Enzymatic Polymerization of Rutin and Esculin and Evaluation of the Antioxidant Capacity of Polyrutin and Polyesculin

oxidase inhibition, chelating and CUPRAC activities of rutin, esculin

As it has been mentioned previously one of the problem in the use of phenolic compounds, was their weak solubility. The first results of enzymatic polymerization reported that the obtained polymers of rutin and esculin were 4200-folds and 189-folds more water soluble than rutin and esculin, respectively [26,28]. The solubility of polyrutin was also increased in dimethylformamide (DMF) and dimethyl sulfoxyde (DMSO) [15].
Enzymatic polymerization of phenolic compounds affected also their biological properties. These properties, including antioxidant activities, might be dependent on the molecular weight of the synthesized polymers and the type and the position of the linkages (M w , PDI, C-C or C-O bridges). Moreover, depending on the used method for determining antiradical activity (AAPH, DPPH,...) of polyphenols, results were controversial.
As an example, rutin polymerized by laccase from Pycnoporus coccineus, Pycnoporus sanguineus or Myceliophthora led to polymers with a better inhibition of AAPH radical, compared to its monomer [15,27]. However, Anthoni et al. [26] reported that polyrutin, obtained by laccase from Trametes versicolor polymerization, had a weaker DPPH radical scavenging activity compared to rutin. This behavior could be due either to the used method of antioxidant activity determination or the degree of polymerization. Oligorutin fractions showed a higher ability of to reduce the genotoxicity induced by H 2 O 2 and antimutagenic effect compared to monomeric rutin [28,29].
Furthermore, the polymerization of 3-methylcatechol by Kawakita et al. [31] led to the formation of polymers with high copper ions adsorption power.
The aim of this work was in one hand, to compare the effect of polymerization on the antioxidant activity of rutin and esculin ( Figure 1) and in other hand, to discuss the structureantioxidant activity relationship. Polyrutin and polyesculin were synthesized by laccase from rutin and esculin, respectively, and carefully separated in different fractions by diafiltration process. Antioxidant activity was evaluated by radical scavenging activity, iron chelating capacity, xanthine oxidase inhibition activity, cupric reducing capacity.

Polymerization reaction
Polymerization reaction was carried out in the same operating conditions described by Anthoni et al [26,28]. Rutin or esculin (50 g/L) was suspended in 1 L of a methanol/ water (30:70 v/v) reaction medium. Laccase solution (3 U/ mL) was added to the mixture. The reaction was stirred at 600 rpm, for 24 h for rutin and 72 h for esculin, at 20 °C. We noticed that rutin polymerization reaction didn't evolve beyond 24 h, whereas, esculin polymerization reaction continued till 72h. The Kinetic of polymerization reaction was followed with seize exclusion chromatography (SEC).

Polymers separation and lyophilization
Final reaction media enriched with rutin and esculin polymers was separated, by successive filtration processes on a 15, 5, 3 and 1 KDa membranes in diafiltration process (INSIDE CeRAM TM ), using a mixture of water/methanol (70:30 v/v) (5 L) as eluent, at 50°C. The transmembranaire pressure (∆P) was fixed at 2 bars. The state permeate flux (F) was in the range of 35 l/h/m 2 .Then, the fractions were lyophilized (Christ Alpha 1-2 LD freeze dryer). Five fractions were thus obtained and characterized (Table 1).

Polymerization reaction
Polymerization reaction was carried out in the same operating conditions described by Anthoni et al [26,28]. Rutin or esculin (50 g/L) was suspended in 1 L of a methanol/ water (30:70 v/v) reaction medium. Laccase solution (3 U/ mL) was added to the mixture. The reaction was stirred at 600 rpm, for 24 h for rutin and 72 h for esculin, at 20 °C. We noticed that rutin polymerization reaction didn't evolve beyond 24 h, whereas, esculin polymerization reaction continued till 72h. The Kinetic of polymerization reaction was followed with seize exclusion chromatography (SEC).

Polymers separation and lyophilization
Final reaction media enriched with rutin and esculin polymers was separated, by successive filtration processes on a 15, 5, 3 and 1 KDa membranes in diafiltration process (INSIDE CeRAM TM ), using a mixture of water/methanol (70:30 v/v) (5 L) as eluent, at 50°C. The transmembranaire pressure (ΔP) was fixed at 2 bars. The state permeate flux (F) was in the range of 35 l/h/m 2 .Then, the fractions were lyophilized (Christ Alpha 1-2 LD freeze dryer). Five fractions were thus obtained and characterized (Table 1).

Seize exclusion chromatography analysis (SEC)
Relative masses of polymers were evaluated by size exclusion chromatography (SEC) (HPLC LaChrom, UV 280 nm LaChrom L-7400, Tosoh TSKgel α 3000 column, 60 °C). Dimethylformamide (DMF) with 1 % LiBr was used as a mobile phase (0.5 mL/min). Molecular mass calibration was obtained using standards of polystyrene and polystyrene sulfonate. The obtained data allowed the determination of number-average molecular mass (Mn), weightaverage molecular mass (M w ), weight-average molecular mass index (I M ) and polydispersity (PDI) as described by Faix et al. [32].

UV analyzes
The UV spectra of rutin, esculin solutions and their obtained polymers fractions were determined using a UV6000LP spectrometer (Spectra System, Thermofinnigan).

FTIR analysis
The IR analyses were conducted by ATR-FT-IR spectroscopy using a FT-IR spectrometer Tensor 27 (Bruker). The analysis was carried out on monomers and polymers lyophilized powders.

Radical scavenging on ABTS +⋅
The assay was conducted according to protocols presented by Re et al. (1999) and van den Berg et al. (2001) [33,34]. To generate the ABTS +⋅ radical, the ABTS stock solution (7 mM) and potassium persulphate (2.45 mM) in water were allowed to stand in the dark at room temperature for 12-16 h before use. For the reaction, 10 µl of each sample at various concentrations (from 800 to 0.25 µM) was added to 990 µl of diluted ABTS +⋅ (absorbance 0.7 at 734 nm) and the absorbance was recorded every min. A standard curve was prepared using a series of concentrations of trolox (from 0 to 15 µM) with 990 µl of diluted ABTS +⋅ solution. The radical scavenging capacity of tested samples was calculated based on the trolox standard curve and expressed as the trolox equivalent antioxidant capacity (TEAC) and as IC 50 .

Radical scavenging activity on DPPH ⋅
The free radical scavenging capacity of the esculin and rutin and their polymers was determined with 2,2-diphenyl-1-picryl-hydrazyl as described by Bruda et Oleszek [35]. A solution of 1 ml of monomers or polymers (from 10 2 to 4 10 4 µM, concentrations were calculated from M w ) in methanol, was mixed with 2 ml of DPPH (10 mg/L in methanol/water, 80:20, v/v). A reference sample was prepared by adding 1 ml of methanol in 2 mL of DPPH solution.
Monomers and polymers absorbance for each concentration was evaluated at 527 nm, after 15 min, at 23 °C. The antiradical activity was calculated as a percentage of DPPH discoloration using the following equation (1).

* 100 Absorbance of the sample Absorbance of polymers Antiradical activity
Absorbance of the reference The results were expressed as IC 50 and TEAC according to the calibration curve (from 0 to 5 µM).

Inhibitory effect on deoxyribose degradation
Inhibitory effects of tested compounds on deoxyribose degradation were determined by measuring the competition between deoxyribose and theses compounds for the hydroxyl radicals generated from the Fe 3+ /ascorbate/EDTA/H 2 O 2 system (referred to non site-specific assay) or Fe 3+/ ascorbate/H 2 O 2 system (referred to site-specific assay which could indicate the hydroxyl scavenging power of tested molecules by iron chelating power) according to the method described by Halliwell et al. [36] with slight modifications.
The tested sample was added to the reaction mixture containing deoxyribose (10 mM), Fe(III) chloride (10 mM), EDTA (1 mM), and H 2 O 2 (10 mM), ascorbic acid (1 mM), 1mM H 2 O 2 and 50 mM potassium phosphate buffer (pH 7.4). The mixture was incubated for 1 h at 37°C, TBA (1%) and TCA (2.8%) were added to the above mixture, and then heated for 90 min on water bath at 80 °C. The absorbance at 532 nm was then measured against a blank containing deoxyribose and buffer. For site-specific hydroxyl radical scavenging activity, the procedure was similar to the above method, except that EDTA was replaced by the equivalent volume of buffer. The gallic acid was used as a standard. The percentage of deoxyribose degradation inhibition was calculated using the equation (2).
where A C is the absorbance of negative control and A S the absorbance of sample solution.
Results of deoxyribose assay in the presence and the absence of EDTA are expressed as IC 50 and as TEAC.

Xanthine oxidase inhibition assay
The tested samples were solubilized in phosphate buffer (pH 7.5, 50 mM), except rutin which was dissolved in a minimum of DMSO (5 µl) and then in buffer. The assay was conducted as described by Kong et al. [37]. Tests solutions were prepared by adding 1600 µL of buffer, 300 µL of tested solutions (from 4 10 -6 to 10 -3 M), 1000 µL of a solution of xanthine (0.15 mM) and 100 µL of a solution of xanthine oxidase (0.2 U/mL). The reaction was monitored for 6 min at 295 nm. Two samples were prepared, the first without tested solutions to determine the total uric acid production, and the second without enzyme to measure the absorbance of tested solutions at 295 nm for the range of concentrations. Results were expressed as the final concentration that results in half-maximal enzyme velocity (IC50) and calculated by standard curve regression analysis and as TEAC according to the calibration curve (from 1 10 -3 to 5 10 -1 µM).

Cupric reducing antioxidant capacity (CUPRAC)
The cupric reducing antioxidant capacity was determined according to the method of Apak et al. [38] To each tube containing 20 µl of tested substrate concentration, we added CuCl 2 to a final concentration of 3.12 mM, ethanolic neocuproine solution and NH 4 Ac buffer solution (pH=7) to final concentrations of 2.34 × 10 -3 M and 312 mM, respectively. The total volume was then adjusted with distilled water to 2 ml and mixed well. Absorbance against a reagent blank containing all reagents except CuCl 2 and neocuproine was measured at 450 nm after 1h. The results were expressed as equivalent of Trolox according to the calibration curve (from 10 to 10 3 µM).

Polymers synthesis, separation and characterization
Kinetics of esculin and rutin polymerization were monitored by SEC-UV at 280 nm. Once the polymerization was achieved, polymers were separated, by successive diafiltration process. Weight-average molecular mass M w , polydispersity (PDI) and weight average molecular mass index (I M ) of obtained fractions (R1-5 and E1-5) were summarized in table 2. These results indicated clearly that the polymerization of the two substrates was occurred and led to polymers of rutin and esculin with high molecular weight ( Figure 2).

UV and FTIR investigations
The UV-visible spectrum of rutin, in methanol/water (30/70 v/v), presented two maxima of absorption at 282 and 359 nm due to the π-π* transition of the aromatic electrons. For polyrutin fractions (R1, R3 and R5) the 359 nm band was larger and presented a hypsochromic shift of 5 nm. Such results could be due to the implication of the B ring of rutin in the formation of polymers. In fact, Anthoni et al. [26] and Marckam [39] observed a similar behavior. The latter stated that the presence of a substitution on the 5, 7 and 4' positions of the phenolic rings led to a hypsochromic shift.
The UV spectra of esculin and polyesculin fractions E2, E3, E4 and E5 presented the same peaks with a maximum of absorption at 345 nm while the peaks correspondent to E5 were broader than those of esculin, which could be attributed to conjugated oligomeric structure [15,40]. The same profile was reported by Anthoni et al. for the esculin polymerization [30].
FTIR spectra of rutin and polyrutin fractions (R1, R3 and R5) (Figure 3), showed a new peaks at 1220 cm -1 and at 1465 cm -1 .The peak at 1220 cm -1 indicated the formation of new ether bonds C-O. The signal at 1465 cm -1 could be attributed to a bond C-C while the absence of a peak at 1747 nm on the R1 spectra compared to rutin spectra could be explained by the disappearance of C-H bonds. These results showed that obtained polyrutin fractions were formed through C-C and C-O linkages. In fact, many authors reported that flavonoid polymers were composed of phenylene units and/or oxyphenylene units [24,26,41]. Uzan et al. [27] reported that the nucleophilicity of the aromatic A-ring seemed to play a major role as the reactive hydroxylated ring in coupling reactions for the formation of a new bond. They suggested that polymerization of rutin by Pycnococus laccases led to formation of polymers through C-C and C-O bonds and more precisely through C8-C8, C6-O4′ and C8-C5′ linkages. A study of the polymerization of quercetin by Bruno et al. [12], with Horseradish peroxidase (HRP), showed that the highest occupied molecular orbital (HOMO) was concentrated on the catechol group. Therefore, these authors expected the polymerization reaction to take place in the two more negative carbons of that group 2′ and 5′.  As for rutin, FTIR spectra of polyesculin fraction E2, E3, E4 and E5 showed a new peak at 1400 cm -1 compared to the spectra of esculin. This could be due to a formation of C-C bonds. In fact, Anthoni et al [30] reported the formation of C-C and C-O linkages, involving both the phenolic and the glucosidic part of the coumarin during the esculin polymerization. Moreover, an in silico structure investigation of oligoesculin by the same authors suggested the preferential formation of C8-C8 linkage between esculin units during the polymerization reaction.
The obtained and reported data of UV and FTIR suggested that different linkages (C-C, C-O) could be achieved depending to monomer, enzyme and operating conditions (pH, temperature, medium). This might affects the antioxidant activity of the polymer.

Evaluation of antioxidant activity of rutin and polyrutin fractions
Different methods were used to evaluate the antioxidant activity (free radicals scavenging activity, iron chelating capacity, xanthine oxidase inhibition activity and cupric reducing capacity) of esculin, rutin and their derivatives. Results were summarized in Table 3.

Free radicals scavenging activity of rutin and polyrutin fractions
Results in Table 3 showed that IC 50 , related to polyrutins, increased progressively versus M w . The fraction R5, presenting the highest M w , led to highest IC 50 values (640, 38 and 38.32 µM) compared to IC 50 values obtained in presence of rutin (320, 1.1 and 18.6 µM) respectively for ABTS, DPPH and hydroxyl radicals. These results suggested that higher is the M w lower is the antiradical activity. The low antiradical activity of polyrutin fractions observed in this study was in accordance with that reported by other authors [15,26,27].

Xanthine oxidase inhibition activity of rutin and polyrutin fractions
For XO inhibition activity (Table 3), the IC 50 values of polymer fractions appeared to be lower than the IC 50 value of rutin (962 µM). The results indicated that IC 50 decreased when the M w arised, which traduced the better ability of polyrutins to inhibit XO compared to monomeric rutin. The fraction R5 illustrated the highest XO inhibition power, 68-folds better than monomeric rutin. The strong XO inhibition observed for polyrutin fractions was in accordance with other studies dialled in enzymatic flavonoid polymerisation [14,15,25,26,42].

Iron chelating properties of rutin and polyrutin fractions
All polyrutin fractions exhibited higher degree of iron chelating ability (Table 3). This activity grow with the increase of M w . The polyrutin fraction R5 presented the highest iron chelating power with an IC 50 value of 36.5 µM compared to 58.3 µM, in presence of the monomer.

Free radicals scavenging activity of esculin and polyesculin fractions
Polyesculin fractions presented lower IC 50 values than those of monomeric esculin which indicated their stronger antiradical activity (Table 3). Polyesculin fraction E5 was the most potent one. It was respectively for ABTS, DPPH and hydroxyl radicals 450, 19 and 80 folds more active than esculin (450, 9200, 5600 µM). Unlike rutin, the antiradical activities increased with M w when ABTS and hydroxyl radical methods were used. However, for DPPH the IC 50 remained constant, about 480 µM, for all tested fractions. So, DPPH scavenging activity seemed to be independent to the degree of polymerisation.

Xanthine oxidase inhibition activity of esculin and polyesculin fractions
Results in Table 3 showed that for all polyesculin fractions, IC 50 were lower than that of the monomer (779 µM). This activity was linked to M w and decreased as M w increased. The fraction E5 presented the lowest IC 50 and therefore the highest XO inhibition activity, 5-folds higher than monomer.

Iron chelating properties of esculin and polyesculin fractions
Polyesculin fractions exhibited high degree of iron chelating activity, according to the sitespecific hydroxyl radical-scavenging assay (Table 3). Results showed that iron chelating capacity was high as the M w increases. The best iron chelating power was observed in the presence of the E5 fraction (IC 50 =180 µM), which was 37-folds better than esculin (IC 50 =6800 µM). Table 3 indicated that polyesculin fractions presented higher TEAC than esculin. This activity rose as the M w increased. Therefore, the best cupric reducing antioxidant capacity was seen with the E5 fraction which was 30-folds more active than esculin.

Structure-antioxidant activity relationship
The structure-antioxidant activity relationship of monomeric flavonoids and coumarins was well investigated. According to many authors [43] free hydroxyl groups on C4', C3' and C7 played a major role in antiradical activity of rutin and esculin. However, few data are available about the behaviour of these activities with polymerization. In this work we observed a decrease of polyrutin antiradical activities with M w increase. This decrease could be attributed to the loss of these groups during the rutin polymerization reaction.
For high iron chelating power and CUPRAC, hydroxyl groups on C5, C3 and the 4 oxo (for flavonoids) and hydroxyl groups and catechol moiety (for coumarins) were essential. So, high iron chelating and cupric reducing antioxidant capacities observed with polyrutin and polyesculin fractions suggested that these groups were not implicated in the linkage occurred in rutin polymerization reaction [44][45][46][47].
For high xanthine oxidase inhibition activity, several works reported the importance of the presence of a double bond between C2 and C3 and free hydroxyl groups on C5 and C7 [26,[48][49][50]. High inhibition of the xanthine oxidase obtained in the presence of polyrutin and polyesculin fractions implicated that these groups are not affected during the polymerization reaction.

Conclusion
Polyphenolic polymers of rutin and esculin were synthesized using a laccase from Trametes versicolor. These polymers were fractioned by diafiltration process.
The analyses of rutin polymers by FTIR showed the presence of new C-C and C-O bonds and the desperation of a C-H bond on monomer. These results suggested that polyrutin were synthesized through phenylene and oxyphenylene units. For polyesculin fraction, FTIR analyses indicated the presence of only C-C bond.
Free radical scavenging activity of rutin was decreased by the enzymatic polymerization while polyesculin fractions showed a high antiradical activity compared to monomeric esculin. This behaviour suggested that the antioxidant activity depend on the position of linkage through the polymerization reaction. For esculin, it seemed that the polymerization didn't affect groups implicated in the antioxidant activity. This could explain the high antioxidant activity values observed for polyesculin.
Both polyrutin and polyesculin fractions exhibited a high XO inhibition activity, iron chelating and cupric reducing antioxidant capacities.