Natural Deep Eutectic Solvents (NADES) to Enhance Berberine Absorption: An In Vivo Pharmacokinetic Study

In the present study results related to the in vivo administration of Natural Deep Eutectic Solvents (NADES)-solubilized berberine are reported for the first time. NADES are mixtures of small natural compounds having a melting point significantly lower than that of any individual component. Such solvents have gained much attention of the scientific community in the green chemistry area, being considered useful alternatives to common organic solvents. NADES can be used also as administration vehicles, and this can be attractive for nutraceutical products when eutectics are formed with food grade ingredients. In this work, different NADES were prepared using mainly food grade constituents and were tested as solvents for the alkaloid berberine. Three selected NADES/berberine solutions and an aqueous suspension were orally administered to mice with in dose of 50 mg/Kg. Blood levels of berberine were measured by a LC-MS/MS method. The pharmacokinetic analysis revealed a 2–20 fold increase in blood concentration of NADES/berberine with significant changes in pharmacokinetic profile. Natural Deep Eutectic Solvents may thus be considered attractive solubilizing agents and may also play a role in the increase of absorption of poorly bioavailable natural products such as berberine.


Introduction
Berberine is a quaternary benzylisoquinoline alkaloid with a very ancient use in the traditional medicine of different countries, especially in China. This compound occurs in the root, rhizome and stem bark of many medicinally important plant species such as Berberis spp., Coptis spp., Hydrastis spp. [1].
Berberine has been used orally in China since ancient times to treat diarrhea [1,2]. This alkaloid also has a very long history of use in Ayurvedic and Chinese medicine as an antimicrobial, the complexation of a hydrogen acceptor and a hydrogen-bond donor. Such solvents are almost non-volatile under ambient conditions, are chemically and thermally stable, non-flammable, and have good solvent properties for several organic compounds [11][12][13][14]. In many cases, bioactive natural compounds like berberine possess limited solubility in water and ethanol mixtures. This can be a severe limitation in the extraction of bioactive compounds for food supplements and nutraceutical ingredients [15,17]. Any alternative ideal solvent should present a high level of safety and eco-sustainability as well as improved extraction performance [11,16,18,19]. NADES can also be considered as 'ingredients' in nutraceuticals or functional foods, and offer the possibility of combining various molecules, leading to the preparation of tailor-made solvents employable for pharmaceutical applications [19]. Being prepared with different ingredients, NADES can be used to solubilize or extract solutes with different properties (polarity, charge, etc.) [10][11][12][13][14][15][16].
During the course of our studies to evaluate NADES as absorption enhancers for poorly bioavailable natural products [20], we considered the possibility of using this approach to increase the bioavailability of berberine. For this purpose, several NADES were prepared and used to solubilize berberine. Three mixtures were selected based on their solubilizing properties towards berberine, their different chemical composition and possible issues related to the toxicity of NADES constituents. The selected mixtures comprised basic compounds (urea) or organic acids (malic and lactic acid), and proline. These NADES with berberine were orally administered to mice and compared to an aqueous berberine suspension. Blood concentrations were monitored by using an ad-hoc developed HPLC-MS/MS method. The comparison of the pharmacokinetics of the three different NADES was used to discuss the role of eutectic solvents as absorption enhancers.

Berberine Solubility in NADES
NADES composed of urea, amino acids, sugars and choline were used to solubilize berberine. The prepared eutectic mixtures are listed in Table 1.
Water and ethanol were used as reference solvents due to their safety and favorable use as solvents in the production for nutraceuticals and functional foods. Water was furthermore selected as a reference, due to the low berberine water solubility (2.1 mg/mL at 22 • C). The solubility of berberine in each solvent is reported in Table 1.
Considering the results of solubility measurements, we can observe that a large number of prepared NADES were not able to increase berberine solubility compared with water or ethanol. In fact, ten different eutectic mixtures showed solubility higher than 2 mg/mL, and many of the prepared solutions suffered instability (as indicated in Table 1) resulting in precipitation of the berberine. Our results showed that choline chloride-based NADES were not efficient in the solubilization of berberine, even though in the literature such NADES have been indicated as optimal mixtures for the extraction of isoflavones from soy [22] and solubilization of rutin [20,23].
The best solubility result (25 mg/mL) was obtained with eutectic 14, (proline-malic acid-lactic acid-water 1:0.2:0.3:0.5), followed by urea-based NADES (mixtures 1 and 2) with solubility higher than 12 mg/mL. The lactic acid-based eutectics presented solubility greater than 9 mg/mL. Due to these results and considering the possible toxicity issues deriving from their composition, the following screening was carried out. The LD 50 of urea in mice is 11,000 mg/kg, thus we selected mixture 2 instead of 1 because it contains less urea per gram. The malic acid LD 50 in rat is 1600 mg/kg, whereas those of lactic acid in rat and mice are 3543 and 4875 mg/kg, respectively. Thus, we selected mixture 5 instead of 6 despite the fact the solubility was 2.8 instead of 6 mg/mL.
Five calibration mixtures prepared mixing different ratios of berberine/ISTD (namely, 0.008654, 0.0108, 0.04326, 0.0649, and 0.108) were used to build a calibration curve (y = area of berberine/area of ISTD; x = quantity of berberine/quantity of ISTD) that was linear and reliable over the considered calibration range. The retention times of berberine (5.7 min) and ISTD (8.0 min) and specific MS/MS transitions allowed the identification of compounds. Limits of quantification (LOQ) and detection (LOD) for berberine were 0.3 ng/mL and 0.9 ng/mL, respectively. Exemplificative multiple reaction monitoring (MRM) chromatograms for berberine-spiked plasma (berberine 45 ng/mL, benzanilide (the internal standard or ISTD, 100 ng/mL) are reported in Figure 1. Selected transitions for quantification were 336 → 291 for berberine and 198 → 105 for the ISTD benzanilide.
Five calibration mixtures prepared mixing different ratios of berberine/ISTD (namely, 0.008654, 0.0108, 0.04326, 0.0649, and 0.108) were used to build a calibration curve (y = area of berberine/area of ISTD; x = quantity of berberine/quantity of ISTD) that was linear and reliable over the considered calibration range. The retention times of berberine (5.7 min) and ISTD (8.0 min) and specific MS/MS transitions allowed the identification of compounds. Limits of quantification (LOQ) and detection (LOD) for berberine were 0.3 ng/mL and 0.9 ng/mL, respectively.

Accuracy and Precision
Spiked samples were assayed for intra-day and inter-day precision and accuracy at concentrations of 0.9, 4.5, 9, 22.5 ng/mL of berberine. Data are summarized in Table 2.

Pharmacokinetics of Selected Eutectic Solubilised Berberin and Berberine Hydrochloride Water Suspensions in Mice
Eutectics formed with A, B and C were selected to evaluate their pharmacokinetic properties in mice and were compared with berberine hydrochloride water suspension. Plasma levels of berberine were determined up to 6 h after its oral administration (50 mg/kg) by gavage in mice. As

Accuracy and Precision
Spiked samples were assayed for intra-day and inter-day precision and accuracy at concentrations of 0.9, 4.5, 9, 22.5 ng/mL of berberine. Data are summarized in Table 2.

Pharmacokinetics of Selected Eutectic Solubilised Berberin and Berberine Hydrochloride Water Suspensions in Mice
Eutectics formed with A, B and C were selected to evaluate their pharmacokinetic properties in mice and were compared with berberine hydrochloride water suspension. Plasma levels of berberine were determined up to 6 h after its oral administration (50 mg/kg) by gavage in mice. As reported in Table 3 and Figure 2, different plasma levels were observed. For water and the three eutectic mixtures, the maximum peak was observed after 30 min.  The observed plasma levels of berberine administered with the three eutectic mixtures were significantly higher than that occurring in water suspension (p < 0.05) after 30 min. For B and C the values at 10, 60 and 180 min were also significantly different (p < 0.05) compared with the water suspension. The increase of the plasmatic peak at 30 min was 3.2, 7.3 and 8.0 times higher for NADES A, B and C, respectively. Table 4 summarize the non-compartmental pharmacokinetic parameters of berberine following oral administration of the same dose of berberine suspended in water or solubilized in the three considered NADES in Balb/c mice.  Table 3 and Figure 2, different plasma levels were observed. For water and the three eutectic mixtures, the maximum peak was observed after 30 min. The observed plasma levels of berberine administered with the three eutectic mixtures were significantly higher than that occurring in water suspension (p < 0.05) after 30 min. For B and C the values at 10, 60 and 180 min were also significantly different (p < 0.05) compared with the water suspension. The increase of the plasmatic peak at 30 min was 3.2, 7.3 and 8.0 times higher for NADES A, B and C, respectively. Table 4 summarize the non-compartmental pharmacokinetic parameters of berberine following oral administration of the same dose of berberine suspended in water or solubilized in the three considered NADES in Balb/c mice.

Discussion
Natural deep eutectic solvents can be attractive options in nutraceutical formulations and for pharmaceutical applications [15,20,24,25]. In this paper, we showed that the preparation of some NADES using mainly food grade materials, namely the eutectics A (proline-malic acid 2:1), B (proline-urea 2:1) and C (lactic acid-proline-malic acid-water) solubilized berberine better than reference solvents (water and ethanol). In particular, for the mixture C, we can explain the good performance in terms of solubility. The eutectic C contains two acids, namely malic and lactic. We selected this mixture due to the fact that malic acid was reported to form soluble berberine malate [21]. The eutectics 36-38 of the present work, containing lactic acid, were able to solubilize berberine at values of 10 mg/mL. Furthermore, the presence of a small amount of water strongly increased the solubility of berberine in the mixture C compared with the eutectics 7 and 8 (higher amounts of water) or 4-6 (without water). Our results on eutectic C are in agreement with previously published data indicating that the properties of NADES can be modulated by water due to changes in hydrogen-bonding interactions that are weakened by water dilution. The same authors also indicated that the solubility of compounds in NADES can be adjusted by changing the water content [14].
The three selected effective eutectics possess proline in their composition. On the other hand, several other tested mixtures (i.e., 3-13) containing proline were not efficient in solubilizing berberine. A further consideration is related to the presence of basic compounds such as urea, in efficiently solubilizing NADES. Also, acidic compound-based eutectics such as the lactic acid-based ones 36-38 showed good solubilization properties. The collected data show the opportunity to create specific eutectic mixtures in order to solubilize specific compounds. On the basis of the prepared eutectics we can postulate a tailor-made NADES for the berberine solubilization requires the presence of an organic acid (lactic or malic), amino acids and water. The presence of these components strongly influences the solubility properties. As a matter of fact, eutectic mixtures containing only proline and malic acid (4)(5) showed better solubilization properties compared with the corresponding ones containing water (7 and 8), but mixture 14, containing 1:0.2:0.3:0.5 proline-malic acid-lactic acid-water, showed improved properties. Thus, small amounts of water are needed to increase the berberine solubilization in organic-acid-proline-based eutectics. The overall results showed a moderate ability (5-to 10-fold) of NADES to solubilize berberine with urea, lactic acid and water-containing NADES.
Considering the pharmacokinetic analysis, the administration by gavage to mice resulted in different behavior comparing water suspension and NADES. An increase of bioavailability of berberine, especially for the eutectic C, was observed as indicated in Table 2, being the AUC four times the one observed with water suspension. The administration of NADES with dissolved berberine yielded in a significant increase of the plasma alkaloid peak at 30 min. All the prepared NADES showed similar pharmacokinetic profiles in the explored range of times, with no differences in the observed t max between the urea-based and the acidic compound-based eutectics, suggesting that the increase of bioavailability observed is mainly related to the solubilization properties of the different eutectic mixtures.

Chemicals
Sugars and polyols were obtained from Carlo Erba (Milan, Italy). Citric and oxalic acids were purchased from Riedel-De-Haen AG (Seelze, Germany), tartaric acid from Codex (Turin, Italy), malic acid from Carlo Erba, proline from Fagron (Bologna, Italy). Choline chloride was purchased from Sigma-Aldrich (Milan, Italy) and urea from Alfa Aesar (Karlsruhe, Germany). Glutamic acid, acetylcarnitine and carnitine were purchased from ACEF s.p.a. (Fiorenzuola D'Arda, Italy). Berberine and the internal standard benzanilide were purchased from Sigma-Aldrich. Solvents such as HPLC grade acetonitrile and methanol were purchased from Scharlau (Barcelona, Spain), and formic acid from Carlo Erba.

Solubility Trials and Quantification of Solubilized Berberine in the NADES by HPLC-DAD
A precisely weighed amount of berberine was suspended in water, ethanol and in the various prepared NADES with an initial concentration of 2.5 mg/mL, then increased according to the solubility capacity of the different solvents (25 mg/mL was the highest concentration reached). Samples were placed on a magnetic stirrer (Stuart, Bibby Scientific Ltd., Stone, Staffordshire, UK) at 22 • C for 1 h, then centrifuged for 21 min at 13,000 rpm with a 5415 R centrifuge (Eppendorf, Hamburg, Germany). For quantitative measurement of solubilized berberine, a portion (100 µL) of the clear supernatant obtained after centrifugation was diluted 1:5 (v/v) in ethanol.
For quantification, stock standard solution of berberine (100 µg/mL) was prepared dissolving the analytical standard in methanol using an ultrasonic bath. A calibration curve was obtained injecting standard solutions of berberine at different concentrations, namely 50, 25, 10, 5 and 1 µg/mL. The following calibration curve was obtained: y = 18,323-50,294 (R 2 = 0.9997). The Limit of Quantification (LOQ) was 1 µg/mL.
For HPLC-DAD analysis, an Agilent series 1260 HPLC instrument (Agilent, Cernusco Sul Naviglio, MI, Italy) equipped with a quaternary pump, a diode-array detector, an auto sampler and a column oven compartment was used. Analyses were performed on Eclipse XDB C 8 column

Animals Blood Collection and Extraction
All experimental protocols involving animals were reviewed and approved by the Ethical Committee for animal Experiments of the University of Padua (CEASA; 49,571). Female, Balb/c mice (8-10 weeks old) were housed (three per cage) in polycarbonate cages and kept on a 12 h light/dark cycle. Food and water were given ad libitum. Mice, randomly divided into groups of 15 animals, one group (n = 15) for each NADES and berberine suspended in water. Each group received 50 mg/kg of berberine by oral gavage as water suspension (group 1) or solubilized in the selected NADES (group 2). A single blood sample was collected by cardiac puncture from each animal at 10, 30, 60, 180, and 360 min after dosing (three animals were used for each time point). The whole blood was heparinized, then three samples were obtained for each time point and each treatment group.

HPLC-MS Plasma Analysis
Standard stock solutions for determination of berberine in mice blood were prepared by dissolving berberine and the ISTD benzanilide in methanol. The calibration curve was obtained mixing 500 µL of 0.8 µg/mL ISTD with different volumes (500, 300, 100, 50, and 25) of 90 ng/mL berberine standard solution in order to obtain different berberine/benzanilide quantity ratios. Mixture of ISTD and berberine were added to blank plasma samples and used for sample and calibration curve preparation. 500 µL of IS solution were added to 400 µL of whole blood (461 ng) in order to precipitate proteins. Three mL of water were added, followed by 7 mL of ethyl acetate. Samples are mixed in a vortexer and subjected to 5 min ultrasound in a bath at room temperature. The sample was then centrifuged and the clear supernatant was concentrated under vacuum at 50 • C. 200 µL of methanol were then used to dissolve the residue and used for HPLC-MS/MS measurements.
For analytical measurements, an Agilent series 1260 HPLC chromatograph equipped with a Varian Prostar 410 autosampler and coupled with a Varian 320 TQD MS spectrometer was used. The mass spectrometer was equipped with electrospray ionization (ESI) source as the interface and analysis was conducted in negative ion mode for both the analytes. Analyses were performed on a Polaris 3 C18-A 150 The ESI source was set in positive ionization mode. Quantification was performed using multiple reaction monitoring (MRM) with m/z 336 > 291 transition for berberine and m/z 198 > 105 transition for the ISTD. The MS parameters were capillary voltage 60 V, needle voltage 4200 V, shield voltage 600 V, collision energy 22 V, Q1 voltage 0.7 V and Q3 voltage 2.8 V, nebulizing gas pressure 50 psi and drying gas pressure 25 psi. Calibration curve using the ratio peak area berberine/peak area ISTD versus quantity berberine/quantity ISTD was y = 0.9701x + 0.0028, R 2 = 0.9899. The limit of detection (LOD) was 0.3 ng/mL and the limit of quantification was 0.9 ng/mL.

Method Validation
Assay specificity was evaluated comparing the chromatograms of standard-spiked plasma with blank plasma from three different sources. Calibration curves were fitted by least square regression analysis to plot peak area ratio of berberine/ISTD relatively to the ratio of the amount of berberine/ISTD. Limit of Quantification (LOQ) was calculated as the lowest amount with a relative standard deviation <20%. Intra-and inter-day stability, extraction recovery, were measured. Precision and accuracy were evaluated using QC samples (n = 5) at concentrations of 0.9, 9, 22.5 and 45 ng/mL on two different days. Different plasma samples were used for intra-and inter-day stability, extraction recovery, with five replicates.

Pharmacokinetic Analysis
Non-compartmental pharmacokinetic analysis was performed using WinNonlin Version 2.1 (Pharsight Corporation, Mountain View, CA, USA) software. The area under the mean plasma concentration-time curve extrapolated to infinity (AUC) was calculated using a linear-up and log-down method. Maximum concentration and the time when it was observed (C max and t max , respectively) were reported as observed. Terminal half-life (t 1/2 ) was calculated as t 1/2 = ln2/λ z , where λ z is the slope of the terminal phase of the plasma concentration-time curve in the semi-log plot calculated by linear regression. Mean residence time (MRT) was calculated as AUC/AUMC, where AUMC is the area under the first moment curve. Relative bioavailability F r of NADES versus suspension was estimated as a ratio of AUC following the administration of NADES and AUC following the administration of the suspension.