In vitro transformation of pearl millet (Pennisetum glaucum (L). R. BR.): Selection of chlorsulfuron-resistant plants and long term expression of the gus gene under the control of the emu promoter

The plasmids pULGU1 and pEmuGN were introduced by biolistics in embryogenic cell suspensions of pearl millet, Pennisetum glaucum. The plasmid pULGU1 contained the mutant acetolactate synthase (ALS) gene of Arabidopsis thaliana, responsible for resistance to the chlorsulfuron herbicide. The plasmid pEmuGN carried the reporter β-glucuronidase (GUS) gene of Escherichia coli. Two months after bombardment of the cells, transformed calli were selected on Murashige and Skoog medium containing 30 nM chlorsulfruron, a completely concentration inhibitory for non-transformed cells. A concentration of 200 nM chlorsufuron did not affect the growth of these selected resistant calli. Genomic DNA of resistant cells presented an electrophoretic pattern revealing the integration of the mutant ALS gene into the genome of Pennisetum glaucum. Expression of the GUS reporter gene was revealed by histochemical assay of the blue coloration of the calli in the presence of the substrate Xgluc (5-bromo-4-chloro-3-indolyl-β-D-Glucuronic acid). The presence of GUS gene was furthermore confirmed, by southern blot hybridization of non-radioactive probes, labelled with digoxigenin, on genomic DNA extracted from selected chlorsulfuron-resistant calli. GUS gene activity observed in all selected calli was high during the first 6 months after bombardment and then decreased. Pearl millet plants were regenerated from cell lines derived from chlorsulfuron resistant calli. Southern blot hybridization of non-radioactive probes with genomic DNA extracted from these regenerated plants showed the presence of the GUS and ALS transgenes, confirming the stable transformation of these plants.


INTRODUCTION
The Poaceae family includes important cereal crops which often need to be improved to face pathogens attacks and other environmental constrains such as drought, flooding and competition of weeds during cropping seasons.Crop plants are usually improved by crossing them with wild relatives.Unfortunately, this approach has some limitations and sometimes genetic engineering methods are used to transfer genes of interest from wild plants to cultivated ones.As compared to dicotyledons plants, graminaceous are recalcitrant to transformation methods using the Ti plasmid as vector (Plaza-Wüthrich and Tadele, 2012;Hussain et al., 2013;Sah et al., 2014).The major obstacles to this gene delivery method were attributed to the limited host-range of Agrobacterium (Hwang et al., 2013;Qamar et al., 2015), and to the different responses of monocots (cell death) and dicots (cell division) to wounds (Potrykus, 1990;Hiei et al., 2014).A variety of techniques such as polyethylene glycol (PEG) method, electroporation method, mild ultrasonication and biolistic have been proposed for genetic transformation of cereals.Some results have been obtained using the direct introduction of DNA into competent protoplasts by polyethylene glycol (PEG) induced uptake (Tiécoura et al., 2001;Bajaj, 2012;Ji et al., 2013), electroporation (Vasil, 2012;Jordan et al., 2013), followed by the growth of calli and in some cases, the regeneration of transformed plants (Shimamoto et al., 1989;Bajaj and Mohanty, 2005;Bahieldin et al., 2005;Danilova, 2007).
Concerning pearl millet, Pennisetum glaucum, the transient expression of the reporter ß-glucuronidase (GUS) gene in transformed calli has been reported (Vasil, 2008;Jha et al., 2011;Ramadevi et al., 2014).Stable calli have been selected and analysed by Lambé et al. (1995) after introducing the neomycin phosphotransferase gene (NPTII).Transgenic plants resistant to phosphinothricin and hygromycin (Goldman et al., 2003;O'Kennedy et al., 2004;Ceasar and Ignacimuthu, 2009;Ramadevi et al., 2014), kanamycin (Plaza-Wüthrich and Tadele, 2012) and mildew (Latha et al., 2006) have been studied.However, to date, no stable P. glaucum transformed calli and plant resistant to the chlorsulfuron herbicide have been reported yet.In the present study, the following were reported: 1) the transient and long term expression of the reporter βglucuronidase gene (GUS) of Escherichia coli in P. glaucum embryogenic cell suspensions transformed using biolistic method; 2) The isolation and analysis of chlorsulfuron herbicide-resistant calli and plants of pearl millet regenerated after introduction of the mutant Acetolactase synthase (ALS) gene of Arabidopsis thaliana and the reporter GUS gene of E. coli.
Author(s) agree that this article remains permanently open access under the terms of the Creative Commons Attribution License 4.0 International License suspensions, two types of calli whose sizes were respectively less than 250 and 630 µm.5 ml of each type of calli at a concentration of 60 to 100 mg/ml were dispersed on a nitrocellulose membrane by vacuum filtration and bombarded with plasmids DNA.

Plasmid DNA
Three types of plasmids (Figure 1) were used to transform the embryogenic cells of P. glaucum.

pULGU1 plasmid
The plasmid pULGU1 has a size of 7.7 kbp (Figure 1A).It contains the NCol-XbI fragment carrying the mutant acetolactate synthase (ALS) gene of Arabidopsis thaliana controlled by the Cauliflower Mosaic Virus (CaMV) 35S promoter and terminator (Haughn and Somerville, 1986).The mutant ALS gene was isolated from A. thaliana plant resistant to chlorsulfuron, a sulfonylurea herbicide.Resistant cells carrying the mutant ALS genes can grow normally in a medium containing the chlorsulfuron herbicide, whereas susceptible cells necrose.

pEmuGN plasmid
The plasmid pEmuGN has a size of 6.504 kb (Figure 1C).It contains also the reporter GUS gene controlled by the Emu promoter and the Nopaline Synthetase (NOS) terminator.The Emu promoter is a recombinant promoter based on truncated maize Adh1 promoter, with multiple copies of the Anaerobic Responsive Element from the maize Adh1 gene and OCS-elements from the octopine synthase gene of A. tumefaciens (Last et al., 1991).

Preparation of DNA-coated tungsten micro-projectiles
10 mg of tungsten micro-projectiles (1.2 µm in diameter) were put in 1 ml of ethanol 95%, sonicated for 2 min and let to stand for 2 h.The microprojectiles were then centrifuged, subjected to two rinses of 2 h each and stored in 200 µl of glycerol 50%.The DNA/microprojectiles mixture was prepared by adding to 20 µl of microprojectiles, 4 µl of 1 µg/µl plasmid DNA, 20 µl of 2.5 M CaCl2 and 8 µl of 0.1 M spermidine.The mixture was then vortexed for 20 s, incubated for 10 min on ice and 30 µl of the supernatant were removed.DNA of plasmid p35SGUS or plasmid pEmuGN was used to transform the cells for transient GUS gene expression.A DNA mixture of the plasmids pULGU1 and pEmuGN (1/1) was used to bombard the cells for their stable transformation.

Assessing the transient expression of GUS gene
Embryogenic calli of less than 250 and 630 µm in diameter were respectively transferred to twenty nitrocellulose membranes and bombarded with the plasmid pEmuGN or the plasmid p35SGUS.Control membranes were bombarded without plasmid.48 h after bombardment, the nitrocellulose membranes were incubated at 37°C for 24 h in a GUS histochemical assay buffer containing 0.1 M NaPO4 pH 7.0, Triton X-100 1% and 0.5 mg/ml X-GLUC.GUS activity was measured by counting GUS positive spots (GUPS) due to cleavage of the X-GLUC substrate which leads to a bluecoloured precipitate (Figure 3A).

Selection of stable transformed calli
Eight nitrocellulose membranes carrying calli of less than 250 µm in diameter were bombarded with a mixture of plasmids pEmuGN and pULGU1.After bombardment, nitrocellulose membranes were transferred to the MS 2.5, 1, 1 medium, that is, Murashige and Skoog medium, supplement with 2.5 mg/l 2,4dichlorophenoxyacetic acid, 1 mg/l napthalene acetic acid, 1 mg/l 6benzylaminopurine.The MS 2.5, 1,1 medium was also supplemented with 30 nM chlorsulfuron, a lethal concentration for non-transformed-cells of Pennisetum glaucum (Tiécoura and Ahoussou, 2001).Two months after bombardment, four calli selected in these conditions and whose growth rate was not affected by a concentration of up to 200 nM chlorsulfuron were considered chlorsulfuron-resistant calli.They were labelled CAL1, CAL2, CAL3 and CAL4 and used for GUS activity assay 2, 6 and 9 months after bombardment.

GUS activity monitoring
0.5 ml of GUS assay buffer, composed of extraction buffer, methanol 20% (v/v) and 1.25 mM 4-MUG (4-methylumbelliferyl β-Dgalactopyranoside), was mixed with 50 µl of β-glucuronidase extracted from P. glaucum cells.The mixture was incubated in a double boiler at 37°C and every 30 min, 100 µl were aliquoted and added to 900 μl of a stop buffer consisting of 0.2 M Na2CO3.The fluorescence of MU (4-methyl umbelliferyl), due to the cleavage of the substrate 4-MUG, was measured with a « KONTRON SFM 25 » fluorometer (excitation at 365 nm, emission at 455 nm).The fluorometer was preliminarily calibrated with MU solutions (100 nM up to 1 µM) GUS protein assay 100 µl of β-glucuronidase, extracted from P. glaucum cells, were added to 3 ml of reagent containing 100 mg/l Coomassie Brilliant Blue G-250, 50 ml/l ethanol 95%, 100 ml/l phosphoric acid 85%.The optical density (OD) of the mixture was measured at 595 nm after incubation for 5 min.A calibration curve was constructed using bovine serum albumin.GUS activity was assessed on the basis of the optical density of the mixture which depends on the quantity in nanomole of MUG cleaved per minute and per mg of GUS protein.

DNA analysis of chlorsulfuron-resistant calli and plants
Six months after bombardment, plants were regenerated from resistant calli.In order to verify the presence of the GUS gene of E. coli and the mutant ALS gene of A. thaliana in the transformed cells of pearl millet, total genomic DNA of chlorsulfuron-resistant calli and plants (Figure 3B) were analysed respectively three and six months after bombardment, by Southern blotting.Genomic DNA extracted from non-transformed calli was used as control.The mutant ALS gene of A. thaliana was detected using a probe of 1.7 kbp obtained by digesting the plasmid pULGU1 with the enzymes NcoI and BglII.The probe contains the 5'coding region of the mutant ALS gene of A. thaliana (Figure 5).The presence of the GUS gene was tested using a probe of 2.797 kbp-long obtained by digesting the plasmid pEmuGN with the restriction enzymes EcoRI and BamHI.This fragment contains the intron 1, the coding region of the GUS gene and the NOS terminator (Figure 6).In each case, the genomic DNA of chlorsulfuron-resistant calli and plants was also digested with the restriction enzymes used to isolate the probes.The probes were labelled with digoxigenin.
Genomic DNA was extracted from 1 g of non-transformed and transformed-cells of P. glaucum calli or plants.10 µg of each DNA sample, non-digested or digested with the appropriate restriction enzymes, was submitted to electrophoresis on 0.8% agar gel for 2 h.DNA fragments were then transferred to Nytran nylon membrane.Non-radioactive probes labelled with digoxigenin were used to detect the target fragments of GUS and ALS genes.

Transient expression of the reporter GUS gene
Results of GUS activity, 48 h after bombardment of the calli, are presented in Figure 2. No blue spot was observed on control nitrocellulose membranes.Bluecoloured GUS positive spots (GUPS) were observed on nitrocellulose membranes bombarded with plasmids p35SGUS and pEmuGN.The average number of GUPS was higher on nitrocellulose membranes carrying calli of less than 250 µm in diameter; 100 and 250 spots were recorded after bombardment with plasmids p35SGUS and pEmuGN, respectively.In the case of nitrocellulose membranes carrying calli of less than 630 µm in diameter, an average of 20 and 25 GUPS were recorded when the bombardment was performed with plamsids   p35SGUS and pEmuGN, respectively.

Long term expression of the reporter GUS gene in selected chlorsulfuron-resistant calli
Figure 4 presents the evolution of GUS activity in four selected chlorsulfuron-resistant calli during nine months after bombardment.Apart from CAL4 (50 nmoles of MU/min/mg GUS Prot.), GUS activity was high in transformed-calli.In CAL1 and CAL2, GUS activity increased during the six first months after bombardment, from 400 to 1500 nmoles of MU/min/mg GUS Prot.and 500 to 900 nmoles of MU/min/mg GUS Prot, respectively.GUS activity decreased thereafter and reached 1100 and 300 nmoles of MU/min in CAL1 and CAL2, respectively, nine months after bombardment.GUS activity was high in CAL3, (1000 nmoles of MU/min/mg GUS Prot.) two months after bombardment and decreased up to 30 nmoles of MU/min/mg GUS Prot., nine months after bombardment.

Detection of the mutant ALS gene in chlorsulfuronresistant calli and plants
Figure 5A-C present the results of DNA analysis of chlorsulfuron-resistant calli, (CAL1, CAL2, CAL3 and CAL4), by southern blotting, three and six months after bombardment.Three months after bombardment (Figures 5A and B), the probe did not hybridize with the control DNA (lane 1).In contrast, it hybridized with nondigested DNA (lanes 2, 3, 4, and 5) and digested DNA (lanes 6, 7, 8 and 9) of CAL1, CAL2, CAL3 and CAL4, respectively.The digested genomic DNA of these calli harbours a 1.7 Kbp fragment equalling the size of the probe obtained from the digestion of the plasmid pULGU1 by the restriction enzymes BglII and PstI (lane 13).Six months post bombardment (Figure 5C), similar results were obtained for non-digested DNA (line 2 and 3) and digested DNA (lanes 4 and 5) of pearl millet plants regenerated from CAL1 and CAL2 Luminescence intensities of the 1.7 Kbp DNA fragments revealed in CAL1, CAL2 CAL3 and CAL4 (Figure 5B: lanes 6, 7, 8 and 12) correspond to the luminescence intensities of 5 to 10 copies of the ALS gene (Figure 5B: lanes 11 and 12).
A DNA fragment of 7.7 Kbp long, corresponding to the linearized plasmid pULGU1, was observed three months after bombardment (Figure 5A) in the genomic DNA of CAL1 and CAL2, non-digested (lanes 2 and 3) or digested by the restriction enzymes BglII and PstI (lanes 10 and 11).And 7.7 Kbp DNA fragment was no longer observed six months after bombardment (Figure 5B: lanes 2 and 3).calli and plants are presented in Figure 6. Figure 6A presents results of DNA analysis of chlorsulfuronresistant calli, (CAL1, CAL2, CAL3 and CAL4) three months after bombardment.The probe did not hybridize with the control DNA (lane 1).The probe hybridized to non-digested DNA (lanes 2, 3, 4, and 5) and digested DNA (lanes 6, 7, 8 and 9) of CAL1, CAL2, CAL3 and CAL4, respectively.The digested DNA of these calli revealed a 2.797 Kbp fragment equal to the size of the probe obtained from the digestion of plasmid pEmuGN by the restriction enzymes BamHI and EcoRI (lanes 10, 11 and 12).Luminescence intensities of the 2.797 Kbp DNA fragments detected in CAL1, CAL2 CAL3 and CAL4 (Lanes 6, 7, 8 and 9), corresponds to 1 to 2 folds the luminescence intensity of one copy of the GUS gene (lane 10).Six months post bombardment (Figure 6B), the probe did not hybridize with non-digested DNA (Line 2) and digested DNA (Line 4) of pearl millet plants regenerated from CAL1.The GUS gene was detected in DNA of plants regenerated from CAL2 but the luminescence intensity was weak (Lanes 3 and 5).

DISCUSSION
GUS positive spots (GUPS) observed on bombarded nitrocellulose membranes indicate that plasmids p35SGUS and pEmuGN have been successfully introduced in P. glaucum cells.The low average number of GUPS (20 and 25, respectively for plasmid p35SGUS and plasmid pEmuGN) observed on nitrocellulose membranes carrying calli of less than 630 µm in diameter could be explained on the bases of calli size.It is possible that, because of their relatively larger size, they hardly stick to the nitrocellulose membrane, and they have probably been scattered by the micro-projectiles.
Regardless of the size of the calli carried by the bombarded nitrocellulose membranes, average numbers of GUPS increased when the bombardment was performed with plasmid pEmuGN (Figure 1).This result is due to the fact that the Emu promoter is stronger than the 35S promoter (Last et al., 1991;Li et al., 1997;Chowdhury et al., 1997;Tiecoura et al.;2001).It has also been observed that other promoters are stronger than the 35S promoter when used for in vitro transformation of different monocot and dicot plants.Indeed, the Cauliflower Mosaic Virus (CaMV) 35S promoter was less effective in driving the GUS gene expression in tomato (dicot) and maize (monocot) plant tissues as compared to the maize promoters Gos-2, Enolase, Actin-2 (Shireen et al., 2002) and Actin-1 (Prakash et al., 2008).By using immature and mature embryos, shoot tips and embryogenic calli for in vitro transformation of S. bicolor, Tadesse et al. (2003) observed that the strength of the CaMV 35S promoter ranked after those of the ubi1, act1D, adh1 promoters.Two switchgrass (Panicum virgatum L.) ubiquitin genes promoters PvUbi1 and PvUbi2, showed strong expression of the GUS gene in switchgrass and rice calli, equaling or surpassing the expression levels of the CaMV 35S promoter (Mann et al., 2011).Furthermore, in these studies, GUS expression levels driven by the different promoters vary according to the plant species and the type of explants used for in vitro transformations.Our results confirmed these between-plant species and between-promoters variabilities of GUS gene expression.They are also consistent with previous in vitro transformation experiments which showed that the Emu promoter is a strong promoter for transient and stable transgene expression in monocot (Singh et al., 2013).
By co-bombarding embryogenic cells of pearl millet with a mixture of plasmids pULGU1 and pEmuGN carrying respectively, the mutant Acetolactate Synthetase gene of A. thaliana and the reporter GUS gene of E. coli, we have obtained chlorsulfuron-resistant calli in which GUS activity was detectable during nine months post bombardment.However, the evolution of GUS activity presented different patterns in four selected chlorsulfuron-resistant calli.Indeed, in two calli, GUS activity increased during the six first months post bombardment and decreased thereafter.In contrast, GUS activity only decreased in two other calli.The variability of GUS activity in transformed tissues is a common feature reported by different authors.For instance, Peng et al. (1990) introduced the GUS gene in protoplasts of different rice (O.sativa) genotypes by polyethylene glycol-mediated transformation and noted that evolution of GUS activity varied according to the rice genotype.Lambé et al. (1995) transformed calli of pearl millet using biolistic and observed between and within-calli variations of long term expression of the GUS gene.At intra calli was longer in meristematic cells than in elongated differentiated cells.These different patterns of GUS activity can be explained by the fact that the used transformation methods do not target any particular region of the genome to be transformed.Thus, expression of the transgene depends on the part of the genome in which it is integrated and regulatory mechanisms to which it will be subjected (Latham et al., 2006).So, whatever the power of the promoters used for in vitro transformations, expression of the transgenes also depends on the genetic backgrounds, that is, the genotypes of the transformed plants (King et al., 2014).On the other hand, when considering the same plant genotype, a high physiological cell activity can lead to a strong expression of the GUS gene in some explants (Jha et al., 2011).In addition, the increase of GUS activity in some calli during the first 6 months post bombardment may be due to the increase of the number of GUS gene copies when transformed-cells undergo mitotic divisions.Whatever the variations of GUS activity, it decreases as the time after the bombardment increases and eventually disappears.
DNA analysis with isoschizomeric enzymes responding differentially to methylation of restriction sites and gene reactivation after treatment with 5-azacytidine, showed that this gradual decrease of the GUS activity is related to a progressive methylation of the transgene in transformed plants.This process is supposedly much higher in monocots (Lambé et al., 1995;Sood et al., 2011).However, in the current study, in contrast to GUS activity, resistance to chlorosulfuron was stably expressed, at least during the first nine months post bombardment, by pearl millet plants regenerated from transformed calli.Thus, since the GUS gene originates from E. coli (a bacteria) and the mutant ALS gene was isolated from A. thaliana (a higher plant), this result illustrates the fact that the origin of the transgène is a parameter determining its more or less rapid inactivation in transformed plants (Stam et al., 1997;Iyer et al., 2000).In addition, six months after the bombardment, the GUS gene was hardly detectable by southern blotting analysis in the genomic DNA of pearl millet plants regenerated from transformed calli.Thus, this suggests that the GUS gene is less stable in the genome of pearl mill than the mutant ALS gene of A. thaliana Genomic DNA analysis of selected chlorsulfuron-resistant calli, by southern blotting (Figure 5A), revealed the presence of the A. thaliana ALS transgene in the genome of pearl millet's transformed cells.This result confirms the fact that the resistance of selected calli to chlorsulfuron is due to the expression of the ALS transgene.Despite the progress and the new interest in using Agrobacteriummediated methods for in vitro transformation of graminaceous (Jha et al., 2011;Sharma et al., 2011;Plaza-Wüthrich and Tadele, 2012;Hiei et al., 2014), the particles bombardment is still an effective technique for the introduction of transgenes in monocot plant cells with an average transformation frequency of more than 50% (Goldman et al., 2003).Thus, our results indicate that the biolistic method is suitable for in vitro transformation of P. glaucum varieties of Côte d'Ivoire because these varieties are recalcitrant to plants regeneration from protoplasts (Tiécoura et al., 2015).
The probe used for southern blotting analysis should not have any homologous sequence in the target genome in order to allow an effective identification of transformed plant materials (Li et al., 1992).The authors observed no cross-hybridization between the probe, consisting of the 5'-half segment of the mutant A. thaliana ALS coding sequences, and the genomic DNA extracted from nontransformed pearl millet calli and plants.This result is in agreement with previous studies which showed that the 5'-half segments of ALS coding sequences are sufficiently different between monocot and A. thaliana to eliminate cross-hybridization with non-transformed material when performing southern blotting analyses (Park et al., 2002;Miki and McHugh, 2004).
Three months post bombardment, a DNA fragment of 7.7 Kbp long, equalling the size of the plasmid pULGU1, was observed in the genomic DNA of two chlorsulfuronresistant calli, non-digested or digested by the restriction enzymes BglII and Pst1 (Figure 5A).However, this fragment of 7.7 Kbp long was no longer observed six months post bombardment.These results indicate that, at least up to three months after bombardment, some plasmids remained free, that is, not integrated in the genomic DNA of the transformed cells.Free plasmid DNA has also been observed in transformed cells of plants such as wheat, barley, maize and Colza (Langridge et al., 1992;Cheng et al., 2004Cheng et al., , 2010)).Langridge et al. (1992) stated, first, that endophytic microorganisms, such as mycoplasma associated with cereals cells, are responsible for the free plasmid-DNA observed in transgenic cells.Then, Cheng et al. (2004Cheng et al. ( , 2010) ) indicated that the presence of free plasmid DNA may be due to the fact that the stable transformation of plant cells is progressive and, cell organelles such as chloroplasts and mitochondria are also transformed.So, in these organelles which are very close to the prokaryotes, the plasmid DNA can be free for a long time before being digested by plant endonucleases.Finally, Rivera et al. (2012) and Sah et al. (2014) supported these two hypotheses indicating that other organelles or cellular mico-organisms can harbor plasmid DNA outside the nucleus of transformed cells.
Six months after bombardment, southern blot analysis revealed the presence of the mutant ALS gene of A. thaliana in the genomic DNA of pearl millet plants regenerated from chlrosulfuron-resistant calli (Figure 5C).This result indicates that the regenerated plants have been stably transformed.Pearl millet transformation has been performed for resistance to some herbicides and some pathogens (Girgi et al., 2002;Goldman et al., 2003;Ceasar and Ignacimuthu, 2009).However, to the knowledge of the authors, the first transgenic plants of P. glaucum resistant to the chlorsulfuron herbicide were obtained Tiécoura et al. 3121 in the current study.
According to luminescence intensities of the DNA fragments revealed by the probes in the genomic DNA of chlorsulfuron-resistant calli, respectively 1 to 2 copies of the GUS gene and 5 to 10 copies of the mutant ALS gene of A. thaliana were integrated in the genomic DNA of pearl millet.Integration of many copies of the transgene, even in the same site or in different chromosomal locations, is an important factor associated with gene silencing in transformed plants (Stam et al., 1997).An increase of single copy insertion has been obtained by optimizing the factors influencing microprojectile bombardment (Jagga-Chugh et al., 2012;Tassy et al., 2014).Application of optimization conditions for further bombardments should thus allow us to obtain pearl millet transformants harboring single copy of the mutant ALS gene of A. thaliana.
In conclusion, in the present study, transgenic plants of Côte d'Ivoire's P. glaucum NE variety, resistant to chlorsulfuron, by co-bombardment of embryogenic calli have been obtained with the mutant ALS gene of A. thaliana and the reporter GUS gene of E. coli.Latha et al. (2006) used a similar approach to develop transgenic pearl millet plants resistant to mildew.So, since several genes of interest can be introduced into the genomic DNA of pearl millet without interfering with their activities, co-transformation could be a helpful means to improve P. glaucum varieties of Côte d'Ivoire.

Figure 1 .
Figure 1.Different plasmids used for the bombardment of P. glaucum embryogenic calli.A) pULGU1 plasmid carrying the gene for resistance to sulfonylurea herbicides (ALS) controlled by the Cauliflower Mosaic Virus 35S (CaMV35S) promoter and terminator.B) p35SGUS plasmid carrying the β-glucuronidase (GUS) gene controlled by the CaMV35S promoter and the Octopine Synthase (OCS) terminator.C) pEmuGN plasmid carrying the β-glucuronidase (GUS) gene controlled by the Emu promoter and the NOS terminator.

Figure 6 .
Figure 6.Southern blot analysis of genomic DNA of pearl millet chlorsulfuron-resistant calli (A) and plants (B) for detection of the βglucuronidase (GUS) gene, respectively three and six months after bombardment.The presence of the GUS gene was checked using a probe of 2.797 kbp-long obtained by digesting the plasmid pEmuGN with the restriction enzymes ECoRI and BamHI.A: DNA analysis of chlorsulfuron-resistant calli (CAL1, CAL2, CAL3 and CAL4) three months after bombardment.Lane 1: Non digested control DNA of non-transformed callus; lanes 2, 3, 4 and 5: non-digested DNA of CAL1, CAL2, CAL3 and CAL4, respectively; lanes 6, 7, 8 and 9: DNA of CAL1, CAL2, CAL3 and CAL4, respectively, digested with the restriction enzymes ECoRI and BamHI; lanes 10, 11 and 12: respectively 1, 5 and 10 copies of plasmid pEmuGN digested with ECoRI and BamHI.B. DNA analysis of pearl millet plants regenerated from two chlorsulfuron-resistant calli (CAL1 and CAL2), six months after bombardment.Lane 1: non digested control DNA of non-transformed callus; lanes 2 and 3: Non-digested DNA of plants regenerated from CAL1 and CAL2, respectively; lanes 4 and 5: DNA of plants regenerated from CAL1 and CAL2, respectively, digested with ECoRI and BamHI.