Phenolic and flavonoid content and in vitro inhibitory effect of some Amazonian fruit juices on CYP3A4 activity

Many foods interact with drugs and may cause changes in the pharmacological effects of the co-administered therapeutic agent. The enzyme CYP3A4, which belongs to the cytochrome P450 enzyme complex, is responsible for the metabolism of most drugs currently on the market and is involved in many drug interactions. Hence, the interaction of this enzyme with juices of some fruits, such as grapefruit, can affect the pharmacokinetics of various drugs. However, native fruits from the Amazon region have not yet been the target of this type of research. We determined total polyphenols and flavonoids of the Amazonian fruits açaí ( Euterpe precatoria ), buriti ( Mauritia flexuosa ), camu-camu ( Myrciaria dubia ), cubiu ( Solanum sessiliflorum ), cupuaçu ( Theobroma grandiflorum ), jenipapo ( Genipa americana ), and taperebá ( Spondias mombin ) and evaluated the effects of each fruit juice on CYP3A4 activity, using the star fruit ( Averrhoa carambola) juice as positive control. Açaí juice presented the highest content of total polyphenols and flavonoids (102.6 ± 7.2 µg gallic acid equivalent (GAE) per mL and 7.2 ± 0.6 µg quercetin equivalent (QE) per mL, respectively). All juices were able to inhibit the activity of CYP3A4. There was no residual activity of the drug-metabolizing enzyme for açai, buriti, cubiu, camu-camu, and taperebá juice, while for cupuaçu, jenipapo and the positive control, the residual activity was 44.3, 54.3 and 20.2%, respectively. Additional studies should identify the phytocompound(s) responsible for this inhibition activity, to clarify the mechanisms involved in this phenomenon.


INTRODUCTION
The Brazilian Amazon has a wide diversity of native fruit species.Many of these fruits are becoming increasingly popular in Brazil and in other countries.Edible fruits marketed in traditional fairs throughout the Amazon region remain to be explored as sources of nutrients for food and pharmaceutical applications, but their chemical profile and effects on health is still poorly known and deserving investigation (Araújo et al. 2021).
Amazonian fruits are usually consumed fresh, frequently in the form of juice (Rabelo 2012).Besides being highly appreciated for their taste, juices facilitate easier access to several bioactive compounds with beneficial effects to human health (Anderle et al. 2004;Chen et al. 2018).Fruits and their juice can also alter the disposition of drugs, mainly by acting on cytochrome P450 enzymes (CYP), especially CYP3A4 (Korhonen et al. 2007;Iwata et al. 2015;Ohkubo et al. 2017).
Several studies have shown that grapefruit (Citrus paradisi Macfad.)juice has an inhibitory effect on CYP3A4 leading to an increased bioavailability of drugs metabolized by this enzyme and, consequently, greater systemic exposure to orally administered drugs (Marco et al. 2002;Yin et al. 2010).The mechanism of this interaction involves irreversible inhibition of CYP3A4 in the small intestine (Girennavar et al. 2006;Girennavar et al. 2007).Other fruits that have been described to inhibit this enzyme are the star fruit (Averrhoa carambola L.) (Hidaka et al. 2004;Zhang et al. 2007;Hosoi et al. 2008), the pomegranate (Punica granatum L.) (Hidaka et al. 2005), and the wild grape (Vitis vinifera L.) (Kim et al. 2006).
According to some authors, the polyphenolic compounds found in fruits are responsible for inhibiting CYP3A4 (Girennavar et al. 2006;Girennavar et al. 2007;Hosoi et al. 2008).Thus, the presence and interaction of these compounds in fruit juices on CYP3A4 must be evaluated properly.The in vitro interaction of dietary polyphenols and CYP3A4 activity is variable due to structural and physicochemical characteristics such as the number of hydroxyl groups, stereostructure, molecular weight and lipophilicity (Basheer and Kerem 2015).The structure-activity relationship of flavonoids towards CYP3A4 has been recently established by computational models, in order to predict some possible drug interactions (Li et al. 2018).
One of the interactions described in the literature is for curmumin, a polyphenol extracted from turmeric (Curcuma longa L.), which can induce pharmacokinetic changes in several drugs by inhibiting cytochrome P450 isoenzymes (Bahramsoltani et al. 2017).Green tea (Camelia sinensis L.) polyphenols may also interact with CYP2C8, CYP2B6, CYP3A4, CYP2D6, and CYP2C19 enzymes, promoting overdose or underdosing of various drugs (Teschke and Xuan 2019).
Despite the wide diversity and amount of native fruit juices consumed by the local population in the Amazon region, to our knowledge, currently there is no study available in the scientific literature which evaluated drug-food interaction by Amazonian fruit juices.Therefore, the aim of this study was to determine the content of phenolic compounds and flavonoids in seven popular Amazonian fruit juices, as well as their in vitro inhibitory effect on CYP3A4 activity.

Materials
Chromatographic grade methanol and acetonitrile were acquired from Panreac (Barcelona, Spain).Diethylamine and hydrochloric acid were purchased from Synth (Diadema, Brazil).The components of the NADPH generation system, nicotinamide adenine dinucleotide phosphate, glucose-6-phosphate and glucose-6-phosphate dehydrogenase enzyme, as well as midazolam (MI) and diazepam (IS), were acquired from Sigma-Aldrich (St. Louis, USA).The 1-hydroxymidazolam (1-HMI) was obtained from the Cayman Chemical Company (Ann Arbor, USA).Gallic acid, Folin-Ciocalteu and quercetin reagents were purchased from Sigma-Aldrich (St. Louis, USA).Human liver microsomes (HLMs) were acquired from Corning Life Science (Arizona, USA) and stored at -80 °C until use.

Fruit samples
The analyzed fruits were chosen among the widely consumed fruit juices in the Amazon region which were harvested and marketed at the time of execution of this study (Rabelo 2012) The star fruit (Oxalidaceae), which is not an Amazonian fruit, was chosen as a standard positive control, as the juice of this fruit has already been reported to inhibit CYP3A4 (Hidaka et al. 2004;Zhang et al. 2007).

Preparation of the juices
The fruits were sanitized in 2% sodium hypochlorite solution (Yuyama et al. 2011), and stored at -20 °C until use.Each sample was subjected to a turbolysis process (2 min), using ultra-pure water as solvent (40% m v -1 ) and the obtained juices were centrifuged at 4000 rpm for 5 min (Girennavar et al. 2007).The supernatant was collected, filtered, and lyophilized (Hidaka et al. 2004).For analysis, the lyophilized ACTA AMAZONICA juices were resuspended in water and centrifuged at 3000 rpm for 10 min (Girennavar et al. 2007).The fruit part used in each case is described in Table 1.

Determination of total polyphenols and flavonoids
The assessment of the phenolic and flavonoid content of edible fruits is important since these compounds are generally linked to biological activities (Becker et al. 2018).Total polyphenol content was determined using Folin-Ciocalteu reagent, as described by Bonoli et al. (2004).For quantification, a standard curve with gallic acid solution was used at the range 7.8125-500 µg mL -1 .The total polyphenol content was assayed in quadruplicates and expressed as µg gallic acid equivalent (GAE) mL -1 .
Total flavonoids were determined according to Chang et al. (2002).The quantification was performed using a standard curve with quercetin solution in the range 7.8125-500 µg mL - 1 .The total flavonoid content was assayed in quadruplicates and expressed as µg quercetin equivalent (QE) mL -1 .
A Beckman Coulter microplate reader, model DTX800 (Brea, USA) operating at the wavelengths of 620 nm and 405 nm was used for the determination of polyphenols and flavonoids, respectively.Values were expressed as the mean ± standard deviation.

Analysis of CYP3A4 activity marker
The CYP3A4 activity marker (1-HMI) was analyzed with an HPLC equipment coupled to a diode-array detector (DAD) acquired from Shimadzu (Kyoto, Japan).The analysis was carried out by using a Gemini C18 analytical column (250 mm x 4.6 mm, 5 µm particle size, Phenomenex, Torrance, USA) and mobile phase composed of acetonitrile: ultra-pure water with 0.1% diethylamine (45:55, v/v).The flow rate was set at 1.2 mL min -1 and the column temperature was kept at 30 °C.The injection volume was 50 µL.Analytes were monitored at 260 nm (for 1-HMI) and 239 nm (for diazepam -IS).The data were analyzed using the LC Solution software, version 1.25 SP1 (Shimadzu).

Enzymatic kinetics of midazolam
The reaction time, protein concentration, and concentration range of MI were determined in advance of sample analysis to ensure that all reactions take place under conditions of V0, which is the initial rate of the enzymatic reaction.This rate corresponds to a linear decrease of substrate concentration in relation to protein concentration and incubation time.

Fruits species
Common ) and total flavonoids equivalent to quercetin (µg QE mL -1 ) in the juice of seven Amazonian fruits (and star fruit as positive control) and their effect on midazolam 1-hydroxylation in human liver microsomes.
Values are the mean ± standard deviation of three replicates; n.d.= not detected; *: not analyzed.
After determining the enzyme kinetics, the K M value -Michaelis-Menten constant (substrate concentration in which V0 is equal to half of V MAX ) was calculated.For the inhibition studies of the fruit juices, midazolam concentration (5 µM) was used, with a value close to its K M value.

Inhibitory effect of fruit juices on CYP3A4 activity
The inhibition of CYP3A4 by the fruit juices was evaluated by the formation of 1-HMI, according to Fujita et al. (2003) and Hidaka et al. (2004).The assay was performed in microtubes using a 5 µM MI solution (5 µL), 5 mg mL -1 lyophilized juice resuspended in PBS solution 0.1 M pH 7.4 (125 µL), NADPH generator system (50 µL) containing 5 mmol L -1 glucose-6-phosphate, 0.25 mmol L -1 nicotinamide adenine dinucleotide phosphate, and 0.5 unit mL -1 glucose-6-phosphate dehydrogenase enzyme.A volume of HLMs of 20 µL was used to a final volume of 200 µL.The samples were pre-incubated in a water bath with stirring for 5 min at 37 °C.The metabolism was initiated by the addition of HLMs, with a reaction time of 10 min at 37 °C.Then, chilled acetonitrile (200 µL), was added to stop the reaction followed by the IS addiction (30 µL of a solution containing 1.4 µg mL -1 ).The mixture was then centrifuged at 3500 rpm for 10 min.The supernatant was collected (200 µL) and samples were analyzed using HPLC-DAD.Under these specific conditions, the term residual activity was regarded as the enzyme activity that remains after exposure of the enzyme to the potential inhibitors.

RESULTS
The determination coefficient obtained for the quantification of total polyphenols was 0.993.Total polyphenol levels varied between 19.57 and 102.56 µg GAE mL -1 .The determination coefficient for the quantification of total flavonoids was 0.999.The levels of total flavonoids were determined only for açaí and cupuaçu (7.16 and 1.64 µg QE mL -1 , respectively) as the results obtained for the other juices were below the limit of quantification of the method used (Table 1).Açaí presented the highest polyphenol and flavonoid concentrations (Table 1).
The total analysis time for the chromatographic separation of 1-HMI, MI and the IS was less than 15 min (Figure 2).The retention time of 1-HMI, MI and IS was 6.2, 9.6 and 11.8 ACTA AMAZONICA min, respectively.The method showed to be selective, since no interferents from the microsomal medium were observed in the retention time of the analytes (data not shown).
The enzyme kinetics of the MI metabolism by HLMs conformed to a Michaelis-Menten kinetic profile (Figure 3).The enzyme parameters were V MAX = 0.59 ± 0.01 nmol mg protein min -1 and K M = 5.0 ± 0.5 µmol L -1 .The control activity of midazolam hydroxylation at 5 µM by HLM determined in the absence of juice was 0.11 ± 0.02 nmol mg protein min -1 .rodent chow for modulation of neurochemical parameters in rat brain.The polyphenol content varied between the two açaí species, with most glycosidic anthocyanins, flavonoids and phenolic acids about two to 50 times higher in E. precatoria, except delphinidin, quercetin and vanillic acids, which were considerably higher in E. oleraceae, resulting in better behavioral outcomes for the diet supplemented with the latter species (Poulose et al. 2017).
Camu-camu, which had the second highest content of polyphenols, is known to have a high concentration of phenolic compounds (Neves et al. 2012;Tauchen et al. 2016).Likewise, buriti, which had the third highest content, is also considered as rich in antioxidant compounds, such as carotenoids, ascorbic acid and phenolic compounds (Melo et al. 2008;Tauchen et al. 2016).The levels of total polyphenols obtained for cubiu and jenipapo were close to those reported by Tauchen et al. (2016), who also reported higher values for these compounds in cupuaçu from Peru.
All evaluated fruit juices interacted with CYP3A4, however only açaí, camu-camu, buriti, cubiu and taperebá juices inhibited 100% of the enzymatic activity (remaining activity 0%).The residual activity of CYP3A4 in the presence of jenipapo and cupuaçu juices was higher than that of the positive control.Despite the low concentration of polyphenols in comparison to other fruits, cubiu and taperebá juices also resulted in the absence of residual CYP3A4 activity.Unlike our result, Hidaka et al. (2004) determined a residual activity of 0.1% for star fruit juice.The procyanidin B1 and B2 stereoisomers and/or catechin/epicatechin found in star fruit are suggested as responsible for the inhibition of CYP3A4 (Hosoi et al. 2008).The discrepancy with the average value of 20.2% obtained in our study is likely owed to variation in the level of phytochemicals, which depends on ripeness, plant part used, geographical origin and environmental conditions of the samples, collection period, harvest conditions, cultivation practices and storage process (Kårlund et al. 2014).Although the levels of phenolic and flavonoid compounds were not determined for our star fruit juice samples, Zhang et al. (2007) observed that large variations in the contents of these chemicals did not result in substantial differences of their inhibitory effect, suggesting the participation of other substances in inhibition mechanisms.
Several limitations might have influenced our results.The fruits were acquired at local fairs, so that we had no precise information regarding geographical origin, ripening and harvesting of the fruits.Furthermore, only one concentration of each juice was analyzed.Further studies should determine the half maximum inhibitory concentration (IC 50 ), i.e., the concentration necessary to decrease the enzyme activity in 50%.Taken together, these facts limit the generalizations that can be drawn from this study, which was devised as a first-step All juices were able to inhibit the activity of the CYP3A4.No residual activity of the drug-metabolizing enzyme was observed for açai, buriti, cubiu, camu-camu, and taperebá, which was below the value observed for star fruit (Table 1).Only cupuaçu and jenipapo showed residual activity above that of star fruit.

DISCUSSION
The highest contents of both phenolic and flavonoid were found in açaí samples.Several studies have reported the antioxidant potential of açaí due to the presence of phenolic compounds, especially flavonoid anthocyanins, both present in both E. precatoria and E. oleracea Mart (Yuyama et al. 2011;Yamaguchi et al. 2015;De Moura and Resende 2016;Martins et al. 2018).
Phenolic and flavonoid contents in fruit juice are generally determined for quality control purposes and not to assess inhibitory effects on CYP450 (Zhang et al. 2007).Some authors evaluated the impact of phenolic and flavonoid contents on other biological activities.For example, freezedried E. oleraceae and E. precatoria fruits was added to NIH-31 ACTA AMAZONICA screening step to determine possible drug-food interactions sensu Burkina et al. (2016).
The interpretation of our results also deserves caution, as in vitro drug interaction studies with plant products, though being relatively easy to perform, inexpensive compared to studies in animals or humans, and allowing controlled test conditions, are not easily reproducible.This is due to insufficient data regarding the absorption process of unknown botanical components, uncertainty about relevant concentrations for the assay, and difficulties in standardizing samples with complex phytochemical profiles, among other aspects.However, despite these limitations, in vitro studies are a viable screening alternative to identify compounds that may represent a risk of drug interactions and, serve as indicators for conducting in vivo studies (Markowitz and Zhu 2012).In this sense, our study provided guiding data on the potential for in vivo drug interactions of several popular Amazonian fruit juices, indicating those requiring further clinical studies (Showande et al. 2019).

CONCLUSIONS
We report the first data on the inhibitory effect of seven popular Amazonian fruit juices on the CYP3A4 enzyme activity.Our results indicate that all the juices may interact with drugs metabolized by CYP3A4, which may cause further changes in the bioavailability of the associated drug.However, additional studies are needed to investigate the clinical relevance of the interaction of these juices with CYP3A4 in vivo and to identify the phytocompound(s) responsible for the observed inhibition effects and to clarify the mechanisms involved.

Figure 2 .
Figure 2. Representative chromatogram of 1-hydroxyimidazolam (1-HMI), midazolam (MI) and diazepam (IS).A -260 nm; B -239 nm.Chromatographic conditions are described in Materials and Methods.This figure is in color in the electronic version.

Table 1 .
Content of total polyphenols equivalent to gallic acid (µg GAE mL -1