Integrated analysis of gut metabolome, microbiome, and exfoliome data in an equine model of intestinal injury.

BACKGROUND: The equine gastrointestinal (GI) microbiome has been described in the context of various diseases. The observed changes, however, have not been linked to host function and therefore it remains unclear how specific changes in the microbiome alter cellular and molecular pathways within the GI tract. Further, non-invasive techniques to examine the host gene expression profile of the GI mucosa have been described in horses but not evaluated in response to interventions. Therefore, the objectives of our study were to (1) profile gene expression and metabolomic changes in an equine model of non-steroidal anti-inflammatory drug (NSAID)-induced intestinal inflammation and (2) apply computational data integration methods to examine host-microbiota interactions. METHODS: Twenty horses were randomly assigned to 1 of 2 groups (n = 10): control (placebo paste) or NSAID (phenylbutazone 4.4 mg/kg orally once daily for 9 days). Fecal samples were collected on days 0 and 10 and analyzed with respect to microbiota (16S rDNA gene sequencing), metabolomic (untargeted metabolites), and host exfoliated cell transcriptomic (exfoliome) changes. Data were analyzed and integrated using a variety of computational techniques, and underlying regulatory mechanisms were inferred from features that were commonly identified by all computational approaches. RESULTS: Phenylbutazone induced alterations in the microbiota, metabolome, and host transcriptome. Data integration identified correlation of specific bacterial genera with expression of several genes and metabolites that were linked to oxidative stress. Concomitant microbiota and metabolite changes resulted in the initiation of endoplasmic reticulum stress and unfolded protein response within the intestinal mucosa. CONCLUSIONS: Results of integrative analysis identified an important role for oxidative stress, and subsequent cell signaling responses, in a large animal model of GI inflammation. The computational approaches for combining non-invasive platforms for unbiased assessment of host GI responses (e.g., exfoliomics) with metabolomic and microbiota changes have broad application for the field of gastroenterology. Video Abstract.

Identification and characterization of the enzymes responsible for the metabolism of the non-steroidal anti-inflammatory drugs, flunixin meglumine and phenylbutazone, in horses.

The in vivo metabolism and pharmacokinetics of flunixin meglumine and phenylbutazone have been extensively characterized; however, there are no published reports describing the in vitro metabolism, specifically the enzymes responsible for the biotransformation of these compounds in horses. Due to their widespread use and, therefore, increased potential for drug-drug interactions and widespread differences in drug disposition, this study aims to build on the limited current knowledge regarding P450-mediated metabolism in horses. Drugs were incubated with equine liver microsomes and a panel of recombinant equine P450s. Incubation of phenylbutazone in microsomes generated oxyphenbutazone and gamma-hydroxy phenylbutazone. Microsomal incubations with flunixin meglumine generated 5-OH flunixin, with a kinetic profile suggestive of substrate inhibition. In recombinant P450 assays, equine CYP3A97 was the only enzyme capable of generating oxyphenbutazone while several members of the equine CYP3A family and CYP1A1 were capable of catalyzing the biotransformation of flunixin to 5-OH flunixin. Flunixin meglumine metabolism by CYP1A1 and CYP3A93 showed a profile characteristic of biphasic kinetics, suggesting two substrate binding sites. The current study identifies specific enzymes responsible for the metabolism of two NSAIDs in horses and provides the basis for future study of drug-drug interactions and identification of reasons for varying pharmacokinetics between horses.

Human Serum Albumin Aggregation/Fibrillation and its Abilities to Drugs Binding.

Human serum albumin (HSA) is a protein that transports neutral and acid ligands in the organism. Depending on the environment's pH conditions, HSA can take one of the five isomeric forms that change its conformation. HSA can form aggregates resembling those in vitro formed from amyloid at physiological pH (neutral and acidic). Not surprisingly, the main goal of the research was aggregation/fibrillation of HSA, the study of the physicochemical properties of formed amyloid fibrils using thioflavin T (ThT) and the analysis of ligand binding to aggregated/fibrillated albumin in the presence of dansyl-l-glutamine (dGlu), dansyl-l-proline (dPro), phenylbutazone (Phb) and ketoprofen (Ket). Solutions of human serum albumin, both non-modified and modified, were examined with the use of fluorescence, absorption and circular dichroism (CD) spectroscopy. The experiments conducted allowed observation of changes in the structure of incubated HSA (HSAINC) in relation to nonmodified HSA (HSAFR). The formed aggregates/fibrillation differed in structure from HSA monomers and dimers. Based on CD spectroscopy, previously absent ßstructural constructs have been registered. Whereas, using fluorescence spectroscopy, the association constants differing for fresh and incubated HSA solutions in the presence of dansyl-amino acids and markers for binding sites were calculated and allowed observation of the conformational changes in HSA molecule.

A critical view on the analysis of fluorescence quenching data for determining ligand-protein binding affinity.

The past decade has seen an increase in the number of research papers on ligand binding to proteins based on fluorescence spectroscopy. In most cases, determination of the binding affinity is made by analyzing the quenching of protein fluorescence induced by the ligand. However, many such articles, even those published in reputed journals, suffer from several mistakes with regard to analysis of fluorescence quenching data. Using the binding of phenylbutazone to human serum albumin as a model, we consider some of these mistakes and show how they affect the values of the association constant. In particular, the failure to correct for the inner filter effect and the use of unsuitable equations are discussed. Ligand binding data presented in these articles should be treated with caution, especially in the absence of data from complementary techniques.

Role of Herborn (K240E) and Milano Slow (D375H) human serum albumin variants towards binding of phenylbutazone and ibuprofen.

Human serum albumin (HSA) is the binding cargo in blood plasma. The binding of drugs to HSA determines the pharmacokinetics and pharmacodynamics of the drugs. There are 67 natural genetic variants of HSA were reported in literature. Studying the effect of albumin modifications on drug binding helps to treat the patients with proper medication. In the present study, we have aimed to understand the effect of two natural variants of HSA, such as Herborn (K240E) and Milano Slow (D375H) on the binding of phenylbutazone and ibuprofen. For this, we have generated K240E and D375H mutants and also double mutant (K240E/D375H) of HSA using site directed mutagenesis. The recombinant HSA and its variants were expressed in Pichia pastoris. The interaction of HSA and its variants to phenylbutazone and ibuprofen was studied using fluorescence spectroscopy. Our results showed that there is no significant effect of K240E and D375H mutations on phenylbutazone and ibuprofen binding. But the effect is significant when both the mutations were there in a single protein (K240E/D375H). Further, the CD spectroscopy data showed that there is no effect of phenylbutazone and ibuprofen binding on the conformation of protein, except in case of D375H, where there is a conformational change in the binding pocket with the ibuprofen binding.

Oxidative Alteration of Trp-214 and Lys-199 in Human Serum Albumin Increases Binding Affinity with Phenylbutazone: A Combined Experimental and Computational Investigation.

Human serum albumin (HSA) is a target for reactive oxygen species (ROS), and alterations of its physiological functions caused by oxidation is a current issue. In this work, the amino-acid residues Trp-214 and Lys-199, which are located at site I of HSA, were experimentally and computationally oxidized, and the effect on the binding constant with phenylbutazone was measured. HSA was submitted to two mild oxidizing reagents, taurine monochloramine (Tau-NHCl) and taurine dibromamine (Tau-NBr2). The oxidation of Trp-214 provoked spectroscopic alterations in the protein which were consistent with the formation of N'-formylkynurenine. It was found that the oxidation of HSA by Tau-NBr2, but not by Tau-NHCl, provoked a significant increase in the association constant with phenylbutazone. The alterations of Trp-214 and Lys-199 were modeled and simulated by changing these residues using the putative oxidation products. Based on the Amber score function, the interaction energy was measured, and it showed that, while native HSA presented an interaction energy of -21.3 kJ/mol, HSA with Trp-214 altered to N'-formylkynurenine resulted in an energy of -28.4 kJ/mol, and HSA with Lys-199 altered to its carbonylated form resulted in an energy of -33.9 kJ/mol. In summary, these experimental and theoretical findings show that oxidative alterations of amino-acid residues at site I of HSA affect its binding efficacy.

Influence of Antipsychotic Drug Risperidone on Human Serum Albumin Affinity to Organic Anions.

BACKGROUND: Risperidone is an antipsychotic drug. In blood, this drug binds mainly to human serum albumin (HSA) and is also transported by HSA. METHOD: To study certain details of the interaction between risperidone and HSA, a fluorescent dye CAPIDAN was used as a reporter. This dye specifically fluoresces from HSA in serum and is highly sensitive to structural changes in HSA including pathology-induced changes. Interaction of CAPIDAN with HSA has been studied using time-resolved fluorescence techniques. RESULT: The addition of phenylbutazone, a marker for the HSA drug-binding site I, leads to displacement of CAPIDAN from this site due to direct competition between phenylbutazone and the dye. The addition of risperidone induces a response of CAPIDAN fluorescence that is highly similar to its response to phenylbutazone. This response depends strongly on ionic strength and is very similar in both cases, phenylbutazone and risperidone. This similarity suggests that risperidone binds to HSA in the region of site I. In this site, the risperidone molecule probably covers the positive charge of Arginine 218 or Arginine 222 preventing their interaction with the CAPIDAN negatively charged carboxyl group. This effect was observed both in isolated HSA and in serum, suggesting similarity of the interaction. CONCLUSION: Thus, risperidone is able to prevent binding of organic anions (i.e. CAPIDAN as a drug-like molecule) to HSA.

Effects of Firocoxib, Flunixin Meglumine, and Phenylbutazone on Platelet Function and Thromboxane Synthesis in Healthy Horses.

OBJECTIVE: Determine the effects of nonsteroidal anti-inflammatory drugs (NSAID) on platelet function and thromboxane synthesis immediately after drug administration and following 5 days of NSAID administration in healthy horses. STUDY DESIGN: Randomized cross-over study. ANIMALS: Healthy adult horses (n=9; 6 geldings and 3 mares). METHODS: Horses received either flunixin meglumine (1.1 mg/kg IV every 12 hours), phenylbutazone (2.2 mg/kg IV every 12 hours), or firocoxib (loading dose of 0.27 mg/kg IV on day 1, then 0.09 mg/kg IV every 24 hours for 4 days) for a total of 5 days. Blood samples were collected prior to drug administration (day 0), 1 hour after initial NSAID administration (day 1), and then 1 hour post-NSAID administration on day 5. Platelet function was assessed using turbidimetric aggregometry and a platelet function analyzer. Serum thromboxane B2 concentrations were determined by commercial ELISA kit. A minimum 14 day washout period occurred between trials. RESULTS: At 1 hour and 5 days postadministration of firocoxib, flunixin meglumine, or phenylbutazone, there was no significant effect on platelet aggregation or function using turbidimetric aggregometry or a platelet function analyzer. There was, however, a significant decrease in thromboxane synthesis at 1 hour and 5 days postadministration of flunixin meglumine and phenylbutazone that was not seen with firocoxib. CONCLUSION: Preoperative administration of flunixin meglumine, phenylbutazone, or firocoxib should not inhibit platelet function based on our model. The clinical implications of decreased thromboxane B2 synthesis following flunixin meglumine and phenylbutazone administration are undetermined.

Analysis of phenylbutazone residues in horse tissues with and without enzyme-hydrolysis by LC-MS/MS.

Phenylbutazone (PBZ) is permitted to be used for the treatment of musculoskeletal pain and inflammation in race horses but it is not approved for use in horses destined for human consumption. In a recent study initiated in our laboratory to study the disposition of PBZ and its oxyphenbutazone (OXPBZ) metabolite in equine tissues, we compared the effect of an additional enzymatic hydrolysis step with ß-glucuronidase on the results of the analysis for PBZ without enzymatic hydrolysis. Incurred tissue samples obtained from a female horse dosed with PBZ at 8.8 mg/kg for 3 days and sacrificed 6 days following the last administration were used for this study. Liver, kidney, and muscle tissues were collected, extracted, cleaned up on a silica-based solid-phase extraction (SPE) preceded by a weak-anion exchange SPE and analyzed with our in-house validated liquid chromatography-tandem mass spectrometry (LC-MS/MS) method for PBZ and OXPBZ. Addition of the hydrolysis step resulted in a significant increase in recovery of both PBZ and OXPBZ residues. © 2016 Her Majesty the Queen in Right of Canada. Drug Testing and Analysis © 2016 John Wiley & Sons, Ltd.

Effects of non-enzymatic glycation in human serum albumin. Spectroscopic analysis.

Human serum albumin (HSA), transporting protein, is exposed during its life to numerous factors that cause its functions become impaired. One of the basic factors --glycation of HSA--occurs in diabetes and may affect HSA-drug binding. Accumulation of advanced glycation end-products (AGEs) leads to diseases e.g. diabetic and non-diabetic cardiovascular diseases, Alzheimer disease, renal disfunction and in normal aging. The aim of the present work was to estimate how non-enzymatic glycation of human serum albumin altered its tertiary structure using fluorescence technique. We compared glycated human serum albumin by glucose (gHSA(GLC)) with HSA glycated by fructose (gHSA(FRC)). We focused on presenting the differences between gHSA(FRC) and nonglycated (HSA) albumin used acrylamide (Ac), potassium iodide (KI) and 2-(p-toluidino)naphthalene-6-sulfonic acid (TNS). Changes of the microenvironment around the tryptophan residue (Trp-214) of non-glycated and glycated proteins was investigated by the red-edge excitation shift method. Effect of glycation on ligand binding was examined by the binding of phenylbutazone (PHB) and ketoprofen (KP), which a primary high affinity binding site in serum albumin is subdomain IIA and IIIA, respectively. At an excitation and an emission wavelength of λex 335nm and λem 420nm, respectively the increase of fluorescence intensity and the blue-shift of maximum fluorescence was observed. It indicates that the glycation products decreases the polarity microenvironment around the fluorophores. Analysis of red-edge excitation shift method showed that the red-shift for gHSA(FRC) is higher than for HSA. Non-enzymatic glycation also caused, that the Trp residue of gHSA(FRC) becomes less accessible for the negatively charged quencher (I(-)), KSV value is smaller for gHSA(FRC) than for HSA. TNS fluorescent measurement demonstrated the decrease of hydrophobicity in the glycated albumin. KSV constants for gHSA-PHB systems are higher than for the unmodified serum albumin, while KSV values for gHSA-KP systems are only slightly lower than that obtained for HSA-KP. The affinity of PHB to the glycated HSA is stronger than to the non-glycated in the first class binding sites within subdomain IIA, in the vicinity of Trp-214. Ketoprofen bound to unmodified human serum albumin stronger than for glycated albumin and one class of binding sites is observed (Scatchard linear plots).

Phenylbutazone Oxidation via Cu,Zn-SOD Peroxidase Activity: An EPR Study.

We investigated the effect of Cu,Zn-superoxide dismutase (Cu,Zn-SOD)-peroxidase activity on the oxidation of the nonsteroidal anti-inflammatory drug phenylbutazone (PBZ). We utilized electron paramagnetic resonance (EPR) spectroscopy to detect free radical intermediates of PBZ, UV-vis spectrophotometry to monitor PBZ oxidation, oxygen analysis to determine the involvement of C-centered radicals, and LC/MS to determine the resulting metabolites. Using EPR spectroscopy and spin-trapping with 5,5-dimethyl-1-pyrroline-N-oxide (DMPO), we found that the spin adduct of CO3(•-) (DMPO/(•)OH) was attenuated with increasing PBZ concentrations. The resulting PBZ radical, which was assigned as a carbon-centered radical based on computer simulation of hyperfine splitting constants, was trapped by both DMPO and MNP spin traps. Similar to Cu,Zn-SOD-peroxidase activity, an identical PBZ carbon-centered radical was also detected with the presence of both myeloperoxidase (MPO/H2O2) and horseradish peroxidase (HRP/H2O2). Oxygen analysis revealed depletion in oxygen levels when PBZ was oxidized by SOD peroxidase-activity, further supporting carbon radical formation. In addition, UV-vis spectra showed that the λmax for PBZ (λ = 260 nm) declined in intensity and shifted to a new peak that was similar to the spectrum for 4-hydroxy-PBZ when oxidized by Cu,Zn-SOD-peroxidase activity. LC/MS evidence supported the formation of 4-hydroxy-PBZ when compared to that of a standard, and 4-hydroperoxy-PBZ was also detected in significant yield. These findings together indicate that the carbonate radical, a product of SOD peroxidase activity, appears to play a role in PBZ metabolism. Interestingly, these results are similar to findings from heme peroxidase enzymes, and the context of this metabolic pathway is discussed in terms of a mechanism for PBZ-induced toxicity.

Phenylbutazone and ketoprofen binding to serum albumin. Fluorescence study.

BACKGROUND: A combination of phenylbutazone (PBZ) and ketoprofen (KP) is popular in therapy of rheumatoid arthritis (RA) but could be unsafe due to the uncontrolled growth of toxicity. METHODS: Quenching fluorescence of serum albumin in the presence of the both drugs has been characterized by dynamic KQ [M(-1)], static V [M(-1)] quenching constants and also association constants Ka [M(-1)]. RESULTS: The quenching of tryptophanyl residues fluorescence by the KP and PBZ indicates the capability of these drugs to accept the energy from Trp-214 and Trp-135. Strong displacement of KP and PBZ bound to albumin cause by the binding of the second drug to SA close to Trp-214 (subdomain IIA) has been obtained. The displacement was also confirmed on the basis of quenching and association constants. CONCLUSIONS: The conclusion, that both PBZ and KP form a binding site in the same subdomains (IIA or/and IB), points to the necessity of using a monitoring therapy owning to the possible increase of the uncontrolled toxic effects.

Pharmacokinetics, pharmacodynamics, metabolism, toxicology and residues of phenylbutazone in humans and horses.

The presence of horse meat in food products destined for human consumption and labelled as beef has raised several concerns of public interest. This review deals solely with one aspect of these concerns; samples of equine tissue from horses destined for the human food chain have tested positive for the non-steroidal anti-inflammatory drug, phenylbutazone. The safety of some or all such foods for human consumers is a major concern, because it was shown many years ago that phenylbutazone therapy in humans can be associated with life threatening blood dyscrasias. As an initial basis for assessing the potential toxicity of foods containing phenylbutazone and its metabolites, this article reviews (1) the pharmacokinetic, pharmacodynamic, metabolic and toxicological profiles of phenylbutazone, with particular reference to horses and humans; (2) toxicity data in laboratory animals; (3) phenylbutazone residues in food producing species, and (4) as a preliminary assessment, the potential hazard associated with the consumption of horse meat containing phenylbutazone and its metabolites. Since phenylbutazone cannot be classified as a carcinogenic substance in humans, and noting that blood dyscrasias in humans are likely to be dose and treatment duration-dependent, the illegal and erratic presence of trace amount residues of phenylbutazone in horse meat is not a public health issue.

Binding of diuretic antihypertensive bendroflumethiazide to human serum albumin studied by ¹9F nuclear magnetic resonance method.

Simultaneous specific and nonspecific binding of bendroflumethiazide (BFZ) to human serum albumin (HSA) and concentration profile of BFZ in HSA buffer (pH 7.40) solution were investigated by ¹9F nuclear magnetic resonance (NMR) method. The ¹9F NMR spectrum of BFZ (200 µM) in a buffer solution showed a sharp signal of its CF3 group at 17.8 ppm from the reference trifluoroethanol. Addition of 0.60mM HSA to the sample solution caused the CF(3) signal splitting into three broadened peaks at 18.4 (A), 17.9 (B) and 17.4 ppm (C). By its chemical shift and spectral behavior, B was assigned to unbound BFZ. Competition experiments with Site I and II ligands lead to C being assigned to Site II bound BFZ. However, the peak intensity (areas) of A was not reduced by these ligands, suggesting that A arises from nonspecific binding. Using the peak intensities at several total concentrations of BFZ, Scatchard plot was performed. The plot for A provided a straight line parallel to the x-axis confirming nonspecific binding and that for C was consistent with specific binding. The binding constants for nonspecific and specific Site II binding were 1.02 and 1.00 × 104 (M⁻¹) (n=1.1), respectively. The presence of 0.10 M Cl⁻ in the sample solution affected the binding constant of Site II binding, but not that of nonspecific binding. The concentration profile of BFZ calculated using the binding constants revealed that nonspecific binding is more effective than Site II binding for the binding of BFZ to HSA. It was also confirmed that considerable amounts of BFZ liberated from Site II by the Site II ligands or Cl⁻ ions bind again nonspecifically.

Physicochemical characterisation, drug polymer dissolution and in vitro evaluation of phenacetin and phenylbutazone solid dispersions with polyethylene glycol 8000.

Poor water solubility leads to low dissolution rate and consequently, it can limit bioavailability. Solid dispersions, where the drug is dispersed into an inert, hydrophilic polymer matrix can enhance drug dissolution. Solid dispersions were prepared using phenacetin and phenylbutazone as model drugs with polyethylene glycol (PEG) 8000 (carrier), by melt fusion method. Phenacetin and phenylbutazone displayed an increase in the dissolution rate when formulated as solid dispersions as compared with their physical mixture and drug alone counterparts. Characterisation of the solid dispersions was performed using differential scanning calorimetry (DSC), Fourier transform infrared spectroscopy (FTIR) and scanning electron microscopy (SEM). DSC studies revealed that drugs were present in the amorphous form within the solid dispersions. FTIR spectra for the solid dispersions of drugs suggested that there was a lack of interaction between PEG 8000 and the drug. However, the physical mixture of phenacetin with PEG 8000 indicated the formation of hydrogen bond between phenacetin and the carrier. Permeability of phenacetin and phenylbutazone was higher for solid dispersions as compared with that of drug alone across Caco-2 cell monolayers. Permeability studies have shown that both phenacetin and phenylbutazone, and their solid dispersions can be categorised as well-absorbed compounds.

Interaction of phenylbutazone and colchicine in binding to serum albumin in rheumatoid therapy: 1H NMR study.

The monitoring of drug concentration in blood serum is necessary in multi-drug therapy. Mechanism of drug binding with serum albumin (SA) is one of the most important factors which determine drug concentration and its transport to the destination tissues. In rheumatoid diseases drugs which can induce various adverse effects are commonly used in combination therapy. Such proceeding may result in the enhancement of those side effects due to drug interaction. Interaction of phenylbutazone and colchicine in binding to serum albumin and competition between them in gout has been studied by proton nuclear magnetic resonance ((1)H NMR) technique. The aim of the study was to determine the low affinity binding sites, the strength and kind of interaction between serum albumin and drugs used in combination therapy. The study of competition between phenylbutazone and colchicine in binding to serum albumin points to the change of their affinity to serum albumin in the ternary systems. This should be taken into account in multi-drug therapy. This work is a subsequent part of the spectroscopic study on Phe-COL-SA interactions [A. Sulkowska, et al., J. Mol. Struct. 881 (2008) 97-106].

Enhancement of transdermal delivery of phenylbutazone from liposomal gel formulations through deer skin.

Phenylbutazone (PBZ) is a nonsteroidal anti-inflammatory drug used in the treatment of chronic pain and arthritis. Topical and transdermal administration of PBZ would be beneficial in large animals in terms of minimizing gastro-intestinal ulcerations and other side effects, easy administration to legs and joints and minimizing the dose to reduce systemic toxicity of the drug. A topical liposomal preparation with different concentrations of a mono-substituted alkyl amide (MSA) and PBZ was formulated. The formulations were evaluated by in vitro skin-permeation kinetics through deer skin using Franz diffusion cells. By increasing drug loading from 1% to 5% w/w, the steady-state flux (microg/cm(2)/h) of PBZ was increased twofold (P < 0.001). Similarly, by increasing the MSA concentration from 0% to 4%, the steady-state flux (microg/cm(2)/h) of PBZ was increased twofold (P < 0.001). Overall, by increasing the drug load and the use of an appropriate amount of the penetration enhancer, the steady-state flux of PBZ through skin was increased fourfold (P < 0.001). MSA at both 2% and 4% w/w concentrations significantly increased the skin levels of PBZ as compared with control (P < 0.05). In conclusion, MSA served as an effective skin-penetration enhancer in the liposomal gel of PBZ for deer.

Rapid characterization of drug-drug interaction in plasma protein binding using a surface plasmon resonance biosensor.

High-throughput characterization of drug-drug interactions in plasma protein binding was demonstrated by using a surface plasmon resonance (SPR) biosensor. The method used in this study enabled the discrimination between the two modes of binding inhibition, direct competition and negative allosteric effect, which was difficult in conventional SPR approaches. Two theoretical equations were used representing SPR binding response for directly competitive binding or for independent binding. The experimental binding data for human serum albumin was processed by non-linear least squared regression of the equations. By this approach, drug-drug interactions were classified into three modes, direct competition, independent binding, and allosteric interaction, which were almost consistent with previous reports. In addition, dissociation constants were also estimated roughly for direct competition and for independent binding. The analytical throughput was almost as high as in the previous reports; three minutes per injection. This method is a powerful tool for the characterization of drug-drug interaction at an early stage of new drug development.