A highly sensitive CE-UV method with dynamic coating of silica-fused capillaries for monitoring of nucleotide pyrophosphatase/phosphodiesterase reactions

Research Article

A highly sensitive CE-UV method withdynamic coating of silica-fused capillariesfor monitoring of nucleotidepyrophosphatase/phosphodiesterasereactions

A new highly sensitive capillary electrophoresis (CE) method applying dynamic coating

and on-line stacking for the monitoring of nucleotide pyrophosphatases/phosphodies-

terases (NPPs) and the screening of inhibitors was developed. NPP1 and NPP3 are

membrane glycoproteins that catalyze the hydrolysis nucleotides, e.g. convert adenosine

50-triphosphate to adenosine 50-monophosphate (AMP) and pyrophosphate. Enzymatic

reactions were performed and directly subjected to CE analysis. Since the enzymatic

activity was low, standard methods were insufficient. The detection of nanomolar AMP

and other nucleotides could be achieved by field-enhanced sample injection and the

addition of polybrene to the running buffer. The polycationic polymer caused a dynamic

coating of the silica-fused capillary, resulting in a reversed electroosmotic flow. The

nucleotides migrated in the direction of the electroosmotic flow, whereas the positively

charged polybrene molecules moved in the opposite direction, resulting in a narrow

sample zone over a long injection time. Using this on-line sensitivity enhancement

technique, a more than 70-fold enrichment was achieved for AMP (limit of detection,

46 nM) along with a short migration time (5 min) without compromising separation

efficiency and peak shape. The optimized CE conditions were as follows: fused-silica

capillary (30 cm effective length� 75 mm), electrokinetic injection for 60 s, 50 mM

phosphate buffer pH 6.5, 0.002% polybrene, constant current of �60 mA, UV detection at

210 nm, uridine 50-monophosphate as the internal standard. The new method was used

to study enzyme kinetics and inhibitors. It opens an easy way to determine the activities

of slowly metabolizing enzymes such as NPPs, which are of considerable interest as

novel drug targets.

Keywords:

Capillary electrophoresis / Dynamic coating / Hexadimethrine bromide /Nucleotide pyrophosphatases / Phosphodiesterases

DOI 10.1002/elps.200800013

1 Introduction

The ecto-nucleotide pyrophosphatase/phosphodiesterase

(E-NPP; EC 3.1.4.1, EC 3.6.1.9) family is constituted of

ubiquitous and conserved eukaryotic proteins that exist both

as membrane glycoproteins, with an extracellular active site,

and as soluble proteins in body fluids [1]. Three nucleotide-

hydrolyzing mammalian NPPs are known: NPP1 (PC-1),

NPP2 (PD-Ia, autotaxin), and NPP3 (gp130RB13-6, B10, PD-

Ib) [2]. NPP1 and NPP3 are more closely related to each

other (50% identity) than to NPP2 (39–41% identity) [3].

NPP1–3 reveal a surprisingly broad substrate specificity,

being capable of hydrolyzing nucleotides, dinucleotides, and

nucleotide sugars; for example, ATP, ADP, NAD1, ADP-

ribose, and diadenosine tetraphosphate all yielding AMP as

a hydrolysis product [4]. Both purine and pyrimidine

nucleotides may serve as substrates [5]. In addition to

nucleotides, NPP2, but not NPP1 and NPP3, also

hydrolyzes lysophosphatididylcholine yielding lysophospha-

tidic acid [6]. Thymidine 50-monophosphate p-nitrophenyl

ester (TMP-pNP) is often used as a synthetic substrate of

NPPs [7, 8]. Phosphorothioate-modified oligonucleotides,

Jamshed Iqbal1

Sebastien A. Levesque2

Jean Sevigny2

Christa E. Muller1

1Pharmaceutical Sciences Bonn(PSB), Pharmaceutical Institute,Pharmaceutical Chemistry I,University of Bonn, Bonn,Germany2Centre de Recherche enRhumatologie et Immunologie,Centre Hospitalier Universitairede Quebec, Universite Laval,Quebec, QC, Canada

Received January 7, 2008Revised February 15, 2008Accepted March 2, 2008

Abbreviations: (E)-NPP, (ecto-)nucleotide pyrophosphatase/phosphodiesterase; RB2, reactive blue 2; TMP-pNP,

thymidine 50-monophosphate p-nitrophenyl ester

Correspondence: Dr. Christa E. Muller, Pharmaceutical SciencesBonn (PSB), Pharmazeutisches Institut, PharmazeutischeChemie I, An der Immenburg 4, D-53121 Bonn, GermanyE-mail: [email protected]: 149228732567

& 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

Electrophoresis 2008, 29, 3685–3693 3685

which are currently developed as antisense therapeutics, can

also be degraded by NPPs [9].

NPP1–3 have been implicated in various biological

processes, including bone mineralization, signaling by

insulin [10] and by nucleotides, and the differentiation and

motility of cells [2, 5, 6]. NPP3 expression is associated with

carcinogenesis and metastasis of cancer cells and has

therefore been proposed as a tumor marker [11–13]. NPP

expression has been reported to be increased in membranes

of aged rat brains (NPP1) and in brain cortex of Alzheimer’s

disease patients (NPP2), and NPP inhibitors have therefore

been proposed as novel therapeutics for neurodegenerative

diseases [7, 14]. In order to explore the potential of selective

NPP inhibitors as novel therapeutics, e.g. as anti-cancer,

anti-metastatic, or anti-neurodegenerative drugs, selective

inhibitors have to be developed. To investigate the inhibitory

potential of a library of compounds, to identify lead struc-

tures, and to optimize them as subtype-selective NPP inhi-

bitors, a fast and easy screening method is required.

Current methods for measuring NPP activities include

(i) radioactive methods applying [3H]- or [32P]-labeled

substrates [8, 15] followed by TLC separation of the radio-

active products (e.g. [32P]pyrophosphate produced from

[g-32P]ATP) [15]; (ii) spectrophotometric assays using TMP-

pNP as a substrate, which yields p-nitrophenolate upon

cleavage [7, 8, 16, 17], and (iii) HPLC methods usually in the

presence of an ion pairing reagent, such as tetra-

butylammonium hydrogen sulfate, coupled with UV or

fluorescence detection [7, 18, 19]. Although low limits of

detection (LODs) are achievable with radioactive methods,

this technique is laborious and time-consuming, radioactive

substrates are required, which are expensive, and – in the

case of 32P – have a short half-life, and safety issues related

to radioactivity can be problematic. Spectrophotometric

assays are restricted to the investigation of artificial

substrates (usually TMP-pNP) and they require large

amounts of material (enzymes, substrates, and test

compounds). HPLC methods require intensive sample

pretreatment in order to remove proteins before injection,

e.g. two centrifugation steps, 5 min each, at 10 000� g [18],

or a 20-min centrifugation at 14 500� g [3]. Furthermore, for

HPLC, large quantities of solvents are required and, in

addition, prices for columns are relatively high. Since the

monitoring of NPPs requires, particularly, high analytical

sensitivity, fluorimetric detection is often preferred to UV

detection. However, this requires artificial fluorescent

substrates, such as etheno-adenine nucleotides (e.g. etheno-

ATP) [7, 19], which limits the application of HPLC-LIF

(laser-induced fluorescence).

An alternative method for assaying enzyme activity with

great potential for drug screening is capillary electrophoresis

(CE) [20]. Recently, we have developed CE-based enzyme

assays for the determination of Michaelis–Menten constants

(Km values), maximal velocities (Vmax) for substrates,

and inhibition constants (IC50 or Ki values) for enzyme

inhibitors of various nucleotide and nucleoside-metaboliz-

ing enzymes, namely ecto-50-nucleotidase [21], nucleoside

triphosphate diphosphohydrolase (ecto-NTPDase) [22],

adenosine kinase [23], and herpes simplex type 1 (HSV-1)

thymidine kinase [24]. However, the previously established

methods were not sensitive enough to quantify the reaction

products of NPP reactions, because their detection limits for

nucleotides such as AMP, the product of NPP-catalyzed

hydrolysis of ATP, are only in the range of 1.6–2.3 mM.

There are several ways to lower the detection limits in

CE analysis: LIF and electrochemical detection are more

sensitive than UV detection, but they require special

substrates. Another solution to improve the LOF is to inte-

grate an on-line preconcentration process for sample

analytes. Several groups have developed on-capillary

preconcentration techniques, including field-amplified

stacking [25], field-enhanced sample injection [26], dynamic

pH junction stacking [27], sweeping [28], and high-salt

stacking [29]. Sample preconcentration offers increased

sensitivity, robust electrokinetic injection schemes, and the

possibility to use detection modes that are less sensitive than

LIF [25, 30]. The electrical current cannot be used only for

separation but also for sample concentration directly in the

capillary [31]. Using the electrokinetic sample injection

method, the sample is concentrated at the tip, thereby

leaving the whole capillary for separation, which results in a

higher plate number [32].

In the current approach a polycationic polymer, hexa-

dimethrine bromide (polybrene), was used as a buffer

additive to form a dynamical coating on the fused-silica

capillary wall. The positively charged capillary wall produced

a reversed electroosmotic flow (EOF) (from cathode to

anode). Since the negatively charged analytes migrate to the

anode under an applied electric field, a reversed polarity of

voltage from cathode to anode had to be employed [33, 34].

By electrokinetically injecting a large amount of sample and

by using the reverse polarity mode, we obtained a sample

stacking effect. Upon application of voltage, the positively

charged polybrene molecules migrated toward the cathode

while the negatively charged analytes migrated toward the

anode resulting in a rapid and high degree of sample

concentration at the capillary tip, while leaving essentially

the whole capillary volume for separation. Thus, the analysis

with this type of injection is rapid, gives a high degree of

concentration, and results in a very good resolution and

sensitivity allowing a fast, inexpensive, and convenient

monitoring of NPP reactions.

2 Materials and methods

2.1 Materials

Adenosine 50-triphosphate (ATP), uridine 50-monopho-

sphate (UMP), adenosine 50-monophosphate (AMP), and

hexadimethrine bromide (polybrene) were obtained from

Sigma (Steinheim, Germany). Magnesium chloride and Tris

(Trizma base) were also from Sigma. Dipotassium hydrogen

phosphate was obtained from Fluka (Neu-Ulm, Germany).

Electrophoresis 2008, 29, 3685–36933686 J. Iqbal et al.


DMEM/F-12, Lipofectamine and fetal bovine serum (FBS)

were purchased from Invitrogen (Burlington, ON, Canada).

Plasmids encoding human NPP1 and NPP3 were obtained

from Dr. Goding and Dr. Sano, respectively [35,36].

2.2 Cell culture and transfection of NPP1 and NPP3

Human NPPs were produced by transiently transfecting

COS-7 cells in 10-cm plates using Lipofectamine, as

previously described [37]. Briefly, 70–90% confluent cells

were incubated for 5 h at 371C in DMEM/F-12 in the

absence of FBS with 6 mg of plasmid DNA and 24 mL of

Lipofectamine reagent. Then, an equal volume of DMEM/F-

12 containing 20% FBS was added, and 40–44 h later cells

were collected.

2.3 Preparation of membrane fractions

NPP1- and NPP3-transfected cells were washed three times

with Tris–saline buffer at 41C, harvested by scraping in

95 mM NaCl, 0.1 mM PMSF, and 45 mM Tris buffer, pH

7.5, and washed twice by centrifugation at 300� g for 5 min

at 41C. Cells were resuspended in the harvesting buffer

containing 10 mg/mL aprotinin and sonicated. Nuclear and

cellular debris were discarded after another centrifugation

step. Glycerol was added to the resulting supernatant at a

final concentration of 7.5%. Samples were kept at �801C

until used. Protein concentration was estimated by the

Bradford microplate assay, with bovine serum albumin as

the standard reference [38].

2.4 Preparation of standard solutions and calibration

curves

Nucleotides (ATP, AMP, and UMP) were dissolved in

deionized water to obtain 10.0 mM stock solutions. These

were further diluted to obtain 1.0 mM solutions in assay

buffer (1 mM MgCl2, 2 mM CaCl2, 10 mM HEPES, pH 7.4).

The 1 mM solutions were further diluted in the same buffer

as required for standard calibration curves and enzyme

assays. Injections of standards were performed in triplicate.

Calibration curves were obtained by plotting the corrected

peak areas of the nucleotide peaks against their concentra-

tions.

2.5 CE apparatus and conditions (standard method)

The experiments were performed on a P/ACE 5500

instrument (Beckman Coulter Instruments, Fullerton, CA)

equipped with a diode-array detection system. The electro-

phoretic separations were carried out using eCAP uncoated

fused-silica capillaries of 37 cm total length (30 cm effective

length)� 75 mm internal diameter (id)� 375 mm outside

diameter (od) obtained from Beckman Coulter. The separa-

tion was performed using an applied constant voltage of

10 kV and a data acquisition rate of 8 Hz. Analytes were

detected using direct UV absorbance at 210 nm. The CE

instrument was fully controlled through a PC, which

operated with the analysis software PACE Station, obtained

from Beckman Coulter. The evaluation of the electropher-

ograms was done using the same software. The capillary

temperature was kept constant at 251C. The temperature of

the storing unit was adjusted to 251C. The running buffer

consisted of 100 mM SDS, 20 mM sodium phosphate, pH

7.5. Samples were introduced into the capillary by hydro-

dynamic injection (0.5 psi for 5 s). Analyses were performed

with normal electrode polarity. Between separations, the

capillary was washed with 0.1 M aqueous NaOH for 2 min,

deionized water for 1 min, and running buffer for 1 min.

2.6 CE apparatus and conditions (new method)

All experiments were carried out using a P/ACE MDQ CE

system (Beckman Instruments) equipped with a UV

detection system coupled with a diode-array detector. Data

collection and peak area analysis were performed by the

P/ACE MDQ software 32 KARAT obtained from Beckman

Coulter. The temperature was as in the standard method

(see Section 2.5). The electrophoretic separations were

carried out using an eCAP fused-silica capillary (40 cm

(30 cm effective length)� 75 mm (id)� 375 mM od obtained

from Beckman Coulter). The separations were performed

using an applied current of �60 mA and a data acquisition

rate of 8 Hz. Analytes were detected using direct UV

absorbance at 210 nm. The capillary was conditioned by

rinsing with 0.1 M aqueous NaOH solution for 2 min, water

for 1 min, and subsequently with buffer (phosphate 50 mM,

polybrene 0.002%, pH 6.5) for 1 min. Electrokinetic sample

injections were made at the cathodic side of the capillary.

The LOD was calculated at a signal-to-noise ratio equal to 3,

while the limit of quantitation (LOQ) was calculated at a

signal-to-noise ratio equal to 10.

2.7 NPP enzyme assays

For the determination of the kinetic parameters (Km and

Vmax) of nucleotide pyrophosphatase/phosphodiesterases,

NPP1 and NPP3 by CE seven different substrate concentra-

tions of ATP were used, 1, 2, 5, 10, 20, 100, and 400 mM. The

enzymatic reaction was performed in a test tube in a final

volume of 100 mL. The reaction buffer consisted of 10 mM

HEPES, 1 mM MgCl2, and 2 mM CaCl2, brought to pH 7.4

by adding the appropriate amount of 1 M aqueous HCl

solution. The reactions were initiated by adding 5 mL of

NPP1 (2.0 mg) or NPP3 (2.1 mg) membrane preparation, and

was then allowed to proceed at 371C for 20 min. The

reaction was stopped by heating the mixture at 991C for

5 min. All nucleotides were stable under these conditions.

Electrophoresis 2008, 29, 3685–3693 CE and CEC 3687


Aliquots of the reaction mixture (50 mL) were then

transferred to mini-CE vials containing 450 mL of a solution

of the internal standard UMP in water (final concentration

0.5 mM). Each analysis was repeated twice (duplicates) in

three separate experiments. The absorbance at 210 nm was

monitored continuously and the nucleotide concentrations

were determined from the area under each absorbance peak.

NPP inhibition assays were carried out at 371C in a final

volume of 100 mL. The reaction mixture contained 1 mM

MgCl2, 2 mM CaCl2, 10 mM HEPES, pH 7.4, and 200 mM

ATP. Solutions (10 mL) of NPP inhibitors (various concen-

trations) in enzyme assay buffer were added, and the reac-

tion was initiated by the addition of 5 mL of appropriately

diluted membrane preparation of either NPP1 or NPP3. The

mixture was incubated for 20 min and terminated by heat-

ing at 991C for 5 min. The reaction mixture was then

transferred to mini-CE vials and injected into the CE

instrument under the conditions described above. Each

analysis was repeated three times.

3 Results and discussion

CE is a separation technique that has emerged as a fast, low-

cost, and powerful separation technique for both charged

and neutral analytes [23]. It has a high mass sensitivity in

relation to the small (typically nanoliter) injection volume

used. However, because of the short optical path length

within the detection cell, the lowest detectable concentration

in CE with UV absorption detection is in the 1–10-mM range

[23, 39, 40]. This concentration sensitivity is lower than

with HPLC. It can usually be improved by using a Z-shaped

or bubble-shaped detection cell or alternatively by using

LIF. All these techniques require rather expensive and

somewhat complex equipment. In contrast, it is very

convenient to use operational modes to improve the

concentration sensitivity. Sample stacking is an inherent

and exclusive feature of CE and through this approach it is

possible to reach higher sensitivity. Recently, on-line

preconcentration methods have gained considerable interest

due to the significant sensitivity improvement that they can

provide [31, 41, 42].

The present study was aimed at developing a fast, easy,

and inexpensive CE method for the monitoring of enzy-

matic reactions of the cell membrane-anchored ecto-

nucleotide pyrophosphatases/phosphodiesterases NPP1 and

NPP3, which showed a rather low enzymatic activity.

Therefore, very high sensitivity for the quantitative deter-

mination of the reaction product(s) was required. The

natural substrate ATP was applied, which is cleaved to the

nucleotide AMP and pyrophosphate by the NPPs. The

enzymatic product AMP was selected for quantification by

CE-UV.

Electrokinetic injection was used, in order to introduce

analytes into the capillary selectively without having to

inject a large volume of sample, which would cause band

broadening. Moreover, sensitivity was greatly increased

by using a capillary dynamically coated with the polycationic

polymer polybrene. A special concentration sweeping

effect was the major physical phenomenon that worked well

Figure 1. Schematic representa-tion of field-amplified samplestacking showing sample ions(in white) and polybrene ions (ingray). Sample ions are stacked bypolybrene ions as they migratetoward the anode. (a) Startingsituation: conditioning of thecapillary with background electro-lyte from both electrode vials. (b)Electrokinetic injection of samplesolution from the cathodic end ofthe capillary. (c) Upon applicationof a constant current, positivelycharged polycationic moleculesof polybrene move toward thecathode and negatively chargedanalyte ions move toward theanode. As a result of the sweep-ing effect of polybrene ions theanalyte ions are stacked to give anarrow band.



for the negatively charged analytes. An extremely narrow

analyte zone was created under the constant electric field.

The capillary wall was coated with the polycationic polymer

polybrene that provided a reverse EOF environment in the

presence of a slightly acidic buffer (pH 6.5). In a previous

study [43], a more than 5000-fold increased sensitivity had

been achieved by using micellar electrokinetic chromato-

graphy and very low pH buffers, which had provided an

almost zero EOF environment. Figure 1 schematically

shows the effect: when a sample is injected into the capillary

and voltage is applied, the positively charged molecules of

polybrene sweep the negatively charged analyte molecules

(e.g. AMP) and concentrate them to make a narrow analyte

zone.

3.1 Effect of polybrene concentration on sensitivity

In order to obtain a short migration time of the nucleotides

along with high sensitivity, different concentrations of

polybrene were added to the running buffer. Initially, the

capillary was flushed with the buffer containing 0.2%

polybrene, which was adsorbed onto the capillary walls to

produce a positively charged inner surface. Then, different

concentrations of polybrene were added to the running

buffer to maintain the positive charge at the inner surface of

the capillary. Dynamic modification of capillaries is easier as

compared with the chemical coating of capillaries [44].

Moreover, washing the dynamically coated capillary with a

strong base (0.1 M aqueous NaOH) and with high salt

concentrations (1 M NaCl) can regenerate the negatively

charged silanolate surface of the capillary. In addition, the

positive coating of the capillary wall reduces protein

adsorption at the surface of the capillary, shortens the

migration time, and enhances the sensitivity.

To investigate the effect of the polybrene concentration

on the sensitivity of ATP, AMP, and UMP determination,

different concentrations of polybrene were added to the

50 mM sodium phosphate buffer (pH 6.5): 0.001, 0.002,

0.004, 0.006, 0.01, and 0.02% m/v polybrene. It was observed

that 0.002% polybrene gave a high sensitivity for all

three measured nucleotides. Raising the polybrene concen-

tration had no effect on the sensitivity (data not shown),

indicating that adsorption of the polyelectrolyte on the

capillary wall had already been saturated, while 0.001% was

not sufficient for complete coating. The migration times of

the nucleotides were not altered by changing the polybrene

concentrations.

3.2 Effect of separation buffer concentration and pH

on sensitivity

As a next step, the effect of background electrolyte

concentration and pH on sensitivity was investigated.

Different concentrations of phosphate buffer were used,

25, 50, 75, and 100 mM at pH 6.5 in the presence

of 0.002% polybrene. The peak areas of the nucleotides

were increased by increasing the concentration of the

separation buffer up to 75 mM (Fig. 2); however, there

was a current drop at 75 and 100 mM phosphate buffer

concentration. Also, at higher buffer concentrations,

longer migration times were observed. A 50 mM buffer

concentration turned out to be a good compromise and

was therefore selected for all subsequent measurements.

Keeping the phosphate buffer concentration constant (at

50 mM), different pH values of phosphate buffer, 5.0, 6.5,

7.0, and 8.0, were investigated. A high sensitivity and short

migration times for the nucleotides were obtained at pH 6.5

(data not shown).

3.3 Effect of capillary length on sensitivity

In subsequent experiments, the sample stacking effect

was investigated using capillaries of different lengths, 30,

40, 50, and 60 cm. The length of the capillary was found to

be inversely proportional to the peak areas of the analytes,

and the longer the capillary was, the lower was the

measured UV response (Fig. 3). This could be due to a

more pronounced stacking effect in the shorter capillaries.

In contrast, in a recently reported on-line drug metabolism

study using CE in a normal polarity mode, an increase in

the capillary length had led to an increase in the

detector response [45]. As expected, a longer capillary also

led to an increase in the migration times. Therefore, a short

capillary length of 40 cm was selected for subsequent

experiments.

Figure 2. Effect of buffer concentration on peak areas. Separa-tion conditions: different concentrations of phosphate asindicated on the X-axis, 0.002% polybrene, pH 6.5, fused-silicacapillary, 40 cm length (30 cm to the detector), 75 mM ID, �60 mA,251C, detection at 210 nm.



3.4 Effect of duration of electrokinetic injection on

sensitivity

In order to improve the sensitivity and the peak shape,

different injection times were investigated (15, 30, and 60 s)

at a constant applied voltage of 6 kV with reverse polarity

mode. The peak areas of the nucleotides were directly

proportional to the injection durations (Fig. 4). At 60 s

injection duration, the highest sensitivity was reached;

therefore, 60 s sample injection time was selected.

In order to obtain reproducible detector responses,

triplicate injections from a single vial containing 200 mL of

the solution were made. Thus, the optimized CE conditions

were 50 mM potassium hydrogen phosphate with 0.002%

polybrene, pH 6.5, fused-silica capillary 75 mM ID, 40 cm

capillary length, �60 mA, and an injection time of 60 s at

6 kV with reverse polarity.

3.5 Quantitative determination of AMP and method

validation

The optimized separation conditions were subsequently

applied to the detection of AMP, the reaction product of

NPPs. AMP was dissolved in enzyme assay buffer (1 mM

MgCl2, 2 mM CaCl2, 10 mM HEPES, pH 7.4) to obtain a

1 mM stock solution. Standard calibration curves were

obtained with final concentrations of 0.0001–10 mM. For

validating the method, 0.5 mM of UMP was used as an

internal standard. The calibration curves were obtained by

plotting the corrected peak area of AMP against its

concentration. A determination coefficient (R2) of 0.9999

was calculated including AMP concentrations from 0.2 to

10 mM. The determined LOD and LOQ were 46 and 178 nM,

respectively.

For comparison, AMP was additionally quantified by a

standard CE method using the same fused-silica capillary,

20 mM sodium phosphate buffer at pH 7.5 with 100 mM

SDS, an applied voltage of 10 kV, normal polarity, and

hydrodynamic injection (5 s, 0.5 psi). A standard calibration

curve was determined essentially as described above using

final AMP concentrations of 5.0, 10, 15, 20, 30, and 40 mM.

UMP was used as an internal standard at a concentration of

20 mM. A determination coefficient (R2) of 0.998 was calcu-

lated. The LOD and the LOQ were 3.3 and 5.0 mM, respec-

tively. These results show that the sensitivity could be

dramatically enhanced by the new method. The LOD for

AMP was 72-fold lower with the new method than that

obtained with a standard CE procedure.

3.6 Biochemical assays

3.6.1 Michaelis–Menten analysis of NPP1 and NPP3

The newly developed highly sensitive method was subse-

quently used to characterize the catalytic properties of

defined members of the E-NPP family. Using the optimized

conditions, Michaelis–Menten constants (Km) and maximal

velocity (Vmax) for nucleotide pyrophosphatases/phospho-

diesterases NPP1 and NPP3 were determined. The Km and

Vmax values were calculated by fitting the initial reaction

rates for the formation of the product AMP as a function of

substrate (ATP) concentration into the Michaelis–Menten

equation. The Michaelis–Menten plots are depicted in

Fig. 5. Estimated Km values of 6.271.5 and 7.473.1 mM

were obtained for E-NPP1 and NPP3, respectively. Vmax

values were 275748 and 147721 nmol/min/mg of protein

(membrane preparation) for NPP1 and NPP3, respectively.

Figure 3. Effect of capillary length on peak areas of ATP (~),AMP (m), and UMP (& ). Separation conditions were: 50 mMsodium phosphate at pH 6.5, 0.002% polybrene, fused-silicacapillary, different capillary lengths, 75 mM ID, �60 mA, 251C,injection voltage �6 kV, detection at 210 nm.

Figure 4. Effect of electrokinetic injection time on peak areas ofATP (~), AMP (m), and UMP (& ). The injection voltage was�6 kV. Separation conditions: 50 mM sodium phosphate bufferat pH 6.5, 0.002% polybrene, fused-silica capillary, 40 cm length(30 cm to the detector), 75 mM ID, �60 mA, 251C, detection at210 nm.



A direct comparison with Km and Vmax values of the same

enzymes for the same substrate (ATP) determined with

another analytical method is not possible because such data

have not been published. The determined Km value for

human NPP1 of 6.2 mM for ATP as a substrate in the present

study was similar to the value obtained for the human NPP1

using an HPLC method and diadenosine triphosphate as a

substrate (5.1 mM) [3]. The literature Km value of rat NPP3 for

diadenosine triphosphate was somewhat higher (50 mM) [3]

than the value determined in the present study for human

NPP3 with ATP as a substrate (7.4 mM). The Vmax values

determined in the present study were roughly in the same

range as Vmax values reported for other NPP preparations

determined with different assays [3, 16, 17].

3.6.2 Enzyme inhibition assay

The newly developed CE method was also used for the

determination of IC50 values and inhibition constants

(Ki values) for inhibitors of recombinant human NPPs.

IC50 values of the standard inhibitors reactive blue 2 (RB2)

[15] and suramin [3, 15, 17] were determined by using a

fixed amount of ATP (200 mM) and a range of concentra-

tions of inhibitors (see Table 1). The obtained concentration

–inhibition curves for RB2 are presented in Fig. 6. A typical

electropherogram for the control assay in the absence of an

inhibitor is shown in Fig. 7A. The percentage conversion of

substrate to product under the applied conditions was less

than 1%. The second electropherogram (Fig. 7B) shows a

typical NPP1 inhibition experiment, in which the inhibitor

RB2 (0.003 mM) was present. In that electropherogram the

peak size for AMP, the product of the enzymatic reaction,

was significantly smaller as compared with the control

assay. The peak for AMP, which was used for the

quantitative determination of the enzymatic reaction,

migrated within less than 5 min. RB2 inhibited both human

enzymes, NPP1 and NPP3, to a similar extent. In contrast,

suramin showed a seven-fold stronger inhibitory effect on

NPP3 than on NPP1 in our hands. Suramin was a stronger

inhibitor of both NPP1 and NPP3 than RB2. Presuming a

competitive mechanism of enzyme inhibition, Ki values

were calculated from the determined IC50 for the inhibitors

and Km values for the NPPs using the Cheng–Prusoff

equation [46] (see Table 1).

4 Conclusions

In conclusion, we have developed a new, simple, fast,

and highly sensitive method for the characterization of

ecto-NPPs and the investigation of potential inhibitors using

CE coupled with UV detection. In order to define the

properties of individual isoforms, we determined the Km

Table 1. Determination of Ki values for standard inhibitors of

human NPP1 and NPP3 with the new CE methoda)

InhibitorNPP1 NPP3

IC507SEM

(mM)

Ki7SEM

(mM)

IC507SEM

(mM)

Ki7SEM

(mM)

RB2 17.071.6 0.5270.03 20.071.3 0.7170.03

Suramin 8.671.3 0.2670.02 1.270.8 0.0470.01

a) Values represent means 7SEM of three separate experi-ments. For CE conditions see Fig. 7. Ki values were calculatedfrom IC50 values (determined with 200 mM ATP) using a Kd

value of 6.2 mM for NPP1 and 7.4 mM for NPP3 (see Fig. 6)assuming a competitive mechanism of inhibition.

Figure 6. Concentration-dependent inhibition of NPP1 (�) andNPP3 (& ) by reactive blue 2 (RB2) determined by CE. Asubstrate concentration of 200 mM ATP, a reaction bufferconsisting of 1 mM MgCl2, 2 mM CaCl2, and 10 mM HEPES, pH7.4, and various concentrations of RB2 were used. Data pointsrepresent means7SD from three separate experiments, eachrun in duplicate.

Figure 5. Michaelis–Menten representation of the hydrolysis ofATP by human NPP1 (�) and NPP3 (& ). Data points representmeans7SD from three separate experiments each run induplicate. For CE conditions see Fig. 7 and Section 2.6.Determined kinetic parameters were: NPP1, Km 5 6.271.5 nmol/min/mg protein, Vmax 5 275748; NPP3, Km 5 7.473.1,Vmax 5 147721 nmol/min/mg protein.



and Vmax values for NPP1 and NPP3 using membrane

preparations of COS7 cells heterologously expressing

the human enzymes. IC50 and Ki values for standard

inhibitors were also determined. By using a dynamic coating

on the fused-silica capillary wall, migration times of

nucleotides were decreased in comparison with a bare

fused-silica capillary, and high sensitivity of analysis was

achieved. The new CE method has multiple advantages in

comparison with standard NPP assays, e.g. no need

for expensive radiolabeled or fluorescent substrates, the

possibility to employ (various) natural substrates rather

than artificial ones, no requirement for sample preparation

prior to analysis, and a minimal use of reagents. The

quantitative analysis of the samples can be carried out

within a few minutes. It will, therefore, allow the screening

of compound libraries in order to identify and

develop selective inhibitors for this pharmacologically

important class of enzymes. In comparison with other CE

methods described for the determination of nucleotides, it

has clear advantages, such as an increased sensitivity

allowing the monitoring of nucleotidases with low enzy-

matic activity, and no requirement of long rinsing

procedures that are typical for the use of fused-silica

capillaries in biological assays.

Financial support by the Deutscher AkademischerAustauschdienst (DAAD, STIBET scholarship) to J. I. isgratefully acknowledged. J. S. was supported by grants from theCanadian Institutes of Health Research (CIHR). S. A. L. was arecipient of a scholarship from ‘‘Fonds de recherche en sante duQuebec’’ (FRSQ) and J. S. of a New Investigator award fromthe CIHR.

The authors have declared no conflict of interest.

5 References

[1] Cimpean, A., Stefan, C., Gijsbers, R., Stalmans, W.,Bollen, M., Biochem. J. 2004, 381, 71–77.

[2] Bollen, M., Gijsbers, R., Ceulemans, H., Stalmans,W.,Stefan, C., Crit. Rev. Biochem. Mol. Biol. 2000, 35,393–432.

Figure 7. Typical electrophero-grams of NPP1 enzyme inhibitionassays. (A) Control without inhi-bitor. (B) Assay in the presence ofRB2 (0.003 mM). The separationconditions were 50 mM phos-phate buffer at pH 6.5, 0.002%polybrene, fused-silica capillary,40 cm length (30 cm to the detec-tor), 75 mM id, �60 mA, 251C,detection at 210 nm.



[3] Vollmayer, P., Clair, T., Goding, J. W., Sano, K. et al.,Eur. J. Biochem. 2003, 270, 2971–2978.

[4] Stefan, C., Jansen, S., Bollen, M., Purinergic Signalling2006, 2, 361–370.

[5] Zimmermann, H., Naunyn Schmiedebergs Arch. Phar-macol. 2000, 362, 299–309.

[6] Stefan, C., Jansen, S., Bollen, M., Trends Biochem. Sci.2005, 30, 542–550.

[7] Asensio, A. C., Rodrıguez-Ferrer, C. R., Castaneyra-Perdomo, A., Oaknin, S., Rotllan, P., Neurochem. Int.2007, 50, 581–590.

[8] Bartkiewicz, M., Sierakowska, H., Shugar, D., Eur. J.Biochem. 1984, 143, 419–426.

[9] Wojcik, M., Cieslak, M., Stec, W. J., Goding, J. W.,Koziolkiewicz, M., Oligonucleotides 2007, 17, 134–145.

[10] Bottcher, Y., Korner, A., Reinehr, T., Enigk, B. et al.,J. Clin. Endocrinol. Metab. 2006, 91, 4948–4952.

[11] Goding, J. W., Grobben, B., Slegers, H., Biochim.Biophys. Acta 2003, 1638, 1–19.

[12] Yano, Y., Hayashi, Y., Sano, K., Shinmaru, H. et al., Int.J. Mol. Med. 2003, 12, 763–766.

[13] Yano, Y., Hayashi, Y., Sano, K., Nagano, H. et al., CancerLett. 2004, 207, 139–147.

[14] Umemura, K., Yamashita, N., Yu, X., Arima, K. et al.,Neurosci. Lett. 2006, 400, 97–100.

[15] Grobben, B., Claes, P., Roymans, D., Esmans, E. L.et al., Br. J. Pharmacol. 2000, 130, 139–145.

[16] Rucker, B., Almeida, M. E., Libermann, T. A., Zerbini,L. F. et al., Mol. Cell. Biochem. 2007, 306, 247–254.

[17] Furstenau, C. R., Trentin, D. S., Barreto-Chaves, M. L.,Sarkis, J. J., Platelets 2006, 17, 84–91.

[18] Nahum, V., Tulapurkar, M., Levesque, S. A., Sevigny, J.et al., J. Med. Chem. 49, 1980–1990.

[19] Joseph, S. M., Pifer, M. A., Przybylski, R. J., Dubyak,G. R., Br. J. Pharmacol. 2004, 142, 1002–1014.

[20] Viglio, S., Zanaboni, G., Luisetti, M., Cetta, G. et al.,Electrophoresis 1998, 19, 2083–2089.

[21] Iqbal, J., Jirovsky, D., Lee, S. Y., Zimmermann, H.,Muller, C. E., Anal. Biochem. 2008, 373, 129–140.

[22] Iqbal, J., Vollmayer, P., Braun, N., Zimmermann, H.,Muller, C. E., Purinergic Signalling 2005, 1, 349–358.

[23] Iqbal, J., Burbiel, J. C., Muller, C. E., Electrophoresis2006, 27, 2505–2517.

[24] Iqbal, J., Scapozza, L., Folkers, G., Muller, C. E.,J. Chromatogr. B Analyt. Technol. Biomed. Life Sci.2007, 846, 281–290.

[25] Jung, B., Bharadwaj, R., Santiago, J. G., Electrophoresis2003, 24, 3476–3483.

[26] Wey, A. B., Zhang, C. X., Thormann, W., J. Chromatogr.A 1999, 853, 95–106.

[27] Horakova, J., Petr, J., Maier, V., Znaleziona, J. et al.,J. Chromatogr. A 2007, 1155, 193–198.

[28] Giordano, B. C., Newman, C. I., Federowicz, P. M.,Collins, G. E., Burgi, D. S., Anal. Chem. 2007, 79,6287–6294.

[29] Palmer, J. F., J. Chromatogr. A 2004, 1036, 95–100.

[30] Jung, B., Bharadwaj, R., Santiago, J. G., Anal. Chem.2006, 78, 2319–2327.

[31] Shihabi, Z. K., Electrophoresis 2002, 23, 2394–2398.

[32] Shihabi, Z. K., J. Chromatogr. A 1999, 853, 3–9.

[33] Yeh, C. F., Jiang, S. J., Analyst 2002, 127, 1324–1327.

[34] Cortacero-Ramirez, S., Segura-Carretero, A., Hernainz-Bermudez, d. C., Fernandez-Gutierrez, A., J. Chroma-togr. A 2005, 1064, 115–119.

[35] Belli, S. I., Goding, J. W., Eur. J. Biochem. 1994, 226,433–443.

[36] Jin-Hua, P., Goding, J. W., Nakamura, H., Sano, K.,Genomics 1997, 45, 412–415.

[37] Kukulski, F., Levesque, S. A., Lavoie, E. G., Lecka, J.et al., Purinergic Signalling 2005, 1, 193–204.

[38] Bradford, M. M., Anal. Biochem. 1976, 72, 248–254.

[39] Shao, X., O’Neill, K., Zhao, Z., Anderson, S. et al.,J. Chromatogr. A 1994, 680, 463–468.

[40] Uhrova, M., Deyl, Z., Suchanek, M., J. Chromatogr. BBiomed. Appl. 1996, 681, 99–105.

[41] Shihabi, Z. K., Electrophoresis 2002, 23, 1628–1632.

[42] Shihabi, Z. K., J. Chromatogr. A 2005, 1066,205–210.

[43] Quirino, J. P., Terabe, S., Science 1998, 282, 465–468.

[44] Tang, Z. M., Kang, J. W., Anal. Chem. 2006, 78,2514–2520.

[45] Kim, H. S., Wainer, I. W., Anal. Chem. 2006, 78,7071–7077.

[46] Cheng, Y., Prusoff, W. H., Biochem. Pharmacol. 1973,22, 3099–3108.



Documents

A highly sensitive CE-UV method with dynamic coating of silica-fused capillaries for monitoring of nucleotide pyrophosphatase/phosphodiesterase reactions