Antioxidants and Phenolic Secretion in Sugarcane Genotypes Shoots Culture

Secretion of phenolic compounds is a major limitation for sugarcane in vitro shoot culture, causing a loss of regenerative capacity and subsequent cell death. In this study, micropropagation and phenolic secretion of four Saccharum genotypes were evaluated in presence of different antioxidants. Aseptic cultures of S. officinarum (PI 184794 and PI 88652), S. sinense (PI 29109) and S. robustum (UNK R65P35) were propagated on medium containing antioxidants, citric acid (100 mg/L), L-cysteine (100 mg/L), polyvynylpirrolidone (300 mg/L) and L-glutathione (50 mg/L) in two consecutive subculture cycles. Interaction between genotypes and antioxidants was significant in both cycles. All genotypes showed good shoot formation, shoot vigor and color, except in PI 88652 which had less shoot development in both the presence and absence of the antioxidants tested. PI 184794 displayed the highest shoot proliferation in the presence of citric acid, and UNK R65P35 produced more shoots per explant in the 2 subculture. For S. sinense (PI 29109), in both subcultures, most shoots were observed in the presence of polyvynylpirrolidone. Medium discoloration due to phenolic secretion was reduced in the presence of citric acid and polyvynylpirrolidone. The type of secreted phenolic compounds differed with genotype as the Principal Component Analysis of cultivation media separated PI 88652 from PI 29109 and UKN R65P35. Phenolic compounds varied in composition and were secreted at various levels as a function of genotype and antioxidant type. Loadings plots indicated the genotype and antioxidant separations were broadly driven by flavonoid compounds.


Introduction
The Saccharum genus belongs to the Andropogoneae tribe of the Poaceae (Gramineae) family (Daniels & Roach, 1987).It is a perennial grass cultivated in tropical and subtropical regions of the world and the global crop production in 2014 was 1900 million metric tons (The Statistics Portal, 2017); in 2013, sugarcane ranked first in commodities (FAO, 2013).
During in vitro organogenesis or somatic embryogenesis, browning of sugarcane culture caused by phenolic secretion can be a major limitation causing a loss of regenerative capacity and subsequent cell death.According to Ndakidemi et al. (2014), the relationship between medium chemical compounds and phenolic exudation influences substantially the intensity of medium discoloration, rooting, explant browning and tissue deterioration.In a culture establishing phase, the phenolic secretion and its oxidation can affect the culture initiation and development (Kerns & Meyer, 1986;Kumari & Verma, 2001).Thus the prevention of culture browning is essential in micropropagation of Saccharum plants.Studies conducted by Qin et al. (1997) indicated that sugarcane has a high content of phenolics, and their oxidation is affected by genotypes, the sources of explants and exogenous growth regulators.Similar observations were recorded for Gossypium spp., cotton (Ozyigit et al., 2007) and Strelitzia reginae, bird of paradise (North et al., 2012).Lux-Endrich et al. (2000) suggested that the composition and synthesis of phenolic compounds in plant tissue may be determined by genetic and environmental conditions.
The possibility of controlling phenolic secretion from sugarcane explants by pretreatment with solutions of ascorbic or citric acid, polyvynylpirrolidone (PVP) and cysteine or culturing the explants on medium with these substances, or with activated charcoal has been suggested (Kumari & Verma, 2001;Lorenzo et al., 2001;Huang et al., 2003;Khan et al., 2007;Lal et al., 2015;Shimelis et al., 2015).However, there is little information on tissue culture responses of various sugarcane species to the type of antioxidants used during culturing and the type of phenolic compounds secreted.Adding from 0.5 to 1 g/L PVP to medium of callus derived sugarcane culture controlled tissue browning but was variety specific (Michael, 2007).Callus derived culture might exhibit somaclonal variation and compromise genetic integrity of propagated material (Nehra et al., 1992;Skrivin et al., 1993;Sahijram et al., 2003;Bairu et al., 2006Bairu et al., , 2008Bairu et al., , 2011)).
The National Plant Germplasm System of the United States Department of Agriculture (USDA) preserves over 400 sugarcane accessions in a clonal form as field plantings.The security backup of the field collection is done by cryopreservation of 0.6-1 mm shoot tips derived from aseptic shoot culture.The shoot cultures are established from apical fragments of field grown canes and it takes usually from 2 to 4 months (genotype dependent; 30-day subculture intervals), after contamination free shoots are obtained, to produce a sufficient number of shoots for cryopreservation.Rooting of the shoot culture is not necessary because the shoots will not be planted before cryopreservation; however, multiplication of a large number of shoots in a short time with a low benzyl aminopurine concentration is desired.Secretion of phenolic compounds and high genotypic variation in the cultivation of tissue culture affects the growth and multiplication of the shoots, and impedes the preservation of the Saccharum genetic resources.Limiting the phenolic occurrence in culture medium might improve shoot multiplication, and limit the number of subcultures needed to produce a large number of shoots, and also it might increase the shoot vigor that supports successful preservation of sugarcane.
The objective of this study was to evaluate in vitro sugarcane shoot performance in the presence of phenolic compounds secreted into medium with selected antioxidants and to characterize the phenolic intensity and composition in the culture medium of three sugarcane species (S. officinarum, S. robustum and S. sinense).

Micropropagation and Antioxidants
Excised shoots (from 15 to 20 mm long), 30±10 days old, 9 shoots per box, were transferred to fresh MS propagation medium (as described before) with the following antioxidants individually or in combination: 100 mg/L of citric acid (C277, PhytoTechnology Laboratories), 100 mg/L L-cysteine (168149, Sigma-Aldrich®), 300 mg/L PVP 40,000 (PX 1300, EM Science) and 50 mg/L L-glutathione (G4251, Sigma-Aldrich®).The selection of the antioxidant type and concentration was based on in-house observations made during micropropagation of other plant species (data not published).After 30 days, cultures were transferred to fresh MS propagation medium with the same antioxidant treatment for another 30 days under same growing conditions, 9 shoots per box and 6 boxes per treatment.

Assessment of Micropropagation and Phenolic Secretion
Genotypic and antioxidant effects were evaluated by observing the number of shoots per explant, their vigor and color, and by assessing the intensity of medium color caused by phenolic secretion during two micropropagation cycles.The two cycles were selected due to previous observations in which the most intense medium discoloration appeared during this time.The shoot vigor was ranked: 1-poor vigor; 2-average good vigor; 3-fully healthy with excellent shoot vigor (adapted from Debnath, 2005).The shoot color was ranked as 1-brown; 2-yellow; 3-light green; 4-dark green and the phenolic secretion intensity (color of medium) was ranked as 1-no phenolic secretion; 2-some; 3-moderate secretion; 4-heavy secretion.
Analysis of secreted phenolic compounds was carried out in propagation media containing antioxidants (1) 100 mg/L citric acid; (2) 100 mg/L L-cysteine; (3) 300 mg/L PVP; or (4) 50 mg/L L-glutathione) separately for the three genotypes (PI 88652, PI 29109 and UNK R65P35) at the end of the 1 st subculture cycle (30 days after inoculation).The assay of phenolic compounds was performed according to Heuberger et al. (2014), with modifications as noted below.

Preparation of Extract
For metabolite profiling, an approximate 200 mg of culture medium pooled from the culture boxes in each treatment (four media combined with three accessions) was collected by a micropipette and kept in Eppendorf® tubes.The samples were freeze dried and homogenized in a bullet blender.A sequential extraction from 7.5 mg of culture medium was performed, first extracting with 1.0 mL methanol and secondarily with 1.0 mL of MTBE/MeOH/Water (6/3/1 v/v/v).The supernatant from each step was collected and pooled.The solvent was evaporated under nitrogen, and the metabolites re-suspended in 100 µL methanol and stored at -80 °C until further analysis.

UPLC-MS Analysis (CSH Phenyl-Hexyl Method)
The amount of each phenolic compound in each extract was determined by an ultra-performance liquid chromatography mass spectrometry (UPLC-MS) analysis (CSH Phenyl-Hexyl method); 5 μL of the methanol extract was injected twice (n = 2 replicates) onto a Acquity UPLC system (Waters Corporation, Milford, MA) in discrete, randomized blocks, and separated using a Acquity UPLC CSH Phenyl-Hexyl column, 1.7 μM, 1.0 x 100 mm (186005402, Waters©), using a gradient from solvent A (water, 0.1% formic acid) to solvent B (Acetonitrile, 0.1% formic acid).Calibration was conducted using sodium iodide with 1 ppm mass accuracy.The capillary voltage was held at 2200 V, source temp at 150 °C, and nitrogen de-solvation temperature at 350 °C with a flow rate of 800 L/hr.

Experimental Design and Statistical Evaluation
Experimental design to evaluate the effect of the antioxidants and genotypes on the growth and phenolic secretion was fully randomized in a 2 × 4 × 5 factorial scheme (2 subculture cycles × 4 genotypes × 5 treatments with or without antioxidants) with six replications per treatment and nine shoots per replication (n = 54).The means were compared by Scott-Knott test at 5% probability.Pearson's correlation coefficient was calculated to denote the relationship between phenolic secretion and number of shoot/explant, their vigor and color, during the 1 st and 2 nd subculture cycles.Pearson correlation coefficient for each dependent variable between actual and predicted values was maximized, which is an indicator of the predictive performance of the algorithm.The hypothesis of the correlation coefficient was tested with a two-sided t test.Statistical analyses were performed with the SAS-9.4 program (SAS Institute, 2013).
To explore the metabolome of the culture medium with emphasis on phenolic compound determination, samples from the first experiment were separated for another experiment and a fully randomized 3 × 4 factorial scheme (3 genotypes × 4 antioxidants) with three replications per treatment was considered.Analysis of variance was carried out for each compound using the Analysis of Variance (AOV) function in R, and p-values were adjusted for false positives using the Bonferroni-Hochberg method in the p adjust function in R (Benjamini & Hochberg, 1995).Principal component analysis (PCA) was carried out on mean-centered and Pareto variance-scaled data using the PCA Methods package in R (Stacklies et al., 2007).The results were plotted as a function of retention time and the -log of the P-value.For all experiments, the differences between the data were considered significant at 5%.

Effect of Genotypes and Antioxidants on Growth Variables (Shoot Number, Color and Vigor)
The subculture cycle factor (S) was significant only for the number of shoots/explant and in combination with the genotype (S×G) for the shoot vigor (Table 1).The genotype and antioxidants were significant for all the factors and presented very high values.A significant interaction was observed between the evaluated genotypes (G) and the applied antioxidants (A).
The evaluated cultures showed good shoot formation, vigor and color, except the PI 88652 (S. officinarum), which displayed low proliferative capacity and shoot vigor in the presence and the absence of the four antioxidants (Table 2).
S. officinarum (PI 184794) showed a significantly higher number of proliferated shoots in the presence of 100 mg/L citric acid (12.53 shoots/explant) and a significantly lower number with 300 mg/L PVP (8.28 shoots/explant).S. officinarum (PI 88652), S. robustum (UNK R65P35) and S. sinense (PI 29109) showed no significant improvement in shoot number with antioxidants in the medium compared to the control.The antioxidants had no effect on the shoot vigor S. officinarum (PI 88652), S. sinense (PI 29109) and S. robustum (UNK R65P35).However, S. officinarum (PI 184794) had significantly higher vigor with three of the antioxidants compared to PVP and the control.Note.ns: not significant; * significant at 5% by F test; ** significant at 1% by F test; df: degree freedom; ms: mean square; Fr: F ratio; VC: variation coefficient.
Shoot color was significantly better for S. officinarum (PI 184794) and S. robustum (UNK R65P35) with some of the antioxidant treatments while the others were not significantly different.All of the genotypes tested were light green to dark green regardless of the antioxidant added to the culture medium.In general an intensive green color was observed on shoots of all genotypes grown on medium with L-cysteine.In all genotypes, the most shoots/explant was observed in the second subculture cycle (Table 3).The GxV interaction was indicated by the high shoot vigor maintained in S. sinense (PI 29109) and S. robustum (UNK R65P35) in the second subculture.Note.Means followed by the same lowercase letter (rows) and the same uppercase letter (columns) are not different from each other according to Scott-Knott test at 5 % probability.

Phenolic Secretion and Correlation With Growth Variables
All the factors and interactions were significant for phenolic secretion (Table 4).Phenolic secretion into the medium and its intensity (measured by the medium discoloration) was significant for the genotype × antioxidant × subculture interaction.Variations in the medium color intensity due to phenolic secretion were observed in both cycles for all genotypes (Table 5).In the 1 st subculture cycle the genotypes showed different phenolic intensity.S. officinarum (PI 184794) and S. robustum (UNK R65P35) presented lowest phenolic secretion in the presence of PVP (2.50; 2.50, respectively).The activated charcoal (AC) induced the lowest discoloration of medium for S. robustum (UNK R65P35).No difference in medium discoloration between the applied antioxidants was observed in S. sinense (PI 29109) culture media.Note.Means followed by the same lowercase letter (rows) and the same uppercase letter (columns) are not different from each other according to Scott-Knott test at 5 % probability.
Except in PI 184794, all other genotypes in the 2 nd subculture cycle did not show any effect of the antioxidants on the medium discoloration.PI 184794 showed the lowest phenolic rank in the presence of L-cysteine (2.00) and L-glutathione (2.50).Antioxidants did not present effect on phenolic rank in both subcultures for PI 88652.
For S. sinense (PI 29109) and S. robustum (UNK R65P35), the average ranking for culture medium discoloration due to phenolic secretion seemed to be slightly higher in the 2 nd subculture than in the 1 st one.The expectation was to observe less phenolic secretion with each subsequent subculture, as it was observed for S. officinarum (PI 184794).However, an increase in phenolic secretion in in vitro cultivation of other sugarcane genotypes was noticed, even in the 4 th and 5 th subculture (data not shown).
In both subculture cycles, the intensity of medium color caused by phenolic secretion was negatively correlated with the number of shoots/explant (-0.2520; -0.2391, respectively, p < 0.01) (   Another observation pertains to the variation in genotypic responses between the two subculture cycles in shoot vigor and phenolic secretion.In PI 184794, during the 1 st cycle, phenolic secretion was higher than in the 2 nd cycle in the presence of L-cysteine and L-glutathione, disagreeing with Lorenzo et al. (2001) that observed an increase of phenol compounds in sugarcane cultivar C-1051-73 when the culture medium was changed.This might be a genotype specific reaction of S. officinarum (PI 184794) considering that S. sinense (PI 29109) and S. robustum (UNK R65P35) had the lowest phenolic ranks on the 2 nd subculture cycle.The make-up and synthesis of phenolics in plant tissue may be determined by genetics and environmental conditions (Lux-Endrich et al., 2000).The results of this study agree with Rodrigues et al. (2011) and Kala et al. (2012), who observed that Saccharum sp. and Psidium sp.genotypes showed variation in secretion of phenolic compounds.Genotype dependence of the in vitro organogenesis responses was also reported by Gandonou et al. (2005) for nine sugarcane (Saccharum sp.) commercial cultivars.Another aspect to consider is the relation between auxin metabolism and phenolic secretion (Lorenzo et al., 2001).In studies on the effect of phenolic acids and their derivatives upon the growth of Avena sativa L. coleoptiles, Wolf et al. (1976) reported different effects of phenolic compounds on the plant growth and inhibition in the presence of indole acetic acid oxidation.
Based on Pearson's correlation coefficient, in both subculture cycles the number of shoots/explant was negatively correlated with phenolic secretion.Kerns and Myer Jr. (1986) emphasized the phenolic secretion and other exudate discharge observed during explant initiation into tissue culture systems lessened with growth and development.However, controversial opinions have been published on the relation between cell and tissue proliferation and low intensity of phenolics found in different species as Medicago sativa L. (Cvikrová et al., 1996), Nicotiana tabacum L. (Chirek, 1990) and Pinus sp.(Herman, 1991).Lorenzo et al. (2001) showed that the increase of phenolic content during the first three days of sugarcane culture is due to the most intensive cell division period, but later the secretion decreased.
The separation by principal components suggests that phenolic compounds, as flavonoids, are secreted at different levels as a function of the genotype and antioxidant.These aspects were observed by other authors for cotton (Ozyigit et al., 2007) and Hawk tea (Tan et al., 2016).These results suggest that additional antioxidant types, not included in this research, might be effective in controlling phenolic secretion in diverse Saccharum genotypes.

Conclusions
The research reinforced the strong interaction between Saccharum genotypes and phenolic secretion into in vitro shoot culture medium.The studies demonstrated that including citric acid (100 mg/L S. officinarum PI 184794) or PVP (300 mg/L S. robustum UNK R65P35 and S. sinense PI 29109) in culture medium reduced the medium discoloration caused by the phenolic secretion.The number of shoots was negatively correlated with the phenolic secretion; hence, antioxidants might promote shoot development.The high levels of flavonoids secretion by S. officinarum PI 88652 had an adverse effect on in vitro shoot development and the culture vigor.The ability of S. officinarum and S. robustum in vitro cultures to produce flavonoid compounds in the presence of antioxidants might be a desired characteristic in producing flavonoids on a larger scale throughout the year under controlled environmental conditions.

Note.
Means followed by the same lowercase letter (rows) and the same uppercase letter (columns) are not different from each other according to Scott-Knott test at 5% probability. 1 Shoot vigor rank: 1-poor vigor; 2-average good vigor; 3-fully healthy with excellent shoot vigor; 4-heavy secretion; 2 Shoot color rank: 1-brown; 2-yellow; 3-light-green; 4-dark green.
Figure supplemen L

Table 1 .
Analysis of variance for number of shoots/explant (NS), shoot color (SC) and shoot vigor (SV) for four Saccharum genotypes Source of variation df MS shoot number Fr shoot number MS shoot color Fr shoot color MS shoot vigor Fr shoot vigor

Table 2 .
Effect of genotypes and antioxidants (GxA) on the number of shoots/explant, shoot vigor and color of four Saccharum genotypes (means of 1 st and 2 nd subculture cycles)

Table 3 .
Effect of subculture cycle on the number of shoots/explant and the shoot vigor of four Saccharum genotypes

Table 4 .
Analysis of variance for phenolic color rank in medium with shoot cultures of four Saccharum species VCNote.ns: not significant; * significant at 5% by F test; ** significant at 1% by F test; df: degree freedom; ms: mean square ; Fr: F ratio; VC: variation coefficient.

Table 5
. Effect of antioxidants (A) and subculture cycles (S) on the phenolic secretion 1 in cultures of four Saccharum spp.genotypes

Table 6
). Negative correlation indicated an inhibitory effect of phenolic compounds on in vitro shoot proliferation.Correlation between phenolic secretion and shoot vigor was not significant in either cycle (p > 0.05); however, in the 2 nd subculture, a other flavonoid sources.Despite the high of dihydroxyflavone-rhamnoside and hydroxyl-methoxyflavone, S. robustum (UNK R65P35) showed good shoot proliferation.