Short Communication

Ultra-fast adenosine 50-triphosphate,adenosine 50-diphosphate and adenosine50-monophosphate detection by pressure-assisted capillary electrophoresis UVdetection

Herein, we report a new CE method to measure adenine nucleotides adenosine

50-triphosphate, adenosine 50-diphosphate, and adenosine 50-monophosphate in red

blood cells. For this purpose, 20 mmol/L sodium acetate buffer at pH 3.80 was used as

running electrolyte, and the separation was performed by the simultaneous application

of a CE voltage of 25 kV and an overimposed pressure of 0.2 psi from inlet to outlet.

A rapid separation of these analytes in less than 1.5 min was obtained with a good

reproducibility for intra- and inter-assay (CVo4 and 8%, respectively) and an excellent

analytical recovery (from 98.3 to 99%). The applicability of our method was proved by

measuring adenine nucleotides in red blood cells.

Keywords:

Adenosine 50-diphosphate / Adenosine 50-monophosphate / Adenosine50-triphosphate / CE / Pressure assisted DOI 10.1002/elps.201000138

Adenosine 50-triphosphate (ATP) is a ubiquitous, energy-

rich compound in all cells of living organisms. It is found

both within organelles, such as mitochondria and chloro-

plasts, and in the cytoplasm of higher organisms. ATP is

produced as an energy source during the processes of

photosynthesis and cellular respiration and consumed by

many enzymes and a multitude of cellular processes

including biosynthetic reactions, motility, and cell division

[1]. The energy content of cells depends on the balance

between the production and the consumption of energy, so

that the levels of ATP, adenosine 50-diphosphate (ADP), and

adenosine 50-monophosphate (AMP) are tightly regulated.

However, some metabolic stresses can cause the alteration

of intracellular energy charge which may fall below the

physiological range [2–7]. Therefore, methods able to

measure adenine nucleotides are needed to monitor the

energy state of cells. ATP, ADP, and AMP have been

commonly measured by using bioluminescence reagents

containing firefly luciferase [8, 9], though HPLC [10] or CE

[11–13] methods have also been described. In all these cases,

tedious preanalytical procedures or long analytical times are

required to perform the analysis. Therefore, we have

recently developed a new method by short-end injection

CE by which the analytes were resolved in 5 min using

methylcellulose as EOF suppressor [14]. However, when it

works on labile metabolites as ATP, the assay quickness is a

crucial factor to obtain useful data. We have already reported

that the water lysis determines a loss of ATP with increased

levels of ADP and AMP due to the ATPase activity that is,

instead, suppressed by acidic treatment [14]. Moreover, we

have found that if the samples were maintained for 24 h at

231C in a TCA solution, a degradation of ATP and an

increase of ADP and AMP levels (perhaps due to the

ATP breakdown) could be evidenced [14]. Therefore, faster

the assay for ATP detection, more accurate will be the

analysis. To further decrease analytical times, by this new

study, we analyze the possibility to apply the technique of

overimposed pressure/voltage, in which a pressure from

inlet to outlet is employed during the electric field

application. The use of a hydrodynamic overimposed

pressure during the whole electrophoretic run is commonly

used in CE-MS systems to enhance the strength of the

method (in particular to provide stable electrospray

conditions for successful coupling) and also to decrease

run times [15, 16]. This is due to the fact that during

the electrophoresis, the buffer in the capillary moves toward

the detection window increasing the mobility of the

analytes. The method applicability was proved by measuring

Angelo Zinellu1,2,3

Salvatore Sotgia1

Bastianina Scanu1,2

Elisabetta Pisanu1,2

Manuela Sanna1

Maria Franca Usai1

Luca Deiana1

Ciriaco Carru1,2�

1Department of BiomedicalSciences and Centre ofExcellence for BiotechnologyDevelopment and BiodiversityResearch, University of Sassari,Sassari, Italy

2National Laboratory of theNational Institute ofBiostructures and Biosystems,Osilo, Italy

3Porto Conte Ricerche Srl,Sassari, Italy

Received March 9, 2010Revised May 19, 2010Accepted May 19, 2010

Abbreviations: ADP, adenosine 50-diphosphate; AMP,

adenosine 50-monophosphate; ATP, adenosine 50-tri-phosphate

�Additional corresponding author: Professor Ciriaco Carru

E-mail: carru@uniss.it

Correspondence: Dr. Angelo Zinellu, Department of BiomedicalSciences and Centre of Excellence for Biotechnology Develop-ment and Biodiversity Research, University of Sassari, Sassari,ItalyE-mail: angelozinellu@libero.itFax: 139-079228275

& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

Electrophoresis 2010, 31, 2854–28572854

analytes’ concentration in red blood cells. To evaluate the

accuracy of CE analysis, the data obtained by the new

method and by the short-end injection CE assay were

compared by statistical Bland–Altman test.

Red blood cells were separated from plasma by centri-

fugation (5000� g for 5 min, at 41C) followed by three

washings with 0.9% w/v NaCl and were immediately

processed. Packed cells (100 mL) were lysed by adding 200 mL

of 10% w/v TFA cold. After centrifugation (3000� g for

5 min, at 41C), 50 mL of supernatant was mixed with 200 mL

of distilled water and subsequently injected in CE. An MDQ

CE system equipped with a DAD was used (Beckman

Instruments, Fullerton, CA, USA). The system was fitted

with a 30 kV power supply with a current limit of 300 mA.

Analysis was performed in an uncoated fused silica capillary

(75 mm id and 30 cm total length), injecting 78 nL of sample

(7 s at 0.5 psi) at the inlet end. Separation was carried out in

a 20 mmol/L sodium acetate buffer pH 3.8, 301C, and 25 kV

(120 mA) at reverse polarity (with the anode at the outlet)

with an overimposed pressure of 0.2 psi from inlet to outlet.

Batch sequence samples were stored in the MDQ auto-

sampler at 101C. After each run, the capillary was rinsed

with 0.5 min of 0.1 mmol/L HCl and equilibrated with run

buffer for 0.5 min. Separation was monitored at 254 nm

wavelength.

Due to their negative charge in a wide pH range, ATP,

ADP, and AMP may be easily resolved in CE. However,

owing to the considerable difference in charge density

among these analytes, their migration times could signifi-

cantly differ, yielding long analytical times. As previously

reported [14], by using a 60 mmol/L sodium acetate run

buffer (pH 3.8) at reverse polarity, the most charged analytes

migrated toward the anode faster than the less charged ones.

The migration order was ATP, ADP, and AMP with this last

analyte taking about 30 min to reach the detection window

sited at 20 cm from the inlet. We have reduced the migra-

tion time of AMP by using methylcellulose, to minimize the

EOF, which in the adopted electrophoretic conditions were

opposite to the migration direction of analytes and though at

pH 3.8 its intensity was low, it delayed the AMP migration.

To further cut the run time, we have reduced the migration

distance as much as possible injecting the analytes at the

end of the capillary, near to the detection window, under the

same electrophoretic conditions (the polarity of the electro-

des was obviously switched over). By using all these tricks,

we were able to reduce the migration times to 5 min

(Fig. 1A) [14]. By this new study, we explore the possibility to

use hydrodynamic overimposed pressure to obtain a more

rapid separation of adenine nucleotides. As shown in

Fig. 1B, by using a 20 mmol/L of sodium acetate buffer, a

normal injection mode and an overimposed pressure of

0.2 psi during the whole electrophoretic run, ATP, ADP, and

AMP were resolved in less than 1.5 min, even without the

0.06AADP

AMP

ATP

AU

0.00

B

ADP

AU ATP

ADPAMP

0 1 2 3 4 5

Migration times (min)

0.06

0.00

Figure 1. Electropherograms of standard adenine nucleotides.Electrophoretic conditions: (A) 60 mmol/L sodium acetate, pH3.8, 0.01% methylcellulose, 10 cm effective capillary length(30 cm total length), outlet injection, 151C cartridge temperature,and 17 kV applied voltage (normal polarity). (B) 20 mmol/Lsodium acetate buffer, pH 3.8, 20 cm effective capillary length(30 cm total length), normal injection, 301C cartridge tempera-ture, and 25 kV applied voltage (reverse polarity), overimposedpressure of 0.2 psi from inlet to outlet. For both: sample injection78 nL (7 s at 0.5 psi), 254 nm detection wavelength.

A

B

AU

ATP

ADP

ATP

AU

ADP

210

Migration times (min)

0.005

0.000

0.005

0.000

Figure 2. Electropherograms of red blood cells sample treatedwith 10% TCA (A) and 10% TFA (B).

Electrophoresis 2010, 31, 2854–2857 CE and CEC 2855

& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

use of methylcellulose as additive. The increase of pressure

intensity or the modification of buffer pH and concentration

caused an overlapping of peaks or an increasing in run

times.

As previously reported [14], we found that the acidic lysis

of red blood cells was advisable to avoid the hydrolysis of ATP,

and in particular we have found that TCA precipitation was

indicated for the sample preparation. However, when the

samples prepared in TCA were used in the new CE condi-

tions, ATP peak resulted overlapped with a contaminant, as

shown in Fig. 2A. Therefore, to find a sample preparation

compatible with the new electrophoretical conditions, we tried

other acidic precipitants such as perchloric acid, metaphos-

phoric acid, sulfosalicylc acid, and TFA. As shown in Fig. 2B,

a good separation could be obtained by using 10% TFA. In

view of our previous results on adenosine 50-triphosphate

unstability in water and TCA-treated red blood cell [14], we

investigate on ATP stability during sample batch processing of

TFA-treated sample. The ATP concentration of the same

sample, stored in the MDQ autosampler at 101C, has been

monitored for 5 h. We found that the sample acidification

allows to stabilize ATP if compared with water lysis, even if we

have also measured a slow disappearance of ATP in acidified

samples of about 0.45% after 1 h, 1.09% after 2 h, 2.22% after

3 h, and 4.13% after 5 h.

The linearity of the detector response to different

analyte concentrations was determined between 40 and

2000 mmol/L for ATP and between 50 and 500 mmol/L for

ADP and AMP. The calibration curves, built plotting

absorbance (Y) versus concentrations (X), Y 5 6.8X1148

(ATP), Y 5 8.1X186 (ADP), and Y 5 10.4X199 (AMP)

showed linear responses over the concentration tested with

regression coefficients R2 5 0.999 for all the analytes.

Within-run precision (intra-assay) of the method was eval-

uated by injecting the same plasma pool ten times conse-

cutively, whereas between-run precision (inter-assay) was

determined by injecting the same sample in ten consecutive

days. Precision tests, summarized in Table 1, indicated a

good repeatability of the method both for intra- and inter-

assay. For the assessment of the analytical recovery, biolo-

gical samples were spiked with standard solutions (2, 1, and

0.5 mmol/L for ATP and 0.2, 0.1, and 0.05 mmol/L for ADP

and AMP) and the means of recovery, evaluated by five

different experiments are summarized in Table 1. To

determine the lowest LOD, serial dilutions (in water) of a

mix standard were injected and the concentrations giving

the smallest observable peak were identified: the LOD for an

S/N of 3 was 12 mmol/L for all nucleotides. The accuracy of

the new CE method was assessed by comparison to the

short-end injection CE assay. Twenty red blood cell samples

were analyzed and the concentrations of ATP and ADP were

determined. The results were evaluated by Bland–Altman

Table 1. Reproducibility and analytical recovery of the new method

Injection reproducibility Method reproducibility Analytical recovery

Migration times (min) Peak area AU Intra-assay (mmol/L) Inter-assay (mmol/L) Recovery (%)

Mean (CV) Mean (CV) Mean (CV) Mean (CV) Mean (CV)

ATP 0.822 (0.39) 6650 (2.7) 1515 (3.6) 1533 (7.8) 98.5

ADP 0.976 (0.41) 813 (2.9) 89 (3.8) 91 (7.8) 99.2

AMPa) 1.468 (0.52) 1100 (3.0) 101 (4.1) 99 (8.0) 98.8

a) AMP standard has been spiked to the sample at a final concentration of 100 mmol/L.

100

+1.96 SD60 8

ATP

.

50

0

PA

CE

- S

EIC

E

Mean

4,1

60,8

-50

-100

PA

CE

- S

EIC

E-1.96 SD

-52,5

1200 1400 1600 1800 2000

AVERAGE of PACE and SEICE

12ADP

8

4

0Mean

0 1

+1.96 SD7,4

-4

-8

0,1

-1.96 SD-7,2

80 120 160 200 240-12

AVERAGE of PACE and SEICE

Figure 3. Bland–Altman plot showing the differences betweenthe new and the reference method. The central line shows themean differences between methods, upper and lower linescorrespond to 71.96 times the SD. PACE, pressure-assisted CEand SEICE, short-end injection CE. Concentrations are expressedas mmol/L.

Electrophoresis 2010, 31, 2854–28572856 A. Zinellu et al.

& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

test, plotting the differences between the analytes’ concen-

trations measured by the two assays against the average of

the two values. The results demonstrated the absence of a

systematic bias both for ATP and for ADP determination

(Fig. 3).

Few applications of pressure-assisted CE are reported in

the literature, due to the difficulty to obtain well-resolved

peaks. In these conditions, in fact, two forces act to move the

analytes toward the detection window, the electrophoretic

mobility due to electrical field, and the run buffer flux

generated by the overimposed pressure. When it is possible

to keep the peaks well resolved, the method allows to obtain

very short migration times. By this approach, we obtain a

rapid separation of adenine nucleotides in less than 1.5 min

without the addition of additives on run buffer. Thus, the

new method allows us to measure 20 samples in 60 versus140 min needed with the previous assay. The quickness of

the method permits to reduce the pitfalls due to the loss of

ATP concentration during the sample batch sequence. Even

if a loss in the sensitivity was observed in comparison to

the short-end injection capillary method (LOD 12 versus6 mmol/L), the applicability of the new assay has been

anyway proved in red blood cells after lysis with TFA,

suggesting that the method should be used as a high-

throughput sample processing tool.

This study was supported by the ‘‘Fondazione Banco diSardegna – Sassari – Italy ’’ and by the ‘‘Ministero dell’Uni-versita e della Ricerca,’’ Italy. The manuscript’s language revi-sion by Maria Antonietta Meloni is greatly appreciated. AngeloZinellu was supported by Regione Autonoma della Sardegna‘‘Ricerca cofinanziata PROGRAMMA OPERATIVO FSESARDEGNA 2007-2013 – L.R.7/2007 – Promozione dellaricerca scientifica e dell’innovazione tecnologica in Sardegna.’’

The authors have declared no conflict of interest.

References

[1] Atkinson, D. E., Adu. Enzyme Regul. 1971, 9, 207–221.

[2] Lowry, O. H., Carter, J., Ward, J. B., Glaser, L., J. Biol.Chem. 1971, 246, 6511–6521.

[3] Miovic, M. L., Gibson, J., J. Bacterial. 1973, 114, 86–95.

[4] Slayman, C. I., J. Bacterial. 1973, 114, 752–766.

[5] Dielzer, D. N., Lais, C. J., Leckie, M. P., Arch. Biochem.Biophys. 1974, 160, 14–25.

[6] Dielzer, D. N., Lais, C. J., Magnani, J. L., Leckie, M. P.,Biochem. Biophys. Res. Commun. 1974, 60, 875–881.

[7] Swedes, J. S., Sedo, R. J., Atkinson, D. E., J. Biol. Chem.1975, 250, 6930–6938.

[8] Sakakibara, T., Murakami, S., Imai, K., Anal. Biochem.2003, 312, 48–56.

[9] Ishii, S., Sato, Y., Terashima, M., Saito, T., Suzuki, S.,Murakami, S., Gotoh, M., Transpl. Proc. 2004, 36,1191–1193.

[10] Pogolotti, A. L., Jr., Santi, D. V., Anal. Biochem. 1982,126, 335–345.

[11] Nguyen, A. L., Luong, J. H. T., Masson, C., Anal. Chem.1990, 62, 2490–2493.

[12] Ng, M., Blaschke, T. F., Arias, A. A., Zare, R. N., Anal.Chem. 1992, 64, 1682–1684.

[13] Liu, J., Banks, F., Novotny, M., J. Microcol. Sep. 1989, 1,136–141.

[14] Zinellu, A., Sotgia, S., Pasciu, V., Madeddu, M., Leoni,G. G., Naitana, S., Deiana, L., Carru, C., Electrophoresis2008, 29, 3069–3073.

[15] Sheeley, D. M., Reinhold, V. N., Anal. Chem. 1998, 70,3053–3059.

[16] Ruiz-Calero, V., Moyano, E., Puignou, L., Galceran,M. T., J. Chromatogr. A 2001, 914, 277–291.

Electrophoresis 2010, 31, 2854–2857 CE and CEC 2857

& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com