1972
Anal. Chem. 1982, 5 4 , 1972-1976
(4) Ohi, M.; Akimoto, Y. Jpn. J . Appl. fhys. 1976, 15, 1177. (5) Bagaev, S. N.; Chebotaev, V. P. Appl. fhys. 1975, 7 , 71. (6) Keller, R. A.; Travis, J. C. "Analytical Laser Spectroscopy", 1st ed.; Wiley: New York, 1979; Chapter 8. (7) Peterson, N. C.; Kurylo, M. Y.; Braun, W.; Bass, A. M.; Keller, R. A. J . Opt. SOC.Am. 1971, 6 1 , 746. (6) Latz, H. W.; Wyles, H. F.; Green, R. B. Anal. Chem. 1973, 45, 2405. (9) Thrash, R. J.; Von Weyssenhoff, H.; Shirk, J. S . J. Chem. fhys. 1971, 5 5 , 4569. (10) Horlick, G.; Codding, E. G. Anal. Chem. 1974, 45, 133. (11) Atkinson, G. H.; Laufer, A. H.; Kuryio, M. J. J . Chem. fhys. 1973, 5 8 , 350. (12) Konjevic, N.; Kokovic, M. Spectrosc. Len. 1974, 7, 615. (13) Dewey, C. F., Jr. Laser Focus 1978, 14 (May), 48. (14) Javan, A.; Bennett, W. R., Jr.; Herriott, D. R. fhys. Rev. Lett. 1966, 6 , 106. (15) Bagayev, N.; Vasiienko, L. S.; Dmitrlyev, A. K.; Gol'dort, V. G.; Savortsov, M. N.; Chebotazev, V. P. Appl. fhys. 1976, 10, 231. (16) Babich, V. M.; Soiov'ev, V. S. I n "Issiedovaniya v Oblasti Kvantovol Elektroniki"; Lelkin, A. Ya., Ed.; Khar'kovskii Gosudarstvennyi Nau-
chno-Issledovatei'skii Institut Meddrologii: Kharkov, 1971; p 27. Ohi, M. Keltyo Kenkyusho Hokoku 1975, 2 4 , 1. Bloom, A. L.; Bell, W. C.; Rempel, R. C. Appl. Opt. 1963, 2 , 317. Sullivan, J. J.; O'Brien, M. J. "Modern Practice of Gas Chromatography", 1st ed.; Why: New York, 1977; Chapter 5.
RECEIVED for review December 15,1981. Accepted June 28, 1982. J.D.P. and R.B.G. acknowledge the support of Matching Fund Grant 14-34-0001-1205 from the Office of Water Research and Technology, Department of Interior, and the University of Arkansas. This work was presented in part at the 8th Annual Federation of Analytical Chemistry and Spectroscopy Societies, Philadelphia, PA, Sept 1981, and the 17th Midwest Regional Meeting, American Chemical Society, Columbia, MO, Nov 1981.
Determination of Glutathione and Glutathione Reductase with a Silver Sulfide Membrane Electrode Saad S. M. Haesan' and G. A. Rechnitz" Department of Chemistty, Universiv of Delaware, Newark, Delaware 1971 1
Oxldlred and reduced glutathlones (GSSG and GSH) are selectlvely measured In mlxtures at levels as low as lo-' M uslng the silver sulflde membrane electrode. The method is based upon removal of GSH and any other associated thlol from one allquot by alkylatlon wlth N-ethylmalelmlde and oxldatlon of a second allquot wlth lodlne. Both are then reduced wlth glutathione reductase enzyme and NADPH, at pH 8 followed by potentlometrlc monltoring of the produced GSH. Samples contalnlng as llttle as 30 ng/mL of each compound (-5 X IO-' to lo-' M) show an average recovery of 98% (standard devlatlon, 1.8 % ) wlthout any slgniflcant Interference from other sulfur compounds. The actlvlty of glutathlone reductase enzyme (0.4-4.0 mIU/mL) Is determlned by a reaction wlth controlled excess concentratlon of GSSG and NADPH, and measurlng the lnltlal rate of GSH productlon.
Glutathione is found in most living cells in the reduced (GSH) and oxidized (GSSG) forms at concentration levels depending upon the activity of the glutathione reductase enzyme which appears to be as ubiquitous in nature as GSH (1). Determination of these substances is essential in many clinical diagnoses. Erythrocytes of normal individuals contain 0.6-0.65 mg of GSH, 30-80 pg of GSSG, and 6.3-39 mIU of glutathione reductase/cm3, and higher concentrations (-3.1 mg of GSH, 255 pg of GSSG, and 4-100 IU of enzyme/g) are commonly found in various biological tissues (1,2). The level of GSH increases in myeloproliferative disorders and sharply decreases in malnutrition, renal and hepatic failure, diabetes, mental diseases as well as with deficiencies of glucose-6phosphate dehydrogenase, 6-phosphogluconate dehydrogenase, and vitamin Bz (3). GSH is also considered to be an important compound for maintenance of the clarity of the eye's lens. The normal lens contains about 4 mg of GSH/g whereas the cataractous lens contains neither GSH nor any protein thiol (4).On the other hand, deficiency of glutathione reductase On leave from Ain Shams University, Cairo, Egypt.
enzyme is one etiology for congenital nonspherocytic hemolytic anemia and is associated with various other clinical states (5). The most commonly used method for determining GSH is based on its reaction with 5,5'-dithiobis(2-nitrobenzoicacid) to form a highly colored anion with maximum absorption at 412 nm (6). Reactions with o-phthalaldehyde (7)and lucigenin (8)have been used for chemiluminscent measurements, but with poor selectivity. Exchange reactions between GSH and certain disulfides to give GSSG followed by reaction with NADPHz in the presence of glutathione reductase and measurement of the decrease in the absorbance of NADPHz at 340 nm have been described (9). These methods, however, suffer from a lack of selectivity and inapplicability to turbid solutions or to samples containing thiols. Selective methods based on the action of GSH on either cis-trans isomerization of malylpyruvic acid followed by monitoring the decrease of the substrate absorption at 330 nm (10) or hydrolysis of methylglyoxal by the glyoxalase enzyme and manometric measurement of the rate of acid formation (11) have been proposed. These methods are also not without limitations; the former requires an excess of a second enzyme to prevent accumulation of the product, the latter is not sensitive enough and both require strictly controlled reaction conditions and enzymes not available commercially. On the other hand, GSSG has traditionally been determined by reduction with NADPHz and glutathione reductase to give GSH and NADP, followed by either the measurement of the decrease in the absorbance of NADPHz at 340 nm (12) or the fluorescence of the product NADP at 460 nm (13) or of the color developed by reaction of the released GSH with 5,5'dithiobis(2-nitrobenzoic acid) (14). Methods based on the same principle have also been utilized for measuring the activity of glutathione reductase (15, 16). Again, thiols significantly interfere and need to be isolated before applying these methods. HPLC has effectively been used for removing the interferences of thiols ( 17-19) and utilized for simultaneous determination of GSSG, GSH, and related thiols (20). Apart from the several derivatization reactions and manipulation steps involved, only 49% of the actual GSSG is recovered (20).
0003-2700/82/0354-1972$01.25/00 1982 American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 54, NO. 12, OCTOBER 1982
The objective of our study was to develop sensitive and selective methods for determining glutathione and glutathione reductase using a simple monitoring system. The sulfide ion selective electrode, whiclh has been thoroughly evaluated and proved to be highly sensitive and selective for sulfide ion (22) and some thiols (22, 23) with fast response time, was used. The response characteriEitdcs of the electrode and the reaction conditions of the system GSSG-glutathione reductaseNADPHz were optimized to allow accurate determination of as low as M oxidized, reduced, and total glutathiones and glutathione reductase emyme activities as low as 0.4 mIU/mL without interferences from other sulfur-containing compounds.
EXPERIMENTAL SECTION Apparatus. All potentiiometric rneasurementa were made with a Corning Model 12 pH/mV meter using a silver sulfide membrane electrode (Orion Model 94-16) in conjunction with a single junction reference elecrode (Orion Model 90-01) and a Heath-Schlumberger SR-210 strip chart recorder. The measurements were made in thermostated 20-mL cellri at 25 f 0.1 "C controlled by Haake Model FS bath. Reagents. All the reagents used were of reagent grade unless otherwise stated and deionized water was used throughout. Glutathione reductase enzyme E.C. 1.6.4.2 (1388 IU/mL) from yeast, nicotinamide adenine dinucleotide phosphate sodium salt (reduced form, NADPH2),reduced and oxidized glutathiones,and N-ethylmaleimide were obtained from Sigma Chemical Co. (St. Louis, MO). Glutathione reductase enzyme working solutions (138.8 and 13.88 IU/mLI were prepared by 10- and 100-fold dilutions, respectively, of the commercially available enzyme. Aqueous stock lo-, M solutions of reduced glutathione, oxidized glutathione, and NADPH2were freshly prepared daily and stored in crushed ice until use. Series of 10-3-10-5 M of GSH and GSSG solutions were prepared by successive dilution of the stock solutions. Iodine and N-ethylmaleimide, lo-, M, were prepared in diethyl ether. Determination of GSSG. Mixtures of 4 mL of 0.1 M TrisHN03buffer of pH 8 and 10 pL of glutathione reductase enzyme (138.8 IU/mL) were transferred to a 20-mL double-jacketed reaction cell thermostated ai, 25 f 0.1 "C and containing a small Teflon-covered spin bar. Then, 10-pL aliquots of 10-2-10-5 M GSSG were added, each in ,B separate experiment, and the solution was stirred. A silver sulfide membrane electrode, in conjunction with a single junction reference elecrode, was immersed in the solution and the potential of the electrode system was allowed to reach a stable reading. 'The speed of the chart recorder was adjusted at 0.1 in./min and a 10-pL aliquot of M NADPH, added. The potential change was recorded until a steady reading was obtained, and a blank experiment was carried out under identical conditions without GSSG. The sulfide electrode was buffer of pH 4 between measuresoaked in 0.1 M Tris-"0, menta. The change of the plotential (AE) was plotted as a function of the logarithm of GSSG concentration and the plot was used for subsequent measurement of unknown GSSG samples in the to concentration range of 30 ng/mL to 60 pg/mL (5 X M) in the absence of GSH or other thiols. GSSG samples containinlg GSH and/or cysteine were similarly determined after removal of the thiol compounds. A series of 50-pL aliquots of 10-2-104 M GSSG was transferred to a 10-mL separatory funnel and diluted to 1mL with deionized water. Then, a 1-mL aliquot of M N-ethylmaleimide in ether was added and the mixture shaken for 8 min. The two phases were allowed to separate and the aqueolus phase was transferred to a 5-mL volumetric flask. The etherial layer was washed with 1-mL portions of deionized water and the aqueous phase was transferred to the flask. The volume was then made to the mark with deionized water, shaken, and placed in a boiling water bath for 3 rnin to remove any traces of ether. After the mixture was cooled, a 100-pL aliquot was transferred to the double-jacketed reaction cell and the GSSG content was enzymatically measured as described above. A blank experiment was carried out under identical conditions without GSSG and the change in the potential (aE) was plotted as a function of the logarithm of GSSG concentration. The graph was used for subsequent measurement of GSSG samples containing GSH or other thiols.
1973
Determination of GSH. A 100-pLaliquot of 10-2-10-4M GSH was transferred to a 10-mL separatory funnel and diluted with 1mL of deionized water. A 1-mLaliquot of lo-, M iodine solution in ether was then added and the mixture shaken for 3 min. The two phases were allowed to separate and the aqueous layer was collected in a 5-mL volumetric flask. The ether layer was washed with 1-mL portions of deionized water and the aqueous phase collected in the volumetric flask. The volume was made to the mark with deionized water and the flask shaken and placed in a boiling water bath for 3 min to remove traces of ether. After cooling, a 100-pL aliquot was transferred to the double jacketed reaction cell and the enzymatic reaction described above was followed. A blank experiment was carried out under identical conditions without GSH. The change of the potential of the was plotted as a function of the logarithm electrode system (a) GSH concentration in the range of 20 ng/mL to 30 pg/mL ( 10-7-10-4 M). The same procedure was used for determining total glutathione (GSSG + GSH). For determining GSH in the presence of GSSG, two separate 100-pL aliquots of 10-2-104 M of GSSG and GSH mixtures were allowed to react with iodine and N-ethylmaleimide as described above followed by enzymatic reduction and monitoring of the GSH produced. The difference in the potential changes caused by both solutions (equivalent to GSH) was compared with the calibration graph of GSH. Determination of Glutathione Reductase Activity. A mixture of 10 pL of M GSSG, 10 pL of lo-, M NADPH2,and 3.70 mL of 0.1 M Tris--HN03buffer of pH 8 was transferred to a 20-mL double-jacketed reaction cell thermostated at 25 f 0.1 "C containing the silver sulfide membrane electrode in conjunction with a single junction reference electrode. After a constant potential reading was obtained, the speed of the chart recorder was adjusted at 0.1 in./min, and 10-pL aliquota containing 0.4-4.0 mIU of glutathione reductase enzyme were added in a series of experiments. The rate curves were recorded, and the maximum initial rate of potential change expressed in millivolts per minute was graphically determined from the recorder by using the rate portion of the curve. The initial rate of GSH formation was plotted as a function of the enzyme activity in this range. A blank experiment was carried out under similar conditions without the enzyme and the calibration graph was used for subsequent enzyme activity measurement. N
RESULTS AND DISCUSSION Glutathione reductase enzyme selectively catalyzes the reductive cleavage of the disulfide linkage of GSSG in the presence of NADPH, (9) according to
GSSG + NADPH2
glutathione reductase
2GSH
+ NADP
(1)
The rate of formation and the overall concentration of the produced GSH were potentiometrically measured as a function of glutathione reductase activity and GSSG concentration, respectively, using the silver sulfide membrane electrode. The mechanism of the electrode response toward thiol compounds has previously been elucidated (22, 23). A preliminary study was made to evaluate the response characteristics of the electrode toward aqueous solutions of GSH at various pH values since most of the previously published work on the use of silver sulfide electrode for direct measurement of thiols has been carried out in either a strongly alkaline or nonaqueous media (22,241. It is fortunate that both GSH and GSSG are highly soluble in aqueous media. Thus, the steady potential reading of the electrode for a series of 0.1 M Tris-HNO, buffer solutions of pH 4-10 was measured a t 25 "C before and after the addition of GSH. In the pH range of 4-9, the relationship between AE and the logarithm of GSH concentration was consistently linear in the range of 10-3-10-7 M with an average slope of 40 mV/concentration decade. At pH values higher than 10, the electrode exhibited different slopes. It has been reported that the response characteristics of the Ag,S membrane electrode toward thiols depend on both the ratio with which the thiol compound reacts
1974
ANALYTICAL CHEMISTRY, VOL. 54,NO. 12, OCTOBER 1982
6ol cIL 750nglmL
70
6ot
I
225ng/mL
60
50
>
>
E
E.
J
30-
Q
lOl; ;$T
30
20 -
O
40
d" f
20
I
i
10
enzyme 0
8
I
16
0.5
Time, min
Figure 1. Typical potentiometric response curves for the reduction of GSSG with 1.4 I U of glutathione reductase enzyme and 1.2 X M NADPH, at pH 8 and 25 OC (1 ng/mL E 0.16 X M GSSG).
lo-'
lo-'
70
t
I
I
1.0
I
I
2.0
1.5
2.5 x 1 6 (
( N A D P H 2 ) ,M
Figure 3. Effect of NADPH, concentration on the reduction of 2.5 X 10" M GSSG and 1.4I U of glutathione reductase enzyme at pH 8 and 25 OC.
I
I
t
I
O'
30 20
25
30
Temperature
2 0 1
35
, oc
40
Flgure 4. Effect of temperature on the reduction of 2.5 X lo-' M GSSG, 1.4 I U of glutathione reductase enzyme and 1.2 X lo-' M NADPH? at pH 8.
0.6
1.2
1.8
2.4
3.0
E n z y m e a c t i v i t y , Iu14 mL
Flgure 2. Effect of glutathione reductase enzyme activity on the M NADPH, at pH 8 M GSSG with 1.2 X reduction of 2.5 X and 25 OC.
with silver ion and the formation constant of the reaction product (24). Since both factors are influenced by the pH change, a study of the effect of pH on the electrode performance was made. Solutions of M GSH in 0.1 M Tris-HN03 buffer of various pH values were titrated with silver nitrate using the Ag2Selectrode. Titration curves having similar potential change at the equivalence point and exhibiting the same stoichiometry were obtained for solutions of pH values in the range of 4-9; at pH values above 10 such behavior is not maintained. However, the electrode showed no significant response in the pH 6-8 range toward GSSG up to M. At lower and higher pHs, M GSSG displayed potential readings equivalent to that obtained for M GSH, probably due to a partial hydrolytic fission of the disulfide linkage. On the other hand, the response time of the electrode
to reach a constant potential reading for GSH decreased with the increase of the pH from 4 to 8, reaching a relatively fast response time ( 5 min) at pH 8-10. Response times of several minutes have also been reported with other thiols (24). Oxidized Glutathione (GSSG). Various concentrations of oxidized glutathione were allowed to initiate the enzymatic reaction shown in eq 1 under the optimized conditions described below, and the steady-state potential of the electrode system due to GSH formation was measured (Figure 1). The optimum activity of glutathione reductase enzyme required to catalyze the reaction of up to M GSSG in the presence of 1.2 X lod M NADPHz was determined in 0.1 M Tris-HN03 buffer of pH 8. An activity of 1.46 IU of the enzyme per 4 mL of the reaction mixture (0.35 IU/mL) was sufficient at 25 "C to effect an optimum reduction (Figure 2). The concentration of NADPHz required to saturate this level of enzyme was 2.5 X lo4 M (Figure 3). The effect of temperature on the reduction of GSSG was also investigated over the range of 20-40 "C with a reaction mixture consisting of 2.5 X M GSSG, 1.2 X M NADPH2, and 1.4 IU of the glutathione enzyme in a total volume of 4 mL of 0.1 M Tris-HN03 buffer of pH 8. The results (Figure 4) showed an increase in the concentration of N
ANALYTICAL CHEMISTRY, VOL. 54, NO. 12, OCTOBER 1982
1975
Table I. Enzymatic Determination of GSSG and GSH in Binary Mixtures with Silver Sulfide Membrane Electrodes anit of GSSG,' ng/mL found recovery, %
taken 30 50 75
29 51 73 98 49 97 196
100 50 100 200 % %
96.7 102.0 97.3 98.0 98.0 97.0 98.0
av recovery std dev a
50
recovery, %
49 97 740 51 197 980 1460
lOOb
750 50 200 1000 1500
98.0 97.0 98.7 102.0 98.5 98.0 97.3
98.1 1.7
1 ng/mL = 0.32 X
M GSH = 0.16 X
lo-'
M GSSG.
the GSH produced with decrease of temperature and a maximum reaction at 201-25 "C. The p H profile (Figure 5) indicated an optimum en:s:pne activity in the pH range of 5-7, which is a good agreement with the findings of other workers (4). However, the response time of the electrode toward GSH within this pH range is of the order of 20 min compared with only 5 min a t p H 8. The GSH concentration released at pH 8 was 95% of that obtainled a t the pH 5-7. Thus, a pH 8 was used throughout our study in order to take advantage of both the fast response of the electrode system and reasonable activity of the enzyme. Under the optimum conditions, a linear relationship was obtained between the change in the steady potential readings of the electrode system and the logarithm of GSSG concentration in the range of l(D-' to 2.5 X lo4 M. The precision and accuracy of the method were calculated from 10 repetitive analyses made on lo+ M samples of GSSG. The average recovery and mean standard deviation were 98.5% and 1.5%, respectively, and no interferences were noticed in the presence of 100-fold excess of other disulfide compounds (e.g., cystine). The high sensitivity of the method was demonstrated by the large potential change a t low GSSG concentration (-25 mV for M). Measurement of GSSG iin the presence of an equal amount and 10-fold excess of GSH or other thiols (e.g., cysteine) showed recoveries of about 90 and 20%, respectively. Since GSSG is usually associated with at least a 10-fold excess of GSH in various biological materials along with cysteine (1-3), attempts were made to eliminate the effect of GSH and any other thiols that might be' present. The use of a cation exchanger in the silver form to remove the thiol-containing compounds failed to give quantitative recovery for the GSSG. However, the effect of both GSH and cysteine was completely tolerated by an alkylation reaction with N-ethylmaleimide (6). The aqueous test solution was treated with M N-ethylmaleimide in ether (eq 2). After removal of the etherial layer, the GSSG in the aqueouci phase was allowed to initiate the enzymatic reaction. Under these conditions, lo4 M GSSG was quantitatively recovered (98 f 2%) in the presence of up to 100-fold excess of GSH and/or cysteine. CH-CO :H-CNC*H5 CHhCO
Reduced a n d Total Glutathiones. Direct potentiometric measurement of GSH at concentration levels of 10-4-10-7 M using the known addition "spiking" technique a t p H 8 (0.1 M Tris-HN03 buffer) showed an average recovery of 97 & 5% without interferences from LIPto M of phosphate, chloride, and iodide ions as well a9 GSSG. However, serious interferences were noticed in the presence of cysteine. The procedure used in the present, work for measuring GSH in the presence of cysteine or any other thiols takes advantage of both the quantitative oxiclation of GSH and any other asso-
98.5 1.6
Mixed with equal amount of cysteine.
5
7
6
6
9
PH
Flgure 5. Effect of pH on the reduction of 2.5 X IO-' M GSSG with 1.4 IU of glutathione reductase enzyme and 1.2 X M NADPH, using 0.1 M Tris-"0, buffer at 25 OC (standard deviation of the potential reading is f l mV).
ciated thiols with iodine into the corresponding disulfides and the selective enzymatic reduction of GSSG into GSH in the presence of other disulfides, according to
2GSH
+ I2
-
GSSG
+ 2HI
(3)
Solutions of GSH alone and mixed with up to a 100-fold excess of cysteine and/or GSSG were allowed to react with M etherial iodine solution, and after separation of the etherial layer, the total GSSG in the aqueous phase was enzymatically measured by monitoring the GSH released with the silver sulfide membrane electrode. The results obtained indicated that 97 f 2% of the originally present GSH or the total GSH plus GSSG were recovered without any noticeable interference from cysteine. Determination of GSH in the presence of GSSG was tried next using two aliquots of the test solution. The first was treated with etherial iodine and the second with etherial N-ethylmaleimide; then the aqueous phases were enzymatically reduced and the GSH produced was measured. The change of the electrode potential caused by the first and second reactions gave the total (GSH + GSSG) and reduced (GSH) glutathiones, respectively. Table I shows results obtained with various synthetic mixtures of GSH and GSSG in the presence of cysteine. The average recoveries were 98.5% (standard deviation 1.6%) and 98.1% (standard deviation
1976
Anal. Chem.
1982,5 4 , 1976-1980
1.7%) for GSH and GSSG, respectively. Glutathione Reductase Activity. In the presence of excess controlled concentrations of both GSSG and NADPH2, the initial rate of GSH formation was measured from the potential-time graph (mV/min) as a function of glutathione reductase activity (25). Best results were obtained with incubation of a mixture of 2.5 X lo4 M GSSG and the enzyme solutions a t 25 OC in a total volume of 4 mL of 0.1 M TrisHNOBbuffer of pH 8 followed by addition of 2.5 X M NADPH2 after a steady potential was attained. The results obtained under these optimal conditions s h w e d that the initial rate of GSH production is linearly related to the enzyme activity in the range of 0.4-4.0 mIU/mL. The sensitivity (1 (mV/min)/mIU) is high enough to measure the entire range of this activity with a precision of k3%. The method is more simple and sensitive than that previously suggested using a cyclic reaction with the pC02 electrode (26). The detection limit is lower than those reported with many other methods in current use (3, 16, 27).
LITERATURE CITED (1) Jocelyn, P. C. "Blochemistry of the SH Group"; Academic Press: London, 1972; Chapter 11. (2) Lazarow, A. I n "Glutathione"; Colowick, S., Schwarz, D. R., Lozarow, A., Stadtman, E., Racker, E., Waelsch, H., Eds.; Academic Press: New York, 1954; pp 231-270. (3) Demetrlou, J. A.; Drewes, P. A,; Gin, J. 9. I n "Clinical Chemistry"; Henry, R., Cannon, D. C., Winkelman, J. W., Eds., Harper and Row: Hagerstown, MD, 1974; Chapter 21.
VanHeynlngen, R.; P i e , A. Biochem. J. 1953, 5 3 , 436-444. Beutler, E. Science 1969, 165, 613-615. Tietze, F. Anal. Biochem. 1969, 2 7 , 502-522. Cohn, V. H.; Lyle, J. Anal. Blochem. 1966, 14, 434-440. Veazey, R. L.; Nleman, T. A. Anal. Chem. 1979, 5 1 , 2092-2095. Davidson, B. E.; Hird, F. J. R. Biochem. J . 1964, 9 3 , 232-236. Lack, L.; Smlth, M. Anal. Biochem. 1964, 8 , 217-222. Krimsky, I.; Racker, E. J. Biol. Chem. 1952, 198, 721-729. Belcher, R. V. Biochem. J . 1965, 9 4 , 705-711. Owens, C. W. I.; Welss, C.; Maker, H. S.; Lehrer, G. M. Anal. Biochem. 1980, 106, 5 12-5 16. Grlffith, 0. W. Anal. Biochem. 1960, 106, 207-212. Suzukl, I.Werkman, C. H. Biochem. J . 1960, 74, 359-365. Racker, E. I n "Methods in Enzymology"; Colowlck, S., Kaplan, N., Eds.; Academlc Press: New york, 1955; Vol. 11, pp 722-725. Takahashi, H.; Nara, Y.; Meguro, H.; Tuzimura, K. Agric. Biol. Chem. 1979, 43, 1439-1445. Adams, R. N. Life Sci. 1978, 23, 1167-1173. Mefford, I.; Reeve, J.; Kuhlenkamp, J.; Kaplowitz, N. J. Chromatogr. 1980, 194, 424-428. Reed, D. J.; Babson, J. R.; Beatty, P. W.; Brodie, A. E.; Ellis, W. W.; Botter, D. W. Anal. Biochem. 1980, 106, 55-62. Ma, T. S.; Hassan, S. S. M. "Organlc Analysis Using Ion-Selective Electrodes"; Academlc Press: London, 1982; Vol. I , Chapter 2. Tseng, P. K. C.; Gutnecht, W. F. Anal. Chem. 1975, 47, 2316-2319. Morf, W. E.; Kahr, G.; Slmon, W. Anal. Chem. 1974, 46, 1538-1543. Peter, F.; Rosset, R. Anal. Chim. Acta 1973, 6 4 , 397-408. Hassan, S . S. M.; Rechnltz, G. A. Anal. Chem. 1981, 5 3 , 512-515. Hassan, S . S.M.; Rechnitz, G. A. Anal. Chem. 1982, 5 4 , 303-307. Vennesland, 9. I n "Methods in Enzymology"; Colowick, S., Kaplan, N., Eds., Academic Press: New York, 1955; Vol. 11, pp 719-722.
RECEIVED for review April 8,1982. Accepted June 29,1982. We are grateful to the National Institutes of Health (Grant GM-25308) for support of this research.
Selectivity Characteristics of Potentiometric Carbon Dioxide Sensors with Various Gas Membrane Materials R. K. Kobos" Department of Chemistry, Virginia Commonwealth University, Richmond, Virginia 23284
S. J. Parks and M. E. Meyerhoff Department of Chemistty, The University of Michigan, Ann Arbor, Michigan 48 109
The selectivity characterlstlcs of potentlometrlc carbon dloxlde sensors with regard to various organlc and lnorganlc acid Interferences have been systematically examined. When used In conjunctlon with a standard slllcone rubber CO, permeable membrane, the sensor displays surprisingly large response to several organlc acids havlng low volatillty, e.g., benzoic, clnnamic, and sallcyllc acids. I f the the outer membrane Is changed to a microporous Teflon materlal, the response to these substances Is dlminished, but poor selectlvity over volatlle organics and acidic gases results. The use of a new homogeneous Teflon-llke membrane materlal is shown to offer dramatic improvement In selectlvity for CO, over all of the compounds tested. The mechanlstlc reasons for this enhanced selectlvity are discussed as are alternate methods for reducing organlc acid interferences when uslng more conventional membrane materlals.
Since the introduction of the pC02 gas sensor in 1957 for the measurement of C 0 2 in blood ( I , 2 ) , this potentiometric
device has also found widespread use in biological and industrial applications. For example, it has been employed to determine C02 in power station waters ( 3 ) ,to be the sensing element in enzyme electrodes (4-9), and to measure decarboxylase enzyme activities (9,IO). The response characteristics of the COz sensor, including the limits of detection, response time, and selectivity, have also been evaluated (3, 4, 11-15). An attractive analytical feature of the C 0 2 sensor is its selectivity. This sensor utilizes a gas-permeable membrane which separates the sample solution from an internal electrolyte solution. Both homogeneous (nonporous) polymer membranes, e.g., silicone rubber, and heterogeneous membranes, e.g., microporous Teflon, have been used ( 1 1 ) . The high selectivity of the sensor results from these hydrophobic membranes which prevent ionic species from entering the internal solution. The C02, however, passes through the membrane causing a pH change in the internal electrolyte, which is measured with a pH glass electrode. Several substances have been reported to interfere with commerical C 0 2 sensors, which utilize silicone rubber mem-
0003-2700/82/0354-1976$01.25/00 1982 American Chemical Society