- Email: [email protected]
Measurement of Carbonic Anhydrase Activity Using a Sensitive Fluorometric Assay
ANALYTICAL BIOCHEMISTRY ARTICLE NO.
252, 190–197 (1997)
AB972305
Measurement of Carbonic Anhydrase Activity Using a Sensitive Fluorometric Assay1 Richard Shingles*,2 and James V. Moroney† *Department of Biology, The Johns Hopkins University, Baltimore, Maryland 21218-2685; and †Department of Plant Biology, Louisiana State University, Baton Rouge, Louisiana 70803
Received March 31, 1997
The dehydration reaction of bicarbonate was measured using the fluorescent pH indicator, 8-hydroxypyrene-1,3,6-trisulfonate (pyranine), in combination with stopped-flow spectrofluorometry. The initial rate of bicarbonate dehydration was measured after mixing a pH 6.0 solution with a pH 8.0 solution containing bicarbonate. Addition of carbonic anhydrase to the pH 6.0 solution enabled the measurement of the initial rate of activity at physiological temperatures with resolution times of 2 ms. This assay was used to resolve differences in activity and sensitivity to sulfonamides by comparing mammalian carbonic anhydrase isoforms. The fluorescent technique used in this study is very sensitive, allowing the determination of initial rates with a protein concentration as little as 65 ng/ml. Pyranine can also be loaded into membrane vesicles to follow carbonic anhydrase activity within vesicles. The change in pH within vesicles is dependent on the concentration of externally added bicarbonate and the presence of carbonic anhydrase on either side of the membrane. Therefore, this assay can be used to measure carbon dioxide flux across membranes and to assess the contribution of carbonic anhydrase to this flux. q 1997 Academic Press
Carbonic anhydrase (carbonic hydrolyase, E.C. 4.2.1.1) catalyzes the reversible dehydration of HCO30 in solution: HCO30 / H/ } CO2 / H2O.
[1]
In the catalysis of CO2 hydration human carbonic anhy1 Supported by U.S. Department of Energy Grant (DE-FG02-92ER 200 280) to Richard E. McCarty and National Sciences Foundation Grant IBN-9632087 to James V. Moroney. 2 To whom correspondence should be addressed. Fax: 1-410-5165213.
drase III (HCA III)3 is one of the least efficient of the known isozymes, having a turnover rate of 1 1 104 s01, about 100-fold less than that of human carbonic anhydrase II (HCA II) (1, 2). These two isoforms also differ in their sensitivity to sulfonamides; HCA II is very sensitive to acetazolamide, while HCA III is peculiarly insensitive to acetazolamide (3). Carbonic anhydrase activity has been measured by various methods (for a review, see Ref. 4). The Wilbur– Anderson method (5), utilizing a pH electrode, is probably the most widely used assay. Unfortunately, this method does not give a linear response over a broad enzyme concentration range and cannot measure small differences in activity. In addition, the slow response times of pH electrodes do not give accurate determinations of the initial rate of the reaction. This assay also requires nonphysiological conditions as the temperature must be near 07C to give the highest sensitivity. A second method involves the use of a mass spectrophotometer to measure isotopic exchange between CO2 and H2O (6). This method is sensitive and accurate, but is time consuming and requires the dedicated use of a mass spectrometer. Radiotracer assays for carbonic anhydrase are also very sensitive (7), but these assays have the disadvantages of the handling and disposal of radioactive compounds, particularly the 14CO2 gas which is evolved during the assay. Colorimetric assays have been developed to follow the esterase activity of carbonic anhydrase (8, 9), but many of the color reagents used have low levels of sensitivity, and esterase activity may be a secondary and perhaps less physiological reaction. Methods that combine the use of pH indicators with stopped flow spectrophotometry have greatly improved the time resolution for analyzing the 3 Abbreviations used: AZA, acetazolamide; BCA II, bovine carbonic anhydrase isoform II; DBS40, dextran sulfonamide; DPX, p-xylene bispyridinium dibromide; DTT, dithiothreitol; HCA II, human carbonic anhydrase isoform II; HCA III, human carbonic anhydrase isoform III; pyranine, 8-hydroxypyrene-1,3,6-trisulfonate.
190
AID
0003-2697/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AB 2305
/
6m45$$$141
09-02-97 01:47:54
aba
FLUOROMETRIC ASSAY FOR CARBONIC ANHYDRASE ACTIVITY
kinetics of carbonic anhydrase (9, 10). Initial rates can be determined by the following relationship: d[x]/dt Å (buffer factor) 1 dA/dt,
191
preparations, cells were grown synchronously with 12h light/12-h dark cycles and harvested 5 h into the light cycle.
[2] Membrane Preparations
where x is the amount of H/ or HCO0 3 formed and A is the absorbance of the indicator. These reactions have to be corrected for the buffer factor caused by having bicarbonate and indicator in the mixture (4, 11). Unfortunately, due to the low sensitivity of the pH indicators used it is difficult to detect small differences in activity due to small changes in enzyme level or at low levels of carbonic anhydrase. Recently, a fluorometric method for detecting carbonic anhydrase using dansylamide was published (12). While this method is very sensitive for detecting the presence of carbonic anhydrase (29 ng/ml) it cannot be used to determine the initial rate of enzyme activity since dansylamide binds to the zinc occupying the active site and results in enzyme inhibition. In this study we describe a method for determining carbonic anhydrase activity using a fluorophore (pyranine) that is highly sensitive to small changes in pH between pH 6.5 and pH 8. Combined with stopped-flow spectrofluorometry this method provides good resolution of the initial rate of carbonic anhydrase activity in solution. We have compared three isozymes of carbonic anhydrase which vary in their catalytic properties and sensitivity to inhibitors to validate this assay. In previous studies (13, 14) pyranine has been loaded into membrane vesicles and the rate of pH change within the vesicles followed using spectrofluorometry. By loading vesicles with pyranine and carbonic anhydrase, this method can be modified to follow bicarbonate and carbon dioxide transport across biological membranes.
Asolectin was prepared from concentrate (Associated Concentrates, Woodside, NY) by suspending 20 mg in 2.0 ml of buffer [0.1 mM K-Hepes (pH 8.0), 5 mM MgCl2 , and 50 mM KCl] followed by sonication for 5 min. Intact chloroplasts were isolated from Chlamydomonas as described by Moroney and Mason (16). Chloroplast envelopes were isolated by modification of the method of Clemetson and Boschetti (17) as described by Ramazanov et al. (18). The chloroplast membranes were resuspended in 100 ml of resuspension buffer (10 mM Tris– HCl, pH 7.8, 1 mM EDTA, 10 mM leupeptin). Erythrocyte ghosts were isolated from horse blood by the method described by Steck and Kant (19). Vesicle Preparation Membrane vesicles were prepared by extrusion as described by Shingles and McCarty (14), in a buffer containing 5 mM pyranine, 0.1 mM K-Hepes (pH 8.0), 5 mM MgCl2 , and 50 mM KCl. Typically 2.0 ml of a membrane suspension containing about 1 mg (protein) of erythrocyte ghosts, 1 mg (protein) of inner envelopes, or 20 mg of asolectin was passed through the extrusion apparatus with a polycarbonate filter (100-nm-pore size) in place. The vesicle preparation was then passed through a 1.6 1 10 cm Sephadex G-50 column equilibrated with 10 mM K-Hepes (pH 8.0), 5 mM MgCl2 , and 50 mM KCl at 47C to remove external pyranine, and the eluant was diluted to 10 ml with the same buffer. The vesicle suspension was allowed to equilibrate for 2 h at 47C before use.
MATERIALS AND METHODS
Reagents and Enzymes
Assay of Carbonic Anhydrase
Pyranine and DPX were purchased from Molecular Probes (Eugene, OR). All other reagents were the highest grades commercially available. Horse blood was obtained from Carolina Biological Supply (Burlington NC). Bovine carbonic anhydrase (BCA II) and HCA II were purchased from Sigma (St. Louis MO). HCA III was a generous gift from David N. Silverman (University of Florida, Gainesville, FL).
Fluorescence measurements were collected with an OLIS-modified SLM-SPF-500C spectrofluorometer equipped with an OLIS USA-SF stopped-flow apparatus. Chamber A contained various concentrations of KHCO3 in 2.0 ml of buffer A (0.5 mM bicine-KOH) at pH 8.0. Chamber B contained carbonic anhydrase in 2.0 ml of buffer B (0.5 mM Hepes-KOH, 100 nM pyranine) at pH 6.0. Mixing of samples was achieved by a nitrogen-driven piston at 80 psi. Fluorescence was followed at an emission wavelength of 512 nm after excitation at 466 nm (Fs) and 413 nm (Fis). All slits were set at 10 nm with a cutoff filter (LP47; Oriel Co., Stamford, CT) placed over the entrance to the emission monochromator. All measurements were taken at 257C. For vesicle preparations, Chamber A contained KHCO3 and 2.0 ml of vesicle suspension in buffer A at pH 8.0 plus 5 mM DPX to quench the fluorescence of
Algal Strains and Culture Conditions The cell-wall-deficient mutant of Chlamydomonas reinhardtii, CC-400 (cw-15 mt/), was obtained from the Duke University Chlamydomonas Culture Collection. In liquid culture, the cells were grown in minimal medium (15), aerated with air, and illuminated with 300 mEinsteins of white light. For chloroplast envelope
AID
AB 2305
/
6m45$$$142
09-02-97 01:47:54
aba
192
AID
SHINGLES AND MORONEY
AB 2305
/
6m45$$2305
09-02-97 01:47:54
aba
FLUOROMETRIC ASSAY FOR CARBONIC ANHYDRASE ACTIVITY
residual external pyranine outside of the vesicles. Chamber B contained carbonic anhydrase in 2.0 ml of buffer B at pH 6.5. Pyranine fluorescence emission was monitored at 512 nm with excitation at 466 nm (Fs) and 418 nm (Fis inner envelopes and erythrocyte ghosts) or 413 nm (Fis asolectin vesicles). Data Reduction and Handling pH can be determined from the Fs/Fis ratio as previously described (17). A calibration curve is generated by adding small aliquots of HCl to the solution or vesicle suspension followed by measuring the relative fluorescence at Fs 466 and Fis 418 (inner envelopes and erythrocyte ghosts) or Fis 413 (in solution or asolectin vesicles) to determine the Fs/Fis ratio. pH values after each addition of base were measured with a pH electrode. The data fit the Henderson–Hasselbalch equation. Curve fitting was carried out using the graphing program Kaleidagraph (Synergy Software, Reading, PA). Data were fit to the single exponential equation Y Å m0*e0m1∗x, where m0 is the extent of the pH change (DpH) and m1 is the rate constant (k). Initial rates of acidification were determined from the relationship Vi Å k∗DpH. The correlation of the fits were greater than 0.95 as determined by the least squares method.
193
Measurement of pyranine fluorescence at a pH-insensitive wavelength (Fis) allows for the assessment of changes in fluorescence unrelated to pH effects. In this manner, the Fis can be used to assess ionic strength effects as the reaction proceeds. However, in these experiments the Fis changes very little during the course of the reaction (traces B and D in Fig. 1). The slight decrease in Fis observed over the duration of the trace is likely due to photobleaching of the fluorophore over time. There were no differences in the Fis traces between the catalyzed and uncatalyzed reactions, indicating that the presence of carbonic anhydrase does not affect pyranine fluorescence. In addition pyranine has no effect on activity as determined by a Wilbur–Anderson pH experiment (data not shown). The ratio of pyranine fluorescence at a pH sensitive wavelength to fluorescence at an insensitive wavelength (Fs/Fis) produces a trace which measures changes in solution pH independent of dye concentration and small changes in ionic strength (Fig. 1, traces E and F). In addition, the Fs/Fis ratio can be used to determine pH changes as HCO30 to CO2 conversion proceeds by the relationship pH Å pKa / log
(Fs/Fis)/(Fs/Fis)max . 1 0 [(Fs/Fis)/(Fs/Fis)max]
[3]
RESULTS
Spectrofluorometric Assay for Carbonic Anhydrase Mixing 5 mM KHCO3 in pH 8.0 buffer with a pH 6.0 buffer causes a pH shift in the equilibrium from HCO30 to CO2 , a process which consumes a proton (Eq. 1). The noncatalyzed reaction proceeds slowly, eventually reaching an equilibrium determined by the final pH after mixing the two solutions. The catalyzed reaction quickly reaches the same equilibrium. In these stopped-flow rapid mixing experiments pyranine is added as a fluorescent indicator to follow pH changes as the reaction proceeds. At the pH-sensitive wavelength (Fs), pyranine fluorescence increases as the conversion of HCO30 to CO2 proceeds consuming a proton (Fig. 1, traces A and C). This rate appears to be linear in the uncatalyzed reaction but is greatly accelerated by the addition of carbonic anhydrase. The catalyzed reaction follows first-order kinetics. If the reaction is allowed to proceed to completion (Ç2 min), the final extent of the fluorescence change (Fs) is equivalent between the catalyzed and uncatalyzed reactions.
The pKa and (Fs/Fis)max are determined in a separate experiment by titrating pyranine with acid and measuring the fluorescence and the pH at each concentration of bicarbonate used (13). When a solution at pH 8.0 is rapidly mixed with a solution at pH 6.0 the final pH is attained within the mixing time of the stopped-flow instrument (Ç4 ms). In the presence of carbonic anhydrase, but absence of bicarbonate there is little change in pH over 10 s (Fig. 2). However, when bicarbonate is added the pH increases in a single exponential manner to reach a new pH equilibrium within 5 s. With increasing bicarbonate concentration up to 10 mM both the extent and the initial rate of the reaction increase. The initial rate of carbonic anhydrase activity was calculated over a wide range of enzyme concentrations (0 – 25 mg) and shown to be linear over the entire range (Fig. 3). The sensitivity of this assay is also evident as initial rates of carbonic anhydrase activity could be detected with concentrations as small as 65 ng/ml protein.
FIG. 1. Fluorescence emission trace for measurement of carbonic anhydrase activity. Changes in pyranine fluorescence after mixing 5 mM KHCO3 in pH 8.0 buffer (0.5 mM bicine-KOH) with a pH 6.0 buffer (0.5 mM Hepes-KOH) was followed by measuring fluorescent emission at 512 nm after excitation at 466 nm (Fs) or 413 nm (Fis). Conditions were (l) catalyzed reaction with 1.0 mg/ml BCA II [(A) Fs , (B) Fis , and (E) Fs/Fis] and (s) uncatalyzed reaction in absence of BCA II [(C) Fs , (D) Fis , and (F) Fs/Fis]. The uncatalyzed reaction reaches final equilibrium in about 2 min. Curve fits were determined using a least squares method.
AID
AB 2305
/
6m45$$$142
09-02-97 01:47:54
aba
194
SHINGLES AND MORONEY TABLE 1
Comparison of Kinetic Parameters for Different Sources of Carbonic Anhydrase Enzyme
Concentration (mgrml01)
Vmax (DpHrs01)
Km (mM)
Kcat (1104 s01)
BCA II HCA II HCA II HCA III HCA III
1.00 1.05 10.55 1.1 11.1
0.33 0.37 3.81 0.03 0.28
3.8 2.2 2.2 22 21
83 85 87 1.3 1.4
Note. Carbonic anhydrase was added in the amounts indicated. The initial rate of the reaction was determined as described in the legend to Fig. 2. Uncatalyzed rates were subtracted from catalyzed rates. From the plot of the initial rate vs bicarbonate concentration the Vmax and Km were determined. Kcat was determined from the Vmax and the enzyme concentration. FIG. 2. Fluorescent assay for carbonic anhydrase activity. BCA II was added at 1.0 mg/ml and bicarbonate was added at a range of concentrations from 0 to 10 mM. Change in pyranine fluorescence upon mixing of a pH 8.0 buffer with a pH 6.0 buffer was followed as described in the legend to Fig. 1. Fluorescence was converted to pH as described under Materials and Methods.
Comparison of Carbonic Anhydrase Isoforms The spectrofluorometric assay used in this study allows for the resolution of the initial rates of carbonic anhydrase activity from which kinetic parameters can be derived. BCA II and HCA II at similar enzyme concentrations (1.0 mg/ml) had similar Vmax
values (Table 1). HCA III at the same protein concentration had a velocity about 10-fold lower than HCA II. Increasing the enzyme concentrations 10-fold resulted in a corresponding increase in velocity for both isozymes (Table 1). The Km for the dehydration reaction of HCO30 could also be determined using this assay (Table 1). BCA II had a Km value of approximately 4 mM, while HCA II had a slightly lower Km value of 2.2 mM for the same reaction. HCA III had a Km value of 22 mM, 10 times larger than the Km for HCA II. Under the conditions employed in this study, with a bicarbonate concentration of 5 mM, a DpH of 1.0 unit is equivalent to a proton concentration change of 8 1 1005 mol. Knowing the protein concentration, the turnover rate (Kcat) can be determined for various isoforms of carbonic anhydrase (Table 1). BCA II and HCA II had similar turnover rates of 83 to 87 1 104 s01. HCA III had a turnover rate of 1.3 to 1.4 1 104 s01 or approximately 65 times slower than BCA II and HCA II. The determinations of Kcat could be calculated at enzyme concentrations differing by 10-fold with reproducible results. Inhibitors of Carbonic Anhydrase Activity Measured with the Fluorometric Assay
FIG. 3. Sensitivity of fluorescent carbonic anhydrase assay. Initial rates were determined as described in the legend to Fig. 1. BCA II at a concentration of 0–25 mg/ml was added with 5 mM bicarbonate.
AID
AB 2305
/
6m45$$$142
09-02-97 01:47:54
Two inhibitors of carbonic anhydrase activity have been successfully employed in this spectrofluorometric assay. Acetazolamide is a widely used inhibitor for atype carbonic anhydrases, although it is reported to have little effect on HCA III activity at low concentrations (3). Dextran sulfonamide (DBS40) is a synthesized compound in which a sulfonamide is linked to a 40,000 MW dextran making the sulfonamide membrane impermeant (20). Both acetazolamide and DBS40 have no effect on pyranine fluorescence in this assay.
aba
FLUOROMETRIC ASSAY FOR CARBONIC ANHYDRASE ACTIVITY
FIG. 4. Comparison of initial rates for two isoforms of carbonic anhydrase. Initial rates were calculated from the pH vs time plots for different concentrations of added bicarbonate. HCA II was used at a concentration of 1 mg/ml and HCA III was added at a concentration of 10 mg/ml. Rates were fit using a least mean squares method.
Acetazolamide added at a concentration of 0.35 mM inhibits the enzyme activity of HCA II (Fig. 4). The inhibited rate of HCA II approximates the uncatalyzed rate for HCO30 to CO2 conversion. By contrast, HCA III is almost completely insensitive to inhibition by this level of acetazolamide. DBS40 was shown to inhibit BCA activity as described in the next section.
195
FIG. 5. Movement of CO2 across a membrane.
A higher pH difference (alkaline inside) results in a higher initial rate of intravesicular acidification. In addition the reaction is dependent upon the presence of carbonic anhydrase on both sides of the membrane to ensure rapid acidification of the intravesicular space (Fig. 6, Table 2). Carbonic anhydrase-mediated CO2 diffusion across
Measuring Carbonic Anhydrase Activity inside Membrane Vesicles Asolectin (soybean phospholipids) was used as a model membrane to test the effectiveness of measuring carbonic anhydrase activity within membrane vesicles. Vesicles were loaded with pyranine and BCA II using previously described methods (14). DPX was added to the vesicle suspension to quench the fluorescence of any residual external pyranine. Vesicles containing 5 mM bicarbonate at pH 8.0 were added to Chamber A of the stopped-flow apparatus. Chamber B contained 10 mg/ml BCA II prepared in buffer with pH ranging between 6.0 to 7.5. Mixing these two solutions drives the hydration of HCO30 to CO2 outside the membrane vesicles as diagrammed in Fig. 5. After CO2 crosses the membrane the higher internal vesicle pH (8.0) drives the dehydration reaction of CO2 to HCO30 liberating a proton in the process. Hence, CO2 transport across the membrane can be measured by the acidification of the intravesicular space and followed by the change in pyranine fluorescence. The initial rate of this process is entirely dependent upon the pH difference imposed across the membranes (Fig. 6).
AID
AB 2305
/
6m45$$$143
09-02-97 01:47:54
FIG. 6. Measurement of CO2 diffusion across membrane vesicles. Asolectin membrane vesicles were loaded with 5 mM pyranine and 10 mg/ml BCA II where indicated. External BCA II was added at 10 mg/ml. KHCO3 was added to 5 mM to the medium and the change in pyranine fluorescence used to monitor internal pH of the vesicles. The initial internal pH was 8.0, while the external pH was varied between 6.0 and 7.5.
aba
196
SHINGLES AND MORONEY TABLE 2
Measurement of Carbonic Anhydrase-Assisted CO2 Diffusion across Membranes 0BCA Inside (DpHrs01)
/BCA inside (DpHrs01)
Additions
Asolectin vesicles
Asolectin vesicles
Chlamydomonas inner envelopes
Erythrocyte ghosts
Control / DBS40 HCO0 3 HCO0 3 / DBS40
0.02 0.04 0.03
0.02 0.37 0.05
0.03 0.41 0.10
0.13 0.72 0.14
Note. The initial rate of the reaction was determined as described under Materials and Methods. DBS40 was added at a concentration of 75 mg/ml outside the vesicles and bicarbonate was added at a concentration of 5 mM. BCA II was present outside the membrane vesicles in all experiments at a concentration of 10 mg/ml.
membranes was measured in asolectin vesicles, isolated Chlamydomonas chloroplast inner envelope vesicles, and erythrocyte membrane vesicles (Table 2). With carbonic anhydrase loaded inside the vesicles there was no significant vesicle acidification in the absence of any added HCO30 and the presence of the inhibitor (DBS40). Erythrocyte membranes had a higher control rate than asolectin or Chlamydomonas inner envelopes. When HCO30 was added outside the vesicles, the rate of vesicle acidification was low when carbonic anhydrase is absent inside the vesicles. However, when enzyme was present inside the vesicles the rate of vesicle acidification was high. When DBS40 was added with HCO30 outside the vesicles, the rate of acidification was similar to control rates in both asolectin and erythrocyte membranes. In Chlamydomonas inner envelopes the initial rate of vesicle acidification was inhibited 75% by DBS40 but the residual rate was still three times higher than control rates. DISCUSSION
Using pyranine as an indicator of the pH changes as the protonation of bicarbonate proceeds has several advantages. As a fluorescent indicator it has a high resolution for pH changes that are nearly linear in the physiological range of 6.5–8.0. Pyranine responds rapidly to changes in pH, has a pKa near neutrality, and lacks toxicity to carbonic anhydrase activity. Pyranine can be detected spectrofluorometrically at very low levels and therefore can be added in low amounts to the reaction mixture. In these experiments pyranine was typically added at concentrations of 100 nM, much less than the 27–50 mM levels of absorbance dyes used in previous studies (11, 21). This low concentration of pyranine reduces any problems that might be encountered as a result of buffering by the indicator dye itself. The sensitivity of pyranine also allows for low levels of carbonic anhydrase to be assayed. Many assays appear to be limited in the range of enzyme that can be mea-
AID
AB 2305
/
6m45$$$143
09-02-97 01:47:54
sured accurately for activity (4), as it is often difficult to measure more than a 10-fold difference in concentration. In previous studies using absorbance dyes the carbonic anhydrase concentration used was typically between 2 and 4 mM (11, 21). In this study the concentration of BCA II ranged between 0.002 and 0.75 mM with 0.03 mM giving good resolution of initial rates for kinetic studies (Figs. 3 and 4). This assay was sensitive enough to measure initial rates of HCA III activity at enzyme concentrations between 0.03 and 0.3 mM (Table 1, Fig. 4) even though HCA III is 100 times less active than HCA II (1, 2, 21). Absorbance dyes have been used in stopped-flow experiments to measure the initial rate of carbonic anhydrase activity (11, 21). The high concentration of dye and bicarbonate resulted in a buffer effect that had to be corrected in order to determine the initial rates of the reaction (Eq. 2). In the studies reported here, the initiation of the carbonic anhydrase catalyzed reaction also caused an offset in the measured fluorescence at the pH sensitive wavelength but not at the pH-insensitive wavelength (Fig. 1). Part of this offset could be due to the reaction proceeding before the first measurements are collected (Ç4 ms), but more likely it is due to the buffering effect of added bicarbonate. Pyranine is a ratiometric indicator allowing quantitative measurements to be made regardless of probe concentration and minor differences in ionic strength as would be experienced in these studies. In these experiments the pH is dependent upon the calculated pKa and the (Fs/Fis)max as described by Eq. 3. Therefore, it is important that both the pKa and the (Fs/Fis)max be determined at each concentration of bicarbonate employed to account for buffering effects. In general, the pH increased with a buffering constant of 0.031 pH units/ mM bicarbonate. Many enzyme inhibitors may be fluorescent themselves or have an effect on the fluorophore such as fluorescent quenching or enhancement. Known inhibitors of carbonic anhydrase activity were used in this study
aba
FLUOROMETRIC ASSAY FOR CARBONIC ANHYDRASE ACTIVITY
to investigate if they could be used as effective inhibitors in a fluorescent assay system. Acetazolamide inhibited HCA II activity very effectively but did not inhibit HCA III activity (Fig. 4), in agreement with previous inhibition studies (3). Carbonic anhydrase is involved in the hydration of CO2 to HCO30 inside erythrocyte membranes (22) and is also involved in the interconversion of HCO30 and CO2 in chloroplasts (16). Isolation of membranes from these organelles and cells allows for the study of carbonic anhydrase-assisted CO2 diffusion across membrane vesicles using this fluorescent assay. In the absence of enzyme the rate of diffusion of CO2 through membranes is limited by the slow hydration–dehydration reaction of CO2 (Fig. 5, Table 2). When adequate levels of carbonic anhydrase are present on both sides of the membrane, the hydration–dehydration reaction is so fast that CO2 diffusion becomes rate limiting (23). Under these conditions, the permeability of membranes to CO2 diffusion can be assessed. Using the fluorescent assay it was possible to measure CO2-dependent pH changes in artificial vesicles as well as chloroplast inner envelope membranes and red cell plasma membranes (Fig. 6, Table 2). Potentially, the presence of membrane-associated carbonic anhydrase or bicarbonate transporter could be detected using inhibitors or by comparing pH changes in the presence and absence of enzyme inside or outside the vesicles. By comparison to other methods using pH indicators (11, 21), the use of pyranine as a fluorescent pH indicator is much more sensitive for detection of carbonic anhydrase activity. This assay is useful for measuring carbonic anhydrase reactions at physiological temperatures and pH ranges and can be applied to measure CO2 diffusion across membranes. Finally, although this study has concentrated on the use of stopped-flow techniques to measure the initial rate kinetics of carbonic anhydrase, it is also possible to follow the reaction in a cuvette assay with any spectrofluorometer that has a reasonable time drive resolution to at least qualitatively assess enzyme activity. ACKNOWLEDGMENTS We would like to thank Zygmunt Schroeder for the construction of the small-volume extrusion apparatus and Catherine Mason for
AID
AB 2305
/
6m45$$$143
09-02-97 01:47:54
197
preparation of Chlamydomonas chloroplast envelopes. We also thank David Husic (Lafayette College, Easton, PA) for his generous gift of DBS40 and David Silverman (University of Florida, Gainesville, FL) for his gift of HCA III. We are grateful to Richard E. McCarty for helpful discussions and critical reading of the manuscript.
REFERENCES 1. Khalifah, R. G., and Silverman, D. N. (1991) in The Carbonic Anhydrases: Cellular Physiology and Molecular Genetics (Dodgson, S. J., Tashian, R. E., Gros, G., and Carter, N. D., Eds.), pp. 49–70, Plenum, New York. 2. Silverman, D. N. (1991) Can. J. Bot. 69, 1070–1078. 3. Dodgson, S. J. (1991) in The Carbonic Anhydrases: Cellular Physiology and Molecular Genetics (Dodgson, S. J., Tashian, R. E., Gros, G., and Carter, N. D., Eds.), pp. 49–70, Plenum, New York. 4. Forster, R. E. (1991) in The Carbonic Anhydrases: Cellular Physiology and Molecular Genetics (Dodgson, S. J., Tashian, R. E., Gros, G., and Carter, N. D., Eds.), pp. 79–98, Plenum, New York. 5. Wilbur, K. M., and Anderson, N. G. (1948) J. Biol. Chem. 176, 147–154. 6. Hock, G., and Kok, B. (1963) Arch. Biochem. Biophys. 101, 160– 170. 7. Stemler, A. (1993) Anal. Biochem. 210, 328–331. 8. Pocker, Y., and Stone, J. T. (1967) Biochemistry 6, 668–678. 9. Abou-Rebyeh, H., Dietsch, P., and Korber, F. (199) in Carbonic Anhydrase (Botre, F., Gros, G., and Storey, B. T., Eds.), pp. 435– 439, VCH Press, New York. 10. Gibson, Q. H., and Milnes, L. (1964) Biochem. J. 91, 161–171. 11. Khalifah, R. G. (1971) J. Biol. Chem. 246, 2561–2573. 12. Li, H-N., and Ci, Y-X. (1995) Anal. Chim. Acta 317, 353–357. 13. Shingles, R., and McCarty, R. E. (1994) Plant Physiol. 106, 731– 737. 14. Shingles, R., and McCarty, R. E. (1995) Anal. Biochem. 229, 92– 98. 15. Sueoka, N. (1960) Proc. Natl. Acad. Sci. USA 46, 83–91. 16. Moroney, J. V., and Mason, C. B. (1991) Can. J. Bot. 69, 1017– 1024. 17. Clemetson, J. M., and Boschetti, A. (1988) Biochim. Biophys. Acta 943, 371–374. 18. Ramazanov, Z., Mason, C. B., Geraghty, A. M., Spalding, M. H., and Moroney, J. V. (1993) Plant Physiol. 101, 1195–1199. 19. Steck, T. L., and Kant, J. A. (1974) Methods Enzymol. 31, 172– 177. 20. Husic, H. D. (1991) Can. J. Bot. 69, 1079–1087. 21. Rowlett, R. S., Gargiulo, N. J., Santoli, F. A., Jackson, J. M., and Corbett, A. H. (1991) J. Biol. Chem. 266, 933–941. 22. Hladky, S. B., and Rink, T. J. (1977) in Membrane Transport in Red Cells (Ellory, J. C., and Lew, V. L., Eds.), pp. 115–135. 23. Gutknecht, J., Bisson, M. A., and Tosteson, F. C. (1977) J. Gen. Physiol. 69, 759–794.
aba
252, 190–197 (1997)
AB972305
Measurement of Carbonic Anhydrase Activity Using a Sensitive Fluorometric Assay1 Richard Shingles*,2 and James V. Moroney† *Department of Biology, The Johns Hopkins University, Baltimore, Maryland 21218-2685; and †Department of Plant Biology, Louisiana State University, Baton Rouge, Louisiana 70803
Received March 31, 1997
The dehydration reaction of bicarbonate was measured using the fluorescent pH indicator, 8-hydroxypyrene-1,3,6-trisulfonate (pyranine), in combination with stopped-flow spectrofluorometry. The initial rate of bicarbonate dehydration was measured after mixing a pH 6.0 solution with a pH 8.0 solution containing bicarbonate. Addition of carbonic anhydrase to the pH 6.0 solution enabled the measurement of the initial rate of activity at physiological temperatures with resolution times of 2 ms. This assay was used to resolve differences in activity and sensitivity to sulfonamides by comparing mammalian carbonic anhydrase isoforms. The fluorescent technique used in this study is very sensitive, allowing the determination of initial rates with a protein concentration as little as 65 ng/ml. Pyranine can also be loaded into membrane vesicles to follow carbonic anhydrase activity within vesicles. The change in pH within vesicles is dependent on the concentration of externally added bicarbonate and the presence of carbonic anhydrase on either side of the membrane. Therefore, this assay can be used to measure carbon dioxide flux across membranes and to assess the contribution of carbonic anhydrase to this flux. q 1997 Academic Press
Carbonic anhydrase (carbonic hydrolyase, E.C. 4.2.1.1) catalyzes the reversible dehydration of HCO30 in solution: HCO30 / H/ } CO2 / H2O.
[1]
In the catalysis of CO2 hydration human carbonic anhy1 Supported by U.S. Department of Energy Grant (DE-FG02-92ER 200 280) to Richard E. McCarty and National Sciences Foundation Grant IBN-9632087 to James V. Moroney. 2 To whom correspondence should be addressed. Fax: 1-410-5165213.
drase III (HCA III)3 is one of the least efficient of the known isozymes, having a turnover rate of 1 1 104 s01, about 100-fold less than that of human carbonic anhydrase II (HCA II) (1, 2). These two isoforms also differ in their sensitivity to sulfonamides; HCA II is very sensitive to acetazolamide, while HCA III is peculiarly insensitive to acetazolamide (3). Carbonic anhydrase activity has been measured by various methods (for a review, see Ref. 4). The Wilbur– Anderson method (5), utilizing a pH electrode, is probably the most widely used assay. Unfortunately, this method does not give a linear response over a broad enzyme concentration range and cannot measure small differences in activity. In addition, the slow response times of pH electrodes do not give accurate determinations of the initial rate of the reaction. This assay also requires nonphysiological conditions as the temperature must be near 07C to give the highest sensitivity. A second method involves the use of a mass spectrophotometer to measure isotopic exchange between CO2 and H2O (6). This method is sensitive and accurate, but is time consuming and requires the dedicated use of a mass spectrometer. Radiotracer assays for carbonic anhydrase are also very sensitive (7), but these assays have the disadvantages of the handling and disposal of radioactive compounds, particularly the 14CO2 gas which is evolved during the assay. Colorimetric assays have been developed to follow the esterase activity of carbonic anhydrase (8, 9), but many of the color reagents used have low levels of sensitivity, and esterase activity may be a secondary and perhaps less physiological reaction. Methods that combine the use of pH indicators with stopped flow spectrophotometry have greatly improved the time resolution for analyzing the 3 Abbreviations used: AZA, acetazolamide; BCA II, bovine carbonic anhydrase isoform II; DBS40, dextran sulfonamide; DPX, p-xylene bispyridinium dibromide; DTT, dithiothreitol; HCA II, human carbonic anhydrase isoform II; HCA III, human carbonic anhydrase isoform III; pyranine, 8-hydroxypyrene-1,3,6-trisulfonate.
190
AID
0003-2697/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AB 2305
/
6m45$$$141
09-02-97 01:47:54
aba
FLUOROMETRIC ASSAY FOR CARBONIC ANHYDRASE ACTIVITY
kinetics of carbonic anhydrase (9, 10). Initial rates can be determined by the following relationship: d[x]/dt Å (buffer factor) 1 dA/dt,
191
preparations, cells were grown synchronously with 12h light/12-h dark cycles and harvested 5 h into the light cycle.
[2] Membrane Preparations
where x is the amount of H/ or HCO0 3 formed and A is the absorbance of the indicator. These reactions have to be corrected for the buffer factor caused by having bicarbonate and indicator in the mixture (4, 11). Unfortunately, due to the low sensitivity of the pH indicators used it is difficult to detect small differences in activity due to small changes in enzyme level or at low levels of carbonic anhydrase. Recently, a fluorometric method for detecting carbonic anhydrase using dansylamide was published (12). While this method is very sensitive for detecting the presence of carbonic anhydrase (29 ng/ml) it cannot be used to determine the initial rate of enzyme activity since dansylamide binds to the zinc occupying the active site and results in enzyme inhibition. In this study we describe a method for determining carbonic anhydrase activity using a fluorophore (pyranine) that is highly sensitive to small changes in pH between pH 6.5 and pH 8. Combined with stopped-flow spectrofluorometry this method provides good resolution of the initial rate of carbonic anhydrase activity in solution. We have compared three isozymes of carbonic anhydrase which vary in their catalytic properties and sensitivity to inhibitors to validate this assay. In previous studies (13, 14) pyranine has been loaded into membrane vesicles and the rate of pH change within the vesicles followed using spectrofluorometry. By loading vesicles with pyranine and carbonic anhydrase, this method can be modified to follow bicarbonate and carbon dioxide transport across biological membranes.
Asolectin was prepared from concentrate (Associated Concentrates, Woodside, NY) by suspending 20 mg in 2.0 ml of buffer [0.1 mM K-Hepes (pH 8.0), 5 mM MgCl2 , and 50 mM KCl] followed by sonication for 5 min. Intact chloroplasts were isolated from Chlamydomonas as described by Moroney and Mason (16). Chloroplast envelopes were isolated by modification of the method of Clemetson and Boschetti (17) as described by Ramazanov et al. (18). The chloroplast membranes were resuspended in 100 ml of resuspension buffer (10 mM Tris– HCl, pH 7.8, 1 mM EDTA, 10 mM leupeptin). Erythrocyte ghosts were isolated from horse blood by the method described by Steck and Kant (19). Vesicle Preparation Membrane vesicles were prepared by extrusion as described by Shingles and McCarty (14), in a buffer containing 5 mM pyranine, 0.1 mM K-Hepes (pH 8.0), 5 mM MgCl2 , and 50 mM KCl. Typically 2.0 ml of a membrane suspension containing about 1 mg (protein) of erythrocyte ghosts, 1 mg (protein) of inner envelopes, or 20 mg of asolectin was passed through the extrusion apparatus with a polycarbonate filter (100-nm-pore size) in place. The vesicle preparation was then passed through a 1.6 1 10 cm Sephadex G-50 column equilibrated with 10 mM K-Hepes (pH 8.0), 5 mM MgCl2 , and 50 mM KCl at 47C to remove external pyranine, and the eluant was diluted to 10 ml with the same buffer. The vesicle suspension was allowed to equilibrate for 2 h at 47C before use.
MATERIALS AND METHODS
Reagents and Enzymes
Assay of Carbonic Anhydrase
Pyranine and DPX were purchased from Molecular Probes (Eugene, OR). All other reagents were the highest grades commercially available. Horse blood was obtained from Carolina Biological Supply (Burlington NC). Bovine carbonic anhydrase (BCA II) and HCA II were purchased from Sigma (St. Louis MO). HCA III was a generous gift from David N. Silverman (University of Florida, Gainesville, FL).
Fluorescence measurements were collected with an OLIS-modified SLM-SPF-500C spectrofluorometer equipped with an OLIS USA-SF stopped-flow apparatus. Chamber A contained various concentrations of KHCO3 in 2.0 ml of buffer A (0.5 mM bicine-KOH) at pH 8.0. Chamber B contained carbonic anhydrase in 2.0 ml of buffer B (0.5 mM Hepes-KOH, 100 nM pyranine) at pH 6.0. Mixing of samples was achieved by a nitrogen-driven piston at 80 psi. Fluorescence was followed at an emission wavelength of 512 nm after excitation at 466 nm (Fs) and 413 nm (Fis). All slits were set at 10 nm with a cutoff filter (LP47; Oriel Co., Stamford, CT) placed over the entrance to the emission monochromator. All measurements were taken at 257C. For vesicle preparations, Chamber A contained KHCO3 and 2.0 ml of vesicle suspension in buffer A at pH 8.0 plus 5 mM DPX to quench the fluorescence of
Algal Strains and Culture Conditions The cell-wall-deficient mutant of Chlamydomonas reinhardtii, CC-400 (cw-15 mt/), was obtained from the Duke University Chlamydomonas Culture Collection. In liquid culture, the cells were grown in minimal medium (15), aerated with air, and illuminated with 300 mEinsteins of white light. For chloroplast envelope
AID
AB 2305
/
6m45$$$142
09-02-97 01:47:54
aba
192
AID
SHINGLES AND MORONEY
AB 2305
/
6m45$$2305
09-02-97 01:47:54
aba
FLUOROMETRIC ASSAY FOR CARBONIC ANHYDRASE ACTIVITY
residual external pyranine outside of the vesicles. Chamber B contained carbonic anhydrase in 2.0 ml of buffer B at pH 6.5. Pyranine fluorescence emission was monitored at 512 nm with excitation at 466 nm (Fs) and 418 nm (Fis inner envelopes and erythrocyte ghosts) or 413 nm (Fis asolectin vesicles). Data Reduction and Handling pH can be determined from the Fs/Fis ratio as previously described (17). A calibration curve is generated by adding small aliquots of HCl to the solution or vesicle suspension followed by measuring the relative fluorescence at Fs 466 and Fis 418 (inner envelopes and erythrocyte ghosts) or Fis 413 (in solution or asolectin vesicles) to determine the Fs/Fis ratio. pH values after each addition of base were measured with a pH electrode. The data fit the Henderson–Hasselbalch equation. Curve fitting was carried out using the graphing program Kaleidagraph (Synergy Software, Reading, PA). Data were fit to the single exponential equation Y Å m0*e0m1∗x, where m0 is the extent of the pH change (DpH) and m1 is the rate constant (k). Initial rates of acidification were determined from the relationship Vi Å k∗DpH. The correlation of the fits were greater than 0.95 as determined by the least squares method.
193
Measurement of pyranine fluorescence at a pH-insensitive wavelength (Fis) allows for the assessment of changes in fluorescence unrelated to pH effects. In this manner, the Fis can be used to assess ionic strength effects as the reaction proceeds. However, in these experiments the Fis changes very little during the course of the reaction (traces B and D in Fig. 1). The slight decrease in Fis observed over the duration of the trace is likely due to photobleaching of the fluorophore over time. There were no differences in the Fis traces between the catalyzed and uncatalyzed reactions, indicating that the presence of carbonic anhydrase does not affect pyranine fluorescence. In addition pyranine has no effect on activity as determined by a Wilbur–Anderson pH experiment (data not shown). The ratio of pyranine fluorescence at a pH sensitive wavelength to fluorescence at an insensitive wavelength (Fs/Fis) produces a trace which measures changes in solution pH independent of dye concentration and small changes in ionic strength (Fig. 1, traces E and F). In addition, the Fs/Fis ratio can be used to determine pH changes as HCO30 to CO2 conversion proceeds by the relationship pH Å pKa / log
(Fs/Fis)/(Fs/Fis)max . 1 0 [(Fs/Fis)/(Fs/Fis)max]
[3]
RESULTS
Spectrofluorometric Assay for Carbonic Anhydrase Mixing 5 mM KHCO3 in pH 8.0 buffer with a pH 6.0 buffer causes a pH shift in the equilibrium from HCO30 to CO2 , a process which consumes a proton (Eq. 1). The noncatalyzed reaction proceeds slowly, eventually reaching an equilibrium determined by the final pH after mixing the two solutions. The catalyzed reaction quickly reaches the same equilibrium. In these stopped-flow rapid mixing experiments pyranine is added as a fluorescent indicator to follow pH changes as the reaction proceeds. At the pH-sensitive wavelength (Fs), pyranine fluorescence increases as the conversion of HCO30 to CO2 proceeds consuming a proton (Fig. 1, traces A and C). This rate appears to be linear in the uncatalyzed reaction but is greatly accelerated by the addition of carbonic anhydrase. The catalyzed reaction follows first-order kinetics. If the reaction is allowed to proceed to completion (Ç2 min), the final extent of the fluorescence change (Fs) is equivalent between the catalyzed and uncatalyzed reactions.
The pKa and (Fs/Fis)max are determined in a separate experiment by titrating pyranine with acid and measuring the fluorescence and the pH at each concentration of bicarbonate used (13). When a solution at pH 8.0 is rapidly mixed with a solution at pH 6.0 the final pH is attained within the mixing time of the stopped-flow instrument (Ç4 ms). In the presence of carbonic anhydrase, but absence of bicarbonate there is little change in pH over 10 s (Fig. 2). However, when bicarbonate is added the pH increases in a single exponential manner to reach a new pH equilibrium within 5 s. With increasing bicarbonate concentration up to 10 mM both the extent and the initial rate of the reaction increase. The initial rate of carbonic anhydrase activity was calculated over a wide range of enzyme concentrations (0 – 25 mg) and shown to be linear over the entire range (Fig. 3). The sensitivity of this assay is also evident as initial rates of carbonic anhydrase activity could be detected with concentrations as small as 65 ng/ml protein.
FIG. 1. Fluorescence emission trace for measurement of carbonic anhydrase activity. Changes in pyranine fluorescence after mixing 5 mM KHCO3 in pH 8.0 buffer (0.5 mM bicine-KOH) with a pH 6.0 buffer (0.5 mM Hepes-KOH) was followed by measuring fluorescent emission at 512 nm after excitation at 466 nm (Fs) or 413 nm (Fis). Conditions were (l) catalyzed reaction with 1.0 mg/ml BCA II [(A) Fs , (B) Fis , and (E) Fs/Fis] and (s) uncatalyzed reaction in absence of BCA II [(C) Fs , (D) Fis , and (F) Fs/Fis]. The uncatalyzed reaction reaches final equilibrium in about 2 min. Curve fits were determined using a least squares method.
AID
AB 2305
/
6m45$$$142
09-02-97 01:47:54
aba
194
SHINGLES AND MORONEY TABLE 1
Comparison of Kinetic Parameters for Different Sources of Carbonic Anhydrase Enzyme
Concentration (mgrml01)
Vmax (DpHrs01)
Km (mM)
Kcat (1104 s01)
BCA II HCA II HCA II HCA III HCA III
1.00 1.05 10.55 1.1 11.1
0.33 0.37 3.81 0.03 0.28
3.8 2.2 2.2 22 21
83 85 87 1.3 1.4
Note. Carbonic anhydrase was added in the amounts indicated. The initial rate of the reaction was determined as described in the legend to Fig. 2. Uncatalyzed rates were subtracted from catalyzed rates. From the plot of the initial rate vs bicarbonate concentration the Vmax and Km were determined. Kcat was determined from the Vmax and the enzyme concentration. FIG. 2. Fluorescent assay for carbonic anhydrase activity. BCA II was added at 1.0 mg/ml and bicarbonate was added at a range of concentrations from 0 to 10 mM. Change in pyranine fluorescence upon mixing of a pH 8.0 buffer with a pH 6.0 buffer was followed as described in the legend to Fig. 1. Fluorescence was converted to pH as described under Materials and Methods.
Comparison of Carbonic Anhydrase Isoforms The spectrofluorometric assay used in this study allows for the resolution of the initial rates of carbonic anhydrase activity from which kinetic parameters can be derived. BCA II and HCA II at similar enzyme concentrations (1.0 mg/ml) had similar Vmax
values (Table 1). HCA III at the same protein concentration had a velocity about 10-fold lower than HCA II. Increasing the enzyme concentrations 10-fold resulted in a corresponding increase in velocity for both isozymes (Table 1). The Km for the dehydration reaction of HCO30 could also be determined using this assay (Table 1). BCA II had a Km value of approximately 4 mM, while HCA II had a slightly lower Km value of 2.2 mM for the same reaction. HCA III had a Km value of 22 mM, 10 times larger than the Km for HCA II. Under the conditions employed in this study, with a bicarbonate concentration of 5 mM, a DpH of 1.0 unit is equivalent to a proton concentration change of 8 1 1005 mol. Knowing the protein concentration, the turnover rate (Kcat) can be determined for various isoforms of carbonic anhydrase (Table 1). BCA II and HCA II had similar turnover rates of 83 to 87 1 104 s01. HCA III had a turnover rate of 1.3 to 1.4 1 104 s01 or approximately 65 times slower than BCA II and HCA II. The determinations of Kcat could be calculated at enzyme concentrations differing by 10-fold with reproducible results. Inhibitors of Carbonic Anhydrase Activity Measured with the Fluorometric Assay
FIG. 3. Sensitivity of fluorescent carbonic anhydrase assay. Initial rates were determined as described in the legend to Fig. 1. BCA II at a concentration of 0–25 mg/ml was added with 5 mM bicarbonate.
AID
AB 2305
/
6m45$$$142
09-02-97 01:47:54
Two inhibitors of carbonic anhydrase activity have been successfully employed in this spectrofluorometric assay. Acetazolamide is a widely used inhibitor for atype carbonic anhydrases, although it is reported to have little effect on HCA III activity at low concentrations (3). Dextran sulfonamide (DBS40) is a synthesized compound in which a sulfonamide is linked to a 40,000 MW dextran making the sulfonamide membrane impermeant (20). Both acetazolamide and DBS40 have no effect on pyranine fluorescence in this assay.
aba
FLUOROMETRIC ASSAY FOR CARBONIC ANHYDRASE ACTIVITY
FIG. 4. Comparison of initial rates for two isoforms of carbonic anhydrase. Initial rates were calculated from the pH vs time plots for different concentrations of added bicarbonate. HCA II was used at a concentration of 1 mg/ml and HCA III was added at a concentration of 10 mg/ml. Rates were fit using a least mean squares method.
Acetazolamide added at a concentration of 0.35 mM inhibits the enzyme activity of HCA II (Fig. 4). The inhibited rate of HCA II approximates the uncatalyzed rate for HCO30 to CO2 conversion. By contrast, HCA III is almost completely insensitive to inhibition by this level of acetazolamide. DBS40 was shown to inhibit BCA activity as described in the next section.
195
FIG. 5. Movement of CO2 across a membrane.
A higher pH difference (alkaline inside) results in a higher initial rate of intravesicular acidification. In addition the reaction is dependent upon the presence of carbonic anhydrase on both sides of the membrane to ensure rapid acidification of the intravesicular space (Fig. 6, Table 2). Carbonic anhydrase-mediated CO2 diffusion across
Measuring Carbonic Anhydrase Activity inside Membrane Vesicles Asolectin (soybean phospholipids) was used as a model membrane to test the effectiveness of measuring carbonic anhydrase activity within membrane vesicles. Vesicles were loaded with pyranine and BCA II using previously described methods (14). DPX was added to the vesicle suspension to quench the fluorescence of any residual external pyranine. Vesicles containing 5 mM bicarbonate at pH 8.0 were added to Chamber A of the stopped-flow apparatus. Chamber B contained 10 mg/ml BCA II prepared in buffer with pH ranging between 6.0 to 7.5. Mixing these two solutions drives the hydration of HCO30 to CO2 outside the membrane vesicles as diagrammed in Fig. 5. After CO2 crosses the membrane the higher internal vesicle pH (8.0) drives the dehydration reaction of CO2 to HCO30 liberating a proton in the process. Hence, CO2 transport across the membrane can be measured by the acidification of the intravesicular space and followed by the change in pyranine fluorescence. The initial rate of this process is entirely dependent upon the pH difference imposed across the membranes (Fig. 6).
AID
AB 2305
/
6m45$$$143
09-02-97 01:47:54
FIG. 6. Measurement of CO2 diffusion across membrane vesicles. Asolectin membrane vesicles were loaded with 5 mM pyranine and 10 mg/ml BCA II where indicated. External BCA II was added at 10 mg/ml. KHCO3 was added to 5 mM to the medium and the change in pyranine fluorescence used to monitor internal pH of the vesicles. The initial internal pH was 8.0, while the external pH was varied between 6.0 and 7.5.
aba
196
SHINGLES AND MORONEY TABLE 2
Measurement of Carbonic Anhydrase-Assisted CO2 Diffusion across Membranes 0BCA Inside (DpHrs01)
/BCA inside (DpHrs01)
Additions
Asolectin vesicles
Asolectin vesicles
Chlamydomonas inner envelopes
Erythrocyte ghosts
Control / DBS40 HCO0 3 HCO0 3 / DBS40
0.02 0.04 0.03
0.02 0.37 0.05
0.03 0.41 0.10
0.13 0.72 0.14
Note. The initial rate of the reaction was determined as described under Materials and Methods. DBS40 was added at a concentration of 75 mg/ml outside the vesicles and bicarbonate was added at a concentration of 5 mM. BCA II was present outside the membrane vesicles in all experiments at a concentration of 10 mg/ml.
membranes was measured in asolectin vesicles, isolated Chlamydomonas chloroplast inner envelope vesicles, and erythrocyte membrane vesicles (Table 2). With carbonic anhydrase loaded inside the vesicles there was no significant vesicle acidification in the absence of any added HCO30 and the presence of the inhibitor (DBS40). Erythrocyte membranes had a higher control rate than asolectin or Chlamydomonas inner envelopes. When HCO30 was added outside the vesicles, the rate of vesicle acidification was low when carbonic anhydrase is absent inside the vesicles. However, when enzyme was present inside the vesicles the rate of vesicle acidification was high. When DBS40 was added with HCO30 outside the vesicles, the rate of acidification was similar to control rates in both asolectin and erythrocyte membranes. In Chlamydomonas inner envelopes the initial rate of vesicle acidification was inhibited 75% by DBS40 but the residual rate was still three times higher than control rates. DISCUSSION
Using pyranine as an indicator of the pH changes as the protonation of bicarbonate proceeds has several advantages. As a fluorescent indicator it has a high resolution for pH changes that are nearly linear in the physiological range of 6.5–8.0. Pyranine responds rapidly to changes in pH, has a pKa near neutrality, and lacks toxicity to carbonic anhydrase activity. Pyranine can be detected spectrofluorometrically at very low levels and therefore can be added in low amounts to the reaction mixture. In these experiments pyranine was typically added at concentrations of 100 nM, much less than the 27–50 mM levels of absorbance dyes used in previous studies (11, 21). This low concentration of pyranine reduces any problems that might be encountered as a result of buffering by the indicator dye itself. The sensitivity of pyranine also allows for low levels of carbonic anhydrase to be assayed. Many assays appear to be limited in the range of enzyme that can be mea-
AID
AB 2305
/
6m45$$$143
09-02-97 01:47:54
sured accurately for activity (4), as it is often difficult to measure more than a 10-fold difference in concentration. In previous studies using absorbance dyes the carbonic anhydrase concentration used was typically between 2 and 4 mM (11, 21). In this study the concentration of BCA II ranged between 0.002 and 0.75 mM with 0.03 mM giving good resolution of initial rates for kinetic studies (Figs. 3 and 4). This assay was sensitive enough to measure initial rates of HCA III activity at enzyme concentrations between 0.03 and 0.3 mM (Table 1, Fig. 4) even though HCA III is 100 times less active than HCA II (1, 2, 21). Absorbance dyes have been used in stopped-flow experiments to measure the initial rate of carbonic anhydrase activity (11, 21). The high concentration of dye and bicarbonate resulted in a buffer effect that had to be corrected in order to determine the initial rates of the reaction (Eq. 2). In the studies reported here, the initiation of the carbonic anhydrase catalyzed reaction also caused an offset in the measured fluorescence at the pH sensitive wavelength but not at the pH-insensitive wavelength (Fig. 1). Part of this offset could be due to the reaction proceeding before the first measurements are collected (Ç4 ms), but more likely it is due to the buffering effect of added bicarbonate. Pyranine is a ratiometric indicator allowing quantitative measurements to be made regardless of probe concentration and minor differences in ionic strength as would be experienced in these studies. In these experiments the pH is dependent upon the calculated pKa and the (Fs/Fis)max as described by Eq. 3. Therefore, it is important that both the pKa and the (Fs/Fis)max be determined at each concentration of bicarbonate employed to account for buffering effects. In general, the pH increased with a buffering constant of 0.031 pH units/ mM bicarbonate. Many enzyme inhibitors may be fluorescent themselves or have an effect on the fluorophore such as fluorescent quenching or enhancement. Known inhibitors of carbonic anhydrase activity were used in this study
aba
FLUOROMETRIC ASSAY FOR CARBONIC ANHYDRASE ACTIVITY
to investigate if they could be used as effective inhibitors in a fluorescent assay system. Acetazolamide inhibited HCA II activity very effectively but did not inhibit HCA III activity (Fig. 4), in agreement with previous inhibition studies (3). Carbonic anhydrase is involved in the hydration of CO2 to HCO30 inside erythrocyte membranes (22) and is also involved in the interconversion of HCO30 and CO2 in chloroplasts (16). Isolation of membranes from these organelles and cells allows for the study of carbonic anhydrase-assisted CO2 diffusion across membrane vesicles using this fluorescent assay. In the absence of enzyme the rate of diffusion of CO2 through membranes is limited by the slow hydration–dehydration reaction of CO2 (Fig. 5, Table 2). When adequate levels of carbonic anhydrase are present on both sides of the membrane, the hydration–dehydration reaction is so fast that CO2 diffusion becomes rate limiting (23). Under these conditions, the permeability of membranes to CO2 diffusion can be assessed. Using the fluorescent assay it was possible to measure CO2-dependent pH changes in artificial vesicles as well as chloroplast inner envelope membranes and red cell plasma membranes (Fig. 6, Table 2). Potentially, the presence of membrane-associated carbonic anhydrase or bicarbonate transporter could be detected using inhibitors or by comparing pH changes in the presence and absence of enzyme inside or outside the vesicles. By comparison to other methods using pH indicators (11, 21), the use of pyranine as a fluorescent pH indicator is much more sensitive for detection of carbonic anhydrase activity. This assay is useful for measuring carbonic anhydrase reactions at physiological temperatures and pH ranges and can be applied to measure CO2 diffusion across membranes. Finally, although this study has concentrated on the use of stopped-flow techniques to measure the initial rate kinetics of carbonic anhydrase, it is also possible to follow the reaction in a cuvette assay with any spectrofluorometer that has a reasonable time drive resolution to at least qualitatively assess enzyme activity. ACKNOWLEDGMENTS We would like to thank Zygmunt Schroeder for the construction of the small-volume extrusion apparatus and Catherine Mason for
AID
AB 2305
/
6m45$$$143
09-02-97 01:47:54
197
preparation of Chlamydomonas chloroplast envelopes. We also thank David Husic (Lafayette College, Easton, PA) for his generous gift of DBS40 and David Silverman (University of Florida, Gainesville, FL) for his gift of HCA III. We are grateful to Richard E. McCarty for helpful discussions and critical reading of the manuscript.
REFERENCES 1. Khalifah, R. G., and Silverman, D. N. (1991) in The Carbonic Anhydrases: Cellular Physiology and Molecular Genetics (Dodgson, S. J., Tashian, R. E., Gros, G., and Carter, N. D., Eds.), pp. 49–70, Plenum, New York. 2. Silverman, D. N. (1991) Can. J. Bot. 69, 1070–1078. 3. Dodgson, S. J. (1991) in The Carbonic Anhydrases: Cellular Physiology and Molecular Genetics (Dodgson, S. J., Tashian, R. E., Gros, G., and Carter, N. D., Eds.), pp. 49–70, Plenum, New York. 4. Forster, R. E. (1991) in The Carbonic Anhydrases: Cellular Physiology and Molecular Genetics (Dodgson, S. J., Tashian, R. E., Gros, G., and Carter, N. D., Eds.), pp. 79–98, Plenum, New York. 5. Wilbur, K. M., and Anderson, N. G. (1948) J. Biol. Chem. 176, 147–154. 6. Hock, G., and Kok, B. (1963) Arch. Biochem. Biophys. 101, 160– 170. 7. Stemler, A. (1993) Anal. Biochem. 210, 328–331. 8. Pocker, Y., and Stone, J. T. (1967) Biochemistry 6, 668–678. 9. Abou-Rebyeh, H., Dietsch, P., and Korber, F. (199) in Carbonic Anhydrase (Botre, F., Gros, G., and Storey, B. T., Eds.), pp. 435– 439, VCH Press, New York. 10. Gibson, Q. H., and Milnes, L. (1964) Biochem. J. 91, 161–171. 11. Khalifah, R. G. (1971) J. Biol. Chem. 246, 2561–2573. 12. Li, H-N., and Ci, Y-X. (1995) Anal. Chim. Acta 317, 353–357. 13. Shingles, R., and McCarty, R. E. (1994) Plant Physiol. 106, 731– 737. 14. Shingles, R., and McCarty, R. E. (1995) Anal. Biochem. 229, 92– 98. 15. Sueoka, N. (1960) Proc. Natl. Acad. Sci. USA 46, 83–91. 16. Moroney, J. V., and Mason, C. B. (1991) Can. J. Bot. 69, 1017– 1024. 17. Clemetson, J. M., and Boschetti, A. (1988) Biochim. Biophys. Acta 943, 371–374. 18. Ramazanov, Z., Mason, C. B., Geraghty, A. M., Spalding, M. H., and Moroney, J. V. (1993) Plant Physiol. 101, 1195–1199. 19. Steck, T. L., and Kant, J. A. (1974) Methods Enzymol. 31, 172– 177. 20. Husic, H. D. (1991) Can. J. Bot. 69, 1079–1087. 21. Rowlett, R. S., Gargiulo, N. J., Santoli, F. A., Jackson, J. M., and Corbett, A. H. (1991) J. Biol. Chem. 266, 933–941. 22. Hladky, S. B., and Rink, T. J. (1977) in Membrane Transport in Red Cells (Ellory, J. C., and Lew, V. L., Eds.), pp. 115–135. 23. Gutknecht, J., Bisson, M. A., and Tosteson, F. C. (1977) J. Gen. Physiol. 69, 759–794.
aba