Send Orders for Reprints to
[email protected] Current Bioactive Compounds 2015, 11, 131-141
131
Secondary Plant Metabolites LogP Determination: the Case of Boropinic and Geraniloxyferulic Acids Isabella Epifanioa, Salvatore Genovesea, Giuseppe Carluccia, Francesco Epifanoa and Marcello Locatellia,b* a
University “G. d’Annunzio” Chieti-Pescara, Department of Pharmacy, via dei Vestini 31, 66100 Chieti, Italy; bInteruniversity Consortium of Structural and Systems Biology, Viale Medaglie d’oro 305, 00136 Roma, Italy Abstract: Lipophilicity, expressed quantitatively through the LogP, is the most important physicochemical property for pharmacodynamic and pharmacokinetic characterization of the tested compound, as well as the most widely used parameter for quantitative structure-activity relationships (QSAR) descriptions. For this purpose three analytical techniques: the classical Shake Flask, the innovative Vortex-Assisted Liquid-Liquid Micro-Extraction (VALLME), and HPLC method, have been employed for the LogP determination of naturally occurring oxyprenylated secondary metabolites such as boropinic and geraniloxyferulic (GOFA) acid. These techniques have allowed us to correctly determine the LogP value for model molecule (ferulic acid) and from the comparison, by means of statistical techniques, of the obtained results it was possible to determine, not only for the first time, the LogP value of these secondary metabolites with interesting biological activity, but it was also possible to highlight the potential and limitations of the used techniques respect to the presence of interfering and / or degradation products.
Keywords: Boropinic acid, lipophilicity evaluation, 3-(4'-geranylxy-3'-methoxyphenyl)-2-trans-propenoic acid, prenyloxyphenylpropanoids, RP-HPLC method, Shake-flask, VALLME method. 1. INTRODUCTION Secondary metabolites of phenylpropanoid biosynthetic origin containing sesquiterpenyl, monoterpenyl and isopentenyl chains attached to a phenol group represent quite a rare class of natural compounds. Only in the last decade these products were studied extensively from a chemical and pharmacological point of view. This group of natural products is mainly found in plants belonging to the Rutaceae, Apiaceae and Compositae families. In particular, prenyloxycinnamic acids were shown to exert promising anti-inflammatory and anti-cancer activity and they seem to play also an important role in the treatment of neurological disorders. Among these, we focused our interest on chemical and pharmacological properties of 4'geranyloxyferulic acid [3-(4'-geranyloxy-3'-methoxyphenyl)2-trans-propenoic acid (GOFA)] (Fig. 1) and boropinic acid [3-(4'-isopentenyloxy-3'-methoxyphenyl)-2-trans-propenoic acid] (Fig. 1). These secondary metabolites are biosynthetically related to ferulic acid (Fig. 1) in which a geranyl and isopentenyl side chain, respectively, are attached to the phenolic group [1]. GOFA was isolated in 1966 from the bark of Acronychia baueri Schott, while boropinic acid was isolated for the first *Address correspondence to this author at the University “G. d’Annunzio” Chieti-Pescara, Department of Pharmacy, via dei Vestini 31, 66100 Chieti, Italy; Tel: +39 0871 3554590; Fax: +39 0871 3554911; E-mail:
[email protected] 1875-6271/15 $58.00+.00
Fig. (1). Chemical structures for the analytes. time in 2000 from the aerial parts of Boronia pinnata, a small tree and a herb, respectively, both belonging to the Australian flora [2]. Only in the last ten years some of the pharmacological properties of these natural metabolites began to be characterized. GOFA showed a series of interesting biological effects such as cancer chemoprevention and other effects closely
© 2015 Bentham Science Publishers
132 Current Bioactive Compounds 2015, Vol. 11, No. 3
related to cancer growth and development as well as a valuable anti-inflammatory and neuroprotective activity [3]. Boropinic acid, instead, showed to exert inhibitory effects both in vitro and in vivo against Helicobacter Pylori [4] with an efficacy comparable to antibiotics used to eradicate the bacterium [5]. Literature presents several articles that describe the phytochemical and pharmacology of these compounds, but there are few data related to their physical-chemical properties. For this purpose, we report herein a comparison between the conventional shake-flask method, the RP-HPLC and the innovative Vortex-Assisted Liquid-Liquid Micro-Extraction (VALLME) for the LogP determination of GOFA and boropinic acid. LogP is the parameter most widely used for the measurement of lipophilicity of a lead compound [6], which describes the partitioning of a compound between an aqueous and a lipid environment. The knowledge of this physicalchemical parameter could be helpful in the determination of many pharmacological parameters such as the ADME profile [7-8], protein binding and Central Nervous System (CNS) penetration [9]. It is also the physical-chemical property most widely used in the drug discovery process to describe the quantitative structure-activity relationships (QSAR) in pharmaceutical field and in biological and environmental analyses [10]. According to the Organization for Economic Cooperation and Development guidelines (OECD), the shake-flask is the standard method for determination of the partition coefficient in a biphasic system composed by water/octanol [11]. This method allows making a direct measurement of LogP in a biphasic system without approximation [12]. However, this procedure is time-consuming, tedious and it is often subject to the emulsions formation [13]. To this aim, in the last decade, new strategies have been developed to speed up, and optimize the LogP measurement. Reversed-phase High Performance Liquid-liquid Chromatography (RP-HPLC) techniques, have proven to simulate, in an optimal way, octanol-water partitioning, and are considered as popular alternatives for lipophilicity assessment [14-15]. The LogP determination by RP-HPLC is based on measurement of the retention factor k for the investigated compound; the extrapolation of Logk values to 100% of water (which leads to Logkw) is a convenient means of standardizing chromatographic lipophilicity parameters [16]. This approach offers several practical advantages, including speed, reproducibility, not affected by the presence of impurities or degradation products, broader dynamic range, on-line detection and reduced sample handling and sample size [17]. Moreover, the mobile/stationary phase interphase models are better for the biological partitioning process than the solute partitioning in the bulk octanol/water phase [18]. RP-HPLC obtained wide acceptance in lipophilicity assessment and has officially been recommended by the OECD [19].
Epifanio et al.
Recently, for the direct determination of LogP, was also proposed a system in microextraction size, known as vortexassisted liquid-liquid microextraction method (VALLME) coupled to a HPLC or UV/Vis analysis. With this approach, micro-volume of octanol is added to the aqueous sample and subsequently dispersed with the help of a vortex. In this way, the formed microdrops ensure a faster partitioning, mainly due to a reduced diffusion distance and to a wider specific surface. Apart from its simplicity and the low solvent consumption, one of the main advantages of this method is the minimum time required to reach the equilibrium [20]. In continuation to our studies on natural products analyses and characterization [21-39], and instrument configurations [40-43], we report herein, for the first time, the LogP of the two secondary metabolites by the means of these three techniques and to highlight the limitations associated with the use of the classic shake flask method. All results, then, were also submitted to a statistical study to evaluate the significativity of the data obtained for each method. 2. MATERIALS AND METHODS 2.1. Reagents and Chemicals 3-(4'-geranyloxy-3'-methoxyphenyl)-trans-2-propenoic acid and boropinic acid were synthesized as previously reported [44] and their purity was assessed by GC-MS (purity > 99%). Ferulic acid (purity 99%), used as a reference compound, was purchased from Sigma Aldrich (Steinheim, Germany) and used without further purification. Analyses were conducted using methanol, acetonitrile, and acetone (all HPLC grade) purchased from Carlo Erba (Milan, Italy) and 1-octanol (HPLC-grade) purchased from Sigma Aldrich (Steinheim, Germany). Water was purified by a Millipore Milli-Q system (Millipore Bedford, MA, USA) and then used for the measurements. For the buffer preparation was employed dihydrogen sodium phosphate and sodium hydrogen phosphate (purity 99%) purchased from Sigma Aldrich (Steinheim, Germany) and ammonium acetate (purity 98%) purchased from Fluka (Steinheim, Germany). 2.2. Instrumentation for Shake-Flask and VALLME Methods The spectrophotometric analyses were conducted on a Cary 50 Scan Instrument (Varian, CA, USA) with Cary WinUV version 3,00 for data acquisition and elaboration. The absorbance measurements were carried out in quartz cells purchased from Hellma (Milan, Italy) with 1 cm cell path length, at a constant temperature of 25°C (±1°C). Samples centrifugation was conducted on a Nuve NF048 (Brussels, Belgium) at 12.000 rpm for 2 minutes. 2.3. Preparation of Standard Solutions for Shake-Flask and VALLME Methods Standard stock solutions of GOFA, ferulic and boropinic acid were prepared by dissolving the pure powder of each analyte in Milli-Q water, in order to achieve a final concentration of 1 mg/mL. By appropriate dilution of the stock solution with Phosphate buffer 0.01 M at pH = 2, pre-saturated with octanol, three working solutions at concentration of 400
Current Bioactive Compounds 2015, Vol. 11, No. 3
Secondary Plant Metabolites LogP Determination
μg/mL were obtained. Then all the aliquots were stored at 4° C for no more than 4 weeks. 2.4. Calibration, Linearity and Quality Control Samples for Shake-Flask and VALLME Methods Spectrophotometric calibration was performed using seven standard solutions prepared by dilution of the working solutions with phosphate buffer at pH = 2, pre-saturated with octanol. The concentration range (0.1, 0.5, 1, 2.5, 3, 4, 6 μg/mL) was calculated in order to obtain optimal absorbance values comprised between 0.1 and 1 (arbitrary units, a.u.). The spectrophotometric readings were conducted at a wavelength of 280 nm for ferulic acid and 268 nm for boropinic acid and GOFA. Calibration curves were plotted using a linear regression model according to the equation y = a + bx, where "y" is the absorbance value, "x" the analyte concentration, while "a" and "b" are the intercept and the slope of the regression line, respectively. All analyses showed a good linearity in the range of concentrations tested (0.1 - 6 μg/mL) and a coefficient of determination R2 0.9579. Precision and trueness were daily evaluated by measuring the absorbance of Quality Control samples (QCs) at three different concentrations (0.25 μg/mL (low level), 2 μg/mL (medium level), and 3.5 μg/mL (high level)), and using standard deviations and BIAS%, respectively. 2.5. LogP Determination by Shake-Flask and VALLME Methods With shake-flask method was prepared 1 L of phosphate buffer solution 0.01 M at pH = 2 to minimize the molecular ionization in the aqueous phase during analysis. Then, a part of solution was introduced in a separator funnel with an equivalent volume of octanol. The resulting two phase’s mixture was repeatedly inverted for 15 minutes and let stand for 24 hours. After separation, solutions of 4 mg/mL for each analyte were prepared diluting 20 μL of working solution (400 mg/mL), with phosphate buffer pre-saturated in octanol. Then a little volume of this solution was added to an equivalent volume of octanol pre saturated in buffer. The resulting two-phase mixtures were placed in a rotating shaker for 1 and 3 hours, respectively. After centrifugation and separation, the concentration of each analyte in the aqueous phase was determined by spectrophotometry UV/Vis. Absorbance values were obtained at 280 nm for ferulic acid and at 268 nm for boropinic acid and GOFA. Calibration curves have allowed obtaining the equilibrium concentrations by the absorbance values reported and the corresponding LogP values, according to the equation 1. analyte analyte in eq LogP = Log analyte eq
eq. 1
Instead, the VALLME method follows a different experimental procedure.
133
In this case solutions at 4 mg/mL were prepared for each analyte, by appropriate dilution of the working solution with phosphate buffer saturated in octanol, and 1 mL of each solution was placed in a centrifuge tube. With the use of a microsyringe, 25 μL or 50 μL of octanol saturated in phosphate buffer were slowly introduced and the mixture was then vigorously shaken using a vortex for 4 minutes at maximum setting. The two phases were subsequently separated by centrifugation at 12.000 rpm for 2 minutes. The octanol phase restored its initial single microdrop shape on the upper surface of the sample solution and the aqueous phase could be collected with the help of a micro syringe and used for spectrophotometric analysis. All the absorbance values were obtained at the same wavelengths used for the shake-flask method. Even in this case calibration curves have allowed obtaining the equilibrium concentrations by the absorbance values reported and so the corresponding LogP values, according to the equation 1. 2.6. Chromatographic Condition RP-HPLC analyses were performed using a Waters liquid chromatograph equipped with a model 600 solvent pump (Waters, Milford, MA USA), an injector loop with a capacity of 20 μL model 7125 Rheodine (Rheodine, Cotati, CA, USA), a degasser Degassex mod. DG-4400 (Phenomenex, Torrance, CA, USA), while a 2996 photodiode array detector and Empower v.2 Software (Waters, Milford, MA, USA) was used for data acquisitions. The UV/Vis acquisition wavelength was set in the range of 200 and 400 nm. Chromatographic analysis was performed using a C18 reverse phase column (GraceSmart RP18, 4.6x 150 mm, 5 mm; Grace Deerfield, IL USA) thermostated at 25°C (±1°C) using a jetstream2 Plus column oven (Aurora Borealis Control, Netherlands). All measurements were achieved under isocratic elution conditions at 1 mL/min flow rate and different percentage of organic modifier. 2.7. LogP Determination by RP-HPLC Solutions at 50 μg/mL of GOFA and boropinic acid and solution at 10 μg/mL of ferulic acid were prepared by appropriate dilution of the stock solutions with 0.01 M phosphate buffer at pH = 2. The samples were analyzed by chromatography separately on an octadecyl silica column (C18). The retention behavior of the cited analytes was investigated using a mobile phase consisting of a mixture of two solvents in different percentages: phosphate buffer 0.01 M at pH = 2 (solvent A) and methanol (solvent B). Under isocratic conditions, three injections were made for each compound at each methanolbuffer ratio and Logk values were calculated from retention time using the formula below:
k=
tr t0 t0
eq. 2
where tr is the retention time of the investigated compound and t0 is the column dead time. Each analysis was performed using a percentage of methanol between 90 and 30% for ferulic acid, between 90 and 50% for boropinic acid and 90
134 Current Bioactive Compounds 2015, Vol. 11, No. 3
Epifanio et al.
and 60% for GOFA. When using methanol, isocratic Logk values are linearly correlated with the organic modifier percentage in the mobile phase, as exhibited by (Fig. 2), and the extrapolation of the Logk to 100% water (0% methanol) allows obtaining the corresponding LogP value. extrapolation to 100% water com position
The results observed for ferulic acid (1.218 ± 0.291) showed that, after one hour, the three compounds reach the distribution equilibrium between the two phases. The significance of the results obtained was confirmed statistically by performing a t-test at 98% confidence level, demonstrating the reliability of the values obtained, as show in Table 1. The negative LogP values of GOFA and boropinic acid could be related to possible degradation compounds in the corresponding ferulic acid and/or to a possible interfering products presence.
1.25
Logk'
With the VALLME approach, in contrast to the previous method, micro volumes of octanol are dispersed into an aqueous sample using vortex mixing, a mild emulsification procedure.
0.25
10
20
30
40
50
60
70
80
% MeOH -0.75
Fig. (2). Linear correlation between Logk and percentage of MeOH for a solution of ferulic acid and extrapolation of the corresponding LogP to 100% water.
3. RESULTS AND DISCUSSIONS 3.1. Methods Comparison Lipophilicity is a physical-chemical parameter that measures the distribution of a compound between two immiscible solvent phases and is experimentally determined by the logarithm of the partition coefficient for single electrical species: analyte oct LogP o = Log analyte w wat
eq. 3
The 1-octanol/water system is the widely accepted reference system for the determination of lipophilicity. The method most commonly used for this type of measurement is the shake-flask, which allows to a direct measure of the concentration of compound between the two phases after equilibration. In this work, the LogP determination of the twooxyprenylated secondary metabolites was initially performed using the classical approach, to establish the most appropriate conditions of analysis, and then with other two methods (VALLME and RP-HPLC). In all cases ferulic acid was used as reference compound. The shake-flask analysis has highlighted the need of a deep pH control through the use of a buffer. Analyzing the compounds only in water saturated with octanol, anomalous values were obtained due to the acid ionization of the three molecules. Thus, according to the definition of LogP, it was necessary the use of a buffer to maintain all three compounds in the undissociated form. The best LogP values have been achieved using phosphate buffer (0.01 M) at pH = 2 for 1 and 3 hours of analysis, respectively.
The major advantage of this method is that the microdroplets formed ensure fast partitioning rates due to the shorter distance and larger specific surface area. This means that the time needed to reach the equilibrium is expected to be faster when using the VALLME approach. In this case, it is recommended that partition coefficient must to be determined using a low solute concentration: in this way it is possible to prevent solute self-association and maintain activity coefficient near or very close to unit [20]. Several experiments were conducted changing the microvolumes of octanol used and the time needed to attain steady state conditions. It was observed that the solutions of ferulic acid treated with 50 and 25 μL reach the equilibrium distribution after four minutes. Even in this case, the t-test at 98% confidence level (Table 2), shows that this method can be used for LogP determination of ferulic acid, while negative values were obtained for GOFA and boropinic acid. From a comparison between these data and those obtained with the previous method, it is evident that the LogP values obtained with the shake-flask are less negative. This is probably due to a consistent degradation process of the two compounds after 1 and 3 hours of analysis. The presence of ferulic acid, which is less lipophilic, could increase the value of LogP calculated with the shake-flask. In any case, the small measurable LogP range (0.5-3), and the high sensitivity of the VALLME method to the presence of impurities or degradation products, explains the anomalous results obtained for GOFA and boropinic acid. Therefore the study was conducted, in the last step, by means of RP-HPLC technique, which is insensitive to the presence of interfering or degradation products due to the fact that in this case only retention times (in order to obtain retention factors) are necessary to extrapolate the LogP value. The chromatographic determinations were carried out with a classical reverse phase column C18, the most used for the evaluation of lipophilicity. However, the interference of silanophilic interactions in the partitioning mechanism in RP-HPLC has been recog-
Current Bioactive Compounds 2015, Vol. 11, No. 3
Secondary Plant Metabolites LogP Determination
Table 1.
t-test comparison with known value for the shake-flask method. 1 hour Known
Mean
Std. Dev.
3 hours Comparison
n.
value
Mean
Std. Dev.
Comparison
n.
known value1
known value 1
Ferulic acid
1.64
1.218
0.291
5
tobs
