The usefulness of the gfp reporter gene for monitoring Agrobacterium ...

Plant Cell Rep DOI 10.1007/s00299-006-0273-8

GENETIC T RANSFORMAT ION AND HYBRIDIZATION

The usefulness of the gfp reporter gene for monitoring Agrobacterium-mediated transformation of potato dihaploid and tetraploid genotypes Elena Rakosy-Tican · Cristian M. Aurori · Camelia Dijkstra · Ramona Thieme · Adriana Aurori · Michael R. Davey

Received: 7 August 2006 / Revised: 6 November 2006 / Accepted: 16 November 2006 C Springer-Verlag 2006

Abstract Potato is one of the main targets for genetic improvement by gene transfer. The aim of the present study was to establish a robust protocol for the genetic transformation of three dihaploid and four economically important cultivars of potato using Agrobacterium tumefaciens carrying the in vivo screenable reporter gene for green fluorescent protein (gfp) and the marker gene for neomycin phosphotransferase (nptII). Stem and leaf explants were used for transformation by Agrobacterium tumefaciens strain LBA4404 carrying the binary vector pHB2892. Kanamycin selection, visual screening of GFP by epifluorescent microscopy, PCR amplification of nptII and gfp genes, as well as RT-PCR and Southern blotting of gfp and Northern blotting of nptII, were used for transgenic plant selection, identification and analysis. Genetic transformation was optimized for the best performing genotypes with a mean number of shoots expressing gfp per explant of 13 and 2 (dihaploid line 178/10 and cv. ‘Baltica’, respectively). The nptII marker and gfp reporter genes permitted selection and excellent visual screening of transgenic tissues and plants. They also revealed the effects of antibiotic selection on organogenesis and transformation

frequency, and the identification of escapes and chimeras in all potato genotypes. Silencing of the gfp transgene that may represent site-specific inactivation during cell differentiation, occurred in some transgenic shoots of tetraploid cultivars and in specific chimeric clones of the dihaploid line 178/10. The regeneration of escapes could be attributed to either the protection of non-transformed cells by neighbouring transgenic cells, or the persistence of Agrobacterium cells in plant tissues after co-cultivation. Keywords Agrobacterium tumefaciens-mediated transformation . Chimeras . Green fluorescent protein reporter gene . Neomycin phophotransferase marker gene . Transgene silencing Abbreviations GA: Gibberellic acid . gfp: Green fluorescent protein gene . GFP: Green fluorescent protein . gus: β-Glucuronidase gene . GUS: β-Glucuronidase . luc: Luciferase gene . MS: Murashige and Skoog (1962) . NAA: Naphthaleneacetic acid . nptII: Neomycin phosphotransferase gene . PCR: Polymerase chain reaction . RT-PCR reverse: Transcriptase-polymerase chain reaction

Communicated by L. Peña E. Rakosy-Tican () · C. M. Aurori · A. Aurori Babes-Bolyai University, Plant Genetic Engineering Group, 400006 Cluj-Napoca, Romania e-mail: [email protected], [email protected] C. Dijkstra · M. R. Davey School of Biosciences, University of Nottingham, Sutton Bonington Campus, Loughborough, LE12 5RD UK R. Thieme Federal Centre for Breeding Research on Cultivated Plants, Institute of Agricultural Crops, Groß Lüsewitz, Germany

Introduction Potato, as the fourth most important crop in the world and a food security crop, is an important target for genetic improvement through classical and biotechnological approaches. In vitro culture and recombinant DNA technologies are available to complement conventional breeding, seed multiplication and disease control. Although potato was one of the first crops to be transformed by Agrobacterium tumefaciens (Stiekema et al. 1988; Sheerman and Bevan 1988) and transgenic plants have been generated in several potato cultivars Springer

Plant Cell Rep

(Heeres et al. 2002), the efficiency of plant regeneration is dependent on genotype. Novel gene sequencing, cloning and insertion, as well as functional genomics, are tools that will benefit the improvement of economically important potato cvs. and dihaploid genotypes through reliable transformation protocols. Moreover, refinement of transformation by a visible reporter gene, such as gfp, should prove useful in transferring genes of interest without the need for selectable markers, such as genes for antibiotic resistance, since the latter raise public concerns, particularly in Europe (Halford 2004). Reporter genes that allow selection and visual detection of transgene expression, such as gus and luc, have been used extensively to monitor plant transformation. The most commonly exploited gus gene with its product GUS (Jefferson et al. 1987), has a crucial drawback because the histochemical assay is destructive for plant tissue and, consequently, cannot be exploited for analysis in vivo of transgene expression. The luc gene that can be monitored in vivo requires an exogenous substrate, luciferin, and sophisticated equipment (Ow et al. 1986). In contrast, the gfp gene from the jellyfish Aequorea victoria (Chalfie et al. 1994) does not need an exogenous substrate, can be monitored in vivo without an additional cofactor, and its expression is visible under fluorescence microscopy or even laboratory fluorescent illumination (Molinier et al. 2000). GFP expression is visible in whole plants, permitting screening of primary transformants and the discrimination of homozygous and hemizygous plants (Molinier et al. 2000; Ghorbel et al. 1999; Zhang et al. 2001). GFP has become useful for studying promoter expression, monitoring of protein trafficking, cell compartmentalisation, virus infection and virus-induced gene silencing (Rizzuto et al. 1995; Kohler 1998; Ruiz et al. 1998), as well as monitoring transformation (Haselof et al. 1997; Rouwendal et al. 1997; Elliott et al. 1999; Monilier et al. 2000; Kaeppler et al. 2001). It has been used as a marker for somatic hybridization in Citrus (Olivares-Fuster 2002) and to produce green fluorescent flowers (Mercuri et al. 2001). However, in potato, gfp has been used only in stable chloroplast transformation (Sidorov et al. 1999). In this report, gfp and the selectable marker gene nptII, were used to monitor Agrobacterium tumefaciens-mediated transformation of dihaploid genotypes and economically important tetraploid cvs. of potato. Gfp expression was reliable as it allowed visual screening of expression from callus to whole plants, with in vivo discrimination of escapes and chimeras, as well as genotype-dependent effects of kanamycin on organogenesis. All genotypes were transformed, but leaf explants of the dihaploid line 178/10 and stem internode explants of the cv. ‘Baltica’ exhibited the greatest transformation frequencies with mean values of 13 and 2 transgenic shoots per initial explant, respectively. The utility of gfp, the limitations of kanamycin selection Springer

and the possible causes of escapes and chimeras, are also discussed.

Material and methods Plant material and transformation Leaf and stem explants of the dihaploid genotypes 178/10, 224/1 and 227/5 were excised from in vitro propagated plants 4 weeks after transfer of the stock plants to RMB5 medium (Menzel et al. 1981). Dihaploids were produced by parthenogenesis after crossing of the tetraploid cvs. ‘Amsel’, ‘Nicola’ and ‘Sante’ with Solanum phureja at the Institute for Potato Research and Production, Bras¸ov, Romania. The tetraploid cvs. ‘Désirée’ (used as reference), ‘Agave’ and ‘Delikat’ (Norika, Gross Lüsewitz, Germany) and ‘Baltica’ (SakaRagis-Pflanzenzucht, Saka-Ragis, Germany), were propagated in vitro on semi-solid MS-based medium enriched with 1.15 g L−1 NH4 NO3 (designed MS.5). Generally, internodal explants from 4-week-old in vitro grown plants of tetraploid cvs. were inoculated with A. tumefaciens LBA4404 harboring pHB2892 (provided by G. Hahne and J. Molinier) for transformation (Fig. 1). Plasmid HB2892 contained the sgfp gene, under the control of a double CaMV 35S promoter, adjacent to the left T-DNA border and the nos-nptII-nos selectable marker gene next to the right T-DNA border (Molinier et al. 2000). The maintenance of bacteria, co-cultivation of explants with bacteria and culture of stem and leaf explants, were as described by Kumar (1995). Since LSR1 followed by LSR2 culture media (Kumar, 1995) induced limited shoot regeneration from tetraploid cvs., MS-based medium with 16 g L−1 glucose, 0.5 mg L−1 folic acid, 0.05 mg L−1 biotin, 40 mg L−1 adenine, 0.02 mg L−1 GA3, 0.02 mg L−1 NAA, 2.0 mg L−1 zeatin riboside, and 0.75% (w/v) agar at pH 5.8 (designed MS-t medium) was assessed. Agrobacterium growth was controlled by addition of 250 mg L−1 cefotaxime (Claforan; Hoechst-Roussel Pharmaceuticals, Frankfurt, Germany) to the culture medium with or without 50 mg L−1 kanamycin (Sigma–Aldrich, Poole, UK). Comparative experiments were undertaken with permanent kanamycin selection or with the elimination of kanamycin from the medium after 5 weeks of culture. Uninoculated explants were cultured in the same way as explants inoculated with Agrobacterium. Cultures were maintained at 22 ± 2◦ C with a 16 h photoperiod (90 µmol m−2 s−1 , Daylight fluorescent illumination). Transgenic tissue, plant selection and analysis Putative transgenic callus and shoots were selected on kanamycin-containing medium followed by visualising gfp expression by fluorescence microscopy (BX-60 microscope

Plant Cell Rep Fig. 1 Structure of the T-DNA of pHB2892 with gfp and nptII genes (Molinier et al. 2000)

NcoI

BamH

MCS

CaMV Enhancer....CaMV Enhancer

NcoI

Ω

BamHI

SGFP

BamHI

LB

MCS CaMV term

with wide band cube unit U-MWB; Olympus GmbH, Hamburg, Germany) to discriminate the green fluorescence of gfp from the native red fluorescence of chlorophyll. Each regenerated shoot exhibiting green fluorescence was considered putatively transgenic; shoots showing red chlorophyll autofluorescence were designated escapes. Shoots with dual green and red fluorescence in different tissues were chimeras. Three replicates of each experiment were set up, each with at least 10–20 individual explants. Control and putative transgenic shoots were analysed for integration of the nptII gene by PCR amplification and Northern blotting, and of gfp by PCR, RT-PCR and Southern blotting. PCR amplification Primers were manufactured and sequenced by MWG Biotech, Ebersberg, Germany. The sequences used for nptII were 5 -AGA CAA TCG GCT GCT CTG AT-3 and 5 -ATA CTT TCT CGG CAG GAG CA-3 and for gfp 5 -AGG GCG ATG CCA CCT A-3 and 5 -GAC TGG GTG CTC AGG TA-3 . Template genomic DNA was extracted using a GenElute Plant Miniprep kit (Sigma–Aldrich). PCR was performed using RED Taq Ready Mix (Sigma) according to the manufacturer’s instructions. Amplification was performed in a DNA Thermal Cycler 480 (Perkin Elmer Applied Biosystems Division, Warrington, UK), with initial denaturing (1 cycle, 94◦ C, 3 min), denaturating (35 cycles, 94◦ C, 1 min), primer annealing [35 cycles, 55◦ C (nptII) or 53◦ C (gfp), 1 min], primer extension (35 cycles, 72◦ C, 90 s), final extension (1 cycle, 72◦ C, 10 min) and holding at 4◦ C (5 min to ∞ ). RT-PCR analysis RNA was extracted from approx. 100 mg leaf samples of putatively transformed and non-transformed plants using an r RNeasy Plant Mini Kit (Qiagen, Crawley, UK). RT-PCR

Lac Z

Nos

nptII

Nos

RB

was peformed using a One Step RT-PCR Kit (Qiagen) according to the manufacturer’s instructions. The amplification programme involved reverse transcription (1 cycle, 50◦ C, 30 min), polymerase activation (1 cycle, 94◦ C, 15 min), denaturation (35 cycles, 94◦ C, 1 min) and primer annealing (35 cycles, 55◦ C, 1 min). Subsequent conditions were as for PCR analysis. NptII gene expression studies using Northern blot analysis Total RNA (15 µg) was isolated from leaves using an r RNeasy Plant Mini Kit (Qiagen). RNA was capillary blotted onto a nylon membrane (Roche Diagnostics Ltd., Lewes, UK). The membrane was prehybridised in DIG-Easy-Hyb buffer (Roche) at 68◦ C for 1 h. For hybridisation, 10 mL of DIG-Easy-Hyb, to which had been added DIG-labelled RNA probes (100 ng mL−1 ; Roche), were used per 100 cm2 of membrane. Hybridisation was at 68◦ C for 16 h. The membrane was washed for 2 × 5 min in 2 × SSC/0.1% SDS and for 2 × 15 min in 0.1 × SSC/0.1% SDS at 68◦ C. The membrane was preincubated for 45 min in blocking solution (Roche) and incubated for 30 min with anti-DIG-AP conjugate (diluted 1:10,000, v:v). Two 15 min washing steps preceeded equilibration in detection buffer for 5 min. Detection involved CDPStar (Roche) with a 20 min exposure to X-ray film (Kodak). Southern blotting analysis for the gfp gene Genomic DNA (10 µg) was isolated from leaves using a GenElute Plant Genomic DNA kit (Sigma–Aldrich). The DNA was digested with EcoRI (Promega) according to the manufacturer’s instructions. Digested DNA was capillary blotted onto nylon membrane (Roche Diagnostics Ltd., Lewes, UK) and prehybridised in DIG-Easy-Hyb buffer (Roche) at 37◦ C for 2 h. For hybridisation, 30 mL of DIGEasy-Hyb, to which had been added DIG-labelled DNA

Springer

Plant Cell Rep 45 178/10

40

A

224/1 35 Shoots per explant

probes (100 ng mL−1 ; Roche), were used at 68◦ C for 16 h. The membrane was washed twice (5 min each) in 2 × SSC/0.1% SDS and twice (15 min each) in 0.1 × SSC/0.1% SDS at 65◦ C. The membrane was preincubated for 45 min in blocking solution (Roche) and incubated for 30 min with anti-DIG-AP conjugate (diluted 1:10,000, v:v). Two 15 min washing steps preceded equilibration in detection buffer for 5 min. Detection involved CDPStar (Roche) with a 5 h exposure to X-ray film (Kodak).

227/5

30 25 20 15 10 5

Results

0 C

Although transformation efficiency was dependent on plant genotype, all genotypes were transformed and regenerated shoots expressed constitutively gfp fluorescence. The dihaploids regenerated well from both leaf and stem explants on LSR1/LSR2 media, as recommended by Kumar (1995) for the cv. ‘Désirée’, a model system in potato transformation. Leaf explants of tetraploid genotypes tended to be overgrown by Agrobacterium and were unable to form callus or shoots. Optimum regeneration was obtained for the four tetraploid cvs. when stem explants were cultured on MS-t medium, as these explants were less affected by Agrobacterium overgrowth. The dihaploid genotypes 178/10 and 224/1 regenerated the most shoots from leaf explants when kanamycin was eliminated from the medium after 5 weeks of culture (Fig. 2A). Interestingly, the dihaploid genotype 227/5, which did not regenerate shoots from uninoculated explants, did regenerate a limited number of gfp-expressing shoots following Agrobacterium inoculation and kanamycin selection (Fig. 2A). Moreover, the percentage of gfp-expressing shoots was dependent on genotype, but was greater with kanamycin selection for the majority of dihaploids (Fig. 2B). A relatively high percentage of shoots not expressing gfp, ranging from 30 to 100%, were also regenerated even under permanent kanamycin selection. The dihaploid 178/10, known to be the most responsive dihaploid genotype in vitro, also regenerated the most shoots after Agrobacterium inoculation. Leaf explants were preferable for the dihaploid genotype 178/10 Table 1 Comparison of shoot regeneration from stem internodes of tetraploid potato cultivars, cultured on LSR1 followed by LSR2 media, or on MS-t medium Potato tetraploid cv.

Shoots per explant LSR1 followed by LSR2a MS-t medium

Agave Baltica Delikat Désirée

0 16 ± 2.0 0 2 ± 0.5

a

Kumar (1995). Data are means ± SEM; n = 30.

Springer

4 ± 0.5 38 ± 4.0 5 ± 0.8 5 ± 0.7

A1 stem

A2

C

A1

A2

leaf

100%

B

90% 80% 70% 60% gfp-

50%

gfp+/gfp+

40% 30% 20% 10% 0% S-224 SS-227 L-178 L-224 L-224 L-227 L-227 A1 227A1 A2 A1 A1 A2 A1 A2

Fig. 2 Mean number of potato shoots regenerated on leaf or stem explants A and the percentage of shoots regenerated on leaf (L) or stem (S) internodal explants B, after co-cultivation of explants with A. tumefaciens LBA4404 pHB2892. Dihaploid potato lines expressing green fluorescence (gfp + ), not expressing fluorescence (gfp − ) or chimeric for this trait (gfp + / − ), were assessed after 3 months of culture on LSR medium with 250 mg L−1 cefotaxime. C = control; A1, selection on 50 mg L−1 kanamycin for the first 5 weeks of culture; A2, permanent selection on 50 mg L−1 kanamycin. (Bars = SEM)

with a mean number of regenerated shoots of 26 from uninoculated explants and 34.5 per explant after Agrobacterium cocultivation. Of the latter, 38% expressed gfp and 15% were chimeras when kanamycin was removed after 5 weeks of culture (Fig. 2A and B). Tetraploid cultivars also exhibited genotype-dependent regeneration, as well as optimum transformation efficiency on MS-t medium (Fig. 3). Previous experiments showed that more shoots were regenerated for all tetraploid cvs. on MS-t medium than on LSR1 medium followed by LSR2 medium (Table 1), although the latter sequence was recommended by Kumar (1995) for the cv. ‘Désirée’. In all experiments, uninoculated (control) explants were unable to regenerate shoots on selection medium containing kanamycin. Preliminary experiments with A. tumefaciens EHA105 pGPTV revealed a tendency of Agrobacterium to over-grow leaf explants (data not shown). Consequently, only stem internodal

Plant Cell Rep 45

A 40 Shoots per explant

Control 35

A1 A2

30 25 20 15 10 5 0 Delikat

Desiree

Baltica

Agave

% Transformation

100 90

A1

80

A2

B

70 60 50 40 30 20 10 0 Delikat

Desiree

Baltica

Agave

Fig. 3 Mean number of potato shoots regenerated on stem internodal explants (A) and the percentage of regenerated shoots expressing gfp (B) in potato tetraploid cvs. following co-cultivation with A. tumefaciens LBA4404 pHB2892 and culture for 4 months on MS-t medium with 250 mg L−1 cefotaxime. A1, selection on 50 mg L−1 kanamycin for the first 5 weeks of culture; A2, permanent selection on 50 mg L−1 kanamycin. (Bars = SEM)

explants were co-cultivated with Agrobacterium LBA4404 pHB2892 on MS-t medium. The most regenerated shoots were obtained for the cv. ‘Baltica’ (Fig. 3A), which showed no significant difference in the percentage of gfp expressing shoots under permanent kanamycin selection, or when kanamycin was eliminated from the medium after 5 weeks (Fig. 3B). The cvs. ‘Désirée’ and ‘Agave’ regenerated most shoots per explant when kanamycin selection was not applied after 5 weeks of culture (Fig. 3A). The percentage of gfp expressing shoots was for all cvs. and, particularly for the ones showing poor shoot regeneration (‘Delikat’ and ‘Agave’), always better under permanent kanamycin selection (Fig. 3B). The tetraploid cvs. showed polarity of regeneration from stem internodal explants, with shoots developing at the apical pole and callus at the basal end, as illustrated for the cv. ‘Baltica’ (Fig. 4). The most responsive genotypes, namely the dihaploid 178/10 and the cv. ‘Baltica’, also regenerated callus and shoots more rapidly, the first shoots developing

within 3–4 weeks of culture in control, uninoculated leaf or stem explants. After Agrobacterium co-cultivation and kanamycin selection, the polarity of regeneration was maintained, but shoot regeneration was delayed by 2–4 weeks, depending on the genotype (Fig. 4). A quite high percentage (30–100%) of regenerated shoots did not express gfp, as in dihaploids, and some chimeras were observed in tetraploid cultivars (Fig. 5). Chimeras were identified in the cv. ‘Désirée’, with low gfp expression in a secondary shoot (Fig. 5D); in the cv. ‘Delikat’, a plant was generated with a green fluorescent stem and meristem, but no GFP in the young leaves (Fig. 5E and F). Interesting chimeric clones were identified in regenerated shoots of the dihaploid genotype 178/10, including silencing of the gfp gene during apical bud development in one clone (Fig. 6), and silencing in collenchymatous cells in another clone (Fig. 7A–D). A chimera exhibiting silencing of gfp in the meristem of the axillary bud compared to a transgenic clone of the same genotype expressing gfp constitutively is shown in Fig. 7E and F. NptII and gfp primers amplified 261 and 497 bp regions of the respective genes. Molecular analysis of four chimeric clones of the cv. ‘Delikat’, showed that all plants had integrated the nptII gene, as shown by PCR amplification (Fig. 8). Northern analysis revealed that nptII expression was less in the chimeric clone T3 as compared to the transgenic clone T4 of the cv. ‘Delikat’ (Fig. 9). PCR amplification, RT-PCR confirmed the integration and expression of the gfp gene in all the clones classified as transgenic by visual screening of green fluorescence, and the absence of the gfp gene in the clones exhibiting the red autofluorescence of chlorophyll, considered as escapes (Figs. 10 and 11). The integration of the gfp gene in putative transgenic clones exhibiting green fluorescence was confirmed by PCR and Southern blot analyses (Fig. 12), the latter showing the copy number to vary between 2 in a chimeric plant and 5 in clones constitutively expressing gfp. Neither gfp nor nptII were detected by PCR analysis in controls or escapes (Figs. 8 and 10). RT-PCR analysis of the gfp gene revealed no differences in its expression in different transgenic clones of dihaploid or tetraploid genotypes of potato (Fig. 11).

Discussion GFP is a useful tool for monitoring the efficiency of transformation, the effects of antibiotics on organogenesis and transgenic plant selection, and the occurrence of escapes and chimeras in potato dihaploid genotypes and tetraploid cultivars. Previously, the cv. ‘Désirée’ was used extensively as a model for potato transformation, but the German commercial potato cultivars or dihaploid genotypes used in this study, were not subjected previously to Agrobacterium-mediated gene transfer. Developing an efficient procedure for genetic Springer

Plant Cell Rep

Fig. 4 Example of polar regeneration of potato shoots and callus on internodal explants of the cv. ‘Baltica’ after 4 weeks of culture on MS-t medium. A Polar regeneration in explants not inoculated with Agrobacterium, showing shoots at the apical pole and callus at the root pole. B Callus development at the root pole and

first developing shoots (arrowed) at the apical pole of an explant on kanamycin selection, following co-cultivation with A. tumefaciens LBA4404 pHB2892. Shoot regeneration is delayed after Agrobacterium co-cultivation. (Bars = 1 cm)

transformation of tetraploid commercial varieties of potato provides the opportunity for efficient transfer and integration of genes to improve traits such as disease resistance or crop quality. A similar study was reported for 16 commercial potato cvs. using the nptII marker gene and an antisense gene for granule-bound starch synthase (Heeres et al. 2002), with low transformation frequencies varying with genotype from 0.02–0.35 transgenic shoots per stem internode explant. Another recent study reported 0.3–0.6 transgenic shoots per stem explant in potato (Rommens et al. 2004).

In comparison to published data, the efficiency of transformation reported in the present paper is high for potato, with some genotypes, such as the dihaploid 178/10 regenerating 13 green fluorescent shoots per leaf explant. The cv. ‘Baltica’, with a mean of 2 transgenic shoots per initial explant, surpassed the best performances of the cv. ‘Désirée’ (Dietze et al. 1995). Importantly, efficient genetic transformation of specific dihaploid genotypes may be useful for further tetraploidization and, consequently, increasing transgene stability in subsequent generations.

Fig. 5 Examples of gfp expression in regenerated potato plants. A Stems of the dihaploid 178/10 with gfp (left) and red chlorophyll fluorescence (right). Tetraploid cvs. ‘Delikat’ B lacking gfp expression (escape shoot) and ‘Désirée’ C expressing gfp; D chimeric lateral shoot primordium (arrowed) emerging on a gfp-expressing shoot of the cv. ‘Désirée’; E shoot of ‘Delikat’ and its tip F showing gfp expression in the stem and apical meristem, but lack of GFP in leaf primordial. (Bars = 1 mm)

Springer

Plant Cell Rep Fig. 6 Expression of gfp during stages of potato bud development (A–F) in a putative transgenic dihaploid clone 178/10. (Bar = 2.5 mm)

A relatively high percentage of shoots not expressing gfp were regenerated under permanent kanamycin selection in all potato genotypes tested. Molecular analysis revealed that all shoots regenerated on kanamycin that did not show green fluorescence, had not integrated the nptII gene. Such a high percentage of these shoots may be possible because auxin pretreatment (10 mg L−1 NAA and 10 mg L−1 zeatin for 2 days), as an initial step of the protocol (Kumar 1995), could act as a signal stimulating the regeneration of escapes. It also accounts for the polarity of regeneration as demonstrated for Citrus transformation (Dom´ınguez et al. 2004). Moreover, as shown in Citrus (Peña et al. 2004), dedifferentiation is required to shift cells to a competent state for stable transformation, and dedifferentiation is triggered by auxins during co-cultivation, as in Citrus, or as a pretreatment in the current experiments. Another possible factor responsible for the escapes is the concentration of kanamycin, which may have been too low, although employed at the concentration recommended previously for routine transformation of the cv. ‘Désirée’ (Kumar 1995). The persistence of agrobacteria during shoot regeneration and the propagation of pu-

tative transgenic clones in vitro over prolonged time after co-cultivation of explants with bacteria, may also account for the high percentage of escapes. Similar results have been reported for Agrobacterium-mediated transformation of Citrus (Dom´ınguez et al. 2004). The present results reinforce the previous assumption (Dom´ınguez et al. 2004) that selection with kanamycin after Agrobacterium co-cultivation did not prevent the regeneration of shoots that failed to integrate the nptII gene. However, in contrast to Citrus (Dom´ınguez et al. 2004), PCR and RT-PCR for potato in the present study did not show any case where escapes had integrated, but not expressed, the nptII gene. A kanamycin concentration of 50 mg L−1 and cocultivation with Agrobacterium showed genotype-dependent effects on organogenesis, reducing shoot yield and delaying shoot regeneration by 2–4 weeks, depending on the genotype. Cefotaxime, used mainly at 250 mg L−1 , also influenced organogenesis with stimulation of shoot regeneration in previous studies (data not presented). Other authors have recommended, for systems that generate many escapes and chimeras, the use of a screenable strategy instead of lethal

Springer

Plant Cell Rep Fig. 7 Different patterns of expression of gfp in stem sections of transgenic potato dihaploid 178/10. A Clone with constitutive gfp expression. B Stem with red chlorophyll fluorescence of a control (non-transformed) clone. C Transgenic clone where GFP is absent from the collenchyma. D Detail of the same clone; arrow indicates collenchyma cells lacking gfp expression. E General view of transgenic clone with constitutive gfp expression in comparison with a clone F where GFP is absent in the collenchyma and the meristem of an axillary bud (arrowed). (Bars = 1 mm, except D, where bar = 100 µm)

selection (Christou and McCabe 1992), with gfp expression being preferred in previous experiments with other plant species (Ghorbel et al. 1999; Zhang et al. 2001). The present results demonstrate that the use of both gfp and nptII genes may reveal the real picture of genotype-dependent transformation efficiency. Observations of the effects of culture media, antibiotics and Agrobacterium strains on organogenesis

Springer

may also be facilitated and permit rapid optimisation of transformation for a particular genotype. Such a transformation protocol could be further used for efficient transfer of other genes (Rakosy-Tican et al. 2004), eventually eliminating the necessity to use genes for antibiotic resistance (De Vetten et al. 2003). The present data, based on PCR screening and Southern blot analyses of transgenic shoots, also reinforce

Plant Cell Rep

M C0 C+ T1 T2 T3 T4 T5 T6 T7 T8 C1 C2 C3 C4 C5

261 bp

Fig. 8 PCR profiles generated using nptII primers for transgenic potato clones and their respective control. M = 100 bp ladder; C0 = negative control (water); C + = positive control (pHB2892) for nptII gene; T1– T4, T5–T6, T7 and T8 = clones of cvs. ‘Delikat’, ‘Désirée’, ‘Baltica’

T1

T2

C1

T3

T4

Fig. 9 Northen blot analysis of transgenic potato clones. T1, T2 = cv. ‘Agave’ expressing gfp; C1 = non-transformed clone of the cv. ‘Agave’; T3 = a chimeric clone partially expressing gfp of the cv. ‘Delikat’; T4 = transgenic clone expressing gfp of the cv. ‘Delikat’

previous reports in potato with 1–5% transformation efficiency (De Vetten et al. 2003), cabbage (Berthomieu et al. 1994) and Citrus (Dom´ınguez et al. 2004), in which PCR analysis was proposed to identify transgenic plants without the use of selectable markers, since the latter raise concerns in relation to the environment and food safety. Furthermore, the present results reveal the key role of the culture medium and auxin pretreatment to increase transgenic shoot regeneration in responsive cvs. In contrast with the results of Dom´ınguez et al. (2004) in Citrus, there was no molecular proof that transformation of potato may occur at a greater frequency than indicated by screening for the gfp reporter gene.

and ‘Agave’ expressing gfp; C1–C5 = DNA from non-transgenic plants of cvs. ‘Agave’, ‘Delikat’, ‘Baltica’ (2) and ‘Désirée’, respectively

The regeneration of chimeras with green fluorescence in some tissues or cells but not in others, was also reported in tobacco and analysed quantitatively, suggesting involvement of a special type of gene silencing (Bastar et al. 2004). Moreover, by analysing the intensity of gfp expression, homozygous and hemizigous states could be visualised in tobacco, with Mendelian segregation being reported in T1 and T2 generations (Molinier et al. 2000). Differences in the intensity of green fluorescence were not observed in the current experiments, and chimeras occurred only in some of the genotypes at low frequency ( < 0.5%), except in the dihaploid 178/10, where 15% of the shoots regenerated from leaf explants were chimeric. Site-specific integration effects, as well as reduction in GFP protein synthesis or gene silencing, may be the causes of the chimeras observed to date. In contrast to tobacco (Molinier et al. 2000), natural fluorescence that interfered with gfp expression was not evident in potato. All transgenic shoots expressing gfp were stable during repeated cloning in vitro, without any adverse effects on growth or development. Variation was observed in gfp expression with silencing during bud development, in

M C+ C0 C1 C2 C+ C3 C4 C5 R1 T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12 T13

497 pb

Fig. 10 PCR profiles generated using gfp primers for transgenic potato clones and their respective controls. M = 100 bp. ladder; C + = positive control (pHB 2892) for gfp; C0 = negative control (water); C1– C5 = non-transgenic clones of the cvs. ‘Agave’, ‘Delikat’, ‘Baltica’, ‘Désirée’ and the dihaploid 178/10, respectively; R1 = clone of the

M C0

C+

T1

T2

T3

cv. ‘Baltica’ not expressing gfp; transgenic clones expressing gfp of the cvs. ‘Agave’ (T1, T2), ‘Delikat’ (T3, T4), ‘Baltica’ (T5–T7), ‘Désirée’ (T8) and the dihaploid 178/10 (T9); T10–T13 = chimeras of the cv. ‘Delikat’

T4 T5 T6

Fig. 11 RT-PCR profiles generated using gfp primers for transgenic potato clones and their respective controls. M = 100 bp. ladder; C0 = negative control (water); C + = positive control (pHB 2892) for gfp; C1 = non-transgenic clone of the cv. ‘Agave’; transgenic clones

T7

T8

T9 T10 T11 C1

expressing gfp of the cvs. ‘Agave’ (T1, T2), ‘Delikat’ (T3, T4), ‘Baltica’ (T5, T6), ‘Désirée’ (T7) and the dihaploid 178/10 (T10); T8, T9 and T11 = chimeras of the cv. ‘Delikat’

Springer

Plant Cell Rep

M

T1

T2

T3

C1

2176 bp 1766 bp 1230 bp 653 bp 234 bp Fig. 12 Southern analysis of DNA from selected transgenic potato clones using a gfp probe. M = DIG-labelled DNA Molecular Weight Marker VI (Roche); T1 = a chimeric clone of the cv. ‘Delikat’ with a copy number of 2; T2, T3 = transgenic clones of the cvs. ‘Agave’ (T2) and ‘Désirée’ (T3) constitutively expressing gfp, both with five copies of the gene; C1 = non-transformed clone of the cv. ‘Agave’

collenchymatous cells in some chimeric clones of the dihaploid 178/10. Silencing of gfp was also seen in lateral branches and leaf primordia of specific tetraploid potato cvs. Such variation in gfp expression during development may be explained by integration site-specific gene inactivation; such plants may represent valuable material for further molecular investigations. Preliminary data from the experiments with potato indicated that Agrobacterium strain EHA105 compared to LBA4404, may also affect explant response to co-cultivation, but further research is required to improve understanding of Agrobacterium–plant tissue interaction, as indicated by Peña et al. (2004). Visual screening of gfp expression may be used to optimise genetic transformation of specific cultivars of potato to eliminate the difficulties associated with kanamycin selection. Selection on kanamycin-containing medium could not discriminate transgenic shoots from escapes and chimeras. Kanamycin has genotype-dependent effects on organogenesis in potato and increasing its concentration for tighter selection may affect the regeneration of transgenic plants. Such effects can be monitored easily using the gfp reporter–nptII marker system that may be suitable for other plant species. In conclusion, a gfp reporter–nptII marker system was used to optimise genetic transformation of dihaploid and tetraploid potato genotypes. Genotype-dependent response to culture media, the effects of kanamycin selection and the transformation protocol, as well as the frequency of escapes, are important factors for establishing strategies to enhance transformation efficiency in this tuberous crop. HowSpringer

ever, it was also found, as indicated previously (Dom´ınguez et al. 2004), that there are several assumptions relating to Agrobacterium-mediated transformation that need to be critically assessed. Transformation efficiency, based on resistance to a selective agent, may not reveal the actual frequency of transgenic plant regeneration in potato. However, screening for the expression of the gfp gene gave the real figure for transformation in the present experiments. Importantly, gfp allows identification of escapes and chimeras in vivo, permitting further investigation of the causes for the regeneration of escapes in some genotypes, together with studies on silencing and site-integration effects. Furthermore, for vegetatively propagated crops that are known to regenerate chimeras and escapes (Dom´ınguez et al. 2004), establishing a robust protocol for transformation, avoiding marker based-selection of transgenic shoots, modifying the antibiotic treatment to reduce Agrobacterium persistence after co-cultivation and screening transgenic events by PCR, are advisable to maximise transformation efficiencies and the generation of marker-free plants. Acknowledgements A German Academic Exchange Service (DAAD) Scholarship and a German–Romanian Bilateral Project are acknowledged for supporting part of this research (ER-T). The EU Erasmus–Socrates programme provided partial support for CD. Dr. G. Hahne and Dr. J. Molinier kindly provided pHB2892.

References ˇ Bastar MT, Luthar Z, Skof S, Bohanec B (2004) Quantitative determination of mosaic GFP gene expression in tobacco. Plant Cell Rep 22:939–944. DOI 10.1007/s00299-004-0782-2 Berthomieu P, Beclin C, Charlot F, Dore C, Jouanin L (1994) Routine transformation of rapid cycling cabbage (Brassica oleracea)— molecular evidence for the regeneration of chimeras. Plant Sci 96:223–235 Chalfie M, Tu Y, Euskirchen G, Ward WW, Prasher DC (1994) Green fluorescent protein as a marker for gene expression. Science 263:802–805. DOI 10.1126/science.8303295 Christou P, McCabe DE (1992) Prediction of germ-line transformation events in chimeric R0 transgenic soybean plantlets using tissue-specific expression patterns. Plant J 2:283–290. DOI 10.1111/j.1365-313X.1992.00283.x De Vetten N, Wolters AM, Raemakers K, Van Der Meer I, Ter Stege R, Heeres E, Heeres P, Visser R (2003) A transformation method for obtaining marker-free plants of a cross-pollinated and vegetatively propagated crop. Nat Biotechnol 21:439–442. DOI 10.1038/nbt801 Dietze J, Blau A, Willmitzer L (1995) Agrobacterium-mediated transformation of potato (Solanum tuberosum). In: Potrykus I, Spangenberg G (eds) Gene transfer to plants. Springer, Berlin Heidelberg New York, pp 24–29 Dom´ınguez A, Cervera M, Pérez RM, Romero J, Fagoaga C, Cubero J, López MM, Juz JA, Navaro L, Peña L (2004) Characterisation of regenerants obtained under selective conditions after Agrobacterium-mediated transformation of citrus explants reveals production of silenced and chimeric plants at unexpected high frequencies. Mol Breed 14:171–183. DOI 10.1023/B:MOLB.0000038005.73265.61

Plant Cell Rep Elliott AR, Campbell JA, Dugdale B, Brettel RIS, Grof CPI (1999) Green fluorescent protein facilitates rapid in vivo detection of genetically transformed plant cells. Plant Cell Rep 18:707–714. DOI 10.1007/s002990050647 Ghorbel R, Juárez J, Navarro L, Peña L (1999) Green fluorescent protein as a screenable marker to increase the efficiency of generating transgenic woody fruit plants. Theor Appl Genet 99:350–358. DOI 10.1007/s001220051244 Halford NG (2004) Prospects for genetically modified crops. Ann Appl Biol 145:17–24. DOI 10.1111/j.1744-7348.2004.tb00355.x Haselhof J, Siermering KR, Prasher DC, Hodge S (1997) Removal of a cryptic intron and subcellular localization of green fluorescent protein are required to mark transgenic Arabidopsis plants brightly. Proc Natl Acad Sci USA 94:2122–2127 Heeres P, Schippers-Rozenboom M, Jacobsen E, Visser RGF (2002) Transformation of a large number of potato varieties: genotypedependent variation in efficiency and somaclonal variability. Euphytica 124:13–22. DOI 10.1023/A:1015689112703 Jefferson RA, Kavanagh TA, Bevan MW (1987) GUS fusion: βglucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J 6:3901–3907 Kaeppler HF, Carlson AR, Menon GK (2001) Routine utilization of green fluorescent protein as a visual marker for cereal transformation. In vitro Cell Dev Biol–Plant 37:120–126 Köhler R (1998) GFP for in vivo imaging of subcellular structures in plant cells. Trends Plant Sci 3:317–320 Kumar A (1995) Agrobacterium-mediated transformation of potato genotypes. In: Gartland KMA, Davey MR (eds) Methods in molecular biology, vol 44. Humana Press Inc, New York Totowa, pp 121–128 Menczel L, Nagy F, Kiss ZsR, Maliga P (1981) Streptomycin resistant and sensitive somatic hybrids of Nicotiana tabacum + Nicotiana knightiana correlation of resistance to N. tabacum plastids. Theor Appl Genet 59:191–195. DOI 10.1007/BF00264975 Mercuri A, Sacchetti A, De Benedetti A, Schiva T, Alberti S (2001) Green fluorescent flowers. Plant Sci 161:961–968 Molinier J, Himber C, Hahne G (2000) Use of green fluorescent protein for detection of transformed shoots and homozygous offspring. Plant Cell Rep 19:219–223. DOI 10.1007/s002990050002 Murashige T, Skoog F (1962) A revised medium for rapid growth and bio-assays with tobacco tissue cultures. Physiol Plant 15:473–497

Olivares-Fuster O, Peña L, Duran-Vila N, Navarro L (2002) Green fluorescent protein as a visual marker in somatic hybridization. Ann Bot 89:491–497. DOI 10.1093/aob/mcf54 Ow DW, Wood KV, DeLuca M, Wet JR de, Helinski DR, Howell SH (1986) Transient and stable expression of the firefly luciferase gene in plant cell and transgenic plants. Science 234:856–859 Peña L, Pérez RM, Cervera M, Juárez JA, Navarro L (2004) Early events in Agrobacterium-mediated genetic transformation of Citrus explants. Ann Bot 94:67–74. DOI 10.1093/aob/mch117 Rakosy-Tican L, Aurori A, Aurori CM, Ispas G, Famelaer I (2004) Transformation of wild Solanum species resistant to late blight by using reporter gene gfp and msh2 genes. Plant Breed Seed Sci 50:119–128 Rommens CM, Humara JM, Ye J, Yan H, Richael C, Zhang L, Perry R, Swords K (2004) Crop improvement through modification of the plant’s own genome. Plant Physiol 135:421–431. DOI 10.1104 Rizzuto R, Brini M, Pizzo P, Murgia M, Pozzan T (1995) Chimeric green fluorescent protein as a tool for vusualizing subcellular organelles in living cells. Curr Biol 5:635–642. DOI 10.1016/SO9609822(95)00128-x Rouwendal GJA, Mendes O, Wolbert EJH, Douwe de Boer A (1997) Enhanced expression in tobacco of the gene encoding green fluorescent protein by modification of its codon usage. Plant Mol Biol 33:989–999. DOI 10.1023/A:1005740823703 Ruiz MT, Voinnet O, Baulcombe DC (1998) Initiation and maintenance of virus-induced gene silencing. Plant Cell 10:937–946 Sidorov VA, Kasten D, Pang S-Z, Hajdukiewicz PTJ, Staub JM, Nehra NS (1999) Stable chloroplast trasnsformation in potato: use of green fluorescent protein as a plastid marker. Plant J 19:209–216. DOI 10.1046/j.1365-313X.1999.00508.x Sheerman S, Bevan MW (1988). A rapid transformation method for Solanum tuberosum using binary Agrobacterium tumefaciens vectors. Plant Cell Rep 7:13–16. DOI 10.1007/BF00272967 Stiekema WJ, Heidekamp F, Louwerse JD, Verhoeven HA, Dijkhuis P (1988) Introduction of foreign genes into potato cultivars Bintje and Désirée using an Agrobacterium tumefaciens binary vector. Plant Cell Rep 7:47–50. DOI 10.1007/BF00272976 Zhang CL, Chen DF, McCormac AC, Scott NW, Elliot MC, Slater A (2001) Use of the GFP reporter as a vital marker for Agrobacterium-mediated transformation of suger beet (Beta vulgaris L.). Mol Biotechnol 17:109–117 (ISSN 1073-6085)

Springer