- Email: [email protected]
An ultra-micro method for the determination of total nitrogen in biological fluids based on Kjeldahl digestion and enzymatic estimation of ammonia
129
Clinica Chimico Acta, 94 (1979) 129-135 0 Elsevier/North-Holland Biomedical Press
CGA 10086
AN ULTRA-MICRO METHOD FOR THE DETERMINATION OF TOTAL NITROGEN IN BIOLOGICAL FLUIDS BASED ON KJELDAHL DIGESTION AND ENZYMATIC ESTIMATION OF AMMONIA
ERIK M. SMIT Department of Medicine, Binnengasthuis, (The Netherlands) (Received October llth,
University of Amsterdam,
Amsterdam
1978)
Summary An ultra-micro method for the determination of the total nitrogen-content of biological fluids and suspensions is described, based on a digestion in sulphuric acid and a enzymatic determination of the ammonia formed with glutamate dehydrogenase (EC 1.4.1.3). The proposed method yields the same results as the classical Kjeldahl procedure, but is less time-consuming. The detection-limit of the nitrogen, without loss of precision and accuracy, is much lower than in the original Kjeldahl procedure, and is in the order of 35 ng N per sample.
Introduction In 1883 Kjeldahl [l] described a new method for nitrogen determination in organic material, based on the estimation of the ammonia formed after treatment with hot sulphuric acid of the sample. The Kjeldahl procedure is nowadays still the standard method for the determination of total nitrogen in biological samples. Although in the century following Kjeldahl’s publication almost every nitrogen-containing compound in biological fluids, such as blood and urine, can be determined separately and quantitatively, the estimation of total nitrogen is far from obsolete. For the study of nitrogen metabolism, total nitrogen determination is an essential tool. The Kjeldahl method consists of two distinct analytical procedures: (1) the digestion of the nitrogen compounds with hot sulphuric acid, and simultaneous reduction to ammonia and (2) isolation and determination of the ammonia formed. Both steps have been intensively investigated [2,3]. However, in spite of all proposed modifications, the Kjeldahl procedure is still a very laborious method, requiring relatively large sample-volumes.
130
This paper deals with the direct estimation of the ammonia in the sulphuric acid digest by a sensitive and specific enzymatic method. The procedures described meets the present day demand for rapid, sensitive, ultra-microanalysis requiring a minimum of technician-time. Methods and materials
The Kjeldahl procedure was carried out as described by De Vries [4] in an all glass apparatus according to Pregl (manufacturer: Brand, Werthein, F.R.G.). In the proposed method, 50 ~1 sample (blank, urine, serum, standard solution or other biological fluid/suspension) is mixed, in 160 X 16 mm standard test tubes, with 1.5 ml digestion mixture. The tubes are placed in a test-tube heater and heated for 1 h at 130°C. After evaporation of the water the temperature is raised to 320°C and the heating at this temperature is continued for another hour. The content of the tubes is then diluted with 20 ml water. 50 ~1 of these dilutions is pipetted into acryl-LKB-cuvettes (LKB-Products Bromma, Sweden) and 1 ml coenzyme-buffer solution is added. After 5 min preincubation at room temperature, the absorbance at 340 nm is measured in a LKB Calculating Absorptiometer at a speed of 100 cuvettes per 3 min. After the first reading 10 ~1 of glutamate dehydrogenase (EC 1.4.1.3) is added with a Hamilton microdispenser. After mixing, the cuvettes are incubated for 10 min at room temperature before the second absorbance measurement is made. The difference between the second and first absorbance measurement is proportional to the ammonia concentration. The ammonia concentration is calculated from standard curves obtained from ammonium sulphate and urea solutions. All determinations were carried out in duplicate; the pipetting was done with automatic pipets and dispensers. All chemicals used were of analytical grade. Glutamate dehydrogenase (EC 1.4.1.3) in buffered 50% glycerol was obtained from Boehringer. Vamin (Vitrum) is a 7% solution of 18 amino acids in 10% glucose and is used as a fluid for intravenous nutrition. Digestion mixture: The digestion mixture contained per litre 200 ml H2S04 (96%), 200 g K,S04 and 7 g HgO. This is a clear and easily dispensable solution at temperatures above 35°C. The effective composition of the mixture, after evaporation of the water, is per ml of H2S04: 1 g KzS04 and 32.4 mg Hg** as recommended by Fleck and Munro [2]. The enzymatic determination of ammonia was carried out either with the Boehringer coenzyme-buffer mixture, Boehringer Monotest (Art. No. 125857) consisting of a coenzyme-buffer vial and a glutamate dehydrogenase vial, or with a homemade coenzyme-buffer solution. The Boehringer coenzyme-buffer solution contains triethanolamine 130 mmol/l, a-oxoglutarate 13 mmol/l, ADP 1.3 mmol/l and NADPH 0.08 mmol/l. The home-made solution consists of triethanolamine 150 mml/l, a-oxoglutarate 15 mmol/l, ADP 1.5 mmol/l, NADH 0.1 mmol/l and EDTA 5 mmol/l. pH of both solutions is 8.6. Results
A major problem of direct ammonia estimation of the Hg2+ concentration on the determination
in the digest is the influence of it [2,3]. In Fig. 1 the
131
0
100
200
Hg”~mol/l
300
in sample
t
D I
Fig. 1. Influence of mercuri-ions on the recovery of ammonia in the enzymic reaction. The experiment is described in the text: 80 Ctmol/l Hg*+ in the cuvette corresponds with 1 ml digestion mixture in the testtube. All points are the means of duplicate determinations. 0, Boehringer coenzyme-buffer solution; A, home-made coenzyme-buffer solution; *, home-made coenzyme-buffer solution minus EDTA.
recovery of the ammonia in the enzymic reaction is given as function of the Hg*+ concentration in the cuvette. For this experiment 50 ~1 of a 150 mmol/l ammonium sulphate solution was mixed with various amounts of digestion mixture (up to 5 ml) and diluted with water to 20 ml. 50 ~1 of these dilutions were enzymatically analyzed as described in Methods. The reaction with the Boehringer-reagent, using NADPH, is inhibited less than the reaction with the home-made coenzyme-buffer solution, using NADH. However addition of EDTA, final concentration 5 mmol/l, easily overcomes the inhibition of the enzymic reaction. When Hg*+ was left out of the digestion mixture no inhibition of the reaction is found with or without EDTA in the coenzyme-buffer solution. The experiments described were performed with the home-made coenzyme-buffer solution because this is less espensive than the Boehringer Monotest. In- Fig. 2 the influence of the digestion temperature and time is demonstrated for urea (400 mmol/l), urine, serum albumin (70 g/l), Autonorm standard serum (Nyegaard and Co. AS. Oslo), Vamin and tryptophan 245 mmol/l in 0.1 mol/l NaOH. The tryptophan solution was included in the experiments because this amino acid is notoriously resistant to destruction by sulphuric acid [51. From Fig. 2 it is clear that at temperatures lower than 300°C not all natural constituents of urine and serum can be converted to ammonia. The apparent 100% conversion to ammonia of the nitrogen in the urine at 290°C is a result of the relatively low content of protein and aminoacids of this particular urine. One hour of heating at 320°C converted all nitrogen-compounds in urine,
132
1
Fig. 2. Digestion efficiency as function of time and temperature. Measurements were performed as described in Methods, except for the differences in time and temperature indicated. Concentrations: urea 400 mmol/l,serumalbumin 70 g/l and tryptophan 245 mmol/l. Bottom 2OO’C; middle 260°C; top 320°C. t = 0 is the moment when the temperature in the test-tube reached the indicated temperature. Each bar represents the mean of three independent experiments; all determinations carried out in duplicate.
serum and Vamin to ammonia. The procedure described in Methods is the resultant of these experiments. In Fig. 3 a representative standard curve is drawn for urea, serum albumin and ammonium sulphate standard solutions. A linear relationship was between the A 340nm and the total nitrogen-content for all these solutions. In the method described the total nitrogen-content in the 50-~1 sample
133
OL
0
10
5
N
g/l
Fig. 3. Standard curve for nitrogen determination. 0, urea 100. 200. 400 mmol/l; 4, serum albumin 30. 50. 70 g/l and 0, ammonium sulphate 150. 250 mmol/l were treated as described in Methods for the proposed method. The AA340 ls measured ln a LKB 2074 calculating absorptiometer.
hl Fig. 4. Comparison of the Kjeldahl method in various determinations were carried urine faecalsuspensions and
TABLE
ultra-micro-semisnsymatic total nitrogen determination with the macrobiological fluids. The Kjeldahl and ultra-micro-semi-enzymatic total nitrogen out as given ln Methods. Determinations of total nitrogen content in serum, fluids from drains are compiled in this figure.
I
PRECISION
OF THE DESCRIBED
METHOD
Coefficient Serum (9.05 g N/I) Urine (5.43 g N/l)
of variation
2.9% (n = 20) 3.4% (n = 201
(day-to-day)
134
amounts to 14-560 pg N. However in the analysis only a very small part (25 X 10m4) of it is used. Other combinations of sample-volume and dilution of the digest are possible, however, but volumes of less than 50 ~1 increase the experimental error of the method. For normal urines and sera this proved to be the practical procedure. The day-to-day variation in the analysis is given in Table I for urine and serum, using ammonium sulphate and urea as standards. The results of the method described were in excellent agreement with the Kjeldahl procedure as shown in Fig. 4. Discussion The Kjaldahl procedure is very time-consuming and the analytical steps involved make the method susceptible to errors, especially when the procedure is carried out precisely. Many investigators have reduced the number of analytical steps by introducing direct chemical determinations of the ammonia formed in the digest. However, presence of mercuri-ions or turbidity of the digest interfere with the chemical colour reactions [2,3]. The proposed enzymatic determination in the digest is not inhibited by either mercuri ions, as shown in Fig. 1, or turbidity, since this does not influence the differential absorbance measurement. In the method described the total number of critical analytical steps is reduced and can be carried out by simple devices like Microsanz pipettes (Beckman Instruments), Eppendorf pipettes and automatic dispensers, since all manipulations are performed identically for sample, blank and standard solutions. Digestion for 1 h at 320°C is sufficient to convert all important constituents of human excreta to ammonia as is demonstrated in Fig. 2. This is in accordance with the temperature used by Dambacher et al. [3]. Meijers and Rutten [6] used a temperature of 200°C for the perchloric acid digestion of urine. Although in normal urine the bulk of the nitrogen is in the form of urea, which is readily converted by the sulphuric acid digestion mixture to ammonia at 200°C protein and substantial amounts of amino acids may be present. A digestion temperature of 200°C is, therefore, too low for the estimation of the total nitrogen-content in urine. It is doubtful if the digestion temperature of 200°C for the perchloric acid, used by Meijers and Rutten [6], is high enough to convert protein and amino acids to ammonia. The precision and accuracy of the method described is good, and is comparable to the macro-Kjeldahl method, as is shown in Table I and Fig. 4. The sensitivity and specificity of the enzymic determination makes it possible to detect minute amounts of nitrogen. In the procedure described, 50 ~1 of sample is used, but this can be scaled down to a few microliters when less than 20 ml water is used for the dilution of the digest. However, using a larger proportion (more than 10 times) of the digest in the enzymic reaction, measures should be taken to overcome the effects of the increased acid and mercuri-ion concentrations in the cuvette. This can be effected either by increasing the buffer capacity and EDTA concentration of the coenzyme-buffer solution or diluting the digest with alkaline EDTA instead of water.
135
The detection limit for the enzymic reaction is about 2.5 nmol ammonia or 35 ng N in the cuvette for a 1 ml reaction volume. Theoretically, this is the amount of nitrogen in the sample which can be detected when the appropriate measures are taken. The proposed method for total nitrogen determination in biological fluids enables clinical laboratories to perform large numbers of determinations with high sensitivity, precision and accuracy using standard laboratory equipment. Acknowledgements The author wishes to thank Mr. H. Sauerwein M.D. for sending so many samples for total nitrogen analysis that it urged us to develop the described method, Mrs. B. Schoemaker and Mr. E. Timmer for expert technical assistance and Mrs. R. van Nieuwenhoven for typing the manuscript. References 1 2 3 4
Kjeldahl. J. (1883) Z. Anal. Chem. 22.366-382 Fleck, A. and Munro, H.N. (1965) Clin. Chim. Acta 11.2-12 Dambacher. M., Gubler, A. and Haas, H.G. (1968) Clin. Chem. 14.615-622 De Vries, L.A. (1955) in Klinische Diagnostiek (Garter, E. and De Graaff. W.C., eds.), PP. 434-437, H.E. Stenfert Kroesse, Leiden 5 Lake, G.R., McCutchan, P.. van Meter, R. and Neel. J.C. (1951) Anal. Chem. 23.1634-1638 6 Meijers. C.A.M. and Rutten. J.C.J.M. (1969) Clin. Chim. Acta 24,308-310
Clinica Chimico Acta, 94 (1979) 129-135 0 Elsevier/North-Holland Biomedical Press
CGA 10086
AN ULTRA-MICRO METHOD FOR THE DETERMINATION OF TOTAL NITROGEN IN BIOLOGICAL FLUIDS BASED ON KJELDAHL DIGESTION AND ENZYMATIC ESTIMATION OF AMMONIA
ERIK M. SMIT Department of Medicine, Binnengasthuis, (The Netherlands) (Received October llth,
University of Amsterdam,
Amsterdam
1978)
Summary An ultra-micro method for the determination of the total nitrogen-content of biological fluids and suspensions is described, based on a digestion in sulphuric acid and a enzymatic determination of the ammonia formed with glutamate dehydrogenase (EC 1.4.1.3). The proposed method yields the same results as the classical Kjeldahl procedure, but is less time-consuming. The detection-limit of the nitrogen, without loss of precision and accuracy, is much lower than in the original Kjeldahl procedure, and is in the order of 35 ng N per sample.
Introduction In 1883 Kjeldahl [l] described a new method for nitrogen determination in organic material, based on the estimation of the ammonia formed after treatment with hot sulphuric acid of the sample. The Kjeldahl procedure is nowadays still the standard method for the determination of total nitrogen in biological samples. Although in the century following Kjeldahl’s publication almost every nitrogen-containing compound in biological fluids, such as blood and urine, can be determined separately and quantitatively, the estimation of total nitrogen is far from obsolete. For the study of nitrogen metabolism, total nitrogen determination is an essential tool. The Kjeldahl method consists of two distinct analytical procedures: (1) the digestion of the nitrogen compounds with hot sulphuric acid, and simultaneous reduction to ammonia and (2) isolation and determination of the ammonia formed. Both steps have been intensively investigated [2,3]. However, in spite of all proposed modifications, the Kjeldahl procedure is still a very laborious method, requiring relatively large sample-volumes.
130
This paper deals with the direct estimation of the ammonia in the sulphuric acid digest by a sensitive and specific enzymatic method. The procedures described meets the present day demand for rapid, sensitive, ultra-microanalysis requiring a minimum of technician-time. Methods and materials
The Kjeldahl procedure was carried out as described by De Vries [4] in an all glass apparatus according to Pregl (manufacturer: Brand, Werthein, F.R.G.). In the proposed method, 50 ~1 sample (blank, urine, serum, standard solution or other biological fluid/suspension) is mixed, in 160 X 16 mm standard test tubes, with 1.5 ml digestion mixture. The tubes are placed in a test-tube heater and heated for 1 h at 130°C. After evaporation of the water the temperature is raised to 320°C and the heating at this temperature is continued for another hour. The content of the tubes is then diluted with 20 ml water. 50 ~1 of these dilutions is pipetted into acryl-LKB-cuvettes (LKB-Products Bromma, Sweden) and 1 ml coenzyme-buffer solution is added. After 5 min preincubation at room temperature, the absorbance at 340 nm is measured in a LKB Calculating Absorptiometer at a speed of 100 cuvettes per 3 min. After the first reading 10 ~1 of glutamate dehydrogenase (EC 1.4.1.3) is added with a Hamilton microdispenser. After mixing, the cuvettes are incubated for 10 min at room temperature before the second absorbance measurement is made. The difference between the second and first absorbance measurement is proportional to the ammonia concentration. The ammonia concentration is calculated from standard curves obtained from ammonium sulphate and urea solutions. All determinations were carried out in duplicate; the pipetting was done with automatic pipets and dispensers. All chemicals used were of analytical grade. Glutamate dehydrogenase (EC 1.4.1.3) in buffered 50% glycerol was obtained from Boehringer. Vamin (Vitrum) is a 7% solution of 18 amino acids in 10% glucose and is used as a fluid for intravenous nutrition. Digestion mixture: The digestion mixture contained per litre 200 ml H2S04 (96%), 200 g K,S04 and 7 g HgO. This is a clear and easily dispensable solution at temperatures above 35°C. The effective composition of the mixture, after evaporation of the water, is per ml of H2S04: 1 g KzS04 and 32.4 mg Hg** as recommended by Fleck and Munro [2]. The enzymatic determination of ammonia was carried out either with the Boehringer coenzyme-buffer mixture, Boehringer Monotest (Art. No. 125857) consisting of a coenzyme-buffer vial and a glutamate dehydrogenase vial, or with a homemade coenzyme-buffer solution. The Boehringer coenzyme-buffer solution contains triethanolamine 130 mmol/l, a-oxoglutarate 13 mmol/l, ADP 1.3 mmol/l and NADPH 0.08 mmol/l. The home-made solution consists of triethanolamine 150 mml/l, a-oxoglutarate 15 mmol/l, ADP 1.5 mmol/l, NADH 0.1 mmol/l and EDTA 5 mmol/l. pH of both solutions is 8.6. Results
A major problem of direct ammonia estimation of the Hg2+ concentration on the determination
in the digest is the influence of it [2,3]. In Fig. 1 the
131
0
100
200
Hg”~mol/l
300
in sample
t
D I
Fig. 1. Influence of mercuri-ions on the recovery of ammonia in the enzymic reaction. The experiment is described in the text: 80 Ctmol/l Hg*+ in the cuvette corresponds with 1 ml digestion mixture in the testtube. All points are the means of duplicate determinations. 0, Boehringer coenzyme-buffer solution; A, home-made coenzyme-buffer solution; *, home-made coenzyme-buffer solution minus EDTA.
recovery of the ammonia in the enzymic reaction is given as function of the Hg*+ concentration in the cuvette. For this experiment 50 ~1 of a 150 mmol/l ammonium sulphate solution was mixed with various amounts of digestion mixture (up to 5 ml) and diluted with water to 20 ml. 50 ~1 of these dilutions were enzymatically analyzed as described in Methods. The reaction with the Boehringer-reagent, using NADPH, is inhibited less than the reaction with the home-made coenzyme-buffer solution, using NADH. However addition of EDTA, final concentration 5 mmol/l, easily overcomes the inhibition of the enzymic reaction. When Hg*+ was left out of the digestion mixture no inhibition of the reaction is found with or without EDTA in the coenzyme-buffer solution. The experiments described were performed with the home-made coenzyme-buffer solution because this is less espensive than the Boehringer Monotest. In- Fig. 2 the influence of the digestion temperature and time is demonstrated for urea (400 mmol/l), urine, serum albumin (70 g/l), Autonorm standard serum (Nyegaard and Co. AS. Oslo), Vamin and tryptophan 245 mmol/l in 0.1 mol/l NaOH. The tryptophan solution was included in the experiments because this amino acid is notoriously resistant to destruction by sulphuric acid [51. From Fig. 2 it is clear that at temperatures lower than 300°C not all natural constituents of urine and serum can be converted to ammonia. The apparent 100% conversion to ammonia of the nitrogen in the urine at 290°C is a result of the relatively low content of protein and aminoacids of this particular urine. One hour of heating at 320°C converted all nitrogen-compounds in urine,
132
1
Fig. 2. Digestion efficiency as function of time and temperature. Measurements were performed as described in Methods, except for the differences in time and temperature indicated. Concentrations: urea 400 mmol/l,serumalbumin 70 g/l and tryptophan 245 mmol/l. Bottom 2OO’C; middle 260°C; top 320°C. t = 0 is the moment when the temperature in the test-tube reached the indicated temperature. Each bar represents the mean of three independent experiments; all determinations carried out in duplicate.
serum and Vamin to ammonia. The procedure described in Methods is the resultant of these experiments. In Fig. 3 a representative standard curve is drawn for urea, serum albumin and ammonium sulphate standard solutions. A linear relationship was between the A 340nm and the total nitrogen-content for all these solutions. In the method described the total nitrogen-content in the 50-~1 sample
133
OL
0
10
5
N
g/l
Fig. 3. Standard curve for nitrogen determination. 0, urea 100. 200. 400 mmol/l; 4, serum albumin 30. 50. 70 g/l and 0, ammonium sulphate 150. 250 mmol/l were treated as described in Methods for the proposed method. The AA340 ls measured ln a LKB 2074 calculating absorptiometer.
hl Fig. 4. Comparison of the Kjeldahl method in various determinations were carried urine faecalsuspensions and
TABLE
ultra-micro-semisnsymatic total nitrogen determination with the macrobiological fluids. The Kjeldahl and ultra-micro-semi-enzymatic total nitrogen out as given ln Methods. Determinations of total nitrogen content in serum, fluids from drains are compiled in this figure.
I
PRECISION
OF THE DESCRIBED
METHOD
Coefficient Serum (9.05 g N/I) Urine (5.43 g N/l)
of variation
2.9% (n = 20) 3.4% (n = 201
(day-to-day)
134
amounts to 14-560 pg N. However in the analysis only a very small part (25 X 10m4) of it is used. Other combinations of sample-volume and dilution of the digest are possible, however, but volumes of less than 50 ~1 increase the experimental error of the method. For normal urines and sera this proved to be the practical procedure. The day-to-day variation in the analysis is given in Table I for urine and serum, using ammonium sulphate and urea as standards. The results of the method described were in excellent agreement with the Kjeldahl procedure as shown in Fig. 4. Discussion The Kjaldahl procedure is very time-consuming and the analytical steps involved make the method susceptible to errors, especially when the procedure is carried out precisely. Many investigators have reduced the number of analytical steps by introducing direct chemical determinations of the ammonia formed in the digest. However, presence of mercuri-ions or turbidity of the digest interfere with the chemical colour reactions [2,3]. The proposed enzymatic determination in the digest is not inhibited by either mercuri ions, as shown in Fig. 1, or turbidity, since this does not influence the differential absorbance measurement. In the method described the total number of critical analytical steps is reduced and can be carried out by simple devices like Microsanz pipettes (Beckman Instruments), Eppendorf pipettes and automatic dispensers, since all manipulations are performed identically for sample, blank and standard solutions. Digestion for 1 h at 320°C is sufficient to convert all important constituents of human excreta to ammonia as is demonstrated in Fig. 2. This is in accordance with the temperature used by Dambacher et al. [3]. Meijers and Rutten [6] used a temperature of 200°C for the perchloric acid digestion of urine. Although in normal urine the bulk of the nitrogen is in the form of urea, which is readily converted by the sulphuric acid digestion mixture to ammonia at 200°C protein and substantial amounts of amino acids may be present. A digestion temperature of 200°C is, therefore, too low for the estimation of the total nitrogen-content in urine. It is doubtful if the digestion temperature of 200°C for the perchloric acid, used by Meijers and Rutten [6], is high enough to convert protein and amino acids to ammonia. The precision and accuracy of the method described is good, and is comparable to the macro-Kjeldahl method, as is shown in Table I and Fig. 4. The sensitivity and specificity of the enzymic determination makes it possible to detect minute amounts of nitrogen. In the procedure described, 50 ~1 of sample is used, but this can be scaled down to a few microliters when less than 20 ml water is used for the dilution of the digest. However, using a larger proportion (more than 10 times) of the digest in the enzymic reaction, measures should be taken to overcome the effects of the increased acid and mercuri-ion concentrations in the cuvette. This can be effected either by increasing the buffer capacity and EDTA concentration of the coenzyme-buffer solution or diluting the digest with alkaline EDTA instead of water.
135
The detection limit for the enzymic reaction is about 2.5 nmol ammonia or 35 ng N in the cuvette for a 1 ml reaction volume. Theoretically, this is the amount of nitrogen in the sample which can be detected when the appropriate measures are taken. The proposed method for total nitrogen determination in biological fluids enables clinical laboratories to perform large numbers of determinations with high sensitivity, precision and accuracy using standard laboratory equipment. Acknowledgements The author wishes to thank Mr. H. Sauerwein M.D. for sending so many samples for total nitrogen analysis that it urged us to develop the described method, Mrs. B. Schoemaker and Mr. E. Timmer for expert technical assistance and Mrs. R. van Nieuwenhoven for typing the manuscript. References 1 2 3 4
Kjeldahl. J. (1883) Z. Anal. Chem. 22.366-382 Fleck, A. and Munro, H.N. (1965) Clin. Chim. Acta 11.2-12 Dambacher. M., Gubler, A. and Haas, H.G. (1968) Clin. Chem. 14.615-622 De Vries, L.A. (1955) in Klinische Diagnostiek (Garter, E. and De Graaff. W.C., eds.), PP. 434-437, H.E. Stenfert Kroesse, Leiden 5 Lake, G.R., McCutchan, P.. van Meter, R. and Neel. J.C. (1951) Anal. Chem. 23.1634-1638 6 Meijers. C.A.M. and Rutten. J.C.J.M. (1969) Clin. Chim. Acta 24,308-310