricinresistant plants were recovered. The integration of the T-DNA was demonstrated by Southern analysis. More-

Communicated by H. Lrz K.V. Krishnamurthy 7 K. Suhasini 7 A.P. Sagare Plant Tissue Culture Division, National Chemical Laboratory, Pune 411 008, India M. Meixner 7 T. Pickardt 7 O. Schieder (Y) Institute for Applied Genetics, Free University of Berlin, Albrecht-Thaer-Weg 6, 14195 Berlin, Germany e-mail: schieder6zedat.fu-berlin.de Fax: c49308384345 A. de Kathen Dep. of Molecular Genetics, University of Hannover, Herrenhauserstr. 2, 30419 Hannover, Germany

236

over, the presence of the nptII-gene in the T1 progeny could be detected by the polymerase chain reaction. Although the current protocol is still labour-consuming due to its relatively low transformation efficiency, it may have general application in the transformation of chickpea.

Materials and methods

Plant material Seeds of chickpea (Cicer arietinum L.) cultivars PG1, PG12 and Chafa were obtained from Mahatma Phule Agricultural University, Rahuri, and cv Turkey was obtained from the local market in Berlin, Germany. Seeds were surface-sterilized according to the protocol described by Suhasini et al. (1994) and soaked overnight (16 h) in sterile distilled water. The embryo axes were subsequently separated from the cotyledons and injured by the removal of the shoot and root meristems (Fig. 1A). Agrobacterium strains and plasmids For inoculation of embryo axes we used A. tumefaciens strains carrying binary vectors: (1) C58C1/GV2260 containing p35SGUSINT and (2) EHA101 (Hood et al. 1986) containing pIBGUS. Plasmid p35SGUSINT (see Fig. 1H) is a pBin19 derivative, which carries the GUS gene with a ST-LS1-derived intron (Vancanneyt et al. 1990). Plasmid pIBGUS was constructed by introducing the HindIII fragment of p35SGUSINT (the GUS gene containing the intron; de Kathen and Jacobsen 1995) into pIB16.41 (Strauch et al. 1988) carrying a procaryotic gentamycin resistance and 35S-PAT and Nos-NPTII as plant selectable markers. A 500-ml aliquot of an Agrobacterium culture taken from glycerol stocks was suspended in 5 ml of YEB medium (Van Larebeke et al. 1977) with 50 mg/l kanamycin (GV2260/ p35SGUSINT) or 40 mg/l geneticin (EHA101/pIBGUS) and grown at 200 rpm on a rotary shaker at 28 7C for 16 h. The culture was subsequently spun at 5000 rpm at RT for 10 min in an Eppendorf centrifuge and the pellet resuspended in 2 ml liquid MS medium with 0.5 mg/l BAP c3% sucrose at a final density of 1!10 5 cells/ml. Cocultivation Thirty injured embryo axes at a time were incubated in 2 ml of Agrobacterium culture for 20 min, blotted dry on a sterile filter paper and cocultivated in 55-mm petri dishes on gelrite-solidified (0.3%) MS medium (Murashige and Skoog 1962) supplemented with 0.5 mg/l BAP and 3% sucrose. Cocultivation was performed for 72 h at 25 7C under a light intensity of 140 mmol m 2 s 1 provided by cool-white fluorescent tubes and a 16-h photoperiod. After cocultivation the explants were rinsed three times with sterile distilled water containing 500 mg/l cefotaxime (Claforan, Hoechst, Frankfurt/Main, Germany), blotted dry on sterile filter paper and cultured in 250-ml glass jars containing shoot regeneration medium (MS salts, 0.5 mg/l BAP, 3% sucrose, 0.3% gelrite) supplemented with 500 mg/l cefotaxime and either 100 mg/l kanamycin (Sigma Chemical, USA) or 10 mg/l PPT (DL-PPT, obtained from Hoechst, Germany) depending on the agrobacterial construct used for treating the embryo axes. Selection and regeneration of plants The responses of plant tissue to kanamycin and PPT and the optimal selection pressure were determined by culturing injured embryo axes (50 per treatment) on shoot regeneration medium

(see above) containing 25, 50, 75, 100, 200, 300 or 400 mg/l kanamycin or 2, 5, 10 or 20 mg/l PPT under the conditions stated above for cocultivation. The cultures were scored after 4 weeks. Inhibition of shoot proliferation from mature embryo axes was observed on medium containing 100 mg/l kanamycin (Fig. 1B) or 10 mg/l PPT. Approximately 160170 explants per experiment were cocultivated, and the experiments were repeated three times. The cultures were subcultured at 3-week intervals on shoot regeneration media to retain the selection pressure under the same conditions as stated above. The concentration of cefotaxime was halfed at each subculture. Green shoots growing in the presence of 100 mg/l kanamycin or 10 mg/l PPT were scored for GUS activity at different intervals of time. Mature plants were generated by grafting shoots of an appropriate length onto 5-day-old etiolated seedlings of chickpea. Grafting was carried out as described by Pickardt et al. (1995). Tissue staining for gus activity Tissues were incubated for 12 h at 37 7C in the presence of 1 mM X-gluc in 0.1 M NaPO4, pH 7.4, 0.1% Triton X-100, 0.5 mM potassium ferricyanide and potassium ferrocyanide and 10 mM EDTA (Stomp 1992), and subsequently bleached in 95% ethanol before observation. The tissue was further processed by passing it through toluene and embedding in wax. The embedded tissue was then cut into 10-mm sections, mounted, dewaxed and viewed under the microscope. Shoots derived from untreated embryo axes served as controls. Southern hybridisation and polymerase chain reaction (PCR) The presence and integration of the NPTII gene was analysed by Southern blot (T0) and polymerase chain reaction (T1). Genomic DNA was isolated from non-transformed plants and putative transformants according to the protocol of Rogers and Bendich (1985) with modifications. For Southern blotting the DNA was digested with HindIII (Boehringer/Germany), separated on 0.8% agarose gels, transferred to Hybond Nc nylon membrane (Amersham) and hybridised with a 0.85-kb NPTII-HindIII-fragment excised from pHP23 (Paszkowski et al. 1988). The probes were radiolabelled according to Feinberg and Vogelstein (1983). After hybridization, the filter was exposed to X-ray film at 70 7C for 5 days. For the detection of the NPTII coding sequence in the progeny genomic DNA was subjected to PCR using the following primers and conditions: forward 5b-TCATCTCACCTTGCTCCTG-3

P Fig. 1 A Entire (a) and decapitated (b) embryo axis used for cocultivation. Bar: 1.25 mm. B Response of embryo axes to increasing concentrations of kanamycin (mg/l). C, D GUS activity in leaf and shoot bud, respectively, of putative transformed shoots. sm Shoot meristem. E (Unstained) histological section of the shoot bud shown in D. Bar:125 mm. F Transformed shoot grafted on the epicotyl of an in vitro-grown seedling. G Grafted shoots transferred to soil at the beginning of the hardening phase. H Southern analysis of putative primary transformants (T0): genomic DNA digested with HindIII probed with a [32P]-labelled 0.85-kbnptII fragment. The different sizes of the hybridising fragments indicate the integration of the T-DNA into the plant genome: the number of hybridising fragments shows that 1 (lines 6/1c1/4) or multiple (2/1c7/1) copies have been integrated. I PCR analysis of 4 T1 progeny plants deriving from lines 15/1 (lane 2), 14/4 (3) and 13/1 (4 and 5). Position 18602222 of Tn5, which is part of the coding region of nptII (363 bp), was amplified. The nptII-containing vector pHP23 served as a positive control

237

238 and reverse 5b-AGCCAACGCTATGTCCTG-3 primers at a concentration of 50 pM were used to amplify a 363-bp region (position 18602222 of Tn5) of the nptII gene (pHP23 served as a positive control) with 0.5 mg genomic DNA of each sample, 2 mM MgCl2 and 2 units of a (native) Taq polymerase (and the recommended buffer) from Gibco in a 50-ml assay. Cycling conditions were: denaturation, 1 min at 94 7C; annealing, 1 min at 50 7C, elongation 2 min at 72 7C (30 cycles) and a final elongation step of 10 min. For detection of the 363-bp fragment DNA was separated on 1% agarose gels.

Results and discussion

Preliminary experiments showed that kanamycin at 100 mg/l (Fig. 1B) and PPT at 10 mg/l totally suppressed shoot development from embryo axes of chickpea grown on MS medium supplemented with 0.5 mg/l BAP. In the present study embryo axes were therefore selected on 100 mg/l kanamycin instead of 50 mg/l kanamycin as reported earlier by Fontana et al. (1993) and Kar et al. (1996). The use of a lower kanamycin concentration resulted in the proliferation of shoots in large numbers, which proved to be cumbersome for screening of the transformants in our system. Between 10 and 15 days after culture on media containing selective pressure generally two shoots appeared from each cotyledonary node and shoot apex region of the embryo axes. Most of the shoots appeared bleached, and some shoots, which were initially green, bleached out gradually, leaving only a few green shoots after 34 months of culture. These green shoots continued to grow (and to propagate) on media containing 100 mg/l kanamycin or 10 mg/l PPT upon further subculture at 3-week intervals. Table 1 summarises the results derived from three independent experiments, each evaluated after 6 months on selection medium. A total of 16 kanamycin-or PPT-resistant shoots/shoot clones could be recovered from approximately 4000 cocultivated explants, giving an overall efficiency of less than 0.4%. Due to the low number of resistant shoots recovered, the differences between cultivars and strains/plasmids used for inoculation may not be significant. Shoots from these putative transformants were analysed for GUS activity. In the histochemical assays activity could be localized in both leaves (Fig. 1C) and the shoot apical meristem region (Fig. 1D). A longituTable 1 Percentage of transformation in different cultivars of chickpea

dinal section shows a uniform distribution of GUS activity (Fig. 1E). Because of difficulties in rooting, several shoots from each of the 15 clones were grafted onto young wild-type seedlings of chickpea (Table 2, Fig. 1F). After 34 weeks grafted plantlets with welldeveloped leaves were put into (autoclaved) soil and, after a hardening phase of approximately 8 days (Fig. 1G), transferred to the glasshouse. Genomic analysis of 4 randomly selected plants from different transformation experiments, designated 1/4 (clone-no./graft-no.), 2/1, 6/1 and 7/1, confirmed the integration of the T-DNA (Fig. 1H). Each clone/plant possesses a different pattern of junction fragments between the T-DNA and the plant genome, depending on the integration site. The number of hybridisation signals led us to assume that 1 (1/4c6/1) or multiple (2/1c7/1) copies of the T-DNA had been integrated. From 36 plants transferred into the glasshouse, only 5 flowered and set seed. Although reduced fertility can generally be observed in tissue culture-derived plants, the fact that wild-type chickpea also failed to grow up properly as well as the T1 seed derived progeny led us to assume that this phenomenon is neither due to tissue culture and selection nor the transgenic status of the plants. Reduced vigour and, consequently, fertility are probably due to suboptimal glasshouse conditions. Therefore, further experiments are necessary to increase the recovery rate. Line 13/1 produced the highest number of seeds, from which, however, only two were able to germinate. These 2 plants and 2 additional T1 seedlings from line 14/4 and line 15/1 were grown up in order to study the inheritance of the foreign DNA. Histochemical X-Gluc assays of leaf and stem tissue revealed that none of them was GUS positive. To preserve the state of the poorly growing T1 plants (and to retain the possibility of obtaining a T2 progeny, a consideration which finally turned out to be futile, since all 4 T1 plants failed to produce seeds), only a few leaves of each plant were subjected to DNA extraction, yielding between 1 and 1.5 mg of genomic DNA. Since this amount was too small for a Southern blot analysis, we used the PCR technique for the detection of the nptII gene, which is located near the right border of both T-DNAs applied in the present study. All 4 plants indeed showed the presence of the expected 363-bp fragment, which is part of the coding region of the nptII gene (see Fig. 1I). In

Cultivar PG1 Chafa Turkey PG12

Plasmid p35SGUSInt pIBGUS p35SGUSInt pIBGUS p35SGUSInt pIBGUS p35SGUSInt pIBGUS

Number of explants treated 510 510 480 480 480 480 510 510

Number of putative transformed lines P 2 3 3 7 P 1

Transformation (%)

P ~0.4 ~0.7 ~0.7 ~1.5 P ~0.2

239 Table 2 In vitro clones of chickpea selected on 100 mg/l kanamycin (p35SGUSINT) or 10 mg/l PPT (pIBGUS) and recovery of T0 plants/T1 progeny In vitro clone 1 (Chafa) Plasmid p35SGUSInt Graft a 1 2 3 4 1 2 3 4 5 (a) (a) (a) (b) (b) (a) 2/ (c) (a) Seeds/germinate GUS (X-Gluc) PCR (NPTII)

2 (Chafa)

p35SGUSInt

3 (Turkey) 4 (PG12) 5 (Turkey) 6 (Turkey) 7 (PG1)

p35SGUSInt pIBGUS p35SGUSInt p35SGUSInt pIBGUS

1 (a) 1 (a) 1 (a) 1 (b) 1 (b) 2 (a) 3 (a) 1 2 3 4 5 (a) (a) (c) (a) (a)

8 (Turkey)

p35SGUSInt

9 (Chafa) 10 (Chafa) 11 (Chafa) 12 (Turkey) 13 (Turkey)

pIBGUS pIBGUS pIBGUS p35SGUSInt p35SGUSInt

1 (c) 2 (a) 1 (c) 2 (a) 1 (a) 1 (a) 1 2 (a) 3 (a) 4 (a) 1 (a) 2 3 (a) 4 1 14/2 PP cc

14 (Turkey)

p35SGUSInt

4/ 5/1 1/1 P P c c

15 (Chafa)

p35SGUSInt

a (a), Grafted plants successfully transferred to soil but died in the glasshouse; (b), subjected to Southern blot; (c), plants flowered in glasshouse but failed to set seed

line 15/1, an additional fragment of approximately 2.5 kbp was amplified, probably due to a rearrangement of two or more incomplete copies. An incomplete transfer/integration of the GUS gene in these lines could be responsible for the absence of GUS activity. On the other hand, according to our own (unpublished) observations, the gus gene containing the STLS1-derived intron tends to exhibit a lower stability of expression in grain legumes when compared to the original gus gene. In the regeneration protocol used in this study shoot formation occurred from preexisting meristems. Thus, chimeric T0 transformants are to be expected, which do not necesserily transmit the foreign DNA to the progeny. Our findings that the progeny of all three tested lines are positive can be attributed to stringent selection preventing the generation of chimeric shoots.

This study has shown for the first time that a recombinant gene in chickpea is transmitted to the progeny following A. tumefaciens-mediated transformation 1. Further optimization of the protocol is necessary to obtain a higher frequency of transformation. The successful application of a directly induced shoot regeneration without an intermediate callus phase leads us to assume that this simple strategy, which is already established in pea (Schroeder et al. 1995), soybean (Di et al. 1996) and peanut (McKently et al. 1995), may play an important role in chickpea transformation as well.

Note added in proof: This statement is related to the time of submission of this manuscript in 1997

240 Acknowledgements The authors are grateful to Dr. R.B. Deshmukh, Mahatma Phule Agricultural University, Rahuri, India, for the supply of seeds. The Research Fellowships awarded by C.S.I.R., Government of India, New Delhi to K.S. and A.P.S. are acknowledged. This research was also supported by NCL-KFA bilateral cooperation programme, No. INI-253. Paszkowski J, Baur M, Bogucki A, Potrykus I (1988) Gene targeting in plants. EMBO J 7 : 40214026 Pickardt T, Saalbach I, Waddell D, Meixner M, Mntz K, Schieder O (1995) Seed specific expression of the 2S albumin gene from Brazil nut (Bertholletia excelsa) in transgenic Vicia narbonensis. Mol Breed 1 : 295301 Rogers SO, Bendich AJ (1985) Extraction of DNA from milligram amounts of fresh, herbarium and mummified plant tissues. Plant Mol Biol 5 : 6976 Schroeder HE, Gollasch S, Moore A, Tabe LM, Craig S, Hardie DC, Chrispeels MJ, Spencer D, Higgins TJV (1995) Bean alpha-amylase inhibitor confers resistance to the pea weevil (Bruchus pisorum) in transgenic peas (Pisum sativum L.). Plant Physiol 107 : 12331239 Siefkes-Boer HJ, Noonan MJ, Bullock DW, Conner AJ (1995) Hairy root transformation system in large-seeded grain legumes. Isr J Plant Sci 43 : 15 Singh KB, Malhotra RS, Halila MH, Knights EJ, Verma MM (1994) Current status and future strategy in breeding chickpea for resistance to biotic and abiotic stresses. Euphytica 73 : 137149 Stomp AM (1992) GUS protocols: using the GUS gene as a reporter of gene expression. Academic Press, San Francisco, Calif. Strauch E, Wohlleben W, Puehler A (1988) Cloning of a phosphinothricin-n-acetyltransferase gene from Streptomyces viridochromogenes tu494 and its expression in Streptomyces lividans and Escherichia coli. Gene 63 : 6574 Suhasini K, Sagare AP, Krishnamurthy KV (1994) Direct somatic embryogenesis from mature embryo axes in chickpea (Cicer arietinum L.). Plant Sci 102 : 189194 Van Larebeke N, Genetello CH, Hernalsteens JP, De Picker A, Zaenen I, Messens E, Van Montagu M, Schell J (1977) Mol Gen Genet 152 : 119124 Van Rheenen HA, Pundir RPS, Miranda JH (1993) How to accelerate the genetic improvement of a recalcitrant crop species such as chickpea. Curr Sci 654 : 14417 Vancanneyt G, Schmidt R, OConnor-Sanchez A, Willmitzer L, Rocha-Sosa M (1990) Construction of an intron-containing marker gene: splicing of the intron in transgenic plants and its use in monitoring early events in Agrobacterium-mediated transformation. Mol Gen Genet 220 : 245250

References
De Kathen A, Jacobsen HJ (1995) Cell competence for Agrobacterium-mediated DNA transfer in Pisum sativum L. Transgen Res 4 : 184191 Di R, Purcell V, Collins GB, Ghabrial SA (1996) Production of transgenic soybean lines expressing the bean pod mottle virus coat protein precursor gene. Plant Cell Rep 15 : 746750 Feinberg AP, Vogelstein B (1983) A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 132 : 613 Fontana GS, Santini L, Caretto S, Frugis G, Mariotti D (1993) Genetic transformation in the grain legume Cicer arietinum L. (chickpea). Plant Cell Rep 12 : 194198 Hood EE, Helmer GL, Fraley RT, Chilton MD (1986) The hypervirulence of Agrobacterium tumefaciens A281 is encoded in a region of pTiBo542 outside of T-DNA. J Bacteriol 168 : 12911301 Islam R, Malik T, Husnain T, Riazuddin S (1994) Strain and cultivar specificity in the Agrobacterium-chickpea interaction. Plant Cell Rep 13 : 561563 Kar S, Johnson TM, Nayak P, Sen SK (1996) Efficient transgenic plant regeneration through Agrobacterium-mediated transformation of chickpea (Cicer arietinum L.) Plant Cell Rep 16 : 3237 McKently AH, Moore GA, Doostdar H, Niedz RP (1995) Agrobacterium-mediated transformation of peanut (Arachis hypogaea L.) embryo axes and the development of transgenic plants. Plant Cell Rep 14 : 699703 Mohapatra ST, Sharma RP (1991) Agrobacterium-mediated genetic transformation of chickpea, Cicer arietinum L. Indian J Exp Biol 29 : 758761 Murashige T, Skoog F (1962) A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol Plant 15 : 473497