ANALYTICAL
75, 447-453 (1976)
BIOCHEMISTRY
Automated WESLEY Department
Method for the Quantitation of Orthophosphate L. TERASAKI
AND GARY BROOKER
of Pharmacology. University of Virginia Charlottesville, Virginia 22903
Medical
School,
Received August 7, 1975; accepted April 28. 1976 A convenient automated method for measuring inorganic phosphate based on the malachite green reaction with a phosphomolybdate complex has been developed. Less than 100 pmol of inorganic phosphate can be readily quantitated by the method which utilizes standard AutoAnalyzer equipment. Inorganic phosphate is measured in sample volumes of less than 0.1 ml and without interference by a number of phosphorylated metabolic intermediates or nucleotides. This methodology is especially useful in the analysis of hydrolytic processes involving phosphorylated substrates.
Phosphorus metabolism plays a rather important role in the many biological reactions which occur in nature. The hydrolysis of phosphorylated compounds has been classically determined by monitoring the inorganic phosphate released by the hydrolytic reaction. A plethora of methods for the determination of inorganic phosphate has been published (l-6). The most widely used procedures involve some modification of methodology capable of detecting a phosphomolybdate complex. In general, these methods are not always extremely sensitive and suffer from variability of the reagent blank. Also, the presence in the sample of substances such as ATP can interfere with color development (7). The variable hydrolysis of labile phosphorylated intermediates poses another major problem in most available methodology. We now present a simple automated system for the analysis of inorganic phosphate at subnanomole levels. This method had been continuously operational in our laboratory for the last several years and is reliable on a day to day basis. MATERIALS
AND METHODS
Ammonium molybdate, (NH,),Mo,O, .4H20, was obtained from Fisher Scientific Company. Malachite green hydrochloride was obtained from J. T. Baker Chemical Company or Matheson, Coleman and Bell. Other reagents were ACS analytical grade or better. The equipment for automation was assembled with standard Technicon AutoAnalyzer components including a proportioning pump (I), and 40-cup sampler. The 447 Copyright Q 1976 by Academic Press. Inc. All rights of reproduction in any form reserved.
448
TERASAKI
AND BROOKER
FIG. 1. Flow diagram for automated phosphate system. All tubes, fittings, and equipment are commercially available items. The tubing sizes illustrated are identified by their approximate flow rates and color codes.
ffow was directed to a Hellma No. 197-0s debubbler flow-cell with a IO-mm path length placed in a Sargent-Welch Model SD spectrophotometer. Tracings were made on a HeathSchIumberger SR-255-B strip chart recorder. The analog output of the spectrophotometer was connected to the recorder through a time constant RC circuit consisting of a 200,000-ohm resistor in series with the output and a IO PF capacitor in paraIle1 connected to ground. The phosphate color reagent was prepared by adding 300 ml of 0.3% malachite green to 120 ml of 3.2% ammonium molybdate in 4.2 N HCI. The solution was then filtered through Whatman No. 1 filter paper and 40 drops of 30% Brij-35 added. Lastly, KH,PO, was added to bring the reagent to a final concentration of 0.5 PM. This solution was used for analyses within a few days. After several days it became difficult to adjust the background absorbance to zero. Before running the phosphate color reagent into the AutoAnalyzer, the appropriate line was washed through with a small volume of 1 M HCI (containing Brij). Also, after determinations were completed, this line was again flushed with HCI solution then with water. This procedure resulted in cleaner lines and coils (less dye deposition) than washing with water only. RESULTS
The basic flow scheme is diagrammed in Fig. 1. Phosphate standards and unknown samples were drawn from sampler cups and added to a bub-
AUTOMATED
PHOSPHATE
449
ASSAY
0.9
0.6
E c
0.7
8
0.6
h s 8
o,s
3
04
Q c? y
0.3
z p”
0.2
/
0.1
d
0
I
20
40 PHOSPHATE
1
1
60 80 ((rM 1
100
FIG. 2. Standard curve. Samples were known concentrations of KHpPO, in water. Each point is the average of duplicate samples with the range less than ? 1% in all cases. A 1 PM solution corresponds to a total amount of 70 pmol of inorganic phosphate or 2.2 ng of phosphate phosphorus.
bled stream of phosphate color reagent. After a brief incubation (3.5 min transit time) the concentration of phosphate in the sample was obtained by reading the peak height absorbance at 660 nm. Standard curves for phosphate solutions in the concentration range of 2 to 80 PM are virtually linear (Fig. 2). At lower concentrations, the curves are noticeably nonlinear but usable provided the unknown samples fall within the range covered by the standards. This lack of linearity was also experienced when we performed the manual method of Itaya and Ui (8) and is therefore not a consequence of the automation design. Less concentrated phosphate color reagent resulted in an exaggeration of this nonlinearity. Samples containing as little as 0.5 PM inorganic phosphate can be routinely assayed (Fig. 2). Since the size of the sample draw is about 70 ~1, this concentration corresponds to a total amount of 35 pmol of phosphate. A typical AutoAnalyzer tracing at high sensitivity is shown in Fig. 3. Full scale absorbance in this case is 0.05 OD unit. Baseline noise is minimal (less than 0.0001 OD unit) and all peaks are smooth and easily read by eye. There is a slight positive baseline drift of about 0.004 OD unit/hr. Repetitive draws of a single sample show good reproducibility with little peak overlap. A typical peak comes to about 80% of its plateau.
450
TERASAKI
AND BROOKER
FIG. 3. Actual strip chart record of automated phosphate method at high sensitivity. The full scale spectrophotometer reading was adjusted to 0.05 OD unit. Samples were 1, 1.5, 2, 3,4 pM KH2POI; four repetitive samples containing 2 PM KH,PO,; and a single prolonged draw of 2 PM KH2POI.
Itaya and Ui have stated that Tween-20 retards color development in the manual malachite greemmolybdate assay procedure, and therefore they added this detergent subsequent to color development. However, we have found that 0.04% Brij-35 in the phosphate color reagent results in the same absorbance whether added before or after the sample is mixed with the reagent. The presence of the detergent is necessary to insure adequate flow characteristics of the solution through the pump tubes and to prevent the spontaneous precipitation of the dye-acid molybdate mixture. With continuous use, the present automated system develops a gradual deposition of fine precipitate on the tubing and coil walls in the immediate area of sample-reagent mixing. In order to eliminate this deposition, Van Belle (6) used methyl green, a dye which does not show this phenomenon, but which is less sensitive than malachite green. We found that the gradual coloration caused by malachite green has no detrimental effect on the assay and, in fact, noticeably enhanced the stability of the recorder baseline. A single platter assembly has been used continuously for well over a year without any apparent need for cleaning or replacement of parts due to dye deposition. In addition, the inclusion of a small amount of KH,PO, in the reagent itself also yields a smoother baseline.
AUTOMATED
PHOSPHATE
ASSAY
4.51
To be of general use in the analyses of inorganic phosphate, it is important that the assay method be specific and free from interference by substances possibly present in sample solutions. A number of standard acids, buffers, salts and phosphorylated compounds of biological interest has been studied to determine the extent of their cross reactivity with the assay (Table 1). In general, there is little interference in the assay by most of the substances tested, and the use of appropriate blanks and controls can easily alleviate most problems encountered. There are three main classes of substances that have noticeable effects on the assay: (i) substances that strongly influence pH, such as fairly high concentrations of acid, buffer, or base; (ii) detergents, which probably act by inhibiting dye-phosphomolybdate complexing; and (iii) viscous solutions, such as 1.2 M sucrose and 60% saturated ammonium sulfate, which do not flow or mix properly through the AutoAnalyzer tubes. Common phosphorylated compounds show no effect on the ability of the assay to measure inorganic phosphate (Table 1) but showed small, variable degrees of cross reactivity (blank value). We presume that the blank seen here is due to the hydrolysis of labile phosphate compounds as a result of the brief (3.5 min at ambient temp) exposure to the acidic phosphate color reagent. Since we took no precautions to purify the phosphorylated compounds prior to subjecting them to analysis, we do not know to what extent the samples originally contained inorganic phosphate as a contaminant. None of the phosphorylated nucleotides or sugars were nearly as effective as phosphate in eliciting color, and pyrophosphate was only about l/ 1000th as effective. DISCUSSION
There are several advantages to the automation of the malachite green-acid molybdate method for measuring inorganic phosphate. Phosphate unknowns and standards can be rapidly analyzed in continuous fashion with the convenience of unattended operation. Sample handling and incubation time from one determination to the next are entirely uniform. This constancy avoids deviations due to pipetting, time-dependent acid hydrolysis of phosphorylated compounds in the sample, and changes in the reagent blank with time. Automated processing also yields a baseline stability (consistency of blank) sufficient to perform analyses at high spectrophotometer sensitivities unapproachable by manual techniques. As a result, the present procedure easily gives precise measurements of subnanomole amounts of inorganic phosphate. To date, we have utilized this method for the quantitation of sodium-potassium ATPase, myosin ATPase, cyclic nucleotide phosphodiesterase, hydrolyzed samples of many phosphorylated intermediates of metabolism, and phosphate efflux of perfused frog ventricles.
TABLE
1
EFFECTOFPOSSIBLE INTERFERING SUBSTANCESON BLANKOFTHEAUTOMATEDPHOSPHATEANALYSW
Interfering substance Water Hydrochloric acid Hydrochloric acid Sulfuric acid Sulfuric acid Perchloric acid Acetic acid Trichloroacetic acid Imidazole-HCl, pH 7.5 Tris-HCI, pH 8.0 Sodium succinate. pH 4.7 Sodium acetate, pH 4.0 Potassium chloride Ammonium sulfate Ammonium sulfate Sodium fluoride Sodium fluoride Calcium chloride Magnesium chloride Disodium EDTA Zinc chloride Manganese chloride Cupric sulfate
Concentration 1.0
M
0.1
M
1.0
M
0.1
M
0.3 M 0.1 M
5% 0.1
M
0.1
M
0.1
M
0.1 M 0.1 M
60% saturationb 10% saturation 0.1 M 10 rnM 10 rnM 10 rnM 10 rnM 0.5
rnM
1 rnM
1 rnM
Tween-20 Sodium dodecyl sulfate Sucrose 2-Mercaptoethanol Bovine serum albumin
1% 1% 1.2 Mb 0.1% 500 pg/ml
ATP ADP AMP GTP Glucose 6-phosphate Fructose 6-phosphate Creatine phosphate Sodium pyrophosphate Sodium pyrophosphate
10 PM 10 /.&M
THE SENSITIVITY
Blank absorbance
10 /.LM 10 /AM 10 /LM
10 rnM 10 /.bM
Decrease in assay sensitivity (%)
0 -0.013
75, 447-453 (1976)
BIOCHEMISTRY
Automated WESLEY Department
Method for the Quantitation of Orthophosphate L. TERASAKI
AND GARY BROOKER
of Pharmacology. University of Virginia Charlottesville, Virginia 22903
Medical
School,
Received August 7, 1975; accepted April 28. 1976 A convenient automated method for measuring inorganic phosphate based on the malachite green reaction with a phosphomolybdate complex has been developed. Less than 100 pmol of inorganic phosphate can be readily quantitated by the method which utilizes standard AutoAnalyzer equipment. Inorganic phosphate is measured in sample volumes of less than 0.1 ml and without interference by a number of phosphorylated metabolic intermediates or nucleotides. This methodology is especially useful in the analysis of hydrolytic processes involving phosphorylated substrates.
Phosphorus metabolism plays a rather important role in the many biological reactions which occur in nature. The hydrolysis of phosphorylated compounds has been classically determined by monitoring the inorganic phosphate released by the hydrolytic reaction. A plethora of methods for the determination of inorganic phosphate has been published (l-6). The most widely used procedures involve some modification of methodology capable of detecting a phosphomolybdate complex. In general, these methods are not always extremely sensitive and suffer from variability of the reagent blank. Also, the presence in the sample of substances such as ATP can interfere with color development (7). The variable hydrolysis of labile phosphorylated intermediates poses another major problem in most available methodology. We now present a simple automated system for the analysis of inorganic phosphate at subnanomole levels. This method had been continuously operational in our laboratory for the last several years and is reliable on a day to day basis. MATERIALS
AND METHODS
Ammonium molybdate, (NH,),Mo,O, .4H20, was obtained from Fisher Scientific Company. Malachite green hydrochloride was obtained from J. T. Baker Chemical Company or Matheson, Coleman and Bell. Other reagents were ACS analytical grade or better. The equipment for automation was assembled with standard Technicon AutoAnalyzer components including a proportioning pump (I), and 40-cup sampler. The 447 Copyright Q 1976 by Academic Press. Inc. All rights of reproduction in any form reserved.
448
TERASAKI
AND BROOKER
FIG. 1. Flow diagram for automated phosphate system. All tubes, fittings, and equipment are commercially available items. The tubing sizes illustrated are identified by their approximate flow rates and color codes.
ffow was directed to a Hellma No. 197-0s debubbler flow-cell with a IO-mm path length placed in a Sargent-Welch Model SD spectrophotometer. Tracings were made on a HeathSchIumberger SR-255-B strip chart recorder. The analog output of the spectrophotometer was connected to the recorder through a time constant RC circuit consisting of a 200,000-ohm resistor in series with the output and a IO PF capacitor in paraIle1 connected to ground. The phosphate color reagent was prepared by adding 300 ml of 0.3% malachite green to 120 ml of 3.2% ammonium molybdate in 4.2 N HCI. The solution was then filtered through Whatman No. 1 filter paper and 40 drops of 30% Brij-35 added. Lastly, KH,PO, was added to bring the reagent to a final concentration of 0.5 PM. This solution was used for analyses within a few days. After several days it became difficult to adjust the background absorbance to zero. Before running the phosphate color reagent into the AutoAnalyzer, the appropriate line was washed through with a small volume of 1 M HCI (containing Brij). Also, after determinations were completed, this line was again flushed with HCI solution then with water. This procedure resulted in cleaner lines and coils (less dye deposition) than washing with water only. RESULTS
The basic flow scheme is diagrammed in Fig. 1. Phosphate standards and unknown samples were drawn from sampler cups and added to a bub-
AUTOMATED
PHOSPHATE
449
ASSAY
0.9
0.6
E c
0.7
8
0.6
h s 8
o,s
3
04
Q c? y
0.3
z p”
0.2
/
0.1
d
0
I
20
40 PHOSPHATE
1
1
60 80 ((rM 1
100
FIG. 2. Standard curve. Samples were known concentrations of KHpPO, in water. Each point is the average of duplicate samples with the range less than ? 1% in all cases. A 1 PM solution corresponds to a total amount of 70 pmol of inorganic phosphate or 2.2 ng of phosphate phosphorus.
bled stream of phosphate color reagent. After a brief incubation (3.5 min transit time) the concentration of phosphate in the sample was obtained by reading the peak height absorbance at 660 nm. Standard curves for phosphate solutions in the concentration range of 2 to 80 PM are virtually linear (Fig. 2). At lower concentrations, the curves are noticeably nonlinear but usable provided the unknown samples fall within the range covered by the standards. This lack of linearity was also experienced when we performed the manual method of Itaya and Ui (8) and is therefore not a consequence of the automation design. Less concentrated phosphate color reagent resulted in an exaggeration of this nonlinearity. Samples containing as little as 0.5 PM inorganic phosphate can be routinely assayed (Fig. 2). Since the size of the sample draw is about 70 ~1, this concentration corresponds to a total amount of 35 pmol of phosphate. A typical AutoAnalyzer tracing at high sensitivity is shown in Fig. 3. Full scale absorbance in this case is 0.05 OD unit. Baseline noise is minimal (less than 0.0001 OD unit) and all peaks are smooth and easily read by eye. There is a slight positive baseline drift of about 0.004 OD unit/hr. Repetitive draws of a single sample show good reproducibility with little peak overlap. A typical peak comes to about 80% of its plateau.
450
TERASAKI
AND BROOKER
FIG. 3. Actual strip chart record of automated phosphate method at high sensitivity. The full scale spectrophotometer reading was adjusted to 0.05 OD unit. Samples were 1, 1.5, 2, 3,4 pM KH2POI; four repetitive samples containing 2 PM KH,PO,; and a single prolonged draw of 2 PM KH2POI.
Itaya and Ui have stated that Tween-20 retards color development in the manual malachite greemmolybdate assay procedure, and therefore they added this detergent subsequent to color development. However, we have found that 0.04% Brij-35 in the phosphate color reagent results in the same absorbance whether added before or after the sample is mixed with the reagent. The presence of the detergent is necessary to insure adequate flow characteristics of the solution through the pump tubes and to prevent the spontaneous precipitation of the dye-acid molybdate mixture. With continuous use, the present automated system develops a gradual deposition of fine precipitate on the tubing and coil walls in the immediate area of sample-reagent mixing. In order to eliminate this deposition, Van Belle (6) used methyl green, a dye which does not show this phenomenon, but which is less sensitive than malachite green. We found that the gradual coloration caused by malachite green has no detrimental effect on the assay and, in fact, noticeably enhanced the stability of the recorder baseline. A single platter assembly has been used continuously for well over a year without any apparent need for cleaning or replacement of parts due to dye deposition. In addition, the inclusion of a small amount of KH,PO, in the reagent itself also yields a smoother baseline.
AUTOMATED
PHOSPHATE
ASSAY
4.51
To be of general use in the analyses of inorganic phosphate, it is important that the assay method be specific and free from interference by substances possibly present in sample solutions. A number of standard acids, buffers, salts and phosphorylated compounds of biological interest has been studied to determine the extent of their cross reactivity with the assay (Table 1). In general, there is little interference in the assay by most of the substances tested, and the use of appropriate blanks and controls can easily alleviate most problems encountered. There are three main classes of substances that have noticeable effects on the assay: (i) substances that strongly influence pH, such as fairly high concentrations of acid, buffer, or base; (ii) detergents, which probably act by inhibiting dye-phosphomolybdate complexing; and (iii) viscous solutions, such as 1.2 M sucrose and 60% saturated ammonium sulfate, which do not flow or mix properly through the AutoAnalyzer tubes. Common phosphorylated compounds show no effect on the ability of the assay to measure inorganic phosphate (Table 1) but showed small, variable degrees of cross reactivity (blank value). We presume that the blank seen here is due to the hydrolysis of labile phosphate compounds as a result of the brief (3.5 min at ambient temp) exposure to the acidic phosphate color reagent. Since we took no precautions to purify the phosphorylated compounds prior to subjecting them to analysis, we do not know to what extent the samples originally contained inorganic phosphate as a contaminant. None of the phosphorylated nucleotides or sugars were nearly as effective as phosphate in eliciting color, and pyrophosphate was only about l/ 1000th as effective. DISCUSSION
There are several advantages to the automation of the malachite green-acid molybdate method for measuring inorganic phosphate. Phosphate unknowns and standards can be rapidly analyzed in continuous fashion with the convenience of unattended operation. Sample handling and incubation time from one determination to the next are entirely uniform. This constancy avoids deviations due to pipetting, time-dependent acid hydrolysis of phosphorylated compounds in the sample, and changes in the reagent blank with time. Automated processing also yields a baseline stability (consistency of blank) sufficient to perform analyses at high spectrophotometer sensitivities unapproachable by manual techniques. As a result, the present procedure easily gives precise measurements of subnanomole amounts of inorganic phosphate. To date, we have utilized this method for the quantitation of sodium-potassium ATPase, myosin ATPase, cyclic nucleotide phosphodiesterase, hydrolyzed samples of many phosphorylated intermediates of metabolism, and phosphate efflux of perfused frog ventricles.
TABLE
1
EFFECTOFPOSSIBLE INTERFERING SUBSTANCESON BLANKOFTHEAUTOMATEDPHOSPHATEANALYSW
Interfering substance Water Hydrochloric acid Hydrochloric acid Sulfuric acid Sulfuric acid Perchloric acid Acetic acid Trichloroacetic acid Imidazole-HCl, pH 7.5 Tris-HCI, pH 8.0 Sodium succinate. pH 4.7 Sodium acetate, pH 4.0 Potassium chloride Ammonium sulfate Ammonium sulfate Sodium fluoride Sodium fluoride Calcium chloride Magnesium chloride Disodium EDTA Zinc chloride Manganese chloride Cupric sulfate
Concentration 1.0
M
0.1
M
1.0
M
0.1
M
0.3 M 0.1 M
5% 0.1
M
0.1
M
0.1
M
0.1 M 0.1 M
60% saturationb 10% saturation 0.1 M 10 rnM 10 rnM 10 rnM 10 rnM 0.5
rnM
1 rnM
1 rnM
Tween-20 Sodium dodecyl sulfate Sucrose 2-Mercaptoethanol Bovine serum albumin
1% 1% 1.2 Mb 0.1% 500 pg/ml
ATP ADP AMP GTP Glucose 6-phosphate Fructose 6-phosphate Creatine phosphate Sodium pyrophosphate Sodium pyrophosphate
10 PM 10 /.&M
THE SENSITIVITY
Blank absorbance
10 /.LM 10 /AM 10 /LM
10 rnM 10 /.bM
Decrease in assay sensitivity (%)
0 -0.013
