43.2.2.1

P450 Enzymatic Transformation

Among the phase I biotransforming enzymatic systems, the “cytochrome P450” system is by far the most versatile and productive. With the largest number of substrates, this system is one of the most important for hepatic detoxification. The “cytochrome P450” enzymes are a large family of heme-thiolate enzymatic proteins located on the endoplasmic reticulum (ER) of hepatocytes. The term cytochrome P450 is derived from the original observation that these enzymes, isolated from hepatocyte ER microsomes, are chemically similar to mitochondrial cytochromes, and are reddish in pigmentation reflecting the heme that is part of their molecular structure. The 450 designation reflects the finding that these microsomes have peak light absorption with 450 nm wavelength light when chemically treated with a reducing agent and bound with carbon monoxide.

The basic reaction catalyzed by the cytochrome P450 is written as follows:

By definition, the reaction is a monooxygenase because of the incorporation of only one of the two oxygen atoms onto the substrate. The reactions catalyzed by P450 proteins are numerous and include: aromatic and side chain hydroxylation; N-, O-, and S-dealkylation; N-oxidation; N-hydroxylation; sulfoxidation; deamination; dehalogenation; and desulfuration. The resulting substrate metabolites are usually stable; however, it is at this step that many toxic metabolites are generated during hepatic drug metabolism.

One of the most important groups of substrates for the cytochrome P450 system is xenobiotics, or those exogenous compounds that are “foreign to life.” This group includes potentially toxic compounds such as therapeutic drugs, food additives, and industrial chemicals or other types of environmental contaminants. Once ingested, the lipophilic quality of these compounds facilitates absorption into the blood. Without removal, they would accumulate in the cells throughout the body and cause cellular dysfunction. However, blood from the gut first enters the liver sinusoids, where xenobiotics can be filtered out by hepatocytes. They either passively diffuse or are actively transported into hepatocytes where they encounter particular P450s able to metabolize them. Some xenobiotics are successfully metabolized by only one P450 and then undergo phase II metabolism prior to excretion. Others, however, require metabolism by more than one P450, and hence undergo more than one phase I reaction prior to undergoing phase II metabolism and excretion ( Fig. 43.1 ). The combination of phase I and phase II metabolic reactions transforms the original lipophilic toxin into a more polar hydrophilic molecule. This water-soluble molecule can be more readily excreted by the kidneys or via the bile into the intestine with a reduced likelihood of enterohepatic cycling.

Enzymatic transformation of toxic compounds in hepatocytes through phase I and II metabolism, and cellular excretion by the ABC transporter proteins. Illustrated are two hypothetical compounds, A and B, bound to proteins when in the blood. After they enter the hepatocyte they diffuse to receptive cytochrome P450 enzymes and undergo phase I metabolism, becoming activated. They are then conjugated during phase II metabolism resulting in polar compounds, A + and B +. These more hydrophilic compounds are then actively transported across the cell membrane for excretion by the ABC transporter proteins.

It is now known that individual P450s represent the products of unique genes, and that the large interindividual differences in the activity of these proteins are in part due to gene polymorphisms. The amino acid sequence of the cytochrome P450 proteins allows them to be classified into families and subfamilies based on homology. Families are designated by a unique Arabic numeral and contain cytochrome P450 proteins that share approximately 40% of their amino acid sequence. Those most important for drug metabolism fall into the gene families CYP1, CYP2, or CYP3. Family members that share at least 55% sequence identity are further classified as a subfamily, and identified by a capital letter appended to the family designation, such as 2A, 2B, or 2C. A final number is added to individually identify a specific protein within a given subfamily, for instance, 2B1, 2B2, or 2B3. The CYP designation is the nomenclature common to all of the cytochrome P450 genes and proteins.

Only a few of the many P450s in the liver are needed to detoxify therapeutic drugs. CYP3A4 is the most important, and is responsible for the metabolism of half of all drugs. The remaining 50% of drugs are metabolized by six additional enzymes: CYP2D6 metabolizes 20%, CYP2C9 and CYP2C19 together another 15%, and CYP1A2, CYP2A6, CYP2B6, and others metabolize the remaining 15%.

Although most substances that undergo phase I P450 metabolism subsequently undergo phase II conjugation prior to elimination from the body, the rate-limiting step is usually the P450 metabolism. Furthermore, the activity of the P450 proteins is genetically variable from individual to individual and can be altered by nongenetic factors. Genetic variability is best exemplified by CYP2D6, the cytochrome responsible for the metabolism of 20% of medications, including some of those used to treat depression, psychosis, and certain cardiac arrhythmias. Individuals with no measurable enzyme activity are termed “poor metabolizers,” and are more susceptible to drug toxicity from these medications. Lack of CYP2D6 activity is an autosomal recessive trait that occurs in 5%–10% of Caucasians of European descent but occurs in only 1%–2% of those of Southeast Asian ancestry. In comparison, 20% of Asians and 5% of Caucasians are “poor metabolizers” with respect to CYP2C19 activity; hence they can develop excessive levels of drugs such as omeprazole that may enhance efficacy in certain circumstances. Inheritance of this defect is similarly autosomal recessive.

Although diminished catalytic activity has been associated with other cytochrome P450 enzymes due to allelic mutations, there is only one example of increased activity from a genetic cause. Gene duplication of CYP2D6 results in exceedingly rapid clearance of some CYP2D6 substrates, which could decrease the efficacy of medications metabolized through this pathway ( Table 43.1 ). Approximately 2% of Caucasians are “ultra-rapid” metabolizers from this defect.

Multiple nongenetic factors can also alter cytochrome P450 activity, including other medications, nutritional changes, and disease. The most notorious of these involve drug-drug interactions ( Table 43.1 ). Since many of the cytochrome P450 enzymes can bind different drug molecules, competition between multiple drugs for the same enzyme will decrease the P450 activity against any one drug. One example of this is decreased clearance of ­cyclosporin A in a patient given erythromycin; an effect caused by inhibition of CYP3A4. Increased activity of P450s can also occur. In most examples of cytochrome P450 induction ( Table 43.1 ), the transcription of the P450 gene is increased, resulting in a higher concentration of the P450 in hepatocytes, and hence higher activity of the enzyme. P450 induction is initiated by binding of the inducer to a specific receptor in the cytoplasm. The complex is then translocated to the nucleus where it interacts with regulatory sites of the P450 gene. For example, phenobarbital may reduce cyclosporin A levels to a subtherapeutic level via induction of CYP3A4 in patients given both drugs.

43.2.2.2

Flavin Monooxygenase

As with the cytochrome P450 enzymes, oxidized flavin adenine dinucleotide (FAD)-containing monooxygenases (FMO) localize to ER microsomes and require Nicotinamide adenine dinucleotide phosphate (NADPH) and O ₂ . In addition, many of the reactions catalyzed by FMO are also catalyzed by the cytochrome P450 enzymes. Such reactions include the oxidation of nucleophilic nitrogen, sulfur, and phosphorous heteroatoms of tertiary amines to N-oxides, secondary amines to hydroxylamines and nitrones, primary amines to hydroxylamines and oximes, and sulfur-containing drugs and phosphines to S- and P-oxides. In contrast to the P450 reactions, however, FMOs do not catalyze N-, S-, or O-dealkylation reactions. Examples of drugs metabolized via these reactions include amphetamine, imipramine, tamoxifen, thiols, and thiocarbamates. Cimetidine is an example of a sulfur-containing drug metabolized by FMO, which is an inhibitor of cytochrome P450.

The five FMO (FMO1-5) enzymes are highly conserved across species and all contain a glycine-rich region that binds 1 mol of FAD near the active catalytic site and also near a second glycine-rich region that binds the NADPH.

In contrast to the cytochrome P450 reactions where oxygenation of the heteroatom is initially with one electron, the structure of FMO allows for two-electron oxygenation. Hence, the N-oxygenation by cytochrome P450 results in N-dealkylation, whereas it results in N-oxide formation when catalyzed by FMO. Overall, the reactions detoxify xenobiotics by reaction with a peracid or peroxide. The FAD is reduced to FADH ₂ by NADPH, after which the oxidized cofactor NADP ⁺ remains bound to the enzyme. FADH ₂ then binds oxygen, producing peroxide. In interaction with a substrate, the flavin peroxide is then transferred, and the FAD is restored via dehydration, releasing NADP ⁺ .

In animals and humans, the five FMO proteins are expressed in liver, kidney, and lung to variable degrees. In humans, FMO3 is by far most highly expressed in the liver and the others are only expressed with low activity. Interestingly, FMO3 is also the flavin monooxygenase responsible for the conversion of ( S )-nicotine to the trans -isomer of ( S )-nicotine N -1′-oxide, which is excreted in the urine of smokers and those using a nicotine patch. Hence, the urinary excretion of trans- ( S )-nicotine N -1′-oxide can be used to probe the activity of FMO3 in vivo in humans. Similarly, the metabolism of cimetidine to cimetidine- S -oxide can be detected in the urine and may reflect enzyme activity. The regulation of the FMOs is independent from that of the P450 enzymes. Many compounds that induce P450 actually inhibit FMO. Indole-3-carbinol, found in Brussels sprouts, induces many of the P450 enzymes but directly inhibits FMO3.

43.2.2.3

Alcohol Dehydrogenase

Among the myriad other enzymes involved in phase I oxidation, alcohol dehydrogenase (ADH), and aldehyde dehydrogenase (ALDH) are of greatest clinical importance. The ADH is a zinc-containing cytosolic enzyme that oxidizes ethanol to acetaldehyde. The ADH is in highest concentration in the liver, although it is also present in the kidney, lung, and gastric mucosa. This enzyme is a dimeric protein comprising two out of six potential 40 kDa subunits and multiple allelic variants. The resultant ADH molecules are grouped into four major molecular classes. Class I enzymes are responsible for the oxidation of ethanol and other small aliphatic alcohols. Class II enzymes, on the other hand, preferentially oxidize larger aliphatic and aromatic alcohols with minimal to no role in the oxidation of ethanol. Class III ADH substrates include the long-chain aliphatic and aromatic alcohols, and this ADH is largely responsible for the detoxification of formaldehyde (class III ADH is actually the same enzyme as formaldehyde dehydrogenase). Class IV ADH is the only ADH not found in the adult human liver. It is, however, present in the mucosa of the upper intestinal tract, especially the mouth, esophagus, and stomach and is most active in oxidizing retinol, which is a vitamin important for cell growth and differentiation. The conversion of ethanol to acetaldehyde, a potential carcinogen, by class IV ADH may be involved in the pathogenesis of upper intestinal cancers associated with chronic alcohol ingestion.

Different isozymes of the class I ADH differ in their affinity and capacity for oxidizing ethanol to acetaldehyde. This explains the ethnic differences in sensitivity to the toxic effects of ethanol that are caused by acetaldehyde. One variant, atypical ADH, is responsible for very efficient conversion in 90% of the Pacific Rim Asian population, but is present in < 20% of Caucasians, < 10% of African Americans, and is not found in Native Americans or Asian Indians. Atypical class I ADH is more active at low ethanol concentrations: it has a K _m of 50 μM compared with 4 mM for “normal” class I ADH. Thus, people with atypical ADH have a lower threshold for the ill effects of alcohol. At inebriating levels of alcohol consumption (a blood ethanol level of 0.1% corresponds to 22 mM) both isoforms are fully active, acetaldehyde production is maximal, and there is little difference in ill effect among ethnic groups.

43.2.2.4

Aldehyde Dehydrogenase

The ALDH oxidizes aldehydes to carboxylic acids (e.g., acetaldehyde to acetate). To date, 19 ALDH genes have been identified in humans, leading to cytosolic, mitochondrial, and microsomal enzymes in different tissue types. ALDH2 is a mitochondrial enzyme found in the liver and mucosa of the upper intestinal tract, among other tissues, and is primarily responsible for the oxidation of simple aldehydes such as acetaldehyde, a potential carcinogen. Approximately 50% of Pacific Rim Asians as well as Taiwanese and Vietnamese are deficient in ALDH2 activity. This same population has a high incidence of atypical ADH activity. The rapid metabolism of ethanol to acetaldehyde in the setting of atypical ADH, combined with ALDH2 deficiency, causes a rapid accumulation of acetaldehyde, resulting in catecholamine release, dilation of facial blood vessels, and flushing. Acetaldehyde accumulation also causes nausea which is likely protective against alcoholism in these individuals. Disulfiram (Antabuse) is used to inhibit ALDH resulting in acetaldehyde accumulation and nausea with alcohol ingestion.

43.2.2.5

Other Phase I Enzymatic Reactions

Reduction and hydrolysis reactions also contribute to phase I metabolism of toxins and drugs. Metals and compounds containing azo- or nitro-groups, and aldehyde, ketone, disulfide, sulfoxide, quinone, alkene, or N-oxide moieties can be reduced enzymatically or nonenzymatically. Although most of the azo- and nitro-reduction is catalyzed by intestinal microflora, the reduction of aldehydes to primary alcohols, ketones to secondary alcohols, and quinones to hydroquinones are catalyzed by carbonyl reductases found in liver and other tissues. In the liver, carbonyl reductase activity is located both in the cytosol and microsomes. The two forms differ in their stereoselectivity. Many other reduction reactions occur to a minor degree in the liver, but one worthy of mention is the reduction of pyrimidines in the cytosol. Dihydropyrimidine dehydrogenase catalyzes the reduction of pyrimidines such as 5-fluorouracil. This enzyme’s importance was exemplified by a tragic drug combination incident in Japan in 1993. In all, 15 patients on Tegafur, a prodrug that is converted to 5-fluorouracil in the liver, were also given the new antiviral drug Sorivudine for the treatment of herpes zoster. Sorivudine is metabolized by gut flora to an intermediate that in turn is metabolized in the liver to a covalent irreversible inhibitor of dihydropyrimidine dehydrogenase. Inactivation of dihydropyrimidine dehydrogenase resulted in the toxic accumulation of 5-fluorouracil in these patients with a fatal outcome in several cases.

Phase I hydrolysis is catalyzed by multiple enzymes located in the serum and numerous tissues, including the liver. Two groups of enzymes of note are the carboxylesterases and epoxide hydrolases. The carboxylesterases are 60 kDa glycoproteins that are predominantly associated with the ER in the liver. They have a central role in the generation of active metabolites from ester and amide prodrugs used as cancer chemotherapeutic agents. They are also active in the metabolism of many endogenous esterified compounds such as palmitoyl-CoA, retinyl ester, and platelet-activating factor. Epoxide hydrolase catalyzes the trans -addition of water to alkene epoxides and arene oxides, both products of cytochrome P450 metabolism. Arene oxides and alkene epoxides are highly reactive to cellular macromolecules such as DNA and proteins and are consequently potently toxic and frequently mutagenic. Although carboxylesterases and epoxide hydrolases share no sequence homology, they both have a catalytic triad with similar topological structure; hence they are categorized with the α/β-hydrolase fold enzymes.

43.2.3.1

Glucuronidation

The most important conjugation reaction, glucuronidation, is catalyzed by uridine diphosphate glucuronosyltransferases (UGT). These enzymes account for roughly 40% of the phase II conjugation of drugs. They catalyze the transfer of glucuronic acid (from uridine diphosphate glucuronic acid) to aromatic and aliphatic alcohols, carboxylic acids, amines, and free sulfhydryl groups of both endogenous and exogenous compounds to form O -, N -, and S -glucuronides, respectively. Glucuronidation also plays an important role in the elimination of endogenous compounds including bilirubin, bile acids, steroids, and fat-soluble vitamins. Although most phase II reactions take place in cytosol, UGT activity is found in microsomes (ER). This unique localization facilitates their access to phase I products and may account for their dominant role in the phase II metabolism of drugs and endogenous compounds.

Besides the liver, UGTs are found in the intestinal epithelium, kidney, and skin. There are 15 known UGTs classified into one of two families, each having > 50% amino acid identity. Members of family 1 are all encoded by a single complex gene, with many different proteins produced by alternative splicing of 12 promoters and exon 1 with exons 2–5. Family 2 can be further divided into three subfamilies based on additional amino acid homology. Although individual isoforms have characteristic substrate specificities there is a large degree of overlap, and multiple isoforms can catalyze any given glucuronidation reaction.

Examples of diseases resulting from UGT deficiency include Crigler-Najjar syndrome Type I (complete absence of hepatic bilirubin-UGT activity), Crigler-Najjar syndrome Type II (intermediate deficiency of hepatic bilirubin-UGT activity), and Gilberťs syndrome (mild deficiency of hepatic bilirubin-UGT activity). Patients with these conditions do not adequately metabolize bilirubin into the water-soluble form for excretion and consequently experience toxic unconjugated hyperbilirubinemia. The resulting tissue deposition of bilirubin then causes functional impairment of the central nervous system, thyroid, and kidneys. In Type II Crigler-Najjar syndrome, the incomplete deficiency of bilirubin-UGT results is less dramatic elevations of unconjugated bilirubin. In addition, in contrast to Crigler-Najjar syndrome Type I, Type II patients respond to phenobarbital therapy with improved glucuronidation of bilirubin. What was previously ascribed to trophic effects on the ER is likely due to induction of bilirubin-UGT gene expression by phenobarbital. Gilberťs syndrome, although relatively mild in comparison to these conditions, may still result in intermittent elevation of unconjugated bilirubin in conditions of fasting or illness.

43.2.3.2

Sulfation

The next most common conjugation reaction is catalyzed by cytosolic sulfotransferases (STs), accounting for approximately 20% of all phase II metabolic reactions. The STs catalyze the transfer of inorganic sulfur from activated 3′-phosphoadenosine-5′-phosphosulfate to the hydroxyl group of phenols and aliphatic alcohols. Aliphatic alcohols and other primary metabolites or parent drugs with hydroxyl groups are frequently both glucuronidated and sulfated. One prominent example is acetaminophen. After therapeutic doses of this medication approximately 50% is excreted as the phenolic O -glucuronide and 30% as the O -sulfate. Both these reactions are important as suggested by animal models of UGT deficiency, in which there is significantly more susceptibility to acetaminophen hepatotoxicity. In humans, increased susceptibility to acetaminophen-induced hepatotoxicity may occur in Gilberťs syndrome with its relative decrease in bilirubin-UGT activity. Generally, however, glucuronidation and sulfation are able to compensate for each other as there is only one example of glucuronidation or sulfation inhibition resulting in acetaminophen toxicity.

43.2.3.3

Glutathione Conjugation and Acetylation

Glutathione S -transferases (GSTs) and N -acetyltransferases (NATs) together contribute approximately 25% of the phase II metabolic activity. The GSTs are localized to many tissues including the liver where > 95% of activity is localized to the cytoplasm and < 5% to the ER. This enzyme catalyzes the conjugation of hydrophobic electrophiles with the tripeptide GSH (consisting of glycine, cysteine, and glutamic acid), forming thioesters. The role of GST in detoxification of drugs in the liver can again be exemplified by the metabolism of acetaminophen. Although at normal doses most acetaminophen is metabolized via glucuronidation and sulfation, a small amount is oxidized by cytochrome P450 proteins to N -acetyl- p -benzoquinone imine (NAPQI; Fig. 43.2 ). This is highly reactive and can cause hepatocellular necrosis of the centrilobular cells, the primary mechanism of acetaminophen-induced hepatotoxicity ( Fig. 43.3 ). Conjugation of NAPQI with glutathione is the main pathway for the removal of NAPQI, but in the setting of acetaminophen overdose GST becomes saturated, resulting in the toxic accumulation of NAPQI and hepatocellular injury. This NAPQI detoxification mechanism has been challenged by evidence from GST knockout mice. These mice are protected from acetaminophen-induced hepatotoxicity despite normal biotransformation of acetaminophen to NAPQI. In addition, transgenic overexpression of glutathione synthetase and elevated hepatic GSH levels are not protective against acetaminophen-induced hepatotoxicity. Finally, acetaminophen-induced hepatotoxicity was reduced by intravenous administration of glutathione peroxidase, which uses GSH to neutralize reactive oxygen species (ROS). Yet transgenic overexpression of glutathione peroxidase in hepatocytes increased the toxicity of acetaminophen. These animal experiments suggest glutathione conjugation is not the sole mechanism for protection against acetaminophen toxicity.

Pathways of acetaminophen metabolism. Acetaminophen in therapeutic doses undergoes sulfation and glucuronidation (phase II reactions). If these conjugation reactions are overwhelmed, however, cytochrome P450 2E1 (phase I reaction) will convert the acetaminophen to hepatotoxic N -acetyl- p -benzoquinone imine (NAPQI). In situations of sufficient glutathione, glutathione- S -transferase is capable of detoxifying NAPQI to mercapturic acid, preventing NAPQI from covalently binding to hepatocellular proteins which results in cellular death.

Histological consequences of acetaminophen overdose. (A) For comparison, a normal hepatic lobule with a portal tract (P) and hepatic venule (V) highlighted. The liver cells appear uniform and cuboid in single-cell plates lining the sinusoids. Very few inflammatory cells are present and no signs of hepatocyte regeneration exist. (B) Acute acetaminophen overdose with resultant liver failure. Moderately severe centrilobular necrosis (around the hepatic venules) is present but viable hepatocytes are seen in the periportal regions (hematoxylin and eosin, × 145).

The two NAT enzymes (NAT1 and 2) catalyze the acetylation of amines, hydrazines, and sulfonamides but are relatively minor contributors to phase II metabolism. Drugs metabolized by NAT include isoniazid, procainamide, dapsone, hydralazine, and caffeine.

There are multiple allelic variants of NAT and GST gene products that are similar to what is seen with the CYP genes. Some forms differ in catalytic activity, and their expression differs among racial and ethnic groups. The slow-acetylator phenotype is seen in 50%–70% of American Caucasians, African Americans, and Northern Europeans, but in only 5%–10% of Southeast Asians. Poor acetylation has been associated with environmental-agent-induced diseases such as bladder and colorectal cancer. A relationship between environmental-agent-induced disease and genetic variability of phase II conjugation has also been proposed for GST.

43.2.3.4

Methylation

Although thiopurine methyltransferase (TPMT) contributes only a few percent to total phase II metabolism it is important in the metabolism of common medications used in the treatment of autoimmune conditions. The TPMT is a cytosolic enzyme that methylates aromatic and heterocyclic compounds such as the thiopurine drugs azathioprine, 6-mercaptopurine (6-MP), and 6-thioguanine (6-TG). Allelic variants in the TPMT gene are associated with variable enzymatic activity and result in 0.3%, 11.1%, and 88.6% of Caucasians having low, intermediate, or high-enzyme activity, respectively. The TPMT catalyzes the conversion of 6-MP to 6-methylmercaptopurine (6-MMP). With no TPMT activity, 6-MP is preferentially metabolized to thiopurines, increasing the risk of myelosuppression. With high activity, 6-MP is preferentially metabolized to 6-MMP, increasing the risk of hepatotoxicity. Recent evidence suggests that TPMT is inducible by mesalamine, a common concomitant drug used in the treatment of inflammatory bowel disease, with potential for altered thiopurine metabolism.

43.2.4.1

MDR3 and BSEP

MDR3 acts as a flippase enzyme, translocating phosphatidylcholine and other phospholipids from the cytosolic face to the extracellular face of the canalicular membrane where they can be inserted into bile salt micelles for excretion. BSEP is the major canalicular export pump for bile salts in normal liver function, which is an important property of the enterohepatic circulation. Given the reliance on bile salt micelles for the excretion of phospholipids, individuals with mutations in the BSEP gene have bile with a minimal phospholipid concentration and an abnormal bile salt profile. Termed progressive familial intrahepatic cholestasis (PFIC) type 2, these patients have severe cholestasis from birth, low γ-glutamyltransferase activity, and suffer bile salt-induced hepatotoxicity. Individuals with mutations in the MDR3 gene, on the other hand, have bile with normal bile salt concentrations but no phospholipids; the free bile salts are potently cytotoxic. Characterized as PFIC 3, the clinical manifestations are similar to those seen in PFIC 2 except that γ-glutamyltransferase activity is elevated in PFIC 3, and the liver histology reveals a more prominent pattern of ductular proliferation.

43.2.4.2

MRP2

MRP2 is also a canalicular efflux pump but for organic anions. It is additionally expressed in the kidney and intestine, likely contributing to its participation in multidrug resistance. The substrates of MRP2 are mostly products of phase II metabolism, as is also true for the other hepatic ABC proteins. MRP2 contributes to bile formation by transporting glutathione, a driving force in bile salt-independent flow of bile. The regulation of MRP2 gene expression is not entirely clear. Substrates of MRP2 and some phase I and phase II enzymes can induce MRP2 expression in a dose- and time-dependent manner, implying positive feedback on gene transcription. MRP2 gene expression is rapidly downregulated in the presence of endotoxins. Dubin-Johnson syndrome is the human clinical phenotype caused by the absence of the transporter due to MRP2 gene mutation. These individuals are unable to excrete nonbile salt organic anions, including bilirubin mono- and diglucuronides, which results in their reflux back into the circulation. This leads to the elevated serum bilirubin characteristic of this disease, approximately half of which is in conjugated form.

43.2.4.3

MRP1 and MRP3

MRP1 and MRP3 are expressed on the basolateral membrane but only under circumstances of canalicular membrane injury. Specifically, MRP1 is expressed during endotoxin-mediated cholestasis and liver regeneration, and MRP3 is increased with cholestasis and hyperbilirubinemia. Both proteins are upregulated under circumstances that downregulate MRP2. As reverse transporters, they help to reduce the intracellular metabolite concentration under circumstances of canalicular membrane transport dysfunction. Primary substrates include GSH conjugates for MRP1 and monovalent glucuronides for MRP3. Interestingly, patients with Dubin-Johnson syndrome have increased hepatic MRP3, suggesting that MRP3 may serve an important role in organic anion export when MRP2 is downregulated.

The potential effects of exogenous compounds on the activity of these membrane transporters must also be mentioned. Given the diversity of endogenous substrates, it is not surprising that many therapeutic medications may also be excreted by these transporters. Similar to what was described for the CYP proteins, drugs may competitively inhibit the transport of other molecules. This inhibition may occur at either the cis , or cytosolic side, or the trans , or canalicular side of the transport molecule. For example, BSEP is inhibited in vitro from the cis side by cyclosporin A, and from the trans side by estradiol-17β-glucuronide.