The Effect of Bile Salts on Carbonic Anhydrase DAVIDE. MILOV,~ WOU-SEOK JOU,' RACHEL B. SHIREhL4N3 AND PAUL w.CHUN' 'Department of Biochemistry and Molecular Biology, 'Department of Pediatrics, Division of Gastroenterology, College of Medicine, Ihiuersity of Florida, Gainesuille, Florida, 32610; and 3Department of Food Science and Human Nutrition, Institute of Food and Agricultural Sciences, University of Florida, Gainesuille, Florida 32611.
consisting of 260 and 261 amino-acid residues (1,2). A zinc ion, bound to three histidyl residues, is located at the base of the enzyme's 15 A deep active site crevice (1). Carbonic anhydrase catalyzes reversible CO, hydration reactions, a process that involves the following: (a) a proton transfer reaction from the zinc-bound water molecule as a result of the lowering of the pKa values of bound H,O and (b) a ligand exchange reaction involving the interconversion between CO, and HC0,-. The catalytic efficiency of this enzyme depends on the efficiency of the proton transfer reaction, which might limit the CO, hydration reaction (3, 4). The enterohepatic circulation of bile salts is a dynamic biochemical process involving hepatocytes, canaliculi and bile ductules as concentrated bile is delivered to the intestinal lumen for fat emulsification. Bicarbonate ions are also crucial to bile flow. For example, enhanced canalicular bicarbonate secretion accounts, in part, for properties of certain bile salt-independent flow inducers (secretin, ~-(2,4-dimethoxy-5-cyclohexylbenzoy1)-propionic acid or SC-2644, salicylate) and certain bile salt-dependent flow inducers (norurodesoxycholate acid and ketolithocholic acid) (5-10). Further, acetazolamideinhibition studies of hepatocytes in culture suggest a role for carbonic anhydraseproduced biocarbonate ions in hepatocyte gluconeogenesis, fatty acid synthesis, urea metabolism and electrolyte transport (9-1 1). Despite the proven and putative roles of carbonic anhydrase in hepatic physiological processes, the effect of bile salt on the enzyme's activity is not known. We speculated that increased tissue concentration of bile salts, as seen in cholestatic TOLOGY 1992;15:288-296.) liver disease, would affect bicarbonate availability by altering HCA activity. In this article, we report on the effect of the bile salt Three isoenzymes of human carbonic anhydrase (HCA) have been described (designated HCA-I, HCA-I1 anion on carbonic anhydrase activity, applying and HCA-111). HCA exists as a single polypeptide chain stopped-flow kinetic measurements to determine the inhibition constant (K.,) and Michaelis-Menten dissociation constant (KJ for HCA-I, HCA-I1 and bovine carbonic anhydrase (BCA) in the presence of four Received November 28, 1990; accepted September 16, 1991. distinct bile-salt molecules. The effect of acetazolamide This work was supported by NSF Grant DMB 83-12101(02),agrant from the Southern Medical ,4ssociation to David E. Milov and in part by a grant from The on bile salt-carbonic anhydrase binding is also reported. Center for Neurobiology of Aging at the University of Florida. We also report the molecular dimensions of the bile Address reprint requests to: P. W. Chun, Department of Biochemistry and salt-carbonic anhydrase complex deduced through Molecular Biology, College of Medicine, Box 100245, J.H.M.H.C., University of scanning molecular sieve chromatography and sedimenFlorida, Gainesville, FL 32610-0245. tation equilibrium measurements. 3111133988
Bile salts are potent inhibitors of bovine carbonic anhydrase and human carbonic anhydrase I and human carbonic anhydrase 11. To further characterize the binding of bile salts to carbonic anhydrase, rate constants for the CO, hydration reaction in the presence of deoxycholate, cholate, glycocholate and taurocholate were determined using stop-flow experiments. Values for the Michaelis-Menton dissociation constant for bovine carbonic anhydrase, human carbonic anhydrase I and human carbonic anhydrase I1 were found to be 5.2,9.2 and 13.2 mmol/L, respectively. The inhibition constant values for the various bile salts tested ranged from 0.1 to 1 mmolb for bovine carbonic anhydrase, 1.6 to 2.4 mmol/L for human carbonic anhydrase I and 0.09 to 0.7 mmol/L for human carbonic anhydrase 11. Our results suggest a mechanism of noncompetitive carbonic anhydrase inhibition for bile salts. Bile-salt binding to carbonic anhydrases as measured by scanning molecular sieve chromatography resulted in an increase in partition radius, molecular volume and surface area. The partition radius increased from 24 A to 28 A in the presence of 2.5 mmol/L sodium deoxycholate at critical micelle concentration. As determined by sedimentation equilibrium measurements, approximately 1gm of carbonic anhydrase will bind 0.03 gm of deoxycholate, suggesting three to six binding sites for bile salt on the carbonic anhydrase molecule. The conformational changes and inhibition of carbonic anhydrases resulting from bile-salt binding may be important to the regulation of enzymatic activity in tissues along the enterohepatic circulation; by limiting bicarbonate availability this interaction may also contribute to the metabolic derangements seen in patients with cholestatic liver disease. (HEPA-
288
Vol. 15, No. 2, 1992
MATERIAL AND METHODS Bile Salts and Enzyme Preparations. Sodium salts of cholic acid, deoxycholic acid, taurocholic acid and glycocholic acid were obtained from Calbiochem, Behring Diagnostics (La Jolla, CA). Bovine erythrocyte carbonic anhydrase (activity 2,800 Wilber Anderson units/mg solid), p-nitrophenyl acetate or p-nitrophenol, HEPES, MES (2-[Nmorpholinolethanesulfonic acid) and MOPS (34Nmorpholino]propanesulfonic acid) were purchased from Sigma Chemical Co. (St. Louis, MO). HCA-I and HCA-11, isolated by an affinity chromatography method (12), were a gift from Dr. David Silverman, University of Florida, (Gainesville, FL). Kinetic Methods. Saturated solutions of CO, were made by bubbling CO, into water a t 25" C; dilutions were prepared in the absence of air by coupling two syringes. CO, concentrations were based on the solubility of 33.97 mmol/L for CO, in water at 25" C (13, 14). All experiments were performed at 25" C. Moreover, all solutions contained 33.3 mmol/L Na,SO, to contribute an ionic strength of 0.1. All kinetic determinations of initial velocity of the CO, hydration reaction were carried out on a Durrum-Gibson stopped-flow spectrophotometer (Model D-110) interfaced with North Star microprocessor (Olis Instruments Model 360). The initial velocity was calculated using the equation v
= -
289
BILE SALTS AND CARBONIC ANHYDRASE
(d[CO,l/dt),,,,,
=
- Qo(dAldt),,t,a,
with (dA/dt) determined by least-square analysis of absorbance vs. time. The buffer factors Q, relating changes in absorption to changes in [H'] were calculated by the method of Khalifah (15) for the buffer indicator pairs listed here. Also given are pKa values, wavelengths used and differences in molar extension coefficients of acidic and basic forms: HEPES (pKa = 7.5) with p-nitrophenol (pKa = 7.1, A = 400 nm, Ae = 1.83 x mol/L-l cm-l) or with phenol red (pKa = 7.5, A = 557 nm, A€ = 550 x lo-* mol/L-l cm-l); MES (pKa = 6.1) with chlorophenol red (pKa = 6.3, A = 574 nm, AE = 1.77 x mo1L-l cm-l); and MOPS (pKa = 7.1) with p-nitrophenol (pKa = 7.1, A = 400 nm, Ae = 1.83 x l o p 4M-' cm-'). The buffer concentration in all kinetic measurements was 20 to 30 mmol/L (16). For every kinetic run, experimentally determined uncatalyzed rates were subtracted to give catalyzed rates. Evaluation of kinetic constants from initial velocities was conducted using the weighted least squares method of Wilkinson (17) and Cleland (18). The I,, (a 50% decrease in enzyme activity, ie., I,, = [actlEIZ) was determined by a previously published method that directly measured the rate of disappearance of O'*-labeled CO, substrate under equilibrium conditions (19,20). BCA activity in the presence of bile salts in HEPES (adjusted to pH 7.61 or 7.19) was also examined by a previously described spectrophotometric method that measured the rate ofp-nitrophenol appearance (348 nm) from the hydrolysis of p-nitrophenyl acetate (21). Scanning Molecular Sieve Chromatogruphy (22-25). All scanning molecular sieve chromatography (SMSC) experiments were conducted using a gel column scanning system that was essentially the same as that described elsewhere with the exceptions of the monochronometer and light source module, which were Heath EU-700 and EU 701-50. The basic operating routine consisted of moving a quartz column (24 cm x 0.90 cm) packed with a suitable gel (i.e., one with minimum scattering) through a horizontally collimated beam of monochromatic light (visible or UV)at a constant rate of 0.19 cmJsecfor 98 sec, during which time transmittance was detected by Aminco 10/267 solid state, blank subtract photo
multiplier microphotometer through an end-on photomultiplier tube (RCA 6903 for U V or RCA C7164 for visible). The direct W gel column scanner built in our laboratory and the stopped-flow (Durrum/Dionex) system were interfaced with a Northstar Horizon Z80-A microprocessor, a Sanyo DM50/2CX video display unit and an Integral Data Systems model 460 printer/plotter. All the data presented here were obtained using a record length of 300 data points. The partitioning properties of varying concentrations of HCA-I, HCA-I1 and BCA in the presence of varying concentrations of bile salt and the partitioning properties of sodium deoxycholate (DOC) itself in Sephadex G-75 (Pharmacia, Piscataway, NJ, 100 to 200 mesh, exclusion limit approximately 7 x lo4), were determined in small zone experiments at 280 nm at 25" C. In this type of experiment, a sample solution at the desired concentration was added continuously to the column until a reproducible baseline was obtained. The column was scanned at regular intervals as the solutionholvent boundary moved through the gel matrix a t a constant flow of 4.30 ml/hr. The baseline records of successive scans were subtracted from each other, yielding difference profiles. The problem of locating the centroid position of each boundary was then reduced t o the simple task of locating the peak position of these difference profiles (22-25). Partitioning calibration parameters for a given column were obtained by determining the rate of movement of a series of sample macromolecules of known molecular radius. The partition coefficient was calculated from u=
(dtidx), - dt/dx), (dt/dx), - (dt/dxl0
where (dtldx) was the slope of a plot of time vs. centroid position for a given sample marker, (dtldx,) was the slope of the void volume marker and (dt/dxJiwas the slope of the internal volume marker. The molecular partition radius of the solute was calculated using the following linear expression: a, = a,
+ b,erfc-'ui
where a, and b, were calibration constants for a set of particles of given radii in a given gel and determined independently and b, was multiplied by the error function complement inverse erfc-l of the partition coefficient u, (22-25). All of the calibration markers -BSA, ovalbumin (OVA), horse cytochrome C and aldolase (r = 34, 31, 17 and 48 A, respectively) -were obtained from Sigma Chemical Co. Sedimentation Equilibrium Measurements. All sedimentation equilibrium experiments were performed with a Beckman Model E analytical ultracentrifuge (Beckman Instruments, Fullerton, CA) equipped with a rotor temperature indicator control unit at 20" C, using a Yphantis (26) three-channel cell with a carbon-filled epoxy centerpiece in an An-D rotor at 12,590 rpm. The calculation of the molecular weight distribution and (dlnC/dX2) was based on the Yphantis method (261, using a computer program that could be modified for use with an Amdahl470 v/6 I1 unit (modified IBM 6000/1800, University of Florida's CIRCA computing facilities), plotting 1nJ or 1nC as a function of X2.The In C vs. X2 data were fitted to a least-squares polynomial, and the values of (dlnC/dX2) were calculated by a modification of the sliding three-point leastsquares quadratic treatment of Yphantis. The preferential interaction term, which in this case was a measure of the degree of sodium DOC binding to BCA, may be evaluated for a three-component system at sedimentation
290
MILOV ET AL.
HEPATOLOGY
TABLE 1. Comparative CO, hydration activities of mammalian carbonic anhydrase isozymes"
HCA-I HCA-I1 BCA (11)
6.33 85.73 0.08
32.6 787.9 1.15
0.54 13.00 1.73
5.15 9.19 13.20
"Temperature = 25" C; pH 7.19 (HEPES buffer), Calculations are based on Michaelis-Menten equation assuming an endogenous CO, concentration of less than 1mmol/L and enzyme concentrations of 16.5 pmol/L, 16.2 pmol/L and 16.5 pmoUL, respectively for isozyme, HCA-I, HCA-I1 and BCA. The k,,, and l i , data for the human isozymes I and I1 have been reported by other researchers. (34, 51, 52).
TABLE 2. Comparative CO, hydration activities inhibited by mammalian carbonic anhydrase isozymes"
Cholic acid (Na salt) Deoxycholic acid (Na salt) Glycocholic acid (Na salt) Taurocholic acid (Na salt)
HCA-I
4 (mmoUL) uncompetitive
& (mmol/L) noncompetitive
Inhibitor type
-
2.600 2.430
1.010 2.120 -
The standard error estimate of the coefficients at these concentrations Na salt of DOC
Cholic acid
[I], mmol/L n 0.5 1.0 0 1.5 A 2.0
0.02420 0.18225 0.04818 0.02389
0.02686 0.21831 0.07336 0.12337
+
~
~
_
_
_
_
Glycocolate
Taurocholic acid
0.01030 0.16879 0.04183 0.02075
0.02190 0.22875 0.03110 0.15386
_
In the HCA-I standard without inhibitor present, the standard error estimate of coefficient was 0.01233. "Kvalues for HCA-I, HCA-I1 and BCA were found to be 5.16, 9.22 and 13.19, respectively. K, = [I]/(ACIDconst/STANDARDconst-1) calculated from ([EIN,,) (1 + [I]/KJ. Temperature = 25" C; pH 7.19 (HEPES buffer). Calculations were based on noncomparative inhibition and inhibitor concentrations of each inhibitor run between 0.5 mmol/L and 2.5 mmol/L. CO, concentration ranged from 2.0 mmol/L to 17.0 mmoUL for all experiments. Values in parentheses obtained from 0"-exchange experiments (31).
equilibrium from the concentration distribution of BCA (CJ, grams of sodium DOC preferentially bound per gram of BCA, as described by the expression Z,, was given by the relationship M, (dpidC,)
=
(dlnC,/dX2)2RT/w2
(1)
where M, is the molecular weight of BCA, (ap/dC,) is the change in solution density with changing concentration of BCA,(alnC,laX2) is the slope of a plot of In C, vs. X2,X i s the radial distance of each fringe from the center of rotation, R is the gas constant in J (moles*K)-l, T is the absolute temperature and o is the angular velocity in radian/sec. The extent of the preferential interaction, which was derived from the three-component theory (27-29) in terms of the number of
(ap/ac,)
=
(1- VZPJ
+ X,(l
-
V,PJ
(2)
in which p, was the solvent density and Vz and ii, were the true partial specific volumes of BCA and sodium DOC, respectively. An alternative approach to determining Ti, from In C vs. X2 plots was based on the assumption that any change in the slope (alnC/aX2)of this plot for a solute in the absence of iigand to that for a solute in the presence of ligand was related to the binding of that ligand.
Vol. 15, No. 2, 1992
BILE SALTS AND CARBONIC ANHYDRASE
291
Avg [KI] = 2.602 bM1, [El = 16.5 IPM]
110
100 55 90 50
>
80
ID
(0
w,
w,
-
6 70
45
60 40 50
35 40
30
I'
30
0
20
40
A
60
80
100
0
120
11c02,
"
l , , , j , , , l , , , I , , , I / / ,
20
60
40
0
80
100
120
11C02,1 M
Avg [K,] = 2.430 ImM1, [El = 16.5 [W1
Avg [Kll= 2.200 lmM1, LEI = 16.5 Iml
110
60
100
55 90
>
50
1 80
0
ID
! - 45 w,
70 I
60
50 40
0 C
20
30
" " " " " '
30 1 " " ' " " " ' 40
60
80
100
lic02,Wl
120
0
20
40
60
D
80
100
120
11C02,l[M]
FIG.1.Double-reciprocalplot showing the inhibition by bile salts of the CO, hydration reaction catalyzed by HCA I in 5 mmoUL HEPES buffer, pH 7.16, at 25" C containing 5 x mol/L nitrophenol. Four different concentrations of four inhibitors (Z] are represented as follows: (A) Cholic acid (Na salt). (B) DOC (Na salt). (C) Glycocolate (Na salt). (D) Taurocholic acid (Na salt).
The effective specific volume,
was determined from
c3
Positive values of indicated preferential binding of sodium DOC, whereas negative values indicated preferential hydration 0', = (V, + ii3?J/(l+ ii,) (3) where p, and pOz were the solvent densities for solvent with and without ligand, respectively. and the preferential interaction term (26, 27) was evaluated Once the true values of molecular weight and v were known, from the radius of an equivalent sphere, r,, was calculated. The ratio ss = (V,O poz - O'OpOc)/(l- V3P0) (4) of the radius determined by SMSC over r, was equivalent to @Ic,
HEPATOLOGY
MILOV ET AL.
292
B
1
C
D
FIG.2. Three-dimensional plot of BCA in 0.041 moliL K-PO, buffer, pH 7.6,on a Sephadex G-75 column. Ten scans were made at 2-min intervals, with the first taken at 2 min. The flow rate was 7.8 mlihr. Samples were equilibrated with 0.041 mol/L K-PO, buffer, pH 7.6, and 0.05 ml of O.D,,,, = 1.253. (A) BCA, no DOC added. (B) BCA in 0.5mmoliL DOC. (C) BCA in 2.5 mmoliL DOC. (D) BCA in 5.0 mmol/L DOC.
the frictional ratio of the particle and was a relative measure of particle asymmetry. For frictional ratios greater than 1.0, the particle was considered ellipsoidal (either prolate or oblate) or even rodlike, and the axial ratio a/b was determined from f/f, according to Van Holde (30).
Because Equation 5 represented only a single binding site rather than preferential interactions at multiple RESULTS sites, it could only provide an approximate value of the Stopped-flow Kinetics Measurements. Comparison of noncompetitive inhibition constants. Noncompetitive and unthe turnover number kcatfor the hydration of CO, and inhibition was characterized by &'lope = yint K, for HCA-I, HCA-I1 and BCA at pH 7.16 is shown in competitive inhibition by a value for K F , which was Table 1. All experiments were conducted in HEPES very large compared with K F . In the pH range of 7.19 buffer adjusted to pH 7.61 or 7.19. The kcatfor HCA-I1 to 7.61, Kiintwas similar in magnitude to K?lope, and the was approximately 30-fold greater than that of HCA-I or data appeared to support noncompetitive inhibition in (1 + BCA under these experimental conditions. The K, value double reciprocal plots. The values of [E]/V,, for HCA-I was about that of HCA-I1 and BCA, which [I]/Kp) were used to evaluate Ki. The kcatand kobswere evaluated based on the following: kc, = V,, [El.,. and were of similar magnitude. The inhibition by bile salts of the hydration of CO, kobs= Ckc,,/K,1 [El,. catalyzed by HCA-I, HCA-I1 and BCA at pH 7.19 was Table 2 shows the inhibition constants for HCA-I, HCA-I1 and BCA (at a concentration of 16.5 pmol/L), evaluated using the following expression:
Vol. 15, No. 2, 1992
BILE SALTS AND CARBONIC A N H Y D W E
293
TABLE 3. Partition radius vs. concentration of BCA in 0.04 moln K-PO, buffer, pH 7.6, plus varying concentrations of sodium DOC
0
0.5
2.5
11.93
24.05 24.07 24.15 24.30 27.13 27.25 27.90 27.90 27.50 27.50 28.10 27.30
25 50 100 200 25 50 100 200 25 50 100 200
0.725
29.9 30.0 30.9 30.5 31.3 31.2 31.5 31.4 31.5 31.1 31.5 31.3
MT=4 TN (radius x 10-'Y/3 5,,assuming globular spherical particle. Radii were calculated based on a linear fit of a vs. u of error function complement of inverse sigma (i.e., a = a, + b,erfc- ui,where a, = 11.83 and b, = 34. Data shown for BCAin buffer alone represent the average of three determinations. S.D. in the fitted radii is 0.26 A in each case (Sephadex G-75).
.,
TABLE 4. Molecular size of BCA and BCA, 2.0 mmol/L sodium DOC complexes Sample
dlnCidX2
BCA (11) BCA (11)
0.62612 0.68322
(ml/gm)
0.725 -
(gm NaDOClgm BCA)
0.030
5s
Whole cell M7
Radius, (A)
0.034
28,486 2 4,500 31,346 c 4,500
24.3 27.8
The whole cell weight average molecular weights and the effective specific volume of BCA, 1 mmol/L NaDOC, were determined by In C vs. X2 plots as previously described, sedimentation equilibrum runs in pH 7.4 PO, buffer of 1 mmol/L sodium DOC buffer at 12,590 rpm at 20" C. The radius is that of an equivalent sphere, calculated from the whole cell weight average molecular weight. Concentration of BCA was approximately 0.8 mgiml in each case. The binding was calculated from equation 3, using V3 = 0.779 mVgm (for NaDOC). The partial specific volume of BCA was taken from Ref 34 and that for BCA, 1 mmol/L NaDOC, was calculated from equation 4.
where CO, concentration varied from 5.1 to 17 mmol/L in the presence of various bile salts. The inhibitor concentration varied from 0.5 mmol/L to 2.5 mmol/L. Although the inhibitory effect was pronounced in HCA-I1and BCA, the effect of these bile salts was greatly reduced in HCA-I. In each case, cholic acid showed the greatest inhibiting effect on these enzymes, where K, for HCA-I1 and BCA was about 0.1 mmol/L and 2.6 mmol/L for HCA-I. The double reciprocal plots for HCA-I and HCA-I1 in the presence of acetazolamide showed Ki values for HCA-I and HCA-I1 of 0.156 and 0.011 kmol/L, respectively. The Ki value for HCA-I1 in the presence of both acetazolamide and cholic acid was 1.02 p.mol/L, a considerable reduction in the inhibitory activity of acetazolamide in the presence of cholic acid. Our stopped-flow obtained Ki values for HCA-I1 in the presence of acetazolamide were in excellent agreement with those obtained by the O'"-exchange method (31). Results obtained by the Ol'-exchange method and plots shown in Figures 1A and 1D indicated that the inhibition of CO, hydration by bile salts was noncompetitive, whereas the cases shown in Figures 1B and 1C were hybrids of uncompetitive and noncompetitive. Inhibition was incomplete in all cases.
The effects of conjugated and unconjugated bile salts on HCA-I1 as determined using the O1"-exchange method are shown in Table 2. The values were quite consistent with those obtained by stop-flow kinetics; however, the scatter of the data makes evaluation of K, based on Durrum stop-flow experiments difficult. Scanning Molecular Sieve Chromatography. Representative small zone profiles of BCA in 0.04 mol/L K-PO, buffer, pH 7.6, on Sephadex G75 at 280 nm containing various concentrations of sodium DOC are illustrated in Figures 2 and 3. The time-dependent centroid movements of varying concentrations of sodium DOC were used to calculate the partition radii of these solutes, based on the centroid movements for column calibration markers as previously described (Table 3). A plot of sodium DOC radius vs. concentration was used to estimate the upper limit of the critical micelle concentration (CMC) of sodium DOC as 2.5 mmol/L. However, although the micelle transition was fairly sharp, no data points were taken between 0.5 and 2.5 mmol/L sodium DOC, indicating that the viscometry value was 1.5 mmol/L in this buffer (Chun, et al., Unpublished observations, 1991). Also, an upper limit value of 21 t 0.5 A was found for the stable sodium DOC micelle in this buffer.
294
MILOV ET AL.
HEPATOLOGY
0.041M PO&, pH 7 . 6
0.5 mn DOCIPOI, 2 . 5 ni-l DOC/PO& 5.0 mfl DOC/PO4
FIG.3. Influence of sodium DOC binding to BCA as examined by direct gel column scanning. Samples in three different concentrations of DOC were scanned on a Sephadex G-75 column in phosphate buffer, pH 7.6. Each experiment consisted of 10 scans and 300 data points/scan. The sixth and seventh scans of each run are shown in the composite figure. By this means, the degree of binding of BCA as influenced by DOC may be detected by changes in particle radius. Using this scanning technique, size changes of 2 to 3 A in radius can be detected analytically. Binding of bile salt to carbonic anhydrase is most pronounced above the CMC.
Table 3 indicates a sodium DOC and BCA concentration dependence on the processes of sodium DOC binding and increase in molecular volume and surface area of the molecule. It was not unreasonable to assume that BCA particles in the presence of 2 mmol/L DOC (below CMC) were swollen, with approximately a 15% increase in radius caused by the binding of sodium DOC monomers. The percent increase in surface volume and surface area was calculated to be 59% and 36%, respectively. This increase in partition radius to 28 A in the presence of DOC remained relatively constant below CMC. (Table 3). When acetazolamide in the presence of 2 mmol/L DOC was incubated with BCA, no noticeable change was seen in the partition radius despite the distinct decrease in inhibitory activity, suggesting the possible formation of an inclusion compound or a significant degree of interaction between DOC and acetazolamide. This finding would be consistent with the observation of De Sanctis (32), who noted that DOC forms inclusion compounds with phenanthrene and (El-p-dimethylaminoazobenzene.Jou and Chun (33) have recently demonstrated, in the simulation of the molecular mechanics of cholic acid micelle formation, that sandwiched multilayer channels are formed that may incorporate guest molecules into the micelle. Other globular proteins exposed to DOC also demonstrated increased partition radius in the SMSC system. We examined the effect of DOC on a number of globular proteins such as BSA, OVA, horse cytochrome C, human low-density lipoprotein (24, 25) and uteroferrin under similar conditions (31). We observed an increase in the partition radius of the following: BSA = 34 to 38 A (0.04 gm DOC/gm protein); OVA = 31 to 33 (0.01 gm DOC/gm protein); horse cytochrome C = no change; uteroferrin = slight change (from 29 to 31 A); and human low-density lipoprotein = 110 to 125 A (0.04 gm DOC/gm protein). This increase in partition radius in
1
I
I
FIG.4. Proposed model for bile-salt inhibition of carbonic anhydrase in the periportal bile canaliculi (the concentration of bile salt in the canaliculus is 5 to 10 mmoUL). Binding of bile salt to carbonic anhydrase may regulate the availability of bicarbonate throughout tissues along the enterohepatic circulation.
these proteins indicated a change in conformation, which was probably caused by partial denaturation that occurred in the presence of bile salts. Such volume changes were not the mere additive dimension of three to six bile-salt molecules. Preferential Binding of Sodium DOC. The 1nC vs. X2 plot of BCA in pH 7.6 buffer revealed no curvature, indicating that very little concentration-dependent selfassociation of BCA particles occurred at 20" C. The whole cell weight average molecular weight calculated from this plot was found to be 29,000 & 4,500 gmlmol (Table 4). Using 8, = 0.725 (341, the radius of an equivalent sphere of that weight would be 24.1 ? 0.5 A, agreeing very closely with SMSC data. The ratio of these radii was equivalent to the frictional ratio and was used to calculate the axial ratio for a prolate ellipsoid as 1.15. The dlnC/dX2 values were used in equations 2 and 3 (27-29; 35, 36) to estimate x3 of the BCA, 1 mmol/L sodium DOC complex as 0.03 gm of sodium DOC bound/= of BCA, resulting in a O r , = 0.747 ? 0.001 mVgm and a preferential interaction term of 6, = + 0.036, indicating preferential binding of sodium DOC. The molecular weight was reevaluated to be 32.5 5 0.01 x lo3 gm/mol. Using O'c = 0.747 ml/gm with this whole cell weight average molecular weight gave the radius of an equivalent sphere as 28.1 ? 0.07 A (see Table 4). A comparison of this radius with that from SMSC indicated an axial ratio of 1.17 for a prolate ellipsoid. Although these radii were determined in two different systems, in which the effects on sodium DOC micelles were apparent, it was assumed that the radii of the BCA, 1mmolb sodium DOC complex would change very little from 2.5 mmol/L to 1.0 mmol/L because only sodium DOC monomers are present.
Vol. 15, No. 2, 1992
295
BILE SALTS AND CARBONIC ANHYDRASE
DISCUSSION
Our results demonstrated bile salts to be potent and rapid inhibitors of in uitro CO, hydrase activity for three carbonic anhydrase isoenzymes. Cholate in particular was the most potent, naturally occurring inhibitor of carbonic anhydrase yet reported, although it has significantly less inhibitory activity than acetazolamide (37, 38). In contrast to the noncompetitive inhibition kinetics observed with acetazolamide and carbonic anhydrase, we found bile salt inhibition of HCA-I and HCA-I1 to be both noncompetitive and uncompetitive, suggesting a distinct inhibition mechanism. The sedimentation equilibrium measurements of the bile-saltbinding to BCA suggest three to six binding sites per molecule of BCA. This is in close agreement with the amount of DOC that binds to human serum albumin, the major bile salt transport protein in human beings (39). Although anion inhibition of carbonic anhydrase occurs through complexes at the active site crevice (401, bile-salt binding must occur at sites in addition t o the active site. The SCMC data showed that sodium DOC induces the swelling of the globular carbonic anhydrase molecule. This may represent a denaturing effect. We hypothesize that bile-salt binding to carbonic anhydrase results in increased molecular volume, or swelling, that inhibits the effectiveness of the crucial proton transfer reaction occurring at the carbonic anhydrase active site during CO, hydration (41,42). A second possibility is that such molecular swelling alters the important conformational relationship between His64, His67 and His200 as shown on x-ray crystallographic data (43, 44). Our studies predict strong inhibition of HCA-I and HCA-I1 in tissues with conjugated or unconjugated bile salt concentration above 2 mmol/L. The physiological relevance of our in uitro studies may be particularly important to patients with cholestatic liver disease. For example, in isolated, perfused rat liver, the addition of acetazolamide results in hyperammonemia; the mechanism involves depletion of the HCO; substrate required for the first enzymes of ureagenesis from NH, and carbamoyl-phosphate synthetase I. (21, 45, 46). Conceivably, hyperammonemia could result from hepatic mitochondrial carbonic anhydrase inhibition during intrahepatic cholestasis. Liver perfusion studies have also shown diminished bicarbonate production during metabolic acidosis (47); this may result in enhanced bile-salt binding to carbonic anhydrase at low pH. This observation may relate to our finding that bile salts bind more avidly to carbonic anhydrase at pH 7.19 than at pH 7.61. At the level of the bile ductule, bile-salt secretion has been shown to require HCO; production because acetazolamide and proton pump inhibitor (N,N'dicyclohexylcarbodiimide) inhibit biliary secretion of HCO 3 (47,481.As such, we speculate that bile salts may, in part, regulate the rate of their own secretion through the feedback inhibition of carbonic anhydrase. Although the precise molar concentrations of bile salts along the
enterohepatic circulation are not known, it can be assumed that carbonic anhydrase inhibition would be most evident at the canaliculus and bile ductule (Fig. 4). Bile-salt inhibition of carbonic anhydrase may be important in other disease states. For example, a recent study has shown that physiological concentrations of bile salts inhibit HCA isolated from gastric mucosa; the authors suggest this to be contributory to gastritis after partial gastrectomy (49, 50).
Acknowledgments: We thank Drs. Silverman and Tu for obtaining I,, values of carbonic anhydrase and for providing the samples of carbonic anhydrase I and 11. REFERENCES 1. Linskog S, Coleman JE. The catalytic mechanism of carbonic anhydrase. Proc Natl Aead Sci USA 1973;70:2505-2508. 2. Deutch HF. Carbonic anhydrases. Int J Biochem 1987;19:101-103. 3. Maren TH, Sanyal G. The activity of sulfonamides and anions against the carbonic anhydrases of animals, plants and bacteria. Ann Rev Pharmacol Toxicol 1983;23:439-459. 4. Pocker Y, Tanaka N. Inhibition of carbonic anhydrase by anions in the carbon dioxide-bicarbonate system. Science 1978;199: 907-909. 5. Hardison WGM, Wood CA. Importance of bicarbonate in bile salt independent fraction of bile flow. Am J Physiol 1978;235:E158E164. 6. Barnhart JL, Combs B. Characterization of SC2644-induced chloresis in the dog: evidence for canalicular bicarbonate secretion. J Pharmacol Exp Ther 1978;206:190-197. 7. Dunmount M, Erlinger S, Uchman S. Hypercholeresis induced by ursodeoxycholic acid and 7-ketolithocholic acid in the rat: possible role for bicarbonate ion. Gastroenterology 1980;79:82-89. 8. Strasberg SM, Ilson RG, Paloheimo JE. Bile salt associated electrolyte secretion and the effect of sodium taurocholate on bile flow. J Lab Clin Med 1983;101:317-326. 9. Dodgson SJ, Forster RE, Storey BT. The role of carbonic anhydrase in hepatocyte metabolism. Ann NY Acad Sci 1984;429: 516-524. 10. Scharschmidt BF, Van Dyke RW. Mechanisms of hepatic electrolyte transport. Gastroenterology 1983;85:1199-1214. 11. Hofmann AF. The liver biology and pathobiology. 2nd ed. In: Arias IM, Jakoby WB, Popper H, Schachter D, Shafritz DA. New York Raven Press, Ltd., 1988:553-569. 12. Osborn WRA, Tashian RE. An improved method for the purification of carbonic anhydrase isozymes by affinity chromatography. Anal Biochem 1975;64:297-303. 13. CRC Handbook. International critical tables. Vol 11. 1980:P260. 14. Pocker Y, Bjorkquist DW. Comparative studies of bovine carbonic anhydrase in H,O and D,O stop-flow studies of kinetics of interconversion of CO, and HCO,. Biochemistry 1977;16:56985707. 15. Khalifah RC. The carbon dioxide hydration activity of carbonic anhydrase. J Biol Chem 1971;246:2561-2573. 16. Kalifah RG, Strader DJ, Bryant SH, Gibson SH. Carbon-13 nuclear magnetic resonance probe of active-site ionizations in human carbonic anhydrase B. Biochemistry 1977;16:2241-2247. 17. Wilkinson GN. Statistical estimations in enzyme kinetics. Biochem J 1961;80:324-332. 18. Cleland WW. The statistical analysis of enzyme kinetic data. Adv Enzymol 1969;29:1-32. 19. Tu C, Wynns GC, Silverman DN. Inhibition by cupric ions of 0'' exchange catalyzed by human carbonic anhydrase 11.J Biol Chem 1981;256:9466-9470. 20. Silverman DN, Tu CK, Lindskog S,Wynns G. Rate of exchange of water from the active site of human carbonic anhydrase C. J Am Chem SOC1979;101:6734-6740. 21. Haussinger D, Kaiser S, Steble T, Gerok W. Liver carbonic
296
HEPATOLOGY
MILOV ET AL.
anhydrase and urea synthesis. Biochem Pharmacol1986;35:33173322. 22. Brumbaugh EE, Saffen EE, Chun PW. Scanning molecular sieve chromatogr,aphy of interacting protein systems. Biophys Chem 1979;9:299-311. 23. Saffen EE, Chun PW. Scanning molecular sieve chromatography of interacting protein systems IV:the difference profile method as applied by the Gibbs-Duhem expression in the analysis of the dimer-tetramer equilibria of oxyhemoglobin A. Biophys Chem 1979;9:329-344. 24. Oswein JQ, Chun PW. Human serum low density lipoproteinsodium deoxycholate interaction. J Biol Chem 1983;258:36453654. 25. Oswein JQ, Chun PW. Interaction of human serum apolipoprotein B with sodium deoxycholate. Biophys Chem 1983;18:35-51. 26. Yphantis DA. Equilibrium ultracentrifugation of dilute solutions. Biochemistry 1964;3:297-317. 27. Reisler E, HaikY, Eisenberg H. Bovine serum albumin in aqueous guanidine hydrochloride solutions: preferential and absolute interactions and comparison with other systems. Biochemistry 1977;16:197-203. 28. Eisenberg 13. Forward scattering of light, X-rays and neutrons. Q Rev Biophys 1982;14:141-172. 29. Casassa EF, Eisenberg H. Thermodynamic analysis of multicomponents solutions. Adv Protein Chem 1964;19:287-395. 30. Van Holde KE. In: Hager L, Wold F, eds. Physical biochemistry. Englewood Cliffs, NJ: Prentice-Hall, Inc., 1971:146-147. 31. Milov DE, Shireman RB, Chun PW. Bile salt-induced swelling of carbonic anhydrase [Abstract]. FASEB J 1990;4:A731. 32. De Sanctis SC. Location of guest molecules in inclusion compounds of deoxycholic acid by means of Van der Waals energy calculations. Acta Crystallogr 1983;B39:366-372. 33. Jou WS, Chun PW. Molecular mechanics of the formation of cholic acid micelles [in press]. J Mol Graphics, 1991. 34. Engberg P, Millquist E, Pohl G, Lindskog S. Purification and some properties of carbonic anhydrase from bovine skeletal muscles. Arch Biochem Biophys 1985;241:628-638. 35. Brumbaugh EE, Hilt ML, Shireman RB, Espinosa AJ, Chun PW. Scanning molecular sieve chromatography of interacting protein systems: least-squares analysis of small zone experiments. Anal Biochem 1986;154:287-304. 36. Chun PW, Baumbaugh EE, Shireman RB. Interaction of human low density lipoprotein and apolipoprotein B with ternary lipid emulsion. Biophys Chem 1986;25:223-241. 37. Maren TH, Sanyal G. The activity of sulfonamides and anions against the carbonic anhydrases of animals, plants, and bacteria. Ann Rev Pharmacol Toxicol 1983;23:439-450. 38. Maren TH. Carbonic anhydrase: chemistry physiology and inhibition. Physiol Rev 1967;47:597-781.
39. Roda A, Cappalleri AR,Roda E, Barbara L. Quantitative aspects of the interaction of bile acids with human serum albumin. J Lipid Res 1982;23:490-495. 40. Edelhoch JL, Bocian DF, Sudmeier JL. Evidence for direct metal-nitrogen binding in aromatic sulfoamide complexes of cadmium (11)-substituted carbonic anhydrases by cadmium. Biochemistry 1981;20:4951-4954. 41. Liang J-Y, Lipscomb WN. Hydration of CO, by carbonic anhydrase: intramolecular proton transfer between Zn ‘-bound H,O and histidine 64 in human carbonic anhydrase 11. Biochemistry 1988;27:8676-8682. 42. Tu C, Silverman DN. Role of histidine 64 in the catalytic mechanisms of human carbonic anhydrase I1 studies with a site-specific mutant. Biochemistry 1989;28:7913-7918. 43. Eriksson AE,Jones TA, Liljas A. Refined structure of human carbonic anhydrase I1 at 2.0 A resolution. In: Leventhal C, ed. Proteins: structure, function, and genetics. New York: Alan R. Liss, Inc., 1988;4:274-282. 44. Kannan KK. Crystal structure of carbonic anhydrase. In: Bauer C, Gross G, Bartels H, eds. Biophysics and physiology of carbon dioxide. New York: Springer-Verlag, 1980:184-196. 45. Hausinger DE, Gerok W. Hepatic urea synthesis and pH regulation role of Co,, HCO-,, pH and the activity of carbonic anhydrase. Eur J Biochem 1985;152:381-386. 46. Cohen NS, Kyan FS, Kyan SS, Cheung CW. The apparent K,,, of ammonia for carbamoyl phosphate synthetase (ammonia) in situ. Biochem J 1985;229:205-211. 47. Garcia-Marin J J , Dumont M, Corbic M, DeCovet G, Erlinger S. Effect of acid-base balance of acetazolamide on ursodeoxycholateinduced biliary bicarbonate secretion. Am J Physiol 1985;248: G20-G27. 48. Grotmol T, Buanes T, Raeder MG, DCCD (N’,N’decyclohexylcarboiimide) inhibits biliary secretion of HCO;. Scand J Gastroenterol 1987;22:207-21. 49. Salomoni M, Zuccato E, Granelli P, Montorsi W, Doldi SB, Germineani R, Mussini E. Effect of bile salts on carbonic anhydrase from rat and human gastric mucosa scan. J Gastroenterol 1989;24:28-32. 50. Mussini E, Salomoni M, Zuccato E. Effect of free conjugate bile salts on human and rat gastric carbonic anhydrase. International workshop on carbonic anhydrase [Abstract]. Palazzo AnainiSpoleto, Italy, 1990. 51. Sanyal G. Comparative carbon dioxide hydration kinetics and inhibition of carbonic anhydrase isozymes in vertebrates. Ann NY Acad Sci 1984;429:165-177. 52. Kararli T, Silverman DN. Inhibition of the hydration of CO, catalyzed by carbonic anhydrase I11 from cat muscle. J Biol Chem 1985;260:3484-3489. +
consisting of 260 and 261 amino-acid residues (1,2). A zinc ion, bound to three histidyl residues, is located at the base of the enzyme's 15 A deep active site crevice (1). Carbonic anhydrase catalyzes reversible CO, hydration reactions, a process that involves the following: (a) a proton transfer reaction from the zinc-bound water molecule as a result of the lowering of the pKa values of bound H,O and (b) a ligand exchange reaction involving the interconversion between CO, and HC0,-. The catalytic efficiency of this enzyme depends on the efficiency of the proton transfer reaction, which might limit the CO, hydration reaction (3, 4). The enterohepatic circulation of bile salts is a dynamic biochemical process involving hepatocytes, canaliculi and bile ductules as concentrated bile is delivered to the intestinal lumen for fat emulsification. Bicarbonate ions are also crucial to bile flow. For example, enhanced canalicular bicarbonate secretion accounts, in part, for properties of certain bile salt-independent flow inducers (secretin, ~-(2,4-dimethoxy-5-cyclohexylbenzoy1)-propionic acid or SC-2644, salicylate) and certain bile salt-dependent flow inducers (norurodesoxycholate acid and ketolithocholic acid) (5-10). Further, acetazolamideinhibition studies of hepatocytes in culture suggest a role for carbonic anhydraseproduced biocarbonate ions in hepatocyte gluconeogenesis, fatty acid synthesis, urea metabolism and electrolyte transport (9-1 1). Despite the proven and putative roles of carbonic anhydrase in hepatic physiological processes, the effect of bile salt on the enzyme's activity is not known. We speculated that increased tissue concentration of bile salts, as seen in cholestatic TOLOGY 1992;15:288-296.) liver disease, would affect bicarbonate availability by altering HCA activity. In this article, we report on the effect of the bile salt Three isoenzymes of human carbonic anhydrase (HCA) have been described (designated HCA-I, HCA-I1 anion on carbonic anhydrase activity, applying and HCA-111). HCA exists as a single polypeptide chain stopped-flow kinetic measurements to determine the inhibition constant (K.,) and Michaelis-Menten dissociation constant (KJ for HCA-I, HCA-I1 and bovine carbonic anhydrase (BCA) in the presence of four Received November 28, 1990; accepted September 16, 1991. distinct bile-salt molecules. The effect of acetazolamide This work was supported by NSF Grant DMB 83-12101(02),agrant from the Southern Medical ,4ssociation to David E. Milov and in part by a grant from The on bile salt-carbonic anhydrase binding is also reported. Center for Neurobiology of Aging at the University of Florida. We also report the molecular dimensions of the bile Address reprint requests to: P. W. Chun, Department of Biochemistry and salt-carbonic anhydrase complex deduced through Molecular Biology, College of Medicine, Box 100245, J.H.M.H.C., University of scanning molecular sieve chromatography and sedimenFlorida, Gainesville, FL 32610-0245. tation equilibrium measurements. 3111133988
Bile salts are potent inhibitors of bovine carbonic anhydrase and human carbonic anhydrase I and human carbonic anhydrase 11. To further characterize the binding of bile salts to carbonic anhydrase, rate constants for the CO, hydration reaction in the presence of deoxycholate, cholate, glycocholate and taurocholate were determined using stop-flow experiments. Values for the Michaelis-Menton dissociation constant for bovine carbonic anhydrase, human carbonic anhydrase I and human carbonic anhydrase I1 were found to be 5.2,9.2 and 13.2 mmol/L, respectively. The inhibition constant values for the various bile salts tested ranged from 0.1 to 1 mmolb for bovine carbonic anhydrase, 1.6 to 2.4 mmol/L for human carbonic anhydrase I and 0.09 to 0.7 mmol/L for human carbonic anhydrase 11. Our results suggest a mechanism of noncompetitive carbonic anhydrase inhibition for bile salts. Bile-salt binding to carbonic anhydrases as measured by scanning molecular sieve chromatography resulted in an increase in partition radius, molecular volume and surface area. The partition radius increased from 24 A to 28 A in the presence of 2.5 mmol/L sodium deoxycholate at critical micelle concentration. As determined by sedimentation equilibrium measurements, approximately 1gm of carbonic anhydrase will bind 0.03 gm of deoxycholate, suggesting three to six binding sites for bile salt on the carbonic anhydrase molecule. The conformational changes and inhibition of carbonic anhydrases resulting from bile-salt binding may be important to the regulation of enzymatic activity in tissues along the enterohepatic circulation; by limiting bicarbonate availability this interaction may also contribute to the metabolic derangements seen in patients with cholestatic liver disease. (HEPA-
288
Vol. 15, No. 2, 1992
MATERIAL AND METHODS Bile Salts and Enzyme Preparations. Sodium salts of cholic acid, deoxycholic acid, taurocholic acid and glycocholic acid were obtained from Calbiochem, Behring Diagnostics (La Jolla, CA). Bovine erythrocyte carbonic anhydrase (activity 2,800 Wilber Anderson units/mg solid), p-nitrophenyl acetate or p-nitrophenol, HEPES, MES (2-[Nmorpholinolethanesulfonic acid) and MOPS (34Nmorpholino]propanesulfonic acid) were purchased from Sigma Chemical Co. (St. Louis, MO). HCA-I and HCA-11, isolated by an affinity chromatography method (12), were a gift from Dr. David Silverman, University of Florida, (Gainesville, FL). Kinetic Methods. Saturated solutions of CO, were made by bubbling CO, into water a t 25" C; dilutions were prepared in the absence of air by coupling two syringes. CO, concentrations were based on the solubility of 33.97 mmol/L for CO, in water at 25" C (13, 14). All experiments were performed at 25" C. Moreover, all solutions contained 33.3 mmol/L Na,SO, to contribute an ionic strength of 0.1. All kinetic determinations of initial velocity of the CO, hydration reaction were carried out on a Durrum-Gibson stopped-flow spectrophotometer (Model D-110) interfaced with North Star microprocessor (Olis Instruments Model 360). The initial velocity was calculated using the equation v
= -
289
BILE SALTS AND CARBONIC ANHYDRASE
(d[CO,l/dt),,,,,
=
- Qo(dAldt),,t,a,
with (dA/dt) determined by least-square analysis of absorbance vs. time. The buffer factors Q, relating changes in absorption to changes in [H'] were calculated by the method of Khalifah (15) for the buffer indicator pairs listed here. Also given are pKa values, wavelengths used and differences in molar extension coefficients of acidic and basic forms: HEPES (pKa = 7.5) with p-nitrophenol (pKa = 7.1, A = 400 nm, Ae = 1.83 x mol/L-l cm-l) or with phenol red (pKa = 7.5, A = 557 nm, A€ = 550 x lo-* mol/L-l cm-l); MES (pKa = 6.1) with chlorophenol red (pKa = 6.3, A = 574 nm, AE = 1.77 x mo1L-l cm-l); and MOPS (pKa = 7.1) with p-nitrophenol (pKa = 7.1, A = 400 nm, Ae = 1.83 x l o p 4M-' cm-'). The buffer concentration in all kinetic measurements was 20 to 30 mmol/L (16). For every kinetic run, experimentally determined uncatalyzed rates were subtracted to give catalyzed rates. Evaluation of kinetic constants from initial velocities was conducted using the weighted least squares method of Wilkinson (17) and Cleland (18). The I,, (a 50% decrease in enzyme activity, ie., I,, = [actlEIZ) was determined by a previously published method that directly measured the rate of disappearance of O'*-labeled CO, substrate under equilibrium conditions (19,20). BCA activity in the presence of bile salts in HEPES (adjusted to pH 7.61 or 7.19) was also examined by a previously described spectrophotometric method that measured the rate ofp-nitrophenol appearance (348 nm) from the hydrolysis of p-nitrophenyl acetate (21). Scanning Molecular Sieve Chromatogruphy (22-25). All scanning molecular sieve chromatography (SMSC) experiments were conducted using a gel column scanning system that was essentially the same as that described elsewhere with the exceptions of the monochronometer and light source module, which were Heath EU-700 and EU 701-50. The basic operating routine consisted of moving a quartz column (24 cm x 0.90 cm) packed with a suitable gel (i.e., one with minimum scattering) through a horizontally collimated beam of monochromatic light (visible or UV)at a constant rate of 0.19 cmJsecfor 98 sec, during which time transmittance was detected by Aminco 10/267 solid state, blank subtract photo
multiplier microphotometer through an end-on photomultiplier tube (RCA 6903 for U V or RCA C7164 for visible). The direct W gel column scanner built in our laboratory and the stopped-flow (Durrum/Dionex) system were interfaced with a Northstar Horizon Z80-A microprocessor, a Sanyo DM50/2CX video display unit and an Integral Data Systems model 460 printer/plotter. All the data presented here were obtained using a record length of 300 data points. The partitioning properties of varying concentrations of HCA-I, HCA-I1 and BCA in the presence of varying concentrations of bile salt and the partitioning properties of sodium deoxycholate (DOC) itself in Sephadex G-75 (Pharmacia, Piscataway, NJ, 100 to 200 mesh, exclusion limit approximately 7 x lo4), were determined in small zone experiments at 280 nm at 25" C. In this type of experiment, a sample solution at the desired concentration was added continuously to the column until a reproducible baseline was obtained. The column was scanned at regular intervals as the solutionholvent boundary moved through the gel matrix a t a constant flow of 4.30 ml/hr. The baseline records of successive scans were subtracted from each other, yielding difference profiles. The problem of locating the centroid position of each boundary was then reduced t o the simple task of locating the peak position of these difference profiles (22-25). Partitioning calibration parameters for a given column were obtained by determining the rate of movement of a series of sample macromolecules of known molecular radius. The partition coefficient was calculated from u=
(dtidx), - dt/dx), (dt/dx), - (dt/dxl0
where (dtldx) was the slope of a plot of time vs. centroid position for a given sample marker, (dtldx,) was the slope of the void volume marker and (dt/dxJiwas the slope of the internal volume marker. The molecular partition radius of the solute was calculated using the following linear expression: a, = a,
+ b,erfc-'ui
where a, and b, were calibration constants for a set of particles of given radii in a given gel and determined independently and b, was multiplied by the error function complement inverse erfc-l of the partition coefficient u, (22-25). All of the calibration markers -BSA, ovalbumin (OVA), horse cytochrome C and aldolase (r = 34, 31, 17 and 48 A, respectively) -were obtained from Sigma Chemical Co. Sedimentation Equilibrium Measurements. All sedimentation equilibrium experiments were performed with a Beckman Model E analytical ultracentrifuge (Beckman Instruments, Fullerton, CA) equipped with a rotor temperature indicator control unit at 20" C, using a Yphantis (26) three-channel cell with a carbon-filled epoxy centerpiece in an An-D rotor at 12,590 rpm. The calculation of the molecular weight distribution and (dlnC/dX2) was based on the Yphantis method (261, using a computer program that could be modified for use with an Amdahl470 v/6 I1 unit (modified IBM 6000/1800, University of Florida's CIRCA computing facilities), plotting 1nJ or 1nC as a function of X2.The In C vs. X2 data were fitted to a least-squares polynomial, and the values of (dlnC/dX2) were calculated by a modification of the sliding three-point leastsquares quadratic treatment of Yphantis. The preferential interaction term, which in this case was a measure of the degree of sodium DOC binding to BCA, may be evaluated for a three-component system at sedimentation
290
MILOV ET AL.
HEPATOLOGY
TABLE 1. Comparative CO, hydration activities of mammalian carbonic anhydrase isozymes"
HCA-I HCA-I1 BCA (11)
6.33 85.73 0.08
32.6 787.9 1.15
0.54 13.00 1.73
5.15 9.19 13.20
"Temperature = 25" C; pH 7.19 (HEPES buffer), Calculations are based on Michaelis-Menten equation assuming an endogenous CO, concentration of less than 1mmol/L and enzyme concentrations of 16.5 pmol/L, 16.2 pmol/L and 16.5 pmoUL, respectively for isozyme, HCA-I, HCA-I1 and BCA. The k,,, and l i , data for the human isozymes I and I1 have been reported by other researchers. (34, 51, 52).
TABLE 2. Comparative CO, hydration activities inhibited by mammalian carbonic anhydrase isozymes"
Cholic acid (Na salt) Deoxycholic acid (Na salt) Glycocholic acid (Na salt) Taurocholic acid (Na salt)
HCA-I
4 (mmoUL) uncompetitive
& (mmol/L) noncompetitive
Inhibitor type
-
2.600 2.430
1.010 2.120 -
The standard error estimate of the coefficients at these concentrations Na salt of DOC
Cholic acid
[I], mmol/L n 0.5 1.0 0 1.5 A 2.0
0.02420 0.18225 0.04818 0.02389
0.02686 0.21831 0.07336 0.12337
+
~
~
_
_
_
_
Glycocolate
Taurocholic acid
0.01030 0.16879 0.04183 0.02075
0.02190 0.22875 0.03110 0.15386
_
In the HCA-I standard without inhibitor present, the standard error estimate of coefficient was 0.01233. "Kvalues for HCA-I, HCA-I1 and BCA were found to be 5.16, 9.22 and 13.19, respectively. K, = [I]/(ACIDconst/STANDARDconst-1) calculated from ([EIN,,) (1 + [I]/KJ. Temperature = 25" C; pH 7.19 (HEPES buffer). Calculations were based on noncomparative inhibition and inhibitor concentrations of each inhibitor run between 0.5 mmol/L and 2.5 mmol/L. CO, concentration ranged from 2.0 mmol/L to 17.0 mmoUL for all experiments. Values in parentheses obtained from 0"-exchange experiments (31).
equilibrium from the concentration distribution of BCA (CJ, grams of sodium DOC preferentially bound per gram of BCA, as described by the expression Z,, was given by the relationship M, (dpidC,)
=
(dlnC,/dX2)2RT/w2
(1)
where M, is the molecular weight of BCA, (ap/dC,) is the change in solution density with changing concentration of BCA,(alnC,laX2) is the slope of a plot of In C, vs. X2,X i s the radial distance of each fringe from the center of rotation, R is the gas constant in J (moles*K)-l, T is the absolute temperature and o is the angular velocity in radian/sec. The extent of the preferential interaction, which was derived from the three-component theory (27-29) in terms of the number of
(ap/ac,)
=
(1- VZPJ
+ X,(l
-
V,PJ
(2)
in which p, was the solvent density and Vz and ii, were the true partial specific volumes of BCA and sodium DOC, respectively. An alternative approach to determining Ti, from In C vs. X2 plots was based on the assumption that any change in the slope (alnC/aX2)of this plot for a solute in the absence of iigand to that for a solute in the presence of ligand was related to the binding of that ligand.
Vol. 15, No. 2, 1992
BILE SALTS AND CARBONIC ANHYDRASE
291
Avg [KI] = 2.602 bM1, [El = 16.5 IPM]
110
100 55 90 50
>
80
ID
(0
w,
w,
-
6 70
45
60 40 50
35 40
30
I'
30
0
20
40
A
60
80
100
0
120
11c02,
"
l , , , j , , , l , , , I , , , I / / ,
20
60
40
0
80
100
120
11C02,1 M
Avg [K,] = 2.430 ImM1, [El = 16.5 [W1
Avg [Kll= 2.200 lmM1, LEI = 16.5 Iml
110
60
100
55 90
>
50
1 80
0
ID
! - 45 w,
70 I
60
50 40
0 C
20
30
" " " " " '
30 1 " " ' " " " ' 40
60
80
100
lic02,Wl
120
0
20
40
60
D
80
100
120
11C02,l[M]
FIG.1.Double-reciprocalplot showing the inhibition by bile salts of the CO, hydration reaction catalyzed by HCA I in 5 mmoUL HEPES buffer, pH 7.16, at 25" C containing 5 x mol/L nitrophenol. Four different concentrations of four inhibitors (Z] are represented as follows: (A) Cholic acid (Na salt). (B) DOC (Na salt). (C) Glycocolate (Na salt). (D) Taurocholic acid (Na salt).
The effective specific volume,
was determined from
c3
Positive values of indicated preferential binding of sodium DOC, whereas negative values indicated preferential hydration 0', = (V, + ii3?J/(l+ ii,) (3) where p, and pOz were the solvent densities for solvent with and without ligand, respectively. and the preferential interaction term (26, 27) was evaluated Once the true values of molecular weight and v were known, from the radius of an equivalent sphere, r,, was calculated. The ratio ss = (V,O poz - O'OpOc)/(l- V3P0) (4) of the radius determined by SMSC over r, was equivalent to @Ic,
HEPATOLOGY
MILOV ET AL.
292
B
1
C
D
FIG.2. Three-dimensional plot of BCA in 0.041 moliL K-PO, buffer, pH 7.6,on a Sephadex G-75 column. Ten scans were made at 2-min intervals, with the first taken at 2 min. The flow rate was 7.8 mlihr. Samples were equilibrated with 0.041 mol/L K-PO, buffer, pH 7.6, and 0.05 ml of O.D,,,, = 1.253. (A) BCA, no DOC added. (B) BCA in 0.5mmoliL DOC. (C) BCA in 2.5 mmoliL DOC. (D) BCA in 5.0 mmol/L DOC.
the frictional ratio of the particle and was a relative measure of particle asymmetry. For frictional ratios greater than 1.0, the particle was considered ellipsoidal (either prolate or oblate) or even rodlike, and the axial ratio a/b was determined from f/f, according to Van Holde (30).
Because Equation 5 represented only a single binding site rather than preferential interactions at multiple RESULTS sites, it could only provide an approximate value of the Stopped-flow Kinetics Measurements. Comparison of noncompetitive inhibition constants. Noncompetitive and unthe turnover number kcatfor the hydration of CO, and inhibition was characterized by &'lope = yint K, for HCA-I, HCA-I1 and BCA at pH 7.16 is shown in competitive inhibition by a value for K F , which was Table 1. All experiments were conducted in HEPES very large compared with K F . In the pH range of 7.19 buffer adjusted to pH 7.61 or 7.19. The kcatfor HCA-I1 to 7.61, Kiintwas similar in magnitude to K?lope, and the was approximately 30-fold greater than that of HCA-I or data appeared to support noncompetitive inhibition in (1 + BCA under these experimental conditions. The K, value double reciprocal plots. The values of [E]/V,, for HCA-I was about that of HCA-I1 and BCA, which [I]/Kp) were used to evaluate Ki. The kcatand kobswere evaluated based on the following: kc, = V,, [El.,. and were of similar magnitude. The inhibition by bile salts of the hydration of CO, kobs= Ckc,,/K,1 [El,. catalyzed by HCA-I, HCA-I1 and BCA at pH 7.19 was Table 2 shows the inhibition constants for HCA-I, HCA-I1 and BCA (at a concentration of 16.5 pmol/L), evaluated using the following expression:
Vol. 15, No. 2, 1992
BILE SALTS AND CARBONIC A N H Y D W E
293
TABLE 3. Partition radius vs. concentration of BCA in 0.04 moln K-PO, buffer, pH 7.6, plus varying concentrations of sodium DOC
0
0.5
2.5
11.93
24.05 24.07 24.15 24.30 27.13 27.25 27.90 27.90 27.50 27.50 28.10 27.30
25 50 100 200 25 50 100 200 25 50 100 200
0.725
29.9 30.0 30.9 30.5 31.3 31.2 31.5 31.4 31.5 31.1 31.5 31.3
MT=4 TN (radius x 10-'Y/3 5,,assuming globular spherical particle. Radii were calculated based on a linear fit of a vs. u of error function complement of inverse sigma (i.e., a = a, + b,erfc- ui,where a, = 11.83 and b, = 34. Data shown for BCAin buffer alone represent the average of three determinations. S.D. in the fitted radii is 0.26 A in each case (Sephadex G-75).
.,
TABLE 4. Molecular size of BCA and BCA, 2.0 mmol/L sodium DOC complexes Sample
dlnCidX2
BCA (11) BCA (11)
0.62612 0.68322
(ml/gm)
0.725 -
(gm NaDOClgm BCA)
0.030
5s
Whole cell M7
Radius, (A)
0.034
28,486 2 4,500 31,346 c 4,500
24.3 27.8
The whole cell weight average molecular weights and the effective specific volume of BCA, 1 mmol/L NaDOC, were determined by In C vs. X2 plots as previously described, sedimentation equilibrum runs in pH 7.4 PO, buffer of 1 mmol/L sodium DOC buffer at 12,590 rpm at 20" C. The radius is that of an equivalent sphere, calculated from the whole cell weight average molecular weight. Concentration of BCA was approximately 0.8 mgiml in each case. The binding was calculated from equation 3, using V3 = 0.779 mVgm (for NaDOC). The partial specific volume of BCA was taken from Ref 34 and that for BCA, 1 mmol/L NaDOC, was calculated from equation 4.
where CO, concentration varied from 5.1 to 17 mmol/L in the presence of various bile salts. The inhibitor concentration varied from 0.5 mmol/L to 2.5 mmol/L. Although the inhibitory effect was pronounced in HCA-I1and BCA, the effect of these bile salts was greatly reduced in HCA-I. In each case, cholic acid showed the greatest inhibiting effect on these enzymes, where K, for HCA-I1 and BCA was about 0.1 mmol/L and 2.6 mmol/L for HCA-I. The double reciprocal plots for HCA-I and HCA-I1 in the presence of acetazolamide showed Ki values for HCA-I and HCA-I1 of 0.156 and 0.011 kmol/L, respectively. The Ki value for HCA-I1 in the presence of both acetazolamide and cholic acid was 1.02 p.mol/L, a considerable reduction in the inhibitory activity of acetazolamide in the presence of cholic acid. Our stopped-flow obtained Ki values for HCA-I1 in the presence of acetazolamide were in excellent agreement with those obtained by the O'"-exchange method (31). Results obtained by the Ol'-exchange method and plots shown in Figures 1A and 1D indicated that the inhibition of CO, hydration by bile salts was noncompetitive, whereas the cases shown in Figures 1B and 1C were hybrids of uncompetitive and noncompetitive. Inhibition was incomplete in all cases.
The effects of conjugated and unconjugated bile salts on HCA-I1 as determined using the O1"-exchange method are shown in Table 2. The values were quite consistent with those obtained by stop-flow kinetics; however, the scatter of the data makes evaluation of K, based on Durrum stop-flow experiments difficult. Scanning Molecular Sieve Chromatography. Representative small zone profiles of BCA in 0.04 mol/L K-PO, buffer, pH 7.6, on Sephadex G75 at 280 nm containing various concentrations of sodium DOC are illustrated in Figures 2 and 3. The time-dependent centroid movements of varying concentrations of sodium DOC were used to calculate the partition radii of these solutes, based on the centroid movements for column calibration markers as previously described (Table 3). A plot of sodium DOC radius vs. concentration was used to estimate the upper limit of the critical micelle concentration (CMC) of sodium DOC as 2.5 mmol/L. However, although the micelle transition was fairly sharp, no data points were taken between 0.5 and 2.5 mmol/L sodium DOC, indicating that the viscometry value was 1.5 mmol/L in this buffer (Chun, et al., Unpublished observations, 1991). Also, an upper limit value of 21 t 0.5 A was found for the stable sodium DOC micelle in this buffer.
294
MILOV ET AL.
HEPATOLOGY
0.041M PO&, pH 7 . 6
0.5 mn DOCIPOI, 2 . 5 ni-l DOC/PO& 5.0 mfl DOC/PO4
FIG.3. Influence of sodium DOC binding to BCA as examined by direct gel column scanning. Samples in three different concentrations of DOC were scanned on a Sephadex G-75 column in phosphate buffer, pH 7.6. Each experiment consisted of 10 scans and 300 data points/scan. The sixth and seventh scans of each run are shown in the composite figure. By this means, the degree of binding of BCA as influenced by DOC may be detected by changes in particle radius. Using this scanning technique, size changes of 2 to 3 A in radius can be detected analytically. Binding of bile salt to carbonic anhydrase is most pronounced above the CMC.
Table 3 indicates a sodium DOC and BCA concentration dependence on the processes of sodium DOC binding and increase in molecular volume and surface area of the molecule. It was not unreasonable to assume that BCA particles in the presence of 2 mmol/L DOC (below CMC) were swollen, with approximately a 15% increase in radius caused by the binding of sodium DOC monomers. The percent increase in surface volume and surface area was calculated to be 59% and 36%, respectively. This increase in partition radius to 28 A in the presence of DOC remained relatively constant below CMC. (Table 3). When acetazolamide in the presence of 2 mmol/L DOC was incubated with BCA, no noticeable change was seen in the partition radius despite the distinct decrease in inhibitory activity, suggesting the possible formation of an inclusion compound or a significant degree of interaction between DOC and acetazolamide. This finding would be consistent with the observation of De Sanctis (32), who noted that DOC forms inclusion compounds with phenanthrene and (El-p-dimethylaminoazobenzene.Jou and Chun (33) have recently demonstrated, in the simulation of the molecular mechanics of cholic acid micelle formation, that sandwiched multilayer channels are formed that may incorporate guest molecules into the micelle. Other globular proteins exposed to DOC also demonstrated increased partition radius in the SMSC system. We examined the effect of DOC on a number of globular proteins such as BSA, OVA, horse cytochrome C, human low-density lipoprotein (24, 25) and uteroferrin under similar conditions (31). We observed an increase in the partition radius of the following: BSA = 34 to 38 A (0.04 gm DOC/gm protein); OVA = 31 to 33 (0.01 gm DOC/gm protein); horse cytochrome C = no change; uteroferrin = slight change (from 29 to 31 A); and human low-density lipoprotein = 110 to 125 A (0.04 gm DOC/gm protein). This increase in partition radius in
1
I
I
FIG.4. Proposed model for bile-salt inhibition of carbonic anhydrase in the periportal bile canaliculi (the concentration of bile salt in the canaliculus is 5 to 10 mmoUL). Binding of bile salt to carbonic anhydrase may regulate the availability of bicarbonate throughout tissues along the enterohepatic circulation.
these proteins indicated a change in conformation, which was probably caused by partial denaturation that occurred in the presence of bile salts. Such volume changes were not the mere additive dimension of three to six bile-salt molecules. Preferential Binding of Sodium DOC. The 1nC vs. X2 plot of BCA in pH 7.6 buffer revealed no curvature, indicating that very little concentration-dependent selfassociation of BCA particles occurred at 20" C. The whole cell weight average molecular weight calculated from this plot was found to be 29,000 & 4,500 gmlmol (Table 4). Using 8, = 0.725 (341, the radius of an equivalent sphere of that weight would be 24.1 ? 0.5 A, agreeing very closely with SMSC data. The ratio of these radii was equivalent to the frictional ratio and was used to calculate the axial ratio for a prolate ellipsoid as 1.15. The dlnC/dX2 values were used in equations 2 and 3 (27-29; 35, 36) to estimate x3 of the BCA, 1 mmol/L sodium DOC complex as 0.03 gm of sodium DOC bound/= of BCA, resulting in a O r , = 0.747 ? 0.001 mVgm and a preferential interaction term of 6, = + 0.036, indicating preferential binding of sodium DOC. The molecular weight was reevaluated to be 32.5 5 0.01 x lo3 gm/mol. Using O'c = 0.747 ml/gm with this whole cell weight average molecular weight gave the radius of an equivalent sphere as 28.1 ? 0.07 A (see Table 4). A comparison of this radius with that from SMSC indicated an axial ratio of 1.17 for a prolate ellipsoid. Although these radii were determined in two different systems, in which the effects on sodium DOC micelles were apparent, it was assumed that the radii of the BCA, 1mmolb sodium DOC complex would change very little from 2.5 mmol/L to 1.0 mmol/L because only sodium DOC monomers are present.
Vol. 15, No. 2, 1992
295
BILE SALTS AND CARBONIC ANHYDRASE
DISCUSSION
Our results demonstrated bile salts to be potent and rapid inhibitors of in uitro CO, hydrase activity for three carbonic anhydrase isoenzymes. Cholate in particular was the most potent, naturally occurring inhibitor of carbonic anhydrase yet reported, although it has significantly less inhibitory activity than acetazolamide (37, 38). In contrast to the noncompetitive inhibition kinetics observed with acetazolamide and carbonic anhydrase, we found bile salt inhibition of HCA-I and HCA-I1 to be both noncompetitive and uncompetitive, suggesting a distinct inhibition mechanism. The sedimentation equilibrium measurements of the bile-saltbinding to BCA suggest three to six binding sites per molecule of BCA. This is in close agreement with the amount of DOC that binds to human serum albumin, the major bile salt transport protein in human beings (39). Although anion inhibition of carbonic anhydrase occurs through complexes at the active site crevice (401, bile-salt binding must occur at sites in addition t o the active site. The SCMC data showed that sodium DOC induces the swelling of the globular carbonic anhydrase molecule. This may represent a denaturing effect. We hypothesize that bile-salt binding to carbonic anhydrase results in increased molecular volume, or swelling, that inhibits the effectiveness of the crucial proton transfer reaction occurring at the carbonic anhydrase active site during CO, hydration (41,42). A second possibility is that such molecular swelling alters the important conformational relationship between His64, His67 and His200 as shown on x-ray crystallographic data (43, 44). Our studies predict strong inhibition of HCA-I and HCA-I1 in tissues with conjugated or unconjugated bile salt concentration above 2 mmol/L. The physiological relevance of our in uitro studies may be particularly important to patients with cholestatic liver disease. For example, in isolated, perfused rat liver, the addition of acetazolamide results in hyperammonemia; the mechanism involves depletion of the HCO; substrate required for the first enzymes of ureagenesis from NH, and carbamoyl-phosphate synthetase I. (21, 45, 46). Conceivably, hyperammonemia could result from hepatic mitochondrial carbonic anhydrase inhibition during intrahepatic cholestasis. Liver perfusion studies have also shown diminished bicarbonate production during metabolic acidosis (47); this may result in enhanced bile-salt binding to carbonic anhydrase at low pH. This observation may relate to our finding that bile salts bind more avidly to carbonic anhydrase at pH 7.19 than at pH 7.61. At the level of the bile ductule, bile-salt secretion has been shown to require HCO; production because acetazolamide and proton pump inhibitor (N,N'dicyclohexylcarbodiimide) inhibit biliary secretion of HCO 3 (47,481.As such, we speculate that bile salts may, in part, regulate the rate of their own secretion through the feedback inhibition of carbonic anhydrase. Although the precise molar concentrations of bile salts along the
enterohepatic circulation are not known, it can be assumed that carbonic anhydrase inhibition would be most evident at the canaliculus and bile ductule (Fig. 4). Bile-salt inhibition of carbonic anhydrase may be important in other disease states. For example, a recent study has shown that physiological concentrations of bile salts inhibit HCA isolated from gastric mucosa; the authors suggest this to be contributory to gastritis after partial gastrectomy (49, 50).
Acknowledgments: We thank Drs. Silverman and Tu for obtaining I,, values of carbonic anhydrase and for providing the samples of carbonic anhydrase I and 11. REFERENCES 1. Linskog S, Coleman JE. The catalytic mechanism of carbonic anhydrase. Proc Natl Aead Sci USA 1973;70:2505-2508. 2. Deutch HF. Carbonic anhydrases. Int J Biochem 1987;19:101-103. 3. Maren TH, Sanyal G. The activity of sulfonamides and anions against the carbonic anhydrases of animals, plants and bacteria. Ann Rev Pharmacol Toxicol 1983;23:439-459. 4. Pocker Y, Tanaka N. Inhibition of carbonic anhydrase by anions in the carbon dioxide-bicarbonate system. Science 1978;199: 907-909. 5. Hardison WGM, Wood CA. Importance of bicarbonate in bile salt independent fraction of bile flow. Am J Physiol 1978;235:E158E164. 6. Barnhart JL, Combs B. Characterization of SC2644-induced chloresis in the dog: evidence for canalicular bicarbonate secretion. J Pharmacol Exp Ther 1978;206:190-197. 7. Dunmount M, Erlinger S, Uchman S. Hypercholeresis induced by ursodeoxycholic acid and 7-ketolithocholic acid in the rat: possible role for bicarbonate ion. Gastroenterology 1980;79:82-89. 8. Strasberg SM, Ilson RG, Paloheimo JE. Bile salt associated electrolyte secretion and the effect of sodium taurocholate on bile flow. J Lab Clin Med 1983;101:317-326. 9. Dodgson SJ, Forster RE, Storey BT. The role of carbonic anhydrase in hepatocyte metabolism. Ann NY Acad Sci 1984;429: 516-524. 10. Scharschmidt BF, Van Dyke RW. Mechanisms of hepatic electrolyte transport. Gastroenterology 1983;85:1199-1214. 11. Hofmann AF. The liver biology and pathobiology. 2nd ed. In: Arias IM, Jakoby WB, Popper H, Schachter D, Shafritz DA. New York Raven Press, Ltd., 1988:553-569. 12. Osborn WRA, Tashian RE. An improved method for the purification of carbonic anhydrase isozymes by affinity chromatography. Anal Biochem 1975;64:297-303. 13. CRC Handbook. International critical tables. Vol 11. 1980:P260. 14. Pocker Y, Bjorkquist DW. Comparative studies of bovine carbonic anhydrase in H,O and D,O stop-flow studies of kinetics of interconversion of CO, and HCO,. Biochemistry 1977;16:56985707. 15. Khalifah RC. The carbon dioxide hydration activity of carbonic anhydrase. J Biol Chem 1971;246:2561-2573. 16. Kalifah RG, Strader DJ, Bryant SH, Gibson SH. Carbon-13 nuclear magnetic resonance probe of active-site ionizations in human carbonic anhydrase B. Biochemistry 1977;16:2241-2247. 17. Wilkinson GN. Statistical estimations in enzyme kinetics. Biochem J 1961;80:324-332. 18. Cleland WW. The statistical analysis of enzyme kinetic data. Adv Enzymol 1969;29:1-32. 19. Tu C, Wynns GC, Silverman DN. Inhibition by cupric ions of 0'' exchange catalyzed by human carbonic anhydrase 11.J Biol Chem 1981;256:9466-9470. 20. Silverman DN, Tu CK, Lindskog S,Wynns G. Rate of exchange of water from the active site of human carbonic anhydrase C. J Am Chem SOC1979;101:6734-6740. 21. Haussinger D, Kaiser S, Steble T, Gerok W. Liver carbonic
296
HEPATOLOGY
MILOV ET AL.
anhydrase and urea synthesis. Biochem Pharmacol1986;35:33173322. 22. Brumbaugh EE, Saffen EE, Chun PW. Scanning molecular sieve chromatogr,aphy of interacting protein systems. Biophys Chem 1979;9:299-311. 23. Saffen EE, Chun PW. Scanning molecular sieve chromatography of interacting protein systems IV:the difference profile method as applied by the Gibbs-Duhem expression in the analysis of the dimer-tetramer equilibria of oxyhemoglobin A. Biophys Chem 1979;9:329-344. 24. Oswein JQ, Chun PW. Human serum low density lipoproteinsodium deoxycholate interaction. J Biol Chem 1983;258:36453654. 25. Oswein JQ, Chun PW. Interaction of human serum apolipoprotein B with sodium deoxycholate. Biophys Chem 1983;18:35-51. 26. Yphantis DA. Equilibrium ultracentrifugation of dilute solutions. Biochemistry 1964;3:297-317. 27. Reisler E, HaikY, Eisenberg H. Bovine serum albumin in aqueous guanidine hydrochloride solutions: preferential and absolute interactions and comparison with other systems. Biochemistry 1977;16:197-203. 28. Eisenberg 13. Forward scattering of light, X-rays and neutrons. Q Rev Biophys 1982;14:141-172. 29. Casassa EF, Eisenberg H. Thermodynamic analysis of multicomponents solutions. Adv Protein Chem 1964;19:287-395. 30. Van Holde KE. In: Hager L, Wold F, eds. Physical biochemistry. Englewood Cliffs, NJ: Prentice-Hall, Inc., 1971:146-147. 31. Milov DE, Shireman RB, Chun PW. Bile salt-induced swelling of carbonic anhydrase [Abstract]. FASEB J 1990;4:A731. 32. De Sanctis SC. Location of guest molecules in inclusion compounds of deoxycholic acid by means of Van der Waals energy calculations. Acta Crystallogr 1983;B39:366-372. 33. Jou WS, Chun PW. Molecular mechanics of the formation of cholic acid micelles [in press]. J Mol Graphics, 1991. 34. Engberg P, Millquist E, Pohl G, Lindskog S. Purification and some properties of carbonic anhydrase from bovine skeletal muscles. Arch Biochem Biophys 1985;241:628-638. 35. Brumbaugh EE, Hilt ML, Shireman RB, Espinosa AJ, Chun PW. Scanning molecular sieve chromatography of interacting protein systems: least-squares analysis of small zone experiments. Anal Biochem 1986;154:287-304. 36. Chun PW, Baumbaugh EE, Shireman RB. Interaction of human low density lipoprotein and apolipoprotein B with ternary lipid emulsion. Biophys Chem 1986;25:223-241. 37. Maren TH, Sanyal G. The activity of sulfonamides and anions against the carbonic anhydrases of animals, plants, and bacteria. Ann Rev Pharmacol Toxicol 1983;23:439-450. 38. Maren TH. Carbonic anhydrase: chemistry physiology and inhibition. Physiol Rev 1967;47:597-781.
39. Roda A, Cappalleri AR,Roda E, Barbara L. Quantitative aspects of the interaction of bile acids with human serum albumin. J Lipid Res 1982;23:490-495. 40. Edelhoch JL, Bocian DF, Sudmeier JL. Evidence for direct metal-nitrogen binding in aromatic sulfoamide complexes of cadmium (11)-substituted carbonic anhydrases by cadmium. Biochemistry 1981;20:4951-4954. 41. Liang J-Y, Lipscomb WN. Hydration of CO, by carbonic anhydrase: intramolecular proton transfer between Zn ‘-bound H,O and histidine 64 in human carbonic anhydrase 11. Biochemistry 1988;27:8676-8682. 42. Tu C, Silverman DN. Role of histidine 64 in the catalytic mechanisms of human carbonic anhydrase I1 studies with a site-specific mutant. Biochemistry 1989;28:7913-7918. 43. Eriksson AE,Jones TA, Liljas A. Refined structure of human carbonic anhydrase I1 at 2.0 A resolution. In: Leventhal C, ed. Proteins: structure, function, and genetics. New York: Alan R. Liss, Inc., 1988;4:274-282. 44. Kannan KK. Crystal structure of carbonic anhydrase. In: Bauer C, Gross G, Bartels H, eds. Biophysics and physiology of carbon dioxide. New York: Springer-Verlag, 1980:184-196. 45. Hausinger DE, Gerok W. Hepatic urea synthesis and pH regulation role of Co,, HCO-,, pH and the activity of carbonic anhydrase. Eur J Biochem 1985;152:381-386. 46. Cohen NS, Kyan FS, Kyan SS, Cheung CW. The apparent K,,, of ammonia for carbamoyl phosphate synthetase (ammonia) in situ. Biochem J 1985;229:205-211. 47. Garcia-Marin J J , Dumont M, Corbic M, DeCovet G, Erlinger S. Effect of acid-base balance of acetazolamide on ursodeoxycholateinduced biliary bicarbonate secretion. Am J Physiol 1985;248: G20-G27. 48. Grotmol T, Buanes T, Raeder MG, DCCD (N’,N’decyclohexylcarboiimide) inhibits biliary secretion of HCO;. Scand J Gastroenterol 1987;22:207-21. 49. Salomoni M, Zuccato E, Granelli P, Montorsi W, Doldi SB, Germineani R, Mussini E. Effect of bile salts on carbonic anhydrase from rat and human gastric mucosa scan. J Gastroenterol 1989;24:28-32. 50. Mussini E, Salomoni M, Zuccato E. Effect of free conjugate bile salts on human and rat gastric carbonic anhydrase. International workshop on carbonic anhydrase [Abstract]. Palazzo AnainiSpoleto, Italy, 1990. 51. Sanyal G. Comparative carbon dioxide hydration kinetics and inhibition of carbonic anhydrase isozymes in vertebrates. Ann NY Acad Sci 1984;429:165-177. 52. Kararli T, Silverman DN. Inhibition of the hydration of CO, catalyzed by carbonic anhydrase I11 from cat muscle. J Biol Chem 1985;260:3484-3489. +
