ArticlePDF Available

Abstract and Figures

Purinergic receptors are important for the regulation of inflammation, muscle contraction, neurotransmission and nociception. Extracellular ATP and its metabolites are the main ligands for these receptors. Occasional reports on beneficial results of ATP administration in human and animal studies have suggested the bioavailability of oral ATP supplements. We investigated whether prolonged daily intake of oral ATP is indeed bioavailable. Thirty-two healthy subjects were randomised to receive 0, 250, 1250 or 5000 mg ATP per d for 28 d by means of enteric-coated pellets. In addition, on days 0 and 28, all thirty-two subjects received 5000 mg ATP to determine whether prolonged administration would induce adaptations in the bioavailability of ATP. ATP supplementation for 4 weeks did not lead to changes in blood or plasma ATP concentrations. Of all ATP metabolites, only plasma uric acid levels increased significantly after the administration of 5000 mg of ATP. Prolonged administration of ATP was safe as evidenced from liver and kidney parameters. We conclude that oral administration of ATP only resulted in increased uric acid concentrations. On the basis of these findings, we seriously question the claimed efficacy of oral ATP at dosages even lower than that used in the present study.
Content may be subject to copyright.
Kinetics and effects of orally
administered ATP
This PhD project was supported by the Graduate School VLAG (Food Technology,
Agrobiology, Nutrition and Health Sciences), accredited by the Royal Netherlands
Academy of Arts and Sciences. The studies presented in this thesis were conducted
at the Department of Epidemiology and the Department of Pharmacology & Toxico-
logy, Maastricht University, Nutrition and Toxicology Research Institute Maastricht,
The Netherlands, in close collaboration with the Laboratory for Physical Chemistry
and Colloid Science, Wageningen University.
Kinetics and effects of orally administered ATP
Thesis, Maastricht University, Maastricht, The Netherlands
ISBN 978 90 86595501
© Copyright Erik Coolen, Maastricht 2011
Design and layout: Erik Coolen
Printed by: Drukkerij Wilco
All rights reserved. No part of this thesis may be reproduced or transmitted in any
form or by any means, electronic or mechanical, including photocopying, recording
or any information storage or retrieval system without permission in writing from
the author, or, when appropriate, from the publisher of the publications.
Financial support for printing of this thesis has kindly been provided by
'Fagron BV',
‘Greiner Bio-one’ and
‘Sanquin Blood Supply, Unit Reagents.’
Kinetics and effects of orally administered ATP
PROEFSCHRIFT
ter verkrijging van de graad van doctor aan de Universiteit Maastricht,
op gezag van de Rector Magnificus,
Prof. mr. G.P.M.F. Mols,
volgens het besluit van het College van Decanen,
in het openbaar te verdedigen
op woensdag 11 mei 2011 om 14.00 uur
door
Johannes Cornelis Maria Coolen
Geboren op 26 april 1981 te Tilburg
Promotores
Prof. dr. A. Bast
Prof. dr. M.A. Cohen Stuart, Wageningen University, Wageningen
Copromotores
Dr. ir. P.C. Dagnelie
Dr. ir. I.C.W. Arts
Beoordelingscommissie
Prof. dr. S. van der Linden (voorzitter)
Prof. dr. R.-J. Brummer (Örebro University, Zweden)
Dr. ir. P.C.H. Hollman (RIKILT – Instituut voor Voedselveiligheid, Wageningen)
Prof. dr. ir. C.P. van Schayck
Prof. dr. F.-J. van Schooten
Chapter 1
5
Contents
Chapter 1 General introduction and aims of the thesis 7
Chapter 2 Simultaneous determination of adenosine triphosphate and
its metabolites in human whole blood by RP-HPLC and
UV-detection
25
Chapter 3 Oral supplements of adenosine 5’-triphosphate (ATP) are
not bioavailable: a randomized, placebo-controlled cross-
over study in healthy humans
41
Chapter 4 Oral bioavailability of ATP after prolonged administration 59
Chapter 5 Time-dependent effects of ATP and its degradation prod-
ucts on inflammatory markers in human blood ex vivo
75
Chapter 6 A comparison of ex vivo cytokine production and antioxi-
dant capacity in whole blood after incubation with uric acid
or monosodium urate crystals
89
Chapter 7 General discussion and conclusions 107
Chapter 8 Summary 121
Chapter 9 Samenvatting 125
Dankwoord 129
Curriculum vitae 131
List of publications 132
Chapter 1
7
Chapter 1
General Introduction
7
General introduction
8
Nucleotides
Nucleotides are compounds that consist of a base, a sugar (either ribose or
deoxyribose), and one to three phosphate groups. They have numerous essential
functions in the mammalian body.
Nucleotides form the building blocks of DNA and RNA (figure 1), making
them vital for encoding all genetic information in the human body. Nucleotides are
also involved in metabolism, cellular signalling (both intra- and extracellular) and
in various cofactors of enzymatic reactions, as will be exemplified below.
One of the best known nucleotides is adenosine triphosphate (ATP) (figure 1).
It can be produced endogenously from glucose through glycolysis, the citric acid
cycle, and oxidative phosphorylation in mitochondria.
Once ATP is produced in the mitochondria, it is transported across the outer
mitochondrial membrane, primarily by voltage-dependent anion channels
(VDAC)(1). ATP can also be salvaged after it is released from cells or it can come
from an exogenous dietary source. The concentration in the cell ranges from 1-10
mM. ATP plays a key role as a carrier of chemical energy. The bounds between
phosphate groups are responsible for the high energy content of ATP. The energy

-30,5 kJ/mol) can for instance be used for synthesis of organic molecules and
transmembrane transport through ATP-binding cassette transporters (ABC-trans-
porters).
The use of ATP as a carrier of chemical energy is very common in eukaryotes.
The incorporation of mitochondria into eukaryotic cells is believed to have occurred
very early in evolution, since it provided cells with the possibility of aerobic respi-
ration with ATP as chemical energy carrier(2).
Next to playing a role in chemical energy supply, ATP is also involved in
intracellular signal transduction, for instance after it is converted to cyclic AMP
(cAMP) by the enzyme adenylyl cyclase, located on the inner side of the plasma
membrane(3). Also several cofactors are derived from ATP, like coenzyme A (CoA),
and the electron acceptors flavin adenine dinucleotide (FAD) and nicotinamide
adenine dinucleotide (NAD), both of which can be used to transfer electrons from
one molecule to another(4).
Besides its intracellular functions, ATP is also well known for its extracellu-
lar functions. Although the concentrations outside the cell are normally below the
micromolar range (between 10 and 100 nM(5)), extracellular ATP influences many
biological processes, including neurotransmission, muscle contraction, inflamma-
tion and cardiac function(6, 7). Whereas intracellular concentrations are maintained
at high levels, ATP released from cells into the extracellular compartment is
tightly regulated and its concentration is kept very low by ecto-enzymes(8-10).
Chapter 1
9
Extracellular ATP degradation
Ecto-enzymes are divided over four families. Adenosine diphosphate (ADP)
and adenosine monophosphate (AMP) are the products that are formed when one
and two phosphates are cleaved off, respectively, by ectonucleoside triphosphate
diphosphohydrolases (NTPDases) (figure 2)(11). Next, ectonucleotide pyrophos-
phate/phosphodiesterase (NPP) catalyzes the hydrolysis of ADP to AMP and of
AMP to adenosine. Alkaline phosphatases (AP) can catalyze the hydrolysis of ATP,
ADP and AMP. The degradation of AMP to adenosine is also done by ecto-5’-nuc-
leotidase (CD73)(12). Adenosine can be broken down further to inosine by adenosine
deaminase (ADA) or to adenine by a purine nucleoside phosphorylase. In the final
stages of purine metabolism, the enzyme xanthine oxidase (XO) is responsible for
the degradation of hypoxanthine and xanthine to the end-product uric acid(13).
Functions of extracellular purinergic metabolites
All of the various degradation products have their own functions, because
they are ligands for different membrane-bound purinergic receptors. Transient
increases in extracellular ATP are vital for cell-to-cell communication in nervous,
Figure 1: Nucleotides consist of a nitrogenous base, a pentose sugar, and one or more phosphate groups.
The nitrogenous base is either a purine or a pyrimidine. Pyrimidine bases are six-membered rings, and
include uracil, cytosine (C) and thymine (T). Purine bases have a second five-membered ring, and include
adenine (A), guanine (G), hypoxanthine and xanthine. A purine or pyrimidine base linked to a pentose
molecule constitutes a nucleoside. A nucleotide is a phosphate ester of a nucleoside, and may occur in
the mono, di- or triphosphate form. The pentose is either ribose or deoxyribose; the ribonucleotides and
deoryribosenucleotides serve as the monomeric units of RNA and DNA, respectively.
General introduction
10
vascular and immune systems(14, 15). In damaged tissues, ATP is a natural endogen-
ous adjuvant which is released from activated immune cells(16), macrophages(17),
microglia(18), and platelets(19) that initiates inflammation, and further amplifies
and sustains cell-mediated immunity through P2 receptor-mediated purinergic
signalling(20).
Purinergic receptors
Purinergic signalling is rapidly becoming recognised as essential for the reg-
ulation of tissue and organ function. Drury & Szent Györgyi discovered the vaso-
dilatory effect of extracellular purines in 1929(21). First, observational studies
helped to develop theories on purinergic signalling. Later, Geoffrey Burnstock pro-
posed extracellular ATP to be a neurotransmitter conducting noradrenergic and
noncholinergic neurotransmission of the gut and urinary bladder in 1972(22). Now,
purinergic receptors have been detected in virtually every type of cell in every type
of tissue, leading to an exponential increase in the number of articles on purines(23).
Purinergic signalling was found to have appeared early in evolution(24) and to be
Figure 2: Metabolism of ATP to uric acid. Numbers in italics represent: 1, ecto-nucleoside triphosphate
diphosphohydrolase (NTPDase) 1; 2, NTPDase 2; 3, NTPDase 3; 4, NTPDase 8; 5, alkaline phosphtase; 6,
ectonucleotide pyrophosphatase/phosphodiesterase (NPP) 1; 7, NPP2; 8, ecto-5’-nucleotidase (CD73); 9,
Adenosine deaminase (ADA); 10, purine nucleoside phosphorylase (PNP) and 11, xanthine osidase (XO).
Chapter 1
11
widespread in most non-neuronal and neuronal cell types(25). Also the recognition
that there is both long-term purinergic signalling in cell-proliferation, differentia-
tion, development, and regeneration and short-term purinergic signalling in neuro-
transmission and secretion(26, 27) has helped the field of purinergic signalling to es-
tablish itself as an important part of various signalling routes.
Purinergic receptors are classified in two families, namely P1 and P2 recep-
tors. P2 receptors have ATP and other nucleotides as their primary ligands, whe-
reas adenosine is the ligand for P1 receptors. The P2 family can be divided into two
subfamilies, P2Y and P2X(25). P2Y receptors are G protein-coupled receptors, of
which twelve subtypes have been described (P2Y1. P2Y2, P2Y4-6 and P2Y8-14(28-30).
P2X receptors are ATP-gated ion channels of which seven subtypes have been
identified (P2X1-7)(31-33). P1 receptors are G-protein coupled receptors which are ac-
tivated by adenosine and are subdivided into A1, A2A, A2B and A3 receptor sub-
types(34, 35). Signal transduction through P2Y receptors occurs via Gq proteins acti-
vating phospholipase C (PLC), which for most receptors induces the release of in-
ositol triphosphate (IP3) and thereby results in mobilization of Ca2+ from intracel-
lular stores(36).
When P2X receptors are activated by ATP, sodium will be transported into
the cell, whereas potassium and calcium will be transported out of the cell through
the formed ion channel. This results in activation of various signalling mole-
cules(37), such as extracellular-signal regulated kinase (ERK), which in turn stimu-
   (38)         
secretion of cytokines and is associated with the development of several diseases,
such as inflammatory arthritis, lung fibrosis, cancer, diabetes and stroke(39).
When LPS-stimulated monocyte-derived dendritic cells are constitutively ex-
posed to low ATP concentrations (100-200 µM), this will lead to inhibition of the
cytokines IL-12, TNF---6(40); P2Y11 receptors are thought to partici-
pate in this reaction. Also other in vitro studies, using for instance macrophages,
have demonstrated the inhibition of inflammatory cytokine release(5, 41, 42). In whole
blood, addition of ATP leads to reduction in release of TNF--
10, thus inhibiting inflammation(43, 44).
Intravenous ATP administration
Several human intervention trials with advance stage cancer patients have
been performed to investigate whether ATP infusion can exert favourable effects.
The main clinical symptoms that were investigated were those that are associated
with the cancer cachexia syndrome: decreased food intake, progressive involuntary
weight loss with depletion of lean body mass, and an impaired immune re-
sponse(45). Cancer-cachexia syndrome has a prevalence of up to 80% in patients
with advanced cancer(46). The reported beneficial effects of ATP infusion ranged
from the prevention of deterioration of muscle mass, functional performance and
General introduction
12
weight(47, 48), to lowering pain and increasing life expectancy(49). In a randomized
clinical trial conducted by Agteresch et al.(47, 48, 50), patients with stage IIIb/IV non-
small cell lung carcinoma (NSCLC) received ATP, up to 75 µg/kg/min (as tolerated)
for 30 hours intravenously at 2- to 4-week intervals. Increased concentrations of
ATP were observed in the blood plasma of these patients following intravenous
infusion. Leij-Halfwerk et al.(51) reported that the ATP concentrations in the livers
of patients with NSCLC were initially lower than those of healthy volunteers, but
recovered back to normal concentrations after intravenous ATP infusions.
Intravenous infusion is, however, quite a burden in particular for patients
with advanced cancer, but also the need for medical supervision and the accompa-
nying high costs are important drawbacks of the intravenous administration me-
thod. Moreover, intravenous administration is limited to infusion rates of approx-
imately 75-100 µg/kg/min to avoid the occurrence of side effects(47, 52). As an alter-
native to infusion, the oral administration of ATP might yield similar effectiveness,
but less side-effects.
Oral route of administration of ATP
ATP is present in substantial concentrations in a number of foods (e.g. meat,
soy, mushroom) and in breast milk. In breast milk, nucleotides are reported to en-
hance the gastrointestinal and immune systems(53-55). Types of food that are rich
sources of dietary nucleotides include organ meat like liver, heart and kidney,
fresh seafood, beer, and some types of beans(54). The amount of nucleotides in an
average meal is by far insufficient to represent the amount that is needed in the
body. In a purine-restricted diet, purines are often restricted to 100-150 milligrams
per day. In normal diets, the amount of nucleotides ingested will be between 500
and 1000 mg/day, but they can reach up to about 5000 mg/day. When ATP is in-
gested with food, there are a number of challenges it has to face before being of use
to the human body. First, the large molecular weight, the negative charge at phy-
siological pH (pH 7.4), and the lack of known nucleotide transporters makes both
passive and active transport of intact ATP over outer cell membranes highly un-
likely. Second, ecto-nucleoside triphosphate diphosphohydrolases (NTPDases)
present on the luminal side of enterocytes will dephosphorylate ATP via ADP to
AMP. AMP, in turn, will be subject to further degradation as ecto-5’-nucleotidase
(CD73) and alkaline phosphatase will degrade it to adenosine(56). For the effective
use of ATP, ATP does not need to be absorbed intact, since adenosine can be taken
up into the enterocytes of the intestinal wall. This occurs through concentrative
(CNT) or equilibrative (ENT) nucleoside transporters(57) on the baso-lateral side of
enterocytes. Provided that adenosine is released intact into the vascular bed, ade-
nosine will come into contact with erythrocytes, which will take up adenosine
through ENTs in their membranes(58, 59). In vivo studies in animals and humans
Chapter 1
13
have shown that adenosine can be used for the synthesis of ATP inside erythro-
cytes.
In humans, there has been no thorough investigation as to whether adeno-
sine originating from intestinal uptake can also be used for ATP production in
erythrocytes. Given adenosine’s own role as an important extracellular signaling
molecule (it facilitates a variety of physiological responses, including coronary va-
sodilation, neuromodulation, and platelet aggregation), its concentration is tightly
regulated(60). Moreover, after adenosine is absorbed from the intestine and released
into the portal vein, it will quickly enter the liver. In the liver, adenosine will be
broken down to uric acid by the enzymes adenosine deaminase and xanthine oxi-
dase(61).
Although the evidence of effectiveness of oral ATP administration in sports
and fitness is not quite solid, commercial marketing of ATP as an aid in in particu-
lar the bodybuilding area of sports is widespread(62, 63).
Evidence from animal studies by Kichenin et al. shows that in rats some of
the molecules from which ATP can be regenerated can be absorbed from the intes-
tinal lumen and secreted into the portal vein(64, 65). ATP, adenine, inosine, adeno-
sine, AMP, ADP and uric acid concentrations in plasma from the portal vein were
increased when, during a surgical procedure, an isolated part of the jejunum was
injected with ATP. This occurred specifically in rats that were treated with ATP
(10 mg/kg/day) for 30 days.
When considering the oral route as an alternative to ATP infusion, some of
the potential limiting factors, like ATP metabolism in the enterocyte, need to be
addressed. One potential way to avoid metabolism of ATP in the upper gastro-
intestinal tract, is to let ATP evade the enzymes and acidic environment in the
stomach that may break it down. An enteric coating would bypass the stomach
with its digestive enzymes and let ATP reach the intestine, where it could be ab-
sorbed intact. Besides enzymes, the acidic environment of the stomach might also
cause the breakdown of ATP, since ATP’s catalytic phosphate (the phosphate lo-
cated furthest away from adenine and which is cleaved off during enzyme-
catalyzed hydrolysis) is acid-labile(66). The gastrointestinal tract is pictured sche-
matically in figure 3.
One possible use of an enteric coating is to target a specific area along the ga-
strointestinal tract. In case of ATP, it is for example known that the proximal
small intestine of mice exhibits higher ATPase activity compared to the distal
small intestine(67). By protecting ATP by a coating consisting of polymers with dif-
ferent pH solubility, the resulting release of ATP in specific areas of the small in-
testine could therefore result in differences in uptake efficiency. This technique
employs the increasing pH along the gastrointestinal tract (figure 3).
General introduction
14
There has been a limited number of studies on oral ATP administration in
humans. In a 14-day study intended to investigate anaerobic performance, Jordan
et al.(68) administered an enterically coated oral ATP supplement containing dosag-
es of up to 225 mg daily. These authors observed small increases in performance of
high-intensity exercise bouts of very short duration, but no statistically significant
change of whole blood or plasma ATP concentrations. The authors suggested that
either the molecular size of ATP might prevent absorption, or that ATP might be
dephosphorylated intraluminally to adenosine. A placebo-controlled RCT was de-
signed to investigate the efficacy of daily administration of 90 mg ATP for 30 days
on subacute low-back pain(69, 70). Results showed that patients in the ATP group
took significantly fewer analgesics than patients in the placebo group. Also within
the department of Epidmiology of Maastricht University some pilot-studies were
performed in small subject-numbers to investigate the effects of oral ATP adminis-
tration when ATP was ingested either dissolved in water or as enteric coated cap-
sules(71).
These pilot studies suggested a possible rise in blood uric acid concentration,
but not ATP after oral ATP administration(71). Explanations for the increased uric
acid concentrations included the metabolism of unprotected ATP dissolved in water
in the stomach, and the premature release of ATP from enteric coated 00-size cap-
Figure 3: A schematic drawing of the human gastrointestinal tract. Using enteric coating, specific parts of
the gastrointestinal tract can be targeted for release of compounds that would otherwise be degraded by
the acidic environment of the stomach or by enzymes that are present in certain parts of the gastrointes-
tinal tract. pH-dependent enteric coating depend on the changes in pH that occur along the gastrointes-
tinal tract. The pH values for the fasted and fed state are taken from Gray and Dressman (1996)97.
Chapter 1
15
sules due to inadequate coating properties. These uncertainties called for a more
thorough and better controlled investigation of oral ATP administration. The next
paragraph will cover the characteristics of uric acid and its diverse functions.
Plasma uric acid
Humans and higher primates have much higher plasma uric acid concentra-
tions than other species. This probably arose from parallel, independent mutations
of the gene coding for uricase in two lineages leading to higher primates in the Mi-
ocene (5-23 million years ago)(72). This caused uric acid to be the final enzymatic
endproduct in purine metabolism, and not the more soluble and readily excreted
allantoin(73). The fact that these mutations were able to maintain themselves,
strongly suggests that there must have been a selection advantage to having high-
er plasma uric acid levels(73). Several hypotheses have been put forward.
Advantages of high plasma uric acid concentrations
The first and most quoted hypothesis involves the important role uric acid
plays as an antioxidant in plasma(74). Proctor(75) and later Ames(74) suggested that
the uric acid mutation may have benefited survival because it maintained ade-
quate plasma antioxidant activity after the loss of ascorbate (vitamin C) synthesis
due to an earlier mutation(76). Consistent with its antioxidant function, uric acid
has been suggested to be protective in various conditions, such as acute stroke,
multiple sclerosis, and Parkinson’s disease(77).
A second hypothesis focuses on uric acid’s role in innate immunity. Specifical-
ly, uric acid may aid in the immune recognition of dying cells(78), help activate the
inflammasome which is responsible for interleukin-1 beta (IL- (79), and
support the immune rejection of tumor cells(80). More recently, uric acid release
from cells in the area directly surrounding a site of vaccin injection was found to
aid the effectiveness of vaccination(81, 82).
A third hypothesis links the higher uric acid concentrations to a survival ad-
vantage in early primates by helping to maintain blood pressure during periods of
dietary change and environmental stress(83). Uric acid has also been proposed to
have neurostimulant properties based on its similarity in chemical structure with
caffeine(84) and due to studies suggesting it may have a role in increasing mental
performance(85).
General introduction
16
Disadvantages of high plasma uric acid concentrations
Although mainly positive properties are ascribed to uric acid when it is
present at normal plasma concentrations of 175-355 µmol/L, uric acid can form
deposits of monosodium urate (MSU) crystals when plasma concentrations are
above 420 µmol/L for prolonged periods of time. These crystals are the causative
agent of gout, which affects about 1% of the population, the majority being male(86).
Gout is a painful inflammation of the joints, which may affect surrounding tissues
and skin. In most cases, patients are hyperuricemic (>420 µmol/L), and a Western
diet (a diet high in fat, fructose and protein) and obesity are often associated with
this condition(87, 88). The prevalence of gout and hyperuricemia has increased in the
last century(89). Gout is a condition exclusive to primates, because all other species
are able to further break down uric acid to allantoin, which is far more soluble and
readily excreted through the kidneys(90). Although more than 10% of the people
with a predominantly Western diet have elevated circulating uric acid concentra-
tions, only a minority of hyperuricemic patients develop gout, suggesting that oth-
er factors are involved in the pathogenesis of gout.
Acute attacks of gout are often triggered by trauma, surgery, intercurrent ill-
ness, excess alcohol intake or drugs that alter serum uric acid levels. Such events
may stimulate de novo generation of MSU crystals (e.g. derived from dying cells) or
may trigger their release from preformed MSU crystal deposits within the joint(91).
In this manner, MSU crystals can act as a ‘danger signal’, resembling exogenous
stimuli (e.g. microbial LPS) that activate the innate immune system, often the first
line of defence protecting the host from invading microbial pathogens(78, 92).
The innate immune system may play a crucial role in triggering MSU crystal-
induced inflammation. For instance, MSU crystals were demonstrated to stimulate
synovial cells, monocyte-macrophages, and neutrophils to produce TNF--8, IL-
-6 and monocyte chemotactic factors, which, in turn, induce acute inflamma-
tion(93). Macrophage activation by MSU crystals was reported to require the recog-
nition of MSU crystals by Toll-like receptors (TLR) 2 and 4, after which IL-o-
duction is stimulated via the downstream TLR adapter protein myeloid differentia-
tion factor 88 (MyD88)(94). The importance of neutrophils in acute gout is supported
by observations that the injection of MSU crystals into the joints of animal models,
which normally reproduces the symptoms of gout, gives a markedly attenuated
response when neutrophils are depleted(95). The neutrophils that are attracted to
the joint actively phagocytose MSU crystals, resulting in membranolysis, genera-
tion of reactive oxygen species and in release of lysosomal enzymes, prostaglandin
E2, leukotrienes, and IL-1(96).
Chapter 1
17
Aims of the thesis
As indicated in this introduction, both intracellular and extracellular ATP
are important in human physiology. ATP has been successfully administered
intravenously to cancer patients in randomized clinical trials, resulting in several
favourable effects, such as reducing pain and increasing life expectancy. The intra-
venous route of administration of ATP is, however, a considerable burden for ad-
vanced stage cancer patients. Therefore, the aim of the work described in this the-
sis was to evaluate whether ATP can be effectively administered orally.
At the start of this project, there was no analytical method available that
could quickly and reliably measure ATP and its metabolites in a single run. There-
fore, we developed an HPLC method for measuring ATP and seven of its metabo-
lites in human blood samples and optimized it for quick analysis in one chromato-
graphic run. This is described in chapter 2. This method was applied to investi-
gate ATP metabolism in human blood in an ex vivo setting and subsequently used
for measuring ATP metabolites in blood samples collected in the human interven-
tion studies described in chapter 3, 4, and 5.
A comparative study using two different oral administration routes of ATP is
described in chapter 3. In eight volunteers, 5000 mg ATP was either administered
dissolved in water directly in the duodenum through a nasoduodenal tube or by
means of enteric coated pellets. Two variants of the pellets were tested next to each
other, each targeting a different section of the small intestine. The study was
blinded and placebo-controlled and between each single administration of ATP was
a period of one week. In this study we measured blood samples collected during a
period up to 8 hours after administration.
Elaborating on the work presented in chapter 3, a study using a prolonged
period of ATP administration is described in chapter 4. This time, 4 groups of
eight volunteers take in a daily dosage of ATP ranging from 0, 250, 1250 to 5000
mg. The enteric coated pellets that gave optimal results in the previous study were
used. On the first and last day of the four-week study period, all volunteers re-
ceived a single dose of 5000 mg ATP. The difference in blood ATP and metabolite
concentrations between the first and last day were compared in order to gain more
insight into the effects of a prolonged period of daily oral ATP administration.
Chapter 5 describes an ex vivo study with LPS-PHA stimulated whole blood.
In this study we looked at the immunomodulatory effects of a 24-hour incubation
period with ATP. In order to investigate the effects of ADP, AMP, adenosine, in-
osine, hypoxanthine and uric acid, the blood was also incubated with these metabo-
lites of ATP. At different time-points after LPS-PHA stimulation several immuno-
modulatory markers were studied. Also the involvement of the transcription factor

In chapter 6, an ex vivo study using fresh blood from healthy volunteers is
described to further investigate the effects of uric acid on inflammation in blood ex
vivo. Blood was incubated in the presence of lipopolysaccharide (LPS), phytohae-
General introduction
18
magglutinin (PHA) and either uric acid or monosodium urate (MSU) crystals. We
looked at the production of a panel of 17 different cytokines after an incubation
period of 24 hours. Besides this investigation into the inflammatory effects, also
the antioxidant capacity was investigated.
In chapter 7, the findings described in the previous chapters will be dis-
cussed. Some further prospects and future perspectives will be presented.
References
1. S. Tornroth-Horsefield and R. Neutze. (2008) Opening and closing the me-
tabolite gate. Proc Natl Acad Sci U S A. 105:19565-19566.
2. M.W. Gray, G. Burger, and B.F. Lang. (1999) Mitochondrial evolution.
Science. 283:1476-1481.
3. F.A. Antoni. (2000) Molecular diversity of cyclic AMP signalling. Front
Neuroendocrinol. 21:103-132.
4. P. Insel, D. Ross, K. MacMahon, et al. Nutrition, Jones & Bartlett Learn-
ing, 2010.
5. F. Di Virgilio, J.M. Boeynaems, and S.C. Robson. (2009) Extracellular nuc-
leotides as negative modulators of immunity. Curr Opin Pharmacol. 9:507-
513.
6. G. Burnstock. Introduction: ATP and its metabolites as potent extracellular
agents. In E.M. Schwiebert (ed.), Extracellular nucleotides and nucleosides:
release, receptors, and physiological and pathophysiological effects, Vol. 54,
Academic press, Amsterdam, 2003, pp. 1-27.
7. G. Burnstock. (2009) Purinergic receptors and pain. Curr Pharm Des.
15:1717-1735.
8. S.B. Coade and J.D. Pearson. (1989) Metabolism of adenine nucleotides in
human blood. Circ Res. 65:531-537.
9. P. Meghji, J.D. Pearson, and L.L. Slakey. (1995) Kinetics of extracellular
ATP hydrolysis by microvascular endothelial cells from rat heart. Biochem
J. 308 ( Pt 3):725-731.
10. K.M. Dwyer, S. Deaglio, W. Gao, et al. (2007) CD39 and control of cellular
immune responses. Purinergic Signal. 3:171-180.
11. G.G. Yegutkin. (2008) Nucleotide- and nucleoside-converting ectoenzymes:
Important modulators of purinergic signalling cascade. Biochim Biophys
Acta. 1783:673-694.
12. G.R. Strohmeier, W.I. Lencer, T.W. Patapoff, et al. (1997) Surface expres-
sion, polarization, and functional significance of CD73 in human intestinal
epithelia. J Clin Invest. 99:2588-2601.
13. C.E. Berry and J.M. Hare. (2004) Xanthine oxidoreductase and cardiovas-
cular disease: molecular mechanisms and pathophysiological implications.
J Physiol. 555:589-606.
14. U. Schenk, A.M. Westendorf, E. Radaelli, et al. (2008) Purinergic control of
T cell activation by ATP released through pannexin-1 hemichannels. Sci
Signal. 1:ra6.
15. K. Atarashi, J. Nishimura, T. Shima, et al. (2008) ATP drives lamina pro-
pria T(H)17 cell differentiation. Nature. 455:808-812.
Chapter 1
19
16. A. Filippini, R.E. Taffs, and M.V. Sitkovsky. (1990) Extracellular ATP in T-
lymphocyte activation: possible role in effector functions. Proc Natl Acad
Sci U S A. 87:8267-8271.
17. A. Sikora, J. Liu, C. Brosnan, et al. (1999) Cutting edge: purinergic signal-
ing regulates radical-mediated bacterial killing mechanisms in macrophag-
es through a P2X7-independent mechanism. J Immunol. 163:558-561.
18. D. Ferrari, P. Chiozzi, S. Falzoni, et al. (1997) ATP-mediated cytotoxicity in
microglial cells. Neuropharmacology. 36:1295-1301.
19. R. Beigi, E. Kobatake, M. Aizawa, et al. (1999) Detection of local ATP re-
lease from activated platelets using cell surface-attached firefly luciferase.
Am J Physiol. 276:C267-278.
20. M.J. Bours, E.L. Swennen, F. Di Virgilio, et al. (2006) Adenosine 5’-
triphosphate and adenosine as endogenous signaling molecules in immuni-
ty and inflammation. Pharmacol Ther. 112:358-404.
21. A.N. Drury and A. Szent-Gyorgi. (1929) The physiological activity of ade-
nine compounds with special reference to their action upon the mammalian
heart. J Physiol (London). 68:213-237.
22. G. Burnstock. (1972) Purinergic nerves. Pharmacol Rev. 24:509-581.
23. G. Burnstock, B.B. Fredholm, R.A. North, et al. (2010) The birth and post-
natal development of purinergic signalling. Acta Physiol (Oxf). 199:93-147.
24. G. Burnstock and A. Verkhratsky. (2009) Evolutionary origins of the puri-
nergic signalling system. Acta Physiol (Oxf). 195:415-447.
25. G. Burnstock and G.E. Knight. (2004) Cellular distribution and functions of
P2 receptor subtypes in different systems. Int Rev Cytol. 240:31-304.
26. M.P. Abbracchio and G. Burnstock. (1998) Purinergic signalling: pathophy-
siological roles. Jpn J Pharmacol. 78:113-145.
27. J.T. Neary and H. Zimmermann. (2009) Trophic functions of nucleotides in
the central nervous system. Trends Neurosci. 32:189-198.
28. I. von Kugelgen. (2006) Pharmacological profiles of cloned mammalian
P2Y-receptor subtypes. Pharmacol Ther. 110:415-432.
29. S.M. Pasternack, I. von Kugelgen, K.A. Aboud, et al. (2008) G protein-
coupled receptor P2Y5 and its ligand LPA are involved in maintenance of
human hair growth. Nat Genet. 40:329-334.
30. M.P. Abbracchio, G. Burnstock, J.M. Boeynaems, et al. (2006) International
Union of Pharmacology LVIII: update on the P2Y G protein-coupled nuc-
leotide receptors: from molecular mechanisms and pathophysiology to
therapy. Pharmacol Rev. 58:281-341.
31. B.S. Khakh and R.A. North. (2006) P2X receptors as cell-surface ATP sen-
sors in health and disease. Nature. 442:527-532.
32. R.A. North. (2002) Molecular physiology of P2X receptors. Physiol Rev.
82:1013-1067.
33. M.F. Jarvis and B.S. Khakh. (2009) ATP-gated P2X cation-channels. Neu-
ropharmacology. 56:208-215.
34. V. Ralevic and G. Burnstock. (1998) Receptors for purines and pyrimidines.
Pharmacol Rev. 50:413-492.
35. B.B. Fredholm. (2010) Adenosine receptors as drug targets. Exp Cell Res.
316:1284-1288.
36. H. Schwalbe and G. Wess. (2002) Dissecting G-protein-coupled receptors:
structure, function, and ligand interaction. Chembiochem. 3:915-919.
General introduction
20
37. R.A. North and A. Surprenant. (2000) Pharmacology of cloned P2X recep-
tors. Annu Rev Pharmacol Toxicol. 40:563-580.
38. D.W. Hommes, M.P. Peppelenbosch, and S.J. van Deventer. (2003) Mitogen
activated protein (MAP) kinase signal transduction pathways and novel
anti-inflammatory targets. Gut. 52:144-151.
39. A.S. Baldwin, Jr. (2001) Series introduction: the transcription factor NF-
kappaB and human disease. J Clin Invest. 107:3-6.
40. F. Wilkin, P. Stordeur, M. Goldman, et al. (2002) Extracellular adenine
nucleotides modulate cytokine production by human monocyte-derived
dendritic cells: dual effect on IL-12 and stimulation of IL-10. Eur J Immu-
nol. 32:2409-2417.
41. G. Hasko, D.G. Kuhel, A.L. Salzman, et al. (2000) ATP suppression of inter-
leukin-12 and tumour necrosis factor-alpha release from macrophages. Br
J Pharmacol. 129:909-914.
42. D. Myrtek, T. Muller, V. Geyer, et al. (2008) Activation of human alveolar
macrophages via P2 receptors: coupling to intracellular Ca2+ increases and
cytokine secretion. J Immunol. 181:2181-2188.
43. E.L. Swennen, A. Bast, and P.C. Dagnelie. (2005) Immunoregulatory ef-
fects of adenosine 5’-triphosphate on cytokine release from stimulated
whole blood. Eur J Immunol. 35:852-858.
44. E.L. Swennen, A. Bast, and P.C. Dagnelie. (2006) Purinergic receptors in-
volved in the immunomodulatory effects of ATP in human blood. Biochem
Biophys Res Commun. 348:1194-1199.
45. J.M. Argiles, R. Moore-Carrasco, G. Fuster, et al. (2003) Cancer cachexia:
the molecular mechanisms. Int J Biochem Cell Biol. 35:405-409.
46. K.A. Nelson. (2000) The cancer anorexia-cachexia syndrome. Semin Oncol.
27:64-68.
47. H.J. Agteresch, P.C. Dagnelie, A. van Der Gaast, et al. (2000) Randomized
clinical trial of adenosine 5’-triphosphate in patients with advanced non-
small-cell lung cancer. J Natl Cancer Inst. 92:321-328.
48. H.J. Agteresch, T. Rietveld, L.G. Kerkhofs, et al. (2002) Beneficial effects of
adenosine triphosphate on nutritional status in advanced lung cancer pa-
tients: a randomized clinical trial. J Clin Oncol. 20:371-378.
49. S. Beijer, P.S. Hupperets, B.E. van den Borne, et al. (2009) Effect of adeno-
sine 5’-triphosphate infusions on the nutritional status and survival of pre-
terminal cancer patients. Anticancer Drugs. 20:625-633.
50. H.J. Agteresch, S.A. Burgers, A. van der Gaast, et al. (2003) Randomized
clinical trial of adenosine 5’-triphosphate on tumor growth and survival in
advanced lung cancer patients. Anticancer Drugs. 14:639-644.
51. S. Leij-Halfwerk, H.J. Agteresch, P.E. Sijens, et al. (2002) Adenosine tri-
phosphate infusion increases liver energy status in advanced lung cancer
patients: an in vivo 31P magnetic resonance spectroscopy study. Hepatolo-
gy. 35:421-424.
52. S. Beijer, E.A. Gielisse, P.S. Hupperets, et al. (2007) Intravenous ATP infu-
sions can be safely administered in the home setting: a study in pre-
terminal cancer patients. Invest New Drugs. 25:571-579.
53. J.D. Carver. (1999) Dietary nucleotides: effects on the immune and ga-
strointestinal systems. Acta Paediatr Suppl. 88:83-88.
Chapter 1
21
54. J.D. Carver and C.I. Stromquist. (2006) Dietary nucleotides and preterm
infant nutrition. J Perinatol. 26:443-444.
55. L.K. Pickering, D.M. Granoff, J.R. Erickson, et al. (1998) Modulation of the
immune system by human milk and infant formula containing nucleotides.
Pediatrics. 101:242-249.
56. K. Synnestvedt, G.T. Furuta, K.M. Comerford, et al. (2002) Ecto-5’-
nucleotidase (CD73) regulation by hypoxia-inducible factor-1 mediates
permeability changes in intestinal epithelia. J Clin Invest. 110:993-1002.
57. M. Molina-Arcas, F.J. Casado, and M. Pastor-Anglada. (2009) Nucleoside
transporter proteins. Curr Vasc Pharmacol. 7:426-434.
58. L.Y. Ngo, S.D. Patil, and J.D. Unadkat. (2001) Ontogenic and longitudinal
activity of Na(+)-nucleoside transporters in the human intestine. Am J
Physiol Gastrointest Liver Physiol. 280:G475-481.
59. M. Pastor-Anglada, E. Errasti-Murugarren, I. Aymerich, et al. (2007) Con-
centrative nucleoside transporters (CNTs) in epithelia: from absorption to
cell signaling. J Physiol Biochem. 63:97-110.
60. S.A. Baldwin, J.R. Mackey, C.E. Cass, et al. (1999) Nucleoside transpor-
ters: molecular biology and implications for therapeutic development. Mol
Med Today. 5:216-224.
61. I.H. Fox. (1981) Metabolic basis for disorders of purine nucleotide degrada-
tion. Metabolism. 30:616-634.
62. http://www.super-smart.eu/nl--PEAK-ATP-125-mg--Energie--0409 (ac-
cessed 03-08-2010 2010).
63. Healthtube. Interviews with Dr. Eliezer Rapaport on the efficacy of the
product Peak ATP that is marketed by his company Technical Sourcing In-
ternational, Inc., 2010.
64. K. Kichenin, M. Seman, S. Decollogne, et al. (2000) Chronic oral adminis-
tration of ATP modulates nucleoside transport and purine metabolism in
rats. J Pharmacol Exp Ther. 294:126-133.
65. K. Kichenin, S. Decollogne, J. Angignard, et al. (2000) Cardiovascular and
pulmonary response to oral administration of ATP in rabbits. J Appl Phy-
siol. 88:1962-1968.
66. A.M. Janski, P.A. Srere, N.W. Cornell, et al. (1979) Phosphorylation of ATP
citrate lyase in response to glucagon. J Biol Chem. 254:9365-9368.
67. K.A. Mohamedali, O.M. Guicherit, R.E. Kellems, et al. (1993) The highest
levels of purine catabolic enzymes in mice are present in the proximal
small intestine. J Biol Chem. 268:23728-23733.
68. A.N. Jordan, R. Jurca, E.H. Abraham, et al. (2004) Effects of oral ATP sup-
plementation on anaerobic power and muscular strength. Med Sci Sports
Exerc. 36:983-990.
69. M. Rossignol, F.A. Allaert, S. Rozenberg, et al. (2005) Measuring the con-
tribution of pharmacological treatment to advice to stay active in patients
with subacute low-back pain: a randomised controlled trial. Pharmacoepi-
demiology and drug safety. 14:861-867.
70. B. Bannwarth, F.A. Allaert, B. Avouac, et al. (2005) A randomized, double-
blind, placebo controlled study of oral adenosine triphosphate in subacute
low back pain. J Rheumatol. 32:1114-1117.
71. C. van den Hurk. Unpublished observations, 2004.
General introduction
22
72. X.W. Wu, D.M. Muzny, C.C. Lee, et al. (1992) Two independent mutational
events in the loss of urate oxidase during hominoid evolution. J Mol Evol.
34:78-84.
73. M. Oda, Y. Satta, O. Takenaka, et al. (2002) Loss of urate oxidase activity
in hominoids and its evolutionary implications. Mol Biol Evol. 19:640-653.
74. B.N. Ames, R. Cathcart, E. Schwiers, et al. (1981) Uric acid provides an
antioxidant defense in humans against oxidant- and radical-caused aging
and cancer: a hypothesis. Proc Natl Acad Sci U S A. 78:6858-6862.
75. P. Proctor. (1970) Similar functions of uric acid and ascorbate in man? Na-
ture. 228:868.
76. M. Nishikimi, R. Fukuyama, S. Minoshima, et al. (1994) Cloning and
chromosomal mapping of the human nonfunctional gene for L-gulono-
gamma-lactone oxidase, the enzyme for L-ascorbic acid biosynthesis miss-
ing in man. J Biol Chem. 269:13685-13688.
77. M.K. Kutzing and B.L. Firestein. (2008) Altered uric acid levels and dis-
ease states. J Pharmacol Exp Ther. 324:1-7.
78. Y. Shi, J.E. Evans, and K.L. Rock. (2003) Molecular identification of a dan-
ger signal that alerts the immune system to dying cells. Nature. 425:516-
521.
79. F. Martinon, V. Petrilli, A. Mayor, et al. (2006) Gout-associated uric acid
crystals activate the NALP3 inflammasome. Nature. 440:237-241.
80. D.E. Hu, A.M. Moore, L.L. Thomsen, et al. (2004) Uric acid promotes tumor
immune rejection. Cancer Res. 64:5059-5062.
81. M. Kool, T. Soullie, M. van Nimwegen, et al. (2008) Alum adjuvant boosts
adaptive immunity by inducing uric acid and activating inflammatory den-
dritic cells. J Exp Med. 205:869-882.
82. B.N. Lambrecht, M. Kool, M.A. Willart, et al. (2009) Mechanism of action of
clinically approved adjuvants. Curr Opin Immunol. 21:23-29.
83. R.J. Johnson, E.A. Gaucher, Y.Y. Sautin, et al. (2008) The planetary biolo-
gy of ascorbate and uric acid and their relationship with the epidemic of
obesity and cardiovascular disease. Med Hypotheses. 71:22-31.
84. E. Orowan. (1955) The origin of man. Nature. 175:683-684.
85. C.M. Barrera, Z.R. Ruiz, and W.P. Dunlap. (1988) Uric acid: a participating
factor in the symptoms of hyperactivity. Biol Psychiatry. 24:344-347.
86. H.M. Kramer and G. Curhan. (2002) The association between gout and
nephrolithiasis: the National Health and Nutrition Examination Survey
III, 1988-1994. Am J Kidney Dis. 40:37-42.
87. H.K. Choi and G. Curhan. (2008) Soft drinks, fructose consumption, and
the risk of gout in men: prospective cohort study. BMJ. 336:309-312.
88. N. Anzai, Y. Kanai, and H. Endou. (2007) New insights into renal transport
of urate. Curr Opin Rheumatol. 19:151-157.
89. G. Nuki and P.A. Simkin. (2006) A concise history of gout and hyperurice-
mia and their treatment. Arthritis Res Ther. 8 Suppl 1:S1.
90. D. Voet and J. Voet. Biochemistry, John Wiley and Sons, New York, 1995.
91. N. Schlesinger. (2005) Diagnosis of gout: clinical, laboratory, and radiologic
findings. Am J Manag Care. 11:S443-450; quiz S465-448.
92. T.H. Mogensen. (2009) Pathogen recognition and inflammatory signaling in
innate immune defenses. Clin Microbiol Rev. 22:240-273, Table of Con-
tents.
Chapter 1
23
93. F.S. Di Giovine, S.E. Malawista, G. Nuki, et al. (1987) Interleukin 1 (IL 1)
as a mediator of crystal arthritis. Stimulation of T cell and synovial fibrob-
last mitogenesis by urate crystal-induced IL 1. J Immunol. 138:3213-3218.
94. R. Liu-Bryan, P. Scott, A. Sydlaske, et al. (2005) Innate immunity con-
ferred by Toll-like receptors 2 and 4 and myeloid differentiation factor 88
expression is pivotal to monosodium urate monohydrate crystal-induced in-
flammation. Arthritis Rheum. 52:2936-2946.
95. W.J. Martin, M. Walton, and J. Harper. (2009) Resident macrophages in-
itiating and driving inflammation in a monosodium urate monohydrate
crystal-induced murine peritoneal model of acute gout. Arthritis Rheum.
60:281-289.
96. R.C. Landis and D.O. Haskard. (2001) Pathogenesis of crystal-induced in-
flammation. Curr Rheumatol Rep. 3:36-41.
97. V. Gray and J. Dressman. (1996) Simulated intestinal fluid TS - Change to
pH 6.8. USP Pharmacopeial Forum. 22:1943-1945.
Chapter 2
Simultaneous determination of adeno-
sine triphosphate and its metabolites in
human whole blood by RP-HPLC and
UV-detection
Erik JCM Coolen
Ilja CW Arts
Els LR Swennen
Aalt Bast
Martien A Cohen Stuart
Pieter C Dagnelie
J Chromatogr B Biomed Appl. 2008 864:43-51.
25
Development of an HPLC method for determining ATP and metabolites
26
Abstract
To obtain insight in mechanisms of action of extracellular ATP and adenosine, a
simple HPLC method has been optimized and applied to investigate ATP metabol-
ism in human whole blood ex vivo. This method provided good chromatographic
resolution and peak shape for all eight compounds within a 19 min run time. The
   
adenine nucleotides and the method demonstrated good linearity. Within-day pre-
cision ranged from 0.7 to 5.9% and between-day from 2.6 to 15.3%. Simplicity and
simultaneous detection of ATP and its metabolites make this method suitable for
clinical pharmacokinetic studies.
1. Introduction
The nucleotide profile of blood is relatively simple compared to that of nucleated
cells. Because no DNA synthesis occurs, only ribonucleotides are present. The ade-
nine ribonucleotide pool, which in metabolically normal erythrocytes mainly con-
sists of adenosine triphosphate (ATP), adenosine diphosphate (ADP), and adeno-
sine monophosphate (AMP), is much larger than that of other nucleotides, such as
the guanine ribonucleotides (GTP, GDP, and GMP)(1). Generally, ADP and AMP
concentrations are only 12.5-20% and 1-2% respectively, of the ATP concentra-
tions(2). Apart from its role as an intracellular energy carrier, extracellular ATP is
involved in processes such as neurotransmission, mechanosensory transduction,
secretory functions and vasodilatation, and long-term (chronic) signalling functions
in development and regeneration of cells(3). Moreover, once ATP is released, for
example through cell lysis during organ injury, it can mediate several inflammato-
ry responses, including the release of cytokines, tumor necrosis factor-r-
leukin-of leukocyte adhesion
to the endothelium(4,5). Proliferation of various in vitro tumor cell lines has been
shown to be suppressed after exposure to ATP(6-8). The pharmacological use of ATP
has received increasing attention following reports of its benefit in pain, vascular
disease and cancer(9). Given our interest in the metabolism of ATP and its beha-
viour after ATP supplementation in humans, we needed a quick and reliable me-
thod for the simultaneous measurement of purine nucleotides, nucleosides, and
bases in human blood. Given the complexity of the nucleotide breakdown cascade,
which involves various enzymes, special care had to be taken to suppress un-
wanted nucleotide degeneration in fresh samples.
Several methods have been published to measure nucleotides, including an
enzyme assay(10), bioluminescence, 31P nuclear magnetic resonance spectroscopy
(NMR), high-performance capillary electrophoresis (HPCE), gas chromatography
(GC), ion-pair reversed phase HPLC(11-16), gradient HPLC(17-19), and ion-exchange
methods(20). RP-HPLC lacks some of the drawbacks of other methods, like, for ion-
Chapter 2
27
exchange HPLC, the need for highly concentrated elution buffers and long analysis
times. An advantage of the method is that it allows simultaneous detection of nuc-
leotides, nucleosides and nucleobases in a single run with short times between in-
jections. Given the high concentrations of adenine nucleotides in human blood
(1500-mol/L)(21), there was no need for measurement in the lower nanomolar
range, which is for instance achievable by bioluminescence kits.
In the present paper we report a reversed-phase HPLC method for the simul-
taneous analysis of ATP, ADP, AMP, adenosine, adenine, inosine, hypoxanthine,
and uric acid in human blood. Sample collection and processing methods were op-
timized. The application of the method is illustrated by an experiment in which the
concentrations of adenine nucleotides are monitored after incubation of human
blood samples with ATP, a model used in our lab to study anti-inflammatory prop-
erties of ATP and its breakdown products.
2. Materials and methods
2.1. Chemicals
Adenosine 5’-triphosphate disodium salt (ATP), adenosine 5’-diphosphate dis-
odium salt (ADP), adenosine 5’-monophosphate sodium salt (AMP), adenine, in-
osine, hypoxanthine and uric acid were purchased from Sigma Chemical Co., St.
Louis, USA. Adenosine was obtained from Bufa B.V., Uitgeest, The Netherlands.
Perchloric acid 70% solution in water (PCA) was purchased from Sigma-Aldrich,
Steinheim, Germany. Potassium hydroxide (KOH), potassium dihydrogen phos-
phate (KH2PO4), potassium carbonate (K2CO3), di-potassium hydrogen phosphate
trihydrate (K2HPO4*3H2O) and sodium hydroxide (NaOH) were obtained from
Merck, Darmstadt, Germany. 0.9% saline was purchased from Braun, Melsungen,
Germany. RPMI 1640 (order number 21875) medium containing L-glutamine, was
obtained from Gibco, Paisly, UK.
2.2. Sample preparation
Blood was collected from healthy volunteers by venipuncture in EDTA-
containing vacutainer tubes (Vacutainer, Becton-Dickinson, New Jersey, USA).
 -mixed    -cold 8% PCA in a 1.5
mL Eppendorf tube. After precipitation of the protein fraction (at 12,000 × g, 10
ed by 2CO3 in 6
    the pH of the samples.
Following centrifugal removal of the perchlorate (12,000 × g, 10 min, 
   ee paragraph 2.4 for its
composition) in HPLC microvials (Agilent Technologies, Palo Alto, CA, USA).
Development of an HPLC method for determining ATP and metabolites
28
2.3. Equipment
ATP and its metabolites were quantified using a series 1100 Agilent HPLC
system (Agilent Technologies, Palo Alto, CA, USA) consisting of a quaternary gra-
dient pump, a variable wavelength detector (set at 254 nm) and a solvent degasser
    
column (150 × 4.6 mm i.d.; Thermo Electron Corp., USA) with pore diameter 120 Å,
protected by a  4 mm i.d.; Thermo Electron
Corp).
2.4. Separation
A 50 mM phosphate buffer (pH 6.0) (mobile phase A), 100% methanol (mobile
phase B) and a flow of 1 mL/min were employed to separate the compounds of in-
terest. Buffers were prepared by diluting 50 mL of a stock solution of 0.6 M
K2HPO4 / 0.4 M KH2PO4, to 1 L of milliQ water. The pH was adjusted to 6.0 with
concentrated phosphoric acid. From 0-2 min, the eluent consisted of 100% mobile
phase A. Between 2-10 min the amount of mobile phase B increased linearly to
12.5%, and then stayed like that for 2 min. Finally, between 12 and 17 min, the
gradient returned linearly to 100 % mobile phase A. The column was equilibrated
between injections for 2 min, leading to a total run time    
sample was    l-
umn was kept at room 
2.5. Standard solutions
All purine standards were treated in a manner similar to the blood samples. First,
the standards were individually dissolved in mobile phase A to a concentration of 4
mM (ATP: 8 mM). Uric acid was dissolved in 0.1 M NaOH instead of mobile phase
A. Next, equal volumes of four compounds were mixed and added to 8% PCA in a
final ratio of mobile phase A to 8% PCA of 1:1. This resulted in two mixtures with
       
subsequently prepared as described in paragraph 2.2 and finally diluted with mo-
bile phase A to acquire a standard range which reflects pre-preparation concentra-
-
of ATP and its metabolites in blood were determined by comparing peak
areas to appropriate standards using Chemstation software (Version A.09.03; Agi-
lent, Palo Alto, CA, USA). A mathematical adjustment is applied to correct for the
experimentally determined 20% increase in metabolite concentration measured in
the supernatant, which occurs due to protein precipitation in blood samples. The
stability of ATP w    
that were a) stored at -u-
 was used
as a measure for the amount of degradation.
Chapter 2
29
2.6. Assay validation
Linearity of the assay was assessed using three calibration curves analyzed
on separate days. The curves were constructed by plotting the peak areas against
the concentration of the sample.
The within-day coefficient of variation (CV) was determined by repeated
analysis of five aliquots of a single volume of whole blood. The blood was spiked
with high and low concentrations of standards. Between-day CV was determined
by analyzing, on five different days, a blood sample that was collected from one
subject, mixed with PCA and then stored in five aliquots at -20 C awaiting further
sample preparation and analysis. The noise was determined by Chemstation soft-
ware in the time range between 10.5 and 13 min after injection of ten separate
whole blood samples. The limit of detection (LOD) and lower limit of quantification
(LLOQ) were calculated by multiplying the SD of the noise by 3 and 10, respective-
ly. The values for all compounds are presented as concentrations in whole blood, in
order to present the actual limits that are achievable by our method.
Recovery data of all eight compounds were obtained in duplicate by adding
known amounts of nucleotide standards (individually dissolved in mobile phase A
(uric acid in 0.1 M NaOH)) or a blank solution (mobile phase A) to blood samples
immediately after collection, at high and low spike concentrations. Preparation
then continued as described above for both these samples and for samples contain-
ing spikes in buffer. Recovery was calculat       
spike)] – -1.
2.7. Optimization
Sample handling: By using the method of Schweinsberg et al.(2) as a starting
point, optimizations were performed with blood samples obtained from healthy
volunteers. The optimization of sample preparation included varying a) the anti-
coagulant in the blood collection tube (EDTA or Li-Heparin), b) the timing of pro-
tein precipitation by PCA after blood collection (directly after blood collection or
after storage at -a-
ture or on ice), d) the precipitation of PCA (separate from, or simultaneous with pH
neutralisation by KOH), and e) the sample pH (pH range 3 to 6). The effects of
changing these parameters were rated optically based on peak shape, yield and
compound separation on the HPLC chromatogram.
Chromatographic conditions: The parameters varied for optimization of the
chromatographic conditions were a) elution buffer pH (range 3 to 6) and phosphate
concentration (range 50- 100 mM), b) timing of the elution gradient, and c) injec-
tion volume.
2.8. Application
After optimization, we applied the method to investigate the metabolism of
ATP added to whole blood ex vivo, and more specifically, the effects of dilution of
the blood with saline or RPMI 1640 medium on the concentration of ATP. This ex-
Development of an HPLC method for determining ATP and metabolites
30
periment was a follow-up to an earlier study by Swennen et al.(22), in which the ex
vivo whole blood model was used to investigate the immunoregulatory effects of
ATP. In this model, whole blood was diluted four-fold with RPMI 1640 medium,
resulting in an approximate ATP half life of 2 hours. We set out to determine the
effect of the dilution factor and type of dilution medium on the ATP metabolism,
since literature reports half times of 15 min in undiluted set-ups(23).
Blood from a healthy volunteer was collected in heparin-containing vacutain-
er tubes and directly put on ice (Vacutainer, 170 IU). Use of EDTA as an anticoa-
gulant was avoided, given our interest in ATP breakdown. Next, appropriate vo-
lumes of blood were put in 6- or 24-well plates and spiked with a known volume of
6 mM ATP in milliQ, saline or medium. Next, the spiked blood was used either
undiluted or diluted two- or four-fold by mixing gently with saline or medium. The

at 0, 2, 4, 8, 15, 30 min and 2, 4, and 24 h after addition of the ATP spike. Plates
were incubated at 5% CO2 es
           -free supernatant was
transferred to a clean Eppendorf tube and stored at -a-
tion products were determined as described above. Standard solutions were pre-
pared, as described in 2.5., and used to calculate the concentrations of ATP and its
degradation products. This resulted in time-dependent degradation profiles for
undiluted, and two- and four-fold diluted blood.
3. Results
3.1. Optimization of the sample handling
The goal of optimizing sample handling was to prevent loss of adenosine nuc-
leotides, which are normally rapidly degraded by ecto-enzymes located on the
plasma membrane. Lower concentrations of nucleotides were present when blood
collection tubes containing the anticoagulant Li-heparin were used instead of ED-
TA-containing tubes (data not shown). Furthermore, the precipitation of proteins
by mixing blood with PCA was found to be most effective in retaining endogenous
nucleotides when done directly following blood collection. This in comparison to the
situation in which the whole blood was mixed with PCA after it had been stored at
-
the latter case. The samples of blood mixed with PCA could either be stored at -80
c-
leotide concentration. After cold storage, equal levels of adenine nucleotides were
found, regardless of whether the samples were thawed on ice or at room tempera-
ture (data not shown).
Chapter 2
31
Figure 1: Chromatogram showing the effects of buffer pH on peak shape and retention time after
separate injections of ATP (100 µmol/L). Peaks are overlaid on the same time-axis, and normalized to
equal height on the Y-axis. Only the relevant part of the X-axis is shown. Widths reported were taken
at half height of the peaks.
Figure 2: Chromatogram of 500 µmol/L of uric acid (UA), ATP, ADP, hypoxanthine (Hyp), AMP, Adenine
(
Ade
),
inosine
(
Ino
)
and adenosine
(
Ado
)
. Conditions of the se
p
aration are as described in Section 2.3.
During sample preparation, it was found that precipitation of PCA in the
samples with K2CO3 could be performed together with neutralization by KOH
without loss of nucleotides. Optimal yield and peak shapes were obtained when all
samples were neutralized with buffer to pH 6.0. At pH 3, 4 and 5, peak shapes not-
ably worsened, affecting separation and yield of the various compounds. An exam-
ple of the effects of buffer pH on the ATP peak is displayed in Figure 1.
3.2. Optimization of chromatographic conditions
Peak shape and compound separation benefited from a buffer pH of 6.0 and a
phosphate concentration of 50 mM (data not shown). The elution gradient used
was a compromise between a fast analysis time and sufficient separation of the
peaks. Higher methanol concentrations resulted in overlapping peaks of the com-
pounds of interest. By returning the methanol concentration to 0 twelve min after
injection, the amount of time needed for column equilibration was limited to two
min. Finall  
volumes resulted in worse separation of ATP and uric acid, particularly in blood, in
which both are the most abundant compounds. A representative chromatogram of
the separation of eight purine standards in one run is shown in Figure 2.
0 2 4 6 8 10 Time (min) 14 16
mAU
1000
800
600
400
200
0
Ino (10,823)
Norm.
200
175
150
125
100
75
50
25
0
2.6 2.8 3 3.2 3.4 3.6 Time (min)
Development of an HPLC method for determining ATP and metabolites
32
3.3. Assay validation
Chromatograms of unspiked and spiked whole blood samples are shown in
Figure 3. Calibration curves were calculated using peak areas at six standard con-
centrations, the range of which was proportional to the concentration of analyte in
the prepared whole blood samples. The concentration range was linear from 1-500
          
greater than 0.999 (Table 1).
Within-day CV (range 0.7 – 5.9%) and between-day CV (2.6 – 15.3%), LOD
and LLOQ are presented for every compound in Table 2. The concentrations re-
ported represent the actual LOD and LLOQ of compounds in blood, since dilution
of the blood necessary for sample preparation is taken into account. The LOD were

TABLE 1. SUMMARY OF LINEARITY (RANGE, SLOPE, R2, AND INTERCEPT VALUES) AND ANALYTE RETENTION TIMES.
FOR EACH ANALYTE N=9 DATAPOINTS WERE USED.
Analyte Linearity Retention
time (min)
Range
(µmol/L)
r2 Slope ± SD Intercept ±
SD
ATP 1-500 0.999 9.59 ± 0.01 17.27 ± 7.27 3.6
ADP 1-500 0.999 8.25 ± 0.01 8.93 ± 6.62 4.5
AMP 1-500 0.999 8.54 ± 0.01 17.02 ± 6.30 6.5
Hypoxanthine 1-500 0.999 5.55 ± 0.01 42.76 ± 4.36 7.4
Uric acid 1-500 0.999 2.91 ± 0.002 5.77 ± 1.74 3.2
Adenine 1-500 0.999 7.04 ± 0.01 14.35 ± 3.96 10.2
Adenosine 1-500 0.999 10.37 ± 0.02 14.88 ± 8.43 15.7
Inosine 1-500 0.999 8.23 ± 0.07 34.34 ± 2.79 10.8
The recovery of the method after addition of known amounts of compounds to
whole blood samples ranged between 58% and 108% as presented in Table 3. The
concentrations of ATP, ADP, and AMP found in the whole blood were 1217 ± 75, 95
      ATP/ADP and ATP/AMP ratios
were 12.8 and 44.9, respectively.
Figure 3: Chromatogram of two whole blood samples. One sample was spiked with a mix of eight
compounds (solid line) followed by immediate sample preparation as described in Section 2.2. In the
fresh unspiked whole blood sample, only uric acid, ATP, ADP and AMP are detected (dashed line).
mAU
120
100
80
60
40
20
0
0 2 4 6 8 10 Time (min) 14 16
Chapter 2
33
The stability of ATP in mobile phase A was assessed for three storage condi-
tions. The proportion of degradation products detected was on average 1% of ATP
 for 1 hour, and 2% for storage and thawing
at -
TABLE 2. SUMMARY OF PRECISION AND DETECTION LIMITS OF ATP, ADP, AMP, HYPOXANTHINE, ADENINE,
ADENOSINE, INOSINE AND URIC ACID (CV = COEFFICIENT OF VARIATION, LOD = LIMIT OF DETECTION, LLOQ =
LOWER LIMIT OF QUANTIFICATION)
Analyte Within-day precision Between-day precision LOD
(µmol/L)
LLOQ
(µmol/L)
Mean
(µmol/L), N=5
CV
(%)
Mean
(µmol/L), N=5
CV
(%)
N=10 N=10
ATP 1003.0 0.7 974.3 3.6 0.049 0.162
ADP 356.5 0.9 357.3 5.5 0.072 0.239
AMP 241.1 1.3 244.5 2.6 0.052 0.172
Hypoxanthine 169.9 5.9 206.1 15.3 0.082 0.272
Uric acid 304.9 1.6 326.4 11.5 0.121 0.405
Adenine 167.5 3.2 194.4 10.4 0.090 0.299
Adenosine 226.1 1.7 221.8 4.3 0.070 0.232
Inosine 206.0 2.1 234.8 15.0 0.059 0.197
3.4 Application
In this experiment, the effects of dilution of blood with saline or RPMI 1640
medium on ATP metabolism were investigated in an ex vivo set-up. Figure 4 dis-
 luted (4a), and two-fold (4b
and 4c) and four-fold (4d and 4e) diluted aliquots of blood. Adenine, adenosine and
inosine were not detected. Besides uric acid, none of the other compounds were
present in samples taken before the addition of ATP.
In the undiluted situation, the ATP concentration declined steadily until
complete degradation after 30 min. The ADP concentration doubled 15 min after
incubation, followed by a decline to zero after two hours. AMP rose between 2 and
30 min, with the strongest increase between 15 and 30 min, followed by a decline
to zero after four hours. Hypoxanthine was first formed after 30 min incubation
and kept rising sharply until the final measurement after 19 hours. In contrast to
the other metabolites, uric acid levels were quite stable with only a minor increase
at the later stages of the incubation.
In the two-fold diluted situation, every compound behaved similarly, but with
a time-frame that was shifted by approximately 90 min. Two notable differences
Between-day precision of five aliquots of a single blood sample, measured on 5 different days spread
over a period of 7 months. Within-day precision is calculated after repeated analysis (N=5) of multiple
aliquots of a single volume of whole blood, spiked with known amounts of standards. Values are the
mean concentrations and the coefficient of variation (CV). The LOD and LLOQ were calculated as 3 and
10 times the SD of the noise of the UV baseline in untreated whole blood samples (N=10) in the time
range of 10.5 -13 minutes after injection.
Development of an HPLC method for determining ATP and metabolites
34
are a) the near simultaneous disappearance of ADP and ATP, instead of ATP pre-
ceding ADP, and b) the smaller rise in hypoxanthine concentration after 19 hours.
TABLE 3. RECOVERY DATA FOR ATP, ADP, AMP, HYPOXANTHINE, AD ENINE, ADENOSINE, INOSINE AND URIC ACID
IN WHOLE BLOOD
Recovery Sample conc
L
Added conc
L
New Conc
/L
%
ATP 844.0 1000 2057.8 121.4%
500 1422.6 115.7%
ADP 124.2 500 749.4 125.0%
250 445.4 128.4%
AMP 36.3 500 548.9 102.5%
250 299.5 105.2%
Adenine 0 500 540.7 108.1%
250 272.3 108.9%
Adenosine 1.53 500 548.4 109.4%
250 275.5 109.6%
Inosine 0 500 562.6 112.4%
250 271.1 108.4%
Hypoxanthine 3.87 500 481.0 95.4%
250 287.9 113.6%
Uric acid 254.6 500 682.1 85.5%
250 465.2 84.2%
Compared to the undiluted situation, the shift in time in the four-fold diluted
situation was approximately 3.5 hours. One difference is that the ATP concentra-
tion does not decline directly after the start of the incubation, but only after 15 min
(Fig. 4d) or 4 min (Fig. 4e). Finally, the increase in hypoxanthine after 19 hours
was even smaller, compared to the twofold diluted situation.
No remarkable differences between medium and saline were found. In the
four-fold diluted situation, complete degradation of ADP took slightly longer in
saline than in medium. Finally, as remarked in the previous paragraph, ATP
seems to start degrading somewhat later when diluted in saline rather than me-
dium.
4. Discussion
In the present paper we have optimized and applied a simple and rapid re-
versed-phase HPLC method for the simultaneous analysis of ATP, ADP, AMP,
adenosine, adenine, inosine, hypoxanthine, and uric acid in human whole blood.
Crucial steps in the collection, handling and preparation of the blood samples
Known concentrations of standard solution were added (Added conc) to a subject’s whole blood sample
with predetermined metabolite levels (Sample conc). The resulting metabolite concentration of the mix-
ture (New conc) was determined. The concentrations presented are averages of n=5 untreated blood
samples and n=5 mixtures.
Chapter 2
35
include collection in EDTA-tubes, immediate PCA addition, and neutralization
before HPLC analysis at pH 6. The faster ATP breakdown in heparin-containing
tubes, compared to EDTA-tubes, can be explained by EDTA’s Ca2+-chelating activi-
ties that inactivate the ATP-degrading enzymes(24). Addition of PCA serves the
same purpose, since PCA precipitates all proteins and thus inactivates enzymes as
well.
0
100
200
300
400
500
600
- 0248153024619
min hours
Ti me after in cubation
ATP
ADP
AMP
Hypoxanthine
Uric acid
(A)
min
Ti me after in cubation hours
0
50
100
150
200
250
300
350
400
450
- 0248153024619
min hours
Ti me after in cubation
(B)
min
Ti me after in cubation
hours
0
50
100
150
200
250
300
350
400
450
- 0248153024619
min hours
Ti me after in cubation
(C)
min
Ti me after in cubation hours
0
50
100
150
200
250
300
350
400
450
- 0248153024619
min hours
Ti me after in cubation
(D)
min
Ti me after in cubation hours
0
50
100
150
200
250
300
350
400
450
- 0248153024619
min hours
Ti me after in cubation
(E)
min
Ti me after in cubation
hours
Figure 4: Degradation profile of ATP a dded t o blood that was (A) undilut ed, (B) dilute d two times wit h
saline, (C) diluted two times with medium, (D) diluted four tim es with saline, or (E) diluted four times with
medium. X-axis (not to scale) plots the time in minutes and hours after incubation at t = 0. Concentrations
in the samples taken before additi on of the ATP spike a re rep resent ed by —. The l egend shown in A applies
to all conditions. The c oncentration (µM) of each of the detected compounds is plotted on the Y-axis.
Curves represent means and error ba rs represent standard deviation (n = 2).
Development of an HPLC method for determining ATP and metabolites
36
For sufficient inactivation of the enzymes, PCA had to be thoroughly mixed
with the blood before freezing the samples. The enzymes were properly inactivated,
as shown by the observation that no significant loss of ATP occurred when samples
were thawed on ice or at room temperature. Further preparation was not neces-
sary, thus making this method suitable for clinical applications (bedside). Chroma-
tographic separation can be difficult for compounds with multiple pKa’s. Neutrali-
sation of the samples with KOH was necessary before HPLC analysis, and eluent
pH values below six resulted in worse shape and separation of the peaks. Finally,
the stability of the samples and the standard compounds was such that large
batches of samples could be analyzed in one run, without risking degradation.
The recovery levels ranged between 82 and 108% for ATP, ADP, AMP, adeno-
sine and inosine, but were below 70% for uric acid, adenine and hypoxanthine.
These low recovery levels may be caused by coprecipitation of these metabolites
together with proteins during the centrifugation step of the acid extraction prepa-
ration of the samples or with the perchlorate precipitate after neutralization(25,26).
In whole blood, precipitate volumes are large, and the 20% adjustment of concen-
tration to correct for this volume change may have been an overcorrection for more
apolar metabolites, such as uric acid, adenine and hypoxanthine, resulting in lower
recoveries. Recovery levels reported by others for methods using perchloric acid for
the extraction of nucleotides from cells or tissues range between 75% and
120.5%(27-30). The recoveries did not differ substantially whether the nucleotides
were spiked at high or low concentrations. The whole blood concentrations of ATP,
ADP, AMP and uric acid reported in table 3 correlate well with those reported by
others(2,31).
   or all compounds) and LLOQ (all below 0.5
 e-
veloped: the quantification of ATP and its metabolites in whole blood. Concentra-
tions of ATP in whole blood are extremely high compared to plasma concentrations,
because ATP present at millimolar concentrations inside the erythrocytes is re-
leased before measuring whole blood. Erythrocyte ATP concentrations exceed
plasma ATP concentrations, which are in the nanomolar range, by over ten-
thousand-fold(32). A frequently used method to measure ATP is the luciferin-
luciferase assay(33,34) This method has the sensitivity needed for the detection of
low plasma ATP levels. A recent similar method is presented by Farthing et al.(35).
For measurement of ATP in whole blood, extreme dilutions are necessary. Howev-
er, its main disadvantage compared to our method is that only ATP can be deter-
mined, whereas we can quantify the complete adenylate pool in one run. Ion ex-
change methods often require more extensive pretreatment of samples and have
difficulties in separating nucleobases, nucleosides and nucleotides in one run, giv-
en the charged nature of the nucleotides in the operating pH range (pH 2-7)(33,36-39).
Adenosine and inosine concentrations in whole blood were below the LOD.
This is in line with results of studies investigating the adenine nucleotide content
in human whole blood and erythrocytes(40,41). Our within-day CV (range 0.7 – 5.9%)
Chapter 2
37
and between-day CV (2.6 – 15.3%) correlate well with the findings of others(42). In
the application of the method, we found that ATP added to whole blood ex vivo was
converted into several intermediate products, which, together, created a typical
degradation profile. The 30 min degradation period of ATP in the undiluted sample
is consistent with the literature data(23,43). In a similar set-up, Heptinstall et al.(23)
found that the degradation of ATP to AMP in plasma can be attributed mainly to
leukocytes, since these possess a high ectonucleotidase activity. When comparing
the whole blood model to the in vivo situation, the main difference is the absence of
vascular endothelial cells in the model. The ectonucleotidases that are active on
the luminal surface of endothelial cells, shorten the ATP half-life to seconds or less
in vivo(44). In an ex vivo set-up, but with endothelial factors present, the half-life of
ATP is longer (5-10 min), probably due to the static nature of this setup, in compar-
ison to the in vivo situation in microvasculature(45-47).
In addition to ATP, the degradation profiles of the other compounds also
showed similarities to those reported before in the literature(23), even though we
employed a longer incubation period and started at a higher ATP concentration.
First, as was also reported by Coade and Pearson(45), this profile indicates the se-
quential catabolism of ATP to ADP and AMP. ADP formation coincided with ATP
degradation, whereas the formation of AMP began four min later. The latter indi-
cates that ADP, while still being formed as a degradation product from its precur-
sor ATP, is simultaneously being degraded into AMP. This also becomes clear from
the observation that, starting 30 min after incubation, the decline in ADP concen-
tration coincides with the complete depletion of its precursor ATP. The combina-
tion of both processes leads to a strong rise in AMP concentration between 30 min
and two hours after incubation. Second, the profile shows a strong increase in hy-
poxanthine concentrations between two and nineteen hours, after which the expe-
riment was terminated. The catabolism of adenine nucleotides beyond AMP has
been shown to involve the enzyme 5’-nucleotidase (CD73) to yield inosine mono-
phosphate (IMP), inosine and hypoxanthine(48). Heptinstall et al.(23,45,49) reported
that this enzyme acts independently from leukocytes or erythrocytes, since break-
down of AMP added to cell-free plasma was similar to that in whole blood. In line
with other studies(23), no other intermediate products (adenosine and inosine), were
detected. Studies with the uptake inhibitor dipyridamole revealed that adenosine
is taken up efficiently by the erythrocytes through equilibrative nucleoside trans-
porters after which rapid sequential conversion into hypoxanthine occurs intracel-
lularly, with hypoxanthine finally being distributed outside of the cell, resulting in
the observed increased hypoxanthine levels(23,50).
In our experiment, hypoxanthine was the final degradation product and we
observed only a very small rise in the concentration of uric acid, which is the
endproduct in vivo(51). This could be explained by the probable absence of xanthine
oxidase (EC 1.2.3.2) in our model, the enzyme responsible for the further degrada-
tion of hypoxanthine to xanthine and finally uric acid. The highest activity of xan-
thine oxidase is reportedly found in the liver and intestinal mucosa(52,53), but it is
Development of an HPLC method for determining ATP and metabolites
38
also present in various organs and vascular endothelial cells(54,55). Both liver and
endothelial cells are absent in our set-up, which explains the lack of change in uric
acid levels.
Besides interest in the breakdown profile of ATP added to whole blood, we
were also interested in the effects of dilution of the whole blood with two different
media: either saline of RPMI 1640 medium. The RPMI 1640 medium is used by
Swennen et al.(22) in the whole blood model to more closely resemble the in vivo
situation. We therefore hypothesized that addition of this medium would help sta-
bilize ATP more than would a similar dilution with saline. However, degradation
profiles in blood diluted with either saline or medium were quite similar. This re-
sult indicates that the medium does not contain any substance that might delay
the degradation of ATP.
In contrast, the extent of dilution does influence the degradation profile of
ATP added to the blood. First, the delay in ATP breakdown got more pronounced
with increasing dilution of the whole blood (Fig. 4). For instance, whereas in undi-
luted blood, ATP completely degraded in 30 min, this process can take up to two
and four hours in two-fold and four-fold diluted blood, respectively. Second, other
features of the degradation profile were delayed as well, such as the decline in ADP
combined with the sharp rise in AMP that coincides with ATP depletion at 2 hours.
Furthermore, the maximum level of AMP was reached after incubation of two
hours (two-fold diluted) or after incubation of four to six hours (four-fold diluted).
5. Conclusion
The HPLC procedure described in this paper allows separation of the main
metabolites of adenine nucleotide metabolism found in human whole blood sam-
ples. Compared to previous methods, it is a simple and rapid procedure, which has
a short run time, good peak shape and high sensitivity. The method has been vali-
dated with respect to accuracy, precision, linearity and limit of detection, recovery
and stability. The method is suitable for purine analysis in vivo or in an ex vivo
set-up, enabling nucleotide metabolic processes to be followed in time. It has been
reported for long that lowered blood ATP levels have been associated with acute
disease states. Decreased levels of adenine nucleotides in erythrocytes were, for
instance, observed in patients with various malignancies(56,57). The method de-
scribed here may be used to monitor changes that occur in these patients and after
therapeutic administration of ATP.
Acknowledgements
The authors thank B. Vriens, M. de Vaan, C. van den Hurk, H. Bos and A.
Skrabanja for their contributions to the method development and M. Fischer and
R. Bartholomé for their excellent technical assistance.
Chapter 2
39
References
1. I. Baranowska-Bosiacka, A.J. Hlynczak, Comp. Biochem. Physiol. C Tox-
icol. Pharmacol. 134 (2003) 403.
2. P.D. Schweinsberg, T.L. Loo, J. Chromatogr. 181 (1980) 103.
3. G. Burnstock, Pharmacol. Rev. 58 (2006) 58.
4. F. Di Virgilio, D. Ferrari, M. Idzko, E. Panther, J. Norgauer, A. La Sala, G.
Girolomoni, Drug Dev. Res. 59 (2003) 171.
5. M.J. Bours, E.L. Swennen, F. Di Virgilio, B.N. Cronstein, P.C. Dagnelie,
Pharmacol. Ther. 112 (2006) 358.
6. H.J. Agteresch, M.H.C. van Rooijen, J.W.O. van den Berg, G.J. Minder-
man-Voortman, J.H.P. Wilson, P.C. Dagnelie, Drug Dev. Res. 60 (2003)
196.
7. T. Yamada, F. Okajima, M. Akbar, H. Tomura, T. Narita, T. Yamada, S.
Ohwada, Y. Morishita, Y. Kondo, Oncol. Rep. 9 (2002) 113.
8. A.D. Conigrave, L. van der Weyden, L. Holt, L. Jiang, P. Wilson, R.I. Chris-
topherson, M.B. Morris, Biochem. Pharmacol. 60 (2000) 1585.
9. H.J. Agteresch, P.C. Dagnelie, J.W. van den Berg, J.H. Wilson, Drugs 58
(1999) 211.
10. D.J. Hearse, Cardiovasc. Res. 18 (1984) 384.
11. K.F. Childs, X.H. Ning, S.F. Bolling, J. Chromatogr. B Biomed. Appl. 678
(1996) 181.
12. B. Pelzmann, S. Hallstrom, P. Schaffer, P. Lang, K. Nadlinger, G.D. Birk-
mayer, K. Vrecko, G. Reibnegger, B. Koidl, Br. J. Pharmacol. 139 (2003)
749.
13. G.A. Cordis, R.M. Engelman, D.K. Das, J. Chromatogr. 386 (1987) 283.
14. A. Jahraus, M. Egeberg, B. Hinner, A. Habermann, E. Sackman, A. Pralle,
H. Faulstich, V. Rybin, H. Defacque, G. Griffiths, Mol. Biol. Cell 12 (2001)
155.
15. P. Peveri, P.G. Heyworth, J.T. Curnutte, Proc. Natl. Acad. Sci. U.S.A. 89
(1992) 2494.
16. V. Reichelova, F. Albertioni, J. Liliemark, J. Chromatogr. B Biomed. Appl.
682 (1996) 115.
17. P.R. Brown, J. Chromatogr. 52 (1970) 257.
18. E. Harmsen, P.P. de Tombe, J.W. de Jong, J. Chromatogr. 230 (1982) 131.
19. D.R.Webster, G.D. Boston, D.M. Paton, J. Pharmacol. Methods 13 (1985)
339.
20. N.G. Anderson, Anal. Biochem. 4 (1962) 269.
21. H.J. Agteresch, P.C. Dagnelie, T. Rietveld, J.W. van den Berg, A.H. Dan-
ser, J.H. Wilson, Eur. J. Clin. Pharmacol. 56 (2000) 49.
22. E.L. Swennen, A. Bast, P.C. Dagnelie, Eur. J. Immunol. 35 (2005) 852.
23. S. Heptinstall, A. Johnson, J.R. Glenn, A.E. White, J. Thromb. Haemost. 3
(2005) 2331.
24. H. Holmsen, I. Holmsen, A. Bernhardsen, Anal. Biochem. 17 (1966) 456.
25. J.F. Huertas-Perez, M. Heger,H. Dekker,H. Krabbe, J. Lankelma, F. Arie-
se, J. Chromatogr. A 1157 (2007) 142.
26. A. Werner, J. Chromatogr. 618 (1993) 3.
27. D. Huang, Y. Zhang, X. Chen, J. Chromatogr. B Analyt. Technol. Biomed.
Life Sci. 784 (2003) 101.
Development of an HPLC method for determining ATP and metabolites
40
28. T. Grune, W.G. Siems, J. Chromatogr. 618 (1993) 15.
29. K.K. Tekkanat, I.H. Fox, Clin. Chem. 34 (1988) 925.
30. R.T. Smolenski, D.R. Lachno, S.J. Ledingham, M.H. Yacoub, J. Chromato-
gr. 527 (1990) 414.
31. G. Crescentini, V. Stocchi, J. Chromatogr. 290 (1984) 393.
32. R.A. Harkness, S.B. Coade, A.D. Webster, Clin. Chim. Acta 143 (1984) 91.
33. E.H. Abraham, A. Salikhova, E.B. Hug, Drug Dev. Res. 59 (2003) 152.
34. M.W. Gorman, E.O. Feigl, C.W. Buffington, Clin. Chem. 53 (2007) 318.
35. D. Farthing, D. Sica, T. Gehr, B. Wilson, I. Fakhry, T. Larus, C. Farthing,
H.T. Karnes, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 854
(2007) 158.
36. H. Daxecker, M. Raab, M. Cichna, P. Markl, M.M. Muller, Clin. Chim. Acta
310 (2001) 81.
37. J.D. Torrance, D. Whittaker, Br. J. Haematol. 43 (1979) 423.
38. D. de Korte, W.A. Haverkort, A.H. van Gennip, D. Roos, Anal. Biochem.
147 (1985) 197.
39. A. Werner, W. Siems, H. Schmidt, I. Rapoport, G. Gerber, R.T. Toguzov,
Y.V. Tikhonov, A.M. Pimenov, J. Chromatogr. 421 (1987) 257.
40. R. Caruso, J. Campolo, C. Dellanoce, R. Mariele, O. Parodi, R. Accinni,
Anal. Biochem. 330 (2004) 43.
41. V. Stocchi, L. Cucchiarini, F. Canestrari, M.P. Piacentini, G. Fornaini,
Anal. Biochem. 167 (1987) 181.
42. N. Cooper, R. Khosravan, C. Erdmann, J. Fiene, J.W. Lee, J. Chromatogr.
B Analyt. Technol. Biomed. Life Sci. 837 (2006) 1.
43. S. Heptinstall, J.R. Glenn, A. Johnson, B. Myers, A.E. White, L. Zhao, Ar-
terioscler. Thromb. Vasc. Biol. 26 (2006) e22.
44. J.W. Ryan, U. Smith, Trans. Assoc. Am. Physicians 134 (1971) 297.
45. S.B. Coade, J.D. Pearson, Circ. Res. 65 (1989) 531.
46. J.D. Pearson, J.L. Gordon, Annu. Rev. Physiol. 47 (1985) 617.
47. A.K. Price, D.J. Fischer, R.S. Martin, D.M. Spence, Anal. Chem. 76 (2004)
4849.
48. A.C. Newby, Biochem. J. 186 (1980) 907.
49. H. Zimmermann, Naunyn Schmiedebergs, Arch. Pharmacol. 362 (2000)
299.
50. D.A. Griffith, S.M. Jarvis, Biochim. Biophys. Acta 1286 (1996) 153.
51. H. Zimmermann, Drug Dev. Res. 39 (1996) 337.
52. U.A. Al-Khalidi, T.H. Chaglassian, Biochem. J. 97 (1965) 318.
53. R.W. Watts, J.E. Watts, J.E. Seegmiller, J. Lab. Clin. Med. 66 (1965) 688.
54. E.D. Jarasch, C. Grund, G. Bruder, H.W. Heid, T.W. Keenan,W.W. Franke,
Cell 25 (1981) 67.
55. Y. Moriwaki, T. Yamamoto, M. Suda, Y. Nasako, S. Takahashi, O.E. Agbe-
dana, T. Hada, K. Higashino, Biochim. Biophys. Acta. 1164 (1993) 327.
56. J. Laciak, S. Witkowski, Otolaryngol. Pol. 20 (1966) 269.
57. H. Wand, K. Rieche, Dtsch Gesundheitsw 27 (1972) 1072.
Chapter 3
Oral supplements of adenosine 5’-
triphosphate (ATP) are not bioavailable:
a randomized, placebo-controlled cross-
over study in healthy humans
Erik J.C.M. Coolen
Ilja C.W. Arts
Martijn J.L. Bours
Nathalie Huyghebaert
Martien A. Cohen Stuart
Aalt Bast
Pieter C. Dagnelie
Submitted
41
Single ATP administration study
42
Abstract
ATP is present in substantial concentrations in certain foods and in breast
milk, it is available as an orally administered drug used for analgesic purposes,
and it is being marketed as a sports aid. Several studies have reported beneficial
effects of oral ATP administration, even though its bioavailability is unclear. We
investigated whether targeted delivery of ATP to the small intestine by using en-
teric-coated pellets would lead to an increase in whole blood concentrations of ATP
and/or its metabolites.
Methods: Eight healthy volunteers participated in a cross-over study. They
were given in random order single doses of 5000 mg ATP or placebo via two types
of pH-sensitive, enteric-coated pellets (targeted at release in the proximal or distal
small intestine), or via a naso-duodenal tube. Blood ATP and metabolite concentra-
tions were monitored by HPLC for 4.5h (duodenal tube) or 7h (pellets) post-
administration. Areas under the concentration vs. time curve were calculated and
compared by paired-samples t-tests.
Results: Except for uric acid, no significant changes in ATP and metabolite
concentrations were observed. Significantly increased uric acid concentrations of
approximately 50% were found for proximal-for duo-
 -release pellets failed to
increase uric acid concentrations.
Conclusion: A single dose of orally administered ATP is not bio available as
such. However, similar increases in its final metabolite uric acid are found when
ATP is administered by proximal-release enteric coated pellets or duodenal tube.
Uric acid itself may have beneficial health effects. Further studies will be neces-
sary to determine whether prolonged daily administration of ATP will enhance its
oral bioavailability.
Introduction
The purine nucleotide adenosine 5’-triphosphate (ATP) is found in all cells of
the human body, where it functions as a source of energy and as a co-factor in cel-
lular metabolism. ATP can also be released from cells to act as a local regulator of
neurotransmission, secretory functions, inflammation, and nociception via interac-
tion with purinergic receptors(1, 2). ATP is present in substantial concentrations in a
number of foods (e.g. meat, soy, mushrooms)(3) and in breast milk(4, 5). Furthermore,
capsules containing ATP are currently registered in France for the treatment of
low back pain of muscular origin, and supplements containing ATP are marketed
on the internet for various purposes including the restoration of energy.
Several studies have found potentially beneficial effects of oral ATP adminis-
tration. In an experimental study by Jordan et al.(6), three groups of nine healthy
men received ATP (150 or 225 mg) or placebo for 14 days. Physical performance
Chapter 3
43
and muscular strength were positively affected, although no significant differences
were observed in whole blood and plasma ATP concentrations between the ATP
and placebo groups. Another study investigated the effects of 30 day supplementa-
tion with an ATP-containing registered drug (Atépadène®, 90 mg daily)(7, 8). The
questionnaire-based outcome indicated that it provided some benefit to patients
with subacute low back pain. The positive results of ATP in the study by Jordan et
al., investigating muscular power, have also promoted the commercial marketing
of ATP as an aid in sports, especially in the area of bodybuilding(6). Animal studies
reported alterations in cardiac, vascular and pulmonary function after 30 days of
oral ATP supplementation(9, 10). Whereas, like in humans(6), no changes in plasma
ATP or metabolites were detected in the systemic circulation, the concentrations of
adenine, inosine, adenosine, adenosine monophosphate (AMP) and uric acid in
plasma from the portal vein were increased. The authors concluded that these pu-
rine nucleosides can be absorbed from the intestinal lumen and secreted into the
portal bloodstream. The identification of a number of nucleoside transporters in
the small intestine further suggested that orally administered ATP can be ab-
sorbed and may be utilized by the human body(11).
We have previously shown that intravenous administration of ATP in several
randomized controlled trials prevented weight loss, fatigue, lowering of serum al-
bumin concentrations, deterioration of muscle strength, and improved functional
performance (walking stairs, household activities etc.) and overall quality of life in
lung cancer patients(12, 13). Suggested mechanisms of action included repletion of
intracellular energy stores(14) and inhibition of the acute phase response(15-17). Al-
though successful, the clear drawbacks of intravenous infusions prompted us to
explore the possibility of an oral route of administration of ATP.
We hypothesized that targeted delivery of ATP to the small intestine using
enteric coated pH-sensitive multi-particulate formulations (pellets) will lead to an
increase in whole blood concentrations of ATP and/or its metabolites. We further
hypothesized that oral administration with pellets is as effective as direct adminis-
tration of ATP into the duodenum using a naso-duodenal tube. Finally, we investi-
gated whether absorption of ATP differs between distinct regions of the small in-
testine.
Materials and methods
Study design
Subjects were examined on five occasions in a cross-over design. On days 0, 7
and 14, subjects received the following ATP doses in random order: 5000 mg ATP
as proximal-release or distal-release pellets, or placebo proximal-release pellets.
The pellets were ingested with approximately 200 ml water acidified to pH < 5
with citric acid. On days 21 and 28, subjects received in random order 5000 mg
             
Single ATP administration study
44
through a naso-intestinal tube. The tube was inserted through the subjects’ nostril
and placed in the stomach. To promote movement of the tube through the pylorus
into the duodenum, subjects were asked to lay down on their right side. To verify
the tube’s position (either stomach or duodenum), gastro-intestinal juice samples
were taken by a syringe and tested for their pH and color. Once pH was above 5
(±180 min after insertion of the tube), and color was yellow, administration started
and the tube was removed 10 min later.
Study population
Male and female subjects (18-60 years) received oral and written information
about the protocol and possible risks before signing informed consent. Exclusion
criteria were a history of lung, heart, intestinal, stomach or liver disease, use of
prescription medication, smoking, drug use, dietary restrictions, and pregnancy.
Subjects abstained from products containing alcohol or caffeine and from purine-
rich foods, such as game, offal, sardines, anchovies and alcohol-free beer for two
days before each test day. Subjects fasted from 10 p.m. the previous day until the
end of the test day (4 p.m.), and refrained from any vigorous physical activity 24 h
before each test. Subjects were allowed to drink water starting 30 min after ATP or
placebo administration. The study was approved by the Medical Ethics Committee
of Maastricht University Medical Centre. The study was carried out according to
the Helsinki Declaration for human experiments.
Materials
ATP disodium salt was purchased from Pharma Waldhof GmbH, Düsseldorf,
Germany. Adenosine diphosphate (ADP) disodium salt, AMP sodium salt, adenine,
inosine, hypoxanthine, uric acid and nitric acid were purchased from Sigma Chem-
ical Co., St. Louis, USA. Adenosine and lithium carbonate (Li2CO3) were obtained
from Fagron BV., Uitgeest, The Netherlands. Perchloric acid (PCA) 70% solution in
water was purchased from Sigma-Aldrich, Steinheim, Germany. KOH, KH2PO4,
K2CO3, K2HPO3*3H2O and NaOH were obtained from Merck, Darmstadt, Germany
and 0.9% saline from Braun, Melsungen, Germany. Bengmark-type naso-duodenal
tubes were from Flocare, Zoetermeer, The Netherlands. NH4NO3 was obtained
from Fluka, Steinheim, Germany. Trichloroacetic acid (TCA) 20% solution in water
was from Serva, Heidelberg, Germany. Citric acid suitable for human consumption
was obtained from the pharmacy of Maastricht University Medical Centre.
Production of pellets
ATP pellets were produced at Ghent University, Faculty of Pharmaceutical
Science, Belgium as described by Huyghebaert et al.(18), with minor modifications
to obtain an ATP concentration of >40% (wt:wt) after coating. Placebo pellets were
produced similarly, but without ATP. To investigate the timing of intestinal re-
lease, Li2CO3 (60 mg per administration) was added to the pellets.
The proximal-release pellets were coated with 30% Eudragit® L30D-55 (ATP
or placebo pellets), and the distal-release pellets (ATP only) were coated with 15%
Chapter 3
45
Eudragit® FS 30 D (Röhm Pharma, Darmstadt, Germany), mixed with anionic
copolymers of methacrylic acid and ethylacrylate (1:1). After coating, the pellets
were cured overnight at room temperature at 60% (proximal-release pellets) or
20% (distal-release pellets) humidity, packed in aluminium foil sachets (VaporF-
lex®, LPS, NJ, USA), sealed at their respective humidity and stored at room tem-
perature. Pellets were used within 3 months after production.
Dissolution testing
To test whether the coating of the pellets was adequate, a dissolution test (n

(USP apparatus 3 from Bio-Dis, VanKel, NJ, USA) at a dip rate of 21 dips per
minute using 3 g pellets per vessel (250 ml) with two consecutive media: 0.1 N HCl
 2PO4    
the proximal-release pellets, and pH 7.4 for the distal-release pellets. Samples
were collected after 2 h in HCl and after 2, 5, 10, 20, 30 and 60 min in buffer as
described in Huyghebaert et al.(18). ATP and metabolite concentrations were meas-
ured by HPLC separation and UV-analysis as previously described(19).
Sample collection during the intervention
Venous blood was collected from the antecubital vein by a 20 gauge intraven-
ous catheter (Terumo-Europe NV, Leuven, Belgium), connected to a three-way
stopcock (Discofix®, Braun Melsungen AG, Melsungen, Germany). Blood was col-
lected into 4 ml EDTA tubes (Venosafe, Terumo-Europe NV) by inserting a 21
gauge multisample needle (Venoject Quick Fit, Terumo-Europe NV) into the mem-
brane of a closing cone (IN-Stopper, Braun Melsungen AG) that was attached di-
rectly to the stopcock. The anticoagulant EDTA inhibits the extracellular hydroly-
sis of ATP by Ca2+- and Mg2+-activated enzymes, like plasma membrane-bound
CD39(20). To avoid clotting after each blood collection, approximately 1.5 ml of he-
parinised (50 I.E./ml) 0.9% saline was used to rinse the blood collection set-up. It
was removed before the next blood collection.
Three baseline blood samples were collected at 30, 20 and 10 min before ad-
ministration. Starting 30 min after pellet administration or 15 min after duodenal
administration, blood samples were collected every 15 min. Between 210 and 420
min (pellets) or 270 min (duodenal tube) after administration, samples were col-
lected every 30 min. Total volume collected per day was 92 ml.
After blood collection, the tubes were inverted three times and put on ice.
Five hundred µl of blood was added to 500 µl ice-cold PCA (8% wt:v), vortex-mixed
and frozen in liquid nitrogen. To determine the concentration of ATP and all meta-
bolites except uric acid in the erythrocyte fraction, while measuring whole blood
samples, the measured concentrations were divided by the hematocrit (determined
using a microhematocrit method(21)). Untreated plasma samples (centrifugation at
     rom the
pellets. All samples were stored at -
Single ATP administration study
46
ATP measurement in whole blood samples by HPLC
Equipment, sample preparation and measurement conditions have been pre-
viously described and validated(19). Briefly, after thawing, the protein fraction was
2CO3 in 6 M KOH was added
to 650 µl supernatant to neutralize the pH. The resulting insoluble perchlorate was
      as
mixed with 160 µl 0.05 M phosphate buffer pH 6.0 in HPLC vials.
Lithium measurement in plasma samples
To investigate the timing of pellet disintegration, plasma concentrations of
the lithium marker were examined in plasma using a modified Trapp protocol(22).
Following thawing on ice, 50 µl plasma was vortex-mixed with 10 µl trichloroacetic
acid (20% v:v) and centrifuged (14,000 rpm, 10 min) to precipitate the proteins. The
supernatant was diluted 20x in 0.1 M nitric acid, which also served as the blank.
Two replicate measurements per sample were performed on a SpectrAA 400 gra-
phite tube atomic absorption spectrophotometer (AAS) (Varian, Palo Alto, CA,
USA) with a lithium hollow-cathode lamp, operated at 5 mA and a 1.0 nm slit.
Peak height measurements at 670.8 nm wavelength were compared with values for
standards of known concentrations (ranging from 2-10 ng/ml). Initially 20 µl sam-
ple and 5 µl modifier solution (1.2 M NH4NO3) were injected into the top hole of the
graphite tube. Then, fluids were evaporat

time. If the obtained signal exceeded the standard concentration range (0-10
ng/ml), samples were diluted with blank and measured again.
To evaluate lithium content of the pellets before use, uncoated placebo-pellets
were mixed thoroughly in buffer and samples were taken at 10, 20, 30 and 60 min
after dissolution. These samples were analysed for lithium content using the above
protocol.
Statistical analysis
The area under the concentration vs. time curve (AUC) was calculated using
the linear trapezoidal rule from time zero until the last time point of sampling t
(AUC0-t). Cmin and Cmax were defined as the minimum and maximum observed con-
centrations, respectively. tmax was the time of Cmax. AUC of the five conditions were
compared and analyzed by paired-samples t-tests. P<0.05 was considered statisti-
cally significant.
Chapter 3
47
Results
Study population
Eight subjects (6 females, 2 males aged 26.9 ± 5.9 (mean ± SD), weighing 70 ±
4.3 kg, with a BMI of 23.6 ± 1.3) completed the trial. No adverse events were ob-
served with both modes of administration (i.e. pellets, solution).
Dissolution testing
First, the lithium content of the uncoated pellets was verified in vitro by AAS
to be 0.6% (wt:wt) (data not shown). Next, the coating properties and ATP-yield of
the pellets were tested in a dissolution experiment lasting 180 min. Fig. 1 shows
the percentage of ATP that was released, with 100% representing the concentra-
tion of the sum of ATP plus its metabolites at 180 min. During 120 min in 0.1 N
HCl, less than 5% ATP and metabolite (5.0 ± 0.6% for the proximal-release pellets
and 3.4 ± 0.4% for the distal-release pellets) was released. Subsequent rapid chang-
ing of the buffer solutions to pH 6.5 or 7.4 for 60 min caused a release of 50% of the
remaining ATP within 5 min (proximal-release pellets) or 25 min (distal-release
pellets), which increased to >80% after sixty min. ATP was partially broken down
to ADP (8.6% for proximal-release pellets and 7.0% for distal-release pellets), AMP
(1.0 and 0.7%, respectively), and uric acid (4.0 and 2.5%, respectively).
Figure 1: Release profiles of ATP and metabolites from proximal-release (closed symbols) and distal-
release (open symbols) pellets after 120 min in 0.1 N HCl (time axis has been shortened between 0 and
120 min) and subsequently 60 min in buffer solutions with either pH 6.5 (proximal-release pellets) or 7.4
(distal-release pellets). The Y-axis represents the percentage of the total amount of ATP and metabolites
released after 60 min. Data were obtained by the reciprocating cylinder method (USP apparatus 3). Val-
ues are means ± SEM, n = 3
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
120 130 140 150 160 170 180
Time (min)
ATP and metabolite release (%
)
proximal-release pellets (ATP)
distal-release pellets (ATP)
proximal-release pellets (ADP + AMP + uric acid)
distal-release pellets (ADP + AMP + uric acid)
Single ATP administration study
48
Lithium in plasma
Fig. 2 depicts mean plasma lithium concentrations in samples collected for 7
h after administration of the coated pellets. The three types of pellets had different
release profiles, as was quantified by measuring the AUC (table I). Comparison of
the AUC of the ATP-containing pellets revealed that the proximal-release pellets
caused a significantly higher increase in plasma lithium than the distal-release
pellets (P -release pellets with
or without ATP, showed that the lithium AUC was significantly lower in the ATP-
containing pellets than in the placebo-containing ones (P   
plasma lithium concentrations are depicted in Supplemental fig. 1 (after refer-
ence section). Lithium Cmax for the proximal release pellets was reached between
135 and 210 min after administration at a mean concentration of 404 ng/ml for the
placebo pellets and 200 ng/ml for the ATP pellets. The highest plasma lithium con-
centration (717 ng/ml) was measured in a volunteer receiving placebo proximal-
release pellets. The distal-release pellets, on the other hand, showed a delayed and
TABLE 1. EFFECT OF AT P VS. PLACEBO ADMINISTRATION ON URIC ACID AND LITHIUM CONCENTRATIONS
Formulation
 AUC uric acid
mmol.min/L
Cmax
mmol/L (range)
tmax
min (range)
AUC Lithium
mmol.min
Duodenal tubeg
ATP 270 min 19.6 ± 4.4 a,b,c 0.31 ± 0.03 135 n.a.
(0.23-0.38) (105-240)
Placebo 270 min -0.4 ± 0.4 0.21 ± 0.03 n.a. n.a.
(0.15-0.33)
Proximal release pelletsh
ATP 270 min 16.1 ± 3.0 n.a. n.a. n.a.
Placebo 270 min 0.8 ± 0.9 n.a. n.a. n.a.
ATP 420 min 25.4 ± 5.7d,e 0.30 ± 0.03 240 65174 ± 7985f
(0.21-0.41) (165-390)
Placebo 420 min 0.9 ± 1.1 0.20 ± 0.02 n.a. 117914 ± 15021f
(0.16-0.31)
Distal release pelletsi
ATP 270 min 1.7 ± 1.1 n.a. n.a. n.a.
ATP 420 min 3.2 ± 1.4 0.22 ± 0.02 390 12575 ± 2832f
(0.17-0.34) (105-420)
Values are group means ± SEM, n = 8 per formulation, paired-samples t-tests.
a Different from duodenal tube placebo (P=0.002), b Different from ATP distal-release pellets 270 min
(P=0.007), c Different from proximal-release placebo pellets 270 min (P=0.007) d Different from ATP distal
release pellets 420 min (P=0.005), e Different from proximal-release placebo pellets (P=0.005), f Different
from each other (P<0.001) g 5000 mg ATP dissolved in 100 ml water or placebo was administered to
healthy volunteers using a naso-duodenal tube. h 5000 mg ATP or placebo was administered to healthy
volunteers by means of an enteric coated pH-sensitive multi-particulate formulation that will release its
content once pH is above 6.5. i 5000 mg ATP or placebo was administered to healthy volunteers by
means of an enteric coated pH-sensitive multi-particulate formulation that will release its content once
pH is above 7.4.
Chapter 3
49
lower release profile, with lithium concentrations starting to rise only approx-
imately 240 min after administration, while a maximum concentration of 103
ng/ml was reached at the final measurement.
ATP and metabolites in blood
HPLC analysis of the whole blood ATP concentration revealed no statistically
significant differences between placebo and ATP for any type of administration
(data not shown). In fact, of the other metabolites (ADP, AMP, adenosine, adenine,
inosine, hypoxanthine and uric acid), only uric acid concentrations changed (Fig.
3). Compared to placebo, the uric acid AUC increased significantly when ATP was
administered by proximal-release pellets (P       P 
0.001). Administration of ATP by distal-release pellets did not lead to a significant-
ly increased uric acid AUC, compared to placebo. The peak uric acid concentrations
(Cmax) were 36% higher (0.28 ± 0.02 mmol/L) for proximal-release pellets compared
to distal-release pellets (0.21 ± 0.01 mmol/L), but 6% lower compared to the admin-
istration via duodenal tube (0.30 ± 0.02 mmol/L) (Fig. 3 and statistics in table I).
The mean time to peak uric acid concentration (tmax) was shorter for duodenal tube
Figure 2: Plasma lithium concentrations after administration of a dose of pellets containing 60 mg
Li2CO3 to healthy volunteers. Following ingestion of the proximal-release pellets without ATP (----),
the plasma lithium concentration rose earlier and higher compared to pellets with the same coating
but with ATP (----). Following ingestion of proximal-release pellets, higher plasma lithium concentra-
tions were attained compared to distal-release pellets (----). Significant differences between the pel-
lets are based on the AUC and can be found in table I. Values are means ± SEM, n = 8.
0
50
100
150
200
250
300
350
400
450
-30 0 30 60 90 120 150 180 210 240 270 300 330 360 390 420
Time (min)
Plasma Lithium (ng/ml)
proximal-release ATP
proximal-release placebo
distal-release ATP
Single ATP administration study
50
administration (tmax ranged from 75 to 195 min with mean ± SD 135 ± 15 min) as
compared to the pellet administration (tmax ranged from 150 to 390 min with mean
± SD 234 ± 32 min). An overview of the inter-subject variability in uric acid concen-
trations following administration of ATP (tube and pellets) is presented in Sup-
plemental Figure 2 (after references).
Discussion
The aim of this study was to investigate whether targeted delivery of ATP to
the small intestine using two types of enteric coated pH-senstive multi-particulate
formulations (pellets) would lead to an increase in whole blood concentrations of
ATP and/or its metabolites, and to compare this to direct administration of ATP
into the duodenum using a naso-duodenal tube. Although the ATP dosages admi-
nistered in our study (corrected for body weight 55.6 - 83.3 mg/kg) exceed those of
other oral administration studies, we observed no changes in erythrocyte ATP con-
centrations. Kichenin et al. orally administered ATP in dosages of up to 20 mg/kg
per day to rabbits and 5 mg/kg per day to rats(9, 10). In humans, ATP dosages rang-
ing from 36 to 108 mg/kg per day were administered intravenously by Haskell et
al. and Agteresch et al.(12, 13, 17). Of the ATP metabolites considered, only uric acid
concentrations increased significantly after administration of the proximal-release
pellets and the duodenoal tube, but not of the distal-release pellets. Assuming that
uric acid is primarily present in the extracellular fluid (the volume of which is ap-
proximately 22% of body weight), that the 5000 mg ATP is completely broken down
to 9.06 mmol uric acid, and that there is no loss of uric acid due to excretion, the
Figure 3: Uric acid concentrations in healthy volunteers supplemented at t=0 with a single dose of
5000 mg ATP or placebo. Data are presented as percentages increase from the mean of three blood
samples taken before administration. Significant differences between the administrations are based
on the AUC and can be found in Table I. Values are means ± SEM, n = 8
-10
0
10
20
30
40
50
60
70
0 30 60 90 120 150 180 210 240 270 300 330 360 390 420
Time after administration (min)
increase from baseline (%)
ATP by duodenal tube
placebo by duodenal tube
ATP by proximal-release pellets
ATP by distal-release pellets
placebo by proximal-release pellets
Chapter 3
51
estimated ‘bioavailability’ of ATP (defined as the observed uric acid increase as a
percentage of the theoretical maximum) was 16.6 ± 2.3% for the duodenal tube,
14.9 ± 2.5% for the proximal-release pellets and 3.2 ± 0.6% for the distal-release
pellets.
When ATP is released in the small intestine, ecto-nucleotidase triphospha-
tase diphosphohydrolases present on the luminal side of enterocytes dephosphory-
late ATP via ADP to AMP(23), after which ecto-5’-nucleotidase (CD73) degrades
AMP to adenosine(24). In mice, the terminal ileum is the site in the intestine with
the lowest ATPase activity(25). Although information on the human intestine is li-
mited, this may explain the difference in plasma uric acid concentrations after in-
gesting the proximal or distal-release pellets. Concentrative (CNT) and equilibra-
tive (ENT) nucleoside transporters are able to transport nucleosides into the intes-
tinal enterocytes and to the capillary bed of the intestinal villi. CNTs exhibit a
proximal-distal gradient with highest transport activities present in the jeju-
num(26). Finally, adenosine is taken up by the erythrocytes through ENTs in the
erythrocyte membrane(27). In vivo studies in animals and humans indicated that
inside the erythrocytes adenosine can be used for the synthesis of ATP(12, 13). In our
study, neither ATP nor adenosine concentrations were increased, suggesting that
instead of being used for ATP synthesis in the erythrocytes, orally administered
ATP is degraded to uric acid by xanthine oxidase, an enzyme which is expressed
mainly in the liver and in endothelial cells of blood vessels(28).
In our study, the increase in plasma uric acid concentration was similar for
the proximal-release pellets and the duodenal tube, indicating complete release of
ATP from the pellets. The delay in uric acid increase of about 1 h following prox-
imal-release pellet administration compared to duodenal tube administration is
probably a combined effect of gastric residence time and the time needed for disso-
lution of the coating of the pellets. We used enteric pH-sensitive coated pellets be-
cause they were previously successfully used for the targeted delivery of various
compounds(29-31). Pellets were preferred over capsules, because they ensure a short
and predictable gastric residence time (mean <1 h in fasted state(32)), are easier to
swallow(33, 34) and distribute more homogeneously in the gastrointestinal tract, im-
proving the absorption possibilities for large molecules, like ATP(31). The pH-
sensitive Eudragit® polymer coating provided sufficient gastroresistance, as un-
wanted in vitro release of ATP from the pellets was within the limits set by the
USP (i.e. <10% drug release in 2 h in 0.1 N HCl)(35). In vivo, the intestinal pH and
transit times are the main factors determining the location where each type of
coating releases its contents. The duodenum has a pH of 6.4 with a mean transit
time to the jejunum of 30 min, while in the ileum, the pH rises to 7.4 with a transit
time to the colon for pellet dosage forms in fasted individuals of approximately 3 ±