Transient gene expression in electroporated banana (Musa spp., cv. ?Bluggoe?, ABB group) protoplasts isolated from regenerable embryogenetic cell suspensions

Plant Cell Reports

Plant Cell Reports (1994) 13:262-266

9 Springer-Verlag1994

Transient gene expression in electroporated banana (Musa spp., cv. 'Bluggoe', ABB group) protoplasts isolated from regenerable embryogenetic cell suspensions Laszlo Sagi 1,., Serge Remy 1.., Bart Panis 2, Rony Swennen 2, and Guido Volckaert 1 x Laboratory of Gene Technology, Catholic University of Leuven, Willern de Croylaan 42, B-3001 Leuven, Belgium 2 Laboratory of Tropical Crop Husbandry, Catholic University of Leuven, Kardinaal Mercierlaan 92, B-3001 Leuven, Belgium * Present address: Laboratory of Tropical Crop Husbandry, Catholic University of Leuven, Kardinaal Mercierlaan 92, B-3001 Leuven, Belgium Received 12 June 1993/Revised version received 28 October 1993 - Communicated by H. Ltrz

S u m m a r y , Electroporation conditions were established for transient expression of introduced DNA in banana ( M u s a spp., cv. 'Bluggoe') protoplasts isolated from regenerable embryogenic cell suspensions. The following p a r a m e t e r s were found to be highly influential: electroporation buffer, polyethylene glycol treatment and its duration before electroporation, use of a heat shock, and chimaeric gene constructs. The maximum frequency of DNA introduction as detected by an in situ assay for transient expression of the uidA gene, amounted to 1.8% of total protoplasts. Since plants have recently been regenerated from banana protoplasts at a high frequency, the present results may contribute to the production of transgenic banana.

Key words: Direct gene transfer - Electroporation - I]Glucuronidase - Banana - Protoplast Abbreviations. AMV = alfalfa mosaic virus, CaMV = cauliflower

mosaic virus, 2,4-D = 2,4-dichlorophenoxyaceticacid, EGTA= ethylene glycoI-O-O'-bis(2-aminoethyl)-N,N,N',N'-tetraaeeticacid, GUS = 15glucuronidase, HEPES = 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid, MES = 2-morpholinoethanesulfonieacid, MS = Mnrashige-Skoog. NOS = nopaline synthase, NFrII = neomycinphosphotransferase,PEG = polyethylene glycol, TGE = transient GUS expression, X-Glue = 5bromo-4-chloro-3-indolyl15-D-glucuronieacid

Introduction Banana and plantain (Musa spp.) are the world's second largest fruit crop with an annual production of "74 million tons (FAO 1991) and the main yield reducing factors are fungal and viral diseases. The annual cost of chemical control of black sigatoka - the major fungal disease of banana caused by Mycosphaerella fijiensis - averages between 0.3 and 1.0 US $ per plant in banana plantations in order to avoid yield losses of 30-50%. Furthermore, plants infected with the banana bunchy top virus may become completely unproductive. In the past 70 years the application of classical methods to breeding for disease resistance resulted in a limited success only, due to long generation times, high sterility and triploidy of most cultivated bananas. The integration of genetic engineering into breeding programmes may provide a powerful tool to overcome these limitations by Correspondence to: G. Volckaert

inducing specific genetic changes within a short period of time. In vitro culture in banana has been extensively used to quickly propagate vegetative clones of many genotypes (Vuylsteke and De Langhe 1985). Recently, a general method has been described for the establishment of regenerable e m b r y o g e n i c cell suspensions f r o m proliferating meristems (Dhed'a et al. 1991) and has been successfully applied to several genetically distant cultivars (Dhed'a 1992). This method has a significant advantage over alternatives using immature zygotic embryos (Escalant and Teisson 1989), since most edible bananas rarely set seed. Protoplasts have also been isolated from an embryogcnie cell suspension and plants have been regenerated through somatic embryogenesis at a high frequency (Panis et al. 1993)- Furthermore, the established embryogenie cell suspensions can be stored by cryoprescrvation without loss of regenerating ability (Panis et al. 1991). Direct DNA introduction by electroporation (Fromm et al. 1985) into viable and highly regenerative protoplasts provides an o p p o r t u n i t y f o r e f f i c i e n t genetic transformation of banana. The technique is effective for genetic transformation of a range of dicot (Riggs and Bates 1986; Lindsey and Jones 1989; Chupeau et al. 1989) and monoeot species (Toriyama et al. 1988; Huang and Dennis 1989). It has been utilized not only for the study of expression of chimaeric gene constructs at the transient l e v d (Fromm et al. 1985; Bates et al. 1988) but also to produce stable transformants (ShiUito et al. 1985; Fromm et al. 1986; Rhodes et al. 1988; Shimamoto et al. 1989). However, there are many variables affecting the efficiency of gene transfer, including capacitance and field strength; duration, shape, number and spacing of electrical pulses; buffer composition; temperature; concentration and form of DNA, etc. In the present work we investigated a number of parameters influencing transient expression of the E. coli uidA gene, a general reporter gene in plants coding for the [3-glucuronidase enzyme (Jefferson et al. 1986), after transformation of banana protoplasts by electroporation. The initial observations of this work have been recently described in a short account (Sagi et al. 1992). This is the first detailed report describing transformation with transient gene expression in banana.

263 Materials and methods Cell suspensions andplasmids. Embryogenic cell suspension lines ofcv. 'Bluggoe' (Musa spp. ABB group) deseribcd by Dhed'a et al. (1991) were maintained and subenltured weekly in MS medium (Murashige and Skoog 1962) supplemented with 5 IxM 2,4-D and 1 IxM zeatin. The cell suspensions were cultured for more than 1 year prior to use in the experiments and consisted of small dusters of isodiametrie, cytoplasmrich cells (Fig. 2a). The structure of the plasmids pBI221 (Clonteeh Laboratories, Inc.), pBI-426, and pBI-505 (kindly provided by William Crosby, Plant Biotechnology Institute, Saskatoon, Sask., Canada) used in this work is shown in Fig. 1. Plasmid DNA was purified using Qiagen columns.

pBi221

I 35S H

I-

uidA H nos-T

pB1-426

[:35S-55S(~ u"idA-neoH nos-T pBI-505 155S-35S(~"' uidA H

nes--i-

Fig. 1. Schematic representation of the chimaeric gene constructs used for electroporation experiments. 35S = CaMV 35S promoter, 35S-35S = tandem repeat CaMV 35S promoter, AMV = Alfalfa Mosaic Virus leader sequence, uidA = GUS reporter gene, uidA-neo = GUS-NPTII fusion gene, nos-T = NOS terminator

Protoplast isolation. Ten or 20 ml of I-week-old suspension cells were transferred to an equal volume of enzyme solution containing 1.0% eellulase, 1.0% macerozyme, 1.0% peetinase, 0.55 M mannitol and 3 mM MES, pH 5.8, for 20-24 hours at 25"C in the dark. The protoplasts were then filtered through 80 Ixm and 25 Itm mesh stainless steel screens and collected by eentrifugation at 90g. The pellet was washed in 30 ml of the protoplast isolation solution but without enzymes followed by a second wash in 20 ml of deetroporation buffer. Finally, the washed pellet was resuspended in eleetroporation buffer at a protoplast density of 106 800 ixi-1. The eleetroporation buffers are described in Table 1. This pr~oeol usually resulted in a protoplast yield of 2 x 106 ml-1 of cell suspension. Protoplasts were counted using a modified Neubauer haemoeytometer.

Complete removal of the cell wall was confirmed by Caleofhor white staining while viability of freshly isolated or eleetroporated protoplasts was controlled by staining either with fluorescein diacetate or Evans' blue. Electroporation. A 800 i~1 aliquot containing 106 protoplasts in elcctroporation buffer was placed into euvettes of 0.4 em gap and with ahminium electrodes glued onto opposite faces (Bio-Rad Laboratories). After addition of plasmid DNA to a concentration of 60 Ixg m1-1, euvettes were stored on ice for 10 min. Just before delivering the pulse, 40% PEG was added to a final concentration of 3-12% for I or 3-4 rain and cuvettes were carefi~lly vortexed to prevent sedimentation of protoplasts. Duration of PEG treatment means the time elapsed between addition of PEG and pulse delivery. The samples were then eleetroporated with a 960 ixF capacitor of a Bio-Rad Gene PulserTM transfeetion apparatus which generates an exponential decay pulse. The time constant of the pulse was monitored after each pulse delivery. After eleetroporation, the cuvettes were placed on ice for 10 rain and then for 10 min at room temperature. Protoplasts were diluted with the MS medium supplemented with 5 IxM 2,4-D, 1 IxM zeatin and 0.55 M mannitol to a density of 105m1-1 and incubated in the dark at 24"C. The following controls were used: (1) samples eleetroporated with pUC19 DNA, (2) samples electroporated without plasmid DNA, (3) nonelectroporated samples incubated with plasmid DNA. Transient ~-glucuronida~e expression assay. For the histochemical in situ assays, protoplasts were eollected at 24 or 4 8 hours after electroporation, resuspended in 50 mM sodium phosphate buffer, pH 7.0 and incubated from overnight up to 10 days at 37"C in the presence of 1 mM X-Gluc as described by Jefferson (1987). Transformation efficiency was assessed by counting blue stained protoplasts and relating to the total number of protoplasts. At least two internal repeats were counted and averaged for each treatment in each experiment. The number of blue-stained cells was counted on average in 105 protoplasts for each repeat. Total number of treated protoplasts was determined independently for each repeat. Since the number of blue protoplasts was significantly increasing after I0 days of incubation in comparison to 1 day incubation (Fig. 2b, c), for statistical evaluation data were obtained after a 10 day incubation in GUS-assay buffer. To avoid overestimation of GUS-expressing protoplasts due to the possible diffusion of the indigo product (Diaz and Carbonero 1992), each blue cluster of protoplasts was scored as a single transformant. Cultures of E. coli were used in parallel as positive controls for the GUS assay. Statistical evaluation was carried out after the arcsin transformation (3" = 2 aresin yl/2) which is necessary to stabilize the variances of the error terms for processing of data which are proportions. The ANOVA and Duncan's multiple range tests were performed using the statistical software package SAS (SAS Institute, Inc. Cary, North Carolina, USA). Where significant differences in the treatment means were found at the 5% probability level of an F-test, means were compared using Duncan's multiple range test at 5% level of significance.

Fig. 2. a cv. 'Bluggoe' banana cell suspension showing a cell aggregate after 3 days of subculture, b Transient GUS expression in banana protoplasts assayed 24 hours after eleetroporation in 5% PEG showing single blue protoplasm after 1 day incubation, c Transient GUS expression in banana protoplasts treated as in b but after 10 days of incubation, dTransient GUS expression in banana protoplasts treated as in b but electroporated without PEG. Bar = 100 ixm

264 Results and discussion

Effect of electroporation conditions on protoplast viability Viability of freshly isolated protoplasts was always over 90% with an average of 93%, as assessed by fluorescein diacetate staining or by the dye exclusion test with Evans' blue. Twenty-four hours after eleetroporation, viability of the control culture (no electroporation) in ASP-buffer (Tada et al. 1990) or in Cl-buffer (Fromm et al. 1985) was reduced from 93% to 75% and 67%, respectively. No clear-cut relation was found between viability and field strength when protoplasts were eleetroporated in CIbuffer. Viability changed irregularly and there was no effect of the capacitance used (500 or 960 pF). Therefore, we tested a chloride-free electroporation buffer containing organic salts as described by Tada et al. (1990). Using this ASP-buffer, protoplast viability dearly correlated with field strength as well as with the two capacitances. Already 4 hours after electroporation, a significant decrease in viability was observed at a field strength of between 700 and 900 V cm -1 (Fig. 3a) referring to a high frequency of irreversible permeation. The same field strength values were found to be critical at 24 hours (Fig. 3b) as wall as at 48 hours after electroporation (data not shown). The field strength optimum for m a x i m u m transient gene expression generally correlates with a protoplast viability of 5 0 % or lower (Hauptmann et al. 1987; Oard et al. 1989) with the exact value depending on the conditions of clcctroporation, the species and even the cell line used (Bekkaoui et al. 1990). Hence. for the banana protoplasts D N A uptake is assumed to be optimal for transient genc expression around a field strength of 800-900 V cm "I. This relatively high value may be due to a protective effect of the ASP-buffer and the small size of the protoplasts.

protoplasts of 15~-2 lxm (Boston et al. 1987) isolated from the W001C cell suspension, Fromm et al. (1985) found that at an estimated capacitance of 960 lxF~ 875 V cm -1 field strength was optimal for transient gene expression which is close to the values obtained here. To determine DNA concentrations for high transient gene expression, plasmid DNA was electroporated at concentrations of 12, 24, 48, and 60 ~g m1-1 and 60 Ixg ml "] resulted in the highest TGE while TGE at 12 and 24 ~tg ml-lwas highly variable (not shown). This value agreed well with that found by others for transient gene expression (Fromm et al. 1985; Sdguin and Lalonde 1988; Dhir et al. 1991) as well as for stable transformation (Fromm et al. 1986; Rhodes et al. 1988). I!

100

[] b b b b []

80

500 micr0F 960 micr0F

c

d

~,60

J~

e

4o

20 ,

|

,

,

400 500 600 700 800 900 1000 125 B [ ] 500 microF 100 ] a [ ] 960 microF a a a aab ab 80

Effect of field strength and DNA concentration on transient GUS expression In initial experiments the expression vector pBI221 was used to monitor gene transfer into banana protoplasts by electroporation at field strengths between 600 and 900 V cm 4 with 50 V cm -1 increments. At a capacitance of 500 IxF, TGE was below 10.4 with all field strengths tested. When a capacitance of 960 pF was used, a slight peak was observed between 750 and 850 V cm -1 at a TGE frequency of 2.5-3 x 10-4 (dam not shown). This range fits well to the critical range of f i d d strength found with viability experiments (Fig. 3). This range of field strength is quite high relative to what has been generally used under comparable conditions. For instance, Planckaert and Walbot (1989) found that at a capacitance of 1550 lxF the optimal field strength was 450 V cm -1 and 600 V cm -t for maize protoplasts isolated from suspension cells and for those isolated directly from callus, respectively. It should be noted, that the diameter of those protoplasts was 60 jam and 20-50 Inn, respectively. A c c o r d i n g to theoretical considerations o f electroporation, protoplast size is inversely related to the field strength causing permeation of membranes (Rouan et al. 1991). The average diameter of the banana protoplasts is 17+3 pm which may account for the relatively high optimal field strength. Indeed, in carrot

a

b

b

b

60 = Ii,

4o

o

400 500 600 700 800 900 10001125 Field strength (Vlcm) Fig. 3. Effect of field strength on viability of banana protoplasts a 4 hours and b 24 hours after electroporation.Columnsmarkedby the same letter are not significantly(P< 0.05) differentfor each capacitanceby Duncan's test after arcsin transformation. Standard deviation did not exceed:L5%and • of the meanfor 500 I~Fand 960 p.F,respectively.

Effect of polyethylene glycol on transient GUS expression In order to increase TGE, the effect of PEG was studied, initially with plasmids pBI221 and pBI-505. With both plasmids, the inclusion of 5% PEG before electroporation resulted in an increased TGE already after 1 day. TGE frequencies increased from 0.02% to 0.11% and from 0.01% to 0.07% for pBI221 and pBI-505, respectively.

265 Similar effects of PEG were detected 2 days after electroporation (Fig. 4). GUS expression was timedependent and assays carried out 2 days after electroporation resulted in higher TGE than those after 1 day. However, 3 days after electroporation, TGE was decreasing both in the presence and in the absence of PEG (not shown). Fig. 4 also demonstrates that without PEG only slight differences were found between the two plasmids tested. However, when 5% PEG was added and transformation frequencies increased, pBI-505 containing the tandem repeat 35S promoter and the AMV leader sequence (Fig. 1) gave significantly higher TGE than pBI221.

..ii,,

1~2 igl NI

A

,~ 0,4 IpI

The duration of PEG treatment has not been studied systematically. In PEG-transformations, Negrutiu et al. (1990) observed an optimal range of 2 to 5 rain incubation time both for transient expression and stable transformation. Combination of electroporation with PEG (Tyagi et al. 1989) resulted in higher stable transformation if a shorter 10 min, rather than a 30 rain PEG treatment, was applied. The present data suggest that duration of PEG treatment is a crucial factor in combination with electroporation, as well.

[] []

1,1r Ill I,_ L

- PEG + PEt

2d

ab bc

1 d~ I=

ld

9 0,2

< 1 rain 5-"I rain

X

~ 0,3 X

[] []

t,gl

I/i L

~0,1

I,i 0 0

I=

3

5

PEG concentration

~, 0,0 pBI221

pBI-505

pBI 221

pBI-505

Fig. 4. Effect of PEG on transient GUS expression in banana protoplasts electroporated at 800 V em -1 and assayed I day (I d) or 2 days (2 d) after electroporation by 10 days of incubation in X-Glue. Columns marked by the same letter are not significantly (pc 0.05) different by Duncan's test after aresin transformation. Percentages based on two replications of 105 protoplasts for each treatment. Standard deviation did not exceed +15% of the mean.

Due to its hydrophilic nature which causes adhesion of molecules and cell membranes followed by endocytosis, PEG is a generally used non-toxic agent for protoplast fusion (Kao and Michayluk 1974) and direct gene transfer to plant protoplasts (Paszkowski et al. 1984). Numerous studies in different species showed a beneficial effect of combining PEG with eleetroporation (Boston et al. 1987; Stguin and Lalonde 1988), while others found decreased transformation efficiencies with the addition of PEG (Tyagi et al. 1989). It is remarkable that transient gene expression was measured in most cases when positive effects were observed and for stable transformation usually negative effects of PEG addition were found. The effect of PEG concentration and incubation time on TGE before electroporation was also investigated using pBI-426. A higher PEG concentration resulted in an increased TGE provided the incubation time before electroporation was less than 1 min (Fig. 5). If, however, PEG was present for a longer time, 3-4 min in the solution, it had a negative effect on the overall TGE and more so at higher concentrations (Fig. 5). Combination of data according to time treatments clearly showed that the short PEG treatment was superior to the longer one. Although 8% PEG resulted in the highest TGE, at this concentration results were highly unreproducible and aggregation of the protoplasts was observed. Therefore, 5% PEG was selected for further use.

8

(X)

Fig. 5. Effectof incubationtime and concentration of PEG on transient GUS expression in banana protoplasts eleetroporated with pBI-426as assayed2 days aftereleetroporationby 10 daysof incubation in X-Glue. Columns marked by the same letter are not significantly (P< 0.05) different by Duncan'stest after aresin transformation. Combinedmeans for 3%, 5% and 8% PEG were 1.03%, 1.23% and 1.35%, respectively, and were not significantlydifferent from each other. Combinedmeans for time treatments were significantlydifferent with a value of 1.29% and 0.59% for the short time treatment and for the longer one, respectively. Standarddeviationdid not exceed+15%of the mean.

Effect of electroporation buffers and heat shock Table 1. Effect of electroporation buffers and heat shock (45"C, 5 rain) on transient GUS expression frequency (%) in banana protoplasts eleetroporated with pBI-426 in 5% PEG and assessed 48 hours after electroporation by 10 days of incubation in X-Glue Treatment a Electroporation buffer ASP1 CI2 Cytomix3

Mean a Control 1.082 b 0.734 b 0.002 c

Shocked 1.868 a 0.758 b 0.018 c

1.468 a 0.746 b 0.009 e

a Entries within these headings followed by the same letter are not significantly (P< 0.05) different by Duncan's test after arcsin transformation. Percentagesbased on two replications of 105protoplasts for eachtreatment. Standarddeviationdid not exceed+20% of the mean except for cytomixtreatments where it was 100%and 54%, respectively. 1_ 70 mM K-aspartate, 5 mM Ca-gluconate, 5raM MES, 0.55 M malmitol, pH 5.8 (Tada et al. 1990) 2. 150 mM NaC1,4 mM CaCI2, 10 mM HEPES, 0.55 M mannitol, pH 7.2 (Frommet al. 1985) 3_ 120 mM KC1,0.15 mM CaC12, 10 mM K2HPO4]KH2PO4,25 mM HEPES, 2 mM EGTA, 5 mM MgCI2, 2 mM ATP, 5 mM glntathione, pH 7.6 (van den Hoffel:al. 1992) The comparison of three different electroporation buffers resulted in significant differences in favour of the ASPbuffer in the presence of 5% PEG (Table 1). Similar

266 trends between electroporation buffers were observed when heat shock (45~ 5 min) was applied to protoplasts after t h e addition of D N A but before PEG. Moreover, heat s h o c k significantly increased T G E in the best performing ASP-buffer only. The ASP-buffer proved already better than Cl-buffer in our studies on the viability of electroporated protoplasts. This is attributed to the elimination of detrimental C12-gas production by substituting organic salts for mineral ones (Tada et al. 1990). Cytomix, a superior electroporation buffer for mammalian cells (van den Hoff et al. 1992) was found to be unsatisfactory, again probably because of the formation of C12-gas. The effect o f heat shock before electroporation is unclear: it might be species-specific and/or dependent on the conditions used. Tautorus et al. (1989) found in electroporated conifer protoplasts that heat shock had a positive effect on transient gene expression, while in t o b a c c o protoplasts a deleterious effect on stable t r a n s f o r m a t i o n was o b s e r v e d (Tyagi et al. 1989). However, there are only a few reports in which the effect o f heat shock before electroporation was studied in combination with P E G treatments. Stguin and Lalonde (1988) found that heat shock had no positive effect in the absence o f P E G but it was synergistic with PEG in alder. Similar results were obtained by Dhir et al. (1991) in s o y b e a n protoplasts. H e a t shock in electroporation appears to be beneficial mainly in combination with PEG. In fact, heat shock for plant transformation was originally used in combination with PEG-mediated transformation (Potrykus et al. 1985) and in the PEG-electroporation procedure (Shillito et al. 1985).

Conclusions This work optimized some electroporation conditions suitable for transient expression of the introduced GUS gene in b a n a n a protoplasts. W h e n using a 960 lxF c a p a c i t o r , the e s t a b l i s h e d c o n d i t i o n s are: (i) electroporation buffer, ASP-buffer, containing 70 m M Kaspartate, 5 m M Ca-gluconate, 5 m M MES, and 0.55 M mannitol (pH 5.8); (ii) electric field strength, 800 V cm-1; (iii) P E G concentration, 5% applied within 1 min before electroporation; (iv) heat shock, 45~ for 5 min before addition of PEG. Since the in situ GUS assay has been infrequently used to estimate the frequency of protoplasts showing Iransient gene expression, the data with which the results presented here could be compared are limited. With PEG-mediated transformation and also using circular plasmids, Zhang and W u (1988) reported that up to 0.05% of rice protoplasts showed T G E at 3 days after gene transfer. A similar 0.05% efficiency o f gene transfer can be calculated from the results given by Diaz and Carbonero (1992). B y using a similar procedure to the one described in this study, Dhir et al. (1991) reported that a maximum of 1.55% of soybean protoplasts showed TGE. Bower and Birch (1990), however, succeeded in achieving 20% T G E in carrot protoplasts. It is expected that further increases in transformation frequency m a y be achieved in banana with the use o f multiple electric pulses, linearised plasmid D N A and carrier D N A though conditions required for transient or stable transformation might be different in field strength, form and concentration of DNA. However, the present method provides a basis to screen various constructs for

expression in banana and together with the plant regeneration procedure from protoplasts m a y contribute to the selection of stable transformants in banana.

Acknowledgements. The authors are grateful to Dr. William Crosby

(Plant BioteclmologyInstitute, Sask., Canada) for the plasmids pBI-426 and pBI-505. We thank Dr. Djailo Dhed'a for the embryogenic cell suspensions and Ines Van den houwe, Ann Janssen, and Willem Dillemans for valuable technical assistance. This work was financed by contract ETC-007 of the "VlaamseAktieprogrammaBiotechnologie"of the Flemish Ministry of Economy and by the International Network for the Improvementof Bananas and Plantains (INIBAP).

References Bates GW, Piastneh W, Riggs CD, Rabnssay D (1988) Plant Cell Tiss Org Cult 12:213-218 Bekkaoui F, Datla RSS, Pilon M, Tautoms TE, Crosby WL, Dunstan DI (1990) Theor Appl Genet 79:353-359 Boston RS, Becwar MR, Ryan RD, Goldsbrough PB, Larkins BA, Hodges TK (1987) Plant Physio183:742-746 Bower R, Birch RG (1990) Plant Cell Rep ~.386-389 Chupeau M-C, Bellini C, Guerche P, Maisonneuve B, Vastra G, Chupeau Y (1989) Bio/Technology7:503-508 Dhed'a D (1992) Ph.D. thesis. Faculty of Agricultural Sciences, Catholic Universityof Leuven, Belgium, 167 pp. Dhed'a D, Dumortier F, Panis B, Vuylsteke D, De Langhe E (1991) Fruits 46:125-135 Dhir SK, Dhir S, Hepburn A, Widholm JM (1991) Plant Cell Rep 10:106-110 Diaz I, CarboneroP (1992) Plant Cell Rep 10:595-598 Esealant JV, Teisson C (1989) Plant Cell Rep 7:665-668 FAO (1992) Production yearbook 1991. Food and Agriculture Organisation of the [Inked Nations. Rome, Italy, 265 pp. Fromm M, Taylor LP, Walbot V (1985) Proe Natl Aead Sci USA 82:5824-5828 Fromm ME, Taylor LP, Walbot V (1986) Nature 319:.791-793 Hauptmann RM, Ozias-Akins P, Vasil V, Tabaeizadeh Z, Rogers SG, Horsch RB, Vasil IK, Fraley RT (1987) Plant Cell Rep 6:265-270 Huang Y-W, Dennis ES (1989) Plant Cell Tiss Org Cult 18:281-296 Jefferson RA (1987) Plant Mol Biol Rep 5:387-405 Jefferson RA, Burgess SM, Hirsh D (1986) Proe Natl Aead Sci USA 83:8447-8451 Kao KN, MichaylukMR (1974) Planta 115:355-367 Lindsey K, Jones MGK (1989) Plant Cell Pep 8:71-74 Murashige T, Skoog F (1962) Physiol Plant 15:473-497 Negmtiu I, Dewulf J, Pietrzak M, Botterman J, Rietveld E, WurzerFigurelli EM, Ye D, Jacobs M (1990) Physiol Plant 79:197-205 Oard JH, Paige D, Dvorak J (1989) Plant Cell Rep 8:156-160 Panis B, Dhed'a D, Swennen R (1991) In: Adams RP, Adams JE (eds) Conservation of plant genes: DNA banking and in vitro biotechnology.Academic Press, New York, pp 183-195 Panis B, Van Wauwe A, Swennen R (1993) Plant Cell Rep 12:403-407 Paszkowski J, Shillito RD, Saul M, Mandak V, Hohn T, Hohn B, Potrykus I (1984) EMBO J 3:2717-2722 Planekaert F, Walbot V (1989) Plant Cell Rep 8:144-147 Potrykus I, Saul MW, Petmska J, PaszkowskiJ, Shillito RD (1985) Mol Gen Genet 199:183-188 Rhodes CA, Pierce DA, Mettler IJ, Masearenhas D, Detmer JJ (1988) Science 240:204-207 Riggs CD, Bates GW (1986) Proc Natl Acad Sci USA 83:5602-5606 Rouan D, Montan6 M-H, Alibert G, Teissi6 J (1991) Plant Cell Rep 10:139-143 Sagi L, Remy S, Panis B, Swennen R, Volckaert G (1992) Banana Newsletter 15:42 S~guin A, Lalonde M (1988) Plant Cell Rep 7:367-370 Shillito RD, Saul MW, Paszkowski J, Muller M, Potrykus I (1985) Bio/Technology3:1099-1103 Shimamoto K, Terada R, Izawa T, Fujimoto H (1989) Nature 338:274276 Tada Y, SakamotoM, Fujimura T (1990) Theor AppI Genet 80:475-480 Tautoms TE, Bekkaoui F, Pilon M, Datla RSS, Crosby WL, Fowke LC, Dunstan DI (1989) Theor Appl Genet 78:531-536 Toriyama K, Arimoto Y, Uehimiya H, Hinata K (1988) Bio/Technology 6:1072-1074 Tyagi S, SpoerleinB, Tyagi AK, Herrmann RG, Koop HU (1989) Theor Appl Genet 78:287-292 van den Hoff MJB, Moorman AFM, Lamers WH (1992) NucleicAcids Res 20:2902 Vuylsteke D, De Langhe E (1985) Tropical Agriculture (Trinidad) 62:323-328 Zhang W, Wu R (1988) Theor Appl Genet 76:835-840