Abstract A highly efficient system for the production oftransgenic maize plants starting from tropical and sub-tropical genotypes was developed. The method is basedon particle bombardment of organogenic calli derivedfrom shoot tips. Six tropical maize genotypes were successfully transformed and regenerated using this pro-tocol. Genetic transformation was confirmed by Southernblot analysis of T0 plants and segregation analysis of theresistance marker in the T1 progeny. Plant transfer intothe greenhouse was 100% successful, and no problemsof fertility were observed with the transgenic plants pro-duced with this transformation protocol.

Keywords Cereal transformation · Organogenesis ·Plant regeneration · Shoot meristem · Transgenic maize

Abbreviations Ade: Adenine · BA: N6-Benzyladenine ·bar: Phosphinothricin acetyl transferase gene · 2,4-D: 2,4-Dichlorophenoxyacetic acid · FW: Freshweight · GUS: β-Glucuronidase · Inbred lines LPC: LíneasPacífico Centro · MPC medium: MS basal medium supplemented with 2 mg/l BA, 1 mg/l 2,4-D, and40 mg/l Ade · PPT: 2-Amino-(methylphosphinyl)-butanoic acid (phosphinothricin) · uidA: β-Glucuronidasegene · WT: Untransformed genotype · ZHM: MS basalmedium supplemented with 2 mg/l BA and 1 mg/l 2,4-D

Introduction

Maize is one of the three most important cereals on aworld production level. Over the last few decades, theannual production tonnages for maize have been similarto those for wheat and rice, and far above the productionnumbers for other amilaceous plants (FAO databasewww.fao.org). Millions of people in the tropical and subtropical zones of the world depend on maize for theirsubsistence. In these areas, crop productivity is oftenlow, due to abiotic stresses, such as drought, aluminiumtoxicity, or scarcity of nutrients, and to biotic stresses,such as pests and diseases. The availability of the mod-ern methodologies of plant genetic engineering holds outthe promise that plants tolerant to these stresses can beproduced, and that the improvements can be achieved ina timely and purposeful manner. Genetic engineering hasalready been successfully used in maize to produce in-sect- and virus-resistant lines (Koziel et al. 1993; Murryet al. 1993) and male-sterile plants (Unger et al. 2001).

An essential requirement for the generation of trans-genic plants is the availability of an efficient plant regen-eration system. Until now, plant regeneration from em-bryogenic type II calli derived from immature embryoshas been the most reproducible transformation/regenera-tion system described for maize (Fromm et al. 1990;Gordon-Kamm et al. 1990; Walters et al. 1992; Pescitelli

Communicated by L. Peña

A. O’Connor-Sánchez · J.L. Cabrera-Ponce · M. Valdez-MelaraP. Téllez-Rodríguez · J.L. Pons-Hernández · L. Herrera-Estrella (✉)CINVESTAV/IPN, Departamento de Ingeniería Genética de Plantas Km 9.6 del Libramiento Norte de la carretera Irapuato-León,Apdo. Postal 629, 36500 Irapuato, Gto., Mexicoe-mail: lherrera@ira.cinvestav.mxTel.: +52-462-39602, Fax: +52-462-45846

Present address:M. Valdez-Melara,Escuela de Biología y Centro de Investigación en Biología Celular y Molecular, Universidad de Costa Rica,San José, Costa Rica

Present address:P. Téllez-Rodríguez,Centro de Ingeniería Genética y Biotecnología, P.O. Box 6162,10 600 La Habana, Cuba

Present address:J.L. Pons-Hernández,Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias, Apdo. Postal 112, Celaya, Gto., México

Plant Cell Rep (2002) 21:302–312DOI 10.1007/s00299-002-0502-8

GENETIC TRANSFORMATION AND HYBRIDIZATION

A. O’Connor-Sánchez · J. L. Cabrera-PonceM. Valdez-Melara · P. Téllez-RodríguezJ. L. Pons-Hernández · L. Herrera-Estrella

Transgenic maize plants of tropical and subtropical genotypes obtained from calluses containing organogenic and embryogenic-like structures derived from shoot tips

Received: 7 February 2002 / Revised: 13 June 2002 / Accepted: 13 June 2002 / Published online: 25 September 2002© Springer-Verlag 2002

303

and Sukhapinda 1995; Russell and Fromm 1996; van derGeest and Petolino 1998; Sidorenko et al. 2000; Srivastavaand Ow 2001). However, type II calli can only be gener-ated starting from a small number of maize genotypes,such as the A-188 line (Armstrong and Green 1985),possibly because the capacity to produce them is undergenetic control (Armstrong et al. 1992).

Introgression of this trait into continuously evolvingelite genotypes by backcrossing is impractical and be-yond the reach of most breeding efforts. A differentsystem, explored by Lowe et al. (1995), involves thebombardment of shoot meristems of immature maizeembryos to produce transgenic sectors from which trans-genic seed can be obtained. However, the efficiency ofgenetic transformation achieved with this system wasvery low. More recently, Zhong et al. (1996) developedan alternative system for plant transformation/regenera-tion by using shoot meristems as initial explants. Thissystem was reported to be generally applicable to tem-perate maize genotypes. Most efforts on maize plant regeneration have used genotypes adapted to temperatezones (Vasil et al. 1984; Duncan et al. 1985) and little attention has been given to the regeneration potential ofmaize germplasm adapted to tropical regions. Prioli andSilva (1989) and Carvalho et al. (1998) described the regeneration of a number of tropical inbred lines startingfrom immature embryos, but until now they have not reported their use to produce transgenic maize plants. Inthis paper, we report the development of a simple andhighly efficient method to produce transgenic tropicaland subtropical maize plants, via particle bombardmentof organogenic/embryogenic-like calluses obtained fromshoot tips.

Materials and methods

Plant material

Mature seeds from six Mexican subtropical genotypes [four inbredlines, LPC13, LPC16, LPC18, and LPC21 (Ramirez-Diaz et al.1995); and two hybrid genotypes, ZM-1 and ZM-7], one Costa Rican inbred line (CR-5), and two tropical varieties from Cuba(FR28-1 and FR28-2), were respectively supplied by the InstitutoNacional de Investigaciones Forestales, Agrícolas y Pecuarias ofMexico, the Oficina Nacional de Semillas of Costa Rica, and theOficina Nacional de Semillas of Cuba. Self-pollinated plantsgrown in 1996–1998 were the source of mature seeds used forplant tissue culture experiments.

Induction of morphogenic calli and plant regeneration

Mature seeds were surface-sterilized in 70% ethanol for 3 min,then in 5% NaClO solution containing 40 µl/100 ml Tween 20 for30 min, and rinsed 6 times in sterile distilled water. Disinfectedseeds were soaked in bidistilled water for 24 h to facilitate the removal of the pericarp before culture. Removal of the pericarpwas done under a dissecting stereomicroscope, using fine scalpelsand forceps. To facilitate germination, seeds were placed with theembryo side up. Genotypes LPC13, ZM1, ZM7, and CR5 wereused to test six culture media: MS basal media (Murashige andSkoog 1962) supplemented with three growth regulators usedalone or in combination: ZHM, 2 mg/l BA+1 mg/l 2,4-D (Zhong

et al. 1992); M1, 1 mg/l 2,4-D; M2, 2 mg/l BA; M3, 1 mg/l 2,4-D+40 mg/l adenine (Ade); M4, 2 mg/l BA+40 mg/l Ade; andMPC, 2 mg/l BA+1 mg/l 2,4-D+40 mg/l Ade. All media were supplemented with 3% (w/v) sucrose and solidified with 2.5 g/lGelrite. The pH of all media was adjusted to 5.8 before autoclav-ing. Fifty seeds of each genotype were germinated on each of thesix different media to determine the efficiency of morphogeniccallus formation. After 1 week of culture (Fig. 1A), 15- to 17-mm-long sections of the mesocotyl-coleoptile, containing the shoot tipin the centre, were excised and subcultured on fresh medium (of the same formulation as before). Cultures were maintainedwith a 16-h photoperiod (50 µmol/m2 per second, provided bycool-white and day light Sylvania fluorescent lamps) at 27°C.Four weeks later, calluses arising from the explants were isolated,fragmented and subcultured again on the same induction medium.One to 2 months later (with 3- to 4-week-old subcultures), well-established morphogenetic regenerable organogenic calluses weremoved into darkness to induce the formation of embryogenic-likestructures. For plant regeneration, calluses were cultured in Magenta boxes, on MS basal medium without growth regulators,with a 16-h photoperiod for 6 weeks. Rooted plantlets were trans-planted to soil, and cultivated under greenhouse conditions.

Growth rates of callus lines

Organogenic/embryogenic-like calluses from the inbred linesLPC13, LPC21, and FR28-1 were chosen to study the growth ratesof the cultures. 250±20 mg of calluses (derived from one originalexplant) from lines LPC13 and LPC21, and 340±35 mg fromFR28-1 were plated on MPC medium, and similar amounts onZHM medium (with ten replicates each). They were cultured indark conditions, aseptically weighed, and subcultured in fresh medium once a week. Data from callus lines and from culture mediawere statistically analysed (by ANOVA) to compare growth rates.

DNA constructs

pBARGUS (Fromm et al. 1990) was used for all transformationexperiments. Thus, the bar gene, expressed from the CaMV 35Spromoter, was used as selectable marker for stable transformation,and the uid A gene, expressed from the maize alcohol dehydroge-nase (Adh1) promoter, as reporter gene. For particle bombardmentexperiments, plasmid DNA was amplified in Escherichia coli, iso-lated by alkaline lysis, and purified by CsCl/ethidium bromidedensity centrifugation. After ethanol precipitation, purified DNAwas resuspended in TE buffer (1 mM TRIS-HCl pH 7.8, 0.1 mMNa2EDTA) and adjusted to 1 µg/µl to prepare particles for bom-bardment experiments.

Microprojectile bombardment

Plasmid DNA was adsorbed onto tungsten particles M10 of an average size of 0.73 µm (Dupont Biolistic Particle Delivery Systems, 1991) as described by Tomes et al. (1995). The DNA-coated microprojectiles were rinsed in 250 µl absolute ethanol. Aliquots of 10 µl (1.66 µg DNA associated with 125 µg tungsten)were loaded on each macrocarrier, air-dried to remove ethanol,and bombarded into organogenic/embryogenic-like calli using thehelium-driven version of PDS-1000 (Sanford et al. 1991). The distance between the rupture membrane and the flying disk was1.2 cm. The macrocarrier travelled another 1.2 cm before impact-ing a steel stopping screen. Well-developed calli cultured in thedark were screened under a dissecting stereomicroscope to selectcell clumps which contained both organogenic and embryogenic-like structures, for bombardment experiments. About 250 mg FWof organogenic/embryogenic-like calluses were located within a 3-cm circle in the centre of a Petri dish filled with solid MPC me-dium, 7 cm away from the launch point. Callus was bombardedonce at 1,200 psi. The sample chamber was evacuated to 0.04 atmbefore the gas acceleration tube was pressurized with helium.

304

GUS expression

To test for GUS expression, samples were immersed for 24 h in anX-GLUC solution at 37°C, as described by McCabe et al. (1988).Three days after bombardment, five samples per bombarded Petridish were tested for GUS transient expression. The number of bluefoci per sample was scored using a dissecting stereomicroscope.

Growth and selection of putative transgenic clones

Twenty-four hours after bombardment, organogenic/embryogenic-like calluses were spread on fresh MPC medium and incubated for2 weeks in the dark at 26°C. Then, calluses were subcultured onselective MPC medium containing 4 mg/l PPT or Bialaphos

(MPCPPT). To ensure effective selection, explants were subcul-tured at 3-week intervals on fresh MPCPPT medium (with carefulmanipulation to prevent formation of siblings), until growth ofmost cells stopped and putative transgenic calluses proliferated(2–3 months). The latter were then subcultured again on MPCPPTand transferred to light conditions to induce organogenic struc-tures.

Osmotic treatment

Three media were compared for their ability to improve the efficiency of transient and stable callus transformation: MPC me-dium (without additional sugar), MPC-0.4 M mannitol (MPCman),and MPC-0.4 M sucrose (MPCsuc). Calluses were transferred to

Fig. 1A–I Induction of organo-genic calli, and organogen-ic/embryogenic-like calli. Selection of putative transgeniccalli. GUS transient and stableexpression in transgenic calli.A Mature seeds of maize cul-tured on MPC medium in light.B Primary organogenic callusderived from the meristemic ar-ea of the maize seedlings, after2 months of culture on MPCmedium in light. C Organogen-ic callus derived from the pri-mary callus, and propagated inthe same conditions as above.D Organogenic/embryogenic-like callus derived from orga-nogenic calli that were culturedin darkness, after 1 month onMPC medium. E Transientgene expression of β-glucuron-idase in organogenic/embryo-genic-like calli, 3 days afterbombardment. F Organogen-ic/embryogenic-like calli12 weeks after bombardment,selected on MPC medium sup-plemented with PPT, in thedark. Calli derived from a bom-bardment experiment (left); un-bombarded (control) calli(right). G Organogenic/embry-ogenic-like calli 14 weeks afterbombardment, selected onMPC medium supplementedwith PPT in light. H Unbom-barded (control) calli in thesame conditions as above. I GUS stable expression intransgenic calli, 14 weeks afterbombardment. Bar represents5 mm

305

each media 24 h before bombardment and maintained in the samedish for 24 h after bombardment. Callus (250 mg FW) of inbredline LPC13 was used for each treatment (with 22 replicates). Theexperiment was repeated 10 times. Blue (GUS expressing) foci persample were counted to evaluate transient expression.

Herbicide assay

Transgenic and control plants (15 cm in height) in soil weresprayed with a 0.5% (v/v) solution of Basta to identify herbicide-resistant lines. To analyse the inheritance of bar gene, 30 T1 seedsfrom each of two independently transformed T0 plants weresprayed, and their survival was scored.

Southern blot hybridization

Total genomic DNA was isolated from leaf tissue of putativetransgenic and wild-type plants as described by Shure et al.(1983). Fifty micrograms of this genomic DNA was digested withBamHI and HindIII endonucleases for bar gene detection, andwith BamHI/BclI for uidA gene detection, electrophoresed in 1.0%agarose, and transferred to positively charged nylon membranes(Hybond-N+, Amersham) using 2×SSC as described by Sambrooket al. (1989). The filter was prehybridized for 24 h at 65°C in2×SCP, Blotto, 0.5% BSA, and hybridized overnight at 65°C(Sambrook et al. 1989). A 1-Kb BamHI/HindIII fragment containingall of the bar gene-nos terminator region from pBARGUS, and a1.8-Kb BamHI/BclI fragment containing all of the uidA gene frompBARGUS were labelled by random priming reactions (Feinbergand Vogelstein 1983; Boehringer Mannheim) to generate labelledprobes for bar gene and uidA gene detection respectively. Filterswere washed in 0.5×SSC, 0.5% SDS at 65°C for 1 h and visual-ized by autoradiography using X-ray film (Dupont, Cronex).

Results

Induction of organogenic/embryogenic-like callus

Six media were tested for their ability to induce organo-genic callus from shoot tips of four maize genotypes (as detailed in Materials and methods). Out of the sixmedia tested, only MPC medium (MS containing 2 mg/lBA, 1 mg/l 2,4-D and 40 mg/l Ade) proved to be effectivein inducing calluses with morphogenetic capacity (Table 1). Media containing only one or two out of thethree growth regulators did not induce callus formation,including the medium reported by Zhong (Zhong et al.1992). ZM-7 shoot tips cultured on M3 (MS containing1 mg/l 2,4-D and 40 mg/l Ade) produced a high prolifer-ation of roots and, at a low frequency (8.0%), non-morphogenic callus. Therefore, only MPC medium wasused to evaluate the remainder of the genotypes. On thismedium, shoot tips from all the tested maize genotypesproduced organogenic primary callus in 4 weeks (Table 2). This callus could be obtained equally well either by excision of the shoot tip as described above, orsimply by trimming the root and leaf tips off the seedlingbefore subculture (Fig. 1B). The primary calluses havemultiple shoot meristems (Fig. 1C), and display a highgrowth rate that is maintained during successive subcul-tures. In the tested genotypes, the percentage of shoottips producing organogenic callus was genotype-depen-

dent, ranging from 40% (in LPC18) to 100% (in LPC16)(Table 2). The highest frequencies were shown by inbredlines LPC16 (100%), LPC 13 (92.5%), and CR-5(82.0%) (Table 2). When 2-month-old organogenic calliwere cultured in the dark, most of them developed somatic embryo-like structures in some regions (Fig. 1D).For this reason, they were termed organogenic/embryo-genic-like calluses. One month after these calluses werecultivated in the dark, it was possible to observe, by simplevisual inspection, that calluses which contained both or-ganogenic structures and embryo-like structures withinthe same piece had a higher growth rate than callusescontaining only one of these types of structures. At thisstage, calluses can be used either to be propagated to create a stock, or to be bombarded.

A high rate of cell division is an important require-ment for a successful transformation. As has been report-ed by Hazel et al. (1998), the cell lines most amenable totransformation also exhibit the highest mitotic index. Totest whether the media formulation influenced growthrate, organogenic/embryogenic-like calluses from theLPC13, LPC21, and FR28-1 genotypes were grown inMPC and ZHM media. The three genotypes that wereevaluated display a statistically significantly highergrowth rate when cultured on MPC medium compared toZHM medium: after 4 weeks, on MPC medium all celllines had about twice the FW, compared to that obtainedon ZHM. This growth rate was found to be similar tothat observed for the highly embryogenic calli type IIfrom line A-188 (data not shown).

Table 1 Percentage of shoot tips producing primary (organogenic)callus as a function of the genotype and the culture medium

Maize genotype Culture medium % Of shoot tips producing callus

LPC13 MS-BA-2,4-D (ZHM) 0MS-2,4-D (M1) 0MS-BA (M2) 0MS-2,4-D-AD (M3) 0MS-BA-AD (M4) 0MS-BA-2,4D-AD (MPC) 78

ZM1 MS-BA-2,4-D (ZHM) 0MS-2,4-D (M1) 0MS-BA (M2) 0MS-2,4-D-AD (M3) 0MS-BA-AD (M4) 0MS-BA-2,4-D-AD (MPC) 44

ZM7 MS-BA-2,4-D (ZHM) 0MS-2,4-D (M1) 0MS-BA (M2) 0MS-2,4-D-AD (M3) 8MS-BA-AD (M4) 0MS-BA-2,4-D-AD (MPC) 54

CR-5 MS-BA-2,4-D (ZHM) 0MS-2,4-D (M1) 0MS-BA (M2) 0MS-2,4-D-AD (M3) 0MS-BA-AD (M4) 0MS-BA-2,4-D-AD (MPC) 82

306

Plant regeneration

When organogenic/embryogenic-like calluses from thenine genotypes used in this study were challenged in regeneration conditions, it was observed that all of themwere able to regenerate plants after 6 weeks. Usually, aprofuse root system develops first (after 3 weeks), andlater (after another 3 weeks) the shoots are generated.The efficiency ranged from 39 (in FR28-2) to 170 (in LPC16) regenerated plants per gram of FW calli (Table 2).

Maintenance of the morphogenic capacity of maize callus lines

To determine whether the calluses derived from shoottips could be subcultured without loosing regenerationcapacity, freshly induced organogenic/embryogenic-likecalluses from the nine genotypes used in this study werepropagated on MPC in the dark (with 3–4 weeks subcul-tures). About 1 g of calluses from each genotype (withthree replicates) was challenged in regeneration condi-tions for 6 weeks, every 4 months, during 1 year. It wasfound that, at least for a 1-year period, over 90% of thecalli produced shoot clumps, from which fertile plantscan be regenerated (data not shown). Therefore, usingthis regeneration protocol, callus lines from all the testedgenotypes can be propagated as stock cultures, at leastfor this length of time.

Effect of osmotic treatments on the efficiency to produce putative transformed clones

To determine the effect of some osmotic treatments onthe efficiency to produce putative transformed (PPT-resistant) clones, organogenic/embryogenic-like callusesderived from LPC13 were cultured on MPC media, or onMPC media containing either 0.4 M mannitol or 0.4 Msucrose, for 24 h before and after bombardment. Whentransient expression was evaluated by counting the num-ber of blue foci (Fig. 1E), no significant difference wasfound between the two osmotic treatments and MPC me-dium: the medium containing 0.4 M mannitol (MPCman)resulted in a mean of 5.3 blue spots per 10 mg of FWcallus, compared with four on MPCsuc and three bluespots on MPC. Nevertheless, supplementing the mediumwith 0.4 M sucrose (MPCsuc) appeared to substantiallyimprove the overall efficiency of stable transformation:about 4 times more PPT-resistant clones could be gener-ated using MPCsuc, compared to the results obtained using MPCman or unsupplemented MPC medium (Table 3). As can be seen in Table 3, the number of PPT-resistant clones obtained using MPCsuc was about17 clones/g FW bombarded callus.

Selection of putative transgenic clones

To select putative transgenic clones, a two-step selectionscheme was followed: (1) selection in the dark for9 weeks, during which some rapidly growing embryo-genic calluses can be observed (Fig. 1F), and (2) a 3-week selection in light, during which some of these

Maize genotype No. of shoot No. of shoot tips No. of regenerated plants tips tested producing organogenic callus per gram of organogenic/

embryogenic-like callus FW tissue

1. LPC13 400 370 (92%) 1502. LPC16 350 350 (100%) 1703. LPC18 300 120 (40%) 1504. LPC21 300 150 (50%) 1005. ZM-1 50 22 (44%) 506. ZM-7 50 27 (54%) 757. CR5 50 41 (82%) 458. FR28-1 400 300 (75%) 459. FR28-2 400 230 (57%) 39

Table 2 Induction of organo-genic callus as a function of thegenotype, and plant regenera-tion per gram of organogen-ic/embryogenic-like callus FWtissue in different subtropicaland tropical genotypes of maizecultured on MPC medium

Medium Total FW gram Total No. of PPT- No. of PPT-resistant of bombarded callus resistant clones clones per gram of FW

bombarded callus

MPC medium 55 223 4.05MPC medium+mannitol 0.4 M 55 230 4.18MPC medium+sucrose 0.4 M 55 940 17.10

Table 3 Effect of osmotictreatments on the production of PPT-resistant clones inmaize inbred line LPC13

307

calluses turn green and the developmental process shiftsto organogenesis (Fig. 1G). During the first step, differ-ent sizes of PPT-resistant calluses are formed, and sus-ceptible ones stop growing and become necrotic. Afterthe second step, PPT-resistant calluses turn green, whilenon-resistant ones (escapes from the first step) remainwhitish. When calluses were tested after the first step forstable expression of GUS, about 60% were positive.However, when green clones were tested after the secondstep, around 80% were found to be GUS positive (datanot shown). As can be seen in Table 4, the number ofgreen PPT-resistant clones per gram of bombarded callusis variable, depending on the maize genotype, rangingfrom 9.2 (in LPC18) to 20.1 (in FR28-2). Nevertheless,it is important to note that all the tested genotypes wereable to produce putative transgenic clones using thismethod.

Plant regeneration

When green PPT-resistant calluses were transferred toplant regeneration conditions (on MS medium withoutPPT to speed up the process) it was observed that almostevery one was able to regenerate at least one full plantafter 6 weeks (and as many as desired, if the clone ispropagated before regeneration). One hundred percent ofthe regenerated plants had roots (Fig. 2A) and were suc-cessfully adapted to greenhouse conditions.

DNA analysis

Several randomly chosen PPT-resistant T0 lines of geno-types LPC13, LPC16, FR28-1, FR28-2, ZM1, and ZM7were analysed by Southern blot hybridization to confirmthe presence of the BamHI-HindIII 1-Kb fragment containing the bar gene in their genome. In the majorityof the T0 plants (22 out of 29) from all the tested geno-types, a 1-Kb positive hybridization signal was observedwith a probe synthesized from the pBARGUS region

containing the bar gene-nos terminator construct(Fig. 3). The presence of positive hybridization signalsof the predicted size shows that the plants contain atleast one intact copy of the expected fragment. In fivecases the 1-Kb band was not clearly detected (lines 1, 4,and 7 of genotype LPC13, line 2 of genotype FR28-1,and line 4 of genotype FR28-2), but higher molecularweight bands are present. Only in two T0 plants no clearhybridization bands were detected (line 3 of genotypeLPC16 and line 3 of genotype ZM1), suggesting thatthey were “escapees” of the selection protocol. The signal intensity for the 1-Kb band showed differencesfrom case to case (for example, in samples 1 and 7 ofgenotype LPC13 and samples 1, 2, and 4 of genotypeZM1 it is very weak) possibly indicating a highly variable copy number. With the exception of genotypesZM1 and ZM7, several bands were detected with molec-ular weights different from that expected for theBamHI/HindIII fragment containing the bar gene-nosterminator, indicating either that multiple independentinsertions occurred, or that the integrated fragments arelong tandem repeats resulting from rearrangements, andthat the frequency of this event seems to be genotype dependent. All the transgenic clones tested of genotypesLPC13, LPC16, FR28-1, and FR18-2 displayed uniquehybridization patterns, indicating that these transgenicplants were derived from independent transformationevents.

To test for cotransformation of the uidA gene and thebar gene, the same eight plants of genotype LPC13 usedto analyse transformation of the bar gene, were used toconfirm the presence of the 1.8-Kb BamHI/BclI fragmentof the uidA gene in their genome. This analysis (Fig. 4)showed for most of the cases (samples 1, 2, 4, 7, 8 andpossibly 5) a positive hybridization signal for the expect-ed 1.8-Kb BamHI/BclI fragment containing the uidAgene, indicating the presence of at least one intact copyof this fragment in the genome of the transgenic plants.Again, in this analysis, a unique uidA hybridization pattern was observed for each sample.

Table 4 PPT-resistant clones derived from six maize genotypes, with osmotic treatment on MPCsuc. No. of Southern blot-positivelines. NT Not tested

Genotype No. of bombardment No. of bombarded No. of PPT- No. of PPT-resistant No. of lines tested experiments plates per experimenta resistant clones clones per gram of FW by Southern blot/no.

bombarded callus of positive lines

LPC13 10 22 940 17.1 8/8LPC16 3 5 52 13.8 4/3LPC18 9 5 103 9.2 NTLPC21 9 5 215 19.2 NTZM1 3 5 NT NT 4/3ZM7 3 5 NT NT 4/3CR5 3 5 NT NT NTFR28-1 9 10 349 15.5 4/4FR28-2 4 15 301 20.1 5/5

a Each plate contained 250 mg FW callus

308

Phenotype of T0 plants and inheritance of the BAR gene

Ninety-five percent of the T0 plants established in soilconditions reached maturity (Fig. 2C). Most of them produced one normal tassel (Fig. 2D) and from one tothree fertile ears. About 30% of them showed typicalcharacteristics for tissue culture-induced stress, such assmall amounts of pollen, and some other morphologicalalteration, such as reduced stature compared with seed-derived plants and/or pistillate flowers on tassels(Fig. 2E). Similar morphological abnormalities in re-generants of maize have been previously reported usingdifferent transformation methods (Rhodes et al. 1988;Gordon-Kamm et al. 1990; D’Halluin et al. 1992; Walters et al. 1992; Zhong et al. 1996).

Thirty T1 putative transgenic plants derived from oneline of genotype LPC13, and 30 others derived from oneline of genotype LPC18, as well as 30 WT control plantsfrom each of the same genotypes, were sprayed withBasta solution to test functional expression of the bargene by their resistance to PPT. After 1 week, all theleaves of the control plants were necrotic (Fig. 2B),while there were 21 surviving plants in the case of theline derived from genotype LPC13, and 23 survivors inthe case of the line derived from genotype LPC18. Statistical analysis shows that the pattern of inheritanceof the bar gene corresponds to a 3:1 segregation ratio,showing that this gene is inherited as a simple-dominantMendelian trait.

Fig. 2A–F Phenotype of T0transgenic plants. A Regenera-tion of putative transgenicplants of maize on MS mediumwithout growth regulators. B Transgenic (left) and controlplants (right), 1 week afterBasta application. C Fertiletransgenic maize plants after3 months, under greenhouseconditions. D A tassel of T0transgenic LPC13 inbred line.E Cobs of a T0 transgenicLPC13 inbred line. F GUS-positive spikelets of a T1Southern blot-positive LPC13inbred line. Upper middle WTspikelet of the same genotype

309

Analysis of GUS expression in T1 plants

To analyse the expression of the uidA transgene in the T1progeny, GUS activity was tested in spikelets of youngear inflorescences. Eight T1 ears (3–6 cm long) from different Southern blot-positive plants derived from twolines of genotype LPC13 and two lines from genotypeFR-2 were tested for GUS activity. As a control, ears offour wild type plants of each of these two genotypeswere tested in the same manner. In every ear tested, except for the wild-type controls, GUS expression wasdetectable in the spikelets (Fig. 2F).

Discussion

For maize genomics to become a reality, it is necessaryto have a routine method which allows the transformation,with a reasonably high success rate, of the numerousmaize genotypes which are available. Such a method

would also be useful for the simpler task of introducing adesirable trait into commercial maize lines. To date, theonly system for transformation/regeneration of maizewhich has been well documented and has been repeatedlyreported as workable is the one based on the generationof type II embryogenic calli (Fromm et al. 1990; Gordon-Kamm et al. 1990; Walters et al. 1992; Pescitelliand Sukhapinda 1995; Russell and Fromm 1996; van derGeest and Petolino 1998). This system, however, onlyappears to work with a few genotypes, e.g. line A-188.Also, even in its successful applications, this methodologyis protracted and laborious. Though the capacity to produce type II calli has been shown to follow a traitwhich is inherited in dominant form (Tomes and Smith1985; Hodges et al. 1986; William et al. 1989; Bronsemaet al. 1997), to transfer this trait by conventional crossesto a maize line of commercial interest, and thereafter toeliminate the undesirable genetic background, is a costlyprocess, as it involves long and complex manipulations(to perform the initial crosses, to isolate immature embryos in each generation, to challenge them in vitro todetermine if they produce type II callus, and to regener-ate plants to continue with the crosses), and the need tohave plants with immature embryos at the right stage inthe greenhouse. Therefore, this methodology is unsuit-able for application to the new lines of interest that arerapidly emerging, and to modern functional genomicprojects.

Another strategy, much less widely used, involvesgermline genetic transformation of immature embryos ofnon-commercial inbreds (A-188, B-73, and some others)

Fig. 3 Southern blot hybridization analysis with a bar probe, ofseveral transgenic maize plants of geotypes LPC13, LPC16,FR28-1, FR28-2, ZM1, and ZM7. For several randomly chosenPPT-resistant maize plants derived from genotypes LPC13,LPC16, FR28-1, FR28-2, ZM1, and ZM7, the DNA was digestedwith BamHI/HindIII, and analysed by Southern blot hybridization.In all of the tested genotypes there were lines with positive hybrid-ization signals, showing integration of the bar gene into theirgenomes

Fig. 4 Southern blot hybridization analysis with a uidA probe ofeight transgenic maize plants of genotype LPC13. The same eightLPC13 plant DNAs examined in Fig. 3, were analysed by diges-tion with BamHI/BclI followed by Southern blot analysis, to testthe presence of the uidA gene into their genome. WT LPC13 un-transformed genotype

310

(Lowe et al. 1995). This approach suffers from low efficiency of genetic transformation, which results in ahigh cost per transgenic clone (see e.g. Gordon-Kamm etal. 1990).

In 1992 and in 1996, Zhong et al. (1992, 1996) reported an alternative method for maize transforma-tion/regeneration, based on obtaining organogenic callifrom shoot tips, thus obviating the need for isolating immature embryos. We tested the protocol of Zhong etal. (1992, 1996) for several tropical maize genotypes,without success. Therefore, it appears to be applicableonly to temperate maize lines. Here, we present a modi-fication of the method of Zhong et al., which resulted insignificant improvements in terms of its applicability totropical maize germplasm. It is a fast, simple, and effi-cient system, starting from shoot tips, which has provensuccessful with all of the tropical and subtropical maizelines tested by us. The main difference from the methoddescribed by Zhong et al. (1992, 1996) is the addition ofadenine to the culture medium. Adenine has been con-sidered a low-activity cytokinin or even a vitamin whichmay have the effects of a cytokinin (Gaspar et al. 1996).In the present work, we observed that adenine, when administered jointly with other growth regulators, playsan important role in the induction of organogenic maizecalluses. Our results show that all the maize lines testedwere capable of producing organogenic calluses, al-though the efficiency to produce them is genotype dependent (Table 2). However, it is important to notethat, once the organogenic calluses are established, theirgeneral behaviour in culture is very similar among thedifferent genotypes. A comparable phenomenon occurswith transformation: the efficiency is different dependingon the genotype, but it was possible to produce transgen-ic maize plants for all the tested genotypes, using thesame protocol. We are currently testing the feasibility ofextending this system not only to other lines of maize,but also to other recalcitrant crops. [Using the MPC medium and a similar protocol, we have been able to induce organogenic calluses amenable for gene transferfrom several Brazilian maize inbred lines, from inbredlines of sorghum, from teosinte (Zea mays ssp. mexicana)(J. L. Cabrera Ponce et al., unpublished results), andfrom the grass Bouteloua gracilis (Aguado-Santacruz etal., in press).] Thus, shoot tips seem to have a strong potential to produce transformation/regeneration profi-cient calluses.

It has been previously reported that osmotic treatmentsenhance the efficiency of both transient expression andstable transformation (Russell et al. 1992; Vain et al.1993; Cabrera-Ponce et al. 1997). It is believed that thisis due to plasmolysis of the target cells. Plasmolysed cellsmay be less likely to extrude their protoplasm followingpenetration of the cell by the particles (Armaleo et al.1990). However, our results indicate that there should beanother factor, additional to osmolarity, affecting thetransformation efficiency, because the treatment with sucrose (a metabolizable sugar) was clearly superior tothe treatment with mannitol (a non-metabolizable sugar).

As can be seen in Fig. 5, using the protocol describedhere, one obtains first (under light conditions) organo-genic calluses (Fig. 1C); subsequently (in the dark),some regions of the calluses generate embryogenic-likestructures (Fig. 1D), which later (changing back to lightconditions) produce organogenic calluses. This provesthat interconversion between organogenic and embryo-genic-like calluses is possible in maize. It would be in-teresting to perform histological studies to determinehow the cells are dividing to bring about these develop-

Fig. 5 Schematic diagram of the protocol to obtain fertile trans-genic maize plants of tropical and subtropical genotypes. D Culti-vated in the dark, L cultivated in light

311

mental changes. Co-generation of organogenic and em-bryogenic-like structures within the same culture hasbeen described for cereals before (Lowe et al. 1985;Wernicke and Milkovits 1986; Bhaskaran and Smith1990), but, to our knowledge, the possibility that thiscould be a suitable material for obtaining transgenicplants has never been suggested.

Southern blot analysis showed that the integrationpattern of the introduced DNA is quite variable and gen-otype dependent: genotypes FR28-1 and FR28-2 showedquite complex integration patterns and highly variablecopy number; genotypes LPC13 and LPC16 showed lesscomplex integration patterns, but still several rearrangedfragments are observed; and ZM1 and ZM2 showed lesscomplex patterns of integrations, showing a maximum ofthree bands. Since genotypes that are genetically moreclosely related (i.e. FR28-1 and FR28-2) showed similarcomplexities in the integration patterns, it is possible thatthere are genetic determinants of maize that influence theintegration pattern. The nature of these genetic determi-nants still remains to be determined, and whether theyinfluence the processing of the introduced DNA beforeor during integration in the maize chromosomes. In afew cases, we observed that bands of the expected sizefor the bar or uidA genes were not present. However,these lines were Basta resistant or GUS positive, indicat-ing that in spite of the rearrangements suffered during integration, the functionality of the chimeric genes wasmaintained.

In assessing the performance of the methodology described here, three features appear particularly note-worthy: (1) most of the lines tested by Southern blotanalysis proved positive for the presence of the trans-genes, indicating the high efficiency of the selectionmethod, (2) in contrast to the low conversion of embryo-genic structures to plants reported by Bohorova et al.(1995) for tropical lines, the calluses obtained with thepresent system show a high capacity for regeneratingplants, and (3) plants regenerated with this method weresuccessfully adapted to greenhouse conditions in 100%of the cases. This latter result is probably due to the rootsbeing well formed before transplantation.

We believe that our protocol could be developed intoa general transformation system suitable for the high-throughput production of transgenic plants which is required for the functional genomic analysis of maize.

Acknowledgements We thank Guadalupe Sánchez, Nancy Villalobos, Omar Gatjens, Rosa María Adame, Silvia Solís, andSilvia Vigil for technical assistance, and Antonio Cisneros for thephotographic work. We are indebted to Agustín Sánchez-Vásquezat the Mexican Embassy in Costa Rica for his kind help. We arealso grateful to Dr Reynaldo Pless-Elling and to Dr June Simpsonfor critically reading this manuscript. This research was supportedin part by grants of the HHMI (75191-526901), the RockerfellerFoundation (AS 9644), and the European Commission (ERBIC-18C-960089) to L. H. E; A. O. S. is indebted to CONACYT for aPhD fellowship (94468) and M. V. M. to UNU/Biolac and UNESCO for a visiting scientist fellowship.

References

Aguado-Santacruz GA, Rascón-Cruz Q, Cabrera-Ponce JL, Martínez-Hernández A, Olalde-Portugal V, Herrera-Estrella L(2002) Transgenic plants of blue grama grass, Bouteloua gracilis (H.B.K.) Lag. ex Steud. from microprojectile bom-bardment of highly chlotophyllous embryogenic cells. TheorAppl Genet 104:763–771

Armaleo D, Ye GN, Klein TM, Shark KB, Sanford JC, JohnstonSA (1990) Biolistic nuclear transformation of Saccharomycescerevisiae and other fungi. Curr Genet 17:97–103

Armstrong CL, Green CE (1985) Establishment and maintenanceof friable, embryogenic maize callus and involvement ofL-proline. Planta 164:207–214

Armstrong CL, Romero-Steverson J, Hodges TK (1992) Improvedtissue culture response of an elite maize inbred through back-cross breeding, and identification of chromosomal regions im-portant for regeneration by RFLP analysis. Theor Appl Genet84:755–762

Bhaskaran S, Smith RH (1990) Cell biology and molecular genetics.Regeneration in cereal tissue culture: areview. Crop Sci30:1328–1337

Bohorova NE, Luna B, Brito RM, Huerta LD, Hoisington DA(1995) Regeneration potential of tropical, subtropical, midalti-tude, and highland maize inbreds. Maydica 40:275–281

Bronsema FBF, Oostveen WJF van, Lammeren AAM (1997)Comparative analysis of callus formation and regeneration oncultured immature maize embryos of the inbred lines A188and A632. Plant Cell Tissue Organ Cult 50:57–65

Cabrera-Ponce JL, Lopez L, Assad-Garcia N, Medina-Arevalo C,Bailey-Moreno AM, Herrera-Estrella L (1997) An efficient par-ticle bombardment system for the genetic transformation of asparagus (Asparagus officinalis L.) Plant Cell Rep 16:255–260

Carvalho CHS, Bohorova N, Bordallo PN, Abreu LL, ValicenteFH, Bressan W, Paiva E (1998) Type II callus production andplant regeneration in tropical maize genotypes. Plant Cell Rep17:73–76

D’Halluin K, Bonne E, Bossut M, De Beuckeleer M, Leemans J(1992) Transgenic maize plants by tissue electroporation.Plant Cell 4:1495–1505

Duncan DR, Williams ME, Zehr BE, Widholm JM (1985) Theproduction of callus capable of plant regeneration from imma-ture embryos of numerous Zea mays genotypes. Planta 165:322–332

Feinberg AP, Vogelstein BA (1983) A technique for radiolabelingDNA restriction endonuclease fragments to high specific ac-tivity. Anal Biochem 132:6–13

Fromm ME, Morrish F, Armstrong C, Williams R, Thomas J,Klein TM (1990) Inheritance and expression of chimeric genesin the progeny of transgenic maize plants. Biotechnology8:833–839

Gaspar T, Kevers C, Penel C, Greppin H, Reid DM, Thorpe TA(1996) Plant hormones and plant growth regulators in planttissue culture. In Vitro Cell Dev Biol Plant 32:272–289

Geest AHM van der, Petolino JF (1998) Expression of a modifiedgreen fluorescent protein gene in transgenic maize plants andprogeny. Plant Cell Rep 17:760–764

Gordon-Kamm WJ, Spencer TM, Mangano ML, Adams TR, Daines RJ, Start WG, O’Brien JV, Chambers SA, AdamsJr, WR, Willets NG, Rice TB, Mackey CJ, Krueger RW, Kausch AP, Lemaux PG (1990) Transformation of maize cellsand regeneration of fertile transgenic plants. Plant Cell 2:603–618

Hazel CB, Klein TM, Anis M, Wilde HD, Parrot WA (1998)Growth characteristics and transformability of soybean embry-ogenic cultures. Plant Cell Rep 17:765–772

Hodges TK, Kamo KK, Imbrie CW, Becwar MR (1986) Genotypespecificity of somatic embryogenesis and regeneration inmaize. Biotechnology 4:219–223

Koziel MG, Beland GL, Bowman C, Carozzi NB, Crenshaw R,Crossland L, Dawson J, Desai N, Hill M, Kadwell S, LaunisK, Lewis K, Maddox D, McPherson K, Meghji MR, Merlin E,

312

Rhodes R, Warren GW, Wright M, Evola SV (1993) Field performance of elite transgenic maize plants expressing aninsecticidal protein derived from Bacillus thuringiensis. Bio-technology 11:194–200

Lowe K, Taylor DB, Ryan PK, Paterson E (1985) Plant regenera-tion via organogenesis and embryogenesis in the maize inbredline B73. Plant Science 41:125–132

Lowe K, Bowen B, Hoerster G, Ross M, Bond D, Pierce D, Gordon-Kamm B (1995) Germline transformation of maizefollowing manipulation of chimeric shoot meristems. Biotech-nology 13:677–682

McCabe DE, Swain WF, Martinell BJ, Christou P (1988) Stabletransformation of soybean (Glycine max) by particle accelera-tion. Biotechnology 11:596–598

Murashige T, Skoog F (1962) A revised medium for rapid growthand bioassays with tobacco tissue cultures. Physiol Plant15:473–497

Murry LE, Elliot LG, Capitant SA, West JA, Hanson KK, ScarafiaHL, Johnston S, DeLuca-Flaherty C, Nichols S, Cunanan D,Dietrich PS, Mettler IJ, Dewald S, Warnick DA, Rhodes C,Sinibaldi RM, Brunke KJ (1993) Transgenic corn plants expressing MDMV strain B coat protein are resistant to mixedinfections of maize dwarf mosaic virus and maize chloroticmottle virus. Biotechnology 11:1559–1564

Pescitelli SM, Sukhapinda K (1995) Stable transformation viaelectroporation into maize type II callus and regeneration offertile transgenic plants. Plant Cell Rep 14:712–716

Prioli LM, Silva WJ (1989) Somatic embryogenesis and plant regeneration capacity in tropical maize inbreds. Rev BrazilGenet 12:553–566

Ramirez-Diaz JL, Ron-Parra J, Cota-Agramont O (1995) H315Híbrido de maíz de ciclo intermedio para la zona subtropical ytropical de México. Instituto Nacional de InvestigacionesForestales y Agropecuarias. Folleto técnico no. 3. Centro deInvestigaciones Pacífico Centro, Campo Experimental Centralde Jalisco

Rhodes CA, Pierce DA, Mettler IJ, Mascarenhas D, Detmer JJ(1988) Genetically transformed maize plants from protoplasts.Science 240:204–207

Russell DA, Fromm ME (1996) Tissue-specific expression intransgenic maize of four endosperm promoters from maize andrice. Transgenic Res 6:157–168

Russell JA, Roy MK, Sanford JC (1992) Major improvements inbiolistic transformation of suspension-cultured tobacco cells.In Vitro Cell Dev Biol 28:97–105

Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning:alaboratory manual, 2nd edn. Cold Spring Harbor LaboratoryPress, Cold Spring Harbor, N.Y.

Sanford JC, DeVit MJ, Russell JA, Smith FD, Harpending PR,Roy MK, Johnston SA (1991) An improved helium-drivenbiolistic device. Technique 3:3–16

Shure M, Wessler S, Fedoroff N (1983) Molecular identificationand isolation of the waxy locus in maize. Cell 35:225–233

Sidorenko LV, Xianggan L, Cocciolone SM, Chopra S, Tagliani L,Bowen B, Daniels M, Peterson T (2000) Complex structure ofa maize Myb gene promoter: functional analysis in transgenicplants. Plant J 22:471–482

Srivastava V, Ow DW (2001) Single-copy transformants of maizeobtained through the co-introduction of a recombinase-expressing construct. Plant Mol Biol 46:561–566

Tomes DT, Smith OS (1985) The effect of parental genotype oninitiation of embryogenic callus from elite maize (Zeamays L.) germplasm. Theor Appl Genet 70:505–509

Tomes DT, Ross MC, Songstad DD (1995) Direct DNA transferinto intact plant cells via microprojectile bombardment. In:Gamborg OL, Phillips GC (eds) Plant cell, tissue and organculture: fundamental methods. Springer, Berlin HeidelbergNew York, pp 197–213

Unger E, Betz S, Xu R, Cigan AM (2001) Selection and orienta-tion of adjacent genes influences DAM-mediated male sterilityin transformed maize. Transgenic Res 5:409–22

Vain P, Mc Mullen MD, Finner JJ (1993) Osmotic treatment enhances particle-mediated transient and stable transformationof maize. Plant Cell Rep 12:84–88

Vasil V, Vasil IK, Lu C (1984) Somatic embryogenesis in long-term callus cultures of Zea mays L. (Gramineae). Am J Bot71:158–161

Walters DA, Vetsch CS, Potts DE, Lundquist RC (1992) Transfor-mation and inheritance of a hygromycin phosphotransferasegene in maize plants. Plant Mol Biol 18:189–200

Wernicke W, Milkovits L (1986) The regeneration potential ofwheat shoot meristems in the presence and absence of 2,4-dichlorophenoxyacetic acid. Protoplasma 131:131–141

William MR, Schroll SM, Hodges TK (1989) Inheritance of so-matic embryogenesis and plantlet regeneration from primary(type I) callus in maize. In Vitro Cell Dev Biol 25(1):95–100

Zhong H, Srinivasan C, Sticklen MB (1992) In-vitro morphogene-sis of corn (Zea mays L.). I. Differentiation of multiple shootclumps and somatic embryos from shoot tips. Planta 187:483–489

Zhong H, Sun B, Warkentin D, Zhang S, Wu R, Wu T, SticklenMB (1996) The competence of maize shoot meristems for integrative transformation and inherited expression of trans-genes. Plant Physiol 110:1097–1107