In this method, DNA of interest is transferred to protoplasts by applying electric pulse to the mixture of DNA and protoplasts. The first successful integration of transgene was performed in maize by transformation of Black Mexican Sweet maize protoplast by uptake of naked DNA through electrochemical method (Fromm et al., 1986). At that time, effective regeneration techniques in maize were not available, therefore full grown transformed plants could not be produced. The first full grown transgenic maize plants were developed in 1988 (Rhodes et al., 1988b). In 1980s, researchers tested different plant materials as explants for development of transgenic maize but only few instances became successful in regenerating whole plants from protoplasts (Rhodes et al., 1988a; Shillito et al., 1989). In order to obtain protoplasts for transformation by direct gene delivery, immature embryos became an excellent choice of explants (Rhodes et al., 1988a). In 1988, Rhodes et al. isolated protoplasts from embryogenic cell suspension culture of inbred line A188 and then transformed the protoplasts successfully by electroporation. The DNA of interest can also be transferred to the protoplasts by the addition of polyethylene glycol (PEG) to the mixture of protoplast and DNA instead of using electric pulse. Fertile transgenic maize plants of He/89 germplasm were obtained by using this method (Golovkin et al., 1993).

In this technique, target DNA is transferred through cell wall penetration by tungsten or gold particles coated with plasmid DNA. Highly accelerated coated particles are used to target into the desired tissue in maize which can be cell suspension culture, Type II callus, Type I callus, organogenic callus from seedlings, immature embryos or shoot meristem cultures. Among the monocotyledonous grains, maize has been one of the leading targets for genetic engineering through particle bombardment technique. In comparison to protoplast transformation, particle bombardment method generated more fertile transgenic events from embryogenic callus. Immature embryos were used as target tissue for particle bombardment mediated transformation of cry1Ab gene (Koziel et al., 1993). Ever since the initial development of biolistic transformation method, several improvements have been made. Songstad et al. (1996) developed a robust biolistic transformation protocol using Hi-II genotype. It was also observed that survival and transformation efficiency can be increased further, if immature embryos were pre-cultured prior to particle bombardment (Vain et al., 1993). Transformation frequency was found to be highly increased when immature embryos were cultured on high osmotic medium after particle bombardment (Brettschneider et al., 1997; El-itriby et al., 2003). A number of other studies have reported successful use of microprojectile bombardment technique for maize transformation (Klein et al., 1989; Fromm et al., 1990; Gordon-Kamm et al., 1990; Genovesi et al., 1992; Walters et al., 1992; Frame et al., 1994; Register et al., 1994; Wan et al., 1995; Brettschneider et al., 1997; Pareddy et al., 1997). An efficient biolistic transformation protocol for organogenic calli was also developed (O’Connor-Sánchez et al., 2002). Zhang et al. (2002) performed transformation of recalcitrant inbred lines B73 and PHTE4 by particle bombardment of shoot meristem culture. In maize, maximum number of commercial events deregulated, were produced using particle bombardment (Table 2).

Usually, the vectors used for transformation consist of a plant expression cassette along with other genetic elements in a bacterial plasmid. However, only the expression cassette is required for transgene expression and not the entire plasmid. In particle bombardment, T-DNA processing and integration steps are not involved. Therefore, the vector backbone is redundant and unnecessary. The so called vectors or plant transformation constructs are used in particle bombardment at best for operational handiness rather than experimental requirement. Fu et al. (2000) designed a strategy of particle bombardment using expression cassette only. In this strategy, all the vector sequences were removed prior to particle loading and the minimal cassette with gene of interest was removed from the plasmid. This linear cassette containing only transcription unit for gene of interest was used for transformation. The results demonstrated that transgene integration and expression were possible using just minimal cassettes also. As an added advantage, the resulting transgenic plants exhibited much simpler transgene integration patterns and lower copy numbers than the plants transformed with equivalent whole constructs. This approach was used for multiple gene transfer in plants by Agrawal et al. (2005), who performed particle bombardment with five separate marker gene cassettes. The majority of the transformed plants showed simple integration patterns with a high proportion of single-copy events, high transgene expression, and inter-generational transgene stability. Thus, the earlier belief that particle bombardment generates large, multi-copy events, prone to instability and silencing may not be true and the refinements in the particle bombardment technology, especially ‘clean DNA transformation’ (Agrawal et al., 2000, 2005) demonstrate the versatility and precision of this method (Altpeter et al., 2005). This development also opened possibility of development of marker free transgenic plants with multiple genes. Generation of high quality transgenic events, in terms of clean site of integration without disrupting any endogenous gene, getting single-copy insertions, and ensuring absence of any vector backbone, etc., is also important from regulation perspective. Refinements in transformation technologies that aid recovery of quality events are very much desirable.

Silicon carbide whiskers are needle like structure having size of 20 μm in length. They penetrate cell wall and plasma membrane of target cell to transfer desired DNA and thus, the transformants are obtained (Southgate et al., 1998). In maize, non-regenerable variety Black Mexican Sweet was transformed by silicon carbide whiskers (Kaeppler et al., 1992). Fertile transgenic maize plants have been developed successfully from Type II callus and cell suspension culture using this method (Frame et al., 1994; Petolino et al., 2000). But this method has certain limitations, such as low transformation frequency and delivery of DNA only to fine cell aggregates. Like silicon carbide whisker, another physical method of gene delivery using an airgun apparatus has also been used for transient gene expression studies in maize. This apparatus utilizes compressed air from a commercial airgun to force macroprojectiles and DNA-coated tungsten particles.

Agrobacterium tumefaciens soil pathogen is a natural genetic engineer which has ability to transform plants. From last two decades, A. tumefaciens has been frequently used for transformation of dicot plants. In comparison to direct DNA transfer methods, Agrobacterium-mediated method has a major advantage that in this method low copy of relatively large DNA fragments can be integrated into host plant genome with minimum rearrangement. This results in high quality transgenic plants. Initially, it was supposed that this technique cannot be used for monocot plants. In mid 1980s, it was demonstrated that Agrobacterium mediated plant transformation can be employed for maize (Graves and Goldman, 1986; Grimsley et al., 1987). Gould et al. (1991) used shoots as target tissue for transformation. The major breakthrough came from Agrobacterium-mediated transformation of maize by Ishida et al. (1996), followed by transformation in other cereals with similar protocols. Ishida et al. (1996) used Agrobacterium strain having super-binary vector pTiBo542 containing vir genes to transform immature embryos. Since Ishida et al. (1996) gave basic Agrobacterium-mediated transformation protocol, it has been greatly improved. The improvements include heat pre-treatment, addition of copper and silver ions to co-cultivation medium and increase in co-cultivation period from 3 to 7 days. Effects of these changes were quite evident. Traditional Agrobacterium-mediated transformation can be limited by host specificity and inability of Agrobacterium to reach cells in target tissues. A new improved Agrobacterium-mediated transformation method has been developed that overcomes above mentioned barriers. It increases transfer of DNA in different plants. This technique is known as Sonication-Assisted Agrobacterium-mediated transformation (SAAT) which involves periodic exposure of target plant tissue to sonication waves in the presence of Agrobacterium (Trick and Finer, 1997).

In planta transformation of Arabidopsis by vacuum infiltration of whole plants (Bechtold et al., 1993) and the floral dip (Clough and Bent, 1998) are now routinely used. However, similar protocols for maize are not feasible. In maize, Otha (1986) and Yang et al. (2009) have depicted the possibility of pollen-tube pathway mediated transformation. However, the transformation frequency was reported to be quite low and the screening of the transformants by PCR analysis was time-consuming. Abhishek et al. (2014b) developed a tissue culture independent protocol for in planta transformation in tropical maize by using plumular meristems of germinating seeds as explants and transforming them using Agrobacterium approach.

Most of the methods described above have been used for developing commercial transgenic events in maize. However, the maximum numbers of commercial transgenic events have been developed using particle bombardment, followed by Agrobacterium-mediated transformation (Table 2).

Neomycin phosphotransferase II gene (nptII) obtained from Escherichia coli Tn5 transposon and kanamycin were used in early maize transformation experiments. So far, 16 events of transgenic maize containing nptII have been commercialized. Another antibiotic selection system is based on hygromycin phosphotransferase (hpt) with hygromycin B as selection agent. Hygromycin inhibits single cell or small clusters of cells and their growth but large clumps of cells are less susceptible to antibiotic selection at later stages of transformation, which require high concentration of antibiotic that may have damaging effect to selected cells or plants. For this reason, hpt and hygromycin are now generally not used widely as selection system in maize and no hpt containing maize transgenic has been released ever.

The use of herbicide resistance offers dual advantage of being a potent selectable marker as well as an important agronomic trait. The widely used selectable markers- bar and pat genes, isolated from Streptomyces hygroscopicus and Streptomyces viridochromogenes, respectively, both encode phosphinothricin acetyltransferase (PAT). PAT eliminates herbicidal activity of glufosinate (phosphinothricin) herbicides by acetylation. The bar gene was the first herbicide selectable marker gene which was used in selection of transformed maize cells. In maize, selections of callus using bar and pat selectable marker have been found to be more efficient than kanamycin. So far, in maize, 83 events have been commercialized with pat gene, five events with a synthetic (syn) version of pat gene and seven events with bar gene. Glyphosate resistance is an important trait for the control of broad spectrum weeds. The different forms of 5-enolpyruvoylshikimate-3-phosphate synthase (epsps) genes have been commonly used for glyphosate resistance. The epsps gene from three sources have been used in maize genetic engineering- the most widely used CP4epsps (aroA:CP4) gene obtained from Agrobacterium tumefaciens strain CP4; the mutated mepsps and the double mutant 2mepsps gene obtained from maize itself and the epsps grg23ace5 gene which was chemically synthesized based on the sequence of epsps grg23 gene from soil bacterium Arthrobacter globiformis. Apart from epsps, glyphosate oxidoreductase (gox) gene- goxv247 obtained from Ochrobactrum anthropi strain LBAA and glyphosate N-acetyltransferase (gat) gene- gat4621 obtained from Bacillus licheniformis can be used to detoxify glyphosate. More than 90 events of transgenic maize containing glyphosate resistance have been released. Improved gat marker gave the transformation frequency of 64% (McCutchen et al., 2007). Mutants of acetolactate synthase (als), also known as acetohydroxy acid synthase (ahas), confer resistance in transformants against ALS inhibitor family of herbicides such as sulfonylurea and imidazolinone (Le et al., 2010). In maize, mutants of als/ahas have been successfully used as selectable markers (Bernasconi et al., 1995). Protoporphyrinogen (ppo) gene is another robust selectable marker, coding for a double mutant which are resistant to butafenacil, which inhibits the activity of PPO enzyme which in turn results in protoporphyrin IX mediated light-dependent membrane damage (Li et al., 2003). Another gene, aad-1 (also known as Rdp A gene) isolated from Sphingobium herbicidivorans encodes aryloxyalkonate dioxygenase which cleaves aryloxy phenoxypropionate (AOPP) herbicides specifically inhibiting the monomeric acetyl-CoA carboxylases from monocots. The aad-1 gene has been efficiently used as selectable marker for maize transformants (Wright et al., 2010).

Mannose-6-phosphate isomerase (PMI) encoding gene, manA allowed selection of maize transformants on mannose containing media (Negrotto et al., 2000; Wang et al., 2000). PMI have higher transformation frequency and powerful selection. Another sugar metabolism based positive selection system based on xyl A gene, similar to PMI, encoding xylose isomerase have been tested in maize and in this system, xylose is used as selection agent (Guo et al., 2007).

Huang et al. (2004) emphasized many variations in the approach for generating marker free transgenics. A strategy referred to as 2T-DNA transformation involves placing the selection marker and the gene(s) of interest on two separate T-DNAs. As the genes of interest and the selection markers are physically separate, their transfer and integration to the plant chromosome are mutually independent. In the cells that have received both the genes, the marker gene can be expected to integrate at an independent site from that of the gene of interest. The plants produced from such cells would produce segregant progenies in the next generation that may be free from the selectable marker but contain the gene of interest.

In cisgenesis, the complete coding sequence (CDS) including introns of a gene originating from the sexually compatible gene pool of the recipient plant along with gene’s own promoter and terminator are used for transformation (Schouten et al., 2006). In this case, the cisgene should be used in its normal sense orientation only. In intragenesis, the full or partial CDS of genes originating from the sexually compatible gene pool of the recipient plant can be used in sense or antisense orientation. In this case, the promoter and terminator could originate from sexually compatible gene pool of the recipient plant and not necessarily from the ‘cisgene’ itself (Rommens et al., 2004). Since, both cisgenics and intragenics would be essentially “marker free,” transformation strategies directed at recovering marker-free plants, had to be essentially employed. In both these cases, particle bombardment using ‘clean DNA transformation’ may be employed. So far, there are no reports of cisgenic or intragenic maize under commercialization or advanced development.

Targeted genome modifications or gene editing with the help of site-specific nucleases (SSNs) have the potential to avoid many regulatory issues regarding transgenics. Site-specific nucleases include Zinc-finger nucleases (ZFNs), Transcription activator-like effector nucleases (TALENs), and Clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated (Cas). ZFNs and TALENs are artificial proteins composed of a specific DNA-binding domain and DNA cleavage domain. In these approaches, double strand breaks (DSBs) are introduced at targeted sites in the DNA. The DSBs are immediately repaired by two mechanisms, viz. non-homologous end joining (NHEJ), which is a type of error prone repair and by Homology direct repair (HDR). With the help of these molecular repair processes, researchers have been able to disrupt specific genes either by inserting exogenous DNA elements into desired genomic sites or by introducing single-nucleotide substitutions. CRISPR together with Cas proteins form CRISPR-Cas system, is the newest genome modification technique. ZFN method has been used to modify endogenous loci in crop plants, like maize. The CRISPR/Cas9 system has been used for gene editing in crop plants like, rice, wheat, sorghum, tomato, tobacco, etc.; with few examples in case of maize as well (Feng et al., 2016). Transformation for gene editing may require co-bombardment of separate DNA vectors containing Cas9 and gRNA. However, stable integration and constitutive expression of gRNAs and Cas9 might lead to somatic mutations and generation of chimeric plants. In this scenario, gRNA may be delivered in form of in vitro synthesized RNA molecule, together with Cas9 as DNA construct. Bombardment of RNA molecules posses further challenges of optimization of particle preparation and gene gun operation protocols. Genetic transformation usually involves transgene integration into the host genome. However, introduction of genes without genomic integration is more desirable for HDR, and other transient expression requiring genome editing tools. Bombardment of single-stranded DNA has been used as one of the approach to evade template integration during HDR-mediated genome editing in maize (Svitashev et al., 2015). Refinements in Agrobacterium-mediated transformation and further development of RNA viruses and geminiviruses based transformation techniques may result in gene transfer protocols with superior genome editing properties (Altpeter et al., 2016). In targeted genome modification by SSNs, modifications performed in organisms’ endogenous gene to develop desired traits do not employ transgene of other species or organism. Targeted genome modified products are essentially mutation based products. Therefore, plants developed using these technologies may be treated at par with mutation breeding products and could be regulated as such. This would ensure faster commercialization and easier availability of this technology for the welfare of the farmers.