General aspects of inhibition
Enzymes and tissue/cell receptors share similar features. A receptor binds a molecule that then acts like a switch to trigger a cascade of molecules to instruct the cell to perform a function. The molecule must fit the receptor precisely and then trigger the cascade, like a key, which first enters a lock, then is successfully turned to open it. A key that fits and enters the lock, but does not turn it, not only fails to open the door but also prevents the correct key from being fitted. The lock is essentially ‘inhibited’.
Although they are highly specialized, CYPs are enzymes like any other in the body and they are inhibited according to the same general principles as other enzymes. How tightly a chemical interacts with a CYP isoform is based on how powerful is the mutual attraction (affinity) between the chemical and the various areas of the active site of the enzyme.
In the case of CYPs and any given enzyme, affinity must be strong enough to ensure the substrate is bound for sufficient time to process it to a product. The quicker this process occurs, the faster the ‘turnover’ of the enzyme and the more efficient it is. It is useful to try to visualize a CYP isoform, or any other human enzyme for that matter, as a three dimensional machine tool, perhaps like a spot welding machine. The enzyme cycles hundreds of times a second. If any single aspect of substrate binding or processing (oxidation or reduction), followed by product release is prevented, the sequential nature of these events means that the enzyme stops functioning. Another analogy might be an automatic paper stapler in a photocopier. Whatever analogy you might use, it is useful to try to visualize enzymes as dynamic micro machines. Broadly, inhibitors of CYPs may frustrate the enzymes’ operating processes in two main ways, with varying impact on drug clearance and the individual enzyme ‘health’ and survival. At high concentrations, many inhibitors might block several CYP subfamilies, but at lower concentrations, they show more selectivity and their potency in blocking individual isoforms can be measured. Inhibition itself can occur through four main processes: competitive, non-competitive, uncompetitive and mechanism-based. Which type of inhibition occurs with various drugs can depend on many factors, such as drug concentration and the characteristics of a particular CYP isoform. Many drugs can act as competitive inhibitors with one CYP and non-competitive with others. Studies with inhibitors of drug metabolism are carried out in vitro with human CYPs, either in human liver or in expressed enzyme systems.These studies do not always reflect what will happen when the drugs are used in patients, but are a reasonable starting point to predict whether a new drug might interfere with the metabolism of another.
Competitive inhibition
This is the simplest form of inhibition, where the substrate (drug) and the inhibitor are very similar in structure and have similar affinities for the same place, i.e. the CYP active site (Figure 5.1). A CYP substrate is normally processed to a different molecule, that is,a metabolite, which then has a much reduced affinity for an active site and is more water-soluble, so it diffuses elsewhere. A competitive inhibitor of a CYP isoform is usually not a substrate and acts like a similar key to the correct key for a doorlock; it may enter and leave the lock freely but does not operate it. As it is not processed into a product, it does not leave the vicinity of the CYP and binds and detaches continually. The CYP might be unable to metabolize the inhibitor, due to particular features of the molecule that might prevent oxidation, but promote binding to the active site. This form of inhibition is common in CYPs and is governed by the law of mass action, which states that the rate of a reaction (in this case enzyme binding) is governed by the concentration of the participants. So for CYP metabolism, whichever agent, drug or inhibitor, is in the greatest concentration, then this will occupy the active site. At low inhibitor concentrations more drug can be added to overcome the inhibitory effects. However, as drug levels must be increased to overcome the inhibitor, this effectively means that the drug’s affinity falls for the site (Km increases) in the presence of the inhibitor. Enzymes are often subject to this process of competitive inhibition because it is usually part of the endogenous feedback control mechanism on product formation. This generally involves enzymes that use cellular energy, or are at the junction of several biosynthetic pathways. When high levels of product are formed, these inhibit the substrate, so limiting the enzyme’s ‘lurnover’, i.e. when the desired product level is reached. This is rather like a thermostat in a heating system, which automatically maintains a preset temperature irrespective of outside temperatures. This is seen in the regulation of vital endogenous molecules like NADPH and glutathione (GSH) and the process avoids unnecessary use of cellular energy. Although the enzyme is temporarily disabled, it is undamaged and has not cycled or used any reducing power. Mathematically, if a Lineweaver-Burk double reciprocal plot is made of competitive inhibition, the Km (inverse of the affinity) changes, but the Vmax does not; in other words, the enzyme will still run at a maximum rate if enough substrate is used, but affinity falls off.
Figure 5.1 Main types of enzyme inhibition that apply to CYP isoforms.
A new drug might be evaluated as a possible inhibitor of a given CYP isoform; if the inhibition of the known CYP substrate yields a Lineweaver-Burk plot as described above, then the new drug is a competitive inhibitor of that CYP and it is likely that the inhibitor is binding the CYP at its active site. There are several examples of competitive inhibitors of CYP isoforms. Indeed, if two drugs of similar affinities are cleared by the same isoform, then competitive inhibition can occur. The major clinically relevant group of competitive inhibitors includes the azole antifungal agents.
Azoles
It is not surprising that these agents are potent human P450 inhibitors, as a great deal of money, time and effort was put into designing them to inhibit fungal CYPs. They prevent the fungal synthesis of ergosterol, by blocking lanosterol alpha-C14-demethylase, so causing the substrates (14-alpha-methylsterols) to accumulate and this disrupts fungal membranes. Interestingly, this is relatively specific; ketoconazole was initially the most commonly used azole agent and this is a potent competitive inhibitor of CYP3A4, as well as a number of other sex steroid-handling CYPs. This meant that the drug was quite toxic,as it caused a significant fall in testosterone levels in blood, which could lead to feminization of males. This could be seen as the appearance of breasts (gynaecomastia), loss of spermatozoa production and impotence. The female menstrual cycle was also disrupted. These effects, coupled with other toxicity, such as GI tract irritation, nausea, vomiting and occasional severe liver toxicity, propelled the continuing development of these agents to less toxic azoles, which would be more potent therapeutically, but with less human CYP impact. These appeared in the 1990s, in the form of itraconazole and fluconazole, which were followed by the third-generation triazoles, such as voriconazole and posaconazole, which were much more effective than ketoconazole as antifungals in vitro, although they do still inhibit CYPs.
Clinically, there is plenty of evidence that azoles inhibit the metabolism of other CYP3A substrates. Fluconazole has caused a patient on simvastatin to develop rhabodomyolysis, which is an uncommon but potentially fatal condition associated with statins. This effect only occurs at high systemic levels of statin and was due to the inhibition of clearance by the azole. Peak plasma concentrations of the CYP3A4 substrate felodipine were increased eightfold and the area under the curve (AUC, or the amount of drug in the plasma) six-fold by the presence of itraconazole. Fluconazole has been shown to increase the half-life of omeprazole by threefold and it’s AUC by a similar value. Other CYP3A4 substrates, such as midazolam, terfenadine and lovastatin, show similar effects with this azole. Clearly, the impact of the inhibition on the pharmacological effects of these drugs is very strong, with significant potentiation of their particular effects. It is interesting that although the inhibitory action of these drugs is simple and reversible, the clinical effect of this process on other drug effects can potentially be extremely serious. However, in a matter of hours after the withdrawal of the azole, the inhibiting effect is lost and substrate clearance resumes. There appears to be no way that the liver CYP nuclear ‘management system’ which is seen operating so successfully with enzyme inducers, can overcome the effects of a drug such as keto-conazole when the agent is taken for a long period of time, although anti-fungals are usually taken for either single topical doses, or relatively short courses and are stopped after the infection is eradicated. The serious disruption of steroid metabolism (gynaecomastia again) testifies to the potential problem posed by longer exposure to these drugs. The inhibition appears to be stable for as long as the drug is administered. This has led to attempts to use inhibitors such as ketoconazole and cimetidine to deliberately block the clearance of certain drugs, of which more later. Voriconazole is cleared by CYP2C19 and CYP3A4 and thus is subject to effects linked to polymorphisms of CYP2C19 and ironically its clearance is significantly reduced by CYP3A4 inhibitors, such as ritonavir.
Azoles and immunosuppressants
Although fluconazole is effective against the common fungal pathogen Candida albicans, it is no use against the loathsome mould species Aspergillus and Cryptococcus which often infect immunosuppressed transplant patients. The third generation azoles were mainly developed to destroy these moulds, so fluconazole therapy is sometimes switched in immunosuppressed patients to voriconazole where fungal breakthrough occurs. Unfortunately, in vitro human liver microsome studies have revealed voriconazole to be capable of blocking CYP2B6 and CYP2C9, CYP2C19 and CYP3A at single figure micromolar concentrations, whilst its effects on CYP2C8, CYP2A6, CYP1A2 and CYP2D6 are relatively weak. This unfortunately translates to the clinical situation, where voriconazole is a potent inhibitor of the metabolism of the newer anti-rejection drugs (tacrolimus, everolimus and sirolimus). Indeed, when voriconazole replaces fluconazole, it has been shown to be a more potent inhibitor of tacrolimus clearance than its predecessor. It is already recommended that the tacrolimus dose should be reduced by a third when voriconazole is used, although a further reduction of around 20 per cent may be necessary when existing fluconazole therapy is replaced by voriconazole. This situation can be complex, depending on whether the azoles are used orally or intravenously. These effects have been ascribed to the high levels of CYP3A4 found in the gut (more than 70 per cent of total CYPs) compared with the generally lower CYP3A hepatic levels (~30 per cent). All azoles, including others such as miconazole and clotrimazole, are generally competitive inhibitors, due to their lone pair of electrons on the azole nitrogen, which temporarily binds to CYP haem groups. This is mostly borne out by in vitro studies with voriconazole, as its most potent inhibitory effects on CYP2B6 and CYP3A are competitive, although there is an element of non-competitive inhibition at slightly higher concentrations. As well as voriconazole’s effects, its sister agent posaconazole is also capable of increasing sirolimus peak concentrations and AUC by nearly nine fold, which is likely to cause severe toxicity to a transplanted kidney.
The use of azoles with immunosuppressant drugs is particularly problematic and it has been recommended that a highly individualized patient approach should be taken due to the complexity of the situation and the risks involved. This should involve close therapeutic monitoring to minimize side effects so retaining good patient morale and compliance. In addition, adequate azole must be present to eliminate the fungal infection quickly and of course it is vital that the organ or graft is not endangered during the treatment of the infection. In contrast, many cancer patients develop fungal infections and the effects of azole antifungals on the clearance and activation of antineoplastic agents does not appear to be well documented. Given that several anticancer agents rely on CYPs for either activation or clearance, it is reasonable to believe that azoles may interfere with these processes.
Non-competitive inhibition
Non-competitive inhibition does not involve the inhibitor and substrate competing for the same active site (Figure 5.1). In non-competitive inhibition, there is another site involved, known as the allosteric site, which is distant from the active site. Once a ligand binds this allosteric site, the conformation of the active site is automatically changed and it becomes less likely to bind the substrate and product formation tails off. This process of allosteric binding is another example of the endogenous control of product formation, perhaps by another product/substrate from a related or similar pathway. The Lineweaver-Burk plot will show a fall-off in Vmax (enzyme cannot run at maximal rate) but Km does not change, that is, the affinity of the substrate for the active site is unchanged.
It has been demonstrated experimentally that many drugs are non-competitive inhibitors of CYP isoforms. This means that the inhibitor is not binding at the active site and must exert some allosteric effect elsewhere. As knowledge of the active sites of many CYPs is still incomplete, we are still not fully aware as to exactly where these allosteric sites are and where they figure in the control of CYPs. It was discussed that CYP3A4 had more than one site available for binding and that various substrates could influence the binding of other substrates, probably connected with hormone metabolism. This potentially provides an hour-by-hour modulation of CYP activity, which is of course necessary during steroidal control of reproductive processes. It is likely that noncompetitive inhibition is a result of drugs fitting these allosteric sites within most CYP isoforms and influencing binding of substrates to the main catalytic site. There are several examples of non-competitive inhibitors of CYPs. St John’s Wort extract (hyperforin) is an inhibitor of CYP2D6 in vitro, although this does not translate to the clinical situation. Omeprazole and lansoprazole are non-competitive CYP3A4 inhibitors in vitro.
Uncompetitive inhibition
This is an unusual form of inhibition, where the inhibitor binds only to the enzyme/substrate complex (Figure 5.1). This has the effect of stimulating enzyme/substrate complex formation so increasing affinity (fall in Km), although the enzyme/substrate/inhibitor complex is non-functional, so the Vmax falls. This appears to be a relatively rare form of inhibition of human CYPs by therapeutic drugs, although the NSAID meloxicam is capable of uncompetitively inhibiting quinidine in vitro, it is not likely to be a significant clinical interaction. Some dietary agents contain inhibitors such as the flavonoid tangeretin, which is found in citrous fruits. Tangeritin is an uncompetitive inhibitor of CYP3A4 in human liver microsomes.
Mechanism-based inhibitors
This type of inhibition is outside the normal classification as outlined with competitive, non-competitive and uncompetitive inhibitions. Mechanism-based inhibition generally involves the same initial steps as a competitive inhibitor, but then the CYP catalytic cycle proceeds, reducing power is consumed and a metabolite is formed, which then occupies the P450 active site for a far longer period than the usual substrate would (Figure 5.2). To extend the ‘machine gun’ analogy of a CYP, mechanism-based inhibition is like failure to extract a spent cartridge. Mechanism-based inhibitors could occupy an allosteric site in a CYP and thus act as non-competitive inhibitors; macrolide antibiotics are sometimes classed as non-competitive inhibitors even though they are mechanism-based. The nearest mechanical analogy to a mechanism-based inhibitor would be the incorrect key turning fully in the lock and not opening, followed by difficult extraction of the key, or even the key breaking off in the lock. This form of inhibition can range from delayed product release, all the way to a violently reactive species-mediated covalent binding of a metabolite, which effectively destroys the active site and terminates the enzyme’s activity. There are degrees of mechanism-based inhibition and moderately potent inhibitors such as the macrolides (like erythromycin and clarithomycin) are eventually removed from the CYP active site, but do not usually damage the enzyme. However, highly potent mechanism-based inhibitors such as the contents of grapefruit juice damage the enzyme to a degree that it is non-functional. This latter process is often termed ‘suicide’ inhibition. Clinically, a competitive inhibitor should wear off after just one or two half-lives, i.e. a few hours to a day or so, depending on a number of factors (inhibitor and substrate dosage, etc). The most extreme form of mechanism-based inhibition, such as grapefruit juice, norfluoxetine or MDMA-mediated ‘suicide’ inhibition, destroys the enzyme from one dose of inhibitor and this takes several days to resolve.
Figure 5.2 Scheme of normal substrate CYP binding (left) and mechanism-based inhibitor on the right, which results in the irreversible binding of product and inactivation of the CYP isoform
The effects of mechanism-based inhibition can be shown very clearly in vitro, where the potency of the inhibition is much greater when the CYP enzymes are incubated with NADPH and the compound prior to the addition of the usual substrate. This enables the enzymes to use the reducing power to run the catalytic cycle, which forms the reactive metabolite, which starts disabling the enzyme. The longer this process goes on, the more enzyme is disabled, so the inhibition becomes more potent over time. Vmax falls and affinity decreases; then obviously the inhibition cannot be overcome by more substrate, as the law of mass action cannot apply because the inhibitor will be already covalently bound to key areas of the CYP active site. If the substrate is present in reasonably high concentrations prior to the appearance of the inhibitor, the substrate can protect the enzyme, although if the inhibitor continues to be present in adequate concentration, this protective effect will eventually be lost. This idea was exploited with neostigmine to protect military personnel against nerve agents. In vivo, mechanism-dependent inhibition lasts for days as previously mentioned, although the inhibition is clinically reversible, but as far as the individual enzyme is concerned, irreversible. This is because the clinical effect is consistent with the time taken for more P450 enzyme to be resynthesized to replace the inactivated enzyme. Obviously this will only be clinically reversible if the inhibitor was only dosed once, or over a short period. Mechanism-based inhibition is often summarized as follows:
• The inhibition becomes stronger over time.
• Inhibition does not progress without co-factors (NADPH).
• Presence of substrate slows the rate of inhibition by protecting the CYP.
• After inhibition, intact enzyme cannot be detected by analytical techniques (irreversibly inactivated).
Although there are examples of suicide inhibitors with other CYPs (such as ticlopidine with CYP2C19 and CYP2B6) CYP3A4 tends to be particularly susceptible to mechanism-based inhibition and although there are many different structures that can cause this effect, agents possessing a tertiary amine, acetylene or furan group are more likely to inhibit in this way. There are many clinically important non- competitive and mechanism- based ‘suicide’ inhibitors, which vary in the intensity of their inhibition. These include the mac-rolides (erythromycin, clarithromycin, oleandomycin), HIV protease inhibitors (ritonavir, indinavir, saquinavir, nelfinavir), the SSRIs (e.g. fluoxetine), anti cancer agents (tamoxifen and irinotecan) some antihypertensives (diltiazem, verapamil) and finally grapefruit juices. Unfortunately, all these compounds are metabolized by, and eventually inhibit, our major CYP, CYP3A4, whilst the illicit amphetamine derivative MDMA irreversibly blocks CYP2D6.There is also evidence that the partially withdrawn second generation antidepressant nefazodone (section 5.3.8- is probably a mechanism-based CYP3A inhibitor, as its effect increases over time and it shows non-linear kinetics consistent with autoinhibition of CYP3A. If a prospective therapeutic agent is found early in drug development to be a potent mechanism-based CYP3A4 inhibitor in vitro, its days are generally numbered, unless it is virtually the only prospect available for a life-threatening disorder.
