AutomatedFluorometricProcedurefor Measurement of ... - CiteSeerX

2 downloads 101 Views 1MB Size Report
Ltd., Chertsey,. Surrey, England). From St. Mary's Hospital, London W.2., England. Reprint requests to Sidney B. Rosalki, M.D., Department of. Diagnostic.
AutomatedFluorometricProcedurefor Measurement of CreatinePhosphokinase Activity J. A. S. Rokos, S. B. Rosalki, and D. Tarlow

An automated procedure is described for measuring creatine phosphokinase activity (in Serum, although the technique is adaptable) with the Technicon AutoAnalyzer. Creatine, formed by the action of the enzyme on phosphocreatine and adenosine diphosphate, is measured by the fluorescence of its reaction product with alkaline ninhydrin solution. Thirty determinations can be made per hour, and only 0.2 ml of sample is required. Additional Keyphrases ninhydrin

AutoAnalyzer

and guanidyl compounds,

violet manual for serum

method,

#{149}

reaction

results compared

#{149}

alkaline ultra-

#{149}

normal

values

Creatine phosphokinase (cPK; ATP: creatine phosphotransferase, EC 2.7.3.2) activity may be measured by various manual procedures (1-7), the most sensitive of which (4, 6, 7) measure formation of ATP or creatine from ADP and phosphocreatine in the reaction: cPK

ADP + phosphocreatine ATP + creatine In 1959, Conn and Davis (8) showed that creatine and other guanidyl compounds react with ninhydrin in alkaline solution to give fluorophors, whereas phosphocreatine reacts to a much lesser extent. This reaction was subsequently used to determine creatine (9) and in the manual determination of CPK (5, 6). We describe here an adaptation of the reaction to the fully automated determination of CPK.

Materials and Methods Principle Creatine formed by the action of creatine phosphokinase on ADP and phosphocreatine reacts with alkaline ninhydrin solution to give a fluorescent product. The fluorescence is measured by use of a fluorometer

connected

(Technicon

Chertsey,

Ltd.,

to

the

Surrey,

AutoAnalyzer

England).

From St. Mary’s Hospital, London W.2., England. Reprint requests to Sidney B. Rosalki, M.D., Department of Diagnostic Chemical Pathology, St. Mary’s Hospital, London W2, U.K. Received Sept. 14, 1971; accepted Nov. 1, 1971.

Reagents Except

where

otherwise

obtained from British Dorset, England. Stock solution.

stated,

Drug

1. N-METHYLMORPHOLINE

all reagents

Houses

Ltd.,

ACETATE

were

Poole,

BUFFER,

0.1 MOL/LITER, pH 6.8. Dilute 1.10 ml of N-methylmorpholine and 8 ml of 1.ON acetic acid (6.0 ml of glacial acetic acid plus 94 ml of distilled water) to about 80 ml with distilled water. If necessary, adjust the pH to 6.8 at 25#{176}C by dropwise addition of N-methylmorpholine or acetic acid, and dilute to 100 ml with distilled water. 2. MAGNESIUM ACETATE, 0.2 MOL/LITEREDTA 7 MMOL/LITER. Dilute 4.29 of magnesium acetate, Mg (CH3.CO2)2.4H2O, and 280 mg of tetrasodium EDTA to 100 ml with distilled water. 3. S’rocic NINHYDRIN, 0.56 M0L/LITER. Dissolve 5 g of 1,2,3-indanetrione monohydrate in 50 ml of methanol. Store the solution in the refrigerator. 4. AQUEOUS NINHYDRIN, 28 MMOL/LITEREDTA, 19 MM0L/LITER. Make a mixture of 10 ml of methanolic ninhydrin stock solution, 1.4 g of disodium EDTA, 80 mg of sodium bicarbonate (final concentration of sodium bicarbonate 5 mmol/ liter), and 0.1 ml of “Brij 35,” and dilute to 200 ml with distilled water. Store in the refrigerator. This reagent is stable for at least three months. 5. AQUEOUS POTASSIUM HYDROXIDE, 1.78 M0L/ LITER. Dissolve 100 g of potassium hydroxide (“AnalaR” grade) in distilled water and dilute to 1 liter. 6. CREATINE STOCK SOLUTION, 10 MMOL/LITER. Dissolve 149 mg of creatine hydrate (Koch-Light Laboratories, Colnbrook, Bucks, England) (or 131 mg of anhydrous creatine), 370 mg of disodium EDTA (final concentration of disoclium EDTA, 10 mmol/liter), and 0.3 ml of 1.ON sodium hydroxide in sufficient distilled water to make a final volume of 100 ml. Refrigerated in a plastic bottle, this reagent is stable for as long as a month. 7. CREATINE STANDARDS. Creatine standards (1, 2, 3,4, 5, and 6 mmol/liter) are prepared as required from the stock 10 mmol/liter solution, and stored in polystyrene tubes. These standards are stable for as long as a month in the refrigerator. Working solutions, prepared freshly as required. 8. MERCAPTOETHANOL, 140 MMOL/LITER. Mercaptoethanol (Sigma Chemical Co., London), 0.1 CLINICAL CHEMISTRY, Vol. 18, No. 3, 1972 193

ml per

10 ml of distilled

water.

Prepare

freshly

every2h.

9. REACTION SOLUTION. Mix four volumes of buffer (stock solution 1), one volume of magnesium acetate-El)TA (stock solution 2), and one volume of mercaptoethanol (solution 8). 10. SUBSTRATE SOLUTION. Mix 8.2 mg of phosphocreatine disodium salt hydrate Sigma, 98100%); 1.6mg of ADP sodium, Grade I (Sigma, 95100%); and 1 ml of reaction solution (solution 9). For each sample assayed, about 1.0 ml of this mixture is required. When the substrate and sample are combined, the final concentrations of reagents (in mmol/liter) are: phosphocreatine, 20; ADP, 2.9; Mg, 28; EDTA, 1; mercaptoethanol, 21; and N-methylmorpholine

buffer,

59.

11. SAMPLE-BLANK SOLUTION. To 1 ml of solution 9 add 4.1 mg of phosphocreatine. For each sample assayed, about 1.0 ml of this solution is required. 12. STANDARD-BLANK SOLUTION: CREATINE, 0.05 MMOL/LITER. Dilute 0.5 ml of creatine (10 mmol/ liter stock solution 6) to 100 ml with water.

Apparatus The AutoAnalyzer manifold is shown diagrammatically in Figure 1. The following points require additional comment: The mixing coils subsequent to potassium hydroxide addition are enclosed in a light-excluding box during fluorescence development, because ambient light has been reported to decrease the fluorescence

(8).

To decrease interference from fluorescenceyielding compounds in serum, we arranged the time delay in these mixing coils so that fluorescence is measured about 3 mm after the potassium

is mixed with the reagent stream. Manual testing showed that alkaline-ninhydrin added to creatine yields maximum fluorescence within this time (Figure 2, curve A), whereas when added to serum alone, fluorescence is well below maximal (Figure 2, curve B). The fluorescenceyielding compounds in serum, which probably represent arginyl residues in the protein, can be removed by protein precipitants. We used a Model LFM/4 fluorometer (The Locarte Co., London), fitted with a quartz flowthrough cuvet, an LF/12 primary filter (peak transmission, 405 nm), and a LF/6 secondary filter (cut off below 470 nm). A 5 k2 resistance was incorporated across the outlet of the fluorometer recorder. With zero fluorescence, a reading of 100 is normally obtained on the Technicon recorder chart. Recorder chart speed was 18 inches (46 cm) per hour.

hydroxide

Proced ure For enzyme activity determinations, the baseline is set by aspirating standard-blank solution (solution 12) into the substrate line, and water into the sample line. Solution 12 contains a low concentration of creatine to ensure that the baseline is on the scale. A series of creatine standards ranging up to 6 mmol/liter is then aspirated through the sample line (standard run), the sensitivity of the fluorometer having been adjusted to give a suitable scale deflection with the highest standard. This sensitivity

is kept

constant.

After the creatine standards have been aspirated, serum or other enzyme-containing samples are aspirated through the sample line with substrate solution (solution 10) at the substrate line (test run). Blanks are then measured by re-running Flue r.te ml, a 0.10 Sample

Air Suhtrat,

SAMP1CRJ

60/h

Fig.

1. AutoAnalyzer

2:1

flow

Ni’shydrin-fDTA Ninhjdria-(DTA

diagram for the fluorometric determination phosphokinase

of

creatine ICOH

(H3, Dl, and Co are connectors,

Water

designated according to their Technicon Catalog reference. SMC, single mixing coil; DMC, double mixing coil) Liqht.ezcluding box

F LU0MtT CR with llow.throu* Cu,, tt, 194

CLINICAL CHEMISTRY, Vol. 18, No. 3,

stir wash

1972

RLCORDtR

the mixing coil before entering the incubation coil. In this study, we used a 20-foot (610 cm) incubation coil in the heating bath, and total incubation time was about 10 miii. Enzyme

activity

is calculated

from

the

formula:

U/iiter=ETBXi000 where U/liter is the activity in International Units (U = micromoles of creatune formed per minute) per liter of enzyme sample; E is the creatune equivalent obtained for the sample run with substrate solution; B is the creatune equivalent obtamed for the sample run with sample-blank solution; and T is the time of incubation in minutes. Under test conditions, sera do not normally pos-

.2 0

sess

A

I 0

___________________

3

I

9

6

significant

activity

of

enzymes

other

than

that can form creatine from phosphocreatine; we obtained identical blank values by running sera with either standard-blank or sample-blank solutions. Therefore, less phosphocreatine is used if standard-blank solution (solution 12) is substituted for sample-blank solution (solution 11) in the blank run. If this is done, the blank is not a true enzyme blank, but only compensates for endogenous sample fluorescence. Water may be aspirated at the sample line during substrate aspiration, to correct for baseline alterations owing to any creatine present as an impurity in the phosphocreatine substrate or formed from it nonenzyCPK

Time in minutes

Fig. 2. Time-course of fluorescence development from mixtures of alkaline-ninhydrin and creatine (A) or alkaline.ninhydrin and serum (B)

matically.

Results the samples with sample-blank 11) at the substrate line.

solution

(solution

Calculation Before enzyme activity is calculated, both the creatine formed by the action of CPK and the time of incubation must be known. Plot a graph of creatine concentration vs. peak height for the creatine standards and obtain the creatine equivalent of the chart peaks for the test and blank runs from this. To determine the period of incubation, measure the time elapsing between mixing of substrate and sample and entry of the mixture into the heatingbath incubation coil and also to the point where the stream is mixed with nunhydrin-EDTA; these times can be measured by timing the progress of suitably colored marker solutions. The time required for the reagent stream to pass between the two latter points in the system represents the incubation period at 37#{176}C. A temperature correction factor (7) may be applied for the short time during which sample and substrate are at room temperature in

A typical standard, test, and blank run is shown in Figure 3. The standard curve for creatine was linear to a creatine concentration of at least 6 mmol/liter (Figure 4). Above this value there was some non-

D

I

Creatine conc. mmollI Standards

Blanks

Fig. 3. Typical AutoAnalyzer blank, and test runs

Tests

chart tracings

of standard,

Samples 1-6 are creatine standards, A-H are patient sera, and I-N are aqueous dilutions of different batches of control sera CLINICAL CHEMISTRY, Vol. 18, No. 3, 1972 195

a

S

$0

6

6

4

I

55 I

ii 60

2

40

I

,,

#{149} ,

,,,

20 Duplicate Creatine

0

Fig. 4.

linearity,

2

3

4 5 Cuotlsg dsadsrd coacsmtrotloa I

Creatine

probably

standard

owing

calibration

7 6 smol/l

precision,

to concentration

determined

quench-

from

results

of duplicate determinations on 48 sera with activities between 22 and 284 U/liter, showed a relative standard deviation (coefficient of variation) of 2.27% (mean, 116 U/liter; SD, 2.63). Duplicate determinations of blank activity of 12 undiluted sera of mean creatine equivalent 1.83 mmol/liter showed a standard deviation of .034 with a relative standard deviation of 1.86%. Between-batch precision was studied for five days, during which time determinations were repeated on 55 sera within 24 h of the initial determination, the samples having been stored at either 4#{176}C or -18#{176}C meanwhile. Activities ranged from 38 to 284 U/liter (mean, 115 U/liter initially, 106 U/liter at repeat determination). Between-batch precision (relative standard deviation) on these samples was 6.83%. Although the difference be196 CLINICAL CHEMISTRY, Vol. 18,

Steady. state

Fig. 5. AutoAnalyzer tracings, showing (left to right) standards run in duplicate, interaction patterns, and steady.state recording

curve

ing. Serum does not interfere with development of the creatine fluorescence; the fluorescence readings obtained by the simultaneous aspiration of serum and creatine standards were additive. Ten replicate determinations, carried out on a serum of mean activity 190 U/liter, showed a coefficient of variation of 1.42% (SD, 2.70). This indicated between-sample carryover to be negligible. In addition, when sera specimens were alternately preceded by sera having high or low activity, nearly identical values were obtained, again indicating negligible sample interaction. Duplicate determinations, interaction patterns, and steadystate recordings obtained with creatine standards are shown in Figure 5. Within-batch

Standards Interaction conc. mmolIl pattern

No. 3, 1972

tween these means was significant at the 0.1% level, the activities obtained for fresh control sera, included in each batch, showed that this was the result of between-batch analytical variation and not of sample deterioration. With a freshly reconstituted CPK control serum (“cPK Control”; Dade Division, American Hospital Supply Corp., Miami, Fla. 33152) included in each daily run of determinations and the control serum itself used as a reference (secondary standard), between-batch precision of these samples improved to 3.64%. It might therefore be valuable to calibrate the method by including dilutions of a control serum in each test run, rather than to calibrate with creatine standards. However, reliance is then placed on the manufacturer’s quality control, product stability, and labeled assay values, and it would be advisable to check the labeled values by a standard assay procedure before calibration. The lyophilized vials of CPK control serum that we used for standardization were stable at 4#{176}C for at least six months, and for five days after reconstitution. Enzyme activities were determined by the present procedure on fivefold and 10-fold dilutions of CPK control sera having nine different batch numbers. The results were compared with values obtained by a standard manual ultraviolet method (7) on undiluted material or material diluted twofold. Distilled water was used as diluent throughout. Correlation was excellent (r = 0.976) between the two procedures (Figure 6), but enzyme activitties as determined by the automated method (calculated from creatine yield in unit time) were 2.20 times as great as those of the ultraviolet procedure. A previous study of 100 normal sera by the ultraviolet

method

suggested

normal

values

up to

0

a

U

1.400 It

I

I I

U 0

I-. -I

3

/

700

e a

0

I;

.

a. ‘I

a

0 CPK

I400

700 .ctivity ( U per literit

37#{176}C). Ultro4olet

method

S

a U

Fig. 6. Comparison of cpx activities of control sera, as determined by the automated fluorometric procedure and a manual ultraviolet method (7)

80 U/liter at 37#{176}C for males and 50 U/liter for females (7). The fluorometric procedure yields apparent activities that are about twice those obtained by the ultraviolet method, so anticipated normal values with the fluorometric method would be as great as 160 U/liter at 37#{176}C for males and 100 U/liter for females. In fact, the range of CPK activity for 20 normal sera (from males and females) was 13 to 124 U/liter at 37#{176}C by the fluorometric method. The higher activities obtained with the fluorometric procedure compared with the ultraviolet method are presumed to result from a combination of the following: (a) the more optimal creatine phosphate concentration (limited in the original ultraviolet procedure by expense), (b) the absence of a coupled enzyme system with elimination of the possibility of interfering sidereactions,

(c) the omission

of adenosine

monophos-

phate (which may slightly inhibit CPK activity, but is required by the ultraviolet procedure for myokinase inhibition), and (d) the avoidance of inhibitory ions. Sera yielding creatine equivalents as great as 6 mmol/liter can be measured by the fluorometric method without dilution. Such sera generally show a blank value equivalent to a creatine concentration of some 2 mmol/liter, and the difference between test and blank runs represents a creatine equivalent of 3 to 4 mmol/liter. Under the laboratory conditions this corresponded to a range of activity up to 300 U/liter at 37#{176}C. This implies that sera with activities greater than twice the upper limit of normal must first be diluted; it also means that the procedure is particularly sensitive for assaying sera having activities in the normal range, or just above it.

Sempi. number

Fig. 7. Effect of various ions on the fluorescence of mixtures of creatine with alkaline-ninhydrin: A, with water aspirated at the substrate line; B, with EDTA, 50 mmol/ liter, at the substrate line; and C with 1DTA, 50 mmol/ liter, and mercaptoethanol, 20 mmol/liter, at the substrate line from all other reaction solutions. Samples creatine, 2 mmol/Iiter. In addition, Sample S contained trietha nolomine hydrochloride, 200 mmol/liter; Sample 3, magnesium chloride, 200mmol/Iiter; Sample 4, manS ganese chloride, 200 mmol/liter; Sample 5, sodium chloride, 200 mmol/liter; Sample 6, aluminum potassium sulfate, 100 mmol/ liter; Sample 7, ammonium chloride, 400 mmol/liter; and Sample 8, calcium chloride, 200 mmol/Iiter. Magnesium, manganese and calcium ions produce marked interference with fluorescence development, which is reversed by EDTA. Mercaptoethanol does not interfere }DTA

was omitted

1 to 8 contained

Sera of high activity were diluted fivefold or 10fold with distilled water or buffer (stock solution 1). Sera with normal activity showed a slight relative increase (less than 10%) of apparent CPK activity on fivefold dilution. This increase was abolished when sera were diluted with heated serum or with a solution of bovine serum albumin (70 g/liter). CLINICAL

CHEMISTRY,

Vol. 18, No. 3, 1972

197

Discussion The concentrations of substrate and reactants that we used were chosen on the basis of published manual procedures (5-7). Enzyme activity did not increase when either phosphocreatine or ADP substrate concentration was increased, and the reactant concentrations chosen yielded linear enzyme values with suitably diluted enzyme samples within the range of the method. N-Methylmorpholine acetate was used as buffer, because its buffering capacity (pKa of N-methylmorpholine, 7.41) is greatest near the optimal pH (6.8) for CPK determination, and, unlike other buffers recommended for use for CPK determination by the fluorescent alkaline-ninhydrin reaction [e.g., imidazole-acetate (5), Tris-acetate (6)], did not interfere with the development of fluorescence from creatine under the conditions of the test. Chloride and sulfate ions were avoided in the substrate, because these inhibit CPK (10). Divalent ions, which must be added to the substrate as enzyme activators, decrease the fluorescence intensity. However, if excess EDTA is added to the ninhydrin solution, as suggested by Conn and Anido (6), this interference is completely eliminated (Figure 7, curves A and B). Mercaptoethanol, added to the substrate as a sulfhydryl donor to activate CPK, did not interfere (Figure 7, curves B and C). Other automated procedures that have been used to measure CPK activity in the direction of creatine formation have depended on fluorometry (11) or colorimetry of creatine after its reaction with diacetyl and o-naphthol (12, 13). All these methods

include

a dialysis

stage,

which

is omitted

here, with resulting procedural simplification. In addition, the a-naphthol/diacetyl methods require suitably pure o-naphthol, which must be protected from auto-oxidation during the test run. The manifold described here is simple and straightforward. No dialysis is required, because under the conditions of assay, protein does not interfere with the development of fluorescence from creatine. Although fluorescence of creatine and protein is additive, blank determinations correct for any fluorescence of the sample that is not the result of its CPK activity. The procedure is extremely sensitive: small

198 CLINICAL CHEMISTRY, Vol. 18, No. 3, 1972

amounts of creatine from sera having low enzyme activity are readily detected, and only small sample volumes (0.2 ml) are required, a feature of special value in research studies. The procedure is also very suitable for enzyme kinetic studies, because, unlike other commonly used procedures (2, 7), it does not depend on a coupled enzyme system.

This work was supported by a grant from the Medical Research Council, London. We also thank Mr. J. H. Glenn and Mr. A. Williams for valued technical advice; Dade Reagents, Inc., for a gift of CPK control sera; and The Locarte Co., London, for the gift of a quartz flow-through fluorometer cuvet.

References 1. Oliver, J. T., A spectrophotometric mination of creatinine phosphokinase

method for the deterand myokinase. Biochem.

J. 61, 116 (1955). 2. Tanzer, M. L., and Gilvarg, C., Creatine and creatine kinase measurement. .1. Biol. Chem. 234, 3201 (1959). 3. Okinaka, S., Kumagai, K., Ebashi, S., Sugita, H., Momoi, H., Toyokura, Y., and Fugie, Y., Serum creatine phosphokinase. Activity in progressive muscular dystrophy and neuro-muscular disease. Arch. Neurol. 4, 520 (1961). 4. Hughes, B. P., A method for the estimation of serum creatine kinase and its use in comparing creatine kinase and aldolase activity in normal and pathological sera. Clin. C/rim. Ada 7, 597 (1962). 5. Sax, S. M., and Moore, J. J., Fluorometric measurement creatine kinase activity. CLIN. CHEM. 11, 951 (1965).

of

6. Conn,

R. B., Jr., and Anido, V., Creatine phosphokinase by the fluorescent ninhydrin reaction. Amer. J. Clin. Pat/wi. 46, 177 (1966).

determination

7. Rosalki, S. B., An improved procedure for serum creatine phosphokinase determination. J. Lab. Clin. Mcd. 69, 696 (1967). 8. Conn, R. B., Jr., and Davis, R. B., Green fluorescence of guanidinium compounds with ninhydrin. Nature 183, 1053 (1959). 9. Conn, R. B., Jr., Fluorimetric determination of creatine. CLIN. CHEM. 6, 537 (1960). 10. Nihei, T., Noda, L., and Morales, M. F., Kinetic properties and equilibrium constant of the adenosine triphosphatecreatine transphosphorylase-catalyzed reaction. J. Biol. Chem. 236, 3203 (1961). 11. Willis, C. E., Nosal, T., and King, J. W., Automated fluorometric determination of serum creatinine phosphokinase by the ninhydrin reaction. In Automation in Analytical Chemi8ty Technicon Symposia 1967, 1; N. B. Scova, et al., Eds. Mediad, White Plains, N.Y., 1968, pp 579-582. 12. Fleisher, G. A., Automated method for the determination of serum creatine kinase activity. CLIN. CHEM. 13, 233 (1967).

13. Siegel, A. L., and Cohen, P. S., An automated determination of creatine phosphokinase. In Automation in Analytical Chemistry, Technicon Symposia 1988, 1; E. Kawerau, et al., Eds. Mediad White Plains, N.Y., 1967, pp 474-476.