A sensitive automated colorimetric method for the determination of serum gamma-glutamyl transpeptidase

43

Clinica Chimica Acta, 69 (1976) 43-51 @ Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands

CCA 7625 A SENSITIVE DETERMINATION

COLORIMETRIC

Laboratories, Fujisawa Pharmaceutical Co., Ltd. ku, Osaka 532 (Japan)

METHOD FOR THE

Kashima, Yodogawa-

(Received October 21, 1975)

summary A highly sensitive and accurate automated method using a Technicon Autoanalyzer AA11 was developed for the determination of human serum y-glutamyl transpeptidase. y-I.,-Glutamyl-p-nitroanilide added as substrate was split .to p-nitroaniline and the y-glutamyl group was trapped by glycylglycine in Ammediol/HCl buffer solution. Liberated p-nitroaniline was diazotized and converted by Tsuda Reagent (NJ-(diethyl)-N’-( 1-naphthyl)ethylendiamine) to an azo-dye possessing an absorbance maximum at 550 nm. This new method of determination, using the conversion of p-nitroaniline to azo-dye, was found to be 7 times more sensitive as compared to the conventional direct p-nitroaniline determination method [l]. This method requires only 0.1 ml serum and permits 60 determinations per hour.

Introduction Assay of y-glutamyl transpeptidase ( y-GTP) enzyme has been studied clinically in recent years and several methods have already been proposed. N-y-D,L-Glutamyl)aniline [ 21, cu-(N-y-D,L-glutamyl)aminonitrile [ 31, y-L-glutamyla-naphthylamide [4,5] and y-L-glutamyl-p-nitroanilide [1,6-g] were used as substrates in these methods. For the automated determination of T-GTP, Haesen et al. [l] developed a method to measure the absorbance at 410 nm of p-nitroaniline (PNA) liberated from y-L-glutamyl-p-nitroanilide by enzymatic cleavage, but the amount of PNA from y-GTP converted by normal human serum was less than 20 pmol/min/l of serum. This method was found to be insufficiently sensitive to apply in an automatic determination, in preliminary experiments. Orlowski and Szewczuk [4] utilized the Bratton Marshall reaction to form ethanol-soluble azodye from liberated a-naphthylamine. To avoid ethanol in

44

the reaction mixture (which is not desirable for an automated procedure) we attempted to form a water-soluble diazo-coupled compound of PNA with Tsuda reagent [lo]. Spectrophotometric determination at 550 nm of this compound gave a suitable result for our purposes. Material and methods Reagents

1. 0.05M Ammediol buffer (pH 8.7). Ammediol (2-amino-2-methyl-propane1,3-dial; Nakarai Chem. Co. Japan), 5.26 g was dissolved in about 700 ml distilled water. The pH of the solution was adjusted to 8.7 + 0.1 with 0.1 M HCl and then it was diluted to 1000 ml. 3 drops of wetting agent Triton X-405 (Technicon Chem. Belgium T21-302-15) were added. 2. Substrate solution. 2.9 mM y-L-glutamyl-p-nitroanilide (Katayama Chem. Co. Japan); and 22.0 mM glycylglycine (Nakarai Chem. Co. Japan). y-L-Glutamylp-nitroanilide, 241.7 mg and glycylglycine, 2.91 g were dissolved under continuous stirring in 1000 ml of Ammediol buffer solution at 50-60°C. The solution was kept at 4°C or -2O”C, and warmed before use in a water bath at 37°C. The pH of the solution was 8.2 f 0.1. 3. 0.5 M hydrochloric acid. 4. 0.1% sodium nitrite. 5. 0.5% ammonium sulfamate. 6. Color reagent (Tsuda reagent). N,N-(Diethyl)-N-( l-naphthyl)ethylendiamine oxalate (Isuzu Pharm. Co. Japan), 50 mg were dissolved in 100 ml of distilled water. 7. Washing medium. Distilled water, containing 5 ml of 8% Brij-35 (Technicon Chem.. Belgium T21-0110-15) per liter, for washing the recipient line. 8. Standard solution, 1.0 mM p-nitroaniline. p-Nitroaniline (Wako Chem. Co. Japan), 96.6 mg was dissolved in 1000 ml of Ammediol buffer solution. The solution was diluted to the appropriate p-nitroaniline concentration with buffer solution (example: 0.08, 0.04 and 0.02 mM). Manual procedure

y-GTP values for standards were obtained by manual assay, performed according to the method of Jacobs [8]. PNA production from substrate y-Lglutamyl-p-nitroanilide was proportional to the y-GTP activity. Automated

procedure

Automated procedure was carried out by the Technicon Autoanalyzer AAII. A schematic diagram of the system is shown in Fig. 1. Samples were added to an air-segmented stream of buffered substrate and mixed thoroughly. The reaction mixture was then incubated at 37°C through 2 turns of a B coil. Immediately after incubation, HCl was introduced into the dialyzer to stop the enzyme reaction. The liberated PNA was then dialyzed into a HCl stream. The PNA was diazotized by the addition of sodium nitrite solution. Excess sodium nitrite was excluded by the addition of ammonium sulfamate solution. After mixing, the buffered Tsuda reagent was added to form a red-

PROPORTIONING

WASTE.

COLOR 2

RECORDERCBASIC

TYPE

2

PEN)

FROM

REAGENT FLOW

CELL

TO SAMPLER WASH RECEPTACLE

M,C.

= MIXING

COIL

: 10

TURN

Fig. 1. Flow diagram for the determination of y-glutamyl transpeptidase.

purple, tubular

water-soluble flowcell.

azo-dye.

This was measured

at 550 nm through

a 15-mm

Calibration curves PNA is not suitable as a standard for an autoanalyzer system, because this compound stains the pumptube. This is why serum with a large number of units of T-GTP activity was used as a standard after the determination of its enzymatic activity by manual assay. Nevertheless, low PNA content (as we used) did not stain so much that it could not be washed out from the tube with detergent. PNA was introduced directly into the substrate line for obtaining a calibration curve. As shown in Fig. 2, the relationship between extinction coefficient e and amount of PNA was linear up to about 0.2 mM (240 U/l).

E & 20 8 EO

0

0.1

0.2

03

Fig_

2. Typical

peaks

OA

mM)

(p-NITROANILINE and

calibration curve obtained with 0.01-O.40

mM P-nitroaniline standard solutions.

36

Calculation of enzyme units One U/l is the amount of enzyme per liter serum that converts 1 pmol of substrate per min at 37°C. Therefore, the enzyme activities were calculated using the following formula:

1 U/l = 50 x

%X

vs

fst x VW x t

E, = extinction coefficient of the sample; eSt = extinction coefficient of 50 PM PNA; V, = total volume: substrate + serum volume (ml/min) = 1.1; V,,, = serum volume (ml/min) = 0.1; t = incubation time in minutes. Results 1. Absbrbance spectrum of the’coloured complex The spectrum of the coloured complex was obtained as shown in the flow diagram. The maximal absorbance of PNA itself was found at 420 nm and that of diazo-coupled PNA was at 550 nm. The maximum peak of serum sample was also found at 550 nm (Fig. 3). These results indicated the liberation of the PNA by the enzyme reaction. 2. Sensitivity The results of standard PNA between 0.01 and 0.4 mM are shown in Fig. 4, indicating 7-fold sensitivity compared with the result of the direct PNA determination method (direct method) by Haesen et al. [ 11. 3. Stability of the substrate solution The substrate was hydrolysed spontaneously at room temperature increasing its absorbance slightly, but the solution was stable for at least 10 days at 4°C or --20°C (Fig. 5). However, the procedure to melt the frozen solution in a water bath resulted in an increase in the absorbance at 410 nm. Daily preparation is recommended for this reason. 4. The absorbance at 550 nm depends on the amount of enzyme Fig. 6 illustrates the influence of increasing amounts of enzyme. tion of PNA was linear with the amount of serum added.

STANDARD

OF

DIRECT

METHOD

SAMPLE

OF DIAZO-COUPLING

METHOD

h:,

400

420

440

460

480

500

520

540

560

580

600

620

(nm) Kg.

3.

Absorbance

spectra of products obtained by the direct and diazo-coupling

methods.

The libera-

47

DIAZO-COUPLING

METHOD

0

NORMAL

0

IIIRECT ME:HoD

./.J./ RANGE

0.1

0.3

OS2 (P-NITROANILINE

CONCENTRATION

m)

Fig. 4. Comparison of calibration curves obtained by the direct and diazo-coupling were obtained with 0.01-0.40 mM p-nitroaniline standard solutions.

A (41C .lR)

AT

0.4

methods.

The curves

ROOM TEMP.

1,o

008

0.6

0.2 AT

0

0

2

4

-2b'C 6

8

10

(DAYS)

Pig. 5. Stability of the substrate solution. The substrate solution was divided into 3 parts and each solution was stored at room temp., 4°C and -ZO”C, respectively. Stability of the substrate was examined by measuremcnt of the absorbance at 410 nm.

SERLIM

!

Fig. 6. Relationship between absorbance and amount of enzyme. The serum enzymes were diluted with saline.

5. The influences of bilirubin and hemoglobin Addition of bilirubin (2.0-10.0 mg%) and hemoglobin (50-300 mg%) to serum, had no effect either on the absorbance at 550 nm or on enzymatic activity. 6. Correlation between the manual and. au tomated analysis Sixty-seven human sera were analyzed both manually and automatically. A close correlation was observed between the two methods. The least-squares regression line was calculated to be Y = 0.946 X + 2.104, and the coefficient of linear correlation was 0.978 (Fig. 7). 7. Precision Fifteen assays of the same sample gave the mean y-GTP value of 14.49 I 0.298 (U/l t S-D.). The coefficient of variation was 2.05%. Reproducibility of automated method was tested on 43 specimens at an interval of 24 h. The coefficient of these two assays was r = 0.992, and the regression line was calculated to be Y = 0.990 X -- 1.215. It was found that our new method has the highest precision and reproducibility. It was also found that serum y-GTP activity was very stable for at least 24 h. 8. Normal value The activity of the y-GTP enzyme was determined in sera of 30 “healthy” donors. The mean value of activity was 23.2 rf: 3.13 (U/l t: SD.). Their GOT, GPT and alkaline phosphatase values were in the normal range.

49

100

80

0

0

40

20 (x:

Fig. 7. Comparison

MANUAL

60

80

100

(11/l_)

METHOD)

of the manual and automated

methods

for determination

of human sera y_GTP activity.

9. Carry-over To check a possible carry-over, low- (20.8 U/l) and high-activity samples (100-250 U/l) were alternated. The samples were processed as demonstrated in Fig. 8 in a direction from right to left of the chart. As can be derived from Fig. 8, carry-over from the low-activity to the highactivity samples was never more than 4.6%. In an abnormally high range of enzyme activity, 4.6% error in measurement seems not to be serious for diagno-

125.0

m

U/L

20.8 U/L

Fig. 8. Carry-over check carry-over.

examination. High-activity and low-activity samples of human Numbers in chart indicate the enzyme activities employed.

serum were alternated

to

50

sis. On the other hand, only the first subsequent sample of low activity after a very high-activity sample (above 125 U/l, 6 times as high as the low-activity sample), did not produce a flat peak, but a shoulder-shaped trace. When we attempted a calculation of the low enzyme activity based on the value of the turning point of the left side of the trace, 23.7 and 26.4 U/l were obtained in the low-activity sample just after the higher-activity sample of 187.5 and 250 U/l, respectively. The former value was 13% higher and the latter was 27% higher in activity than that of the sample employed (20.8 U/l). Thus, a re-examination of the sample would be required in such a case. Discussion According to Jacobs [S], the enzyme activity increases with the concentration of y-L-glutamyl-p-nitroanilide and reaches a maximum when the substrate sofution is saturated. Saturation was obtained at 2.9 mM concentration and the substrate did not precipitate. When raising the concentration over 3 mM, precipitation appeared. In addition, Szasz reported [ 71 that at a higher concentration of glycylglycine the enzyme activity increases, the maximum being reached at 50 mM but glycylglycine usually precipitated at more than 30 mh4 after 24 h. The enzyme activity, using 20 mM of glycylglycine, was 82% of that determined with 50 mM glycylglycine. After considering these descriptions, 22.0 mM glycylglycine with 2.9 mM y+glutamyl-p-nitroanilide was adopted for our method according to Haesen et al. [l]. Some workers reported [7,8] that the solubility of the substrate was promoted with MgClz which did not activate the enzyme. Neither the activation of enzyme nor the effect on substrate solubility was obtained by addition of MgClz in our experiments. Therefore, we did not use MgC12. The y-GTP enzyme is an excellent indicator of liver diseases, especially in hepatobiliary disease where it indicates the changed y-GTP activity induced by the slight tissue damage much more sensitively than leucine aminopeptidase or alkaline phosphatase [ 11,121. An automated assay of leucine aminopeptidase in serum was recently developed by us (unpublished result). PNA liberated from a substrate L-leu~yl-~-nitroanilide by the attack of this enzyme, was converted to the colored complex. We can combine the two assay methods, namely we can determine y-GTP and leucine aminopeptidase with the same aliquot serum samples only by replacing the substrate. Both enzyme assays are important for the diagnosis of liver diseases, because the classification of liver diseases is possible by the combined data of these two enzyme activities in human serum. References 1

Haesen,

2

Goldbarg,

J.P.,

(1960)

Berends,

J.A., Arch.

3

Szewczuk,

4

Orlowski,

5

Kulhanek,

6

Orlowski,

7

Ssasz,

Biochem.

A.

and

M. and V.

G.T.

Friedman,

and

M. and

G. (1969)

and

Zondag,

O.M.,

Pineda,

Biophys.

G.A. E.P.,

(1972) Smith,

Clin.

Chim.

Orlowski.

M.

(1960)

Clin.

Chim.

Acta

5. 680-688

A.

flb’i2)

Clin.

Chim.

Acta

7, 755-760

Clin.

Chim.

Aeta

Dimov, Clin.

K.M. A.

Chem.

(1966;

(1963) 15.

BiociLT.

124-136

Biaphys.

14, Acta

619--623 73.

37,

463-470

R.. Stein,

91.61-70

Szewczuk, Meister,

Acta

E.E.,Chatterji,

679+681

E.H.

and Rutcmburg,A.M.

51 8 9 10 11 12

Jacobs, W.L.W. (1971) Clin. Chim. Acta 31.175-179 Sericchio. M.. Furedi, R. and Klotzsch. S. (1972) Adv. Autom. Anal. Technicon Int. Congr. Tsuda, K. and Matsunaga. I. (1942) J. Pharm. Sot. Jap. 62.362-364 Lukasik. S.. Richterich. R. and Colombo, J.-P. (1968) Schweiz. Med. Wochenschr. 98. 81-83 Szasz. G., Rosenthal, P. and Fritzsche, W. (1969) Dtsch. Med. Wochenschr. 94. 1911-1917

1, 93-95