ACTAUNIVERSITATIS
UPSALIENSISUPPSALA
2012
Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 805
Clinical Pharmacogenetics ofOlanzapine
with Focus on FMO Gene Polymorphisms
MAO MAO SÖDERBERG
ISSN 1651-6206ISBN 978-91-554-8454-5urn:nbn:se:uu:diva-179957
Dissertation presented at Uppsala University to be publicly examined in Enghoffsalen,Akademiska sjukhuset, ingång 50, bv, Uppsala, Monday, October 15, 2012 at 09:30 for thedegree of Doctor of Philosophy. The examination will be conducted in Swedish.
AbstractMao Söderberg, M. 2012. Clinical Pharmacogenetics of Olanzapine: with Focus onFMO Gene Polymorphisms. Acta Universitatis Upsaliensis. Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty of Medicine 805. 81 pp. Uppsala.ISBN 978-91-554-8454-5.
Pharmacogenetics is the study of variability in drug response attributed to genetic variation.Olanzapine (OLA) is a widely used antipsychotic drug for schizophrenia treatment. Thepharmacokinetics of OLA display large inter-individual variation leading to multiple-folddifferences in drug exposure between patients at a given dose. This variation in turn gives rise tothe need of individualized dosing in order to avoid concentration-dependent adverse effects andtherapeutic failure. The observed variability has been partially explained by environmental andphysiological factors. Genetically determined differences in drug metabolism represent a lessstudied source of variability. Precluded contribution by cytochrome P450 (CYP) 2D6 calls forevaluation of the other major OLA metabolizing enzymes. The objective of this thesis was tostudy pharmacogenetic influence of flavin-containing monooxygenase (FMO) 1 and 3, CYP1A2and uridine diphosphate-glucuronosyltransferase (UGT) 1A4 on therapeutic OLA exposure. Weconducted genetic association studies applying gene re-sequencing and genotyping of candidateand tagging SNPs.
Patients carrying the FMO1*6 allele displayed increased dose-adjusted concentrations (C/Ds) of OLA, in serum as well as cerebrospinal fluid. Patients who were homozygous for theFMO3 K158-G308 compound variant showed reduced C/Ds of OLA N-oxide metabolite, butno alteration in OLA exposure. This compound variant is expected to have clinical relevanceprimarily for non-African populations, since low frequencies were detected among nativeAfricans. Deviation in OLA exposure was observed in carrier of a rare FMO3 mutation,predicted in silico to affect gene splicing. Reduced OLA exposure was observed in UGT1A4*3carriers. The CYP1A2 -163(A) (CYP1A2*1F) variant was not associated with increase inCYP1A2-catalyzed OLA metabolism or reduction in OLA exposure. Correlations were detectedfor two cis-acting variants within the inter-genetic region of the CYP1A cluster and a trans-actingvariant located upstream the locus encoding aryl hydrocarbon receptor. The inconsistent datareported for CYP1A2*1F could be explained by presence of ethnic specific haplotype structuresincorporating the -163(A) variant.
A continuously improved understanding of the wide range of factors that can influencepharmacokinetics and pharmacodynamics will increase the likelihood of achieving optimaltreatment response for individual patients.
Keywords: olanzapine, pharmacogenetics, drug metabolism, schizophrenia, therapeutic drugmonitoring, FMO1, FMO3, CYP1A2, UGT1A4
Mao Mao Söderberg, Uppsala University, Department of Medical Sciences, Clinicalpharmacogenomics and osteoporosis, Akademiska sjukhuset, SE-751 85 Uppsala, Sweden.
© Mao Mao Söderberg 2012
ISSN 1651-6206ISBN 978-91-554-8454-5urn:nbn:se:uu:diva-179957 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-179957)
Grant me the serenity to accept the things I
cannot change, courage to change the
things I can, and wisdom to know
the difference.
-Reinhold Niebuhr
To my mother
List of Papers
This thesis is based on the following papers, which are referred to in the text by their Roman numerals.
I Mao, M., Matimba, A., Scordo, M.G., Gunes, A., Zengil, H.,
Yasui-Furukori, N., Masimirembwa, C., Dahl, M-L. (2009) Flavin-containing monooxygenase 3 polymorphisms in 13 eth-nic populations from Europe, East Asia and sub-Saharan Africa: frequency and linkage analysis. Pharmacogenomics, 10(9):1447–55
II Mao, M., Skogh, E., Scordo, M.G., Dahl, M-L. (2012) Interin-dividual variation in olanzapine concentration influenced by UGT1A4 L48V polymorphism in serum and upstream FMO polymorphisms in cerebrospinal fluid. Journal of Clinical Psy-chopharmacology, 32(2):287-9
III Mao, M., Haslemo, T., Molden, E., Dahl, M-L. (2012) Influ-ence of FMO1 and 3 polymorphisms on serum olanzapine and its N-oxide metabolite in psychiatric patients. Submitted to The Pharmacogenomics Journal (under revision)
IV Mao, M., Haslemo, T., Molden, E., Dahl, M-L. (2012) Influ-ence of CYP1A1/CYP1A2 and AHR polymorphisms on sys-temic olanzapine exposure. Submitted to Pharmacogenetics and Genomics (under review)
Reprints were made with permission from the respective publishers. N.B. The authors Mao Mao Söderberg and Mao Mao refer to the same per-son after name change due to marriage.
Contents
Introduction ................................................................................................... 13 Metabolism as determinant in drug response ........................................... 13 Concept of pharmacogenetics .................................................................. 14 From phenotype to genotype .................................................................... 15
Debrisoquine and CYP2D6 ................................................................. 15 Trimethylamine and FMOs .................................................................. 17
Genetics .................................................................................................... 19 Genetic polymorphisms ....................................................................... 19 Genetic association studies .................................................................. 20
Antipsychotics for schizophrenia ............................................................. 22 Two generations of drugs .................................................................... 22 The need for treatment individualization ............................................. 23
Olanzapine, an atypical antipsychotic ...................................................... 24 PD and PK ........................................................................................... 24 Concentration-effect relationship ........................................................ 25
Metabolic pathways of olanzapine ........................................................... 25 N-glucuronidation and UGT1A4/2B10 ............................................... 27 N-demethylation and CYP1A2 ............................................................ 27 Hydroxylation and CYP2D6 ................................................................ 27 N-oxidation and FMOs ........................................................................ 28
Inter-individual variation in olanzapine metabolism ................................ 28 Non-genetic sources ............................................................................. 28 Pharmacogenetics of olanzapine metabolizing enzymes ..................... 30
Aim of the thesis ........................................................................................... 34
Material and Methods ................................................................................... 35 Subjects .................................................................................................... 35 Methodological overview ......................................................................... 37
Analysis of drug and metabolite concentrations .................................. 37 DNA extraction .................................................................................... 37 SNP genotyping ................................................................................... 38 Gene re-sequencing ............................................................................. 39 Bioinformatics ..................................................................................... 40
Statistics ................................................................................................... 41
Results and Discussion ................................................................................. 43 Paper I ...................................................................................................... 43
Inter- and intra-ethnic variation in FMO3 SNP distribution ................ 43 The FMO3 K158-G308 compound variant prevalent among non-Africans ............................................................................................... 44
Paper II. .................................................................................................... 46 FMO and UGT1A4 SNP detection and haplotype construction .......... 46 Serum concentration variation and UGT1A4*3 ................................... 46 CSF concentration variation and upstream FMO SNPs ...................... 48 Elevated OLA exposure in the carrier of a novel FMO3 mutation ...... 48
Paper III. ................................................................................................... 50 Quantification of OLA N-oxide ........................................................... 50 Reduced serum OLA N-oxide concentrations in the homozygous FMO3 G308 carriers ............................................................................ 51 FMO1 SNPs associated with increased serum OLA concentrations ... 52
Paper IV ................................................................................................... 53 No influence by CYP1A2*1F on OLA disposition in vivo.................. 54 Decreased systemic OLA exposure in carriers of a Caucasian specific CYP1A1/1A2 haplotype ....................................................................... 54 Indication of trans-acting effect on OLA desmethylation by a SNP upstream AHR ...................................................................................... 57
General discussion ........................................................................................ 58 Relevance of the findings ......................................................................... 58 Obstacles and limitations ......................................................................... 59
Conclusions ................................................................................................... 62
Acknowledgements ....................................................................................... 64
References ..................................................................................................... 66
Abbreviations
AHR aryl hydrocarbon receptor ANCOVA analysis of covariance ANOVA analysis of variance C/D dose-adjusted concentration CSF cerebrospinal fluid CYP cytochrome P450 monooxygenase DMO 4'-desmethyl olanzapine DNA deoxyribonucleic acid EPS extra pyramidal symptom FGA first-generation (or typical) antipsychotics FMO flavin-containing monooxygenase FMO1 flavin-containing monooxygenase isoform 1 FMO3 flavin-containing monooxygenase isoform 3 HWE Hardy-Weinberg equilibrium LD linkage disequilibrium MAF minor allele frequency OLA olanzapine PD pharmacodynamics PK pharmacokinetics SGA second-generation (or atypical) antipsychotics SNP single nucleotide polymorphism TDM therapeutic drug monitoring UGT uridine diphosphate-glucuronosyltransferase
13
Introduction
Metabolism as determinant in drug response Drug response is a result of interaction between an administrated drug and its target site (such as receptors and enzymes) in the body. Two concepts are central while studying this drug-target interaction, pharmacodynamics (PD) and pharmacokinetics (PK). The former describes the mechanisms of drug action, i.e. interactions with drug target(s) in the body and the events elicited by these interactions, whereas the latter displays drug disposition in the body including absorption, distribution, metabolism and excretion (Fig 1).
Sufficient drug exposure at target site is a prerequisite for successful pharmacotherapy. Lipophilicity of a drug enables passage through biological membranes, e.g. the blood-brain-barrier for psychotropic agents before reaching the target sites in the brain. Prior to final excretion from the body, lipophilic drugs need to undergo, metabolism, structural modification to become more water-soluble. Compared to the parent compound, the products generated under metabolic processes are commonly inactive, but may also be pharmacologically active (e.g. morphine as metabolite of codeine), show altered property (e.g. salicylic acid lacking the antiplatelet activity of aspi-rin), or even induce toxicity (e.g. hepatotoxicity of paracetamol due to its alkylating metabolite).1
Biochemical reactions during metabolism are often classified as Phase I (oxidation, reduction and hydrolysis) and Phase II (conjugation with polar endogenous molecules) reactions (Fig 1). They occur mainly in the liver but also in other organs such as intestine, lungs and kidney. Although Phase I reaction usually precedes Phase II reactions, it is not necessarily so in all cases. The cytochrome P450 monooxygenase (CYP) system is most exten-sively studied for its role in Phase I reactions and the enzyme family of uridine diphosphate-glucuronosyltransferases (UGT) in Phase II reactions. However, not all metabolic reactions involve CYPs and UGTs, examples of other enzyme groups are the flavin-containing monooxygenase (FMO) sys-tem and monoamine oxidases for Phase I reactions and N-acetyltransferases and glutathione S-transferases for Phase II reactions.1
For drugs undergoing extensive metabolism, the activity of the enzymes involved will influence the systemic drug exposure. For individual treat-ments, systemic exposure that exceeds or falls below the therapeutic window will lead to suboptimal drug response or induce adverse effects in patients.
14
Metabolic capacity varies in a population and is influenced by factors of various origins, i.e. environmental (e.g. smoking habits, diet), physiological (gender, age), epigenetic (histone modification and DNA methylation) and genetic (missense mutation, gene deletion and duplications). The observation of inherited traits in drug metabolism was the first example showing the importance of genetics for drug response, a field known today as “pharma-cogenetics”.2
Figure 1. An overview of drug response and the major processes involved. PD, pharmacodynamics; PK, pharmacokinetics
Concept of pharmacogenetics Pharmacogenetic studies aim to identify and quantify the effect of genetic markers that are associated with or are direct causes of the observed drug response differences between individuals, upon receiving the same treat-ment. In the long run, a continuously improved understanding on the wide range of factors, that can influence PK and PD, will increase the likelihood of achieving optimal treatment response for individual patients.
Inherited differences in individuals’ capacity to conduct chemical trans-formations were understood long before the discovery of the DNA double helix. In his book “Inborn error of metabolism” from 1909, Sir Archibald Garrod, an English physician, described alcaptonuria as a result of “chemical individuality” and concluded that this was due to inherited conditions.3 Be-ing ahead of his time, Garrod also proposed the idea that toxic manifestation is caused by deviation from normal metabolism due to overproduction, an incomplete chain of catalytic reactions or formation of abnormal product.4 In 1950s, Arno Motulsky refined these ideas on the role of genetics in drug
[ Drug ]
Drug
Body
Distribution
Metabolism- Phase I reaction- Phase II reaction
Excretion[ Drug +Target]
Absorption therapeuticeffect
adverseeffect
15
response5 and the term “pharmacogenetics” was first introduced by Friedrich Vogel6. Genetic influence on metabolic traits can be demonstrated with stud-ies in twins or family pedigrees. To study variability in a population at large, frequency distribution of a phenotypic parameter (e.g. a metabolic ratio, see Fig 2) in a population of biologically unrelated individuals can be plotted. If multimodal distribution of the data is obtained, individuals with deviating response can be identified at the far right or left of the frequency distribution (Fig 2).With the emergence of molecular genetics and decades of clinical observations on inherited differences in drug response, pharmacogenetics became a recognized science in the period spanning from the1950s to the 1990s.7
From phenotype to genotype Early pharmacogenetic research was characterized by phenotype-driven assessment of variation in drug metabolizing enzymes. Discovery of glu-cose-6-phosphate dehydrogenase deficiency was made in American black soldiers suffering from primaquine-induced hemolysis.8 Prolonged apnea and paralysis observed in some individuals treated with the muscle relaxant suxamethonium resulted in detection of serum cholinesterase deficiency.9,10 Polymorphic acetylation was demonstrated by bimodal frequency distribu-tion of plasma isoniazid concentrations among treated subjects, and slow acetylators were determined to be carriers of an autosomal recessive trait.11
Debrisoquine and CYP2D6 First indication of polymorphic metabolism by CYPs was reported in the late 60s for nortriptyline and imipramine, later known as substrates of CYP2D6.12 The observed difference in steady-state plasma concentrations between two groups of patients after a fixed oral dose was subsequently showed to be genetic in nature.13 In the mid-1970s, the urinary ratio of the antihypertensive drug debrisoquine to its 4-hydroxy metabolite, termed “metabolic ratio” (Fig 2) was found bimodally distributed in Caucasian populations.14,15.The subjects with highest metabolic ratios, above the anti-mode of 12.6, were classified as poor metabolizers (PM, Fig 2) of debriso-quine and displayed intensified effect of the drug. The majority of individu-als, with metabolic ratios below the antimode, were classified as extensive metabolizers (EM, Fig 2). Subsequent studies revealed that the PMs of de-brisoquine were also deficient in metabolizing other compounds such as sparteine, metoprolol, codeine, tricyclic antidepressants and dextromethor-phan.16-18
Prior to elucidation of the underlying molecular mechanisms, these com-pounds, all being substrates of CYP2D6, were used as “probe drugs” to clas-
16
sify subjects regarding their metabolic capacity, so called phenotyping. The metabolic ratio, based on concentration measurement in urinary or plasma samples, was used to assess absolute or partial deficiency of the enzyme. The genetic basis of the two metabolizer phenotypes was revealed 10-15 years after the initial clinical observations, with cloning of the human CYP2D6 cDNA19 and characterization of the altered restriction fragment length polymorphism (RFLP) patterns in PMs20-22. In the 1990s, multiplica-tion of the CYP2D6 genes was detected in individuals with extremely high metabolic capacity, termed ultra-rapid metabolizers (UM, Fig 2).23,24 Identi-fication of gene variants associated with various levels of CYP2D6 activities enabled prediction of the phenotype based on genetic constitution of an indi-vidual.25,26 The inter-individual variation in CYP2D6 activity was also found to differ between populations.27,28 Caucasians metabolize debrisoquine on average faster than Orientals do, while the incidence of PM is higher among Caucasians than in other major ethnicities.29
Figure 2. Concept illustration of metabolic ratio, its distribution in two different populations and three categories of metabolizers; UM= ultra-rapid metabolizer, EM=extensive metabolizer, PM=poor metabolizer. The average level of metabolic activity may differ between populations (Population 2 > Population 1). (Adapted from Bertilsson et al. 1992 28)
Metabolic ratio = parent drug / metabolite
Num
ber
of s
ubje
cts
Population 1
Population 2
EM
Cutoff
EM
UM PM
17
Trimethylamine and FMOs Trimethylamine (TMA), a volatile tertiary amine, is naturally present at high concentrations in marine fishes, but can also be produced through intestinal bacterial degradation from dietary choline.30 Trimethylamine was early shown to undergo hepatic oxidation that forms the N-oxide metabolite prior to urinary excretion.31,32 Case reports on deficiency in converting the malo-dorous trimethylamine (TMA) to its odorless N-oxide have been docu-mented since 1970.33 This rare condition, termed trimethylaminuria (TMAuria), was shown to be genetic in nature.34 As direct elimination of TMA in urine, sweat and breath gives the characteristic fishy odor, TMAuria is also known as the fish-odor syndrome. TMA N-oxidation ability was sub-sequently characterized to be polymorphic in a British population.35 The finding has been verified in various populations of different ethnic origins, e.g. Caucasian Canadians36, Oriental groups as Thai37 and Chinese38, and also in Jordanian, Ecuadorian and New Guinean populations39. Reduced, but not absent, capacity was recognized in a subject if proportion of TMA ex-creted as its N-oxide fell within the range of 50 - 80%, compared to more than 90% normally excreted . Occurrence of mild deficiency was reported to vary from 0.6% up to 11% among the studied populations.
Prior to the 1960s, oxidation of the NADPH-dependent heteroatom-containing compounds was thought to be mediated essentially by micro-somal CYPs. FMO was first described in 1964, a novel monooxygenase containing flavin adenine dinucleotide (FAD) as redox cofactor, not heme as CYPs do.40 But like CYPs, FMO also utilizes molecular oxygen and the reducing agent NADPH while catalyzing oxidation. After its purification from pig liver in 197141, identification and characterization of multiple forms of FMO were carried out from the mid 1980s to early 1990s42-46. Of the five active human FMO isoforms (FMO1 – 5) identified (Table 1)47,48, FMO3 was the major hepatic form49,50 and also exhibited a clear substrate prefer-ence for tertiary amines51. In 1997, defective FMO3 was established as un-derlying cause of TMAuria with identification of nonsense and missense mutations in affected subjects.52-54
Although FMOs are capable of metabolizing an exceptionally wide range of xenobiotics with very little structural features in common, they are con-sidered to play a minor role compared to CYPs in drug metabolism.55,56 This view is rooted in the problems experienced while studying FMOs. FMOs and CYPs share a number of overlapping characteristics including expres-sion pattern in tissues and cells, substrate specificity as well as type of cata-lytic reaction.57 Specific inhibitors of FMOs are not available. Together with thermal instability of FMO activity shown in absence of NADPH58 and the frequent inter-conversion between the parent amines and the N-oxide me-tabolites59, these features have contributed to the difficulty and uncertainty in interpreting the relative contribution of FMO from CYP in drug metabolism.
18
Less therapeutic compounds are known FMO substrates compared to those of CYPs.56 The FMO-mediated oxidation of nitrogen (N) or sulfur (S) is seldom recognized as the predominant metabolic reaction for a substrate.56 Phenotyping for a given FMO is therefore difficult to perform.
Table 1. Features of human FMO enzyme family
Isoform Major sites of expression Clinical relevance
FMO1 Kidney, intestine, liver (fetal)
Alteration in expression associated with amyotro-phic lateral sclerosis60,61 Exaggerated drug response in FMO1 knockout mouse62 SNPs as risk factors for nicotine dependence63
FMO2 Lung Functional protein expressed mainly in Africans Ethnic difference in response to tuberculosis treat-ment64,65
FMO3 Liver (adult) Deleterious variants cause trimethylaminuria66,67 FMO4 low expression levels in all
studied tissuesPoorly understood
FMO5 Liver (adult), intestine Poorly understood
However, FMOs do possess features that are distinct from CYPs and desir-able for drug development.57.The enzymes are not known to be as easily induced by xenobiotics as the CYPs are. FMOs are unusual for not having the substrate binding as the rate-limiting step in its catalytic cycle. Due to the stability of the oxidant, i.e. 4a-hydroperoxyflavin of FAD, the enzyme is present in an activated form. FMO operates on full speed once its substrate has gained access to the catalytic site.55,58 In general, FMOs produce inactive and water-soluble metabolites, but a few exceptions to this have also been reported.55,56 Among the known substrates, many are active in the central nervous system (CNS), including chlorpromazine, clozapine, OLA, perazine, imipramine, fluoxetine, xanomeline, amphetamine, methamphetamine and nicotine.56
Results from in vitro studies support the observation of large variability between individuals, at expression levels, for several FMOs.68,69 Expression of FMO1 is highest in adult kidney (47±9 pmol/mg protein), and is detected to a less extent also in intestines.70 The expression level of FMO3 in adult liver (60±43 pmol/mg protein) is comparable to that of CYP3A4 and CYP2C.68 On the other hand, similar gene expression levels of all five FMO isoforms have been reported in human brain tissue.49 FMO activity has also been characterized in mammalian brain microsomes including human71-73 and detected in various brain regions in animal studies.74,75 FMO1 gene dele-tion in mice treated with imipramine, a FMO1 isoform-specific substrate, was shown to result in elevated systemic concentrations of the parent com-pound and increased cerebral levels of CYP-mediated metabolite, desip-ramine.62 Exaggerated adverse pharmacologic response, as body tremor and
19
spasm, was also observed in the knockout mice.62 On the contrary, the seda-tive effect of imipramine was abolished in the knockout mice, suggesting a pharmacological role of the N-oxide metabolite.62
Genetics Genetic polymorphisms The hereditary information of any human being is coded by 3.2 billion nu-cleotides base pairs and stored on 23 chromosome pairs.76 Although only monozygotic twins have completely identical genomes, the difference is not large between individuals (estimated 99.9% identical between two ge-nomes).77 Still, this small fraction contributes to the observed genetic herita-bility among individuals. DNA sequence changes, when located in sequence elements of functional relevance, can alter expression, translocation or func-tionality of the encoded gene products, proteins, which are essential for all biological process.76
When a DNA mutation occurs with a frequency greater than 1% in the general population, it is a genetic polymorphism. Categories of polymor-phism include single nucleotide polymorphism (SNP) as variation of a single base pair at a fixed position (Fig 3), insertion/deletion, duplication, inversion and copy number variation. SNP is shown to be an exceedingly common form of polymorphism and observed once every 1000-2000 base pairs in the genome.78 For CYP2D6 described above, genetic variants causing the wide spectrum of enzyme activity include, in PMs, SNPs causing frame shift, SNPs affecting gene splicing, insertion, and deletion of the whole gene whereas gene duplications and multi-duplications are identified in UMs.29 The frequencies of these alleles vary between populations, explaining the previously reported inter-ethnic differences in CYP2D6 activity.29
20
Figure 3. (A) One individual has two alleles (one paternal and one maternal) at each autosomal locus. DNA sequence for each allele is shown base-paired. Variation at one nucleotide locus when comparing allele 1 with allele 2 is a SNP. (B) Haplotypes are adjacent SNPs inherited together and they can be identified by only assessing the tagSNPs located within.
Genetic association studies The SNPs that influence drug response and clinical outcome make up only a tiny fraction of all the SNPs existing in the human genome (Fig 4). To iden-tify this group of SNPs without having to assess all the SNPs, patterns of genetic structure are applied in designing genetic association studies. When a group of adjacent SNPs on the same chromosome tend to be inherited to-gether, they form a haplotype (Fig 3). The size of a haplotype depends on the extent of recombination that have occurred between the loci.76 Non-random association between SNPs is commonly observed in the human genome and is expressed by degree of linkage disequilibrium (LD).79 The LD structure of one specific genomic region varies between regions as well as ethnic popula-tions.79 Measure of LD is under the assumption of Hardy-Weinberg equilib-rium (HWE). For a SNP in HWE, it states a stable distribution of genotypes (AA, Aa and aa) and their constant association with frequencies of the two alleles, p and q(=1-p) (AA=p2, Aa=2pq and aa=q2) through generations within a large population.80 The correlation assumes random mating and lack of genetic influences including natural selection, mutations and random drift.80 Assessment of HWE services also as a quality control for potential genotyping errors.
--------
--------
--------
--------
--------
--------
--------
--------
T A A G T A C C C G A A G T A
A T T C A
A
T G G G C T T C A T
T A A G C
T
A C C T G A A G G A
A T T C G
A
T G G A C T T C C T
Allele 1
Allele 2
T
G A A G G
G G G T A
G A G G A
C G G G G
G C
A C
G C
A T
C G G G GA C
Haplotype 1
Haplotype 2
Haplotype 3
Haplotype 4
Haplotype 5
tagSNPs
SNP SNP SNP(A)
(B)
21
Figure 4. Hierarchy of SNPs based on function and clinical relevance.
LD is assessed in its simplest form between two SNP loci and expressed by two parameters, D’ and r2, with different properties. D’ shows the prob-ability for historical recombination between the loci in a given population whereas r2 is used to determine the sample size required to detect association between a trait and the causal locus by using an associated SNP.81 Two SNP loci are in complete LD with each other if r2 =1. They are totally independent of each other when r2=0.81 The lower the value of r2 is, the larger the sample size that is required for association studies. Instead of assessing individual markers, a subset of the associated SNPs, called tagSNPs (Fig 3), can be selected to represent other SNPs nearby that are in strong LD with them. Compared to single marker analysis, a small number of tagSNPs can capture haplotype variation within a larger genomic region and increase statistical power.82
Pharmacogenetic studies can be designed as hypothesis generating or hy-
pothesis testing. Hypothesis generating studies, genome-wide association studies (GWAS), are assumption free and assess genetic variants on genome scale, with aim to detect novel correlations to clinical parameters of interest. In hypothesis testing studies, candidate gene association studies, one or more gene(s) and genetic variant(s) are selected with presumed involvement in the predefined endpoint, based on an acceptable level of biological under-standing (i.e. annotated functionality or previous implicated association). For both study designs, the ability to detect variant-endpoint association is influ-enced by the factors described in above sections including LD, sample size, SNP frequency and the effect size of the gene product in generating the drug response.
22
Antipsychotics for schizophrenia The term schizophrenia was coined around 1911 by Eugen Bleuler (1857-1939), a Swiss psychiatrist, and defined with four main symptom descrip-tions: affective disturbance, autism, associative disturbances and ambiva-lence (Bleuler’s 4A).83 They are now considered so called negative symp-toms, one of two categories of symptoms often used to describe schizophre-nia today. The other category is so called positive symptoms, resembling manifestations of psychosis, i.e. hallucinations and delusions.83 Current di-agnostic criteria for schizophrenia are described thoroughly in The Diagnos-tic and Statistic Manual of Mental Disorders 4th Edition (DSM-IV-TR).84
Schizophrenia has near 1% prevalence with an early onset in life (teens, early adulthood). Sustained recovery is reported in less than 14% of patients within the first five years after a psychotic episode.85,86 The disorder is often chronic with high risk for relapse, more frequently observed in patients with poor compliance than in those showing good drug adherence.87 It impairs social functioning and reduces lifespan of affected patients, as much as 15-25 years.85
Two generations of drugs Antipsychotics, the cornerstone of modern treatment for schizophrenia, started with the introduction of chlorpromazine (Fig 5) in the early 1950s. As the first antipsychotic medication, it revolutionized pharmacotherapy of schizophrenia by greatly improving quality of life for patients and contribut-ing to a decrease in institutionalization, the then standard treatment. This was followed by intensive research and development of new structurally related compounds over the past 60 years (Table 2). Treatment with chlorpromazine and its successors, however, was complicated by high rates of neurologic side effects of involuntary movements, such as tardive dyskinesia and Park-insonism, broadly termed extra pyramidal symptoms (EPS).88,89
Figure 5. Chemical structures of chlorpromazine and clozapine, milestones in drug treatment and drug development for schizophrenia
A new line of research started with the development of clozapine (Fig 5) in the 1960s, an effective compound with low propensity for EPS. Instead clozapine was challenged with a different side effect, agranulocytosis, caus-
Chlorpromazine Clozapine
23
ing acute decrease in white blood cell account, which restricted its clinical use. Clozapine was re-introduced in the 1990s as a result of documented superiority over other antipsychotic agents in treatment resistant patients and markedly reduced risk for the potentially fatal idiosyncratic side effect through safety routine of monitoring white blood cells. It too became a pro-totype for development of a new class of antipsychotics, often termed as second-generation or atypical antipsychotics (SGA), the earlier ones being referred to as first-generation or typical antipsychotics (FGA) (Table 2). Over time, FGAs have been gradually replaced by the growing number of SGAs.88,89
Table 2. Two classes of antipsychotic drugs
First generation antipsychotics (typical)
1950s Chlorpromazine1960s Haloperidol Fluphenazine Perphenazine Trifluoperazine Loxapine 1970s Molindone Pimozide
Second generation antipsychotics (atypical)
1990s Clozapine Risperidone Olanzapine Quetiapine2000s Ziprasidone Aripiprazole Paliperidone Iloperidone Asenapine Lurasidone
The need for treatment individualization Antipsychotics are generally effective in treating positive symptoms and reduce risk of relapse, but are weaker in managing negative symptoms.89 Outcome of schizophrenia treatment, however, is highly variable and indi-vidual, with 30-40% of patients showing no initial response.90 Results from large pragmatic non-industry-sponsored clinical trials (CATIE, CUtLASS, EUFEST studies) in recent years failed to provide clear support for SGAs being more effective than FGAs.90-92 Two exceptions were observed. The first was clozapine having increased efficacy in patient who were resistant to other treatments and the other was olanzapine (OLA), a successor of clozap-ine, having the lowest rate in treatment discontinuation.90 On the other hand, a meta-analysis based on 150 double-blind studies concluded better overall efficacy with small to medium effect size and lower incidence of EPS for four SGAs (clozapine, OLA, amisulpride and risperidone) compared to FGAs.93
Although the superiority of SGAs in drug efficacy remains a topic of de-bate, views on their side-effect profile are rather aligned. Instead of having EPS as the primary concern, metabolic disturbance (i.e. weight gain, hyper-lipidemia, and diabetes) are commonly seen in patients treated with SGAs.94,95 Having cardiovascular disease as long term consequence, the
24
metabolic adverse effects may thus increase both morbidity and mortality for schizophrenia.94,95 Treatment with clozapine and OLA are burdened with highest incidence/risk for developing metabolic disturbances.90-92
Results from clinical trial and meta-analyses define an average level of observed treatment outcome in each patient population studied but generally overlook individual differences between patients. To achieve better clinical use of available drugs, it is desirable to establish sub-group specific treat-ments and doses and, thus, identify determinants of individual variation in drug disposition and response.
Olanzapine, an atypical antipsychotic Olanzapine (OLA), (2-methyl-4-(4-methyl-1-piperazinyl)-10H-thieno[2,3-b] [1,5]benzodiazepine) is a SGA (atypical) compound launched as Zyprexa by Eli Lilly and Company (Indianapolis, USA) in 1996 96 (Fig 6) . It is now a commonly prescribed antipsychotic for acute and maintenance treatment of schizophrenia with estimated sales of 4.6 billion dollar in total at the end of 2011 (Lilly 2011 Annual Report, http://investor.lilly.com/annuals.cfm). OLA resembles the precursor compound, clozapine, both structurally and pharma-cologically, but is not associated with agranulocytosis.97,98
PD and PK Meta-analysis studies assessing dose-response correlation in OLA treatment have suggested a near-maximal effective dose to be >16mg daily.99 Addi-tional therapeutic advantage in dosing >20mg daily is reported only for treatment resistant or markedly ill patients, not apparent in non-treatment-resistant patients with diagnose chronic schizophrenia.100 In a clinical study with fixed-doses, increased adverse events were reported for a dose at 40mg daily.101
Through neuroimaging techniques such as PET (positron emission tomo-graphy), OLA has shown binding affinity for a range of neuron receptors, dopamine, serotonin, cholinergic muscarinic, α1-adrenergic and histamine receptors. Although the precise mechanism of drug action is still unclear for antipsychotic drugs, blockade of dopamine- (D2) receptors is a shared feature and believed to control positive symptoms of schizophrenia.102 A range of 60-80% striatal D2 receptors occupancy is regarded as optimal to gain satis-factory response and reduce risk for EPS.103 In patients which have reached steady-state, OLA induced 43%-80% D2 blockade within the recommended clinical dose range of 5-20mg daily, 83%-88% within 30-40mg daily as well as a near complete occupancy of serotonin receptors at low dose of 5mg daily.104
25
OLA is well absorbed orally as > 65% of the dose is detected in urine and fecal samples.105 No data on absolute oral bioavailability based on intrave-nous administration is available. When given orally, plasma concentration increases proportionally with dose, (also known as linear or first-order kinet-ics) throughout the clinical dosage range.106 Pharmacokinetic characteristics of OLA at steady-state are reported to be consistent with those seen after a single dose.106 The maximum plasma concentration is reached within 5-8 hours and the average time required for the plasma concentration to fall by half (half-life, t1/2) is estimated to 33 hours (90% within 21-54 hours) in adult healthy individuals.106 The steady-state concentration can thus be reached in about a week.106
Concentration-effect relationship Based on the 60-80% D2 receptor occupancy, a plasma concentration inter-val between 20 to 40ng/mL for 12-hour post-dose sampling has been esti-mated as an optimal therapeutic range for OLA.104,107 A threshold at 23ng/mL 10-16 hours post dose (or 9ng/mL predose or trough) has been suggested as the lower end of the therapeutic window.107,108 However, the predictive strength of the breakpoint of 23ng/mL is only moderate and even at a given serum concentration, the antipsychotic effect is largely individual. High median plasma OLA concentrations have been shown in patients with adverse effects compared to those without.101,109
Drug concentrations in the central nervous system would be expected to be more closely related to therapeutic effects of centrally acting drugs than serum concentrations are. A strong linear correlation between serum and cerebrospinal fluid (CSF) concentrations of OLA has been reported in pa-tients on long-term treatment with OLA.110 However, a significantly faster decline of serum OLA concentration has been observed as compared to re-duction of D2 receptor occupancy in healthy volunteers, suggesting dissocia-tion in kinetic profile between the two compartments.111
Metabolic pathways of olanzapine Metabolite characterization in urine samples identified 9 compounds.105 Di-rect glucuronidation is the primary route of OLA biotransformation and also a to humans unique metabolic pathway.112,113 OLA also undergoes hepatic oxidative metabolism through N-demethylation, N-oxidation and 2-alkyl hydroxylation105 (Fig 6). The metabolites identified are considerably less active than the parent compound in vivo.114 Approximately 50-60% of an orally administrated dose is eliminated via urine and about 30% via feces.105
26
F
igur
e 6.
Met
abol
ic p
athw
ays
of o
lanz
apin
e in
hum
ans
(ada
pted
fro
m K
assa
hun
et a
l., 1
997
105 )
Ola
nza
pin
e
Glu
c
Ola
nza
pin
e-10
-N-g
lucu
roni
de
Glu
c
Ola
nzap
ine
-4’-N
-glu
curo
nid
e
Ola
nza
pin
eN
-oxi
de
N-d
esm
ethy
lola
nzap
ine
Glu
c
Ola
nza
pin
eN
-oxi
de-2
-ca
rbox
y g
lucu
roni
de
N-d
esm
ethy
l-2-c
arbo
xy
olan
zapi
ne
2-ca
rbox
y ol
anz
apin
e
Glu
c
2-ca
rbo
xy o
lanz
apin
eg
lucu
roni
de
2-h
ydro
xym
ethy
l ola
nza
pin
e
UGT
1A4
UGT
1A4,
U
GT2B
10FM
O
CYP2
D6
CYP1
A2
27
N-glucuronidation and UGT1A4/2B10 The attachment of glucuronic acid to the target compound is catalyzed by UGTs. The subfamilies 1A and 2B are most important for drug metabolism. Two N-glucuronide conjugates have been identified, the tertiary OLA 10-N-glucuronide (exists in two isomers) being the main circulating metabolite in plasma and the quaternary OLA 4’-N-glucuronide produced in minor amounts105,113 (Fig 6). In humans, N-glucuronidation is a major route of me-tabolism for tertiary amines, including many clinically used psychoactive drugs such as tricyclic antidepressants and antipsychotics, and usually result-ing in the formation of quaternary N-glucuronides.115 OLA is an exception, additionally having a rare tertiary N-glucuronide metabolite.113 By quantify-ing the metabolites detected in urine and fecal samples, about 21-25% of a single dose is estimated to be eliminated as OLA 10-N-glucuronide.105 This metabolite is thus a quantitatively more important N-glucuronide.
To date, all known human UGT1A and 2B enzymes except UGT1A5 and 2B28 have been screened for their ability to glucuronidate OLA.116,117 The observed catalytic activity was first attributed to UGT1A4 and later also to UGT2B10, an orphan isoenzyme with high N-glucuronation activity discov-ered in 2007.116-118 However, UGT2B10 is shown to be most active in the formation of the minor OLA 4’-N-glucuronide, in contrast to UGT1A4 pro-ducing both N-glucuronides.117
N-demethylation and CYP1A2 The formation of N-desmethyl OLA (DMO) was best correlated with the catalytic activity of CYP1A2 in vitro (Fig 6), although small amounts formed by CYP3A4 and CYP2D6 were also detected.119 The plasma ratio of DMO to OLA is significantly correlated to OLA clearance (r2=0.35, P<0.0002).106 Hepatic CYP1A2 activity can be estimated in vivo using caf-feine as probe drug given that this enzyme catalyzes more than 90% of its systemic clearance.120-122 Indeed, significant correlations have been reported between caffeine metabolic ratios and OLA clearance in healthy volunteers123, as well as between caffeine metabolic ratios and dose-adjusted concentrations (C/Ds) of plasma OLA in psychiatric patients124 (Table 3).
Hydroxylation and CYP2D6 The formation of 2-hydroxy OLA was attributed to the polymorphic CYP2D6 activity in vitro (Fig 6). CYP2D6 has earlier been shown to be of importance for the metabolism of most FGAs.125 However, it appears to be of minor importance for OLA disposition in vivo (Table 3). CYP2D6 PMs
28
showed no deviating pharmacokinetic characteristics compared to individu-als displaying normal CYP2D6 activity110,124,126 (Table 3).
N-oxidation and FMOs The formation of OLA N-oxide is primarily a product of FMO activity (Fig 6).119 Limited formation at low rate has also been shown for CYP450 en-zymes such as CYP3A4, 1A2, 2D6, 2E1 and 2C9.119 Although the catalytic activity was correlated to immunoquantified levels of the FMO isoform 3 (FMO3) in human liver microsome as well as to the FMO3-mediated forma-tion of nicotine N-oxide in vitro119, OLA is not known to be an isoform-specific substrate for FMO3 over other known FMO isoenzymes. The role of FMO in OLA disposition is insufficiently studied, and warrants in vivo evaluation.
Inter-individual variation in olanzapine metabolism Elimination of OLA has been reported to vary nearly 10-fold within studied populations and exhibits larger variability between individuals than within.106,127 Data from various TDM (therapeutic drug monitoring service) studies revealed larger than 25 fold difference in plasma OLA C/Ds.109,128-130 Inter-individual variation in OLA exposure is attributed to individual charac-teristics of the patients.
Non-genetic sources Smoking Schizophrenic patients are prone to smoke, a behavior significantly more prevalent than in the general population.131 This is of clinical significance as smoking is consistently identified as strong predictor of reduced OLA expo-sure in vivo, independent of dose levels and other factors.109,127,130,132-136 This has been attributed to increased metabolic clearance following enzyme in-duction caused by polycyclic aromatic hydrocarbons in tobacco smoke.137 As much as 50% lower median plasma OLA C/D as well as higher prescribed dose has been reported for smokers while compared to non-smokers.109,132,133 Population kinetic modeling has estimated that 26% of observed variability in elimination of OLA can be explained by smoking.127 Adverse clinical outcomes related to increased drug concentrations following smoking cessa-tion have also been documented.138
29
Gender OLA clearance is lower in females than in males, a difference estimated to account for 12% of the overall variability by population kinetic model-ing.106,127 Women display higher median/mean OLA C/Ds than men at a given dose (30% to 60%), a difference remaining significant after adjustment for body-weight and/or smoking status.109,130,132,139 Hence, the female non-smoker group is predisposed to elevated OLA exposure whereas the male smoker group to suboptimal treatment. Oral contraceptives have been as-sessed as a potential underlying factor for the observed gender difference. They showed no clinically relevant influence on serum OLA concentrations despite an inhibitory effect on CYP1A2-mediated OLA N-demethylation by ethinyl estradiol.140
Concomitant medication Inhibitors or inducers of CYP1A2 or UGT1A4 can cause significant, though not necessarily clinically relevant, changes in OLA exposure in vivo. Co-administration of the CYP1A2-inhibitor, fluvoxamine resulted in 2-3 fold increase in plasma OLA C/Ds.129,141 Carbamazapine is known as a strong inducer of CYP1A2. Data from TDM studies reveal about 40 to 70% reduc-tion in median OLA C/Ds in carbamazapine treated patients compared to patients on OLA monotherapy.109,128,142 Lamotrigine may exert inhibitory effect on OLA metabolism as both undergo UGT1A4-mediated glucuronida-tion. However, impact on OLA pharmacokinetics has been inconsistent, from no effect over the range of 50mg to 200mg lamotrigine daily in healthy volunteers143, significant but mild (16%) increase in mean plasma OLA con-centration at 200mg daily in patients144, to a finding of 35% increase in plasma OLA C/Ds observed in smoking patients145. Significant decrease in plasma OLA C/Ds has been reported with co-administration of valproic acid.146,147 Studies based on TDM data, however, have so far shown no influ-ence of valproic acid on plasma OLA.129,132
Age, weight and ethnicity Age and body weight have been identified as significant factors for plasma OLA concentrations. However, the contribution of age and body weight to increased OLA exposure has proven to be small in the overall concentration variability.109,129,130,132 In two recent studies, increased clearance and reduced plasma concentrations have been reported for patients of African origin compared to other major ethnic groups.127,135
30
Pharmacogenetics of olanzapine metabolizing enzymes UGT1A4/2B10 Both UGT1A4 and 2B10 are highly abundant in human liver but not de-tected in brain tissue.148-151 The UGT1A subfamily is encoded by a single gene locus on chromosome 2q37 with multiple first exons but shared exon 2-5. Of the 13 first exons exist in man, each is spliced to exon 2-5 to produces a unique UGT1A gene/pseudogene. UGT1A4 is one of the nine functional UGT1A isoforms.
Of the ten non-synonymous amino acid changes reported for UGT1A4 (www.pharmacogenomics.pha.ulaval.ca/cms/ugt_alleles/), two common polymorphisms, p.P24T and p.L48V, are most studied. Although, differen-tial effect on in-vitro-glucuronidation have been reported for the two variants depending on substrates tested, increased efficiency is most often seen with the V48 variant and reduced/no activity alteration by the T24 variant.117,152-
156 In addition, upstream promoter variants of UGT1A4 (g.-219T, g.-204A and g.-163A) have been shown to modulate transcriptional activity.155,157 In clinical studies with OLA treated patients, UGT1A4 has been included as candidate gene in few studies in which only the V48 variant was assessed (Table 3). Significant correlation with reduced serum OLA concentrations is reported in Caucasians, but not in Japanese patients.134,136,158
UGT2B family is encoded by separated genes located on chromosome 4q13. The UGT2B10 gene consists of 6 exons separated by 5 introns. Few non-synonymous changes are reported with one missense variant, p.D67Y identified to confer reduced level of N-glucuronides of several substrates including OLA117,159 (Table 3).
CYP1A2 About 13% of total hepatic CYP content can be accounted by CYP1A2.160 Hepatic mRNA level of CYP1A2 is reported to vary 40-fold among indi-viduals.161 The encoded gene, CYP1A2, together with CYP1A1 which is not constitutively expressed in human liver161, forms the CYP1A cluster on hu-man chromosome 15q24.1162. The two genes are located in the head-to-head orientation and thus share a common region scattered with regulatory ele-ments.162,163 The CYP1A2 gene comprises 7 exons separated by 6 introns.164 Although numerous polymorphisms are reported for CYP1A2 (http://www.cypalleles.ki.se/cyp1a2.htm), only a few of them have been targets of extensive studies and displayed considerable variation in popula-tion distribution.165
31
Table 3. Data summary on variability in olanzapine pharmacokinetics influ-enced by genotype and phenotype of olanzapine metabolizing enzymes
(Nozawa et al., 2008 134; Ghotbi et al., 2010 136; Erickson-Ridout et al., 2011 117; Haslemo et al., 2012 158; Laika et al., 2009 166; Skogh et al., 2011 110; Hägg et al., 2001 126; Carrillo et al., 2003 124; Shirley et al., 2003 123; Cashman et al., 2008 167)
SNP / Phenotyping marker Effect Parameters assessed Study design(a) Sample size (M/F) Smoker Y/N Ethnic origin Reference
UGT1A4 no influence OLA, DMO, OLA-10-NG, OLA/DMO, OLA/OLA-10-NG
psychiatric patients
51(34/17) 16/35 Japanese Nozawa et al.2008
reduction OLA C/D psychiatric patients
121(77/44) 50/71 Swedish Ghotbi et al.2010
Increase for both glucuronides
Vmax/Km for formation of OLA-10-NG and OLA-4'-NG
in vitro Erickson-Ridout et al.2011
no influence on OLA C/D, decrease in OLA-10-NG C/D, increase in Vmax/Km
OLA C/D, OLA-10-NG C/D, Vmax/Km for formation of OLA-10-NG
TDM data from psychiatric patients and in vitro
OLA: 407 (55%/45%) OLA-10-NG: 129
60%/40% Norweigian Haslemo et al. 2012
UGT2B10 D67Y 199G>T Decrease for both glucuronides
Vmax/Km for formation of OLA-10-NG and OLA-4'-NG
in vitro Erickson-Ridout et al.2011
CYP1A2 no influence OLA, DMO, OLA-10-NG, OLA/DMO, OLA/OLA-10-NG
psychiatric patients
51(34/17) 16/35 Japanese Nozawa et al.2008
reduction OLA C/D bodyweight psychiatric patients
73(36/37) 30/43 German Laika et al. 2009
no influence OLA C/D psychiatric patients
121(77/44) 50/71 Swedish Ghotbi et al.2010
no influence OLA, DMO psychiatric patients
37(25/12) 10/27 Swedish Skogh et al.2011
*1C, -3860G>A no influence OLA, DMO, OLA-10-NG, OLA/DMO, OLA/OLA-10-NG
psychiatric patients
51(34/17) 16/35 Japanese Nozawa et al.2008
*1D, -2467delT no influence OLA C/D psychiatric patients
121(77/44) 50/71 Swedish Ghotbi et al.2010
paraxanthine/caffeine (saliva, 10hr)
no influence CL(oral)=dose/AUC adjusted for weight
healthy volunteers
17(17/0) 0/17 Swedish Hägg et al.2001
(AFMU+1U+1X+17U+17X)/137X (urine, day15)
r=0.89, p<0.0001 correlation
OLA C/D psychiatric patients
17(9/8) 8/9 Spanish Carrillo et al.2003
17X/137X (plasma, 4hr, saliva, 6hr and 10hr, urine, 8hr), (17X+17U)/137X (urine, 8hr), (AAMU+1X+1U)/17U (urine, 8hr)
R=0.701, P<0.005, correlation
CL(oral) healthy volunteers
14(13/1) no info American Shirley et al.2003
CYP2D6 *10A, 100C>T no influence OLA, DMO, OLA-10-NG, OLA/DMO, OLA/OLA-10-NG
psychiatric patients
51(34/17) 16/35 Japanese Nozawa et al.2008
*3, *4, *5, *6, *41 no influence OLA, DMO psychiatric patients
37(25/12) 10/27 Swedish Skogh et al.2011
PM, EM by dextromethorphan (saliva)
no influence AUC and CL(oral)=dose/AUC adjusted for weight
healthy volunteers
17(17/0) 0/17 Swedish Hägg et al.2001
4-OH-debrisoquine/debrisoquine (urine, 8hr)
no influence OLA C/D psychiatric patients
17(9/8) 8/9 Spanish Carrillo et al.2003
FMO3 K158, M257, G308, K158-G308
3.2 fold decrease
with K158-G308(b)
Vmax/Km for formation of OLA N-oxide
in vitro Cashman et al. 2008
a) steady state obtained in psychiatric patients whereas a single dose given in healthy volunteers
b) no statistic p-values were provided for kinetic analysis in this study
OLA-10-NG, OLA 10-N-glucuronide; OLA-4'-NG, OLA 4’-N-glucuronide
L48V, 142T>G
*1F, -163C>A
32
Significant association with OLA disposition is known for merely one variant, g.-163A (also known as CYP1A2*1F, rs762551C>A, Table 3).110,166 None of the other CYP1A2 alleles studied for influence on OLA disposition was found as a significant contributor to inter-individual variation (Table 3). In Caucasian psychiatric patients, the -163A/A genotype carriers displayed on average 22% lower serum OLA concentrations (dose and body weight adjusted), independent of inducing factors.166 This C to A nucleotide change in intron 1 is recognized to confer a higher inducibility as well as an elevated basal enzyme activity, mainly in Caucasian cohorts.168-171 Replications of the finding in patient cohorts of Asian origin repeatedly failed, despite the high population frequencies of this variant among both ethnicities (60-70%).134,169,172-174 Worth noting is the fact that CYP1A2 g.-163C>A was not identified as a representative haplotype tagging SNP across major ethnic groups in CYP1A1/1A2 haplotype construction.175
GWAS studies on searching genetic determinants for habitual caffeine consumption have identified several SNPs located within the inter-genic spacer region of the CYP1A cluster (CYP1A1/CYP1A2, rs2470893C>T and rs2472297C>T), at CYP1A2 locus (rs2472304A>G) as well as within the region upstream the aryl hydrocarbon receptor (AHR) coding locus (rs6968865T>A and rs4410790C>T).176-178 In addition to the well estab-lished use of caffeine for CYP1A2-phenotyping, the key role of AHR in regulation of CYP1A2 gene expression179 further supports the biological plausibility of the association reported. These SNPs represent new candidate markers for assessment of the CYP1A2 genotype-phenotype relationship.
FMO3 and FMO1 Each FMO isoform is encoded by a single gene. The regional location of FMO genes has been refined to chromosome 1q23-25 for FMO1, 2, 3, 4 and 6, which form a gene cluster, whereas FMO5 is located on chromosome 1q21.1.47 The sixth isoform, FMO6, is a pseudogene.180 In addition to these 6 genes, a second FMO gene cluster, located between FMO5 and the cluster with FMO1 to 6 has also been described in humans. This cluster however contains five pseudogenes, FMO7P, 8P, 9P, 10P and 11P.47 They do not encode functional proteins.
In general, clinical pharmacogenetic studies including FMO polymor-phisms are rare. Prior to our studies, none of the FMO SNPs have been as-sessed for association with variability in OLA disposition in vivo. Due to the pattern of tissue distribution relevant for drug metabolism and existing knowledge on clinical importance of each FMO isoform (Table 1), genetic variants of FMO3 and FMO1 are the main points of interest for association studies in this thesis.
Three common FMO3 coding variants, p.E158K, p.V257M, and p.E308G, as well as the K158-G308 compound allele have been best charac-terized for functional impact on enzyme activity. Variable degrees of re-
33
duced enzyme activity in a substrate dependent manner were observed.181-184 In subjects homozygous for K158 and G308, a transient or mild form of TMAuria can be triggered under exposure to TMA and hormonal influence.185,186 In vivo correlation between carriage of both variant alleles and reduced catalytic efficiency has been demonstrated for ranitidine and sulindac.187-190 Regarding psychoactive substrates, altered metabolic effi-ciency associated with K158 has been shown in vitro for amphetamine and methamphetamine.191 Reduction in OLA N-oxidation was reported in vitro for the protein variant expressing the K158-G308 compound allele, though supported only by raw data with no data on statistical assessment167 (Table 3). On the contrary, the common variants were not associated with inter-individual variation in clozapine metabolism in German patients.192
Additional known coding variants affecting enzyme activity are ethnic specific and/or TMAuria-causing mutations.193-196 FMO3 p.D132H and p.L360P were only found in African Americans with reduced in vitro cata-lytic efficiency associated for the former and enhanced efficiency for the latter.197 Among the non-coding gene variants identified, two common up-stream variants have been associated in vitro with alteration in promoter activity affecting FMO3 gene expression, g.-2177G>C causing an activity increase and g.-2106G>A causing an activity reduction.198
The FMO1 gene locus is more conserved than other FMOs.193 Four of the five non-synonymous FMO1 variants detected in gene re-sequencing studies have been studied, p.H97Q, p.I303V, p.I303T and p.R502X. These genetic variants have been characterized with modest activity impact and are rare alleles in general populations.193,199,200 Of the common upstream variants identified, only FMO1 g.-9536C>A, termed FMO1*6, is considered to have potential functional impact as it eliminates the binding site of a transcription factor, leading to reduced gene expression in vitro.199
34
Aim of the thesis
The overall aim of this thesis was to study genetic influence of the drug me-tabolizing enzymes on inter-individual variability in therapeutic olanzapine exposure, by exploring the significance of FMOs and further evaluating the polymorphic effect of UGT1A4 and CYP1A2.
Specific aims were:
Paper I • To assess inter- and intra-ethnic variation in distribution of five non-
synonymous FMO3 SNPs and their haplotypes in 13 defined ethnic groups from Europe, East Asia and sub-Saharan Africa
• To assess the appropriateness of extrapolating frequency data to corre-sponding ethnic populations from other continents
Paper II: • To investigate potential influence of FMO1, FMO3 and UGT1A4 poly-
morphisms on steady-state concentrations of olanzapine and desmethyl olanzapine in serum and cerebrospinal fluid
Paper III: • To quantify the steady-state concentrations of olanzapine N-oxide in
patients treated with olanzapine • To further evaluate the correlation between olanzapine concentrations
and the two FMO SNPs observed in Paper II • To search for additional SNP markers within the FMO3 genomic region
Paper IV: • To validate the reported association of CYP1A2*1F with reduced sys-
temic olanzapine exposure • To investigate the potential impact of four additional candidate markers
at CYP1A1/CYP1A2 and AHR loci on systemic olanzapine exposure
35
Material and Methods
Subjects The ethnicity of all study subjects (healthy volunteers and patients) was de-fined by self-reporting. Informed consent was obtained from the subjects included in Paper I and II. The study presented in Paper I was approved by the local ethics committees of the countries from which the samples were collected. The study presented in Paper II was approved by the Ethics Committee of the Medical Faculty of Linköping University, Sweden, the Swedish Medical Products Agency, and the Swedish Data Inspection Board. The studies presented in Paper III and Paper IV were approved by the fol-lowing Norwegian authorities: the Regional Committee for Medical and Health Research Ethics, the Privacy Ombudsman and the Investigational Review Board at Diakonhjemmet Hospital.
In Paper I, a total of 2152 unrelated adult subjects (healthy volunteers) were included. They represented the three major ethnicities (Caucasian, Asian and African) and formed 13 defined ethnic subgroups (Table 4). The Caucasian population consisted of Swedes, Italians and Turks, and the Asian population of Japanese. From the African continent, nine ethnic groups from three regions of Sub-Saharan Africa were included comprising in total 863 subjects. The Swedish volunteers were anonymous blood donors. Volunteers were recruited among university students and employees for the Italian, Turkish and Japanese groups. The African samples were retrieved from the biobank at African Institute of Biomedical Science and Technology, Zim-babwe.201
Table 4. Sample size and country origin of the study subjects included in the 13 ethnic groups in Paper I.
Caucasian Asian African
Northern Europe East Asia North-Western AfricaSweden 410 Japan 300 Nigeria 100 Hausa 99 Ibo 100 Yoruba
Southern Europe Central-Eastern AfricaItaly 279 Kenya 99 Luo 97 Kikuyu 143 Masaai
Eurasia Southern AfricaTurkey 300 Zimbabwe 99 Shona 63 San
South Africa 63 Venda
36
In Paper II, a group of 37 Caucasian outpatients (25 males and 12 fe-males; age range, 23-50 years) at Linköping University Hospital, with diag-nosis of schizophrenia or schizoaffective disorder according to DSM-IV criteria was included. These patients were initially recruited for investigation of the relationship between steady-state serum and CSF concentration of OLA and the DMO metabolite.110 They were not first-episode patients and all but three had been medicated earlier with an antipsychotic drug other than OLA. At the time of study, all patients were on OLA as the only antip-sychotic drug treatment with stable daily dose (dose range, 2.5 to 25mg/day) for at least 14 days. The length of OLA treatment varied between 0.2 and 11 years (median, 2 years). The concentration ranges for OLA were 3.5-102 ng/mL in serum, 0.56-9.43ng/mL in CSF and for DMO was 2.6-17.4ng/mL in serum, 0.12-2.79ng/mL in CSF.
For Paper III and Paper IV, serum OLA concentrations in 379 psychiatric patients (198 males and 177 females, no data on gender for 4 patients, Table 5), were obtained from routine TDM service conducted at the Center for Psychopharmacology in Diakonhjemmet Hospital, Oslo, Norway, during the period July 2007 - December 2010. One sample from each patient was in-cluded for analysis.
Table 5. Characteristics of patients included in Paper III and IV.1
OLA2 OLA N-oxide DMO3
Patient number 379 123 342
Age (years) 45 (16 - 89) 39 (19 - 89) 45(16-89)
Gender/Smoking status Non-smoker
Smoker Non-smoker
Smoker Non-smoker
Smoker
Female 81 96 22 35 76 85 (22%) (26%) (18%) (29%) (22%) (25%)
Male 80 118 25 41 71 109 (21%) (31%) (20%) (33%) (21%) (32%) Valproic acid co-medication (Yes/No)
29/350 12/111 26/316
Dose (mg/day) 15 (2.5 - 60) 20 (5 - 40) 15 (2.5 - 60) Serum conc. (nmol/L) 101 (6 - 481) 5.8 (0.8 - 21) 23 (2 - 97) Serum C/Ds (nmol/L/mg) 7.6 (2.4 - 38.6) 0.31 (0.08 - 1.8) 1.7 (0.25 - 7.0) OLA N-oxide/OLA ratio (%) 4.2 (1.1 - 11.1)
DMO/OLA ratio (%) 23.6(4.0 - 70.3)
1) Data are median (range) or n (%), 2) No data on gender for 4 patients, 3) No data on gender for 1 patient
Criteria for inclusion were: a) the time between the last drug intake and se-rum sampling was 10-30 hours, b) measured OLA serum concentrations above the lower limit of quantification, and c) no concurrent use of CYP1A2
37
or UGT1A4 inducers/inhibitors except for tobacco smoking and co-medication with lamotrigine or valproic acid. The patients included were of Caucasian origin. Due to biobank regulation, serum samples for drug con-centration determination can only be stored for 3 months post analysis. Re-analysis to measure the metabolites was therefore only possible for sub-groups of the study population. Of the 379 patients, quantification of DMO could be done for 342 patients and OLA N-oxide for 123 patients (Table 5).
Methodological overview Analysis of drug and metabolite concentrations The molecular mass of OLA is 312.4g/mol. For unit conversion to ng/mL, multiply the concentration value in nmol/L with a factor of 0.3124. For unit conversion to nmol/L, multiply the concentration value in ng/mL with a fac-tor of 3.2.
In Paper II, fasting blood samples for the analysis of OLA and DMO were collected at 9 to 14.5 hours (median 12 hours) after the evening dose. Lum-bar puncture was carried out at close connection to blood sampling at the minimum of eight hours in the fasting state. The procedure was, however, unsuccessful in 8 patients (dose range, 7.5-15mg/day; median, 10mg/day) and CSF samples were thus available for 29 of the 37 patients. OLA and DMO concentrations in serum and CFS were analyzed using a liquid chro-matography/tandem mass spectrometry (LC-MS/MS) method developed by Josefsson et al.202
In Paper III and Paper IV, serum concentrations of OLA, DMO and OLA N-oxide were analyzed by a validated and certified UPLC-MS/MS method developed for routine TDM analyses at the Center for Psychopharmacology in Diakonhjemmet Hospital, Oslo, Norway. The experimental procedure is described in details in the papers included.
DNA extraction In Paper I and II, genomic DNA was purified from peripheral leukocytes and stored at -20⁰C, with the exception of the Swedish blood donors, from whom buffy coat was available. Standard extraction was performed using QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany), according to the guide-lines of the manufacturer. In Paper III and IV, DNA was extracted from pe-ripheral blood leukocytes using E.Z.N.A® Blood DNA Mini Kit (Omega Bio-tek, USA). DNA quantitation of study samples was evaluated using Nanodrop® in Paper I and II, and Picogreen® in Paper III and IV.
38
SNP genotyping Candidate SNPs was genotyped using 1) TaqMan® allelic discrimination with Fluorogenic 5´ nuclease assays (Paper I, II and III) and 2) Illumina GoldenGate® genotyping assays with discriminatory DNA polymerase and ligase (Paper III and IV). The candidate SNPs included and the method(s) used in each paper are listed in Table 6. For both methods, pre-designed and validated assays for the SNP loci of choice were purchased, from Applied Biosystems for TaqMan® assays and from Illumina® for GoldenGate® as-says. TaqMan® assay was custom designed for one novel FMO3 mutation identified in the gene re-sequencing study (Paper II), FMO3 g.18129T>C. Forward primer: GTCTTCTGACACCACTTTCTGC Reverse primer: TGTTTTGGGCTTACAGGACA
The TaqMan® assays were performed with ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, USA) and conducted at the department of clinical chemistry and pharmacology, Uppsala University Hospital. The Illumna® assays were performed with Illumina BeadXpress and conducted at The SNP&SEQ Technology Platform, Uppsala University Hospital. All analysis was carried out according to the guidelines of the manufacturer.
Table 6. The candidate SNPs genotyped in the thesis
Gene Accession No.
Alternative terminology
Genotyping method TaqMan®
Assay ID Paper No.
FMO3 rs10910879 Illumina III
rs16863990 Illumina III rs1736554 Illumina III rs12404183 Illumina III rs1736560 Illumina III rs3754491 FMO3 g.-2177 Illumina & TaqMan C__27511936_10 III rs10911192 Illumina III rs12072582 FMO3 p.D132H TaqMan C__30633935_10 I rs2266782 FMO3 p.E158K Illumina & TaqMan C__2461179_30 I, III ss470259352 FMO3 g.18129 TaqMan AH20TVW II rs10797894 Illumina III rs2235192 Illumina III rs1736557 FMO3 p.V257M Illumina & TaqMan C__8698544_30 I, III rs2075992 Illumina III rs909529 Illumina III rs909530 FMO3 p.N285N Illumina III rs2266780 FMO3 p.E308G Illumina & TaqMan C__2220257_30 I, III rs28363581 FMO3 p.L360P TaqMan C__30633936_20 I FMO1 rs12720462 FMO1*6 TaqMan C__31270820_10 II, III rs7877 TaqMan C__ 14532_30 III CYP1A rs2470893 CYP1A1-4011 Illumina IV rs2472297 CYP1A1-12441 Illumina IV
rs762551 CYP1A2-163 / CYP1A2*1F
Illumina IV
rs2472304 CYP1A2+2159 Illumina IV
AHR rs4410790 Illumina IV
39
Gene re-sequencing Detection of FMO3 and UGT1A4 polymorphisms in Paper II was done using gene re-sequencing analysis. The gene regions of interest were amplified by PCR reaction. The nucleotide sequences were determined by sequencing the purified PCR products using BigDye® Terminator v3.1 cycle sequencing kit (Applied Biosystems, Foster City, USA) and analyzed on ABI 3730XL DNA Sequencer (Applied Biosystems, Foster City, USA). The same primer pairs were used for PCR amplification as for gene sequencing. Amplicons were sequenced on both forward and reverse strands (Table 7).
The PCR reaction was carried out on a Peltier thermal cycler with a final reaction volume of 50 µl containing a) DNA template (60 ng – 120 ng), b) 1x PCR Buffer (20mM Tris-HCL and 50mM KCl), c) 1.5 mM MgCl d) 0.2 mM dNTP, e) 1 unit Platinum Taq polymerase and f) 0.2 µM each primer. Primer sequences and specific annealing temperatures and durations are summarized in Table 7. The cycling protocol started with an initial denatura-tion at 94⁰C for 2 minutes followed by 31 cycles for FMO3 and 29 cycles for UGT1A4 of denaturation at 94⁰C for 30 seconds, annealing with specific temperature and time duration, and extension at 72⁰C for 1 minute, com-pleted with final extension at 72⁰C for 7 minutes. The PCR products were then purified using QIAquick PCR purification kit (QIAGEN Ltd, Hilden, Germany), sequenced using BigDye® Terminator v3.1 cycle sequencing kit (Applied Biosystems, Foster City, USA) and analyzed on ABI 3730XL DNA Sequencer (Applied Biosystems, Foster City, USA).
Table 7. Primer pairs and PCR conditions used in FMO3 and UGT1A4 sequence analysis.
Product size Annealing
Gene Region Sequence (5' to 3') (bp) ⁰C Time (s)
FMO3 Upstream_F1 TGAATTGCAGCCCTAGAAAA 733 56 45
Upstream_R2 AGCCCAATAAGGAGGATGAC
exon1_F TGACTGGGAGTTAGGTGCAGT 940 60 50
exon1_R AGTTTCGCTTTTATCCAGGTTG
exon2_F AGCCAAAGAGCGAAATCAAA 406 61 45
exon2_R GCAAGGAGGTTACTTTCCAAC
exon3_ F GATTCAACCCACCATTGATT 929 55 45
exon3_R GCTTCCTATGTTTCCCACAA
exon4-5_F CCATATCTCCAAGGGGTGTC 990 60 50
exon4-5_R TTGCTGATGTTCTGCCTTTG
exon6_F GTCTTCTGACACCACTTTCTGC 407 58 45
exon6_R TGTTTTGGGCTTACAGGACA
exon7_F TGGAGATACTCTATGCCTCAGAAA 596 57 30
exon7_R TTCAAAACTGAAGGGGACCT
exon8_F GGTGTCTGTCTGAAAATGAACA 498 59 30
exon8_R TCCTTTCTGGGTCTCCTTTC
exon9_F TGTCAGGGTAGTGTGGGAAA 994 60 50
exon9_R GGGCAGAGAAGAGGAGGTCT
40
UGT1A4 Upstream_F CAAATTATGCAGCCCGTTCT 1317 62 30
Upstream_R GAACCCTTGAGTGTAGCCCA
exon1_F GCTGATTTGCTAGGTGGCTC 1171 63 30
exon1_R CTCATCCGTGCCTCTTCTTC
1) F as forward primer; 2) R as reverse primer
Bioinformatics Primer design In Paper II, reference nucleotide sequences were obtained from the online database of the National Center for Biotechnology Information (NCBI, http://www.ncbi.nlm.nih.gov/) (AY895830.1 for FMO3 and AF297093.1 for UGT1A4). Primer design and sequence specificity check for gene re-sequencing of FMO3 and UGT1A4 were carried out using the online tool, Primer-BLAST (http://www.ncbi.nlm.nih.gov/tools/primer-blast/). Primer and probe design for the custom designed TaqMan® genotype assay (AH20TVW, Table 6) were done using the software program ABI PRISM Primer Express (Applied Biosystems).
tagSNP selection In Paper III, genotype data was retrieved from the International HapMap Project (http://hapmap.ncbi.nlm.nih.gov/cgi-perl/gbrowse/hapmap28_B36/, release #28) for the SNPs located within the genomic region of FMO3 plus 10kb in 5’ and 5kb in 3’ flanking regions (chr1:169316660 to 169358581). The data was obtained for the CEPH group (Utah residents with ancestry from Northern and Western Europe). Selection and evaluation of the tagSNPs were done using the Tagger function implemented in Haploview v4.2203 based on the empirical patterns of linkage disequilibrium obtained from the International HapMap Project.
Sequence variant analysis Post-sequencing analysis in Paper II included fragment assembling, base calling as well as SNP and insertion/deletion detection, using the software package novoSNP204. Reference sequences for the sequenced regions were downloaded from the GRCh37 (Genome Reference Consortium Human genome build 37) assembly via ENSEMBL (http://www.ensembl.org). All identified SNPs were compared with reported SNPs for FMO3 and UGT1A4 listed in NCBI Single Nucleotide Polymorphism database (dbSNP build 131, http://www.ncbi.nlm.nih.gov/projects/SNP/) at the time of the study.
The Haploview v4 (or higher) software was used for calculating MAF and probability of deviation from HWE due to chance for each polymorphism genotyped or detected in a defined study population. It was also used for assessment of pair-wise LD, characterized by D´ and r2 values and visualized by LD plots (Paper I, II, III and IV).
41
Haplotype inference and frequency determination were carried out by the software package PHASE v2.1 in Paper I, Haploview v4.1 in Paper II, III and IV. Population differentiation in haplotype distribution was evaluated with Arlequin v3.1205 in Paper I. The UNPHASED v3.1.5 software206 was used for haplotype association analysis in Paper IV.
Putative effect prediction For the novel intronic FMO3 g.18129T>C detected in Paper II, the online tool, Automated Splice Site Analysis (https://splice.uwo.ca/)207 was used to assess possible disruption of consensus sequences for splicing and detect potential formation of cryptic splice sites (i.e. acceptor site, donor site and branch point). The prediction is based on the fact that sequences immedi-ately adjacent to natural splice sites contain information that dictate the strength of recognition for spliceosome to utilize the splice site. Mutations located within such regions may alter strength of existing natural sites or putative cryptic sites as well as form new cryptic splice sites.
Statistics Two-tailed P≤0.05 was considered statistically significant. For comparison between ethnic populations in Paper I, allele distribution was compared us-ing two tailed Chi-square test or Fisher’s exact test (GraphPad Prism v 4, Graphpad Software, CA). For genotype association studies in Paper II, III and IV, statistical analysis was performed using STATISTICA 9.1 or higher (StatSoft. Inc, Tulsa, OK). All variables (C/Ds and the DMO/OLA ratio) were tested for normality by Shapiro-Wilk test (Paper II, III and IV). When significant deviation was observed, log-transformation was applied to attain normality (Paper III and IV). Degree of correlation between the concentra-tion variables was assessed using Spearman rank correlation test (coefficient: ρ) (Paper III and IV).
In Paper II and IV, concentration comparisons between genotypes were carried out using ANOVA (analysis of variance) or ANCOVA (analysis of covariance). Planned comparisons of least squares means (contrast analysis) were used to test the statistical significance of predicted specific differences associated with the variant alleles. The combined effect of SNPs and non-genetic covariates on OLA C/Ds was assessed by multiple linear regression analysis. Best-subset regression with Mallow’s cp as criterion was applied in Paper II whereas backward stepwise regression was used in Paper IV. Due to the small sample size in Paper II, no more than four variables for serum con-centrations and three variables for CSF concentrations were included in each analysis. In Paper IV, evaluation of CYP1A1/CYP1A2 haplotype association to DMO/OLA ratios and OLA C/Ds were carried out using UNPHASED (v3.1.5).
42
In Paper III, multiple linear regression analysis applying best-subset method (Mallow’s cp as criterion) was used to identify the subset of FMO3 SNPs that best describe concentration variability. ANCOVA was used to assess the two FMO1 polymorphisms. Bonferroni test was used as post hoc analysis. The combined effect of the significantly associated SNPs and non-genetic covariates was further assessed in the final regression models apply-ing backward stepwise regression.
43
Results and Discussion
Paper I
Considerable variability in distribution of non-synonymous FMO3 SNPs across 13 ethnic populations
Majority of the reported population studies were based on cohorts of indi-viduals from Northern America. Due to the long immigration history and large admixture in the region, the genetic backgrounds of the socially de-fined ethnic groups in these countries are not always as clear as they appear to be. Variability in distribution of FMO3 polymorphisms among subpopula-tions within the same ethnicity has too been overlooked. In this study, we investigated population distribution of five functionally relevant FMO3 SNPs: three common variants E158K, V257M and E308G; two African spe-cific variants D132H and L360P.
Inter- and intra-ethnic variation in FMO3 SNP distribution We confirmed inter-ethnic variability in allele distribution (detailed fre-quency data summarized in Table 1 in Paper I). Population comparison re-vealed the highest frequency of K158 variant to be found in native Africans (42 – 52% except the San group, 33%), followed by European Caucasians (34 – 44%) and East Asians (Japanese, 23%). A reversed order was shown for the M257 variant with highest frequency in East Asians (Japanese, 15%), followed by European Caucasians (6 – 7%) and native Africans (0 – 5%). The G308 variant was also rare among native Africans (0 – 1.6%) but much more prevalent in Japanese (21%) and European Caucasians (6 – 22%).
Concerning the two variants originally identified in African-Americans, occurrence of H132 was confirmed whereas P360 variant was not observed in any of the native Africans samples. This may be due to the P360 variant being region specific or less common than previously reported.
The published data on FMO3 polymorphisms in African populations is largely based on African Americans. Here, only the K158 variant was com-mon in all native African groups whereas the other variant alleles were rare. The ancestral history of African Americans is complex as they are not only descendants of many different populations within Africa but also have on
44
average 20% European ancestry208. Estimates of allele frequencies in African Americans should therefore not be uncritically taken as representative of African populations in other parts of the world.
Intra-ethnic differences were detected for E158K and E308G among the three European Caucasian populations (Swedish, Italian and Turkish). The Swedes (K158, G308; 44%, 22%) had higher allele frequencies of both K158 and G308 than the Italians (34%, 11%), and the Turks (36%, 6%, P<0.005 in all group comparisons). The reported studies with Northern American cohorts displayed frequency values, for all three sites, either be-tween the Swedes and the Italians/Turks or higher than our groups (data summarized in Table 1 in Paper I). The differences between European and American studies can be due to the heterogeneous ancestral background of American Caucasians from different parts of Europe.
Among the native African groups, regional differences were detected for H132 with higher frequencies in Western Africa groups (5.0 – 8.2%) com-pared to the groups from Eastern and Southern Africa (1.1 – 3.5%, p<0.01 in both cases) with no difference between the latter two regions. The intra-ethnic difference observed for K158 variant (P=0.03) and M257 variant (P=0.02), was on the other hand entirely due to deviating low frequencies in single groups, the San group in particular.
The FMO3 K158-G308 compound variant prevalent among non-Africans We further confirmed the cis-linkage between K158 and G308 variants, and observed no LD for additional pair-wise combinations of the polymor-phisms. The r2 values were consistently close to zero for all combinations across all groups, except for the combination of K158 and G308 variants in the non-African groups. As the r2 value denotes the power to predict preva-lence of one variant by measuring another variant, the group order of in-crease in the r2 value went hand in hand with the decreasing difference in allele frequencies between the K158 and G308 variants in these groups. The r2 values varied from being low for the Turkish (r2=0.1) and the Italian (r2=0.2) groups, to being moderate for the Swedish (r2=0.4) and to high for the Japanese (r2=0.9).
The variant G308 was inferred to co-occur always with K158 variant in hap-lotype construction (haplotype A-G-G as K158-V257-G308 in Table 2 in Paper I). The proportion of individuals carrying this construction was higher in non-African groups (12 – 38%) compared to the Africans (1.3%, Figure 1). Since r2=0.9 between E158K and E308G among Japanese, the Japanese carrying the K158-G308 compound variant were either heterozygous or ho-mozygous for both sites (158-308: GA-AG and AA-GG) (Fig 7). A third
45
type of carrier (AA-AG) was identified in the Caucasian groups being ho-mozygous mutated for site 158 but heterozygous for site 308. Substantial proportions of all three types of carriers (GA-AG, 23.4%; AA-AG, 8.3%; AA-GG, 6.6%) were observed in the Swedish group compared to in the other populations (Fig 7). Given the impaired FMO3 activity reported for the K158-G308 compound variant and the frequency variability shown here, inter- and intra-ethnic differences in drug metabolism catalyzed by FMO3 can be expected. The Swedes are detected here with relatively high frequen-cies of the variant compound genotypes and might thus in theory be more prone to decreased activity of the enzyme.
Figure 7. Genotype distribution of the FMO3 K158-G308 compound variant in ethnic populations (carrier fractions colored in purple, green and red).
0,0
10,0
20,0
30,0
40,0
50,0
60,0
70,0
80,0
90,0
100,0
Swedes Italians Turks Japanese Africans158-308 Wt/Ht 0,0 0,0 0,0 0,3 0,0158-308 Mut/Wt 5,4 5,0 9,7 0,0 20,5158-308 Ht/Wt 24,6 31,2 36,3 3,3 49,4158-308 Mut/Mut 6,6 1,1 0,0 4,7 0,0158-308 Mut/Ht 8,3 4,3 4,0 0,3 0,4158-308 Ht/Ht 23,4 15,1 8,0 32,0 0,9158-308 Wt/Wt 31,7 43,3 42,0 59,4 28,7
Com
poun
d ge
noty
pe fr
eque
ncy (
%)
46
Paper II.
Inter-patient variation in OLA exposure is correlated to UGT1A4 in serum and FMO in cerebrospinal fluid
To investigate the potential polymorphic influence of FMO1, FMO3 and UGT1A4 on inter-patient variation observed for OLA exposure, we per-formed a pilot study consisting of 37 Swedish Caucasian patients treated with OLA as the only antipsychotic drug. Steady-state concentrations of OLA and the metabolite DMO were available for both serum and CSF, pro-viding estimations on systemic and local drug exposure, respectively. From the previous evaluation of the same patient cohort, smoking and age were identified as significant covariates for C/Ds of OLA and DMO, respectively, whereas functional polymorphisms of CYP1A2 and CYP2D6 displayed no major influence other than in association with smoking.110
FMO and UGT1A4 SNP detection and haplotype construction One novel intronic mutation and 25 known variants were identified within the FMO3 gene region and 11 known variants for UGT1A4 by sequencing analysis (detailed SNP data summarized in Supplemental Table A in Paper II).
Four FMO3 haplotype structures were deduced in addition to the wild type haplotype (Supplemental Figure A in Paper II). Each of them incorpo-rated one of the four variants with reported functional impact, g.-2177G>C, E158K, V257M and E308G. For FMO1, we chose to genotype one candi-date SNP, FMO1*6, and identified five carriers who were all heterozygous mutated.
Of the two non-synonymous UGT1A4 SNPs identified, P24T was not in LD to any of other UGT1A4 SNPs detected whereas L48V was in complete LD (r2=1.0) with 4 upstream variants (Supplemental Figure A in Paper II). Our data is in agreement with the earlier reported linkage pattern for UGT1A4 polymorphisms. Six UGT1A4 haplotypes were inferred and further grouped into UGT1A4*1, UGT1A4*2 carrying T24, and UGT1A4*3 carrying V48 (Supplemental Figure A in Paper II). Patients identified carrying UGT1A4*2 and *3 were all heterozygous carriers.
Serum concentration variation and UGT1A4*3 No significant correlation with C/Ds of OLA was found for any of the FMO3 haplotypes or FMO1*6 in serum. The influence of homozygosity for the K158-G308 compound variant was however not possible to assess since only one carrier was identified. On the other hand, we observed lower serum
47
OLA C/Ds in carriers of UGT1A4*3 suggesting increased OLA glucuronida-tion activity (Table 8). The effect of reduced UGT1A4 gene expression re-ported in vitro for two of the linked upstream variants g.-219(T) and g.-163(A)157 seems to be limited in vivo. The impact associated with this haplo-type is suggested to be primarily attributed to the V48 variant.
Mean OLA C/D in the non-smoking UGT1A4*3 carriers were 35% lower than those of wild type non-smokers (P=0.06), comparable to the mean C/D in wild type smokers (Table 8). Although the association suffered in statisti-cal power due to the small sample size, the observation was in line with find-ings in earlier studies assessing this variant.117,152-156 The hypothesis of in-creased OLA clearance by UGT1A4-mediated glucuronidation was further supported by a 37% lower serum DMO C/Ds observed in the non-smoking UGT1A4*3 carriers compared to wild type non-smokers (Table 8). The ef-fect of UGT1A4*3 on serum DMO C/D remained significant even after ad-justment of age (age, P=0.002; UGT1A4, P=0.005). Smoking alone had no significant influence on DMO C/D (Table 8). However, in a recently pub-lished study with 407 TDM patients, the effect of UGT1A4 L48V on OLA disposition was shown to be evident only on the formation of OLA N-10-glucuronide, but not on systemic OLA exposure.158 Whether there is any pharmacogenetic influence of UGT1A4*3 on OLA exposure in vivo will need to be demonstrated in future studies.
Table 8. Serum OLA and DMO concentrations (ng/mL/mg) in relation to UGT1A4 genotype and smoking habit 1
Smoker UGT1A4 genotype No. OLA C/D P value2 DMO C/D P value2
No *1/*1 or *1/*2 24 3.45 ± 1.09 0.75 ± 0.19 Yes *1/*1 or *1/*2 9 2.28 ± 0.96 0.006 0.68 ± 0.25 >0.1 No *1/*3 3 2.24 ± 0.43 0.06 0.47 ± 0.16 0.04 Yes *1/*3 1 1.81
1) The data are given as mean ± SD. 2) The P values refer to comparisons to the non-smoking *1/*1 or *1/*2 group
The UGT1A4*2 was not associated with any significant change in C/Ds of OLA or DMO, which is in line with data from in vitro studies. The UGT1A4 g.-204A variant (UGT1A4*1k, Supplemental Figure A in Paper II) was previously associated with reduced glucuronide formation in vitro. The effect of this variant could, however, not be statistically assessed in the cur-rent study due to limited number of carriers identified, two heterozygous patients with one of them also being a UGT1A4*3 carrier.
48
CSF concentration variation and upstream FMO SNPs Carriers of UGT1A4*3 did not show deviating OLA C/D in CSF. Given the lack of data supporting UGT1A4-specific expression in brain tissue, our findings in serum and CSF seem to support the UGT1A4-mediated glu-curonidation of OLA to be restricted to the liver.
On the other hand, a two-fold difference in OLA C/Ds was observed be-tween carriers of the FMO3 g.-2177C variant and carriers of FMO1*6 (Fig 8). The comparatively high and low OLA C/Ds observed for the carriers of these two upstream SNPs, respectively, is in line with in vitro function data reported for them.198,199 Given that comparable expression levels for all FMO isoforms have been reported in human brain, our data suggest that FMOs might mediate OLA metabolism locally in the central nervous system.
Figure 8. Boxplot with median and interquartile range of CSF OLA C/D, in relation to combined FMO1/FMO3 genotypes and smoking habits (○ non-smoker; ● smoker).
Elevated OLA exposure in the carrier of a novel FMO3 mutation The novel FMO3 intronic mutation (g.18129T>C) was located at 11 nucleo-tides upstream exon 6 and thus adjacent to the intron 5 splice acceptor site (Fig 9A). Assessing the individual information content of natural splice sites in FMO3, intron 5 has the lowest Ri value for its splice acceptor site (3.3 bits) and is the only intron having low Ri values at both splice donor and acceptor sites compared to the other introns. The nucleotide change, g.18129T>C, caused a decrease in Ri value of the intron 5 splice acceptor site from 3.3 to 2.6 bits. As the information content remains above the lower threshold of recognition, 1.6 bits, the impact of the mutation was inferred to
49
weaken the natural splice acceptor site for intron 5 and cause leaky pre-mRNA splicing.
Interestingly, the one patient carrying the novel variant displayed the highest serum OLA C/D within the study group, together with one of the highest OLA C/Ds in CSF (Fig 9B). Whether there is a cause-effect associa-tion warrants further studies. However, when population distribution of the variant allele was assessed in a group of 300 Swedish blood donors, no addi-tional carrier was identified. Hence, this nucleotide change seems to be a rare mutation among Swedes.
Figure 9. A) A chromatogram showing the position of the novel FMO3 g.18129T>C mutation in intron 5. The splice acceptor site at the end of intron 5 is underlined. B) Boxplot with median and interquartile range of serum and CSF OLA C/Ds in 37 schizophrenic patients on long-term olanzapine treatment. The arrows indicate OLA C/Ds of the patient heterozygous for the g.18129T>C mutation.
g.18129 T Referencecontrol
Heterozygouscarrier
Exon 6Intron 5
g.18129 T>C
Exon 6Intron 50,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
CS
F O
LA
C/D
(n
g/m
L/m
g)
0
1
2
3
4
5
Se
rum
OL
A C
/D (
ng
/mL
/mg
)
Serum CSF
50
Paper III.
FMO3 and FMO1 SNPs as genetic factors influencing OLA N-oxidation and systemic OLA exposure in vivo
In the following study, statistical power was improved with a larger patient group (n=379) compared to the cohort in Paper II. The subgroup with avail-able concentration data on OLA N-oxide did not differ from the study popu-lation regarding OLA C/Ds and patient characteristics such as gender, smok-ing status and co-medication with valproic acid (Table 5). We analyzed two candidate SNPs for FMO1 (FMO1*6 and rs7877) and 15 tagSNPs for FMO3 (detailed SNP data summarized in Table 2 in Paper III). The FMO1 SNPs displayed little pair-wise LD with each other, or to any of the FMO3 SNPs (for LD plot see Supplementary Figure 1 in Paper III).
Quantification of OLA N-oxide The levels of OLA N-oxide detected in patient serum ranged from 1% to 11% of those of OLA, with an average of 4% (Table 5, Fig 10). This is in accordance with a minor quantitative role of this metabolic pathway reported for the elimination of OLA in human. We observed however a relatively high correlation between the serum concentrations of OLA N-oxide and the parent compound (ρ=0.72, P<0.001, Fig 10) which could suggest an equilib-rium between them.
Figure 10. Serum concentrations of OLA and the metabolite, OLA N-oxide, in 123 patients on long term treatment.
0 50 100 150 200 250 300 350 400 450 500
Serum OLA (nmol/L)
0
2
4
6
8
10
12
14
16
18
20
22
Se
rum
OL
A N
-oxi
de
(n
mo
l/L)
51
Reduced serum OLA N-oxide concentrations in the homozygous FMO3 G308 carriers We found FMO3 E308G (MAF123 =0.23) to be the only significant genetic factor for OLA N-oxide C/D (P=0.0005). The association remained signifi-cant (P=0.001) after further adjustment for the non-genetic covariates, smok-ing and gender. The finding supports the in vitro data on FMO3 catalyzing N-oxidation of OLA. Homozygous carriers of the minor G-allele (n=6) had a median C/D OLA N-oxide approximately 50% lower compared to both het-erozygous carriers and non-carriers (Fig 11). All the homozygous carriers of FMO3 E308G in this study were also homozygous mutated for FMO3 E158K, in accordance with the observation from Paper I. Our data support the impact of this variant on FMO3 enzyme activity, but no effect on the C/Ds of OLA was found.
Figure 11. Serum OLA N-oxide C/Ds in 123 OLA-treated patients stratified by FMO3 rs2266780A>G (p.E308G). Genotype composition and sample size are given for each subgroup. The median concentrations with interquartile range are shown both numerically (raw data) and graphically (log-transformed). Non-outlier range is also displayed in the boxplot. Comparisons to G/G genotype are denoted by P values (Bonferroni test).
No additional FMO3 polymorphisms were identified as genetic markers for variability in C/Ds, despite the large genomic region studied. Up to now, the majority of reported FMO3 variants with functional impact are rare nonsense and missense mutations detected in patients affected with TMAuria.67 Few markers have been associated with polymorphic variation of enzyme activity within the general population.67 Inter-individual variability in FMO3 gene expression and protein content has been reported to be correlated in vitro to the levels of transcription regulatory factors, not affected by FMO3 haplo-types.209 Although our data indirectly support the direction of incorporating
A/A A/G G/G-3,0
-2,5
-2,0
-1,5
-1,0
-0,5
0,0
0,5
1,0
log
(C
/D O
LA
N-o
xid
)
Group size (n). 74 43 6
Median (25%-75%) (nmol/L/mg)
0.34 (0.24─0.52) 0.32 (0.24─0.47) 0.16 (0.08─0.19)
P (compared to G/G) <0.001 0.002 ─
52
trans-acting factors into future studies, this approach is challenged as FMO3 regulation is exceptionally complex and poorly characterized at the current stage.
FMO1 SNPs associated with increased serum OLA concentrations We observed no significant association of the two FMO1 SNPs with C/Ds of OLA N-oxide. However, FMO1*6 and rs7877T correlated significantly (P<0.05 for both) to serum OLA C/Ds after correction for smoking. The smokers carrying both variants had 45% higher median OLA C/D than wild type smokers (Ppost hoc=0.04) (Fig 12). Among carriers of both variant alleles, no significant effect of smoking on C/Ds of OLA was found (Ppost hoc=0.82). This is in contrast to the other groups where smokers consistently had lower OLA C/Ds than non-smokers (Ppost hoc<0.001 for all three group compari-sons) (Fig 12).
Figure 12. Serum OLA C/Ds in 379 OLA-treated patients stratified by FMO1*6C>A, FMO1rs7877C>T and smoking status. Genotype composition and sample size are given for each subgroup. The median concentrations with interquar-tile range are shown both numerically (raw data) and graphically (log-transformed). Non-outlier range is also displayed in the boxplot. Comparisons between smokers and non-smokers carrying the same genotype composition are denoted by P(smoker vs non-smoker) values. For multiple pair-wise comparisons post hoc analysis with Bonferroni test was performed.
These findings suggest a reduced influence of smoking on OLA clearance in patients carrying both FMO1*6 and rs7877T. FMO1 has been shown to be
log
(C/D
OLA
)
0,5
1,0
1,5
2,0
2,5
3,0
3,5
4,0
P= 0.04
FMO1*6C>A C/C C/C A carrier A carrier C/C C/C A carrier A carrier
FMO1rs7877C>T C/C T carrier C/C T carrier C/C T carrier C/C T carrier
Group size (n). 82 45 23 11 101 72 20 25
Median (25%-75%) (nmol/L/mg)
9.2 (7.4─12.3)
10.8 (8.0─13.5)
13.4 (7.9─14.6)
10.1 (9.0─15.0)
5.3 (4.3─7.5) 5.4 (4.8─8.5) 5.5 (4.3─7.7)
7.7 (7.0─10.0)
P(smoker vs non-smoker) <0.001 <0.001 <0.001 0.82
Non-smokers (n=161) Smokers (n=218)
53
an efficient catalyst of nicotine N-oxidation and rs7877C>T was originally identified as a determinant for maintenance of nicotine addiction. The re-ported odds ratio of 0.77 for heavy smoking associated with rs7877C>T suggests a prolonged duration of nicotine effect.63 A hypothetical explana-tion for our observations could be that if carriers of this polymorphism smoke less heavily, there would also be lower induction of drug metabolic enzymes such as CYP1A2, resulting in higher than expected OLA C/Ds.
Assessing impact on the overall concentration variability, the non-genetic factors (age, gender, smoking status and co-medication with valproic acid) and FMO1*6 (P=0.02) remained significant factors whereas rs7877C>T did not (P=0.08). Approximately 37% (adjusted R2=0.37) of the variability in serum OLA C/Ds in the study cohort (n=379) could be explained by the five factors, FMO1*6 accounting for 2%. The significant correlation between FMO1*6 and elevated serum C/Ds of OLA is in line with the increased OLA C/Ds in CSF of treated patients carrying FMO1*6 observed in Paper II. The influence of the SNP observed supports the reported in vitro data on the reduced promoter activity (2- to 3-fold) associated with this variant.199
Paper IV Candidate SNPs of CYP1A1/2 and AHR loci are novel genetic factors influ-encing desmethylation of OLA and systemic OLA exposure The plasma DMO/OLA ratio has been shown to be significantly correlated to OLA clearance.106 This ratio was chosen as a pathway-specific marker to study the effect of 5 candidate markers on the CYP1A2-mediated OLA me-tabolism (Table 9).
Table 9. Characteristics of the candidate SNPs studied in Paper IV
SNP rs2470893 rs2472297 rs762551 rs2472304 rs4410790
Allele C>T C>T C>A A>G C>T Chromosome Genomic position
Chr15: 75019449
Chr15: 75027880
Chr15: 75041917
Chr15: 75044238
Chr13: 17284577
Gene CYP1A1/1A2 CYP1A1/1A2 CYP1A2 CYP1A2 AHR SNP type inter-gene inter-gene intron 1 intron 4 upstream MAF from HapMap database CEU (n=226) 0.323 0.248 0.721 0.336 0.420 JPT (n=172) 0 0 0.605 0.797 0.680 YRI (n=226) 0 0 0.566 0.991 0.500
MAF in this study 0.278 0.225 0.674 0.433 0.387 Abbreviation: n, sample size. Data retrieved from HapMap database is sorted by population descrip-tor. CEU: Utah residents with Northern and Western European ancestry from the CEPH collection JPT: Japanese in Tokyo, Japan. YRI: Yoruban in Ibadan, Nigeria.
54
No influence by CYP1A2*1F on OLA disposition in vivo We found no influence of CYP1A2-163(A) (CYP1A2*1F, rs762551A) alone on either the DMO/OLA ratio or OLA C/Ds. Our data contradict the re-ported finding of a 22% reduction of serum OLA concentrations in Cauca-sian patients carrying CYP1A2-163(A/A) genotype, independent of inducing factors.166 The allele frequency of 67% in our study cohort was similar to that reported for Caucasians in the HapMap database (72.1%, Table 9). As our sample size is much larger than that in the previous study (n=342 vs. n=73), our analysis should not suffer in power to validate the effect of this highly frequent allele.
However, CYP1A2-163(A) was deduced in four (H2 to 5) of the five ma-jor haplotype constructions (Table 10). Haplotypes H3 and H4 were signifi-cantly associated with variability in DMO/OLA ratio and/or OLA C/D (Ta-ble 10). This finding is consistent with the fact that CYP1A2-163(A) is pre-sent in a number of reported CYP1A2 haplotype constructions (http://www.cypalleles.ki.se/cyp1a2.htm) with different functional impact, but displays lack of mechanistic impact in functional evaluation in vitro.210
Table 10. CYP1A1/1A2 haplotype inference (>1%) and effect association
Haplotype % rs2470893 rs2472297 rs762551 rs2472304 DMO/OLA OLA C/D
P P1
H1 32.4 C C C G Ref Ref H2 28.6 C C A A ns ns H3 21.9 T T A A ↑ 0.004 ↓ 0.04 H4 10.7 C C A G ↓ 0.001 ns H5 5.7 T C A A ns ns rare 0.7
1) Effect displayed for patients who were non-smokers and without co-medication with val-proic acid (n=135) ns: not significant. The direction of impact displayed by the markers
Decreased systemic OLA exposure in carriers of a Caucasian specific CYP1A1/1A2 haplotype The H3 carriers (tagged by rs2472297C>T), compared to non-carriers, dis-played 20-30% increase in median DMO/OLA ratio independent of smoking habit (Fig 13A) and 17% reduction in median C/D of OLA, though among non-smokers only (Fig 13B).
This haplotype was further confirmed as a significant contributor (P=0.01) to the overall variability in OLA exposure together with the non-genetic factors (smoking, gender, co-medication with valproic acid and age). In total, 36% of the variability (adjusted R2=0.36) could be explained by the factors. As the two CYP1A1/1A2 variants included in the haplotype H3 are Caucasian specific according to population diversity data publically avail-
55
able (Table 9), they may thus, at least partly, explain the conflicting findings on the impact of CYP1A2-163(A) between ethnicities.
Figure 13. Serum DMO/OLA ratios (A) and OLA C/Ds (B) in relation to CYP1A1/1A2 rs2472297C>T genotypes (tagging the H3 haplotype). Data are cate-gorized by smoking status in (A) and by smoking status and co-medication with valproic acid (Co-med) in (B). Group sizes are displayed (n). The medians with interquartile range are shown graphically and numerically. Non-outlier range is also displayed in the boxplots. P-values were determined by Kruskal-Wallis tests. ns, not significant.
DM
O/O
LA-r
atio
Non-smokers Smokers0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8A
P=0.004
P=0.01
n=85 n=61 n=115 n=79
C/C C/T, T/T C/C C/T, T/T
0.17 (0.12─0.23) 0.22 (0.13─0.29) 0.26 (0.19─0.34 0.31(0.23─0.38)
CYP1A1/1A2rs2472297
OLA
C/D
(nm
ol/L
/mg)
Non-smokers Smokers0
5
10
15
20
25
30
35
40
45
ns
B
P=0.03
n=78 n=56 n=7 n=5 n=107 n=73 n=8 n=6
C/CC/TT/T
C/CC/TT/T
C/CC/TT/T
C/CC/TT/T
─ ─ Co-med Co-med ─ ─ Co-med Co-med
11.2(8.6─14.5)
9.3 (7.2─12.3)
8.4 (5.7─12.3)
9.2 (7.1─12.4)
6.1 (4.4─8.9)
5.8 (4.8─7.5)
4.3 (4.0─5.8)
5.1 (2.9─5.4)
CYP1A1/1A2rs2472297
56
Indication of trans-acting effect on OLA desmethylation by a SNP upstream AHR The effect association between DMO/OLA ratio and CYP1A2rs2472304(G) (H4 in Table 10) did not remain after taking AHRrs4410790C>T into ac-count. The influence of AHRrs4410790C>T was apparent only among non-smokers with carriers showing a 20% decrease in DMO/OLA ratios (Fig 14A) and a 17% increase in OLA C/Ds (Fig 14B). As AHR holds a central role in regulating the induction of CYP1A enzymes, the association observed suggests potential trans-acting influence.
Figure 14. Serum DMO/OLA ratios (A) and OLA C/Ds (B) in relation to of AHRrs4410790C>T genotypes, categorized by smoking status. Group sizes are dis-played (n). The medians with interquartile range are shown both graphically and numerically. Non-outlier range is also displayed in the boxplots. P-values were de-termined by Kruskal-Wallis tests. ns, not significant.
DM
O/O
LA-r
atio
Non-smokers Smokers0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
Ans
P=0.002
AHRrs4410790
n=59 n=86 n=65 n=116
C/C C/T, T/T C/C C/T, T/T
0.20 (0.16─0.27) 0.16 (0.11─0.23) 0.28 (0.20─0.36) 0.27(0.21─0.36)
OL
A C
/D (
nm
ol/L
/mg)
Non-smokers Smokers0
5
10
15
20
25
30
35
40
45B
ns
P=0.06
AHRrs4410790
n=59 n=86 n=65 n=116
C/C C/T, T/T C/C C/T, T/T
9.2 (7.0─12.4) 10.8 (8.0─14.2) 5.6 (4.8─7.8) 5.7 (4.4─8.6)
58
General discussion
The importance of pharmacogenetics in drug metabolism is best character-ized for the enzymes catalyzing the oxidative metabolism of a large number of drugs in clinical use, e.g. CYP2D6, CYP2C9 and CYP2C19.211 When contribution by any of these polymorphic enzymes is limited in metabolism of a drug compound, potential polymorphic influence of the enzymes re-sponsible for the remaining metabolic routes should not be disregarded. With respect to OLA, CYP2D6 was early precluded as major predictor of variabil-ity in OLA exposure whereas influence by polymorphic CYP1A2 and UGT1A4 has been suggested.123,124,126,136 The main objective in the current thesis was to assess pharmacogenetic influence of FMO in OLA disposition. It had its rationale in the FMO-mediated OLA N-oxidation pathway being overlooked and the functional polymorphisms of FMO3 (hepatic isoform) and FMO1 (extra-hepatic isoform) being poorly evaluated for their potential effects in vivo.
Relevance of the findings Catalytic activity of FMO3 towards OLA is identified in vitro using liver microsomes119, but its effect in vivo is unknown. Our data provided, for the first time, in vivo support for the role of FMO3 in the formation of OLA N-oxide. The median C/D of OLA N-oxide was shown approximately 50% lower in homozygous carriers of the FMO3 K158-G308 compound allele compared to those of heterozygous carriers and non-carriers (Figure 11, Pa-per III). The role of FMO1 in metabolism of many psychoactive compounds is well supported by existing literature56, but its role in OLA metabolism is unknown as there is no assessment available on FMO isoform specificity towards OLA N-oxidation. Since FMO1 is known to be the FMO isoform with broadest substrate range and efficient towards tertiary amines55, we hypothesized that it could also contribute to the metabolism of OLA. In Pa-per III, the FMO1*6 allele was identified as a significant factor, together with the non-genetic covariates, influencing the overall variability in sys-temic OLA exposure. Patients carrying FMO1*6 displayed increased C/Ds of OLA, in serum (Fig. 12, Paper III) as well as cerebrospinal fluid (Fig. 8, Paper II). The effect in serum was further enhanced in smokers carrying an additional FMO1 SNP, rs7877 (Fig. 12, Paper III). Our data suggests an
59
increased systemic (and central nervous system) exposure of OLA in FMO1*6 carriers, possibly due to decreased intestinal first-pass metabolism catalyzed by FMO1. As FMO1 is highly expressed in kidney, potential role of renal metabolism in systemic elimination of OLA is yet another intriguing aspect for future studies.
Based on published data and our findings, CYP1A2 seems to be the most important enzyme for systemic exposure of OLA. Hepatic CYP1A2 activity estimated by caffeine metabolic ratios has been correlated to OLA clearance as well as plasma C/Ds of OLA.123,124 Significant correlation between plasma DMO/OLA ratio and OLA clearance has too been documented.106 Although the reported influence by CYP1A2*1F (CYP1A2-163(A))166 could not be verified in Paper IV, we made a novel observation. Significant influence on both DMO/OLA ratio and OLA C/Ds was detected for a Caucasian specific CYP1A1/1A2 haplotype structure incorporating the -163(A) variant (Table 10, Paper IV). Considering the role of AHR in regulating the induction of CYP1A enzymes and the well characterized effect of smoking on OLA ex-posure, our finding of a significant association between the SNP located upstream AHR locus and the DMO/OLA ratio (Fig. 14A, Paper IV) further support importance of CYP1A2 mediated OLA desmethylation.
Although direct glucuronidation by UGTs is considered to be the primary route of OLA biotransformation due to the large amount of OLA N-glucuronides detected in plasma, fecal and urinary samples105,112, inconsis-tent results have been reported for the effect of UGT1A4*3 on systemic OLA exposure.134,136 We observed lower serum C/Ds of OLA in UGT1A4*3 carri-ers compared to non-smoking non-carriers (Table 8, Paper II), but our study suffered from having a small sample size. A recently published large study, with assessments both in vitro and in vivo, found the SNP effect to be only apparent on formation of the N-glucuronide metabolite, but cause no altera-tion in serum C/Ds of OLA.158 Verification of this finding and/or identifica-tion of novel genetic markers of UGT1A4 are warranted in future studies for clarifying the importance of UGT1A4 polymorphism in OLA metabolism.
Obstacles and limitations For any pharmacogenetic study, characteristics of the SNPs and the clinical parameters chosen are the core. Multiple factors determine the study power including effect size of the causative locus on the clinical parameter, accu-racy in measurement of phenotype and genotype, the degree of LD between the causative locus and the SNP marker, MAFs and sample size. A number of these aspects can be discussed in relation to the studies included in this thesis.
To estimate the role of FMO enzyme activity in OLA disposition in vivo is rather a difficult task. In contrast to CYP2D6 and CYP1A2 having de-
60
brisoquine, respective, caffeine as well characterized probe drug, phenotyp-ing of FMO is underdeveloped. Although TMA and ranitidine have been used to assess FMO3 activity in few clinical studies35,187,190, their suitability as probe drugs for FMO3 has not been systematically evaluated. Phenotyp-ing of the other FMO isoforms has not been reported.
The steady-state serum levels of OLA N-oxide were found to be low in relation to those of the parent compound (Table 5, Figure 10, Paper III). As cyclic inter-conversion is known to occur between tertiary amines and their N-oxide metabolites59,212,213, it is unclear whether the OLA N-oxide quanti-fied reflects the activity of the oxidative pathway, the reduction or a combi-nation of both. Whether this inter-conversion actually occurs between OLA and OLA N-oxide has not been shown so far. The quantitative contribution of N-oxidation in OLA metabolism is thus difficult to determine on the basis of current knowledge. As OLA biotransformation involves multiple path-ways, catalyzed by a number of different enzymes (Fig. 6), multiple sources of variability in drug exposure can be expected. The influence of any spe-cific pathway on the overall OLA exposure could thus be expected to have limited effect size.
Our SNP selection was thorough ranging from known functional SNPs, to the SNPs that can help to capture pattern of inheritance among the SNPs scattered within the genomic region of interest: • candidate SNPs with reported functional impact in vitro and/or in vivo • candidate SNPs identified as hits in GWAS studies assessing clinical
parameters relevant for xenobiotic metabolism • SNPs detected by gene re-sequencing analysis • SNPs tagging haplotype structure in the genomic region of interest The FMO SNP hits identified by GWAS studies have only been reported for FMO1 at the time the current studies were designed. Sequencing and tagSNP based analysis was only carried out for FMO3 given that FMO1 was re-ported to be a more conserved gene compared to FMO3 214, hepatic metabo-lism was considered more important than extra-hepatic clearance for OLA disposition105 as well as the matter of cost to conduct analysis for both genes.
MAFs set the start point for the sample size. MAF of one SNP might dis-play population variation related to both ethnicity and geography, as demon-strated in Paper I. If the effect of the SNP is autosomal recessive, rather than being dominant, and the MAF is low, a large sample size will be needed in order to find an adequate number of homozygous carriers of the minor allele. The number of SNPs included as variables in association studies further influences the samples sizes required, due to the need to reduce the likeli-hood of false positive results. Large sample size is, thus, always desirable in order to identify sufficient carriers and obtain confidence in statistical sig-nificance. However, the challenges for collecting large data sets from well-characterized patient populations are considerable, especially in psychiatry. The study design in Paper II can be criticized for being underpowered. The
61
study was designed primarily to assess the relationship between serum and CSF concentrations of OLA and its metabolites. For this purpose, it was an adequate number of patients.
Phenotype-driven assessment is vital in clinical pharmacogenetic re-search, with observation of outliers in clinic foregoing elucidation of geno-type.4 When in vivo phenotype data is not available, in vitro characterization of genetic variants will provide indication of functional relevance. Signifi-cant associations obtained here support the functional impact of FMO1*6 and the FMO3 K158-G308 compound allele, earlier reported in vitro. How-ever, when functional characterization of genetic variants has been con-ducted in limited extent, such as for FMO SNPs, an alternative approach to select candidate SNPs has to be applied. In Paper III and IV, our hypothesis was generated by correlations detected in exploratory GWAS studies assess-ing clinical parameters relevant for xenobiotic metabolism. In Paper III, the candidate SNP FMO1rs7877 was identified as determinant for nicotine de-pendence.63 In paper IV, genetic determinants for caffeine consumption were chosen for evaluation of the role of the CYP1A2 mediated metabolic path-way.176-178 As CYP1A2 is better characterized than FMO1 regarding varia-tion in enzyme activity and its molecular mechanism, the hypothesis was biologically plausible and able to assess both cis- and trans-acting influence. Hence, taking the exploratory data from GWAS studies into consideration can be helpful in selecting potential SNP markers for genes that are short of SNPs with characterized functional impact.
Implementation of personalized drug prescribing requires adequate knowl-edge on factors contributing to variability in drug response. The magnitude of effect by variables of physiological, environmental or genetic origin var-ies between patients. Each category of factors needs to be evaluated for a complete evidence base. Improved knowledge on the clinical relevance of less well characterized drug metabolic pathways and the polymorphic char-acteristics of the responsible enzymes will help in the process of improving the clinical use of existing drug treatment, as well as be valuable in future drug development.
62
Conclusions
The pharmacokinetics of olanzapine display large inter-individual variation leading to multiple-fold differences in drug exposure between patients at a given dose. This variation in turn gives rise to the need of individualized dosing in order to avoid concentration-dependent adverse effects and thera-peutic failure. In addition to environmental factors (smoking habit, co-medications) and physiological characteristics (gender, age and body weight), genetic variants of drug metabolizing enzymes are a potential source of variability in OLA exposure. The current thesis has evaluated the pharmacogenetic effect of OLA metabolizing enzymes FMO3, FMO1, CYP1A2 and UGT1A4, and their impact on OLA and metabolite kinetics in patients. The major conclusions are the following, • The metabolite OLA N-oxide is presented in low concentrations in se-
rum, but displays high correlation to OLA concentrations.
• The role of FMO3 in the formation of OLA N-oxide is supported. The reduced enzyme activity associated with the FMO3 K158-G308 com-pound variant in vitro is too supported by our data in vivo. The effect seems to be autosomal recessive.
• The FMO3 K158-G308 compound variant can be attributed to the fact
that G308 variant co-occurs always with K158 variant, but not vice versa. The compound variant is expected to have potential clinical im-portance primarily in non-African populations due to its low prevalence in Africa.
• No additional FMO3 SNP marker was associated with variation in C/Ds
of OLA N-oxide, despite the large genomic region studied. Common FMO3 polymorphisms do not seem to influence serum C/Ds of OLA. Rare point mutations with potential functional impact on gene expres-sion might contribute to altered OLA exposure.
• The upstream polymorphism FMO1*6 (g.-9536C>A) is a novel factor
associated with increased C/Ds of OLA in serum and CSF. The effect in serum might be further enhanced among smokers carrying FMO1
63
rs7877C>T located in 3’UTR of the gene. The role of FMO1 in OLA metabolic clearance, concerning the first-pass metabolism in the intes-tine as well as local metabolism in the brain, need to be clarified further.
• The CYP1A2-catalyzed OLA demethylation is influenced by both cis-
and trans-acting variants; the CYP1A2 haplotype [rs2470893(T)-rs2472297(T)-rs762551(A)] with increased OLA demethylation and AHR rs4410790C>T with decreased activity.
• The inconsistent data on the impact of CYP1A2*1F (-163A) in different
populations might be explained by ethnic specific CYP1A2 haplotype structures incorporating the -163(A) variant.
• The UGT1A4 L48V (UGT1A4*3) polymorphism was correlated with
increased systemic OLA clearance but showed no influence on the cere-bral exposure. No additional UGT1A4 SNP marker was identified with significant impact. Due to low statistic power, further examination is needed to verify the finding.
In summary, novel observations on polymorphic influence of FMOs on vari-ability in OLA exposure are presented. The importance of CYP1A2-mediated OLA demethylation on systemic OLA exposure is further sup-ported whereas the effect of UGT1A4 SNPs cannot be concluded. Although identified as significant genetic factors, the SNPs described in the current studies displayed minor contribution to the overall variability in OLA me-tabolism at population level. For individuals who harbor undesirable combi-nations of multiple determinants, they are nerveless expected to be relevant.
64
Acknowledgements
The work in this thesis was supported by grants from the Swedish Research Council, Uppsala University, Fredrik och Ingrid Thurings Stiftelse, and through the regional agreement on medical training and research (ALF) between Uppsala County Council and Uppsala University. The studies were conducted at the Division of Clinical Chemistry and Pharmacology, Department of Medical Sciences, Uppsala University, Sweden.
I am very grateful to all the experiences that I have gained during the past years spent on conducting this project, from including “pharmacogenetics” into my vo-cabulary to having this thesis in its printed form. A Chinese idiom says, being in the company of three, at least one of the other two is good enough to be my teacher. So it has, indeed, been for me. I THANK YOU ALL for having contributed to this process in your own way and helped to bring the best out of me.
Especially I want to express my sincere gratitude to: my supervisors Prof. Marja-Liisa Dahl, and Dr. Gabriella Scordo, for taking me as your student, giving me freedom to proceed and space to grow, having trust in my ability and patience for my development. Many thanks also for your valuable ad-vices and encouragement, for your caring friendship and delightful company.
my co-authors/collaborators, Prof. Espen Molden, Dr. Tore Haslemo, Elisabeth Skogh, Dr. Alice Matimba, Dr. Norio Yasui-Furukori, Dr. Collen Ma-simirembwa, Prof. Haka Zengil, for your constructive help and fruitful collabora-tions. Without you, this thesis would not have been possible.
Prof. Håkan Melhus, Doc. Mia Wadelius, Doc. Pär Hallberg, for being inspiring scientists and sharing senior expertise in the field of clinical pharmacology and pharmacogenetics
Lena Fredriksson, Gunilla Frenne, Hugo Kohnke and David Munro, for your continuous help, advices and nice company in the lab ever since my very first day at Klin Farm.
Dr. Jenny Alfredsson, Helena Vretman, Kristin Blom, Doc. Lena Douhan Håkansson and Maria Rydåker, for teaching me valuable lab techniques, and having endless patience with me and my questions.
65
Anna Foyer, Kristin Bryon, Eva Prado, and Elisabeth Harryson, for being the helping hands with many economic and administrative issues, and Enrique Vegas, for efficiently providing emergency IT support.
former group-mates, Arzu Gunes Granberg and Laura Magliulo, for walking the crooked PhD-path with me, for inspiring me with your strength, energy and creativ-ity, for all the fun memories at and off work. Lena Lenhammar, Christina Leek and Nasrin Najafi, for being great roomies, filling the office with loads of laughter and much thoughtfulness. Ilma Bertulyte, for your splendid humor regarding every-thing and the emergency espresso supply. Jessica Schubert, Martin Dahlö, Sebas-tian DiLorenzo, Markus Mayrhofer, Christofer Bäcklin, Henning Karlsson, Michael Brennan, Agnes Knight and Martha Wadelius, for “den sena fika grup-pen” combing marvelous cakes with exciting nerdy topics, always nicely blended. Dr. Caroline Haglund, for sharing my very first conference trip and the teaching experience over the years. Dr. Mårten Fryknäs, for the good teaching team. Anna-Karin Hamberg, for your frankness. Anna Lundberg, for the heart-warming sur-prise kladdkaka for my half-time. Sofie Schwan, Elisabeth Stjernberg, Dr. Anders Isaksson, Hanna Göranson Kultima, Prof. Rolf Larsson, Prof. Peter Nygren, Sara Strese, Anna Eriksson, Dr. Saadia Bashir Hassan, Dr. Malin Jarvius, Prof. Mats Gustafsson, Dr. Claes Andersson, Obaid Aftab, Kashif Muhammad, Anna-Karin Lannergård and Nadja Lundström, for the friendly smiles on your faces and the interesting chats over the years. The late Daniel Laryea, absent in body, present in spirit.
my senseis at Uppsala Kendo Hokushinkan, Jessica Fröberg, Jimmy Cedervall, Martin Agback, Kavoos Nasoudi-Shoar, Anna Lagercrantz and fellow members of Uppsala Kendo, for sharing your passion, for your guidance and encouragement on the path of continuous improvement, to always accept challenges with high spirit and not be stopped by fear of failure. Now I have more time to join the fight!
my friends through the years (you know who you are!), near or far, old or new, for continuously enriching my life and expanding my world. Special thanks to Ellen Chen and Amy Tong, you two are the closest to sisters for me, to Katarina Stark, for just being you! to the Zebrowski family: Anna, Stefan, Elisabet and Janek, for your kindness, care and support during all my years in Uppsala, to Tala Jalilian, for your wonderful laughter and sense of humor, to members of my Syssy-Family, for all the fabulous dinners we have created together, and to Lina Kittel, for your big heart!
my parents, Yang Li-Ming and Mao Xin, and my in-laws, Lillemor and Thomas, for your unconditional love to me and Gabriel.
finally, my husband, Gabriel, for the hours that have been and many more to come, thank you for standing by me.
66
References
1. Rang, H.P. Pharmacology, (Churchill Livingstone, New York, 2001).
2. Weinshilboum, R. Inheritance and drug response. N. Engl. J. Med. 348, 529-537 (2003).
3. Garrod, A.E. Inborn errors of metabolism, (H. Frowde and Hodder & Stoughton, London,, 1923).
4. Weber, W.W. Pharmacogenetics, (Oxford University Press, Oxford ; New York, 2008).
5. Motulsky, A.G. Drug reactions enzymes, and biochemical genetics. J Am Med Assoc 165, 835-837 (1957).
6. Vogel, F. Moderne probleme der humangenetik. Ergeb. Inn. Med. Kinderheilkd. 12, 52-125 (1959).
7. Ingelman-Sundberg, M. Pharmacogenetics of cytochrome P450 and its applications in drug therapy: the past, present and future. Trends Pharmacol. Sci. 25, 193-200 (2004).
8. Alving, A.S., Carson, P.E., Flanagan, C.L. & Ickes, C.E. Enzymatic deficiency in primaquine-sensitive erythrocytes. Science 124, 484-485 (1956).
9. Kalow, W. & Genest, K. A method for the detection of atypical forms of human serum cholinesterase; determination of dibucaine numbers. Can J Biochem Physiol 35, 339-346 (1957).
10. Kalow, W. & Gunn, D.R. The relation between dose of succinylcholine and duration of apnea in man. J. Pharmacol. Exp. Ther. 120, 203-214 (1957).
11. Evans, D.A., Manley, K.A. & Mc, K.V. Genetic control of isoniazid metabolism in man. Br. Med. J. 2, 485-491 (1960).
12. Hammer, W. & Sjoqvist, F. Plasma levels of monomethylated tricyclic antidepressants during treatment with imipramine-like compounds. Life Sci. 6, 1895-1903 (1967).
13. Alexanderson, B., Evans, D.A. & Sjoqvist, F. Steady-state plasma levels of nortriptyline in twins: influence of genetic factors and drug therapy. Br. Med. J. 4, 764-768 (1969).
14. Mahgoub, A., Idle, J.R., Dring, L.G., Lancaster, R. & Smith, R.L. Polymorphic hydroxylation of Debrisoquine in man. Lancet 2, 584-586 (1977).
15. Tucker, G.T., Silas, J.H., Iyun, A.O., Lennard, M.S. & Smith, A.J. Polymorphic hydroxylation of debrisoquine. Lancet 2, 718 (1977).
67
16. Eichelbaum, M., Bertilsson, L., Sawe, J. & Zekorn, C. Polymorphic oxidation of sparteine and debrisoquine: related pharmacogenetic entities. Clin. Pharmacol. Ther. 31, 184-186 (1982).
17. Kupfer, A., Schmid, B., Preisig, R. & Pfaff, G. Dextromethorphan as a safe probe for debrisoquine hydroxylation polymorphism. Lancet 2, 517-518 (1984).
18. Lennard, M.S., Silas, J.H., Freestone, S. & Trevethick, J. Defective metabolism of metoprolol in poor hydroxylators of debrisoquine. Br. J. Clin. Pharmacol. 14, 301-303 (1982).
19. Gonzalez, F.J., et al. Characterization of the common genetic defect in humans deficient in debrisoquine metabolism. Nature 331, 442-446 (1988).
20. Heim, M. & Meyer, U.A. Genotyping of poor metabolisers of debrisoquine by allele-specific PCR amplification. Lancet 336, 529-532 (1990).
21. Johansson, I., et al. Genetic analysis of the Chinese cytochrome P4502D locus: characterization of variant CYP2D6 genes present in subjects with diminished capacity for debrisoquine hydroxylation. Mol. Pharmacol. 46, 452-459 (1994).
22. Yue, Q.Y., et al. Disassociation between debrisoquine hydroxylation phenotype and genotype among Chinese. Lancet 2, 870 (1989).
23. Johansson, I., et al. Inherited amplification of an active gene in the cytochrome P450 CYP2D locus as a cause of ultrarapid metabolism of debrisoquine. Proc. Natl. Acad. Sci. U. S. A. 90, 11825-11829 (1993).
24. Aklillu, E., et al. Frequent distribution of ultrarapid metabolizers of debrisoquine in an ethiopian population carrying duplicated and multiduplicated functional CYP2D6 alleles. J. Pharmacol. Exp. Ther. 278, 441-446 (1996).
25. Dahl, M.L., Johansson, I., Bertilsson, L., Ingelman-Sundberg, M. & Sjoqvist, F. Ultrarapid hydroxylation of debrisoquine in a Swedish population. Analysis of the molecular genetic basis. J. Pharmacol. Exp. Ther. 274, 516-520 (1995).
26. Dahl, M.L., et al. Genetic analysis of the CYP2D locus in relation to debrisoquine hydroxylation capacity in Korean, Japanese and Chinese subjects. Pharmacogenetics 5, 159-164 (1995).
27. Kalow, W. Ethnic differences in drug metabolism. Clin. Pharmacokinet. 7, 373-400 (1982).
28. Bertilsson, L., et al. Pronounced differences between native Chinese and Swedish populations in the polymorphic hydroxylations of debrisoquin and S-mephenytoin. Clin. Pharmacol. Ther. 51, 388-397 (1992).
29. Ingelman-Sundberg, M. Genetic polymorphisms of cytochrome P450 2D6 (CYP2D6): clinical consequences, evolutionary aspects and functional diversity. Pharmacogenomics J 5, 6-13 (2004).
68
30. de la, H.J. & Popper, H. Urinary excretion of choline metabolites following choline administration in normals and patients with hepatobiliary diseases. J. Clin. Invest. 30, 463-470 (1951).
31. Baker, J.R. & Chaykin, S. The biosynthesis of trimethylamine-N-oxide. J. Biol. Chem. 237, 1309-1313 (1962).
32. Higgins, T., Chaykin, S., Hammond, K.B. & Humbert, J.R. Trimethylamine N-oxide synthesis: a human variant. Biochem. Med. 6, 392-396 (1972).
33. Humbert, J.A., Hammond, K.B. & Hathaway, W.E. Trimethylaminuria: the fish-odour syndrome. Lancet 2, 770-771 (1970).
34. Al-Waiz, M., Ayesh, R., Mitchell, S.C., Idle, J.R. & Smith, R.L. Trimethylaminuria (fish-odour syndrome): an inborn error of oxidative metabolism. The Lancet 329, 634-635 (1987).
35. Al-Waiz, M., Ayesh, R., Mitchell, S.C., Idle, J.R. & Smith, R.L. A genetic polymorphism of the N-oxidation of trimethylamine in humans. Clin. Pharmacol. Ther. 42, 588-594 (1987).
36. Lambert, D.M., et al. In vivo variability of TMA oxidation is partially mediated by polymorphisms of the FMO3 gene. Mol. Genet. Metab. 73, 224-229 (2001).
37. Thithapandha, A. A pharmacogenetic study of trimethylaminuria in Orientals. Pharmacogenetics 7, 497-501 (1997).
38. Lee, C.W., Tomlinson, B., Yeung, J.H., Lin, G. & Damani, L.A. Distribution of the N-oxidation of dietary-derived trimethylamine in a male Chinese population. Pharmacogenetics 10, 829-831 (2000).
39. Mitchell, S.C., Zhang, A.Q., Barrett, T., Ayesh, R. & Smith, R.L. Studies on the discontinuous N-oxidation of trimethylamine among Jordanian, Ecuadorian and New Guinean populations. Pharmacogenetics 7, 45-50 (1997).
40. Pettit, F.H., Orme-Johnson, W. & Ziegler, D.M. The requirement for flavin adenine dinucleotide by a liver microsmal oxygenase catalyzing the oxidation of alkylaryl amines. Biochem. Biophys. Res. Commun. 16, 444-448 (1964).
41. Ziegler, D.M., Poulsen, L.L. & McKee, E.M. Interaction of primary amines with a mixed-function amine oxidase isolated from pig liver microsomes. Xenobiotica 1, 523-531 (1971).
42. Williams, D.E., Ziegler, D.M., Nordin, D.J., Hale, S.E. & Masters, B.S. Rabbit lung flavin-containing monooxygenase is immunochemically and catalytically distinct from the liver enzyme. Biochem. Biophys. Res. Commun. 125, 116-122 (1984).
43. Tynes, R.E., Sabourin, P.J. & Hodgson, E. Identification of distinct hepatic and pulmonary forms of microsomal flavin-containing monooxygenase in the mouse and rabbit. Biochem. Biophys. Res. Commun. 126, 1069-1075 (1985).
44. Atta-Asafo-Adjei, E., Lawton, M.P. & Philpot, R.M. Cloning, sequencing, distribution, and expression in Escherichia coli of
69
flavin-containing monooxygenase 1C1. Evidence for a third gene subfamily in rabbits. J. Biol. Chem. 268, 9681-9689 (1993).
45. Dolphin, C.T., Shephard, E.A., Povey, S., Smith, R.L. & Phillips, I.R. Cloning, primary sequence and chromosomal localization of human FMO2, a new member of the flavin-containing mono-oxygenase family. Biochem. J. 287 ( Pt 1), 261-267 (1992).
46. Ozols, J. Multiple forms of liver microsomal flavin-containing monooxygenases: complete covalent structure of form 2. Arch. Biochem. Biophys. 290, 103-115 (1991).
47. Hernandez, D., Janmohamed, A., Chandan, P., Phillips, I.R. & Shephard, E.A. Organization and evolution of the flavin-containing monooxygenase genes of human and mouse: Identification of novel gene and pseudogene clusters. Pharmacogenetics 14, 117-130 (2004).
48. Lawton, M.P., et al. A nomenclature for the mammalian flavin-containing monooxygenase gene family based on amino acid sequence identities. Arch. Biochem. Biophys. 308, 254-257 (1994).
49. Zhang, J. & Cashman, J.R. Quantitative analysis of FMO gene mRNA levels in human tissues. Drug Metab. Dispos. 34, 19-26 (2006).
50. Lomri, N., Gu, Q. & Cashman, J.R. Molecular cloning of the flavin-containing monooxygenase (form II) cDNA from adult human liver. Proc. Natl. Acad. Sci. U. S. A. 89, 1685-1689 (1992).
51. Lomri, N., Yang, Z. & Cashman, J.R. Expression in Escherichia coli of the flavin-containing monooxygenase D (form II) from adult human liver: determination of a distinct tertiary amine substrate specificity. Chem. Res. Toxicol. 6, 425-429 (1993).
52. Cashman, J.R., et al. Human flavin-containing monooxygenase form 3: cDNA expression of the enzymes containing amino acid substitutions observed in individuals with trimethylaminuria. Chem. Res. Toxicol. 10, 837-841 (1997).
53. Dolphin, C.T., Janmohamed, A., Smith, R.L., Shephard, E.A. & Phillips, I.R. Missense mutation in flavin-containing mono-oxygenase 3 gene, FMO3, underlies fish-odour syndrome. Nat. Genet. 17, 491-494 (1997).
54. Treacy, E.P., et al. Mutations of the flavin-containing monooxygenase gene (FMO3) cause trimethylaminuria, a defect in detoxication. Hum. Mol. Genet. 7, 839-845 (1998).
55. Cashman, J.R. Structural and catalytic properties of the mammalian flavin-containing monooxygenase. Chem. Res. Toxicol. 8, 166-181 (1995).
56. Krueger, S.K. & Williams, D.E. Mammalian flavin-containing monooxygenases: Structure/function, genetic polymorphisms and role in drug metabolism. Pharmacol. Ther. 106, 357-387 (2005).
57. Cashman, J.R. Some distinctions between flavin-containing and cytochrome P450 monooxygenases. Biochem. Biophys. Res. Commun. 338, 599-604 (2005).
70
58. Ziegler, D.M. Flavin-containing monooxygenases: Catalytic mechanism and substrate specificities. Drug Metab. Rev. 19, 1-32 (1988).
59. Bickel, M.H. Liver metabolic reactions: Tertiary amine N-dealkylation, tertiary amine N-oxidation, N-oxide reduction, and N-oxide N-dealkylation : I. Tricyclic tertiary amine drugs. Arch. Biochem. Biophys. 148, 54-62 (1972).
60. Cereda, C., et al. Increased incidence of FMO1 gene single nucleotide polymorphisms in sporadic amyotrophic lateral sclerosis. Amyotroph Lateral Sc 7, 227-234 (2006).
61. Gagliardi, S., et al. Flavin-containing monooxygenase mRNA levels are up-regulated in als brain areas in SOD1-mutant mice. Neurotox Res 20, 150-158 (2011).
62. Hernandez, D., et al. Deletion of the mouse Fmo1 gene results in enhanced pharmacological behavioural responses to imipramine. Pharmacogenet Genomics 19, 289-299 (2009).
63. Hinrichs, A.L., et al. Common polymorphisms in FMO1 are associated with nicotine dependence. Pharmacogenetics and Genomics 21, 397-402 (2011).
64. Whetstine, J.R., et al. Ethnic differences in human flavin-containing monooxygenase 2 (FMO2) polymorphisms: detection of expressed protein in African-Americans. Toxicol. Appl. Pharmacol. 168, 216-224 (2000).
65. Francois, A.A., Nishida, C.R., de Montellano, P.R.O., Phillips, I.R. & Shephard, E.A. Human Flavin-Containing Monooxygenase 2.1 Catalyzes Oxygenation of the Antitubercular Drugs Thiacetazone and Ethionamide. Drug Metabolism and Disposition 37, 178-186 (2009).
66. Akerman, B.R., et al. Trimethylaminuria is caused by mutations of the FMO3 gene in a North American cohort. Mol. Genet. Metab. 68, 24-31 (1999).
67. Phillips, I.R. & Shephard, E.A. Flavin-containing monooxygenases: mutations, disease and drug response. Trends Pharmacol. Sci. 29, 294-301 (2008).
68. Overby, L.H., Carver, G.C. & Philpot, R.M. Quantitation and kinetic properties of hepatic microsomal and recombinant flavin-containing monooxygenases 3 and 5 from humans. Chem. Biol. Interact. 106, 29-45 (1997).
69. Koukouritaki, S.B., Simpson, P., Yeung, C.K., Rettie, A.E. & Hines, R.N. Human hepatic flavin-containing monooxygenases 1 (FMO1) and 3 (FMO3) developmental expression. Pediatr. Res. 51, 236-243 (2002).
70. Yeung, C.K., Lang, D.H., Thummel, K.E. & Rettie, A.E. Immunoquantitation of FMO1 in Human Liver, Kidney, and Intestine. Drug Metabolism and Disposition 28, 1107-1111 (2000).
71. Bhagwat, S.V., Bhamre, S., Boyd, M.R. & Ravindranath, V. Cerebral metabolism of imipramine and a purified flavin-containing
71
monooxygenase from human brain. Neuropsychopharmacology 15, 133-142 (1996).
72. Bhagwat, S.V., Bhamre, S., Boyd, M.R. & Ravindranath, V. Further characterization of rat brain flavin-containing monooxygenase. Metabolism of imipramine to its N-oxide. Biochem. Pharmacol. 51, 1469-1475 (1996).
73. Kawaji, A., Isobe, M. & Takabatake, E. Differences in enzymatic properties of flavin-containing monooxygenase in brain microsomes of rat, mouse, hamster, guinea pig and rabbit. Biol. Pharm. Bull. 20, 917-919 (1997).
74. Kawaji, A., Yamaguchi, T., Tochino, Y., Isobe, M. & Takabatake, E. Assessment of benzydamine N-oxidation mediated by flavin-containing monooxygenase in different regions of rat brain and liver using microdialysis. Biol. Pharm. Bull. 22, 1-4 (1999).
75. Kawaji, A., et al. Flavin-containing monooxygenase mediated metabolism of benzydamine in perfused brain and liver. Biochim. Biophys. Acta 1425, 41-46 (1998).
76. Strachan, T. & Read, A.P. Human molecular genetics 3, (Garland Press, London ; New York, 2004).
77. Kruglyak, L. & Nickerson, D.A. Variation is the spice of life. Nat. Genet. 27, 234-236 (2001).
78. Chakravarti, A. Single nucleotide polymorphisms: . . .to a future of genetic medicine. Nature 409, 822-823 (2001).
79. Reich, D.E., et al. Linkage disequilibrium in the human genome. Nature 411, 199-204 (2001).
80. Mayo, O. A century of Hardy-Weinberg equilibrium. Twin Res Hum Genet 11, 249-256 (2008).
81. Mueller, J.C. Linkage disequilibrium for different scales and applications. Brief Bioinform 5, 355-364 (2004).
82. Akey, J., Jin, L. & Xiong, M. Haplotypes vs single marker linkage disequilibrium tests: what do we gain? Eur. J. Hum. Genet. 9, 291-300 (2001).
83. Tandon, R., Nasrallah, H.A. & Keshavan, M.S. Schizophrenia, "just the facts" 4. Clinical features and conceptualization. Schizophr. Res. 110, 1-23 (2009).
84. American Psychiatric Association. Diagnostic criteria from DSM-IV-TR, (American Psychiatric Association, Washington, D.C., 2000).
85. Lehman, A.F., et al. Practice guideline for the treatment of patients with schizophrenia, second edition. Am. J. Psychiatry 161, 1-56 (2004).
86. Robinson, D.G., Woerner, M.G., McMeniman, M., Mendelowitz, A. & Bilder, R.M. Symptomatic and functional recovery from a first episode of schizophrenia or schizoaffective disorder. Am. J. Psychiatry 161, 473-479 (2004).
72
87. Thieda, P., Beard, S., Richter, A. & Kane, J. An economic review of compliance with medication therapy in the treatment of schizophrenia. Psychiatr. Serv. 54, 508-516 (2003).
88. Grunder, G., Hippius, H. & Carlsson, A. The 'atypicality' of antipsychotics: a concept re-examined and re-defined. Nat Rev Drug Discov 8, 197-202 (2009).
89. Tandon, R. Antipsychotics in the treatment of schizophrenia: an overview. J. Clin. Psychiatry 72 Suppl 1, 4-8 (2011).
90. Lieberman, J.A., et al. Effectiveness of Antipsychotic Drugs in Patients with Chronic Schizophrenia. N. Engl. J. Med. 353, 1209-1223 (2005).
91. Lewis, S., et al. First generation versus second generation (non-clozapine) antipsychotic drugs versus clozapine in schizophrenia: The CUtLASS trials. Neuropsychopharmacology 30, S31-S32 (2005).
92. Kahn, R.S., et al. Effectiveness of antipsychotic drugs in first-episode schizophrenia and schizophreniform disorder: an open randomised clinical trial. The Lancet 371, 1085-1097 (2008).
93. Leucht, S., et al. Second-generation versus first-generation antipsychotic drugs for schizophrenia: a meta-analysis. The Lancet 373, 31-41 (2009).
94. De Hert, M., Detraux, J., van Winkel, R., Yu, W. & Correll, C.U. Metabolic and cardiovascular adverse effects associated with antipsychotic drugs. Nat Rev Endocrinol 8, 114-126 (2012).
95. Melkersson, K. & Dahl, M.L. Adverse metabolic effects associated with atypical antipsychotics: literature review and clinical implications. Drugs 64, 701-723 (2004).
96. Baldwin, D.S. & Montgomery, S.A. First clinical experience with olanzapine (LY 170053): results of an open-label safety and dose-ranging study in patients with schizophrenia. Int. Clin. Psychopharmacol. 10, 239-244 (1995).
97. Bymaster, F.P., et al. In vitro and in vivo biochemistry of olanzapine: a novel, atypical antipsychotic drug. J. Clin. Psychiatry 58 Suppl 10, 28-36 (1997).
98. Moore, N.A., Leander, J.D., Benvenga, M.J., Gleason, S.D. & Shannon, H. Behavioral pharmacology of olanzapine: a novel antipsychotic drug. J. Clin. Psychiatry 58 Suppl 10, 37-44 (1997).
99. Davis, J.M. & Chen, N. Dose response and dose equivalence of antipsychotics. J. Clin. Psychopharmacol. 24, 192-208 (2004).
100. Citrome, L. & Kantrowitz, J.T. Olanzapine dosing above the licensed range is more efficacious than lower doses: fact or fiction? Expert Rev Neurother 9, 1045-1058 (2009).
101. Kinon, B.J., et al. Standard and higher dose of olanzapine in patients with schizophrenia or schizoaffective disorder: a randomized, double-blind, fixed-dose study. J. Clin. Psychopharmacol. 28, 392-400 (2008).
73
102. Howes, O.D., et al. Mechanisms underlying psychosis and antipsychotic treatment response in schizophrenia: insights from PET and SPECT imaging. Curr. Pharm. Des. 15, 2550-2559 (2009).
103. Nordstrom, A.L., et al. Central D2-dopamine receptor occupancy in relation to antipsychotic drug effects: a double-blind PET study of schizophrenic patients. Biol. Psychiatry 33, 227-235 (1993).
104. Kapur, S., et al. 5-HT2 and D2 receptor occupancy of olanzapine in schizophrenia: a PET investigation. Am. J. Psychiatry 155, 921-928 (1998).
105. Kassahun, K., et al. Disposition and biotransformation of the antipsychotic agent olanzapine in humans. Drug Metab. Dispos. 25, 81-93 (1997).
106. Callaghan, J.T., Bergstrom, R.F., Ptak, L.R. & Beasley, C.M. Olanzapine. Pharmacokinetic and pharmacodynamic profile. Clin. Pharmacokinet. 37, 177-193 (1999).
107. Perry, P.J., Lund, B.C., Sanger, T. & Beasley, C. Olanzapine plasma concentrations and clinical response: acute phase results of the North American Olanzapine Trial. J. Clin. Psychopharmacol. 21, 14-20 (2001).
108. Fellows, L., et al. Investigation of target plasma concentration-effect relationships for olanzapine in schizophrenia. Ther. Drug Monit. 25, 682-689 (2003).
109. Skogh, E., Reis, M., Dahl, M.L., Lundmark, J. & Bengtsson, F. Therapeutic drug monitoring data on olanzapine and its N-demethyl metabolite in the naturalistic clinical setting. Ther. Drug Monit. 24, 518-526 (2002).
110. Skogh, E., Sjodin, I., Josefsson, M. & Dahl, M.L. High correlation between serum and cerebrospinal fluid olanzapine concentrations in patients with schizophrenia or schizoaffective disorder medicating with oral olanzapine as the only antipsychotic drug. J. Clin. Psychopharmacol. 31, 4-9 (2011).
111. Tauscher, J., Jones, C., Remington, G., Zipursky, R.B. & Kapur, S. Significant dissociation of brain and plasma kinetics with antipsychotics. Mol. Psychiatry 7, 317-321 (2002).
112. Mattiuz, E., et al. Disposition and Metabolism of Olanzapine in Mice, Dogs, and Rhesus Monkeys. Drug Metabolism and Disposition 25, 573-583 (1997).
113. Kassahun, K., Mattiuz, E., Franklin, R. & Gillespie, T. Olanzapine 10-N-Glucuronide. Drug Metabolism and Disposition 26, 848-855 (1998).
114. Calligaro, D.O., Fairhurst, J., Hotten, T.M., Moore, N.A. & Tupper, D.E. The synthesis and biological activity of some known and putative metabolites of the atypical antipsychotic agent olanzapine (LY170053). Bioorganic & Medicinal Chemistry Letters 7, 25-30 (1997).
74
115. Green, M.D. & Tephly, T.R. Glucuronidation of Amine Substrates by Purified and Expressed UDP-Glucuronosyltransferase Proteins. Drug Metabolism and Disposition 26, 860-867 (1998).
116. Linnet, K. Glucuronidation of olanzapine by cDNA-expressed human UDP-glucuronosyltransferases and human liver microsomes. Hum Psychopharmacol 17, 233-238 (2002).
117. Erickson-Ridout, K.K., Zhu, J. & Lazarus, P. Olanzapine metabolism and the significance of UGT1A448V and UGT2B1067Y variants. Pharmacogenet Genomics 21, 539-551 (2011).
118. Kaivosaari, S., et al. Nicotine Glucuronidation and the Human UDP-Glucuronosyltransferase UGT2B10. Mol. Pharmacol. 72, 761-768 (2007).
119. Ring, B.J., et al. Identification of the human cytochromes P450 responsible for the in vitro formation of the major oxidative metabolites of the antipsychotic agent olanzapine. J. Pharmacol. Exp. Ther. 276, 658-666 (1996).
120. Kalow, W. & Tang, B.K. The use of caffeine for enzyme assays: a critical appraisal. Clin. Pharmacol. Ther. 53, 503-514 (1993).
121. Fuhr, U. & Rost, K.L. Simple and reliable CYP1A2 phenotyping by the paraxanthine/caffeine ratio in plasma and in saliva. Pharmacogenetics 4, 109-116 (1994).
122. Carrillo, J.A., et al. Evaluation of caffeine as an in vivo probe for CYP1A2 using measurements in plasma, saliva, and urine. Ther. Drug Monit. 22, 409-417 (2000).
123. Shirley, K.L., et al. Correlation of cytochrome P450 (CYP) 1A2 activity using caffeine phenotyping and olanzapine disposition in healthy volunteers. Neuropsychopharmacology 28, 961-966 (2003).
124. Carrillo, J.A., et al. Role of the smoking-induced cytochrome P450 (CYP)1A2 and polymorphic CYP2D6 in steady-state concentration of olanzapine. J. Clin. Psychopharmacol. 23, 119-127 (2003).
125. Dahl, M.L. Cytochrome p450 phenotyping/genotyping in patients receiving antipsychotics: useful aid to prescribing? Clin. Pharmacokinet. 41, 453-470 (2002).
126. Hägg, S., Spigset, O., Lakso, H.A. & Dahlqvist, R. Olanzapine disposition in humans is unrelated to CYP1A2 and CYP2D6 phenotypes. Eur. J. Clin. Pharmacol. 57, 493-497 (2001).
127. Bigos, K.L., et al. Sex, race, and smoking impact olanzapine exposure. J. Clin. Pharmacol. 48, 157-165 (2008).
128. Olesen, O.V. & Linnet, K. Olanzapine serum concentrations in psychiatric patients given standard doses: the influence of comedication. Ther. Drug Monit. 21, 87-90 (1999).
129. Gex-Fabry, M., Balant-Gorgia, A.E. & Balant, L.P. Therapeutic drug monitoring of olanzapine: The combined effect of age, gender, smoking, and comedication. Ther. Drug Monit. 25, 46-53 (2003).
130. Patel, M.X., et al. Plasma olanzapine in relation to prescribed dose and other factors data from a therapeutic drug monitoring service, 1999-2009. J. Clin. Psychopharmacol. 31, 411-417 (2011).
75
131. de Leon, J. & Diaz, F.J. A meta-analysis of worldwide studies demonstrates an association between schizophrenia and tobacco smoking behaviors. Schizophr. Res. 76, 135-157 (2005).
132. Weiss, U., Marksteiner, J., Kemmler, G., Saria, A. & Aichhorn, W. Effects of age and sex on olanzapine plasma concentrations. J. Clin. Psychopharmacol. 25, 570-574 (2005).
133. Haslemo, T., Eikeseth, P.H., Tanum, L., Molden, E. & Refsum, H. The effect of variable cigarette consumption on the interaction with clozapine and olanzapine. Eur. J. Clin. Pharmacol. 62, 1049-1053 (2006).
134. Nozawa, M., et al. The relationship between the response of clinical symptoms and plasma olanzapine concentration, based on pharmacogenetics: Juntendo University Schizophrenia Projects (JUSP). Ther. Drug Monit. 30, 35-40 (2008).
135. Citrome, L.M.D.M.P.H., et al. Olanzapine Plasma Concentrations After Treatment With 10, 20, and 40 mg/d in Patients With Schizophrenia: An Analysis of Correlations With Efficacy, Weight Gain, and Prolactin Concentration. J. Clin. Psychopharmacol. 29, 278-283 (2009).
136. Ghotbi, R., et al. Carriers of the UGT1A4 142T>G gene variant are predisposed to reduced olanzapine exposure--an impact similar to male gender or smoking in schizophrenic patients. Eur. J. Clin. Pharmacol. 66, 465-474 (2010).
137. Ma, Q. & Lu, A.Y.H. CYP1A Induction and Human Risk Assessment: An Evolving Tale of in Vitro and in Vivo Studies. Drug Metabolism and Disposition 35, 1009-1016 (2007).
138. Lowe, E.J. & Ackman, M.L. Impact of tobacco smoking cessation on stable clozapine or olanzapine treatment. Ann. Pharmacother. 44, 727-732 (2010).
139. Kelly, D.L., Conley, R.R. & Tamminga, C.A. Differential olanzapine plasma concentrations by sex in a fixed-dose study. Schizophr. Res. 40, 101-104 (1999).
140. Haslemo, T., Refsum, H. & Molden, E. The effect of ethinylestradiol-containing contraceptives on the serum concentration of olanzapine and N-desmethyl olanzapine. Br. J. Clin. Pharmacol. 71, 611-615 (2011).
141. Weigmann, H., et al. Fluvoxamine but not sertraline inhibits the metabolism of olanzapine: Evidence from a therapeutic drug monitoring service. Ther. Drug Monit. 23, 410-413 (2001).
142. Linnet, K. & Olesen, O.V. Free and glucuronidated olanzapine serum concentrations in psychiatric patients: influence of carbamazepine comedication. Ther. Drug Monit. 24, 512-517 (2002).
143. Sidhu, J., et al. Pharmacokinetics and tolerability of lamotrigine and olanzapine coadministered to healthy subjects. Br. J. Clin. Pharmacol. 61, 420-426 (2006).
76
144. Spina, E., et al. Effect of adjunctive lamotrigine treatment on the plasma concentrations of clozapine, risperidone and olanzapine in patients with schizophrenia or bipolar disorder. Ther. Drug Monit. 28, 599-602 (2006).
145. Botts, S., et al. Estimating the effects of co-medications on plasma olanzapine concentrations by using a mixed model. Prog. Neuropsychopharmacol. Biol. Psychiatry 32, 1453-1458 (2008).
146. Bergemann, N., Kress, K.R., Abu-Tair, F., Frick, A. & Kopitz, J. Valproate lowers plasma concentration of olanzapine. J. Clin. Psychopharmacol. 26, 432-434 (2006).
147. Spina, E., et al. Effect of Valproate on Olanzapine Plasma Concentrations in Patients With Bipolar or Schizoaffective Disorder. Ther. Drug Monit. 31, 758-763 (2009).
148. Strassburg, C.P., Oldhafer, K., Manns, M.P. & Tukey, R.H. Differential expression of the UGT1A locus in human liver, biliary, and gastric tissue: identification of UGT1A7 and UGT1A10 transcripts in extrahepatic tissue. Mol. Pharmacol. 52, 212-220 (1997).
149. Zhang, W., Liu, W., Innocenti, F. & Ratain, M.J. Searching for tissue-specific expression pattern-linked nucleotides of UGT1A isoforms. PLoS One 2, e396 (2007).
150. Nakamura, A., Nakajima, M., Yamanaka, H., Fujiwara, R. & Yokoi, T. Expression of UGT1A and UGT2B mRNA in human normal tissues and various cell lines. Drug Metab. Dispos. 36, 1461-1464 (2008).
151. Ohno, S. & Nakajin, S. Determination of mRNA expression of human UDP-glucuronosyltransferases and application for localization in various human tissues by real-time reverse transcriptase-polymerase chain reaction. Drug Metab. Dispos. 37, 32-40 (2009).
152. Wiener, D., Fang, J.L., Dossett, N. & Lazarus, P. Correlation between UDP-glucuronosyltransferase genotypes and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone glucuronidation phenotype in human liver microsomes. Cancer Res. 64, 1190-1196 (2004).
153. Mori, A., Maruo, Y., Iwai, M., Sato, H. & Takeuchi, Y. UDP-glucuronosyltransferase 1A4 polymorphisms in a Japanese population and kinetics of clozapine glucuronidation. Drug Metab. Dispos. 33, 672-675 (2005).
154. Sun, D., et al. Characterization of tamoxifen and 4-hydroxytamoxifen glucuronidation by human UGT1A4 variants. Breast Cancer Res 8, R50 (2006).
155. Benoit-Biancamano, M.O., et al. A pharmacogenetics study of the human glucuronosyltransferase UGT1A4. Pharmacogenet Genomics 19, 945-954 (2009).
156. Argikar, U.A. & Remmel, R.P. Effect of Aging on Glucuronidation of Valproic Acid in Human Liver Microsomes and the Role of UDP-
77
Glucuronosyltransferase UGT1A4, UGT1A8, and UGT1A10. Drug Metabolism and Disposition 37, 229-236 (2009).
157. Erichsen, T.J., et al. Genetic variability of aryl hydrocarbon receptor (AhR)-mediated regulation of the human UDP glucuronosyltransferase (UGT) 1A4 gene. Toxicol. Appl. Pharmacol. 230, 252-260 (2008).
158. Haslemo, T., et al. UGT1A4*3 Encodes Significantly Increased Glucuronidation of Olanzapine in Patients on Maintenance Treatment and in Recombinant Systems. Clin. Pharmacol. Ther. 92, 221-227 (2012).
159. Chen, G., Dellinger, R.W., Gallagher, C.J., Sun, D. & Lazarus, P. Identification of a prevalent functional missense polymorphism in the UGT2B10 gene and its association with UGT2B10 inactivation against tobacco-specific nitrosamines. Pharmacogenetics and Genomics 18, 181-191 (2008).
160. Shimada, T., Yamazaki, H., Mimura, M., Inui, Y. & Guengerich, F.P. Interindividual variations in human liver cytochrome P-450 enzymes involved in the oxidation of drugs, carcinogens and toxic chemicals: studies with liver microsomes of 30 Japanese and 30 Caucasians. J. Pharmacol. Exp. Ther. 270, 414-423 (1994).
161. Schweikl, H., et al. Expression of CYP1A1 and CYP1A2 genes in human liver. Pharmacogenetics 3, 239-249 (1993).
162. Corchero, J., Pimprale, S., Kimura, S. & Gonzalez, F.J. Organization of the CYP1A cluster on human chromosome 15: implications for gene regulation. Pharmacogenetics 11, 1-6 (2001).
163. Ueda, R., et al. A common regulatory region functions bidirectionally in transcriptional activation of the human CYP1A1 and CYP1A2 genes. Mol. Pharmacol. 69, 1924-1930 (2006).
164. Ikeya, K., et al. Human CYP1A2: sequence, gene structure, comparison with the mouse and rat orthologous gene, and differences in liver 1A2 mRNA expression. Mol. Endocrinol. 3, 1399-1408 (1989).
165. Gunes, A. & Dahl, M.L. Variation in CYP1A2 activity and its clinical implications: influence of environmental factors and genetic polymorphisms. Pharmacogenomics 9, 625-637 (2008).
166. Laika, B., Leucht, S., Heres, S., Schneider, H. & Steimer, W. Pharmacogenetics and olanzapine treatment: CYP1A2*1F and serotonergic polymorphisms influence therapeutic outcome. Pharmacogenomics J 10, 20-29 (2010).
167. Cashman, J.R., Zhang, J., Nelson, M.R. & Braun, A. Analysis of flavin-containing monooxygenase 3 genotype data in populations administered the anti-schizophrenia agent olanzapine. Drug Metab Lett 2, 100-114 (2008).
168. Sachse, C., Brochmoller, J., Bauer, S. & Roots, I. Functional significance of a C -> A polymorphism in intron I of the cytochrome P450 CYP1A2 gene tested with caffeine. Br. J. Clin. Pharmacol. 47, 445-449 (1999).
78
169. Ghotbi, R., et al. Comparisons of CYP1A2 genetic polymorphisms, enzyme activity and the genotype-phenotype relationship in Swedes and Koreans. Eur. J. Clin. Pharmacol. 63, 537-546 (2007).
170. Gunes, A., et al. Influence of genetic polymorphisms, smoking, gender and age on CYP1A2 activity in a Turkish population. Pharmacogenomics 10, 769-778 (2009).
171. Dobrinas, M., et al. Impact of smoking, smoking cessation, and genetic polymorphisms on CYP1A2 activity and inducibility. Clin. Pharmacol. Ther. 90, 117-125 (2011).
172. Obase, Y., et al. Polymorphisms in the CYP1A2 gene and theophylline metabolism in patients with asthma. Clin. Pharmacol. Ther. 73, 468-474 (2003).
173. Shimoda, K., et al. Lack of impact of CYP1A2 genetic polymorphism (C/A polymorphism at position 734 in intron 1 and G/A polymorphism at position −2964 in the 5′-flanking region of CYP1A2) on the plasma concentration of haloperidol in smoking male Japanese with schizophrenia. Prog. Neuropsychopharmacol. Biol. Psychiatry 26, 261-265 (2002).
174. Chen, X., et al. The G-113A polymorphism in CYP1A2 affects the caffeine metabolic ratio in a Chinese population. Clin. Pharmacol. Ther. 78, 249-259 (2005).
175. Jiang, Z., et al. Toward the evaluation of function in genetic variability: characterizing human SNP frequencies and establishing BAC-transgenic mice carrying the human CYP1A1_CYP1A2 locus. Hum. Mutat. 25, 196-206 (2005).
176. Amin, N., et al. Genome-wide association analysis of coffee drinking suggests association with CYP1A1/CYP1A2 and NRCAM. Mol. Psychiatry (2011).
177. Cornelis, M.C., et al. Genome-wide meta-analysis identifies regions on 7p21 (AHR) and 15q24 (CYP1A2) as determinants of habitual caffeine consumption. PLoS Genet 7, e1002033 (2011).
178. Sulem, P., et al. Sequence variants at CYP1A1-CYP1A2 and AHR associate with coffee consumption. Hum. Mol. Genet. 20, 2071-2077 (2011).
179. Nebert, D.W., Dalton, T.P., Okey, A.B. & Gonzalez, F.J. Role of aryl hydrocarbon receptor-mediated induction of the CYP1 enzymes in environmental toxicity and cancer. J. Biol. Chem. 279, 23847-23850 (2004).
180. Hines, R.N., Hopp, K.A., Franco, J., Saeian, K. & Begun, F.P. Alternative processing of the human FMO6 gene renders transcripts incapable of encoding a functional flavin-containing monooxygenase. Mol. Pharmacol. 62, 320-325 (2002).
181. Brunelle, A., et al. Characterization of two human flavin-containing monooxygenase (form 3) enzymes expressed in Escherichia coli as maltose binding protein fusions. Drug Metab. Dispos. 25, 1001-1007 (1997).
79
182. Cashman, J.R., Akerman, B.R., Forrest, S.M. & Treacy, E.P. Population-specific polymorphisms of the human FMO3 gene: Significance for detoxication. Drug Metab. Dispos. 28, 169-173 (2000).
183. Stormer, E., Roots, I. & Brockmoller, J. Benzydamine N-oxidation as an index reaction reflecting FMO activity in human liver microsomes and impact of FMO3 polymorphisms on enzyme activity. Br. J. Clin. Pharmacol. 50, 553-561 (2000).
184. Shimizu, M., et al. Effect of genetic variants of the human flavin-containing monooxygenase 3 on N- and S-oxygenation activities. Drug Metab. Dispos. 35, 328-330 (2007).
185. Zschocke, J., et al. Mild trimethylaminuria caused by common variants in FMO3 gene. Lancet 354, 834-835 (1999).
186. Shimizu, M., Cashman, J.R. & Yamazaki, H. Transient trimethylaminuria related to menstruation. BMC Med Genet 8(2007).
187. Kang, J.H., et al. Phenotypes of flavin-containing monooxygenase activity determined by ranitidine N-oxidation are positively correlated with genotypes of linked FMO3 gene mutations in a Korean population. Pharmacogenetics 10, 67-78 (2000).
188. Hisamuddin, I.M., et al. Genetic polymorphisms of human flavin monooxygenase 3 in sulindac-mediated primary chemoprevention of familial adenomatous polyposis. Clin. Cancer Res. 10, 8357-8362 (2004).
189. Hisamuddin, I.M., et al. Genetic polymorphisms of flavin monooxygenase 3 in sulindac-induced regression of colorectal adenomas in familial adenomatous polyposis. Cancer Epidemiol. Biomarkers Prev. 14, 2366-2369 (2005).
190. Lee, K.H., Song Ki, S., Shin, S.G. & Cha, Y.N. Correlation between FMO3 genotypes and FMO activity measured by using trimethylamine N-oxidation and ranitidine N-oxidation. Journal of Korean Society for Clinical Pharmacology and Therapeutics 8, 44-59 (2000).
191. Cashman, J.R., Xiong, Y.N., Xu, L. & Janowsky, A. N-oxygenation of amphetamine and methamphetamine by the human flavin-containing monooxygenase (form 3): role in bioactivation and detoxication. J. Pharmacol. Exp. Ther. 288, 1251-1260 (1999).
192. Sachse, C., et al. Flavin monooxygenase 3 (FMO3) polymorphism in a white population: Allele frequencies, mutation linkage, and functional effects on clozapine and caffeine metabolism. Clin. Pharmacol. Ther. 66, 431-438 (1999).
193. Furnes, B., Feng, J., Sommer, S.S. & Schlenk, D. Identification of novel variants of the flavin-containing monooxygenase gene family in African Americans. Drug Metabolism and Disposition 31, 187-193 (2003).
194. Shimizu, M., Fujita, H., Aoyama, T. & Yamazaki, H. Three novel single nucleotide polymorphisms of the FMO3 gene in a Japanese
80
population. Drug metabolism and pharmacokinetics. 21, 245-247 (2006).
195. Allerston, C.K., Vetti, H.H., Houge, G., Phillips, I.R. & Shephard, E.A. A novel mutation in the flavin-containing monooxygenase 3 gene (FMO3) of a Norwegian family causes trimethylaminuria. Mol. Genet. Metab. 98, 198-202 (2009).
196. Teresa, E., et al. A spectrum of molecular variation in a cohort of Italian families with trimethylaminuria: identification of three novel mutations of the FM03 gene. Mol. Genet. Metab. 88, 192-195 (2006).
197. Lattard, V., et al. Two new polymorphisms of the FMO3 gene in Caucasian and African-American populations: Comparative genetic and functional studies. Drug Metabolism and Disposition 31, 854-860 (2003).
198. Koukouritaki, S.B., Poch, M.T., Cabacungan, E.T., McCarver, D.G. & Hines, R.N. Discovery of novel flavin-containing monooxygenase 3 (FMO3) single nucleotide polymorphisms and functional analysis of upstream haplotype variants. Mol. Pharmacol. 68, 383-392 (2005).
199. Hines, R.N., et al. Genetic variability at the human FMO1 locus: significance of a basal promoter yin yang 1 element polymorphism (FMO1*6). J. Pharmacol. Exp. Ther. 306, 1210-1218 (2003).
200. Furnes, B. & Schlenk, D. Evaluation of Xenobiotic N- and S-Oxidation by Variant Flavin-Containing Monooxygenase 1 (FMO1) Enzymes. Toxicol. Sci. 78, 196-203 (2004).
201. Matimba, A., et al. Establishment of a biobank and pharmacogenetics database of African populations. Eur. J. Hum. Genet. 16, 780-783 (2008).
202. Josefsson, M., Roman, M., Skogh, E. & Dahl, M.L. Liquid chromatography/tandem mass spectrometry method for determination of olanzapine and N-desmethylolanzapine in human serum and cerebrospinal fluid. J. Pharm. Biomed. Anal. 53, 576-582 (2010).
203. Barrett, J.C., Fry, B., Maller, J. & Daly, M.J. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 21, 263-265 (2005).
204. Weckx, S., et al. novoSNP, a novel computational tool for sequence variation discovery. Genome Res. 15, 436-442 (2005).
205. Excoffier, L., Laval, G. & Schneider, S. Arlequin (version 3.0): an integrated software package for population genetics data analysis. Evol Bioinform Online 1, 47-50 (2005).
206. Dudbridge, F. Likelihood-based association analysis for nuclear families and unrelated subjects with missing genotype data. Hum. Hered. 66, 87-98 (2008).
207. Nalla, V.K. & Rogan, P.K. Automated splicing mutation analysis by information theory. Hum. Mutat. 25, 334-342 (2005).
81
208. Shriver, M.D. & Kittles, R.A. Genetic ancestry and the search for personalized genetic histories. Nat Rev Genet 5, 611-618 (2004).
209. Nagashima, S., et al. Inter-individual Variation in Flavin-containing Monooxygenase 3 in Livers from Japanese: Correlation with Hepatic Transcription Factors. Drug Metab. Pharmacokinet. 24, 218-225 (2009).
210. Aklillu, E., et al. Genetic polymorphism of CYP1A2 in Ethiopians affecting induction and expression: Characterization of novel haplotypes with single-nucleotide polymorphisms in intron 1. Mol. Pharmacol. 64, 659-669 (2003).
211. Ingelman-Sundberg, M., Sim, S.C., Gomez, A. & Rodriguez-Antona, C. Influence of cytochrome P450 polymorphisms on drug therapies: Pharmacogenetic, pharmacoepigenetic and clinical aspects. Pharmacology & Therapeutics 116, 496-526 (2007).
212. Bickel, M.H. The Pharmacology and Biochemistry of N-oxides. Pharmacol. Rev. 21, 325-355 (1969).
213. Krueger, S.K., VanDyke, J.E., Williams, D.E. & Hines, R.N. The role of flavin-containing monooxygenase (FMO) in the metabolism of tamoxifen and other tertiary amines. Drug Metab. Rev. 38, 139-147 (2006).
214. Koukouritaki, S.B. & Hines, R.N. Flavin-containing monooxygenase genetic polymorphism: Impact on chemical metabolism and drug development. Pharmacogenomics 6, 807-822 (2005).
Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 805
Editor: The Dean of the Faculty of Medicine
A doctoral dissertation from the Faculty of Medicine, UppsalaUniversity, is usually a summary of a number of papers. A fewcopies of the complete dissertation are kept at major Swedishresearch libraries, while the summary alone is distributedinternationally through the series Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty ofMedicine.
Distribution: publications.uu.seurn:nbn:se:uu:diva-179957
ACTAUNIVERSITATIS
UPSALIENSISUPPSALA
2012
![Effect of Olanzapine on Clinical and …downloads.hindawi.com/journals/schizort/2018/3968015.pdfnights.M¨ulleretal.[ ]reportedanincreaseinSE,SWS, andREMsleepaerfour-weekadministrationofolanzapine](https://static.documents.pub/doc/80x56/5f4abc30d174a336e86b10e2/effect-of-olanzapine-on-clinical-and-nightsmulleretal-reportedanincreaseinsesws.jpg)
