a) Obtaining Embryogenic Calli

Young stems of plants from the nursery aged from 25 to 30 days were used. From these stems, the external and mature leaves had been removed to leave a section approximately 5 cm long from the last node, wrapped in the youngest leaves. This material was washed with 20% soapy solution, and disinfected with 1% sodium hypochlorite for 20 minutes (^{Marcano et al. 2002}). Two washes with distilled water were carried out in a laminar flow cabinet, each one lasting ten minutes. Then, a last layer of leaves was removed to obtain a stem with a diameter of approximately 1 cm. The explants were divided at a distance of 2 cm to 5 cm from the shoot apex in the form of 2 mm-thick discs for a total of 30 discs per variety. These discs were planted at the ratio of six explants per flask with induction medium, comprised of: complete Murashige and Skoog (1962) salts, 100 mg/l inositol, 1 mg/l thiamine, 0.5 mg/l pyridoxine, 0.5 mg/l nicotinic acid, 2 mg/l glycine, 50 mg/l arginine, 50 mg/l cysteine, 0.15 mg/l citric acid, 30 g/l sucrose, 50 ml/l coconut water, and 8 g/l agar, supplemented with 6.63 mg/l 3,6 dichloro2-methoxybenzoic acid (dicamba) or 3 mg/l 2,4-dichlo-rophenoxyacetic acid (2,4-D), adjusted to pH 5.7 with solutions of 0.1 N NaOH and 1 N HCl. They were cultivated in a room in complete darkness at 25 "C for one month (^{Marcano et al 2002}).

b) Regeneration of Vitro Seedlings

For the development of seedlings, the embryogenic calli were subcultured in a regeneration medium, which is comprised as follows: half Murashige and Skoog (1962) salts, 100 mg/l inositol, 1 mg/l thiamine, 60 g/l sucrose, 200 ml/l coconut water, and 8 g/l agar, adjusted to pH 5.7 with solutions of 0.1 N NaOH and 1 N HCl, and they were incubated in conditions of continuous light in a room at 25 "C for two months (^{Marcano et al. 2002}). Five flasks were planted each with six calli for a total of 30 calli for each variety.

c) Statistical Analysis

The multifactor analysis of variance was used on the interaction between the varieties of sugarcane and the hormones applied with the statistical program Stat-graphics Centurion 12®, with dependent variables (percentage of callus formation and percentage of vitro seedling formation) and independent variables (hormones, varieties and replicas). The aim was to evaluate whether there are significant differences between the percentages of embryogenic callus formation and between the percentages of development of vitro seedlings from the somatic embryos obtained. To identify which of the sugarcane varieties and which of the hormone treatments show significant differences, the comparison of means was applied using Tukey's range test with a confidence level of 95%.

a) Obtaining Embryogenic Calli

Different genetic and physiological factors influence the induction of somatic embryogenesis, leading to the replacement of the gene expression pattern existing in the explant's tissue with a new expression of embryogenic genes. This process is only possible if the cells are competent and receive the appropriate inducing stimuli. Among the influencing factors, plant hormones, especially auxins, are key components, because their exogenous application recapitulates the embryogenic potential of the mitotically inactive somatic cells. Several studies have indicated that signposting by auxins, as well as stress, can be key events in genetic reprogramming of somatic cells during early embryogenesis, which is necessary for de-differentiation and to acquire the embryogenic condition (^{Silveira et al. 2013}).

The different stages of the process of somatic embryo-genesis in the V99245 sugarcane variety can be observed in Figure 1, representing all the varieties studied in this work. In Part A of the figure, the inoculated leaf discs can be observed in the callus induction medium. After 25 days, in Part B of the figure, the formation of a compact callus can be observed with a whitish color; typical characteristics of the embryogenic callus of sugarcane (^{Oropeza et al. 2001}). ^{Taylor et al. (1992}) provide a precise description of the different types of callus that originate from segments of young sugarcane leaves: Type I appears first, after five days of culture at the ends of the cutting and it is characterized by being semi-translucent and consisting of long, large cells. Type II is developed in the tissue of Type I after 13 days of culture, and it is smooth, slimy, yellow and contains highly disassociated round, vacuolated cells. Type III appears after 13 to 21 days at the ends of the explants around the vascular tissue, consisting of a compact, hard callus with globular surface nodules, which is yellowish at the start but then turns a whitish color and is comprised of small, round cells with thick cytoplasm. Type IV is yellow, friable, presents agglomerates of semi-organized structures, and consists of small, round cells with thick cytoplasm. Type I and Type II calli are non-morphogenic, while Type III and IV calli can regenerate plants at a high frequency through somatic embryogenesis, and at a low frequency through organogenesis, respectively. The Type III callus is referred to as an embryogenic callus and the heterogeneity is due to the explant coming from the transversal section of different leaves with different degrees of maturity. ^{Guiderdoni and Demarly (1988)} described only two types of callus: the friable callus and the compact callus, which apparently coincide with Type I and Type III of ^{Taylor et al. (1992)}. Similar results were obtained by ^{Oropeza et al. (2001)} and by ^{Marcano et al. (2002)}, only observing the compact and friable calli in accordance with the description by ^{Guiderdoni and Demarly (1988)}. ^{Silveira et al. (2013)} studied the differences in the acquisition of competencies and the later maturation of the somatic embryo in embryogenic calli and non-embryogenic calli of the SP791011 variety of sugarcane through histological analysis, growth curves, number of somatic embryos, and polyamide content. The embryogenic calli could be distinguished from the non-embryogenic calli because they presented round tissue with a dry and compact aspect, and an opaque, whitish color with less oxidation, as well as the appearance of green spots after several days; while the non-embryogenic calli presented long, smooth and friable cells, were mucilaginous, and had a translucent brown color due to oxidation in the presence of light.

Figure 1 Induction of somatic embryogenesis with dicamba in the V99205 variety of sugarcane. A) Discs of surrounding leaves at the start of induction. B) Callus generated after 25 days of culture in induction medium. C) Calli showing embryogenic cell aggregates after 32 days of incubation in induction medium (110X). 

In Figure 1C, the aggregates of embryos formed from the embryogenic calli are observed 32 days after planting, showing the characteristic whitish color. The small amount of non-embryogenic callus obtained presented a smooth and friable aspect, showing cell oxidation with translucent brown coloring. The embryogenic calli showed a more compact aspect, and an opaque and whitish coloring with numerous somatic embryos in their periphery, which after five days started to germinate, observing a large amount of stems and roots. The rest of the varieties showed the same type of growth. Although no histological study was conducted to determine the embryogenic origin of the vitro seedlings, the culture conditions and the morphology of the callus obtained were similar to those reported by ^{Oropeza et al. (2001)} and by ^{Marcano et al. (2002)}, who through an anatomical study, corroborated the observations at a morphological level in the embryogenic calli. Therefore, given the morphological characteristics of the calli obtained in this research and the high frequency of regeneration of seedlings from these calli, it is very probable that the main process of in vitro regeneration that is occurring is embryogenic.

In Table 1, the percentages of formation of embryogenic calli for each variety are observed with both hormone treatments, which show significant differences between 2,4-D and dicamba. In terms of the response by variety, a particularly low response is observed for the RB855546 variety with both treatments, which corroborates the fact that every embryogenic response depends on the genotype. In general, for the V0050 variety, the formation of embryogenic calli is greater with 2,4-D. A very similar response was observed between the varieties CC8475, V756 and V781, but the response oscillates between 40 and 90%, all with a greater response to dicamba; just like between the varieties CC8592 and V99245. These last two varieties have the highest response for the formation of embryogenic calli with both treatments.

Table 1 Percentages of embryogenic callus formation of each sugarcane variety. 

Comparison of means with Tukey's range test with a confidence level of 95%. Letters a and b indicate highly significant differences between the hormone treatments. Letters c, d, e and f represent highly significant differences between the varieties' responses in callus formation. Varieties with the same letter do not present significant differences between them.

The start and maintenance of embryogenic callus is an important step for the development of somatic embryos and regeneration of seedlings (^{Silveira et al. 2013}). Traditionally as well as currently, somatic embryogenesis is induced with 2,4-D in many species of gramineae. For the first time, ^{Ho and Vasil (1983)} reported that calli were formed in the cut ends of monocotyledon explants cultivated in Murashige and Skoog (1962) medium supplemented with 2,4-D (^{Guiderdoni and Demarli, 1988}; Guiderdoni et al. 1995; ^{Suprasanna et al. 2005}; ^{Lakshmanan et al. 2005}; ^{Raza et al. 2012};

^{Taparia et al. 2012}; ^{Kaur and Kapoor, 2015}). In our research, the induction of embryogenic calli and somatic embryos was slightly greater in the explants cultivated in media supplemented with dicamba (70.83% of embryogenic callus per variety), although the majority of the varieties responded to 2,4-D (62.08% of embryogenic callus per variety). The percentage of embryogenic callus formation of each variety for each treatment is observed in detail in Table 1. Rapid development of embryogenic callus was observed during the induction phase under the test conditions in culture media supplemented with dicamba, a compound that is used in field conditions as a selective herbicide. Its mode of action is to modify the transport of indole-3-acetic acid, destroying the cambium and the parenchyma at the height of the nodes or above them. However, its chemical structure is that of an artificial auxin, which when used for in vitro cultures of sugarcane, permits the anaplasia of the tissue in the discs of surrounding leaves.

The use of explants of young leaves in our research has been validated in different protocols established for the in vitro morphogenesis of sugarcane using explants of different parts of the plant, including for example, meristems, whorls of internal leaves, segments of inflorescences, immature inflorescences, segments of young leaves, and discs of immature leaves. However, the most used one is the disc of immature, surrounding leaves close to the apical meristem inoculated horizontally, with its distal ends in contact with the medium, because the best results in the production of somatic embryos have been achieved with this kind of explant (^{Guiderdoni et al. 1995}; ^{Oropeza et al. 2001}; ^{Marcano et al. 2002}; ^{Suprasanna et al. 2011}; ^{Joshi et al. 2013}). ^{Kaur and Sandhu (2015)} established an efficient system for the direct production of adventitious buds, also from the discs of young leaves, demonstrating the potential of this explant for mass industrial production of sugarcane plants.

^{Manchanda and Gosal (2012)} studied the effect of activated carbon, the sources of carbon and gelling agents on the somatic embryogenesis of sugarcane, and largely recommend the addition of activated carbon to the culture medium to reduce the blackening of the tissue by the exudation of phenolic compounds, and therefore, increase the embryogenic potential of the tissue. Additionally, they recommend supplementing the medium with maltose, although the use of sucrose continues to provide good results. Finally, they recommend gelling the culture media with agar, because they observed that with its use, the formation of somatic embryos increased, as well as the regeneration frequency of sugarcane compared to Gelrite. Blackening was not observed, so it was not necessary to use activated carbon. The small amount of oxidation observed was controlled with the addition of cysteine to the culture medium.

^{Suprasanna et al. (2005)} found that embryogenic callus formation was higher with the combination of 2,4-D and zeatin in the Co91010 variety, while for the CoC671 variety, this combination was made efficient by adding proline to the culture medium. Upon analyzing the response obtained in our research, particularly with the RB855546 variety, it could be recommended to supplement the M&S (1962) salts used for inducing embryogenic calli (which contain the amino acids arginine and cysteine) with proline in order to improve the response of this variety. Furthermore, no synthetic cy-tokinins were used in our research in combination with the auxins tested for inducing the process of somatic embryogenesis. Apparently, the amount of zeatin that the coconut water contains is sufficient to induce the processes of cell division and elongation, which in combination with the auxins, provide the cytokinins during this process.

^{Marcano et al. (2002)} optimized the process of somatic embryogenesis in Venezuelan varieties of sugarcane, using the synthetic auxins 2,4-D and dicamba, both of which are substances used in the field as herbicides, because they affect the transport of natural auxins and cell elongation. They demonstrated that after 45 days in the medium supplemented with dicamba, the highest percentage of embryogenic callus formation was observed for the tested Venezuelan varieties V756 and V781.

As previously mentioned, numerous authors have reported that somatic embryogenesis in sugarcane is induced by the auxin 2,4-D (^{Ho and Vasil, 1983}; ^{Brisibe et al. 1994}; ^{Falco et al. 1996}; Oropeza and García, 1996; Oropeza et al. 2001; ^{Watt et al. 2009}; ^{Snyman et al. 2011}). However, in the majority of the varieties analyzed in this research, the embryogenic callus percentage obtained is greater when dicamba is used. This may be due to the structural difference of the molecules of 2,4-D and dicamba, as the latter can be absorbed and degraded by the plant tissue more quickly.