PlantCeU Reports

Plant Cell Reports (1992) 11:11-15

9 Springer-Verlag1992

Transformation of Solanum integrifolium Poir via Agrobacterium tumefaciens: Plant regeneration and progeny analysis G. L. Rotino 1, D. Perrone

2,

p. Ajmone-Marsan 3, and E. Lupotto 3

1 Istituto Sperimentale per l'Orticoltura, P.O. Box 48, 84098 Pontecagnano (SA) z Istituto Sperimentale per l'Orticoltura, Via Paullese 28, 20075 Montanaso L. (MI) 3 Istituto Sperimentale per la Cerealicoltura, Via Stezzano 24, 24100 Bergamo, Italy Received June 11, 1991/Revised version received October 15, 1991 Communicated by H. Ltrz

Summary. The wild species Solanum integrifolium represents a source of pest and disease resistance genes for breeding strategies of the cultivated species Solanum melongena. Somatic hybridization via protoplast fusion between the two species may provide a valuable tool for transferring polygenic traits into the cultivated species. The availability of S.integrifolium cells carrying dominant selectable markers would facilitate the heterokaryon rescue. An appropriate methodology for in vitro culture and plant regeneration from leaf explants of S.inte~rifolium is reported. Efficient leaf-disk transformation via co-cultivation with Agrobacterium tumefaciens led to the regeneration of transformed plants carrying the reporter genes GUS and NPT-II. Transformed individuals were obtained through selection on kanamycin-containing medium. Stable genetic transformation was assessed by histochemical and enzymatic assays for GUS and NPT-II activity, by the ability of leafdisks to initiate callus on Kin-containing medium, Southern blot analyses of the regenerated plants, and genetic analysis of their progenies. Selfed-seed progeny of individual transformed plants segregated seedlings capable to root and grow in selective condition, while untransformed progeny did not. Genetic analyses of progeny behaviour showed that the reporter gene NPT-II segregated as single as well as two independent Mendelian factors. In two cases an excess of kanamycin-sensitive seedlings was obtained, not fitting into any genetic hypothesis. Key w o r d s : Agrobacterium integrifolium, transformation.

tumefaciens,

Solanum

Abbreviations: MS, Murashige and Skoog (1962) medium; NOS, nopaline synthase; NPT-II neomycin phosphotransferase; GUS, beta-glucuronidase; LB, Luria and Bertani medium; KIN, 6-furfurylaminopurine; B A P , 6-benzylaminopurine; 2iP, N6-(2-isopentyl)adenine; ZEA, zeatin; TDZ, Thidiazuron. Introduction

The development of systems for transferring valuable agronomic traits from wild Solanum species into cultivated Solanum melongena, represents a major requirement in eggplant breeding programs. The wild species Solanum integrifolium is a source of genes for resistance to Fusarium wilt (Yamakawa and Mochizuki 1979) and mite Tetranychus urticae (Dikii and Voronina 1985). It is also currently used in Japan as rootstock for eggplant cultivation (Okimura e t a ] . 1986). Somatic hybridization between S.integrifolium and S.melongena may represent an alternative approach to sexual hybridization for transferring valuable polygenic Offprint requests to. G. L. Rotino

traits into the cultivated species (Akamatsu et al. 1989). The availability of S.integrifolium cells carrying dominant selectable markers would result extremely useful, by improving the efficiency of heterokaryon selection. In S.melongena both genes NPT-II (neomycin phosphotransferase) and CAT (chloramphenicol acetyl transferase), have been recently inserted (Rotino and Gleddie 1990). Species belonging to the Solanum genus are susceptible to infection by Azrobacterium tumefaciens. Availability of tissue culture techniques for plant regeneration from i n vitro cultured S.integrifolium tissues, represents an essential requirement for the applicability of the leafdisk transformation system. The present paper reports an appropriate methodology for in vitro culture and plant regeneration of S.integrifolium, and the development of an efficient procedure for genetic transformation via co-cultivation of leaf-disks with A.tumefaciens. Analyses of tranformants were performed on R0 regenerated plants and their R1 progenies through biochemical assays for the reporter gene activities (GUS and NPT-II), leaf-disk culture on selective medium, and Southern blot analyses. Materials and Methods

Plant material and in vitro plant regeneration. Seeds of S.integrifolium were kindly supplied by Kamimura S. (lwate Prefectural Horticultural Experiment Station, Japan). Seeds were sterilized in commercial NaOCI (1% active chloride) for 30 min, extensively rinsed in sterile demineralized water and aseptically germinated onto filter paper in I00 x 15 mm petri dishes, and incubated at 26~ in complete darkness. The germinated seeds were singly transferred to GA7 boxes (Magenta Corp.) containing 40 ml of 0.25% gelrite (Kelco) solidified V3 medium (Chambonnet 1985). V3 medium contains MS (Murashige and S](oog 1962) macrosalts, Heller's (1953) microsalts and More] and Wetmore (1951) vitamins, 2% (w/v) sucrose. The pH was adjusted to 5.8 before autoclaving (121~ for 20' rain). For in vitro plant multiplication, the apical bud with 2-3 internodes were monthly subcultured on the same medium. All cultures were incubated at 25 + 2~ under 50 pEm-2sq 16/8 h day/night cycle. For plant regeneration experiments, leaf-disks (I cm 2) were prepared from young fully expanded leaves and placed with the lower side in contact with medium in 100x15 mm petri dish sealed with household plastic wrap. Cytokinins: 6-benzylaminopurine (BAP), 6-furfurylaminopurine (KIN), zeatin (ZEA), N6-(2-isopentyl)adenine (2iP) and Thidiazuron (TDZ) (Schering, FRG) at 0.5, 2, 5 and 10 mg/l singly added

12 to MS basal medium, were tested for their ability to regenerate shoots. A total of 27 leaf-disks (9/plate) were employed for each treatment; after 4 weeks the number of shoots and shootbuds was scored with the aid of a dissecting microscope. Shoots 1.5-2 cm high were excised from the callus and rooted into V3 medium. Plantlets were transferred to sterile peat under high humidity condition for i0 days and maintained in the greenhouse. Bacterial strains. Agrobacterium tumefaciens binary vector p35S GUS INT (Vancanneyt et al. 1990) carrying a chimeric NPT-II gene and a GUS-gene with a ST-LSI derived intron in pBIN19, was used. The helper plasmid was pGV2260 in a C58 chromosomal background (Deblaere et al. 1985). Both the strain GV3101, a C58 type chromosome, with the plasmid pGV3850 providing the vir-function (Zambryski et al. 1983), and the strain LBA 4404 (Hoekema et al. 1983), a derivative of Ach 5 harbouring the helper plasmid pAL 4404, were also employed in a binary system with a pBI121derived plasmid (Jefferson et al. 1987) containing both NPT-II and GUS-gene under the control of 35S CaMV promoter. The A.tumefaciens strains were grown overnight at 28~ in the darkness at 200 rpm in LB liquid medium supplemented with the appropriate antibiotics. For leaf explant infection the bacteria were diluted 1:20 with MS liquid basal medium, 2% sucrose, pH 5.8 and supplemented with acetosyringone 20 mM final concentration. Transformation and regeneration procedure. Whole fully expanded leaves of in vitro grown plants were preincubated in 100x15 mm petri dish with the lower side in contact with the medium. Both MS hormone-free medium, 2% (w/v) sucrose, and the regeneration medium (MS basal medium supplemented with 0.5 mg/l TDZ plus 0.1 mg/l IAA), were tested as preincubation and co-cultivation medium. After 72 h the leaf-disks were prepared from the preconditioned leaves and dipped for 5 rain in 20 ml of diluted bacterial suspension, blotted dry onto sterile filter paper and placed back onto the same plates. After 48 h of co-cultivation the leaf pieces were transferred to the selection medium (regeneration medium as above) containing 500 mg/1 cefotaxime (Hoechst) and I00 rag/1 Kanamycin sulphate. Control leaf-disks were subjected to the same procedure with the exception of Agrobacterium infection. Every three weeks the explants were transferred to fresh selective medium. Calli 5 mm in diameter were excised from the leaf-disks and subcultured in selective medium (i00 mg/l kanamycin) containing a lower concentration of cefotaxime (300 mg/l). Calli with shoots and shoot-buds were transferred to GA7 boxes containing 40 ml MS basal medium supplemented with 1 mg/l ZEA and 100 mg/l kanamycin for shoot elongation. Shoots from each single callus were considered to come from independent transformation events. Shoots 1,5-2 cm high were excised and transferred to V3 medium containing 50 mg/l kanamycin. Transgenic plantlets were grown in the greenhouse, the flowers were handpollinated and covered with paper bags. GUS and NPT-II expression assays. The GUS activity was determined by both histochemical and fluorimetric assay as reported by Jefferson (1987). For the fluorimetric assay one young leaf for each putative transgenic plant grown in the greenhouse, was homogenized, the protein concentration standardized to 40-50 ug for each sample and incubated with the substrate in triplicate for each assay. After 2 hour reaction the samples were analyzed under UV light. In vivo staining of intact A~robacteriuminfected explants was performed in the experiments in which the vector p35S GUS-INT was used as the GUS gene is expressed upon transfer to the plant cell and not expressed in Agrobacterium (Vancanneyt et al. 1990). The NPT-II assay was performed as dot-blot assay according to

McDonnell et al. (1987). A mean amount of 20 ug total proteins were reacted each time and spotted onto prewashed P81 WhatmanTM paper, and each sample tested in triplicate. Unspecific background was removed by a treatment with I mg/ml proteinase-K (Boheringer), in 1% SDS, at 65~ for 45 minutes, followed by a 5 rain wash with distilled water at 80~ Recallusing assay. Leaf-disks taken from fully expanded leaves of each putative transformed regenerated plant were cultured on regeneration medium (MS basal medium, 0.5 mg/l TDZ, 0.i mg/l IAA) containing I00 mg/l kanamycin. DNA isolation and Southern hybridization. Total genomic DNA was isolated from fully expanded leaves of transgenic and untransformed control greenhouse-grown plants according to the CTAB method (Saghai-Maroof et al. 1984). Eight ~ug undigested and HindIII digested genomic DNA samples were subjected to electrophoresis in 0.8% agarose gels and blotted onto Hybond N nylon membranes (Amersham) by the method of Southern (1975). Conditions for Southern capillary blotting and hybridization were according to the manufacturer (Amersham). Probes were prepared from purified HindIII (NPT-II) and EcoRI-HindlII (GUS) fragments from a plasmid containing a chimeric NPT-II gene (CaMV 35S promoter/NPT-II coding sequence/nos 3'), and a chimeric GUS-gene (CaMV 35S promoter/GUS coding sequence/nos 3'), kindly provided by Ceriotti A. and Bernacchia G. (unpublished). The probes were radiolabelled with p32 by the random priming technique (Amersham). Following overnight hybridization at 65~ the membranes were washed with 2xSSC, 0.1% SDS (twice), and 0.5xSSC, 0.1% SDS (once) at 65~ for 30 minutes. Dryed membranes were exposed to Kodak X-OMat film at -80~ I to 3 days using an intensifying screen for autoradiography. Genetic analysis of transgenic progeny. Seeds from eleven transgenic selfed transformed plants and from one control plant were sterilized and incubated as described above. Hypocotyls with fully expanded cotyledons were excised from the primary roots and inserted into GA7 boxes (9/box) containing V3 medium with 50 mg/l kanamycin. After 6 weeks, resistance or sensitivity was scored for the presence/absence of normal roots and shoot growth. Chi square test was used for testing hypotheses concerning a different number of T-DNA insertions, and a different possible transmission of Km resistance to the progeny. R e s u l t s and D i s c u s s i o n

Plant regeneration. In the preliminary experiments all the cytokinins tested induced shoots from leaf-disks of S.integrifolium. 6-BAP, KIN, and 2iP gave the lowest regeneration efficiency, with 5-10 shoots per leaf-disk; higher values were reached in the case of 5 and 10 mg/1 6-BAP. Also zeatin supported efficient shoot differentiation at the lower concentration (0.5 mg/1) but promoted fast shoot elongation. The highest frequency of shoot regeneration was obtained with TDZ regardless of the concentration used. A mean value of 33 shoot-buds per leaf-disk was obtained with 2 m g / 1 T D Z . Shoots and shoot-buds differentiated from the callus developed on the cut edges of the explants, while hard green callus developed close to the leaf-veins. Regenerated plants appeared phenotypically normal and set viable seeds. On the basis of these results, medium supplemented with 0.5 mg/l TDZ plus 0.I mg/l IAA was chosen for the transformation experiments. The auxin IAA was added to the regeneration media to facilitate cell division of the wounded explants. Zeatin (I mg/1) containing medium was employed for shoot elongation. Leaf-disk

transformation. The preculture of the entire

]3 leaf combined the positive effect of explant preconditioning to the fresh wounding for A.tumefaciens infection. Explants precultured and co-cultivated in hormone-free medium showed a higher transformation efficiency. B o t h t h e frequency of transformed leaf-disk (93%) and the frequency of independent transformation events per single leaf-disks, were increased by using the hormone-free medium (MShf) with respect to the hormone containing medium (Table 1). Other authors discussed the importance of the preculture medium and of the duration of preincubation on transformation efficiency (examples given in Draper et al. 1988). Tab. 1

frequency of leaf-disks using r A. tomefaciens strains with vsrioes incubation sad co-cultivation media.

- ~ortaaUon

Pre-culture No. leafAgrolmeterium and co-cultivation disks strain medium inoculated pGUs-nqT

0.5TDZ-tO.11AA M~tf

No. d/sire Total No. of No. of forming Km resistant mature calli a calli a plants

90

22 (24)

29 (32)

90

84 (93)

283 (314)

23 71

pGV3850 pBI121

Ml~,bf

79

31 (44)

59 (82)

34

LBA4404 pBII21

M~tf

70

21 (30)

29 (41)

20

a = in parenthesis the percentage over the total number of leaf-disks is indicated

After three weeks of incubation on selective medium containing 500 mg/l cefotaxime and I00 mg/l kanamycin, green callus grew along the cut surface of the infected leaves whereas the negative control leaves turned brown and died (Fig. la). After five weeks of selection, calli showed meristematic centres and shoot-buds and were transferred to elongation medium in selective conditions. All the selected calli regenerated shoots. Shoots were excised from 148 independent Km resistant calli and were transferred to the rooting medium containing 50 mg/l kanamycin. All the putative transformed shoots rooted on this medium while negative controls bleached and died (Fig. Ib). Leaf-disks taken from fully expanded young leaves were used as explants in the recallusing assay from plants rooted in the presence of kanamycin. They indeed produced callus and shoot-buds in regeneration medium containing 100 mg/l kanamycin while leaf-disks taken from non-transgenic plants failed to produce callus. One rooted plant per each single selected callus was maintained in the greenhouse for further analyses.Transgenic plants appeared phenotypically normal and set seeds after self-pollination. This procedure worked equally well for the three A.tumefaciens strains employed (Table 1).

Tab. 2 -

G U S a c t i v i t y in e a l l i a n d p l a n t s t r a n s f o r m e d w i t h d i f f e r e n t A.tumefaeiens str~

GUS activity (No. of iedividua]s) Abn~metorium strain

ealli

p I s n t s

histoehemical t p

histoehemieal t p

fluorimetrieal t p

GUS-4NT

18

16

34

33

94

93

pGV3850 p B l l 2 1

15

15

20

20

34

34

L B 4 4 0 4 pBI121

10

10

13

13

20

20

t = t o t a l number of calti or plants tested; p = number of poedtive individuals

Fig. 1 - Leaf-disk transformation of S.integrifolium. a) Leaf tissues cultured on 100 mg/] Km containing medium after a 3-week culture period. Petri dish 1 and 2 show no callus development from non-inoculated tissues, and petri dish 3 and 4 show extensive green callus proliferation at the edges of the tissues, c) GUS assay performed on leaf tissue 2 weeks after infection with A.tumefaciens p35S GUSiNT, showing the GUS-expressing ealli developed.

Detection of NPT-II and GUS activities. Early histoehemieal GUS assay performed on a sample of leaves after one week from the infection w i t h A.tumefaciens p35S GUS-INT, provided the first evidence for transformation of ca]li growing on selective medium (Fig. Ic) A large number of Km resistant calli and all the greenhouse-grown plants were also tested for GUS activity. Untransformed calli or leaves never displayed GUS activity in both histochemical and fluorimetrical assays. Two out of 48 calli tested histochemically did not show GUS activity (Table 2). One plant did not express GUS activity both with histochemical+ and fluorimetrical assays. Because all the plants rooted in the presence of 50 mg/l Km, it was likely to suppose for the GUS-negative plant that: i) the GUS gene was lost; ii) the GUS gene was turned off because of internal deletions or rearrangements or methylation; iii) a very weak expression could not be detected. Loss of reporter genes activity has also been described in petunia (Horsch et al. 1985) and tomato (McCormick et al. 1986) and, m o r e recently, in trans interactions with negative effects on gene expression in transgenic tobacco plants have been described (Matzke and Matzke 1991). The NPT-II dot-blot assay was used for testing calli and plants growing on kin-containing medium. Callus and leaf-derived extracts showed various degrees of activity. In all calli growing on km-containing medium the NTPII assay was positive (results not shown). In some cases, the activity of leaf-derived extracts was not detected by the NPT-II assay, although plants grew on km-containing medium. In order to verify whether an endogenous inhibitor in the S.integrifolium extract could be present and inactivate

14 the enzyme in the in vitro assay reaction, samples of plants M1 and M3 (Fig. 2, spot 1 and 2) were mixed with the extract of a strong-signalling transformed tobacco plant (Fig. 2, spot 3). Indeed, mixtures in various ratios (1:1, 1:2, and 2:1) of the extracts of plants MI and M3 with transformed tobacco, resulted negative in the NPT-II assay (Fig. 2, spots 4 to 9), thus suggesting that a strong inhibition of the tobacco signal had occurred. Plants M1 and M3 were however considered positive as they rooted promptly on Km containing medium, and the NPT-II gene insertion was confirmed by Southern blot analyses. The other transformed S.integrifolium plants all showed a well detectable signal at the NPT-II assay (e.g. plants M6 and M7, spots 10 and II in Fig. 2), whilst untransformed plants did not show any unspecific background (Fig. 2, spot 12).

Fig. 2 - NPT-II dot-blot assay of four regenerated KmR, Solanum integrifolium plants. A mean value of 20 ug total proteins were reacted each time. The autoradiogram was exposed 16 hours at -80~ with intensifying screens. 1, 2 = absence of signal in two Solanum plants (M] and M3) grown on Km containing medium; 3 = signal of a N.tabacum transformed control plant; 4, 5, 6 = mix of Solanum M1 and transformed N-tabacum extracts. 1~1, 1:2. 2:1 (v:v)respectively;7, 8, 9 = mix of Solanum M3 and transformed N.tabacum extracts 1:1, 1:2, 2:1 (v:v) respeeziveiy; I0, 11 = S.integrifolium transformed M6 and M7 plants, respectively; 12 = S.integrifoliumuntransformedplant.

Southern hybridization analysis of transformants. Southern blot analyses confirmed that the NPT-II and GUS genes were stably integrated in the nuclear genome (Fig. 3a, b). Hybridization of undigested DNA of transformed plants occurs as high molecular weight fragment detected bY the NPT-II probe (Fig. 3a, lanes 3 to 10) while a 6.7 kb band is evidenced on the EcoRI digested DNA of plant M5 (Fig. 3a, lane 1), and no signal is detected for a control untransformed plant (Fig. 3a, lane 2). In Fig. 3b genomic DNAs of transgenic plants M1 to M9 were digested with HindlII restriction enzyme and probed with the GUS gene (HindIII-EcoRI fragment of pBI 221-derived plasmid). A single hybridization fragment was present in plants M2, M4, MS, and M9 (Fig. 3b, lanes 3, 5, 9, 10, respectively), whilst more than one copy of the GUS gene were integrated in plants MI, M3, and M6 (lanes 2, 4, 7, respectively). The strong signal present in plants M5, M6 and M7 was probably caused by tandem multiple insertion of the gene, as confirmed by the size of the detected fragment (about 3 kb, Fig. 3b, lanes 6, 7, 8, respectively). Complex insertion patterns such as reported here in S.integrifolium transformed plants have been reported previously in other systems of transformation with A.tumefaciens (Schmidt and Willmitzer 1988). Genetic analysis of transgenic progenies. Sailed-seed progeny of individual transformed plants were able to root and to grow normally in selective conditions (50 mg/l kanamycin) while untransformed progenies did not. The segregation ratio for Km resistance of seedlings derived from self-

Fig. 3 - Southern blot analyses of genomic DNA from S.integrifolium transformed plants, a) Southern blot of genomic DNA (8 ug) from transformed plants probed with the NPT-II gene (HindIII fragment of pBl 221-derived plasmid). Lane 1: HindIIIdigestedtransformedplant M2showing a unique hybridization band at 6.7 kb; lane 2: undigested untransformed control plant; lane 3 to 10: undigestedtransformed plants M1 to MS; b) Southern blot of genomicDNA (8 ug per lane) from control and transformed plants probed with GUS gene (HindIII-EcoRI fragment of pBI 221-