ANALYTICAL
BIOCHEMISTRY
loI),
339-342 (1979)
A Microassay
for Analysis
D. E. C. COLE, F. MOHYUDDIN,
of Serum
Sulfate
AND C. R. &RIVER
Departments of Biology (Faculty of Science) and Pediatrics (Faculty of Medicine) and Human Genetics Center, McGill University, Montreal, Quebec, Canada Received April 30, 1979 A simple assay for analysis of sulfate in a 1OOql volume of serum is described. Selective precipitation of l=Ba after removal of protein with trichloroacetic acid is reproducible and accurate. In particular, no major interference from calcium or phosphate within the physiological range of concentration is found.
The measurement of inorganic sulfate in serum has been accomplished in the past by selective precipitation of sulfate with barium (1) or be&dine (2). However, these methods require large sample volumes and they are unsatisfactory for clinical investigation. We have developed a reliable method to measure sulfate in a small volume (100 ~1) of serum, utilizing the technique of precipitation with radiolabeled barium (3). MATERIALS
AND METHODS
Materials Chemicals and isotopes. Reagent-grade chemicals were purchased from Fisher Chemical Company. 133BaC12(10 mCi/pmol) and Na, 35S04 (117 mCi/pmol) were purchased from New England Nuclear Corporation, Boston, Massachusetts. Disposable polyethylene tubes (1.5-ml Microtest tubes, Brinkman) and pipet tips (Fisher Co.) were used to transfer all solutions. Serum. Whole blood was obtained at 0900 h from the antecubital vein of healthy adult volunteers and drawn into a tube without anticoagulant; subjects were fasted overnight. Mouse serum was obtained in the morning after normal night feeding from C57BY6J and Swiss albino strains by decapitation and collection of whole blood directly into Microtainer separator tubes 339
for centrifugation (Becton and Dickinson, N. J.). All sera were stored at -20°C until analysis if not used immediately. Methods Principle of assay. The radioisotope precipitation method of sulfate analysis uses 133BaC12 (3). To aqueous solutions containing various amounts of inorganic sulfate standard, a known amount of labeled BaCl, is added and mixed, allowed to stand at 4°C for 180 min, and centrifuged at 6000g for 60 min. The supematant fraction is then counted in a Beckman Biogamma II y-emission spectrometer and the loss of ‘=Ba from the supematant solution determined. Assuming stoichiometric precipitation of BaSO,, a standard curve can be constructed, relating removal of 133Ba to the amount of inorganic sulfate in the sample. Storage of samples. Aliquots of serum samples containing known amounts of added sulfate were stored at -20°C for 1, 7, and 100 days to determine loss of sulfate during storage. Protein precipitation: Comparative study. Aliquots (300 ~1) of five different reagents for deproteinization were added to aliquots (600 ~1) of serum containing tracer amounts of 35S@4- (5 ~1 Na, 35S04, 7.2 pmol, 2.5 x 105 dpm). Recovery of sulfate was monitored in the supematant after the centrifugation step. ooO3-2697/79/180339-04$02.00/O Copyright 0 1979 by Academic Press, Inc. All rights of reproduction in any form reserved.
340
COLE, MOHYUDDIN, TABLE RECOVERY
1
OF %O:FROM SERUM PROTEIN PRECIPITATION
Final concentration (%I
Precipitant
AFTER
Percentage recovery of =so:-= (mean * SD, n = 4)
Ethanol (95%, v/v)
15 (v/v)
55 2 9
Trichloroacetic
acid
I (w/v)
loo + 5
Trichloroacetic acid in methanol”
7 (w/v) 15 (v/v)
98 k 4
Perchloric acid
2 (w/v)
107 2 2
Uranyl acetate
0.3 (w/v)
81 + 4
0.3 (w/v) (107 mM, pH 3.8)
78 +- 3
Uranyl acetate, buffered in acetic acid
a A tracer amount of 35SOi- in the form of 5 ~1 N~?SO, (containing 2.5 x lo5 dpm) was added to 600 ~1 serum. Recovery of sulfate was measured by counting ?S in the supematant after precipitation of proteins. b Added subsequently to lower the solubility of the BaSO, complex.
Eficiency of precipitation step. Two volumes of N&SO, (0.02-0.2 mM) in trichloroacetic acid (20%, w/v) were mixed with 1 vol 133BaC1, (1.25 mM containing 5 x lo3 dpm; final volume 600 ~1). The mixture was kept at 4°C for 3 h and then centrifuged at 4°C for 1 h at 6000g. Efficiency of precipitation was calculated by measuring loss of radioactivity from the supernatant. Serum Sulfate Analysis in Presence of Other Ions Human serum was dialyzed against 0.9% NaCl for 24 h at 4°C. Known amounts of Na$O,, CaC&, MgCl,, sodium phosphate (pH 7.4), or an organic sulfate ester (dehydroepiandrosterone sulfate) were added to aliquots (500 ~1) of dialyzed serum and deproteinized with 0-5 vol of trichloroacetic acid (20%, w/v) in methanol (50%, v/v). Protein-free supematant (100 ~1) was then taken, to which was added 10 ~1 of 5 mM
AND SCRIVER
N&SO, and 15 ~1 of 10 mM 133BaC12. The mixture was left at 4°C for 3 h, then centrifuged as described under Methods, and the supematant counted. RESULTS AND DISCUSSION
Analysis of sulfate in protein-free water solutions by the original method of barium radioisotope precipitation (3) was successful in our hands. We had no dficulty in applying the method to the analysis of urinary inorganic sulfate. However, analysis of blood sulfate in the small volumes compatible with clinical investigation and small animal research was more difficult with the original method. We found that analysis of plasma was generally unsatisfactory-and, in fact, impossible, when heparin was used as the anticoagulant. Accordingly, we examined the various steps in processing serum to determine their effect on sulfate analysis in the protein-free samples. Method of Protein Precipitation The recovery of 35SOi- from serum samples was examined in the presence of five different reagents used for precipitation of serum proteins (Table 1). Ethanol (95%, v/v) gave very poor recoveries of sulfate in the supematant solution. There was a significant loss of sulfate in the presence of uranyl ion, an agent in current use to remove phosphate anion and protein together during sulfate analysis (1,3). Loss of sulfate during uranyl acetate precipitation has also been reported by others (4). Precipitation of proteins with trichloroacetic acid (TCA),’ or TCA in methanol, gave acceptable recoveries of sulfate, and this agent was used in all subsequent studies. Efficiency of 133Ba Precipitation in the Presence of Sulfate and Other Ions in Serum , The reliability of the method depends on stoichiometric removal of lmBa2+ from solu* Abbreviation
used: TCA, trichloroacetic
acid.
MICROASSAY
0
FOR SERUM SULFATE
0.1
0.05
Sulfate
341
ANALYSIS
0.15
Concentration(mM)
FIG. 1. Completeness of *33BaS0, precipitation
at low sulfate concentrations.
tions, to form insoluble isotopic BaSO, in “ethereal” sulfate had no consistent or the presence of SO,‘-. Resolubilization of significant effect on the determination of the precipate constitutes interference in the sulfate in serum. However, the quantity of assay as does significant removal of 133Ba heparin employed for anticoagulant in by anions other than SO;-. microcapillary tubes and receiving tubes in Precipitation of sulfate from a water use in hospitals was sufficient to nullify any solution (0.9% saline) in the presence of effort to measure sulfate accurately in the plasma prepared from such samples. 0.4 mM BaCl, (final concentration) is incomplete at low physiological concentrations of sulfate (co.150 mM) (Fig. 1). Accordingly, Handling of Serum Samples a known amount of exogenous sulfate was We compared the recovery of sulfate from added to the serum samples and the con- freshly prepared serum samples and from centration of BaCl, increased to 1.2 mM. aliquots stored at -20°C for long intervals. Under these conditions, complete precipitaThe recovery was 98 f 5% (mean + SE, tion and the expected stoichiometry were obtained (Fig. 2). Recovery of sulfate from serum was 0.4 measured by first dialyzing sulfate away, ‘: z then reconstituting the sample by adding 0.3 z known amounts of inorganic sulfate, and Q) Q) finally proceeding with TCA precipitation '0 ‘0 of protein and analysis of sulfate by the 0.25 VI coprecipitation method. Recovery of sulfate m from human serum was complete and E accurate even at low concentrations. 0 : The presence of phosphate in physioa logical serum concentrations did not inter0 0.1 0.2 0.3 0.4 fere with sulfate analysis (Table 2); howSulfate Added( mM) ever, at levels of phosphate in serum above the physiological range (l-2 mM), a modest FIG. 2. Recovery of sulfate added to dialyzed human inhibition of sulfate analysis was apparent. serum. Vertical bars indicate SD of fcur independent Divalent cations (Ca*+ and Mg*+) and an experiments, each done in quadruplicate.
2
342
COLE, MOHYUDDIN, TABLE
2
EFFECTOF OTHERIONSON ASSAYOFSULFATE Ion concentration Ion
fm@
mgldl”
Sulfate recovery, % k SD (n)
Sodium phosphate (at pH 7.4)
0.1 0.5 1.0 2.5 5.0
0.3 1.6 3.1 7.8 16
103 101 100 95 90
f + f f f
7 (5) 5 (5) 5 (5) 11 (5) 5 (5)
W+
1.0 2.0
2.3 4.6
Ca*+
1.25 2.5 5.0
5 10 20
loo -e 10 (10) 109? 8(10) 107 f 8 (10)
DHEA-SO,”
0.001 1.0
-
105 * 2 (5) 102 2 4 (4)
107 f 15 (10) 109 * 10 (10)
a Equivalent concentrations of element in common clinical usage (e.g., inorganic phosphate as milligrams phosphorus per deciliter). * Dehydroepiandrosterone sulfate (A5-androstene3/3-ol-17-one-3-sulfate).
n = 5) after 24 h at -2O”C, and 95 & 6% after a week of frozen storage; after 3 months the recovery was 107 & 7%. We observed, in the aforementioned study with dialyzed serum, that repeated freezing and thawing caused the release of measurable amounts of sulfate, presumably from “bound” forms. On two occasions, 0.046 and 0.064 mM of additional sulfate (i.e., about 15% of the normal serum concentration) were recovered from dialyzed samples after four cycles of freeze-thaw. This finding indicates that serum samples should be handled prudently prior to analysis. Sulfate Concentration in Serum Human. The morning fasting serum sulfate concentration in healthy human adults was 0.297 ? 0.038 mM (mean + SD, n = 19). The individual values in the popula-
AND SCRIVER
tion sample were the mean of measurements in quadruplicate. The coefficient of variation calculated from 19 sets of quadruplicates was 7.4%. Miller and colleagues (3), using larger volumes of serum (> 1 ml) and the same isotope coprecipitation assay, showed that serum sulfate in man was 0.323 + 0.085 mM (mean + SD, n = 88); their coefficient of variation was also about 7%. However, consistency in time of sampling and alimentation was not reported in the early study and inconsistency of sampling may account for the slight difference between our “normal” serum sulfate value and the earlier report. Murine. Mouse serum contains more sulfate (> 1 mM) than human serum; the finding is analogous to that of phosphate content in sera of the two species. The sulfate concentration in C57B1/6J mouse serum was 1.00 ? 0.09 mM (mean + SD, n = 13) and in the serum of the Swiss albino mouse, 1.24 ? 0.16 mM (n = 7). The difference between inbred strains is significant (P < 0.001; Student t test), suggesting a genetic basis for the interstrain variation. The findings in human and murine sera offer a basis for further investigation that are now feasible with our microassays. ACKNOWLEDGMENTS This work was supported by the Medical Research Council of Canada and the Quebec Network of Genetic Medicine. D.E.C.C. is a Fellow of the MRC.
REFERENCES 1. Loeb, R. F., and Benedict, E. M. (1927) J. Clin. Invest.
4, 33-48.
2. Letanoff, T. V., and Reinhold, J. G. (1936) J. Biol. Chem. 114, 147-156. 3. Miller, E., Hlad, C. J., Levine, S., Holmes, J. H., and Ehick, H. (1961) J. Clin. Lab. Invest. 58, 656-661. 4. Berglund, F., and Sorbo, B. (1960) Stand. J. Clin. Lab. Invest. 12, 147- 153. 5. Hoffman, W. S., and Cardon, R. (1935) J. Biol. Chem. 109,711-127.
BIOCHEMISTRY
loI),
339-342 (1979)
A Microassay
for Analysis
D. E. C. COLE, F. MOHYUDDIN,
of Serum
Sulfate
AND C. R. &RIVER
Departments of Biology (Faculty of Science) and Pediatrics (Faculty of Medicine) and Human Genetics Center, McGill University, Montreal, Quebec, Canada Received April 30, 1979 A simple assay for analysis of sulfate in a 1OOql volume of serum is described. Selective precipitation of l=Ba after removal of protein with trichloroacetic acid is reproducible and accurate. In particular, no major interference from calcium or phosphate within the physiological range of concentration is found.
The measurement of inorganic sulfate in serum has been accomplished in the past by selective precipitation of sulfate with barium (1) or be&dine (2). However, these methods require large sample volumes and they are unsatisfactory for clinical investigation. We have developed a reliable method to measure sulfate in a small volume (100 ~1) of serum, utilizing the technique of precipitation with radiolabeled barium (3). MATERIALS
AND METHODS
Materials Chemicals and isotopes. Reagent-grade chemicals were purchased from Fisher Chemical Company. 133BaC12(10 mCi/pmol) and Na, 35S04 (117 mCi/pmol) were purchased from New England Nuclear Corporation, Boston, Massachusetts. Disposable polyethylene tubes (1.5-ml Microtest tubes, Brinkman) and pipet tips (Fisher Co.) were used to transfer all solutions. Serum. Whole blood was obtained at 0900 h from the antecubital vein of healthy adult volunteers and drawn into a tube without anticoagulant; subjects were fasted overnight. Mouse serum was obtained in the morning after normal night feeding from C57BY6J and Swiss albino strains by decapitation and collection of whole blood directly into Microtainer separator tubes 339
for centrifugation (Becton and Dickinson, N. J.). All sera were stored at -20°C until analysis if not used immediately. Methods Principle of assay. The radioisotope precipitation method of sulfate analysis uses 133BaC12 (3). To aqueous solutions containing various amounts of inorganic sulfate standard, a known amount of labeled BaCl, is added and mixed, allowed to stand at 4°C for 180 min, and centrifuged at 6000g for 60 min. The supematant fraction is then counted in a Beckman Biogamma II y-emission spectrometer and the loss of ‘=Ba from the supematant solution determined. Assuming stoichiometric precipitation of BaSO,, a standard curve can be constructed, relating removal of 133Ba to the amount of inorganic sulfate in the sample. Storage of samples. Aliquots of serum samples containing known amounts of added sulfate were stored at -20°C for 1, 7, and 100 days to determine loss of sulfate during storage. Protein precipitation: Comparative study. Aliquots (300 ~1) of five different reagents for deproteinization were added to aliquots (600 ~1) of serum containing tracer amounts of 35S@4- (5 ~1 Na, 35S04, 7.2 pmol, 2.5 x 105 dpm). Recovery of sulfate was monitored in the supematant after the centrifugation step. ooO3-2697/79/180339-04$02.00/O Copyright 0 1979 by Academic Press, Inc. All rights of reproduction in any form reserved.
340
COLE, MOHYUDDIN, TABLE RECOVERY
1
OF %O:FROM SERUM PROTEIN PRECIPITATION
Final concentration (%I
Precipitant
AFTER
Percentage recovery of =so:-= (mean * SD, n = 4)
Ethanol (95%, v/v)
15 (v/v)
55 2 9
Trichloroacetic
acid
I (w/v)
loo + 5
Trichloroacetic acid in methanol”
7 (w/v) 15 (v/v)
98 k 4
Perchloric acid
2 (w/v)
107 2 2
Uranyl acetate
0.3 (w/v)
81 + 4
0.3 (w/v) (107 mM, pH 3.8)
78 +- 3
Uranyl acetate, buffered in acetic acid
a A tracer amount of 35SOi- in the form of 5 ~1 N~?SO, (containing 2.5 x lo5 dpm) was added to 600 ~1 serum. Recovery of sulfate was measured by counting ?S in the supematant after precipitation of proteins. b Added subsequently to lower the solubility of the BaSO, complex.
Eficiency of precipitation step. Two volumes of N&SO, (0.02-0.2 mM) in trichloroacetic acid (20%, w/v) were mixed with 1 vol 133BaC1, (1.25 mM containing 5 x lo3 dpm; final volume 600 ~1). The mixture was kept at 4°C for 3 h and then centrifuged at 4°C for 1 h at 6000g. Efficiency of precipitation was calculated by measuring loss of radioactivity from the supernatant. Serum Sulfate Analysis in Presence of Other Ions Human serum was dialyzed against 0.9% NaCl for 24 h at 4°C. Known amounts of Na$O,, CaC&, MgCl,, sodium phosphate (pH 7.4), or an organic sulfate ester (dehydroepiandrosterone sulfate) were added to aliquots (500 ~1) of dialyzed serum and deproteinized with 0-5 vol of trichloroacetic acid (20%, w/v) in methanol (50%, v/v). Protein-free supematant (100 ~1) was then taken, to which was added 10 ~1 of 5 mM
AND SCRIVER
N&SO, and 15 ~1 of 10 mM 133BaC12. The mixture was left at 4°C for 3 h, then centrifuged as described under Methods, and the supematant counted. RESULTS AND DISCUSSION
Analysis of sulfate in protein-free water solutions by the original method of barium radioisotope precipitation (3) was successful in our hands. We had no dficulty in applying the method to the analysis of urinary inorganic sulfate. However, analysis of blood sulfate in the small volumes compatible with clinical investigation and small animal research was more difficult with the original method. We found that analysis of plasma was generally unsatisfactory-and, in fact, impossible, when heparin was used as the anticoagulant. Accordingly, we examined the various steps in processing serum to determine their effect on sulfate analysis in the protein-free samples. Method of Protein Precipitation The recovery of 35SOi- from serum samples was examined in the presence of five different reagents used for precipitation of serum proteins (Table 1). Ethanol (95%, v/v) gave very poor recoveries of sulfate in the supematant solution. There was a significant loss of sulfate in the presence of uranyl ion, an agent in current use to remove phosphate anion and protein together during sulfate analysis (1,3). Loss of sulfate during uranyl acetate precipitation has also been reported by others (4). Precipitation of proteins with trichloroacetic acid (TCA),’ or TCA in methanol, gave acceptable recoveries of sulfate, and this agent was used in all subsequent studies. Efficiency of 133Ba Precipitation in the Presence of Sulfate and Other Ions in Serum , The reliability of the method depends on stoichiometric removal of lmBa2+ from solu* Abbreviation
used: TCA, trichloroacetic
acid.
MICROASSAY
0
FOR SERUM SULFATE
0.1
0.05
Sulfate
341
ANALYSIS
0.15
Concentration(mM)
FIG. 1. Completeness of *33BaS0, precipitation
at low sulfate concentrations.
tions, to form insoluble isotopic BaSO, in “ethereal” sulfate had no consistent or the presence of SO,‘-. Resolubilization of significant effect on the determination of the precipate constitutes interference in the sulfate in serum. However, the quantity of assay as does significant removal of 133Ba heparin employed for anticoagulant in by anions other than SO;-. microcapillary tubes and receiving tubes in Precipitation of sulfate from a water use in hospitals was sufficient to nullify any solution (0.9% saline) in the presence of effort to measure sulfate accurately in the plasma prepared from such samples. 0.4 mM BaCl, (final concentration) is incomplete at low physiological concentrations of sulfate (co.150 mM) (Fig. 1). Accordingly, Handling of Serum Samples a known amount of exogenous sulfate was We compared the recovery of sulfate from added to the serum samples and the con- freshly prepared serum samples and from centration of BaCl, increased to 1.2 mM. aliquots stored at -20°C for long intervals. Under these conditions, complete precipitaThe recovery was 98 f 5% (mean + SE, tion and the expected stoichiometry were obtained (Fig. 2). Recovery of sulfate from serum was 0.4 measured by first dialyzing sulfate away, ‘: z then reconstituting the sample by adding 0.3 z known amounts of inorganic sulfate, and Q) Q) finally proceeding with TCA precipitation '0 ‘0 of protein and analysis of sulfate by the 0.25 VI coprecipitation method. Recovery of sulfate m from human serum was complete and E accurate even at low concentrations. 0 : The presence of phosphate in physioa logical serum concentrations did not inter0 0.1 0.2 0.3 0.4 fere with sulfate analysis (Table 2); howSulfate Added( mM) ever, at levels of phosphate in serum above the physiological range (l-2 mM), a modest FIG. 2. Recovery of sulfate added to dialyzed human inhibition of sulfate analysis was apparent. serum. Vertical bars indicate SD of fcur independent Divalent cations (Ca*+ and Mg*+) and an experiments, each done in quadruplicate.
2
342
COLE, MOHYUDDIN, TABLE
2
EFFECTOF OTHERIONSON ASSAYOFSULFATE Ion concentration Ion
fm@
mgldl”
Sulfate recovery, % k SD (n)
Sodium phosphate (at pH 7.4)
0.1 0.5 1.0 2.5 5.0
0.3 1.6 3.1 7.8 16
103 101 100 95 90
f + f f f
7 (5) 5 (5) 5 (5) 11 (5) 5 (5)
W+
1.0 2.0
2.3 4.6
Ca*+
1.25 2.5 5.0
5 10 20
loo -e 10 (10) 109? 8(10) 107 f 8 (10)
DHEA-SO,”
0.001 1.0
-
105 * 2 (5) 102 2 4 (4)
107 f 15 (10) 109 * 10 (10)
a Equivalent concentrations of element in common clinical usage (e.g., inorganic phosphate as milligrams phosphorus per deciliter). * Dehydroepiandrosterone sulfate (A5-androstene3/3-ol-17-one-3-sulfate).
n = 5) after 24 h at -2O”C, and 95 & 6% after a week of frozen storage; after 3 months the recovery was 107 & 7%. We observed, in the aforementioned study with dialyzed serum, that repeated freezing and thawing caused the release of measurable amounts of sulfate, presumably from “bound” forms. On two occasions, 0.046 and 0.064 mM of additional sulfate (i.e., about 15% of the normal serum concentration) were recovered from dialyzed samples after four cycles of freeze-thaw. This finding indicates that serum samples should be handled prudently prior to analysis. Sulfate Concentration in Serum Human. The morning fasting serum sulfate concentration in healthy human adults was 0.297 ? 0.038 mM (mean + SD, n = 19). The individual values in the popula-
AND SCRIVER
tion sample were the mean of measurements in quadruplicate. The coefficient of variation calculated from 19 sets of quadruplicates was 7.4%. Miller and colleagues (3), using larger volumes of serum (> 1 ml) and the same isotope coprecipitation assay, showed that serum sulfate in man was 0.323 + 0.085 mM (mean + SD, n = 88); their coefficient of variation was also about 7%. However, consistency in time of sampling and alimentation was not reported in the early study and inconsistency of sampling may account for the slight difference between our “normal” serum sulfate value and the earlier report. Murine. Mouse serum contains more sulfate (> 1 mM) than human serum; the finding is analogous to that of phosphate content in sera of the two species. The sulfate concentration in C57B1/6J mouse serum was 1.00 ? 0.09 mM (mean + SD, n = 13) and in the serum of the Swiss albino mouse, 1.24 ? 0.16 mM (n = 7). The difference between inbred strains is significant (P < 0.001; Student t test), suggesting a genetic basis for the interstrain variation. The findings in human and murine sera offer a basis for further investigation that are now feasible with our microassays. ACKNOWLEDGMENTS This work was supported by the Medical Research Council of Canada and the Quebec Network of Genetic Medicine. D.E.C.C. is a Fellow of the MRC.
REFERENCES 1. Loeb, R. F., and Benedict, E. M. (1927) J. Clin. Invest.
4, 33-48.
2. Letanoff, T. V., and Reinhold, J. G. (1936) J. Biol. Chem. 114, 147-156. 3. Miller, E., Hlad, C. J., Levine, S., Holmes, J. H., and Ehick, H. (1961) J. Clin. Lab. Invest. 58, 656-661. 4. Berglund, F., and Sorbo, B. (1960) Stand. J. Clin. Lab. Invest. 12, 147- 153. 5. Hoffman, W. S., and Cardon, R. (1935) J. Biol. Chem. 109,711-127.
