Trimethylamine 7V-Oxidation in Chinese: Lee, Chi-Wai

Trimethylamine 7V-Oxidation in
Chinese
.',’& •
•〜？. \
A Thesis submitted to The Chinese University of Hong Kong:fb《 the degree of Master
ofPhilosophy in Pharmacology by .务 \
；-
Lee, Chi-wai
B.Sc. (Hons), Cert. Pharmacol. & Pharm. Mgt.
Department ofPharmacology
Faculty ofMedicine
The Chinese University ofHong Kong
Shatin, N e w Territories
Hong Kong S A R
China
April 1999
• The Chinese University ofHong Kong 1999

y ^ ^ ^ ^ s ^
Z ^ l i S A S XHV^^^^m
^ : 1 ! ^ 1
g 6561 W * 1 Jsl
^W
…一一：
^:^..-.-..
‘：
i•'<
^
i ‘
Declaration
N o portion ofthe work referred to in this thesis has been submitted in support of an
application for another degree or qualification of this or any other university or other
institution oflearning.
i
Acknowledgements
I wish to express m y sincere gratitude to m y supervisors Profs. Jim L. A. Damani,
John H . K . Yeung, and Brian Tomlinson for their invaluable intellectual support
during the course of this study.
I also thank Prof. Kenneth Raymond in the Department ofPharmacy for his advice
in statistical analysis and Prof. G. Lin in the Department ofPharmacology for her
cooperation throughout the period of this study.
Finally, I a m also grateful to Mr. Sam K . S. Chung for his technical assistance in
the Department ofPharmacy and Ms. Karen S. Y . W o n g for her help in collecting
urine samples from patients in the Department of Medicine and Therapeutics,
Prince ofWales Hospital.
ii
Abstract
Inter-ethnic differences in drug metabolism have been reported in numerous studies,

which compared Asian or Chinese to Caucasian populations. A discontinuous
distribution of the iV-oxidation oftrimethylamine ( T M A ) to trimethylamine A^-oxide
( T M A O ) has been described in population samples from Britain (Caucasians), Jordan,
Ecuador, N e w Guinea and Thailand which appear to be due to the genetic
polymorphism oftheflavin-containingmonooxygenase (FMO). In view ofthe lack of
published data for Chinese populations, this study was conducted to determine the
distribution of T M A A^-oxidation in the Chinese population in Hong Kong. Other
factors that may affect TMAA^-oxidation in man have also been examined in this study.
Chinese male subjects (n = 270) were recruited. Urine samples were collected for the
determination of T M A and T M A O by gas chromatography. The percentage of total
T M A excreted as T M A O (calculated as T M A ) in urine was determined. The frequency
distribution for the percentage of total T M A excreted as T M A O showed a similar
pattern to previous studies in British Caucasian populations. Most subjects excreted
more than 90 % o f T M A as T M A O but the distribution was skewed with an asymmetric
tail extending toward lower values. Totally, the combined result from 24-hour and spot
urine samples showed seven subjects (3 % ) who excreted 80-90 % o f T M A as T M A O
and another two subjects (1 % ) who excreted 74 % and 75 % of T M A as T M A O
respectively.
Data were also analyzed to investigate the effect of age and diets on the ability o f T M A
A^-oxidation. From the results obtained, regarding the activity of F M O on T M A N-
oxidation, there seems to be no direct relationship (correlation) to the age of male
volunteers, although a U-shaped relationship for the % T M A oxidation at different ages
was obtained. The difference in the amounts o f T M A and T M A O excreted within the
urine collection periods might be due to other factors including dietary changes. The
dietary study showed that seafood is the major source of T M A and T M A O , such that
iii
intake of seafood (as high intakes of T M A and T M A O ) stimulated the initial F M O
activity and led to higher percentage T M A A^-oxidation. It also produced persisting
effects on the rate o f T M A O excretion and the percentage o f T M A excreted as T M A O
in urine.
Effects of diseases and medications were also reported in this study. Damage ofthe
liver (cirrhosis) impaired the liver function and reduced the liver F M O activity, which
led to a lower percentage value o f T M A iV-oxidation. Therefore, cautious prescription
is needed for the drugs that are metabolized by F M O in these patients. The effects of
two different drugs, atorvastatin and hydroxychloroquine (HCQ), were also studied.
Atorvastatin had no effect on the percentage of T M A 7V-oxidation. For H C Q , the mean
values for percentage TMAA^-oxidation were similar in patients treated with it to those
who were not. However, three patients treated with H C Q showed reduced percentage
of T M A 7V"0xidati0n (78 - 88 %). The effects of other drugs on T M A 7V~oxidation
should be systematically studied.
In conclusion, at least 1 % ofthe Chinese population studied may have mild deficiency
in TMAA^-oxidation, which may represent impairment due to genetic or other factors.
There appears to be no clear relationship between age and the degree of T M A N-
oxidation but dietary factors such as a high intake o f T M A and T M A O for example in
seafood appear to influence the F M O activity as demonstrated by T M A TV-oxidation.
Cirrhosis of the liver is associated with reduced T M A A/-oxidation, but diabetes,
hyperlipidaemia and hypertension had no significant effect, and treatment with the
drugs atorvastatin or hydroxychloroquine also had no significant influence on T M A
iV"Oxidation.
iv
槪要
許多硏究指出藥物之新陳代謝在種族間，包括亞洲人，華人以至白人，都存在
不同的差異。由三甲胺aMA)轉化至7V-氧化三甲胺aMAC¾之7^-氧化作用之硏究
顯示出英國白人，約旦人，厄瓜多爾人及泰國人間出現不連貫和不同程度的轉
化及分佈，這些都是由含黃素單氧化（FMC¾的遺傳多型變態所弓丨致°本硏究描
述三甲胺之7^-氧化作用及其於華人體內的分佈情況和其他可能改變上述#氧化
作用的因素之影響，這些資料在以前是從未有文獻發表過的°
本硏究共有二百七十位華裔男性參與並使用氣相色譜技術來對他們的尿液樣本
進行三甲胺與iY-氧化三甲胺含量檢定，數據以尿液內7^-氧化三甲胺含量(計算
時以三曱胺表示)除以總三甲胺含量(三甲胺+^^-氧化三甲胺(計算時以三甲胺表
示))的百分比來表示(三甲胺A^-氧化百分比）。結果顯示三甲胺A^-氧化百分比的
頻數分配與英國白人的頻數分配大致相同。大部份男性排泄出的三曱胺斤-氧化
百分比均在百分之九十或以上，並產生偏態分佈°總括而言’在廿四小時及單
獨尿液樣本測試中發現七人(佔總人數百分之三)的三甲胺A^-氧化百分比只在百
分之八十至九十之間’另外二人(佔總人數百分之一)分別只錄得百分之七十四及
七十五的三甲胺7^-氧化百分比。
上述數據亦用作分析年齡與飲食習慣對三甲胺A^-氧化作用之影響o結果顯示年
齡與三甲胺7V-氧化百分比爲一 U形曲線關係，但卻找不出任何明顯的關係(關
聯）°參加者在尿液收集期間之三甲胺及7^-氧化三甲胺排泄量的變化極可能基於
其他的因素包括食物改變而引起。在飮食習慣分析方面發現海產類食物是三甲
胺及A^-氧化三甲胺的主要來源，進食海產類食物(亦即攝取大量三甲胺及N-氧
化三甲胺)激化初期性含黃素單氧化的活動力，並使三甲胺7^-氧化百分比進一
XV
步增高，亦對7^-氧化三甲胺排泄率及三曱胺7^-氧化百分比產生持續性影響o
本文亦同時對疾病及藥物於三甲胺A^-氧化作用之影響進行分析。硏究發現肝臟
倉丨』傷(肝硬化)損害肝功能，並減弱了含黃素單氧化的活動力，三甲胺之N-m
化百分比亦因而降低°因此，處方此類需要含黃素單氧化新陳代謝作用的藥
物於上述病人時必須十分小心°兩種藥物，包括atorvastatin及hydroxycWoroquine
(HCQ)，對於三曱胺7^-氧化作用之影響均進行硏究。硏究發現atorvastatin對三
甲胺7^-氧化百分比並無明顯的負面影響。在HCQ方面，雖然病人受HCQ治療
及在不使用HCQ下的三甲胺^¥氧化百分比平均値並無明顯變化，但是有三位
病人在HCQ治療下降低了三曱胺A^-氧化百分比(78 - 88 %)。因此，其他藥物對
三甲胺A^-氧化作用之影響必需進行系統性的分析°
總括而言，接受硏究的華人內最少有百分之一對三甲胺A^-氧化作用有輕微不足
現象，這現象可能是由遺傳或其他因素所引致。年齢與三甲胺7^-氧化作用則無
任何明顯的關係。另外’進食含有大量三甲胺及7^-氧化三曱胺的食物(例如海產
類食物)能激化含黃素單氧化的活動力，此程可由對三曱胺A^-氧化作用之分
析而得悉°肝硬化妨礙了三甲胺的A^-氧化作用，糖尿病、高血脂及高血壓則對
三甲胺A^-氧化作用無明顯的影響°另外，使用atorvastatin及hydroxycWoroquine
治療對於三甲胺A^-氧化作用亦無明顯的影響。
vi
Contents
Declaration i
Acknowledgements ii
Abstract iii
槪要 V
Contents vii
List ofFigures xi
List ofTables xiv
List of Abbreviations xvi
Chapter 1 Introduction 1
1.1 Overview of pharmacogenetics 2
1.2 Clinical importance of nitrogen oxidations 8
1.3 Characteristics of trimethylamine ( T M A ) 9
1.4 Genetic polymorphism of trimethylamine 7V-oxidation 12
1.5 Enzyme systems mediating trimethylamine A^-oxidation 15
1.6 Aims and objectives 19
Chapter 2 Development of an Analytical Method for Trimethylamine

& Trimethylamine AA-Oxide 20
2.1 Introduction 21
2.2 Materials 26
2.3 Methods 27
2.3.1 Preparation of stock solutions 27
2.3.2 Preparation of G L C glass column 28
vii
2.3.3 Optimizing G L C conditions for the separation of
trimethylamine 29
2.3.4 Sample pretreatment for G L C analysis 29
2.3.5 Construction of calibration curve 31

2.3.6 Determination of the required amounts of titanium (III)
chloride for trimethylamine A^-oxide reduction 31
2.3.7 Intra- and inter-assay variations 31

2.3.8 Equations used in the determinations of free T M A ,
total T M A , and percentage T M A excreted as T M A O 32
2.3 Results 33
2.3.1 G L C chromatogram 33

2.3.3 Determination of the required amounts of titanium (III)
chloride for trimethylamine iV-oxide reduction 33
2.4 Discussion 37
Chapter 3 Trimethylamine A^-Oxidation in Chinese 41
3.1 Introduction 42
3.2 Experimental protocols 43
3.2.1 Whole day urine collections 43
3.2.2 Spot urine collections 44
3.3 Results 44

3.3.3 Comparison between whole day urine collections and
spot urine collections protocols 50
3.3.4 Comparison between smokers and non-smokers 52
viii
3.3.5 Comparison the results obtained in between four
individual periods of whole day urine collections 52
3.4 Discussion 56
Chapter 4 Effect ofAge and Diet on Trimethylamine A^-oxidation 64
4.1 Introduction 65
4.2.1 Effect of age on T M A A^-oxidation 73
4.2.2 Effect of diet on T M A 7V-oxidation 73
4.2.3 Effects of control diet on T M A 7V-oxidation 75
4.3 Results 76
4.3.1 Effect of age on T M A A^-oxidation 76
4.3.2 Effect of diet on T M A 7V-oxidation 80
4.3.3 Effects of control diet on T M A A^-oxidation 83
4.4 Discussion 89
Chapter 5 Effect ofDisease on Trimethylamine TV-Oxidation 101
5.1 Introduction 102
5.3 Results 109
5.4 Discussion 116
Chapter 6 General Discussion 119
References 124
ix
Appendices
Appendix A
A. 1 Sample record sheet for food intake and activity
(English) A-1
A.2 Sample record sheet for food intake and activity

(Chinese) A-3
Appendix B
Food intake and activity record for volunteer A in

controlled diet experiment B-1
XV
List of Figures
Figure 1.1 The conversion of trimethylamine (TMA) to 9

trimethylamine A^-oxide (TMAO).
Figure 1.2 Some drugs which undergo nitrogen oxidations by F M O . 10
Figure 1.3 Precursors of trimethylamine in foods. 11
Figure 2.1 Block diagram of a gas chromatograph. 25
Figure 2.2 Typical headspace G L C chromatogram of T M A 34

analysis.
Figure 2.3 Calibration curve for T M A determination. 35
Figure 2.4 Determination of the amount of titanium (III) chloride 36

for the reduction ofurine T M A O to T M A .
Figure 2.5 Intra-assay variation for T M A and total T M A + T M A O 38

(as T M A ) .
Figure 2.6 Inter-assay variation for T M A and total T M A + T M A O 39

(as T M A ) over three months.
Figure 3.1 Frequency distribution histogram for the percentage total 47

trimethylamine ( T M A ) excreted as trimethylamine N-
oxide ( T M A O ) in urine during 24 h period in a Chinese
population (n = 159).

trimethylamine ( T M A ) excreted as trimethylamine N-
oxide ( T M A O ) in urine by spot urine test in a Chinese
population (n = 111).

trimethylamine ( T M A ) excreted as trimethylamine TV-
oxide ( T M A O ) in urine in a Chinese population (n =
270). Data combined from whole day urine collection
and spot urine test.
Figure 3.4 Amounts of T M A and T M A O (as T M A ) excreted in 54

urine at the four collection periods (n = 159).
Figure 3.5 Percentage T M A excreted as T M A O (calculated as 55

T M A ) in urine at the four collection periods (n = 159).
xi
Figure 3.6 Frequency distribution histogram for the logarithmic 60
ratio of total trimethylamine ( T M A ) excreted as
trimethylamine A^-oxide ( T M A O ) in urine in a Chinese
population (n = 270). Data combined from whole day
urine collection and spot urine test.
Figure 4.1 Inhibitors of the flavin-containing monooxygenase. 12
Figure 4.2 Excreted T M A amounts in different periods and in 11

different age groups.
Figure 4.3 Excreted T M A O amounts (as T M A ) in different periods 78

and in different age groups.
Figure 4.4 Percentage T M A excreted as T M A O in urine (combined 79

data for spot urine test and in whole day urine collection
test).
7
Figure 4.5 Percentage T M A excreted as T M A O at Period 1 (after 81

1st dinner).
Figure 4.6 Percentage T M A excreted as T M A O at Period 4 (after 82

2nd dinner).
Figure 4.7 Excreted amounts of T M A and T M A O for volunteer A 87

in the three-day control experiment.
Figure 4.8 Excretion rates of T M A and T M A O for volunteer A in 88

the three-day control experiment.
Figure 4.9 Percentage T M A TV-oxidation for volunteer A in the 90

three-day control experiment.
Figure4.10 Average excretion rates of T M A and T M A O for five 91

volunteers in three-day control experiment.
Figure4.il Average percentage T M A excreted as T M A O for five 92

volunteers in three-day control experiment.
Figure 5.1 The free exchange of fluid and substrate(s) between the 103
sinusoidal lumen and the space of Disse in a healthy
liver.
Figure 5.2 Chemical structures of a) atorvastatin calcium, and b) 107

hydroxychloroquine sulphate.
Figure 5.3 Percentage TMAA/-oxidation in four patient groups. 110
Figure 5.4 Scattered plots for the percentage of T M A excreted as 112

T M A O in urine in the four patient groups.
xii
Figure 5.5 Percentage T M A A^-oxidation for patients treated with 114
atorvastatin and at off drug.
Figure 5.6 Effect of hydroxychloroquine on the percentage T M A 115

A^-oxidation.
Xlll
List of Tables
Table 1.1 Levels of T M A precursors in some major foodstuffs. 13
Table 1.2 Factors affecting Tish-odour syndrome，. 16
Table 1.3 Relative abundance of different isoforms of F M O in the 18

organs of the human body.
Table 2.1 Chromatographic methods. 23
Table 3.1 Descriptive statistics of urinary trimethylamine (TMA), 45

trimethylamine A^-oxide ( T M A O ) (calculated as T M A ) ,
and percentage T M A excreted as T M A O during a 24-hr
period in a male Chinese population (n = 159) and in a
male British Caucasian population (n = 221).
Table 3.2 Descriptive statistics of the percentage T M A excreted as 48

T M A O by whole day urine collections, spot urine tests
and the combined data (spot urine tests+whole day urine
collections) in a male Chinese population (n = 270).
Table 3.3 Amounts of urinary T M A and T M A O (calculated as 53

T M A ) excreted and the percentage urinary T M A
excreted as T M A O in four continuous urine collection
periods in Chinese population (n = 159).
Table 3.4 Comparative incidence of dysfunctional TV-oxidation of 62

trimethylamine in various populations.
Table 4.1 Previous studies of the relation of age to the clearance of 66

drugs cleared by hepatic biotransformation.
Table 4.2 Previous studies on the effect of age on liver drug 67

metabolizing enzymes in animals.
Table 4.3 Schedules for urine collection in Periods 1 and 4 (24-hr 74

urine collection protocols).
Table 4.4 Reasons that the data should be omitted in the statistical 74
analysis for dietary effect.
Table 4.5 Major types of food intakes at dinners during the three- 75
day control experiment.
Table 4.6 Spearman rank order correlation analysis for the factors 84
to the percentage T M A A^-oxidation.
xiv
Table 4.7 Multiple regression analysis for factors on % T M A N- 84
oxidation (% T M A excreted as T M A O ) at Periods 1 & 4.
Table 4.8 Multiple linear regression for the effects of foods on the 85
five factors: 1) % T M A excreted as T M A O ; 2) Amount
of T M A excreted (TMA); 3) Amount of T M A O
excreted (TMAO); 4) Rate o f T M A excreted (TMA/hr);
5) Rate o f T M A O excreted (TMAO/hr), at Period 1, n =
98.
Table 4.9 Multiple linear regression for the effects of foods on the 86
five factors: 1) % T M A excreted as T M A O ; 2) Amount
of T M A excreted (TMA); 3) Amount of T M A O
excreted (TMAO); 4) Rate o f T M A excreted (TMA/hr);
5) Rate o f T M A O excreted (TMAO/hr), at Period 4, n =
88.
Table 4.10 Spearman rank order correlation analysis for the factors 93
to the percentage T M A 7V-oxidation in three-day control
experiment.
Table 4.11 Factors affecting drug disposition in elderly patients. 96
Table 5.1 Details for 20 patients with chronic liver disease. 113
XV
List of Abbreviations
% percentage
< smaller than or equal to
> greater than or equal to
< smaller than
= equal to
> greater than
°C degree Celsius
|ig microgram
1° primary
V' first
2° secondary
2"d second
a.m., A M before noon
ADH Alcohol dehydrogenase
AE Aldrin epoxidase
ALDH Aldehyde dehydrogenase
ALT Alanine transaminase
ANVOA Analysis of variance
b.p. boiling point
BPC Bounded-phase chromatography
CoV coefficient of variation
CYP Cytochrome P-450
e.g. for example
EC Exclusion chromatography
ECD Electron-capture detector
EOR Ethoxyresofufin-0-deethylase
et al. and other people
etc. etcetera
FK) Flame-ionization detector
Fig. Figure
FMO Flavin-containing monooxygenase
xvi
g gram
G6PD Glucose-6-phosphate dehydrogenase
GC Gas chromatography
GLC Gas-liquid chromatography
GSC Gas-solid chromatography
GSH Glutathione
GST Glutathione *S-transferase
HCQ Hydroxychloroquine
HPLC High-performance liquid chromatography
hr hour
hrs hours
i.e. in other words
HDL Intermediate density lipoprotein
JEC Ion-exchange chromatography
JK Infra-red
Km enzyme affinity
L Litre
1-1 per litre
LDL L o w density lipoprotein
LLC Liquid-liquid chromatography
LSC Liquid-solid chromatography
M Molar
M Metre
mg milligram
ml millilitre
ml] per millilitre
mm millimetre
mol mole
mRNAs messenger ribonucleic acids
MS Mass spectroscopy
MSD Mass spectroscopic detector
n frequency
NADPH Nicotinamide adenine dinucleotide
phosphate, reduced form
NDMAd 7V-Nitrosodimethylamine demethylase
xvii
NMR Nuclear magnetic resonance
NPD Nitrogen-pho sphorus detector
p.m., P M aftemoon
PAH Polynuclear aromatic hydrocarbons
ppm part per million
PROD Pentoxyresorudin 0-dealkylase
r regression coefficient
RT retention time
S.D., s.d. standard deviation
SEM standard error of mean
SGPT Serum glutamic pyruvic transaminase
SLE Systemic lupus erthematosus
tV2 half-life
TCD Thermal-conductivity detector
TMA trimethylamine
TMAO trimethylamine A^-oxide
TSD Thermionic specific detector
UDP- Uridine diphospho-
UV/VIS Ultra-violet/visible
VLDL Very low density lipoprotein
v/v volume by volume
w/w weight by weight
a alpha
j3 beta
士 plus/minus
xviii
Chapter 1
Introduction
1.1 Overview of pharmacogenetics 2

1.2 Clinical importance of nitrogen oxidations 8
1.3 Characteristics of trimethylamine ( T M A ) 9
1.4 Genetic polymorphism of trimethylamine iV-oxidation 12
1.5 Enzyme systems mediating trimethylamine A^-oxidation 15
1.6 Aims and objectives 19
1
1.1 Overview of pharmacogenetics
Pharmacogenetics is the study of individual variations in both the rate and the
pattern of drug metabolism in human. It involves the investigation of a large
number of rare genetic traits that show unusual drug reactions because of the
effects of impaired enzymatic activity. The concept of pharmacogenetics was
first described by Sir Archibald Garrod who recognized the causal
relationships between the inheritance of defective genes and deficiencies of
specific enzymes that produced several “inborn errors of metabolism"
including albinism (Garrod, 1909). According to the marked differences of
genetic constitution in each human body, it was suggested that individual
differences in drug and chemical reactions could be predicted (Garrod, 1931).
The concepts were further developed and recognized as a special field of study
called pharmacogenetics by Motulsky (1957) and Vogel (1959). Therefore,
studies in pharmacogenetics nowadays include a combination of
pharmacology, pharmacy, biochemistry, genetics, and molecular biology in
order to identify and study unusual responses with drugs or xenobiotics due to
genetic variations.
A number of enzymes and/or their enzyme-related systems including

esterases, transferases, dehydrogenases, cytochrome P-450s, and other
oxidoreductases, showed genetic variations in their response to drugs and
xenobiotics Enzyme variants showed abnormal activities when compared to
the normal enzymes There are two examples of esterases exhibiting genetic
variations The first example is butyrylcholinesterase (cholinesterase), which
hydrolyzes succinylcholine to form succinylmonocholine and choline
Succinylcholine was commonly used in the 1950's as a muscle relaxant to
produce a prompt and complete paralysis, followed by rapid recovery, with
little toxic side efTects However, the muscle paralysis for people with
atypical butyrylcholinesterase could last for hours rather than the usual 2 lo 6
minutes. These patients were maintained on mechanical ventilators because
they were unable to breathe This atypical butyrylcholinesterase variant which
failed to hydrolyze succinylcholine was observed by Kalow (1956), and
Kalow and Staron (1957) also showed that atypical butyrylcholinesterase was
9
^
a heritable trait. Population studies showed that the frequency of the atypical
allele of butyrylcholinesterase varied from 0.017 in European populations
(0.017 is equal to 1 out of3500 people with the homozygous atypical allele) to
0.076 for Jews in Iran (Whittaker, 1986). The second example is human
serum paraoxonase, which catalyzes the hydrolysis of paraoxon and a number
of other organophosphates, aromatic carboxylic acid esters and carbamates.
Paraoxon is the metabolite of parathion (an insecticide) by cytochrome P450,
in which the sulphur atom in parathion is replaced by oxygen atom. It has two
genetically determined polymorphic forms. Type A has lower paraoxonase
activity, while type B has much higher paraoxonase activity. These two types
of isozyme showed a bimodal distribution in the first Caucasian
pharmacogenetic study (Krisch, 1968). However, unimodal distributions of
serum paraoxonase levels were found in some African and Oriental
populations including Nigerians and Chinese (Playfer et al., 1976). These
results also showed that the paraoxonase activity was different in various
populations, resulting from genetic polymorphism. For transferases, a
polymorphism in A^-acetyltransferase in man was discovered when isoniazid
was used in tuberculous patients, in w h o m 68 % of patients excreted the
administered dose in urine in a conjugated microbiologically inactive form,
and the remaining patients excreted it in an un-conjugated form (Bonicke and
Reif, 1953). This genetic variation was confirmed as a bimodal distribution
(slow metabolizers and rapid metabolizers) by Biehl in 1957. The slow
acetylators of certain therapeutic drugs (isoniazid, sulphamethazine, etc.) were
more susceptible to adverse effects than the rapid acetylators because less
drugs were cleared from the body and thus remained in the body at higher
concentrations. Although the distribution of A^-acetyltransferase activity was
bimodal in the previously described study, Sunahara et al. (1963) showed a
trimodal distribution of A^-acetyltransferase in a Japanese population, in w h o m
three genotypes (slow homozygotes, intermediate heterozygotes, and rapid
homozygotes) were recognized by microbiological assay. Moreover, the
'rapid' acetylators represented over 90 % in Eskimo and Japanese populations,
but some Arab, Indian, and European populations showed 70 % of slow
acetylators. Those results showed that genetic variations occurred with
different frequencies in various populations (Price Evans, 1992). Another
3
typical transferase is human glutathione ^-transferase (GST). It contains six
different isozyme groups named G S T 1 , GST2, G S T 3 , G S T 4 , G S T 5 , and
GST6. G S T s catalyze the conjugation reaction between reduced glutathione
(GSH) and electrophiles, including some mutagens and carcinogens such as
ethylene oxide, styrene oxide and benzo[a]pyrene, etc. (Seidegard et al., 1985;
Warholm et al., 1981) All the isozymes are the products from at least three
genetic loci: GST1, GST2, and GST3 (Board, 1981; Laisney et al, 1984;
Strange et al., 1984). The products from the GST1 loci show genetic
polymorphism and four phenotypes: G S T 1 1, G S T 1 2-1, G S T 1 2, and G S T 1
0; are exhibited. Moreover, the phenotype G S T 1 0 indicates a deficiency
allele at the GST1 locus and it had been found to be the most c o m m o n type in
various populations, while the other three phenotypes occur in different
frequencies in different ethnic groups. This indicated that individuals w h o are
at high risk of cancer are presented in various ethnic groups but in different
proportions. Alcohol dehydrogenase and aldehyde dehydrogenase are two
well-known dehydrogenases. Alcohol dehydrogenase ( A D H ) participates as
the principal oxidative pathway for ethanol metabolism in mammals. It
catalyzes the oxidation of primary and secondary alcohols to aldehydes and
ketones. A D H contains at least five different gene loci (ADHi to ADH5)
which code for human A D H which is formed arising from a random dimeric
association from eight different type of subunits. A D H is also divided into
three classes (I, n, and III), while atypical A D H arises from the ADH2 locus
(Von Wartburg and Schurch, 1968) and contains a variant B2 subunit instead
ofthe usual Bi subunit. Moreover, this atypical enzyme has a higher catalytic
activity than the normal enzyme at p H 8.8. The occurrence of the atypical
A D H enzyme was around 5-10 % in British Caucasian populations (Edwards
and Price Evans, 1967; Smith et al., 1971 & 1972), but it occurred in up to 85
% in Asian populations (Borson and Li, 1986; Fong et al., 1989; Fukui and
Wakasugi, 1972; Harada et a!., 1980; Ikuta et al., 1988; Ogata and Mizohata,
1973; Stamatoyannopoulos et al, 1975; Teng et al., 1979). Aldehyde
dehydrogenase ( A L D H ) acts to detoxify acetaldehyde and other aldehydes in
human organs and tissues. T w o major isozymes, A L D H 1 and A L D H 2 occur
in humans and A L D H 2 shows genetic polymorphism. Pervious studies
described that A L D H 2 showed a lack of activity in Asian populations
4
(Agarwal et al., 1981; Goedde et al., 1979; Fong et al., 1989; Goedde and
Agarwal, 1981; Harada et al., 1980), instead of Caucasian populations
(Goedde and Agarwal, 1981; Harada, et al., 1978). The reason was a point
mutation that occurred at the 14^ position from the C-terminus, where the
glutamic acid was substituted in the deficient isozyme by lysine (Hempel et
al., 1984 & 1985). The cytochrome P-450s (CYPs) are a family of enzymes
that can be found throughout the animal and plant kingdoms. It contains many
gene families: I，II, III, IV, VI, VII, XI, XVII, XIX, XXI, X X V I , LI, L E ,
L X X I , XI, XII, adrenodoxin，and adrenodoxin reductase. The P-450s within
the P-450I, P-450n, and P-450in gene families metabolize drugs and
carcinogens. Some of the C Y P s exhibit genetic polymorphism resulting in
altered drug metabolism. One of the best examples is the hydroxylation by
C Y P 2 D 6 . This is a major metabolic pathway for debrisoquine and some other
cardiovascular agents, including the beta-blockers propranolol and metoprolol
and antiarrhythmics such as flecainide. In Caucasian populations about 5-10%
of subjects were found to be poor metabolizers through this isozyme, whereas
Oriental populations had almost no poor metabolizers (Lou, 1990). However,
in Chinese the extensive metabolizers appeared to have a lower average
metabolic capacity for debrisoquine hydroxylation than Caucasians. Poor
metabolizers are homozygous for an autosomal recessive gene but the reduced
metabolism in extensive metabolizers appears to be due to two other
polymorphisms in C Y P 2 D 6 , C/Tigg and G/C4268, which are highly associated
and the Tigg and C4268 alleles are more common in Chinese. Moreover, a
major metabolic pathway for propranolol metabolism was 4-hydroxylation by
C Y P 2 D 6 and the Tigg allele was associated with reduced metabolism of
propranolol, particularly S-propranolol (Lai et al., 1995). However, the area
under the curves were lower for propranolol and higher for 4-
hydroxypropranolol in Chinese compared to white American subjects in an
earlier study (Zhou et al., 1989), although the beta-blocking and hypotensive
effects of propranolol were greater in the Chinese subjects. It appeared that
there was a larger interindividual variation in propranolol disposition in
Chinese than in white subjects so it should be used more judiciously in
Chinese patients (Lai et al., 1995) which is usual in clinical practice. Other
cytochrome P450 enzymes showing genetic polymorphisms with different
5
frequencies in Asians and Caucasians include impaired S-mephenytoin
hydroxylation (CYP2C19), which was more common in Chinese and Japanese
than Caucasians. Nifedipine is also subject to extensive first-pass oxidative
metabolism through CYP3A4. Plasma levels of nifedipine were found to be
higher in Asians than in Caucasians (Ahsan et al., 1991) and this may result in
a greater hypotensive effect. C Y P 3 A 4 is an important metabolic pathway for
other calcium channel blockers and many other drugs, so ethnic differences in
these might be expected although no polymorphism in the gene for C Y P 3 A 4
has been discovered yet to account for these differences. Other important
enzymes, for example, flavin-containing monooxygenase also exhibit genetic
polymorphism in several populations as demonstrated by the measurement of
the extent of TV-oxidation of trimethylamine ( T M A ) to trimethylamine 7V-oxide
( T M A O ) in man (Al-Waiz et al., 1987b; Hadidi et al, 1995; Mitchell et al.,
1997; Thithapandha, 1997; Zhang et aL, 1996a). F M O includes five isoforms:
F M 0 1 , F M 0 2 , F M 0 3 , F M 0 4 , and F M 0 5 . Deficiency o f F M O activity leads
to impaired 7V>oxidation of some drugs, e.g. nicotine, chlorpromazine, etc.
(Ayesh and Smith, 1992). The sufferer of homozygous deficiency o f F M 0 3
has a fishy-body smell, due to the severe impairment of A^-oxidation of T M A
to T M A O . This genetic disease is called 'fish-odour syndrome'.
Most pharmacogenetic conditions have been seen from the responses to drugs
in a large number of patients under standard dosages. Some patients appeared
to be undertreated, while others required smaller doses to obtaining the normal
pharmacological responses. Therefore, this wide range of individual
responses to a standard dose was reproducible for each individual, and would
be expected for an inherited trait (La Du, 1992). Moreover, many
pharmacogenetic studies related to drug responses are carried out now because
unusual drug responses may be harmful to patients. O n the other hand, certain
compounds or their metabolites in foods, e.g. T M A , may also be affected by
or interact with the metabolic polymorphism, due to the structural similarities,
inhibitory effects, etc. Carsin et al. (1956) found that treatment with drugs,
such as primaquine, caused haemolysis in primaquine sensitive red blood
cells, due the deficiency of glucose-6-phosphate dehydrogenase (G6PD).
Moreover, a similar G 6 P D defect was suspected to be the cause of erythrocyte
6
haemolysis in some people in the Mediterranean area after the ingestion of
Fava beans (Favism), because an unknown compound or its metabolites from
the beans acted like primaquine in these susceptible individuals.
The wide variation of genetically controlled individual also has a profound

effect on pharmacokinetics and the rates of drug removal from the body.
Therefore, the primary objective of pharmacogenetic study has been to learn
the genetic basis of adverse drug reactions which arise as the consequence of
abnormal or defective proteins, for example enzymes, binding proteins
(Alexanderson and Borga, 1972), and receptors. It is particularly related to
genetic mutations of defective or isozymic forms of drug metabolism enzymes
produced by allelic genes, which may be dominant or recessive.
Both animal models and human may be utilized in pharmacogenetic study.

Animal models may be useful to elucidate the molecular basis and
physiological mechanisms of human disease. Also, the pharmacological and
toxicological consequences of the trait can be determined by animal studies in
responders of differing, known genotypes and phenotypes under carefully
controlled conditions that are unattainable in human studies. There are many
animal models used in the study of genetic polymorphism, which include
genetic polymorphisms of the A h receptor, of P450 variants of C Y P 2 D 6
(debrisoquine hydroxylation polymorphism) and of C Y P 2 B 1 in rats; genetic
polymorphism of A^-acetyltransferase in rats, mice, hamsters and rabbits;
thiopurine methyltransferase polymorphism in mice; and the occurrence of
sensitivity and resistance to bleomycin in mice (Weber, 1997). In human
studies, general population surveys by phenotyping have often been utilized.
Affected individuals are first identified in the population study, their genetic
traits can then be detected in the family members and classified as
homozygous (affected or unaffected) or heterozygous phenotypes. The
phenotyping would often make use of the ratio of drug/xenobiotic to
dmg/xenobiotic metabolites in body fluids. One good example is the current
phenotyping study of flavin-containing monooxygenase ( F M O ) activity using
trimethylamine ( T M A ) TV-oxidation. Other methods such as autopsy liver
samples were used for the glutathione S-transferase polymorphism (Harada et
7
al., 1987) and in atypical alcohol dehydrogenase investigation (Von Wartburg
and Schurch, 1968). Percentage of inhibition of dibucaine was also used for
phenotyping of typical and atypical butyrylcholinesterases in human, this
percentage was called dibucaine number (Kalow and Genest, 1957).
Differences in response to salt were also utilized in the examination of

paraoxonase activity (Smolen et al., 1982). Molecular biology techniques
have also been used increasingly to identify the variants of enzymes, which
participate in the metabolic polymorphisms (Skoda et al., 1988; McGurie et
al., 1989).
1.2 Clinical importance of nitrogen oxidations
Many common drugs and xenobiotics, e.g. imipramine, 7V,7V-dimethylaniline,

trimethylamine, tamoxifen, nicotine, pargyline, A^-ethyl-7V-methylaniline,
guanethidine contain nitrogen in their chemical structures. Nitrogen is one of
the three major elements, the other two being carbon and hydrogen, commonly
present in xenobiotics. The most common nitrogen containing groups in drugs
are primary amines, secondary amines, tertiary alkylamines or arylamines
(Cholerton and Smith, 1991). Flavin-containing monooxygenases,
cytochrome P-450s and several other enzymes are involved in the oxidation
(7V"Oxidation) of the nitrogen containing groups in there drugs converting them
into the corresponding A^-oxides. One good example is the conversion of
trimethylamine ( T M A ) to its A^-oxide, trimethylamine 7V-oxide (TMAO). The
formation of a dative bond where the nitrogen atom donates two electrons to
form a single saturated bond with oxygen atom, and the reduced basicity are
two characteristics for this 7V-oxide (Fig. 1.1). Figure 1.2 also summarizes
some typical examples of drugs and their metabolites, which undergo nitrogen
oxidations by F M O .
Generally speaking, the metabolites of xenobiotics by nitrogen oxidation

should be ‘less toxic’ to the body. Therefore, A^-oxidation may be classified as
a detoxification process in most instances.
8
CH3 CH3
H3C - N : • H3C — N ~ ~ • 〇
CH3 CH3
Fig. 1.1 The conversion of trimethylamine ( T M A ) to trimethylamine TV-

oxide ( T M A O )
1.3 Characteristics of trimethylamine (TMA)
Trimethylamine ( T M A ) is a volatile tertiary amine (boiling point, 3°C), which

has malodorous characteristics (Fig. 1.1). It has a low olfactory threshold (< 1
ppm) and resembles the smell of rotting fish. T M A occurs normally in its
odourless 7V-oxide form, trimethylamine iV-oxide ( T M A O ) (Fig. 1.1), in fish,
especially in marine fish (Dyer, 1962; Shewan, 1951). In mammals, T M A is
additionally derived from the degradation by gut bacteria of common dietary
components, such as choline, carnitine, and lecithin, which are present in
major foodstuffs including eggs, liver, and soybean (Cholerton and Smith,
1991) (Fig. 1.3). Most of the T M A absorbed in man is metabolized by N-
oxidation to T M A O by a flavin-containing monooxygenase ( F M O ) and the
T M A O is excreted in the urine along with a small amount of T M A (Al-Waiz
etal., 1987a).
9
p r ^ f Y ^
、一 k 、一 � , V
Nicotine Nicotine-l'-A^^xide
f^^^Ss/^^Y^^ f/H^s^|/^
^ N ^ N ^ ^ ^ C 1 • ^N^N'^^^C1
CH2CH2CH2N〔CH3 CH2CH2CH2N(CH3
CH3 I CH3
0
Chlorpromazine Chlorpromazine A^-oxide
0
& C \ /CH3 H3C^f/CH3
N N
r^ ^ r^
^^ k^
A^JV-Dimethylaniline A/^,A^-Dimethylaniline
rCn
W ^ N - ^ N ^
I CH
——cCO I
CH2CH2CH2N:^
zCH3
CH2CH2CH2N( 3 ’�
CH3
CH3 0
tnipramine inipramine "-oxide
Fig. 1.2 Some drugs which undergo nitrogen oxidations by F M O
10
(CH3)3N+ (CH3)3N+-CH2
CH2 HO一CH
CH2OH CH2
Choline C00_
Camitine
HOOC N
Y ^ y-sH 聊 +
~"N CH2COO"
(CH3)3)N
Betaine
Ergothioneine
CH2OCOR
CHOCOR
0'
CH2O~~P——OCH2CH2N(C^)3
0
Lecithin
Fig. 1.3 Precursors of trimethvlamine in foods
11
T M A O is formed by the oxidation of T M A together with the formation of a
dative bond in between the nitrogen and oxygen atoms. It is an odourless and
non-volatile compound (melting point, 255-257 °C). With respect to the
absorption and disposition of T M A O , 50 % of a dose of this A^-oxide appeared
to be well absorbed and excreted unchanged when given to normal volunteers.
The remainder was reduced, by gut floral enzymes to trimethylamine, which
was absorbed and subsequently metabolized to the iV-oxide (Al-Waiz et al.,
1987b). This process has been named as "metabolic retroversion".
T M A and/or its precursors are present in many major foodstuffs (Table 1.1).
A major source of T M A is from the marine fishes and the amount of T M A O
can comprise up to 0.5 % of the body weight in marine fish. O n the other
hand, beef and other meats do not contain T M A and its oxide, but with
significant amounts of carnitine and choline in red and white meats
respectively. Carnitine and choline are two major precursors o f T M A .
1.4 Genetic polymorphism of trimethylamine A^-oxidation
Normally, T M A undergoes A^-oxidation to T M A O in the body by flavin-

containing monooxygenase. The odorless and less volatile A^-oxide is excreted
mainly in the urine but also to a slight extent in the sweat and the breath.
Normal individuals excrete around 50 m g of total trimethylamine each day,
and most of the trimethylamine (> 90 % ) is excreted as its TV-oxide. The
extent of A^-oxidation of T M A to T M A O in man had been described in
pervious studies in various ethnic groups and showed a polymorphic
distribution (Al-Waiz et al., 1987b; Hadidi et al, 1995; Mitchell et al., 1997;
Thithapandha, 1997; Zhang et al., 1996a). Although the majority of subjects
excreted more than 9 0 % of the total T M A in the form o f T M A O , the incidence
of T M A A^-oxidation deficiency varied from at least 1 % (British Caucasian) to
probably 11% QS[ew Guinean) within the studied ethnic groups.
Deficiency of T M A 7V-oxidation showed various degrees of severity. The

most severe form is an inherited metabolic disorder. In affected individuals,
most ofthe T M A is excreted through urine, sweat, breath and other secretions
12
Table 1 • 1 Levels o f T M A precursors in some major foodstuffs (mg/100 g)
Food Trimethylamine A^-oxide Choline Lecithin Carnitine
Dog Fish 830 - -- --
Skate 1103 - -- -
Salmon 83 — -- —
“ Trout 66 - 580 --
Ox muscle -- -- -- 60
Lamb muscle -- - -- 209
Soybeans - 273 1480 --
Cauliflower - 78 2 < 1
A"" ohtainc(/from A vesli and Smith (1992)
in thefreeT M A form rather than in its oxide form. Subjects suffering from
this condition have an odour of rotting fish and this may result in severe
psychosocial problems for the individual. According to these characteristics,
this genetic disorder is called ‘Trimethylaminuria' or 'Fish-odour Syndrome’
(Humbert etal., 1970; Ayesh et al., 1993; Mitchell, 1996). It appears that the
A^-oxidation of T M A is controlled by a single autosomal diallelic locus in
which one allele induces extensive T M A TV-oxidation while the uncommon
variant allele results in impairment of this process. Individuals with the fish-
odour syndrome are homozygous for the allele determining impaired
metabolism. Since this syndrome is primarily a genetic defect, it is called
‘primary trimethylaminuria' as distinguishable from other forms of
trimethylaminuria, caused by external non-genetic factors.
The rotting-fish smell from sufferers of ‘trimethylaminuria‘ was detected at

different ages. Some sufferers were confirmed by their parents to have a fishy
smell just after birth (Lee, et aL, 1976; Pike et al., 1988)，while others were
only recognized at puberty (Danks et al., 1976; Marks et al., 1976). The
condition became more severe during menstruation in female sufferers (Ayesh
et cd., 1993). The intensity of the smell may vary in different subjects, often
enhanced through exercise and emotion when sweating is increased.
Moreover, the serious body smell caused social difficulties at school in the
childhood years, and led to aggressive behavior and educational disadvantages
for the sufferers (Todd, 1979).
Moderate impairment of metabolism occurred when the subject has one

variant allele for T M A A^-oxidation. These heterozygous unaffected subjects
appeared normal under usual dietary conditions, but they developed a fish-
odour under heavy exposure to T M A , including an oral challenge with T M A ,
due to saturation of T M A iV-oxidation ability (Ayesh et al., 1993). The
subjects suffering with this abnormality were 'secondary trimethylaminurics,,
who had this underlying genetic abnormality but required additional
environmental factors to manifest the phenotype. Factors that contribute to
different extents to the impairment of T M A 7V-oxidation in man include
14
precursor overload (e.g. choline), liver and renal disease, hormonal effects,
inhibitors of the T M A oxidizing system, vaginal disease, and drug treatments.
The significance of these factors are discussed further in Chapters 4 and 5.
The factors affecting this syndrome and the severity or their manifestations are
summarized in Table 1.2.
1.5 Enzyme systems mediating trimethylamine 7V-oxidation
Many drugs and xenobiotics are substrates for both the flavin-containing
monooxygenases (FMOs) and the cytochrome P-450-dependent
monooxygenases, F M O s are responsible for the TV-oxidation of trimethylamine
( T M A ) to its 7V"Oxide ( T M A O ) (Ziegler, 1980). The flavin-containing
monooxygenase ( F M O ) was first purified in 1972 from pig liver (Ziegler and
Mitchell, 1972). The enzyme was difficult to sequence because of the blocked
N-terminus and other technical difficulties. With the recent development of
molecular biology techniques, isozymes of F M O were identified and cloned
(Gasser et al., 1990; Lawton et aL, 1990). According to the amino acid
sequences, five distinct F M O isozymes were identified and classified as:
F M 0 1 , F M 0 2 , F M 0 3 , F M 0 4 , and F M 0 5 , all of which have different
functions in the human body (Gasser, 1996).
F M 0 1 is found mainly in human kidney and not the liver. Therefore, it should
not be involved in the genetic disease of trimethylaminuria. O n the other
hand, Northern blot analysis showed F M 0 1 m R N A was abundant in the fetal
stage, indicating that F M 0 1 is required for regulation in body growth. F M 0 2
is abundant in human lung, and also found in liver and in kidney. In rabbit
lung, F M 0 2 is predominantly localized in the nonciliated bronchiolar
epithelial (Clara) cell, in addition to ciliated, endothelial, type I, and type II
cells and in the tracheal lining layer (Overby et al., 1992). It is almost 85 %
similar to the human F M 0 2 . It catalyzes 7V^-oxygenation of alkylamines to
oximes but has no effect on the primary arylamines (Poulsen et al., 1986).
F M 0 3 was first isolated from human liver by Lomri et al. in 1992 and it was
15
Table 1.2 Factors affecting Tish-odour svndrome’
、T ^ ^ , Genetic 。 .
No. Factors Syndrome effect Severity
Homozygous for the

1 allele determining 1° Trimethylaminuria Yes Most severe
impaired metabolism
TT p , Moderate,
Heterozygous for the depends on
2 allele determining 2° Trimethylaminuria Yes other non
impaired metabolism genetic factor
Moderate,
3 Precursor overload 2° Trimethylaminuria No associated with
factor (2)
4 Liver disease 2° Trimethylaminuria No 二 ^ factor^!^)^
5 Renal disease 2° Trimethylaminuria No 二 ^ Ta(^^v^(3)
6 Vaginal disease 2° Trimethylaminuria No Moderate
， T^. ‘ ,o rp . 丄 1 . . 、T Mild, associated

7 Dietary enzyme 2° Trimethylaminuria No • /^ .^.
inhibitors
Enhance the
8 Hormonal effect N/A No syndrome
caused by
factors (1) & (2)
Enhance the
9 Drug effect N/A No syndrome
caused by
factors (1) & (2)
16
found to exhibit distinct stereoselectivity for oxygenation of cimetidine and S-
nicotine (Cashman et al., 1995). Moreover, F M 0 3 is abundant in human liver
and now it is generally believed to be the major isoform for the A^-oxidation of
T M A (Cashman, 1995). Another minor isoform of F M O is F M 0 4 . It is
present in human brain and liver. However, its function and substrate
specificity remain unknown. F M 0 5 also occurs in human liver (Phillips et al.,
1995). It is inactive with methimazole, a general substrate, but highly active
with n-octylamine. The relative abundance of different isoforms of F M O in
the organs of the human body is summarized in Table 1.3.
The enzyme responsible for T M A TV-oxidation was initially thought to be

F M 0 1 (Dolphin et al., 1991), but subsequent shown that only a small amount
o f F M 0 1 was found in human liver (Burnett et al., 1994; Cashman et al.,
1993; Lomri et al., 1992; Sadeque, et al., 1992; Wrighton et al, 1993). Liver
F M 0 3 was generally thought to be the major enzyme for T M A TV-oxidation in
humans (Cashman, 1995; Zhang et al., 1996a). It appears that the isoform
F M 0 3 of the flavin-containing monooxygenases (FMOs) is the major enzyme
involved in the A^-oxidation of trimethylamine into trimethylamine 7V-oxide.
Moreover, its gene is located on human chromosome lq (Shephard et al.,
1993), and contains one non-coding and eight coding exons (Dolphin et al.,
1997a). A C ^ T missense mutation, corresponding to nucleotide 551 in the
c D N A , was identified in exon 4 of both F M 0 3 alleles of the primary
trimethylaminurics (homozygous affected subjects). The mutation changes a
C C C proline triplet at codon 153 to a C T C leucine triplet (Prol53^Leul53).
The parents (heterozygous unaffected subjects) were heterozygous for triplets
at this position, while the non-trimethylaminurics (homozygous unaffected
subjects) were homozygous for a C C C triplet at the same position (Dolphin et
a/.，1997b). It was possible that the proline to leucine substitution in codon
153 caused protein tertiary structural changes that result in abnormal protein
folding and an inactive human F M 0 3 (Treacy et al., 1998). This finding may
be useful to provide a new picture of the genetic polymorphism in T M A N-
oxidation.
17
Table 1.3 Relative abundance of different isoforms of F M O in the organs of
the human body
Isoform Organ Abundance*
FM01 Kidney +++
Liver ++
FM02
Kidney ++
FM03 Liver +++++
Brain +
FM04
Liver ++
FM05 Liver ++++
* + sign represent the relative abundance of the corresponding isozyme in the organ
18
1.6 Aims and objectives
Inter-ethnic differences in drug metabolizing enzymes have been reported in

numerous studies involving Asian and Chinese populations, which constitute
one of the largest ethnic groups in the world (Kiniron et al., 1996; Lee at el,
1988; Lou, 1990; Lou et al., 1987; Sowunmi et al., 1995; W a n g et al., 1997;
Xie et al., 1996; Zhou et al., 1989), and these may be caused by genetic
variations. For example, the metabolism of debrisoquine showed an incidence
of poor metabolizers of about 3-10% in various European populations, but less
than 1 % in Chinese (Lee at el., 1988; Lou, 1990; Lou et al., 1987). However,
the distribution of T M A A^-oxidation had been studied in a number of ethnic
groups (British Caucasian, Ecuadorian, Jordanian, and Thai), but not in the
Chinese population. Therefore, the first objective in this study was to
determine the distribution of this metabolic trait in the Chinese population in
Hong Kong. Moreover, the effect of age on the ability for T M A TV-oxidation
in Chinese male subjects was also examined as the second objective. The
third objective in this study was to determine the effects of diet on total T M A
and T M A O excretion in Chinese male subjects. Since T M A and T M A O , and
their precursors are present in many types of foods, including fish, beef, etc.,
the effects of diet and the dietary habits in southern Chinese was examined to
see if there was any significant difference in T M A and T M A O excretion in
Chinese compared to other ethnic groups. The fourth objective of the study
was to examine if any disease factors affected T M A and T M A O excretion in
certain patient categories including liver disease, renal disease, diabetes, and
patients treated with lipid-lowering drugs.
19
Chapter 2
Development of an Analytical Method for
Trimethylamine & Trimethylamine 7V-Oxide
2.1 Introduction 21
2.2 Materials 26
2.3 Methods 27
2.3.1 Preparation of stock solutions 27
2.3.2 Preparation of G L C glass column 28
2.3.3 Optimizing G L C conditions for the separation of trimethylamine 29
2.3.4 Sample pretreatment for G L C analysis 29
2.3.6 Determination of the required amounts of titanium (III) chloride
for trimethylamine 7V-oxide reduction 31
2.3.8 Equations used in the determinations of free T M A , total T M A ,
and percentage T M A excreted as T M A O 32
2.4 Results 33
2.4.1 G L C chromatogram 33
2.4.3 Determination of the amounts of titanium (III) chloride required
for trimethylamine TV-oxide reduction 33
2.5 Discussion 37
20
2.1 Introduction
Sensitive and specific analytical techniques have been used in many metabolic
studies to characterize drugs and their metabolites. M a n y of these techniques
have been described in detail for drug and metabolite characterizations,
including colorimetry, ultra-violet/visible (UV/VIS) spectrophotometry, infra-
red (IR) spectrophotometry, mass spectrometry (MS), nuclear magnetic
resonance QSfMR) spectroscopy, high-performance liquid chromatography
(HPLC), gas chromatography (GC), etc. These methods are useful in the
determination of different drugs and their metabolites, according to their
physical and chemical properties. Quantitation and characterization of drugs
and their metabolites by colorimetry and UV/VIS spectrophotometry have
been popular in the past, and methods based on formation of coloured or U V
absorbable complexes with chemical reagents are available for many
metabolic reactions, e.g. the determination of the A^-oxides of N’N-
dialkylanilines by Ziegler and Pettit (1964). However, the colorimetric and
UV/VIS spectrophotometric methods are limited in their use because of their
low sensitivity and lack of specificity (Damani, 1985). JR spectrophotometry
is also used for drug characterization, but it has its own limitations for drug
and metabolite analysis including low sensitivity, and the difficulties in
spectrum interpretation. Therefore, JK spectrophotometry is generally used in
qualitative drug identifications. N M R spectroscopy and M S are two modern
techniques in qualitative and quantitative d m g and metabolite analyses. N M R
phenomenon occurs when nuclei, e.g. ^H, ^^C, in the molecule under
investigation aligned with an applied field are induced to absorb energy and
change their spin orientation with respect to the applied field. O n the other
hand, M S is a result of the alternation of path, by the application of magnetic
field, of the pieces of ions in which breakdown from the molecule under
investigation after the electron bombardment or chemical ionization process.
Both techniques can produce specific and sensitive results in d m g and
metabolite analysis including T M A and T M A O determinations (Abeling et al.,
1995; DaCosta et aL, 1990; Hirama et aL, 1993; Tjoa and Fennessey, 1991).
However, these two sophisticated techniques require expensive equipment and
high maintenance cost, and therefore are not common in many laboratories.
21
Chromatography, including G C and H P L C , is a very popular technique in
drug and metabolite determinations in modern laboratories, so it needs to be
described clearly. The primary characteristic of chromatography is that two
mutually immiscible phases are brought into contact; one phase is stationary
and the other mobile. W h e n a sample is introduced into the column, it will be
carried by the mobile phase. The sample components interact with the mobile
phase and stationary phase to different extents. These repeated interactions of
the sample components with the stationary and mobile phases cause
separation. Since the interactions are different for different components, the
least interactingy^retarded component will leave the column first and the most
interacting one will stay in the column for the longer period. The separated
components, in the form ofbands, are directed to the end of the column, where
they can be detected by some type of detector or collected by suitable
apparatus. The most important part in chromatography is the column. It
provides flexibility for different analysis. Different stationary/mobile phases,
length of column, temperature, pressure, additives, and the combinations of
these factors can further extend the range and depth of the component
separations. This versatility has helped chromatography to become the most
favoured separation technique in research and in product development.
Moreover, the use of different types of detectors, e.g. thermal-conductivity
detector (TCD), flame-ionization detector (FDD), electron-capture detector
(ECD), nitrogen-phosphorus detector ( W D ) , mass spectroscopic detector
(MSD), etc; also helps the researcher for specific analysis of the sample
components. The mobile phase can be a gas or a liquid, but the stationary
phase can be only a liquid or a solid. Table 2.1 describes the general
classification of the present chromatographic techniques. Although it can be
subdivided into many variations, the two most common chromatographic
techniques in analysis are gas-liquid chromatography (GLC) and liquid-liquid
chromatography (LLC). Liquid-liquid chromatography performed under high
pressure increased the performance of separation. It is now defined as high-
performance liquid chromatography (HPLC). Before the invention of gas
chromatography, drug analysis and metabolism studies used column, paper or
thin-layer chromatography for the separation and characterization of drugs and
22
Table 2.1 Chromatographic methods
Chromatography
Gas Liquid
- Chromatography Chromatography
Gas-liquid Gas-solid Liquid-liquid Liquid-solid Ion-exchange Exclusion

(GLC) (GSC) (LLC) (LSC) (IEC) (EC)
Bonded-phase Ion-pair
(BPC) Chromatography
their metabolites. One major problem was the detection limits and
sensitivities with the above techniques. Drugs and their metabolites in sample
fluids are mostly found in very low levels, and this led to the development of
more sophisticated techniques for the analysis. In the last twenty years, the
rapid development of gas chromatography (GC) and high-performance liquid
chromatography (HPLC) helped to solve most of the above problems.
H P L C is one of the most common techniques for drug and metabolite analysis,
especially when the molecules are large, non-volatile, or thermally unstable.
Moreover, H P L C can separate macromolecules, ionic species, labile and
polymeric materials, etc, and the components can be quantitative recovered.
M a n y drug characterizations are based on the H P L C techniques rather than the
old colormetric techniques (Szepesi, 1990). Separation and quantitative
determination of many sulphur-containing compounds have also been
performed by H P L C . One good example is the separation and quantitative
determination of sulindac sulphide and its oxygenated metabolites by reversed
phase H P L C (Swanson and Boppana, 1981). Another example is the
characterizations of sulphinpyrazone and its sulphoxide and sulphone
metabolites by M<-Bondapak C18 H P L C column (Renwick et al., 1982).
Although H P L C is effective for drug and metabolite characterizations, gas
chromatography (GC), especially gas-liquid chromatography (GLC), is still
very useful for the separation of thermally stable and volatile organic and
inorganic compounds. It is especially effective because the flow restriction in
a G C column is lower than in the H P L C column. Therefore, the time of
analysis can be greatly reduced when a suitable column is chosen. Other
advantages, including the simple operating principle and the possibility of
coupling it to a mass spectrometer or an infrared spectrophotometer, result in
G L C still being used in many laboratories.
G L C is commonly used in the determination of amines and their derivatives in

metabolic studies. Many methods have been developed throughout the last
decade. The samples can be analyzed after the addition of alkali, either by
direct injection or by headspace technique, or they are extracted into an
24
Injector Detector Recorder
\ 1 ^
-^~~^ n^^^^"
n M In W
' w [_
/
Fig-
Carrier
Supply
I
2.1GasBlock diagram of a Oven
\“
gas chromatographColumn
\
Detector
\SupplyGas
organic solvent before analysis. Marzo et al. (1990) described the direct
injection method for aliphatic amines by G C using a flame-ionization detector
(FK)) and thermionic specific detector (TSD). Lundh and Akesson (1993)
used derivatization with isobutyl chloroformate to determine primary and
secondary low-molecular-mass aliphatic amines in urine. Pfundstein et al.
(1991) also employed gas chromatography and chemiluminescence detection
with a modified thermal energy analyzer for primary and secondary amines.
Some researchers also employed the headspace method for amine analysis. A
recent published article by Kataoka (1996) also reviewed the methods of
amine determination by gas chromatography, and this research group also
utilized derivatization and flame photometric detection for primary and
secondary amine determinations (Kataoka et al., 1992 & 1993). O n the other
hand, Krzymien and Elias (1990) used the headspace G C method to determine
fish freshness by T M A measurement. Zhang et al. (1992) also used a very
sensitive and simple headspace method to determine amines in urine samples.
This sensitive headspace assay with suitable modifications was applied in this
research study. Details are described in the following sections.
In this research study, the headspace G L C method was employed to determine

the amounts of trimethylamine ( T M A ) in urine samples and the reduced
trimethylamine TV-oxide ( T M A O ) as T M A after treatment with titanium (III)
chloride.
2.2 Materials
Gas-liquid chromatography was performed on an H P 5890 Series II

chromatograph together with an H P 3396 Series II Integrator (Hewlett-
Packard Company, Avondale, U.S.A.), equipped with a flame-ionization
detector and a glass made packed column (2.4 m x 4 mm). The stationary
phase, Carbowax 20 M , was purchased from Chrompack International B.V.
(Middelburg, The Netherlands). Graphitized carbon support (60-80 mesh)
Carbopack B and the 5 ml gas-tight G C syringe with sample lock was
purchased from Supelco (Bellefonte, PA, U.S.A.). Potassium hydroxide for
the stationary phase preparation and for sample treatment, hydrochloric acid
26
for the preparation of standard solutions, and chloroform for dissolving G C
stationary phase, were purchased from B D H Chemical Limited (Poole, Dorset,
U.K.). Trimethylamine hydrochloride salt (primary standard), isopropylamine
(99 % ) solution (internal standard), and dimethyl-dichlorosilane solution
(silanizing agent) were obtained from Sigma Chemical Company (St. Louis,
M O , U.S.A.). High-purity nitrogen (>99.99 % ) (carrier gas), compressed air
and hydrogen (for F K ) ignition) were supplied from Chun W a n Industrial Gas
Company (Hong Kong, China). 5 ml glass sample vials, rubber stoppers,
aluminium caps, and the capping/decapping apparatus for urine sample
holding were purchased from Wheaton Company (Millville, NJ, U.S.A.). The
water bath for urine samples heating was bought from Cole-Parmer
International (Vernon Hills, JL, U.S.A.).
2.3 Methods
2.3.1 Preparation of stock solutions
0.01 M hydrochloric acid solution for the preparation of stock

solutions was prepared from the 0.1 mol 1"^ hydrochloric acid ampoule
by suitable dilution with deionized-distilled water. In the preparation
of trimethylamine standard solution, trimethylamine stock solution at 1
m g ml—i was firstly prepared from the trimethylamine hydrochloride
salt in 0.01 M hydrochloric acid solution. For convenience,
trimethylamine intermediate solution at 50 îg ml'^ was also prepared
from the stock solution in 0.01 M hydrochloric acid solution. The
solutions were kept in a refrigerator at 4。C. In the preparation of
isopropylamine internal standard solution, isopropylamine stock
solution at 2 % (v/v) was prepared from the isopropylamine (99 %
w/w) solution in 0.01 M hydrochloric acid solution and also kept at
4。C.
27
2.3.2 Preparation of GLC glass column
A n empty pre-coiled G C glass column (2.4 m x 4 m m ) was first

cleaned with chromic acid followed by rinsing three times with
deionized water and methanol. Then, it was cleaned with silanizing
agent—dimethyldichlorosilane followed by rinses with deionized water
and methanol. It was dried in an oven at 60 °C.
To coat the stationary phase on solid support, a suitable amount of

Carbowax 20 M and potassium hydroxide were weighed in a round
bottom flask. This was then dissolved in chloroform. Solid support
(Carbopack B, 60-80 mesh) was added to the chloroform solution and
the mixture was allowed to stand overnight at room temperature. This
resulting slurry was then evaporated to dryness on a rotary film
evaporator (J. Bibby, Stone, U.K.) at 70 °C. To ensure complete
dryness for the coated support, it was further heated in an oven at 100
°C for 1 hour. The resulting solid support was 4 % Carbowax
20 M/0.8 % potassium hydroxide on Carbopack B (60-80 mesh).
The Carbowax 20 M coated Carbopack B solid support was packed

into the empty glass column with the aid of a vacuum pump (J. Bibby,
Stone, U.K.) with gentle vibration by hand. Silanized glass wool was
plugged into one end of the column where the vacuum was applied.
The other end of the column was plugged with a small funnel for ease
of material transfer. W h e n the column wasfiillypacked with solid
support, the funnel was removed and silanized glass wool was plugged
into this end. The column was then installed into the gas
chromatograph. To remove volatile impurities and to ensure tight
packing, the column was conditioned overnight below the
recommended maximum operating temperature (220 °C) of the
stationary phase with the carrier gas (high purity nitrogen) flowing.
After overnight conditioning, the column was inspected by eye for any
loose packing. It was then detached from the gas chromatograph, the
glass wool plug at the inlet was removed and an extra amount of solid
28
support was then added to ensure tight packing. W h e n the glass wool
was re-plugged into the column, it was then reinstalled into the gas
chromatograph. The above conditioning procedures were repeated until
no loose packing was observed.
2.3.3 Optimizing GLC conditions for the separation of trimethylarnine
The optimum G L C operating conditions for trimethylamine analysis

were:
Injection port temperature: 170 °C

FID detector temperature: 200 °C
Oven temperature: 70 °C
Carrier gas (nitrogen) flow rate: 32 mi min"'
Glass column dimension: 2.4 m x 4 m m
Solid packing: 4 % Carbowax 20 M/
on potassium hydroxide on
Carbopack B (60-80 mesh)
Hydrogen flow rate for FlD 55 ml min"'
Compressed air t1ovv rate for FID 250 ml min''
Auxiliary nitrogen flow rate for FID 60 ml min''
2.3.4 Sample prctreatment for (;L(，ana/ysis
To analyze irimelhylamine ( T M A ) and trimethylamine A,-oxide

( T M A O ) in urine, 0 2 % (v/v) isopropylamine was prepared as the
internal standard by 1 in 10 dilution from 2 % (v7v) isopropylamine
stock solution with 0 01 M hvdrochloric acid The workinu 0 2 %
‘ w
(v/v) isopropylamine internal standard solution was used within one
week
For the determination o f T M A in urine. 2 ml of urine sample was put

into a 5 ml glass vial immersed in an ice baih 80 ul 0 2 % (v/v)
isopropylamine internal standard solution was then added into the vial
29
1.2 g potassium hydroxide powder was added to bring the p H > 10 to
convert the T M A in urine to its free base form. The vial was stoppered
with a rubber stopper immediately and capped with an aluminium cap.
The sample-in-vial was then shaken with a vortex to ensure all K O H
powders were dissolved. It was put back into the ice bath until the
temperature had appropriately dropped. The sample-in-vial was then
put into a water bath at 90 °C for 20 minutes. After 20 minutes, 2 ml
of headspace gas in glass vial was removed by a gas tight syringe. It
was held in the syringe by closing the sample lock in the syringe. The
syringe was then inserted into the G C injection port and the sample
lock was released. The gas was then totally injected into the column
and the integration was started at the same time.
T M A O was not analyzed by G L C directly, it was first reduced to T M A

by titanium (III) chloride. Then, the total amount of T M A (free
T M A + T M A from the reduction o f T M A O ) was detected by G L C , and
the T M A O (as T M A ) amount was determined by subtraction the value
of total amount of T M A (i.e. T M A + T M A O (as TMA)). For the
determination of total T M A , 1 ml of urine sample was put into a plastic
sample vial together with 200 ^1 of 15 % (w/v) titanium (III) chloride
(reducing agent). The vial was capped and left at room temperature for
30 minutes. Then, 4 ml of deionized water was added into the vial and
it was shaken well. After this, 2 ml of this solution was added into the
glass vial in ice bath and the same procedure was followed as in T M A
determination.
The peak area ratio for T M A to isopropylamine and the amount of

T M A in the sample solution were determined and calculated by
suitable mathematical calculations (Section 2.3.8). If the peak area
ratio exceeded the range of the calibration curve, the urine sample was
diluted with deionized water appropriately and it was re-analyzed.
Section 2.3.8 describes the calculations o f T M A (free T M A ) and T M A
+ T M A O (as T M A ) (total T M A ) in urine sample.
30
2.3.5 Construction of calibration curve
To construct the calibration curve, trimethylamine ( T M A ) intermediate

solution (50 îg ml"^) was used to prepare a series of T M A standard
solutions. This T M A intermediate solution was kept for 3 months at
4°C.
T M A standard solutions (0.1 ^g ml"^ to 10 ^g ml"^) were prepared by

dilution with 0.01 M hydrochloric acid. Nine different concentrations
of T M A standard solutions were employed to construct this calibration
curve: 0.1, 0.2, 0.3, 0.5, 0.8, 2.0, 5.0, 8.0 ,10.0 îg m r \ All points
were performed in duplicate and on three consecutive days, and
therefore six sets of data were obtained.
2.3.6 Determination of the amount of titanium (III) chloride required for

trimethylamine N-oxide reduction
The amount of titanium (III) chloride required for trimethylamine TV-

oxide ( T M A O ) reduction to trimethylamine ( T M A ) in the urine
samples was determined. T w o urine samples were chosen for
determination in the same day. One sample contained high amounts of
T M A and T M A O and another sample contained little amounts o f T M A
and T M A O .
2.3.7 Intra- and inter-assay variations
Intra-assay variation was calculated by multiple analysis of the same

urine sample in a single assay, while inter-assay variation was
determined by repeated analysis of aliquots of the same urine samples
over three months.
31
2.3.8 Equations used in the determinations of free TMA, total TMA, and
percentage TMA excreted as TMAO
Equations 2.1 & 2.2 describe the mathematical determinations of free

T M A and total T M A in urine sample.
Amount of T M A in urine sample (free T M A ) (mg)=
Concentration of T M A (free T M A ) in urine sample for G C analysis

([igAnl) X (dilution factor) x Volume of urine excreted (ml) + 1000
Note: (dilution factor) = 1 when sample dilution is not required. For

1 in 10 dilution, (dilution factor) = 10，etc.
(Equation 2.1)
Amount o f T M A + T M A O (as T M A )
in urine sample (total T M A ) (mg)=
Concentration o f T M A + T M A O (as T M A ) (total T M A ) in urine sample

for G C analysis (|igAnl) x (dilution factor) x 5.2 x Volume of urine
excreted (ml) + 1000
Note: (dilution factor) = 1 when sample dilution is not required. For

1 in 10 dilution, (dilution factor) = 10, etc.
(Equation 2.2)
A n equation to determine the percentage of T M A excreted as T M A O

in urine was also developed, in order to classify the volunteers and
patients as to whether they suffered from trimethylaminuria (fish-odour
syndrome) or an intermediate deficiency in T M A A^-oxidation.
32
Percentage T M A excreted as T M A O =
Total T M A - Free T M A
X100%
Total T M A
(Equation 2.3)
2.4 Results
2.4.1 GLC chromatogram
The typical G L C chromatogram is shown in Fig. 2.2. Peaks of

trimethylamine ( T M A ) and the internal standard, isopropylamine, were
well separated and without any interference from unwanted peaks.
The retention times for T M A and isopropylamine were 5.2 mins and
7.9 mins. respectively. The total run time for one injection was less
than 10 minutes. The conditions are listed at Section 2.3.3.
2.4.2 Construction of calibration curve
The calibration curve was constructed combining all these sets of data
as described in Section 2.3.5 (Fig. 2.3).
2.4.3 Determination of the amount of titanium (III) chloride required for

trimethylamine N-oxide reduction
One sample contained high amounts of T M A and T M A O and the final

amount o f T M A + T M A from reduced T M A O was located near the high
end of the calibration curve. O n the contrary, another sample
contained little amounts of T M A and T M A O and the final amount of
T M A + T M A from reduced T M A O was located at the low end of the
calibration curve (Fig. 2.4).
33
Trimethylamine
< (RT = 5.19min)
Isopropylamine
^ (RT = 7.88 min)
ULIL
Time
Fig. 2.2 Typical headspace G L C chromatogram o f T M A analysis
34
0.6 ^
0^ 0.5 - ^ ^
I j /
I 0.4 - ^ ^
I 。 3 - ^ ^
� H ^ ^
^ o Z
1 02 - ^ ^
cd Z
CD Z
\。1 - ^ ^
<D Z
� _>*^
0.0 - 广
I 1 1 1 H
0 2 4 6 8 10
Trimethylamine (^g/ml)
Fig 2.3 Calibration curve for T M A determination (each point represents the mean and standard deviation ofthree
separate davs of determinations^
60 " I
50 -
I- h^\~^
麵 30- /
, 2 � - /
10 - /
/ ,,靈 Hfc 彖- 4t X
' T 1 1 1 1——
0 100 200 300 400
Amount of 15 % w/v TiCl^ Used (îl)
Fig. 2.4 Determination the amount of titanium fIII) chloride for the
reduction of urine T M A O to T M A ： ---•— low T M A + T M A O
sample. ~ • 一 high T M A + T M A O sample (each points represents
the mean and standard deviation of three separate determinations)
36
2.4.4 Intra- and inter-assay variations
Intra-assay variation determination produced a coefficient of variation

ofl.3 % for T M A (0.93 土 0.01 ^g/ml, mean 土 s.d.，n = 5) and 3.7 %
for T M A + T M A O (as T M A ) (98.19 士 3.64, mean 土 s.d., n = 5) (Fig.
2.5). Inter-assay variation resulted in a coefficient of variation of 3.9
% for T M A (0.43 土 0.02 pgAnl, mean 土 s.d., n = 6) and 2.4 % for
T M A + T M A O (as T M A ) (7.83 士 0.19 ^g/ml, mean 士 s.d., n = 6) (Fig.
2.6).
2.5 Discussion
In this study, headspace sampling was employed rather than the traditional
liquid injection technique. The reason was the low boiling points for T M A
(b.p. 3 °C) and the internal standard-isopropylamine (b.p. 32 °C), and that the
partition coefficient allowed sufficient amounts of them into the gaseous
phase. Therefore, time-consuming extraction of T M A from the urine sample
could be avoided Most of the impurities with high-boiling points in urine
were not vaporized and the headspace gas in the glass vial mainly consisted of
vaporized T M A and isopropylamine.
The calibration curve ranging from 0.1-10 ^g ml'^ was established with good
linearity. It was non-linear beyond 10 ^g ml'V Therefore, only the linear
range from 0.1 \ig/m\ to 10 |ig/ml of T M A was utilized to create one
calibration curve because samples with large amounts of T M A and T M A O
would be subjected to suitable dilution. Moreover, a flame-ionization detector
was used in G L C rather than the other specific detectors, e.g. nitrogen-
phosphorus detector, because it was well suited for the range of T M A
detection.
37
1.2. p 110
98.19±3.64，CoV = 3.7o/o
1.15- ^ " \ ^ ^ ^ ^ ^ ^ - 100
1.1 - -90
/ ^ ^
S
1.05- — 80 g
/^ w
a §
> §
3 1- - 70 ^
i I
u 0.93±0.01，CoV=1.3o/o P
^0.95— . •、默 - 6 0 J
H ��Z � s ^ 0
� � , z 1
0.9 - - 50 p
+
^
g
0.85- __40 —
O.s] 1 1 1 1 [- 30
1 2 3 4 5
Number of Attempts
Fig. 2.5 Intra-assav variation for T M A (-m-) and total T M A + T M A O (as

T M A ) ( ~ • ~ ~ ) (legend represents mean 士 s.d., and coefficient of
variation (CoY))
38
1 10
G9 7.83±0.19,CoV = 2.4o/o ^
0.8 I f •• I 8
-"^^..,r,.^-*"----ii----ii……II -
0.7 1 丄丄丄 7
0.6 6
/^
s^ ^
^ 0.5 5 3
^ —- -— 妄
i 0.4 _— $^""""^"^^^"^^^^^"i^^_^j_^^^î——• __ 4 I
0 忽
1 0.3 __ 0.43i0.0ACoV = 3.9% —_ 3 |
H
0.2 — __ 2 I
0.1 1
M ^ “ I ^ I ^ ~ I ^ I ~ ~ I ^ ^ [ o
DayO Day 1 Day7 Wk2 lMth 3Mth
Time
Fig. 2.6 Inter-assay variation for T M A (~~•~~) and total T M A + T M A O fas

T M A ) ( — • — ) over three months (legend represents mean 士 s.d.,
and coefficient of variation rCoV) over three months: each point
represents the mean and s.d. of three separate determinations)
39
The optimal amount of titanium (III) chloride for T M A O reduction was
determined. Fig. 2.4 described the effect of different amounts of titanium (III)
chloride on the results of T M A + T M A O in the samples. T w o samples were
employed. For the low T A M + T M A O sample, it showed no significant
difference of T M A + T M A O amounts at different concentration of titanium
(III) chloride added (50-400 ^g ml"^). Although the determined
T M A + T M A O amounts varied slightly at the sample with high T M A + T M A O
amounts, it was still reasonable to use 200 )ig ml'^ titanium (III) chloride
because this amount showed an optimal result for T M A + T M A O .
The exact values of intra- and inter-assay variations are described in Figures
2.5 and 2.6. Coefficients of variations were smaller than 4 % . Therefore, it
was possible to analyze urine samples within three months after the collection,
without significant change in T M A and T M A O amounts. Therefore, a rapid
and sensitive analytical method by headspace gas chromatograph was
developed for the analysis of urinary T M A and T M A O . It was used for the
experiments described in the other chapters.
40
Chapter 3
Trimethylamine 7V-Oxidation in Chinese
3.1 Introduction 42
3.3 Results 44
3.3.3 Comparison between whole day urine collections and spot
urine collections protocols 50
3.3.4 Comparison between smokers and non-smokers 52
3.3.5 Comparison of the results obtained in four individual
periods of whole day urine collections 52
3.4 Discussion 56
41
3.1 Introduction
The genetic polymorphism of trimethylamine ( T M A ) #-oxidation, as

described in Chapter 1，can be divided to three categories. The first category
is the normal group, which constitutes the majority of the population. These
people excrete more than 90 % of T M A as T M A O in urine and the flavin-
containing monooxygenase ( F M O ) is fully functional. The second category
includes people w h o excreted 60-90 % of T M A as T M A O in urine. This
condition is known as 'secondary trimethylaminuria', which can result from
genetic or non-genetic factors (Chapter 1). The third category is the
homozygous inherited metabolic disorder, in which most of the T M A is
excreted through urine, sweat, breath and other secretions in the free T M A
form rather than in its oxide form. This genetic disorder is termed 'primary
trimethylaminuria' or 'fish-odour syndrome，(Ayesh et al,, 1993; Humbert et
al., 1970; Mitchell, 1996)，and the percentage of T M A as T M A O in the urine
is less than 60 % .
The distribution of this genetic polymorphism was first investigated in a

British Caucasian population (n = 169) by Al-Waiz et al. (1987). In this
study, no primary trimethylaminuric was found except one subject (1 % ) who
showed moderate impairment in T M A TV-oxidation. The distribution was
skewed with a tail extending towards the lower values in the negative direction
from the median. The polymorphic distribution of T M A 7V-oxidation has been
studied in some other ethnic groups. Zhang et al. (1996a) also extended the
study (n 二 421) in a British Caucasian population and they found a similar
result as in the initial study (1 % ) with impairment in T M A A/-oxidation.
Hadidi et al. (1995) carried out a study in a Jordanian population. They found
that about 8 of 82 subjects (10 % ) had impaired ability in T M A A^-oxidation.
Mitchell et al. (1997) studied the discontinuous A^-oxidation oftrimethylamine
among Jordanian (n =116，1/3 day urine collection), Ecuadorian (n = 80, spot
urine) and N e w Guinean (n = 100，spot urine) populations based on a 1/3 day
urine collection or a spot urine study. It was found that the proportion of
subjects with a relative deficiency iniV-oxidation were 1.7 % in Jordanian, 3.8
% in Ecuadorian, and 11.0 % in N e w Guinean.
42
Recently, a study has been carried in Thailand (Thithapandha, 1997).
Although the study recruited 5 probands with clinical trimethylaminuria for
urine analysis, the regular population study (n = 103) only showed around 4 %
subjects had impaired T M A A^-oxidation (with the aid of T M A load test only).
The other non-genetic factors such as age, diet, and disease, affecting
secondary trimethylaminuria will be discussed in Chapters 4 & 5.
3.2 Experimental protocols
Chinese male subjects (age range from 18 to 64 years) were recruited for the
study. All the subjects were fully informed in writing of the study
requirements and their written consents were obtained. The subjects
completed a questionnaire (Appendices A.1 and A.2), providing details of
fluid/food consumption over the periods of urine collection, with details of any
medication taken within the previous week. None of the volunteers reported
having any organic disease. T w o concurrent protocols: 1) whole day urine
collections; and 2) spot urine collections, were carried out simultaneously.
The whole day urine collection protocol was used in most of the previous
studies (Al-Waiz et al., 1987b; Hadidi et al, 1995; Mitchell et al., 1997;
Thithapandha, 1997; Zhang et al., 1996a), but a spot urine collection protocol
was used in only one recent study (Mitchell et al., 1997). Also, no comparison
was done before between these two protocols. In this study, these two
protocols were used simultaneously and compared together, in order to
determine the differences and similarities for these two protocols.
3.2.1 Whole day urine collections
The 24 hour urine samples were collected in four contiguous periods:
1) midnight (bedtime) to 7:00 a.m. (before breakfast);

2) 7:00 a.m. (after breakfast) to 1:00 p.m. (before lunch);
3) 1:00 p.m. (after lunch) to 7:00 p.m. (before dinner); and
4) 7:00 p.m. (after dinner) to midnight (before bedtime).
43
Urine from each individual period was collected in a plastic bottle
already containing hydrochloric acid (5 ml; 3 M). The corrected
volume (volume of acid was subtracted) of urine excreted in each
period was measured and a 50 ml aliquot was stored at -20 °C until
analysis of T M A and T M A O .
3.2.2 Spot urine collections
For the spot urine collection, as a more convenient and straightforward

protocol than 24-hr urine collection for primary trimethylaminuric
screening, the volunteer was asked to provide only one urine sample
after waking (7:00 a.m.) and before taking any food or drink. The
urine sample was collected in a plastic bottle already containing
hydrochloric acid (5 ml; 3 M ) and a 50 ml aliquot was stored at -20。C
for analysis o f T M A and T M A O .
3.3 Results
3.3.1 Whole day urine collections
Results were obtained for 159 Chinese male subjects aged 28.0 士 9.6
years (mean 士 s.d.) of w h o m 13 subjects were regular smokers (5 to 25
cigarettes per day) and the results are shown in Table 3.1. Results
obtained in British male subjects from a previous study (Zhang et al.,
1996a) were included for comparison. The amount o f T M A O excreted
(79.4 土 118.6 m g in Chinese, 53.7 m g in British, calculated as T M A )
was much higher than that offreeT M A (1.4 士 1.5 m g in Chinese, 1.6
m g in British). The large variation of T M A O (calculated as T M A )
excreted in both Chinese (3.5-947.0 mg) and British (4.0-965.7) might
be caused by the variations of dietary intake among the subjects. Both
studies showed similar results for % of T M A A^-oxidation.
44
Table 3.1 Descriptive statistics of urinary trimethylamine (TMA), trimethylamine iV-oxide ( T M A O ) (calculated as T M A ) , and
percentage T M A excreted as T M A O during a 24-hr period in a male Chinese population (n = 159) and in a male British
Caucasian population (n = 221) (Zhang et al�1996a)
T M A (mg) T M A O (calculated as T M A ) (mg) % T M A excreted as T M A O
Chinese Male British Male Chinese Male British Male Chinese Male British Male
Range 0.2-9.5 0.1 - 13.5 3.5-947.0 4.0 — 965.7 74.4-99.8 83.4-99.9
5 Mean 1.4 1.6 79.4 53.7 96.7 96.0
Median 0.8 1.0 36.8 25.8 97.6 96.6
S.D. 1.5 N o Data 118.6 N o Data 3.6 N o Data
SEM 0.1 N o Data 9.4 N o Data 0.3 N o Data
Skewness 3.48 3.1 3.87 6.3 -3.61 - 1.6
Kurtosis 13.91 11.5 20.55 43.5 17.33 3.3

Figure 3.1 showed the frequency distribution of the percentage of total
T M A excreted as T M A O in urine for the subjects under whole day
urine collection. The majority of subjects excreted more than 90 % of
T M A as T M A O in urine with mean and median values of 96.7 % and
96.6 % respectively. The distribution was skewed (skewness -3.61,
Table 3.1) with an asymmetric tail extending toward lower values.
Moreover, kurtosis gave a value of 17.33 (Table 3.1), which indicated
it was a relatively peaked distribution when compared with the normal
distribution. The pattern of the frequency distribution was very similar
to that in the British male subjects, as described by skewness (-1.6) and
kurtosis (3.3) in Table 3.1. In the present study of Chinese male
subjects, five subjects (3 % ) excreted 80-90 % of T M A as T M A O and
another two subjects (1 % ) excreted 74 % and 75 % o f T M A as T M A O
respectively.
3.3.2 Spot urine collections
Results were obtained for 111 Chinese male subjects aged 38.0 士 12.3
years (mean 士 s.d.) of w h o m 23 subjects were regular smokers (2 to 30
cigarettes per day) and these are shown in the third column of Table
3.2. The last column of combined data will be described in Section
3.3.3. Figure 3.2 showed the frequency distribution of the percentage
of total T M A excreted as T M A O in urine for the subjects in spot urine
samples. The results were similar to those obtained in the whole day
collection, the majority of subjects excreted more than 90 % of T M A
as T M A O in urine with mean and median values of 96.7 % and 97.3 %
respectively. The distribution was skewed (skewness -1.15) with an
asymmetric tail extending toward lower values. Although kurtosis
gave a value of 1.04 because of missing extreme outliners (< 70 % of
T M A excreted as T M A O ) , a relatively peaked distribution was still
presented when compared with the normal distribution. T w o subjects
(1 % ) excreted 80-90 % of T M A as T M A O but no subject excreted
less than 80 % o f T M A as T M A O respectively.
46
80l
70" ~ "
60"
I 50- 厂
0^
^
s
cr •
fe 40
30"
20"
10" ~
_ n r ~ h - r ~ r H
0
74 76 78 80 82 84 86 88 90 92 94 96 98100
% total TMA excreted as TMAO
Fig. 3.1 Frequency distribution histogram for the percentage total trimethylamine
( T M A ) excreted as trimethvlamine A/"-oxide ( T M A O ) in urine during 24 h
period in a Chinese population (n = 159V
47
Table 3.2 Descriptive statistics of the percentage T M A excreted as T M A O by whole day urine collections, spot urine tests and the
combined data (spot urine tests+whole day urine collections) in a male Chinese population (n = 159, whole day urine
collections: n = 111, spot urine test: n = 270, combined data)
% T M A excreted as T M A O
Whole Day Urine Spot Urine Combined Data
Range 74.4-99.8 88.9-99.5 74.4_99.8
Mean 96.7 96.7 96.7

4：^
00
Median 97.6 97.3 97.4
S.D. 3.6 3.1 2.2
SEM 0.3 0.2 0.2
Skewness -3.61 -1.15 - 3.46
Kurtosis 17.33 1.04 18.70

45i
40-
35~ ~
30-
^ -
S;:3 25
c^
^
fe 20-
15-
10-
5 -
0 ‘ ‘ ‘ I I I 1 "^"“^ ^^^^~ ^~^~ ^^^^~ ^"~^~~ ~^~~^"
74 76 78 80 82 84 86 88 90 92 94 96 98 100
% TMA excreted as TMAO
( T M A ) excreted as trimethylamine iV-oxide ( T M A O ) in urine by spot urine
test in a Chinese population (n = 111).
49
3.3.3 Comparison between whole day urine collections and spot urine
collections protocols
Statistical tests were performed to compare the whole day urine

collections and spot urine collections protocols. In spot urine
protocols, as samples were collected in the morning, the percentage
values o f T M A excreted as T M A O from all age groups were compared
to the values obtained at Period 1 (12:00 a.m. to 7:00 a.m.) in the
whole day urine collection, and also to the whole day combined
percentage values. In this comparison, Kruskal-Wallis One W a y
Analysis of Variance on Ranks was performed since the distributions
were skewed and not in a Gaussian form (normality test failed, p <
0.001). It was found that the p value was greater than 0.05, so the
differences in the median values among the treatment groups were not
great enough to exclude the possibility that the difference was due to
random sampling variability. The unpaired t-test or Mann-Whitney U-
test also produced no statistical difference (p > 0.05) when performed
to compare the morning session (Period 1) of the 24 hr collection
protocol and the spot urine data within all age groups combined.
Therefore, it appeared to make no statistical difference to choose spot
urine collections or whole day urine collections for the investigation of
the general distribution of T M A iV-oxidation. It was considered that
this statistical insignificance permitted us to combine data from the
spot urine tests and data from whole day urine collections for the
investigation of the general distribution of T M A #-oxidation, and the
combined data are plotted in Figure 3.3.
Figure 3.3 showed the frequency distribution of the percentage of total

T M A excreted as T M A O in urine for the subjects in spot urine tests
and in whole day urine collections (n = 270). The majority of subjects
excreted more than 90 % of T M A as T M A O in urine with mean and
median values of 96.7 % and 97.4 % respectively (Table 3.2).
Moreover, the distribution was skewed (skewness -3.46) with an
50
1201
100一
8 0 一
^
^
fl
I^
u
60-
40-
20"
0 ^—r=l 1 1 1 ^^^^^=H=4—f—r—r—
74 76 78 80 82 84 86 88 90 92 94 96 98 100
% TMA Excreted as TMAO
( T M A ) excreted as trimethylamine A^-oxide ( T M A O ) in urine in a Chinese
population (n = 270). Data combined from whole day urine collection and
spot urine test.
51
asymmetric tail extending toward lower values. Kurtosis also had a
value of 18.70. Totally, the combined result had seven subjects (3 % )
who excreted 80-90 % of T M A as T M A O and another two subjects (1
% ) who excreted 74 % and 75 % o f T M A as T M A O respectively.
3.3.4 Comparison between smokers and non-smokers
Mann-Whitney Rank S u m Test had been performed for the smokers

and non-smokers in this population study and it showed no statistical
significance (p > 0.05) for the effect of smoking on the percentage
TMAA^-oxidation (smoker: 96.4 ± 2.6 % , n = 36; non-smoker: 96.7±
3.2%,n = 234; mean±s.d.).
3.3.5 Comparison of the results obtained in four individual periods of whole

day urine collections
Data for individual periods of urine collection are described in Table

3.3 and the curves are plotted in Figures 3.4 & 3.5. The data showed
the highest amounts of T M A and T M A O excreted and the highest
percentage value of T M A excreted as T M A O in Period 1. W h e n
compared to the corresponding data during the other three periods,
significant results were obtained (p < 0.01, T M A ; p < 0.005, T M A O
(calculated as TMA); p < 0.05, % T M A excreted as T M A O ) . Period 2
showed the lowest values within the collection periods, and when
compared the values to other three periods, significant results obtained
for T M A Q? < 0.001) and T M A O (calculated as T M A ) (p < 0.005).
W h e n the % of T M A excreted as T M A O at Period 2 was compared to
other periods, only Period 1 showed statistical difference {p < 0.05).
52
Table 3.3 Amounts of urinary T M A and T M A O (calculated as T M A ) excreted and the percentage urinary T M A excreted as T M A O in four
continuous urine collection periods in Chinese population (n = 159).
T M A (mg) T M A O (calculated as T M A ) (mg) % T M A excreted as T M A O
Period 1 0.5±0.8* 35.3±68.4** 96.8±3.4*
ui Period 2 0.3±0.8 10.2±21.1 95.7±6.3

U)
Period 3 0.3±0.2 14.5±27.9 95.3 ±6.1
Period 4 0.3±0.4 19.4±31.3 95.8 ±5.2
N.B. Data is presented as mean 士 s.d. */? < 0.05, **/? < 0.005 compared to other three periods.
0.6 n r 50
T -- 45
0.5 - 氺仏
0、 - 40
\
V
0.4— * * f s ^ \ - - 3 5 |
I \ � � T\ _ - i - 30 I
I 0.3 - \ 、、丄 i — 25 ^
: 2 - —
- 1 5 ,
. \ \ ^ - ^ ^ __^ 1
^ 0.卜 > ^ -]o
--5
0 H 1 1 1 h 0
Period 1 Period 2 Period 3 Period 4
Figure 3.4 Amounts o f T M A ( — • — ) and T M A O (as T M A ) ( — • — ) excreted in urine at the four collection periods. Each point and error
bar represents mean value and S.E.M. respectively, n = 159. *p < 0.05. **p < 0.005 compared to other three periods.
97.5n
97- 氺 T
霞 9 6 . 5 - ^ \ T T
S 96 \ ^ ^ ^
|95.5- ^^^^^__-^^^"""^"""""^^"""^^"^^"^^
1^ 丄
0^
I 95- J-
^ b^
^ ^ 94.5-
94-
93.5 4— 1 , ,
Figure 3.5 Percentage T M A excreted as T M A O (calculated as T M A ) in urine at the four collection periods. Each point and error bar
represents mean value and S.E.M. respectively, n 二 159. *p < 0.05 compared to other three periods
3.4 Discussion
In the present study, both whole day urine collections and spot urine test
protocols were utilized concurrently. They were compared to determine the
differences and similarities for these two protocols by statistical tests, as
described in Section 3.3.3, on the results of percentage T M A A^-oxidized to
T M A O . In this comparison, Kruskal-Wallis One W a y Analysis of Variance
on Ranks found that no statistical difference for these results and therefore
there appeared to be no statistical difference to choose spot urine collections or
whole day urine collections for the investigation of the general appearance of
the distribution of T M A A^-oxidation. Figure 3.3 showed the combined data
from these two protocols for the primary trimethylaminuria investigation. The
whole day, spot, and combined percentage values of T M A A^-oxidized are
already described in Table 3.2. Moreover, the whole day urine collection
protocol in the present study had been modified from the original method. In
previous studies (Al-Waiz et al., 1987; Hadidi et al., 1995; Thithapandha,
1997; Zhang et al., 1996a), the excreted urine within the 24 hr period was
collected in one large bottle. The matrix composite was then sampled at the
end of this period. This was inconvenient for the volunteer to handle a large
bottle and to keep it with them for a whole day. It also lost information
regarding the amounts and rates of change of T M A and T M A O excretion, and
the change of the percentage T M A excreted as T M A O during the collection
period. Therefore, the current 24-hr urine collection was divided into four
contiguous periods (see Section 3.2.1), and for each period the subject was
provided with a smaller sized urine collection bottles. This made it possible
for the volunteers to continue their normal duties and activities during the
weekdays whilst making the urine collection. Furthermore, subdivided
collection periods provided useful data on the effects of different types of diet
and activities on the T M A and T M A O excretion. In Chapter 4, it is described
more clearly about the changes of amounts/rates of T M A , T M A O , and
percentage T M A excreted as T M A O in urine according to the changes of diet.
Although spot urine tests could not display any persisting effect of dietary
intake on T M A and T M A O excretion, it was not necessary to reject spot urine
tests as another useful and convenient protocol for homozygous affected
56
subjects (primary trimethylaminurics) investigation. Moreover, a similar spot
urine collection protocol had already been performed in Ecuadorian and N e w
Guinean populations (Mitchell et al., 1996), and it was possible to develop
interpretable polymorphic frequency distribution histograms for those
populations.
The amounts of T M A and T M A O excreted in the urine, and the percentage of

T M A excreted as T M A O , were similar to those reported in previous studies
(Al-Waiz et aL, 1987; Hadidi et aL, 1995; Zhang et al., 1996; Thithapandha,
1997). Table 3.1 showed the amounts of T M A and T M A O , and the
percentage o f T M A excreted as T M A O , in British male subjects (n = 221) in a
recent study (Zhang et al.) 1996). The higher mean and median T M A O values
in Chinese may be related to differences in dietary habits in southern China
compared to Great Britain. For % T M A excreted as T M A O , it also showed
similar results (96.0-96.7, mean; 96.3-97.6, median) in these British male
subjects and in the Chinese male subjects under spot urine tests, 24 hr urine
collections, and combined data.
The frequency distribution for the degree of T M A 7V>oxidation obtained in this

Chinese population was very similar to the distributions obtained in British
Caucasian populations (Al-Waiz et al., 1987b; Zhang et al., 1996). None of
the subjects in our study or in these two other studies were identified as
primary trimethylaminurics under normal conditions. Intra-individual
variation was not determined but a previous study showed a negligible effect
(coefficient of variation < 1%) (Al-Waiz et al., 1987b). The majority of
subjects in this and other studies excreted greater than 9 0 % of the total T M A
as T M A O in urine, and only few individuals (< 1%) excreted less than 80%
T M A as T M A O in urine. In the British Caucasian population study (n 二 169)
(Al-Waiz et al, 1987b), the subjects who had only slightly impaired ability in
T M A A^-oxidation under normal dietary conditions were identified as
heterozygous carriers by a T M A load test. In the two subjects who excreted
less than 80 % of T M A as T M A O , it may be reasonable to conclude that this
moderate impairment of A^-oxidation was consistent with carrier status for the
57
allele resulting in poor metabolism. However, previous studies of
heterozygotes for the allele for impaired A^-oxidation of T M A , who were
identified as being the parents of subjects with the homozygous condition,
showed that they often excreted more than 9 0 % of T M A as T M A O under
normal dietary conditions (Ayesh et al., 1993), so some heterozygotes could
be missed by this screening test. Also, it is not certain whether subjects who
excreted 80 to 90 % T M A as T M A O in urine are heterozygote carriers for the
allele for poor metabolism. However, it is reasonable to assume that (without
the T M A load test) subjects who excreted less than 80 % of T M A as T M A O
in urine are the carriers of the allele for impaired T M A TV-oxidation (Mitchell
etal., 1997), and a genetic deficiency in T M A A^-oxidation presumably should
occur in different populations.
The second British Caucasian population study (n = 421) (Zhang et al., 1996)
showed that the six subjects (1%) who excreted less than 8 0 % of T M A as
T M A O in urine were females. The over-representation by females in this
study in Britain may be due to the interaction of metabolism and renal
clearance, especially for compounds with high renal clearance (Jackson et al.,
1986). This was because of differences in the efficiency of secretory
mechanisms or as a result of dietary habits as they affected urine p H and hence
tubular reabsorption. Also, females who were suffering from bacterial
vaginosis showed higher T M A levels in vaginal discharge than control
females (Brand and Galask, 1986; Sardas et al, 1996), and it might affect the
T M A level in urine samples taken from female volunteers. Also increased
urinary T M A levels have been found in patients with cervical cancer as well
as with vaginitis (Brewster and Schedewie, 1983). Although nicotine is a
substrate for F M O , according to the information provided from the volunteers,
apparently smoking did not alter the pattern of T M A iV-oxidation in this study
{p>0.05) (Section3.3.4).
U-shaped curves appeared for the results obtained for T M A and T M A O

(calculated as T M A ) excretion and the percentage T M A excreted as T M A O in
the four continuous periods (Figs. 3.4 & 3.5). Period 1 showed the highest
values and period 2 and 3 produced relatively lower values than the first and
58
the last periods. The highest values for T M A and T M A O excretion at Period
1 may be explained due to the longest duration for urine collection at this
period than that at the other three periods. However, those values were
increased again at Period 4. The high amounts of T M A and T M A O excreted
in the urine during Periods 1 and 4 may be related to relatively high intake of
food at the dinners before periods 1 and 4, since many foods contains T M A ,
T M A O and T M A precursors (e.g. choline, carnitine). The relative high
percentages of T M A excreted as T M A O in urine at Periods 1 and 4, when
compared to Period 2 and 3, may also be related to the high intakes of T M A ,
T M A O and T M A precursors. The study on the effects of foods will be
described more clearly in Chapter 4.
One subject who excreted less than 8 0 % of T M A as T M A O in urine reported

that he had eaten fried rice containing cauliflowers and sprouts at dinner just
before starting the urine collection, and therefore as the percentage value of
T M A excreted as T M A O in urine at Period 1 was lower than that at the
subsequent periods, it may have been related to this food intake. Cauliflowers
and sprouts are known to contain progoitrin, which is a potential inhibitor of
T M A 7V-oxidation (Fenwick et al, 1982). The effect of progoitrin on T M A TV-
oxidation in man due to the consumption of these vegetables is still not clear
(Fenwick et al., 1983). One previous study found Brussel sprouts and swede
had no effect on the capacity to iV-oxidize T M A (Al-Waiz, 1988). Therefore,
the effect of those vegetables on the low percentage T M A oxidation in this
subject must be considered. These inhibitors may contribute to the reduction
of the percentage of T M A excreted as T M A O to a certain extent, which might
lead to a lower percentage value for T M A as T M A O in urine.
Combined data for the percentage T M A #-oxidized (Fig. 3.3) were also
transformed by logarithmic calculations to allow clearer visualization of
whether there was the presence of a bimodal distribution. Figure 3.6 showed
the frequency distribution of the logarithmic ratio of T M A excreted as T M A O
in urine. In this logarithmic distribution histogram, no distinct bimodal
distribution appeared. It is therefore more likely that the distribution is just
skewed with a long tail.
59
140
——120
100
80 ^
a
Qj
s
a^
QJ
60 ^
40
20
~,~~F=i~,~",~~,~I I ) r ^ ~ ~ ~ ~ i o
-15 -14-13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1
Log (Ratio o f T M A excreted as TMAO) X 100
Fig. 3.6 Frequency distribution histogram for the logarithmic ratio of total
trimethylamine ( T M A ) excreted as trimethylamine A^-oxide ( T M A O ) in
urine in a Chinese population (n — 270). Data combined from whole day
urine collection and spot urine test.
60
One common problem for researchers is how to classify primary
trimethylaminurics or secondary trimethylaminurics according to the
percentage of T M A excreted as T M A O . The commonly accepted value to
classify normal volunteers was > 90 % based on previous experience.
However, there is some controversy about how to classify primary (frank)
trimethylaminurics (homozygous affected subjects, severe impairment in
TMA A^-oxidation,) and secondary trimethylaminurics (including
heterozygous unaffected subjects, moderate impairment in T M A A^-oxidation).
The borderline for separation of mildly affected cases (secondary
trimethylaminuria) and the normal subjects (homozygous unaffected subjects)
falls in the 80 to 90 % range, and the margin for categorizing of homozygous
affected subjects was < 50 - 60 % . It should not be difficult to identify
homozygous affected subjects because they should produce a fish-like
unpleasant body odour. However, it is sometimes difficult to identify
secondary trimethylaminurics because they may appear normal under normal
dietary conditions. However, the % T M A conversion will be reduced if the
intake of T M A and its precursors are high enough to saturate the liver F M O
activity since the F M O activity is already impaired by genetic or other
environmental or disease conditions (e.g. liver diseases). Although the T M A
load test (Zhang et al., 1995) can be performed to identify these heterozygous
subjects, it was not undertaken in this study, and it provides no background
information about the effects of other environmental factors (e.g. diseases,
drugs, and/or dietary effects). This study focused on the effects of these
factors on T M A and T M A O excretion in a Chinese population. It will be
discussed in more detail in the next chapter.
Table 3.4 describes the results of impaired T M A #-oxidation in various

populations. Initially, sufferers of Fish-odour syndrome were found in the
Caucasian populations. However, recent studies also showed the existence of
impaired T M A iV-oxidation in other ethnic groups including Southern Asians
(Thithaoandha, 1997). In a Thai population, 5 sufferers were found clinically
to have primary trimethylaminuria, but no sufferer was identified in a normal
population survey. Moreover, only two subjects excreted < 90 % (84 % and
61
Table 3.4 Comparative incidence of dysfunctional iV-oxidation of trimethylamine in various populations
Subjects excreting < 80 % total material

as the #-oxide
Population Number Number Percentage Reference
Jordanian 116 2 1.7 Mitchell etal, 1997
Ecuadorian 80 3 3.8 Mitchell et al, 1997

ON
^ NewOzinean • n 11.0 Mitchellda/.,1997
Jordanian 82 8 9.8 Hadidi etal, 1995
British 169 1 0.6 Al-Waiz etal, 1987
Thai 103 0* 0* Thithapandha, 1997

Chinese 270 2 0.7 Present study
Only two subjects excreted the TMAA^-oxide at 84 % and 86 % before T M A load test was utilized.
86 % ) of T M A as T M A O in urine, but some subjects (who originally had
values > 90 % ) also showed impaired T M A A^-oxidation after the T M A load
test (> 80 % T M A excreted as T M A O ) . Zhang et al. (1995) also showed
similar findings for the T M A load test in the obligate heterozygous unaffected
subjects. This may mean that the margin (90 % T M A excreted as T M A O )
proposed by the researchers in the studies of British Caucasian populations is
not a good and appropriate definition to classify homozygous unaffected
subjects and heterozygous clinically unaffected subjects. Moreover, it is not
clear whether all cases with a genetic cause have the same mutation and there
are probably additional polymorphisms of the gene, which may have some
effect. Therefore, the only way to resolve the problem is to examine a large
population sample with phenotypes and genotypes. If a clear bimodal
distribution is found this would suggest heterozygotes are showing reduced
activity and an 'arbitrary' cut-off point can be made. It is more likely that the
distribution is just skewed with a long tail due to environmental factors,
diseases or one or more other polymorphisms.
In conclusion, the statistical insignificance permitted us to combine data from

these two protocols for the investigation of primary trimethylaminuria (Fig.
3.3). Also, this study has shown that about 1 % of the Chinese population
studied showed mild deficiency in T M A #-oxidation. To determine whether
this represents a genetic variation or whether it is due to environmental or
dietary factors will be the subject of further studies. Moreover, the results
obtained in this study are very similar to the results obtained in the
pharmacogenetic studies of another important liver enzyme: cytochrome P-
450 system, i.e. C Y P 2 D 6 activity in the hydroxylation of drugs. Initial studies
distinguished two groups of subjects (poor metabolizers and extensive
metabolizers) in Caucasian populations but not in Chinese. Therefore, it is
reasonable to conclude that genetic polymorphism of enzymes may vary in
different ethnic groups, and population studies should be carried out in various
populations in order to obtain thefiillpicture of the polymorphism.
63
Chapter 4
Effect of Age and Diet on
Trimethylamine A^-Oxidation
4.1 Introduction 65
4.2.1 Effect of age on T M A 7V-oxidation 73
4.2.2 Effect of diet on T M A A^-oxidation 73
4.2.3 Effect of control diet on T M A 7V-oxidation 75
4.3 Results 76
4.3.1 Effect of age on T M A 7V>oxidation 76
4.3.2 Effect of diet on T M A A^-oxidation 80
4.3.3 Effect ofcontrol diet on TMA7V-oxidation 83
4.4 Discussion 89
64
4.1 Introduction
The elderly represents a significant proportion of the population in many

developed countries, e.g. in Hong Kong, European and North American
countries. Elderly people are also one of the most medicated segments in
those societies. In the U S A , the elderly segment is the most medicated and
accounts for 25 % of all prescription drugs dispensed (Christopher et al., 1979;
Eckhardt, 1978). The cost of these prescribed drugs is more than $15 billion
per year. Schmucker (1985) mentioned that this marked increase in drug
treatments for the elderly was also linked to a greater incidence of adverse
drug reactions in geriatric patients in comparison to younger subjects
(Hurwitz, 1969; Schmucker etal., 1984).
Although many studies reported that the effects of drugs are changed as a
function of age, other studies did not show any significant effect of age on
certain drug metabolism (summarized in Table 4.1). Moreover, studies also
showed that drug metabolism by Phase II conjugation reactions was
unchanged with normal aging (Greenblatt et al, 1982c), but the Phase I
oxidation reactions declined with aging and resulted in the reduction of drug
clearance in the elderly (Greenblatt etal., 1986; Montamat et al., 1989; Richey
and Bender, 1977; Vestal, 1982). Phase I enzymes include both the
cytochrome P450 enzymes and flavin-containing monooxygenase. Many
investigations had been carried out on the effect of age on hepatic drug
metabolizing enzymes in animal and in human. The results for animal liver
drug metabolizing enzymes are listed in Table 4.2. O n the other hand, the
investigations of the effect of age of drug metabolizing enzymes in human
liver had different results from the data obtained in animal studies. Most
studies showed drug-metabolizing enzymes, e.g. cytochrome P450s are not
affected by age. Wynne et al. (1988) studied the kinetics of the microsomal
monooxygenase 7-ethoxycoumarin-O-de-ethylase in 17 human liver biopsy
specimens, and no correlation was observed between age and microsomal
protein recovery, maximal enzyme activity or apparent enzyme affinity. It
was assumed that a fall in liver volume or blood flow should be the major
cause for the decline in clearance of many oxidized drugs which occurs with
65
Table 4.1 Previous studies of the relation of age to the clearance of drugs cleared bv
hepatic biotransformation (Data modified from Greenblatt et aL, 1982c)
Drug or Metabolite Initial Pathwayof References

b Biotransformation
Evidence suggesting age-related reduction in clearance:

0'Malley et al., 1971, Liddell et al., 1975,
Antipyrine Oxidation Wood et al., 1979, Vestal et al., 1975,
Greenblatt et al., 1982b
Kanto et al., 1979, Macklon et al, 1979，
Diazepam Oxidation Greenblatt et al., 1982a, Klotz et al., 1975，
Ochs etal., 1981
Chlordiazepoxide Oxidation Shader et al.，1977, Roberts et al., 1978
^ ^, , , . r~77 Allen etal., 1980, Shader etal., 1981，
Desmethyldiazepam 0>adatKm 馳 and Mtlller-Seydlitz, 1979
DesaUcymurazepam Oxidation Greenblatt et al., 1981b
Clobazam Oxidation Greenblatt et al., 1981a
Alprazolam Oxidation Greenblatt etal., 1983
Quinidine Oxidation "Ochs et al., 1978，Drayer et al., 1980
Theophylline — Oxidation Jusko et al.，1979, AntaI et al., 1981
D 11 TTT" VQSi2\etal, 1979,
Pr—olol Oxidation Castleden and George, 1979，Feely et al, 1981
Nortriptyline Oxidation Dawlingg/^/., 1980
Small or negligible age-related change in clearance:
“ ~~“ ^ Greenblatt et al., 1980，
Oxazepam Glucuromdation ^ 騒 et al., 1976, Ochs et al., 1981
T ^, ., ,. Karuse/fl/., 1978,
Lorazepam GlucuromdaUon 。職触 et al., 1979
Temazepam Glucuronidation Divoll et al., 1981
Warfarin Oxidation Sherpherd et al., 1977
Lidocaine Oxidation Nation et al., 1977
Nitrazepam Nitroreduction Castleden et al., 1977, Kangas et al., 1979
^1 . “ Oxidation, Kantos etal., 1981
Fluitrazepam .备 , “ ，
mtroreduction
Isoniazid Acetylation Farahg^Q/., 1977
Ethanol Oxidation ^ e s t a l et al., 1977
Metoprolol Oxidation Quartermangfa/.� 1 981
Digitoxin Oxidation Donovang^ al., 1981
Prazosin Oxidation Ruhinetal., 1981
Data conflicting or not definitive:
Meperidine Oxidation Metherg/ al., 1975
Phenylbutazone Oxidation 0'Malley etal., l91l,Thggsetal., 1975,
Fnenylbutazone OxidaUon whittaker and Evans, 1970
Phenytoin Oxidation Hayes et al., 1975, Shewin et al.，1974
Imipramine Oxidation Nies etal., 1977
Amitriptyline Oxidation Nies etal., 1977
Paracetamol Glucuronidation, Triggs et al., 1975, Briant et al., 1976
(acetaminophen) sulphation Fluton et al., 1979, Divoll et al., 1982
Amobarbital Oxidation lwmQetal, 1974
66
Table 4.2 Previous studies on the effect of age on liver drug metabolizing enzymes
in animals
Enzyme studied Effect of aging References

Aldrin epoxidase (AE) Decreases activity W y n n e etal, 1987
Catalase Less resistant to heat Ando et al, 1997
CYP1A1 Decreased inducibility of Horbach etal., 1990
the corresponding m R N A
CYPlA2 Decreased inducibility of Horbach^/a/., 1990
the corresponding m R N A
rvP2Rl Decreased expression rate Agrawal and Shapiro,
ofits m R N A 1 ^
CYP2B2 Decreases activity Shimada et al., 1995
CYP2C11 Decreases activity Nakajima./a/.199^^
Shimada etaL, 1995
CYP2C12 Increases activity Shimada et al.，1995
CYP2C13 Increases activity Bandiera et al., 1986
CYP2C7 Increases activity Bandiera et al., 1986
CYP2E1 Decreases activity Nakajima etal, 1992
CYP3A Decreased inducibility by Lee and Werlin, 1995
dexamethasone
CYP3A2 Decreases activity Shimada etal, 1995
— CYP2P1 Decreases activity Shimada et al., 1995
^CytochromeP450^ Decreasesactivity Guo,,a/., 1993
reductase
Cytochrome P450s Less resistant to heat Ando et al., 1997
DT-Diaphorase N o effect — G u o etal, 1993
Ejhoxyresof^j- ^ D e c r e a s e s activity""" Wynne^l987
deethylase (EOR)
Glutathione ^-transferase ^ ^ff^^t Guo etaL, 1993
(GST) —
A/-nitrosodimethylamine Decreases activity in G u o etal, 1993
demethylase C N D M A d ) female mice
Pentoxyresorufin 0- Decreases activity in G u o etal, 1993
dealkylase ( P R O D ) female mice
Sulphotransferase N o effect Guo etal., 1993
UDP-Glucuronosyl- N o effect Guo etal., 1993
transferase
87
aging. Hunt et al. (1990) also sought to define potential aging-related
alterations in CYP2E1, by examining the activity of this isoform in vitro, in
liver microsomal samples obtained at surgery from patients age 60-75, and
comparing these results to those obtained in liver microsomes prepared in
younger subjects, age 30-59. The activity of hepatic C Y P 2 E 1 was again
unaffected by normal aging. Another extended study (Schmucker et al., 1990)
also demonstrated that there was no significant difference in the microsomal
concentrations of several cytochrome P450 isozymes with aging. Also, the
activity of the human hepatic cytochrome P450, C Y P 3 A , was quantified in
vitro as erythromycin N-demethylation in microsomes prepared from 43
resected human liver specimens obtained from patients, age 27 to 83, with
normal liver function. It was not surprising that the activity of C Y P 3 A did not
correlate with age and other factors: smoking status, ethanol consumption, etc.
(Hunt et aL, 1992). Moreover, one study showed a negative correlation
between age and total cytochrome P450 content, NADPH-cytochrome c
reductase activity and levels of CYP2E1 and 3 A proteins, although the
C Y P l A 2 and 2C proteins were unaltered with advancing age (George et al.,
1995). These results were different to the results in animal studies, and
therefore more extensive study for the d m g metabolizing enzymes in human
liver is needed. Moreover, no study has been done on the effect of age on the
activity of flavin-containing monooxygenase--another important drug
metabolizing enzyme, in human liver. F M O may show similar or different
patterns of the effects of age when compared to cytochrome P450s, which has
not been considered before. Therefore, w e reported the results of the age
effect on liver F M O activity in a Chinese population, using urinary T M A as
the probe.
Besides the age factor, the metabolism of drugs and other xenobiotics could
also be influenced by a variety of factors including dietary intakes (Campbell
and Hayes, 1974; Peterson and Holtzman, 1980). The effects of diet on d m g
metabolism have been studied in both animal models and recently, in humans.
Those studies include: the effects of food on the bioavailability of drugs, the
identification of special dietary components that can influence drug
metabolism, and the effects of those dietary components on enzyme activities.
68
Since the major enzyme systems responsible for drug metabolism also
transform other exogenous chemicals and xenobiotics, e.g. carcinogens
(Conney, 1986; Kuntzman et al., 1964), change of enzyme activities for those
substances can indicate the diseases or deficiencies of factors in individuals
with altered drug metabolism and disposition.
Differences in food intake, both in amount and type, have been shown to alter
the bioavailability of several antihypertensive drugs such as propranolol and
metoprolol (Melander et al., 1977; Melander, 1978). However, other studies
had shown that food reduced the absorption of many antibiotics, e.g. isoniazid,
rifampicin, penicillin, ampicillin and tetracycline, and had no effect on the
bioavailability of other drugs (Melander, 1978; Welling, 1977). Factors that
influence the bioavailability of drugs include direct binding of drugs by food,
or by altering luminal pH, gastric emptying, intestinal transit, mucosal
absorption, splanchnic-hepatic blood flow, etc. (Anderson, 1988). Dietary
influences on the microflora in the intestine may be another factor to affect the
bioavailability of drugs when the drugs undergo metabolic transformation by
these microorganisms (Scheline, 1973). Dietary macronutrients such as
proteins, lipids, and carbohydrates also showed significant effects on drug
metabolism. The metabolisms of antipyrine, theophylline, and propranolol in
high carbohydrate, in high protein, and/or in high fat diets were studied in
human subjects. It was found that substitution of protein for either fat or
carbohydrate increased d m g oxidation rates (Alvares et aL, 1976; Anderson et
al., 1979 & 1982; Fagan et aL, 1987; Kappas et al., 1976). Other reports
showed that increased amounts of polyunsaturated oils in diets increased the
cytochrome P450 and mixed function oxidases activities, while the lipid-free
diets decreased their activities (Davidson and Wills, 1974; Marshall and
McLean, 1971; Norred and Wade, 1972). The reason is the B-position fatty
acid in phosphatidylcholine is highly unsaturated, and this phospholipid is
essential to maintain the integrity of microsomal membrane. Amino acids,
e.g. tryptophan increased liver protein synthesis and induced the mixed
function oxidase system in animals and in liver cell cultures (Evarts and
Mostafa, 1981; Paine, 1976; Sidransky，1986). High carbohydrate intakes
resulted in a decreased cytochrome P450 activity and rate of drug metabolism
69
in animals (Campbell and Hayes, 1974), but had an inhibitory effect on a-
aminolaevulinate synthetase (Tschudy, 1978). Ascorbic acid deficiency
impaired drug metabolism and reduced cytochrome P450 activity in laboratory
animals (Holloway and Peterson, 1984; Horio et aL, 1986). Brodfuehrer and
Zannoni (1986 and 1987a) found that deficiency in ascorbic acid in the diet
also led to a reduction in F M O activity. They also confirmed that reduction in
hepatic F M O activity was not related to stress (Brodfuehrer and Zannoni,
1987b). Moreover, polynuclear aromatic hydrocarbons (PAH) are commonly
present in processed food due to incomplete combustion and pyrolysis of
petroleum and coal; some of them are carcinogens. Charcoal-broiled meat is a
typical food containing P A H . Since drug oxidations in smokers are enhanced
by these chemicals (Jenne et al., 1975; Kuntzman et al., 1977; Pantuck et al.,
1974), it seems the charcoal-broiled beef might also increase the rate of drug
metabolism. This hypothesis was confirmed in the studies of antipyrine,
theophylline, and phenacetin metabolisms, in which the subjects were fed with
charcoal-broiled beef (Conney et al., 1976; Kappas, A., 1978). Certain
indoles, present in cruciferous vegetables such as brussels sprouts, cabbage,
and cauliflower, had been found to induce oxidations (Loub et al., 1975;
Pantuck et al., 1976). Those effects on drug oxidations and conjugations have
been investigated in human subjects, where the metabolism of antipyrine and
phenacetin were markedly increased and the mean ratio of conjugated N-
acetyl-p-aminophenol was also increased (Pantuck et al., 1979). Moreover,
ingestion of cabbage and brussels sprouts also increased the glucuronidation
and metabolic clearances of paracetamol (Pantuck et al., 1984). However,
paracetamol metabolism due to drug conjugation was not affected in subjects
fed with charcoal-broiled beef (Anderson et al., 1983).
Although many studies have been carried out to investigate the effects of diet
and dietary components on enzyme activity involved in drug metabolism (see
last paragraph), few investigations studied the effects on the F M O activity.
70
Therefore, it is not known whether the activity of F M O may be affected by
dietary factors in normal subjects or in unaffected subjects with heterozygous
trimethylaminuria. Since many types of food contain large amounts of
trimethylamine (TMA), trimethylamine iV-oxide ( T M A O ) or their precursors,
high intake of those foods may alter the activity of F M O in A^-oxidation of
T M A in these subjects. As described in Table 1.1, fish is a major source of
T M A and T M A O , while meat and meat products, e.g. beef, lamb, also contain
carnitine. Cauliflower, egg, soybean, and their products also contain choline
and lecithin, which are precursors of T M A , generating T M A in vivo through
the conversion of gut bacteria (Fig. 1.3) (De La Huerga and Poper, 1951;
Prentiss et al., 1961). High intakes of T M A , T M A O , and their precursors
produce high levels of free T M A in the gut by enterobacterial degradation.
Conversely, some Brassica species and cruciferous vegetables, including
cabbage, cauliflowers, Brussel sprouts, and swedes also contain progoitrin,
which is an inhibitor of F M O enzyme (Fig. 4.1). Intake of these vegetables
may lead to the inhibition of T M A A^-oxidation (Fenwick et aL, 1982; Oginsky
et al., 1965). It was uncertain whether the consumption of these vegetables
led to secondary trimethylaminuria in man (Fenwick et cd., 1983; Al-Waiz,
1988), although Pearson et al (1978) demonstrated in a breed of chicken with
a genetically reduced ability to A/-oxidize T M A that further impairment in
T M A 7V-oxidation occurred when fed with Brassica rapeseed meal. Another
report showed that a high intake of milk from cows fed from a wheat pasture
with significant levels of T M A O produced trimethylaminuria in a 10-year-old
boy (Rothschild and Hansen, 1985). The present study was carried out to
investigate the effects of diets and the dietary components on F M O activity by
measuring the T M A and T M A O in a Chinese population. The F M O activity
was then correlated with the subjects' age and dietary intake during the urine
sampling period.
71
H2C N H H2N
H
H 2 C = C — H C
I I
C = S /
/ C = S
\ / H^N
0
^ .‘ . Thiourea
Goitrin
H N — — C
/? H3C
/ \ z N
S = C CH C ^ \
\ / /.C-SH
H N — C H C ^ “
\ ^ N
CH2CH2CH3
Propylthiouracil Methimazole
Fig. 4.1 Inhibitors of theflavin-containingmonooxygenase
72
4.2.1 Effect of age on TMA N-oxidation
Urine sample collection procedures were carried out during the

population study as described in Section 3.2. Data were collected after
G C analysis. Both 24-hr and spot samples were analyzed. In the 24-hr
urine collection, urine samples were collected during the four periods
(midnight to wake up and before breakfast; after breakfast and before
lunch; after lunch and before dinner; after dinner and before bedtime).
These samples were used to determine the effect of age on the changes
of urinary T M A , T M A O and percentage T M A A^-oxidized during the
24-hr period. However, the whole day values of percentage T M A N-
oxidized in 24-hr urine collection were combined with the values of
percentage T M A 7V-oxidized in spot urine tests for the investigation of
the effect of age on T M A iV-oxidation only. Data were analyzed by
suitable statistical analysis tool pack (SigmaStat).
4.2.2 Effect of diet on TMA N-oxidation
The effect of diet on T M A 7V-oxidation was also based on the urine

samples obtained from the population study. The information provided
by the volunteers in the questionnaires in the 24-hr urine collection
protocols (Section 3.2.1) and data obtained from Period 1 and Period 4
in 24-hr urine collection protocols were used for analysis, details of
which were described in Table 4.3. Based on the major types of food
intake at the dinner, individuals were classified into several food
groups: seafood, beef, chicken, and low-TMA diets. Some data had
been omitted in the dietary study due to the reasons listed in Table 4.4.
Amounts and rates of T M A and T M A O excretion in urine and the
73
Table 4.3 Schedules for urine collection in Periods 1 and 4 (24-hr urine
collection protocols)
Period 1 Period 4
Dining Time 8:00 p.m., Day 0 8:00 p.m., Day 1
Urine Collection 12:00 a.m., Day 1 to 8:00 p.m., Day 1 to

Time 7:00 a.m., Day 1 ll:59p.m.,Day 1
Duration 7 hours 3 hours
Table 4.4 Reasons that the data should be omitted in the statistical
analysis for dietary effect
Examples
Reason Types of food taken Amounts
hcomplete (Blank) (Blank)

questionnaires for diets
Information is too rough Seafood dinner Heavy

and general Hot-pot dinner Little bit
Beef (Blank)
Amounts of food taken T^. , ^ , ^
Fish Enough amount
werenotspecified chicken Very little
Dietary intake was too .QvêftetfTalad Too complicate to

complicated ^^^'^^‘‘ b e ^ saiad,
garoupa, jelly, ice- ^
cream)
Insufficient size for Roast Lamb (n = 1) N/A
statistical analysis Beancurd (n 二 2) N/A
74
percentage of T M A iV>oxidation were determined. Moreover, an
arbitrary number between 1 to 10 (i.e. 1 = very little amount, 10 = very
heavy amount) was assigned to each dietary component, according to
the amounts of dietary components taken by the volunteers. The
assigned number helped to select and group the subjects into suitable
food groups.
4.2.3 Effect of control diet on TMA N-oxidation
Five volunteers were recruited in this study for three consecutive days.
Urine samples were collected and volumes of excretion were measured
over this period. Different types of foods were taken at each dinner to
evaluate the effects on T M A 7V-oxidation. Only foods containing
insignificant amounts of T M A , T M A O , or their precursors were taken
by the volunteers between the dinners (Appendix B). Table 4.5
describes the major foods to be taken by the volunteers during the
experiment. The amounts and rates o f T M A , T M A O excreted, and the
percentages T M A oxidation in urine samples were analyzed.
Table 4.5 Major types of food intakes at dinners during the three-day
control experiment
Dinning Time Major Food Intake
Day 0 Shrimp, 400 g;

Garoupa, 200 g
Day 1 Beef, 500 g
Day 2 Chicken Meat, 500 g
75
4.3 Results
4.3.1 Effect of age on TMA N-oxidation
The amounts of T M A excreted in the 24-hr urine collection protocols

in different age groups are described in Figure 4.2. Results of T M A O
excreted (as T M A ) in the 24-hr urine collection protocols in different
age groups are described in Figure 4.3. The results indicated that some
age groups showed statistical differences at certain periods for T M A (jf
< 0.05: age 18-30 vs age 41-50 at Periods 1 and 2, age 18-30 vs age
51-64 at Period 3, age 31-40 vs age 41-50 at Period 4; p < 0.01: age
18-30 vs age 41-50 at Period 4) and/or T M A O (as T M A ) Q? < 0.05:
age 18-30 vs 41-50 at Periods 2 and 4, age 18-30 vs age 51-64 at
Period 3, age31-40 vs age 51-64 at Periods 3 and 4; p < 0.001: age31-
40 vs age 41-50 at Period 4) excretion. However, those differences
appeared randomly rather than following a specific trend. The % T M A
excreted as T M A O in urine from spot urine tests and whole day urine
collections were also combined together. The combined data for %
T M A excreted as T M A O in different age groups (Fig. 4.4) showed that
age 31-40 produced the lowest value (96.2 土 2.9 % ; mean 士 s.d.) and
age 41-50 produced the second lowest value (96.4 士 2.5 % ) in the
percentage T M A 7V-oxidation. W h e n Mann-Whitney U-test was
performed, age 31-40 showed statistical difference compared to age
18-30 and age 51-64 {p < 0.05). Also, age 41-50 showed statistical
difference when compared to age 51-64 {p < 0.05).
Spearman rank order correlation was also performed to investigate the

effect of age on the whole day T M A and T M A O (as T M A ) amounts.
It was found that no apparent correlation occurred between age and
T M A (correlation coefficient = 0.136, p > 0.05) or T M A O (as T M A )
(correlation coefficient = 0.012, p > 0.05) amounts (n 二 159). O n the
other hand, Spearman rank order correlation was also performed on the
76
1 n
0.9 T
0.8

’ 1 | ^ B • Age 18-30 P I Age31-40 _ Age41-50 _ Age 51-64
• (n=lll) U (n = 32) ™ (n = 9) “ (n = 7)
Fig. 4.2 Excreted T M A amounts in different periods and in different age groups. Each bar represents mean 士 S.EM, ^p < 0.05. age 18-30 vs
age 41-50： **p < 0.01 age 18-30 vs age 41-50: > < 0.05. age 18-30 vs age 51-64: ^p < 0.05. age31-40 vs ase 41-50.
180 n
160 T
140
B 120
st>
_ 爾 ^

^ ,:3kiir^îlU^ •
Age 18-30
(n=lll)
~ ~ Age 3140
__ (n = 32)
^ ^
^ S
Age 41-50
(n = 9)
•
_
Age 51-64
(n = 7)
Fig. 4.3 Excreted T M A O amounts fas T M A ) in different periods and in different age groups. Each bar represents mean 士 S.E.M. *p < 0.05.
age 18-30 vs age 41-50: **p < 0.001 age31-40 vs age 41-50: > < 0.05. age31-40 vs age 51-64: ^p < 0.05. age 18-30 vs age 51-64.
o * X
2 100 ^ ^
^ T T
H • ,
(/j
c«
节 QC ______
^^ 乂 w^ •
c^
芯
u
o
1^
W 90
^ I
H
、© «s
^ OJ
80j , 1 ,
18-30 31-40 41-50 51-60

(n = 150) (n = 50) (n = 44) (n = 26)
Age
Fig. 4.4 Percentag;e T M A excreted as T M A O in urine (combined data for spot urine test and in whole dav urine collection testl Each bar
represents mean 士 S.D. *p < 0.05: ase31-40 vs ase 18-30 and age 51-64: > < 0.05: age 41-50 vs age 51-64.
combined data of the percentage T M A excreted as T M A O in urine to
the age of volunteers (normality test failed, p < 0.001, n = 270). It was
found that there was no observed correlation between age and the
percentage value (correlation coefficient = -0.08,p > 0.05).
4.3.2 Effect of diet on TMA N-oxidation
According to the major types of food intake, four groups of diets were
classified for analysis: seafood, beef, chicken, and low-TMA diet.
'Low-TMA' diet represents the type of food taken that does not contain
seafood, beef or chicken and no foods with T M A , T M A O , and T M A
precursors. The percentage T M A excreted as T M A O at Periods 1 and
4 in the four food groups are shown in Figs. 4.5 and 4.6, respectively.
Seafood diet showed the highest percentage T M A excreted as T M A O
in urine at both periods (98.2 土 1.9 % at Period 1, 97.6 土 2.5 % at
Period 4; mean 土 s.d.), while the low-TMA diet showed the lowest
values (96.6 士 2.5 % at Period 1, 95.2 土 2.9 % at Period 4). At Period
1, Mann-Whitney U-test showed that the percentage of T M A excreted
as T M A O in the seafood diet group was significantly higher than the
other three diet groups Q? < 0.05). At Period 4, the percentage value of
the seafood diet group only showed significant difference when
compared to the 'Low-TMA' diet Q? < 0.005), but no difference to the
beef and chicken diet groups {p > 0.05). Spearman rank order
correlation analysis was also performed to assess any correlation
between the amount of T M A excreted (TMA), the amount of T M A O
excreted (TMAO), the rate of T M A excreted (TMAy'hr), and rate of
T M A O excreted (TMAO/hr) to the % T M A A^-oxidation (% T M A
excreted as T M A O ) (Table 4.6). All four factors ( T M A , T M A O ,
TMA/hr, TMAO/hr) showed positive correlation to the % TMA
excreted as T M A O . However, only the factors T M A O , and TMAO/hr
showed outstanding positive correlation (correlation coefficient > 0.76,
p < 0.001) to the % T M A excreted as T M A O , and the other two
80
0 100 =F ^ * * *
< ^ **
1 * T T
i I I
Fish Beef Chicken Low-TMA Diet
(n = 44) (n = 9) (n = 22) (n = 21)
i : : i H Food
Fig. 4.5 Percentage T M A excreted as T M A O at period 1 (after 1^ dinner). Shaded bar and error bar represent mean 士 s.d. p values
obtained through Mann-Whitnev U-test where *p < 0.01. **p < 0.005 and ***p < 0.001 compare to seafood diet.
o
^ 100 T
% •氺*
乂^ ^ I I I
Fish Beef Chicken Low-TMA Diet
(n = 47) (n = 8) (n=19) (n=14)
• i : S S Food
Fig. 4.6 Percentage T M A excreted as T M A O at period 4 (after 2"^ dinner). Shaded bar and error bar represent mean 士 s.d. p values
obtained through Mann-Whitney U-test where ***p < 0.005 compare to seafood diet.
factors, T M A and TMA/hr only had limited correlation coefficients (<
0.5).
Multiple linear regression analysis was performed on those factors to

detect any simple positive or negative effect on the % T M A N-
oxidation (Table 4.7). Except for the factors T M A and TMAy^r Q?=
0.1) at Period 1, all other factors had significant negative/positive
effects (p < 0.001) on the % T M A iV-oxidation (% T M A excreted as
T M A O ) . In order to predict the effects of these four food groups on
the % T M A 7V"Oxidation, multiple linear regression analysis was
performed again and an arbitrary number 1 or 0 was assigned on the
subjects (Period 1, n = 98; Period 4, n = 88), to represent the intake of
those foods or not (Tables 4.8 & 4.9). The results showed that seafood
diet produced prominent positive effects on the percentage T M A
excreted as T M A O , amount of T M A excreted, amount of T M A O
excreted, rate of T M A excreted, and rate of T M A O excreted. The p
values were less than 0.005 in all the five regression equations at either
period. Regression coefficients for seafood also produced higher
values when compared to beef diet or chicken diet. Therefore, seafood
appeared to be the major source o f T M A and T M A O .
4.3.3 Effect of control diet on TMA N-oxidation
Five volunteers (A to E) took part in this three-day experiment. The

following graphs describe the amounts of T M A and T M A O excreted
(Fig. 4.7), rates of T M A and T M A O excreted (Fig. 4.8) and the
percentage T M A A^-oxidation (Fig. 4.9), in one typical volunteer (A).
The results showed that the amounts of T M A (0.6 mg) and T M A O
(117.9 m g as T M A ) excreted were suddenly increased after intake of
seafood (Fig. 4.7). The rate of T M A O excreted (15.7 mg/hr as T M A )
showed a sudden increase while the rate of T M A excreted (0.1 mg/hr)
was decreased after intake of seafood (Fig. 4.8). Other foods did not
show significant effect on the amounts of T M A and T M A O excreted
and on the rates of T M A and T M A O excreted. The percentage T M A
83
Table 4.6 Spearman rank order correlation analysis for the factors to the
percentage T M A 7V-oxidation (Each box contained correlation
coefficient and the p value)
Period 1 Period 4
(n = 98) (n = 88)
™ 續 0.212
{p < 0.005) Qp < 0.05)
^ i ^ C.812 0.783
Qr7<0.001) 0?<0.001)
TA/TA/û 0.323 0.187
画虹 (p<0.005) (p = 0.08)
TMAO/hr ( 0.8:、 ( 0.:、
O^<0.001) Qp<0.001)
Table 4.7 Multiple regression analysis for factors on % T M A A^-oxidation (%

T M A excreted as T M A O ) at Periods 1 & 4
Period 1 Period 4
(n = 98) (n = 88)
% T M A excreted as T M A O = % T M A excreted as T M A O =
97.12 — (0.74 X T M A ) + (0.01 x 96.76 - (2.26 x T M A ) + (0.04 *
T M A O ) (r = 0.435) T M A O ) (r = 0.586)
TMA TMAO TMA TMAO
O = 0.1) O<0.001) (^<0.001) (^<0.001)
% T M A excreted as T M A O = % T M A excreted as T M A O 二
97.19 - (6.59 X TMA/hr) + (0.11 96.72 - (9.12 x TMA/hr) + (0.19
X TMAO/hr) (r = 0.413) x TMAO/hr) (r = 0.590)
TMAy^hr TMAO/hr TMA/hr TMAO/hr
(p = 0.1) (p<0.001) O<0.001) (/7<0.001)
84
Table4.8 Multiple linear regression for the effects of foods on the five factors: 1) % T M A excreted as T M A O : 2) Amount of T M A
excreted (TMA): 3) Amount o f T M A O excreted (TMAQ): 4) Rate o f T M A excreted fTMA/hr): 5) Rate of T M A O excreted
(TMAO/hrV at Period L n 二 98. fEach box contains the corresponding x-coefficient andp value)
% TMA excreted as
_ , AA TMA TMAO TMAM TMAOy^r
TMAO
Constant 96.59(p<0.001)* 0.03 0^<0.001)* 14.41(p<0.001)* 0.03 Q? < 0.001)* 1.51(^<0.001)*
00 Seafood 1.64 (p<0.005)* O.O5 0^<O.OOl)* 66.41(p<0.005)* 0.05(p<0.001)* 8.00 (p < 0.005)*
L/1
Beef 0.06(^>0.5) 0.004 Q? > 0.5) -1.91(p>0.5) 0.004(p>0.5) -0.07 Q? > 0.5)
Chicken 0.341(p>0.5) 0.02(p>0.3) 5.54 0^>O.5) 0.02 Q? > 0.3) O.64 0^>O.5)
rvalue 0.351 0.368 0.398 0.376 0.384
N.B. Food 'Low-TMA' was removed in the analysis because it was co-linear.
*Represents statistical significance
Table 4.9 Multiple linear regression for the effects of foods on the five factors: 1、% T M A excreted as T M A O : 2) Amount of T M A
excreted (TMA): 3、Amount o f T M A O excreted (TMAO): 4) Rate o f T M A excreted (TMA/hr): 5) Rate of T M A O excreted
(TMAO/hrl at Period 4. n 二 88. (Each box contains the corresponding x-coefficient and p value)
% TMA excreted as
… … TMA TMAO TMA/hr TMA(Vhr
TMAO
Constant 95.24(><0.001)* 0.17 0^<0.001)* 4.90 0^<0.001)* 0.04 {p < 0.001)* 1.14(p<0.001)*
00 Seafood 2.40 (p<0.005)* 0.35 (p<0.05)* 37.00(p<0.001)* 0.09 (^ < 0.05)* 8.77^^<0.001)*
a^
Beef 1.40(p>0.5) 0.010^>0.5) 7.310^>O.5) 0.03 (p > 0.5) 2.80(p>0.3)
Chicken 2.18(p< 0.05)* 0.11(p>0.3) 9.91{p>03) 0.03(p>0.3) 2.78 0^>O.3)
rvalue 0.355 0.332 0.419 0.297 0.403
N.B. Food 'Low-TMA' was removed in the analysis because it was co-linear.
*Represents statistical significance
140 ~|
A -- 0.6
120 Seafood Intake /-\ Vegetah1e Tntake
Oînner，8:00 PM) / • '1 Oûnch, 1: 5 PM) ^^
: \ Pork Intake __ ^ 5
100 M——A gLunch, 12:45 PM)
I 80 /_^î ^^¢^ A 04 ^
」 : '\ ODinner, 8:30 PM) / \ Chicken Intake g
§ V '\ / \ ODimer, 8:00PM) Q3 蒼
^ ,::t ^r:lc^:,._..y^:
0 •--•-•-••,•• ’• •• -•• • 0
^ H 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I 1 1 r~~T 1 I !• I 1 1 ：— U
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 00 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 00 0
,,« .爾.，.._ ._. , • • •‘霧.，• • _, • , • • •‘ • _, • • _, . , ； . • • • • '• •_. • • ,_. • _. • • •_. • • •_.暴 _, _. • • • •‘ '• • • • ••暑• • . •‘ • , . , . •‘ • • • • • ._ • • • •_. • •, • , , ,_ ,•, , • ._ •_. • • • ••箭• • •. ._ • • , ._ ••, . , ._ ._• , • ._ . • , • -m . ••• • •• ••,
¢^" ^ ¢7 ^ ¢^ bi r- o^ ^ 0 u"i r- ^ ^ 0 ^ 0 LA r^ ^ *- co ui r-o^ »- o ^ o u ^ r^

••“ CU 01 ••“ TT- ^ •«" 1~ OJ 01 T- T- T— *— ••“ Cfl OJ
Time
-_ - • • - - Excreted T M A O (detected as T M A ) ^ Excreted T M A
Fig. 4.7 Excreted amounts o f T M A and T M A O for volunteer A in the three-dav control experiment
25 . — 0.12
m \ _ •• Pork Intake Vegetable Intake ^ 0.1

zU V"" ^
^ \. 、、 OLunch, 12:45 PM) 0-unch, 1:15 PM)
J3 \' \ 感 BeefIntake
^0 \ V ^ • Chicken Intake -- 0.08
d A ^/"^_X^ pinner, 8:30 PM) W 一
S 15 , _ Y y Z — — “ • • /4 Oînner, 8:00 PM) ^
S ：^ , \ T \ -- 0.06 I
？ 10 - ~ ~ " ^ k ^ ^ ^ ^ l ^ ^ - 0 . 0 4 麵
^^ Seafood Intake ^^ \ ^>J^\^
0 ODinner, 8:00 PM) • N^X'^ ^\^^
00 _ 5 ir ^ ^ r _ 0.02
^ H \ V ••• T >
. ^ “ [ _ i . _ . .
v^ I I / I I I / I I I I / I { I f 1 / / / 1 I I 1 ( I I I I I I v/
o o o o o o o o o o o o o o o o o o o o o o o o o o o o oo o
o o o o o o o o o o o o o o o o o o o o o o o o o o o o oo o
..* .*. • . .•. » • ..._._...‘•._. ..‘•._.. ..•._....•.•. ..._._.. . . k .i••.._.星...._.k k. • . , .i. b k. M . ..._.•...._.•....•.•. . . . • . » . .::•••::::•:•: ：； k： :M ：：： :t: M ：：；•； M ：：： :ri:i: ；：；：
M :ri :::_:k:::::_:k:: :::•:•:::::_:«'::•:•:::::_:•«:
cn ^ ¢0 T- « iTi r- o^ "i" ¢^ u~i r- ^ ^ n ^ «• in h- cr^ »- <o u^ r--⑶ ^ n ^ o u i r-
T- oi cy 1" 1" ••“ i~ T- oj oj -r-产 TT- T- ^ oj cy
Time
• - • • - • - TMAO (detected asTMA)/hr « TMA/hr
Fig. 4.8 Excretion rates o f T M A and T M A O for volunteer A in the three-day control experiment
excreted as T M A O was also increased (> 99.5 % ) after intake of
seafood (Fig. 4.9).
The mean excretion rates of T M A and T M A O at different periods in

this three-day experiment were plotted in Figure 4.10. Intake of
seafood not only increased the rate of T M A O excreted at the first
period (9.6 mg/hr.), it also produced a persisting effect over the
following periods (gradually decreased from 4.4 mg/hr at Period 2 to
0.7 mg/hr at Period 7; p < 0.05 when compared Period 1 with Periods 3
to 7). O n the other hand, there was no significant change for the rate
of T M A excretion (p > 0.05). Figure 4.11 also described the mean
percentage of T M A excreted as T M A O at different periods in this
three-day experiment. Similarly, Period 1 (intake of seafood) showed
the highest percentage T M A excreted as T M A O in urine (99.2 士 0.3
% ) during the experiment. O n the other hand, Table 4.10 described the
results obtained from Spearman rank order correlation. All four
factors: amounts of T M A and T M A O excreted, rates of T M A and
T M A O excreted showed positive correlation with the percentage T M A
excreted as T M A O in urine {p values < 0.05).
4.4 Discussion
Based on the data obtained in this study, it was found that some age groups
showed statistically significant differences for T M A or T M A O (as T M A )
excretion at certain periods, when compared to other age groups (see footnotes
in the figures). However, the differences appeared randomly rather than
following a specific trend. For example, the youngest age group (age 18-30)
showed a statistically significant difference in the T M A O amount excretion
(Fig. 4.3) at Period 2 and at Period 4 when compared with the group: aged 41-
50，but this difference did not occur at Period 1 and at Period 3. Moreover, the
pattern of differences in T M A O excretion could not be reproduced in T M A
excretion, i.e. another but random pattern occurred in T M A excretion (Fig.
89
。 ‘ ； : r ~ ^ T " w - ^ " T " > ^
I
^ 90 Porirftitake BeefIntake
c«
73 (Lunch, 12:45 PM) Gînner, 8:30 PM) Chicken Intake
"S Vegetable Intake ODmner, 8:00 PM)
o 85 (Lunch, l:15PM)
^
M
^ o^ Seafood Intake
5 oU
^ P (Dinner, 8:00 PM)
o ^
75
^r 1r
V
70 J：： T
’ ^ T / i I I I I I I 1 " ^ ^ ^ ^ ^ " " " ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ " ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ " ^ " I ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ " " f i I ^ ^ ^ ^ ^ ^ ^ ^ ^
o o o o o o o o o o o o o o o o o oo o o o o o o o o o o oo
o o o o o o o o o o o o o o o o o oo o o o o o oo o o o oo
_ • • • a • • • ‘• • • • • • • •_ i_ It a • • • • • • I • _ •• • • • • •_ t _ t« _ • _fe _ k _ k M g • • •署 _ __ •
<ji »- <r> ^ o yi N 0^ 1" o u^ h- 0^ *i" ¢0 ^ o ui r- ^ ^ c^ ui r^ o^»- co *" 0 bVN>
^ 01 cu p ^ ^ T- ^ cy cu — ^ — i~ i"Cy cy
Time
Fig. 4.9 Percentage T M A A^-oxidation for volunteer A in the three-day control experiment
0.1 — r 18
0.09 — — 16
0.08 X ： .
14
0.07 - . — — 口 - - - ；、 “ ^ ：- 12 一
I 0.06 ^ ：• 10 I
W) 0.05 j<* X J J 8 S
!0.04 • _ _ P ^ \、二-- 一 1 “ … - - j 丨： : : §
g 0.03 ^ ^ 4 g
0.02 r^^>^ =� ^
^ 0.01 ^ " ^ 、 • — =- 2
^ * A： 1" ―^ A
oJ 1 1 象 ^ ^ 象 I ‘ h
Dinner, Fish- Breakfast-Lunch Lunch-Dinner, Dinner, Beef- Breakfast-Lunch Lunch-Dinner, Dinner, Chicken-
Breakfast Beef Breakfast Chicken Breakfast
Oêriodl) (Teriod 2) (?enod 3) (?ehod 4) OPeriod 5) (Teriod 6) (Period7)
Time
__1__ TMA/hr (mg/hr) ^ _ TMAO(as TMAyhr (mgy%)
Fig. 4.10 Average excretion rates of T M A and T M A O forfivevolunteers in three-day control experiment. Each point
represents mean and the standard error of mean, n = 5. For TMAO/hr, *p < 0.05 compared with Periods 3 to 7.
本
100
？ ^ " " • " • " " " " " " " " " " ^ > > > > _ _ 丁
^ 98 ^^^^^-^ f
IH _ T T T
96
1 ^ ^^ . ^ ^ ^ ^ “
«3 ^^> ~<^ *^
坊 94 ± f
""S
u 92 J-
o 丄
><
<o ^
^ I 90
一 88
容
86 "I 1 1 1 1 1 1
Dinner, Fish- Breakfast-Lunch Lunch-Dinner, Beef Dinner, Beef- Breakfast-Lunch Lunch-Dinner, Dinner, Chicken-
Breakfast Breakfast Chicken Breakfast
(Period 1) (Period 2) (Period 3) (Period 4) (Period 5) (Period 6) (Period 7)
Time
Fig. 4.11 Average percentage T M A excreted as T M A O forfivevolunteers in three-day control experiment. Each point
represents mean and the standard error of mean, n =5. *p < 0.05, compare Period 1 to other two periods after
dinners, i.e. Periods 4 & 7.
Table 4.10 Spearman rank order correlation analysis for the factors to
the percentage T M A #-oxidation in three-day control
experiment (Each box contained correlation coefficient and
thep value).
0.389
TMA
0^<O.O5)
0.879
TMAO
(/7<0.001)
0.482
TMA/hr
(p< 0.005)
0.900
TMAO/hr
(p<0.001)
93
4.2). One possible reason was the over-representation of the youngest age
group, since most volunteers were belonged to this age group (〜70 % ) and
only 10 % of the volunteers were at age 41 or above. Also, the effect of age
on T M A and T M A O excretion might be masked by other factors, e.g. dietary
factors or other unknown factors, and the true patterns may then not be seen in
this study. Therefore, a study with controlled diet should be carried out for the
investigation of the effect of age on T M A A^-oxidation.
Since Kruskal-Wallis One W a y A N O V A on Ranks was performed (Chapter

3), the data showed no difference for spot urine collections and for whole day
urine collections. The sets of data were combined and the percentages of
T M A excreted as T M A O in different age groups are plotted in Figure 4.4. It
produced interesting results that the two middle-age groups (age 31-40, and
age 41-50) produced the lowest percentage of T M A 7V-oxidation when
compared to the youngest and oldest age groups (p < 0.05, refer to the footnote
in Fig. 4.4). A U-shaped change for the % T M A oxidation versus age
occurred. O n the other hand, as described at the result section, the Spearman
rank correlation showed no correlation between age and the percentage of
T M A 7V"Oxidation (Section 4.3.1). This showed that the effect of age on the
ability of T M A 7V>oxidation might not increase or reduce in a linear or
logarithmic format. Since T M A is the substrate for flavin-containing
monooxygenases (FMOs), especially for liver F M O , it is reasonable to
extrapolate the data to human F M O activity. Therefore, the effect of age on
the F M O activity might not reduce or increase linearly or logarithmically.
However, other factors, possibly dietary factors, or other unknown factors,
may be present to influence the T M A TV-oxidation in the two middle-aged
groups.
M a n y studies have shown that the ability for drug metabolism by Phase I
enzymes (oxidation) decreased with aging (Table 4.1). In this study, there was
no direct relation of F M O activity as shown by T M A A/-oxidation with age.
The results were contradictory to the results obtained in animal studies. Many
studies in animals showed that aging can decrease the activity of enzymes (e.g.
cytochrome P450s) and in drug metabolism by Phase I oxidation (Tables 4.1
94
& 4.2). However, it was not surprising that when comparing the results to the
data obtained in previous studies of human enzymes including cytochrome
P450s (Section 4.1, paragraph 2). M a n y studies showed that drug
metabolizing enzymes including cytochrome P450s (e.g. C Y P 3 A 2 , C Y P 2 C 7 )
did not alter with advancing age. The decline in drug clearance might be
caused by a fall in liver volume, blood flow, renal insufficiency, etc. (Table
4.11) (Vestal and Dawson, 1985; W y n n e et al., 1988). Therefore, the
extrapolation of animal data to human might not be applicable. Although the
current study showed that the liver F M O activity (using T M A #-oxidation as
the probe) may change with age in a U-shape, more extensive study is needed
because only limited studies have been done before on the effect of age on
F M O activity and the numbers of subjects in some age groups were limited.
In conclusion, there was no direct correlation between age and F M O activity

in the Chinese population studied. The reduced ability for drug metabolism
might not be related to enzyme activity or affinity directly, but may be related
to other factors including changes in liver blood flow, mass, renal
insufficiency, etc. (Table 4.11). To date, there has been relatively little
information and work carried out on F M O s . Therefore, other types of study,
such as measurement of liver F M O activity and its amount in liver biopsy,
should be carried out. A longitudinal study may be carried out to look at the
change of the F M O ability for T M A iV-oxidation throughout the lifetime of the
volunteers, although this type of study would be time-consuming and difficult
to perform.
In the study of diet on the T M A iV-oxidation and in analyzing the data

obtained in 24-hour urine collection protocols, Period 1 and Period 4 were
studied since urine collections were started after the dinners. Although Period
1 was started four hours after the first dinner but Period 4 was started just after
the second dinner (Table 4.3)，results were useful to show the prolonged
effects and excretion characteristics after intake of foods (and intakes of T M A ,
T M A O , and the precursors). O n the other hand, Period 4 showed the
immediate effects and excretion characteristics of the above factors. It was
clear that intake of seafood produced the highest percentage of T M A N-
95
Table 4.11 Factors affecting drug disposition in elderly patients (modified
from Vestal and Dawson, 1985)
Cause Factors
1) Achlorhydria
2) Cancer
3) Congestive heart failure
4) Dehydration
5) Diarrhea
6) Edema or ascites
7) Fever
Disease 8) Hepatic failure
9) Hepatic insufficiency
10) Hypovolemia
11) Malabsorption syndromes
12) Mahiutrition
13) Pancreatitis
14) Postgastrectomy
15) Renal Failure
16) Renal insufficiency
17) Thyroid disease
18) Viral infection or immunization
1) Decreased absorptive surface

2) Decreased cardiac output
3) Decreased gastrointestinal motihty
4) Decreased glomerular filtration rate
5) Decreased lean body mass
6) Decreased liver blood flow
7) Decreased liver mass
Physiological change 8) Decreased renal blood flow
9) Decreased serum albumin
10) Decreased splanchnic blood flow
11) Decreased total body water
12) Decreased tubular secretion
13) Increased body fat
14) Increased gastric pH
15) Increased x j -acid glycoprotein
1) Antacids
2) Anticholinergics
3) Cholestyramine
Therapeutic and 4) Dietary composition
environmental factor 5) Drug interactions
6) Food or meals
7) Insecticides
8) Protein-binding displacement
9) Smoking
96
oxidation in urine in both periods (98.2 % at Period 1; 97.6 % at Period 4).
The results obtained in Figures 4.5 and 4.6 showed some differences, although
the low-TMA diet showed the lowest percentage of T M A A^-oxidation, and
beef and chicken diets produced intermediate values. In Period 1, a statistical
difference (Mann-Whitney U-test) was observed when the seafood diet was
compared to all other three types of foods. Thus, high intake of T M A and
T M A O through seafood consumption increased the percentage of T M A TV-
oxidation and had persisting effects on this oxidation and excretion several
hours after the diet intake. High amounts of T M A and T M A O from seafood
took time to inter-convert in the gut (by bacteria) and to oxidize by liver
F M O s . O n the other hand, it has been suggested that under heavy T M A
exposure conditions (consumption of seafood) and the consumption of
vegetables containing inhibitors of the T M A oxidase, features of the fish-
odour syndrome may occur in carriers on a sporadic basis (Ayesh and Smith,
1992). Therefore, apart from oral T M A dosage, intake of seafood may be a
good alternative to investigate the deficiency of T M A A^-oxidation, the activity
of F M O s on TMA 7V-oxidation, and the genetic condition of
trimethylaminuria. It may be useful for the subjects who had mild impaired
ability in T M A A^-oxidation, i.e. secondary trimethylaminuria by genetic
problem, disease, and other environmental factors. In Period 4, however, no
short term statistical difference was obtained for seafood, beef, and chicken
intakes. It could be explained because urine samples at Period 4 were
collected just after the intake of food. Beef and chicken diets, although not
containing high amounts of T M A and T M A O as in seafood, contain certain
amounts of carnitine and choline. These precursors were converted to T M A O
by gut bacteria and excreted through the urine. Therefore, the percentages of
T M A 7V-oxidation were similar in seafood, beef, and chicken diets at the four
hour periods after taking those foods, because the initial intakes of T M A ,
T M A O or the precursors stimulated the initial activity of F M O . After these
four hours, the excreted amounts of T M A O from beef and chicken were
expected to reduce since most of the carnitine and choline were already
depleted. At that moment, only the seafood diet (containing very high
amounts of T M A and T M A O ) still produced high T M A O excretion in urine
97
and a high percentage of T M A A^-oxidation. This explanation was supported
by the results obtained in Period 1 (Fig. 4.5).
In Spearman rank order correlations for the amounts of T M A ( T M A ) and

T M A O ( T M A O ) excreted, rates of T M A (TMAy^hr) and T M A O (TMAO/hr)
excreted, to the % T M A iV-oxidation (% T M A excreted as T M A O ) in both
periods, as described in Section 4.3.2, all four factors ( T M A , T M A O , T M A y % ,
TMAO/hr) showed a positive correlation to the % T M A excreted as T M A O ,
but only the factors T M A O and TMAO/hr showed outstanding positive
correlation (correlation coefficient > 0.76, p < 0.001) to the % T M A excreted
as T M A O . Therefore these two factors should be useful to monitor the
percentage change of T M A TV-oxidation in urine. In addition, multiple linear
regression analysis showed that those factors (except T M A and TMA/hr at
Period \,p = 0.1) had significant positive effects on the percentage of T M A TV-
oxidation {p < 0.001) (Table 4.7). The equations showed reasonable results
that a decreased amount of T M A excreted and a decreased rate of T M A
excreted at both periods should increase the percentage of T M A A^-oxidation.
The reason was most of the T M A was converted to T M A O , and therefore the
percentage value o f T M A excreted as T M A O in urine was increased (Equation
2.3). The effects of the four food groups on the percentage of T M A N-
oxidation were shown in Tables 4.8 & 4.9, using multiple linear regression
analysis. According to the results shown in Tables 4.8 and 4.9, seafood diet
produced prominent positive effects on the percentage T M A excreted as
T M A O , amount of T M A excreted, amount of T M A O excreted, rate of T M A
excreted, and rate of T M A O excreted {p < 0.005). Also, the regression
coefficients for seafood also produced the highest values when compared to
beef diet or chicken diet. It can be concluded that seafood was the major
source of T M A and T M A O and this result also suggested that intake of
seafood may be a good source to investigate the deficiency of T M A N-
oxidation, the activity of F M O s on T M A , and the genetic condition of
trimethylaminuria.
In the effect of control diet on T M A #-oxidation, the amounts of T M A and

T M A O excreted were suddenly increased after the intake of seafood (Fig. 4.7).
98
The rate of T M A O excreted was increased but the rate of T M A excreted was
decreased (Fig. 4.8). This feature was not reproduced after the intakes of other
foods, which further support the hypothesis that F M O activity could be
stimulated by high amounts of T M A and T M A O intake (from seafood).
Therefore, the highest percentage of T M A excreted as T M A O in urine
occurred immediately after the intake of seafood (Fig. 4.9). The raw data was
not shown for the other four volunteers because the times for urination were
different for the volunteers. However, all the volunteers showed similar
results for T M A TV-oxidation. The mean values for the rates of T M A and
T M A O excreted (Fig. 4.10) and the percentage of T M A A^-oxidation (Fig.
4.11) further confirmed that intake of seafood increased the rate of T M A O
excretion and the percentage of T M A TV-oxidation (see data at Period 1).
Also, a high intake of T M A O produced a persisting effect on the rate of
T M A O excretion, and a decreased rate of T M A excretion. The results of
Spearman rank order correlation in this control experiment also showed that
the amount of T M A O excreted and the rate of T M A O excreted had
outstanding positive effects on the percentage of T M A excreted as T M A O in
urine (correlation coefficients > 0.87, p values < 0.001).
From the results of the effects of diet and control diet on T M A A/-oxidation, it
can be concluded that seafood is the major source o f T M A and T M A O . Intake
ofseafood (as high intakes o f T M A and T M A O ) can stimulate the initial F M O
activity and lead to a higher percentage of T M A 7V-oxidation. It also produced
persisting effects on the rate of T M A O excretion and the percentage of T M A
excreted as T M A O in urine. Therefore, it may be used as a medium, instead
of direct T M A dosage, to investigate the subjects with impaired ability in
T M A iV-oxidation, i.e. secondary trimethylaminuria. Moreover, there have
been few studies conducted on the effect of diet on enzyme activity and drug
metabolism together with a dietary survey. Only one similar study has been
carried out before to determine the relationship of oxidative drug metabolism
to food preference (Britto et al, 1991). In this study, the poor metabolizers of
dextromethorphan showed a diminished stated preference for cauliflower and
coconut, and poor metabolizers of mephenytoin showed a diminished stated
preference for spinach and cabbage. Moreover, the 'unfavorable responses' to
99
those foods might be due to the presence of chemical components in those
foods that caused inhibition to the corresponding cytochrome P450s.
Therefore, the current study may be an additional source to show the effect of
normal dietary intake on enzyme activity.
According to the data obtained in this study, there was no direct correlation
between age and F M O activity on the percentage T M A iV-oxidation, but the
dietary study (dietary survey and controlled diet) showed that intake of
seafood increased the F M O activity on the percentage T M A iV-oxidation.
Thus, intake of seafood may be useful to investigate the subjects with
impaired ability in T M A TV-oxidation.
100
Chapter 5
Effect of Disease on
Trimethylamine A^-Oxidation
5.1 Introduction 102

5.3 Results 109
5.4 Discussion 116
101
5.1 Introduction
The effect of disease on the pharmacodynamics and pharmacokinetics of drugs

has been well documented. Diseases such as chronic liver disease and renal
disease affect drug metabolism and lead to changes in the efficacy and
duration of action of many drugs. Chronic liver disease has a major effect on
drug metabolism in patients. Figure 5.1 describes the free exchange of fluid
and substrates between the sinusoidal lumen and the space of Disse in a
healthy liver. The mechanisms of impairment of hepatic drug elimination in
chronic liver disease can be described by four different theories: the sick cell
theory; the intact hepatocyte theory; the impaired drug uptake theory; and
the oxygen limitation theory. Dosage requirements for drugs of both high and
low hepatic clearance may be affected (Morgan and McLean, 1995). The sick
cell theory envisaged a reduction in the content and activity of the hepatic
drug metabolizing enzymes while blood flow is maintained. It was supported
by the studies of a decrease in the hepatic extraction ratio of indocyanine
green by functional hepatocytes (Perlik et al., 1992), and a correlation between
systemic and intrinsic hepatic clearances of pethidine and morphine in
cirrhotic rats (Callaghan et al., 1993). The second theory is the intact
hepatocyte theory and it shows a reduced mass of cells, which function
relatively normally and are normally perfused. It assumes the same fractional
reduction in hepatocellular volume, with which the hepatic clearance of low
clearance drugs is proportional, as the fractional reduction in functional
hepatic blood flow rate (i.e. flow through remaining functional tissue), with
which the hepatic clearance of high clearance drugs is proportional. It was
proposed by the measurement of the liver volume in patients with cirrhosis by
computerized topography (Kwasaki et al., 1992) and by liver biopsies from
patients with cirrhosis (Meyer et al., 1991)，which showed unchanged
clearance per hepatocyte in cirrhosis. The third theory is the impaired drug
uptake theory. It proposes that the most significant feature of cirrhosis is the
process of sinusoidal capillarization. Capillarization involves the formation of
continuous filtration and diffusion barriers, which hinder bidirectional
102
Sinusoid
j ^ ^ y Sinusoid epithelium
c r g - ^ ^ N ^ c ^ = Z 3 C Z = = > / < r = Z T Z ' ^

\ / ^ Space ofDisse
广 � ^j >^^^ J^ ^ ^^^^^^ Hepatic substrate
• 丨 •
^"N^7^
Hepatocytes
Fig. 5.1 The free exchange offluidand substrate(s) between the sinusoidal lumen and the space ofDisse in a healthy liver.
macromolecular exchange (Martinez-Hernandez and Hernandez, 1991). The
barriers are caused from loss of fenestration of the sinusoidal endothelium,
development of basal laminae and deposition of complex macromolecules in
the space of Disse. The fourth theory is the oxygen limitation theory due to
the impaired uptake of oxygen across the capillarized endothelium (McLean
and Morgan, 1991; Morgan and McLean, 1991). This theory was supported
by the increased theophylline clearance in cirrhotic rats, but not in healthy rat,
with inhaled oxygen supplementation (Hickey et al., 1995). Uptake of
oxygen becomes the rate limiting step，rather than the drug uptake in impaired
drug uptake theory. As a result, the impaired drug uptake theory is generally
accepted to describe the impairment of hepatic drug elimination in chronic
liver disease. O n the other hand, studies also showed that moderate degrees of
liver impairment can impair the renal drug clearance (McLean and Morgan,
1991; Westphal and Brogard, 1993), and the renal elimination of drug
metabolites was impaired by decreased renal blood flow and glomerular
filtration. Renal disease is another major disease which interferes with the
elimination of many drugs and it also affects drug metabolism. It not only
reduces the renal clearance of drugs but also affects the nonrenal clearance of
drugs (Farrell, 1987; Gibson, 1986, Touchette and Slaughter, 1991). Studies
showed that multiple microsomal enzyme activities in animals were decreased
with acute and chronic renal failure (Anders, 1980; Hogan et al, 1979;
Patterson and Cohn, 1984; Terner et al., 1978). The decrease in total
metabolic activity in intact animals with renal insufficiency may be due to the
impairment of hepatic metabolic activity as well as the loss of renal
metabolism. Therefore, drug clearance will be altered due to the effects on
glomerular blood flow and filtration, tubular secretion, reabsorption and renal
parenchymal mass. It leads to the need for alternations in dosage regimens to
optimize therapeutic outcome and minimize the risk of toxicity.
In view of the impairment of drug metabolism by liver and renal disease,

decreased liver function including decreased liver F M O activity (sick cell
theory) or the formation of continuous filtration and diffusion barriers
(impaired drug uptake theory) may retard the TV-oxidation of T M A to T M A O
in the liver (Ayesh and Smith, 1992; Marks et al., 1978; SimenhofF et al.,
104
1977; Wills and Savory, 1981; Wranne, 1956). One piece of evidence was
that liver cirrhosis increased the urinary excretion of T M A because of the
impaired hepatocellularfiinctionor the presence of portosystemic venous
shunts, which might increase the systemic availability of T M A due to the
decreased first-pass metabolism (Marks et al., 1978). Moreover, increased
blood levels o f T M A and dimethylamine have been suggested to be one of the
etiological factors in the development of hepatic coma in cirrhotic patients.
Another piece of evidence was the increased excretion of T M A in a 3-year-old
female with congenital intrahepatic portocaval shunt (Fernandez et al, 1997).
This also confirmed that such a venous shunt is another factor to increase the
systemic availability of T M A . Other diseases, such as diabetes may also alter
the F M O activity and lead to a change in the percentage of T M A 7V-oxidation.
Rouer et al (1988) reported that liver microsomal F M O activity assessed by
thiobenzamide metabolism was increased by two-fold in streptozotocin-
induced diabetic (insulin deficient) rats and mice and, to a lesser degree in
congenital insulin resistant Ob/Ob mice. It might therefore be expected that
the percentage of T M A TV-oxidation in diabetic patients should have a
different pattern when compared to normal subjects. According to the
previous findings of T M A iV-oxidation and in F M O activity, it was possible to
assume that patients with retarded liver function would show a low percentage
value of T M A 7V-oxidation. Therefore, the present study evaluated patients
with chronic liver disease and retarded liver function, and patients with
diabetes, on the percentage of T M A 7V-oxidation, which may result from a
change in F M O activity. The patterns of these diseases on T M A TV-oxidation
were also compared with other diseases in patients with normal liver function
(hyperlipidaemia treated with statins, hypertension treated with anti-
hypertensive drugs). Patients with hyperlipidaemia were chosen in this study
because hyperlipidaemia is a disease associated with high plasma triglyceride
and cholesterol levels caused by high intake of fat and cholesterol from the
diet. The metabolism of fat involves a rather complicated pathway including
the esterification of fatty acids to triglycerides in the liver. Triglycerides in
very low density lipoprotein (VLDL) particles are also metabolized by
lipoprotein lipase resulting transiently in intermediate density lipoprotein
(HDL) particles which are taken up by the liver or converted by hepatic lipase
105
into low density lipoprotein (LDL) particles. Some lipid-lowering drugs, e.g.
most of the lipid-soluble statins, that are used for the treatment of
hyperlcholesterolaemia are metabolized by C Y P 3 A 4 (Stern et al., 1997). The
drug interactions that have been described are mainly with other agents that
inhibit this isoenzyme such as the macrolide antibiotics, the azole antifungals
and cyclosporine. These interfere with statin metabolism and may increase
the risk of myopathy. Although statins do not appear to have a major
inhibitory effect on this enzyme, it is not unreasonable to consider that statins
might affect other metabolic pathways and interfere with the function of other
enzymes. Therefore, it is reasonable to examine the effect of hyperlipidaemia
itself on the percentage of T M A 7V-oxidation, and additionally the effect of
statin treatments. Moreover, one type of the statin: atorvastatin (Fig. 5.2 a),
which is the most potent lipid-lowering drug was being evaluated in a clinical
study at the time of this work. It can reduce the blood cholesterol by almost
50 % . Currently, there is no information about the effects of this drug on
F M O s and on T M A 7V-oxidation. It was therefore considered opportune to
investigate the effects of this drug on T M A A^-oxidation as the patient cohort
was already available. Another drug, hydroxychloroquine ( H C Q ) (Fig. 5.2 b),
which is one of the antimalarial agents, is commonly used in the treatment of
systemic lupus erthematosus (SLE) because of its anti-inflammatory activity
and steroid sparing effect (Hodis et aL, 1993). SLE is a chronic multi-organ
systemic disease thought to be of autoimmune origin that affects young men
and w o m e n in their twenties and thirties. H C Q is also used for the treatment
of rheumatoid arthritis and it has been suggested that it might be useful as a
treatment for hyperlipidaemia by inhibiting cholesterol synthesis through its
effect on 2,3-oxidosqualene-lanosterol cyclase. O n the other hand, there is no
information about its effect on F M O activity and T M A 7V-oxidation. Similar
to atorvastatin, the use of this drug might cause other unexpected side effects
in other metabolic pathways, e.g. T M A iV-oxidation. A study had also been
initiated to examine the effects of H C Q on lipid levels in a group of patients
with SLE and the opportunity was taken to obtain urine samples from these
subjects to study whether H C Q had any effect on T M A A^-oxidation.
106
HsC ,CH3
\ / OH OH 0
CH „ • •
^ ^ CH .CH 人
/=\ i 人八 / V \ 八 -
ly V N H C ^ J n, H, H, • Ca2+
V^ V=<^ . 3H.0
—
0 认」 2
(a) atorvastatin calcium
C 1 X ^ ^ ^ ^ N 广
^ = = : : : / ^ ^ ^ f ^ 、 N E t 32SO4
™^^^^^^
H^ Me
(b) hydroxychloroquine sulphate
Fig. 5.2 Chemical structures of a) atorvastatin calcium, and b)

hydroxychloroquine sulphate
107
In the studies in this chapter, T M A 7V-oxidation was measured in groups of
Chinese patients with various diseases (chronic liver disease, diabetes,
hyperlipidaemia with statin treatments, and hypertension), and in two study
cohorts to determine whether certain medications (atorvastatin,
hydroxychloroquine) would alter the F M O activity by the measurement of the
degree of T M A iV>oxidation. Although patients with hypertension have not
been reported to have any change in T M A A^-oxidation, they have not been
studied generally and they were included in this clinical screening study
because a large number ofhypertension cases are seen in the hospital and they
were readily available to study.
In the study of the effects of diseases and medications on T M A 7V"Oxidation,

urine samples were collected from male and female patients at the Prince of
Wales Hospital, Shatin, N T , Hong Kong. Spot urine samples or 24 hr urine
samples were collected for analysis. The sample collection procedures were
already described in Section 3.2, Chapter 3. As described in Section 5.1, four
patient groups were studied: chronic liver disease (cirrhosis), diabetes,
hypertension, hyperlipidaemia. Patients treated with atorvastatin or
hydroxychloroquine were also investigated. Patients treated with atorvastatin
had severe resistant hyperlipidaemia which had not responded adequately to
previous therapy. They had been on treatment for 16 weeks at the time the
first urine samples were collected. The dose of atorvastatin had been
increased from 10 m g to 20, 40 then 80 m g in most patients at intervals of 4
weeks. The second urine samples were collected in these patients after they
had been off treatment for 4 weeks when plasma lipid levels had returned to
pretreatment levels. The effect of hydroxychloroquine treatment was studied
in a cohort of patients with SLE who were being studied to examine the effects
of this drug on the lipid profile. The patients with S L E were attending the
rheumatology clinic at the Prince of Wales Hospital. The diagnosis of SLE
had been made according to the American Association of Rheumatology
criteria. None of the patients had significant renal disease as determined by
plasma creatinine levels. The patients attended in the morning fasting for
108
blood sampling. Spot urine collections were made at these visits. The results
for T M A A^-oxidation in those subjects w h o were maintained on
hydroxychloroquine were compared with control subjects w h o were not taking
hydroxychloroquine but were maintained on anti-inflammatory doses of
corticosteroids (prednisone or prednisolone).
Data were analyzed using the SigmaStat software (Jandel Scientific).

Different groups were compared by analysis of variance, unpaired t-tests or
non-parametric tests if the data were not normally distributed. Results for
patients treated with atorvastatin were compared with the results after stopping
treatment by paired t-tests.
5.3 Results
In the study ofthe effect of disease on T M A A^-oxidation, the percentages of

TMA7V-oxidation for four different patient groups are described in Figure 5.3.
Hyperlipidaemic patients treated with statins (n = 19; male = 9, female = 10),
were of age from 23 to 70 years (46.7 士 13.3; mean 土 s.d.). They showed the
highest percentage o f T M A excreted as T M A O in urine (96.3 士 2.8 % ; mean 士
s.d.). Hypertensive patients (96.0 士 2.8 % ; n = 26; male = 11, female 二 15),
age range from31-64 years (49.5 士 9.7), and diabetic patients (94.3 土 6.0 % ; n
= 2 2 ; male 二 13，female = 9)，age range from 24 to 81 years (55.5 士 17.1)
produced comparable percentage values to the hyperlipidaemia patients and
there was no significant difference between these groups. However, patients
with chronic liver disease (n = 20; male = 16, female = 4), age range from 37
to 76 years (60.0 士 11.8) showed retarded ability in T M A A^-oxidation. This
group produced the lowest percentage value (85.8 士 5.3 % ) when compared to
the other three patient groups and the data obtained from the normal subjects
(Chapter 3) (p < 0.001, Mann-Whitney U-test). However, the other three
patient groups (hypertension, diabetes, and hyperlipidaemia) did not show any
statistical significance when compared with each other and to the normal
subjects (Chapter 3) (p > 0.05, Mann-Whitney U-test). Moreover, the
percentage of T M A A^-oxidation for patients with diabetes was not increased
109
100 T “
i 卯—11 “^^“H _
云t = H = _ _
：丨• I • I厂丨I
Hypertension Diabetes Hyperlipidaemia Chronic Liver Disease
n = 26 n = 22 n=19 n = 20
Fig. 5.3 Percentage T M A A^-oxidation in four patient groups fEach bar represents mean and S.D. and * denotesp < 0.001).
markedly when compared to normal subjects. Scattered plots for the
percentage of T M A excreted as T M A O in urine in the four patient groups are
shown in Figure 5.4. O n the other hand, the details for the 20 patients with
chronic liver disease are listed in Table 5.1. Some of these patients were also
diabetic on oral hypoglycaemic drugs or insulin. Most subjects w h o excreted
less than 90 % of T M A as T M A O showed abnormal high levels of total
bilirubin (> 15 îmol 1"^) and alanine aminotransferase (ALT) (> 58 RJ 1'^).
The mean values of total bilirubin and alanine aminotransferase in these 20
patients were 65.8 士 67.5 îmol 1"^ and 58.0 士 44.5 RJ 1"^ respectively (mean 士
s.d.).
Urine samples from 19 patients (male = 5, female = 14), age range from 32 to
69 years (50.8 士 13.0, mean 士 s.d.), treated with atorvastatin were examined
and compared with the samples from the same patients 4 weeks after stopping
treatment. There was no significant effect on the percentage of T M A N-
oxidation (treated with atorvastatin: 94.3 士 3.4 % , ofF drug: 95.3 士 2.0 % ;
mean 士 s.d., n = \9,p > 0.05) (Fig. 5.5).
Urine samples from SLE patients treated (age range from 21 to 52 years, 38.6
士 7.6, mean 士 s.d.; male = 1, female = 45; n = 46) or without treatment (age
ranged from 18 to 66, 40.5 土 11.4，mean 土 s.d.; male = 1, female = 23; n 二 24)
with H C Q were analyzed. The treatment with H C Q for patients with SLE also
had no significant effect on the mean percentage of T M A #-oxidation (with
H C Q : 94.2±5.2o/o,mean±s.d.,n = 46;withoutHCQ: 94.5±2.5o/o,n = 24.
p > 0.05) (Fig. 5.6). However, three female patients, w h o were on treatment
with H C Q showed a relatively low percentage of T M A A^-oxidation, the
percentage values were 78.3 % , 82.5 % , and 88.4 % respectively. These
patients will be studied further to determine whether these findings are
consistent and if so to try to determine the cause.
111
00 I I I •
0 95 I • ： .
_ • ： • «
H 90 - • 1
Qf) • 5
^ X
% 85 - • • •
0>
^
- 运
f^ s 80 -
1 •
H 75 -
^ • •
70 - J
65 ^ ‘ ‘ ‘ ‘
Hypertension Diabetes Hyperlipidaemia Chronic Liver
(n = 26) (n = 22) (n = 19) Disease (n = 20)
Fig. 5.4 Scattered plots for the percentage o f T M A excreted as T M A O in urine in the four patient groups.
Table 5.1 Details for 20 patients with chronic liver disease
r “ |Albumin|TotalBilimbin| AUcaline |ALT*|%mAExcreted

No. Sex Age Diagnosis Dmg Treatment (g y^) (nmol 1'') Phosphatase (K； 1'^) (KJ 1'^) as TMAO (%)
— W ~ ~ — Normalsubject - 36-58 <15 40-100 < 58 > 90
~ \ " " • M ~ ~ W Hepatic encephalopathy, puhnonary tuberculosis, hepatitis B, Actrapid, insulin, vitamin K, aldactone, lactulose, 15 117 149 53 68#
alcoholic liver cirrhosis. colchicine, minidiab
飞~~M~~^ Liver cirrhosis, chronic subdural haematoma Lactulose, thiamine, zinacef ^ ^ ^ 1^ ™
~1~~M~~56 Hepatic encephalopathy, liver cirrhosis Lactulose, intravenous fluid ^ ^ ^ ™ 謂
~ 4 ~ ~ M ~ ~ ^ Post hepatitis B & alcoholic liver cirrhosis, abdominal distension ML 25 55 166 158 94
& anaemia
~S F ^ Liver cirrhosis, spontaneous bacterial peritonitis, non-insuHn Aldactone, vitaminK, gliclazide 20 101 155 71 86#
dependent diabetes meUitus
~6 W 37 Liver cirrhosis, ascites. Slow K, zinacef, aldactone, inderal, lasix, lactulose 23 83 151 36 %
~"~M 40 Liver cirrhosis “ ML 25— 28 90 58 93
~8 K ^ ^ Liver cirrhosis, portal hypertension, chronic hepatitis B Sprionolactone, lasix, lactulose, vitaminK, albumin 19 10 ^ ^ ^
~~9 F~~48 Liver cirrhosis, pleural effiision Aldactone, vitaminK, sucralfate, inderal, zinnat 23 45 130 28 ^
10 M " ^ Liver cirrhosis, end-stage renal failure, tuberculosis Isoniazid, mithramycin, pyrazinamide, ethambutol 20 7 398 21 ^
二 ~J\ F ^ Child C hepatitis C cirrhosis, ascites Acetomenaphthone vitamin K, lasix, aldactone, 33 59 102 40 89#
lactulose, zinacef
12 M ~48 Hepatitis B cirrhosis, gross ascites Cefuroxime, vitamin K, lactulose 25 299 465 116 86#
13 M 76 Alcoholic liver cirrhosis, diabetes melHtus on diet, chronic Prednisolone, voknax, ventolin preparation 30 15 199 17 94
obstructive airway disease
l4~~W 百 Liver cirrhosis Aldactone, lasix ^ M Vfl ^ W
15 M 63 Sepsis, chest infection, benign essential hypertension, alcoholic Adalat preparation 25 31 116 49 91
liver cirrhosis, acute renal failure, non-rheumatic mitral
regurgitation
16 M 53 Alcoholic liver disease, diabetes meUitus, hypertension Lactulose, aldactone, vitamin K, daonil, augmentin 15 22 356 32 68#
17 M 49 Child C alcoholic cirrhosis Lactulose, folate, cefuroxine, thiamine, aldactone, 18 95 115 53 73#
vitamin B Co, vitamin A
18 M 75 Litrahepatic biliary obstruction Vitamin K, ampiciUin, flagyl, cefuroxine, thiamine, 23 116 1059 169 94
digoxin
19 M 67 Cirrhosis, macrocytic anaemia ML 34 4 91 36 92
20 F 74 Acute respiratory infection, liver cirrhosis Vitamin K, aldactone, lasix, lactulose 22 119 142 51 88#
*ALT: alanine aminotransferase, # indicates < 90 % ofTMA excreted as TMAO in urine. Subject no. 0 is created for comparison.
140
98
96
94 nnMnmnmnmnmnmnmnnnnnnnnnnnnmi
92 I II I
I 90 III I II
_ • •
82 II II I
I I I I
80 |隱^ l______^
I
Treated with atorvastatin Off atorvastatin
Fig. 5.5 Percentage T M A 7V-oxidation for patients treated with atorvastatin and at off
drug. Each bar represents mean 士 s. d. (n 二 19. p > 0.05)
114
100
*ym
95
I
i—
90
.
‘85 — 一
丨
11丨
1丨
丨
11丨
丨丨丨
1丨
丨丨丨
丨
1丨丨
丨丨丨
1丨
丨丨丨
丨丨
1丨
丨丨丨
丨
1丨丨
丨丨丨
丨
1丨丨
丨丨
1丨
1丨
11 丨：:::::::::::::::::::::::::::::::::::::::::::::::::::::::::1
Treated with HCQ no HCQ

Fig. 5.6 Effect of hydroxychloroquine on the percentage T M A A^-oxidation. Each
bar represents mean 士 s.d. (n =46 for H C O , n = 24 for no H C Q : p > 0.05).
Three patients showed relatively low percentage T M A excreted as T M A O
in urine when treated with H C Q , the values being 78.3 % , 82.5 % , and
88.4 %•
115
5.4 Discussion
Patients with chronic liver disease showed an impaired ability of T M A N-

oxidation in this study Q? < 0.001), while the other patient groups appeared
normal (p > 0.05), when compared to normal volunteers. Although the
patients with chronic cirrhosis had different extents of liver disease, most of
them had severe liver damage. The damage produced elevated levels of
bilirubin and alanine aminotransferase (ALT) (see Section 5.3, paragraph 1).
Table 5.1 showed the clinical diagnoses, liver function tests and drug
treatment information for this patient group. More than half the number of
patients showed elevations of total bilirubin and alkaline phosphatase.
Bilirubin comes mainly from red cell precursors within the bone marrow and
from the breakdown of myoglobin. It is then removed from the body by the
liver. Hyperbilirubinaemia is the result of severe hepatocellular damage
during hepatitis or cirrhosis. Raised levels of A L T , which is a cytosolic
enzyme, also known as serum glutamic pyruvic transaminase (SGPT), in the
chronic liver disease patient suggests liver cell damage resulting in this
enzyme leaking into the blood stream. The reasons for impaired T M A N-
oxidation in cirrhotic liver disease may be numerous pathophysiologic
changes in the liver, similar to the impaired uptake theory described in Section
5.1. The development of cirrhosis starts with initial hepatocellular damage
that produces inflammation; dead or necrotic cells are removed by
phagocytosis. Then, fibroblasts secret collagen to help maintain the integrity
of the liver architecture, while the damaged or dead cells are repaired or
replaced. Collagen fibrils accumulate in the sinusoidal space, including the
space ofDisse, and produced bands of connective scar tissue characteristic of
cirrhosis. Also, a basement membrane lacking in microvilli is produced at the
sinusoidal surface of the hepatocyte. The collagen barrier between hepatocyte
and sinusoid interferes with the exchange of nutrients, drugs, and the
metabolites, between the blood and hepatocytes. Liver nodules might occur
due to the regeneration ofhepatocytes and the compensation for hepatocellular
damage. Therefore, the vascular resistance and portal venous pressure are
increased, and extra- and inter-hepatic shunts develop. Blood could then
bypass thefiinctioninghepatocytes and flow through these shunts into the
116
central veins. As a result, xenobiotics, drugs, and metabolites, e.g. T M A ,
might be less accessible to functional hepatocytes, which contain enzymes for
normal metabolism, and the mean percentage of T M A 7V-oxidation in patients
with liver disease was lower than that in the normal volunteers. Moreover,
damage to hepatocytes might cause reduction of the amount and activity of
liver F M O , which could also lead to a low mean percentage of T M A N-
oxidation.
Although a previous study in rats showed that diabetes might affect F M O

activity (Rouer et al., 1988)，the current study showed no apparent change in
the percentage of T M A A^-oxidation in patients with non-insulin dependent
diabetes. Moreover, patients with hypertension who were treated with anti-
hypertensive drugs and patients with hyperlipidaemia w h o were treated with
statins also showed no effect on the T M A A^-oxidation. This suggests that, as
expected, these diseases and these types of drugs are unlikely to have a major
effect on hepatic F M O activity and the liver function for these patients was
also normal. Although atorvastatin previously showed effects on cytochrome
P450 activity, the present study with atorvastatin in patients with normal liver
function showed that this statin did not affect the percentage of T M A TV-
oxidation {p > 0.05) (Fig. 5.5). As this is the most potent of the statins, it is
unlikely that lowering of cholesterol by inhibition of hydroxymethyl-glutamyl-
C o A reductase per se will affect F M O activity, but it may not be appropriate
to extrapolate these results to other statins which could have different direct
effects on F M O .
In the hydroxychloroquine (HCQ) study, the mean values for percentage T M A

A^-oxidation were similar in patients treated with H C Q to those that were not
on H C Q treatment {p > 0.05) (Fig. 5.6). However, three female patients
treated with H C Q showed a reduced percentage of T M A 7V>oxidation (78 - 88
%). This raises the question as to why should these three patients have a lower
percentage values than the other patients treated with H C Q . Did they have
any other diseases or previous record of clinical problems? It is possible that
they are heterozygous for a mutation in the F M O gene. Perhaps there are
other factors in these patients which affect T M A TV-oxidation, such as the
117
vaginal conditions and menstruation. Females w h o are suffering from
bacterial vaginosis showed higher T M A levels in vaginal discharge than
control females (Brand and Galask, 1986; Sardas et al., 1996). Moreover,
Zhang et al. (1996b) showed that in normal females as menstruation began,
there was a marked depression in urinary T M A A^-oxide output with a
reciprocal increase in the unchanged amine, signifying a decrease in T M A TV-
oxidation. These changes might be caused by hormonal modulation.
Therefore, more clinical details will be obtained for the three female outliers
in the H C Q study initially determining the menstrual history and whether there
was evidence ofbacterial vaginosis.
According to the results obtained in this study, patients with impaired liver
function, i.e. chronic liver disease, showed an impaired ability in T M A N-
oxidation and this impairment may be caused by a reduction of liver F M O
activity and/or shunting of blood from the portal to the systemic circulation.
Other patients with normal liver function but a variety of other diseases and
receiving various commonly used dmgs did not show any significant change
in the ability of T M A 7V>oxidation. Therefore, it can be concluded that the
impairment of T M A A^-oxidation is principally related to chronic liver disease.
Therefore, cautious prescribing is needed for the drugs that are metabolized by
F M O in these patients.
The two drugs studied, atorvastatin and hydroxychloroquine, showed no

significant effect on T M A A^-oxidation. These drugs were chosen for study
because clinical studies with these agents were already planned and urine
samples were readily available. It is possible that other drugs are more likely
to have an effect on T M A 7V-oxidation and this should be studied
systematically.
118
Chapter 6
General Discussion
119
Inter-ethnic differences in drug metabolizing enzymes such as flavin-containing
monooxygenase have been reported in numerous studies involving different ethnic
groups (Chapter 1). Most ofthe studies on F M O and T M A 7V-oxidation were carried
out in Caucasian populations and in several non-Caucasian populations. In this study,
T M A 7V"0xidati0n was investigated in a Chinese population, to determine the F M O
activity by measuring T M A A^-oxidation.
A simple and sensitive gas chromatographic method for T M A and TMAO

(trimethylamine 7V-oxide) analysis in urine samples was developed (Chapter 2). B y
using the gas chromatograph with flame-ionization detector, the linear range of
calibration curve was shown to be from 0.1 |ig/ml to 10 îg/ml of T M A . The time for
one injection was less than 10 minutes and this was an improvement in the time for
analysis when compared to the old methods, i.e. more than 40 urine samples could be
analyzed per day. The analytical method was validated. Intra-assay variation of the
same urine sample in a single assay and inter-assay variation of the same urine
samples over three months showed excellent results in repeatability (CoV % = 1.3 %
for T M A , 3.7 % for T M A + T M A O (as T M A ) ) and in reproducibility (CoV % = 3.9 %
for T M A , 2.4 % for T M A + T M A O (as TMA)). The % T M A conversion (Equation
2.3) was chosen to represent the extent of T M A iV-oxidation, since the % TMA
conversion was the accepted value to represent the ability of T M A A^-oxidation in
similar pharmacogenetic studies.
The population study of T M A A^-oxidation in a Chinese population was carried out

using whole day urine collection method (24-hr urine samples). In addition, a spot
urine collection method was also employed in which the subject provided one spot
urine sample only. The results were analyzed in parallel in which the spot urine
method showed comparable results to the whole day urine collection method {p <
0.05). Therefore, the spot urine collection procedure is a simpler, but equally accurate
method for the investigation of primary trimethylaminuria. The current study showed
that none of the volunteers suffered from fish-odour syndrome (trimethylaminuria),
but seven subjects (3 % ) excreted 80-90 % of T M A as T M A O and another two
subjects (1 % ) excreted 74 % and 75 % of T M A as T M A O respectively. These
subjects may have impaired ability of T M A A^-oxidation due to genetic or other
factors. The T M A load test was not performed in this study, and as such, remains a
120
limitation of the study because the T M A load test in these people with heterozygous
defects in F M O (50 - 90 % T M A excreted as T M A O in urine) would be a useful
confirmation.
The effects of age and diet on T M A A^-oxidation were also investigated in this study.
Although it was found that some age groups showed statistical differences for T M A
or T M A O (as T M A ) excretion at certain periods, when compared to other age groups,
those differences appeared randomly rather than in a specific trend. Therefore, the
amounts of T M A and T M A O excreted in the urine did not show any specific pattern
related to age. O n the other hand, a U-shaped change for the % T M A A^-oxidation
versus age occurred. Furthermore, no correlation was found between age and the
percentage ofTMAiV-oxidation. This showed that the effect of age on the ability of
T M A A^-oxidation might not show any linear or logarithmic relationship. Since T M A
is a substrate for flavin-containing monooxygenases (FMOs), especially for liver
F M O , it is possible to extrapolate the data for human F M O activity. Therefore, the
effect of age on the F M O activity might not show any linear or logarithmic
relationship. The effects of age obtained in this study were similar to the results for
many other enzyme studies in human and showed no special correlation for F M O
activity with aging. Since only limited studies have been done before on the effect of
age on F M O activity, other types of study, including the measurement of liver F M O
activity and amount in liver biopsy samples, and longitudinal study to look at the
change ofthe F M O activity assessed by T M A A^-oxidation throughout the life-time of
the volunteers, should be carried out.
In the study of the effects of diet on T M A iV-oxidation, the four types of food were
compared based on the percentage of T M A excreted as T M A O in urine at Periods 1
and 4. It was clearly shown that intake of seafood (i.e. food rich in T M A , T M A O , or
T M A precursors) but not the other types of food (beef, chicken, and others) increased
the amounts of T M A and T M A O excreted, and it produced a higher percentage of
T M A excreted as T M A O in urine. Moreover, the high T M A O amount and high rate
of T M A O excreted in urine showed a significant positive correlation to the percentage
value 0fTMA7V"0xidized. The correlation showed that high intake of T M A , T M A O ,
or T M A precursors couldfortherstimulate the liver F M O activity and as a result there
was a higher T M A O excreted and higher percentage of TMAA^-oxidized. Moreover,
121
the above results were further confirmed by the results obtained from the five
volunteers with controlled dietary intake, intake of seafood produced the highest
percentage o f T M A excreted as T M A O in urine, and the highest amounts o f T M A and
T M A O excreted. The effects were prolonged for several hours then the values
dropped gently until most o f T M A and T M A O were excreted.
In this study, patients with chronic liver disease showed impaired ability of T M A N-
oxidation {p < 0.001), while the other patient groups (hypertension, diabetes,
hyperlipidaemia) appeared normal Q? > 0.05), when compared to normal volunteers.
The lower percentage of T M A A^-oxidized in patients with chronic liver disease was
due to the severe liver damage, characterized by the elevated levels of bilirubin (65.8
士 67.5 îmol l_i) and alanine aminotransferase (ALT) (58.0 士 44.5 RJ 1"^) in these
patients. The results clearly showed that the low percentage of T M A 7V-oxidation in
these patients was caused by the impaired liver function. Although a previous study
showed that diabetes might affect F M O activity, this study did not show any change
in the percentage of T M A A^-oxidation. Moreover, patients with hypertension who
were treated with anti-hypertensive drugs and patients with hyperlipidaemia w h o were
treated with statins also showed no effect on the T M A A^-oxidation. These suggested
that these types of drugs are unlikely to have a major effect on hepatic F M O activity
because the liver function of these patients did not change on these treatments. T w o
drugs, atorvastatin and hydroxychloroquine also had no effect on the mean percentage
of T M A excreted as T M A O in urine. However, three female patients treated with
H C Q showed a reduced percentage ofTMAA^-oxidation (78 - 88 %). This raises the
question why these three patients had a lower percentage value than other patients
treated with H C Q . The changes might be caused by hormonal modulation but more
clinical details are needed to assess these cases.
Taken together,fromthe results obtained in this study, it can be concluded that about
1 % ofthe Chinese population may have impaired ability of T M A A^-oxidation (< 90
% of T M A excreted as T M A O in urine). The result was similar to previous studies in
British Caucasian populations. The percentage of T M A A/-oxidized was not affected
by age, but intake of high-TMA foods (e.g. seafood) could increase the percentage
value of T M A A^-oxidation and the liver F M O activity in normal subjects. Therefore,
122
intake of seafood might be useful to characterize the presence of impaired ability of
TMAA^-oxidation in man, due to the presence of large amounts of T M A and T M A O
in seafood. Chronic liver disease lowered the percentage value of T M A 7V-oxidized
due to the damage of liver and the reduction of liver function, while other diseases
(diabetes, hypertension, and hyperlipidaemia) with normal liver function did not
affect the percentage value of T M A A/-oxidation. Furthermore, atorvastatin and
hydroxychloroquine also showed no apparent effect on the percentage of T M A N-
oxidation but other drugs should also be studied systematically.
123
References
Abeling, N . G . G . M., Van Gennip, A . H., Bakker, H . D , Heerschap, A., Engelke, U.,
Wevers, R. A . (1995) Diagnosis of a new case of trimethylaminuria using direct
proton N M R spectroscopy of urine. J. Inher. Metab. Dis. 18: 182-184.
Agrawal, A . K., Shapiro, B. H . (1996) Phenobarbital induction of hepatic C Y P 2 B 1
and CYP2B2: pretranscriptional and post-transcriptional effects of gender, adult age,
and phenobarbital dose. Mol Pharmacol 49: 523-531.
Agarwal, D . P., Harada, S., Goedde, H . W . (1981) Racial differences in biological
sensitivity to ethanol: the role ofalcohol dehydrogenase and aldehyde dehydrogenase
isozymes. Alcoholism: Clin. Exp. Res. 5: 12-16.
Ahsan, C. H., Renwick, A . G., Macklin, B., Challenor, V . F., Waller, D . G., George,
C. F. (1991) Ethnic differences in the pharmacokinetics of oral nifedipine. Br. J.
Clin. Pharmacol 31: 399-403.
Alexanderson, B., Borga, 0. (1972) Interindividual differences in plasma protein
binding ofnortriptyline in m a n — a twin study. Eur. J, Clin. Pharmacol 4: 196-200.
Allen, M . D.，Greenblatt, D . J., Harmatz, J. S., Shader, R. I. (1980)
Desmethyldiazepam kinetics in the elderly after oral prazepam. Clin. Pharmacol
Ther. 28: 196-202.
Alvare, A . P., Anderson, K . E., Conney, A . H., Kappas, A . (1976) Interactions
between nutritional factors and biotransformations in man. Proc. Natl Acad. Sci.
U.S.A. 73: 2501-2504.
Al-Waiz, M . (1988) Genetic polymorphism of trimethylamine N-oxidation in man
and its relationship to the fish-odour syndrome. P h D Thesis. University of London.
172-173.
Al-Waiz, M., Mitchell, S. C., Idle, J. R., Smith, R. L. (1987a) The metabolism of
^^C-labelled trimethylamine and its A^-oxide in man. Xenobiotica. 17: 551-558.
Al-Waiz, M., Ayesh, R., Mitchell, S. C., Idle, J. R.，Smith, R. L. (1987b) Disclosure
ofthe metabolic retroversion of trimethylamine A^-oxide in humans. Clin. Pharmacol
Ther. 42: 602-612.
Anders, M . W . (1980) Metabolism of drugs by the kidney. Kidney Int. 18: 636-
647.
124
Anderson, K . E. (1988) Influences of diet and nutrition on clinical pharmacokinetics.
Clin, Pharmacokinet. 14: 325-346.
Anderson, K . E., Conney, A . H., Kappas, A . (1979) Nutrition and oxidative drug
metabolism in man: relative influence of dietary lipids, carbohydrate, and protein.
Clin. Pharmacol Ther. 26: 493-501.
Anderson, K . E., Conney, A . H., Kappas, A . (1982) Nutritional influences on
chemical biotransformations in humans. Nutr. Rev. 40: 161-171.
Anderson, K . E., Schneider, J., Pantuck, E. J., Pantuck, C. B., Mudge, G . H., Welch,
R . M., Conney, A . H., Kappas, A . (1983) Acetaminophen metabolism in subjects fed
charcoal-broiled beef. Clin. Pharmacol Ther. 34: 369-374.
Ando, M., Katagiri, K., Yamamoto, S., Wakamatsu, K., Kawahara, I., Asanuma, S.,
Usuda, M., Sasaki, K . (1997) Age-related effects of heat stress on protective
enzymes for peroxides and microsomal monooxygenase in rate liver. Environ. Health
Prespect 105: 726-733.
Antal, E. J., Karmer, P. A., Mercik, S. A., Chapron, D . J., Lawson, I. R. (1981)
Theophylline pharmacokinetics in advanced age. Br. J. Clin. Pharmacol. 12: 637-
645.
Ayesh, R., Mitchell, S. C., Zhang, A., Smith, R. L. (1993) The fish odour syndrome:
biochemical, familial and clinical aspects. Br. Med. J. 307: 655-657
Ayesh, R., Smith, R. L. (1992) Genetic polymorphism of trimethylamine N-
oxidation. In Pharmacogenetics of Drug Metabolism, Ed. Kalow, W . Pergamon
Press, N e w York, U.S.A. 315-332.
Bandiera，S., Ryan, D . E., Levin, W.，Thomas, P. E. (1986) Age- and sex-related
expression of cytochromes P450f and P450g in rat liver. Arch Biochem. Biophy.
248: 658-676.
Biehl, J. P. (1957) Emergence of drug resistance as related to the dosage and
metabolism of isoniazid. Trans. 1&^ Conf. Chemother. Tuherc. Washington D . C.,
U . S. Veterans A d m . A r m y Navy, U.S.A. 108-113.
Board, P. G . (1981) Biochemical genetics of glutathione *S'-transferase in man. Am.
J. Hum. Genet. 33: 36-43.
B6nicke, R., Reif, W. (1953) Enzymatische inaktivierung von
isonicotinsaurehydrazid im menschlichen und tierischen organismus. Arch Exp.
Pathol Pharmak. 220: 321-333.
125
Borson, W . F., Lee, T K . (1986) Genetic polymorphism of human liver alcohol and
aldehyde dehydrogenases, and their relationship to alcohol metabolism and
alcoholism. Hepatology. 6: 502-510.
Brand, J. M., Galask, R. P. (1986) Trimethylamine: The substance mainly
responsible for the fish odor often associated with bacterial vaginosis. Obstet.
Gynecol 68: 682-685.
Brewster, M . A., Schedewie, H . (1983) Trimethylaminuria. Ann. Clin. Lab. Sci.
13: 20-24.
Briant, R. H . Dorrington, R. E., Cleal, J., William, F. M . (1976) The rate of
acetaminophen metabolism in the elderly and the young. J. Am. Geriatr. Soc. 24:
359-361.
Britto, M . R., McKean, H . E., Bruckner, G. G., Wedlund, P. J. (1991)
Polymorphisms in oxidative drug metabolism: relationship to food preference. Br. J.
Brodfuehrer, J. I., Zannoni, V. G. (1986) Ascorbic acid deficiency and the flavin
containing monooxygenase. Biochem. Pharmacol 35: 637-644.
Brodfuehrer, J. L, Zannoni, V . G. (1987a) Flavin-containing monooxygenase and
、 ascorbic acid deficiency. Qualitative and quantitative differences. Biochem.
Pharmacol 36: 3161-3167.
Brodfuehrer, J. I., Zannoni, V . G. (1987b) Modulation of the flavin-containing
monooxygenase in guinea pigs by ascorbic acid and food restriction. J. Nutr. 117:
286-290.
Burnett, V . L., Lawton, M . P., Philpot, R. M . (1994) Cloning and sequencing of
flavin-containing monooxygenases FM03 and FM04 from rabbit and
characterization o f F M 0 3 . J. Biol Chem. 269: 14314-14322.
Callaghan, R., Desmond, P. V.，Paull, P., Mashford, M . L. (1993) Hepatic enzyme
activity is the major factor determining elimination rate of high-clearance drugs in
cirrhosis. Hepatology. 18: 54-60.
Campbell, T. C.，Hayes, T. R. (1974) Role of nutrition in the drug-metabolizing
enzyme system. Pharmacol. Rev. 26: 171-197.
Carson, P. E., Flanagan, C. L., Ickes, C. E., Aliving, A. S. (1956) Enzymatic
deficiency in primaquine sensitive erythrocytes. Science. 124: 484-485.
Cashman, J. R. (1995) Structural and catalytic properties of the mammalian flavin-
coating monooxygenase. Chem. Res. Toxicol 8: 165-181 •
126
Cashman, J. R.，Park, S. B , Berkman, C. E., Cashman，L. E. (1995) Role ofhepatic
flavin-containing monooxygenase 3 in drugs and chemical metabolism in adult
humans. Chem. Biol Interact 96: 33-46.
Cashman, J. R., Yang, Z. C., Yang, L., Wrighton, S. (1993) Stereo- and
regioselective N- and *S'-oxidation of tertiary amines and sulfides in the presence of
adult human liver microsomes. DrugMetab. Dispos. 21: 392-501.
Castleden, C. M., George, C. F. (1979) The effect of ageing on the hepatic clearance
of propranolol. Br. J. Clin. Pharmacol 7: 49-54.
Castleden, C. M., George, C. F.，Marcer, D., Hallett, C. (1977) Increased sensitivity
to nitrazepam in old age. Br. Med. J. 1: 10-12.
Chen, K., Robbins, E. (1944a) Influence of age of mice on the toxicity of alcohol. J.
Am. Pharm. Assoc. 33: 62-65.
Chen, K., Robbins, E. (1944b) Age of animals and drug action. J. Am. Pharm.
Assoc. 33: 80-85.
Cholerton, S., Smith R. L. (1991) H u m a n pharmacogenetics of nitrogen oxidations.
In N-Oxidation ofDrugs: Biochemistry, Pharmacology, Toxicology, Eds. Hlavica, P.,
Damani, L. A . Chapman and Hall, London, U.K. 107-131.
Christopher, B., Ballinger, R., Shepherd, A. (1979) A survey ofhospital prescribing
for the elderly. In Drugs and the Elderly, Eds. Crooks, J., Stevenson, I. University
I
Park Press, Baltimore, NJ, U S A . 321-338.
Conney,A.H. (1986) Induction of microsomal cytochrome P-450 enzymes: the first
Bernard B. Brodie lecture at Pennsylvania State University. Life Sci. 39: 2493-2518.
Cormey, A . H., Pantuck, E. J., Hsiao, K . C., Garland, W . A., Anderson, K . E.,
Alvares, A . P., Kappas, A. (1976) Enhanced phenacetin metabolism in human
subjects fed charcoal-broiled beef. Clin. Pharmacol Ther. 20: 633-642.
Crooks, J., 0'Malley, K., Stevenson, 1. (1976) Pharmacokinetics in the elderly. Clin.
Pharmacokinet. 1: 280-296.
DaCosta, K . A., Vrbanac, J. J., Zeisel, S. H . (1990) The measurement of
dimethylamine, trimethylamine, and trimethylamine A^-oxide using capillary gas
chromatography-mass spectrometry. Anal Biochem. 187: 234-239.
Damani, L. A. (1985) Colorimetric and chromatographic assay of tertiary amine N-
oxides. In Biological Oxidation of Nitrogen in Organic Molecules: Chemistry，
Toxicology and Pharmacology, Eds. Gorrod, J. W., Damani, L. A . Ellis Horwood
Ltd., Chichester, U.K. 33-39.
127
Danks, D . M.，Hammond, J., Schlesinger, P., Fach, K.，Burke, D., Halpern, B. (1976)
Trimethylaminuria: diet does not always control the fishy odour. N. Eng. J. Med
295: 962.
Davidson, S. C., Wills, E. D . (1972) Studies on the lipid composition of the rat liver
endoplasmic reticulum after induction with phenobarbital and 20-methylcholanthrene.
Biochem.J. 140: 461-468.
Dawling, S., Crome, P., Braithwaite, R. (1980) Pharmacokinetics of single oral
doses of nortriptyline in depressed elderly hospital patients and young healthy
volunteers. Clin. Pharmacokinet. 5: 394-401.
Dearing, W., Barnes, A.，Essex, H . (1944) Experiments with calculated therapeutics
and toxic doses of digitalis. Am. Heart. J. 27: 96-107.
D e La Huerga, J., Popper, M . (1951) Urinary excretion of choline metabolites
following administration in normals and patients with hepatobiliary diseases. J. Clin.
Investig. 30: 364-470.
Divoll, M., Abernethy, D . R., Ameer, B., Greenblatt, D . J. (1982) Acetaminophen
kinetics in the elderly. Clin. Pharmacol Ther. 31: 151-156.
Divoll，M.，Greenblatt, D . J, Harmatz, J. S , Shader, R. 1. (1981) Effect of age and
gender on disposition of temazepam. J. Pharm. Sci. 70: 1104-1107.
Dolphin, C., Sherpard，E. A., Povey, S., Palmer, C. N., Ziegler, D. M., Ayesh, R.,
Smith, R . L.，Phillips, I. R. (1991) Cloning, primary sequence, and chromosomal
mapping ofahuman flavin-containing monooxygenase (FM01). J. Biol Chem. 266:
12379-12385.
Dolphin, C. T, Riley, J. H., Smith, R. L., Shephard, E. A., Phillips, I. R. (1997a)
Structural organization of the human flavin-containing monooxygenase 3 gene
(FM03), the favored candidate for fish-odor syndrome, determined directly from
genomic DNA. Genomics. 46. 260-267.
Dolphin, C. T.，Janmohamed, A . , Smith, R. L., Shephard, E . A . , Phillips, 1. R.
(1997b) Missense mutation in flavin-containing mono-oxygenase 3 gene, F M 0 3 ,
underlies fish-odour syndrome. Nat. Genet. 17: 491-494.
Donovan, M . A., Castleden, C. M., Pohl, J. E. F., Karft，C. A . (1981) The effect of
age on digitoxin pharmacokinetics. Br. J. Clin. Pharmacol 11: 401-402.
Drayer, D . E., Hughes, M.，Lorenzo, B., Reidenberg, M . M . (1980) Prevalence of
high (3 S)-3-hydroxyquinidine/quinidine ratios in semm, and clearance of quinidine in
cardiac patients with age. Clin. Pharmacol Ther. 27: 72-75.
128
Dyer, W . J. (1962) Amines infishmuscle. VI. Trimethylamine oxide content of
fish and marine vertebrates. J. Fish. Res. Bd Can. 8: 314-324.
Eckhardt, M . (1978) Consequences of alcohol and other drug use in the aged. In The
Biology ofAging, Eds. Behnke, J., Finch, C., Moment, G . Plenum Press, N e w York,
N Y , U S A . 191-204.
Edwards, J. A., Price Evans, D . A . (1967) Ethanol metabolism in subjects possessing
typical and atypical liver alcohol dehydrogenase. Clin. Pharmacol Ther. 8: 824-
829.
Evarts, R. P., Mostafa, M . H . (1981) Effects of indole and tryptophan on cytochrome
P-450, dimethylnitrosamine demethylase, and arylhydrocarbon hydroxylase activities.
Biochem. Pharmacol 30: 517-522.
Fagan, T. C., Walle, T., Oexmann, M . J., Walle, U . K., Bai, S. A . Gaffney, T. E.
(1987) Increased clearance of propranolol and theophylline by high-protein compared
with high-carbohydrate diet. Clin. Pharmacol Ther. 41 402-406.
Farah, F., Taylor, W.，Rawlins, M . D., James, 0 . (1977) Hepatic drug acetylation
and oxidation: effect of aging in man. Br. Med. J. 2: 155-156.
Farrell, G . C. (1987) Drug metabolism in extrahepatic diseases. Pharmacol Ther.
35: 375-404.
Feely, J., Crooks, J., Stevenson, I. H . (1981) The influence of age, smoking, and
hyperthyroidism on plasma propranolol steady state concentration. Br. J. Clin.
Pharmacol 12: 73-78.
Fenwick, G . R., Butler, E. J., Brewster, M . A. (1983) Are brassica vegetables
aggravating factors in trimethylaminuria (fish-odour syndrome)? Lancet. 1 1983.
Fenwick, G . R , Heaney, R K., Mullin, J. W . (1982) Glucosinolates and their
breakdown products in food and food plants. CRC Critical Review ofFood, Science,
andNutrition. 18: 123-201.
Fernandez, M . S.，Gutierrez, C , Vila, J. J., L6pez, A., Ibanez, V.，Sanguesa, C ,
Lluna, J., Barrios, J. E. (1997) Congenital intrahepatic portocaval shunt associated
with trimethylaminuria. Pediatr. Surg. Int. 12: 196-197.
Fong, W . P., Ho, Y W., Lee, C Y., Keung, W M. (1989) Liver alcohol and
aldehyde dehydrogenase isozymes in a Chinese population in Hong Kong. Hum.
Hered 39 185-191.
Fukui, M , Wakasugai, C (1972) Liver alcohol dehydrogenase in Japanese
population. Jap. 1 Legal Med, 26 46-51.
129
Fulton, B., James, 0., Rawlins, M . D . (1979) The influence of ace on the
pharmacokinetics of paracetamol. Br. J. Clin. Pharmacol 7: 418P.
Garrod, A . E. (1909) Inborn Errors of Metabolism. Oxford University Press.
London, U.K.
Garrod, A . E. (1931) The Inborn Factors in Disease. Oxford University Press.
London, U.K.
Gasser, R.. (1996) The flavin-containing monooxygenase system. Exp. Toxic.
Pathol 48: 467-470.
Gasser, R.，Tynes. R. E., Lawton, M . P., Korsmeyer, K . K.，Ziegler, D . M., Philpot, R.
M . (1990) The flavin-containing monooxygenase expressed in pig liver: primary
sequence，distribution, and evidence for a single gene. Biochemistry. 29: 119-124.
George, J., Byth, K.，Farrel, G. C. (1995) Age but not gender selectively affects
expression of individual cytochrome P450 proteins in human liver. Biochem.
Gibson, T. P. (1986) Renal disease and drug metabolism: an overview. Am. J.
KidneyDis. 8: 7-17.
Goedde，H. W., Agarwal, D . P (1986) Pharmacogenetics of aldehyde dehydrogenase.
In Pharmacogenetics of Drug Metabolism, Ed. Kalow, W . Pergamon Press. N e w
York, U.S.A. 281-311.
Goedde, H . W., Harada, S., Agarwal, D . P. (1979) racial differences in alcohol
sensitivity: a new hypothesis. Hum. Genet. 51: 331-334.
Greenblatt, D . J., Abernethy, D . R., Shader, R. 1. (1986) Pharmacokinetic aspects of
drug therapy in the elderly. Ther. DrugMonit. 8: 249-255.
Greenblatt, D . J., Allen, M . D., Harmatz, J. S., Shader, R. I. (1982a) Diazepam
disposition determinants. Clin. Pharmacol Ther. 27: 301-312.
Greenblatt, D . J., Allen, M . D., Locniskar, A., Harmatz, J. S., Shader, R. I. (1979)
Lorazepam kinetics in the elderly. Clin. Pharmacol Ther. 26: 103-113.
Greenblatt, D . J.，Divoll, M , Abernethy, D . R.，Harmatz, J. S., Shader, R. I. (1982b)
Antipyrine kinetics in the elderly: prediction of age-related changes in benzodiazepine
oxidizing capacity. J. Pharmacol Exp. Ther. 220: 120-126.
Greenblatt, D . J., Divoll, M , Abernethy, D . R., Moschitto, L. J.，Smith, R. B., Shader,
R. I. (1983) Alprazolam kinetics in the elderly: relation to antipyrine disposition.
Arch. Gen. Psychiatry. 40: 287-290.
130
Greenblatt, D . J:，Divoll, M , Harmatz, J. S., Maclaughlin, D . S., Shader, R. I. (1981b)
Kinetics and clinical effects of flurazepam in young and elderly noninsomniacs. Clin.
Pharmacol Ther. 30: 475-486.
Greenblatt, D . J., Divoll, M , Harmatz, J. S., Shader, R. I. (1980) Oxazepam kinetics:
effects of age and sex. J. Pharmacol Exp. Ther. 215: 86-91.
Greenblatt, D . J., Divoll，M., Puri, S. K., Ho, I, Zinny, M . A.，Shader, R. I. (1981a)
Clobazam kinetics in the elderly. Br. J. Clin. Pharmacol 12: 631-636.
Greenblatt, D . J., Sellers, E. M., Shader, R. I. (1982c) Drug disposition in old age. N.
Eng. 1 Med. 306: 1081-1088.
G u o , Z.，Wang, M.，Tian, G., Burger, J.，Gochfeld, M.，Yang, C . S. (1993) Age- and
gender-related variations in the activities of drug-metabolizing and antioxidant
enzymes in the white-footed mouse {Peromyscus leucopus). Growth Develop. Aging.
57: 85-100.
Hadidi, H . F., Cholerton, S , Atkinson, S., Irshaid, Y . M., Rawashdeh, N . M., Idle, J.
R. (1995) The #-oxidation oftrimethylamine in a Jordanian population. Br. J. Clin.
Hall, G., Ettinger, G., Banting, F. (1936) A n experimental production of coronary
thrombosis and myocardial failure. Can. Med. Assoc. J. 34: 9-15.
Harada, S., Abei, M., Tanaka, N., Agarwal, D . P., Goedde, H . W . (1987) Liver
glutathione 5'-transferase polymorphism in Japanese and its pharmacogenetic
importance. Hum. Genet. 75: 322-325.
Harada, S., Misawa, S, Agarwal, D . P., George, H . W . (1980) Liver alcohol and
aldehyde dehydrogenase in the Japanese: isozyme variation and its possible role in
alcohol intoxication. Am. J. Hum. Genet. 32: 8-15.
Hayes, M . J., Langman, M . J. S., Short, A. H . (1975) Changes in drug metabolism
with increasing age. 2. Phenytoin clearance and protein binding. Br. J. Clin.
Pharmacol 2: 73-79.
Hempel, J., Kaiser, R., Jornvall, H . (1985) Mitochondrial aldehyde dehydrogenase
from human liver: primary structure, differences in relation to the cytosolic enzyme
andfiinctionalcorrelations. Eur. J. Biochem. 153: 13-28.
Hempel, J., Von Bahr-Lindstrom, H., Jornvall, H . (1984) Aldehyde dehydrogenase
from human liver. Primary structure ofthe cytoplasmic isoenzyme. Eur. J. Biochem.
141: 21-35.
131
Hickey, P. L., Angus, P. W., McLean, A . J., Morgan, D . J. (1995) Oxygen
supplementation restores theophylline clearance to normal in cirrhotic rats.
Gastroenterology. 108: 1504-1509.
Hirama, Y , Shinriki, N., Fujino, T., Uedaira，H. (1993) ^ H - N M R spectroscopic
analysis of human urine: periodic variation of trimethylamine oxide excretion.
Physiol Chem. Phys. 25: 209-218.
Hodis, H . N., Quismorio Jr., F. P., Wickham, E., Blankehorn, D . H . (1993) The lipid,
lipoprotein, apolipoprotein effects of hydroxychloroquine in patients with systemic
lupus erthematosus. J. Rheumatol 20: 661-665.
Hogan, E. M., Nicholls, P. J,. Yoosuf, A . (1979) Hepatic microsomal oxidative N-
demethylation in rats with renal failure. Br. J. Pharmacol 66: 74P-75P.
Holloway, D . E., Peterson, F. J. (1984) Ascorbic acid in drug metabolism. In Drugs
and Nutrients: The Interactive Effects, Eds. Roe, D . A., Campbell, T. C. Marcel
Dekker, Inc., N e w York, N Y , U.S.A. 225-295.
Horbach, G . J. M . J., Van Asten, J. G., Van Bezooijen, F. A . (1990) The influence of
age on the inducibility of rat liver cytochrome P450IA1 (CYPIA1) and P450IA2
(CYPIA2) m R N A s . Mutat Res. 237: 117-121.
Horio, F., Ozaki, K., Kohmura, M., Yoshida, A., Makino, S., Hayashi, Y . (1986)
Ascorbic acid requirement for the induction of microsomal drug-metabolizing
enzymes in a rat mutant unable to synthesize ascorbic acid. J. Nutr. 116: 2278-2289.
Humbert, J. R., H a m m o n d , K . B., Hathaway, W . E., Marcoux, J. G., O'Brien, D .
(1970) Trimethylaminuria: the fish-odour syndrome. Lancet. 2: 770-771.
Hunt, C. M.，Starter, S., Stave, G. M . (1990) Effect of normal aging on the activity
ofhuman hepatic cytochrome P450IEE1. Biochem. Pharmacol 40: 1666-1669.
Hunt, C. M.，Westerkam, W . R., Stave, G. M . (1992) Effect ofage and gender on the
activity ofhuman hepatic C Y P 3 A . Biochem. Pharmacol 44: 275-283.
Hurwitz, N . (1969) Predisposing factors in adverse reactions to drugs. Br. Med. J.
1: 536-539.
Ikuta, T., Shibuya, A., Yoshida, A. (1986) Direct determination of usual (Caucasian-
type) and atypical (Oriental-type) alleles of the class I human alcohol dehydrogenase-
2 locus. Biochem. Genet 26: 519-525.
Irvine, R. E., Grove, J.，Toseland, P. A., Trounce, J. R. (1974) The effect of age on
the hydroxylation of amylobarbitone sodium in man. Br. J. Clin. Pharmacol 1: 41-
43.
132
Jackson, P. R., Tucker, G . T., Lennard, M . S.，Woods, H . F. (1986) Polymorphic
drug oxidation: pharmacokinetic basis and comparison of experimental indices. Br.
J. Clin. Pharmacol 22: 541-550.
Jenne, J.，Nagasawa, H., M c H u g h , R., MacDonald, F., Wyse, E. (1975) Decreased
theophylline half-life in cigarette smokers. Life Sci. 17: 195-198.
Jusko, W . J., Gardner, M . J., Mangione, A., Schentag, J. J., Koup, J. R., Vance, J. W .
(1979) Factors affecting theophylline clearance: age, tobacco, marijuana, cirrhosis,
congestive heart failure, obesity, oral contraceptives, benzodoazepines, barbiturates,
and ethanol. J. Pharm. Sci. 68: 1358-1366.
K a l o w , W . (1956) Familial incidence oflow pseudocholinesterase level. Lancet. 2:
576-577.
Kalow, W.，Genest, K . (1957) A method for the detection of atypical forms of
human serum cholinesterase. Determination of dibucaine numbers. Can. J. Biochem.
Physiol 35: 568-574.
Kalow, W., Staron, N . (1957) O n distribution and inheritance of atypical forms of
human serum cholinesterase as indicated by dibucaine numbers. Can. J. Biochem.
Physiol 35: 1305-1320.
Kangas, L., Iisalo, E., Kanto, J., Lehtinen, V., Pynnonen, S., Ruikka, I., Salminen, J.,
Sillanpaa, M., Syvalahti, E. (1979) H u m a n pharmacokinetics of nitrazepam: effect
ofage and diseases. Eur. J. Clin. Pharmacol 15: 163-170.
Kanto, J.，Kangas, L., Aaltonen, L., Hilke, H . (1981) Effect of age on the
pharmacokinetics and sedative effect offlunitrazepam.Int. J. Clin. Pharmacol 19:
400-404.
Kanto, J. Maenpaa, M., Mantyla, R., Sellman，R., Valovirta, E. (1979) Effect of age
on the pharmacokinetics of diazepam given in conjugation with spinal anesthesia.
Anesthesiology. 51: 154-159.
Kappas, A., Alvares, A. P., Anderson, K . E., Pantuck, E. J., Pantuck, C. B., Chang,
R., Conney, A . H . (1978) Effect of charcoal-broiled beef on antipyrine and
theophylline metabolism. Clin. Pharmacol Ther. 23: 445-450.
Kappas, A., Anderson, K . E., Conney, A. H., Alvares, A . P. (1976) Influence of
dietary protein and carbohydrate on antipyrine and theophylline metabolism in man.
133
Kataoka, H . (1996) Derivatization reactions for the determination of amines by gas
chromatography and their applications in environmental analysis. J. Chromatogr. A.
733: 19-34.
Kataoka, H.，Eda, M , Makita, M . (1993) Selective determination of secondary
amines as their A^-diethylthiophosphoryl derivatives by gas chromatography with
flame photometric detection. Biomed. Chromatogr. 7: 129-133.
Kataoka, H.，Ohrui, S., Miyamoto, Y., Makita, M . (1992) Determination of low
molecular weight aliphatic primary amines in urine as their benzenesulphonyl
derivatives by gas chromatography with flame photometric detection. Biomed
Chromatogr. 6: 251-254.
Kiniron, M . T., Lang, C. C., He, H . B., Ghebreselasie, K., Shay, S., Robin, D . W.,
W o o d , A . J. J. (1996) Triazolam pharmacokinetics and pharmacodynamics in
Caucasians and southern Asians: ethnicity and C Y P 3 A activity. Br. J. Clin.
Klotz, U , Avant, G. R., Hoyumpa, A., Schenker, S., Wilkinson, G. R. (1975) The
effects ofage and liver disease on the disposition and elimination ofdiazepam in adult
man. J. Clin. Invest 55: 347-359.
Klotz, U., Muller-Seydlitz, P. (1979) Altered elimination of desmethyldiazepam in
the elderly. Br. J. Clin. Pharmacol 7: 119-120.
Kraus, J. W., Desmond, P. V., Marshell, J. P., Johnson, R. F., Schenker, S.,
Wilkinson, G. R. (1978) Effects of aging and liver disease on disposition of
lorazepam. Clin. Pharmacol Ther. 24 411-419.
Krisch, W . (1968) Enzymatische hydrolyse van diathyl-/?-nitro-phenylphosphat
(E600) durch menschliches serum. Z.Klin.Chem. 6: 41-45.
Krzymien, M . E., Elias, L. (1990) Feasibility study on the determination of fish
freshness by trimethylamine headspace analysis. J. FoodSci. 55: 1228-1232.
Kuntzmann, R, Jacobson, M., Schneidman, K., Conney, A. H . (1964) Similarities
between oxidative drug-metabolizing enzymes and steroid hydroxylases in liver
microsomes. J. Pharmacol Exp. Ther. 146: 280-285.
Kuntzmann, R, Pantuck, E. J., Kaplan, S. A., Conney, A. H . (1977) Phenacetin
metabolism: effects of hydrocarbons and cigarette smoking. Clin. Pharmacol Ther.
22: 757-764.
134
Kawasaki, S., Imamura, H., Bandai, Y.，Sanjo, K., Idezuki, Y . (1992) Direct
evidence for the intact hepatocyte theory in patients with liver cirrhosis.
Gastroenterology. 102: 1351-1355.
La Du, B. N . (1992) Overview ofpharmacogenetics. In Pharmacogenetics ofDrug
Metabolism, Ed. Kalow, W . Pergamon Press. N e w Y o r k , U.S.A. 1-12.
Lai, M . L., Wang, S. L., Lai, M . D.，Lin, E. T., Tse, M., Huang, J. D . (1995)
Propranolol disposition in Chinese subjects of different C Y P 2 D 6 genotypes. Clin.
Pharmacol Ther. 58: 264-268.
Laisney，V.，Cong, N . V., Gross, M . S., Frezal, J. (1984) H u m a n genes for
glutathione *S'-transferase. Hum. Genet. 68: 221-227.
Lawton, M . P., Gasser, R., Tynes, R. E., Hodgson, E.，Philpot, R. M . (1990) The
flavin-containing monooxygenase enzymes expressed in rabbit liver and lung are
products ofrelated but distinctly different genes. J. Biol Chem. 265: 5855-5861.
Lee, C. W . G., Yu, J. S., Turner, B. B., Murray, K . E. (1976) Trimethylaminuria:
Fishy odors in children. N. Eng. J. Med. 295: 937-938.
Lee, E. J. D., N a m , Y . P., Hee, G. N . (1988) Oxidation phenotyping in Chinese and
Malay populations. Clin. Exp. Pharmacol Physiol 15: 889-891.
Lee, P. C.，Werlin, S. L. (1995) The induction of hepatic cytochrome P450 3 A in
rats: effects ofage. Proc. Soc. Exp. Biol. Med 210: 134-139.
Liddell, D . E., Williams, F. M., Brian, F. H . (1975) Phenazone (antipyrine)
metabolism and distribution in young and elderly adults. Clin. Exp. Pharmacol
Physiol 2: 481-487.
Lomri, N.，Gu, Q., Cashman, J. R. (1992) Molecular cloning of the flavin-containing
monooxygenase (form II) c D N A from adult human liver. Proc. Natl. Acad. Sci.
U.S.A. 89: 1685-1689.
Lou, Y., Ying, L., Bertilsson, L., Sjoqvist, F. (1987) L o w frequency of slow
debrisoquine hydroxylation in a native Chinese population. Lancet. 2: 852-853.
Lou, Y. C. (1990) Differences in drug metabolism polymorphism between Orientals
and Caucasians. DrugMet Rev. 22: 451-475.
Loub, W . D., Wattenberg, L. W., Davis, D . W . (1975) Arylhydrocarbon hydroxylase
induction in rat tissues by naturally occurring indoles of cruciferous plants. J. Natl.
Cancer Inst. 54: 985-988.
135
Lundh, T.，Akesson, B . (1993) Gas chromatographic determination of primary and
secondary low-molecular-mass aliphatic amines in urine using derivatization with
isobutyl chloroformate. J. Chromatogr. 613: 191-196.
Macklon, A . F., Barton, M., James, O.，Rawlins, M . D . (1980) The effect of age on
the pharmacokinetics of diazapam. Clin. Sci. 59: 479-483.
Marks, R., Dudley, F., W a n , A . (1978) Trimethylamine metabolism in liver disease.
Lancet. 1: 1106-1107.
Marks, R., Greaves, M . W., Danks, D., Pummer, V . (1976) Trimethylaminuria or
fish odour syndrome in a child. Br. J. Dermatol 95: 11-12.
Marshall, W . J., McLean, A. E. M . (1971) A requirement for dietary lipids for
induction of cytochrome P-450 by phenobarbitone in rat liver microsome fraction.
Biochem. J. 122: 569-577.
Martinez-Hernandez, A., Martinez, J. (1991) The role of capilliarization in hepatic
failure: studies in carbon tetrachloride-induced cirrhosis. Hepatology. 14: 864-874.
Mather, L. E., Tucker, G . T.，Pflug, A. E., Lindop, M . J., Wilkerson, C. (1975)
Meperidine kinetics in man: intravenous injection in surgical patients and volunteers.
McGurie, M . C., Nogueira, C. P., Bartels, C. F.，Lightstone, H., Hajra, A.，Van der
Spek, A . F. L., Lockridge, 0., La Du, B. N . (1989) Identification of the structural
mutation responsible for the dibucaine-resistant (atypical) variant form of human
serum cholinesterase. Proc. Natl Acad. Sci. U.S.A. 86: 953-957.
McLean , A . J., Morgan, D . J. (1991) Clinical pharmacokinetics in patients with liver
disease. Clin. Pharmacokinet 21: 42-69.
Melander, A . (1978) Influence of food on the bioavailability of drugs. Clin.
Pharmacokinet. 3; 337-351.
Melander, A., Danielson, K., Schersten, B., Wahlin, E. (1977) Enhancement of the
bioavailability of propranolol and metoprolol by food. Clin. Pharmacol Ther. 22:
108-112.
Meyer, B.，Luo, H . S., Bargetzi, M., Renner, E. L., Stalder, G. A. (1991) Quantitation
of intrinsic drug-metabolizing capacity in human liver biopsy specimens: support for
the intact-hepatocyte theory. Hepatology. 13: 475-481.
Mitchell, S. C. (1996) The fish-odour syndrome. Perspect. Biol. Med 39: 514-526.
136
Mitchell, S. C.，Zhang, A . Q.，Barrett, T., Ayesh, R., Smith, R . L. (1997) Studies on
the discontinuous A^-oxidation of trimethylamine among Jordanian, Ecuadorian and
N e w Guinean populations. Pharmacogenetics. 7: 45-50.
Montamat, S. C., Cusack, B . J., Vestal, R . E. (1989) Management of drug therapy in
the elderly. N. Eng. J. Med 321: 303-309.
Morgan, D . J., McLean, A . J. (1991) Therapeutic implications of impaired hepatic
oxygen diffusion in chronic liver disease. Hepatology. 14: 1280-1282.
Morgan, D . J., McLean, A . J. (1995) Clinical pharmacokinetic and
pharmacodynamic considerations in patients with liver disease. Clin. Pharmacokinet.
29: 370-391.
Motulsky, A . G . (1957) Drug reactions, enzymes and biochemical genetics. J. Am.
Med. Assoc. 165: 835-837.
Nakajima, T.，Wang, R. S., Katkura, Y.，Kishi, R., Elovaara, E , Park, S. S., Gelboin,
H . V., Vainio, H . (1992) Sex-, age- and pregnancy-induced changes in the
metabolism oftoluene and trichloroethylene in rat liver in relation to the regulation of
cytochrome P450Iffil and P450nCll content. J. Pharmacol Exp. Ther. 261 869-
874.
Nation, R. L.，Triggs, E. J., Selig, M . (1977) Lignocaine kinetics in cardia patients
and aged subjects. Br. J. Clin. Pharmacol 4: 439-448.
Nies, A., Robinson, D . S.，Friedman, M . J., Green, R., Cooper, T. B., Ravaris, C. L.,
Ives, J. 0 . (1977) Relationship between age and tricyclic antidepressant plasma
levels. Am. J. Psychiatr. 134 790-793.
Norred, W . P., Wade, A. E. (1972) Dietary fatty acid induced alternations of hepatic
microsomal metabolism. Biochem. Pharmacol 21: 2887-2897.
Ochs, H . R.，Greenblatt, D . J., Divoll, M., Aberaethy, D R., Feyenbend, H., Dengler,
H.J. (1981) Diazepam kinetics in relation to age and sex. Pharmacology. 23: 24-
30.
Ochs, H . R , Greenblatt, D . J., Otten, H . (1981) Disposition of oxazepam in relation
to age, sex, and cigarette smoking. Klin. Wochenschr. 59: 899-903.
Ochs, H . R., Greenblatt, D . J., W o o , E., Smith, T. W . (1978) Reduced quinidine
clearance in elderly persons. Am. J. Cardiol 42: 481-485.
Oginsky, E. L., Stein, A. E., Greer, M . A. (1965) Myrosinase activity in bacteria as
demonstrated by the conversion of progoitrin to goitrin. Proc. Soc. Exp. Biol. Med
119: 360-364.
137
0'Malley, K., Crooks, J., Duke, E., Stevenson, I. H . (1971) Effect of age and sex on
human drug metabolism. Br. Med. J. 3: 607-609.
Overby, L., Nishio, S. J., Lawton, M . P., Plopper, C. G., Philpot, R. M . (1992)
Cellular localization offlavin-containingmonooxygenase in rabbit lung. Exp. Lung
Res. 18: 131-144.
Paint, A . J. (1976) Effect of amino acids and inducers on the activity of the
microsomal mono-oxygenase system in rat liver cell culture. Chem. Biol Interact.
13: 307-315.
Pantuck, E. J., Hsiao, K . C., Loub, W . D., Wattenberg, L. W., Kuntzman, R., Conney,
A . H . (1976) Stimulatory effect of vegetables on intestinal drug metabolism in the
rat. J. Pharmacol Exp. Ther. 198 278-283.
Pantuck, E. J., Hsiao, K . C., Maggio, A., Nakamura, K., Kuntzman, R., Conney, A . H .
(1974) Effect of cigarette smoking on phenacetin metabolism. Clin. Pharmacol
Ther. 15: 9-17.
Pantuck, E. J., Pantuck, C. B., Anderson, K . E., Wattenberg, L. W . (1984) Effect of
brussels sprouts and cabbage on drug conjugations in humans. Clin. Pharmacol
Ther. 35: 161-169.
Pantuck, E. J.，Pantuck, C. B., Garland, W . A., Min, B. H., Wattenberg, L. W.,
Anderson, K . E., Kappas, A., Conney, A. H . (1979) Stimulatory effect of brussels
sprouts and cabbage on human drug metabolism. Clin. Pharmacol Ther. 25: 88-95.
Patterson, S. E., Cohn, V . H . (1984) Hepatic drug metabolism in rats with
experimental chronic renal failure. Biochem. Pharmacol 33: 711-716.
Pearson, A. W., Butler, E. J., Curtis, R. C , Fenwick, G . R., Hobson-Frohock, A.,
Land, D . G., Hall, S. A. (1978) Effect of rapeseed meal on laying hens {Gallus
dmesticus) in relation to fatty liver-haemorrhogic syndrome and egg taint. Res. Vet
Sci. 25: 307-313.
Perlik, F., Janku, I., Jedlicka, J. (1992) Functional and non-functional liver blood
flow in patients with liver cirrhosis measured by indocyanine green. Meth. Find. Exp.
Peterson, F. J.，Holtzmann, J. L. (1980) Drug metabolism in the liver—A
perspective. In Extraphepatic Metabolism of Drugs and Other Foreign Compounds,
Ed.Gram,T.E. Medical and Scientific Books, N e w York, N.Y., U.S.A. 1-122.
138
Pfundstein, B., Tricker, A . R., Preussmann, R. (1991) Determination of primary and
secondary amines in foodstuffs using gas chromatography and chemiluminescence
detection with a modified thermal energy analyser. J. Chromatogr. 539: 141-148.
Phillips, I. R., Dolphin, C. T., Clair, P., Hadleym M . R., Hutt, A.J., McCombie, R. R.,
Smith, R. L,. Shephard，E. A . (1995) The molecular biology of the flavin-containing
monooxygenases of man. Chem. Biol Interact. 96: 17-32.
Pike, M . G., King, G . S., Pettit, B. R., Leonard, J. V., Atherton, D . J. (1988)
Lactulose in trimethylaminuria, the fish-odour syndrome. Helv. Paediat. Acta. 43:
345-348.
Poulsen, L. L., Taylor, K., Williams, D . E., Masters, B. S., Ziegler, D . M . (1986)
Substrate specificity ofthe rabbit lung flavin-containing monooxygenase for amines:
oxidation products ofprimary alkylamines. Mol Pharmacol. 30: 680-685.
Playfer, J. R., Eze, L. C., Bullen, M . F., Evans, D . A . P. (1976) Genetic
polymorphism and interethnic variability of plasma paraoxonase activity. J. Med.
Genet. 13: 337-342.
Prentiss, P. G , Rosen, H., Brown, N., Horwitz, R. E., Malm, 0., Levenson, S. M .
(1961) The metabolism ofcholine by the germ free rat. Arch Biochem. Biophys. 94:
424-429.
Price Evans, D . A. (1992) A^-Acetyltransferase. In Pharmacogenetics of Drug
Metabolism, Ed. Kalow, W . Pergamon Press, N e w York, U.S.A. 95-178.
Quarterman, C. P., Kendall, M . J., Jack, D . B. (1981) The effect of age on the
pharmacokinetics of metoprolol and its metabolites. Br. J. Clin. Pharmacol 11:
287-294.
Renwick, A. G.，Evans, S. P., Sweatman, T. W., Cumberland, J., George, C. F.
(1982) The role of the gut flora in the reduction of sulphinpyrazone in the rat.
Biochem. Pharmacol 31: 2649-2656.
Richey, D., Bender, A. (1977) Pharmacokinetic consequences of aging. Ann. Rev.
Pharmacol Toxicol 17: 49-65.
Roberts, R. K., Wilkinson, G . R., Branch, R. A., Schenker, S. (1978) Effect of age
and parenchymal liver disease on the disposition and elimination of chlordiazepoxide
(Librium). Gastroenterology. 75: 479-485.
Rona, G., Chapel, C, Balazs, T., Gaudry, R. (1959) The effect ofbreed, age, and sex
on myocardial necrosis produced by isoproterenol in the rat. J. Gerontol. 14: 169-
173.
139
Rothschild, J. G., Hansen, R. C. (1985) Fish odor syndrome: trimethylaminuria with
milk as chiefdietary factor. Pediatr. Dermatol 3: 38-39.
Rouer, E., Rouet, P., Delpech，M.，Lerous, J. P. (1988) Purification and comparison
of liver microsomal flavin-containing monooxygenase from normal and
streptozotocin-diabetic rats. Biochem. Pharmacol 37: 3455-3459.
Rubin, P. C., Scott, P. J. W.，Reid, J. L. (1981) Prazosin disposition in young and
elderly subjects. Br. 1 Clin. Pharmacol 12 401-404.
Sadeque，A. J. M., Eddy, A . C.，Meier, G . P., Rettie, A . E. (1992) Stereoselective
sulfoxidation by human flavin-containing monooxygenase. DrugMetab. Dispos. 20:
832-839.
Sardas, S., Akyol, D.，Green, R. L.，Mellon, T., G6kmen, 0., Cholerton, S. (1996)
Trimethylamine A^-oxidation in Turkish women with bacterial vaginosis.
Pharmacogenetics. 6: 459-463.
Scheline, R. R . (1973) Metabolism of foreign compounds by gastrointestinal
microorganisms. Pharmacol Rev. 25: 451-523.
Schmucker, D., Vessey, D., Wang, R., James, J., Maloney, A . (1984) Age-dependent
alternations in the physiochemical properties of rat liver microsomes. Mech. Ageing
Dev. 27: 207-217.
Schmucker, D . L., Woodhouse, K . W., Wang, R. K., Wynne, H., Hjames, 0 . F.,
McManus, M., Kremers, P. (1990) Effects of age and gender on in vitro properties of
human liver microsomal monooxygenases. Clin. Pharmacol Ther. 48: 365-374.
Schmucker, D . L. (1979) Age-related changes in drug disposition. Pharmacol Rev.
30: 445-456.
Schmucker, D . L. (1985) Aging and drug disposition: an update. Pharmacol Rev.
37: 133-148.
Seidegard，J., DePierre, J. W., Pero, R. W . (1985) Hereditary interindividual
difference in the glutathione transferases activity towards trans-stiVoQnQ oxide in
resting human mononuclear leukocytes are due to a particular isozyme(s).
Carcinogenesis. 6: 1211-1216.
Shader, R. I.，Greenblatt, D . J., Ciraulo, D . A., Divoll, M., Harmatz, J. S., Georgotas,
A . (1981) Effect ofage and sex on disposition of desmethyldiazepam formed from in
precursor clorazepate. Psychopharmacology (Berlin). 75: 193-197.
140
Shader, R. I., Greenblatt, D . J., Harmatz, J. S., Franke, K., Koch-Weser, J. (1977)
Absorption and disposition ofchlordiazepoxide in young and elderly male volunteers.
J. Clin. Pharmacol 17: 709-718.
Shepherd, A . M . M., Hewick, D . S., Moreland, T. A.，Stevenson, 1. H . (1977) Age as
a determinant ofsensitivity to warfarin. Br. J. Clin. Pharmacol 4: 315-320.
Shephard, E. A., Dolphin, C. T., Fox, M . F., Povey，S., Smith, R., Phillips, I. R.
(1993) Localization of genes encoding three distinct flavin-containing
monooxygenases to human chromosome lq. Genomics. 16: 85-89
Sherwin, A . L., Loynd, J. S., Bock, G. W., Sokolowski, C. D . (1974) Effects of age,
sex, obesity, pregnancy on plasma diphenylhydantoin levels. Epilepsia. 15: 507-
521.
Shewan, J. M . (1951) The chemistry and metabolism of the nitrogenous extracts in
fish. Biochem. Soc. Symp. 6: 28-48.
Shimada, M., Murayama, N., Yamazoe, Y., Hashimoto, H., Ishikawa, H., Kato, R.
(1995) Age- and sex-related alternations of microsomal drug- and testosterone-
oxidizing cytochrome P450 in Spragu-Dawley strain-derived dwarf rats. J.
Pharmacol Exp. Ther. 275: 972-977.
Shull, H . J. Jr., Wilkinson, G . R., Johnson, R.，Schenker, S. (1976) Normal
disposition ofoxazepam in acute viral hepatitis and cirrhosis. Arm. Intern. Med. 84:
420-425.
Sidransky, H . (1986) Effects of tryptophan on protein synthesis by liver. In
Nutritional Diseases: Research Directions in Comparative Pathobiology, Eds.
Scarpelli, D . G., Migaki，G. Alan R. Liss, Inc., N e w York, N Y , U.S.A. 81-90.
SimenhofF, M . L., Burke, J. F, Sankkonen, J. J., Ordinario, A . T., Doty, R. (1977)
Biochemical profile ofuremic breath. N. Eng. J. Med 297: 132-135.
Skoda, R. C., Gonzalez, F. J., Demierre, A., Meyer, U . A . (1988) T w o mutant alleles
of the human cytochrome P-450dbl gene (P-450C2D1) associated with genetically
deficient metabolism of debrisoquine and other drugs. Proc. Natl. Acad. Sci. U.S.A.
85: 5240-5243.
Smith, M.，Hopkinson, D . A., Harris, H . (1971) Developmental changes and
polymorphism in human alcohol dehydrogenase. Ann. Hum. Genet. 34: 251-271.
Smith, M., Hopkinson, D . A., Harris, H . (1971) Alcohol dehydrogenase isozyme in
adult human stomach and liver: evidence for activity of the ADH3 locus. Ann. Hum.
Genet. 35: 243-253.
141
Smolen，A., Eckerson, H . W . ， L a Du, B . N . (1982) Purification of arylesterases A
and B from human serum. Am. J, Hum. Genet. 34: 63A.
Sowunmi, A., Rashid，T. J., Akinyinka, 0 . 0., Renwick，A. G . (1995) Ethnic
differences in nifedipine kinetics: comparisons between Nigerians, Caucasians and
south Asians. Br. J. Clin. Pharmacol 40: 489-493.
Stamatoyannopoulos, G., Chen, S. H , Fukui, F. (1975) Liver alcohol dehydrogenase
in Japanese: high population frequency of atypical form and its possible role in
alcohol sensitivity. Am. J. Hum. Genet. 27: 789-796.
Stern, R., Abel, R., Gibson, G . L., Besserer，J. (1997) Atorvastatin does not alter the
anticoagulant activity of warfarin. J. Clin. Pharmacol 37: 1062-1064.
Strange, R. C., Faulder, C. G., Davis, B. A., H u m e , D . R., Brown, J. A . H.，Cotton,
W., Hopkinson, D . A . (1983) The human glutathione ^-transferases; studies on the
tissue distribution and genetic variation of the GST1, G S T 2 and G S T 3 isozymes.
Ann. Hum. Genet 48: 11-20.
Sunahara, S.，Urano, M., Ogawa, M.，Yoshida, S., Mukoyama, H., Kawai, K . (1963)
Genetical aspect ofisoniazid metabolism. Jap. J. Hum. Genet. 8: 93-111.
Swanson，B. N., Boppana, V . K . (1981) Measurement of sulindac and its metabolites
in human plasma and urine by high-performance liquid chromatography.
J. Chromatogr. 225: 123-130.
Szepesi, G . (1990) H P L C in Pharmaceutical Analysis. Vol. I & II. C R C Press, Boca
Raton, FL, U.S.A.
Terner, U . K., Wiebe, L. I., Noujaim, A . A., Dssetor, J. B., Sanders, E. J. (1978) The
effects of acute and chronic uremia in rats on their hepatic microsomal enzyme
activity. Clin. Biochem. 11: 156-158.
Thithapandha, A . (1997) A pharmacogenetic study of trimethylaminuria in orientals.
Pharmacogenetics. 7: 497-501.
Tjoa，S., Fennessey, P. (1991) The identification of trimethylamine excess in man:
quantitative analysis and biochemical origins. Anal Biochem. 197: 77-82.
Todd, W . A . (1979) Psychosocial problems as the major complication of an
adolescent with trimethylaminuria. J. Pediat. 94: 936-937.
Touchette, M . A., Slaughter, R. L. (1991) The effect of renal failure on hepatic drug
clearance. DICP. 25: 1214-1224.
Treacy, E. P., Akerman, B. R.，Chow, L. M . L.，Youil, R., Bibeau, C., Lin, J., Bruce,
A . G., Knight, M., Danks, D . M., Cashman, J. R., Forrest, S. M . (1998) Mutations of
142
the flavin-containing monooxygenase gene {FMOS) cause trimethylaminuria, a defect
in detoxication. Hum.Mol.Genet. 7: 839-845.
Triggs, E., Nation, R., Long, A.，Ashley, J. J. (1975) Pharmacokinetics in the elderly.
Eur. 1 Clin. Pharmacol 8: 55-62.
Triggs, E., Nation, R. (1975) Pharmacokinetics in the aged: a review. J.
Pharmacokinet. Biopharm. 3: 387-418.
Tschudy，D. P. (1978) The influence of hormonal and nutritional factors on the
regulation of liver heme biosynthesis. In Handbook ofExperimental Pharmacology:
Heme and Hemoproteins, Vol. 44, Eds. D e Matteis, F., Norman, A . W., Springer-
Verlag, Berlin, Germany. 255-271.
Vestal, R . E., Dawson, G . W . (1985) Pharmacology in aging. In Handbook ofthe
Biology ofAging, Eds. Finch, S. E., Schneider, E. L. 2"^ ed., Van Nostrand Reinhold，
N e w York, N Y , U.S.A. 744-819.
Vestal, R. E., McGuire, E. A., Tobin, J. D., Andres, R., Norris, A . H., Mezey, E.
(1977) Aging and ethanol metabolism. Clin. Pharmacol Ther. 21: 343-354.
Vestal, R. E., Norris, A . H., Tobin, J. D., Cohen, B. H., Shock, N . W., Andres, R.
(1975) Antipyrine metabolism in man: influence of age, alcohol, caffeine, and
smoking. Clin. Pharmacol Ther. 18: 425-432.
Vestal, R. E., W o o d , A . J. J., Branch, R. A., Shand, D . G., Wilkinson, G . R. (1979)
Effect of age and cigarette smoking on propranolol disposition. Clin. Pharmacol
Ther. 26: 8-15.
Vestal, R . F. (1982) Pharmacology and aging. J. Am. Geriatr. Soc. 30: 191-200.
Vogel, F. (1959) Moderne Probleme der Humangenetik. Ergeb. Inn. Med.
Kinderheilk 12: 52-125.
V o n Wartburg, J. P., Schurch, P. M . (1968) Atypical human liver alcohol
dehydrogenase. Ann. N. Y. Acad. Sci. 151: 936-946.
Wang, J. S., Wang, W., Xie, H . G., Huang, S. L., Zhou, H . H . (1997) Effect of
troleandomycin on the pharmacokinetics of imipramine in Chinese: the role of
C Y P 3 A . Br. J. Clin. Pharmacol 44: 195-198.
Warholm, M., Guthenberg, C., Mannervik, B., V o n Bahr, C. (1997) Purification of a
new glutathione 5"-transferase (transferase //) from human liver having high activity
with benzo[a]pyrene-4,5-oxide. Biochem. Biophys. Res. Commun. 98: 512-519.
W e b e r ， W . W . (1997) Pharmacogenetics, Oxford University Press, N e w York,
N.Y.,U.S.A. 93-122.
143
Welling, P. G . (1977) Influence offood and diet on gastrointestinal drug absorption:
a review. J. Pharmacokinet. Biopharm. 5: 291-334.
Westphal, J. F., Brogard, J. M . (1993) Clinical pharmacokinetics of newer
antibacterial agents in liver disease. Clin. Pharmacokinet. 24: 46-58.
Whittaker, M . (1986) Cholinesterase. Monographs in Human Genetics, Vol. 11, Ed.
Beckman, L. Karger, Basel, Switzerland.
Whittaker, J. A.，Evans, D . A . P. (1970) Genetic control of phenylbutazone
metabolism in man. Br. Med. J, 4: 323-328.
Wills, M . R., Savory, J. (1981) Biochemistry of renal failure. Ann. Clin. Lab. Sci.
11: 292-299.
W o o d , A . J. J., Vestal, A . E., Wilkinson, G . R., Branch, R. A., Shand, D . G . (1979)
Effect of aging and cigarette smoking on antipyrine and indocyanine green
elimination. Clin. Pharmacol Ther. 26: 16-20.
Wrighton, S. A., Vandenbranden, M., Stevens, J. C., Shipley, L. A., Ring, B. J.,
Rettie, A . E., Cashman, J. R. (1993) In vitro methods for assessing human hepatic
drug metabolism: their use in drug development. DrugMetab. Rev. 25: 453-484.
Wranne, L. (1956) Urinary excretion of trimethylamine and trimethylamine oxide
following trimethylamine administration to normal and patients with liver disease.
ActaMed Scand. 153: 433-441.
Wynne, H., Mutch, E., James, 0 . F. W., Rawlins, M . D., Woodhouse，K. W . (1987)
The effect of age on monooxygenase enzyme kinetics in rat liver microsomes. Age
Ageing. 16: 153-158.
Wynne, H . A., Mutch, E.，James, 0. F., Wright, P., Rawlins, M . D., Woodhouse, K .
W . (1988) The effect of age upon the affinity of microsomal mono-oxygenase
enzymes for substrate in human liver. Age Ageing. 17: 401-405.
Xie, H . G., Xu, Z H., Luo, X ’ Huang, S. L., Zeng, F. D., Zhou, H . H . (1996)
Genetic polymorphisms of debrisoquine and iS'-mephenytoin oxidation metabolism in
Chinese populations: a meta-analysis. Pharmacogenetics. 6: 235-238.
Zhang, A . Q., Mitchell, S. C., Ayesh, R., Smith, R. L. (1992) Determination of
trimethylamine and related aliphatic amines in human urine by head-space gas
chromatography. J. Chromatogr. 584: 141-145.
Zhang, A . Q., Mitchell, S., Smith, R. (1995) Fish odour syndrome: Verification of
carrier detection test. J. Inher. Metah. Dis. 18: 669-674.
144
Zhang, A. Q., Mitchell, S. C., Smith, R. L. (1996a) Discontinuous distribution ofN-
oxidation ofdietary-derived trimethylamine in a British population. Xenobiotica. 26:
957-961.
Zhang, A. Q.，Mitchell, S. C., Smith, R. L. (1996b) Exacerbation of symptoms of
fish-odour syndrome during menstruation. Lancet. 348: 1740-1741.
Zhou, H . H., Koshakji, R. P., Silbertein，D. J., Wilkinson, G . R., W o o d , A . J. J.
(1989) Racial differences in drug response: Altered sensitivity to and clearance of
propranolol in m e n of Chinese descent as compared with American Whites. N. Eng.
J.Med. 320: 565-570.
Ziegler, D . M . (1980) Microsomal flavin-containing monooxygenase: oxygenation
of nucleophilic nitrogen and sulphur compounds. In Enzymatyic Basis of
Detoxification, Ed. Jakoby, W . B. Academic Press, N e w York, U.S.A. 201-227.
Ziegler, D . M., Mitchell, C. H . (1972) Microsomal oxidase. IV Properties of a
mixed ftmction amine oxidase isolated from pig liver microsomes. Arch Biochem.
Biophys. 150: 116-125.
Ziegler, D . M., Pettit, F. H . (1964) Formation of an intermediate #-oxide in the
oxidative demethylation of A^,A^-dimethylaniline catalyzed by liver microsomes.
Biochem. Biophys. Res. Commun. 15: 188-193.
145
Appendix A.1: Sample record sheet for food intake and activitv (English)
Activity/Meal/Drink Record Sheet
Name of Volunteer: Age:
Address:
Phone/Pager:
Period ofRecord (From): (To):
Are you smoker?# Yes No

If 'Yes,, please indicate how many cigarettes you smoke per day:
Important Notice:
Please start to recordyour activities/meals/drinks from the dinnerjust before the experiment
begins (12:00 a,nt, midnight). For example, ifyou have dinner at 8:00p,nt, 1st May, 1996，
please start to recordyour activities/meals/drinks from 8:00 p.m., 1st May, 1996, but urine
collecting activity is still begins from this midnight (12:00 a,m, 2nd May, 1996).
Description of
Date |^工騰 AM./PM. Activity Meals/Drinks/Medication Approx. Amount Taken*
(hh:mm) Taken/Exercise Performed
P.M. Dinner
#please V the appropriate

*e.g. a cup oftea « 100 ml; a glass of water ~ 200 ml; abowl of soup « 200 ml; abowl ofrice « 200 g
( fP.2)
A-1
Description of
Date 丁血® A.M.fPM. Activity MealsA)rinksMedication Approx. Amount Taken*
恤讓） Taken/Exercise Performed
*e.g. a cup oftea « 100 ml; a glass of water ~ 200 ml; a bowl of soup « 200 ml; a bowl of rice « 200 g
Name of Analyst:
Signature ofVolunteer: Signature of Analyst:
Date: Date of Analysis:
A-2
Appendix A.2: Sample record sheet for fnnd intake and activity (Chinese)
活動，進食及飮料紀錄
參加者姓名：年齡：
地址：
聯絡電話/傳呼機：
記錄時間油)：（至)：
你是否吸煙人士?# 是__ 否 —
如果‘是’，請說出你每天吸多少枝煙：
mmmm：
請由實驗前(深夜十二時正)的晚飯開始記錄你所有的活動及飮食資料o舉例’如果你的
晚飯時間是在95年5月1日’晚上八時_舌，請於翻(95年5月1日晚上八時)開始
識尔所有的活動及飮食餓o但請謹記,小便樣本仍是在當晚深夜開始收集o (95年5
月2日’深夜十二時）。
曰期（ S n ) ^ / 活動（働/飮料^勿/運動等）大繊量/份量
下午. 晚飯
#請在適當位置加上“號
* - f m « 100毫升;一玻璃杯水« 200毫升；一碗湯« 200毫升；一碗飯« 200克
( /P.2)
A-3
• tH 時間上午/ 、法動簡述大約數量/份量
曰期（hh:mm) 下午估勤（食物/飮料/藥物/運動等）入彳」数
* -iff^ « 100毫升;一玻璃杯水« 200毫升；一碗湯« 200毫升；一碗飯« 200克
分析員姓名：
參加者簽名：分析員縛
日期：分析曰期：
A-4
Appendix B: Food intake and activity rernrd for volunteer A in COntrpl diet
experiment
Date Time (AMyTM) Activity Food Intake 一 Amount

16/8/97~~8:00 PM Dinner Rice lBowl
Shrimp 300 g
Garoupa 190 g
Broccoli LitUe bit
Beer (Blue Ribbon) 355 ml
Water 1 Glass
^ : 4 5 PM Urination Sample 1 380 ml

TO:OOPM Water ^ Glass
17/8/97 l ^ : 1 5 A M Urination — Sample 2 190 ml
T 2 : 3 0 AM Bedtime
7:45 AM Wake up & Sample 3 600 ml
Urination
一 —Chinese Tea % Glass
8:30 AM Breakfast Rice Roll ~~ 1 Dish (-200 g)
Chinese Tea 1 Glass
T 0 : 3 0 AM Urination Sample 4 350 ml
一 —Chinese Tea 乂 Glass
11:15 AM " c W e s e Tea % Glass
T2:30 PM Urination Sample 5 380 ml
12:45 PM Lunch Pork Steak 300 g
Rice 1 Bowl
Bitter Gourd Little Bit
Cream Soda 355 ml
Chinese Tea 么 Glass
T : 0 0 PM — Urination Sample 6 360 ml
—Water 1/2 Glass
~6:00 PM Urination Sample 7 360 ml
7:00 PM Chinese Tea 1 Glass
8:00 PM Urination Sample 8 l^Q^
8:30 PM Dinner Beef 600 g
Rice /2 Bowl
8:45 PM 一 Water 1/2 Glass
ll:OOPM Urination Sample 9 360 ml
11:10 PM Water V2 Glass
l8/8/97l:lOAM Urination Sample 10 120 ml
^ 1:15 AM — Water 1/2 Glass
1:30 AM —— Bedtime
Urination
8:50 AM Chinese Tea 1 Glass
9:45 AM Breakfast Congee (w/ 70 g pork, some edible 3 Bowl
flowers, ginger)
10:00 AM — Chinese Tea /2 Glass
11:15 AM Urination & Sample 12 560 ml
Defeacation
12:00 PM Chinese Tea 1/2 Glass
~l ： 10 PM Urination Sample 13 160 ml
1:15 PM Lunch Lettuce 75g
Apple 1
Rice 1 Bowl
Coke 355 ml
B-1
2:30 PM ChineseTea '/2Glass
"^:00 PM Urination Sample 14 540 ml
7:15 PM Soup (w/ tomato, chicken's liver, 2 Bowl
kidney)
T"55PM Urination — Sample 15 110 ml
8:00 PM Dinner Chicken's Meat 150g
Rice % Bowl
l0:45PM Urination Sample 16 — 300 ml
H:30PM — Water % Glass
ll:55PM Urination 一 60 ml
19/9/97 l2:OOAM — Bedtime —
Urination
B-2
171
ttfti
1
: ^
m
m
m
I
i通
i
I
i
资
ff
j
'.5
1¾
^1
；
.
mmuwi
CUHK L i b r a r i e s
0037D515^

Trimethylamine 7V-Oxidation in Chinese: Lee, Chi-Wai

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Trimethylamine 7V-Oxidation in Chinese: Lee, Chi-Wai

Uploaded by

Copyright:

Available Formats

Trimethylamine 7V-Oxidation in

B.Sc. (Hons), Cert. Pharmacol. & Pharm. Mgt.

• The Chinese University ofHong Kong 1999

I wish to express m y sincere gratitude to m y supervisors Profs. Jim L. A. Damani,

during the course of this study.

Inter-ethnic differences in drug metabolism have been reported in numerous studies,

List ofTables xiv

List of Abbreviations xvi

1.1 Overview of pharmacogenetics 2

1.2 Clinical importance of nitrogen oxidations 8

1.3 Characteristics of trimethylamine ( T M A ) 9

1.4 Genetic polymorphism of trimethylamine 7V-oxidation 12

1.5 Enzyme systems mediating trimethylamine A^-oxidation 15

1.6 Aims and objectives 19

Chapter 2 Development of an Analytical Method for Trimethylamine

2.3.1 Preparation of stock solutions 27

2.3.2 Preparation of G L C glass column 28

2.3.4 Sample pretreatment for G L C analysis 29

2.3.5 Construction of calibration curve 31

2.3.7 Intra- and inter-assay variations 31

2.3.2 Construction of calibration curve 33

2.3.4 Intra- and inter-assay variations 37

Chapter 3 Trimethylamine A^-Oxidation in Chinese 41

3.2 Experimental protocols 43

3.2.1 Whole day urine collections 43

3.2.2 Spot urine collections 44

3.3.1 Whole day urine collections 44

3.3.2 Spot urine collections 46

3.3.4 Comparison between smokers and non-smokers 52

Chapter 4 Effect ofAge and Diet on Trimethylamine A^-oxidation 64

4.2 Experimental protocols 73

4.2.1 Effect of age on T M A A^-oxidation 73

4.2.2 Effect of diet on T M A 7V-oxidation 73

4.2.3 Effects of control diet on T M A 7V-oxidation 75

4.3.1 Effect of age on T M A A^-oxidation 76

4.3.2 Effect of diet on T M A 7V-oxidation 80

4.3.3 Effects of control diet on T M A A^-oxidation 83

Chapter 5 Effect ofDisease on Trimethylamine TV-Oxidation 101

5.1 Introduction 102

5.2 Experimental protocols 108

5.3 Results 109

5.4 Discussion 116

Chapter 6 General Discussion 119

A.2 Sample record sheet for food intake and activity

Food intake and activity record for volunteer A in

Figure 1.1 The conversion of trimethylamine (TMA) to 9

Figure 1.2 Some drugs which undergo nitrogen oxidations by F M O . 10

Figure 1.3 Precursors of trimethylamine in foods. 11

Figure 2.1 Block diagram of a gas chromatograph. 25

Figure 2.2 Typical headspace G L C chromatogram of T M A 34

Figure 2.3 Calibration curve for T M A determination. 35

Figure 2.4 Determination of the amount of titanium (III) chloride 36

Figure 2.5 Intra-assay variation for T M A and total T M A + T M A O 38

Figure 2.6 Inter-assay variation for T M A and total T M A + T M A O 39

Figure 3.1 Frequency distribution histogram for the percentage total 47

Figure 3.2 Frequency distribution histogram for the percentage total 49

Figure 3.3 Frequency distribution histogram for the percentage total 51

Figure 3.4 Amounts of T M A and T M A O (as T M A ) excreted in 54

Figure 3.5 Percentage T M A excreted as T M A O (calculated as 55

Figure 4.1 Inhibitors of the flavin-containing monooxygenase. 12

Figure 4.2 Excreted T M A amounts in different periods and in 11

Figure 4.3 Excreted T M A O amounts (as T M A ) in different periods 78

Figure 4.4 Percentage T M A excreted as T M A O in urine (combined 79