**3. Somatic embryogenesis and transformation**

Somatic embryogenesis is a process by which a plant somatic cell develops into a whole plant without gametic fusion but undergoes developmental changes as that of zygotic em‐ bryogenesis [71, 72]. The first demonstration of *in vitro* somatic embryogenesis was reported in *Daucus carota* by Reinert [73]. The concept of embryogenesis has drawn a lot of attention because of its significance in theory and practice. Primarily, somatic embryos can be pro‐ duced easily and quickly, so that it provides an economical and easy way to study plant de‐ velopment. Secondly, synthetic seeds developed from somatic embryos open the possibility of developing high quality seeds and may allow us to produce seeds from those plants that require a long period for seed production. Somatic embryogenesis is also useful in plant ge‐ netic engineering since regeneration *via* somatic embryogenesis is frequently single of cell origin, resulting in a low response of chimeras and high a number of true transgenic regen‐ erants [74, 75].

#### **3.1. Somatic embryogenesis**

The first record of soybean somatic embryogenesis was reported by Beversdorf & Bingham [76], followed by Christianson et al. [77] who regenerated plants through the method. The immature cotyledon is the preferred explant for soybean somatic embryogenesis as it has pre-determined embryogenic cells. Somatic embryogenesis is a multi-step regeneration process starting with the formation of proembryogenic cell mass, followed by somatic em‐ bryo induction, their maturation, desiccation and finally plant regeneration [78].

Soybean somatic embryos were induced from immature cotyledon explants cultured on me‐ dium containing high levels of 2,4-D [79]. Even though NAA induced somatic embryogene‐ sis from immature cotyledons, the mean number of embryos produced on 2,4-D was significantly higher [80]. Explant orientation, pH, solidifying agent, and 2,4-D concentration have a synergic effect on somatic embryo induction [81]. The early-staged somatic embryos can be maintained and proliferated by subculturing the tissue on either semi-solid medium [79] or liquid suspension culture medium [82]. Somatic embryos incubated in a medium containing NAA do not proliferate so well as those produced on a medium containing 2,4-D [83]. Somatic embryos initiated on NAA are more advanced in embryo morphology than those induced on 2,4-D and the efficiency of somatic embryo induction was highest with a medium containing 2-3% sucrose. Cultures initiated on lower sucrose concentrations tended to produce a higher amount of friable embryos, while increased concentrations of this sugar impaired embryo induction [80,84-86]. Histodifferentiation and maturation of somatic em‐ bryos doesn't need exogenous auxin or cytokinins [87]. Indeed, poorly developed meristem or swollen hypocotyls may be an undesired outcome of the application of exogenous auxins and cytokinins, respectively. Moon and Hildebrand, [88] investigated the effects of prolifera‐ tion, maturation, and desiccation methods on conversion of soybean somatic embryos to plants. Somatic embryos proliferated on solid medium showed a higher regeneration rate when compared with the embryos proliferated in liquid medium. The growth period of so‐ matic embryo development can be reduced one month by culturing in a medium devoid of 2,4-D and B5 vitamins. Carbon source is critical for embryo nutritional health and improves somatic embryo maturation. The effects of carbohydrates on embryo histodifferentiation and maturation on liquid medium were analyzed by Samoylov et al. [89]. FNL medium sup‐ plemented with 3% sucrose (FNL0S3) or 3% maltose (FNL0M3) were compared. Data indi‐ cated that sucrose promotes embryo growth and significantly increases the number of cotyledon-stage embryos recovered during histodifferentiation and maturation. However, the percentages of plants recovered from embryos differentiated and matured in FNL0S3 was lower than those grown in FNL0M3 (Samoylov et al. 1998b). The quality of somatic em‐ bryos can be positively influenced by a low osmotic potential in maturation medium [90, 91]. Carbohydrates can act as an osmotic agent. Polyethylene glycol 4000, mannitol and sor‐ bitol were tested as supplements to a liquid Finer and Nagasawa medium-based histodiffer‐ entiation/maturation medium FNL0S3, for soybean (*Glycine max* L. Merrill) somatic embryos of 'Jack' and F138 or 'Fayette'[90]. Overall, 3% sorbitol was found to be the best of the os‐ motic supplements tested. The ability of histodifferentiation and conversion of somatic em‐ bryo have been improved by the use of ethylene inhibitor aminoethoxyvinylglycine [92]. The effects of ethylene on embryo histodifferentiation and conversion were genotype-specif‐ ic. The germination frequency of soybean embryos is very low [93], and therefore, partial desiccation of somatic embryos was emphasised with a view to improving the germination frequency in soybean [87,94&95]. Desiccation induced a physiological state there by increase the germination ability of somatic embryos [87].

### **3.2. Genotype**

**3.1. Somatic embryogenesis**

Relationships

428

The first record of soybean somatic embryogenesis was reported by Beversdorf & Bingham [76], followed by Christianson et al. [77] who regenerated plants through the method. The immature cotyledon is the preferred explant for soybean somatic embryogenesis as it has pre-determined embryogenic cells. Somatic embryogenesis is a multi-step regeneration process starting with the formation of proembryogenic cell mass, followed by somatic em‐

A Comprehensive Survey of International Soybean Research - Genetics, Physiology, Agronomy and Nitrogen

Soybean somatic embryos were induced from immature cotyledon explants cultured on me‐ dium containing high levels of 2,4-D [79]. Even though NAA induced somatic embryogene‐ sis from immature cotyledons, the mean number of embryos produced on 2,4-D was significantly higher [80]. Explant orientation, pH, solidifying agent, and 2,4-D concentration have a synergic effect on somatic embryo induction [81]. The early-staged somatic embryos can be maintained and proliferated by subculturing the tissue on either semi-solid medium [79] or liquid suspension culture medium [82]. Somatic embryos incubated in a medium containing NAA do not proliferate so well as those produced on a medium containing 2,4-D [83]. Somatic embryos initiated on NAA are more advanced in embryo morphology than those induced on 2,4-D and the efficiency of somatic embryo induction was highest with a medium containing 2-3% sucrose. Cultures initiated on lower sucrose concentrations tended to produce a higher amount of friable embryos, while increased concentrations of this sugar impaired embryo induction [80,84-86]. Histodifferentiation and maturation of somatic em‐ bryos doesn't need exogenous auxin or cytokinins [87]. Indeed, poorly developed meristem or swollen hypocotyls may be an undesired outcome of the application of exogenous auxins and cytokinins, respectively. Moon and Hildebrand, [88] investigated the effects of prolifera‐ tion, maturation, and desiccation methods on conversion of soybean somatic embryos to plants. Somatic embryos proliferated on solid medium showed a higher regeneration rate when compared with the embryos proliferated in liquid medium. The growth period of so‐ matic embryo development can be reduced one month by culturing in a medium devoid of 2,4-D and B5 vitamins. Carbon source is critical for embryo nutritional health and improves somatic embryo maturation. The effects of carbohydrates on embryo histodifferentiation and maturation on liquid medium were analyzed by Samoylov et al. [89]. FNL medium sup‐ plemented with 3% sucrose (FNL0S3) or 3% maltose (FNL0M3) were compared. Data indi‐ cated that sucrose promotes embryo growth and significantly increases the number of cotyledon-stage embryos recovered during histodifferentiation and maturation. However, the percentages of plants recovered from embryos differentiated and matured in FNL0S3 was lower than those grown in FNL0M3 (Samoylov et al. 1998b). The quality of somatic em‐ bryos can be positively influenced by a low osmotic potential in maturation medium [90, 91]. Carbohydrates can act as an osmotic agent. Polyethylene glycol 4000, mannitol and sor‐ bitol were tested as supplements to a liquid Finer and Nagasawa medium-based histodiffer‐ entiation/maturation medium FNL0S3, for soybean (*Glycine max* L. Merrill) somatic embryos of 'Jack' and F138 or 'Fayette'[90]. Overall, 3% sorbitol was found to be the best of the os‐ motic supplements tested. The ability of histodifferentiation and conversion of somatic em‐ bryo have been improved by the use of ethylene inhibitor aminoethoxyvinylglycine [92].

bryo induction, their maturation, desiccation and finally plant regeneration [78].

Soybean somatic embryogenesis is highly genotypic when compared to organogenesis. The existence of strong genotype specificity in the regeneration capacity of the different cultivars represents a major limiting factor for the advancement of soybean biotechnology. The em‐ bryogenic efficiency of soybean was shown to be different among cultivars at each stage (in‐ duction, proliferation, maturation, germination) of somatic embryogenesis [92,96] and it is very challenging to identify genotypes highly responsive to all stages. Simmonds and Do‐ naldson, [97] screened 18 short season soybean genotypes for proliferative embryogenesis. Five genotypes produced embryogenic cultures which were proliferative for at least 6 months. Yang et al. [98] screened 98 Chinese soybean varieties for somatic embryogenesis and selected 12 varieties based on their embryogenic capacity. The greatest average number of plantlets regenerated per explant (1.35) was observed in N25281. Bonacin et al. [99] dem‐ onstrated the influence of genotype on somatic embryogenic capability of five Brazilian cul‐ tivars. Droste et al. [100] reported somatic embryo induction, proliferation and transformation of commercially grown Brazilian soybean cultivars for the first time. Soy‐ bean somatic embryo conversion is genotype dependent; germination frequency of H7190 was approximately three fold lower than that of PI 417138 [101]. Hiraga et al. [102] exam‐ ined the capacity for plant regeneration through somatic embryogenesis in Japanese soy‐ bean cultivars and identified Yuuzuru and Yumeyutaka as having high potential for somatic embryogenesis. Several cultivars were identified as uniformly embryogenic at the primary induction phase at all locations, among which Jack was the best [103]. Kita et al. [104] evalu‐ ated somatic embryogenesis, proliferation of embryogenic tissue, and regeneration of plant‐ lets in backcrossed breeding lines derived from cultivar Jack and a breeding line, QF2. The backcrossed breeding lines exhibited an increased capacity for induction and proliferation of somatic embryos and were used successfully to generate transgenic plants.

### **3.3.** *Agrobacterium* **mediated transformation**

Recovery of the first transgenic plant *via* somatic embryogenesis in soybean was reported by Parrott et al. [105]. Immature cotyledon tissues were inoculated with *Agrobacterium* strain which contained 15 kD zein gene and the neomycin phosphotransferase gene. The explants were placed on medium containing high auxin for somatic embryo induction. Three transgenic plants containing the introduced 15 kD zein gene were regenerated. Un‐ fortunately, these plants were chimeric and the 15 kD zein gene was not transmitted to the progeny. Sonication-assisted *Agrobacterium*-mediated transformation (SAAT) of imma‐ ture cotyledons tremendously improves the efficiency of *Agrobacterium* infection by intro‐ ducing large numbers of micro wounds into the target plant tissue [48]. The highest GUS

expression was obtained when immature cotyledons were sonicated for 2s in the pres‐ ence of *Agrobacterium* followed by co-cultivation for 3 days. Trick and Finer, [108] success‐ fully employed Sonication-assisted *Agrobacterium*-mediated transformation of embryogenic suspension culture tissue and when SAAT was not used, no transgenic clones were ob‐ tained. Yan et al. [109] demonstrated the feasibility of *Agrobacterium* mediated transforma‐ tion of cotyledon tissue for the production of fertile transgenic plants by optimising the *Agrobacterium* concentration, using co-cultivation time and selecting proper explant. Ko and Korban, [110] investigated optimal conditions for induction of transgenic embryos fol‐ lowed by *Agrobacterium* mediated transformation. Using cotyledon explants from imma‐ ture embryos of 5-8mm length, a 1:1 (v/v) concentration of bacterial suspension and 4 day co-cultivation period significantly increased the frequency of transgenic somatic embryos. The *Agrobacterium* tumefaciens strain KYRT1 harboring the virulence helper plasmid pKYRT1 induces transgenic somatic embryos at a high frequency from infected immature soybean cotyledons [111]. Recently, the successful recovery of a high number of soybean transgenic fertile plants was obtained from the combination of DNA- free parti‐ cle bombardment and *Agrobacterium*-mediated transformation using proliferating soybean somatic embryos as targets [112].

#### **3.4. Particle bombardment**

Particle bombardment is a widely used technique for transformation of embryogenic cul‐ tures of soybean; the major advantage of this technique over *Agrobacterium* is the removal of biological incompatibilities. Particle bombardment in soybean was first reported by Fi‐ ner and McMullen [64], in which embryogenic suspension culture tissue of soybean was bombarded with particles coated with plasmid DNAs encoding hygromycin resistance and β-glucuronidase. Analysis of DNA from progeny plants showed genetic linkage for multiple copies of introduced DNA. Using particle bombardment, fertile plants could be routinely produced from the proliferating transgenic embryogenic clones. Hazal et al. [113] studied growth characteristics and transformability of embryogenic cultures and found that cultures bombarded between 2-6 days after transfer to fresh medium showed more transient expression of the reporter gene. Histological analysis showed that the most transformable cultures had cytoplasmic-rich cells in the outermost layers of the tissue. Maughan et al. [114] bombarded embryogenic cultures with plasmid containing 630-bp DNA fragment encoding a bovine milk protein, β-casein. Hadi et al. [115] co-transformed 12 different plasmids into embryogenic suspension culture by particle bombardment. Hy‐ bridization analysis of hygromycin resistance clones verified the presence of introduced plasmid DNAs. Santarem and Finer [116] investigated the effect of desiccation of target tissue, period of subculture prior to bombardment and number of bombardments per tar‐ get tissue for enhancement of transient expression of the reporter gene. Desiccation of proliferating tissue for 10 min, subculture on the same day prior to bombardment and three times bombardment on a single day enhanced the transient expression of β-glucuro‐ nidase [116]. Dufourmantel et al. [117] successfully transformed chloroplasts from em‐ bryogenic tissue of soybean using DNA carrying spectinomycin resistance gene (*aadA*) by bombardment. All transplastomic T0 plants were fertile and T1 progeny was uniformly spectinomycin resistant, showing the stability of the plastid transgene. *Droste* et al. [100] successfully transformed embryogenic cultures of soybean cultivars recommended for commercial growing in South Brazil by bombardment, and this opened the field for the improvement of this crop in this country by genetic engineering.

### **3.5. Genes for trait improvement**

expression was obtained when immature cotyledons were sonicated for 2s in the pres‐ ence of *Agrobacterium* followed by co-cultivation for 3 days. Trick and Finer, [108] success‐ fully employed Sonication-assisted *Agrobacterium*-mediated transformation of embryogenic suspension culture tissue and when SAAT was not used, no transgenic clones were ob‐ tained. Yan et al. [109] demonstrated the feasibility of *Agrobacterium* mediated transforma‐ tion of cotyledon tissue for the production of fertile transgenic plants by optimising the *Agrobacterium* concentration, using co-cultivation time and selecting proper explant. Ko and Korban, [110] investigated optimal conditions for induction of transgenic embryos fol‐ lowed by *Agrobacterium* mediated transformation. Using cotyledon explants from imma‐ ture embryos of 5-8mm length, a 1:1 (v/v) concentration of bacterial suspension and 4 day co-cultivation period significantly increased the frequency of transgenic somatic embryos. The *Agrobacterium* tumefaciens strain KYRT1 harboring the virulence helper plasmid pKYRT1 induces transgenic somatic embryos at a high frequency from infected immature soybean cotyledons [111]. Recently, the successful recovery of a high number of soybean transgenic fertile plants was obtained from the combination of DNA- free parti‐ cle bombardment and *Agrobacterium*-mediated transformation using proliferating soybean

A Comprehensive Survey of International Soybean Research - Genetics, Physiology, Agronomy and Nitrogen

Particle bombardment is a widely used technique for transformation of embryogenic cul‐ tures of soybean; the major advantage of this technique over *Agrobacterium* is the removal of biological incompatibilities. Particle bombardment in soybean was first reported by Fi‐ ner and McMullen [64], in which embryogenic suspension culture tissue of soybean was bombarded with particles coated with plasmid DNAs encoding hygromycin resistance and β-glucuronidase. Analysis of DNA from progeny plants showed genetic linkage for multiple copies of introduced DNA. Using particle bombardment, fertile plants could be routinely produced from the proliferating transgenic embryogenic clones. Hazal et al. [113] studied growth characteristics and transformability of embryogenic cultures and found that cultures bombarded between 2-6 days after transfer to fresh medium showed more transient expression of the reporter gene. Histological analysis showed that the most transformable cultures had cytoplasmic-rich cells in the outermost layers of the tissue. Maughan et al. [114] bombarded embryogenic cultures with plasmid containing 630-bp DNA fragment encoding a bovine milk protein, β-casein. Hadi et al. [115] co-transformed 12 different plasmids into embryogenic suspension culture by particle bombardment. Hy‐ bridization analysis of hygromycin resistance clones verified the presence of introduced plasmid DNAs. Santarem and Finer [116] investigated the effect of desiccation of target tissue, period of subculture prior to bombardment and number of bombardments per tar‐ get tissue for enhancement of transient expression of the reporter gene. Desiccation of proliferating tissue for 10 min, subculture on the same day prior to bombardment and three times bombardment on a single day enhanced the transient expression of β-glucuro‐ nidase [116]. Dufourmantel et al. [117] successfully transformed chloroplasts from em‐ bryogenic tissue of soybean using DNA carrying spectinomycin resistance gene (*aadA*) by bombardment. All transplastomic T0 plants were fertile and T1 progeny was uniformly

somatic embryos as targets [112].

**3.4. Particle bombardment**

Relationships

430

Li et al. [6] attempted to transform two antifungal protein genes (*chitinase* and ribosome-in‐ activating protein) by co-transformation. Transgenic soybeans expressing the Yeast SLC1 Gene showed higher oil content [118]. They reported that, compared to controls, the average increase in triglyceride values went up by 1.5% in transgenic somatic embryos and also found that a maximum of 3.2% increase in seed oil content was observed in a T3 line. Trans‐ fer of Δ6 desaturase, fatty acid elongase and D5 desaturase into soybean under seed specific expression produced arachidonic acid (ARA) in seeds of soybean [119]. In an attempt to en‐ hance soybean resistance to viral diseases, several groups successfully generated transgenic plants by expressing an inverted repeat of soybean dwarf virus SbDV coat protein (*CP*) genes [120], or soybean mosaic virus (SMV) coat protein gene [121]. The nutritional quality of soybean has been improved for enhanced amino acid, proteins and vitamin production by transgenic technology [114, 122, 123, 124, and 125]. The feasibility of genetically engineer‐ ing soybean seed coats to divert metabolism towards the production of novel biochemicals was tested by transferring the genes phbA, phbB, phbC from *Ralstonia eutropha*. Each gene was under the control of the seed coat peroxidase gene promoter [126]. The analysis of seed coats demonstrated that polyhydroxybutyrate (PHB) was produced at an averge of 0.12% seed coat dry weight.

### **4. Conclusion and future prospects**

As demands increase for soybean oil and protein, the improvement of soybean quality and production through genetic transformation and functional genomics becomes an important issue throughout the world. Modern genetic analysis and improvement of soybean heavily depend on an efficient regeneration and transformation process, especially commercially important genotypes. The transformation techniques developed until now till date do not al‐ low high-throughput analyses in soybean functional genomics; though significant improve‐ ments have been made in the particle bombardment of embryogenic culture and *Agrobacetrium* mediated transformation of the cotyledonary node over the past three deca‐ des. However, routine recovery of transgenic soybean plants using either of these two trans‐ formation systems has been restricted to a few genotypes with no reports of transformation on other locally available commercial genotypes. Therefore, development of an efficient and consistent transformation protocol for other locally available commercial genotypes, will greatly aid soybean functional genomics and transgenic technology.
