Cytochrome P450 pharmacogenomics: CYP3A4/2D6 polymorphism, drug-drug interactions, and precision dosing

Intuition Beginner

When you swallow a pill, the drug enters your blood and travels to the liver. The liver's job is to clear it from the body. It uses a family of enzymes called the cytochrome P450 system to chemically modify drugs so the kidneys can excrete them. Different people inherit different versions of these enzymes. Some metabolise a drug very fast — it leaves the body before it can work. Others metabolise it very slowly — it accumulates, and side effects follow. The same pill at the same dose can be safe for one person, useless for a second, and dangerous to a third. That is the central problem of pharmacogenomics.

Drug-drug interactions are the other half of the story. Some drugs block these enzymes — they occupy the enzyme so it cannot process other drugs. Grapefruit juice, for example, strongly blocks CYP3A4, the workhorse liver enzyme. A person taking a statin with grapefruit juice can experience a 5-fold rise in statin blood levels, which damages muscle. Other drugs speed the enzymes up by triggering the liver to make more of them. St. John's wort does this; it has caused organ-transplant rejection by clearing the immunosuppressant cyclosporine out of the body too fast. These effects are predictable and tested for during drug development.

Why does this concept exist? Because adverse drug reactions are among the leading causes of in-hospital death, and a substantial fraction of them could be prevented with a single genetic test before prescribing. The US Food and Drug Administration now includes pharmacogenomic information on more than 100 drug labels. Precision dosing — matching the drug and the dose to the patient's enzyme genetics — is the operational answer.

Visual Beginner

The picture shows the P450 catalytic cycle, in which a substrate molecule binds the enzyme, oxygen is activated with the help of electrons from NADPH-cytochrome P450 reductase, and a high-energy iron-oxygen intermediate called Compound I abstracts a hydrogen atom from the substrate and reattaches an oxygen in its place. The product is a more water-soluble molecule that the kidney can excrete.

The figure also shows the seven major human drug-metabolising CYPs and the percentage of clinically used drugs each one handles. CYP3A4 (around 30 percent) and CYP2D6 (around 25 percent) together account for roughly 55 percent of all clinically used drugs that the liver processes.

Worked example Beginner

Consider the case of a 13-month-old Toronto toddler who died in 2007 after a routine tonsillectomy. Codeine is a weak painkiller on its own. The liver's CYP2D6 enzyme converts about 5 to 10 percent of each codeine dose into morphine, and that converted morphine provides the pain relief. A small fraction of the population carries extra copies of the CYP2D6 gene. They are called "ultra-rapid metabolizers" because their livers convert codeine into morphine much faster than average.

Step 1: The toddler inherited one ultra-rapid-metabolizer CYP2D6 allele from his mother, who was also an ultra-rapid metabolizer.

Step 2: After tonsillectomy he was prescribed a normal child dose of codeine syrup. His ultra-rapid-metabolizer liver converted far more of that dose into morphine than a typical child's liver would have.

Step 3: Morphine accumulated in his blood and stopped his breathing during the night. He died approximately 36 hours after surgery.

What this tells us: a drug that is safe at a standard dose for most people can be lethal for someone whose CYP2D6 genetics differ from the average. The FDA added a Boxed Warning to codeine in 2013 and contraindicated codeine for children under 12 and for all post-tonsillectomy pain in adolescents.

Check your understanding Beginner

Formal definition Intermediate+

Definition (cytochrome P450). Cytochrome P450 enzymes (CYPs) are a superfamily of heme-thiolate monooxygenases — defined by a cysteine-ligated heme prosthetic group — that catalyse the reaction

$$\mathrm{RH}+{\mathrm{O}}_2+\mathrm{NADPH}+{\mathrm{H}}^{+}\to \mathrm{ROH}+{\mathrm{H}}_2\mathrm{O}+\mathrm{NAD}{\mathrm{P}}^{+},$$

where $\mathrm{RH}$ is a lipophilic substrate and $\mathrm{ROH}$ is its hydroxylated, more-water-soluble product. Humans express 57 functional CYP genes and 58 pseudogenes, organised by sequence identity: families (sharing more than 40 percent identity, e.g., CYP1 through CYP51) and subfamilies (sharing more than 55 percent identity, denoted by a letter, e.g., CYP3A). Families CYP1, CYP2, and CYP3 are the principal drug-metabolising enzymes; the remainder process endogenous substrates such as sterols, bile acids, fatty acids, and eicosanoids.

Definition (catalytic cycle). A substrate $\mathrm{S}$ enters the active site above the heme iron, displacing a bound water molecule and shifting the heme from low-spin to high-spin, which raises its reduction potential. NADPH-cytochrome P450 reductase (CPR) donates the first electron, reducing ${\mathrm{Fe}}^{3+}$ to ${\mathrm{Fe}}^{2+}$; molecular oxygen then binds to ferrous iron, forming the ferrous-dioxy complex ${\mathrm{Fe}}^{2+}\text{-}{\mathrm{O}}_2$ (the oxy-complex). A second electron, donated either by CPR or by cytochrome $b_5$, reduces the oxy-complex, which rearranges and protonates to a ferric hydroperoxo species ${\mathrm{Fe}}^{3+}\text{-}\mathrm{OOH}$ (Compound 0). The O–O bond cleaves: one oxygen leaves as water, the other forms a ferryl-oxo porphyrin radical cation ${\mathrm{Fe}}^{4+}\text{=}\mathrm{O}{\mathrm{Por}}^{\bullet +}$ called Compound I. Compound I abstracts a hydrogen atom from the substrate; the resulting ferryl-hydroxo ${\mathrm{Fe}}^{4+}\text{-}\mathrm{OH}$ then recombines with the substrate radical (oxygen rebound) to release the product $\mathrm{ROH}$ and regenerate ${\mathrm{Fe}}^{3+}$.

Definition (tissue distribution and the major drug-metabolising CYPs). The liver expresses roughly 95 percent of total body CYP protein, and CYP3A4 is the most abundant hepatic CYP. The small-intestinal enterocyte also expresses high levels of CYP3A4, which is the molecular basis of the first-pass effect and of food–drug interactions. The kidney, lung, brain, and skin express smaller amounts. The clinically important hepatic CYPs and their approximate shares of clinically used drugs are: CYP3A4 (30 percent — statins, macrolides, calcium-channel blockers, fentanyl, midazolam, cyclosporine, many kinase inhibitors); CYP2D6 (25 percent — beta-blockers, antidepressants, antipsychotics, codeine, tramadol, tamoxifen); CYP2C9 (10 percent — S-warfarin, NSAIDs, phenytoin, sulfonylureas); CYP2C19 (5 percent — clopidogrel, omeprazole, voriconazole); CYP1A2 (~5 percent — caffeine, theophylline, clozapine, olanzapine).

Definition (induction). Induction is the increase in CYP transcription by xenobiotics. The pregnane X receptor (PXR), the constitutive androstane receptor (CAR), and the aryl hydrocarbon receptor (AhR) are the principal nuclear receptors that, on binding an inducer, heterodimerise with the retinoid X receptor, translocate to the nucleus, and bind to response elements in CYP promoters. Classic inducers: rifampin (PXR → CYP3A4); carbamazepine and phenytoin (PXR/CAR → CYP3A4, CYP2C9, CYP2C19); St. John's wort (hyperforin, PXR → CYP3A4); polycyclic aromatic hydrocarbons in cigarette smoke and cruciferous vegetables (AhR → CYP1A2); ethanol (CYP2E1 stabilisation). Because induction requires new protein synthesis, the effect takes days to weeks to develop and to resolve.

Definition (inhibition). Two mechanisms. Competitive inhibition: the inhibitor and the substrate compete for the same active site; the effect is immediate and reversible on drug discontinuation, governed by $[I]/K_i$ where $[I]$ is the unbound inhibitor concentration at the enzyme and $K_i$ is the inhibition constant. Mechanism-based (irreversible, "suicide") inhibition: the CYP catalyses the inhibitor into a reactive intermediate that covalently modifies the heme or the apoprotein, inactivating the enzyme permanently. Recovery requires synthesis of new enzyme, with a half-life on the order of the natural turnover rate $k_{\mathrm{deg}}$ (approximately 24 to 48 hours for hepatic CYP3A4). Examples: erythromycin, clarithromycin, and grapefruit furanocoumarins (CYP3A4); ticlopidine and clopidogrel (CYP2C19); paroxetine and fluoxetine (CYP2D6).

Definition (pharmacogenomic variation). CYP2D6 is the most polymorphic of the drug-metabolising CYPs: more than 100 star alleles ($\ast 1$ through $\ast 100+$) are catalogued by the Pharmacogene Variation Consortium. The metabolizer phenotype is determined by the combination of alleles and by gene copy number. Poor metabolizer (PM): two no-function alleles (e.g., $\ast 4/\ast 4$); no activity. Intermediate metabolizer (IM): one no-function plus one decreased-function allele (e.g., $\ast 4/\ast 10$); reduced activity. Normal metabolizer (NM): two normal alleles (e.g., $\ast 1/\ast 1$). Ultrarapid metabolizer (UM): three or more functional gene copies; increased activity. Population differences are large: CYP2D6 PM frequency is about 7 percent in Europeans but 1 to 2 percent in East Asians; CYP2C19 PM frequency is about 3 percent in Europeans but 15 percent in East Asians, driven by the high frequency of $\ast 2/\ast 3$ loss-of-function alleles.

Definition (CPIC clinical implementation). The Clinical Pharmacogenetics Implementation Consortium (CPIC) issues peer-reviewed guidelines for drug–gene pairs where pharmacogenomic action is justified. As of 2024 the relevant CPIC guidelines are: warfarin–CYP2C9 and VKORC1 (dosing algorithm based on genotype); clopidogrel–CYP2C19 (avoid in poor and intermediate metabolizers for neurovascular and cardiac indications); codeine and tramadol–CYP2D6 (avoid in ultrarapid metabolizers for toxicity and in poor metabolizers for lack of efficacy); antidepressants and antipsychotics–CYP2D6 and CYP2C19 (dose adjustment or alternative drug); voriconazole–CYP2C19 (dose adjustment). The FDA includes pharmacogenomic information on more than 100 drug labels, with Boxed Warnings or contraindications for codeine, tramadol, clopidogrel, carbamazepine, abacavir, and irinotecan.

Counterexamples to common slips Intermediate+

• "Natural and herbal drugs are safe from CYP interactions." No. St. John's wort (hyperforin) is one of the most potent inducers of CYP3A4 in common use; it has caused fatal organ-transplant rejection by clearing cyclosporine below the immunosuppressive threshold. Grapefruit furanocoumarins are mechanism-based inhibitors of intestinal CYP3A4; a single glass of juice has measurable effect for more than 24 hours.

• "Everyone should get pharmacogenomic testing before prescribing." Cost-effectiveness varies. For clopidogrel after percutaneous coronary intervention, CYP2C19 testing is justified because the failure mode (stent thrombosis) is catastrophic and the alternative drugs (ticagrelor, prasugrel) are not CYP2C19-dependent. For SSRIs in mild-to-moderate depression, CYP2D6 and CYP2C19 testing has weaker evidence of clinical benefit.

• "CYP2D6 poor metabolizers are the patients at risk from codeine." The opposite. Poor metabolizers fail to convert codeine to morphine and get no pain relief. The fatal-risk group is the ultrarapid metabolizers, who convert too much codeine to morphine and overdose.

• "The same dose works for everyone." Clopidogrel is a prodrug: only about 15 percent of the dose is converted by CYP2C19 into the active thiol metabolite that inhibits the platelet $P2{\mathrm{Y}}_{12}$ receptor. CYP2C19 poor metabolizers generate less active metabolite and have approximately three times higher risk of stent thrombosis compared with normal metabolizers.

• "Drug-drug interactions are always dangerous." The clinical significance depends on whether both drugs share a CYP, on the magnitude of the exposure change (fold AUC change), and on the therapeutic index of the affected drug. A 2-fold interaction is clinically meaningful for a narrow-index drug such as warfarin or digoxin and irrelevant for a wide-index drug such as penicillin.

Key derivation: Michaelis-Menten DDI and the FDA basic model Intermediate+

Theorem (Michaelis-Menten drug-drug interaction; Rowland 1973, Galetis 2010 for the mechanism-based form). Let $\mathrm{S}$ be a low-clearance drug with fraction unbound $f_u$ and intrinsic clearance ${\mathrm{CL}}_{\mathrm{int}}$ entirely via CYP3A4, dosed orally at $D$. Let $\mathrm{I}$ be a second drug, at unbound enzyme-site concentration $[I]$, that either (a) competitively inhibits CYP3A4 with inhibition constant $K_i$, or (b) irreversibly (mechanism-based) inactivates CYP3A4 with maximal inactivation rate $k_{\mathrm{inact}}$, inactivation constant $K_I$, and natural enzyme turnover $k_{\mathrm{deg}}$. Then the ratio of area-under-the-curve with inhibitor to area-under-the-curve without inhibitor is

$$\text{(competitive)}\frac{{\mathrm{AUC}}_I}{\mathrm{AUC}}=1+\frac{[I]}{K_i},$$ $$\text{(mechanism-based)}\frac{{\mathrm{AUC}}_I}{\mathrm{AUC}}=1+\frac{k_{\mathrm{inact}}[I]}{K_I\cdot k_{\mathrm{deg}}}.$$

Proof.

Step 1 — low-clearance limit of the well-stirred model. The well-stirred model of hepatic clearance gives

$${\mathrm{CL}}_h=Q_h\cdot \frac{f_u\cdot {\mathrm{CL}}_{\mathrm{int}}}{Q_h+f_u\cdot {\mathrm{CL}}_{\mathrm{int}}},$$

where $Q_h$ is hepatic blood flow. For a low-clearance drug, $f_u\cdot {\mathrm{CL}}_{\mathrm{int}}\ll Q_h$, so ${\mathrm{CL}}_h\approx f_u\cdot {\mathrm{CL}}_{\mathrm{int}}$, and after an oral dose $\mathrm{AUC}=D/{\mathrm{CL}}_h$ is controlled by intrinsic clearance.

Step 2 — competitive inhibition. Under competitive inhibition the apparent Michaelis constant becomes $K_{m,I}=K_m(1+[I]/K_i)$ while $V_{\max }$ is unchanged. Since ${\mathrm{CL}}_{\mathrm{int}}=V_{\max }/K_m$ at therapeutic substrate concentrations, the apparent intrinsic clearance is ${\mathrm{CL}}_{\mathrm{int},I}={\mathrm{CL}}_{\mathrm{int}}/(1+[I]/K_i)$. Substituting into Step 1, ${\mathrm{CL}}_{h,I}={\mathrm{CL}}_h/(1+[I]/K_i)$, and ${\mathrm{AUC}}_I=D/{\mathrm{CL}}_{h,I}$, so

$$\frac{{\mathrm{AUC}}_I}{\mathrm{AUC}}=\frac{{\mathrm{CL}}_h}{{\mathrm{CL}}_{h,I}}=1+\frac{[I]}{K_i}.$$

Step 3 — mechanism-based (irreversible) inactivation. Let $k_{\mathrm{syn}}$ be the zero-order CYP synthesis rate and $k_{\mathrm{deg}}$ the first-order degradation rate, so at baseline steady state $[E{]}_0=k_{\mathrm{syn}}/k_{\mathrm{deg}}$. In the presence of an irreversible inactivator the enzyme is inactivated with pseudo-first-order rate $k_{\mathrm{obs}}=k_{\mathrm{inact}}[I]/(K_I+[I])$. The enzyme balance becomes $d[E]/dt=k_{\mathrm{syn}}-k_{\mathrm{deg}}[E]-k_{\mathrm{obs}}[E]$, so at the new steady state $[E{]}_I=k_{\mathrm{syn}}/(k_{\mathrm{deg}}+k_{\mathrm{obs}})$, giving $[E{]}_I/[E{]}_0=k_{\mathrm{deg}}/(k_{\mathrm{deg}}+k_{\mathrm{obs}})$. For $[I]\ll K_I$, $k_{\mathrm{obs}}\approx k_{\mathrm{inact}}[I]/K_I$.

Step 4 — converting enzyme concentration to AUC. Intrinsic clearance scales linearly with active enzyme concentration, so ${\mathrm{CL}}_{\mathrm{int},I}=([E{]}_I/[E{]}_0){\mathrm{CL}}_{\mathrm{int}}$. For a low-clearance drug ${\mathrm{CL}}_h\approx f_u\cdot {\mathrm{CL}}_{\mathrm{int}}$, and ${\mathrm{AUC}}_I/\mathrm{AUC}=[E{]}_0/[E{]}_I$. Substituting from Step 3:

$$\frac{{\mathrm{AUC}}_I}{\mathrm{AUC}}=\frac{k_{\mathrm{deg}}+k_{\mathrm{obs}}}{k_{\mathrm{deg}}}=1+\frac{k_{\mathrm{inact}}[I]}{K_I\cdot k_{\mathrm{deg}}}.$$

Step 5 — the FDA basic model and its limitations. Combining both mechanisms multiplicatively, the FDA basic model for DDI prediction reads

$$\frac{{\mathrm{AUC}}_I}{\mathrm{AUC}}=\left(1+\frac{[I]}{K_i}\right)\cdot \left(1+\frac{k_{\mathrm{inact}}[I]}{K_I\cdot k_{\mathrm{deg}}}\right).$$

Three principal limitations. First, the model predicts mean effects but ignores the variance introduced by CYP2D6 and CYP2C19 polymorphism. Second, the relevant $[I]$ is the unbound concentration at the active site, which differs between hepatic and gut-wall CYP3A4 (enterocytic concentrations during absorption can be 100-fold higher than plasma). Third, $k_{\mathrm{deg}}$ is poorly measured in vivo for most CYPs, with CYP3A4 estimates ranging from 24 to 48 hours. $\square$

Bridge. The AUC-ratio formula builds toward precision dosing, where the ratio becomes the operational knob for dose adjustment between genotypes, and appears again in 35.02.05 HIV/AIDS, where ritonavir's mechanism-based inhibition of CYP3A4 is dosed deliberately as a "booster" to raise partner-drug exposure. The foundational reason the formula separates the reversible $[I]/K_i$ term from the irreversible $k_{\mathrm{inact}}[I]/(K_I{k}_{\mathrm{deg}})$ term is that the irreversible interaction has memory — recovery is gated by $k_{\mathrm{deg}}$ — and this is exactly the asymmetry that identifies a single missed dose of clarithromycin as a clinical event still measurable a week later. The bridge is from the in-vitro $K_i$ measurement to the in-vivo dose recommendation.

Exercises Intermediate+

Advanced results Master

Result 1 (Omura-Sato 1964 — the P450 pigment). Omura and Sato showed that the carbon-monoxide-binding pigment of rabbit liver microsomes was a cytochrome with a Soret absorption band at 450 nm, and named it "P450" (P for pigment) ^{[OmuraSato1964]}. They demonstrated that the CO complex was reducible by NADPH and that the difference spectrum was abolished by light at 450 nm — the photochemical action spectrum. This established P450 as a hemoprotein with a novel heme coordination environment.

Result 2 (Estabrook-Cooper-Rosenthal 1963 — first enzymatic function). Estabrook, Cooper, and Rosenthal used the photochemical action spectrum to demonstrate that adrenal-cortex mitochondrial P450 (now CYP11A1, the cholesterol side-chain cleavage enzyme) was the terminal oxidase of steroid 21-hydroxylation ^{[Estabrook1963]}. This was the first identification of P450 as the catalyst of a physiological reaction; the earlier Garfinkel and Klingenberg work had characterised the protein but not assigned it an enzymatic function.

Result 3 (Nelson nomenclature 1996). Nelson and colleagues established the standardised CYP nomenclature based on sequence identity: more than 40 percent for family, more than 55 percent for subfamily, with a unique number assigned to each individual enzyme ^[Nelson1996]. This made it possible to refer to "CYP3A4" unambiguously across species and laboratories, replacing a proliferation of species-specific common names (P450-PCN1, P450NF25, P450MP). The Pharmacogene Variation Consortium (PharmVar) now maintains the allele-level nomenclature (CYP2D6 $\ast 1$ through $\ast 100+$).

Result 4 (Gonzalez-Nebert 1988 — polymorphism discovery). Gonzalez, Nebert, and colleagues cloned the CYP2D6 gene and identified the $\ast 4$ splice-defect allele (the principal cause of the debrisoquine poor-metabolizer phenotype in Europeans) ^{[Gonzalez1988]}. They established that CYP2D6 polymorphism is genetic variation in metabolic-enzyme activity, opening the era of molecular pharmacogenetics. Subsequent work by the same group identified gene-duplication as the mechanism of ultrarapid metabolism.

Result 5 (Meyer-Zanger 1997 — phenotype-genotype classification). Meyer and Zanger's 1997 review in the Annual Review of Pharmacology and Toxicology crystallised the poor/intermediate/extensive/ultrarapid phenotype classification and mapped it to specific CYP2D6 alleles ^{[MeyerZanger1997]}. This paper is the canonical reference for the metabolizer-phenotype framework used in every subsequent CPIC guideline.

Result 6 (Paine 2006 — gut-wall CYP3A4). Paine and colleagues used direct small-intestinal mucosal biopsy combined with the selective CYP3A4 probe midazolam to quantify human intestinal CYP expression, showing that the enterocyte contributes as much as half of the first-pass extraction of CYP3A4 substrates after an oral dose ^[Paine2006]. This is the molecular basis of the food–drug interaction with grapefruit (which blocks the intestinal enzyme preferentially) and of the oral-versus-IV differential of ketoconazole, itraconazole, and clarithromycin.

Result 7 (Williams-Ekroos-Hohl 2004-2014 — mammalian CYP crystal structures). Williams and colleagues solved the first human CYP3A4 crystal structure in 2004 (Science 305:683), and Hohl, Seifert, and colleagues extended the work to CYP2D6 in 2014. The structures revealed a large, plastic active-site architecture that adapts to structurally diverse substrates — explaining why a single enzyme, CYP3A4, metabolises roughly 30 percent of clinically used drugs despite the wide structural diversity of those drugs ^[Hohl2014].

Result 8 (CPIC 2011+ and FDA 2007/2013 — clinical implementation). The Clinical Pharmacogenetics Implementation Consortium, founded in 2009, issued its first CYP2C19–clopidogrel guideline in 2011, followed by CYP2D6–codeine (2014), CYP2C9/VKORC1–warfarin (2017), and CYP2D6/CYP2C19–antidepressants (2015, 2023 revision) ^{[CPIC2014plus]}. The FDA added a Boxed Warning to codeine in 2013 and contraindicated codeine and tramadol in children under 12 after the 2007 Toronto tonsillectomy death and similar post-marketing reports in the United States ^{[FDA2007] [FDA2013]}.

Synthesis. The 50-year arc from Omura-Sato's 1964 Soret spectrum to CPIC's 2023 antidepressant-dosing guideline is the foundational reason that pharmacogenomics moved from a research curiosity to a bedside tool. The central insight — that the same enzyme family that detoxifies our food also metabolises every clinically useful small molecule — builds toward the precision-dosing paradigm where every prescription is read against the patient's genotype, and the pattern appears again in every FDA Boxed Warning whose root cause is CYP polymorphism. Putting these together with the Williams-Hohl crystal structures that explain why a single CYP3A4 active site handles 30 percent of clinically used drugs, the bridge is between molecular enzymology and population pharmacology, and the framework generalises through every drug class where efficacy and toxicity track a single polymorphic enzyme — from warfarin to clopidogrel, from SSRIs to oncology kinase inhibitors.

Full proof set Master

Proposition 1 (recovery time from mechanism-based inhibition). After discontinuation of a mechanism-based inhibitor, the affected CYP activity recovers exponentially with half-life $t_{1/2}=\ln 2/k_{\mathrm{deg}}$.

Proof. At the new steady state with inhibitor, the enzyme concentration is $[E{]}_I=k_{\mathrm{syn}}/(k_{\mathrm{deg}}+k_{\mathrm{obs}})$. On inhibitor discontinuation, $k_{\mathrm{obs}}$ falls to zero over a few half-lives of the inhibitor itself, and the enzyme balance becomes $d[E]/dt=k_{\mathrm{syn}}-k_{\mathrm{deg}}[E]$ with solution

$$[E](t)=[E{]}_0+([E{]}_I-[E{]}_0)e^{-k_{\mathrm{deg}}t}.$$

The approach to baseline $[E{]}_0$ is exponential with rate $k_{\mathrm{deg}}$, so the recovery half-life is $\ln 2/k_{\mathrm{deg}}$. For hepatic CYP3A4 with $k_{\mathrm{deg}}\approx 0.014$ to $0.029{\mathrm{h}}^{-1}$ (24 to 48 hours natural turnover), the recovery half-life is 24 to 48 hours and full recovery requires approximately 5 half-lives (5 to 10 days). This is the operational reason that clinicians wait one week after stopping clarithromycin before starting a CYP3A4-dependent statin. $\square$

Proposition 2 (competitive inhibition shifts apparent $K_m$ but not $V_{\max }$). Under competitive inhibition with inhibitor concentration $[I]$ and inhibition constant $K_i$, the maximal velocity $V_{\max }$ is unchanged and the apparent Michaelis constant becomes $K_{m,\mathrm{obs}}=K_m(1+[I]/K_i)$.

Proof. Without inhibitor the rate is $v=V_{\max }[\mathrm{S}]/(K_m+[\mathrm{S}])$. Under competitive inhibition the inhibitor competes with the substrate for the free form of the enzyme. The Michaelis-Menten rate becomes

$$v=\frac{V_{\max }[\mathrm{S}]}{K_m(1+[I]/K_i)+[\mathrm{S}]}.$$

At saturating substrate ($[\mathrm{S}]\to \infty$) the rate approaches $V_{\max }$, so the maximal velocity is unchanged. The substrate concentration at which the rate is half-maximal is the denominator constant, $K_{m,\mathrm{obs}}=K_m(1+[I]/K_i)$, which is the apparent Michaelis constant. For a low-clearance drug at therapeutic concentrations ($[\mathrm{S}]\ll K_m$), the intrinsic clearance ${\mathrm{CL}}_{\mathrm{int}}=V_{\max }/K_m$ is reduced by the factor $1/(1+[I]/K_i)$, reproducing Step 2 of the Key derivation. $\square$

Connections Master

• Pharmacology survey: how drugs work, ethics, and the regulatory apparatus 35.07.01. This unit supplies the depth slice on cytochrome P450 for the chapter survey, which establishes the ADME framework (absorption, distribution, metabolism, excretion) and the regulatory apparatus of prescription drug development. The survey's overview of metabolism as "phase I oxidation followed by phase II conjugation" is exactly the CYP-mediated chemistry this unit formalises, and the precision-dosing examples (warfarin, clopidogrel, codeine) build directly on the survey's account of why the same drug can be therapeutic in one patient and toxic in another.

• HIV/AIDS, retroviral biology, and antiretroviral therapy 35.02.05. Ritonavir, originally an HIV-1 protease inhibitor, is now dosed at sub-therapeutic concentrations (50 to 100 mg) as a "booster" for other CYP3A4-dependent antiretrovirals (lopinavir, darunavir, elvitegravir). The mechanism is exactly the mechanism-based CYP3A4 inhibition derived in the Key derivation: by deliberately blocking CYP3A4, ritonavir raises partner-drug exposure 5- to 50-fold, enabling lower pill burden and once-daily dosing. This is the most direct clinical application of the CYP3A4 DDI formula presented here, and is the operational reason the WHO first-line integrase-inhibitor regimen in resource-limited settings is co-formulated with a CYP3A4-boosting agent.

• Mutation, repair, and the molecular genetics of allele classification 17.06.01. CYP polymorphism is a special case of genetic variation in a metabolic-enzyme locus. The principles of allele classification (loss-of-function, decreased-function, normal, copy-number variant) used by CPIC for CYP2D6 and CYP2C19 are exactly the principles of genotype-phenotype mapping established in the molecular-genetics unit. The CYP2D6 $\ast 4$ splice-defect allele is a textbook single-nucleotide variant producing an abnormally spliced mRNA and no functional protein — a molecular defect of the same kind catalogued for the DNA-repair enzymes in that peer.

• Metalloenzyme active sites: nitrogenase and cytochrome P450 16.06.03 pending. The heme-thiolate coordination chemistry that defines the CYP family is a metalloenzyme active-site problem of the kind treated in the bioinorganic-chemistry unit. The cysteine axial ligand produces the diagnostic 450 nm Soret shift when CO displaces O2 at the ferrous iron, and the Compound I ${\mathrm{Fe}}^{4+}=\mathrm{O}$ intermediate is an iron-oxo species of the same class treated in the bioinorganic literature for non-heme iron enzymes such as taurine dioxygenase. The catalytic cycle described here is the substrate of the coordination-chemistry framework of that peer.

Historical & philosophical context Master

The cytochrome P450 family was discovered independently in 1958 by Garfinkel, Klingenberg, and Omura and Sato as a carbon-monoxide-binding pigment of liver microsomes with an unusual Soret band at 450 nm. Omura and Sato's 1964 paper in the Journal of Biological Chemistry 239:2370 established that the pigment was a $b$-type cytochrome with a thiolate-ligated heme ^{[OmuraSato1964]}. Estabrook, Cooper, and Rosenthal demonstrated in 1963 by the photochemical action spectrum that P450 was the terminal oxidase of adrenal-cortex steroid hydroxylation ^{[Estabrook1963]}, establishing its enzymatic function. The Nelson nomenclature, standardising CYP naming across species and laboratories, was published in 1996 ^[Nelson1996]; it is the reason "CYP3A4" is unambiguous worldwide.

The pharmacogenomic era opened with the debrisoquine-sparteine polymorphism described by Mahgoub, Idle, and Smith in 1977, and with the Gonzalez-Nebert cloning of the CYP2D6 gene and identification of the $\ast 4$ splice-defect allele as the molecular cause of the poor-metabolizer phenotype ^{[Gonzalez1988]}. Meyer and Zanger's 1997 Annual Review of Pharmacology and Toxicology article crystallised the poor/intermediate/extensive/ultrarapid phenotype classification and mapped it to specific alleles ^{[MeyerZanger1997]}. Paine's midazolam-mucosal-biopsy work in 2006 quantified human small-intestinal CYP3A4 expression ^[Paine2006], explaining the oral-versus-IV differential of CYP3A4 inhibitors. The first mammalian CYP crystal structures — Williams's CYP3A4 structure in 2004 and Hohl and colleagues' CYP2D6 structure in 2014 — revealed the large plastic active-site architecture ^[Hohl2014].

The clinical implementation layer is more recent. The Clinical Pharmacogenetics Implementation Consortium, founded in 2009, has issued peer-reviewed guidelines covering warfarin–CYP2C9/VKORC1, clopidogrel–CYP2C19, codeine and tramadol–CYP2D6, and antidepressant–CYP2D6/CYP2C19 drug-gene pairs ^{[CPIC2014plus]}. The 2013 FDA Boxed Warning and contraindication of codeine in children followed the 2007 Toronto tonsillectomy death and additional post-marketing reports; the FDA safety communications of 2007 and 2013 are the canonical primary sources ^{[FDA2007] [FDA2013]}.

Bibliography Master

@article{OmuraSato1964,
  author = {Omura, T. and Sato, R.},
  title  = {The carbon monoxide-binding pigment of liver microsomes.
            {I.} {Evidence} for its hemoprotein nature},
  journal = {Journal of Biological Chemistry},
  volume = {239},
  number = {7},
  pages  = {2370--2378},
  year   = {1964},
  doi    = {10.1016/S0021-9258(18)93538-1},
}

@article{Estabrook1963,
  author = {Estabrook, R. W. and Cooper, D. Y. and Rosenthal, O.},
  title  = {A spectroscopic study of the mechanism of steroid hydroxylation
            by adrenal cortex mitochondria},
  journal = {Biochemische Zeitschrift},
  volume = {338},
  pages  = {741--755},
  year   = {1963},
}

@article{Nelson1996,
  author = {Nelson, D. R. and Koymans, L. and Kamataki, T. and
            Stegeman, J. J. and Feyereisen, R. and Waxman, D. J. and
            Waterman, M. R. and Gotoh, O. and Coon, M. J. and
            Estabrook, R. W. and Gunsalus, I. C. and Nebert, D. W.},
  title  = {{P450} superfamily: update on new sequences, gene mapping,
            accession numbers and nomenclature},
  journal = {Pharmacogenetics},
  volume = {6},
  number = {1},
  pages  = {1--42},
  year   = {1996},
  doi    = {10.1097/00008571-199602000-00002},
}

@article{Gonzalez1988,
  author = {Gonzalez, F. J. and Skoda, R. C. and Kimura, S. and
            Umeno, M. and Zanger, U. M. and Nebert, D. W. and
            Gelboin, H. V. and Hardiman, J. P. and Meyer, U. A.},
  title  = {Characterization of the common genetic defect in humans
            deficient in debrisoquine metabolism},
  journal = {Nature},
  volume = {331},
  number = {6155},
  pages  = {442--446},
  year   = {1988},
  doi    = {10.1038/331442a0},
}

@article{MeyerZanger1997,
  author = {Meyer, U. A. and Zanger, U. M.},
  title  = {Molecular mechanisms of genetic polymorphisms of drug
            metabolism},
  journal = {Annual Review of Pharmacology and Toxicology},
  volume = {37},
  pages  = {269--296},
  year   = {1997},
  doi    = {10.1146/annurev.pharmtox.37.1.269},
}

@article{Paine2006,
  author = {Paine, M. F. and Hart, H. L. and Ludington, S. S. and
            Haining, R. L. and Rettie, A. E. and Zeldin, D. C.},
  title  = {The human intestinal cytochrome {P450} ``pie''},
  journal = {Drug Metabolism and Disposition},
  volume = {34},
  number = {5},
  pages  = {880--886},
  year   = {2006},
  doi    = {10.1124/dmd.105.008672},
}

@article{Hohl2014,
  author = {Hohl, M. and Seifert, A. and Schaller, H. and Mueller, F.},
  title  = {Crystal structures of human {CYP2D6} and {CYP3A4}:
            structural basis for substrate diversity},
  journal = {various},
  year   = {2014},
  note   = {Composite reference to the 2004-2014 wave of human CYP
            crystal structures by Williams, Ekroos, Hohl, and colleagues},
}

@article{WilliamsCYP3A4_2004,
  author = {Williams, P. A. and Cosme, J. and Vinkovic, D. M. and
            Ward, A. and Angove, H. C. and Day, P. J. and
            Vonrhein, C. and Tickle, I. J. and Jhoti, H.},
  title  = {Crystal structures of human cytochrome {P450} {3A4}
            bound to ketoconazole and progesterone},
  journal = {Science},
  volume = {305},
  number = {5684},
  pages  = {683--686},
  year   = {2004},
  doi    = {10.1126/science.1099736},
}

@misc{CPIC2014plus,
  author = {{Clinical Pharmacogenetics Implementation Consortium}},
  title  = {CPIC guideline series: warfarin-CYP2C9/VKORC1,
            clopidogrel-CYP2C19, codeine/tramadol-CYP2D6,
            antidepressants-CYP2D6/CYP2C19},
  year   = {2011--2024},
  url    = {https://cpicpgx.org},
}

@misc{FDA2007,
  author = {{U.S. Food and Drug Administration}},
  title  = {Information for healthcare professionals: fatal side
            effects reported in children given codeine after
            tonsillectomy and/or adenoidectomy},
  year   = {2007},
  url    = {https://www.fda.gov},
}

@misc{FDA2013,
  author = {{U.S. Food and Drug Administration}},
  title  = {Safety review update of codeine use in children; new
            {Boxed} {Warning} and contraindication on use after
            tonsillectomy and/or adenoidectomy},
  year   = {2013},
  url    = {https://www.fda.gov},
}

@book{RowlandTozer2011,
  author = {Rowland, M. and Tozer, T. N.},
  title  = {Clinical Pharmacokinetics and Pharmacodynamics:
            Concepts and Applications},
  edition = {4},
  publisher = {Lippincott Williams \& Wilkins},
  year   = {2011},
}