Plant Cell Reports

Plant Cell Reports (1993) 12:74-79

9 Springer-Verlag 1993

Genetic transformation of green bean callus via Agrobacterium mediated D N A transfer Chandra I. Franklin, Tony N. Trieu, Brandt G. Cassidy, Richard A. Dixon, and Richard S. Nelson Plant Biology Division, The Samuel Roberts Noble Foundation, P.O. Box 2180, Ardmore, OK 73402, USA Received August 19, 1992/Revised version received September 22, 1992 - Communicated by J.K. Vasil

Summary. Kanamycin resistant callus was produced from leaf disc or hypocotyl explants of green bean (Phaseolus vulgaris L.) when cultured on a defined medium containing 50 rag/1 kanamycin after 4 days of co-cultivation with Agrobacterium tumefaciens strain EHA101 containing the binary vector pKYLX71GUS. The presence of neomycin phosphotransferase II (NPT-II) in crude cellular extracts from the kanamycin resistant callus was confirmed by ELISA. The expression of the t~ -glucuronidase (GUS) reporter gene was confirmed by histochemical and fluorimetric analyses. Southern blot border analysis confirmed the integration of the foreign DNA. In addition to the evidence obtained from Southern analysis, the absence of Agrobacterium in the transformed callus cultures was confirmed by microscopic observation and through test cultures. Using the above protocol, bean callus cultures were also transformed with a bean chalcone synthase promoter-GUS fusion. These cultures, when treated with the elicitor glutathione, showed higher levels of GUS expression than the unelicited callus clumps. Key words. Green bean - Phaseolus vulgaris L. - Genetic transformation- Stable integration chalcone synthase. INTRODUCTION Green bean (Phaseolus vulgaris L.) has been used as a model system to study the molecular mechanisms underlying induced defense against fungdal pathogens (Dixon et al. 1986). A number of efense response genes have been cloned from this source, encoding enzymes for the biosynthesis of antimicrobial isoflavonoid phytoalexhls, lignin, hydroxyproline-rich glycoproteins and antimicrobialhydrolases(see Dixon and Harrison 1990 for a review). These genes are much more Correspondence to: C.I. Franklin

rapidly activated in incompatible as compared to compatible interactions with the anthracnose fungus Colletotrichurn lindemuthianum (Dixon et al. 1986), and are induced in suspension culture cells in response to a range of elicitor molecules, including the tripeptide glutathione (Wingate et al. 1988). In order to study further the mechanisms underlying fungal activation of defense genes in green bean, it will be important to be able to introduce homologous gene promoterreporter constructs into specific bean cultivars of defined resistant or susceptible genotypes. The lack of an efficient genetic transformation protocol has hindered such studies in green bean. In recent reports, transient gene expression systems for bean ( Leon et al. 1990; Bustos et al. 1991; Genga et al. 1991) and the susceptibility of green bean to Agrobacterium infection and transformation of cotyledonary tissue (McClean et al. 1991) have been reported. The latter report did not provide molecular evidence to confirm stable genomic integration in the transformed cotyledonary tissue. Attempts in our laboratory to produce transgenic bean plants using a published regeneration protocol (Franklin et al. 1991) have not yet been successful. We have, however, obtained transformed bean callus and cell cultures from leaf discs or hypocotyls. Results from molecular and biochemical analyses confirm the stable integration and expression of the foreign gene and the absence of Agrobacterium in the transformed callus cultures. This callus transformation system can be used for some of the studies described above. In this paper, a protocol to obtain transgenic bean callus and cell cultures is described, and the stable integration and expression of a bean defense response gene promoter-reporter fusion in the homologous system is reported.

75

MATERIALS A N D METHODS Binary vector and Agrobacterium. The binary vector pKYLX71GUS was constructed (Fig. 1) by inserting the 1.8 kb GUS reporter gene (Jefferson 1987) between the HindIII and SacI sites of the binary vector pKYLX71 (Schardl et al. 1987; Berger et al. 1989). Following the direct transformation m e t h o d of An et al. (1988), Agrobacterium tumefaciens strains EHA101 (Hood et al. 1986) and EHA105 [EHA101 lacking kanarnycin resistance (E. Hood, personal communication)] were t r a n s f o r m e d w i t h the b i n a r y vectors pKYLX71GUS and CHS15-GUS, respectively. The construct CHS15-GUS contains the bean chalcone synthase (CHS15)promoter-GUS fusion in the binary vector pBI101.1(Stermer et al. 1990). Callus initiation and culture. Leaf discs 8 mm in diameter from 7 day-old aseptically grown bean (Phaseolus vulgaris L. cv DarkRed Kidney) seedlings, or hypocotyl segments (1-2 mm in thickness) from 3-4 day old aseptically germinated seeds were used as explants. The seeds were surface sterilized and germinated as described previously (Franklin et al. 1991). The explants were immersed in a suspension ofA. tumefaciens strain EHA101, containing the binary vector pKYLX71GUS, or EHA105 containing CHS15GUS, for 30 sec. to I min. and then co-cultivated for 4 days on callus culture (CC) medium. The CC medium contained all the components of the modified Schenk and Hildebrandt medium (Dixon 1985) plus 10 pM p-chlorophenoxyacetic acid, 2 ~M 2,4-dichlorophenoxyacetic acid and 0.5 pM kinetin. After co-cultivation, the explants were transferred to fresh CC medium containing 500 mg/1 carbenicillin, 50 mg/1 vancomycin and 50 mg/1 kanamycin. Four weeks after culture initiation, the explants along with the newly initiated callus were transferred to fresh CC medium containing the same levels of antibiotics. Eight weeks after culture initiation, the calli were subcultured at 4 week intervals to fresh CC medium containing the same levels of antibiotics. Cell suspension cultures were initiated from the kanamycin resistant callus. Suspension culture (SC) medium contained all the components of CC medium except agar.

Enzyme assays. The presence of the NPT-II protein in the transformed callus was detected by ELISA using a NPT-II ELISA kit (5 prime -') 3 prime, Inc., Boulder CO, USA). GUS activity in the transformed callus was determined by both histochemical and fluorimetric assays (Jefferson 1987). For the histochemical assay, callus tissues were incubated overnight at 37~ in 1 mM X-

~

SP6 Promoter Hind III Pst I Sail Acc I Hinc II _ .. Xbal . uamH= GUS Barn HI t Ava I Sma I

S sacl I

GUS

psP GOs

Sacl

I

oo

Xba I

\%,;

~ ' ~ ~ "

BarnHI

4~176 Ld~

Hindllll

pKYLX71GUS

Fig. 1 Construction of pKYLX71GUS. A BamHISacI fragment from pBI121 (Clonetech, USA) containing the GUS gene was ligated into pSP64 (Promega, USA) restricted with BamHI and SacI to form pSP64GUS. A HindIII-SacI fragment containing the GUS gene was then isolated from pSP64GUS and ligated into pKYLX71 restricted with HindIII and SacI to form pKYLX71GUS. Solid arrows indicate the direction of transcription of the expressed gene. Asterisks indicate restriction enzyme sites used for subcloning. LB and RB represent left border and right border of T-DNA, respectively. Gluc solution in 50 mM sodium phosphate buffer, pH 7.0. After incubation, the samples were plasmolyzed using 95% ethanol prior to microscopic observation. Using crude cellular extracts, the amount of 4-methyl umbelliferone (4MU) produced by GUS in the transformed callus tissues was quantified fluorimetrically. The amount of total protein present in the crude cellular extracts was determined by the Bradford assay (Bradford 1976).

Elicitor induction of the CHS15 promoter. Cell clumps 1-3 mm in size from clonally derived callus cultures transformed with the CHS15GUS construct were incubated overnight with 30 ~tlof l m M glutathione in SC medium in microtiter plates under aseptic conditions. Cell clumps from the same callus clones incubated under identical conditions in the SC medium without glutathione served as controls. After incubation, the cell clumps were rinsed with distilled water, and the GUS expression measured fluorimetrically.

76

Fig. 2 Proliferation of transformed callus on kanamycin selection medium from a bean leaf disc (A) and histochemical assay for ~-glucuronidase (GUS) in the transformed bean calli. B; callus clumps showing GUS activity. C and D; GUS activity observed in different cell types of the callus. Note the location of blue dye inside the plasma membrane of the plasmolyzed cell in D. Bars represent I mm in A, 200 ~tm in B, 100 ~tm in C and 25 ~tm in D.

Southern blot analysis. Genomic DNA was isolated from transformed and untransformed callus tissues as described by Guillemaut and Mardchal-Drouard (1992). Two to 3 btg of DNA for each callus sample and 100 pg of plasmid DNA were digested with restriction enzymes and electrophoresed through a 0.8% agarose gel. The DNA was transferred onto a nylon membrane (Gene Screen Plus, NEN Research Products, Boston, Massachusetts, USA) and probed with a [32p] labelled GUS fragment from the plasmid pKYLX71GUS following the procedure described by the supplier of the nylon membrane. RESULTS AND DISCUSSION Based on results from dose response experiments (data not shown), kanamycin at a concentration of 50 rag/1 was chosen for selecting trans-

formed callus from leaf or hypocotyl explants of green bean cultivar Red Kidney. This cultivar is ideal for studying the mechanisms underlying fungal activation of defense genes in green bean, because it exhibits clearly defined resistance or s u s c e p t i b i l i t y to different races of C. lindemuthianum. Callus initiation occurred from all leaf disc and hypocotyl explants by 2 weeks after culture initiation. However, further proliferation of callus resistant to kanamycin occurred only from those explants treated with A. tumefaciens containing pKYLX71GUS (Fig. 2A). The callus was produced as clumps from the wounded areas of the explant. Each clump could have been produced from a single transformation event. In order to maintain the potentially clonal origin of the callus from a single transformation event, the callus clumps were separated from each other 4 weeks after culture initiation

77

and cultured independently. Eight weeks after culture initiation, the transformed callus proliferated rapidly on kanamycin selection medium and doubled in volume during a 4 week subculture period. Assays for NPT-II protein levels and GUS activity indicated the expression of the NPTII selectable marker gene and the GUS reporter gene in the transformed bean calli. Relative levels of NPT-II protein in transformed callus measured by ELISA were at least twice as high as the b a c k g r o u n d level p r e s e n t in the untransformed callus (data not shown). Expression of the GUS reporter gene in callus transformed with the pKYLX71GUS construct was confirmed by both histochemical and fluorimetric assays. Results from the GUS histochemical assay are shown in Fig. 2 B-D. Using the GUS fluorimetric assay, transformed bean calli were shown to express high levels of GUS activity (Fig. 3). In a callus transformation system, cells of Agrobacterium that may survive the antibiotic treatment (ie. carbenicillin and vancomycin) but which are unable to produce visible colonies, may interfere with molecular and biochemical analyses and lead to the erroneous conclusion that stable integration and expression of the foreign gene has taken place. In order to discount such a phenomenon, two different tests to verify the absence of Agrobacterium in the transformed callus cultures were conducted. During tissue preparation for the GUS histochemical assay, samples of transformed callus were plasmolyzed in order to test that the blue dye indicating GUS activity was inside the cells and not produced by Agrobacterium in the intercellular spaces of the callus cultures. Such contamination should be revealed as bacterial cells staining blue around the outer surface of the plant cells in a uniform pattern. As shown in Fig. 2D, the blue dye was present only within the plasma membrane of the lasmolyzed cell, indicating that free livingAgrocterium is not responsible for the GUS activity. As a further test to confirm the absence of Agrobacterium in the transformed callus cultures, an inoculation loop was stabbed several times in and around putatively transformed callus (after the 4th or 5th subculture) and then streaked on LB plates + appropriate levels of antibiotics. LB plates inoculated with A. tumefaciens strains EHA101 and EHA105 carrying the binary vectors pKYLX71GUS and CHS15-GUS respectively served as controls. Bacterial growth was not observed in the test plates even after several days of incubation, whereas normal bacterial growth was observed in controls (data not shown). This result further confirms the absence of Agrobacterium in the transformed bean callus.

90

70 tIm

G) O I.-

~.

N

50 i

O)

E

"e ~

E

30

=E

i

4 m

O

E C

10 m

m m

Untrans- 1 2 3 4 5 formed Callus Transformed Calli

Fig. 3 Relative levels of l~-glucuronidase activity in bean calli transformed with pKYLX71GUSmeasured by fluorimetric assay. Each callus analyzed was from an independent transformation event and was clonally derived. Mean + S.D. from two replicates per sample is given. Callus clones 1 (Fig. 4B lanes 5 and 6), 3 (Fig. 4B lanes 3 and 4), 4 (Fig. 4A lane 1) and 5 (Fig.4A lane 2) were analyzed by Southern blot. Southern blot border analyses were conducted to verify the integration of the GUS gene into the bean genome (Fig. 4). Border fragments from all transformed cell lines hybridized with a [32p] labeled GUS fragment (Fig. 4A, lanes I and 2; 4B, lanes 3, 5 and 7 and 4C, lane 3), and all hybridizing fragments were greater in length than the maximum distance from the particular unique restriction site within the vectors and the respective T-DNA borders (pKYLX71GUS: 4.7 kb from the HindIII site to the right border, CHS15-GUS: 4.5 kb from the EcoRI site to the right border). Genomic D N A from untransformed bean calli did not hybridize with the labeled GUS fragment (Fig. 4B and 4C, lane 2). These results confirm the stable integration of the foreign DNA into the plant genome. The

78

Fig. 4 Southern blotborder analyses of bean callus and suspension culture celllines. A and B represent results from the analyses of bean cell lines transformed with pKYLX71GUS. Approximately 2-3 ~g of genomic DNA from the nontransformed bean callus (B, lane 2), four clones of kanamycin resistant bean callus (A, lane 1; B, lanes 3, 5 and 7) and one clone of kanamycin resistant bean cell suspension culture (A, lane 2), as well as 100 pg of pKYLX71GUSplasmid DNA (B, lane 1)were digestedwith restriction enzymesand electrophoresed through 0.8% agarose. The genomic DNA was digested with HindIII, and the plasmid DNA was digested with both HindIII and SacI. DNA was blotted onto nylon membranes and probed with a [32p] labeled GUS fragment from pKYLX71GUS. Undigested DNA from the kanamycin resistant calli is represented by a single band in B, lanes 4, 6 and 8. Undigested DNA loaded in B, lanes 4, 6 and 8 were isolated from the callus clones represented in B, lanes 3, 5 and 7, respectively. Results from the analysis of a bean cell line transformed with CHS15-GUSare represented in C. Two ~g of genomic DNA from the callus transformed with CHS15-GUS (C, lane 3) and from the nontransformed callus (C, lane 2) were digested with BamHI. CHS15-GUSplasmid DNA was digested with both BamHI and EcoRI(C, lane 1). Undigested DNA from this transformed cell line and from CHS15-GUSwas run in C, lanes 4 and 5 respectively. binary vectors pKYLX71GUS, when digested with HindIII and SacI and CHS15-GUS with BamHI and EcoRI, yielded a 1.8 kb GUS fragment and linearized plasmid fragments (14 kb for pKYLX71GUS and 12.4 kb for CHS15-GUS) verifying the uniqueness of these restriction sites (Fig. 4B and 4C, lane 1). The linearized fragments were produced due to partial digestion. Undigested CHS15-GUS plasmid DNA produced multiple bands (Fig. 4C, lane 5). Undigested DNA (Fig. 4B, lanes 4, 6 and 8 and 4C, lane 4) from the transformed cell lines hybridizing to the GUS probe is represented by only a single band at 23 kb or greater indicating the insert is within the undigested plant genomic DNA. The absence of multiple bands and the position of the single band representing undigested plant DNA verifies the absence of Agrobacterium containing the vectors in these cultures. The expression of the GUS reporter gene in the callus transformed with the CHS15-GUS construct was assayed fluorimetrically. The

CHS15 promoter-driven GUS expression in the callus elicited with I mM glutathione was higher than the background levels of GUS expression in the unelicitedcallus, demonstrating the elicitor inducibility of the CHS15 promoter in the stably transformed bean ceils (Fig. 5). The increase in GUS expression in response to elicitation varied among the different callus clones tested; these differences may be attributed to position effects on the insert(s) a n d / o r the number of copies of the foreign gene in the plant genome. It is possible that the somewhat elevated levels of GUS expression in unelicited transformed cultures compared with untransformed cultures may be due to the induction of the CHSI5 promoter due to the stress caused by tissue culture conditions. It is known from previous studies (Ryder et al. 1987) that bean chalcone synthase is induced in response to both biotic (eg. fungal infection) and abiotic (eg. wounding) stress. The callus transformation system described here can be used for quantifying the

79 REFERENCES

10

8 i

7

6

/

t

/

4

9* o.1 9- 0.138 0.06 0.04 0.02 0

~ Untransformed 1 Callus

2

3

4

5

Transformed Calll

Fig. 5 Relative levels of t~-glucuronidaseactivity in unelicited ( ~ ) and elicited ( ~ ) bean calli transformed with CHS15-GUS measured by fluorimetric assay. Each transformed callus analyzed was clonally derived. Callus clone #4 was analyzed by Southern blot (see Fig. 4C, lanes 3 and 4). i

expression of a foreign gene stably integrated into bean cells. Similar callus transformation systems have been reported in a few other species such as grapevine (Baribault et al. 1989), Chenopodium quinoa (Komari 1990) and parsley (Douglas et al. 1991). In parsley, cell cultures derived by pooling five or more clones of transformed calli were used for studying elicitor and light activation of the plant defense response gene 4-coumarate:CoA ligase (4CL-l), using 4C L1-promoter-GUS fusions. Thus far, studies involving defense response genes isolated from bean have been conducted in heterologous systems such as tobacco (Stermer et al. 1990) or alfalfa (Harrison et al. 1991). Further studies with the bean callus transformation system will focus on the expression of defense response genes in the homologous system. ACKNOWLEDGEMENTS

We thank Drs. Abraham Oommen and Marilyn Roossinck for helpful suggestions and critical reading of the manuscript, Cuc Ly and Jackie Brightwell for graphics and Allyson Wilkins for preparation of the manuscript.

An G, Ebert PR, Mitra A, Ha SB (1988) (eds.) Binary vectors. In Plant Molecular Biology Manual, Kluwer Academic Publishers, Dordrecht, The Netherlands, pp 1-19 Baribault TJ, Skene KGM, Scott NS (1989) Plant Cell Reports 8:137-140 Berger PH, Hunt AG, Domier LL, Hellmann GM, Stram Y, Thornbury DW, Pirone TP (1989) Proc Nail Acad Sci USA 86:84028406 Bradford MM (1976) Anal Biochem 72:248-254 Bustos MM, Battraw MJ, Kalkan FA, Hall TC (1991) Plant Mol Biol Rep 9:322-332 Dixon RA (1985) Plant Cell Culture: A Practical Approach. Oxford, WashingtonDC, IRL Press Dixon RA, Bailey JA, BellJN, Bolwell GP, Cramer CL, Edwards K, Hamdan MAMS, Lamb CJ, Robbins MP, Ryder TB, Schuch W (1986) Phil Trans Royal Soc Lond B314: 411-426 Dixon RA, Harrison M (1990) Adv Genet 28: 165-234 Douglas CJ, Hauffe KD, Ites-Morales M-E, Ellard M, Paszkowski U, Hahlbrock K, Dangl JL (1991) EMBO J 10:1767-1775 Franklin CI, Trieu TN, Gonzales RA, Dixon RA (1991) Plant Cell Tissue and Organ Culture 24:199-206 Genga A, Ceriotti A, Bollini R, Bemacchia G, Allavena A (1991) J Genet Breed 45:129134 Guillemaut P, Mar6chal-Drouard L (1992) Plant Mol Biol Rep 10:60-65 Harrison MJ, Choudhary AD, Dubery I, Lamb CJ, Dixon RA (1991) Plant Mo 1Bio116:877890 Hood EE, Helrner GL, Fraley RT, Chilton M-D (1986) J Bacteriol 168:1291-1301 Jefferson RA (1987) Plant Mol Biol Rep 5:387-405 Komari T (1990) Plant Cell Rep 9:3030-306 Leon P, Planckaert F, Walbot V (1990) Plant Physiol 95:968-972 McClean P, Chee P, Held B, Simental J, Drong RF, Slightom J (1991) Plant Cell Tissue Organ Culture 24:131-138 Ryder TB, Hedrick SA, Bell JN, Liang X, Clouse SD, Lamb CJ (1987) Mol Gen Genet 210:219-233 Schardl CL, Byrd AD, Benzion G, Altschuler MA, Hildebr DF, Hunt AG (1987) in Gene 61:1-11 Stermer BA, Schmid J, Lamb CJ, Dixon RA (1990) Mol Plant-Microbe Interact 3:381-388 Wingate VPM, Lawton MA, Lamb CJ (1988) Plant Physiol 87:206-210