Pharmacogenomics and Herb-Drug Interactions: Merge of Future and Tradition

The worldwide using of herb products and the increasing potential herb-drug interaction issue has raised enthusiasm on discovering the underlying mechanisms. Previous review indicated that the interactions may be mediated by metabolism enzymes and transporters in pharmacokinetic pathways. On the other hand, an increasing number of studies found that genetic variations showed some influence on herb-drug interaction effects whereas these genetic factors did not draw much attention in history. We highlight that pharmacogenomics may involve the pharmacokinetic or pharmacodynamic pathways to affect herb-drug interaction. We are here to make an updated review focused on some common herb-drug interactions in association with genetic variations, with the aim to help safe use of herbal medicines in different individuals in the clinic.


Introduction
Herbal medicinal products (HMPs) belong to a main part of complementary and alternative medicine (CAM). The National Center for Complementary and Alternative Medicine (NCCAM) has defined CAM as those health care practices not currently considered an integral part of conventional medicine [1]. During the past two decades, the world has witnessed an increased rate of conventional medicines integrated with CAM for treatment. It was estimated that in 1990 about one-third of the US population used unconventional therapies and the expenditures on this amount to $13.7 billion [2], whereas in 2007 approximately 38 percent of American adults (83 million) are using some form of CAM for their health, and the expenditures reached $34 billion [1]. Correspondingly the funding for NCCAM has reached $124.125 million in 2013, much more than that was in 1998 ($19.5 million) when it was established [1]. In eastern Asia, countries that have a tradition use of oriental herb medicine for centuries, the prevalence of CAM use is much higher. In Japan, the use of CAM in the general population had been reported to be 76% [3]. In China, 93.4% of cancer patients reported having used CAM in 2009-2010 [4]. In South Korean population reported a range from 29% to 83% [5].
The prevalence of HMPs and other forms of CAM for treatment have its reasons. Conventional medicines are relatively more expensive, more difficult, and inconvenience of access and sometimes more likely to get side effects, toxicities, and incomplete efficacy. CAM, such as traditional Chinese medicine (TCM), though empirical, has been used for centuries and proved to be effective. In developing countries like China patients, especially those in rural areas, are more likely to choose HMPs as main medicine for treatment. In developed countries or areas like Japan, the increased use of CAM has promoted the government approval of 148 traditional Japanese (or Chinese) herbal medicines listed on the "National Health Insurance Drug Tariff " [6]. Western countries, however, in most cases use herb prescriptions as dietary supplements or as part of a traditional medicine, and this trend is still rising. This increased the chance of concurrently use of herbs and conventional drugs. Similar to drug-drug interactions, the multicomponent herbs can exert their effects on pharmacokinetics and consequently lead to herb-drug interactions (HDIs) [7]. Understanding the exact 2 Evidence-Based Complementary and Alternative Medicine mechanism of each extract of an herb that affects drug's absorption, distribution, metabolism, and excretion (ADME) properties is incomplete at the moment. A review of previous studies indicated that pharmacogenetic factors may help to explain the HDIs effect and give us instructions for rational drug administration.

Herb-Drug Interactions
The incidences of herb products to affect drug responses maybe more often in the clinic than expected for it is more easy for patients to conduct self-administration without informing their health care providers. Some of the interactions may have a beneficial effect by increasing drug efficacy or diminishing potential side effects. For example, combined therapy of garlic (250 mg/kg) with captopril demonstrated higher synergistic action with respect to fall in blood pressure and ACE inhibition [8]; patients received silymarin (140 mg three times daily) in combination with conventional desferrioxamine therapy showed beneficial effects on thalassemia patients [9]. However, more often the potential side-effect and outcome influence may be brought about by combination use of HMPs and conventional drugs, and these circumstances are always difficult to predict. Previous studies identified various mechanisms of pharmacokinetic HDIs, mainly mediated by drug-metabolizing enzymes and transporters [10]. For example, St. John's wort (Hypericum perforatum) significantly reduced the area under the plasma concentration-time curve (AUC) and blood concentrations of cyclosporine, midazolam, tacrolimus, amitriptyline, digoxin, indinavir, warfarin, phenprocoumon, and theophylline. Most of them are substrates of cytochrome P450s (CYPs) and/or P-glycoprotein (P-gp) [11]. An important fact is that most of these affected drugs have very narrow therapeutic indices. So, adverse drug reactions, toxicities, and treatment failure are more likely to occur when they are integrated with herbs. To date an increasing number of studies in evaluating HDIs have been reported [11].
CYP enzymes can be induced by various herbs. It was reported that Danshen-Gegen formula (DGF) could induce the liver phase I metabolism of warfarin, especially CYP1A1 and CYP2B1, as a result, it increased the intestinal absorption of warfarin by >30% and decreased its plasma protein binding by >11.6% [12].
A clinic case reported that preadministration of St. John's wort significantly decreased stable plasma level of clozapine from 0.46-0.57 mg/L to 0.19 mg/L, after discontinuation of St. John's wort for a month, the clozapine concentration restored to normal [13]. As the author discussed, St. John's wort may pharmacokinetically interact with clozapine through inducing a number of CYP450 enzymes such as CYP3A4, CYP1A2, CYP2C9, and CYP2C19, which are responsible for clozapine metabolism. In another case it is reported that Ginkgo biloba can negatively influence the effect of antiretroviral drug Efavirenz (EFV) in a HIV infected male patient. This may due to the inducing effect of Ginkgo biloba on the EFV metabolism enzymes CYP2B6 and CYP3A4 [14].
Recently, an in vivo study showed that coadministration of a single dose of baicalin, a Chinese medicine isolated from Scutellaria baicalensis, resulted in a dose-dependent decrease of midazolam in its clearance (CL) from 25% to 34%, while it increased in AUC 0-∞ from 47% to 53%. Pretreatment of baicalin also reduced midazolam CL by 43%, with an increase in AUC 0-∞ by 87%. Meanwhile, multiple doses of baicalin decreased the expression of hepatic CYP3A2 by approximately 58%. This indicated that baicalin induced changes in the pharmacokinetics of midazolam may be due to its inhibition of CYP3A in the liver [15]. There is a letter that analyzed a number of separated studies and reported that a well-known Chinese medicine ginseng may have interaction with anticoagulant warfarin by reducing warfarin's plasma levels and anticoagulant effect in ischemic stroke patients [16]. However, the exact mechanism of ginseng-warfarin interaction is unknown and needs to be thoroughly investigated.
Most of study reported that HDIs mechanisms are associated with pharmacokinetics and pharmacodynamics related genes, mainly including CYP450 enzymes and/or Pglycoprotein (P-gp) [11,17] and also UDP-glucuronosyltransferases (UGTs) [18]. For more details we can refer to the reviews [11,19,20]. Some studies found that herb induced expression changes of genes may have some specificity; for instance, it was reported that coadministration of Marsdenia tenacissima extract (MTE) with gefitinib significantly decreased the in vitro intrinsic clearance (Clint) of gefitinib by 2.6-and 4.0-fold for CYP2D6 and CYP3A4, respectively, but did not affect other CYP450s [21]; another study reported that hyperoside could selectively inhibit CYP2D6 activity in a dose dependent manner and might cause herb-drug interactions when coadministrated with CYP2D substrates [22].
However, whole herbs like garlic, St. John's wort, Gingko, and so forth are complex compounds. Various confounding factors driving the pharmacokinetic profiles of them are considered to be a crucial factor when evaluating HDIs. In these cases we should make sure the confounding factors are at the same levels in the research situations. While evaluating the pharmacokinetic synergy effects of different ingredients in a whole herb on HDIs remains a big challenge so far. Previous studies have made us believe that the genetic involvement in the pharmacokinetic pathways may help us to understand this effect. Recent pharmacogenomics studies in the following and cases listed in Tables 1 and 2 of these pharmacokinetics and pharmacodynamics related genes may help explain the individual differences in HDIs.

Pharmacogenomics in HDIs Mechanisms
Genetic polymorphisms have long been studied and believed to involve pharmacokinetic and pharmacodynamic pathways that cause individual difference in drug responses [47]. Herbal medicines are actually a combination of potentially biologically active compounds possessing various inherent pharmacological activities. The metabolism of these compounds usually occurs by the same mechanisms as that of drugs. Pharmacokinetic interactions mediated by drugmetabolizing enzymes or transporters are involved in many herb-drug interactions [48], and as the concentrations of the Evidence-Based Complementary and Alternative Medicine 3 The AUC ratio of hydroxybupropion to bupropion tended to be more pronounced lower in * 6/ * 6 compared with * 1/ * 1 genotype patients (5.3 versus 8.0) after baicalin treatment. [23] Echinacea Dextromethorphan CYP2D6

Extensive metabolizers
Poor metabolizer Echinacea dosing reduced the oral clearance of dextromethorphan by 28% and increased AUC by 42% in the CYP2D6 poor metabolizers, while extensive metabolizers were not affected. [24]
herbal extract and the drug may determine the degree of DHI, polymorphisms in the drug-metabolizing enzyme genes (listed in Table 1) and transporter genes (listed in Table 2) that alter the systemic exposure to the substrate drugs or active components of herbs may affect the risk of interaction [10,48,49]. On the other hand, drug metabolism or transport has alternative pathways. When poor metabolizer genotype or inhibition of an enzyme may lead to the drug metabolized by another pathway enzymes, which may be more sensitive to their native substrates, then it may cause the competition metabolism of drugs or herbs and eventually HDIs difference. So conducting pharmacogenomics studies on HDIs may help to illuminate the fundamental mechanisms underlying the HDIs. In the following, we reviewed pharmacogenomics studies, mainly from clinic study, on some important herb products that have been reported to have HDIs.

St. John's Wort.
St. John's wort is an herb most commonly used for depression and conditions that sometimes go along with depression such as anxiety, tiredness, loss of appetite, and trouble in sleeping. There is some strong scientific evidence that it is effective for mild to moderate depression. Currently it is displayed that 95 drugs (296 brand and generic names) are known to have a major interaction with St. John's wort. After administration of St John's wort, the AUC 0-∞ of nifedipine and dehydronifedipine decreased by 42.4% and 20.2% in PXR H1/H2 genotype and 47.9 and 33.0% in H2/H2 genotypes, whereas for the H1/H1 the AUC 0-∞ of nifedipine decreased by 29.0%, but the AUC 0-∞ of dehydronifedipine increased by 106.7% [30]. St. John's wort treatment significantly increased phenytoin clearance in CYP2C19 EMs ( * 2, * 3) but not in PMs and decreased the plasma concentrations of omeprazole in a CYP2C19 genotype-dependent manner [33]. Subjects harbouring the ABCB1 haplotype comprising 1236C>T, 2677G>T/A, and 3435C>T polymorphisms had lower intestinal MDR1 mRNA levels and showed an attenuated inductive response to St. John's wort as assessed by talinolol disposition [45]. The AUC of voriconazole was decreased by 59% with St. John's wort treatment, with a 144% increase in oral clearance of voriconazole. The baseline apparent oral clearance of voriconazole and the absolute increase in apparent oral clearance were smaller in CYP2C19 * 2 carriers than those with CYP2C19 * 1/ * 1 genotype [32].

Ginkgo.
Ginkgo is often used for memory disorders including Alzheimer's disease. It is also used for conditions that seem to be due to reduced blood flow in the brain, especially in older people. These conditions include memory loss, headache, ringing in the ears, vertigo, difficulty concentrating, mood disturbances, and hearing disorders. Ginkgo enhanced omeprazole hydroxylation in a CYP2C19 genotype-dependent manner. The decrease was greater in CYP2C19 PMs ( * 2, * 3) than in EMs [25].

Grapefruit
Juice. Grapefruit juice, and grapefruit in general, is a potent inhibitor of the cytochrome P450 CYP3A4 enzyme, which can affect the metabolism of a variety of drugs, increasing their bioavailability. In some cases, this can lead to a fatal interaction with drugs like astemizole or terfenadine. Grapefruit juice treatment significantly increased total AUC of lansoprazole in CYP2C19 PMs ( * 2, * 3), and the total AUC of lansoprazole sulfonic/lansoprazole was 6 Evidence-Based Complementary and Alternative Medicine significantly decreased in CYP2C19 homozygous EMs ( * 1/ * 1) [26]. Homozygous wild types of ABCB1 3435C>T but not the other genotypes showed a significant decrease in the active metabolite carebastine urinary excretion after grapefruit juice [43].
3.6. Cranberry Juice. Cranberry juice is the juice of the cranberry. Cranberry juice contains phytochemicals, which may help prevent cancer and cardiovascular disease. Cranberry juice is high in oxalate and has been suggested to increase the risk for developing kidney stones, although more recent studies have indicated it may lower the risk. Cranberry significantly increased the area under the INR-time curve by 30% when administered with warfarin without altering pharmacokinetics or plasma protein binding of S-or Rwarfarin, and this effect is dependent on VKORC11173T>C polymorphism. Subjects with CT and TT genotypes coadministered with cranberry juice extract significantly reduced S-warfarin EC50 (concentration of S-warfarin that produces 50% inhibition of prothrombin complex activity) by 22% and 11%, respectively [50]. This case gives us an example of pharmacodynamic pathway gene polymorphisms that can be involved in the herb-drug interaction. A total of 194 impaired glucose tolerance (IGT) subjects treated with Tianqi Jiangtang for 12 months were genotyped for 184 mutations in 34 genes involved in drug metabolism or transportation. The rs1142345 (A>G) SNP in the thiopurine S-methyltransferase (TPMT) gene was significantly associated with the hypoglycemic effect of the drug ( = 0.001, FDR = 0.043). The "G" allele frequencies of rs1142345 in the healthy (subjects reverted from IGT to normal glucose tolerance), maintenance (subjects still had IGT), and deterioration (subjects progressed from IGT to T2D) groups were 0.094, 0.214, and 0.542, respectively. Rs1142345 was also significantly associated with the hypoglycemic effect of the drug between the healthy and maintenance groups ( = 0.027, OR = 4.828) and between the healthy and deterioration groups ( = 0.001, OR = 7.811). Therefore, rs1142345 was associated with the clinical effect of traditional hypoglycemic herbs. This is the first study to utilize the ADME gene chip in the pharmacogenetic study of traditional herbs [35].

Conclusion and Future Perspectives
Due to the limitations of conventional drugs, the use of herbal medicinal products (HMPs) as a CAM integrated with drugs for treatment has been increasing, this eventually will contribute to a rising incidence of herb-drug interactions. However, because of variability in herbal product composition, uncertainty of the causative constituents, and often scant knowledge of causative constituent pharmacokinetics, the mechanisms underlie herb-drug interactions remain an understudied area, and evaluation of herbal product interaction liability is still a big challenge [19]. Studies showed that these HDIs are mediated mainly by metabolism enzymes and transporters in the pharmacokinetic pathways. Polymorphisms in these pharmacokinetic pathways related genes can contribute to the individual difference of HDIs effect. We make an updated review with respect to pharmacogenomics related HDIs on some important herb products and call for a vigilance of using these drug for therapy when the patient is taking medication with narrow therapeutic indices.
Currently, there is still a lack of effective method to predict HDIs in the clinic; the main obstacles are laid by the multiconstituents of an herb. In the near future years, scientist in this area should focus on identifying individual constituents from herbal products, characterizing pharmacokinetics and pharmacodynamics of each individual constituents, discovering the fundamental genetic bases that eventually facilitate prospective identification of herb-drug interactions. Here we reviewed a definite involvement of pharmacogenomics in HDIs; some other mechanism, such as epigenetic regulations, that is, a recent hotspot area, is also supposed to take part in it.
The National Institute of Health (NIH) and similar organizations in other countries have increased the financial support for scientific investigation on safety use of herb medicines. This trend will promote the herb medicines more normalization as that of drugs, and the progress may promote finding new effective drugs from extracts of herbs. With respect to the wide and increasing use of herb drugs and their benefits, and with the progress in pharmacogenomics and pharm-chemistry, we are still optimistic on this area.