ΔCW = + 1.43078–0.22065*(2,4-D) + 0.056268*(IBA) + 0.010482*(2,4-D)*(IBA) + 0.012465*(2,4-D)2 -0.00869*(IBA) (4)
The equations 3 and 4 predict the percentage of samples with embryogenic callus and the weight gain on callus produced respectively, based on the values of 2,4-D and IBA; furthermore, the models used a good significance level (p = 0.0005 and 0.004, respectively). The multiple determination coefficients (R2) obtained were 0.8472 and 0.8772 respectively, indicating a satisfactory fitting on the models predicted to the experimental data.
According to Myer and Montgomery (2002), a good prediction model must have a determination coefficient (R2) of at least 0.80 and a significance level (p) lower of 0.05. All these parameters have been used to decide the level of satisfaction of the models.
Induction of sugarcane embryogenic callus has been already reported mainly using 2,4-D, IBA and NAA in a wide range of concentrations and combinations (Bomfim de Alcantara et al. 2014; Arjun and Rao 2015; Solangi et al. 2016; Naz and Hayat 2017; Jamil et al. 2017; Passamani et al. 2020). In this study, a clear tendency to increase the %SEC as the 2,4-D concentration decreased was observed. Meanwhile, in the case of IBA the best results for %SEC were obtained in the concentration range from 2.5 to 5.5 mg L− 1 (Fig. 2A-B); thus, the best treatment to embryogenic callus induction (reaching a value of %SEC = 100.0), confirmed by largest red region in the surface response, was the combination of 0.5 mg L− 1 2,4-D + 5.25 mg L− 1 IBA.
Razdan (2003) suggested the use of 2,4-D in the ranged of 0.5 to 2.5 mg L− 1 for somatic embryogenesis induction in monocotyledons as S. officinarum, because higher doses could inhibit the embryogenic callus induction. Similar results have been reported in sugarcane cultivar CP-841198, where the percentage of embryogenic callus induction gradually decreased as the concentration of 2,4-D was increased (Chengalrayan and Gallo-Meagher 2001; Zamir et al. 2014). Passamani et al. (2019), demonstrated that long-term culture of sugarcane callus with 2.2 mg L− 1 2,4-D decreased gradually the embryogenic competence, through affecting the polyamine metabolism, the regulation of late embryogenesis proteins and the accumulation of putrescine and spermidine as well. Orłowska and Kępczyńska (2020), in a study conducted with Medicago truncatula, reported that higher concentrations of 2,4-D causes an increase in O2 accumulation and antioxidant enzyme activity, altering formation of callus and somatic embryos. In this work, the lowest SEC percentages (10 to 35.7%) were obtained in treatments in which high concentrations of 2,4-D (8.61 and 10.0 mg L− 1) were used (Table 1).
On the other hand, for the ΔCW variable it was observed that combinations of 2,4-D and IBA in the range of 5.25-10.0 mg L− 1 and 0.5–5.25 mg L− 1, respectively, were not adequate for callus growth, corresponding to the lowest values shown in the blue area (Fig. 2C-D). Similar results were reported by Naz and Hayat (2017), obtaining decreases in the weight of induced callus up to 86% less as the concentrations of 2,4-D increased above 3 mg L− 1 in two varieties of sugarcane.
Although there is not a clear effect of the growth regulators in ΔCW, the region of maximum response values was obtained through combinations of 2,4-D in the range of 0.5 mg L− 1 to 1.89 mg L− 1 along with IBA in the range between 4.55 mg L− 1 to 5.25 mg L− 1 (Fig. 2C-D). Van der Vyver (2010) obtained a high rate of embryogenic callus from leaf disks of sugarcane (Black Cheribon genotype) in medium containing 0.5 mg L− 1 2,4-D. Khan et al. (2004) reported that 2.0 mg L− 1 2,4-D, proved to be an excellent dose for embryogenic callus induction and rapid proliferation in ten sugarcane genotypes. Solangi et al. (2016) evaluated the effect of different doses of the auxins 2,4-D, NAA and picloram on young meristems of three varieties of sugarcane, reported that 2,4-D at 3.0 mg L− 1 concentration proved to be the most effective auxin for callus mass induction of all the varieties tested. Moreover, Naz and Hayat (2017) evaluated the effect of several doses of 2,4-D (0.5 to 4 mg L− 1) on three sugarcane varieties, finding that the concentrations of 1.0 and 2.0 mg L− 1 were the best conditions for greatest callus induction, and also reported that variations in 2,4-D concentration could adversely influenced the ratio of callus development and growth.
Differences on embryogenic callus induction through the application of different doses of 2,4-D in sugarcane could be due to several factors such as genotype, explant type, and/or the combination with other phytohormones into the medium (Khan et al. 2004; Dinesh et al. 2017; Getnet 2017; Naz and Hayat 2017).
On the other hand, desirability function-based method was applied for the optimization of embryogenic callus induction, considering the maximum response values for %SEC and ΔCW (Fig. 3). To optimize the concentration of auxins needed to induce embryogenic callus, the desirability value obtained was 0.945, which is close to the maximum possible global desirability (value = 1); being the best predicted response corresponding to 2,4-D (0.5 mg L− 1) and IBA (3.53 mg L− 1); this values correspond to independent variables associated with the maximum overall desirability, where the predicted values of the model were 100.0% of SEC and 1.4323 g of ΔCW (Table 2).
Table 2
Regression coefficients and analyses of variance of the experimental prediction models showing in coded factors the relationship among process variables (X) and response variables (Y).
Coefficient | %SEC | ΔCW |
Intercept | |
ß0 | 65.97 | 0.96 |
Linear | |
ß1 | -31.61 | -0.12 |
ß2 | -0.41 | 0.67 |
Quadratic | |
ß11 | | 0.14 |
ß22 | -9.42 | -0.098 |
Interaction | |
ß12 | | 0.12 |
Lack of fit | 0.6965 | 0.1308 |
R2 | 0.8472 | 0.8775 |
R2adj | 0.7962 | 0.7900 |
P | 0.0005 | 0.0043 |
C.V. | 21.85 | 9.43 |
To corroborate the optimized conditions for the embryogenic callus induction, a validation assay using 0.5 mg L− 1 2,4-D + 3.53 mg L− 1 IBA was performed, from which a value of 94.0% was obtained for %SEC, against 100% predicted by numerical optimization. Meanwhile, for ΔCW, the average weight of the samples was of 1.510 g, against the predicted value for the model of 1.4323 g. The experimental values obtained with the validation assay developed experimentally, were very close to the predicted ones generated with the RSM optimization. About this topic, Andressen et al. (2009) indicates that the RSM allows evaluating the substances concentrations which can modify the growth and development of different organs and analyze their effects. Also, Abbasi et al. (2016) and Aghayeh et al. (2020), suggested that this statistical modeling can be applied to estimate the best combinations of the phytohormones needed for successful in vitro plant regeneration, maximizing the responses and minimizing the application of hormonal treatments. In agreement with those previous reports, the results obtained in the present work indicate that the use of this statistical tool can be applied with a high degree of certainty in the development of tissue culture tools, such as embryogenic callus induction of elite cultivars of sugarcane.
Related to the induced callus type, through the optimization tests and even under optimized conditions, different types of callus were obtained; thus, fluffy and soft non-embryogenic callus were obtained from 8.61 mg L− 1 2,4-D + 8.61 mg L− 1 IBA (Fig. 4A); compact non-friable callus were induced from 5.25 mg L− 1 2,4-D + 0.5 mg L− 1 IBA (Fig. 4B); and combinations of non-embryogenic and embryogenic callus induced over the same apical meristem were also obtained (Fig. 4C). Finally, under the optimized conditions (0.5 mg L− 1 2,4-D + 3.53 mg L− 1 IBA) embryogenic callus was mainly obtained (Fig. 4D-F). It has been reported that as the formulation of the medium is modified, different types of callus are induced, which are classified based on their macroscopic characteristics (Ikeuchi et al. 2013). As starting point, the type of induced callus could be originated by specific and distinct gene expression profile; which in turn, could be related to the interaction between growth regulators present in the culture medium and those produced by the explant itself (Iwase et al. 2011; Maulidiya et al. 2020).