Variabilität der hepatischen Glukuronidierung: neue funktionelle Polymorphismen der UDP-Glukuronosyltransferase UGT1A4 : novel functional polymorphisms of the UDP-Glucuronosyltransferase UGT1A4

Aus dem Zentrum Innere Medizin

Abteilung für Gastroenterologie, Hepatologie und Endokrinologie (Direktor: Prof. Dr. med. M.P. Manns)

der

Medizinischen Hochschule Hannover

Variabilität der hepatischen Glukuronidierung:

Neue funktionelle Polymorphismen der UDP- Glukuronosyltransferase UGT1A4

Dissertation

zur Erlangung des Doktorgrades der Medizin an der Medizinischen Hochschule Hannover

vorgelegt von Ursula Ehmer aus Bad Harzburg

Hannover, 2005

Angenommen vom Senat der Medizinischen Hochschule Hannover am 20.9.2006 Gedruckt mit Genehmigung der Medizinischen Hochschule Hannover

Rektor: Prof. Dr. Dieter D. Bitter-Suermann Betreuer: Prof. Dr. Christian. Straßburg

Referent: Prof. Dr. Siegurd Lenzen Korreferent: Priv.-Doz. Dr. Michael Melter

Tag der mündlichen Prüfung: 20.9.2006

Promotionsausschussmitglieder: Prof. Dr. Reinhold Ernst Schmidt Prof.’in Dr. Anke Schwarz Prof.’in Dr. Bettina Wedi

1

Inhalt

1. Inhaltsverzeichnis 1

2. Einleitung 2

3. Dissertation: Variation of Hepatic Glucuronidation: Novel Functional Polymorphisms of the UDP-Glucuronosyltransferase UGT1A4

6

4. Zusammenfassung 14

5. Literaturverzeichnis 17

6. Lebenslauf 20

7. Erklärung 24

2

Einleitung

In seinem Kontakt mit der Außenwelt ist der Mensch einer Vielzahl von Fremdstoffen, sog.

Xenobiotika, ausgesetzt. Insbesondere lipophile Substanzen werden aufgrund ihrer guten Membrangängigkeit leicht vom Körper – sei es über die Haut, die Atemwege oder den Gastrointestinaltrakt – aufgenommen. Eine direkte Elimination dieser Stoffe erfordert jedoch metabolische Modifikationen, da die Exkretion von hydrophoben Substraten über Urin oder Galle nicht effektiv ist (1). Um eine Akkumulation der z.T. deutlich toxischen Substanzen im Organismus zu vermeiden, ist ein Mechanismus, der eine Ausscheidung trotz Lipophilie er- möglicht, von lebensnotwendiger Bedeutung. Welche fatalen Folgen ein Fehlen dieser Exkre- tionsmöglichkeit hat, zeigt eines der klinisch bekanntesten Beispiele, das den Bilirubinmeta- bolismus betrifft: durch Ausfall oder die starke Verminderung der Funktion der UDP- Glukuronosyltransferase UGT1A1 kommt es beim Crigler-Najjar-Syndrom zu einer Unfähig- keit des menschlichen Körpers, das Abbauprodukt des Häm-Stoffwechsels zu glukuronidieren (2), mit konsekutiver Akkumulation von indirektem Bilirubin in den Geweben des Körpers einschließlich des zentralen Nevensystems. Verschiedene Mutationen des UGT1A1-Gens, die zu diesem schweren Krankheitsbild mit Kernikterus und dessen Folgen führen, konnten defi- niert werden (2-4).

Die Biotransformation ist ein in höheren Organismen hochspezialisierter Stoffwechselweg.

Durch die Einführung (Konjugation) hydrophiler chemischer Reste in primär lipidlösliche Verbindungen (Glukuronidierung) ermöglicht er deren effiziente biliäre und renale Exkretion, wobei nicht nur Xenobiotika, sondern auch körpereigene lipophile Substanzen wie das er- wähnte Bilirubin aber auch Steroidhormone der Elimination zugeführt werden können (5-7).

Der Ablauf des Reaktionsweges der Biotransformation erfolgt nach gängiger Definition in zwei Phasen (8): Initial wird in der Phase I ein reaktiver Metabolit durch oxidativen Metabo- lismus erzeugt, der in der Phase II (Konjugation) die Einführung unterschiedlicher stark pola- rer Gruppen in das Molekül gestattet und dieses somit wasserlöslich und für den Körper aus- scheidbar macht.

Eine besondere Bedeutung kommt im Phase II-Stoffwechsel den UDP-Glukuronosyl- transferasen (UGT) zu, die in der als Glukuronidierung bezeichneten Reaktion die Kopplung von Uridindiphoshat-(UDP)-Glukuronsäure an verschiedene Substrate katalysieren (9).

UGT glukuronidieren eine Vielzahl von Substanzen, unter denen sich so unterschiedliche befinden wie z.B. Phenole (10), Flavone, Coumarine, Retinoide, Steroide (11, 12), Bilirubin (13) und Gallensäuren, alkoholische Verbindungen, primäre, sekundäre und tertiäre Amine (14, 15), verschiedene Heterozyklen, polyzyklische aromatische Kohlenwasserstoffe und an- dere Karzinogene (16-18), sowie zahlreiche Arzneistoffe (7, 19). Ein solch weites Substrat- spektrum wird durch verschiedene Enzyme der humanen UGT-Familie gewährleistet, von

3

denen bisher 15 funktionelle Mitglieder identifiziert wurden (7, 20). Diese weisen im einzel- nen eine sehr differentielle Spezifität gegenüber den unterschiedlichen Stoffen auf, werden allerdings auch durch ein überlappendes Substratspektrum gekennzeichnet, was biologisch die Aufgabe erfüllt, eine effiziente Entgiftungsleistung – auch bei Ausfall einzelner UGT- Unterformen – sicherzustellen (7). Diese Redundanz in der Glukuronidierung ist jedoch nicht in allen Fällen gegeben, wie das Beispiel des bereits oben erwähnten Crigler-Najjar-Syndroms mit einem Ausfall der Bilirubinglukuronidierung zeigt.

Die 15 humanen UGT-Enzyme lassen sich nach der Homologie ihrer Gensequenzen in zwei Hauptgruppen gliedern: UGT1A mit acht (UGT1A1, UGT1A3, UGT1A4, UGT1A6, UGT1A7, UGT1A8, UGT1A9 und UGT1A10) und UGT2B mit sechs Subtypen (Isoformen) (UGT2A1, UGT2B4, UGT2B7, UGT2B10, UGT2B11, UGT2B15 und UGT2B17) (7, 21).

Der einzige Vertreter der UGT2A-Familie, UGT2A1, konnte bisher nur im olfaktorischen Epithel und auf geringem Niveau auch im Gehirn und in der fetalen Lunge nachgewiesen werden (22). UGT-Proteine sind, ebenso wie andere Enzyme der Biotransformation (z.B. die der Cytochrom-P450-Familie), nicht nur in der Leber, sondern in gewebespezifischer Vertei- lung auch in zahlreichen extrahepatischen Organen nachweisbar, wobei sich eine besonders hohe Expression in den Epithelien des Gastrointestinaltraktes (23-25), der Lunge und der Nie- ren (UGT1A und UGT2B), sowie in den endokrinen und steroidabhängigen Geweben (UGT2B) finden (7). Hieraus resultiert eine spezifische Glukuronidierungskapazität der ein- zelnen Organe, deren exprimiertes Enzymprofil eine optimal auf spezifische Funktion und anatomische Gegebenheiten des entsprechenden Gewebes zugeschnittene Biotransformation ermöglicht (z.B. vornehmlich Steroid-metabolisierende UGT2B in Prostata, Brustdrüse, Ute- rus (26) und hohe Expression von UGT1A im Gastrointestinaltrakt, die durch ihre hohe Ent- giftungsleistung gegenüber phenolischen und anderen xenobiotischen Substanzen eine Barrie- refunktion des Körpers gegen diese Stoffe darstellen (27, 28)), unabhängig davon, ob es sich bei den Substraten um Karzinogene, Pharmaka o. ä. handelt.

Jedes Gen der in der vorliegenden Arbeit näher charakterisierten UGT1A-Genfamilie besteht aus fünf Exons, wobei nur das erste Exon für den jeweiligen Subtyp spezifisch ist (da es ver- antwortlich für die Substratbindung des Enzyms ist), hingegen sind Exon 2-5 (Glukuronsäu- rebindung und katalytisches Zentrum) bei allen Mitgliedern dieser Genfamilie nach dem Prin- zip des „Exon sharing“ identisch (29). Um die einzelnen UGT1A-Gene zu untersuchen, ist daher besonders das spezifische Exon 1 von großer Bedeutung. Veränderungen in Exon 2-5, die zu Verlust oder Abschwächung der Enzymaktivität führen, würden unweigerlich in dem oben beschriebenen Criggler-Najjar-Syndrom resultieren, da hier auch das für den Bilirubin- abbau essentielle UGT1A1 betroffen wäre. Funktionsänderungen anderer UGT1A-Enzyme führen auf den ersten Blick nicht zu solch gravierenden Folgen, können aber durch ihre Wir- kung auf den Pharmaka-Metabolismus und die Entgiftungsleistung gegenüber Karzinogenen eine entscheidende Bedeutung erlangen.

4

Insbesondere die Glukuronidierungsfähigkeit der Enzyme der UGT1A-Familie gegenüber verschiedenen häufig angewendeten pharmazeutischen Substanzen wie z.B. Morphin (30, 31), Acetaminophen (Paracetamol) (32, 33) und trizyklischen Antidepressiva (15) macht einen bedeutenden Einfluss auf den Arzneimittelmetabolismus mit klinischer Relevanz wahrschein- lich. Ihre Lokalisation in den Organen, die den ersten Kontakt zu oral zugeführten Arzneimit- teln haben (Epithelien des Gastrointestinaltraktes), und in dem primären Entgiftungsorgan des Körpers – der Leber – legt den Schluss nahe, dass sie von nicht unerheblichen Einfluss auf die Pharmakokinetik der von ihnen metabolisierten Substanzen sind.

Die in der vorliegenden Arbeit untersuchten UGT1A3- und UGT1A4-Enzyme sind gerade deshalb besonders interessant, weil sie hepatisch hoch exprimiert werden und damit in hohem Maße an der Biotransformation im Körper zirkulierender Stoffe teilnehmen (im Gegensatz zur der im Vordergrund stehenden Barrierefunktion hauptsächlich epithelial exprimierter UGTs) und weil sie eine große Anzahl pharmazeutisch häufig verwendeter Substanzen gluku- ronidieren (Antidepressiva (15, 34), Neuroleptika wie z.B. Clozapin (35), wie auch eine große Zahl weiterer Pharmaka), ebenso wie zahlreiche Steroide (Dihydrotestosteron, Östrogene (36)).

Zu Beginn der vorliegenden Arbeit waren nur wenige UGT-Varianten bekannt, bereits oben erwähnt wurde das Crigler-Najjar-Syndrom, ein zwar seltenes, aber klinisch um so dramati- scher imponierendes Krankheitsbild. Die ursächlichen Mutationen hierfür können sowohl im Promotorbereich, als auch in jedem der 5 Exons liegen, jeweils mit unterschiedlicher Ausprä- gung der Symptomatik (3). Häufiger tritt der ebenfalls einen Defekt der Bilirubinglukuroni- dierung betreffende Morbus Gilbert auf, der auf einem Polymorphismus in der Promotorregi- on des UGT1A1 beruht (37) und zu einer ca. 70%igen Verminderung der UGT1A1- Enzymaktivität führt. Eine klinische Relevanz wurde lange Zeit nicht gesehen, jedoch gibt es neue Hinweise auf Veränderungen der Suszeptibilität für Pharmaka bei Patienten mit Morbus Gilbert (38).

Die in den letzten Jahren begonnene Suche nach weiteren Polymorphismen brachte interes- sante neue Aspekte im Hinblick auf die zytoprotektive Potenz der UGT, wie die Assoziation von Tumoren des Gastrointestinaltraktes und der Leber mit einer neuen UGT1A7-Variante (UGT1A7*3) zeigt, welche eine im Vergleich zum Wildtyp signifikant verminderte Glucuro- nidierungsaktivität vor allem gegenüber kanzerogenen Hydroxy-Benzo[a]pyrenen aufweist (39-42).

Zu Beginn der vrliegenden Arbeit lag eine vollständige Charakterisierung des UGT1A-Lokus noch nicht vor, doch die bereits vorhandenen Daten (43-46) gaben Hinweise darauf, dass es möglicherweise noch weitere Polymorphismen an diesem Genlokus geben könnte. Im Hin- blick auf die zahlreichen Pharmaka, die durch UGT metabolisiert werden erschien ein Ein- fluss solcher Polymorphismen auf die Pharmakokinetik hochwahrscheinlich, wobei UGT1A3 und UGT1A4 aufgrund ihres breiten Substratspektrums und Expressionmusters besonders

5

interessante Kandidatengene für die weitere Suche nach Polymorphismen waren. Vor allem die Glukuronidierung klinisch relevanter Pharmaka und die mögliche Enzymaktivitätsände- rung bei Vorliegen polymorpher Enzymvarianten könnte klinische Relevanz in Bezug auf die individuellen Unterschiede der Pharmakogenetik gewinnen (47).

Ein Zusammenhang mit Medikamentenunverträglichkeiten und erhöhter Nebenwirkungsrate bei einer Funktionsminderung der UDP-Glukuronosyltransferasen konnte bereits kürzlich gezeigt werden (signifikant vermehrtes Auftreten schwerer Diarrhöen nach Irinotecan-Gabe bei Patienten mit Morbus Gilbert, bei dem eine Verminderung der UGT1A1-Aktivität vorliegt (37, 38, 48, 49)) und stellt ein interessantes Modell für die Erklärung der interindividuellen Variabilität des Arzneimittelmetabolismus dar. Das Feld der Pharmakogenetik beschäftigt sich mit solchen Mechanismen, die zunehmend auch Bedeutung für die Medikamentenzulas- sung erlangen (50, 51). Das volle Ausmaß der Bedeutung von UGT-Polymorphismen für die individuelle Medikamentenverträglichkeit wird erst durch die in letzter Zeit veröffentlichten Daten bekannt und ein weiteres Verständnis erfordert die Analyse des gesamten UGT1A- Genlokus (52). Abschließendes Ziel ist die Erstellung eines individuellen Variantenprofils, das Aussagen über Karzinomdisposition und Arzneimittelverträglichkeiten machen könnte.

Ziel der vorliegenden Arbeit war die Charakterisierung neuer Polymorphismen des UGT1A- Genlokus und eine Untersuchung auf deren mögliche Assoziationen mit Tumoren am Beispiel des Hepatozellulären Karzinoms (HCC), sowie die funktionelle Charakterisierung neuer po- lymorpher Enzymvarianten.

Variation of Hepatic Glucuronidation:

Novel Functional Polymorphisms

of the UDP-Glucuronosyltransferase UGT1A4

Ursula Ehmer,* Arndt Vogel,* Jan Karl Sch ¨utte, Britta Krone, Michael P. Manns, and Christian P. Strassburg

UDP-glucuronosyltransferases are a family of drug metabolizing enzymes contributing to hepatic drug metabolism and protection against environmental toxins. The aim of this study was to identify polymorphisms at the humanUGT1Agene locus and to characterize their function and potential association with hepatocellular carcinoma (HCC). Genomic DNA from the blood of 363 subjects (128 patients with HCC, 235 blood donors) was analyzed for polymorphisms of the UGT1A3, UGT1A4, UGT1A8, UGT1A9, UGT1A10 genes using polymerase chain reaction, sequencing analysis. Recombinant variant UGT protein was analyzed by activity assays. In theUGT1A8gene an A173G variant and a conserved G to A exchange at position 765 were detected in 25% and 15%. UGT1A9 exhibited two variants C3Y and M33T in 1% and 3%.UGT1A10exhibited conserved nucleotide exchanges (128 G3A and 696 C3T) in 2% and 13%. In theUGT1A3gene a W11R, a V47A variant, and a conserved G to A exchange at position 81 with an incidence of 65%, 58%, and 65%, respectively, were identified.UGT1A4exhibited a P24T and an L48V variant in 8% and 9%.

UGT1A SNPs were not associated with HCC. UGT1A4 P24T and L48V exhibited reduced glucuronidation activities:␤-naphthylamine 30% and 50%, and dihydrotestosterone 50%

and 0%, respectively. In conclusion, the high prevalence of SNPs throughout the human UGT1Agene locus illustrates a genetic basis of interindividual variations of hepatic metab- olism. Two polymorphisms of the hepatic UGT1A4 protein show a differential metabolic activity toward mutagenic amines and endogenous steroids, altering hepatic metabolism and detoxification.(HEPATOLOGY2004;39:970 –977.)

G

on chromosome 2 include chemically divergent com- pounds which include bilirubin, steroid hormones, and bile acids, as well as polyaromatic hydrocarbons, hetero- cyclic amines, and drugs such as morphine, antidepres- sants, nonsteroidal antiinflammatory drugs, and anticancer drugs.¹A stable glucuronidation function is ensured by redundancy encoded at the human UGT1A gene locus.² This is the result of the transcription of at least nine functional members of theUGT1Agene family in addition to overlapping substrate specificities of these individual UGTs.³ The example of fatal unconjugated hyperbilirubinemia present in Crigler-Najjar’s syndrome type 1 illustrates the catastrophic effects of a complete absence of UGT1A1 activity toward a single compound, namely, bilirubin.⁴ For the majority of all other UGT substrates differences of glucuronidation are the result of differential transcriptional regulation,^5,6polymorphic in- terindividual expression patterns,^7,8and the combination of single nucleotide polymorphisms (SNP).^{9 –12}SNPs of theUGT1A1gene (bilirubin transferase) are the genetic basis of variable phenotypes of unconjugated hyperbiliru-

Abbreviations: HCC, hepatocellular carcinoma; UGT, UDP-glucuronosyltrans- ferase.

From the Department of Gastroenterology, Hepatology and Endocrinology, Han- nover Medical School, Hannover, Germany.

Received October 13, 2003; accepted January 12, 2004.

Supported by Deutsche Forschungsgemeinschaft SFB621, Deutsche Krebshilfe (to C.P.S.), and the Heisenberg program of the Deutsche Forschungsgemeinschaft (to C.P.S.).

*A.V. and U.E. shared equally in this work.

Address reprint requests to: Christian P. Strassburg, M.D., Associate Professor of Experimental Gastroenterology, Department of Gastroenterology and Hepatology, Hannover Medical School, Carl-Neuberg-Str. 1, 30625 Hannover, Germany. E- mail: strassburg.christian@mh-hannover.de; fax:⫹49-511-532 2093.

Copyright © 2004 by the American Association for the Study of Liver Diseases.

Published online in Wiley InterScience (www.interscience.wiley.com).

DOI 10.1002/hep.20131 970

binemia evident in Gilbert’s disease and Crigler-Najjar’s disease types 1 and 2.⁴These same SNPs also impact the glucuronidation of drugs such as the anticancer drug iri- notecan (SN38)¹⁰and determine a predisposition toward unwanted side effects.¹³Based on these considerations the identification of SNPs at the humanUGT1Agene locus are of great importance for understanding the alterations in human hepatic glucuronidation and elucidating poten- tial genetically determined mechanisms of interindividual differences in metabolism as well as predisposition to drug and xenobiotic associated diseases.

Moreover, recent research has provided evidence that UGTs play a role in cellular protection.^2,7Catalytic activ- ity profiles of isoforms such as UGT1A7,¹⁴ UGT1A8, and UGT1A10^{12,14 –16} include the detoxification of human tobacco smoke-borne mutagens such as benzo(␣)pyrene metabolites and heterocyclic amines.^1,17 In addition to the low activity detoxification allele UGT1A7*3, which has been identified as a cancer risk factor,^9,11,18other UGT gene products could be expected to play a similar role due to their overlapping substrate specificities with UGT1A7.¹ However, one factor that determines the organ tissue-specific profile of glucu- ronidation is a tissue-specific expression profile of UGT1A gene products. The human liver expresses UGT1A1, UGT1A3, UGT1A4, UGT1A6, and UGT1A9.¹⁹In contrast, UGT1A7 is expressed in proxi- mal gastrointestinal epithelia,^19,20while UGT1A8 is ex- pressed primarily in the colon,^3,14and UGT1A10 in all nonhepatic gastrointestinal tissues.^14,19,20 Functional polymorphisms of selectively expressed UGTs have been identified for theUGT1A7andUGT1A8genes.^11,12,21In addition, allelic variants of UGT1A9 and UGT1A10 have been described.^22–24The low catalytic activity poly- morphism UGT1A7*3 has been linked to gastrointestinal cancer of the colon,⁹ the mouth,^18,25 the esophagus,²⁵ stomach,²⁵and the liver.²¹In these cancers epidemiolog- ical observations suggest that cancer risk is modulated by the exposure to environmental mutagens. Typical muta- gens such as benzo(␣)pyrene metabolites and heterocyclic amines such as PhIP are substrates of UGT1A7,^1,14,17il- lustrating the relationship of environmental exposure and genetically determined metabolic predisposition for the development of disease. The UGT1A3, UGT1A4, UGT1A8, UGT1A9,andUGT1A10genes also represent attractive candidate genes for cytoprotection based on the analysis of their detoxification activity toward multiple human mutagens.¹ Polymorphisms of the hepatic UGT1A3 and UGT1A4 genes have not been described to date. The present study analyzes the presence and fre- quencies of SNPs of theUGT1A3, UGT1A4, UGT1A8, UGT1A9,andUGT1A10genes in blood donors and pa-

tients who have developed hepatocellular cancer (HCC) to provide an insight into SNPs of theUGT1Agene fam- ily.

In our study we demonstrate a high frequency of SNPs at the human UGT1A gene locus in addition to the iden- tification of functional polymorphisms of the hepatic UGT1A4 gene leading to differential alterations of xeno- biotic and steroid metabolism.

Patients and Methods

Patients. Blood samples were collected from patients seen in the Department of Gastroenterology, Hepatology and Endocrinology, Hannover Medical School, Ger- many, between 1999 and 2002. All patients were of Ger- man Caucasoid ancestry. Written consent was obtained from all patients prior to blood sampling and the Ethics Committee of Hannover Medical School approved the study protocol.

Two patient groups were studied: As controls, healthy blood donors were studied (n⫽235, female: 98, male:

137, age: 45⫾16 years, range: 19 – 81 years). HCC pa- tients (n ⫽ 129, female: 26, male: 103, age: 61 ⫾ 11 years, range: 22– 84 years) were diagnosed by ultrasound and abdominal computed tomography scans. In each case the diagnosis was cytologically confirmed by ultrasound- guided needle biopsy or by histological examination of resection specimens obtained upon surgical resection of the tumor. None of the HCC patients exhibited other tumors detectable by diagnostic procedures performed for tumor staging, which included abdominal computed to- mography, chest X-ray, and bone radionuclide scans.

Genomic DNA. Full blood samples were used to pre- pare genomic DNA using the QiaAmp isolation system according to the recommendations of the manufacturer (Qiagen, Hilden, Germany). Concentrations were deter- mined by spectrophotometry at 260 and 280 nm. All samples were stored in 10 mM Tris/EDTA buffer (pH 8.0) at –20°C until analysis.

Polymerase Chain Reaction Amplification of UGT1A3, UGT1A4, UGT1A8, UGT1A10 exon 1 se- quences. The humanUGT1Agene locus consists of di- vergent, individually regulated exon 1 sequences at the 5⬘ end followed by one copy of exons 2–5 at the 3⬘end. Each individual UGT1A transcript therefore comprises unique exon 1 sequences combined with exon 2–5 sequences present on all UGT1A transcripts. Therefore, in order to study a specificUGT1Agene, examination of exon 1 se- quence is required.

UGT1A3, UGT1A4, UGT1A8, UGT1A10exon 1 se- quences were amplified by polymerase chain reaction.

The forward and reverse primers were located outside the

HEPATOLOGY, Vol. 39, No. 4, 2004 EHMER ET AL. 971

open reading frame to obtain specificity of the amplifica- tion reaction because exon 1 sequences of UGT1A locus share a very high homology of up to 93%. PCR products were amplified in a volume of 100␮l containing 10 mM KCl, 20 mM Tris-HCl (pH 8.8), 10 mM ammonium sulfate, 2 mM magnesium sulfate, 0.2 mM each dNTP, 500 ng of genomic DNA, 2␮M of primers and 1 unit of Taq DNA polymerase (Promega, Madison, WI). After a hot start at 94°C for 3 minutes, 30 cycles were run on a Perkin Elmer (Norwalk, CT) GeneAmp PCR 2400 sys- tem. Sense and antisense primers and annealing temper- atures are summarized in Table 1.

Sequence Analysis. The PCR products were visual- ized by 2% agarose gel electrophoresis and purified using QiaQuick Columns according to the manufacturer’s rec- ommendations (Qiagen). The sequences of the PCR products were determined by automated fluorescence se- quencing (MWG-Biotech, Ebersbach, Germany). The sequence data was analyzed using the PC Gene software package (Oxford Molecular, Campbell, CA). For se- quencing of UGT1A3, a separate sequencing primer was used to ensure specificity (5⬘-TTACACGTTGATTT- GCTAAGTGGCTC-3⬘).

Expression ofUGT1A4 P24TandUGT1A4 L48V.

Wildtype UGT1A4 cDNA was used to generate a C to A transversion at position 1 of codon 24 and a T to C transversion at position 2 of codon 48 by PCR mutagen- esis as previously described in detail.⁹ The variant UGT1A4 cDNAs were inserted into pCDNA3.1 (In- vitrogen, Karlsruhe, Germany) and pEYFE (BD Bio- sciences, Palo Alto, CA) vectors for expression and localization studies. Transfection was performed using HEK293 cells as published⁹followed by in vitro glucu- ronidation activity assays as previously reported.⁷ The substrates ␤-naphthylamine, benzidine, trans-andros- terone, and dihydrotestosterone were purchased from ICN (Irvine, CA).

Cellular Localization of UGT1A4 P24T and UGT1A4 L48V. HEK 293 cells were seeded to 60%

confluence in 12-well plates containing coverslips pre- treated with 86% potassium hydroxide. Transfection was performed with 0.8␮g DNA per well and after 48 hours cells were fixed with 4% paraformaldehyde, washed with phosphate-buffered saline, and treated with the “Prolong Antifade Kit” (MoBiTec, Go¨ttingen, Germany) accord- ing to the protocol of the manufacturer. Fluorescence was visualized with an Olympus IMT 2 immuno-fluorescence microscope (Tokyo, Japan). Transfection with empty pEYFE and no vector was performed as controls.

Statistical Analysis. The statistical analyses were cal- culated using the two-tailed Fisher’s Exact Test as well as the Epicalc2000 software package to determine odds ra- tios and confidence intervals.

Results

Detection and Prevalence of UGT1A3 Polymor- phisms. Sequence data of exon 1 portions identified three novel single nucleotide polymorphisms (all identified polymorphisms are graphically summarized in Fig. 1): At codon 11 a T to C transversion in position 1 resulting in an amino acid exchange from tryptophan to arginine (W11R) was identified. 2) At codon 27 a conserved nu- cleotide G to A transversion was detected affecting nucle- otide position 81. 3) Codon 47 was found to be characterized by a T to C transversion at position 2 lead- ing to an amino acid change from valine to alanine (V47A). Interestingly, W11R and 81G3A were always found on the same allele and based on our data appear to be in linkage disequilibrium (submitted to GenBank as UGT1A3*3, accession number (acn.) AF465193). In ad- dition, V47A was also found to be present together with the other two SNPs on a single allele (submitted to Gen- Bank as UGT1A3*2, acn. AF465194). The prevalence of Table 1. Primer Sequences Used for the Detection of UGT1A Polymorphisms

Gene Primer Sequence

Annealing Temperature

UGT1A3 F: 5⬘-CCAAAACCACATAGCCAGCCTCCACG-3⬘ 63°C

R: 5⬘-TGGAACATTGATTGGATGAAGGCACC-3⬘

UGT1A4 F: 5⬘-GTTGGGCCCATAACGAAAGGCAGTT-3⬘ 62°C

R: 5⬘-TGGAACATTGATTGGATGAAGGCACC-3⬘

UGT1A8 F: 5⬘-GGTTTTGTGCCTGTAGTTCTTCCG-3⬘ 60°C

R: 5⬘-GCGGATATCCATAGGCACTGGCTTTCCCTGATG-3⬘

UGT1A9 F: 5⬘-CGGCTCGAGACTACTGTATCATAGGAGCTTAGATTCCC-3⬘ 62°C

R: 5⬘-GCGGATATCCATAGGCAACGGCTTTCCCTGATGGCA-3⬘

UGT1A10 F: 5⬘-GGGCTGCAGTTCTCTCATCG-3⬘ 62°C

1^stpart R: 5⬘-TGAGAACCCTAAGAGATCAT-3⬘

UGT1A10 F: 5⬘-CCTCTTTCCTATGTCCCCAATGA-3⬘ 62°C

2^ndpart R: 5⬘-AGGCACTGGCTTTCCCTGATGACAGTTGA-3⬘

972 EHMER ET AL. HEPATOLOGY, April 2004

these polymorphisms was studied in blood donors (n⫽ 81) and HCC patients (n⫽90). W11R, 81G3A, and V47A were detected in 65%, 65%, and 58% of blood donors, respectively, and in 75%, 75%, and 67% of HCC patients, respectively. The prevalence of these polymor- phisms was found to be high and in excess of those iden- tified in the other UGT1A genes. Differences between blood donors and HCC were not statistically significant (Table 2).

Detection and Prevalence of UGT1A4 Polymor- phisms. Polymorphisms were detected at two codons within the UGT1A4 first exon: 1) At codon 24 a C to A transversion at position 1 was observed leading to a pro- line to threonine amino acid change (P24T, submitted to GenBank as UGT1A4*2 acn. AF465196). 2) At codon 48 a T to G transversion at position 1 leading to a leucine to valine amino acid change (L48V, submitted to Gen- Bank as UGT1A4*3 acn. AF465197). Interestingly, in most cases both polymorphisms were present in the same patient, indicating that these two polymorphisms are in linkage disequilibrium. To determine whether UGT1A4 polymorphisms contribute to an increased risk for hepa- tocellular carcinoma we examined the prevalence of UGT1A4 polymorphisms in normal controls (n ⫽83) and cancer patients (n ⫽76) (Table 2). P24T was ob- served in 13 controls (8%), and L48V in 14 (9%). This distribution did not differ significantly from that found in the cancer patients. The cancer patients exhibited the presence of P24T in 10 (7%) and L48V in 15 (10%) of the cases.

Fig. 1. Schematic of the localization of polymorphisms identified at the humanUGT1Agene locus on chromosome 2. Polymorphisms within the unique exon 1 sequences affect individual isoforms, while alterations of the common exons 2 to 5, which are shared by all gene products encoded at the locus, would affect every UGT1A isoform. UGT1A2 is a pseudogene and is not expressed. In previous studies, UGT1A6²⁷and UGT1A7^18,21 polymorphisms have been described. na, nucleic acid;⫹, 36 polymorphisms of the UGT1A1 first exon and exon 2–5 sequences have been described and are linked to nonhemolytic unconjugated hyperbilirubinemia.⁴Sequence data has been deposited in GenBank under the following accession numbers (acn.): UGT1A7*2, AF296126; UGT1A7*3, AF296127; UGT1A7*4, AF296128; UGT1A3*2 AF465193, UGT1A3*3 AF465194, UGT1A4*2, AF465196; UGT1A4*3, AF465197; UGT1A8*2, AF465199; UGT1A8*3, AF 465200.

Table 2. Prevalence ofUGTPolymorphisms in Hepatocellular Carcinoma Patients and Normal Controls

Alleles (n)

Normal Controls

HCC Patients

P Value UGT1A

1A3

11^trp 181 91 (0,56) 90 (0,5)

11^arg 141 71 (0,44) 70 (0,5) n.s.

81^G 181 91 (0,56) 90 (0,5)

81^A 141 71 (0,44) 70 (0,5) n.s.

47^val 215 108 (0,65) 107 (0,59)

47^ala 127 54 (0,35) 73 (0,41) n.s.

1A4

24^pro 294 153 (0,92) 141 (0,93)

n.s.

24^thr 24 13 (0,08) 11 (0,07)

48^leu 286 150 (0,91) 136 (0,90)

n.s.

48^val 30 16 (0,09) 16 (0,10)

1A8

173^ala 298 148 (0,74) 150 (0,75)

n.s.

173^gly 102 52 (0,26) 50 (0,25)

765^G 337 167 (0,84) 170 (0,85)

n.s.

765^A 66 33 (0,16) 30 (0,15)

1A9

3^cys 399 200 (1,00) 199 (0,99)

n.s.

3^tyr 1 0 (0,00) 1 (0,01)

33^met 397 199 (0,99) 198 (0,99)

n.s.

33^thr 3 1 (0,01) 2 (0,01)

1A10

128^G 195 98 (0,98) 97 (0,97)

n.s.

128^A 5 2 (0,02) 3 (0,03)

695^C 171 87 (0,87) 84 (0,84)

n.s.

695^T 29 13 (0,13) 16 (0,16)

Numbers in brackets refer to allelic prevalence. Figures in italics indicate non-coding polymorphisms based on nucleotide sequence numbering.

n.s., not statistically significant.

HEPATOLOGY, Vol. 39, No. 4, 2004 EHMER ET AL. 973

Detection and Prevalence of UGT1A8 Polymor- phisms. Sequence data on 200 patients and controls identified two polymorphisms within the UGT1A8 first exon: 1) At nucleotide position 518 a C to G transition was observed leading to an alanine to glycine amino acid change at codon 173 (A173G) (deposited in GenBank as UGT1A8*2 acn. AF465199). 2) At nucleotide position 765 an A to G transversion was identified leading to a silent mutation at codon 255 (deposited in GenBank as UGT1A8*3 acn. AF465200). A173G was observed in 52 (26%), and 765 A3G in 33 (16%) of the controls, re- spectively. This distribution did not differ significantly from that of the cancer patients. The cancer patients ex- hibited the presence of A173G in 50 (25%) and 765 A3G in 30 (15%) of the cases (Table 2). A single nucle- otide polymorphism at position 277 as reported in a pre- vious publication analyzing 69 samples¹² was not identified in any of the 200 subjects studied here.

Detection and Prevalence of UGT1A9 Polymor- phisms. Sequence analyses of 100 HCC patients and 100 blood donor controls identified two rare coding polymor- phisms. At nucleotide position 8 a G to A transversion led to an amino acid exchange from cysteine to tyrosine at codon 3 (C3Y). This was observed in none of the controls and in one of the HCC patients on one allele (1%). In addition, a T to C transversion at nucleotide position 98 led to an amino acid exchange from methionine to thre- onine at codon 33 (M33T). This SNP was present in one blood donor and in two HCC patients (2%). The D256N SNP detected among Japanese subjects was not detect- ed.²³

Detection and Prevalence of UGT1A10 Polymor- phisms. Sequence analyses of 50 HCC patients and 50 blood donors demonstrated the absence of coding poly- morphisms in all cases. However, two silent polymor- phisms were identified: a G to A transversion at nucleotide position 128, and at nucleotide position 695 a C to T transversion. Both polymorphisms were equally distributed between controls (2% and 13%, respectively), and HCC patients (3% and 16%, respectively) (Table 2).

The previously described SNPs T202I and M59I found in Japanese patients were not detected.²⁴

UGT1A4 P24T and L48V Lead to Altered Cata- lytic Activity. The analysis of recombinant UGT1A4 variants showed a differential profile of glucuronidation activity with xenobiotic (␤-naphthylamine and benzi- dine) and steroid (trans-androsterone and dihydrotestos- terone) substrates compared to wildtype UGT1A4 activities. While both polymorphic variants had decreased activities with all four substrates, UGT1A4 L48V re- tained higher specific activities with the mutagens benzi- dine and␤-napthylamine, and UGT1A4 P24T retained

higher activities with the steroids trans-androsterone and dihydrotestosterone. Dihydrotestosterone activity was completely lost by UGT1A4 L48V. UGT activities are summarized in Table 3.

Cellular Expression of Yellow Fluorescent Protein Fusion Proteins of UGT1A4 P24T and UGT1A4 L48V. Fluorescence microscopy of UGT1A4 variant proteins expressed in HEK293 cells showed a normal pe- rinuclear distribution of the wildtype UGT1A4 protein and the two polymorphic variants. Empty vector controls showed a distribution of yellow fluorescent protein throughout the cell and contrasts that of UGT protein (data not shown). These experiments indicate normal UGT targeting of wildtype as well as polymorphic vari- ants to the endoplasmic reticulum.

Discussion

The identification and characterization of single nucle- otide polymorphisms of major drug-metabolizing en- zymes is crucial for understanding differences in drug metabolism, therapeutic efficacy, and also for the predis- position toward diseases such as cancer. In this study anal- yses of theUGT1A3, UGT1A4, UGT1A8, UGT1A9, and UGT1A10genes were undertaken, which—together with published data on UGT1A1 and UGT1A6 —provides an analysis of the entire humanUGT1Agene locus.12,21,26,27

In addition to analyzing healthy blood donors, a goal of this study was to analyze UGT1A polymorphisms in pa- tients with hepatocellular carcinoma.

The analyses of the human UGT1A3, UGT1A4, UGT1A8, UGT1A9, andUGT1A10genes led to the de- tection of nine SNPs, seven of which were previously not reported. The rationale to analyze these UGT1A genes was based on catalytic activity profiles and tissue distribu- tion patterns. Analysis of theUGT1A8gene was based on its similar catalytic activity profile with UGT1A7. Re- combinant protein analyses have shown that UGT1A8 is capable of glucuronidating complex phenols, heterocyclic amines, anthraquinones, and coumarins.¹ Our study showed two polymorphisms (A173G and 765 A3G). A

Table 3. Differential Activities of UGT1A4 and Its Polymorphic Variants

Substrate

UGT1A4 Wild Type

UGT1A4 P24T

UGT1A4 L48V Activities in pmol/mg/min

␤-naphthylamine 31.8 9.7 18.2

Benzidine 1.9 n.d. 1.3

Trans androsterone 86 54 1.5

Dihydro testosterone 27 18 n.d.

All determinations were performed in triplicate.

n.d., no activity detectable.

974 EHMER ET AL. HEPATOLOGY, April 2004

recent analysis of 69 healthy samples demonstrated three SNPs, the two also found in our cohort and an additional silent SNP at codon 277. The A173G variant was func- tionally characterized, with only minimal glucuronida- tion activity differences.¹²In our analysis SNPs at codon 277 were not detected. The reason for this discrepancy may be the analysis of European Caucasoid patients ver- sus an American collective. This may indicate ethnic dif- ferences in the distribution of UGT1A polymorphisms in humans. Despite the high sequence similarity of UGT1A8 with UGT1A7, a significant association with HCC was not demonstrable. The analysis of UGT1A9 identified two novel coding SNPs. Both UGT1A9 C3Y and M33T were found to be rare and affected only about 1% of studied patients. This low frequency in our Euro- pean cohort is the reason they were not recognized in an earlier study.²¹Finally, theUGT1A10gene was studied, which has been detected to be expressed in biliary epithe- lium and throughout the gastrointestinal tract, with the exception of the liver.^7,19,20UGT1A10 has been charac- terized to glucuronidate heterocyclic amines and aromatic hydrocarbons in addition to steroids and complex phe- nols.¹ No UGT1A10 nucleic acid changes leading to amino acid substitutions were identified. Interestingly, our European cohort did not show polymorphisms de- tected by a Japanese group, further indicating evolution- ary differences regarding the incidence and prevalence of UGT SNPs.^23,24

Of particular interest are SNPs of the hepatic UGT1A3 and UGT1A4 proteins, variants of which have not previously been described or analyzed in detail, but which are of high relevance to hepatic glucuronidation activity.^{28 –31} The humanUGT1A3andUGT1A4genes are primarily expressed in the human liver, both in the hepatocyte and the biliary epithelium, but are absent from proximal gastrointestinal tract tissues: mouth and esoph- agus. UGT1A3 is present in about 20% of gastric tissue samples.^14,19 Compared to other UGT1A proteins UGT1A4 catalytic activity is most specific for primary and secondary amines, steroid hormones, as well as for sapogenins that are commonly present in therapeutic drugs.¹The UGT1A3 protein exhibits a substrate spec- trum including steroids, amines, and phenolics. Our anal- ysis detected five SNPs of the UGT1A3 and UGT1A4 genes: UGT1A3 W11R, UGT1A3 V47A, and a noncod- ing SNP 81G3A, as well as UGT1A4 P24T and UGT1A4 L48V. To our surprise, UGT1A3 SNPs showed a prevalence of 58 – 65% in our cohort, which exceeds the frequency of any of the other UGT1A variants identified to date.²⁶ Neither UGT1A3 nor UGT1A4 SNPs were more frequent in HCC patients. To study the functional relevance of the identified polymorphisms,

UGT1A4 variants were expressed and catalytically char- acterized for their ability to glucuronidate xenobiotics with mutagenic properties and steroids. Using the human carcinogens␤-naphthylamine and benzidine, a differen- tial pattern was seen, leading to a considerable decrease of glucuronidation to 30% and 0% by P24T and a mild decrease by L48V (Table 3). The pattern was reversed when steroid substrates were analyzed. Trans-andros- terone and dihydrotestosterone glucuronidation were al- most absent using the L48V variant, but only reduced to more than 50% of activity using the P24T variant.

This finding is significant since it illustrates that the choice of substrates determines the functional impact of a UGT SNP. This specificity directly influences drug me- tabolism or disease predisposition associated with specific chemical agents. The presence of UGT1A4 L48V would have a greater impact on steroid metabolism than on xe- nobiotic glucuronidation, while UGT1A4 P24T would preferentially affect mutagen metabolism, making the UGT1A4 gene an interesting candidate risk gene. The data obtained from this analysis raises two interesting points regarding hepatic metabolism and genetic predis- position: 1) SNPs of UGT proteins do not result in an overall alteration of specific glucuronidation activity, and 2) UGT SNP alter the specific catalytic profile of individ- ual UGT1A isoforms. This leads to the conclusion that every SNP of a drug-metabolizing enzyme requires indi- vidual and complete analysis to elucidate its potential im- pact on drug, xenobiotic, or endobiotic metabolism. It also raises the interesting question whether the previously defined UGT1A (wildtype) isoforms and their catalytic activity profiles are indeed biologically strictly relevant.

Characterizations have been undertaken with the pre- sumed wildtype enzymes.¹⁴Our analysis now shows that polymorphic genetic variants, some of which occur at surprisingly high allelic frequencies (Table 2), have an altered glucuronidation activity profile which is clearly distinct from the wildtype isoform. Although by defini- tion this does not constitute a novel isoform for lack of sequence disparity exceeding about 5%, it does lead to functionally distinct enzymes.³² Applied to the medical management of patients, our data illustrates a genetic ba- sis for a multitude of potential combinations leading to highly individualized metabolic specificities between in- dividuals and the often-observed individual unforeseeable reactions to drugs.^33–35In addition to the demonstrated alterations of UGT1A4 variants, functional variants have also been identified for UGT1A1 and UGT1A7, expand- ing the scope of this observation to a very high number of possible combinations between different UGT1A SNPS in an individual person. These considerations provide ar- guments for the fact that UGT1A polymorphisms have

HEPATOLOGY, Vol. 39, No. 4, 2004 EHMER ET AL. 975

been retained throughout the course of evolution and provide the basis for human hepatic variability of metab- olism by glucuronidation. In our study we were not able to demonstrate an association with hepatocellular carci- noma, as previously shown for UGT1A7 SNPs. However, the elucidated catalytic variations of UGT1A4 are likely to impact steroid-dependent disease entities^36,37and fur- ther studies will be required to analyze this potential as- sociation.

Combined, the present analysis of all major UGT1A genes and the catalytic characterization of the novel he- patic UGT1A4 variants is remarkable, since it demon- strates that polymorphisms of the humanUGT1A gene locus on chromosome 2 are surprisingly frequent (Fig. 1) and exceed those detected for other enzyme systems, in- cluding the cytochrome P450 supergene family. In view of their frequency, most individuals can be expected to carry a number of the identified variations. This possibil- ity has recently been discussed with respect to UGT1A1 and UGT1A7 polymorphisms and the irinotecan metab- olite SN38, which are frequently detected in hu- mans.^10,13,33

In summary, our analysis demonstrates genetic vari- ability and functional consequences for hepatic glucu- ronidation illustrated by the identification of a functionally distinct P24T and a L48V variant of the human hepatic UGT1A4 protein in addition to a surpris- ingly high number of SNPs of the humanUGT1Agene locus. These data are likely to represent the genetic basis for the considerable spectrum of interindividual varia- tions of hepatic glucuronidation in humans.

References

1. Tukey RH, Strassburg CP. Human UDP-glucuronosyltransferases: me- tabolism, expression, and disease. Annu Rev Pharmacol Toxicol 2000;40:

581– 616.

2. Tukey RH, Strassburg CP. Genetic multiplicity of the human UDP-glu- curonosyltransferases and regulation in the gastrointestinal tract. Mol Pharmacol 2001;59:405– 414.

3. Strassburg CP, Nguyen N, Manns MP, Tukey RH. UDP-glucuronosyl- transferase activity in human liver and colon. Gastroenterology 1999;116:

149 –160.

4. Strassburg CP, Manns MP. Jaundice, genes and promoters. J Hepatol 2000;33:476 – 479.

5. Strassburg CP, Manns MP, Tukey RH. Differential down-regulation of the UDP-glucuronosyltransferase 1A locus is an early event in human liver and biliary cancer. Cancer Res 1997;57:2979 –2985.

6. Strassburg CP, Strassburg A, Kneip S, Barut A, Tukey RH, Rodeck B, Manns MP. Developmental aspects of human hepatic drug glucuronida- tion in young children and adults. Gut 2002;50:259 –265.

7. Strassburg CP, Kneip S, Topp J, Obermayer-Straub P, Barut A, Tukey RH, Manns MP. Polymorphic gene expression and interindividual varia- tion of UDP-glucuronosyltransferase activity in human small intestine.

J Biol Chem 2000;275:36164 –36171.

8. Strassburg CP, Nguyen N, Manns MP, Tukey RH. Polymorphic expres- sion of the UDP-glucuronosyltransferase UGT1A gene locus in human gastric epithelium. Mol Pharmacol 1998;54:647– 654.

9. Strassburg CP, Vogel A, Kneip S, Tukey RH, Manns MP. Polymorphisms of the UDP-glucuronosyltransferase (UGT) 1A7 gene in colorectal cancer.

Gut 2002;50:851– 856.

10. Gagne JF, Montminy V, Belanger P, Journault K, Gaucher G, Guillemette C. Common human UGT1A polymorphisms and the altered metabolism of irinotecan active metabolite 7-ethyl-10-hydroxycamptothecin (SN-38).

Mol Pharmacol 2002;62:608 – 617.

11. Guillemette C, Ritter JK, Auyeung DJ, Kessler FK, Housman DE. Struc- tural heterogeneity at the UDP-glucuronosyltransferase 1 locus: functional consequences of three novel missense mutations in the human UGT1A7 gene. Pharmacogenetics 2000;10:629 – 644.

12. Huang YH, Galijatovic A, Nguyen N, Geske D, Beaton D, Green J, Green M, et al. Identification and functional characterization of UDP-glucurono- syltransferasesUGT1A8*1, UGT1A8*2andUGT1A8*3.Pharmacogenet- ics 2002;12:287–297.

13. Tukey RH, Strassburg CP, Mackenzie PI. Pharmacogenomics of human UDP-glucuronosyltransferases and irinotecan toxicity. Mol Pharmacol 2002;62:446 – 450.

14. Strassburg CP, Manns MP, Tukey RH. Expression of the UDP-glucu- ronosyltransferase 1A locus in human colon. Identification and character- ization of the novel extrahepatic UGT1A8. J Biol Chem 1998;273:8719 – 8726.

15. Cheng Z, Radominska-Pandya A, Tephly TR. Cloning and expression of human UDP-glucuronosyltransferase (UGT) 1A8. Arch Biochem Biophys 1998;356:301–305.

16. Mojarrabi B, Mackenzie PI. Characterization of two UDP glucuronosyl- transferases that are predominantly expressed in human colon. Biochem Biophys Res Commun 1998;247:704 –709.

17. Nowell SA, Massengill JS, Williams S, Radominska-Pandya A, Tephly TR, Cheng Z, Strassburg CP, et al. Glucuronidation of 2-hydroxyamino-1- methyl-6-phenylimidazo[4,5-b]pyridine by human microsomal UDP-glu- curonosyltransferases: identification of specific UGT1A family isoforms involved. Carcinogenesis 1999;20:1107–1114.

18. Zheng Z, Park JY, Guillemette C, Schantz SP, Lazarus P. Tobacco carcin- ogen-detoxifying enzyme UGT1A7 and its association with orolaryngeal cancer risk. J Natl Cancer Inst 2001;93:1411–1418.

19. Strassburg CP, Oldhafer K, Manns MP, Tukey RH. Differential expres- sion of the UGT1A locus in human liver, biliary, and gastric tissue: iden- tification of UGT1A7 and UGT1A10 transcripts in extrahepatic tissue.

Mol Pharmacol 1997;52:212–220.

20. Strassburg CP, Strassburg A, Nguyen N, Li Q, Manns MP, Tukey RH.

Regulation and function of family 1 and family 2 UDP-glucuronosyltrans- ferase genes (UGT1A, UGT2B) in human oesophagus. Biochem J 1999;

338:489 – 498.

21. Vogel A, Kneip S, Barut S, Ehmer U, Tukey RH, Manns MP, Strassburg CP. Genetic link of hepatocellular carcinoma with polymorphisms of the UDP-Glucuronosyltransferase UGT1A7 gene. Gastroenterology 2001;

121:1136 –1144.

22. Villeneuve L, Girard H, Fortier LC, Gagne JF, C G. Novel functional polymorphisms in the UGT1A7 and UGT1A9 glucuronidating enzymes in Caucasian and African-American subjects and their impact on the me- tabolism of 7-ethyl-10-hydroxycamptothecin and flavopiridol anticancer drugs. J Pharmacol Exp Ther 2003;307:117–128.

23. Jinno H, Saeki M, Saito Y, Tanaka-Kagawa T, Hanioka N, Sai K, Kaniwa N, et al. Functional characterization of human UDP-glucuronosyltrans- ferase 1A9 variant, D256N, found in Japanese cancer patients. J Pharmacol Exp Ther 2003;306:688 – 693.

24. Jinno H, Saeki M, Tanaka-Kagawa T, Hanioka N, Saito Y, Ozawa S, Ando M, et al. Functional characterization of wild-type and variant (T202I and M59I) human UDP-glucuronosyltransferase 1A10. Drug Metabol Dispos 2003;31:528 –531.

25. Vogel A, Ockenga J, Ehmer U, Barut A, Kramer F-J, Tukey RH, Manns MP, et al. Polymorphisms of the carcinogen detoxifyingUDP-glucurono- syltransferase UGT1A7in proximal digestive tract cancer. Z Gastroenterol 2002;40:497–502.

976 EHMER ET AL. HEPATOLOGY, April 2004

26. Mackenzie PI, Miners JO, McKinnon RA. Polymorphisms in UDP glu- curonosyltransferase genes: functional consequences and clinical relevance.

Clin Chem Lab Med 2000;38:889 – 892.

27. Ciotti M, Marrone A, Potter C, Owens IS. Genetic polymorphism in the human UGT1A6 (planar phenol) UDP-glucuronosyltransferase: pharma- cological implications. Pharmacogenetics 1997;7:485– 495.

28. Gall WE, Zawada G, Mojarrabi B, Tephly TR, Green MD, Coffman BL, Mackenzie PI, et al. Differential glucuronidation of bile acids, androgens and estrogens by human UGT1A3 and 2B7. J Steroid Biochem Mol Biol 1999;70:101–108.

29. Green MD, King CD, Mojarrabi B, Mackenzie PI, Tephly TR. Glucu- ronidation of amines and other xenobiotics catalyzed by expressed human UDP-glucuronosyltransferase 1A3. Drug Metab Dispos 1998;26:

507–512.

30. Cheng Z, Rios GR, King CD, Coffman BL, Green MD, Mojarrabi B, Mackenzie PI, et al. Glucuronidation of catechol estrogens by expressed human UDP-glucuronosyltransferases (UGTs) 1A1, 1A3, and 2B7. Toxi- col Sci 1998;45:52–57.

31. Mojarrabi B, Butler R, Mackenzie PI. cDNA cloning and characterization of the human UDP glucuronosyltransferase, UGT1A3. Biochem Biophys Res Commun 1996;225:785–790.

32. Mackenzie PI, Owens IS, Burchell B, Bock KW, Bairoch A, Belanger A, Fournel-Gigleux S, et al. The UDP glycosyltransferase gene superfamily:

recommended nomenclature update based on evolutionary divergence.

Pharmacogenetics 1997;7:255–269.

33. Ando M, Ando Y, Sekido Y, Shimokata K, Hasegawa Y. Genetic poly- morphisms of the UDP-glucuronosyltransferase 1A7 gene and irinote- can toxicity in Japanese cancer patients. Jpn J Cancer Res 2002;93:

591–597.

34. Burchell B. Genetic variation of human UDP-glucuronosyltransferase: im- plications in disease and drug glucuronidation. Am J Pharmacogenomics 2003;3:37–52.

35. Peters WH, te Morsche RH, Roelofs HM. Combined polymorphisms in UDP-glucuronosyltransferases 1A1 and 1A6: implications for patients with Gilbert’s syndrome. J Hepatol 2003;38:3– 8.

36. Belanger A, Hum DW, Beaulieu M, Levesque E, Guillemette C, Tchernof A, Belanger G, et al. Characterization and regulation of UDP-glucurono- syltransferases in steroid target tissues. J Steroid Biochem Mol Biol 1998;

65:301–310.

37. Schroder FH. Antiandrogens as monotherapy for prostate cancer. Eur Urol 1998;34(Suppl 30:12–17.

HEPATOLOGY, Vol. 39, No. 4, 2004 EHMER ET AL. 977

14

Zusammenfassung

Die Glukuronidierung durch UDP-Glukuronosyltransferasen (UGT) repräsentiert einen der bedeutensten Stoffwechselwege zur Elimination unterschiedlichster exogener und endogener Substanzen aus dem menschlichen Körper.

Die renale oder biliäre Ausscheidung lipidlöslicher Verbindungen wird erst durch die Reakti- onen der Biotransformation möglich, welche durch Oxidation, Reduktion oder Hydrolyse (Phase I) und anschließende Konjugation mit polaren Molekülen (Phase II) die Umwandlung lipophiler in hydrophile Stoffe katalysiert. UGT sind Enzyme der Phase II der Biotransforma- tion, die durch Kopplung verschiedener Substrate mit UDP-Glukuronsäure diese für den Kör- per ausscheidbar machen. Zu dem weiten Spektrum der auf diese Weise glukuronidierten Substanzen zählen endogene Steroide, Bilirubin und Gallensäuren sowie eine große Anzahl an Xenobiotika, darunter karzinogene Amine und polyzyklische Kohlenwasserstoffe, aber auch zahlreiche Pharmaka (z.B. Morphine, Antidepressiva und nichtsteroidale Antiphlogisti- ka). Die hohe Effizienz dieses Metabolisierungsweges und Breite des Substratspektrums wird durch unterschiedliche UGT-Unterformen gewährleistet, die sich in ihrer Glukuronidierungs- fähigkeit gegenüber den verschiedenen Substanzen zum Teil deutlich unterscheiden. Die hier untersuchten UGT1A-Enzyme finden sich (außer in der Leber) vor allem in den Epithelien des Gastrointestinaltraktes, wo sie durch ihre Entgiftungsleistung einen protektiven Faktor gegenüber kanzerogenen Stoffen darstellen können. Die charakteristische Verteilung der UGT-Enzyme – einige Subtypen kommen zum Beispiel nicht in der Leber vor – resultiert in sehr spezifischen Metabolisierungsspotentialen der einzelnen Organe. Von dem auf Chromo- som 2 lokalisierten UGT1A-Genlokus sind neun verschiedene Subtypen bekannt (UGT1A1, UGT1A3, UGT1A4, UGT1A6, UGT1A7, UGT1A8, UGT1A9 und UGT1A10), die sich jeweils durch ein individuelles Exon1 (verantwortlich für die Substratbindung) unterscheiden, wäh- rend Exon2-5 bei allen UGT1A identisch sind. Genpolymorphismen, die nur eine UGT1A- Unterform betreffen, müssen daher im Bereich des spezifischen Exon1 liegen. Mehrere sol- cher „single nucleotide polymorphisms“ (SNPs), die zum Teil zu erheblichen Funktionsmin- derungen der polymorphen Enzymvarianten führen, sind in letzter Zeit charakterisiert worden.

Einer der klinisch bekanntesten UGT-Enzymdefekte betrifft die Bilirubinglukuronidierung:

durch Ausfall oder starke Verminderung der UGT1A1-Funktion kommt es beim Crigler- Najjar-Syndrom zu einer Unfähigkeit, das Abbauprodukt des Häm-Stoffwechsels zu glukuro- nidieren. Verschiedene Mutationen, die zu dem schweren Krankheitsbild mit Kernikterus und dessen Folgen führen, sind bekannt. Funktionelle Polymorphismen der anderen UGT1A- Enzyme haben auf dem ersten Blick nicht so gravierende Auswirkungen, jedoch können sie möglicherweise Veränderungen im endogenen Hormonhaushalt (Steroide), im Arzneimittel- stoffwechsel oder in der Resistenz des Organismus gegenüber kanzerogenen Stoffen nach sich ziehen.

Ziel dieser Arbeit war die Suche nach neuen Polymorphismen der UGT1A-Familie sowie der Nachweis einer möglichen Assoziation mit dem hepatozellulären Karzinom (Untersuchung