Abstract
Natural products, such as herbal medicinal products, food supplements, and functional food, are widely used to support well-being and for promoting health. In general, the effects of using these products are desired and beneficial, but unexpected adverse effects might also occur, especially when natural products are used with medication. Consumers do not often even know that they are exposed to bioactive compounds that might interact with the body and have effects on their well-being. One of the objectives of the World Health Organization is to promote herb–drug interaction monitoring, and it is important to know how marketed preparations and compounds from common foods interact when they are absorbed. In this review, we describe how a Caco-2 cell absorption model has been used to study how natural products, such as flaxseed, rapeseed, purple loosestrife, pine, echinacea, certain berries and herbs, anthranoid laxatives, and traditional medicinal plants, affect the absorption of co-administered drugs. We discuss the types of interactions and adverse effects that might occur and their possible reasons. Overall, we conclude that the Caco-2 cell absorption model is a useful tool for studying the absorption of natural products with drugs; and that to enable the safe use of natural products with medicines, concomitant use should be studied.
Keywords
Introduction
In the past, when no pharmaceutical industry existed, the only cure for illness was found from nature. 1 The use of certain herbs or mixtures of compounds was transferred from generation to generation.1,2 This knowledge is still valuable, although most of the drugs used in modern societies are now industrial products. 1 In developing countries, natural products are still used as the prevalent medications.1,3,4 In general, these compounds are secondary metabolites that are produced by organisms as a response to some external stimuli such as infection, wounding, drought, or nutritional changes.5,6 One herb contains numerous compounds, and very often it is a particular combination of compounds that is effective instead of a single compound. 7
The interest toward preventive healthcare is popular and vigorous.8-10 People use vitamins, herbal supplements, and functional foods to maintain their health and support their well-being, and they do not realize that sometimes these products may also cause problems in the body. Even common foods may cause serious undesired effects, especially if certain food ingredients are consumed solely and/or excessively. In the case of medications, the risk for adverse effects accumulates. 11 Therefore, it is important to know how commercial herbal medicinal products and compounds from common foods are absorbed and interact in the human body.
In real life, it is not always easy to notice interactions. Effects on one's well-being might take time to appear, so linking them with the actual source can be challenging. There might also be other life changes that could be confused as being responsible for one's changed condition. To overcome these difficulties, potential interactions should be studied beforehand. The human colon adenocarcinoma cell line (Caco-2) cell model is a human-derived in vitro model for studying gastrointestinal (GI) absorption. 12 This model is very suitable for primary screening of absorption and drug interactions, 13 and its use can decrease the number of animals needed for experimental studies. Multiple studies can be conducted with small amounts of a sample, and the cells are easy to handle. Natural product extracts are not the easiest matrix to handle due to the large number of interfering compounds such as chlorophyl, sugars, and color. 14 In addition, the heterogeneity of natural product samples adds difficulties to sample preparation and analysis. 15 Nevertheless, studies with Caco-2 cells have shown that this cell model is suitable for assaying natural products as well.16-18 The absorption of a single natural compound or the effect of an extract on the absorption of other compounds can be studied.16-22
One of the objectives of the World Health Organization (WHO) is to monitor natural product–drug interactions and that way promote human health. 23 In this review, we focus on studies on natural product–drug interactions using a Caco-2 cell absorption model. We discuss the use of select natural products and food supplements and the reasons that might cause their possible interactions with the drugs studied.
Natural Extracts, Food Supplements, and Interference with Drugs
Natural products are used widely to manage overall physical and mental health. People pay more and more attention toward well-being and issues related to pure food and health.8-10,24 Often these herbs are self-administered along with therapeutic drugs.9,25 Patients might not regard these products as medicines, so they usually do not tell their physician about the use of such products. According to 1 study, as much as 95% of herbal supplement users and 75% of nonusers regarded these products safe. 26 However, numerous studies in vitro and in vivo have shown that herbal products and dietary supplements interact with many drugs.8,9,25,27 The interactions often involve a change in pH or effects on transporter systems and drug-metabolizing enzymes, especially cytochrome P450 enzyme family enzyme 3A4 (CYP3A4) which is the major drug-metabolizing enzyme.28-31 These things affect also drug absorption and can either increase or decrease the potency of therapeutic drugs. In both cases, serious side effects may occur due to changed bioavailability of drugs.
The interactions between herbal medicines and drugs are emphasized in older people. They usually have prescribed medicines and in addition, use various herbal products. 32 Natural products are used to support, e.g., general health, heart function, and mental disorders or to relieve menopause symptoms and infections.32-34 Frequently, these people use more than one natural product and other nutritional supplements. 32 The people likely to use natural therapies often also consume a great number of prescription drugs. 35 Approximately one-third of herbal-medicine users have a risk of drug interactions.36-38 Commonly used natural products ginkgo (Ginkgo biloba L.), St. John's wort (Hypericum perforatum L.), garlic (Allium sativum L.), and ginseng (Panax ginseng C.A.Mey.) all cause different kinds of interactions if administered contemporaneously with medication.25,39 The interactions could increase or reduce drug concentrations, alter blood pressure, increase the risk of bleeding, and alter the body's glucose and electrolyte homeostasis.38,40 In addition, the changes in the function of metabolic enzymes or elongation of residence time in the body may lead to the formation of toxic metabolites.40,41 For example, St. John's Wort, which is used for the relief of depression and anxiety, induces both the metabolizing enzyme CYP3A4 as well as the P-glycoprotein (P-gp) efflux protein.27,42-44 When these act synergistically, serious adverse effects are likely to occur with drugs that are also substrates of these. Such compounds include, for example, imatinib, digoxin, and cyclosporine.27,42,43
A tricky interaction inducer is grapefruit juice. It inhibits the activity of the P-gp efflux transporter as well as the CYP3A4 enzyme.45-47 In the small intestine, CYP3A4 and P-gp act synergistically and serious side effects might occur. This is dangerous since people may not realize that they are exposed to a substance that may cause interactions. The main component responsible for the interaction is shown to be a coumarin compound, 6′,7′-dihydroxybergamottin, but the effects of naringenin, a flavonoid compound, should not be entirely excluded either.29,46 Very common especially among older people is the concomitant usage of Ginkgo biloba and aspirin.26,48 Ginkgo is used to treat peripheral vascular disorders and to enhance memory function, and its usage with aspirin may increase bleeding time significantly.
The exact composition of natural products is often uncharacterized. 8 These products consist usually of complex mixtures of bioactive substances that may or may not possess therapeutic activity. In some cases, it is uncertain which compound causes the pharmacological effect.27,39 The chemical composition of products may also vary depending on seasonal alterations. Several factors, including soil and climate, ripeness at the time of harvest, extraction, processing, and storage, affect the polyphenol content of plants. 49 Different batches of the same product can thus have different effects, highlighting the importance of chemically characterized samples. Moreover, comparing different products containing the same herb is often impossible due to the poor characterization of the products. These factors complicate the evaluation of the effects of certain natural products in combination with medication.
Food can be regarded as functional if it is satisfactorily demonstrated that one of the components, or a combination of these, affects targeted functions in the body in a specific, positive way, producing a physiological effect beyond the food's nutritional value.50,51 The food industry produces not only added-value products with minerals and vitamins, but also functional foods with various ingredients 31 which could be fractions of herbal extracts, berries, or other plant materials proven to have a beneficial effect on health. 52 However, such enrichment could result in increased adverse effects. Some drugs appear to be safe when taken with regular food, but less so if taken with fortified foods.31,53
When developing new natural-derived products, the components should be characterized well. 54 About 12% of plant extracts currently available on Western markets have not been extensively studied for their properties, and they are likely to cause adverse effects. 55 The effects of single compounds as well as whole products and their possible effects on metabolizing enzymes should be studied. The effects on the absorption of different kinds of medicines could be used to evaluate their interactions in the human body. 17 The more concentrated a product is, the more important its evaluation is. The clinical importance of herb–drug interactions in humans is likely highly variable depending on interindividual differences in health status, age, food habits, genetic makeup, and metabolizing capacity.9,25,56 Exposure duration is also critical in possible interactions. However, in vitro studies can provide crucial information on what kind of interactions are likely to occur.
Absorption in the Body
When a molecule enters the body, it faces various circumstances. The prevailing situation depends on other compounds that have entered the body at about the same time.57,58 The type of food consumed simultaneously may have a significant effect on the absorption of a compound. 59 Absorption may be reduced, delayed, or accelerated. In order to have an effect on the body, a molecule first has to dissolve into the body's fluids. It also has to be able to pass through biological membranes. Either dissolution or absorption may be the rate-limiting factor for the effects of the molecule. 60 The contents of the GI tract have an influence on the pH value in the intestinal lumen, which greatly affects the dissolution process.57,61 An acidic molecule is in ionic form when pH is over the molecule's pKa value, and the opposite is true for basic substances. 31 pH is lower in a fed state compared to a fasted state, and for some molecules, the state may have significant effects on the bioavailability of a compound.
Absorption from the gut lumen can occur by passive or active mechanisms (Figure 1). A molecule can passively diffuse through epithelial cells (transcellularly) or pass through intracellular spaces (paracellularly). The transcellular pathway is the most popular pathway since the cellular absorbing surface is significantly greater than in paracellular space, although the importance of the paracellular way is also important, especially for slowly transported hydrophilic compounds.62,63 A paracellularly permeating molecule has to be polar or highly soluble and relatively small since the route is restricted by intracellular tight junctions. A transcellular permeating molecule has to be able to permeate biological barriers such as intestinal mucosa and epithelia. 64
Drug transportation pathways across the intestinal epithelium. (A) Passive transcellular, (B) passive paracellular, and (C) active carrier-mediated route, and (D) an efflux transporter diminishing absorption.
A molecule has to be lipophilic enough to pass through the epithelium but not too lipophilic to avoid retention in the cells. Absorption may also be enhanced by carrier molecules such as peptide transporters. 65 On the other hand, efflux transporters, such as multidrug resistance protein 1 (MDR1) and breast cancer resistance protein (BCRP), diminish the fraction absorbed by pumping their substrate molecules out back into the lumen.9,66,67 Efflux transporters may significantly affect drug absorption through intestinal mucosa after oral administration. They can also indirectly enhance intestinal CYP3A4 metabolism by increasing the intestinal residence time of a drug and by preventing CYP3A4 product inhibition by the removal of primary metabolites.68-70 Inhibition or induction of efflux transporters by co-administered drugs, food, or herbal products may result in pharmacokinetic interactions leading to unexpected toxicities or under treatment.
Caco-2 Cell Model in Absorption Studies
The Caco-2 cell model is a useful tool in predicting drug absorption and in determining the roles of various physical and biochemical barriers to drug absorption.71-73 This model is regulatory-accepted and widely utilized for the prediction of absorption and permeability of compounds in humans.12,74-76 The use of cell lines in the study of compounds’ absorption decreases the number of animals being exposed to compounds unnecessarily, and the same time, it enables screening with live material. 77 Caco-2 cells, originated from human colonic adenocarcinoma, exhibit morphological and functional similarities to intestinal enterocytes. When cultivated on a permeable membrane, the cells form tight junctions and a brush-border membrane with microvilli on the apical side as enterocytes do on their luminal side. 78 But since Caco-2 cells originate from the colon, the tight junctions are tighter than in the small intestine and the mucus layer normally covering enterocytes is absent.79,80 Transporter expression may also vary between the small intestine and different Caco-2 clones.81,82 This is important to consider when evaluating compounds that are transported actively across the membrane. 83 There are uptake transporters, e.g., the monocarboxylic acid transporter (MCT1) transporter, which enhances the absorption of their substrates, and efflux transporters, e.g., the MDR1 transporter, which transports their substrates to the lumen. These transporters have crucial roles in the pharmacokinetic profiles of their substrates. 84
A study with 2 different Caco-2 cell clones indicated that profiles of transporter expression correlated significantly between the 2 clones and the human small intestine. 85 Multidrug resistance-associated proteins multidrug resistance-associated protein 2 (MRP2), MRP3, MRP4, MRP5, MRP6, organic anion transporter polypeptide A (OATP-A), OATP-B, organic cation transporter 1 (OCT1), and MCT1 transporters were nevertheless quantitatively expressed at higher, and BCRP at lower, levels in Caco-2 cells compared to the small intestine. The relative expression of MDR1, MRP2, OATP-A, and PEPT1 mRNA was found to differ in the 2 analyzed Caco-2 cell clones. Another study indicated that the expression levels of MDR1, MRP2, MPR3, BCRP, OCTN2, PepT1, OATP8, OCT1, OCT2, OAT1, OAT2, OAT3, and MCT1 were lower in Caco-2 cells than in the human jejunum. 84 The relative order of expression levels correlated between the jejunum and Caco-2 cells. In another study, no correlation was found in transporters between the duodenum and Caco-2 cells. 86 In addition, levels of the metabolizing CYP3A4 enzyme are generally low in Caco-2 cells, which may lead to overestimating the absorption of its substrates.60,87 Thus, if needed, enzyme expression may be induced. 88 The transporter expression levels vary also between different parts of the intestine and between different culturing conditions of Caco-2 cells. 89 Besides that, transporter function may also vary due to prevalent drug concentrations, pH, and nutritional constituents. 90 These findings underline the interest in carefully characterizing transporter levels in Caco-2 cells. We cannot mimic the exact situations in the lumen, but we can try to get as close as possible. Variations of Caco-2 cell permeabilities between different laboratories are often quite significant, but intralaboratory variation is generally low. The results between laboratories can be compared when standard compounds with known permeabilities are used. 91 This emphasizes the use of standard compounds with known permeabilities. When the characteristics of the Caco-2 cells used are known, the cell line is very suitable for permeation and uptake experiments as well as for the determination of absorption mechanisms and toxicity studies.
For permeation experiments, Caco-2 cells are seeded onto filter membranes on multiwell plates (Figure 2). Traditionally cells are grown on 12-well plates for 21–27 days, but accelerated methods on 96-well plates also exist.20,92-94 Transepithelial electrical resistance measurements are used to assure monolayer integrity.75,76,93,95,96 In addition, various markers representing different modes of absorption are used for reference compounds (Table 1). For example, fluorescent Lucifer yellow and radioactive [14C]-mannitol could be used to evaluate tight-junction formation, and verapamil and rhodamine 123 could be used as markers for efflux transportation.17,72,93,97 Transportation across a cell membrane could be studied both in the apical-to-basolateral (AP-BL) as well as basolateral-to-apical (BL-AP) directions by taking samples from different sides of the membrane. When samples are collected at various timepoints, apparent permeability coefficients Papp (cm/s) could be calculated to illustrate and compare the results from various experiments.
Insert membrane on which cells are grown for permeation experiments.
Model Compounds Suggested by the US Food and Drug Administration 91 for use in Establishing the Suitability of a Permeability Method.
Evaluation of the Effects of Natural Extracts on the Permeability of Drugs
Caco-2 monolayers are a suitable tool for studying drug–natural product interactions in vitro.17,98-101 However, it must be kept in mind that a natural product may change the pH of the test solution and affect the permeabilities of the drugs. 17 Acidic compounds are transported more effectively in acidic environments, and basic compounds behave the opposite way. pH monitoring and buffered solutions are thus essential. Also, colored test substances may interfere with the concomitant evaluation of marker substances in absorption-based detection. Such compounds are, for example, Lucifer yellow and rhodamine 123 used to detect monolayer integrity and efflux activity. 93 The use of precolumn is also recommended for samples analyzed with high-performance liquid chromatography.
The United States Food and Drug Administration (FDA) 91 has suggested model drugs suitable as method-validation references for permeability experiments (Table 1). These compounds have also been used to evaluate natural product effects on their permeabilities, and examples of such studies are discussed in the following sections.
Flaxseed (Linum usitatissimum L., Linaceae)
Flaxseed is a rich source of fibers and lipids, such as alpha-linoleic acid and linoleic acid. 102 Flaxseed contains 35% to 45% fiber, of which 2/3 is insoluble and 1/3 is soluble fibers. The soluble fiber portion contains polysaccharides and forms mucilage in the intestine at pH 6–8. A traditional use of flaxseed is for relieving constipation, but nowadays its value has been recognized in other areas too. The fibers of flaxseed improve glycemic control and are associated with the reduction of GI cancers. 103 The lignan precursors of flaxseed, secoisolariciresinol, and matairesinol, are converted by the microbial flora in the colon to enterodiol, and further to enterolactone. These phytoestrogens could reduce the risk of hormone-derived cancers such as prostate and breast cancer.104-106
The study of flaxseed effects on the permeation of a set of model drugs showed that a flaxseed methanol extract reduced the permeation of metoprolol, ketoprofen, paracetamol, and mannitol drugs. 17 No effects were seen on the permeation of efflux substrates verapamil and rhodamine. The reduction in permeabilities was likely due to the mucilage, soluble fibers, and lignans in the flaxseed extract. The reduction was greatest on the transcellularly permeating metoprolol. In the intestine, soluble fibers absorb fluids and cause swelling of intestinal contents. Dissolved drugs present in the intestine may thus be trapped in the mucilage. In addition, mucilage disturbs the contact with intestinal enterocytes. Concomitant drug intake with a great amount of flaxseed could thus lead to a reduction in the fraction absorbed of the drug.
Rapeseed (Brassica napus L., Brassicaceae)
Rapeseed oil is widely used for cooking. It has a beneficial fatty acid composition and is recommended to be used instead of more-saturated fats such as animal fats. 107 It is a rich source of essential omega-6 and omega-3 fatty acids which the human body cannot synthesize. These fatty acids are associated with reduced cholesterol levels and a lower risk of coronary heart disease.108,109 The main phenolics in rapeseed meal extract, sinapine and sinapic acid, have antimicrobial and antioxidative properties.110,111 A rapeseed oil extract containing 87% vinylsyringol has an excellent anti-inflammatory capacity and inhibits effectively the formation of inflammatory mediators nitric oxide (NO) and prostaglandin E2 (PGE2). 18
A rapeseed oil extract of 1 mg/mL had no effects on model drugs’ (metoprolol, ketoprofen, verapamil, paracetamol) transportation across a Caco-2 cell monolayer, whereas a rapeseed meal extract of 1 mg/mL increased ketoprofen and verapamil transportation. 18 Ketoprofen permeation is facilitated by the MCT1,112,113 and verapamil is a known substrate of P-gp.114-116 Rapeseed meal contains 64% sinapic acid, whereas rapeseed oil extract contains only 6% sinapic acid.111,117 Instead, the oil extract contains 87% vinylsyringol. This difference could explain their different effects on ketoprofen permeation. The effects might also be due to interference by active absorptive and efflux transport.112,113
Purple loosestrife (Lythrum salicaria L., Lythraceae)
Purple loosestrife contains a significant amount of plant phenolics such as flavonoids and tannins.118,119 It is traditionally used as tea to treat diarrhea but has been shown to possess antimicrobial and antifungal activity. 118
Purple loosestrife methanol extract has been shown to reduce the Caco-2 permeation of basic drugs metoprolol and verapamil and to increase the permeation of acidic ketoprofen and paracetamol. 17 This is assumed to be due to the tannins and phenolic acids in the extract, which could lower the pH at the site of absorption. According to the pH-partition theory, relative contributions of the unionized forms of acids are bigger in acidic environments; meanwhile, basic compounds behave in the opposite way. When the unionized portion is increased, absorption is also increased.
Pine (Pinus sylvestris L., Pinaceae)
Pine bark water extract contains phenolic compounds such as phenolic acid glucosides, catechins, and catechin derivatives, taxifolins and taxifolin derivatives, lignan glucosides, and procyanidins.120,121 Pine bark has been shown to possess antibacterial, antioxidant, and anti-inflammatory activity.18,118,120 Its anti-inflammatory activity is associated with inhibition of the production of NO and PGE2, which act as anti-inflammatory mediators.18,118,120
Pine bark water extract decreased metoprolol and verapamil transportation across a Caco-2 cell membrane, while it did not have any significant effects on acidic compounds’, paracetamol and ketoprofen, permeability, nor on paracellularly permeating mannitol. 17 As was described in the earlier section on purple loosestrife, these effects could be due to tannins and phenolic acids in the extract which lower the pH in the site of absorption. However, in another study, a pine bark acetone extract and its chloroform-extracted subfractions enhanced the transportation of metoprolol. 18 This clearly shows that with different extraction methods, different kinds of compounds could be obtained, and despite the common origin of the extracts, the effects on concomitant ingestion of drugs may vary. This example indicates that concomitant ingestion of concentrated pine bark products with medication should be avoided or at least discussed with a physician.
Echinacea
The echinacea species Echinacea purpurea (L.) Moench, Echinacea angustifolia DC., and Echinacea pallida (Nutt.) Nutt. (all Asteraceae) are widely used and sold herbal medicinal products for treating upper respiratory infections and common colds.100,122-124 They are known for immunostimulatory, anti-inflammatory, antioxidant, antibacterial, antiviral, and larvicidal activity, and they are also used for treating toothache, bowel pain, snake bites, skin disorders, seizures, chronic arthritis, and cancer. Echinacea species contain different classes of pharmacologically active secondary metabolites such as caffeic acid derivatives, polysaccharides, alkamides, and glycoproteins. Caco-2 studies with echinacea-containing products have shown that echinacea could increase P-gp-mediated efflux at low concentrations but decrease it at high, from 0.4 dry weight/mL, concentrations.99,100 The results indicate that an increase in the absorption of P-gp substrates, e.g., digoxin and indinavir, could occur at high concentrations of echinacea obtained from commercial products. On the other hand, in another Caco-2 cell study with E purpurea, derived active compounds cichoric acid, echinacoside, and tetraenes did not induce P-gp. 125 The results indicate that the compounds in an echinacea extract might have synergistic effects. Moreover, clinically, the effect of a concomitant medication could increase due to echinacea's effects on metabolizing CYP enzymes.41,126 Thus, the amount of the drug in the body may increase also this way. In addition, echinacea's effect may also lead to the formation of toxic metabolites. To further complicate the situation, it has been shown that the inhibitory action against metabolizing CYP enzymes may differ widely and is associated with the extract's alkylamide content. 127 These results show that caution is needed when using echinacea products with a medication.
Black Ginger (Kaempferia parviflora Wall. ex. Baker)
Black Ginger (Kaempferia parviflora Wall. ex. Baker) is widely used herb in Asia and Thailand. 98 It is used wide variety of indications such as to stimulate sexual performance, relieve depression, 128 and anti-obesity.128-130 It is also reported to possess neuroprotective, anti-inflammatory, anticancer, antiallergic, and cardio-protective properties.34,131-136 It contains secondary metabolites, especially flavonoids and flavonoid glycosides, among which 5,7-dimethoxyflavone (DMF) is the major constituent.98,137
A Caco-2 cell study showed that black ginger 10 µg/ml ethanol extract and 5,7-DMF 5 µM were able to increase digoxin, estrone stulfate, and paclitaxel absorption. 98 The studies suggested that the observed increase was due to the inhibition of P-gp and BCRP. The results were confirmed with Madin-Darby canine kidney cell line II (isolated from a high passage parental cell line) (MDCKII) cell lines. It remains unclear if effective concentration could be obtained in gut lumen, so clinical studies are needed to verify the situation in vivo. Nevertheless, black ginger could be a potent herb to act as an adjuvant to improve the bioavailability of low-absorptive drugs. 138
Berries
Berries are a rich source of flavonoids, ellagitannins, anthocyanins, and vitamins.18,139,140 Ellagitannins are not absorbed in the gut lumen as such but are metabolized by intestinal flora into urolithins.141,142 Caco-2 cell studies with ellagic acid, complexed in the form of ellagitannins, have shown that AP-BL transport across the cell membrane is very low. 143 Anthocyanins are colorful water-soluble pigments, found extensively in blue, red, and purple berries. 144 A study of anthocyanin absorption across Caco-2 cell monolayers indicates that anthocyanins can be transported across the cell membrane in intact glycone form, although the transportation efficiency was quite low.145,146 Moreover, anthocyanins are metabolized by the CYP3A4 isoform, which increases the possibility of interactions.145,147 Berries are good antioxidants, and as they have a rich flavonoid content, they have a supportive role in cancer prevention.144,148-153 Bilberry, cowberry, and raspberry possess antimicrobial activity against Gram-negative bacteria but not against Gram-positive bacteria. 154 Gram-positive and Gram-negative bacteria have different cell surface structures, which could explain the difference in the activity. The outer membrane of Gram-negative bacteria functions as a barrier against hydrophobic compounds, and the degree of hydroxylation correlates with the antimicrobial activity of phenolic compounds.154,155
Components of berries might have effects on the absorption of drugs. Glycosidic and aglycosidic bilberry (Vaccinium myrtillus L., Ericaceae), cowberry (Vaccinium vitis-idaea L., Ericaceae), and raspberry (Rubus idaeus L., Rosaceae) extracts exhibited only very minor effects on the permeation of the model drugs metoprolol, ketoprofen, verapamil, and paracetamol. 17 On the contrary, a raspberr yacetone extract with an ellagitannin content of 50% of total phenolics (554 mg/g of freeze-dried extract) and its concentrated ellagitannin fraction (80% ellagitannins) and concentrated anthocyanin fraction exerted effects on the permeability of these drugs.16,18 The raspberry extract and ellagitannin fraction showed very similar effects on permeation, whereas the effect of the concentrated anthocyanin fraction was smaller and in the opposite direction. 16 Metoprolol permeability decreased slightly, but the effect was so small that severe adverse effects are unlikely to occur with this medicine and these kinds of raspberry samples. The permeability of verapamil decreased considerably when co-administered with the raspberry extract, and the ellagitannin fraction and permeability of ketoprofen were markedly decreased by the ellagitannin fraction in a dose-dependent manner. On the contrary, the anthocyanin fraction slightly increased the permeabilities of ketoprofen and verapamil. The permeability of paracetamol was not affected by any of the raspberry samples. This clearly shows that when components are used in a more concentrated form, adverse effects may occur even though a milder mixture of the same components lacks an effect. In some cases, the effects might disappear.
Ellagitannins and ellagic acid have been indicated to accumulate in cells.143,156 The interaction with cell membranes could explain also the reduced permeation of ketoprofen and verapamil. The CYP activity in Caco-2 cells is generally low, but the metabolism of anthocyanins might contribute to the permeability changes of model drugs.14,147 The regional change in the membrane due to slowly permeating anthocyanins could have an effect on the permeability of verapamil and ketoprofen as well.
Herbs
Herbs are rich sources of flavonoids. Flavonoids are polyphenolic secondary metabolites that are universally represented throughout the plant kingdom. They are found to some extent in almost all plant families in the leaves, stems, flowers, roots, and seeds of plants. 157 The basic structure of flavonoids comprises 15 carbon atoms which form 2 benzene rings joined by a linear 3-carbon chain representing a C6–C3–C6 system.151,158 The 3-carbon chains usually form a 6-membered or 5-membered heterocyclic ring involving an oxygen atom, which is one of the criteria in the classification of flavonoids. Flavonoids are categorized into several subgroups; flavonols (e.g., quercetin, myricetin, kaempferol) and flavones (luteolin, apigenin) are the most common groups in plant-based foods. 154 In nature, flavonoids occur mainly in glycoside form.158,159
Plants containing flavonoids have been used as disease-preventive and therapeutic agents for thousands of years in Eastern medicine. 149 Green tea, berries, herbs, vegetables, black chocolate, onions, and red wine are very rich sources of flavonoids. A number of studies on flavonoids have focused on their antioxidant properties, e.g.,160-163 and on how the consumption of natural products rich in flavonoids has beneficial health effects.158,164,165 Flavonoids are suggested to have beneficial effects on various diseases, including cancer, neurogenerative disorders, and cardiovascular disease.160,166-172 Flavonoids modulate cell signaling pathways, but their precise mechanism of action is unclear. They are strong antioxidants, but this is not the sole explanation for their beneficial effects. The hydroxylation pattern of the B-ring in flavones and flavonols seems to have an important influence on antitumor activity, 149 while the number of hydroxyl groups is shown to be essential for antimicrobial activity. 154
Caco-2 permeability studies with water extracts of oregano (Origanum vulgare L., Lamiaceae), sage (Salvia officinalis L., Lamiaceae), and rosemary (Rosmarinus officinalis L., Lamiaceae) indicated that these herbs cause partial opening of the paracellular space. 17 The transportation across a Caco-2 cell monolayer of the paracellular marker mannitol was clearly increased in the presence of water extracts of these herbs. The transportation of slowly paracellularly permeating furosemide was also markedly increased in the presence of a 1-mg/mL oregano and rosemary extract. On the contrary, no effects were seen on the transportation of rapidly transcellularly permeating compounds, nor on the efflux marker rhodamine. In another study, with MCF-7 breast cancer cell line, a rosemary methanolic extract was found to inhibit P-gp. 173 In the presence of rosemary extract, doxorubicin and vinblastine were accumulated in the cells overexpressing P-gp. No accumulation of the drugs in the presence of rosemary extract was seen in the cells that lacked P-gp. A recent study with the rosemary phytochemicals carnosic acid, carnosol, and ursolic acid showed also an inhibitory effect against P-gp, which was seen in the accumulation of rhodamine and daunorubicine. 97 The controversial results from different studies could be due to distinct extraction methods that yield various compounds. Also, the starting material might contain unequal amounts of compounds due to different growing conditions. The results also show that pure compounds could have an inhibitory effect that may cause interactions, even though milder extracts lack effects.
Anthranoid Laxatives
Anthranoid-structured compounds from senna plants (Cassia senna L., Fabaceae; Cassia angustifolia M.Vahl, Fabaceae) are widely used laxatives. The core structure of the compounds is an anthracene ring where a hydroxyl or carbonyl group is in position C-9, and a C-8 hydroxyl enhances the laxative function.174,175 Anthranoid laxatives have the ability to enhance fluid secretion to the small and large intestine by disrupting the tight junctions between epithelial cells.176-178
Caco-2 absorption studies with senna anthranoids rhein, danthron, sennidins A/B, sennosides A/B, and senna-leaf infusion have indicated that absorption interactions of highly permeable drugs are unlikely to occur. 101 Rhein and danthron affected tight junctions and opened the paracellular space. Thus, interactions with slowly paracellularly permeating compounds might occur, especially when doses are high. Another study with A/B sennosides and their aglycon A/B sennidins indicated that these compounds are poorly transported in the apical-to-basolateral (AP-BL) direction but that transportation is higher in the BL-AP direction. 96 The result indicates the involvement of efflux transporters which pump the compounds back to the lumen. Thus, no extensive absorption of senna laxatives is likely to occur.
Traditional Medicinal Plants
“Cancer bush” (Sutherlandia frutescens L., Fabaceae) and “African potato” (Hypoxis hemerocallidea Fisch., C.A.Mey. & Avé-Lall., Hypoxidaceae) are herbal medicines used to treat HIV/AIDS.179-181 These African plants have traditionally been used to treat various disorders such as cancer, diabetes, influenza, rheumatoid arthritis, and stomach problems.182-186 The principal compounds in the plants assumed to be responsible for the pharmacological properties are L-canavanine in S. frutescens and sterols and sterolins in H. hemerocallidea.179,181 In Caco-2 cell experiments, a 20-mg/mL water extract of H. hemerocallidea and a DPBS solution with 1.22 mg of added L-canavanine reduced statistically significantly the BL-AP transport of 5 mM nevirapine, an antiviral compound. 179 The inhibitory effect of a 10-mg/mL water extract of S. frutescens was not statistically significant. The result indicates that the amount of L-canavanine S. frutescens extract is less than 1.22 mg and that the inhibition of efflux is concentration-dependent. The selection of extract concentrations was based on the traditional use of the herbs. Nevirapine acts as a reverse transcriptase inhibitor and has been reported to be a P-gp substrate.187,188 H. hemerocallidea and L-canavanine interact with efflux proteins, and thus a controlled concomitant administration of H. hemerocallidea and L-canavanine protease inhibitors may have beneficial effects on the absorption of drugs. 179 On the other hand, if H. hemerocallidea and L-canavanine are used, uncontrolled side effects might occur due to the higher bioavailability of nevirapine.
The traditional medicinal plants bitter leaf (Vernonia amygdalina Del., Asteraceae), neem leaf (Azadirachta indica A.Juss., Meliaceae), brimstone bark (Morinda lucida Benth., Rubiaceae), lemongrass (Cymbopogon citratus Stapf, Poaceae), pawpaw leave (Carica papaya L., Caricaceae), African mistletoe leave (Tapinanthus sessilifolius Blume, Loranthaceae), and turmeric (Curcuma longa L., Zingiberaceae) are used to treat various disorders such as malaria, cancer, fever, wounds, and hypoglycemia, and they are used for their antimicrobial effects.189-195 These plants contain various phytochemicals such as flavonoids, tannins, coumarins, alkaloids, glycosides, terpenes, and saponins that are responsible for the plants’ medicinal activity. Some of these plants are also used for their nutritional value or as common kitchen herbs and spices, so it is important to know their effects on concomitant medication. A Caco-2 cell study investigated the influence of these herbals on the transport of a known efflux substrate, digoxin. 192 The extract's effect on AP-BL and BL-AP transport was further investigated in the presence of known P-gp inhibitor verapamil. Based on the study, bitter leaf (Vernonia amygdalina), African mistletoe leave (Tapinanthus sessilifolius), and pawpaw leave (Carica papaya) inhibited P-gp-mediated efflux at concentration levels 0.02–20 mg/mL. It was suggested that the flavonoids might have a major role in the inhibition. The results indicate that interactions might occur also with other P-gp-substrate drugs when absorption is increased. These drugs include, for example, many anticancer and some antimalarial drugs such as tamoxifen, topotecan, imatinib, etoposide and chloroquine, quinidine, respectively.
Perspectives
The use of natural products is vivid and as discussed by this article they have many health-promoting effects. Besides, the concerns about their effects to concomitant medication and unwanted effects, natural products can also be used to lower the dose of prescribed medicine and relieve possible side effects.138,196 For example, flavonoids may act as sensitizers when used in combination with radiation and anticancer drugs. 151 Hence, lower doses of radiation or chemotherapeutic drugs could be used and less adverse effects could be obtained with similar efficacy.138,151 In cancer, cell proliferation and differentiation have been disturbed. Dietary polyphenolic compounds have been shown to have various anticancer effects such as cell growth and kinase activity inhibition, apoptosis induction and tumor-invasive behavior. 149 Dietary phenolics can work in ways that arrest or reverse carcinogenic changes before the appearance of malignant disease. 151
These facts encourage us to consume natural products straight from nature as such and store-bought in even more concentrated forms. The products are in general safe, but the consumer has to report product use to their physician, especially when various drugs are used. The examples in Table 2 and reviewed in this paper clearly show that when compounds are used in more concentrated form, adverse effects may occur. Also, a mixture with synergistically acting substances can exhibit activity even though no single compound in the mixture has an effect by itself.
Natural Products and Their Suggested Effects on the Model Drug Permeation.
↑indicates increase of absorption of the drug in the presence of natural product extract or fraction and ↓ indicates decrease of absorption of the drug in the presence of natural product extract or fraction. Abbreviations: BCRP, breast cancer resistance protein; Caco-2, human colon adenocarcinoma cell line; MCT1, monocarboxylic acid transporter 1; WE, water extract; AE, acetone extract; CS, chloroform extracted subfractions; EE, ethanol extract; SI, Senna infusion; D, danthron; RH, rhein.
Caco-2 cell monolayers are a useful tool for studying the effects on the absorption of natura lproduct compounds as well as their effects on concomitant medication. When we have information about their absorption, beneficial natural compounds are safer to use and possible interactions can be taken into account.
The struggle against interactions requires effort from both consumers and industry. Patients should always discuss the products that they want to use with their physician. Clinicians should not judge their patient's wish to use a certain product 35 as this might scare a patient with strong beliefs and hinder open conversation in the future. To make correct decisions on whether a product is safe to use, manufacturers should specify its composition. 197 Products have to be examined case-by-case and too deep generalizations should not be made. Despite the possible effects, consumers should not be too scared of using natural products. When people tell their physician or pharmacist about the use of natural compounds and certain medications, they can be advised on whether concomitant use is safe or not. 196 Also, pharmacists should be active in informing customers about herbal medicinal products. This highlights the need for co-operation between various healthcare professionals. 197
Conclusions
The Caco-2 cell model is a useful, regulatory-accepted tool for predicting drug absorption and permeability of substances. Using this model reduces animal experimentation, but enables screening with live material. The Caco-2 cell absorption model has also been used to study how natural products, such as flaxseed, rapeseed, purple loosestrife, pine, echinacea, certain berries and herbs, anthranoid laxatives, and traditional medicinal plants, affect the absorption of co-administered drugs, and interactions and adverse effects that might occur. Overall, we conclude that the Caco-2 cell absorption model is valuable in studying natural product—drug interactions and can provide supporting information for the safe use of natural products with medicines.
Footnotes
Abbreviations
Author contributions
AG made the literature search and wrote the manuscript, which was then critically reviewed and discussed with PT. Both authors reviewed and approved the final manuscript.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Helsinki University Library.
