determination of inorganic phosphate by flow injection method with immobilized enzymes and...

ANALYTICALBIOCHEMISTRY 182,366-370 (1989)

Determination of Inorganic Phosphate by Flow Injection Method with Immobilized Enzymes and Chemiluminescence Detection

Hideki Kawasaki, Katsumi Sato, Jyunko Ogawa, Yukio Hasegawa, and Hidetaka Yuki Department of Clinical Chemistry, School of Pharmaceutical Sciences, Toho University, M&ama 2-2-i, Funabashi, Chibd 274, Japan

Received April 24,1989

A flow injection method with chemiluminescence detection has been developed for the enzymatic determination of inorganic phosphate. Purine nucleoside phosphorylase, xanthine oxidase, and urate oxidase were immobilized on controlled-pore glass beads. Hydrogen peroxide released by the enzymatic reactions of phosphate and inosine in carrier buffer was detected by the luminol-microperoxidase system in a flow cell. The calibration graph was linear over the range of 5 to 250 pmol, and reproducibility was 1.75% at 10 pmol. The detection limit was 500 fmol of phosphate in 50 ~1 of sample injected. The phosphate content in deoxyribo- nucleic acid was measured by this method. o lsss A~~-

demic Press, Inc.

Many phosphate-containing substances, such as nucleic acids, lipids, and others, exist in our body, and they are involved in many important metabolic reactions. The inorganic phosphate level in body fluids pro- vides significant information in the clinical diagnosis of hyperparathyroidism (l), vitamin D deficiency (2), Fan- coni syndrome (3), and so on. In addition, the determination of phosphoric acid derivatives has become important in environmental hygiene today (4).

In most cases, determination of inorganic phosphate in biological materials is achieved by the widely used Fiske-Subbarow method (5) or its modifications based on calorimetry of a phosphomolybdate complex. These chemical methods have the disadvantages of being carried out under strong acid conditions, poor sensitivity, and laborious operations. Therefore, several enzymatic methods have been introduced in clinical analysis to overcome these limitations (6-8). However, still higher sensitivity is required in order to apply this method to biochemical research. In these enzymatic methods for the determination of inorganic phosphate, several reaction products can act as substrates for the chemilumi-

366

nescence detection system, thereby offering an ultrahigh sensitivity for chemical detection (9).

At the beginning of this investigation, we coupled chemiluminescence detection with the enzymatic method by the sole use of chemiluminescence reagents in place of calorimetric reagents in a tube reaction system, but this method did not offer a sensitive detection because of the high background level due to the unknown con- tamination of the reagents used, especially commercial purine nucleoside phosphorylase (PNP)l preparations. Actually, such a chemiluminescence method has not been reported for the determination of inorganic phosphate in biochemical analysis.

Lukovskaya and Bilochenko (10,ll) reported a chemiluminescence determination of phosphorus based on the oxidation of luminol by the molybudovanadophosphoric acid complex and detection by a spectrophotometric or a photographic method. However, no application on a biological specimen has been reported. Here the investigation was initiated to establish a sensitive and practical chemiluminescence method for the determination of inorganic phosphate by the use of enzymatic reactions. In addition, as the immobilization of enzymes reduces in- terference of unknown substances in enzyme preparations, the flow injection system equipped with the column of immobilized enzymes was used in this study. For sensitive detection by chemiluminescence, the flow injection system must have a stabilized pulseless delivery system, and this exact operating condition is attained by the use of a high-pressure pulseless micropump. Thus, the system should be operated under high-pressure conditions. Therefore, controlled-pore glass beads, which have good mechanical durability, were employed as sup- port media for the immobilization of the enzymes.

1 Abbreviations used: PNP, purine nucleoside phosphorylase; XOD, xanthine oxidase; UOD, urate oxidase; mPOD, microperoxidase; aminopropyl-CPG, aminopropyl-controlled-pore glass; FIA, flow injection analysis; CV, coefficient of variation.

0003~2697/89 $3.00 Copyright 0 1989 by Academic Press, Inc.

All rights of reproduction in any form reserved.

ENZYMATIC DETERMINATION OF INORGANIC PHOSPHATE 367

chemk;ceff;, yagent -jcv 1 ) ‘r:iF’]

0.8pM mPOD 1 ml/min waste

carbonate buffer(pH 10.5)

FIG. 1. FIA manifold used in the determination of inorganic phosphate samples.

In this investigation, PNP, xanthine oxidase (XOD), and mate oxidase (UOD) were immobilized. As is seen in the sequence of reactions [l]-[4], one molecule of inorganic phosphate and one molecule of inosine reacts to give three molecules of hydrogen peroxide, which are quantified by a luminol and microperoxidase (mPOD) system.

Pi + inosine z hypoxanthine + ribose l-phosphate

ill

Hypoxanthine + 2Hz0 + 202 “2 uric acid + 2H202 [2]

Uric acid + 2Hz0 + O2 “2 allantoin + H202 [3]

HzOz + luminol mPoD - aminophthalate + Nz + light [4]

MATERIALS AND METHODS

Materials

Unless otherwise specified, deionized distilled water was used for the preparation of reagents or buffer solu- tions. Water of fluorometric purity (Dojindo Lab., Ku- mamoto, Japan) was used for the preparation of samples or reagents for incineration of DNA. Hepes, 4-(2-hy- droxyethyl)-1-piperazineethanesulfonic acid was also a product of Dojindo Lab. Luminol and glutaraldehyde (25% in water, for electron microscopy) were obtained from Tokyo Kasei Co. Hydrogen peroxide (31% H,O,) was a product of Mitsubishi Gas Chemicals Co. Tris(hy- droxymethyl)aminomethane was obtained from Na- karai Tesque Co. (Kyoto, Japan). Inosine was purchased from Wako Pure Chemicals (Osaka, Japan). Microper- oxidase (MP-ll), PNP (bacterial, lyophilized powder, 16.7 units/mg), XOD (grade I from buttermilk, suspen- sion in 2.3 M ammonium sulfate, 10.8 units/ml), and DNA (type I, from calf thymus) were obtained from Sigma Chemical Co. (St. Louis, MO). UOD (type II, from Candida sp. 6.22 units/mg) was obtained from To- yobo Biochemicals (Osaka, Japan). All other reagents used were of reagent grade.

Methods

Immobilization of enzymes. Immobilization was carried out according to the method described by Hayashi et al. (12). The enzymes were immobilized by the glutaraldehyde method on aminopropyl-controlled-pore glass beads (aminopropyl-CPG, AMP-500, 120-200 mesh, mean pore diameter 491 nm, Electra Nucleonics, Funa- koshi, Japan). Precisely, the aminopropyl-CPG beads (0.5 g) were mixed with 2 ml of 5% glutaraldehyde in 0.1 M phosphate buffer solution (pH 7.0) and reacted for 1 h by rotating the vial under reduced pressure by the use of an aspirator. The remaining reagent was removed by suction and the beads were then washed with 50 ml of water. Seventy units (4 mg) of PNP was dissolved in 0.1 M phosphate buffer solution (pH 7.0) and mixed with the glutaraldehyde-treated CPG beads. The mixture was then agitated by a rotary shaker for 24 h at 4°C. The reacted CPG beads were then washed with 100 ml of water, 100 ml of 1 M NaCl, and finally 1000 ml of water. In the immobilization of XOD and UOD, 2.7 units (4 mg of protein) of XOD was dialyzed initially against 0.1 M

phosphate buffer solution (10 mM EDTA, pH 7.2) over- night at 4°C to remove the commercial preservations. The XOD was mixed with 25 units of UOD (4 mg of protein) dissolved in 2 ml of 0.1 M phosphate buffer solution (pH 7.0), and the mixture was reacted with the glutaraldehyde-treated CPG beads in a manner similar to that of PNP. The yields of immobilization of the enzymes were estimated from the protein concentration of each reaction supernatant determined by the Lowry method (13) before and after the reaction. When not in use, the immobilized enzymes were stored in a 0.1 M phosphate buffer solution (pH 7.0) at 4°C.

i .cn 8 % d

r I,0 % 0.5

go I I 1 IV Y 0

6.5 7.0 7.5 0.5 6.5

PH

FIG. 2. The pH profiles of immobilized XOD and UOD in 10 mM

Tris buffer (A), 10 mM Hepes buffer (O), 5 mM Hepes buffer (m), and 10 mM carbonate buffer (V). Five picomoles of hypoxanthine in 50 ~1 of buffer solution was injected as a sample (solid symbols), and 50 pl of the same buffer solution was injected as a blank (open symbols). Each point is the average of three runs.

368 KAWASAKI ET AL.

12 3 4 5

Concentration(pmol/5Olrl)

FIG. 3. Calibration curves for standard hypoxanthine from FIA with immobilized XOD and UOD in 10 mM Tris buffer (A), 10 mM

Hepes buffer (O), and 10 mM carbonate buffer (V). The pH of Tris buffer, Hepes buffer, and carbonate buffer were 7.5, 7.5, and 8.5, respectively.

Flow injection apparatus and procedure. The ar- rangement of the flow injection manifolds used in this study is shown in Fig. 1. Polytetrafluoroethylene tubing (0.25mm i.d., Gasukuro Kogyo Co., Tokyo) was used throughout the manifold. The carrier solution, which was 25 PM inosine in 10 mM Hepes-NaOH buffer solution (pH 7.5), was pumped with an Atras NP-DX-2 pump (Niti-On, Tokyo). The flow rate was 1 ml/min. The sample solution, from 10 to 50 ~1 in loading size, was injected through a rotary valve injector (50~~1 sample loop, VMD-350-EIE, Shimamura Keiki, Tokyo) into the carrier stream. The reactor consisted of two glass mini- columns (3-mm id., g-mm o.d., 5-cm length, Omnifit, UK), packed with the immobilized PNP, and coimmobilized XOD and UOD, respectively. The sample solution was passed through the columns in this order. The reagent solution, which was 10 PM luminol and 0.8 PM

mPOD in 50 mM carbonate buffer solution (pH 10.5), was pumped at a flow rate of 1 ml/min. The sample carrier stream from the reactor was merged with the reagent stream in the flow cell unit (spiral-shaped glass cell with a Y-shaped merging port, lOO-~1 internal volume) in the detector (Lumiflow LF-800, Niti-On, To- kyo) and the chemiluminescence emission was detected directly by the photomultiplier unit attached to the cell unit. The signal from the photometer is recorded with a plotter-integrator (Chromatopak C-R3A, Shimadzu, Kyoto) as flow injectogram, peak height, and peak area.

Preparation of inorganic phosphate sample from DNA. Incineration of DNA was carried out according to the method described by Chen et al. (14). Calf thymus DNA (6.5 ng in 0.5 ml of water) was placed in a test tube

(12 X 105 mm), and 0.2 ml of 2.5% ethanol solution of magnesium nitrate was added. The tube was heated to dryness by direct application of a gas flame, and the DNA was incinerated. The residue was dissolved by addition of a small amount of 1 N HCl and subsequent addition of 1 ml of water. Next, the solution was neutralized with 1 N NaOH, and finally the volume was adjusted to 2 ml by addition of 50 I.LM inosine in 20 mM Hepes buffer (pH 7.5). Fifty microliters of this sample solution was calculated to contain 14.5 pmol of inorganic phosphate based on a P content of 9% of calf thymus DNA. The sample was then injected into the flow injection analysis (FIA) system.

RESULTS AND DISCUSSION

Immobilization of Enzymes and Optimization of the Flow Injection System

Various methods for immobilization of enzymes were first examined to establish the reaction system from [l] to [4]. In the case of PNP, although PNP from calf spleen was previously reported to require purification prior to immobilization (12), the bacterial PNP used without purification did not interfere with the chemiluminescence detection in this experiment. Therefore, the purification step was omitted to simplify the procedure. In addition, by the use of commercial CPG beads and the glutaraldehyde method, more than 93% of the yields for immobilization were easily attained throughout this experiment, including the immobilization of XOD and UOD.

Optimization of the reaction conditions of enzymes was first examined at the stage of reactions [2] and [3]. When coimmobilized XOD and UOD were used, the chemiluminescence peak height of the flow injectogram generated from 50 pmol of hypoxanthine was 1.43 times that of immobilized XOD alone. Therefore, XOD and UOD were used in coimmobilized form in the subsequent experiments.

, I I , I I

01 5 10 15 20 25

Concentration (~Ml

FIG. 4. Dependence of chemiluminescence peak area on inosine concentration for an inorganic phosphate standard (250 pmol). FIA conditions are given under Methods.

ENZYMATIC DETERMINATION OF INORGANIC PHOSPHATE 369

ti 0 20 40 60 80 100

Time (mid

FIG. 5. Flow injectograms for the inorganic phosphate standards. Concentrations are indicated at the peaks of each sample. Blank peak is indicated by B. FIA conditions are given under Methods and in Fig. 1.

Until this stage, phosphate buffers were used as the carrier for chemiluminescence detection, but these buffers could not be used for the detection of inorganic phosphate; therefore, buffers suitable for the FIA system were examined using a XOD-UOD column and 5 pmol of hypoxanthine as a substrate. The variations in the peak area of hypoxanthine and the peak height of the blank under varying pH conditions in several buffers are shown in Fig. 2. For Tris buffer, the peak height was maximum but the blank was high around the optimum pH of enzymes, and for carbonate buffer, the sensitivity was low and it was difficult to maintain the optimum pH of enzymes. In contrast, for 10 mM Hepes buffer, the peak height was high and reproducible in the pH range examined and the blank was low around pH 7.5, but 5 mM Hepes buffer gave the lower sensitivity and higher blank value. In addition, the calibration curves from 250

0

/

0 50 100 150 200 250

Concentration (pmoli50pI)

FIG. 6. Calibration curve from the FIA method for inorganic phosphate. Details are given under Results and Discussion.

FIG. 7. Flow injectograms of samples obtained by incineration of DNA. a, b, and c correspond to 36.2 pmol P, 14.5 pmol P, and blank, respectively. Experimental details are given in the text.

fmol to 5 pmol of hypoxanthine were made for each buffer (Fig. 3). For Tris buffer, although the sensitivity was high, a straight calibration line was not obtained, and the sensitivity was low for carbonate buffer (pH 8.5). A linear calibration line (r = 0.9993) was obtained for 10 mM Hepes buffer (pH 7.5), the reproducibility (coefficient of variation, CV) was 2.7% at 5 pmol (n = 4), and the limit of detection (S/N = 2) was 250 fmol. Conse- quently, 10 mM Hepes buffer (pH 7.5) was used for the carrier buffer solution. Other pH conditions did not give a better result.

The entire enzymatic reaction condition was examined by the addition of the PNP column in front of the XOD-UOD column for reaction [l]. As reaction [l] requires inosine as a substrate, the optimum inosine concentration in the carrier buffer was examined by inject- ing 250 pmol of dipotassium hydrogen phosphate as a sample over the concentration range from 1 to 25 PM

(Fig. 4). The peak area of chemiluminescence became constant over 10 PM. Thus the concentration of 25 PM,

which is more than adequate, was chosen as the operating concentration. In some cases, impurities such as hypoxanthine were contained in commercial inosine, and significant increases in baseline and noise level were observed. In such cases, the detection of phosphate at low level was difficult.

Calibration Curve, Reproducibility, and Sensitivity

Under the optimized conditions specified above for the entire enzymatic system, the calibration curve of dipotassium hydrogen phosphate was prepared over the concentration of 5 to 250 pmol. The typical flow injectograms are shown in Fig. 5. The calibration curve (Fig. 6) was linear over the concentration range examined (r = 0.9979), and the least-squares regression equation, Y (X106 pV.s) = -0.132 + 0.041X (pmol), was obtained. The symbol Y represents the area of chemiluminescence peak and the symbol X represents quantities of inorganic phosphate. The CV within-batch at 250, 50, 10, and 5 pmol were 1.03,2.30,1.75, and 9.36%, respectively. Except CV at 5 pmol, good reproducibility was obtained, and the blank peak was scarcely observed under this condition. The limit of detection was 500 fmol (S/N

370 KAWASAKI ET AL.

= 2). These data suggest the capability of the system for the sensitive detection of inorganic phosphate.

Application to DNA Sample

In order to test the applicability of this method to biological samples, the phosphate in DNA was measured. The sample was prepared by incineration according to the method of Chen et al. (14). As described under Meth- ods, calf thymus DNA was incinerated with magnesium nitrate to yield inorganic phosphate. The typical flow injectograms of the samples are shown in Fig. 7. In every sample prepared by incineration, a negative dip was observed in front of the sample peaks. This was probably due to the suppression of chemiluminescence by reagent remaining in the sample, which was excessively high in concentration, but the detection of the peak was not affected. In the case of the blank, a peak that corresponded to an amount of about 7 pmol of inorganic phosphate appeared. This was probably due to the contami- nation from the experimental procedures or reagents used. However, the amount of inorganic phosphate estimated by this method agreed with the theoretical value calculated from the phosphorus content (about 9%) of the calf thymus DNA, and the peak height increased with increasing concentration of DNA in the sample solution. Although only raw data were examined in this experiment, these facts indicated that the system was able to determine the phosphate in DNA samples. These results also suggested the applicability of this system to the analysis of other biological materials. The sensitivity of this method was 160-fold higher than those reported for the conventional spectrophotometric techniques (14). When the immobilized enzymes were stored at 4°C in phosphate buffer, more than 80% of the activity re- mained over a period of 2 months for PNP and 3 months for XOD and UOD.

Results from the Batch Method

As mentioned earlier, the enzymes were used without purification in the preparation of immobilized enzymes, but the enzymes tend to contain other enzymes, such as adenosine, deaminase, or guanase, which convert some nucleic acid metabolites to xanthine or hypoxanthine. Although the conditions employed in this experiment were not affected by these enzymes, samples from biological materials are likely to contain the substrates for these enzymes as well as other interfering factors. These effects could result in an increase in noise and baseline level and, consequently, make sensitive detection difficult. This must be taken into consideration for further studies of this system. We also manually measured the phosphate by performing reactions [l]-[4] in an aqueous solution system in a glass tube, instead of in the flow system with the immobilized enzyme columns. In brief, after additions of PNP, phosphate, and inosine

into a microtube, the mixture was incubated for 10 min and XOD and UOD were added and further incubated for 10 min. The hydrogen peroxide generated was determined as chemiluminescence by the luminol-mPOD system. The reaction mixture without phosphate was used as the blank. The blank itself produced consider- able chemiluminescence, which corresponded to 21% of 1 nmol phosphate. Since the reaction mixture without PNP decreased the chemiluminescence intensity of the blank to one-tenth, the major noisy background may be due to the impurity of PNP preparation. It was consid- ered necessary to purify PNP and other reagents to achieve lower detection limits. These are still under investigation. These results also suggested the effective- ness of the flow system with immobilized enzymes, which minimized the background of chemiluminescence for determination of phosphates.

CONCLUSION

The proposed new FIA system with immobilized enzymes and chemiluminescence detection offers high sensitivity and precision for the determination of inorganic phosphate. Moreover, this method requires no compli- cated technique, can be accomplished in short operating times and needs no expensive instrumentation. Thus this method could be applied widely in biomedical inves- tigations in which inorganic phosphate is a concern. In addition, our preliminary experiment revealed that the reaction of PNP could be amplified by coupling with al- kaline phosphatase, which reproduces inorganic phosphate from ribose l-phosphate formed by PNP. We are now investigating such a system and further applica- tions for biomedical analysis.

REFERENCES

1. Bronsky, D., Kushner, D. S., Dubin, A., and Snapper, I. (1958) Medicine 37,317-352.

2. Singer, M. (1954) J. Bone Jt. Surg. 36B, 633-636.

3. Ogura, Y., Imamura, N., and Tsukui, I. (1982) Japan J. Clin. Med. 40,2680-2688.

4. Adachi, A., and Kobayashi, T. (1986) Eisei Kagaku 32,39-45.

5. Fiske, C. H., and Subbarow, Y. (1925) J. Biol. Chem. 66,375-400.

6. Schulz, D. W., Passonneau, J. V., and Lowry, 0. H. (1967) Anal. Biochem. 19,300-314.

7. Scopes, R. K. (1972) Anal. Biochem. 49,88-94.

8. Hwang, W. I., and Cha, S. (1973) Anal. Biochem. S&379-387.

9. Shevlin, P. B., and Neuteld, H. A. (1970) J. Org. Chem. 35, 2178- 2182.

10. Lukovskaya, M. N., and Bilochenko, V. A. (1977) Ukr. Khim. Zh. 43,756-759.

11. Lukovskaya, M. N., and Bilochenko, V. A. (1974) Zavod. Lab. 40, 936-937.

12. Hayashi, Y., Zaitsu, K., and Ohkura, Y. (1986) Anal. Chim. Acta 186.131-137.

13. Lowry, 0. H., Rosebrough, N. J., Far+, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193,265-275.

14. Chen, P. S., Jr., Toribara, T. Y., and Warner, H. (1956) Anal. Chem. 28.1756-1758.

determination of inorganic phosphate by flow injection method with immobilized enzymes and...

Documents