hydroxybutyrate and the reaction ismonitored for 60 S with ... of semi-pure 3-hydroxybutyrate ..... determination of beta hydroxybutyrate and acetoacetate. Clin.
CLIN.CHEM.23/10, 1893-1897 (1977)
A Kinetic Spectrophotometric Assay for Rapid Determination of Acetoacetate in Blood C. P. Price, B. Lloyd, and K. G. M. M. Alberti
We describe an automated kinetic assay for acetoacetate in blood. Acetoacetate is enzymatically reduced to D-flhydroxybutyrate and the reaction is monitored for 60 S with a reaction-rate analyzer. This technique allows low concentrations of acetoacetate to be measured with good precision and overcomes many of the problems associated with other automated techniques. Our studies on the stability of acetoacetate emphasize the need for care in handling specimens. The use of a reaction-rate analyzer, an item of equipment common to most laboratories, allows for rapid handling of samples in small or large batches, depending on the needs of the laboratory. AdditIonal Keyphrases: enzymatic methods . intermediary metabolism . reaction-rate method . sample stability diabetes nel!itus
Interest in the metabolic disturbances associated with diabetes mellitus and other catabolic states has led to the need for simple analytical procedures to determine intermediary metabolites, in particular the ketone bodies. Enzymatic determinations of the ketone bodies, 3-hydroxybutyrate and acetoacetate, became possible after the production of semi-pure 3-hydroxybutyrate dehydrogenase (EC 1.1.1.30) (1). Use of this enzyme provided enhanced specificity over the earlier chemical methods (2, 3). The original method involved spectrophotometry, which is time consuming and not particularly sensitive and does not allow for large batch processing. Attempts to automate the procedure by using a conventional continuous-flow analyzer linked to a fluorometer have been successful for 3-hydroxybutyrate (4), but in our hands have not been satisfactory for reproducible and accurate estimation of the small amounts of acetoacetate often present in blood. Recently enzymatic methods have been developed in which pseudo-zero-order or first-order kinetics are used to determine certain substrates (5). These depend on the substrate concentration being many times smaller than the Km value of the enzyme and therefore lend themselves to the measurement of substances at very low concentrations. This paper describes a method for the kinetic determination of acetoacetate, and seeks Dept. of Chemical Pathology and Human Metabolism, ampton General Hospital, Southampton S09 4XY, U. K. Received May 17, 1977; accepted July 28, 1977.
South-
to overcome some of the constraints of using fluorometric and spectrophotometric assays in the presence of very low substrate concentration.
Materials and Methods Principle
of the Method
The method is based on the fact that 3-hydroxybutyrate dehydrogenase catalyzes the following reaction: Acetoacetate
+ NADH + W
3-hydroxybutyrate
+ NAD
At pH 7.0 and with a suitable excess of NADH, at least 98% of the acetoacetate is reduced to 3-hydroxybutyrate (6). Reagents Sodium phosphate buffer, 0.3 mol/liter, pH 6.9. Prepare this by dissolving 29.6 g of Na2HPO4.2H20 and 20.9 g NaH2PO42H2O in 1 liter of distilled water. Sodium phosphate buffer, 0.3 mol/liter, pH 8.2. Prepare this by dissolving 51.8 g of Na2HPO4.2H20 and 1.4 g of NaH2PO4.2H20 in 1 liter of distilled water. Working buffer/coenzyme mixture. To 100 ml of the buffer (pH 6.9 for plasma sample, pH 8.2 for acid extracts) add 30 mg of NADH (Grade II; Boehringer Corporation, Lewes, U.K.). Prepare freshly each day. Hydroxybutyrate dehydrogenase solution. 3-Hydroxybutyrate dehydrogenase prepared from Rhodopseudomonas spheroides, specific activity about 3 kU/g, was obtained from the Boehringer Corp., Ltd. The enzyme was diluted 10-fold for plasma samples and sixfold for acid extracts. Acetoacet ate standard. A stock standard, 10 mmol/ liter, was prepared by dissolving 108 mg of lithium acetoacetate (Sigma Chemical Co., Ltd.) in 100 ml of distilled water. Working standards were prepared over the range 0-0.5 mmol/liter in distilled water or 0.5 mol/liter perchloric acid. It has been our experience that the stock lithium acetoacetate standard (10 mmol/liter) is stable for at least a week. Working standards should be prepared daily. The stability of the stock standard is checked by CLINICALCHEMISTRY,Vol. 23, No. 10, 1977
1893
assaying it by the manual spectrophotometric method in which the concentration of acetoacetate is related to the extinction coefficient of NADH (6).
Table 1. Working Reagents for Kinetic Determination of Acetoacetate in Aqueous, Plasma, and Perchloric Acid Extracts of Whole Blood
Instrumentation All spectrophotometric measurements were made at 37 #{176}C with an LKB 8600 Reaction Rate Analyser (LKB Instruments Ltd., 232 Addington Rd., South Croydon, Surrey, U.K.) The analyzer was set up on the 0-0.05 absorbance scale with background absorbance set to accommodate the initial absorbance of the various solutions used. Table 1 gives detailed machine settings for the method protocols. Procedure Preparation of sample. Several workers have described the measurement of acetoacetate in plasma and deproteinized extracts of whole blood by spectrophotometry and fluorometry (3, 7,8). Three types of samples were therefore investigated: (1) Plasma-Venous blood was collected into commercially prepared tubes (Searle Diagnostics, High Wycombe, Bucks, U.K) containing lithium heparin, mixed by inversion, and centrifuged without delay. (2) Acid extract of deproteinized whole bloodWhole blood was deproteinized by adding about 1.5-2 ml of venous blood to a preweighed bottle containing 5 ml of ice-cold 0.8 mol/liter perchioric acid. The sample was well mixed by inversion and the bottle was reweighed, centrifuged, and the acid supernatant fluid removed for analysis. The dilution of blood was calculated from the weight change observed. (3) Neutral extracts of deproteinized whole blood. By including a laborious neutralization step it is possible to overcome the problems associated with perchioric acid. The perchioric acid supernate was taken into a weighed tube, which was then reweighed. Two drops of Universal Indicator were added and the sample was titrated to about pH 7.0 with 2 mol/liter potassium hydroxide. The tube was then reweighed and centrifuged to remove the potassium perchiorate precipitate. The further dilution was calculated from the weight differences. Assay protocol. The method described allows for measurement of acetoacetate in plasma samples and in whole blood after deproteinization. Pipette 0.2 ml of sample or standard into a reaction cuvette and add 0.8 ml of the appropriate buffer/NADH mixture, depending on the type of sample. Load the reaction cuvettes into the racks and place them into the input magazine of the instrument. Having set up the instrument as indicated in Table 1, and primed the reagent pump, feed the reaction cuvettes through according to the manufacturer’s instructions. Determine the absorbance change per minute from the recorder trace, using the comparator provided, and, having subtracted the reagent blank, determine the sample acetoacetate concentrations by reference to a calibration curve. 1894
CLINICAL CHEMISTRY,
Vol. 23, No. 10, 1977
Concn(acty) In reaction mixture
Reagent
Volume, ml
Plasma system
Sample 240 mol
Sodium phosphate buffer (pH 6.9) NADH 3-Hydroxybutyrate
dehydrogenase
Whole blood/perchioric system
0.34 moI 0.15U/ca.50 Lg of protein
0.2 0.8
0.1
acid
Sample Sodium phosphate buffer (pH 8.2) NADH 3-Hydroxybutyrate dehydrogenase
240 izmol 0.34 .tmol 0.25 U/ca.80 ig of protein
0.2 0.8
0.1
Machine settings: decrease mode, 0.05 absorbance range, delay on, background correction 0.7, measure period, 1 mm
Results Analytical
Variables
Buffer type and pH. The pH optimum is about 7.0 for the reaction (6). Buffers with pKa values in this region-including N- (2-acetamido)-2-aminoethane sulfonic acid, N- (2-acetamido)aminodiacetic acid, tris(hydroxymethyl)methylamine, triethanolamine, and sodium phosphate-were compared at pH 6.9 to determine which buffer supported the highest enzymatic activity. Because of the known lability of NADH in acid solution, the absorbance changes caused by nonenzymatic breakdown of NADH were also assessed for each buffer. It was found that the phosphate buffer offered a marginally more sensitive system. Using a 0.1 mol/liter sodium phosphate buffer, we investigated the effect of pH over the range 6.1-7.5. It can be seen from Figure 1 that the nonenzymatic breakdown of NADH increases with decreasing pH. However, the pH optimum appears to be near 6.9. Investigation of buffer concentration at pH 6.9 showed that nonenzymatic breakdown of NADH increased with molarity of buffers, with a small increase in the rate of enzymic conversion of acetoacetate up to a buffer molarity of 0.3 mol/liter. Coenzyme and enzyme concentration. This type of pseudo-zero-order or first-order reaction, in which the rate is a function of substrate concentration, depends on the relation that exists between the enzyme and its substrates and cofactors. Therefore the effect of varying
0.04
0.05 Iz
0.0L1
0.03 ,D
0.03
0.02
‘C
5 0.02 ‘C
0.01 0.01 -
--._-_-._
6.5
6,7
6.9 PH
AT
370
7.1 C
7.3
0.2
0.1
S
7.5
Fig. 1. pH optimum of 3-hydroxybutyrate dehydrogenase I0.1 mmoi/Iiter acetoacetate, #{149} #{149} reagent blank - -
0,3
CONCENTRATION OF P.C.A. IN
0.4 SAMPLE
0.5 (M0L/L)
Fig. 3. Effect of various perchloric acid concentrations in the sample on the catalytic activity of 3-hydroxybutyrate dehydrogenase
0.40
0.30 C
0.20 C
0
0.10
0.1
0.2 ACETOACETATE
0.3
0.4
0.5
CONCENTRAI!ON(*‘oL/L)
Fig. 2. Variation in absorbance change per minute at different concentrations of enzyme and coenzyme over a range of ace-
toacetate (0-0.5 mmol/liter) Concentrations in cuvette: .-0.34 tmol NADH; - - - - 0.28 moi NADH; 0.23 imol NADH; #{149}, 0.25 U of enzyme. x, 0.15 U of enzyme per cuvette
the NADH and enzyme concentrations was investigated at several acetoacetate concentrations. The NADH range was limited in part by the reaction rate analyzer, which can accommodate a maximum reaction-mixture absorbance of 1.8. A preliminary study suggested that reaction rates were satisfactory at about 0.3 tmol of NADH and 0.2 U of hydroxybutyrate dehydrogenase per cuvette. In a further study, we varied the NADH concentrations in each cuvette from 0.23 to 0.34 umol at enzyme activities of 0.15 and 0.25 U/cuvette over a range of acetoacetate concentrations up to 0.5 mmol/liter. The results (Figure 2) demonstrate that the absorbance change per minute was increased at the higher enzyme concentration, as would be expected. However, linearity of the calibration curve at the highest substrate concentration (0.5 mmol/liter) was a function of NADH concentration.
An enzyme concentration equivalent to 0.15 U per cuvette and an NADH concentration of 0.34 tmol per cuvette were chosen for routine use in the plasma assay system. This enzyme activity gave a satisfactory response relative to the clinical range of acetoacetate concentrations studied. Effect of perch loric acid on plasma assay system. Titration of 0.3 mol/liter perchloric acid with various phosphate solutions showed that it was necessary to increase the pH of the buffer in order to maintain the reaction mixture pH at 6.9 on adding the acid sample. Use of a 0.3 mol/liter sodium phosphate buffer with an initial pH of 8.2 led to a reaction pH of 6.9 with the volumes given in the original assay protocol. To test the effect of perchloric acid on the assay system, we added 0.1 mmol/liter solutions of acetoacetate, prepared in concentrations of perchloric acid ranging from 0 to 0.5 mol/liter, to 0.3 mol/liter phosphate buffers, all prepared to given final reaction mixtures of pH 6.9. The results (Figure 3) show that the sensitivity of the reaction decreases with increasing concentrations of perchioric acid. It was necessary to increase the enzyme concentration when acid extracts were used, because the perchloric acid decreased the catalytic activity. By neutralizing the acid extract as described earlier, it was possible to use the same assay system used for plasma samples. Precision
and Recovery
We studied plasma samples and perchloric acid extracts with and without neutralization. Calibration standards were prepared over the range 0 to 0.5 mmol/ liter in distilled water for the plasma and neutralized samples and in 0.5 mol/liter perchioric acid for the acid samples (Figure 4). Within-batch precision was determined by analyzing 20 replicates of standards (20 to 100 mol/liter) and samples containing low and high concentrations of acetoacetate. The results (Table 2) show that in each instance the precision is considerably better for the higher concentrations of acetoacetate, the inCLINICAL CHEMISTRY,
Vol. 23, No. 10, 1977
1895
0.25
Table 2. WithIn-batch Precision for Samples and Standards at Two Acetoacetate Concentrations Type of sample
PCA standard Aqueous standard
PCA sample Neutral sample
Plasma sample
Mean A change ± SD
CV, %
6.1
Low
0.0062 ± 0.0004 0.0308 ± 0.0012 0.006 ± 0.0006 0.032 ± 0.0012 0.0053 ± 0.0005
High
0.0361±0.0011
Low
0.006 ± 0.0005
3.1 7.6
High
0.087 ± 0.0020
2.3
Low High
0.0 139 ± 0.0006 0.0483 ± 0.0005
4.5 1.1
Concn, mmol/llt.r
0.02 0.10 0.02 0.10
0.20
3.9 9.4 3.6 8.9
0.15
0.05
0.1
Perchioric acid.
0.2
0.3
0.4
ACEIOACETATE CONCENTRATION
0.5
(,v4oL/L)
Fig. 4. Calibration curves for plasma/neutral extract and perchloric acid extract systems (#{149} - - - -
creased coefficient of variation at the low concentrations being mainly due to the very small changes in absorbance. Analytical recovery of acetoacetate was determined by adding a known amount of the stock solution to each type of sample. The results for each system (Table 3) show a variation in the mean recovery for each type of sample. Stability
of Sample
A 200-mi sample of venous blood was taken from each of seven subjects after a 15-h fast. Two-thirds of each sample was placed in lithium heparmn-containing tubes and the remainder was diluted with twice its volume of 0.8 mol/liter perchioric acid in a preweighed flask. Both samples were centrifuged without delay. Half of the acid supernatant fluid was apportioned whilst the remainder was taken through the neutralization procedure before being divided into aliquots for storage. The plasma samples were assayed 10 mm after taking the blood, the perchloric acid-treated samples were assayed 40 mm after taking the blood, and the neutralized samples within an hour. All three types of samples were stored at room temperature, 4#{176}C, and -20 #{176}C immediately after aliquoting when the first measurements were made, then assayed daily for another five days. Figure 5 shows the mean for all samples. Clearly, if the sample is not to be analyzed immediately the supernatant plasma and the perchloric acid extract should be stored at -20 #{176}C. The neutral extracts
should be analyzed
within 24 h and perchloric
(#{149} I) -
acid ex-
tracts within 48 h. Comparison
of Results
Fifty samples of blood were collected into 0.8 mmol/ liter perchloric acid and aliquots of the acid extracts were assayed with the LKB Reaction Rate Analyser. The remaining acid extracts were then neutralized and assayed by the manual spectrophotometric method of Mellanby and Williamson (6). The results (Figure 6) show a good correlation between the two methods. A further 30 samples of blood were taken and the acetoacetate in whole-blood acid extract and plasma were assayed with the Reaction Rate Analyser. The acetoacetate concentrations in the whole-blood samples were within the range 70-95% of the plasma values.
Discussion Since the successful purification of 3-hydroxybutyrate dehydrogenase (1), many variations have been described for the enzymatic estimation of acetoacetate. The methods depend on monitoring the co-enzyme NADH spectrophotometrically or fluorometrically. Spectrophotometric monitoring of NADH is less sensitive and therefore larger sample volumes are required to achieve sensitivity comparable to that of fluorometric assays. The larger sample volumes make it necessary to neutralize acid extracts, so as not to overload the buff-
Table 3. Analytical Recovery of Acetoacetate Added to Plasma, Perchloric Acid (PCA) Extracts, and Neutral Extracts of Whole Blood Type of sample
PCA extract
InitIal concn
Acetoacetate added mmol/Ilter
0.012
0.200
Neutral extract
0.0 17
0.200
Plasma
0.045
0.200
Mean concn after additIon
Mean recovery,
0.213
103 (range 96-115) 95 (range 92-98) 106 (range 102-113)
(range 0.207-0.237) 0.203 (range 0.198-0.2 10) 0.247
(range 0.24-0.263) 1890 CLINICALCHEMISTRY,Vol. 23, No. 10. 1977
%
B,
A.
C.
100 0.4
90 80 70
0.3
60 50 0.2
40
30 20
0.1
10 12345 DAYS
12345 DAYS
12345
0.1
DAYS
Fig. 5. Effect of different storage conditions or) acetoacetate plasma and whole blood perchloric acid extracts
0.2
0.3
BLOOD ACEIOACETATE
in
(A) plasma,(B) perchloric acid extracts, (C) neutralizedperchioric acid extracts. room temperature, - - . -4 Cc, and.... -20 CC
MANUAL
0.5
METHOD (MM0L/L)
Fig. 6. Correlation of results for acetoacetate
by the proposed kinetic procedure and by manual spectrophotometric proce-
dures Regression line: kinetic method
ering capacity of the assay system. Furthermore because of the low concentrations of acetoacetate frequently found in normal subjects and the characteristics of the enzyme used, the reaction requires a lengthy incubation stage to reach completion. In our hands the fluorometric assay, although suitable for measuring acetoacetate concentrations in ketotic conditions, has been unsatisfactory at the normally low substrate concentrations because of the native fluorescence of the sample blanks. Although the spectrophotometric procedure is less sensitive, it can be adapted for use on modern reaction-rate analyzers. These instruments have in many cases been designed with NADH-linked reactions in mind and incorporate electronic initial-absorbance blank-correction facilities to allow accurate and reproducible monitoring of small absorbance changes. This development in instrumentation opens up the possibilities for development of analytical techniques based on kinetic measurements. Consideration of Michaelis-Menten kinetics will show that when the substrate concentration is many times smaller than the Michaelis constant, the reaction rate depends mainly on substrate concentration (9). The kinetic approach has been applied to the measurement of acetoacetate in whole blood and plasma. Acetoacetate may be present in the blood in very low concentrations, and consequently the absorbance changes are small; however, acceptable precision has been achieved without resorting to large sample volumes. The reaction was monitored for 1 mm; longer measuring periods were deemed unsatisfactory because of a loss of linearity of the reaction rate and the instability of the sample. Earlier end-point methods have yielded good precision in many cases as a result of large sample volumes. However it has been shown in this work that when using a perchloric acid extract the effective catalytic activity of the enzyme is reduced and higher levels of 3-hydroxybutyrate dehydrogenase are required. One ap-
0,4
-
=
(manual method)+ 0.005. r
0.935
proach is to neutralize the samples; however, this adds a further step to the method that affects the overall precision of the method. A survey of the literature indicates that there is a wide variation of views on the stability of acetoacetate. Our experience with attempts to increase the period of monitoring of reaction rates led us to conclude that acetoacetate is unstable at room temperature. From the data shown in Figure 5, we conclude that samples of blood must be collected into ice-cold perchloric acid and separated as quickly as possible, with the supernatant fluid being stored at -20 #{176}C no longer than three days. Plasma samples are satisfactory if they are stored at -20 #{176}C, but the analyses must be done within 24 h of specimen collection.
References 1. Gavard, R., Combre, C., and Tuffet, A., Study of the D(-)-/3-hydroxybutyrate dehydrogenase of Bacillus megathenum. C. R. Acad. Sci. 251, 1931 (1960). 2. Williamson, D. H., Mellanby, J., and Krebbs, H. A., Enzymatic determinations of D(-)-fl-hydroxybutyric acid and acetoacetic acid in blood. Biochem. J. 82,90 (1962). 3. Gibbard, S., and Watkins, P. J., A micromethod for the enzymatic determination of beta hydroxybutyrate and acetoacetate. Clin. Chim. Acta, 19,511(1968). 4. Cramp, D. G., Automated fluorimetric techniques based on NAD and NADP-linked enzyme systems in the study of intermediary carbohydrate metabolism. J. Med. Lab. Technol. 27, 359 (1970). 5. Tiffany, T. 0., Jansen, J. M., Burtis,C. A., et al., Enzymatic kinetic rate and end-point analyses of substrate, by use of a GEMSAEC Fast Analyzer. Clin. Chem. 18, 829 (1972). 6. Mellanby, J., and Williamson, D. H., Acetoacetate. In Methods of Enzymatic Analysis. H. U. Bergmeyer, Ed. Academic Press, New York, N. Y., 1974, pp 1840-1943. 7. Wildenhoff, K. E., A micromethod for the enzymatic determination of acetoacetate and 3-hydroxybutyrate in blood and urine. Scand. J. Clin. Lab. Invest. 25, 171 (1970). 8. Ozand, P. T., Hawkins, R. L., Collins, J. M., et al., A microautoanalytic procedure developed for the determination of ketone bodies, gluconeogenic aminoacids, pyruvate, lactate and glucose in metabolic studies. Biochem. Med. 14, 170 (1975). 9. Bergmeyer, H. U., Determinations of concentrations by kinetic methods. In Methods of Enzymatic Analysis, H. U. Bergmeyer, Ed. Academic Press, Inc. New York, N.Y., 1974, pp 131-134. CLINICAL CHEMISTRY,
Vol. 23, No. 10, 1977
1897
