Enzymatic synthesis of 1,3-dihydroxyphenylacetoyl-sn-glycerol: Optimization by response surface methodology and evaluation of its antioxidant and antibacterial activities

Bioorganic Chemistry 75 (2017) 347–356

Contents lists available at ScienceDirect

Bioorganic Chemistry journal homepage: www.elsevier.com/locate/bioorg

Enzymatic synthesis of 1,3-dihydroxyphenylacetoyl-sn-glycerol: Optimization by response surface methodology and evaluation of its antioxidant and antibacterial activities Nadia Kharrat a, Imen Aissa a, Youssef Dgachi b, Fatma Aloui a, Fakher Chabchoub b, Mohamed Bouaziz c, Youssef Gargouri a,⇑ a

Laboratory of Biochemistry and Enzymatic Engineering of Lipases, ENIS, Route of Soukra, P.O. Box 1173, 3038, University of Sfax, Tunisia Laboratory of Applied Chemistry: Heterocycles, Lipids and Polymers, Faculty of Sciences of Sfax, University of Sfax, P.O. Box 802, 3000 Sfax, Tunisia c Electro-chemical Environmental Laboratory, ENIS, Route of Soukra, P.O. Box 1173, 3038, University of Sfax, Tunisia b

a r t i c l e

i n f o

Article history: Received 21 June 2017 Revised 4 October 2017 Accepted 22 October 2017 Available online 23 October 2017 Keywords: Biocatalysis Esterification 1,3-Dihydroxyphenylacetoyl-sn-glycerol Structural identification Antibacterial and antioxidant activities

a b s t r a c t In this study, the enzymatic synthesis of phenylacetoyl glycerol ester was carried out as a response to the increasing consumer demand for natural compounds. 1,3-dihydroxyphenylacetoyl-sn-Glycerol (1,3-diHPA-Gly), labeled as ‘‘natural” compound with interesting biological properties, has been successfully synthesized for the first time in good yield by a direct esterification of glycerol (Gly) with p-hydroxyphenylacetic acid (p-HPA) using immobilized Candida antarctica lipase as a biocatalyst. Spectroscopic analyses of purified esters showed that the glycerol was mono- or di-esterified on the primary hydroxyl group. These compounds were evaluated for their antioxidant activity using two different tests. The glycerol di-esters (1,3-di-HPA-Gly) showed a higher antiradical capacity than that of the butyl hydroxytoluene. Furthermore, compared to the p-HPA, synthesized ester (1,3-di-HPA-Gly) exhibited the most antibacterial effect mainly against Gram + bacteria. Among synthesized esters the 1,3-di-HPA-Gly was most effective as antioxidant and antibacterial compound. These findings could be the basis for a further exploitation of the new compound, 1,3-di-HPA-Gly, as antioxidant and antibacterial active ingredient in the cosmetic and pharmaceutical fields. Ó 2017 Published by Elsevier Inc.

1. Introduction Over the last few years, a commercial relevance for glycerol has been marked by a significant increase because of its rising inevitable formation as a by-product of biodiesel production (10%, w/w referred to biodiesel) [1–3]. The use of glycerol as feedstock has awakened the interest in the development of new green synthesis processes. Several studies have reported the chemistry of glycerol, starting from the classical esters to generate new products like glycerol carbonate, propanediols telomers and epoxides via chemical and biocatalytic reactions [3]. The phenolic compounds derived from by-product plants are highly regarded for their antioxidant capacity [4]. Recently, growing evidence has shown that phenolic compounds in olive oil could play a part in protection against cancer and more attention is focalized to its phenolic compounds [5–7]. Olive oil extraction generates solid and liquid wastes ⇑ Corresponding author at: Laboratoire de Biochimie et de Génie Enzymatique des Lipases, ENIS, Université de Sfax, Route de Soukra, 1173 Sfax, Tunisia. E-mail address: [email protected] (Y. Gargouri). https://doi.org/10.1016/j.bioorg.2017.10.011 0045-2068/Ó 2017 Published by Elsevier Inc.

(olive leaf and waste olive oil), which represents a big environmental pollution problem [8,9]. Nevertheless, these wastes are also promising sources of phenolic compounds that can be recovered and converted to useful products by synthesis prcesses. These by-products include especially 4-hydroxy phenylethanol (tyrosol), 3,4-dihydroxyphenylethanol (hydroxytyrosol), caffeic acid and 4-hydroxyphenylacetic acid (p-HPA) used for the preparation of non-steroidal anti-inflammatory drugs, p-coumaric acid, protocatechuic acid and many others [10–12]. These bioactive phenolic acid compounds exhibit powerful biological activities such as anti-oxidation [13,14], inhibition of human immune deficiency virus (HIV) [15], anti-inflammatory [16] and joint-degenerative effects [17]. Interestingly, two molecules, 1-feruloyl-sn-glycerol (FG) and 1,3-diferuloyl-sn-glycerol (F2G), have been isolated from natural sources, from Solanum tuberosum (potato) [18]. These ferulic acid derivatives have been motivated by the feruloyl moiety which exhibited antioxidant properties and ultraviolet (UV) light absorbing [19,20]. In fact, these ferulic acid derivatives may also be an active ingredient attractive to the food and cosmetic industries.

348

N. Kharrat et al. / Bioorganic Chemistry 75 (2017) 347–356

Besides, numerous studies have shown that F2G has been synthesized using chemical catalysts which make the processes industrially and economically infeasible [18]. Such antioxidants have been used as natural substitution for their synthetic counterparts, such as butylated hydroxytoluene [21]. However, it is well known that chemical synthesis generates secondary compounds which must be eliminated by purification processes. In addition, chemical catalysis does not meet the requirements for food applications. For these reasons, enzymatic catalysis seems to be the ideal method which offers various advantages like specificity, milder reaction conditions and without producing secondary compounds. The enzymatic esterification of glycerol and soybean oil mono- and diacylglycerols with ethyl ferulate using Candida antarctica, generated the production of 1-feruloyl-sn-glycerol and 1,3-diferuloyl-snglycerol compounds, respectively [21]. In addition, the esterification of hydroxycinnamic acids to glycerol using feruloyl esterases, has been previously reported [22]. Vafiadi et al. [23] reported the esterification of sinapic acid to one of the primary hydroxyl groups of glycerol. The authors showed that the synthesized compound (glycerol sinapate) retained, after esterification, 89.5 ± 1.1% of its antioxidant activity against low-density lipoprotein oxidation. In this context, the main focus of this study is to improve the biological activity of a phenolic compound, by fixing two or three molecules of p-HPA onto a hydroxyl group of glycerol (Gly) using Novozym 435 as a biocatalyst. Then, the antioxidant and antibacterial activities of the newly synthesized esters will be investigated. 2. Materials and methods 2.1. Chemicals and reagents Chloroform and methanol were purchased from Scharlau (Spain); acetonitrile and acetic acid from Pharmacia (Uppsala, Sweden); 2-methyl-2-propanol from Fluka (Germany); The p-HPA; 2,2-azino-bis-3ethylbenzothiazoline-6-sulfonic acid (ABTS); 4,5-dimethylthiazol-2-yl-2,5-diphenyltetrazolium bromide (MTT) and 2,2-diphenyl-1-picrylhydrazyl (DPPH) were purchased from Fluka (Switzerland); butylated hydroxytoluene (BHT) (purity
99%) from Sigma. Candida antarctica lipase (Novozym 435) from Novo Nordisk. 2.2. Experimental designs and esterification reactions 2.2.1. Box-Behnken designs and response surface analysis Response surface methodology (RSM) is an empirical optimization technique for evaluating the relationship between experimental responses and coded factors called X1, X2, X3. This method is usually used in combination with factorial design methods such as central-composite designs and Box–Behnken designs. Adopting Box–Behnken designs allowed reducing the number of experimental sets without decreasing the optimization accuracy compared with traditional factorial design methods which are based on a large number of experimental assays. Three factors (enzyme amount, p-HPA/glycerol molar ratio and temperature) have been selected according to preliminary assays, for enhancing the conversion yield of the synthesized product. These factors were investigated at three different levels (1, 0, +1) in terms of coded and uncoded symbols as shown in Table 1. An empirical model related to the factors has been obtained using a second-order polynomial equation and a multiple regression of data [24]. The general form of the second-order polynomial equation is:

Y ¼ b0 þ

X

bi Xi þ

X

bii X2i þ

X

bij Xi Xj

Table 1 Levels of the factors tested in the Box-Behnken design. Factors

Symbol

p-HPA/Gly molar ratio Temperature (°C) Enzyme amount (mg)

X1 X2 X3

Coded levels 1

0

+1

2 35 10

5 45 30

8 55 50

where Y is the predicted response, Xi and Xj are independent factors, b0 is the intercept, bi is the linear coefficient, bii is the quadratic coefficient and bij is the interaction coefficient. 2.2.2. Data analysis and software Design-expert, version 7.0 (STAT-EASE Inc., Minneapolis, USA) was used for the experimental designs and the statistical analysis of the experimental data. The analysis of variance (ANOVA) was used to estimate the statistical parameters. 2.2.3. Esterification reactions Glycerol contains three hydroxyl groups in its structure, and therefore three esterified derivatives with p-HPA were expected. HPLC analysis of the reaction mixture after 48 h showed the production of two new compounds by the enzymatic reaction. Each new compound was purified and identified by NMR analysis as described below. The purified compounds are used as external standard. RSM was performed to optimize the enzymatic synthesis of hydroxyphenylacetoyl glycerol derivatives. Several conditions with a total of 17 runs were tested including different combinations of three factors (p-HPA/glycerol molar ratio, temperatures and enzyme amounts) (Table 2). In fact, the production of hydroxyphenylacetoyl glycerol derivatives was performed by the direct esterification of glycerol with p-HPA in screw-capped flasks. Different molar ratios of p-HPA to glycerol were dissolved in 3 mL of tertbutanol/acetonitrile volume ratio 1:1 (v/v). Since, we have shown that the adequate solvent to solubilize the two substrates is a mixture of solvent ratio ter-butanol/acetonitrile 1:1 (v/v). We have shown also that the addition of a more hydrophobic solvent in the reaction mixture, like tert-butanol (Log P = 0.8), trigger the enzyme activity. The mixture was stirred at different temperatures in an orbital shaker at 220 rpm and in the presence of different amounts of Novozym 435 (Candida antarctica lipase). Controls were run in parallel, under the same conditions, without enzyme addition. After 48 h of reaction incubation, an aliquot of the mixture reaction was withdrawn and filtered to be used for HPLC analysis. The conversion yield of synthesized compounds was calculated as the ratio of the number of moles of synthesized compound (determined by an external standard range previously established for each compound) per the total number of moles of p-HPA. The esterification study has been carried out using four microorganism lipases (Rhizopus oryzae, Staphylococcus aureus, and Staphylococcus xylosus) produced in our laboratory and immobilized on CaCO3 and a commercial one (Candida antarctica lipase immobilized on macroporous acrylic resin). Novozym 435 (Candida antarctica lipase) showed the highest yield conversion (data not shown) and was selected for the rest of our work. 2.3. Purification and identification of synthesized compounds: 1hydroxyphenylacetoyl-sn-glycerol and 1,3-dihydroxyphenylacetoylsn-glycerol The purification of the newly synthesized compounds was achieved by chromatography on a silica gel60 column (Merck)

N. Kharrat et al. / Bioorganic Chemistry 75 (2017) 347–356 Table 2 The Box-Behnken design of RSM for optimization of the conversion yield of 1,3-diHPA-Gly. Experimental conditions

Conversion yield (%)

Run

X1

X2

X3

Calculated

Predicted

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

2 8 2 8 2 8 2 8 5 5 5 5 5 5 5 5 5

35 35 55 55 45 45 45 45 35 55 35 55 45 45 45 45 45

30 30 30 30 10 10 50 50 50 50 10 10 30 30 30 30 30

76.91 70.32 66.8 54.15 59.83 43.30 60.41 46.55 60.02 41.14 49.87 38.92 70.11 67.15 70.73 68.14 65.12

78.75 69.37 67.75 52.31 57.37 43.63 60.08 49.01 58.51 40.52 50.49 40.43 68.25 68.25 68.25 68.25 68.25

(25 cm  2 cm) according to Kharrat et al. [25]. The reaction mixture resulting from the esterification of glycerol with p-HPA contains a mixture of the new synthesized compounds and residual substrates. After enzyme removal by centrifugation at 5009g for 15 min, the solvent (2-methyl-2-propanol and acetonitrile) was evaporated at 45 °C under vacuum. One hundred mg of target product was taken up in 1 mL of acetonitrile. Elution was carried out using hexane/diethyl ether/acetic acid mixtures (50:49:1). The collected solvent fractions were analyzed by TLC. The spots were identified under evaporated iodine. Purified fractions of each compound were pooled and solvents evaporated at 45 °C under vacuum. Final purity of each compound obtained was checked and analyzed using HPLC, FT-IR, LC-MS/MS and RMN analysis. The purified compounds are stored at 20 °C until uses in antioxidant and antibacterial activities. 2.4. HPLC analysis The identification and conversion yields of hydroxyphenylacetoyl glycerol esters were carried out by HPLC analysis (Ultimate 3000, Dionex, Germany). The HPLC system was equipped with a pump (LPG-3400SD), column oven, and diode-array ultraviolet– visible detector (DAD-3000RS). The output signal of the detector was recorded using Dionex Chromeleon chromatography Data System. Separation was executed on an Inertsil ODS-4 C-18 column (5 lm, 4.6 mm  250 mm; Shimpack) maintained at 30 °C. The flow rate used was 1.5 mL/min, and the detection UV wavelength was set at 280 nm. The used mobile phase was water (A) versus 0.1% acetic acid in acetonitrile (B) for a total running time of 26 min, and the following proportions of solvent (B) were used for elution: 0–5 min, 10–30%; 5–10 min, 30–70%; 10–24 min, 70% and 24–26 min, 70–10%. 2.5. LC–MS/MS analysis An analytical LC–MS/MS analysis was performed on a Phenomenex Luna C18 (2) column (150  4.6 mm i.d., 5 lm particle size), using a 1 mL/min linear mobile phase gradient of 20–50% aqueous MeOH (containing 1% acetic acid) in 30 min, and mass spectra were recorded by a Thermo Scientific ‘‘LCQ Classic” ion trap mass spectrometer fitted with an electro-spray ionization (ESI) source. Accurate mass measurements were performed on Finnigan MAT900 XLT or Thermo Scientific LTQ Orbitrap XL mass spectrometers in negative and positive ESI mode.

349

2.6. NMR and FT-IR experiments 1

H NMR and 13C NMR spectra were recorded in deuterated chloroforme on a Bruker spectrometer operating at 400 MHz and 100 MHz respectively. The chemical shifts are reported in parts per million (ppm), using tetramethylsilane (TMS) as internal reference. The multiplicities of the signals are indicated by the following abbreviations: s, singlet; d, doublet; t, triplet; q, quadruplet; and m, multiplet coupling constants are expressed in Hz. IR spectra were recorded on Agilent Cary 600 FT/IR (Malaysia). 1-HPA-Gly. 1H NMR (400 MHz, CDCl3) d 3.38 (2H, s, (H40 ), ACH2ACOAOA), 3.76 (2H, m, (H200 ), AOH CH2ACHACH2AOA), 3.92 (1H, m, (H1), ACH2ACHACH2A), 4.38 (2H, d, (H20 ), ACH2AOACO), 6.64 (2H, d, (H70 and H90 ), J = 9 Hz), 7.08 (2H, d, (H60 and H100 ), J = 9 Hz). 13C NMR (100 MHz, CDCl3) d 172.5 (C30 (C@O)), 156.6 (C80 ), 130.7 (C60 and C100 ), 124.9 (C50 ), 115.5 (C70 , C90 ), 71.9 (C1), 67.4 (C20 ), 66.3 (C200 ). 1,3-di-HPA-Gly. 1H NMR (400 MHz, CDCl3) d 3.39 (4H, s, (H40 et H400 ), 2 ACH2ACOAOACH2A), 3.90 (1H, m, (H1), ACH2ACHACH2A), 4.39 (4H, m, (H20 et H200 ), 2 ACHACH2AOC@O), 6.62 (4H, d, (H70 , H60 , H700 and H900 ), J = 9 Hz), 7.09 (4H, d, (H60 , H100 , H600 et H1000 ), J = 9 Hz). 13C NMR (100 MHz, CDCl3) d 172.1 (C30 and C300 (C@O)), 155.1 (C80 and C800 ), 130.3 (C60 , C100 , C600 , and C1000 ), 125.3 (C50 and C500 ), 115.3 (C70 , C90 , C700 and C900 ), 76.1 (C1), 67.8 (C20 and C200 ). IR (liquid) cm1: 1-HPA-Gly. 3500–3400 (HO-Ø), 2982 (CAH), 1710(C@O), 1598 (C@C) and 1220 (CAO). 1,3-di-HPA-Gly. 3500– 3400 (HO-Ø), 2980 (CAH), 1707(C@O), 1598 (C@C) and 1220 (CAO).

2.7. Reusability of the enzyme Under optimal conditions, Novozym 435 (Candida antarctica lipase) was reused for consecutive cycles in a fresh reaction medium. In fact, at the end of each batch, Novozym 435 was eliminated by filtration and washed thoroughly with tert-butanol/acetonitrile in order to remove any substrate or product retained in the support. The values are presented as the means of triplicate analyses.

2.8. Antioxidant activity 2.8.1. DPPH assay The free radical scavenging activities of each pure compound pHPA, 1-HPA-Gly and 1,3-di-HPA-Gly were determined using the method of Brand-Williams et al. [26] with a little modification. An aliquot of ethanol absolute solution (0.1 mL) containing different concentrations (1:2 serial dilutions from the initial sample) of p-HPA and hydroxyphenylacetoyl glycerol esters was added to 3.9 mL of DPPH solution (0.06 mM in ethanol). The mixture was vortexed vigorously and incubated at room temperature (25 °C) in darkness. After 60 min of incubation, the absorbance was read at 517 nm against ethanol blank using a spectrophotometer (Uvi Light XT5). The IC50 values denote the concentration of tested compounds, which is required to scavenge 50% of DPPH free radicals. The corresponding inhibition percentages were calculated according to the following Eq. (1):

Radical scavenging activity ð%Þ ¼ ½ðAcontrol  Asample Þ=Acontrol   100

ð1Þ

where Acontrol is the absorbance of the control (prepared in the same manner without test compound), and Asample is the absorbance of the test compound. BHT was used as standard control. The values are presented as the means of triplicate analyses.

350

N. Kharrat et al. / Bioorganic Chemistry 75 (2017) 347–356

2.8.2. ABTS radical scavenging assay The ABTS radical-scavenging activity was determined according to Bernini et al. [27]. The ABTS radical cation was prepared by reacting an aqueous solution of ABTS (7 mM) with potassium persulfate (2.45 mM, final concentration) which was kept in the dark at 25 °C for 12–16 h. The obtained solution was diluted in ethanol to an absorbance of 0.70 (±0.020) at 734 nm before use. A volume of 10 mL of trolox or sample in ethanol was mixed with 990 mL of this diluted solution and the absorbance was determined at 734 nm after 6 min of initial mixing. The antioxidant activity was calculated relative to the equivalent trolox concentration. The activity of each antioxidant was determined at three concentrations, within the range of the dose–response curve of trolox. The radical-scavenging activity was expressed as the trolox equivalent antioxidant capacity (TEAC), defined as mM of trolox per gram of sample. BHT was used as a reference compound. The values are presented as the means of triplicate analyses. 2.9. Lipophilicity evaluation The lipophilic character of hydroxyphenylacetoyl glycerol derivatives was evaluated as described by Viskupicova et al. [28] by determining the partition coefficient (miLog P). The miLog P values of hydroxyphenylacetoyl glycerol derivatives were calculated using Molinspiration software, which is based on fragmental methods (available online www.molinspiration.com). 2.10. Determination of the minimum inhibitory concentration (MIC) and the minimum bactericidal concentration (MBC) The antibacterial activities of each pure compound p-HPA, 1HPA-Gly and 1,3-di-HPA-Gly were tested against bacterial strain Gram (+) and Gram (). The minimum inhibitory concentration (MIC) values, which correspond to the lowest compound concentration that completely inhibits the growth of microorganisms, were determined by a microwell dilution method as previously described by Eloff [29] using 3-(4,5-dimethylthiazol-2-yl)-2,5-dip henyltetrazolium bromide (MTT). The lowest concentration that yielded no growth after this sub-culturing will be taken as the minimum bactericidal concentration (MBC) [30]. The inoculum of each bacterium was prepared and the suspensions were adjusted to 106 CFU/mL. Each pure compound p-HPA, 1-HPA-Gly and 1,3-di-HPA-Gly was dissolved in ethanol, and then dilution series were prepared in a 96-well plate, ranging from 6.25 mg/mL to 4 mg/mL. Each well of the microplate contained 185 mL of the growth medium, 5 mL of inoculum and 10 mL of the diluted samples. Ethanol was used as a negative control. The plates were incubated at 37 °C for 24 h, then 40 mL of MTT (0.5 mg/mL freshly prepared in sterile water), was added to each well. After incubation for 30 min, the change of medium to purple color indicated that the bacteria were biologically active. So, the MIC was taken where no change of colour of MTT was observed in the well. For the determination of MBC, a portion of liquid from each well that showed no change in color will be placed on solid LB and incubated at 37 °C for 24 h. All experiments were made in duplicate. Several bacterial strains were used: gram (+) Bacillus cereus (BC), Micrococcus luteus (ML), Staphylococcus aureus (Sa), Staphylococcus xylosus (Sx), Staphylococcus epidermidis (SEp), gram (–) Enterobacter cloacae (EC), Pseudomonas aeruginosa (Ps), Klebsiella pneumonia (KP) and Escherichia coli (Ecoli). 2.11. Statistical analysis All analyses were carried out in triplicate. Results were expressed as mean values ± standard deviation (SD) (n = 3). The differences were calculated using one-way analysis of variance

(ANOVA), and statistically significant differences were reported at P < .05. Data analysis was carried out using the SPSS 10.0 software. 3. Results and discussion 3.1. Preliminary study The synthesis of 1,3-dihydroxyphenylacetoyl-sn-glycerol by esterification of glycerol with p-HPA, catalyzed by Novozym 435, in organic media was carried out under several experimental conditions using the different factors mentioned below. In a preliminary study, we showed that the solvents ratio tert-butanol/ acetonitrile to the reaction medium was not necessary to improve the conversion yield. We also found that neither the increase of reaction time up to 48 h, nor the addition of water or a crude molecular sieve 4 Å at the beginning of reaction improved the conversion yield (data not shown). Taking into account these results, we choose the p-HPA/glycerol molar ratio (X1), temperature (X2) and enzyme amounts (X3), as the most effective operating variables on the response and we assigned two level limits (high and low levels) for each factor as shown in Table 1. The other parameters were fixed as follows: reaction time: 48 h, solvent ratio tertbutanol/acetonitrile 1:1 (v/v) with a final volume of 3 ml and stirring speed: 220 rpm. 3.2. Model determination A response surface methodology was performed to optimize the enzymatic synthesis of phenyl ester by a direct esterification of glycerol with p-hydroxyphenylacetic acid using Novozym 435 as a biocatalyst. Several conditions with a total of 17 runs were tested including different combinations of p-HPA/glycerol molar ratio, temperatures and enzyme amounts (Table 2). The experimental and the predicted results are presented in Table 2. The standard ANOVA was used to analyze the experimental results and the Box Behnken design was fitted with the second order polynomial equation in term of coded factors:

^ ¼ 68:25  6:20375X 1  7:01375X 2 þ 2:025X 3  1:515X 1 X 2 y þ 0:6675X 1 X 3  1:9825X 2 X 3 þ 1:915X 1 X 1  3:12X 2 X 2  17:6425X 3 X 3

ð2Þ

where X1, X2 and X3 are the coded factors of p-HPA/glycerol molar ratio, temperatures and enzyme amounts respectively. 3.3. Analysis of variance and validation of the models Table 3 shows the statistical summary for each model that was output by Design Expert V 8.0.6. A good model fit should have an R2 of at least 0.8 [31]. In our case, the response model evaluated in this study can explain the reaction, with an R2 of 0.979 and an Adj-R2 of 0.951 at a confidence level of 95%. As shown in Table 3, these results indicated that 97.9% of the variability in the response could be explained by the model. The quadratic model was therefore selected to fit the experimental responses. The statistical significance of the model equation was evaluated by the F-test for ANOVA. The model F-value of 35.57 implied that the model was significant. There was only a 0.01% chance that the model F-value could occur due to noise. The P-value was very low (P < 0.0001) indicating that the model was significant. In addition, the lack of fit F-value of 8.74 implied that there was a 30.46% chance that the lack of fit F-value could occur due to noise (Table 3). These values indicate that the linear relationship between factors is adequate for the experimental data. An accuracy check is

351

N. Kharrat et al. / Bioorganic Chemistry 75 (2017) 347–356 Table 3 Analysis of Variance (ANOVA) for the fit of experimental data to response surface model.

a

Sum of squares (SS)

Degrees of freedom (df)

Mean square

F value

p-value Prob > F

Regression Residual Lack of fit Pure error Cor total R-squared Adj R-squared

2143.15 46.86 26.23 20.63 2190.01 0.979 0.951

9 7 3 4 16

238.13 6.69 8.74 5.16

35.57

<0.0001a

1.69

0.3046b

Statistically significant at 95 % of confidence level. Indicates non significant at the level 95%

A

Predicted conversion yield (%)

b

Source of variation

Experimental conversion yield (%) Fig. 1. Comparison of predicted and experimental conversion yield.

B

necessary to get an adequate model. The checking of model accuracy was evaluated by comparing the experimental and predicted conversion yield, which shows a linear relationship (Fig. 1). 3.4. Graphical interpretation of the response surface model The response surface curves are used to explain the relationship between the variables and to determine the optimum value of each variable to get the maximum conversion yield. The response surface curves are presented in Fig. 2. Each figure demonstrates the effect of two factors while the others are fixed at zero level. Fig. 2A shows the effect of varying p-HPA/Gly molar ratio and enzyme amount on the conversion yield checked at a constant temperature of 45 °C. It can be seen from Fig. 2A that at any given p-HPA/Gly molar ratio, the conversion yield increases considerably (from 52.3 to 74%) when the enzyme amount increases from 10 to 30 mg but decreases slightly when the enzyme amount increases from 30 to 50 mg. We can conclude from Fig. 2A that the major effect on the synthesis yield was provided by the enzyme amount. This phenomenon has also been observed for the lipase-catalyzed transesterification of triolein with ethyl ferulate [32]. These results can be explained by a mass transfer limitation at a high enzyme amount, which could affect the diffusion of the substrate to the enzyme active site and its availability for the reaction [33]. In addition, the presence of high enzyme amount could alter its active conformation and decrease its catalytic efficiency [34]. In fact, Torres et al. [35] reported that the decrease of the synthesis yield may be related to the enzymes agglomeration phenomenon that may lead to masking the active sites which cannot be exposed to the substrates. We can notice that the increase of the substrate molar ratio and the temperature leads to a decrease in the conversion yield (Fig. 2B). These results could be explained by the toxic effect of the high substrate concentrations on the enzyme activity. In fact,

C

Fig. 2. 2D diagrams showing: (A) the effect of the and enzyme amount (mg) on the 1,3-di-HPA-Gly conversion yield. Temperature is equal to 45 °C, (B) the effect of the p-HPA/glycerol molar ratio and temperature (°C) on the 1,3-di-HPA-Gly conversion yield. Enzyme amount is equal to 30 mg and (C) the effect of the enzyme amount (mg) and temperature (°C) on the 1,3-di-HPA-Gly conversion yield. p-HPA/glycerol molar ratio is equal to 5.

352

N. Kharrat et al. / Bioorganic Chemistry 75 (2017) 347–356

steric hindrance of the enzyme and the substrates may contribute to a decrease in the enzymatic activity [35,36]. As can be seen from Fig. 2B, at any given substrates molar ratio there is a slight increase in the conversion yield when the temperature passes from 35 to 45 °C and then it decreases slightly when it exceeds 45 °C. Fig. 2C shows the variation of the reaction yield in the plane enzyme amount (10–50 mg) versus temperature (35–55 °C) at a constant p-HPA/Gly molar ratio (equal to 2). This figure confirms the major effect of the enzyme amount on the synthesis yield shown in Fig. 2A. It also illustrates that at constant substrates molar ratio, the 1,3-dihydroxyphenylacetoyl-sn-glycerol synthesis carried out at a moderate reaction temperature (approximately 40 and 45 °C) and at an enzyme amount of 30 mg, gave a high conversion yield (74%). These results could be explained by the thermostability of Novozym 435 which is known by its tolerance for high temperatures (40–60 °C). In general, an elevated temperature within a certain range could influence both the solubility of reactants, the reaction rate and improve the conversion yield [37]. Conversely, treatment at high temperature may disrupt enzyme tertiary structure, making it inactivated or decreasing its catalytic efficiency [38].

Fig. 4. Effect of reusability of biocatalyst on the conversion yield of 1,3-dihydroxyphenylacetoyl-sn-glycerol. Reaction conditions: p-HPA/glycerol molar ratio of 2.2, acetonitrile/tert-butanol 1,1 (v:v) 3 ml, Novozym 435 amount of 30 mg, temperature 36 °C, agitation speed of 220 rpm, reaction time 25 h. Conversion yield was estimated using HPLC system. Bars indicate standard deviation.

3.6. Reusability of the enzyme

3.5. Optimal reaction conditions The optimization function in the Design Expert Software was used to obtain the optimal values of the three factors predicted by the model. These factors are X1 = 2.2, X2 = 36 °C and X3 = 30 mg referring respectively to p-HPA/Gly molar ratio, temperature and enzyme amount. The maximum predicted conversion yield was 78.375 ± 2.38%. To confirm the validity of the predicted optimal response, three additional experiments were carried out under the optimal conditions. The good agreement between the experimental and predicted results verifies the validity of the model. The experimental conversion yield of 1,3-di-HPA-Gly (77.12 ± 0.516%) was very close to the predicted estimated value (78.375 ± 2.38%) at a reaction time of 25 h (Fig. 3). These results indicate that RSM can easily describe the relationship between the different variables and their influence on the p-HPA ester synthesis. In fact, RSM is a very effective tool for optimizing individual factors in a new process. The experimental conversion yield of 1-HPA-Gly in the same optimal condition of 1,3-di-HPA-Gly synthesis was very low about 25.5% (data not shown).

Predicted yield: 78.375% Calculated yield : 77.12%

Fig. 3. Production under optimal conditions of 1,3-di-hydroxyphenylacetoyl-snglycerol during esterification reaction. Reaction conditions: enzyme amount of 30 mg, p-HPA/glycerol molar ratio of 2.2 and temperature of 36 °C. The p-HPA ester yield was estimated using HPLC system.

The reusability of enzyme for several reactions is an essential factor to allow a significant reduction in the cost of industrial production. Under the optimum reaction conditions, the reusability studies of Novozym 435 were carried out to evaluate their stability. In fact, Novozym 435 was removed from the organic solvent after each reaction by filtration and washed with ter-butanol/ acetonitrile twice, then reused. As shown in Fig. 4, Novozym 435 was successfully reused for 4 cycles without a significant decrease in the conversion yield of 1,3-di-HPA-Gly. Then, a remarkable decrease in the conversion yield was noticed from the fifth cycle. The result showed that Novozym 435 had a good stability in the reaction medium and had great potential for 1,3-dihydroxyphenylacetoyl-sn-glycerol production. 3.7. Structure determination of p-HPA derivatives Glycerol contains three hydroxyl groups in its structure, and therefore different esterified derivatives with p-HPA were expected. Fig. 5 shows that the enzymatic reaction of glycerol and p-HPA generates two new products eluted at a retention time of 6.89 min and 12.937 min. On the basis of the NMR, LC/MS-MS and FT-IR data, it was possible to determine the structure of these two new compounds. With regard to 13C NMR and 1H NMR data of the mono- or diesterified compound, one can say that using Novozyme 435 as biocatalyst, glycerol was esterified only on the primary hydroxyl group in the sn-1 and sn-3 position (Fig. 6B and D). In fact, we have found only 1-HPA-Gly and 1,3-di-HPA-Gly as products without the presence of 1,2-di-HPA-Gly. The absence of the last compound is may be due to the steric hindrance of the substrate p-HPA which prevents Novozyme 435 to reach and react with sn-2 position. Consequently, the 1,2-di-HPA-Gly cannot be occurred. In addition, the NMR analysis confirmed that the di-esterified glycerol was a symmetric compound. In addition, the NMR analysis showed that the di-esterified glycerol was a symmetric compound. The LC-MS/MS analysis (Fig. 6) allowed determining the structure of the new compounds. The MS1 analysis in the negative mode of the first peak (eluted at 6.89 min) (Fig. 6A) showed a molecular ion at m/z = 225.1694 [MH] attributed to the molecular weight of calculated 1-hydroxyphenylacetoyl-sn-glycerol using Molinspiration software (226.228 g/mol) shown in Table 4. The MS1

N. Kharrat et al. / Bioorganic Chemistry 75 (2017) 347–356

353

Fig. 5. HPLC chromatograms of the reaction medium before and after synthesis reaction. The separation was made on C18 reverse-phase HPLC. Flow 1.5 mL/min, and UV detection was at 280 nm. (A) Time of esterification = 0 h. (B) Time of esterification = 25 h. 1: p-HPA. 2: 1-HPA-Gly. 3: Gly-1,3-di-HPA. Synthesis was performed under optimal conditions.

experiments focusing on the fragment generated from the first peak in m/z = 225.1694 [MH] revealed a fragment corresponding to the molecular ion at m/z = 151.1554 attributed to p-HPA and a molecular ion at m/z = 271.0367 [M+COOH-H] attributed to the 1-hydroxyphenylacetoyl-sn-glycerol associated to a molecule of formic acid ion. The MS2 analysis in the negative mode of the fragment m/z = 271.0367 generated a molecular ion at m/z = 225.1694 which was formed by the neutral losses of formic acid linked to the hydroxyphenylacetoyl-sn-glycerol (Fig. 6B). In fact, the MS2 confirmed the structure of 1-hydroxyphenylacetoyl-snglycerol showed in Fig. 6B. However, the MS1 analysis in the negative mode of the second peak (eluted at 12.937 min) (Fig. 6C) showed a molecular ion at m/z = 359.2149 [MH] attributed to the molecular weight of calculated 1,3-dihydroxyphenylacetoylsn-glycerol using Molinspiration software (360.362 g/mol) shown in Table 4 and a molecular ion at m/z = 718.6716 [2MH] corresponding to the dimeric molecule. The MS2 analysis in the negative mode focusing on the fragment at m/z = 359.2149 [MH] revealed a fragment corresponding to the molecular ion at m/z = 225.1173 [Glycerol + p-HPA] attributed to the glycerol linked to one molecule of p-HPA and a molecular ion at m/z = 150.9878 attributed to p-HPA which confirms the linking of two molecules of p-HPA to glycerol (Fig. 6D). These profiles demonstrate that final compounds generated by enzymatic esterification are the 1-hydroxyphenylacetoyl-sn-glycerol (1-HPA-Gly) and 1,3dihydroxyphenylacetoyl-sn-glycerol (1,3-di-HPA-Gly). The production of the monoesterified glycerol (1-HPA-Gly) and the di-esterified glycerol (1,3-di-HPA-Gly) was confirmed by comparing their FT-IR spectra with those of the glycerol. The FT-IR spectra showed a peak at 1710 cm1 and at 1707 cm1 for the 1-HPA-Gly and 1,3-di-HPA-Gly, respectively. These peaks are attributed to a vibration of carbonyl group (C@O) and peak at 2982 cm1 attributed to CAH vibration of methylene groups for the two esters. In addition, a large band around 3500–3400 cm1 was attributed to the hydroxyl group linked to the aromatic ring. Other peaks were detected at 1598 and 1220 cm1 attributed to CAC and CAO respectively. The chemical change in functional groups of the simples indicated the formation of hydroxyphenylacetoyl glycerol esters.

3.8. Antioxidant activity The antioxidant activity was evaluated using two in vitro assays: the DPPH test, which measured the ability of the compounds to scavenge the free stable radical 2.2-diphenyl-1-

picrylhydrazyl, and the ABTS assay, which evaluated the activity toward a stable free radical, 2,2-azino-bis-3ethylbenzothiazoline6-sulfonicacid (ABTS+). The radical-scavenging activities of the evaluated antioxidants pHPA and their corresponding purified esters (1-hydroxyphenylacetoyl-sn-glycerol and 1,3-dihydroxyphenylacetoyl-sn-glycerol) and BHT are summarized in Table 4. The 1,3-dihydroxyphenylacetoyl-sn-glycerol presented interesting antioxidant properties with an IC50 about 9.115 ± 0.042 mg/mL, lower than the IC50 value of the initial substrate p-HPA (24.315 ± 0.085 mg/mL) and similar to those exhibited by the synthetic antioxidant, BHT, with an IC50 about 8.565 mg/mL. The same result was also observed in the ABTS test. In fact, the TEAC measured with synthesized ester is higher (2.93 mM Trolox equivalent) than the value obtained for each substrate phenolic acid or the monoesterified esters (1-HPAGly) (9.85 mM TEAC). In addition, the compounds scavenging activity of ABTS and DPPH radicals decreased in the following order: 1,3di-HPA-Gly
BHT > 1-HPA-Gly = p-HPA with all values significantly different at p < 0.05 (Table 4). These results indicate that this tested ester 1,3-di-HPA-Gly present a higher radical scavenging activity due to their hydrogen-donating ability and the presence of 2 phenol rings. Bendary et al. [39] showed that the antioxidant activity increased with increasing the number of the active group (OH or NH2) linked to the phenol ring. The findings of the present study could be the basis for a further exploitation of the new compound, 1,3-dihydroxyphenylacetoylsn-glycerol, as antioxidant active ingredient in the cosmetic and pharmaceutical fields, especially skin care as well as human inflammatory and aging processes.

3.9. Determination of the minimum inhibitory concentration (MIC) and the minimum bactericidal concentration (MBC) The antibacterial activity of p-HPA and the synthesized esters (1-hydroxyphenylacetoyl-sn-glycerol, 1,3-dihydroxyphenylacetoyl-sn-glycerol) were checked against Gram-positive (Bacillus cereus (BC), Micrococcus luteus (ML), Staphylococcus xylosus (Sx), Staphylococcus aureus (Sa) and Staphylococcus epidermidis (SEp)) and Gram-negative (Enterobacter cloacae (EC), Pseudomonas aeruginosa (Ps), Klebsiella pneumonia (KP), and Escherichia coli (Ecoli)) bacteria by the determination of MIC and MBC values. As can be seen from Table 5, p-HPA has not an bactericidal effect until a concentration of 4 mg/mL especially against Gram (). Compared to p-HPA and monoesterified p-HPA (1-HPA-Gly), synthesized ester 1,3-di-HPA-Gly exhibits the highest effect against bacteria,

354

N. Kharrat et al. / Bioorganic Chemistry 75 (2017) 347–356 100

225.1694

(A) MS1

225.1694 [M-H]271.0367

95 90

271.0367 [M+HCOO--H]-

Relative Abundance

85 80 75 70 65 60 55 50 45 40 35 30 25 20

p-HPA 151.1554

151.1554

15 10 359.1495

5

730.1743

504.0587549.7999 179.1319

325.0257

0 150

200

250

373.1709

300

496.4049

451.0155

350

400

450

576.0893

500

600

766.1273

676.3778

638.2916

550

650

700

797.5552

750

927.1020

891.7495

800

850

900

m/z 95

(B) MS2 of 271

225.1003 [M-H]-

90 85 80 75

Relative Abundance

70

OH

65

2" CH 2' H 2C 1 CH 2

55

OH

50 45

O

40

6'

4'

60

3' C

H 2C

O

35

7'

5'

8' 10'

9'

30 25

OH

1-Hydroxyphenylacetoyl-sn-Glycerol (1-HPA-Gly)

20 15 10 5 0

151.1916

140

178.8015

160

197.1798

180

251.3557

200

220

240

260

280

300

320

340

360

380

400

420

440

460

480

500

m/z 95

(C) MS1

359.2149 [M-H]-

90 85 80

Relative Abundance

75 70 65 60 55 50 45 40

[2M-H]-

718.6716

35

718.6716

30 25 20 15 10

395.0421 151.1270

5 0

427.0178

225.1760

522.1534

200

300

400

95

500

150.9878

90

772.1352

630.3422

600

700

907.5168

800

900

1 1 9 7 . 9 6 0 4

984.1694

1000

m/z

1100

1200

1307.6979

1300

1 4 8 7 . 0 4 8 5

1400

1 7 9 7 . 6 7 9 9

1666.5414

1500

1600

1700

1800

1 9 7 6 . 9 4 5 2

1900

2000

(D) MS2 of 359

[p-HPA]

85 80

7"

Relative Abundance

75

OH

6"

4"

70 65

HO

60

5"

8"

3" C O

H2C

6'

2" CH 2' H 2C 1 CH2

O

3' C

4' CH2

7'

5'

8'

OH

55 50

9"

45

O

10"

O

10'9'

40

1,3-di-Hydroxyphenylacetoyl-sn- Glycerol (1,3-di-HPA-Gly)

35 30 25 20

225.1173 [Glycerol + p-HPA]-

15

225.1173

10 5

207.0450

107.1425

121.0776

0 100

241.1128

133.0481 163.0717

120

140

160

177.0652

285.2801

180

200

220

240

260

280

300

320

340

360

m/z

Fig. 6. LC-MS/MS (negative mode) profiles of peak identified as 1-HPA-Gly and the other peak identified as 1,3-di-HPA-Gly. (A) and (B) MS1 and MS2 spectra of the first eluted peak (C) and (D) MS1 and MS2 spectra of the second peak.

355

N. Kharrat et al. / Bioorganic Chemistry 75 (2017) 347–356 Table 4 The antioxidant activity of the hydroxyphenylacitoyl Glycerol derivatives and physico-chemical parameters of these compounds related to their structures. Compounds

Abbreviations

MW (g/mol)a

miLog Pa

IC 50 (mg/mL)b,*

TEAC (mM)c,*

Glycerol p-hydroxyphenyl acetic acid 1-hydroxyphenylacetoyl-sn-Glycerol 1,3-di-hydroxyphenylacetoyl-sn-Glycerol Butylhydroxytoluene

Gly p-HPA 1-HPA-Gly 1,3-di-HPA- Gly BHT

92.094 152.149 226.228 360.362 220.356

1.596 0.883 0.223 2.042 5.435

– 24.315 ± 0.085 24.855 ± 0.073 9.115 ± 0.042 8.565 ± 0.074

– 9.86 ± 0.031 9.85 ± 0.026 2.93 ± 0.007 4.25 ± 0.091

The results are the average of 3 repetitions ± SD. * Statistically significant data (p < 0.05). a miLog P and MW values calculated using Molinspiration program. b Antioxidant activity of the different compounds determined by DPPH method. c Antioxidant activity of the different compounds determined by ABTS assay. Results are expressed as Trolox equivalent antioxidant capacity (TEAC) in units of mmol Trolox/L.

Table 5 Antibacterial activities of p-HPA, 1-HPA-Gly et 1,3-di-HPA-Gly. Bacteria strain

Bacillus cereus Micrococcus luteus Staphylococcus aureus Staphylococcus xylosus Staphylococcus epidermidis Pseudomonas aeruginosa Enterobacter cloacae Klebsiella pneumoniae Escherichia coli

Gram

+ + + + + – – – –

MIC (mg/mL)

MBC (mg/mL)

p-HPA

1-HPA-Gly

1,3-di-HPA-Gly

p-HPA

1-HPA-Gly

1,3-di-HPA-Gly

2 1 1 1 2 4 4 4 4

2 1 1 1 2 >4 4 4 >4

0.5 0.25 0.25 0.25 0.25 1 1 2 2

4 2 2 2 4 >4 >4 >4 >4

4 2 2 2 4 >4 >4 >4 >4

2 1 1 1 1 4 4 >4 >4

especially Gram (+) ones. In fact, with Gram (+) bacteria, MIC values varied from 0.25 to 0.5 mg/mL and the MBC values are between 1 and 2 mg/mL, while Gram () bacteria appear to be less sensitive to the tested compounds. In fact, this higher resistance may be explained by the difference compositions in the membrane bacteria cell. As can be seen from Table 5, MIC and MBC values showed that 1,3-dihydroxyphenylacetoyl-sn-glycerol exhibited an interesting inhibitory activity and more bactericidal effect against Gram (+) bacteria than the original substrate (p-HPA). This effective antibacterial activity of the synthesized compound, 1,3-di-HPAGly, could be attributed to the presence of two molecules of pHPA with their two benzene rings which present two hydroxyl groups. These results indicate that we have significantly ameliorated the antibacterial activity of the p-HPA after esterification. Similar results were reported by Fu et al. [40] who showed that caffeic acid amides with electron donating groups at p-position of benzene ring had a better inhibitory activity against bacteria. In fact, the enzymatic synthesis of phenol derivatives can be used as a tool to improve their biological and therapeutic functions. 4. Conclusion In the present study, a Box-Behnken design of RSM was employed to optimize the synthesis conditions of 1,3dihydroxyphenylacetoyl-sn-glycerol. Under optimal conditions, a conversion yield of 77.12 % was obtained. The 1,3-di-HPA-Gly showed interesting antioxidant and antibacterial activities. Moreover, this ester, 1,3-dihydroxyphenylacetoyl-sn-glycerol, was more efficient than p-HPA, 1-HPA-Gly and BHT. In fact, the use of glycerol showed that this by-product could be converted into healthimproving functional compounds. Furthermore, the findings of the present study could be the basis for a further exploitation of the new synthesized compound, 1,3-dihydroxyphenylacetoyl-snglycerol, as antioxidant and active ingredients in the cosmetic and pharmaceutical fields, especially skin care as well as human inflammatory and aging processes.

Acknowledgements This work is part of a doctoral thesis by Kharrat Nadia. We are grateful to Dr. Bouaziz Mohamed for LC-MS/MS analysis, to Dr. Dghachi Youssef for NMR analysis and to Amina Zineddine for FT-IR analysis. We thank Professor Robert Verger for his fruitful discussion. The authors are indebted to Professor Hafedh Bejaoui (FSS) and Anouar Fendri (FSG) for English language correction. This study received financial support from the ‘Ministry of Higher Education and Scientific Research’ – ‘Tunisia’.

References [1] M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi, C.D. Pina, From glycerol to valueadded products, Angew. Chem. Int. Ed. 46 (2007) 4434–4440. [2] M. Pagliaro, M. Rossi, The future of glycerol, second ed., RSC Green Chemistry Royal Society of Chemistry, Cambridge, UK, ISBN No. 8, (978-1-84973-046-4), 2010. [3] A. Behr, J. Eilting, K. Irawadi, J. Leschinski, F. Lindner, Improved utilisation of renewable resources: new important derivatives of glycerol, Green Chem. 10 (2008) 1463–9262. [4] V. Kristinova, R. Mozuraityte, I. Storro, T. Rustad, Antioxidant activity of phenolic acids in lipid oxidation catalyzed by different prooxidants, J. Agric. Food Chem. 57 (2009) 10377–10385. [5] I. Casaburi, F. Puoci, A. Chimento, R. Sirianni, C. Ruggiero, P. Avena, V. Pezzi, Potential of olive oil phenols as chemopreventive and therapeutic agents against cancer: a review of in vitro studies, Mol. Nutr. Food Res. 57 (2013) 71– 83. [6] S. Schaffer, W.E. Müller, G.P. Eckert, Cytoprotective effects of olive mill wastewater extract and its main constituent hydroxytyrosol in PC12 cells, Pharmacol. Res. 62 (2010) 322–327. [7] S. Sanchez-Fidalgo, L. Sánchez de Ibargüen, A. Cárdeno, C. Alarcón dela Lastra, Influence of extra virgin olive oil diet enriched with hydroxytyrosol in a chronic DSS colitis model, Eur. J. Nutr. 51 (2012) 497–505. [8] J. Fernández-Bolaños, G. Rodríguez, R. Rodríguez, R. Guillén, A. Jiménez, Potential use of olive by-products: extraction of interesting organic compounds from olive oil waste, Grasas y Aceites. ENERO-MARZO 57 (2006) 95–106. [9] R. Tunalıog˘lu, Ç. Koç, T. Bektasß, A multiperiod location-routing problem arising in the collection of Olive Oil Mill Wastewater, J. Oper. Res. Soc. 67 (7) (2016) 1012–1024.

356

N. Kharrat et al. / Bioorganic Chemistry 75 (2017) 347–356

[10] L. Abaza, W. Taamalli, M. Zarrouk, et al., Natural antioxidant composition as correlated to stability of some Tunisian virgin olive oils, Riv. Ital. Sostanze Gr. 132 (2005) 12–18. [11] C. Calabriso, M. Massaroa, E. Scodittia, S. D’Amoreb, A. Gnonic, M. Pellegrino, C. Storellid, R. De Caterinae, G. Palascianob, M.A. Carluccioa, Extra virgin olive oil rich in polyphenols modulates VEGF-induced angiogenic responses by preventing NADPH oxidase activity and expression, J. Nutr. Biochem. 28 (2016) 19–29. [12] M. Vazquez-Velasco, L. Esperanza Díaz, R. Lucas, S. Gómez-Martínez, S. Bastida, A. Marcos, F.J. Sánchez-Muniz, Effects of hydroxytyrosol-enriched sunflower oil consumption on CVD risk factors, Br. J. Nutr. 105 (2011) 1448–1452. [13] Z. Bouallagui, J. Han, H. Isoda, S. Sayadi, Hydroxytyrosol rich extract from olive leaves modulates cell cycle progression in MCF-7 human breast cancer cells, Food Chem. Toxicol. 49 (2011) 179–184. [14] G. Rodríguez, R. Rodríguez, J. Fernández-Bolaños, R. Guillén, A. Jiménez, Antioxidant activity of effluents during the purification of hydroxytyrosol and 3,4-dihydroxyphenyl glycol from olive oil waste, Eur. Food Res. Technol. 224 (2007) 733–741. [15] M.T. Hassan Khan, A. Ather, Potentials of phenolic molecules of natural origin and their derivatives as anti-HIV agents, Biotechnol. Annu. Rev. 13 (2007) 223–264. [16] V. Carito, M. Ceccanti, L. Tarani, G.N. Ferraguti, G. Chaldakov, M. Fiore, Neurotrophins’ modulation by olive polyphenols, Curr. Med. Chem. 23 (2016) 3189–3197. [17] A. Iacono, R. Gómez, J. Sperry, J. Conde, G. Bianco, R. Meli, et al., Effect of oleocanthal and its derivatives on inflammatory response induced by LPS in chondrocyte cell line, Arthritis Rheum. 62 (2010) 1675–1682. [18] J. Graca, H. Pereira, Suberin structure in potato periderm: glycerol, long-chain monomers, and glyceryl and feruloyl dimers, J. Agric. Food Chem. 48 (2000) 5476–5483. [19] J.A. Laszlo, K.O. Evans, K.E. Vermillion, M. Appell, Feruloyl dioleoylglycerol antioxidant capacity in phospholipid vesicles, J. Agric. Food Chem. 58 (2010) 5842–5850. [20] J. Rozema, A.J. Noordijk, R.A. Broekman, A. Van Beem, B.M. Meijkamp, N.V.J. De Bakker, J.W.M. Van de Staaij, M. Stroetenga, S.J.P. Bohncke, M. Konert, S. Kars, H. Peat, R.I.L. Smith, P. Convey, (Poly) phenolic compounds in pollen and spores of Antarctic plants as indicators of solar UV-b a new proxy for the reconstruction of past solar UV-b, Plant Ecol. 154 (2001) 9–26. [21] K. Warner, J. Laszlo, Addition of ferulic acid, ethyl ferulate, and feruloylated monoacyl- and diacylglycerols to salad oils and frying oils, J. Am. Oil Chem. Soc. 82 (2005) 647–652. [22] M. Tsuchiyama, T. Sakamoto, T. Fujita, S. Murata, H. Kawasaki, Esterification of ferulic acid with polyols using a ferulic acid esterase from Aspergillus niger, Biochim. Biophys. Acta 1760 (2006) 1071–1079. [23] C. Vafiadi, E. Topakas, V.R. Nahmias, C.B. Faulds, P. Christakopoulos, Feruloyl esterase catalysed synthesis of glycerol sinapate using ionic liquids mixtures, J. Biotechnol. 139 (2009) 124–129. [24] R. Myers, D.C. Montgomery, Response Surface Methodology: Process and Product Optimization Using Designed Experiments, second ed., Wiley, New York, 2002.

[25] N. Kharrat, I. Aissa, M. Sghaier, M. Bouaziz, M. Sellami, D. Laouini, Y. Gargouri, Lipophilization of ascorbic acid: a monolayer study and biological and antileishmanial activities, J. Agric. Food Chem. 62 (2014) 9118–9127. [26] W. Brand-Williams, M. Cuvelier, C. Berset, Use of a free radical method to evaluate antioxidant activity, Lebenson-Wiss Technol. 28 (1995) 25–30. [27] R. Bernini, F. Crisante, M. Barontini, D. Tofani, V. Balducci, A. Gambacorta, Synthesis and structure/antioxidant activity relationship of novel catecholic antioxidant structural analogues to hydroxytyrosol and its lipophilic esters, J. Agric. Food Chem. 60 (2012) 7408–7416. [28] J. Viskupicova, M. Danihelova, M. Ondrejovic, T. Liptaj, E. Sturdik, Lipophilic rutin derivatives for antioxidant protection of oil based foods, Food Chem. 123 (2010) 45–50. [29] J.N. Eloff, A sensitive and quick microplate method to determine the minimal inhibitory concentration of plant extracts for bacteria, Planta Med. 64 (1998) 711–713. [30] V.C. Okore, Evaluation of chemical antimicrobial agents. Bacterial resistance to antimicrobial agents. Pharmaceutical Microbiology, El’Demark publishers, Nsukka, 2005, pp. 55–120. [31] A.M. Joglekar, A.T. May, Product excellence through design of experiments, Cereal Food World 32 (1987) 857. [32] Y. Zheng, X.M. Wu, C.B. White, J. Quan, L.M. Zhu, Dual response surface optimized process for feruloylated diacylglycerols by selective lipasecatalyzed transesterification in solvent free system, Bioresour. Technol. 100 (2009) 2896–2901. [33] T. Palmer, Kinetics of multi-substrate enzyme-catalysed reactions, in: UnderStanding Enzymes, fourth ed. Hertfordshire, Great Britain, Prentice-Hall/Ellis Horwood, 1995, pp. 155–174. [34] G.D. Yadav, P.S. Lathi, Kinetics and mechanism of synthesis of butyl isobutyrate over immobilized lipases, Biochem. Eng. J. 16 (2003) 245–252. [35] M. Torres, P.G. Del, M.L. Foresti, M.L. Ferreira, Cross-linked enzyme aggregates (CLEAs) of selected lipases: a procedure for the proper calculation of their recovered activity, AMB Express. 3 (2013) 25. [36] N.M. Hadzir, M. Basri, M.B.A. Rahman, C.N.A. Razak, R.N.Z.A. Rahman, A.B. Salleh, Enzymatic alcoholysis of triolein to produce wax ester, J. Chem. Technol. Biotechnol. 76 (2001) 511–515. [37] P.Y. You, E.Z. Su, X.P. Yang, D.B. Mao, D.Z. Wei, Carica papaya lipase catalyzed synthesis of terpene esters, J. Mol. Catal. B – Enzym. 71 (2011) 152–158. [38] W. Chun, F. Xian-Feng, W. Zhao, Y. Xin-Jun, Z. Yin-Jun, Z. Jian-Yong, Efficient synthesis of vitamin E intermediate by lipase-catalyzed regioselective transesterification, J. Mol. Catal. B – Enzym. 106 (2014) 90–94. [39] E. Bendary, R.R. Francis, H.M.G. Ali, M.I. Sarwat, S. El Hady, Antioxidant and structure–activity relationships (SARs) of some phenolic and anilines compounds, Ann. Agric. Sci. 58 (2013) 173–181. [40] J. Fu, K. Cheng, Z. Zhang, R. Fang, H. Zhu, Synthesis, structure and structure– activity relationship analysis of caffeic acid amides as potential antimicrobials, Eur. J. Med. Chem. 45 (2010) 2638–2643.