KINETICS AND INHIBITION OF CARBONIC ANHYDRASE III:
COMPARISON WITH CARBONIC ANHYDRASE II





By



TUGRUL T. KARARLI


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1984
















This dissertation is dedicated to my parents, Fatma and Mehmet,

whose love, support and sacrifices have brought me to where I am.














ACKNOWLEDGMENTS

I am grateful to Dr. David N. Silverman for his excellent guidance

throughout my graduate education to become an independent scientist.

I would like to thank Dr. C. K. Tu for helping me with his valuable

technical expertise and thoughtful advice in my research problems. I

acknowledge Dr. R. Rowlett with whom I have had valuable discussions.

I would like to extend my gratitude to Mr. Dave Godman for his very

valuable help with the perfusion techniques. I would like to thank

Dr. W. Kem, Dr. Gautam Sanyal, Mr. George C. Wynns and Mr. Ness Pessah

for their contributions in protein purification. I especially

acknowledge Mrs. Judy Adams for her excellent job of typing this

dissertation. Finally, I would like to acknowledge the Scientific and

Technical Research Council of Turkey for their financial support

during my doctoral education.












PREFACE

This dissertation is composed of an introductory chapter, four

chapters written in standard manuscript style, a conclusions chapter

and an appendix section. Due to this format, some material may appear

redundant. It is hoped that the reader will accept the repeated

material as being important in the scope of this dissertation.










TABLE OF CONTENTS

PAGE

ACKNOWLEDGMENTS ..........................****..........***.... iii

PREFACE ...................................... ....... ...... iv

ABSTRACT ....................................................... vii

CHAPTER ONE INTRODUCTION .................................... 1

CHAPTER TWO INHIBITION BY CUPRIC IONS OF HYDRATION OF CO2
AND DEHYDRATION OF HCO3 CATALYZED BY HUMAN
CARBONIC ANHYDRASE II .......................... 8

Introduction ................................. 8
Experimental Procedure ...................... 9
Results .............. ........................ 13
Discussion ................................... 23

CHAPTER THREE HYDRATION, DEHYDRATION AND ESTERASE ACTIVITY
OF CAT CARBONIC ANHYDRASE III ................... 29

Introduction ................................. 29
Experimental Procedure ....................... 30
Results ...................................... 33
Discussion ......................... .......... 40

CHAPTER FOUR RATE-CONTRIBUTING EVENTS IN THE HYDRATION
OF CO2 CATALYZED BY CARBONIC ANHYDRASE III
FROM CAT SKELETAL MUSCLE ........................ 45

Introduction ................................. 45
Experimental Procedure ...................... 46
Results ..................... ............ 46
Discussion .................. ..... .......... 53

CHAPTER FIVE INTERACTION OF CAT CARBONIC ANHYDRASE III
WITH ANIONS AND SULFONAMIDES .................... 59

Introduction ............................. 59
Experimental Procedure .................... 60
Results ...................................... 62
Discussion ....... ........... ............ 75

CHAPTER SIX CONCLUSIONS .................................. 81








Table of Contents--Continued



APPENDIX A ..................................................... 83

APPENDIX B .................................................. 90

REFERENCES ..................................................... 100

BIOGRAPHICAL SKETCH ............................................ 106













Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


KINETICS AND INHIBITION OF CARBONIC ANHYDRASE III:
COMPARISON WITH CARBONIC ANHYDRASE II

By

Tugrul T. Kararli

December, 1984

Chairman: David N. Silverman
Major Department: Pharmacology and Therapeutics

In order to get more information about the rate-contributing

intramolecular proton transfer step in the catalytic mechanism of

carbonic anhydrase II, the inhibition by Cu2+ of the hydration of CO2

and dehydration of HCO3 catalyzed by human isozyme II was measured over

the pH range 6.0 to 7.6 by using a stopped-flow technique. Also, the

interaction of Cu2+ with the active site sphere of human carbonic

anhydrase II was evaluated by a fluorescence method which utilizes

fluorescent dansylsulfonamide. The inhibition of both hydration and

dehydration reaction by Cu2+ were mixed/noncompetitive at a fixed high

buffer concentration. The corresponding pattern was uncompetitive when

concentration of CO2 was fixed and buffer varied. The inhibition

constant for Cu2+ was affected by a group with a pKa close to 7.0.

These results, which were also verified by simulation of the catalysis

with computer methods, suggest that Cu2+ binds to His 64 of carbonic

anhydrase II and inhibits an intramolecular proton transfer.









To determine the kinetic mechanism of isozyme III, several kinetic

parameters of this isozyme from cat were measured by using stopped-

flow, fluorescence, 018exchange and spectrophotometric techniques. The
C02
pH profiles of kcat and kenz were consistent with the isozyme having a

zinc-bound H20 with a pKa < 6 and a basic residue with a pKa '9.
HC03 HC03
Increasing values of kcat /K, as pH was decreased, also supported such
CO2
a low pKa for zinc-bound H20. The isotope effects of magnitude 2.6 in kcat
C02 CO2
and 1.3 in kcat/Km were indicative of a rate-limiting proton transfer
transfer located outside the interconversion of CO2 and HCO. Buffers

did not have any affect in the catalysis of CO2 by isozyme III. The uncom-

petitive/mixed patterns of inhibition of the hydration reaction at low as

well as at high pH by N5 and I- further supported a pKa < 6 for zinc-

bound H20 and a rate-limiting event outside the interconversion of CO2

and HCO3. Inhibition constants of N3 and binding constant of

chlorzolamide were dependent on the ionization of a group with a pka > 8.

Anions and sulfonamides were shown to compete for the same or an overlapping

binding site on isozyme III.


viii











CHAPTER ONE
INTRODUCTION


Carbonic anhydrase is a zinc-containing metalloenzyme which

catalyzes the reversible hydration of CO2 as given in equation 1.

CO2 + H20O- HCO3 + H (1)
In this process, there is a net release of one proton and the HCO3

ion. The reaction 1 is catalyzed by all three mammalian carbonic

anhydrase isozymes and follows the Michaelis-Menten kinetics. The

turnover number of this reaction for each isozyme is given in Table 1.

The understanding of the catalytic groups and rate-limiting events in

the kinetic mechanism of carbonic anhydrase II has been a main focus of

intense research in the past. (See Appendix A for history, distribution

and general properties of isozymes I and II.)

Steiner et al. (1) for the human and Pocker and Bjorkquist (2) for

the bovine isozyme II found that while the Michaelis-Menten constants
CO2 HC03
K, and Km for the forward and reverse reactions are independent of
CO2 HC03
pH, the turnover numbers kcat and kcat are correspondingly dependent
on the basic and acidic form of an activity controlling group titrating

with a pKa close to 7. Experimental evidence suggests that this

group is the water molecule on the zinc (3). It was shown by Jonnson

et al. (5) at steady state and by Silverman and Tu (6) at chemical

equilibrium that reaction 1 catalyzed by isozyme II is limited by the

release of the product proton to the solution in the absence of or at

low concentration of buffers. Addition of buffers up to 5-10 mM









TABLE 1
MICHAELIS-MENTEN PARAMETERS FOR THE ISOZYMES OF CARBONIC ANHYDRASE


CO2 CO2 CO2 C02
kcat (sec-1) Km (mM) kcat/Km
(M-1 sec-1)

Isozyme Ia 1.2 x 105 4.2 2.9 x 107
Isozyme IIb 1.0 x 106 8.3 1.2 x 108
Isozyme IIIc 1.0 x 104 30.0 3.3 x 105

abData from the high pH (pH > 8) of Lindskog et al. (7) (25C).

CData from this study at pH 9.5 and 250C.








concentration enhances the rates of reaction 1 by increasing the rate

of this proton release (6). Understanding of the rate-determining

events in the presence of high buffer concentration came from the

measurement of the rates of reaction 1 catalyzed by carbonic anhydrase

II in H20 and D20. Again, Steiner et al. (1) for the human and Pocker

and Bjorkquist (2) for the bovine carbonic anhydrase II found that
CO2 CO2
while kcat and Km have isotope effects of magnitude as big as 3 to 4,
CO2 CO2
kcat/Km has an isotope effect of unity. A similar pattern was also
observed for the reverse reaction in these studies. All the reaction
CO2 CO2
steps occurring before the departure of HCO3 are included in kcat/K .
CO2 CO2
This isotope effect of unity on kCt /K2 lead Steiner et al. (1) to

propose the enzymic mechanism given in Scheme I, with its

rate-contributing intramolecular proton transfer located outside the

CO2-HCO3 interconversion. There is some experimental evidence to
0
suggest that histidine 64 residue of isozyme II which is located 6 A

away from the zinc (8) shuttles protons between the catalytic group and

solvent in this rate-contributing event.

Later, 180 exchange experiments by Silverman et al. (9) and Tu and

Silverman (10) at chemical equilibrium have been very useful in further

understanding of this rate-contributing proton transfer. In the 180

exchange technique, one can measure the rate of the CO2-HCO0
exch CO2
interconversion, kcat /Keff, separate from rate of the release of

180-labelled ligand water from the active site, kH20. Measurement of
exch CO2
these rates in H20 and D20 showed that while kcat /Keff does not

have any isotope effect, kH20 has an isotope effect of magnitude as big

as 8.0 at pH 6.6 (9,10). Tu et al. (11) further showed that Cu2+ is a

























Proposed mechanism of human carbonic anhdyrase II
catalysis. EX refers to the transitory complexes ES and
EP, and HEX refers to HES and HEP, respectively. S and P
stand for CO2 and HCO3. Numerals indicate the steps to
which rate constants with the corresponding subscripts
refer. H following E (as in EH) denotes the enzyme with a
protonated activity-controlling group, probably a
zinc-bound water molecule. H preceding E (as in HEX)
denotes the enzyme with a protonated group involved in the
shuttling of protons between the activity-controlling
group and the external medium. Histidine-63 is a likely
proton shuttle group. (With permission from Dr. Roger
Rowlett.)


Scheme 1.

















HEXt


rr

a, 0 3


EEX
EH~







specific inhibitor of this intramolecular proton transfer with an IC50

value of 1 x 10-7 M at pH 7.3. All these results are certainly in

favor of the catalytic mechanism of isozyme II given in Scheme I.

After 50 years since its discovery, there are still some unanswered

questions about the kinetic mechanism of carbonic anhydrase II. One

is the mechanism of the intramolecular proton transfer, especially the

group responsible for it. For this purpose, I have chosen Cu2+ as a

tool and measured its inhibition of the hydration and dehydration

reaction catalyzed by human isozyme II as a function of pH by using a

stopped-flow technique. Also by using a fluorescence technique, I have

measured the interaction of Cu2+ with the active site sphere of isozyme

II. The results of these experiments are presented in the second

chapter of my dissertation.

Carbonic anhydrase III is the newest and the slowest isozyme found

in mammals (12,13). (See Appendix A for history, distribution and

physiological function of isozyme III.) Sanyal et al. (14) found that

while cat isozyme III binds anions tighter than isozyme II, it inter-

acts 700 to 30,000 times less tightly with the sulfonamide inhibitors.

Tu et al. (15) showed that the values of the kinetic parameters for the

hydration reaction catalyzed by cat isozyme III are independent of pH

in the pH region 6.0 to 8.5, where in the same pH region, isozyme I and

II have pH dependent activity. Tu et al. (15) also by measuring the

180-exchange parameters in H20 and D20 found that while the

rate constant kH20 has an isotope effect of magnitude 2.5,

kexch /KC2 has one of unity. This suggested a rate-contributing proton
cat eff
transfer located outside the interconversion of C02-HCO5 in

reaction 1 catalyzed by cat isozyme III.










All these intriguing findings caused further interest in the

kinetic properties of this isozyme, namely the identification of the

groups which control its activity and binding of inhibitors, and the

elucidation of the rate-contributing proton transfer. An understanding

of these events will then enable us to draw a more complete picture

about the general mechanism of the catalysis by all carbonic anhydrase

isozymes. For this purpose, I have measured several kinetic properties

of cat isozyme III by using stopped-flow, fluorescence, and

180-exchange techniques. The third chapter of my dissertation

describes the pH profiles of the hydration, dehydration, and esterase

activity of cat isozyme III. In the fourth chapter, the solvent

deuterium isotope and buffer effects on CO2 hydration are presented.

The fifth chapter includes the results of the inhibition by anions and

sulfonamides of the hydration of CO2 catalyzed by cat isozyme III.











CHAPTER TWO
INHIBITION BY CUPRIC IONS OF HYDRATION OF CO2
AND DEHYDRATION OF HC03 CATALYZED BY HUMAN CARBONIC ANHYDRASE II

Introduction

The inhibition by cupric ions of the hydration of CO2 catalyzed by

carbonic anhydrase II is interesting because of the recent results of

Tu et al. (11) suggesting that Cu2+ is specifically inhibiting a proton
shuttle mechanism in the catalytic pathway. Tu et al. based this

conclusion on their observation that Cu2+ as well as Hg2+ in the range
of 10-7 to 10-6 M inhibited the release from the enzyme of H20 bearing

substrate oxygen, but did not affect the actual interconversion of

CO2 and HC03 at chemical equilibrium. It has been known for some time

that Cu2+ is an inhibitor of the catalysis measured at steady state

(16). The purpose of this study is to examine this inhibition by Cu2+

at steady state in detail in order to understand more fully the

implication of inhibition by copper of this catalysis. For this

purpose, the inhibition by CuSO4 of hydration of CO2 catalyzed by human

isozyme II was measured, observing the rate of change of a pH-indicator

in a stopped-flow spectrophotometer. These results are used to devise

a model for the inhibition of carbonic anhydrase, a model which is then

supported by simulations of the catalysis (Appendix B). It is the

overall conclusion that copper inhibits intramolecular proton transfer
in carbonic anhydrase II by binding to a residue in the active-site
cleft, probably the imidazole ring of His 64, which participates in the

transfer of protons between the active site and solution.









Experimental Procedure

Materials. Human carbonic anhydrase II was obtained from human

erythrocytes of outdated blood. The erythrocytes were washed three

times with isotonic saline and then lysed by addition of a volume of

water equal to the volume of packed cells. The separation of membranes

was achieved by centrifugation and human carbonic anhydrase II was

prepared from the supernatant by the affinity chromatography procedure

of Khalifah et al. (17) in the following way. The pH of supernatant

was brought to 8.7 with solid Tris and then gently stirred overnight

with the affinity chromatographic resin containing CM-Bio-Gel A

(Bio-Rad Laboratories, Richmond CA) coupled to p-aminomethylbenzene

sulfonamide (pAMBS). This gel was prepared by the method of Khalifah

et al. (17) and washed several times with 0.1 M Tris-sulfate/0.2 M

sodium sulfate, pH 8.7, prior to its use. The following day, the fluid

mixture was passed through a sintered glass filter into a vacuum flaks.

The affinity gel on the filter was thoroughly washed with the same

buffer to remove any nonspecific protein which might adhere to the

affinity gel. To elute the enzyme, 0.1 M Tris-sulfate/0.4 M sodium

azide, pH 7.0 was used. The azide containing protein mixture was

concentrated and then dialyzed for a week at 2C against 9 changes of a

large volume of water.

Enzyme concentrations were estimated from absorbance measurements

at 280 nm by using a molar extinction coefficient of 5.34 x 104 M-1

cm-1 (18). The buffers 4-(2-hydroxyethyl)-l-piperazineethanesulfonic

acid (Hepes), 4-morpholinepropanesulfonic acid (Mops), and









1-4-piperazinebis(ethanesulfonic acid) (Pipes) and indicators

chlorophenol red, bromocresol purple, phenol red, and p-nitrophenol

were obtained from Sigma Chemical Co. 5-Dimethylaminonaphthalene-1-

sulfonamide (DNSA) was synthesized from dansyl chloride by treatment

with ammonium hydroxide by the method of Weber (19) and recrystallized

from ethanol three times. The extinction coefficient of DNSA at 280 nm

was 6.75 x 103 M-1 cm-1. Distilled, deionized water was passed through

an ion-exchange column (Cole-Parmer 1506-35) prior to use to avoid

contamination with adventitious metals. All glassware was soaked in

solutions of EDTA, rinsed thoroughly, and dried before use.

Kinetic methods. Saturated solutions of CO2 were made by bubbling

CO2 gas into water in a vessel maintained at 25C. Solutions of CO2
were prepared in the absence of air by diluting saturated solutions

with degassed water using two connected syringes. Concentrations of CO2

were calculated based on the concentration of 33.8 mM at 25C for a

saturated solution (20).

Measurements of the initial velocity of hydration of CO2 and

dehydration of HCO3 were carried out on a Durrum-Gibson (Model D-110)

stopped-flow spectrophotometer equipped with a Nicolet Explorer Model

206 digital oscilloscope. All experiments were performed at 250C. One

drive syringe contained a solution of CO2, and the second contained

enzyme, buffer, indicator, and 33 mM Na2SO4 (to contribute an ionic

strength of 0.1 after mixing). The initial velocity of hydration (v)

was determined according to Eq. (2), with dA/dt, the rate of change of

absorbance of indicator, obtained by least-squares analysis of









absorbance versus time plots using an interfaced Hewlett-Packard 9835 B
calculator.

initial velocity= '0 -Q. i- (2)
Sdt initial \dt initial
Initial velocities were obtained from the first 5 to 10% of the

reaction with each velocity an average of four absorbance versus time

traces.
The buffer factors Q. relate changes in absorbance to changes in
[H'] and were calculated by the method of Khalifah (21) which uses the

equation
[Buff] K Buff (Ind + [H 2
Of = x x (3)
[Ind] LAe Ind Buff +[H
a K fa [
where [Buff] and [Ind] represent total buffer and indicator
Buff Ind
concentrations, respectively, Ka and K are the buffer and

indicator ionization constants, Ac is the difference in molar extinction
coefficient between the acidic and basic forms of the indicator, and L

is the reaction-cell optical path length in cm. In stopped-flow

studies, the following buffer-indicator pairs were used (also listed
pKa and Ae values): Pipes (pKa = 6.8) with bromocresol purple (pKa =
6.8, X = 588 nm, AE = 3.00 x 104 M-1 cm-1); Mops (pKa = 7.1) with
p-nitrophenol (pKa = 7.1, X = 400 nm, As = 1.83 x 104 M-1 cm-1);
Hepes (pKa = 7.5) with phenol red (pKa = 7.5, X = 557 nm, AE = 5.59 x
104 M-1 cm-1).

For each catalyzed, initial velocity reported, the experimentally
determined uncatalyzed rates have been subtracted. Estimation of
kinetic constants from initial velocities was done using the










least-squares method of Wilkinson (22) and Cleland (23) with v4

weights. Such a weight is necessary when the kinetic data are analyzed

in double-reciprocal form of the Michaelis-Menten equation in which the

variance in (1/v) as a function of 1/s is not a constant value.

Fluorescence measurements. Solutions of human carbonic anhydrase

II (4.0 x 10-7 M) containing various concentrations of CuSO4 and DNSA

were irradiated at 280 nm and the fluorescence emission at 460 nm

caused by the DNSA bound to enzyme was measured (see Chen and Kernohan

[24) for details of the spectra and energy transfer). DNSA is a

sulfonamide inhibitor of carbonic anhydrase. Fluorescence measurements

were made at 25C with a Perkin-Elmer MPF-44B Spectrofluorimeter.

Solutions also contained 50 mM Mops; no Na2SO4 was added (although pH

was adjusted with NaOH and H2S04). Copper was added first at the

desired concentration followed by repeated additions of DNSA. In all

cases, dilution caused by addition of DNSA was less than 2% of the

total volume of enzyme solution (2.5 ml).

In all experiments, the dissociation constants of Cu2+ and DNSA

were much greater than the enzyme concentrations; hence, no correction

was made for the amount of inhibitor bound to enzyme. The observed

intensities of fluorescence were corrected for absorption of incident

light at 280 nm by DNSA using the relationship
2.36280C
Fcorr = Fobs --(4)
(1 10-E280C)

where e280 is the molar extinction coefficient of DNSA and C is its

concentration; Fcorr and Fobs are the corrected and observed

fluorescence intensities. The magnitude of this correction is very









small, less than 3% of Fobs in all cases. The fluorescence of unbound

DNSA at 460 nm is negligible (24).

Data are presented (Figure 5) using the Eadie-Hofstee form of the

Langmuir equation based on the intensity of the corrected fluorescence

at 460 nm. The equilibrium dissociation constant of DNSA was estimated

from the Langmuir equation using the nonlinear, least-squares method of

Wilkinson (22) and Cleland (23) with no weighting factors.
Results

The inhibition by CuSO4 of the hydration of CO2 catalyzed by human

carbonic anhydrase II at pH 6.0 was noncompetitive (Figure 1); at pH

7.2 it was mixed (Figure 2), and at pH 7.4 it was noncompetitive (data

not shown). Studies at values of pH greater than 7.6 were not

successful because of the complication of formation of hydroxides of

copper. Data such as those in Figures 1 and 2 were fit by

least-squares methods to Equation (5).

1 =_1 1 + + + -- (5)
v Vmx Kint Vmax slope [C02]

The inhibition constants Kslope and Kint are obtained from the

slopes and the intercepts of double reciprocal plots. Any inhibition

pattern with values of the ratio, Knt/Klope, close to 1 and 0 within
experimental errors are called noncompetitive and uncompetitive,

respectively. If this value is very large, approaching infinity,
then the pattern is described as competitive. A mixed inhibition is
the case when the values of this ratio are other than the values

mentioned above. The values of pKnt and pK lope for inhibition of CO2
hydration by Cu2+ are given in Figure 3 (standard derivations for Knt






























>>0 ,)



4) C:J: 4 4 0
4 -) O o -

0 0 4 E- C -4
>: n .)| 4- S :E




C O C-I-) V) X
-0 o I- M 4 I

oX 1 C')O 4 C
*. 0 = 0 - I
4- L IO 4 r-i D M

c: u 4n- o
0- 4- C ) 4-
0r -o I e
4 O to.0 -CN4
(0S4-) ) -0 4-C
L O Ji -" U II
L -C 3 'r- a II
000 S a1) aC -V


4- *c u *r- 4 C- 0

1- U0 1 o 0 0
04 0 0 (U 1 0



e OO L3S tO L --4
= t1 AL X*- 4- S
0- 0 O E > r r- X-


C) WO 4-4-o a) -
0.'- a 0 0L I 3

.0 4 0 r = U) 0U +1 %D
MC O (D to L
.0aJ 0 a. 0 to 0 *L <
SlX-- *r- o 9 +1






- .0 O O k w O C
JC C ) a) b
'q <0 <-f 4J M 0 to








C X *r- C5 OC y-4 +.

. a *= 4- LO Cr E ( AU
-. tMho Ln lh cn U) :r



1- 0
*r-I


LL-




































_Olx (oes ) I_) A
I


r()


I
0
x
X



-| 0





























>) 10 (A -
C ) I

W 4- i -
N to r- C C00
,- "Z 0 C -"
4- 0 2- l -
C -4 th a)
,) o Z -)
0 I -- 2< I
0OO E '- 0
C. e-- C -- C
o (1) xfl
** 4- JLO
o .- .0OC 0 r
C S *XW X
04-' 0 03
4-) 0 Cn> -
M 4-/) 0 *
L- C 3 *r- O
"a C) 0 0U
0 0- C t
0 ,- 0 .- 0

4-0 .- Cr 4-0 W U
-'o 0 0 -
4- *- 0 .C II
0 4- q 4-'
0) t CJ 4-
W/ S 0 1- 4- )0
C S- C C) 0 U



o 4-u o J
JC C )
.0 *r- tO

*o r- C0 0 "
4-) o C
*r- LO LDl --th ( 0U
-0 S-(i I CL
*- -o CDC 0 Wd
C to X CC >
i- c S. to
(aU E('jo5 'U
S * 4- L C
r-- 1C`, -C W -4 ca


X



0
C-)






II
4)
Q.








-4
o











-4
x



+1
4-1




5-4
I I

C:






4-



:0)
+IO

.0
LO ,-4

























CCj


x
c'J





CU'
-I



-10
JQ





























It



Co 4-3 .- 0.
40 C-) =f 48-3.
0. -$- In

0o 4- 0 q) -i >, :3 C
r- O r cg to
lfl.-4 (.4- E 4- o 1
b e c =5 r o C


*,(. U ) I=O
V4-@ 0r I n *1*

c c Lo c Q. ri CA +1
co S- C) S-- I

4-) =y C)
LA Cc a) tD
r- r ) AC

.0 = -44- 4-) CO
> 4 C) -- to.0 L
u) 4-( u= E 0
&-L. LO (A2
0, C04-.- E (o
0 >) 0 : 0 r
.,- 4.4C W f4- a) C)
4- Eto C: C

4.) .C.-~000Vfl(n
CL CCC
4J 0 'a0 4. tn"'





-b 2 r
0 to 4C-) Q, Cf

CL 0) 90 +1C
4- 4-) U Cn (v = -
0 3CL-i -'a0








Q) P) Ec~- L ) 0I
aj o t o 'a = r


U.r 4- E S- 4-+
r- ,10 M .3M O
.C0 '0 0 0 4J SC







C'r)


a,
1-

a-































Q.


co V-S
ILi uif

1Iqd







and Kslope are given for selected points in the Figure). At pH 6.0,

KInt was equal to K lpe (noncompetitive). The values of KI were
smallest near pH 7 where inhibition was mixed. At pH near 7.4, the
inhibition was again close to noncompetitive. All of the inhibitory
effects of Cu2+ could be reversed by addition of EDTA. The line drawn
through pK lope was obtained by using the equation
K O
slope K1 (6
K = (6)

KEl [H+]

where KI is the lowest value of Kslope and KE and KE are the acid
dissociation constants for the first and second group which control the
binding of Cu2+ (see Figure legend 3 for the values of these
parameters).
Linear Dixon plots were observed for the inhibition of CO2
hydration at pH 7 and of dehydration at pH 6.7 up to concentrations of
1.2 x 10-5 M Cu2+. No difference in inhibition was observed when Cu2+
was placed either in the drive syringe with enzyme or in the drive
syringe with CO2. Also, I have used buffers of the type studied by
Good et al. (25) that have only weak binding interactions with many
metals (however, see Nakon and Krishnamoorty [26]). In this study, we
have made no correction for the possible binding of Cu2+ to buffer.

The inhibition by Cu2+ of the dehydration of HCO3 catalyzed by

carbonic anhydrase II at pH 6.8 was mixed with K nt = (9 2) x 10-7 M
and Kslope = (4 1) x 10-6 M (data not shown).
The inhibition caused by Cu2+ at fixed CO2 concentration and varied
buffer concentration gave patterns indicating uncompetitive inhibition
in the pH range 6.8 to 7.4. Figure 4 is a typical result at pH 6.8
using the buffer Pipes.






























0
4-)
S, X
*T- in *
co an o C)

C E I
4- "" E -- 0
4, E S O o Co .-
4- r- O


4- C> 0-

CNO> +0
C N +> 0 W +1
0 >)- X-..- U
o CO C. *


C) t "n (/) n

-C O c


C 0 0 0- --1
m0 cO c *

4- *- E0 4- t




L) "- 4.-'4),..
S! - r*-1 C) -



>)C r- ".O 05


C 0 M 0
C) t.O C 0 t
4 0 4-) *--1 t







0) 4- C-O -0 CC
'C *- 4) )
0 0 .U 41:1- *r-
*r- (n 0 W




(a*- 00 C-0 ..









S.
*-
.- cu JC: C) E



JSh~c ~ -i <













l2'



























ro
re)





I
0









The interaction between Cu2+ and DNSA binding to carbonic anhydrase

II was noncompetitive as determined by fluorescence quenching at pH 7.5

(Figure 5). Cupric ions had no effect on the fluorescence of DNSA in

the absence of carbonic anhydrase, at the concentrations studied here.

From the data of Figure 5, the value of the equilibrium dissociation
constant for Cu2+ is (2.0 0.2) x 10-6 M, and that for DNSA is (1.5
0.1) x 10-6 M at pH 7.5. I have not determined whether Cu2+ and DNSA

are able to bind to carbonic anhydrase II simultaneously or whether
their binding is mutually exclusive. When the binding of DNSA to

bovine red cell carbonic anhydrase II was determined by fluorescence

measurements at pH 7.5 using 20 mM phosphate buffer as solvent, I found

the equilibrium dissociation constant to be 3 x 10-7 M, in agreement

with the results of Chen and Kernohan (24).

Discussion
First, I comment that the range of concentration for the inhibition
by Cu2+ of the hydration of CO2 catalyzed by carbonic anhydrase II is

comparable for many different methods of measurement. This study

(Figure 3) has shown strongest inhibition at pH near 6.8 with KI ~ 8 x
10-7 M. Magid (16) measured the initial velocity of catalyzed

dehydration of bicarbonate using a pH-titration (pH-stat) procedure at

pH 7.3. He reported 50% inhibition of catalysis near 5 x 10-7 M CuSO4
(see Figure 2 of Magid [16]). In the fluorescence quenching

experiments, the maximal binding of Cu2+ occurs above pH 7.3 with Kdiss
~10-6 M. Tu et al. (11) have reported the inhibition by Cu2+ and

Hg2+ of the exchange of 180 between CO2 and H20 catalyzed by human

carbonic anhydrase II. They found that cupric ions and mercuric ions






























44--
I X

O -4 O4. O .
.- 0 0 V) S-
C) a 0
E o .. L. ,

to L-.o E ) c(o
( t 0 O- 0 -)
>if0 Xn
*4-) 2 S- 0 27 0 L)
O C 4- 4- to .
oL OC o t 1 .V-
WU U I r-
U CJ ** 0 4-
C C C U -4 S- 0

c. 0 l aw) 0 C: c<
S<- 0 CL. -* e-- 0
SOU 0 CM C'
S (.) 0 S.
r- E B -~ C: o





(U C C O tI

ruo t U M- X *4-
S -M q0 O -- *r-



.- oEI S- (


4- L .00 C0 *W5
0 () U 2 C)
S4- U 0 *.-



C a c(to r-I
<0 S- EU c X C
a 0 C I to
CC.nC v0




0- 4. *-O C)













































0 0 0 I0
0 N-- I0 N
JJooJ %


(O



0




O





0
o m

CJ
a.









inhibit the release from the enzyme of H2180, but not the

interconversion of CO2 and HC03. The data of Tu et al. (11) show 50%

inhibition of the release H2180 at 1 x 10-7 M Cu2+ at pH 7.3. This

value was obtained using solutions containing no buffers and avoided

any competition for copper binding between buffer and enzyme. Hence,

there is reasonable agreement among these various methods in the

concentration of Cu2+ that inhibits the hydration of CO2 catalyzed by

isozyme II. It is useful to note that no inhibition of isozyme I is

observed at pH 7 for Cu2+ present in the micromolar range (16,11).

Both the stopped-flow studies (Figure 3) and the results for 180

exchange show maximal inhibition of hydration catalyzed by carbonic

anhydrase II at pH near 7, with inhibition much weaker at either pH 6.0

or 7.6. These observations are consistent with the binding of Cu2+ to

the suggested proton shuttle, the imidazole ring of His 64 which is a

favorable metal interaction site (27). This ring has been shown by NMR

studies to have a pKa of 7.1 (28). Thus, as pH is lowered from 7, this

imidazole ring becomes protonated and loses its capacity to bind copper

tightly. As pH is increased above 7, inhibition by Cu2+ also becomes

weaker. There is no clear explanation of this although Tu et al. (11)

suggested that another residue of the active site cleft, not bound by

Cu2+, could act as a proton shuttle at pH > 7. Another possibility is

the formation of hydroxides of copper as pH approaches 8, causing less

copper to be bound to enzyme.

What is clear from the fluorescence studies (Figure 5) is the

noncompetitive interaction between Cu2+ and DNSA in binding to carbonic

anhydrase II. This is quite consistent with the suggestion of Tu et










al. (11) that Cu2+ binds to the proton shuttle group, His 64, and

not at the site of binding of sulfonamides, which is the coordination

shell of zinc. Further, Liljas has found, from x-ray diffraction data,

that Hg2+ coordinates to the imidazole ring of His 64 in human enzyme

II.1 In the crystal structure of carbonic anhydrase, the imidazole
0
ring of His 64 is about 6 A from the zinc (8), and sulfonamide nitrogen

binds directly to the zinc (29,30).

Finally, I comment on the mode of inhibition by Cu2+ of catalyzed

hydration. The present understanding of this catalysis by isozyme II

is that both intra- and intermolecular proton transfers can be

rate-limiting (31). This means that the concentrations of the various

enzyme forms (e.g. E, EH+, HEH2+, . ) during catalysis at steady

state can be quite different from the concentrations of these forms

during catalysis under equilibrium conditions (such as 180 exchange).

Tu et al. (11) in their equilibrium experiments found evidence that

Cu2+ inhibits the proton shuttle mechanism but not the catalytic

interconversion of CO2 and HCO3. If it is assumed that all forms of

the enzyme are in their equilibrium ionization states, then the result

of Tu et al. predicts uncompetitive inhibition of CO2 hydration by Cu2+

at steady state. That this is not observed (Figures 1-3) indicates

that the interpretation of present steady-state data cannot rely on

equilibrium protonation states of all residues involved in the

catalysis. This conclusion has already been reached in the explanation

of the inhibition of CO2 hydration by anions at pH > 8 (7,32). Thus,



1Anders Liljas, University of Uppsala, personal communication.








to interpret the mode of inhibition of CO2 hydration by Cu2+ requires

that a specific model of the binding of Cu2+ to the enzyme be

constructed and the effect of this binding on catalysis be calculated

by simulation at steady-state conditions. This is done in Appendix B

which confirms that the steady state data reported here are consistent

with the inhibition by Cu2+ of the proton shuttle mechanism.

One aspect of copper inhibition is subject to a simple

interpretation. The mode of inhibition of the protonation of buffer

during the hydration reaction (Figure 4) is uncompetitive. This is

consistent with the suggestion that His 64 is a proton shuttle residue

which must be deprotonated by transfer of H+ to buffer before Cu2+ can

bind. Thus the binding of Cu2+ depends on the prior interaction of

enzymes with substrate buffer before Cu2+ can bind, consistent with

uncompetitive inhibition.

In summary, these studies confirm the inhibition by cupric ions of

the hydration of CO2 catalyzed by carbonic anhydrase II with a KI for

binding of Cu2+ near 1 x 10-6 M. The pH dependence of this binding is

consistent with a binding site on the enzyme with a pKa near 7, such as

the imidazole side chain of His 64. This binding site is distinct from

the coordination sphere of zinc in which the actual interconversion of

CO2 and HCO3 and the binding of sulfonamides takes place.













CHAPTER THREE
HYDRATION, DEHYDRATION AND ESTERASE ACTIVITY
OF CAT CARBONIC ANHYDRASE III


Introduction

The results of experiments by Tu et al. (15) at steady state and

chemical equilibrium which showed pH independent values of the kinetic
CO2 CO2 exch CO2
parameters, kcat, Km and kcat /Keff for the hydration of CO2

catalyzed by cat carbonic anhydrase III in the pH range 6.0 to 8.5

suggested that the pKa of the activity controlling groups) in this

isozyme is outside of this pH range. However, Koester et al. (33)

found that the hydrolysis of p-nitrophenyl acetate catalyzed by isozyme

III from rabbit was dependent on a catalytic group titrating with a pKa

close to 7.5. Although the sequence of neither of the enzymes is

available yet, if one assumes that they are comparable, then the

results of studies by these investigations suggest that CO2 hydration

and esterase activities of isozyme III might depend on the ionization

of different active site groups. Since hydration of CO2, dehydration

of HCO3 and hydrolysis of esters catalyzed by carbonic anhydrase II

depend on the ionization of the same catalytic group, zinc-bound H20

with a pKa around 7.0, then the question is what active site groups)

control the same reactions catalyzed by cat carbonic anhydrase III.

For this purpose, I have extended the studies of the hydration of

CO2 catalyzed by cat carbonic anhydrase III to a wider region of pH and









measured dehydration of HCO3 and hydrolysis of p-nitrophenyl acetate

catalyzed by cat isozyme III in the pH region 5.7 to 9.7. The kinetics

of catalysis of CO2 hydration and HC03 dehydration were observed by a

stopped-flow method (21) in which velocities were determined from the

rate of change in absorbance of a pH indicator. The kinetics of the

hydrolysis of p-nitrophenyl acetate were determined by following the

appearance of the product p-nitrophenoxide at 400 nm using a

spectrophotometer.

I found that the hydration, dehydration and esterase pH-rate

profiles of cat isozyme III are consistent with this enzyme having an

activity controlling group with a pKa < 6. The results also indicate

an additional active site group with a pKa ~9.0, the ionization of

which increases the catalytic activity of isozyme III above pH 8.

Experimental Procedure

Methods. Carbonic anhydrase III was prepared from the hind limbs

of fully anesthetized cats pentobarbitall, 25 mg/kg, i.v.). After

administration of heparin (3000 units/kd, i.v.), the hind limb was

perfused with 0.9% NaCl until the muscles appeared pale and the venous

effluent was colorless. The muscle was excised, homogenized, and the

homogenate centrifuged. Carbonic anhydrase III was isolated from the

supernatant by the method of Sanyal et al. (14) which uses two affinity

gels: first with p-amninomethylbenzene sulfonamide coupled to

carboxymethyl (CM)-Bio-Gel A (Bio-Rad) to remove erythrocyte carbonic

anhydrase I and II, and second with 2-(4-aminobenzene)sulfonamide-

1,3,4-thiadiazole-5-sulfonamide coupled to (CM)-Bio-Gel A to isolate

carbonic anhdyrase III. The second gel was then rinsed clean with Tris









sulfate buffer at pH 7.5, following which carbonic anhydrase III was

recovered by elution with 0.4 M sodium azide. This solution was

dialyzed against several changes of a large volume of 1 nmf

dithiothreitol (DTT) solution. To separate the monomer from the dimer,

the purified enzyme was passed through an Ultra Gel (LKB AcA 44)

molecular sieve column pre-equilibrated with 0.05 M Tris sulfate,

0.025 M Na2S04, and 1 mM dithiothreitol at pH 7.5. The monomeric

enzyme fraction was collected and dialyzed against 1 m'M dithiothreitol

and stored at 4C in the same reducing medium. Only the monomeric form

of cat isozyme III was used in these studies. The purity of the sample

was estimated by polyacrylamide gel electrophoresis and with the

inhibitor ethoxzolamide, which is a potent inhibitor of human isozymes
I and II (KI = 2 x 10-9 M) and a weak inhibitor of cat isozyme III

(KI= 5 x 10-5 M) (measured at OC, see reference 14). By neither of

these methods, no contamination by isozyme I or II was detected. The

concentration of cat isozyme III was determined from the molar
extinction coefficient of 4.8 x 104 M-1 cm-1 (14).

Kinetic method. Solutions of CO2 were made up in the same way as

described in Chapter Two. For the dehydration reaction, appropriate
concentrations of KHCO3 were prepared by diluting freshly prepared

0.2 M stock solutions of KHCO3 with C02-free H20. All the experiments
were performed at 250C. Moreover, all solutions contained 33 mM Na2SO4
to give an ionic strength of 0.1.

Initial velocity measurements of CO2 hydration and HC03 dehydration
were carried out on a Durrum-Gibson stopped-flow spectrophotometer in

the same way as described in Chapter Two. In stopped-flow studies, the









following buffer-indicator pairs were used (also given are pKa values,

wavelengths used and differences in molar extinction coefficients of

acidic and basic forms): 4-morpholineethanesulfonic acid (Mes) (pKa =

6.1) and 3,5-lutidine (pKa = 6.2) with chlorophenol red (pKa = 6.3, X =

574 nm, AE = 1.77 x 104 M-1 cm-1); 1,4-piperazinebis(ethanesulfonic

acid) (Pipes) (pKa = 6.8) with bromocresol purple (pKa = 6.8, X =
588 nM, As = 3.0 x 104 M-1 cm-1); 4-morpholinepropanesulfonic acid

(Mops) (pKa = 7.1) and 1-methylimidazol (pKa = 7.1) with p-nitrophenol

(pKa = 7.1, X = 400 nm, As = 1.83 x 104 M-1 cm-1); 4-(2-hydroxethyl)-l-
piperazineethanesulfonic acid (Hepes) (pKa = 7.5) with phenol red

(pKa = 7.5, X = 557 nm, As = 5.59 x 104 M-1 cm-1); 1,2-dimethylimidazole
(pKa = 8.2) with m-cresol purple (pKa = 8.3, X = 578 nM, Ae = 3.81 x
104 M-1 cm-1); 1,4-diazabicyclo-[2.2.2]octane (Ted) (pKa = 9.0) with
thymol blue (pKa = 8.9, X = 590 nM, As = 2.43 x 104 M-1 cm-1).

Measurements of esterase activity were performed at room

temperature similar to previously described procedures (34,35) using
p-nitrophenyl acetate as substrate. The hydrolysis of p-nitrophenyl

acetate was followed by measuring the appearance of p-nitrophenoxide
and p-nitrophenol at 400 nM by using a Beckman spectrophotometer Model
UV 5260. At this wavelength, e = 1.83 x 104 M-1 cm-1 for

p-nitrophenoxide and E = 200 M-1 cm-1 for p-nitrophenol were used (34).
A pKa of 7.14 for p-nitrophenol was assumed (36).
The following procedure was used for activity measurements. A
fresh 40 nmM solution of p-nitrophenol acetate was prepared in 10 ml
acetone and was kept in ice during the experiment. First, enzyme was
mixed with 50 mM buffer/33 mM Na2SO4 solution in a 1.5 ml volume









cuvette. Then to that 50 1l of substrate solution was added to give

2 mM final concentration in 1 ml total reaction volume. This procedure

was repeated at 5 different enzyme concentrations from 0.6 x 10-5 to

2.2 x 10-5 M at each pH, also the uncatalyzed rates were determined

under the same conditions without enzyme and subtracted from the

catalyzed rates. Data were plotted as the rates on the Y and enzyme

concentrations on the X coordinate and the slopes of these lines which

are kenz [S] were obtained by using a least squares method (22,23).

kenz is equivalent to kcat/Km only when the concentration of substrate
is lower than the values of Km.

Results
CO2
Hydration activity. The turnover number kct and Michaelis-Menten
CO2
constant Km2 for the hydration of CO2 catalyzed by cat carbonic

anhydrase III were independent of pH in the pH range from 5.7 to 8.0 as

was also observed by Tu et al. (15). When pH was increased above 8,
CO2 CO2
both kcat and K2 increased and did not reach plateau at pH 9.6
CO2 CO2
(Figure 6, open symbols). Although the increases in kcat/Km above pH
CO2
8 were not as large as the increase in kcat, at pH 9.6 the value of
CO2 CO2
kcat/Km was more than doubled with reference to its value in the pH
region 5.7 to 8.0 (Figure 6). In Table 2, the average values of these

kinetic parameters at the indicated pH values are given. The line

drawn through the experimental points in Figure 6 was obtained by using
the equation
S (kCO2 )O K
kcat = (kcatl ) x cat2 x E2 (7)
KEI + [H ] KE2 + [H+]




























e-
0 4-) X CO Wt
NCi .C- 0 r- r- r-
C3:= S- 1 ) c 1 m : .- U O o
C,4-.o0 C =C ) 3 (4.) .0 C


OCE ( E 4 0 >4- U
0o r- 0 M a C2 o ea
O C U E E C:O EC-- ---- U U CL.
0 0 I C = C+14- E 0 3
4- 4- E n C COJ L- 00 E 3 4- 0 ,-
o e CN O 'C-,i c L o Ou
S- 4-3 *e -- I- Ar-- l4---X.
c i aOEU . V L >UM CC (V 4- -0O
C r i-- < E *-*r *. ..C 1-4 CIC
W E O O r- at -- -C --4 0] -
W l 0 *- 00 0
c -9-- S E a c <1 u (..) E &a n
Ci4-4ME- > l N CX 0 -- W. ) c-- : to tf.- 0-. .O O
-Q) 4- c) 4- -, > C C 4-l S C -U
-0 0 --0 *, 4- E C 0 ,U E-
iA n *r- -0 0 U>4 C C
SS O Q J- *a a- O O C L, 4-)

E c 0 *- E a O .0 W c
.C (X .-) 3: E E 5- 0 "r- 0
>,oCo cia -4 UC._ai)
L o "o o r--4 0 CO EL
)C U Or-( 0C II C O i, Iu
tu (n OE D (E AL
Sc u o a oj E 0 4 a aC u c a
*r 0) co i3 O *- CI CU


r c *Or- *O 0 -r "0
r- J= 00 a -4 a )O C
CD 0> C (* V. C N N CJO
a4- 00 0 -0 00. .L-,
-O .0E- C E (A 4- C
.a o + *- Er *E U. *r- C
E- E *C ar *- i- to- a (o
d)lJ E >OU .S-'- a U1 )
C. 4- N VI0 4 0 4-> >, -. r- 0 4- O 0)
.-r- > LO aC Ll r X cu U) co CMI-i-
Slr- COJ W ECJ > W0 C3 0
a EU 0 E U 0 (1 3 t) 4-O. W E c ftt 10 < E
= 0 O ( O *r- : rCO C4- a )
F- r- U : UM "43 4-3 -4 o ER(














































OC c.


0 0


"I


('4
N
oZ








TABLE 2
THE VALUES OF THE KINETIC PARAMETERS FOR THE HYDRATION,
DEHYDRATION AND p-NITRO PHENYLACETATE ESTERASE ACTIVITY
OF CARBONIC ANHYDRASE III AT THE INDICATED pH VALUES


Parameter Value at Low pH Value at High pH Dimension


a C02
kcat

b(C02)


c C02 C02
kcatKm


d C02
dcat

d C02
-Km


(cat) (1.8 0.5) x 103


(cat)


12 5


(cat) (1.4 0.5) x 105


(bovine) (1.2 0.1) x 103


(bovine)


9 1


(1.0 0.3) x


( 30 10 )


( 3 0.3) x


(1.2 0.2) x 104


26 + 6


105 M-1 sec-1


sec-1


e HCO3 HCO3
4cat /Km (cat)


7.5 x 104


(cat) 0.6 0.2


4.6 x 103

4.5 0.8


M-1 sec-1

M-1 sec-1


M-1 sec-1


(rabbit) 1.2


abcValues at low pH and high pH were calculated by taking the
average values of the kinetic parameters in the pH region from 5.7 to
7.8, and at pH 9.5 and presented with their standard deviations.
-Data at pH 7.4 (in 30 mM Hepes) and pH 9.3 (in 30 mM Ted).

-Low pH value is 6.0 and high pH value is 7.4.
f
-k z esterasee reaction) from this study: low pH = 6.7; high
pH = 8.
k from Koester et al. (33): low pH = 6.7; high pH = 8.7.
enz


kenz

kenz









where (kc ) and (kct 2)" are the maximum values of kca in the pH
cat cat2 CO2 in the pH
regions 6.0 to 8.0 and greater than 9.6, KE1 and KE2 are the acid
dissociation constants of the first and second enzymic group which
control the hydration activity of isozyme III (see Figure 6 for the
values of these parameters). A result similar to that given in figure
6 (open symbols) was also observed by using bovine carbonic anhydrase
CO2 CO2
III. The values of kcat and K, at the indicated pH values are given
in Table 2.
HCo3 HCo3
Dehydration activity. The pH dependence of kcat /K 3 for the
dehydration of HCO3 catalyzed by cat isozyme III is presented in Figure
7. Because of the difficulties in getting sufficiently accurate data
due to the uncatalyzed rates being as high as 25-40% of the catalyzed
HCO3
rates, and also high values of K, with high uncertainties
especially at increasing pH values, no attempt was made to interpret
HC0o3 HCO3
kcat and Km values individually. The data shown in Figure 7 are in
approximate accordance with the following expression which describes
the observed increase in activity when pH is lowered:

HC03 HCO3
kcat kcat [H+]
x (8)
HC05 HC03
Km [H+ + KE1
where the subscript reflects to a maximum value (see Figure legend 7
for the values of parameters used in this fitting). In Table 2, the
HC03 HC03
values of kcat /Km at low and high pH values are given.
The observed Michaelis-Menten parameters given in Figures 6 and 7
can be related through Haldane relationship (equation 9) to the
equilibrium constant of CO2 and HCO3 (equation 10).

























0
LD
C.j

4 V) 1 C
in e in *) en
o 4 C') I-- ..

O Q ) M ( A
Q. 4- -C 3 C) C4
4- 0 o e-
a) o) <(a M
C C) 0 C rIo II
0 S- C fI *r- L

,- = n', 0 4- .-
P > U) U) . .. L.

-H L 4- 0
> 0 .4-- o,






0 a- 0 -
4- r- 0 M 0
oU E t 0 C)
C) c ci- C:
L U -- '-Q .C

o t-0 c o *4-
*r- 0 S r- *3-
0u E 0 o oC
tr *r- 1 o + r-- rE



C- C f 4- W S- C
. O1 ,. 0 O X-

I- E ac E- E
4-) ( S S *r- X II
S- U C0 4 C) 4--
co U. 3- 4 C- 0



M4- CO (0 C i o
0 j 0 3 o. -
u- Go o .o








LLO
y + o aO anc co
C> rcO 'O
E5 U0 C :1 E



*a AI *i- *r- 2 ra I ca

-J U Z Zr- *I ) J -- -^

*



-r

LL-











0
CO




0
I-







DV (


OOH
V V





IIj







CO2 HCO3
kcat Km Ke
x I - eg- a (9)
CO2 HC03
K, kcat [H+]
where Keq is

S= [HCO][H ]
K 3 (10)
eq [C02]

Insertion of the data of Figure 6 and 7 in equation 9 yields Keq
1.7 x 10-6 M which is comparable to the literature value 9.0 x 10-7 M

(37).

Esterase activity. The pH profile of the hydrolysis of
p-nitrophenyl acetate catalyzed by cat isozyme III is shown in Figure

8. Because of the limited solubility of p-nitrophenyl acetate, kenz was

measured. As can be seen, the shape of Figure 8 looks similar to that

of Figure 6 in that kenz is also independent of pH in the pH range 6.2

to 7.6 and increases above pH 7.6. The line drawn through the
experimental points in Figure 8 was obtained using equation 7 by taking

the same values of KE1 and KE2 listed in the legend to Figure 6 for CO2

hydration. The values of kenz at the indicated pH values are given in

Table 2 together with the values obtained by Koester et al. (33) using

rabbit isozyme III.

Discussion
CO2 CO2
The observation of a plateau for kcat and Km in the pH region 6.0
to 8.0 is the same as Tu et al. have found (15). This plateau of

activity could very well be the high pH arm of a pH-rate profile having
an activity controlling group with a pKa lower than 6.0. The
dehydration pH profile (Figure 7) certainly indicates such a low pKa.






























) Q) II

4 Cr-- O r-*-
i- +CS = s a CL.
-- t4- -J-- C '_. r Q M-

0 a Q. )U E .: I-
4- U O s- I- "a )U *,
L ,I- 0 < E-.
C.0 *r- *r- S U
W Z U"aO (o tc rc
-. "O I -- C -) 0. ut
0 Q. C 4) 0 .0
4- N 0 1 0 ) -4
C) 4 -V) r

o- t*o L to te 0
>0 u C cI (n
0 CO O c to I
,-O I C)
4M) 4-1 ) 0 -- o
4-) M -4 r-'
0M4-)' Cn C co "
L C x. X c to > M4-)4->
u r- (0
-. 0 t 0 t-) 0 0
S0r-- J + c .-..--
S- *M Q-' C- I
r- (A 1 *r- C
O C )U l EO *-
0 > C *r- r--4
O .0 3 -.. -- -
c: Q. ,, i o
OLU OS X U)
0 0 *- C 0
/) *- W) +) ( 4 EI
ct Lw 0o Cu ) u +- i
S to S- ..C= .
QS- 4- 4J 4- 0
41 0C 4- to *
4- >c i >4- .C *.- C
4- o .c u .0 03: o
o CC .0 05 11
ti o x o. II
E-- 00 U C l
Lc) 0 C C.C Co L
4- *>- *r- *r- 0 2
*r- C 4-C E : .r- --IQ.
5- 0 0 0 S- 0 C 41+4-4
S0 L. .0 4-) > -- r t0co )A *O
o > o U C) 0 L- 0- -, *
-J LC u 0 t4- -0 0) LO


CO
D)
S-


LL.















0











IcD


CL







(D

0 LO 0
0 0

ZN31 Bol









So it appears that in the pH region 6 to 8, CO2 hydration activity of

cat isozyme III depends on the basic and the dehydration activity

depends on the acidic form of this catalytic group with pKa lower than

6.0. Therefore, if the velocities of the hydration reaction catalyzed

by isozyme III are measured above pH 6.0, only the high plateau region

of pH profile would be seen (Figure 6); where, in the same pH region,

the dehydration pH-rate profile would be pH dependent with increasing

activities as the pH decreases (Figure 7). This result is similar to

what is observed for isozyme II, except its activity controlling group

has a pKa of 7.0.

I suggest that this ionizable group with a pKa lower than 6.0 is

zinc-bound H20. Further supportive evidence for this hypothesis will

be presented in the following Chapters from the anion inhibition and

solvent deuterium isotope effect studies. Moreover, Engberg and

Lindskog (38) have found that Co(II)-substituted bovine carbonic

anhydrase III has a visible spectrum at pH > 6.0 very similar to that
CO2
of Co(II)-substituted isozyme II at pH > 8. The pH profiles of kcat for

isozyme III from bovine and cat look similar, which suggests that the

active site regions of both enzymes are comparable. This low pKa for

H20 in isozyme III is perhaps a result of the positively charged

environment created by lysine 64 and arginine 91 which are found in the

active site cleft for bovine isozyme III (39), which could stabilize a
CO2 CO2
zinc bound hydroxide. The cause of the increase in kcat and Km at pH

> 8 could be due to the ionization of these basic amino acid residues,

especially lys 64 which would become a proton shuttle group as pH is

increased, and enhance activity by providing a more efficient pathway









for transfer of protons from the active site to solution. In this

respect, the function of lys 64 in isozyme III at pH 9 resembles that

of His 64 in isozyme II at pH 7 (1). It is also possible that this
CO2
increase in kcat could be caused by an effect of hydroxide ion in

solution acting as a proton acceptor in this catalysis that produces a

proton in addition to HC03. A turnover number 103 sec-1 is consistent

with the maximal rate of proton transfer from a group of pK, around 5

(zinc bound water in isozyme III) to H20 in solution. Hydroxide ion in

solution, being a much better proton acceptor, would be expected to
CO2
increase the rate at alkaline pH. The appearance of pH dependent kcat
CO2 CO2
and pH independent kcat/Km above pH 8 suggests that the change in
CO2
ionization state observed for kcat occurs in an enzyme substrate

complex not in the free enzyme (40).

As shown in Table 2, the values of kenz obtained in this study by

using cat isozyme III agree with the corresponding values obtained by

Koester et al. (33) using rabbit isozyme III at the indicated pH

values. But, in contrast to their result which showed a pKa 7.5 for

the activity controlling group, the present study shows that the

hydrolysis of p-nitrophenyl acetate catalyzed by cat isozyme III depends
CO2
on the ionizations of the same catalytic groups which effect kcat

(Figure 6) with pKa values lower than 6.0 and 9.0.
In summary, the results presented in this chapter are consistent
with carbonic anhydrase III from cat having at least two active site
groups with pKa values lower than 6.0 and around 9.0. I suggest that

the catalytic group with the low pKa is zinc-bound H20.













CHAPTER FOUR
RATE-CONTRIBUTING EVENTS IN THE HYDRATION OF
CO2 CATALYZED BY CARBONIC ANHYDRASE III
FROM CAT SKELETAL MUSCLE

Introduction
In a recent study, Tu et al. (15) by using 180 exchange technique
showed that the magnitudes of the solvent hydrogen isotope effects on
exch CO2
kinetic parameters kcat /Keff and kH2 (see Chapter One for their

meaning) obtained from the hydration of CO2 catalyzed by cat isozyme

III were unity and 2.5, respectively. This result suggested that the

rate-contributing event in hydration of CO2 catalyzed by cat carbonic

anhydrase III could also be a proton transfer event located outside the

interconversion of CO2 and HC03 as observed for isozyme II (1,2).
exch CO2 CO2 CO2
Since kc /Keff and kH20 are described to be equivalent to kcat/Km,
cat /eff an k a
CO2
and kcat at steady state (9), such a result requires that the
CO2 CO2 CO2
magnitudes of isotope effects on kcat/Km and kcat measured at steady

state should be close to those obtained in 180 exchange experiments.

The effect of buffers on the release of product proton in the
hydration reaction catalyzed by isozyme II was discussed in Chapter

One. It was not known until this work what effect buffers have on
hydration of CO2 catalyzed by isozyme III.

In order to understand the rate-contributing events, the rates of

CO2 hydration catalyzed by cat carbonic anhydrase III were measured in
H20 and 98% D20 in the pH interval from 6.2 to 9.5 by using









stopped-flow technique. The effect of buffers on the hydration
reaction was studied by using stopped-flow and 180 exchange techniques.
The results of my experiments support the hypothesis of Tu et al. (15)
that the rate determining event in the catalysis of the hydration
reaction by cat isozyme III is a proton transfer located outside the

interconversion of CO2 and HC03. The rate of the release of product
proton to solution is not rate limiting at any concentrations of
buffers.

Experimental Procedure
The materials and kinetic methods were similar to that described in
Chapter Three. The concentration of saturated solutions of CO2 in D20
was taken as 3.8 x 10-2 M (20).
The values of pH(D) reported here are uncorrected pH meter
readings. The correction of a pH meter reading in 100% D20 CpD =
(meter reading) + 0.4] (41) is approximately offset by the change in
.the ionization constant of buffer in D20 (pKD pKH = 0.5 0.1) (42).
Consequently, one expects that all reactive species maintain
approximately the same ratio of acidic to basic forms in two
experiments, one in H20 and one in D20 at the same value of the pH
meter reading. The magnitudes of solvent deuterium isotope effects on
kinetic parameters were determined by simply dividing the value of the
parameter in H20 by that in D20.
Equilibrium 180-exchange experiments were performed as described by
Silverman et al. (9).

Results
CO2 CO2 C02
The values of kat and kcat/Km obtained in H20 and 98% D20 in the
pH range from 6.0 to 9.5 are presented in Figure 6. In Figure 9, one































C > N NC,
0 C* +
*- 0 -o CMJ C"4-
u 4-) a) *C 4-) +1 (_) u
O ---
0a E =3
L- .o Ln a 0 0 Rt
(O LO -f 4-) S- -
< CJ J 40) 4-)
-= 4-2 O0 0 IIC1I


0n o a
to V) II u E
W :0 -*0 c'C
MV) 3 N--O < CO

r- A0 V4->C --

.iE 0 ( N
U w 0 m 11E
-0 0 .In I I

*e I) C ) ) *
<0O S 1 0 M
4- O C. i Vn C. )




4'I 0 CM C C'. 0
OcO ..C oC C\I
3C a a u
E l- i a 1 CM


M CA ro (A X *
V) 0O U o ( 1
O C4- C >- *0C
M 3 ca U *r- .--





CL- ( OOCC -
0 C o
MS:6 (JO4 Ob- C C



L 0r 0 L-- Na N)
a Qoo o M PO 04
4m cr> c = (nl U-







00 C0 = Q) C M





rC O S- CO C
C0- CO) O4 E C_
O OC 0 o) a (A N N d-
0 4o-3 4-1 M
L- C:) o

1) W a (1 At 0 C-) E














LL






48










0
-0
K)



-0

ro
CJM



S0
^L -









of those hydration experiments performed at pH = 6.6 in H20 and 98% D20

is given in the double reciprocal form. The average values of the
CO2 CO2 CO2 CO2
solvent hydrogen isotope effects on kcat, Km and kcat/K obtained

from the data in the pH range 6.2 to 9.5 are tabulated in Table 3. The
CO2
ionizations which determine the behavior of kcat for cat isozyme III

are consistent with a change of pK, of (PKa)D20 (pKaH20 = 0.5,
which is the value commonly observed for weak acids with values of pKa

in the range of 5 to 10 and close to that for the ionization constant

for dissociation of H20 (0.6) (43).

By using the stopped-flow technique, it was observed that

concentrations of 2 to 50 mM of Mops at pH 7.1, Hepes at pH 7.5, and

Ted at pH 9.0 caused no change in the rate of hydration of CO2

catalyzed by cat isozyme III. In agreement with this, there was no

change in the catalyzed 180-exchange rates in the presence of up to

50 mM mops and hepes compared with the 180-exchange in the absence of

buffers at pH 7.1.

In contrast to the buffers mentioned above, imidazole caused an

increase in the rates of CO2 hydration studied by both methods. In the
180-exchange experiments, the effect of imidazole was mainly on kH20o

the rate constant corresponding to the release of 180-labeled H20 from

the active site (Figure 10). The values of kH20 in the presence of
30 amMt imidazole were almost doubled with reference to its value in the

absence of imidazole at pH 7.1. The value of the other 180-exchange
exch CO2
parameter kcat /Keff showed a consistent 15 to 25% drop at all

concentrations of imidazole with reference to its value in the absence
of imidazole (data not shown). The effect of imidazole on hydration of







TABLE 3
SOLVENT HYDROGEN ISOTOPE EFFECTS FOR STEADY-STATE AND
EQUILIBRIUM EXCHANGE CONSTANTS IN THE CATALYSIS OF CO2 HYDRATION


D C02a D C02 CO2 D exch CO2 D
kcat cat/Km ) (kcat /Keff) kH20


Human
Isozyme II 3.8b 1.0b 1.0 0.1c 4.0 0.5C

Cat
Isozyme III 2.6 0.3d 1.3 0.2d 1.0 0.e 2.4 0.4e

a. The superscript D indicates solvent hydrogen isotope effect. Thus,
D CO2 ( cat (kCO2
kcat is kcat)H20/kcat)D20
b. Data from the high pH plateau (pH > 8) of Steiner et al. (1).
c. Data at uncorrected pH meter reading 8.0 from reference 10.
d. The average values of solvent hydrogen isotope effects obtained
from Figure 6 (this work) presented with their corresponding
standard deviations.
e. Data at uncorrected pH meter reading 8.0 from reference 15.























r-




co 0 0 *l1
*r- 0

*-I 4- .0 *r- 10
0 S- 4-) Z
C rt- .
r-0 0 O-
C 4- U) *r-
*r- to0 3:
0to 0 0 t-- *
4-) S- 0 L Mj
C 4- 4- I ()
0 C: 0 CO

COE tC
0) 0 *r- ) +

0 S- **(- 0=


O *o C- --
0 0 "C.'-j
O OJ C 4-




c- ..) I '-
C C 0 c 0 o

rO *- uO





Q &- m-
C *M I
re r> .ci
-- Uf 0 E
r N O
-J= ) i0 ) C-
4- d *r- No 4-)
0 0 Ca X a
C"a r10 J= a)
3I S- 0 u U

U *0 r- o* 0
4-J ru u
c E w c
(a 0o a (a to 4)
4- I- i. rE, 4-
(n 4- (0
0 +> 1 CJ S..
0O toJ0-0 4- C
U ) l-" 4-S (n
L r3- (1) C 3
a) 0 0 CU

43 to-

L CL 4-




0S
S w o




*r-





L-




































OI xF13 s


0
(0




o,,
O2
0
N
Q


0






0


OZH ]









CO2 catalyzed by cat carbonic anhydrase III at steady state was similar
CO2 C02
to that shown in Figure 10. The values of kcat and Km obtained in

20 mM imidazole at pH 7.1 were 1.5 x 104 sec-1 and 50 mM, respectively.

These values are 5 to 6 times higher than the values listed in Table 2

using other buffers. By using fluorescence, a quenching method which

will be described in Chapter Five in detail, imidazole, up to 70 mM

concentrations at pH 7.1, did not affect the fluorescence of the

chlorzolamide-cat carbonic anhydrase III complex at 330 nM. Similar

to this, imidazole up to 0.2 M concentrations at pH 9.0 did not affect

the fluorescence of DNSA-human isozyme II complex at 460 mM. However,

the fluorescence of DNSA-human isozyme I complex at pH 9.0 was

decreased by imidazole with a value of IC50 -50 mM.

Discussion
There is a very large difference in maximal turnover for hydration

of CO2 between isozymes II and III at high pH with a difference of
CO2 CO2
similar magnitude between kcat/K, (Table 1). Despite this difference,

I would like to point out similarities in the kinetic features of these

isozymes II and III which strongly suggest that they have in common a

rate-contributing event occurring outside the steps involved in the

actual interconversion of CO2 and HCO3. This event is suggested to be

a proton transfer for isozyme II and may be a proton transfer or

conformational change for isozyme III (1,2). These conclusions are

drawn from two types of experiments: 1) a comparison of equilibrium
180-exchange data with steady-state results and 2) solvent hydrogen

isotope effects on the 180-exchange and steady-state constants.









Comparison of 180 with steady state. Table 4 compares two

constants obtained by equilibrium 180 exchange with the steady-state
HC03 exch
turnover number for dehydration, kcat The rate constant kcat which

was obtained from the rate of exchange of 180 between CO2 and H20 (9),

and has also been determined from the exchange broadening of the 13C

resonances of H13C03 and 13C02 (44), is the maximal rate constant for
interconversion of CO2 and HCO3 at chemical equilibrium. Table 4 shows

that this rate constant is greater than either the steady-state
HCO3
turnover number for dehydration kcat or kH20 (The rate constant kH20

measures the release from the enzyme of oxygen abstracted from
substrate bicarbonate [9].) This implies that the turnover rate at

steady state is limited by a step outside those involved in the
HCO3
interconversion of CO2 and HCO3. The constants (kHct )m and
2- 3 cat max
(kH20)ma are approximately equal, suggesting that both are determined
by this rate-limiting step. For isozyme II, Simonsson et al. (44)
noted that such data are consistent with an intramolecular proton
transfer between zinc-bound water and an ionizable group in the
active-site cleft. That this step involved proton transfer was
indicated from the solvent hydrogen isotope effects.

Solvent hydrogen isotope effects. For the hydration of CO2
catalyzed by human carbonic anhydrase II and bovine carbonic anhydrase
II, Steiner et al. (1) and Pocker and Bjorkquist (2) showed that the
CO2 D CO2
solvent hydrogen isotope effect of kcat written kcat, was close to 3.8

for the high pH arm of hydration (pH > 8). However, there was a
CO2 CO2
solvent hydrogen isotope effect of unity for kcat/Km The ratio
CO2 CO2
kcat/Km contains rate constants for steps up through the first









TABLE 4

MAXIMAL VALUES FOR SOME STEADY STATE AND EQUILIBRIUM
EXCHANGE CONSTANTS IN THE CATALYSIS OF CO2 HYDRATIONa


ex HCOl
(kcat max (kH20)max (kcat3)ax

(x10-5)s-1 (x10-5)s-1 (x10-5)s-1


Human isozyme II 18 10 6

Human isozyme I 0.4b 0.3 0.4b

Cat isozyme III >0.1 0.02 0.02


a.
pH 6 to


kcat is independent of pH for isozymes I and II in the range of
8 (44,45) The value of kcat for isozyme III was determined at


pH 6.6 and 7.1. Maximal values of kH20 occur near pH 6.5 for isozymes
II (9) and I. kH20 is independent of pH between 6 and 8 for isozyme
HCO3
III. kcat has a maximum near pH 6 for isozyme II and I. For isozyme
HCO3
III the value of kcat was obtained at pH 6.0.
b. From reference 45.









irreversible step which, in the initial velocity experiments, is the
D CO2 CO2
release of HCO3 from the enzyme. The value of unity for (k t/Km )

for isozyme II strongly suggests no rate-contributing proton transfer

in the steps involved in the interconversion of CO2 and HC03. The 180
D exch CO2
exchange experiments confirm this in that (kcat /Keff) is also unity
exch CO2 CO2 CO2
(the ratio kcat /Keff is equivalent to kcat/Km ). However, the value
of kcat of 3.8 indicates a primary proton transfer somewhere in the

catalysis. The 180 exchange experiments reveal a large solvent
hydrogen isotope effect on kH20 for isozyme II, as great as 8 at pH
6.6, indicating that the rate-limiting proton transfer is also involved

in the steps that release 180 from the enzyme.

There is a rather close parallel between the results described here
for isozyme II (at pH > 8) and those for cat isozyme III as shown in
D CO2 CO2 D exch CO2 D CO2
Table 3: (kcat/Km and (kcat /Keff) are close to unity, kcat and
D CO2
kH20 are very significantly different from unity. We therefore
suggest, in anology with isozyme II, that the hydration of CO2

catalyzed by cat isozyme III is limited in turnover rate by a step

occurring outside the interconversion of CO2 and HCO3 and that this
step involves the changing of bonding to hydrogen.

There is no evidence to suggest the nature of the rate-limiting
event for hydration of CO2 catalyzed by cat isozyme III. A turnover
number of 104 sec-1 at pH 9 seems a bit slow to be determined in rate

by an intramolecular proton transfer, although this possibility cannot
be eliminated (31). It is part of the hypothesis of Steiner et al. for
isozyme II that buffers are not able to transfer a proton directly with
the active site group but require an intermediate proton shuttle group









(1). One possibility for isozyme III is that there is no suitable

residue near the active site which can act as a proton transfer group

between the active site and buffers in solution. Then, isozyme III

would have to transfer a proton to water at the slow rate of 103 to 104

sec-1 which can very well be the rate-limiting step in the overall

hydration reaction in H20 and D20. This would also produce no buffer

effect in the catalyzed hydration reaction.

The effect of imidazole on the hydration of CO2 catalyzed by cat

isozyme III is not clear yet. Tu et al. (46) also observed that the

phosphate buffer at 25 to 50 mM concentrations caused 100% increase in

the 180 exchange parameter kH20 obtained using cat isozyme III. It is

therefore possible to suggest that buffers such as imidazole and

phosphate which have relatively smaller sizes than buffers Mops, Hepes

and Ted can get into the active site and by accepting the proton from

zinc-bound H20 can increase the rate of the hydration reaction. The

rates of hydration of CO2 obtained by using isozyme II are also

several-fold faster in imidazole buffer (47). Rowlett and Silverman

(32) suggested that this could be due to the ability of imidazole to

get into the active site of isozyme II and short-circuit the proton

shuttling step. Imidazole is a specific inhibitor of CO2 hydration

catalyzed by isozyme I with a KI of 18 mM (21) and is shown to have a

specific binding site within the active site (8). It is apparent from

the fluorescent experiments that the effect of imidazole on the

hydration of CO2 catalyzed by isozymes II and III is not due to its

binding anywhere in the active site region of these isozymes. However,

the decrease caused by imidazole in the fluorescence of isozyme I-DISA






58


complex suggests that inhibition of isozyme I is due to the binding of

imidazole to a specific site in the active site region, consistent with

x-ray data (8).

In summary, carbonic anhydrase III from cat has a rate-limiting

proton transfer which comes after the interconversion of CO2 and HCO3

as found for isozyme II. However, in contrast to isozyme II, isozyme

III is not enhanced by buffer in its catalysis.













CHAPTER FIVE
INTERACTION OF CAT CARBONIC ANHYDRASE III
WITH ANIONS AND SULFONAMIDES

Introduction

Anions and sulfonamides are inhibitors of carbonic anhydrase

isozymes which have been shown to coordinate to the active site metal
ion, zinc (8,48,49). In the past, they have been very useful to
elucidate the kinetic mechanisms of isozyme I and II. Anions were
shown to inhibit CO2 hydration catalyzed by human carbonic anhydrase I
in a noncompetitive fashion at all pH values (50,51). The
corresponding pattern for carbonic anhydrase II was noncompetitive at
low pH values but was changed to uncompetitive at pH > 8 (50,52). This
difference in the inhibition patterns is perhaps the reflection of the
different rate limiting steps in the kinetic mechanisms of isozymes I
and II. An uncompetitive type of inhibition for isoyzme II was
explained to be consistent with the enzymic model given in Scheme 1
(7,32).
Sulfonamides are potent inhibitors of isozymes I and II (53).
Therefore they equilibrate rather slowly with the enzyme while the

catalysis precedes and at steady state behave as pseudo-irreversible
inhibitors and give rise to noncompetitive patterns with respect to

CO2. But a competitive behavior with respect to HCO3 has been reported
(54).

Sanyal et al. (14) previously reported that the IC50 values of many

anions obtained from the inhibition of the hydration of CO2 catalyzed








by cat isozyme III were lower than the corresponding numbers obtained

for isozyme II. In their study, they have also shown that the IC50

values of many sulfonamides, such as ethoxzolamide and acetazolamide

for cat isozyme III, were much higher than the corresponding number for

isozyme I and II (IC50 value of acetazolamide for isozyme II is 10-8 M

and for isozyme III is 3.0 x 10-4 M [14]. These results suggested that

the active site of cat carbonic anhydrase III must be quite different

than the active sites of the other isozymes.

To understand further the kinetic mechanism of cat isozyme III and

also the residues which control the binding of anions and sulfonamides,

the inhibition by these inhibitors of the hydration of CO2 catalyzed

by cat isozyme III was measured and the pattern of inhibition and

magnitude of KI as a function of pH was determined by using a

stopped-flow technique. Direct interaction of anions and sulfonamides

with the active site of cat isozyme III was investigated by using a

fluorescence method.

The results of experiments further support the hypothesis that the

pKa of the catalytic group in cat isozyme III, most likely a zinc bound

water, is lower than 6.0 and that the catalysis by this enzyme is

limited in rate by a step occurring outside the interconversion of CO2

and HC03. The results also show that another active site residue with

pKa higher than 8 affects the kinetics of inhibition by anions.

Experimental Procedure

Materials. Chlorzolamide (5-[o-chlorophenyl]-1,3,4-thiadiazole-

2-sulfonamide) was obtained from Lederle Laboratories (Pearl River,

NY). Its molar extinction coefficient was determined to be 1.5 x 104









M-1 cm-1 at 280 nm and 1.0 x 103 M-1 cm-1 at 330 nm. Ethoxzolamide was

obtained from Upjohn Drug Company. The salts, NaN3 and KI, were

obtained from Aldrich Chemical Company. The rest of the materials were

the same as described in Chapter Four.

In inhibition experiments, N3 and I- were chosen as inhibitors of

the hydration of CO2 catalyzed by cat isozyme III because of their IC50

values which were reported to be relatively higher than some of the

other anions such as CN~ and OCN- (14) and also higher than the enzyme

concentrations employed in these experiments ("2.0 x 10-6 M). This

precaution was taken to avoid any tight binding by those inhibitors

which will equilibrate with the enzyme slowly while the catalysis is

proceeding at steady state and therefore would effect the results of

inhibition experiments. In stopped-flow procedures, these inhibitors,

including sulfonamides, were mixed with the solution containing enzyme,

buffer and indicator prior to the mixing with substrate. When the

inhibitor iodide was mixed with substrate solution prior to the mixing

with the enzyme, there was no difference in inhibition results.

Fluorescence Quenching. We observed the quenching of the

tryptophan fluorescence of cat carbonic anhydrase III caused by the

binding of the inhibitor chlorzolamide, a sulfonamide. Solutions of

purified enzyme (at 2.4 x 10-7 M) were irradiated at 280 nm and the

emissions were recorded at 330 nm (Perkin-Elmer MPF-44B

spectrophotometer using cells of 1.0 cm pathlength). Solutions

contained 50 mM buffer (see Figure Legends), 1.0 mM dithiothreitol to

prevent dimer formation, and no sodium sulfate. Titration with










chlorzolamide was carried out by repeated additions of a chlorzolamide

solution into enzyme solution; dilution at the end of the titration was

less than 2% of the total volume of the enzyme solution (2.5 ml). The

observed intensities of fluorescence were corrected for self-absorption

of incident light at 280 nm by chlorzolamide using the relationship in

Equation 4. The correction due to absorption by chlorzolamide of light

emitted at 330 nm was negligible. In experiments involving both

chlorzolamide and anion (CNO-), the anion was added first at the

desired concentration prior to repeated additions of chlorzolamide. In

all experiments the dissociation constant of inhibitor is much greater

than the enzyme concentration; hence, no correction was made for the

amount of chlorzolamide bound to the enzyme's active site.

Data are presented and were analyzed using this Eadie-Hofstee form

of the Langmuir equation
AF Fmax AF chlor + anionn] (11)
AF=AF .K 1+ (11)
max C diss Kanion
diss

in which AF is the difference in fluorescence intensity of protein in
chlor
the absence and presence of chlorzolamide at concentration C, Kdiss
anion
is the apparent dissociation constant for chlorzolamide and Kdiss

for the anion. Values of Kdiss were determined by least-squares fits

of the slopes of plots of AF versus AF/C.
Results

The inhibition by sodium azide of the hydration of CO2 catalyzed by

cat carbonic anhydrase III at pH 6.0, 7.0 and 9.5 is shown in Figures

11, 12 and 13. These data were fit by a least-squares method (22,23)

to Equation 5.
































l-C
O 0
4- u j4- "r X
0 4- 4- -m

2 *- 4J
c0 o -C


t t: 0
O C r- E 0
"o c" E LO, )
"r 0 (a 1-
4-3 JQ *- "M X ,-
S* C 0 m r- ) X
C* fuo 0C CO .
.C C XJ 0 3 0 (O
C: 4-> ( C) J C) 0
C O 4.-) 4 =f=l (. C
) o 0 Co ) C +1

01. 0 ) < =') 4U-
$ CML O *
4-- O 4) Q) 4-- = 0

C L ( 1 CT O CM
*r- N ( A U 4- It
>- I0 4- 0 II
o a-- 0 9U) r-- eCJ4-
= to -. ) 0-) r-0 0 4-
U) 43 O Z 3r 0(. (JC T-
4- u I- / O S
o cj 0)) 0) (D
-- CM 0 .M.C -C 4J
1 (A "a fu C E
- : o C C 0
') 4-) X 4
o I U)- 0 0
. C00 30 t > o 1

Ca *- ) U a 0
U43> X30 ** *
S*L,) oC -o L
V0 .060 *r-S S.-
S S S 0 II
O I 0 C O

CO 1. 4-l (0 C0
(1) 4JCI a 4-) 0
0-3 *-4 4M 06).C M v-E



'-4


t-


LL





64








0





S Io

0 O
o !

O Z-0
\U
\


0
x
V)


-I>































4-
0
4- a)
O'nE

ro t o
C.V)..





0 r-
4 .0 +J
CU O


C 4-
U '


t0 C



0 JC
r- Ci 01





.C O




o to yi



0

0 0

0 --
0- *






U O -
*r *- 0


0 *C
S.0
QI- C


X

n r~r-i C

C) (D
CL X )*


3S t) II cM

0-Q L)
U* CT M *-


4/) I II
c 3E E C



co 0r
- -c -

) 0 < -

S4-) C ,.-.
0 *0-> .-I
SO0

.- 0 +)
O C **
C O N,

0 o hI
W .to -i



,L O) r
.0. Co) -OC) -



-
'O 0 0 I


0 0 C ""
*4, 0 t r
0 ) > II


CM .0 C 0

+*>r as
c v> s:




















Q
rc)


I
c'J
'Cj



cm
I

Q ri


'Cj
bI



XI





























3
4- C.
0 --
-- Q 0 II)
OWE O>
0 E = .


l0 (D CM t0 > e-
Z -C C (0 (e
> < Z S c)
S" < S v ---
*r + ft 0 0
o c: en c) >,
4-* .0C <0
S,-- M ,-4I r-. "
4-) -0 4-) 0 no
*r- $. C" X C --
ur 0 m a0 O
C-: S U II
C 4--' 4-. -O v-4 4U
O C-- r N
r. U = 0- E
"= : > CTC..",.
4-> .0 C I- W
0 C 0 4- 0 I
*r- N 3 4-3 U
3: >> j a, < u
0 r-- 0 -i l
.= to o 4
o "C S.- 0
4-) U 0 'L -4
0o 4 CM Z
r- C 0 t 0 x
0-O > to Z
0 c-- J-- S d
r- 4- CM I
r 4- X. CmO .* 0
U O. 0 U)1 O -4
0 1 4-) CO 4)
S.C0rU == H X
C. O 0 3 C:
-O --4 .0 C
U 4- X S. E. C *C
U 03 4-> t0 o *
L -LCO 0 (- O

U) >i U U) II +1
r- .C- 0 C C
3 4-3 4-+ CMJ4- LI
o M-' -C 0 CO U Ca
M 4- -) 3 *+ -. 1 --*

*

S-4

r-
cn
L.


















N





-8
U




-I


10
X
U)


-I>








1 nt slope
The values of pK and pKI are given for sodium azide in

Figure 14 and and for potassium iodide in Table 5. At pH values 6 and

9.5, the inhibition patterns were close to uncompetitive (Figures 11

and 13) with characteristic parallel lines in a double reciprocal plot.

In the pH range 7.7 to 9.0 however, the patterns of inhibition were

mixed/noncompetitive (Figure 12). The pattern of inhibition by KI at

pH 6.0 was also close to uncompetitive. Above pH 8.0, it was

mixed/uncompetitive. The values of KInt for these anionic inhibitors

were independent of pH from 6.0 to 8.0, although Kslope varies greatly.

Both Klope and Knt increased as pH was raised above 8 (Figure 14 and

Table 5). From the analysis of data in Figure 14 using equation 12 by

a least-squares method (22,23), the pKa of the group which affects the

inhibition by NaN3 of the hydration of CO2 catalyzed by cat isozyme III

was obtained as 8.4 0.3

inntrinsic
K nt = (12)
[H+]
[H+] + K
a
where K}ntrinsic is the lowest plateau value of Knt. Equation 12

assumes that the inhibitor can bind to enzyme when a basic amino acid

residue of isozyme III is protonated. The inhibition by NaN3 of CO2

hydration catalyzed by cat isozyme III gave a linear Dixon plot at pH

7.1 for concentrations up to 1.0 x 10-4 M in NaN3; however, the Dixon

plot was nonlinear at higher concentrations of NaN3 (data not shown).
chlor
In Figure 15, the values of Kdiss the equilibrium dissociation

constant of chlorzolamide and cat carbonic anhydrase III are given as a

































%c c* O
So o >
f- a (D
4- w .- r- 0) 0 C
) 00 0 > <4- 0
C. **- N 0 -C
0 -4 C*'J E: 4 .C
r- a wmm a3 L

Q. -, L-- 01 ,
4-- M ( - c 4-
,r M ,- >) "a r I-
-",O" I E *"- C *
c o 0r 0 c o

to c >, -- +- 0 4-


0 > *T- *- I -C *-
.0 .0 t o 0 (;


r- LU 4- 0 r-
S- U*r- 0 i-n C (L

C--0 C th r- C
4 (V W"- U
SQ.o C *r- In



*r- > N0 r0- 0 L-
=- .C W r- -C (- C C
0. 4- 0U
0 4-* 3 a co
OL OL- aS ) + E


S-C' "" C- CO 4-)4-I
C v- 0 4- O > C
C) 0 G) 4)
-0 0 Oe 4J N
a4- U C 3 *- -0

U t-- CC r Q. (0+> -
C 00 4- 0 4-

S*- In 0 Ln a) *E -

.L- (0 "" -- a) 04-)
S-L O 0 -= cc
*- 4J L 0.r-
.c4- e>a 0 1 M E ,
-u- -- cn-- cL






o*-


LL.











I I I I I













-!





V
I I IPo
__ ______ I______ 1 ______ I ______-----


EH
a-


a,
a

a-


3:
0a









TABLE 5

THE INHIBITION BY IODIDE OF THE HYDRATION OF
CO2 CATALYZED BY CAT CARBONIC ANHYDRASE III


int slope
H pKI pKI




6.2 3.3 1.6

6.7 3.0 2.1

7.3 3.0 2.5

7.7 2.9 2.1

9.5 1.9 1.2


aExperiments were performed
containing 0.033 M Na2SO4. The
lutidine (pH 6.2), 20 mM Mops (
15 mM Hepes (pH 7.7) and 20 mM
are given in the units of molar


I at 25"C in solutions
Suffers were 20 mM 3,5-
pH 7.3), 15 mM Pipes (pH 6.7),
Ted (pH 9.5). Values of KI






























0 **
So **> 0

0) I E N
a) C W 2E I(
4-0 -
OX 0 E
.0 0) *r-
C .3 >,

'.-- t. ~ 0 ) 4 -
4-) E 4- 0)

to i QC N 3 *r-
S- E C .0 '

c- t o *
o- (u P & O-l





o C00)toN
"- 0) C +' -o
>- aq 0 0


O C t0
4-) -C -- 4-


C- 0 0 C 0
*S- t to 0
4- ) C CT )0

r- .0 *r- 0- *

=Q- m. 0 0
> to C 0



to C- 0


C t 04- U W-
- rr-

O 0) > 0
SC .0 E *

) f- 0*3 >*
a "a .0 C r-> 4
O O- O3 *r- 3


0 01 1 wC







--
0)
S.-


LL-


















































I.O O u
?L LOm


0 T

C)


iirTT


a1










function of pH in the pH interval from 6.0 to 8.5. As can be seen, the

values of Kdsor were independent of pH in the pH range 6.0 to 7.5 but

increased above pH 7.5. The pKa of the group which affects the binding

of chlorzolamide to cat isozyme III was obtained as 7.5 0.1 by using

Equation 12. The inhibition by the sulfonamide ethoxzolamide of the

hydration of CO2 catalyzed by cat isozyme III was measured at pH 7.7

and observed to be indistinguishable from noncompetitive (data not

shown). The inhibition pattern was also noncompetitive when

chlorzolamide was used as the inhibitor at pH 7.7 with value of Knt =

1.5 x 10-6 M. The binding of cyanate at pH 6.0 was inhibited in a

competitive manner by binding of chlorzolamide (Figure 16). When

experiments were performed using human isozyme II, a competitive

pattern of interaction was observed at pH 6.2 between the fluorescent

sulfonamide dansylsulfonamide and azide with values of Kdiss 2.0 x

10-6 M and 5.0 x 10-4 M for dansylsulfonamide and N3 respectively (data

not shown). The binding of dansylsulfonamide to cat isozyme III was

too weak to be useful in these studies.
Discussion

The uncompetitive pattern of inhibition by N3 and I of the

hydration of CO2 catalyzed by cat carbonic anhydrase III at pH 6.0 can

be compared with the observation of Pocker and Deits of uncompetitive

inhibition at pH > 8 by anions of the hydration of CO2 catalyzed by

isozyme II (52). In the previous Chapters, I have argued that isozyme

III from cat similar to isozyme II has a rate-limiting proton transfer

occurring outside the interconversion of CO2 and HC03. The

uncompetitive mode of inhibition by anions of carbonic anhdyrase II was































) E-

w 4v >) Lr-
oU N *
w -. a C 4 X

$1 "- 10 0 I-Xt"
a C E 0 o -


S o
(a fu B 0 0
) ~ ) O 4 - -
I -- 1C 4 4 *- -
U 0 0 *, *r- O
000'rr O
- *r- N *r- 0)**
0) C r- 4 0 L-
4- 0 0 10 M C: -
U 0 r- S 4-) 0
SS- mO 4 C 0 -
S- to0 0 C 4-3

S4- OU 4-. L-


*- O L I "
l- *r- CM EC C

C S ~ I- o <-
S4-" 4-) 4-" r-4 a ) --
SX Ca *- 0 Vo

W) C .C C 4-
C < 0 Cr t *rc to
4-) U *r- (0 04- I
WI 3 +-00
4- - C 1-4
O 0 *- C0 WO
5 -4- 0- UC X

4- Cr- O
0 4 U 0 W u*s
r- 0 *- WSC r- +
L 4. 0 a
0x I 4- C
CO u tor *r- O
) C I =E C -
4J M c- r- Z *
W (0 0 :z
4- >4 t0 X V
0 *-* S- "a to
En o) | ca = <1 "O M,
u e: C uLO E .-<
mC = m C0 C)
Lli .-. ( o Zi CJ -O


40
r-4


E)

*r-
37


















6







m

E
0
0
0
C)







0
IL?

0


I-
O
<8
L<










explained by using the enzymic model in Scheme 1 with its rate-limiting

proton transfer (7,32). I suggest that the uncompetitive type of

inhibition observed in this study could also be the result of such

rate-limiting proton transfer in the kinetic mechanism of cat isozyme

III. Since an uncompetitive pattern of inhibition for isozyme II was

observed only at pH values greater than the pKa of the activity

controlling group (pKa 7 for zinc-bound H20), at pH 6.0 carbonic

anhydrase III must be in its high pH form to give the same pattern of

inhibition (Figure 11). In other words, zinc-bound H20 in cat isozyme

III could have a pKa lower than 6.0. The change in the mode of

inhibition toward mixed/noncompetitive between pH 7.7 and 9 is not

understood. One possibility is that ionization of a group such as

Lys 64 is responsible.

It is interesting to note that although isozyme III and II share a

similar type of inhibition in their high pH arms, the magnitude of

inhibition constants for anions is similar for isozyme I and III (see

Table 2 in ref. 14). It is found for isozymes I and II, as pH

increases the values of KI for anions increase, perhaps because of

the competition between OH" and anions at the active site (55). When

pH is lowered, anions can bind more strongly to the zinc H20 form of

the active site. It is therefore curious that low values of KI are

observed for the inhibition by anions of the hydration of CO2 catalyzed

by isozyme III at pH values where the zinc-OH form the active site is

suggested to exist (Figure 14). This result could be due to the

stabilization of the negative charge on zinc-bound hydroxide by the

positive charges of the basic amino acid groups such as lys 64 and










arg 91. Therefore, at pH values lower than the pKa of these residues

(pH < 8), anions could coordinate to a fifth ligand site on zinc

without displacing OH" and give rise to pH independent values of K nt.
BH+ BH+ B

Zn OH + I"- -.- Zn H Zn OH (13)
I-

However, at pH > 8, the ionizations of these positive charges will

leave the negative charge on hydroxide ion unstabilized, therefore the

values of KI will increase (Figure 14). A value of pKa = 8.4 0.3

obtained from the analysis of Figure 14 is consistent with the pKa of a

group such as lysine. Further, this group may be the same one which
CO2
causes the increase in kcat above pH 8 (Figure 6, upper panel).
The Dixon plot of azide inhibition is nonlinear above 1.0 x 10-4 M

concentration of NaN3, meaning that the inhibitor bound enzyme still

has a residual activity. Tibell et al. (55) for isozyme II also

observed that anions give rise to nonlinear Dixon plots at

concentrations much higher than their corresponding KI values. They

explained this result by proposing a fifth coordination site for anions

on isozyme II. Such a hypothesis was also suggested above to explain

the pH profile of K nt for azide (Figure 14). Therefore, the nonlinear

Dixon plot could be further evidence for the presence of a five

coordinated anion-isozyme III complex.
chlor
The values of Kdiss are also independent of pH in the pH region

6.0 to 7.5 but increases above pH 7.5 (Figure 15). This result could

also be explained by the same hypothesis given above for Figure 14.
The value of pKa 7.5 1 obtained from Figure 15 is quite lower than










the value of pKa 8.4 3 obtained from Figure 14. This difference
could be due to ionizations of the inhibitor itself (pKa values of

chlorzolamide are 6.6 and 8.8 [56]), in addition to ionizations on the

enzyme which could alter the interaction of chlorzolamide with the

active site zinc and hydrophobic region.

The noncompetitive pattern of inhibition by ethoxzolamide and

chlorzolamide of the hydration of CO2 catalyzed by cat isozyme III is

similar to that observed using isozymes I and II. This noncompetitive

pattern of inhibition may be the result of tight binding by

sulfonamides Also similar to that found for isozyme I and II
(57,58), anions and sulfonamides compete for the same binding site,

metal ion in the active site of cat isozyme III.

In summary, the uncompetitive pattern of inhibition by N3 and I- of
the hydration of CO2 catalyzed by cat isozyme II is consistent with the

hypothesis given in Chapter IV in that this isozyme was suggested to

have an activity controlling group with a pKa value lower than 6.0, and

rate-limiting step located outside the interconversion of CO2 and HCO3.

The pH profiles of K nt and Kdisso suggested that cat isozyme III could
have a fifth coordination site for anions and sulfonamides. They also

indicated the presence of a basic amino acid group with a value of pKa

higher than 8.0.













CHAPTER SIX
CONCLUSIONS


I have shown further evidence to support the hypothesis by Tu et

al. (11) that Cu2+, through interacting possibly with the histidine 64

residue of human carbonic anhydrase II, inhibits only the

intramolecular proton transfer in the hydration and dehydration

reactions catalyzed by carbonic anhydrase II. This conclusion further

implies that the histidine 64 residue of isozyme II is also the group

which shuttles protons between the water ligand on zinc and the buffers

in solution during the catalysis of the hydration reaction.

I have shown that, while carbonic anhydrase III from cat has a

rate-limiting proton transfer located outside the interconversion of

CO2 and HCO3, it does not have any buffer effect in its catalysis of

the hydration of CO2 as found for isozyme II (1). The pH-rate profiles

of the hydration, dehydration and esterase activity of cat isozyme III

suggested that the pKa of water, which is normally 14, can be lower

than 6.0 when it is bound to the active site of cat isozyme III. I

have also discovered that a basic amino acid residue near the active

site with a pKa near 9.0 affects catalysis and also the binding of

anions and sulfonamides. Although there is no experimental evidence, I

suggest that lys 64 and arg 91, residues which are found in the active

site of bovine isozyme III which also seem to exist in cat isozyme III,

can be responsible for these effects. These basic residues, by









stabilizing the negative charges at the active site, can also decrease

the pKa of zinc-bound H20 to a value lower than 6.0.

Perhaps one of the reasons for carbonic anhydrase III being the

slowest isozyme of carbonic anhydrase relates to the fact that it has

these basic amino acid residues at its active site which cannot accept

a proton from the zinc-bound H20 as fast as histidine 64 can in isozyme

II, at least at pH < 8 during its catalysis. These positive charges at

the active site can also be the reason why isozyme III can bind tighter

to anions and not to sulfonamides.

Finally, I hope that this study, which establishes the similarities

and differences between the kinetic mechanisms of carbonic anhydrase II

and III, will be useful for the future studies of investigators in this

field.













APPENDIX A
History, Distribution and General Properties of
Carbonic Anhydrase from Red Cells

In the early 1930's it was already known that the release of CO2

from hemolyzed blood occurs more rapidly than expected from the uncat-

alyzed rate of conversion of HC03 to CO2. Calculations based

on the uncatalyzed rates indicated that one passage time of blood through

lungs would only permit the release of 17% of CO2 known to be excreted

(59). In 1933 Meldrum and Roughton discovered carbonic anhydrase in red

blood cells as a discrete protein different than hemoglobin (60). The

presence of carbonic anhydrase activity in pancreas and muscle was also

reported by the same investigators (60). Anions, cyanide, sulfide and

metal ions Cu2+ and Hg2+ were shown to be inhibitors of this

enzyme (60). The role of carbonic anhydrase in respiration was elegantly

explained by Roughton in his 1935 review article (59). In 1940 Keilin

and Mann, after purifying carbonic anhydrase some 200 times, showed the

presence of zinc, hence made carbonic anhydrase the first zinc enzyme to

be recognized (61). Mann and Keilin also discovered that sulfanilamide

is a specific inhibitor of carbonic anhydrase with a KI 10-6 M

(62). Mann and Keilin went on further and showed that all unsubstituted

aromatic sulfonamides R-S02-NH2 are active (62). Roblin and Clapp

(63) and Miller et al. (64) synthesized dozens of unsubstituted hetero-

cyclic sulfonamide compounds which have higher inhibitory activities than

sulfanilamide. From these compounds, acetazolamide, ethoxzolamide, meth-

azolamide and benzolamide became the most valuable tools in research in

83







84

determining the role of carbonic anhydrase in transport of metabol-

ically formed C02 and production and secretion of H+ and HC03

in blood, kidney, pancreas, stomach, liver, CSF, eye, lung and many

other organs (65). Maren discovered that for a sulfonamide inhibitor

to exert any pharmacological effect, it must be present at concentra-

tions high enough to inhibit 99.9% of enzyme in kidney and many

tissues, because the concentration of carbonic anhydrase in tissues is

in excess and C02 hydration is never rate-limiting (65,66).

In 1960 Nyman (67) and Lindskog (68) purified for the first time

the isozymes of carbonic anhydrase from erythrocytes. Human erythro-

cytes contain the B (or I) and C (or II) forms of carbonic anhydrase.

Isozyme C is one of the fastest enzymes in nature and has about five-

fold larger kC2 for hydration of CO2 (Table 1) than that for
cat
isozyme B. Both isozymes have about 260 amino acid residues in a

single polypeptide chain (isozyme C has one less) and M.W. of 30,000

(69). There is a 59% homology between the amino acid sequences of

isozymes B and C (70). Two forms of carbonic anhydrase were isolated

from bovine erythrocytes (68) and both have the same kinetic and

molecular properties, similar to that of isozyme C from human red cells

(68). Although letters B and C and Roman numerals I and II were

interchangeably used in the past to name the red cell isozymes, after

the discovery of the third isozyme in muscle now only Roman numeral

notations are used in the genetic literature (71).

Distribution of isozymes I and II in mammalian tissues has been

reviewed by Maren (65) and recently by Tashian et al. (72). It appears

that isozyme II occurs more widespread than isozyme I especially in

secretary tissues. Carbonic anhydrase I is the most abundant protein





85

in red blood cells of man, also found in primates and rodents but mis-

sing in the red cells of ruminants and felines (65). Maren et al. (73)

have pointed out the very powerful inhibitory action of Cl- ions on

carbonic anhydrase I, at the chloride concentrations present in the red

cells. Therefore the function of isozyme I in red cells is not clear

yet. Besides in vertebrates, carbonic anhydrase is also found in

invertebrates, higher plants, algae and bacteria (75). Carbonic

anhydrase in cells can appear either as cytosolic protein or bound to

membrane (65,74).

Although physiological substrates for carbonic anhydrase isozymes

are CO2 and HC03, they can also catalyze the hydrolysis of

esters and hydration of aldehydes (75).

Studies on the kinetic and spectroscopic properties of mammalian

carbonic anhydrase I and II have been very intense in the laboratories

of Lindskog, Pocker, Coleman, Silverman, Koenig and Bertini. Here I

shall not describe any of their findings, but lead readers to excellent

recent reviews by Pocker and Sarkanen (75), Lindskog (76) and Silverman

and Vincent (31).

History, Distribution and Physiological Function
of Carbonic Anhydrase III

The presence of CO2 hydration activity in the watery extracts of

mammalian muscles which were perfused well with saline solution to

remove the blood was reported as early as in 1933 by Meldrum and

Roughton (60). Despite this observation, Roughton stated that "in such

a location carbonic anhydrase would be an enemy to the organism, rather

than a friend: in its absence the CO2 is able to diffuse away

rapidly without loitering appreciably by the wayside in the form of





86

HC03 ions" (59). This statement perhaps delayed for several

decades the discovery of carbonic anhydrase in muscle. Another factor

which also contributed to such a delay was the expectation that the

hydration of CO2 catalyzed by carbonic anhydrase should be inhibit-

able by 10-7 to 10-8 M concentrations of sulfonamide inhibitors

such as acetazolamide. We now know that carbonic anhydrase III is

rather resistant to inhibition by several sulfonamide inhibitors. Lack

of inhibition by 10-7 to 10-8 M of acetazolamide was inter-

preted as the absence of carbonic anhydrase in muscle. Maren et al. in

1966 (77) reported that the supernatant solutions of male rat liver

extracts have carbonic anhydrase activity which is about 10 to 3000

times less sensitive to inhibition by sulfonamides than seen in the red

cell or kidney preparations. In 1974 purification of that sulfonamide-

resistant carbonic anhydrase from the male rat liver was reported by

King et al. (78). It was found that this enzyme is not only quite

resistant to inhibition by sulfonamide inhibitors but also has a turn-

over number for hydration of CO2 which is about 1% of that red cell

isozyme II (78). At that time it was clear that this liver carbonic

anhydrase is different than red cell isozymes I and II. Holmes in 1976

reported the discovery of carbonic anhydrase III from the skeletal

muscles of sheep, chicken, rabbit and mice (79), and subsequently puri-

fied this enzyme from the skeletal muscle of sheep and chicken (12).

In the same year in Noltmann's group identification of the so called

"basic muscle protein" as carbonic anhydrase III was established (13).

This basic muscle protein was first recognized in 1972 during the puri-

fication procedures of phosphoglucose isomerase from rabbit muscle and






87

at that time its function was not understood (80). So far, purifica-

tion of this third isozyme of carbonic anhydrase has been reported in

sheep, chicken, horse, rabbit, bovine, cat, pig, human/gorilla skeletal

muscles (12-14,81-84,39). Isozyme III has also a zinc ion essential

for its activity and a molecular weight of about 30,000 similar to

other isozymes (12,13). It originates from a genetic locus distinct

from those responsible for the synthesis of red cell isozymes I and II

(39). The amino acid composition of isozyme III showed some similar-

ities to red cell isozymes (81), moreover Tashian's group has shown

that for bovine isozyme III there is considerable sequence homology

with the isozymes I and II, though there appear to be two arginyl resi-

dues and one lysyl residue substituting for three of the active site

residues in isozymes I and II (39). Partial sequences of muscle iso-

zyme from bovine, human, chicken and rat revealed almost identical res-

idues at positions from 79 to 92 (85). Depending on the species its

isolated from, isozyme III has 2 to 6 cysteine residues (81) which have

been shown to cause dimerization through disulfide bonding formation

(12,13). The concentration of isozyme III in skeletal muscle in the

cat is high, about 100 Um constituting 1 to 2% of the total soluble

cytosolic proteins (14). Histochemical methods confirm its presence in

the cytoplasm of muscle cells (86,87).

An interesting study by Carter et al. showed that carbonic anhy-

drase enzymes isolated from male rat liver and skeletal muscle have

identical amino acid sequence, immunodiffusion and electrophoretic

patterns (85). Moreover, the inhibitory activities of both anions and

sulfonamides against purified male rate liver and cat muscle carbonic

anhydrase III were shown to be similar by Sanyal et al. (14).






88

The physiological function of muscle carbonic anhydrase III is un-

known. Holmes suggested that this enzyme may function in C02/HCO3

exchange and transport (12). However, the results of Sanyal et al.

which showed KC02 values for cat isozyme III as high as 30 mM

do not support such an hypothesis (14). Its possible function as a phos-

phatase is also unlikely, because of the extremely low values of Km for

this reaction (33). Its role in intracellular buffering during acidosis

and alkalosis was also considered (14), but is not supported due to its

absence in tissues such as cardiac muscle (88) which has considerable

buffering capacity (89,90), and its presence in tissues such as skeletal

muscle (12,13) where carbonic anhydrase III occurs at high concentra-

tions, but there is very little buffering capacity (91).

Despite Roughton's statement which I have quoted at the beginning of

this section, Sanyal et al. (14) suggested that isozyme III in muscle

might facilitate diffusion of metabolically formed CO2 out of muscle.

There is now a good body of evidence to support such an hypothesis. Gros

et al. in their study showed that carbonic anhydrase is necessary for

facilitated diffusion of CO2 to occur in phosphate buffer solutions at

low pCO2 gradients (92). Moreover Kawashiro and Scheid showed that

facilitated diffusion of CO2 indeed occurs in muscle preparations (93).

But unfortunately they have not shown the effect of carbonic anhydrase

inhibition on this facilitated diffusion (93). Studies in Maren's lab

showed that the role of carbonic anhydrase on the lens (94) and rectal

gland (95) appears to be for the carriage of metabolically formed CO2.

Inhibitors of carbonic anhydrase cause accumulation of CO2 in these

tissues indicating that the transport of CO2 is diminished in the







89

absence of carbonic anhydrase. These studies indicate that a possible

role of carbonic anhydrase III in muscle for the facilitated diffusion of

CO2 is quite likely.

Enns and Hill (96) and Geers and Gros (97) in a similar kind of

study showed that the washout of 14C from the bloodless hindlimb of

cats and rabbits considerably increases when carbonic anhydrase inhib-

itors are present in the perfusate. This effect was due to decreased

entry of 14C labelled CO2 and HCO3 into tissues after carbonic

anhydrase inhibition. Geers and Gros showed that the onset of action by

sulfonamides, methazolamide and benzolamide, with differing lipid solu-

bility was less than 1 minute, also the IC50 of these inhibitors were

inconsistent with their known KI values for pure isozyme III (97)

(IC50 those effects seen with sulfonamides were not due to inhibition of the

intracellularly located carbonic anhydrase III, but rather due to inhibi-

tion of another type of enzyme bound from the outside to muscle cell mem-

brane or endothelium (97).

So far these studies bring little insight to what carbonic anhydrase

III might be doing in skeletal muscle. Further experiments are necessary

in which selective carbonic anhydrase II and III inhibitors should be

used in resting and exercising muscle to understand the role of isozyme

III.












APPENDIX B


The computer simulation experiments conducted here are an extension

of the previous work by Rowlett and Silverman (32) which uses the

enzymic model shown in Scheme 1. The details of these computer

simulation experiments are given by Rowlett and Silverman. Here, a

brief description will be presented.

The change in the concentrations of each enzyme form with time can

be described by a set of differential equations. Equation 14

demonstrates one of those differential equations for the E form of the

enzyme in Scheme 1.

S= k2[EX] + k7[EH] [B] kl[S] + k8[BH][E] (14)

Then this set of differential equations is solved by the Euler method

of numerical integration (98) to obtain the steady-state concentrations

of each enzymic species.

This method involves taking very small increments of time dt to

calculate the rate of change of each enzyme species at time t and then

updating the concentration of each enzyme form after each iteration.

When all of the enzyme forms have reached their steady-state

concentrations, the initial rate of hydration is calculated from the

rate of appearance of HC03 as given by Equation (15).

d[HC03 k3E + HEX] -
dt = k3[EX+E k4[HCO31]([EH] + [HEH]) (15)








These simulation experiments by Rowlett and Silverman (32)

explained almost all the kinetic features of carbonic anhydrase II,

some of which are mentioned in Chapter One.

In this study, I have modified the computer programs by Rowlett and

Silverman on the basis of the new model given in Scheme 2. This model

is the same as the one given in Scheme 1, except it incorporates Cu2+

as a specific inhibitor of carbonic anhydrase II (bottom cage of

Scheme 2) and assumes that Cu2+ inhibits only the intramolecualr proton

transfer event through binding to the histidine 64 residue of this

2+
isozyme. The values of the on and off rate constants of Cu2 were

calculated as 1.0 x 1010 M-1 sec-1 (diffusion limited value) and 5.0 x
3 -1 Cu2+ 7
10 sec1 on the basis of Kdis = 5.0 x 107 M.

The results of these simulation experiments which are conducted

under different reaction conditions are summarized below.

1. At fixed buffer concentration (15 mM) when the concentration of

CO2 was varied from 6 to 30 mM, the resulting patterns of inhibition by

Cu2+ of the hydration of CO2 catalyzed by isozyme II were

noncompetitive for values of pH from 6.0 to 8.0 (Figure 17).

2. The values of K nt or Kslope which decreased as pH increased

gave a pKa value 7.0 for the group to which Cu2+ binds (Figure 18).

3. The corresponding pattern of inhibition was uncompetitive when

CO2 concentration was fixed (8.5 mM) and buffer concentrations were

varied from 1 to 10 mM at pH 7.0 (Figure 19).


























Proposed model for Cu2+ inhibition of human carbonic
anhydrase II. It was assumed that Cu2+, through binding
only to the unprotonated form of proton shuttle group,
inhibits the intramolecular proton transfer process
(k5, kS). (See Scheme 1 for details of the model without
Cu2+.) (With permission from Dr. Roger Rowlett.)


Scheme 2.