Flavonoids and Antimicrobial Properties of Begonia fischeri var. palustris in vitro Plantlets

Corresponding Author: Evgeniya A. Karpova, Central Siberian Botanical Garden Siberian Branch of Russian Academy of Sciences, Novosibirsk, Russia E-mail: karpova@csbg.nsc.ru Abstract: The objective of this study was to design a protocol of the successful establishment of plants of Begonia fischeri var. palustris obtained in the in vitro seeds culture, to evaluate flavonoid content and antimicrobial properties of in vitro plantlets. The significant increase of percentage of seed germination (92.5%) was recorded in the half strength MS medium and 1% agar medium (83.3%) in comparison with 50-60% in greenhouse conditions. Flavonoid composition of the leaves of in vitro plantlets and greenhouse stock plants had no substantial differences. Significant differences (p<0.05) were not found between flavonoid contents of the leaves (13.6 and 15.5 mg·g −1 of Dry Weight (DW), respectively). Aqueous ethanol extracts of plants showed antimicrobial effects against reference strains of Bacillus subtilis, Streptococcus pyogenes and Staphylococcus aureus. Concentration of flavonoids in acetone and ethanol extracts of exudative compounds of the leaves of in vitro plantlets was 0.02 and 2.0 mg·g −1 DW, respectively. Our results indicated that this protocol can be used for the production of a large number of plants of B. fischeri var. palustris with rich flavonoid composition and high quantities of flavonoids, including exudative compounds, with healing and air cleaning properties.


Introduction
Begonia fischeri Schrank var. palustris (Hartw. ex Benth.) Irmsch. (section Begonia) is ornamental shrublike plant with a vigorous growth. It is implemented in ornamental gardening in greenhouses throughout the world, including Siberia. Its natural habitat is the Atlantic coastal forest in Brazil. It grows mostly in marshy places and forms large thickets under the shadow of forest canopy. Flowering continues throughout the year.
Vegetative multiplication of the species by means of traditional cultivation techniques seems to be highly laborious. In vitro raising of many begonias is an alternative efficient system widely used by commercial nurseries and institutes for rapid plants propagation, germplasm conservation, elimination of pathogens and for phytochemical screening of medicine-useful compounds (Murugan et al., 2016). The seed populations in vitro can be stored for a long time lacking the problems of bacterial or fungal infections ordinarily associated with long-term maintenance of Begonia clonal stock.
In addition to high decorative, B. fischeri var. palustris contains phenolic compounds, including flavonoids and have prospects for use as medicinal . Nowadays, a large number of studies are devoted to antimicrobial activity of plants. Flavonoid-rich plant extracts from many species were shown to possess antibacterial and antifungal activity (Cushnie and Lamb, 2005;Xie et al., 2015).
Despite the increased attention to the biology and pharmacognosy of the genus Begonia in the world, data on the chemical composition and biological activity of the species of the genus are still sparse. Antimicrobial properties of the extracts of several species (B. erythrophylla, B. heracleifolia, B. semperflorens, B. fuchsioides, B. picta and B. malabarica) were investigated (Frei et al., 1998;Ramesh et al., 2002;Nisha et al., 2016).
Representatives of the genus Begonia are widely presented in the plant design indoor. In this regard, it is important to assess the composition of the surface exudative compounds of the leaves as air cleaners. At this moment, exudative constituents of the leaves, located on the leaf surface, are extensively investigated (Muravnik et al., 2016). Meanwhile, these compounds of representatives of the genus Begonia are at the very beginning of study (Ensemeyer and Langhammer, 1982).
The objective of this study was to design a protocol for the plant production using in vitro seeds culture and to evaluate flavonoid content and antimicrobial properties of in vitro plantlets B. fischeri var. palustris. This includes initiation and inoculation of aseptic cultures of seeds, examination of the effect of different media on the germination rate of B. fischeri var. palustris.

Seed Sterilization and Germination
The mature fruits of B. fischeri var. palustris were harvested three months after hand pollination. Seed capsules were surface disinfected with 70% ethanol for 1 min, followed by immersion in 0.1% solution of HgCl 2 for 10 min and finally rinsed thoroughly with three times in sterile distilled water.
The nutritious media were enriched by 2.5% sucrose and solidified with 0.6% agar. The pH in all variants of media were adjusted to 5.8 prior to autoclaving at 121°C and 103.4 kPa for 20 min. The cultures were incubated at 22±2°C with light/dark 16/8 hours photoperiod provided by white fluorescent tubes (1500 lx).
Seeds were inoculated onto four variants of imbibition media: 1.0% agar, ½ MS (half strength MS salts) (Murashige and Skoog, 1962), MS and MS supplemented with 0.5 mg·L −1 6-benzylaminopurine (BAP) and 0.3 mg·L −1 of α-Naphthaleneacetic Acid (NAA). In each flask, 100±5 seeds were inoculated, using the modified technique, where the disinfected seeds were firstly scattered on a sterile paper and transferred individually to the agar medium with a dissecting needle. In every variant of medium, three replicates were fulfilled.
B. fischeri var. palustris seed germination was defined here as hypocotyl emergence (1/3 the length of the seed size). Germination was monitored every day for until no further germination was recorded. Final germination percentage (FG) and Mean Daily Germination (MDG) were calculated using the formulas given below: Where: n = The number of germinated seeds N = The total number of seeds D = The number of days to final germination

Plant Growth Conditions in Greenhouse
Hardened in vitro plantlets of B. fischeri var. palustris after transfer from sterile substrate grew in greenhouse on soil mixture consisted of the rotted leaf litter mixture with river sand (1:1) in the greenhouse conditions (temperature 20-23°C, humidity 50-60%, illumination 40-60 µmol m −2 s −1 ).
Simultaneously with the beginning of the seed germination in vitro, B. fischeri var. palustris seeds gathered from greenhouse stock plants were sown directly in greenhouse on the same soil mixture under the similar greenhouse conditions (greenhouse plants). Further, experimental design consisted in comparison of the greenhouse plants and in vitro plantlets.

Flavonoid Identification and Quantification
The greenhouse-grown plants and in vitro plantlets of B. fischeri var. palustris were grown in the composition of "Collections of living plants indoors and outdoors" USU 440534 of Central Siberian Botanical Garden, Siberian Branch of Russian Academy of Sciences, Novosibirsk, Russian Federation (CSBG SB RAS) under greenhouse conditions with a temperature of + 15-25 С 0 , humidity of 30-60% and illumination of 40-60 µmol m −2 s −1 .
Leaves from ten of greenhouse plants (L) and ten in vitro plantlets (LR) were collected for detection of flavonoids in the beginning of generative state (at 60th day after transfer ex vitro).
For detection of flavonoids, precisely weighed samples of fresh plant material (0.5 g) were exhaustively extracted with an ethanol: water mixture (70:30, v/v) in a water bath at 60-70°C. Aqueous ethanol crude extract was filtered, diluted with the ethanol:water mixture (70:30, v/v) up to the volume of 25 mL in a graduated flask and used for quantification of flavonoids and evaluation of antimicrobial activity. Dry-weight concentration in the samples was calculated by the gravimetric method. The method was reported in detail previously (Karakulov et al., 2018).
For studying the exudate flavonoids of the leaves of in vitro plantlets, acetone and ethanol extracts of leaf surface constituents were prepared. For acetone extraction, fresh (not ground) precisely weighed leaves were briefly rinsed with acetone to dissolve the surface constituents (Exudate A). Ethanol extract (Exudate E) was made with analogous wash of fresh leaves with 95% ethanol. The acetone and ethanol solutions were evaporated to dryness and each residue was dissolved in a small volume of 95% ethanol. These ethanol extracts were used for quantification of compounds in leaf exudates and for evaluation of antimicrobial activity. Dry-weight concentration in the samples was calculated by the gravimetric method.
The HPLC system for absolute quantification of flavonoids consisted of an Agilent 1200 with a Diode Array Detector (DAD) and the ChemStation software (Agilent Technologies, Santa Clara, CA, USA) for data processing. The chromatographic separation was conducted at 25°C on a Zorbax SB-C18 Column (4.6×150 mm, 5 µm i.d.) with the Agilent Guard Column Hardware Kit (p.n. 820888-901).
The mobile phase consisted of MeOH (solvent A) and 0.1% orthophosphoric acid dissolved in water (solvent B). Separation of glycosides of flavonoids in crude extracts was performed with gradient. The run via gradient was started with a solvent A-solvent B mixture at 32:68 (v/v) followed by a linear gradient to 33:67 (v/v) for the first 27 min, then to 46:64 (v/v) from minute 28 to minute 38, then to 56:44 from minute 39 to minute 50, then to 100:0 from minute 51 to minute 54. The mobile phase was returned to 32:68 (v/v) from minute 55 to minute 56. The flow rate was set to 1 mL⋅min −1 . The sample injection volume was 10 µL and the absorbance was measured at 210, 255, 270, 290, 325, 340, 360 and 370 nm. The quantification of phenolic compounds was conducted by the external-standard method. Chemical reference standards of quercetin, kaempferol, luteolin, kaempferol rutinoside, vitexin and isovitexin were purchased from Sigma (St. Louis, MO, USA) and hyperoside, isoquercitrin, astragalin were obtained from Fluka (Fluka, Sigma-Aldrich Chemie GmbH, Munich, Germany). Validation of the analytical procedures was performed in accordance with ICH guidelines (2005).
Anthocyanin content was determined by spectrophotometric method on UV-Vis spectrophotometer "Agilent 8453" (Agilent Technologies, Santa Clara, CA, USA). The concentration of anthocyanins (mg·g −1 of dry weight) was calculated using the molar absorbtion coefficient of cyanidin-3-rutinoside in 1% HCl in water equal to 28 840. The absorbance of fresh leaves extract (in 1% HCl in water) was measured at wavelengths of 529 and 650 nm .

Evaluation of Antimicrobial Activities
The disk diffusion method was chosen for preliminary evaluation of the antimicrobial activities of crude extracts and exudates of the greenhouse plants and the in vitro plantlets of B. fischeri var. palustris. The strains of American Type Culture Collection were used as reference strains: gram-positive bacteria -Staphylococcus aureus ATCC 25923 (SA), Streptococcus pyogenes ATCC 12344 (SP) and Bacillus subtilis ATCC 6633 (BS); gramnegative bacteria -Escherichia coli ATCC 25922 (EC), Klebsiella pneumonia ATCC 13883 (KP), Enterobacter aerogenes ATCC 35029 (EA), Pseudomonas aeruginosa ATCC 27853 (PA) and Enterococcus faecalis ATCC 19433 (EF) and fungi -Candida albicans ATCC 10231 (CA). Ciprofloxacin (5 µg) served as a standard antibacterial agent and itraconazole (10 µg) was used as a standard antifungal agent (positive control). Ethanol served as a negative control.
Seven reference strains (S. aureus, S. pyogenes, B. subtilis, E. coli, E. arerogenes, E. faecalis and C. albicans) were selected for the determination of Minimal Inhibitory Concentrations (MICs) of the extracts and the exudates, based on their disc-diffusion susceptibility data. MICs of the leaf extracts of the greenhouse plants (L) and the in vitro plantlets (LR) and the acetone (A) and the ethanol (E) exudates together with ciprofloxacin and itraconazole were determined using the broth microdilution assay (Andrews, 2001). The bacteria reference strains and the strain of C. albicans were grown overnight on nutritive agar for microorganisms and on mycobiotic agar (respectively) from Laboratorios Conda and diluted equivalent to a 0.5 McFarland standard (Andrews, 2001). Microtiter plate wells (final total volume of 200 µL), each containing 90 µL of Mueller-Hinton (MH) broth were inoculated with 10 µL of cell suspension and serial dilutions of the extracts and the exudates (100, 50, 20, 10 and 5 µg·m L −1 ) and antimicrobial agents ciprofloxacin (0.025 µg·mL −1 to 128 µg·m L −1 ) and itraconazole (0.008 µg·m L −1 to 4 µg·m L −1 ). Test range of itraconazole was established in accordance with European Committee for Antimicrobial Susceptibility Testing (Arendrup et al., 2017). The plates were incubated at 30°C for 24 h without shaking. The negative control wells contained MH broth only and the positive control wells contained the respective cell suspensions with no plant extract (exudates)/antimicrobial agents added. This was done in triplicate (Andrews, 2001). The MIC was the lowest concentration of antimicrobial agent, which inhibited visible growth of organism.

Statistical Analysis
All the data were expressed as mean ± standard error (SE) of three replicates %. The data were analyzed using one-way analysis of variance (ANOVA) and means were compared using Duncan multiple range test. The values of P<0.05 were considered significant.

Results
B. fischeri var. palustris seed germination in the greenhouse conditions generally occurred within 3 weeks and final germination percentage (FG) was in the range of 50-60%.
When evaluating the final germination percentage of the species in vitro, significant differences were observed between all the variants tested. The highest percentages of seed germination (92.5%) were recorded in the half strength MS basal medium (½ MS) followed by 1% agar medium (83.3%). In the tissue culture conditions, germination rate of seeds was the highest in Var.1: seed emergence occurred within 11 days (Table 1).
In Var. 2 and 2A the germination rate was poor, it took more than 15 days for seed emergence and seedlings developed more slowly (Fig. 1).
Half strength MS medium produced optimal results for the B. fischeri var. palustris seedling development suggesting that this medium was the most suitable for the germination of the species.
After 30 days, all the germinated seeds were transferred to a sterilized medium containing fullstrength MS salts, 2% sucrose and 0.6% agar. For acclimatization, plantlets with well-developed roots and shoots (Fig. 1c) were removed from the tissue culture, transferred to small plastic pots and exhibited 85% survival in the mixture of sterile garden soil and river sand (1:1). Seedlings were incubated in a room condition at 20°-22°C and 12 h photoperiod under cool-white light (40 µmol m −2 s −1 ) for 40 days. Then the hardened plantlets, as well as the greenhouse plants, were investigated for antimicrobial activity and flavonoid composition (Fig. 2).    0.2 ± 0.0 6.1 ± 0.4 *Significant differences between L and LR (p>0.05) a the leaves of greenhouse plants (L) and in vitro plantlets (LR); acetone (A) and ethanol (E) extracts of exudate compounds of the leaves b In parentheses is ratio "concentration of the constituent in exudate A or E/its total concentration in the leaf of in vitro plantlets"
Numbers of flavonoids in the leaves and leaf exudates slightly varied mainly due to free aglycones and minor C-glycosyl flavones. Minimal number of flavonoids was found in the exudate E, whereas concentrations of the constituents were substantially higher than in exudate A. In exudate A, number of flavonoids was maximal due to the presence of isovitexin, kaempferol rutinoside, free kaempferol and the others minor compounds.
Proportions of the constituents in the leaves and leaf exudates were also equal. Concentrations of total flavonoids in the leaves of in vitro plantlets were insignificantly lower in comparison with greenhouse plants. Both values were substantial (above 13 and 15 mg·g −1 , respectively). Significant differences were found between the concentrations of isovitexin, hyperoside, isoquercitrin, rutin, kaempferol rutinoside and free quercetin. Flavonoids were more than half of the concentration of total phenolic compounds. Isoquercitrin was about 60% of total flavonoids concentration.
Concentrations of phenolic compounds in the exudates A and E were < 0.02 mg·g −1 and > 2 mg·g −1 , respectively. In the exudates A this concentration was 0.08% and concentration of flavonoids was 0.10% of their concentration in the leaves as a whole. In the exudates E, comparable percentages were 3.22% and 9.21%. Certain constituents were extracted with acetone better, for example, free quercetin and kaempferol glycosides. Thereby, extraction of exudate flavonoids with acetone caused the enrichment of the extract by free quercetin and kaempferol glycosides and extraction with ethanol promoted the increase of total flavonoids in the extract of exudate compounds. Some amount of anthocyanins in exudate E indicates possible extraction of definite amounts of vacuolar constituents.
Despite of some decrease of total flavonoids concentration in the leaves of the in vitro plantlets in comparison with greenhouse plants, the yield of total flavonoids per plant in the plantlets was significantly higher ( Table 2).
The examination of antimicrobial properties of the leaf extracts showed that crude extracts of the leaves of B. fischeri var. palustris inhibited growth of seven from nine reference strains (Table 3). Both extracts were active against gram-positive bacteria, S. aureus (MICs of 50 µg/mL), S. pyogenes (MICs of 20 µg·mL −1 ) and B. subtilis (MICs of 20 and 50 µg·mL −1 ), as well as E. aerogenes (MICs of 50 µg·mL −1 ). The extract L inhibited the growth of E. coli, E. faecalis and C. albicans (MICs of 50 µg·m L −1 ). None of the extracts showed activity against K. pneumonia and P. eaeruginosa.   50  20  20  -a  50  50  -50  50  Extract LR  50  20  50  -50  - The absence of activity b L -extract from the leaves of greenhouse plants; LR -extract from the leaves of in vitro plantlets; exudates A and E-acetone and ethanol extracts of exudate compounds of the leaves, respectively Therefore, the spectra of antimicrobial activity of two extracts differed. The spectrum of extract L from the leaves of greenhouse plants (7 reference strains) was wider than the spectrum of extract LR from the leaves of in vitro plantlets (4 reference strains). Only extract L from the leaves of greenhouse plants was active against C. albicans.
Antimicrobial effects of the exudates was reduced. The exudate E exhibited inhibition of the growth of two reference strains of gram-positive bacteria, S. aureus and B. subtilis (MICs of 50 µg·mL −1 ) and the exudate A inhibited the growth of only one reference strain, B. subtilis (MIC of 50 µg·m L −1 ).

Discussion
The result suggests that a low profile of macro-and micronutrients is effective for the species seed germination. This is consistent with the earlier report of Chamandoosti (2016), where among of different basal media that were evaluated for their effects on sesame seed germination, half strength MS was found to be the best medium for germination.
During the plants establishment in greenhouse, we note that plant status and size of the leaves of in vitro plantlets were higher in comparison to greenhouse plants ( Fig. 2b, a, c). The difference in phenolic and flavonoid contents in the leaves reflected the difference in the morphological traits. Thus, the plantlets with more vigorous growth had lower flavonoid content. This corresponds with many reports about increase of flavonoids in the stress or unfavorable conditions .
On account to the significantly higher yield of flavonoids per plant and a shorter (by 2-3 months) growing cycle, the in vitro plantlets of B. fischeri var. palustris can be regarded as a promising source of flavonoids.
We showed that antimicrobial activity of the extracts of B. fischeri var. palustris varied with various flavonoid composition and content. The spectra of antimicrobial activity of the extracts of in vitro plantlets and the exudates were substantially reduced. At the same time, inhibition effect of the extracts on the growth of the strains of B. subtilis and S. aureus manifested in most cases. Unstable inhibition effects of the extracts of plants of B. fischeri var. palustris in relation to the strains of E.
coli, E. faecalis and C. albicans indicates that antimicrobial activity against these pathogens requires higher concentrations of flavonoids. Further works should be designed in order to assess the impact of concentration of the main constituents and some varying components of extracts (quercitrin, free aglycons and kaempferol glycosides) on the antimicrobial action.
Another direction of the study can be dealt with the selection of clones with high flavonoid content. Our results are in accordance with the reports on high flavonoid content of micropropagated plants (Rad et al., 2014;Goda et al., 2017), but the level can be increased.
Of great importance are compounds located on the leaf surface. Recently, these compounds are the objectives of new direction of investigations, the metabolic engineering concerning with in vitro synthesis and secretion of large amounts of target compounds by leaf surface structures. The works in this field increase, but at this point, they are still quite fragmentary (Huchelmann et al., 2017). We showed that the flavonoid composition of exudates of B. fischeri var. palustris differs slightly from the composition of the leaves as a whole, but the concentrations of most of components are insignificant. This caused a weak antimicrobial activity of exudates. However, the direct contact of the flavonoids of the leaf surface with the air creates the base for antimicrobial properties sufficient to reduce a colony-forming units in the air. This reduction was previously shown in experiments with intact plants in kindergartens and libraries . One of the further directions of the investigation should be the nutritious medium optimization to get a significant increase of exudative flavonoids amounts for high antimicrobial activity (Kim et al., 2015).
Thus, in our study we established an efficient in vitro propagation and regeneration protocol of B. fischeri var. palustris plants, with rich flavonoid content, obtained from the seeds culture, for the aims of interior design and health benefits.

Conclusion
In our study, we developed the protocol of micropropagation and acclimatization of in vitro plantlets of B. fischeri var. palustris with high flavonoid content in the leaves (13 mg·g −1 of dry weight) insignificantly differed from the leaves of the greenhouse plants and with higher yield of total flavonoids per plant (189.1 g in comparison with 174.0 g). The highest percentages of seed germination were revealed in the half strength MS medium (92.5%) and in 1% agar medium (83.3%). The extract of the leaves inhibited the growth of four from nine reference strains, S. aureus (MIC of 20 µg·mL −1 ), B. subtilis, S. aureus and E. aerogenes (MICs of 50 µg·mL −1 ). Concentrations of flavonoids in acetone and ethanol extracts of exudative compounds of the leaves of in vitro plantlets were 0.02 and 2.0 mg·g −1 of dry weight, respectively. These extracts inhibited the growth of one (B. subtilis) and two reference strains (B. subtilis and S. aureus). The in vitro derived B. fischeri var. palustris plants are promising for indoor design of children institution and public places and for air cleaning. The purpose of further investigation is the improvement of culture conditions for the production of plants with genetically stable high level of flavonoid content.