A Semiautomated Procedure for Determination of Adenosine Triphosphate L. DUFRESNE Department
H. J. GITELMAN
of Medicine, University of North Carolina School of Medicine, Chapel Hill, North Carolina 37’614 Received February
Firefly luminescence has provided the basis for a specific and sensitive assay system for the estimation of adenosine triphosphate (ATP) in biological materials (1, 2). However, manual techniques, employing the luciferin-luciferase reaction, are tedious and demanding. We have developed a semiautomated modification of the firefly method to circumvent these disadvantages. Interference from inorganic salts has been eliminated by extreme dilution. Additionally, a simplified method of sample preparation for erythrocyte ATP assays has been devised. METHOD I. Apparatus
Standard Technicon AutoAnalyzer equipment is used with a Packard flow detector, series 317. A Pyrex glass coil serves as a flow cell (Fig. 1) , and the flow detector is connected to a Packard Tri-Carb liquid scintillation spectrometer, series 314F. The manifold contains a diluting segment in which the sample is diluted approximately 1: 70 with water and buffer. A portion of the diluted sample is pumped into an airsegmented buffer stream and firefly ‘enzyme is added immediately before the diluted sample mixture enters the Pyrex glass coil, achieving a final sample dilution of approximately 1: 1000 in the flow detector. II. Reagents A. Firefly enzyme. This solution is prepared from a commercial firefly preparation (Worthington) by dissolving the contents of one vial of firefly extract in 75 cc of distilled water at least 12 hr before uze. The refrigerated solution is filtered through a Millipore filter (0.45 p) prior to use in an ice bath. 403
UPPER LOWER I LEVEL LEVEL
WASTE ? I
PE 100 FIO.
B. Buffer. The buffer (pH 7.5) contains 50 m&I glycylglycine and 5 mM magnesium. Prepare by dissolving 6.6 gm glycylglycine, 1.23 gm MgSO,*7H,O, and 0.3 ml 10N sodium hydroxide in 1 liter distilled water. Add 0.2 ml NP-27 (Union Carbide) and mix. C. 1 x 1O-3 M ATP stock. Prepare from disodium ATP (Sigma) by dissolving approximately 650 mg in 1 liter distilled water and freeze in 10 ml aliquots. The ATP concentration of the stock solution is standardized by the hexokinase method (3). Working standards are prepared daily from the stock solution by appropriate dilution with distilled water and stored in an ice bath. D. Spectrometer Calibrations. Linear calibration of the spectrometer was achieved for ATP concentrations ranging between 1CV and 1O-4M. The optimal settings for a variety of ATP concentrations and the standard calibration lines are shown in Figure 2. No loss of precision was noted when analyzer rather than coincident settings were employed. E. Sample Preparation. (1) Hemolyzing solution containing 5 mM EDTA (pH 11) : Dissolve I.86 gm disodium EDTA in 500 cc distilled water. Add 2 ml 10N sodium hydroxide and dilute to 1 liter. Add 1 cc NP-27 (Union Carbide) and mix. (2) Sample preparation: Using disposable polypropylene tubes, pipet 1 ml of a washed erythrocyte suspension of known hematocrit, between 20 and 30%, into 4 ml of the hemolyzing solution and mix thoroughly. A clear homogenous solution is obtained with a pH of 8.5-9. The diluted sample may be analyzed
FIO. lo4 M samples settings
2. Spectrometer settings and Al?? calibration line is not containing less than 1 X lo4 adjusted for lo-’ or lo-'M
calibration lines for lo-’ to lo-‘M ATP. The linear between 0 and 1 X lo-” M ATP, and M ATP should be analyzed with spectrometer Al??.
immediately for ATP or frozen for later analysis. We have observed no fall in the ATP concentrations of EDTA hemolyzates after 10 days of frozen storage. III.
A. Factors Affecting the Procedure. (1) The assay conditions in the automated system approximate those employed by previous investigators with manual techniques (1, 2) and are provided by the pH 7.5 buffer which contains the optimal concentration of magnesium for this studies demonstrated system. (9) Hemolyzing solution: Preliminary stable ATP concentrations in EDTA hemolyzates incubated for 1 hr at 37°C. The concentration of EDTA employed does not interfere with
the luciferin-luciferase assay system. A detergent, NP-27, was incorporated in the hemolyzing solution after demonstrating complete lysis of erythrocytes, in isotonic media, with this compound. NP-27 does not interfere with the firefly enzyme assay. The adequacy of ATP extraction by the EDTA method was examined by preparing erythrocyte extracts from identical red cell samples using a perchloric acid precipitation technique (4). The results obtained by both methods are shown in Table 1. No significant differences between methods were observed. (3) Firefly preparation. Maximum stability was achieved by preparing firefly solutions lo-12 hr prior to use. Millipore filtration of the firefly solution was found essential to assure adequate reproducibility. With the concentration of crude firefly employed, linearity is achieved for concentrations of ATP varying between low7 and lo-‘M ATP. Higher concentrations of firefly enzyme can be TABLE ATP
Methods of extraction EDTA
31.2 17.6 29.1 14.8 33.2 19.1 26.6 12.1 25.9 11.6
76.5 62.5 81.1 66.9 62.3 48.4 81.2 67.1
30.6 16.8 28.9 14.6 33.1 19.0 26.6 11.8 25.8 11.4
76.5 62.6 81.8 68.1 62.1 48.4 81.2 67.1
employed to achieve greater sensitivity for smaller concentrations of ATP. B. Interferences. (1) Ionic effects. A variety of ions may interfere with the luciferin-luciferase reaction. This potential interference was examined by preparing ATP at 1O-4 and 1O-5 M concentrations in the following media: 0.161M sodium chloride, 0.16 M potassium chloride, 0.17 M tetramethyl ammonium chloride, 0.12 M disodium phosphate, 10 mM calcium chloride, and 10 n& magnesium chloride. No ionic interferences were noted. (2) Nucleotide interference. Commercial adenosine diphosphate was dissolved in 0.04 N HCl and incubated with Dowex AG-1 resin (5) to remove contaminating ATP. A 1 m&f solution of the purified ADP yielded counts equivalent to a 5 X KF6 M ATP solution, indicating a maximum interference ‘of 0.5%. One millimolar solution of AMP and cyclic AMP prepared from commerically available materials contained no measurable ATP. The presence of ADP, AMP and cyclic AMP did not interfere with the recovery of added ATP. IV. Application of Method A. Sampling and recording are performed manually using a preset time, 0.3 min, for counting. It is assumed that the sample has reached the flow cell after two similar counts have been obtained. The sample line is then inserted into the second sample. During the time it takes the second sample to reach the flow cell, at least one additional count from the initial sample will be obtained. In this fashion, each sample is read in triplicate. The introduction of an air segment into the sample line may be helpful in distinguishing samples of similar ATP content. Standardization is performed in an identical fashion. When the procedure is operating satisfactorily, variance in standards will be less than 2%. Maximum stability is achieved by sampling a high standard for a period of several minutes at the start of a given run. B. Calculation of results. ATP concentrations are calculated from the total counts by either graphic or mathematical interpolation and corrected for the dilution factor and hematocrit of the red cell suspension. Our results are expressed in mmoles ATP per liter of red blood cells. C. Reproducibility. Results of triplicate analysis of 24 different samples are shown in Table 2. In this study, the standard deviation of an individual value from the mean ranged from -+0.005 to 20.034. Poor reproducibility can be related to worn pump tubes or to a malfunction of the spectrometer. D. Recovery. Eleven recovery experiments were performed by adding 14 x KF6 M ATP to erythrocyte hemolyzates prepared from normal and ATP-depleted red blood cells, obtained by incubating normal erythrocytes in deoxyglucose for 3 hr prior to the recovery experiments.
of 22 Samples
ATP concentration Sample 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
8.81 8.78 8.13 8.12 7.71 7.61 7.30 7.29 6.72 6.67 3.39 3.37 3.14 3.10 2.67 2.64 1.95 1.93 1.78 1.75 1.22 1.21
8.80 8.80 8.14 8.12 7.65 7.61 7.33 7.33 6.71 6.70 3.37 3.34 3.10 3.10 2.65 2.67 1.93 1.93 1.77 1.73 1.21 1.21
8.81 8.83 8.10 8.11 7.65 7.64 7.29 7.33 6.71 6.70 3.37 3.37 3.13 3.10 2.68 2.66 1.94 1.94 1.76 1.74 1.21 1.20
8.82 8.80 8.12 8.11 7.67 7.62 7.30 7.31 6.71 6.69 3.37 3.36 3.13 3.10 2.66 2.65 1.94 1.93 1.77 1.74 1.21 1.20
f0.028 kO.025 f0.020 f0.005 kO.034 xko.017 fO.020 kO.023 kO.005 zko.017 +0.011 kO.017 +0.010 kO.004 kO.015 kO.015 +0.010 kO.005 fO.O1O ~0.010 kO.005 z!zo.o05
These results are listed in Table 3. The percentage ATP recovered with normal erythrocytes ranged from 98.2 to lOO.O”/o with an average of 98.9%, and with depleted erythrocytes from 97.1 to 100.7% with an average of 99.1%. DISCUSSION
This method is rapid, accurate, and precise, permitting 30 samples per hour to be estimated with relative convenience. Ionic interferences have been eliminated by extreme dilution which, in turn, has permitted a simplified sample preparation and standardization technique. Although designed primarily for use in the assay of ATP in erythrocytes, the inherent sensitivity and specificity of the assay system should permit the immediate application of this technique to a variety of biological samples. SUMMARY
We have presented a rapid, accurate, and convenient semiautomated procedure to measure ATP in erythrocytes. The technique employs con-
TABLE 3 ATP Recovery ATP cone. of RBC in hemolyzates
ATP added x 10-d M
ATP recovered x lo-“M
1.3 to 1.5 mmoles/l RBC (nondepleted cells)
14.2 14.1 14.5 14.3 14.9 14.1
14.2 13.9 14.2 14.4 14.9 13.9
100.0% 98.5% 98.1% 98.2% 100.0% 99.0% Av. 98.9%
0.2 to 0.3 mmoles/l RBC (depleted cells)
14.0 14.2 14.4 14.4 14.4
13.6 14.3 14.1 14.5 14.3
97.1% 100.7% 97.9% 100.7% 99.3% Av. 99.1%
ventional autoanalyzer equipment in conjunction with a liquid scintillation spectrometer to detect ATP by firefly luminescence. The sensitivity of the detecting system has permitted potential interferences from inorganic ions to be eliminated by dilution. Additionally, a simplified manual technique for the preparation of erythrocyte samples for ATP assay has been devised. ACKNOWLEDGMENT This work was supported by research granta from the United States Public Health Service (AMa58, AM-95054, AM-08685) and by a fellowship from the Medical Research Council of Canada. REFERENCES 1. STI~EHLER, B. L., 2. 3. 4. 5.
Bioluminescence Assay: Principles and Practice, Methods 16, 99-181 (1968). B. L., Adenosine 5’-Triphosphate and Creatine Phosphate. Determination with Luciferaxe, in “Methods of Enzymatic Analysis” (H.-V. Bergmeyer, ed.),pp. 563-568. Verlag-Chemie, Berlin, and Academic Press, New York, 1963. LAMPRECHT, W., AND TRAUTSCHOLD, I., ATP Determination with Hexokinase and Glucose 6-Phosphate Dehydrogenase, in “Methods of Enzymatic Analysis,” pp. 543-551. BUCHEFC, T., Estimation of Substrates: Pyruvate in “Methods of Enzymatic Analysis,” pp. 254-255. BARTLETT, G. R., J. Biol. Chem. 239, 459 (1959).