Analytical approaches to determine cytochrome P450 inhibitory potential of new chemical entities in drug discovery

https://doi.org/10.1016/j.jchromb.2006.12.024Get rights and content

Abstract

The use of a cassette incubation of probe substrates with human liver microsomes (HLM) – also known as the ‘cocktail’ approach – is becoming a widely accepted approach to determine the interaction of new chemical entities (NCEs) with cytochrome P450 enzymes (CYP450) in early drug discovery. This article describes two LC–MS/MS-based analytical methods used at the high-throughput (HT) stage and late discovery (LD) stage for analysis of ‘cocktail’ incubates to analyze the probe metabolites 1′-hydroxymidazolam (CYP3A4), 4′-hydroxydiclofenac (CYP2C9), dextrorphan (CYP2D6), 1′-hydroxytacrine (CYP1A2) and 4′-hydroxymephenytoin (CYP2C19). The analytical methods are advantageous over currently reported methods due to their sensitivity, shorter analyses times (<2 min/sample for the HT method and 4 min/sample for the LD method) and their ability to monitor a unique set of clinically relevant probe metabolites from a biological incubate containing low microsomal protein (0.1 mg/mL). The analytical methods employ the same mobile phase, acetonitrile and 0.1% formic acid, under similar LC–MS/MS conditions. In the HT method, the chromatographic method consists of a short robust step-gradient where the probe metabolites are simultaneously and quickly eluted to enhance throughput. The probe metabolites are chromatographically resolved in the LD stage by utilizing a true linear gradient to obtain optimal peak separation. The IC50 data generated by both analytical methods using single incubations versus cocktail incubations for various test compounds are in good agreement (correlation coefficient (r2)  0.98). The scientist conducting the analysis is provided with a choice of method selection depending on the stage of the test compound and on whether throughput or minimizing interference from other co-eluting metabolites is the most important criterion.

Introduction

The pharmaceutical industry continues to face an overall high attrition rate, primarily due to lack of efficacy and safety. Hence, there is a demand on pharmaceutical companies to meet their business objectives and to think about ways to achieve efficiencies [1]. Currently, a vast number of new chemical entities (NCEs) are being generated by combinatorial chemistry and tested using in silico and in vitro technologies. These tools are widely used in the industry in various disciplines to predict in vivo behavior in an effort to minimize resource consumption and enhance throughput. Drug–drug interaction (DDI) is one such area where in vitro results are used to predict a compound's inhibitory potential towards the cytochrome P450 enzymes, which serve as the major clearance pathway for hepatically cleared compounds. There are serious ramifications to the marketability when a compound's CYP inhibitory potential is identified during clinical development [2]; hence it is essential to utilize in vitro technologies in early discovery for DDI risk assessment.

Although more than 40 human cytochrome P450 enzymes have been identified, 5 enzymes are responsible for more than 87% of human drug metabolism: CYP3A4, CYP2D6, CYP1A2, CYP2C9 and CYP2C19 [3]. The cocktail DDI approach is an in vitro tool that is becoming widely accepted in the early drug discovery stages among researchers. Early on Breimer and Schellens demonstrated the feasibility and utility of the cocktail strategy [4]. It deviates from the traditional singlet incubation assay by pooling multiple CYP450 probe substrates into a single human microsomal incubation under the same biological conditions [5], [6], [7], [8], [9]. The cocktail assay has allowed researchers rapid and simple assessment of the potential inhibitory effect of test compounds on CYP450s.

Throughout drug discovery, analytical methodology is designed to provide the necessary speed and quality predicated on demand. In early discovery several hundreds to thousands of compounds are screened per week. To assess CYP inhibitory potentials, fluorescent techniques are typically used at this stage for high-throughput. For example, CYP3A4 interaction potential is assessed using 7-benzyloxy-4-trifluoromethylcoumarin (BFC)/vivid red as substrates [10]. Fluorescent assays often require the use of expressed CYP450/recombinant enzymes instead of human liver microsomes (HLM) due to poor specificity of the substrates [11]. Typically, in vitro DDI data (% inhibition from fluorescent methods) at this stage is used in conjunction with other in vitro parameters such as potency, selectivity and safety to bin and effectively screen high-risk compounds.

After this initial stage, the number of compounds is often in the few hundreds or less per week. During this phase, CYP inhibitory potential data are used to guide structural modification of compounds to mitigate risk. CYP inhibitory potentials can usually be determined using non-fluorescent techniques or clinically relevant probes in a human liver microsomal incubation by LC–MS/MS. Either % inhibition or IC50 can be determined using extrapolation from single concentration data in this phase. Yet, determination of IC50 involving incubations at various concentrations of test compound would be resource consuming due to the large number (several hundreds) of compounds in this stage gate. In the late stages of drug discovery where the number of compounds is fewer, definitive assays are conducted to determine IC50. The test compound is incubated at different concentrations (determined based on projected free efficacious concentrations) with the probe substrates in human liver microsomes. The IC50 value in conjunction with the compounds free efficacious concentration is used for risk assessment in the clinic. At this stage, analyses of the microsomal incubations require robust, reliable analytical techniques to separate the probe metabolite from the parent compound and other metabolites at higher concentrations of the test compound to prevent ion suppression and to ensure quality. Typically, this data can be used up to IND submissions, to predict clinical risk of DDIs, to increase patient safety and to design early phase DDI studies in healthy volunteers. Additionally, throughput is not a concern, considering only a small percentage of compounds progress to this stage of testing.

Traditionally, prior to the development of cocktail biology, singlet enzyme incubation in conjunction with a single LC–MS/MS method has been used for determination of CYP inhibitory potential [12]. Upon the development of a substrate cocktail biological incubation assay there was a need for a cocktail analytical assay with appropriate throughput and sensitivity to determine a test compound's CYP450 inhibitory potential. Several LC–MS/MS-based cocktail analytical assays have been reported in the literature [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28]. However, many of these methods suffer from limitations such as use of recombinant CYPs, clinically irrelevant probe substrates, intensive sample preparation, use of sample preparation HPLC columns, longer run times and higher protein (HLM) content in biological incubations that provides higher substrate turnover; however, could result in non-specific binding. Reported herein are two LC–MS/MS-based analytical methods that are stage gate aligned (high-throughput (HT)/late discovery (LD)) to enable appropriate throughput, have shorter run times, possess increased sensitivity and enable analysis of metabolites for a unique set of clinically relevant probe substrates (Table 1). The appropriate choice of analytical conditions and hindrances that were faced during method development are discussed.

Section snippets

Materials

Diclofenac sodium salt, sulfaphenazole, ketoconazole, testosterone, 6β-hydroxytestosterone, phenacetin, acetamidophenol, furafylline, quinidine, dextrorphan, ticlodipine, fluconazole, fluoxetine, benidipine HCl, potassium phosphate buffer (pH 7.4) and the reduced form of NADPH were purchased from Sigma–Aldrich (St. Louis, MO). Human liver microsomes (HLM), midazolam hydrochloride, 1′-hydroxymidazolam, S-mephenytoin, 4′-hydroxydiclofenac and 4′-hydroxymephenytoin were obtained from Gentest Corp.

Results and discussions

The methodology reported here consists of a cocktail biological incubation of five probe substrates (most of which are recommended by Pharmaceutical Research and Manufactures of America (PhRMA)) for the five major CYPs in HLM with the NCE and monitors for their respective probe metabolites using LC–MS/MS as listed in Table 1 [29]. The development of cocktail biology will be reported separately. Fig. 1 depicts the appropriate technologies used to determine DDI potential of NCEs at the various

Conclusions

We have presented analytical approaches that can be combined with an in vitro cocktail incubation to assess in vitro CYP inhibitory potential for NCEs. The two LC–MS/MS-based methods are robust and provide reliable data fit for the stage (HT and LD) at which they are used. These methods provide exemplary sensitivity for the 4′-hydroxymephenytoin probe metabolite for the 2C19 isozyme over other existing methods. Both methods employ the same combination of acetonitrile and water (0.1% formic

References (32)

  • F.P. Guengerich

    Adv. Pharmacol.

    (1997)
  • D.D. Breimer et al.

    Trends Pharmacol. Sci.

    (1990)
  • C.L. Crespi et al.

    Methods Enzymol.

    (2002)
  • H.Z. Bu et al.

    J. Chromatogr. B

    (2001)
  • H.Z. Bu et al.

    J. Pharm. Biomed. Anal.

    (2001)
  • E.S.M. Lutz et al.

    J. Chromatogr. B

    (2002)
  • T. Zhang et al.

    J. Chromatogr. B

    (2002)
  • S.A. Testino et al.

    J. Pharm. Biomed. Anal.

    (2003)
  • M. Turpeinen et al.

    Eur. J. Pharm. Sci.

    (2005)
  • I. Kola et al.

    Nat. Rev. Drug Discov.

    (2004)
  • R.F. Frye et al.

    Clin. Pharmacol. Ther.

    (1997)
  • R.J. Scott et al.

    Rapid Commun. Mass Spectrom.

    (1999)
  • D.S. Streetman et al.

    Clin. Pharmacol. Ther.

    (2000)
  • B. Zhu et al.

    Clin. Pharmacol. Ther.

    (2001)
  • E.A. Dierks et al.

    Drug Metab. Dispos.

    (2001)
  • Cited by (56)

    • Breaking the sensitivity limitations of cytochrome P450 oxidation product: Dansyl chloride derivatisation of 4-OH mephenytoin, a CYP2C19 metabolite and its application to in vitro CYP inhibition assay

      2015, Journal of Chromatography B: Analytical Technologies in the Biomedical and Life Sciences
      Citation Excerpt :

      S-mephenytoin has very low turnover when compared to other substrates [6–10]. Various strategies such as use of alternate probe substrates [7], lower DMSO percentages [8], higher sensitive instruments [8], prolonged incubation time [9], higher protein concentration [9] were investigated by the researchers to enhance the sensitivity of 4-OH mephenytoin, but the low turnover of S-mephenytoin poses analytical difficulties. Alternate probe substrate might not be an ideal choice as the magnitude of drug–drug interactions may vary depending on the choice of probe substrate [6].

    • A cocktail approach for assessing the in vitro activity of human cytochrome P450s: An overview of current methodologies

      2014, Journal of Pharmaceutical and Biomedical Analysis
      Citation Excerpt :

      Additionally, it should be noted that OMZ is unstable in aqueous acidic conditions relative to (S)-MP, which requires appropriate sample preparation [137]. Conversely, when OMZ was incubated in suitable conditions instead of (S)-MP, 5-hydroxyomeprazole showed better sensitivity in ESI positive mode and less sensitivity to the matrix effect compared with 4′-hydroxymephenytoin [57,60]. The CYP2C19-mediated O-demethylation of OMZ showed similar characteristics compared to the 5-hydroxylation, but it was still considered a minor pathway [138] that was not officially recommended.

    View all citing articles on Scopus
    1

    Both authors contributed equally to this work.

    View full text