Source of the catalytic differences between human carbonic anhydrases III and II as determined by site-specific mutagenesis

MISSING IMAGE

Material Information

Title:
Source of the catalytic differences between human carbonic anhydrases III and II as determined by site-specific mutagenesis
Physical Description:
xi, 138 leaves : ill. ; 29 cm.
Language:
English
Creator:
Lograsso, Philip Victor, 1963-
Publication Date:

Subjects

Subjects / Keywords:
Carbonic Anhydrases -- genetics   ( mesh )
Mutagenesis, Site-Directed   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1992.
Bibliography:
Includes bibliographical references (leaves 132-137).
Statement of Responsibility:
by Philip Victor Lograsso.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 27132999
ocm27132999
System ID:
AA00011159:00001

Full Text













SOURCE OF THE CATALYTIC DIFFERENCES BETWEEN HUMAN CARBONIC
ANHYDRASES III AND II AS DETERMINED BY SITE-SPECIFIC
MUTAGENESIS















BY


PHILIP VICTOR LOGRASSO


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



UNIVERSITY OF FLORIDA


1992































To my father--for teaching me his principles: caring,

dedication, and honesty

To my mother--for her unconditional love

To my friends--for the good times
















ACKNOWLEDGEMENTS

I am very grateful to many people for helping me in my

work. Dr. David Silverman has been a true mentor. He has

taught me to be a thorough and diligent scientist. The data

analysis, writing, and organizational skills I have learned

from him are invaluable.

Dr. C.K. Tu warrants a special mention because a great

many problems were solved with a simple, "Hey Tu." I have

learned a great deal about enzyme kinetics from Tu, but it is

his generous spirit and willingness to help that I most hope

I retain. Thanks Tu.

Three people have been a great help on the molecular

biology aspect of this work: Drs. Phil Laipis, Susan

Tanhauser, and David Jewell. Each has contributed to my

knowledge in this field, and for that I am grateful.

George Wynns taught me a great deal about protein

purification and deserves many thanks. Our conversations

about baseball, Florida, and life always brightened the day.

It was also a pleasure working with all of the staff in the

Pharmacology Department. Special thanks go to Judy Adams for

formatting this thesis and all her help.

Finally, no acknowledgement would be complete without

thanking my parents and friends for their encouragement and

love.


iii
















TABLE OF CONTENTS


paae

ACKNOWLEDGMENTS .......................................... iii

LIST OF TABLES ............................................vi

LIST OF FIGURES ..........................................vii

LIST OF ABBREVIATIONS ....................................ix

ABSTRACT ................................................... x

CHAPTER 1: INTRODUCTION .................................. 1
Investigating the Catalytic Mechanism of Carbonic
Anhydrase using Site-Directed Mutagenesis ..............1
Genetic Differences and Tissue Distribution of
Various Carbonic Anhydrase Isozymes .................... 5
Biological Role of Carbonic Anhydrase III ..............7
Structure of Carbonic Anhydrase ....................... 8
General Shape of the Molecule ...................... 9
Secondary Structure ................................12
Catalytic Mechanism ....................................18
Interconversion of C02 and HCO3- ................... 19
Proton Transfer ....................................20
Dehydration.. ......................................21

CHAPTER 2: THE MUTAGENESIS, EXPRESSION, PURIFICATION,
AND KINETIC METHODS USED TO CREATE AND CHARACTERIZE
HCA III MUTANTS ........................................25
Mutagenesis ............................................25
Growth of Phage ....................................29
Isolation of ss template DNA........................30
Phosphorylation and Hybridization of the
Site-Specific Oligonucleotide...................31
In Vitro Double Strand DNA Synthesis ...............32
Screening Potential Mutants ........................32
Cassette Mutagenesis ..................................33
Creation of Remaining Mutants .........................35
Expression of HCA III Mutants in E coli ...............35
Purification of Human Carbonic Anhydrase Mutants
Expressed in E coli ................................42
Steady-State Kinetic Methods .........................52
Stopped-Flow Spectrophotometry.....................52
4-Nitrophenyl Acetate Hydrolysis ...................54
Equilibrium Kinetic Methods ...........................58











CHAPTER 3: CATALYTIC ENHANCEMENT OF HUMAN CARBONIC
ANHYDRASE III BY REPLACEMENT OF PHE 198 WITH LEU ...... 61
Introduction ...........................................61
Results ................................................62
Discussion .............................................77
Interconversion of C02 and HC03- ...................... 77
Proton Transfer ....................................81
Inhibition..........................................84
Conclusion .............................................86

CHAPTER 4: EFFECTS OF THE MOLECULAR PROPERTIES OF
RESIDUE 198 ON CATALYSIS AND INHIBITION OF HCA III .... 87
Introduction .......................................... 87
Results ................................................88
Discussion ............................................. 98
Interconversion of CO2 and HCO3- .................. 100
Proton Transfer ...................................111
Inhibition.........................................121
Conclusion ............................................123

CHAPTER 5: DISCUSSION AND CONCLUSIONS .................. 125

LIST OF REFERENCES ...................................... 132

BIOGRAPHICAL SKETCH ..................................... 138













LIST OF TABLES


1. Comparison of maximal steady-state rate constants
and values of the apparent pKa for the hydration
of CO2 and the hydrolysis of 4-nitrophenyl
acetate catalyzed by HCA III and HCA II .............. 3


2. Comparison of amino acid differences near the
active sites of HCA III and HCA II ................... 11

3. Maximal (pH-independent) steady-state constants
and values of apparent pKa for the hydration of
C02 and the hydrolysis of 4-nitrophenyl acetate
catalyzed by carbonic anhydrase and mutants .......... 63

4. Inhibition constants KI (micromolar) determined
from R1, the catalyzed rate of interconversion
of C02 and HCO3- at chemical equilibrium ............. 75

5. Solvent hydrogen isotope effects on R1 and RH20
catalyzed by HCA III, HCA II, and two mutants ........ 76

6. Maximal (pH-independent) steady-state constants
and values of the apparent pKa for the hydration
of C02 and rate of release of H2180 catalyzed by
mutant and wild-type carbonic anhydrases ............. 89

7. Inhibition constants KI (micromolar, ethoxzolamide),
(millimolar, cyanate), determined from R1, the
catalyzed rate of interconversion of CO2 and HC03-
at chemical equilibrium ............................. 99
















LIST OF FIGURES


Figure page

1 Comparison of amino acid sequences for human
CA II and CA III ........................................10

2 X-ray crystal structure of part of the active
site cavity of bovine carbonic anhydrase ...............13

3 Side-by-side representation of some of the amino
acid side chains found in the active site cavity of
carbonic anhydrase III and II ...........................16

4 Oligonucleotide-directed mutagenesis using single-
stranded template containing uracil residues ........... 28

5 Location of key features on the pET81fl HCA III
expression vector .......................................38

6 Time course of F198V HCA III synthesis after
induction with 40 1M IPTG ..............................41

7 Flow chart detailing the purification of human
carbonic anhydrase III mutants expressed in E coli ..... 44

8 Elution profile of K64H-R67N HCA III after
purification on DEAE-sephacel ..........................48

9 UV absorbance spectrum for purified F198A HCA III ...... 49

10 12% polyacrylamide gel detailing the levels of purity
of K64H-R67N HCA III during the different stages of
carbonic anhydrase purification ........................51

11 Typical data analysis used to calculate steady-state
kinetic constants .......................................56

12 Carbonic anhydrase catalyzed 180-exchange at chemical
equilibrium ........... ..... ............................ 60

13 Comparisons of kcat/Km for hydration of CO2 catalyzed
by variants of HCA III obtained by site-directed
mutagenesis at positions 64, 67, and 198 in the
active site cavity of HCA III ......... ...............66


vii









14 The pH dependence of RH20/[E] for F198L HCA III;
R67N-F198L HCA III; and K64H-R67N-F198L HCA III ........ 69

15 Comparison of the water-off rates, RH20/[E] for
variants of HCA III at positions 64, 67, and 198
in the active site cavity of HCA III ................... 71

16 The dependence on the imidazole concentration of
RH20/[E] and R1/[E] catalyzed by F198L HCA III;
R67N-F198L HCA III; and K64H-R67N-F198L HCA III ........ 73

17 The pH dependence of the logarithm of kcat/Km for
the hydration of CO2 catalyzed by F198D HCA III,
and F198N HCA III ................... ................... 92

18 The pH dependence of the logarithm of kcat for the
hydration of CO2 catalyzed by F198D HCA III, and
F198N HCA III ...........................................94

19 The pH dependence of RH20/[E] for F198D HCA III,
and F198N HCA III ...................................... 97

20 Plot of the logarithm of kcat/Km for CO2 hydration
versus pKa of zinc-bound water for six amino acid
replacements at position 198 and for wild-type
HCA III ................................................102

21 Plot of the logarithm of kcat/Km for CO2 hydration
versus the logarithm of KIOCN- for six amino acid
replacements at position 198 and for wild-type
HCA III ................................................107

22 Plot of the logarithm of kcat versus AG transfer
of the amino acid side chain from nonaqueous
solution to water ......................................118


viii














LIST OF ABBREVIATIONS


BCA
BSA
CA
CD
cDNA
CHES
DEAE-sephacel
DNA
DNase
dsDNA
DTT
EDTA
g
HCA
HEPES

IPTG
kDa
LB
MOPS
MW
NMR
OD
PEB
PEG
SDS
SSC
ssDNA
TAPS

TBE
Tris


bovine carbonic anhydrase
bovine serum albumin
carbonic anhydrase
circular dichroism
complimentary DNA
2[N-cyclohexyl amino] ethanesulfonic acid
diethylaminoethyl sephacel
deoxyribonucleic acid
deoxyribonuclease
double-stranded DNA
dithiothreitol
ethylenediaminetertaacetic acid
gravitational force
human carbonic anhydrase
N-2-hydroxyethylpiperazine-N'-
2-ethanesulfonic acid
isopropylthio-B-D-galatopyranoside
kilodalton
Luria broth
3-(N-morpholino)propanesulfonic acid
molecular weight
nuclear magnetic resonance
optical density
phenol extraction buffer
polyethylene glycol
sodium dodecyl sulfate
standard sodium citrate buffer
single-stranded DNA
N-{[Tris(hydroxymethyl)methyl] amino}
propane sulfonic acid
tris-borate EDTA
tris(hydroxymethyl)aminomethane















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


SOURCE OF THE CATALYTIC DIFFERENCES BETWEEN HUMAN CARBONIC
ANHYDRASES III AND II AS DETERMINED BY SITE-SPECIFIC
MUTAGENESIS

By

Philip V. LoGrasso

May, 1992




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


The objective of this work was to determine which of the

unique amino acids near the active site of the low activity

human carbonic anhydrase (HCA III) is the molecular source of

the 500-fold difference in catalytic and inhibition

properties between HCA III and human carbonic anhydrase II

(HCA II).

Mutants of HCA III were made by oligonucleotide-directed

mutagenesis replacing three residues near the active site

with amino acids known to be at the corresponding positions

in HCA II (Lys 64 -- His, Arg 67 ->Asn, and Phe 198 Leu).

Emphasis was placed on Phe-198 where replacement residues

varied in size, charge, and hydrophobicity and were selected










to test how these physical properties would affect C02

hydration, 4-nitrophenyl acetate hydrolysis, and inhibitor

binding. Catalytic properties were measured by stopped-flow

spectrophotometry and 180 exchange between CO2 and water using

a mass spectrometer.

Replacing Phe-198 with Ala, Asn, Asp, Leu, Tyr, and Val

caused an increase in kcat/Km with Phe 198 -4 Asp having the

largest increase, approximately 130-fold greater than wild-

type HCA III. In general, non-hydrogen bonding residues at

position 198 (Ala, Val, Leu) had a maximal velocity 10-fold

greater than hydrogen bonding residues (Asn, Asp, Phe, Tyr)

where Phe 198 has been shown to be hydrogen bonded to

active-site water. In addition, a correlation between

hydrophobicity and the turnover number was found.

The triple mutant, with the replacements Lys 64 -- His,

Arg 67 -+ Asn, and Phe 198 -+ Leu had a C02 hydration

activity, a turnover number for CO2 hydration, and an

activity for 4-nitrophenyl acetate hydrolysis which

approximated within 5-fold that of HCA II in magnitude and pH

profile.

These data suggest that Phe 198 is a major contributor

to the low C02 hydration activity, the low pKa of the zinc-

bound water, and is a major interactive binding site of

sulfonamides and cyanate in HCA III. Moreover, the

correlation between the turnover number and hydrogen-bonding

capability or hydrophobicity suggests an altered proton

transfer pathway in the mutants.














CHAPTER 1
INTRODUCTION



Investigating the Catalytic Mechanism of Carbonic Anhydrase
Using Site-Directed Mutaaenesis


The carbonic anhydrases (CA) are zinc-containing metal-

loenzymes which catalyze the following reversible reaction:

C02 + H20 --7 HCO3- + H+. Carbonic anhydrase is a

relatively small protein existing as a monomer and having an

approximate molecular weight of 30 kDa. It contains one zinc

atom/molecule protein which is located at the bottom of a

conical cavity 15 A from the surface of the protein. The

zinc ion is coordinated to the imidazole ring of three

histidine residues, numbers 94, 96, and 119, in a slightly

distorted tetrahedral geometry (Eriksson et al., 1988). A

fourth coordination site harbors a water molecule. It is

this metal center which is believed to be directly involved

in the conversion of CO2 to HCO3-. Accordingly then, the

ionization of the zinc-water is critical in the catalytic

pathway.

Of the eight genetically-distinct isozymes of carbonic

anhydrase (Kato, 1990; Tashian, 1989) known to catalyze the

above reaction, two of them, CA III and CA II found primarily

in red skeletal muscle and red blood cells, respectively,










provide a useful venue for investigations of catalytic

mechanism using site-directed mutagenesis because they have

large differences in enzymatic activity with very similar

backbone structures. The high activity form, isozyme II, has

a steady-state turnover number for CO2 hydration of 1.4 x 106

s-1 (Khalifah, 1971) and is indeed one of nature's most

catalytically efficient enzymes. Conversely, isozyme III has

a turnover number of 1 x 104 s-1 (Jewell et al., 1991;

Silverman and Lindskog, 1988) that is approximately 140-fold

smaller than that of CA II (Table 1). In conjunction with

the sizable differences in turnover number between the two

isozymes, they also show significantly different kcat/Km

values. For example, CA II has kcat/Km at 1.5 x 108 M-1s-1

(Khalifah, 1971) (Table 1), close to the diffusion controlled

limit, whereas CA III has kcat/Km at 3 x 105 M-1s-1 (Jewell et

al., 1991). To date, it remains unclear as to what primary

structural features are the source for this large catalytic

difference between these two isozymes. It is one purpose of

this study to investigate which amino acids in the active-

site of isozyme III may account for this considerable

difference. Particular emphasis will be placed on phenyl-

alanine (Phe) 198.

In addition to their different catalytic efficiencies,

their ability to be inhibited by sulfonamides is different.

The binding constant for inhibition by acetazolamide (Sanyal

et al., 1982; Tu et al., 1983; Engberg et al., 1985; Karali

and Silverman, 1985) is much weaker for isozyme III compared


















44 4) O
04 0 "ON
4 4 o 0-
O O




4P 4J




) O 2 VI

Oo o a O 'II
-40




"0(0 I o V CO
44 (0 -H *-



.o -o
O -- 3 -I
(ao














- ( 0 .
C) ( 0
>21 D M









0 x (0 0 (



O O
,4- N 4-)d nI






4-1 Cp U)0<43)

-'-I U.,-I~ (0-
>- 1 ( O 01 4 t

,a 0 )- A0
,D Ica ro E-4

4 V H OI o
4-X J .O



^ GO a
S- 0 4 0



0 C >. r-Iu
) 0 ( (D



0- L i) -4 No ) (0-







w 43 0 -
O 4 4 ) ( P 0 -
>1 to 6D to 4) + o
4( H m U4-11 3 00 4)
0a p aI (a 0)alO
o>1 P- I>










with isozyme II. Similarly, ethoxzolamide binds CA II more

tightly than CA III (Table 1). A third major difference that

exists between these two isozymes is in their ability to

hydrolyze esters such as p-nitrophenyl acetate. CA II

possesses considerable activity (Steiner et al., 1975) for

this process while CA III has a negligible activity (Tu et

al., 1986) (Table 1). Another major difference is the pKa of

the main activity-controlling group, which is zinc-bound

water for both CO2 hydration and ester hydrolysis. Isozyme

III has a pKa for zinc-bound water that is at least one pKa

unit lower than that of isozyme II (Engberg and Lindskog,

1984; Kararli and Silverman, 1985; Ren et al., 1988). Values

for these pKas are given in Table 1.

Despite the considerable differences in catalytic and

inhibition properties between CA III and CA II, the two

isozymes have backbone conformations which are very similar.

Comparison of the x-ray crystal structures from human

carbonic anhydrase II (HCA II) and bovine carbonic anhydrase

III (BCA III) shows a root mean square difference in location

of main chain atoms of 0.92 A (Eriksson, 1988). This

difference is even lower for many residues near the active

site. Since human carbonic anhydrase III (HCA III) is 89%

identical in primary sequence to BCA III, it is believed that

its structure is nearly identical to that of BCA III.

With these catalytic differences and structural

similarities in mind, one intent of this study is to gain a

mechanistic understanding of carbonic anhydrase using site-







5



specific mutants of isozyme III to determine the molecular

source of these considerable differences.


Genetic Differences and Tissue Distribution of Various
Carbonic Anhydrase Isozymes


Eight genetically distinct isozymes of carbonic

anhydrase have now been identified in mammals (Tashian, 1989;

Kato, 1990). Four of these (coding for CA I, CA II, CA III,

and CA VII) have been fully or partially characterized

(Venta et al., 1985; Lloyd et al., 1986; Brady et al., 1987;

Yoshihara et al., 1987; Montgomery, 1988) with regard to

their intron and exon structure. The human CA I, CA II, and

CA III genes are linked on chromosome 8 (Davis et al., 1987)

whereas the genes coding for CA VI and VII have been located

on chromosomes 1 and 16, respectively. In addition, the

nucleotide sequence of a full length cDNA clone which encodes

the human muscle-specific carbonic anhydrase (CA III) has

been reported (Lloyd et al., 1986). This cDNA identifies a

1.7-kb mRNA coding for the protein. To date, the chromosomal

positions of CA IV and CA V have yet to be identified.

It has been demonstrated by a variety of biochemical

methods (Siffert and Gros, 1982; Shima et al., 1983;

Dermietzel et al., 1985) that at least three types of

carbonic anhydrase are present in striated muscle. CA III is

found in rather large quantities in the cytosol of red (i.e.

predominantly slow-oxidative) skeletal muscle and can com-

prise as much as 20% of all cytosolic protein (Gros and











Dodgson, 1988). It is found in contrastingly low quantities

in white (i.e. fast-glycolytic) muscles. In comparison, the

high activity CA II isozyme is either absent or low in red

muscles, exemplified by soleus and vastus intermedius, and

high in white muscle such as tibialis anterior and

gastronemius. In addition to isozymes II and III, a high

activity CA, identified as CA IV, could be extracted from

membrane fractions.

CA III is also found in liver of non-primates, salivary

glands, red cells, lung, and kidney. CA I and CA II, on the

other hand, are widely distributed in red blood cells (where

CA I is the most abundant protein next to hemoglobin in these

cells) and secretary tissues such as the lens and cilliary

body of the eye. CA V is a mitochondrial carbonic anhydrase

which has been found in the kidney and liver (Tashian, 1989).

Lastly, it has been shown (Carter et al., 1984; Jeffery

et al., 1984; Jeffery et al., 1986) in rat liver that CA III

is low in females and high in males and is apparently under

control of growth hormone. However, there is no evidence to

suggest that levels of CA III in muscle are different between

males and females (Shiels et al., 1984). Moreover, Wistrand

et al. (1987) using immunofluorescence microscopy showed that

denervation of fibers of rat soleus, tibialis anterior, and

exterior digitorum longus muscles markedly raised CA III

levels in all three fibers, especially in type II fibers

which normally lack CA III.










Biological Role of Carbonic Anhydrase III


Three primary biological activities of CA III have been

studied. They include: C02 hydratase activity; ester

hydrolysis activity; and a phosphatase activity. To date,

the only one of these activities which is believed to have

any physiological significance is the C02 hydratase activity.

In 1935, Roughton reasoned that carbonic anhydrase in

the muscle would be an enemy to the cell rather than an aid

due to its fast kinetics of CO2 hydration (Roughton, 1935).

Conversely, after the discovery of isozyme III, many

questioned whether the role of CA III was as a CO2 hydratase

at all due to its low specific activity. In fact, it has

been shown (Gros and Dodgson, 1988) that carbonic anhydrase

in the muscle accelerates the rate of CO2 hydration in the

rabbit soleus by a factor of 400. This significant increase,

in spite of the enzyme's low turnover number, can largely be

attributed to the relatively high concentration of CA III in

red muscle fibers which has been shown to be 0.5 mM in rat

soleus (Carter et al., 1982). In addition, Gros et al.

(1987) have shown that C02 diffusion in rat abdominal muscle

can be suppressed if the muscle is incubated with

acetazolamide concentrations which are high enough to

indicate the involvement of the sulfonamide-resistant

carbonic anhydrase (CA III). Thus, they concluded that the

primary role of CA III is for the facilitation of C02

diffusion.











Tu et al. (1986) have shown that BCA III has a

negligible esterase activity, which probably occurs at a site

different than the active site of CO2 hydration. This

evidence comes from experiments in which the zinc atom was

removed from the enzyme, and others in which high

concentrations of known inhibitors of CA such as N3-, CNO-,

and some sulfonamides had no effect on the activity. Other

reports (Engberg et al., 1985; Jewell et al., 1991) have also

shown this to be the case for BCA III and HCA III,

respectively.

There is also evidence for a phosphatase activity in CA

III (Koester et al., 1981; Nishita and Deutsch, 1986).

However, it has been suggested that the difference in pH

dependence between the phosphatase activity and the CO2

hydratase activity, and the difference in inhibition potency

using phosphate, molybdate, fluoride, and acetazolamide for

these two activities, are the active site is different for

these two activities. Furthermore, and perhaps most

importantly, no physiological substrate for the phosphatase

activity is known (Gros and Dodgson, 1988).


Structure of Carbonic Anhydrase


Figure 1 presents a comparison of the amino acid

sequences for human CA II, and CA III. Isozymes II and III

show a 56% identity in their primary sequences (Tashian,

1989). Of the 36 residues postulated to occur within the

active-site cavity of carbonic anhydrase (Tashian, 1989), 25










are invariant between HCA III and HCA II. Some notable

differences of amino acids near the active sites of HCA III

and HCA II are compared in Table 2. Three major differences

which have been the focus of considerable work are the

residues, lysine (Lys) 64, arginine (Arg) 67, and phenylala-

nine (Phe) 198. With the exception of horse CA III, which

has Arg 64, the other isozymes of vertebrate carbonic

anhydrase sequenced to date have these residues unique to

isozyme III. In HCA II these residues are histidine (His)

64, asparagine (Asn) 67, and leucine (Leu) 198.


General Shape of the Molecule


The crystal structures for both HCA II (Liljas et al.,

1972; Eriksson et al., 1988) and BCA III (Eriksson, 1988)

have been determined at 2.0 A resolution. The general shape

of HCA II can be described as an ellipsoid with approximate

dimensions 54x42x39 A3 measured between extremes of the

backbone. Since the structures of BCA III and HCA II are

nearly identical (Eriksson, 1988), the dimensions for HCA II

are an excellent approximation for BCA III. Indeed, Engberg

et al. (1985) have shown that the hydrodynamic properties of

isozyme III are very similar to isozymes I and II suggesting

that isozyme III, like isozymes I and II, is a compact nearly

spherical molecule.












































n 3
r n



rn >





C-)
a


o p




-e
0









0- Or
LA U
0









S1



0


('


u

--0










:: )





..
** t
"i
** o
--0






uz
..p


0)





** 0






3


-4
0-
C-, a
















0i








IX.

C)
r4

00
o-








-4 0
CE
C0

S5









C,
U
0






4o-

















C'J
cy
Sp



u

g- a



r-


0-1

** 0





-I0
.. -

.o



o0
:: p












- -
*. I*

p4
** Q








.. z


:: E





-a

:: E






** o1
.. u

.-
sc
ui C






**


0-3







O- P>



o-
O i










vp




0 -0






-4 H
o-4
o -







o- 0
r-- M4







- 0











a-1


1-1



ufl
** td

U
C,























-1-
** P








4:











** 4
0







CL)






-4


-4
" 4
** C




::














H


U- l


a c

U a) C41

0) U) .
S0 a .U





0 4 3 -4
() 4 4
oC .MQ




0 ca 0
0) 4-J4-r



a a)0
: 0^ CO -0


0) ca



0 0 ( O
















4C




HaH
o H 0





c 1 1 -4


U) 0 m













E-4 44 4
-O' 0) U-


3cI c*i-4










E-i -1 -4-
0) a-, --
U~-~ O'(H


a
0 E
\0 4








c.i a











cyi
S3



En





0 -
Pd5


La


-4

** cy












--
C4



04


z


V)







-a
-U)


H
r-H


c-40

I-H
cL1


CM Ci
Su
U
I I














Table 2. Comparison of amino acid differences near the active
sites of HCA III and HCA II.



Amino acid position HCA III HCA II


64 Lys His

65 Thr Ala

66 Cys Phe

67 Arg Asn

91 Arg Ile

141 Ile Leu

198 Phe Leu

204 Glu Leu

207 Ile Val

211 Leu Val










Secondary Structure


The high resolution x-ray data show the major secondary

structure of HCA II to be a central B-pleated sheet made up

of ten segments involving approximately 30% of the amino acid

residues (Kannan, 1980). There are a number of minor pleated

sheet structures as well as two well defined a- helical

regions. Other a-helices are of the distorted 310 type.

Overall, the secondary structure of carbonic anhydrase has

been shown to be 17% helix and 40% b-sheet. These

estimations are in close agreement with Raman scattering

studies (Craig and Garber, 1977) which indicate HCA I is 19%

helix and 39% 8-sheet. Moreover, infrared absorption spectra

(Timasheff et al., 1967) and circular dicroism (CD) spectra

(Beychok et al., 1966; Coleman, 1968) both confirm the

presence of antiparallel 1-structures in carbonic anhydrase.

The active site cavity of carbonic anhydrase lies on a

B-pleated sheet structure consisting of four antiparallel

strands designated y, 8, e, and (Liljas et al., 1972). The

active site is easily located by the presence of the

essential zinc atom which is found at the apex of the conical

active site cavity. Three histidyl residues, two of them His

94 and His 96 reside on the 8 strand, and one His 119 resides

on the E strand of the 1-structures, along with a water

molecule ligand to the zinc forming a distorted tetrahedral

geometry. Figure 2 presents the x-ray cyrstal structure of

part of the active site of BCA III (Eriksson, 1988). The












Y7


Q 2 E106


K64 H--
N "0-
S2 o' NH96
o V' 199
SH o T200 '. Z n
23 NH94
N62 ...... 3 1 2

,1% "'3 N F198
NH2,,
NH
R67 .


Figure 2. X-ray crystal structure of part of the active site
cavity of Bovine Carbonic Anhydrase III (Eriksson, 1988).










water ligand (water #263) of the zinc is hydrogen bonded to

the side-chain hydroxyl group of threonine (Thr) 199 which in

turn is hydrogen bonded to the carboxylate of glutamate (Glu)

106. Other active site water molecules are presented as well

as the unique residues Lys 64, Arg 67, and Phe 198.

Aside from the central zinc atom the active site cavity

of carbonic anhydrase can be characterized by its hydrophilic

and hydrophobic portions. The hydrophilic region of isozyme

III contains the unique residues Lys 64 and Arg 67 as well as

the three His ligands (94, 96, 119) to the zinc.

Additionally, Thr 199, Thr 200, and Tyr 7 are also found in

this portion of the molecule. The most important residue in

terms of this work that appears in the hydrophobic region of

the active site cavity is Phe 198. Other potentially

important residues that reside in this region are Arg 91, Val

121, Phe 131, Val 143, and Glu 204.

The closeness in backbone structures between BCA III and

HCA II is illustrated in Figure 3 which presents the side-by-

side structures of these two isozymes. The root mean square

fit for 95 main chain residues situated in the central 8-

sheet structures of BCA III and HCA II (numbers 39-41, 89-98,

115-123, 140-149, 168-176, 187-218, and 256-260) was found to

be 0.39 A (Eriksson, 1988) indicating nearly identical

structures for those residues.

Despite this closeness in backbone structure between BCA

III and HCA II, the x-ray crystal data indicate that the

























Figure 3. Side-by-side representation of some of the amino
acid side chains found in the active site cavity of carbonic
anhydrase III (right panel) and II (left panel).

The atoms are colored by type: oxygen is red; nitrogen is
blue; and carbon is green. Due to film exposure, some
carbon atoms photographed yellow. Central to the BCA III
structure (right panel) are the three histidine ligands
coordinated to the zinc (not pictured); and the unique
residues Phe 198, Lys 64, and Arg 67. Additionally Tyr 131
can be found orthogonal to Phe 198. For comparison, the
left panel shows the corresponding residues of HCA II which
are Leu 198, His 64, Asn 67, and Phe 131. Both structures
present identical molecular orientations. These structures
are taken from the x-ray crystal data of Eriksson (1988) for
BCA III and Eriksson et al.(1988) for HCA II.













active site cavity of BCA III is much more sterically

constrained than HCA II (Eriksson, 1988). This has largely

been attributed to two residues in BCA III: Arg 67 and Phe

198. These residues are Asn and Leu respectively in HCA II.

The distance between the N-C atom of Arg 67 and the zinc ion

is 8.7 A. The distance between the C-a atom of the Phe 198

side-chain and the zinc ion is 7.38 A (Eriksson, 1988).

Eriksson gives no quantitative measure as to how much the

active site cavity volume is reduced by these unique

residues, but comparison of the volumes of these two residues

shows Phe to be 203.4 A3 and Leu to be 167.9 A3 (Richards,

1977).

Lastly, it should be noted that several other features

of the Phe 198 side-chain could contribute to structural,

catalytic and inhibition differences between the two

isozymes. For example, water 335 has been shown to hydrogen

bond to the x-electron system of the benzene ring of Phe 198.

In turn water 335 is involved in a hydrogen bonding network

through the hydroxyl group of Thr 200 and water 318 to the

zinc-bound water of the active site (Eriksson, 1988) (Figure

2). Since no analogous pattern is possible for HCA II, which

contains leucine at position 198, this unique water structure

could be responsible for the unique catalytic properties of

HCA III. Furthermore, Eriksson et al. (1988) have shown Leu

198 to be in van der Waals' contact with sulfonamide and

anion inhibitors. Therefore, the steric bulk of Phe 198











could indeed be the reason why sulfonamides are such poor

inhibitors of CA III.


Catalytic Mechanism


There is a wide body of evidence (Silverman et al.,

1979; Simonsson et al., 1979; Tu et al., 1990; Rowlett et

al., 1991) which supports a catalytic mechanism for carbonic

anhydrase that occurs in two distinct and separate processes.

The first stage of catalysis for hydration of C02 is the

nucleophilic attack of EZnOH- on C02 to yield HCO3-. The

release of HCO3- from the enzyme results in water bound to the

metal (eq. 1). The second stage of catalysis involves a rate-

contributing proton transfer from EZnOH2 to buffer to

regenerate the active form of the enzyme, EZnOH-, thus

completing the catalytic cycle (eq. 2).





kl k2[H20]

EZnOH- + C02 EZn-HC03- __ EZn-OH2 + HC03- (eq. 1)

k-1 k-2





k3

EZn-OH2 + B: EZnOH- + BH+ (eq. 2)

k-3










Interconversion of CO2 and HCO3-


As seen in equation 1, the interconversion of CO2 and

HC03- contains rate constants in the pathway up to and

including release of HC03- from the enzyme. In terms of the

Briggs-Haldane steady-state approximation, the apparent

second order rate constant, kcat/Km, is equal to klk2/(k-i +

k2) for this simplified pathway. For HCA II, it has been

shown that kcat/Km is dominated by ki (Lindskog, 1984). On

the other hand, Rowlett et al. (1991) have shown that, unlike

isozyme II, isozyme III probably has significant contribu-

tions for kcat/Km from all three constants, ki, k-1, and k2.

For the hydration of CO2 catalyzed by CA II, kcat/Km has

been shown to have a pH-profile which can be described by a

simple titration curve having a single ionization with a pKa

near 7 (Khalifah, 1971; Steiner et al., 1975). This pKa has

been attributed to zinc-bound water and has much experimental

support (Simonsson and Lindskog, 1982). CA III, on the other

hand, shows pH-profiles for kcat/Km which are independent of

pH in the range of 6 to 9 (Tu et al., 1983; Engberg and

Lindskog, 1984; Karali and Silverman, 1985; Ren et al.,

1988).

Data from solvent hydrogen isotope effect studies on HCA

II (Steiner et al., 1975; Venkatasubban and Silverman, 1980)

in H20 and D20 show an isotope effect near unity for kcat/Km

suggesting CO2 and HC03~ interconversion occurs in a step

independent of a rate-limiting proton transfer step. These










data, taken together with the pH-rate profile data, are

highly indicative of two separate and independent processes

for this catalytic mechanism. Further support for the

veracity of these conclusions will be presented when the

dehydration aspect of this mechanism is considered.


Proton Transfer


As described in equation 2, the catalyzed hydration of

CO2 requires that a proton be transferred from the active

site of carbonic anhydrase to the surrounding medium. Since

it was known that proton transfer from a group of pKa 7 to

water could be transferred at a maximal rate of 103-104 s-1

(Eigen, 1964), it was initially a dilemma to explain how CA

II could have a turnover number of 106 s-1. In 1973, several

workers (Khalifah, 1973; Lindskog and Coleman, 1973)

suggested that proton transfer was to buffer in solution

rather than water.

For isozyme II, there are many studies (Steiner et al.,

1975; Tu et al., 1989) which show that His 64 acts as a

proton shuttle enhancing activity by increasing the rate

described in equation 2. Data from solvent hydrogen isotope

effects (Venkatasubban and Silverman, 1980) indicate that

proton transfer proceeds through at least two hydrogen-bonded

water molecules which are suggested to be the ones situated

between the zinc-water and the imidazole of the side-chain of

His 64.










For isozyme III, it has been observed that the turnover

number is 103-104 s-1 (Karali and Silverman, 1985; Jewell et

al., 1991) consistent with proton transfer to water.

However, Rowlett et al. (1991) and Jewell et al. (1991) have

shown that HCA III is capable of intramolecular proton

transfer at high pH through lysine 64. This approximate 5-

fold increase in kcat at pH values > 8 can be correlated with

the pKa of lysine. Moreover, Tu et al. (1990) have shown

that proton transfer in catalysis by HCA III can be enhanced

6-fold by the addition of 100 mM imidazole. These data

suggest that the imidazole side-chain of the buffer can act

in lieu of the imidazole side-chain of His 64 which is

naturally present in CA II, but absent in CA III.

Finally, it should be noted that there is a body of data

from solvent hydrogen isotope effects for both isozymes

(Steiner et al., 1975; Karali and Silverman, 1985; Tu et al.,

1990) which indicate that the proton-transfer step described

by kcat is rate-limiting and separate from the steps involved

in CO2 and HCO3- interconversion.


Dehydration


The mechanism can also be studied in the dehydration

direction at chemical equilibrium. In this method, the

exchange of 180 between CO2 and H20, and between 12C and 13C

labeled CO2 can be followed (Silverman and Tu, 1975;

Silverman et al., 1979). In these experiments, two rates for











catalysis at chemical equilibrium are determined. The first

is R1, the rate of interconversion of CO2 and HCO3- (eq. 3).



HCO0180- + EZnOH2 -- EZnl80H- + C02 + H20 (eq. 3)
-7



In this dehydration reaction (eq. 3), the 180-label from

substrate bicarbonate transiently labels the active site of

carbonic anhydrase. Since 13CO2 is also present at chemical

equilibrium, one possible fate of the 180-label is hydration

of 13CO2. (The experimental details of this method will be

discussed in chapter 2). A second possible fate of this 18-0

label is release to water, a step which can occur

independently of the first step. From these processes, two

experimental rate constants can be determined. Ocat is the

rate constant which describes the exchange of 180 between 12C
and 13C species of CO2. 8cat is the rate constant describing

the loss of 18-0 label to water. (The rate of release of 180-

labeled water, RH20 will be discussed shortly.) Silverman et

al. (1979) have described these rate constants in detail and

have derived equations which express R1 and RH20 in terms of

Ocat and Ocat.

The measurement of 13C NMR line widths also gives the

catalytic rate at chemical equilibrium between C02 and HCO3-

(Simonsson et al., 1979). These researchers found that the

rate of CO2 and HC03- interconversion at chemical equilibrium

can be expressed in the form of the Michaelis-Menten











equation. Similarly, Silverman et al. (1979) found that R1

could also be expressed in this form (eq. 4):



R1 = kcatexch [Stot] [Etot]/(Keff + [Stot]) (eq. 4)



where Keff is the apparent binding constant between substrate

and enzyme. Finally, kcatexch/Keff has been shown to be

equivalent to kcat/Km.

As mentioned earlier, two rates for catalysis are

measured at chemical equilibrium. The second of these two,

RH20, is the rate of release of 180-labeled water from the

active site. This rate is given by equation 5:



H20

EZn18OH- + BH+ z--= EZn18OH2 + B -==7 EZnOH2 + H2180 + B (eq. 5)



In terms of the mechanism presented in equation 5, it is seen

that RH20 involves a proton transfer from BH+ (which can be

buffer in solution, water in the active site, or a group on

the enzyme such as histidine) to the zinc-bound hydroxide in

the active site. This forms a zinc-bound water which is

readily exchangeable with solvent water (Tu and Silverman,

1985). Thus, RH20 describes the proton transfer dependent

rate of release of 180-labeled water from the active site of

carbonic anhydrase.

Much experimental evidence exists for two distinct and

separate aspects of catalysis by carbonic anhydrase at










chemical equilibrium. The bulk of this evidence comes from

buffer dependence and solvent isotope experiments conducted

by 13C NMR (Simonsson et al., 1979) and 180-exchange using

mass spectrometry (Silverman et al., 1979; Tu et al., 1990).

The 13C NMR studies show that the exchange of C02 and

HC03- is equally rapid in H20 and D20 indicating this process

is not dependent on, and separate from, a rate-contributing

proton transfer step. Furthermore, they show that the

exchange is equally rapid in the presence or absence of

buffers indicating that no buffer dependent steps are located

on the pathway of C02 and HC03- exchange.

Similarly, the 180-exchange studies (Silverman et al.,

1979; Tu et al., 1990) show that R1 is independent of buffer

indicating no buffer dependent steps for CO2 and HCO3-

exchange. Conversely, they found RH20 was indeed dependent

on buffer concentration indicating a proton transfer step

facilitated by the presence of buffers.

Finally, it has been shown for both HCA III (Karali and

Silverman, 1985) and HCA II (Tu and Silverman, 1982) that R1

is essentially the same in H20 and D20 indicating that CO2 and

HC03- interconversion is independent of a rate contributing

proton transfer, whereas RH20 is highly dependent (especially

for HCA II) on this process.

Thus, just as in the case as the steady-state, there is

a wealth of experimental evidence at chemical equilibrium

which suggests that the catalytic mechanism of carbonic

anhydrase occurs in two distinct and separate steps.














CHAPTER 2
THE MUTAGENESIS, EXPRESSION, PURIFICATION AND KINETIC METHODS
USED TO CREATE AND CHARACTERIZE HCA III MUTANTS



Mutagenesis


The molecular biological effort of this project is the

creation and expression of site-specific mutants of human

carbonic anhydrase III. Two primary means of mutagenesis

were used. The first was an oligonucleotide-directed

mutagenesis approach developed by Kunkel (1985) and Kunkel et

al. (1987). The second was a cassette mutagenesis approach

developed by Wells et al. (1985).

Kunkel (1985) and Kunkel et al. (1987) have described a

method using the E coli strain BW313 (dut- ung-) to create a

DNA template which contains a small number of uracil residues

in place of thymine. E coli dut- mutants lack the enzyme

dUTPase. This causes an elevation of dUTP levels allowing

effective competition with TTP for incorporation into DNA. E

coli ung- mutants lack the enzyme uracil N-glcosylase which

normally excises uracil from DNA. Thus, the combination of

the two mutants allows successful incorporation of uracil

into DNA in place of thymine. After the in vitro extension

reaction, the double stranded plasmid composed of one uracil

containing strand and one strand containing the introduced










mutation, is transformed into a dut+ ung+ E coli host (TG-1)

which selectively degrades the uracil containing strand.

Upon subsequent bacterial replication a mutant plasmid is

produced. This scheme is illustrated in Figure 4.

The mutants K64H-R67N HCA III, R67N-F198L HCA III, and

R67N-F198A HCA III were created using the following

procedure: Initially, 5 p1 of a 50% glycerol stock of E coli

BW313 were grown in 2 ml of Luria broth (LB). The media

consists of 10.0 g tryptone (Difco Laboratories); 5.0 g yeast

extract (Difco Laboratories); 5.0 g NaCI per liter of

deionized water. The pH of this medium is raised from 7.2 to

7.4 with 12 M NaOH and the medium is autoclaved for

sterility. All cultures were grown for 12 to 16 hours at

370C and agitated for oxygenation at 280 rpm.

Competent E coli BW313 for transformation were prepared

as follows: 1 ml of the overnight culture was transferred to

100 ml LB. The cells were grown until the OD600 = 0.2-0.4.

Cells were centrifuged at 5000 x g for 10 minutes. The cells

were resuspended in 1/2 volume (50 ml) of ice-cold 50 mM

CaCl2 and incubated on ice for 20 minutes. The cells were

centrifuged as before and resuspended in 1/10 volume (1 ml)

of ice-cold 50 mM CaC12. These cells were incubated on ice

for no less than 20 minutes and no more than 4 hours prior to

transformation.





























Figure 4. Oligonucleotide-directed mutagenesis using a
single-stranded template containing uracil residues (Kunkel
et al., 1987) is prepared from a dut- ung- E coli strain.

A synthetic oligonucleotide containing a desired mutation is
annealed to the template and treated with T4 DNA polymerase
and T4 DNA ligase to produce double-stranded product
composed of one strand containing uracil residues.
Transformation into a dut+ ung+ E coli host allows for
recovery of mutant DNA containing only the desired mutation.















A WC mutated oligonucleotide




single-stranded M13 template
containing uracil residues

U




T4 DNA polymerase
4 dNTP
T4 DNA ligase





double-stranded product
uracil residues are
not mutagenic






transform
wild-type E. coli






strong selection for
M13 phages containing
mutation of interest










Transformation of competent E coli BW313 was carried out

as follows: 1 il of pET81fl HCA III R67N plasmid (these

vectors containing an origin of replication for single strand

(ss) DNA will be described in the section of this chapter on

expression) was incubated with 0.1 ml of competent E coli

BW313 on ice for 45 minutes. The cells were then heat

shocked at 420C for 2 minutes to maximize plasmid absorption.

Two milliliters of LB media was added, and the cells were

incubated for 1 hour at 370C and 280 rpm. Plasmids were

selected by plating on LB plates which contained 20 gg/ml

ampicillin (Amp).


Growth of Phage


E coli BW313 colonies selected from the LB/Amp plates

were grown so that phage infection could be used to produce

template DNA. Briefly, 20 ml of the transformed BW313 cells

were grown to mid-log phase. Five milliliters of this

culture was transferred to 100 ml LB/Amp containing 10 lg/ml

uridine and 108-109 plaque forming units (pfu) of the ss-DNA

inducing helper phage R408 (Stratagene, La Jolla, CA), an M13

derivative, and grown at 370C, 280 rpm for 18 hours. After

incubation, the culture was centrifuged at 5000 x g for 30

minutes to collect the supernatant.










Isolation of ss Template DNA


Bacteriophage were precipitated from the clear super-

natant by adding 1 volume of 5X PEG/NaCI (15% w/v

Polyethylene glycol 8000; 2.5 M NaC1) to 4 volumes of

supernatant. The solution was mixed well and incubated on

ice for 1 hour. The precipitate was collected by centri-

fugation at 5000 x g for 15 minutes. The pellet was well-

drained, and resuspended in 5 ml of phenol extraction buffer

(PEB) (100 mM Tris-HC1, pH 8.0; 300 mM NaCl; 1 mM EDTA).

This mixture was vortexed vigorously for resuspension, and

incubated on ice for 1 hour. Next, the sample was centri-

fuged again as above and the previously described steps were

repeated. At this point, the supernatant was extracted two

times with 5 ml of phenol (which was equilibrated with PEB),

centrifuged for 5 minutes at 5000 x g, and the aqueous layer

was extracted 2X with 5 ml chloroform/amyl alcohol (24:1).

Next, the DNA was precipitated from the aqueous layer by

adding 0.1 volume of 3 M sodium acetate (pH 5.0) and 2.5

volumes of 100% ethanol. After incubating the above mixture

at -200C for 20 minutes, the ssDNA was collected by centri-

fugation for 1 hour, dried under vacuum, and resuspended in

200 gi of water.









Phosphorylation and Hybridization of the Site-Specific
Oligonucleotide


A 17-base oligonucleotide, complimentary to nucleotides

1145-1161 of the ssDNA template was synthesized for the

creation of the mutant R67N-F198L HCA III. This 17-mer had

the following nucleotide sequence:



5' G GGC TCA QT_ ACC ACG C 3'

N -Gly-Ser-Leu-Thr-Thr- C



The two mutant nucleotides are underlined, and reside in the

codon for amino acid 198. In the wild-type sequence, they

would be T and C, respectively, and therefore code for Phe.

In this mutant sequence, Leu is substituted for Phe at amino

acid 198 and therefore the codon is CTG. The entire amino

acid sequence coded for by the 17-mer is presented beneath

the DNA sequence. (A similar approach was used to create the

other mutants made in this manner.) Phosphorylation of the

5' hydroxyl group of the 17-mer was carried out in a 10 gi

reaction containing 50 mM Tris-HC1 (pH 7.5); 10 mM MgCl2;

5 mM dithiothreitol; 1 mM ATP; 2.0 gg Bovine serum albumin

(BSA); 2.0 gg oligonucleotide; and 2 units T4 polynucleotide

kinase (New England Biolabs). The mixture was incubated at

370C for 1 hour and terminated by adding 3 gl of 100 mM EDTA

and heating to 650C for 10 minutes.

The hybridization of the phosphorylated nucleotide to

the ssDNA template was carried out in a primer:template molar










ratio of 10:1. The total volume of the reaction was 25 .1

and contained 1.2 .l of 20X SSC (3 M NaCl; 300 mM sodium

citrate). The annealing reaction mixture was incubated at

700C for 10 minutes and allowed to cool slowly to room

temperature.


In Vitro Double Strand DNA Synthesis


To create a covalently closed dsDNA molecule, 20 p. of

the above reaction was added to 10 .l of the following

buffer: 20 mM Hepes (pH 7.8); 2 mM dithiothreitol; 10 mM

MgCl2; 5 .1 each of 10 mM dATP, dTTP, dGTP, and dCTP; 10 p.

of 10 mM ATP; 36 1i H20; 2 units (2 pl) T4 DNA ligase (New

England Biolabs) and 2.5 units T4 DNA polymerase (U.S.

Biochemical). The reaction mixture was incubated at 0C for

5 minutes, room temperature for 5 minutes, and then 370C for

2 hours.


Screening Potential Mutants


Plasmid DNA was sequenced by the dideoxysequencing

method developed by Sanger et al. (1977), using the Taq Track

system of Promega. DNA sequencing reactions were subjected

to electrophoresis on 0.05 cm thick gels, 54 x 33 cm,

consisting of 6% (w/v) polyacrylamide, 7 M urea, and 0.05

TBE, pH 8.3. Gels were typically run at 1400 V for 7 hours.

Upon completion of electrophoresis, the gel was transferred

to 3 M paper and covered with saran wrap for autoradiography










on XAR film (Kodak) at -200C. The film was generally exposed

for 12 to 16 hours.


Cassette Mutagenesis


Wells et al. (1985) have described the use of cassette

mutagenesis to create multiple mutations at defined sites.

Briefly, this method introduces unique restriction sites into

a plasmid using oligonucleotide mutagenesis. These sites are

in close proximity to a target codon and often do not alter

the final amino acid coding sequence. Thus, we are able to

insert a mutagenic cassette into restriction sites flanking

the region to be mutated, maintaining the wild-type coding

sequence, and introducing a mutant codon at the desired codon

or codons. Design of the restriction sites is also intended

to eliminate one or in some cases both restriction sites.

This facilitates selection of clones containing the mutagenic

cassette because they can simply be screened by restriction

digests.

The mutants F198D HCA III, F198N HCA III, F198V HCA III,

and F198Y HCA III were created using the following procedure:

Tanhauser et al. (unpublished) have created unique restric-

tion sites for Nru I and Stu I in the pET81fl HCA III

expression vector (this vector will be discussed in the next

section). The Nru I site was introduced at nucleotide 1126

and the Stu I site was introduced at Nucleotide 1160. This

allows for mutants within a 34-base pair region.










For example, the mutant F198Y HCA III was created by

synthesizing the following single-stranded oligonucleotides:


5' C GAC TAC TGG ACC TAC CAG GGC TCA TAC ACC ACG 3'

5' CGT GGT GTA TGA GCC CTG GTA GGT CCA GTA GTC G 3'



These two oligonucleotides were 5'-phosphorylated in separate

10 p. reaction mixtures containing approximately 3 jg ss-

oligonucleotide; 1 mM ATP; 1.0 U T4 kinase (New England

Biolabs); 70 mM Tris-HCl (pH 7.6), 10 mM MgCl2, and 5 mM DTT,

and then annealed by mixing, heating to boiling, and letting

cool to room temperature.

The double-stranded 34-mer was ligated to a pET81fl HCA

III backbone which had the wild-type Nru I Stu I cassette

region excised. Ligation reactions typically used

backbone:oligonucleotide molar ratios of 1:10 and were

carried out at 140C for 16 hours using 1 mM ATP; 1.0 U T4 DNA

ligase (New England Biolabs); 70 mM Tris-HCl (pH 7.6), 10 mM

MgCl2, and 5 mM DTT. Ligation reactions were terminated with

25 mM EDTA. The ligation mixture was then transformed into

competent E coli GM119 dcm- for the purpose of screening for

positive clones with Nru I and Stu I restriction digests.

Screening for potential positive clones was first

accomplished by restriction digests using Nru I and Stu I.

Creation of a positive clone which contains the desired

mutation left the Nru I site intact but deletes the Stu I

site. Thus, potential positive clones yielded a single











linear band after restriction with Nru I, but yielded a wild-

type double-stranded DNA pattern when cut with Stu I. These

results were then corroborated with DNA sequencing of the

plasmid (Sanger et al., 1977). Positive clones had just the

desired mutant sequence in the region of interest and the

wild-type sequence everywhere else.


Creation of Remaining Mutants


The mutants F198A HCA III and F198L HCA III were made by

recombination of DNA from wild-type HCA III expression vector

backbone with a 625 base-pair fragment containing the 198

region excised and the 625 base-pair region containing the

198 region of R67N-F198A and R67N-F198L, respectively. K64H-

R67N-F198L HCA III was made by recombination of the K64H-R67N

backbone with a 625 base-pair fragment of the 198 region

excised and the 625 base-pair region of F198L. Basically,

this was accomplished by digesting both pET81fl HCA III and

pET81fl HCA III mutants with BamH I and Bcl I. This removes

a 625 base-pair region from nucleotides 866-1491. Religation

of the proper fragments created the desired plasmid. Clones

were then sequenced to confirm mutant assignment.


Expression of HCA III Mutants in E coli


The expression system used in this study was based on

bacteriophage T7 RNA polymerase as first described by Studier

and Moffatt (1986) This expression system has many

advantages for it provides high-level transcription from the










T7 promoter and high selectivity due to the absence of a

natural promoter in E coli.

Figure 5 presents some key features of a modified form

of this expression vector termed pET81fl HCA III (Tanhauser

et al., unpublished). Notable features include: an fl

origin for single-strand phage replication; T7 promoter and

terminator regions; the HCA III gene which contains the

unique Nru I and Stu I restriction sites; and a gene which

confers ampicillin resistance.

Expression of the HCA III gene is carried out in the E
coli lysogen BL21(DE3) which carries a Xlysogen encoding

the gene for T7 RNA polymerase. This T7 RNA polymerase gene

is under control of the lac UV5 promoter (Studier and

Moffatt, 1986). Thus, upon induction of the culture with

isopropyl-B-D-thiogalactopyranoside (IPTG), T7 RNA polymerase

is synthesized with subsequent translation of the target HCA

III gene.

The E coli strain BL21(DE3)plysS (chloramphenicol

resistant) was transformed with the pET81fl HCA III mutant

plasmid. In addition to conferring chloramphenicol

resistance, the plysS plasmid carries the gene for lysozyme.

This is used as an aid for cell lysis in the purification

procedure. Resultant cells which contained both plasmids

were selected on LB plates containing both 200 mg/L

ampicillin and 34 mg/L chloramphenicol. A single colony from

this plate was used to inoculate 2 ml of LB containing the

above concentrations of antibiotics. The 2 ml culture was













0 (D

P-




0w
S(D


Ow












cn

C:
O


(D I-

H"
H- 0







rt
0I)

u IZ
ow






I--j




SH-0
P 0






0

0
r+












P- H-







PH
0

0






H- H
rt
(O



I-3
h o



























o0 2


_ Is
-3, U.'
"x____ ZY


I -
m -










grown at 370C, 280 rpm to log-phase. One ml of this culture

was used to inoculate a 1 liter expression culture containing

the above concentrations of antibiotics and 0.2% (w/v)

glucose. This culture was grown at 370C, 280 rpm in 2800 ml

Fernbach flasks until the OD600 = 0.2 to 0.4. At this time,

the culture was induced for HCA III synthesis with 40 pM IPTG

(Studier and Moffatt, 1986; Rosenberg et al., 1987). In

addition, 2 lM ZnS04 was added. The cells were incubated an

additional 3 hours after induction, pelleted at 5000 x g for

10 minutes, frozen at -700C for 16 to 20 hours, and thawed at

room temperature to allow cell lysis so that carbonic

anhydrase activity could be tested.

The synthesis of carbonic anhydrase was monitored by

taking one ml aliquots from the bacterial culture at hourly

time points during expression. These 1 ml aliquots were

centrifuged and the pellet was frozen for future SDS-PAGE

analysis. The zero hour time point was just prior to

induction with IPTG. Subsequent time points at 1, 2, and 3

hours were also taken.

Figure 6 shows a typical time course of expression of

HCA III. The 1 ml pellets were resuspended in 200 il of SDS-

reducing buffer (62.5 mM Tris-HC1, pH 6.8; 10% glycerol; 2%

SDS; 0.05% bromophenol blue; and 5% (v/v) mercaptoethanol),

heated to boiling for 3 minutes, and loaded in 20 p. aliquots

on a 12% polyacrylamide mini-gel. The samples were

electrophoresed at 200 V for 45 minutes. The gel was stained

with 0.1% Comassie Blue R-250 containing 40% methanol, 10%



























Figure 6. Time course of F198V HCA III synthesis after
induction with 40 pM IPTG.

One milliliter aliquots were taken from the E coli
expression culture at 0, 1, 2, and 3 hours after IPTG
induction (Lanes 1-4, respectively). Lane 5 is a standard
sample of Bovine carbonic anhydrase from Sigma Chemical
Company used as a molecular weight marker (MW = 30 kDa).
Samples were electrophoresed for 45 minutes at 200 V on a
12% SDS-polyacrylamide gel and stained with Coomassie blue.






41










Time (hours)

0 1 2 3






S. 4=


MW

30 kDa


'S










acetic acid for 16 hours. Gels were destined in 10% acetic

acid and 40% methanol solutions.


Purification of Human Carbonic Anhydrase Mutants
Expressed in E coll


The method used for purification of HCA III mutants

created for our studies was based on that of Tu et al.

(1986). However, it is important to note that many

modifications of their procedure were necessary due to the

unique properties of each of the mutant enzymes. This

statement particularly applies to the steps involved in ion-

exchange chromatography where Tris buffer concentration and

pH varied depending on the mutant being purified.

Figure 7 presents a flow chart outlining the overall

purification scheme of HCA III mutants. Cells were freeze-

thawed for lysis and then treated with 1 mg/20 ml DNase I for

digestion of nucleic acid. Centrifugation at 24,000 x g

removed cell debris from the supernatant. Elution on the

Ultrogel-44 column separated approximately 20% of all

proteins from carbonic anhydrase. Cation exchange

chromatography on DEAE-sephacel separated carbonic anhydrase

based on its unique pi from all other proteins in the

mixture. It is important to note that 15 mM Tris buffer (pH

8.4) was used for dialysis and DEAE-sephacel elution of all

mutants except: F198D HCA III which needed 0.1 M Tris (pH

8.0); and K64H-R67N-F198L HCA III which needed 75 mM Tris (pH
































Figure 7. Flow chart detailing the purification of human
carbonic anhydrase III mutants expressed in E coli










1 L E. coli BL21(DE3) plysS pET81fl HCA III mutant pellet


Freeze at -700C for 16 hours
Thaw
+ 2 mg DNase I; 20 .1 10 mM ZnS04
+ 2 volumes 25 mM Tris buffer (pH 8.1)
containing 3 mM mercaptoethanol


Stir mixture at 40C for one hour





Centrifuge at 24,000 x g for 90 minutes


Load clear supernatant on 5 cm X 50 cm
if Ultrogel-44 gel filtration column


Elute Ultrogel-44 column at one ml/min at 40C with
25 mM Tris buffer (pH 8.1) containing 3 mM mercaptoethanol


I Collect 15 ml fractions in tubes



Assay every fifth tube for carbonic anhydrase activity
monitoring 180-exchange between C02 and H20 using gas analyzer


I Pool enzyme active fractions



Concentrate pooled fractions to 30 ml on an
Amicron filtration system at 40C




Dialyze 3X-vs- 4L of 15 mM Tris buffer (pH 8.4)
containing 3 mM mercaptoethanol at 40C



Load dialyzed supernatant on a
2.5 cm X 25 cm DEAE-sephacel,
Pharmacia ion-exchange column
















Elute DEAE-sephacel column at one ml/min with
15 mM Tris buffer (pH 8.4) at 40C


Collect 15 ml fractions in tubes


Assay every second tube for protein absorbance at 280 nm


fractions


Fia. 7 Cont.


Assay all 280 nm absorbance peaks for carbonic anhydrase
activity monitoring 180-exchange between C02 and H20
using a gas analyzer










8.1). Figure 8 presents a typical elution profile of K64H-

R67N HCA III after DEAE-sephacel purification.

Figure 9 shows a typical UV spectrum recorded in the

range of 310 to 210 nm of purified F198A HCA III. The

concentration of human carbonic anhydrase III mutants was

determined from the molar absorptivity of 6.2 x 104 cm-1 at

280 nm (Sanyal et al., 1982). For the mutants F198L, R67N-

F198L, and K64H-R67N-F198L HCA III, this molar absorptivity

was confirmed by titration with the tight binding inhibitor

ethoxzolamide by estimating the active-site concentration

using a Henderson plot (Segel, 1975). The two methods

yielded an enzyme concentration which differed by

approximately 10-15%. Furthermore, for the triple mutant

K64H-R67N-F198L HCA III we were able to corroborate the

enzyme concentration determined from the Henderson plot with

the Bradford assay. The calculated enzyme concentration from

these two methods differed by less than 5% suggesting that

all of the protein present is pure, active carbonic

anhydrase.

Figure 10 presents a 12% polyacrylamide gel which

follows the entire purification scheme. Lane 4 shows the

final purified carbonic anhydrase which is estimated to be

greater than 98% pure. Further description of other lanes is

found in the legend to Figure 10. For the single mutants

F198V and F198D HCA III this enzyme purity was verified with

silver staining. Enzyme purity was estimated to be greater

than 95% pure by this method (data not shown). Finally, in




























Figure 8. Elution profile of K64H-R67N HCA III after
purification on DEAE-Sephacel.

The column was eluted at a flow rate of 1 ml/min with 15 mM
Tris (pH 8.4) at 40C. Fractions were collected in a volume
of 15 ml/tube. Enzyme activity was found in tubes 46-61.
These fractions were pooled for use in kinetic experiments.









































0 20 40 60 80


Tube Number


0.8


0.6


0.4





0.2


0.0


100






















0.500








0







0.000
210 nm 310nm
Wavelength














Figure 9. UV absorbance spectrum for purified F198A HCA III.

Spectra were recorded on a Beckman DU-7 spectrophotometer in
the range of 310-210 nm. The spectrophotometer was blanked
with 15 mM Tris (pH 8.1).























Figure 10. 12% polyacrylamide gel detailing the levels of
purity of K64H-R67N HCA III during the different stages of
carbonic anhydrase purification.

Lane 1: 1 ml crude cell lysate of E coli BL21(DE3) plysS
pET81fl HCA III pellet

Lane 2: Supernatant after centrifugation at 24,000 x g

Lane 3: Pooled enzyme active tube fractions after Ultrogel-
44 gel filtration chromatography

Lane 4: Pooled enzyme active tube fractions after DEAE-
Sephacel cation exchange chromatography

Lane 5: Standard sample of bovine carbonic anhydrase from
Sigma Chemical Co. used as a molecular weight marker
(MW= 30 kDa)

















Lanes


1 2 3 4 5


MW

30 kDa


*m -










the substantial history of carbonic anhydrase literature

there are no published or anecdotal reports for HCA II, HCA

III, or any mutants of these two isozymes which show any

evidence for the isolation of non-active enzyme fractions or

inclusion bodies. Thus, it is most likely that our measures

of enzyme concentration reflect solely on pure, fully active

HCA III mutants. Any trace of inactive enzyme would result

in an active enzyme concentration which is lower than

reported. Thus, the reported enzyme concentrations reflect a

lower limit for enzyme concentration.


Steady-State Kinetic Methods


Stopped-Flow Spectrophotometry for C02 Hydration


Saturated solutions of C02 were prepared by bubbling C02

gas into deionized water contained in a vessel maintained at

250C. Dilutions were prepared in the absence of air by

coupling two syringes as described by Khalifah (1971). C02

concentrations were calculated based on a saturated C02

solution concentration of 33.8 mM at 250C (Poker and

Bjorkquist, 1977). After dilutions were made, final C02

concentrations ranged from 0.85 to 17 mM.

Initial velocities for C02 hydration were determined by

stopped-flow spectrophotometry (carried out on an Applied

Photophysics stopped-flow spectrophotometer) using a pH-

indicator method (Khalifah, 1971). The pH-indicator method

takes advantage of buffer-indicator pairs which have nearly










identical pKa values. The buffer-indicator pairs used in

this study, their pKa values, and the wavelength observed

were as follows: Mops{3-(N-morpholino)propanesulfonic acid}

(pKa 7.2) and p-nitrophenol (pKa 7.1, 400 nm); Taps {N-

tris[hydroxymethyl]methyl-3-amino propanesulfonic acid} (pKa

8.4) and m-cresol purple (pKa 8.3, 578 nm); Ches {2[N-

cyclohexyl amino] ethanesulfonic acid} (pKa 9.3) and thymol

blue (pKa 8.9, 590 nm). The difference in molar extinction

coefficients between the acid and base forms of the indicator

was determined by spectrophotometric titration at the buffer

concentration, ionic strength, temperature, and wavelength

used in the stopped-flow experiments. Final buffer

concentrations were 50 mM. All steady-state experiments were

performed at 250C with a total ionic strength of the solution

maintained at a minimum of 0.1 M using Na2SO4.

To record the initial velocity of hydration (v), the

rate of change of the absorbance of indicator versus time was

measured by rapidly mixing the contents of two syringes. One

drive syringe contained a CO2 solution, and the second

contained enzyme, buffer and indicator. The initial velocity

of hydration was determined by least-squares analysis of a

minimum of six traces of indicator absorbance versus time,

each comprising less than 10% of the complete reaction

(Rowlett and Silverman, 1982). Initial velocities were

corrected for by subtracting the uncatalyzed rate for C02

hydration.










The initial velocity (v) can be calculated as described

by Rowlett and Silverman (1982) and is related to the buffer

factor, Q, described by Khalifah (1971). Briefly, Q is a

function of the state of ionization of the buffer and

indicator, and relates changes in absorbance in indicator to

changes in concentration of H+. It is calculated as

described by Khalifah (1971).

The main panel of Figure 11 presents a typical

Michaelis-Menten, velocity-vs-substrate profile for CO2

hydration catalyzed by the mutant K64H-R67N-F198L HCA III.

The inset shows the Lineweaver-Burke plot for these data.

The program Enzymatic Analysis in Enzfitter (Elsevier

Biosoft, Cambridge, UK) was used to calculate steady-state

constants from these data. The enzyme concentrations used in

these studies varied depending on the C02 hydration activity

of the enzyme but were generally in the range of 93 nM to

1 pM.


4-Nitrophenyl Acetate Hydrolysis Kinetics


Initial velocities of the hydrolysis of 4-nitrophenyl

acetate catalyzed by carbonic anhydrase III mutants were

measured (using a Beckman DU7 Spectrophotometer) by the

method of Verporte et al. (1967) in which the increase in

absorbance was followed at 348 nm. Absorbance-vs-time traces

were recorded for 3 minutes. To investigate the hydrolysis

activity over a range of pH values, solutions containing the





























Figure 11. Typical data analysis used to calculate steady-
state kinetic constants.

The main panel shows a plot of the velocity-vs-substrate
curve for the hydration of CO2 catalyzed by the mutant K64H-
R67N-F198L HCA III at pH 6.35 and 250C. The reaction was
carried out in 50 mM MOPS, ionic strength 0.1 M. The inset
shows a double-reciprocal plot of these data.





















K64H-R67N-F198L HCA III

10-1 .t,


2.40


2.00


1.601


IA





0.80-
1i






ii

0.40 i
i


0.00
o.oo 00
0.00


_..o'
.--'

A-.

f


.-"
./

L4
/


111


J-V


.---


,I..


1.00


0.90 -


0.80.


0.70-


0.60-


0.50o

i
0.40 i ...


2.00
1/ E C02


0.20 0.40 0.60 0.80
CIC023


1.00
mM


1.20


4.00
(cM--1)


1.40


10-1


10 1


I I


A~-
,

+/
i j
a

i
i/


/

./
,.=
. .










following buffers were used: Mes (2-N[morpholino]ethane-

sulfonic acid); Mops; Hepes (N-2-Hydroxyethylpiperazine-N'-2-

ethanesulfonic acid); and Taps. Reaction solutions contained

33 mM of one of the above buffers whose total ionic strength

was maintained at 0.1 M with Na2SO4. All measurements were

made at 250C.

The molar absorptivity for 4-nitrophenyl acetate is 5.4

x 103 M-1cm-1 (Verporte et al., 1967). The cell path length

was 1 cm and the final substrate concentration in the

reaction mixture was 1 mM. Enzyme concentrations varied

depending on the catalytic activity of the mutant being

studied (the general range was from 0.5 pM to 5 pM). The

enzyme catalyzed rate was determined by subtracting the

uncatalyzed rate of hydrolysis from the catalyzed rate. The

steady-state rate constant, kcat/Km, termed kenz, for 4-

nitrophenyl acetate hydrolysis was calculated using the

following equation (eq. 6):



kenz = kcat/Km = AR (eq. 6)
(E) (b) [S] [E]



where AR = catalyzed rate uncatalyzed rate of 4-nitrophenyl

acetate hydrolysis, b = path length, e = molar absorptivity of

4-nitrophenol, [S] = substrate concentration, and [E] =

enzyme concentration.










Equilibrium Kinetic Methods


180-Exchange at Chemical Equilibrium


The measurement of the rate of hydration-dehydration
cycles of CO2 by the exchange of 180 between CO2 and water is

based on the method of Mills and Urey (1940). Oxygen-18

labeled bicarbonate and carbon-13 labeled bicarbonate were

prepared as described by Silverman et al. (1979). Isotope

exchange experiments were performed on a Dycor N-100 residual

gas analyzer at 250C in the pH range 6 to 9. Adjustments in
pH were made with NaOH and H2SO4. In all experiments, the

total ionic strength of solution was maintained at 0.2 M by
the addition of Na2SO4.

Each experiment was started by placing into the vessel

8.0 ml of 100 mM Hepes (pH 7.5 to 7.7) in which 180- and 13C-

labeled bicarbonate had been dissolved. A period of several

minutes elapsed to allow approach to chemical equilibrium.
At this point, measurements of the isotopic content of CO2

were taken. The uncatalyzed rate of CO2 and HC03- inter-

conversion was subtracted from the enzyme catalyzed rate for

all experiments.

The fate of the 180-12C-labeled bicarbonate atoms and the
160-13C-labeled bicarbonate atoms for carbonic anhydrase

catalyzed 180-exchange at chemical equilibrium is illustrated

in Figure 12. 180-labeled atoms are represented by the filled

oxygens. This technique allows the measurement of the atom
fraction of 180 in all C02. Silverman and Tu (1976) and

Silverman et al. (1979) have described in detail how the rate






59


constant, 0, describing the exchange of 180 between CO2 and

H20, and the rate constant, 0, describing the exchange of 180

between 12C- and 13C-containing CO2 species can be obtained

from experiment. These rate constants, in turn, describe R1

and RH20 (see Chapter 1). The standard deviations in RH20

were 10 to 25% with the poorest precision at higher values of

RH20. The standard deviations in R1 were less than 10%.





















+HZn+H2
+ H .


+H20
--- 3ZnOH2
+H20


HCOO@-+ ;ZnOHz # ZnOH- + COz + H20


+13C02 ZnOH2 + H13CO00-
+H20


Figure 12. Carbonic anhydrase catalyzed 180-exchange at
chemical equilibrium.

This technique allows for the measurement of 180 in all C02
species, and thus R1 and RH20. The 180-label is designated by
the filled oxygen atoms.













CHAPTER 3
CATALYTIC ENHANCEMENT OF HUMAN CARBONIC
ANHYDRASE III BY REPLACEMENT OF PHE 198 WITH LEU



Introduction


The replacements Lys 64 -+ His and Arg 67 -9 Asn in HCA

III, as reported by Jewell et al. (1991), result in mutants

with modest (3-fold or less) enhancement in kcat/Km for

hydration of CO2. Neither of these replacements at positions

64 and 67 causes the pKa for the zinc-bound water to increase

into the pH range above 6 or to enhance the very weak

catalytic hydrolysis of 4-nitrophenyl acetate. Histidine at

position 64, which occurs naturally in HCA II and has been

placed in HCA III, is necessary for maximal turnover in the

hydration of CO2 by providing a pathway for proton transfer

between zinc-bound water and buffers in solution (Tu et al.,

1989; Jewell et al., 1991).

There have been no previous kinetic experiments that

reflect on the specific function or role of Phe 198 in

catalysis by HCA III. We have approached this problem using

the site-specific mutant in which Phe 198 is replaced with

Leu, the residue that appears in this position in HCA II. In

addition, the double and triple mutants with replacements

also at positions 64 and 67 were studied to detect










interactions between these positions in catalysis. Catalysis

of the hydration of CO2 was measured by stopped-flow

spectrophotometry and the exchange of 180 between CO2 and

water by mass spectrometry. The replacement Phe 198 -+ Leu in

HCA III caused major increases in kinetic constants for CO2

hydration and 4-nitrophenyl acetate hydrolysis and in the

tightness of binding of some sulfonamide inhibitors. Thus,

Phe 198 is a significant contributor to some of the unique

properties of carbonic anhydrase III.


Results

The steady-state parameters for the hydration of CO2

catalyzed by mutants of HCA III with replacements at

positions 64, 67, and 198 were measured by stopped-flow

spectrophotometry. The pH dependence of kcat/Km could be

described by a single ionization with a maximum at high pH;

values of this ratio measured by 180 exchange at low total

substrate concentration were identical to values measured at

steady state. The replacement Phe 198 -) Leu resulted in a

large enhancement of kcat/Km for the hydration of C02, about

20-fold compared with wild-type HCA III (Table 3). The

further replacement Arg 67 to form the double mutant R67N-

F198L HCA III resulted in a 70-fold increase compared with

the value of 3 x 105 M-1s-1 for wild-type HCA III. There was

no further enhancement in this parameter caused by the

replacement Lys 64 -4 His to form the triple mutant (Table 3).
















t C

p 0
0 U

O'0




0
S-H
(0c
0,40







(0
4-1 4


a )

( N

-4

M (
i0)



4-)


01

4-




>OI


4j-)

-4-4
8 0


SU)







i>1

0
Srl












tO -H 4
rdl

C U

















(0 (0
1 4-1
Q T 4-
(0 <
E-iu


-H
U)

r-4
0

nO (a

>,4

(D

4-)
40








0

-4 I






i4-)













>1
0
0'
TO
'-
















x


0 )
O 0
0) 4-4


CD
r-
Q.0






04



a






o
C.4
a-,



> .


0
44
o,-



.4-
x c


4 0


O
C4-

-H 04


0) -H



0



CON

(00



4-)
-H


0)r






4-)

oa
( S
40)4
4-

lo-H


TO


0




4



0
O




0




-H
C m
0
o 0

M


'O


00 >




M 4
o












41 0)
> 0



















-H-
o (0






C

0
0)
--H


.'-)
'0 >


-H
o ,-


0 (N

> >




0 0) 4J
-)r





0



U)
-4-4
9 co


3 -H


- 3 0


0)
4-)







>1
0



0
(A
-4





0




-M
-i-l
or





4-J
0
41
















-4
(0




























CA
o
rl









0

4-1






-0


-H



'O




0



-41



-4
Ur
-r




-H 0
i0 >
4- o









Maximal values of kcat/Km for these and two additional mutants

are given in Figure 13. Each of the mutants containing Leu

198 showed changes in kcat/Km consistent with an ionization of

pKa near 7, assumed to be that of the zinc-bound water with

similarities to HCA II (Table 3; Khalifah, 1971).

The presence of an ionization affecting catalysis was

confirmed by the measurement of catalytic hydrolysis of 4-

nitrophenyl acetate. The resulting pH-rate profiles can be

described as dependent on a single ionization with values of

pKa between 6.4 and 6.9 (Table 3). The catalytic rate of

this hydrolysis by the wild-type bovine isozyme III is very

small (the maximal value of kcat/Km is 11 M-1s-1, Tu et al.,

1986). By this measure, the double and triple mutants of HCA

III listed in Table 3 have activity nearly equivalent to that

of isozyme II.

The steady-state turnover number kcat for hydration

catalyzed by the single mutant F198L (Table 3) did not vary

in the pH range 6.5 to 8.5 and was approximately 10-fold

higher than the wild-type HCA III which is 2 x 103 s-1 in this

pH range (Jewell et al., 1991). The double mutant R67N-F198L

HCA III had values of kcat similar to the single mutant,

except at pH > 8.5 for which kcat increased with pH. This

effect may be caused by proton transfer involving Lys 64. A

similar effect was observed in the single mutant R67N HCA III

(Jewell et al., 1991). There was further enhancement

























Figure 13. Comparisons of kcat/Km for hydration of CO2
catalyzed by variants of HCA III obtained by site-directed
mutagenesis at positions 64, 67, and 198 in the active site
cavity of HCA III.

Values of kcat/Km in M-1s-1 appear beneath each designated
mutant. The values in kcal/mol adjacent to the arrows are
the changes in free energy barriers for the catalysis cor-
responding to the designated mutations. Free energy changes
are determined using AG = -RTln[(kcat/Km)mut2/(kcat/Km)mutl]






















Lys64 Arg67 Phe198
3 x 105 M-is-1


(kcat/Km)C02


-1.9
kcal/mol


Lys64 Arg67 Leu198
7 x 106


-0.7




Lys64
1 x



0.0


-0.6


Asn67 Phe198
106


-1.8
S Lys64 Asn67 Leu198
2 x 107


0.0


His64 Asn67 Phe198
1 x 106


-1.8
-1 His64 Asn67 Leu198
2 x 107










observed for the triple mutant for which kcat attained values

comparable to those for HCA II of 1.4 x 106 s~1 (Khalifah,

1971).

The proton transfer-dependent rate of release of water
from the active site, RH20/[E], was measured in the absence

of buffers other than the substrate itself, CO2 and its

hydrated forms. RH20/[E] varied with pH for the triple mutant

(Figure 14) in a manner similar to HCA II (Tu and Silverman,
1985). The maximal value of RH20/[E] for the triple mutant,

9 x 104 s-1, can be compared to the maximal value, near 5 x
105 s-1, for HCA II. The values of RH20/[E] for F198L and

R67N-F198L HCA III are similar near 104 s-1 and rather

independent of pH (Figure 14). For wild-type HCA III, the

values of RH20/[E] are independent of pH at 2 x 103 s-1.

Maximal values of RH20/[E] for additional mutants are given

in Figure 15.

Upon addition of the buffer imidazole, there was
considerable enhancement of RH20/[E] for the single and

double mutants (Figure 16). Enhancement of RH20 by addition

of imidazole is a characteristic of HCA III and the K64H and

R67N single mutants (Tu et al., 1990) as well as HCA II and

some mutants (Tu et al., 1989). This has been interpreted as

evidence for proton transfer from the imidazolium cation to
the zinc-hydroxide increasing the rate of release of H2180 as

shown in equation 4. For three mutants of isozyme III

containing the replacement Phe 198 -- Leu, the rate R1 of

























Figure 14 The pH dependence of RH20/[E] for (A) F198L HCA
III; (*) R67N-F198L HCA III; (*) K64H-R67N-F198L HCA III. The
values for HCA II are represented by the solid line (see Tu
and Silverman, 1985).

Solutions contained 100 mM total concentration of CO2 and
HC03- and no buffers were used. Experiments were carried out
at 250C with the total ionic strength of solution maintained
at 0.2 M with Na2SO4.










106


105







104


7 8


pH


103

























Figure 15. Comparisons of the water-off rates RH20/[E]
(maximal values under the conditions of Figure 14) for
variants of HCA III at positions 64,67, and 198 in the active
site cavity of HCA III.

Values of RH20/[E] in s-1 appear beneath each designated
mutant and the values in kcal/mol adjacent to the arrows are
the free energy changes for the barriers in catalysis for the
corresponding variants.
















RH20/[E]


Lys64 Arg67 Phe198
2 x 103 s-1


-0.7




Lys64
6x


-1.3
kcal/mol
---m His64 Arg67 Phe198
2 x 104


-0.6


Asn67 Phe198
103


-1.2
S His64 Asn67 Phe198
5 x 104


-0.3


-0.4


Lys64 Asn67 Leu198
1 x 104


-1.3
S His64 Asn67 Leu198
9 x 104

















(0 c
(0i

H
OH



N

OH

H
s-I
ct



rH





H
ico
rD






-H
-O

4-)





C 1


44


-x1 H1
C4H






H


O







0 .~Z

4 u) I.

0 lw
0 M
0

















-U M


E-

o


o

0
+O

,j Z
(0
4r
O

O4
O



4-d


>1
H0




O *-1
4-)
O

-H 4






0
0
41
(0
0 -







0
C
















O

- (0
03











.4 0
oE 4J
U O



rd C











-HCO







73

O



00

oo




00
0 0
1z


Wo


C o


0

CD

] E


o I
'- N


m


N(0

SP 0
CM

< *<

0










L-S (<-O. x) [3]IL pue [h]OzHU









interconversion of CO2 and HC03- at chemical equilibrium

appears to be slightly inhibited by the addition of

imidazole, as determined from the negative slopes of the

plots in Figure 16.

The inhibition of two mutants and wild-type carbonic

anhydrases II and III by two anions and the sulfonamides

acetazolamide and ethoxzolamide are compared in Table 4. The

reported values of K, were obtained from 180-exchange

experiments and are the concentrations of inhibitor required

to reduce R1 to 50% of its uninhibited value under the

condition that the total substrate concentration ([CO21 +

[HC03-] = 25 mM) is much less than the apparent binding

constant for substrate Keff. At the pH of these experiments,

7.2, Keff is about 200 mM for HCA II (Simonsson et al., 1979)

and greater than 200 mM for HCA III and the two mutants in

Table 3 as determined by substrate dependence studies (data

not shown). The values of KI determined from RH20 in the

same manner agreed within experimental uncertainty ( 25%)

with the values determined from R1. For the tightly bound

inhibitors ethoxzolamide and acetazolamide, we accounted for

the mutual depletion of free enzyme and inhibitor (Segel,

1975).

Solvent hydrogen isotope effects (SHIE) observed for
R1/[E] are unity for isozymes II and III and two mutants of

HCA III (Table 5). This is consistent with a wide body of

data indicating no rate-contributing proton transfer in the

steps of the interconversion of CO2 and HCO3- shown in













Table 4. Inhibition constants KI (micromolar) were
determined from R1, the catalyzed rate of interconversion of
CO2 and HC03- at chemical equilibrium. Measurements were made
at pH 7.2 with conditions as described in the legend to
Figure 14. No buffers were used.



KI
Inhibitor
HCA III F198L R67N-F198L HCA II
HCA III HCA III


IM
Acetazolamide 40 0.6 0.03 0.06
(0.4)a (0.08) (0.007) (0.02)

Ethoxzolamide 8 0.1 0.004 0.008
(0.08) (0.02) (0.0007) (0.003)

OCN- 30 50 30 30
(0.3) (7) (7) (10)

I- 30,000 30,000 30,000 30,000
(300) (5,000) (6,000) (9,000)


a Values in parenthesis are the estimated, pH-independent
values for KI describing the binding of inhibitors to the
zinc-bound water form of the enzymes. These values were
calculated as explained in the text using the values of pKa
determined from the esterase activities (Table 3). For HCA
III these values were calculated assuming a pKa of 5.2 for
the zinc-bound water.













Table 5. Solvent hydrogen isotope effects on R1 and RH20
catalyzed by HCA III, HCA II, and two mutants. Data were
obtained at an uncorrected pH meter reading of 7; other
conditions were as described in the legend to Figure 14.



HCA III F198L K64H-R67N- HCA II
HCA III F198L-HCA III


(R1) H20
1.03 0.05a 0.96 0.08 1.05 0.07 1.05 0.03b
(R1)D20


RH20
0 2.4 0.4a 2.7 0.7 4.5 0.5 8.0 0.7b
RD20



a Kararli and Silverman (1985).

b Tu and Silverman (1982), this measurement at an uncorrected
pH meter reading of 6.6.










equation 1 (Silverman and Lindskog, 1988). The SHIE for

RH20/[E] are greater than 2.0 for each case in Table 5,

consistent with proton transfer as a rate-contributing step

in the pathway for the release of water from the active site

(eq. 4). For comparison, kcat for hydration of CO2 has a SHIE

of 3.8 and 2.5 for HCA II and III, respectively (Steiner et

al., 1975; Kararli and Silverman, 1985).


Discussion


The most significant structural differences near the

active sites of isozymes II and III of human carbonic

anhydrase are at positions 64, 67, and 198. We have prepared

mutants of isozyme III in which each of these residues is

replaced by the amino acid present at the corresponding

position of isozyme II. The aim of this work is to determine

the catalytic role of these residues and the source of the

large differences in catalytic and inhibition properties

between HCA II and III.


Interconversion of CO2 and HCQ3-


The replacements Arg 67 -+ Asn and Phe 198 -4 Leu

affected kcat/Km independently; the lowered energy barrier for

the double mutant was the algebraic sum of the energy changes

for each of the single mutants. This is demonstrated in

Figure 13 in which the change in energy barriers for kcat/Km

of the catalysis is given by the value adjacent to the

arrows, in the manner described by Carter et al. (1984). The










replacement Lys 64 -* His had no effect on kcat/Km (except at

pH > 8, see Jewell et al., 1991). These results can be

understood by the rather widely separated positions of these

residues in the active site cavity of isozyme III (Right

panel of Figure 3). Phe 198 is on the hydrophobic side of

the active-site cavity and Lys 64 and Arg 67 are on the

hydrophilic side. Moreover, these two positively charged

residues are separated by 8.8 A, the distance between the

side-chain N-C of Lys 64 and the C-C of Arg 67 in the crystal

structure of bovine CA III (Eriksson 1988). The distance

between these positions and the zinc is 12.6 and 9.2 A,

respectively.

We have also observed a pH dependence in kcat/Km for

hydration of CO2 in the HCA III mutants containing the Phe

198 -+ Leu replacement. This is in contrast to the values of

kcat/Km for wild-type CA III which are independent of pH in

the range of pH 6 to 8 (Tu et al., 1983; Engberg et al. 1985;

Kararli and Silverman, 1985). The lack of a pH dependence is

attributed to the low pKa of the zinc-bound water. The

visible absorption spectrum of Co(II)-substituted bovine CA

III (Engberg and Lindskog, 1984) provides additional evidence

that for wild-type HCA III the pKa of the zinc-bound water is

less than 6.0. The pH dependence of kcat/Km in the Leu 198-

containing mutants suggests that the pKa of the zinc-water is

above 6 and closer to the value of pKa 7 for HCA II.

The pH dependence of catalysis for the Leu 198-

containing mutants was supported by observation of the










hydrolysis of 4-nitrophenyl acetate in which the effect of an

ionization with pKa near 6.5 is apparent. There is very

strong evidence that the pKa obtained from ester hydrolysis

reflects the ionization of the catalytically important metal-

bound water (Pocker and Sarkanen, 1978; Simonsson and

Lindskog, 1982). We conclude that the Phe 198 -+ Leu

replacement has increased the pKa of the zinc-bound water

from a value below 6.0 in wild-type HCA III (the exact value

has not been determined because the enzyme denatures before

the ionization is reached) to a value near 6.5 for the Leu
198-containing mutants. The values of pKa from the esterase

data are a more precise estimate than those obtained from

kcat/Km for CO2 hydration.

The increases in kcat/Km for the mutants of HCA III

containing Leu 198 (Table 3 and Figure 13) can be attributed

to the steps in equation 1. The ratio kcat/Km for hydration

of CO2 contains rate constants in the pathway up to and

including the release of HC03- from the enzyme; in terms of

the simplified pathway of equation 1, kcat/Km = klk2/(k_1 +

k2). This ratio is unaffected by the intra- and inter-

molecular proton transfer steps described in equation 2 (see

for example Jonsson et al., 1976; Tu et al., 1990). Unlike

HCA II in which kcat/Km is dominated by k1 (Lindskog, 1984),

isozyme III probably has significant contributions to kcat/Km

of all three constants k1, k-1, and k2 (Rowlett et al., 1991).

Thus, the Phe 198 -+ Leu replacement could enhance kcat/Km at

one or more of three steps: CO2 binding, the interconversion










on the enzyme, and the dissociation rate of product HC03-. To

emphasize the complexity in explaining these changes, we note

that the mutant of the efficient carbonic anhydrase II made

by the replacement Leu 198 -4 Phe (that is, L198F HCA II) has

a value of kcat/Km that is about as great as that observed for

wild-type HCA II (Ren et al.,1991; Taoka et al., 1991).

The increased value of the apparent pKa of the catalysis

upon the replacement Phe 198 -4 Leu is not due significantly

to changes involving the two positively charged side chains

of Lys 64 and Arg 67 which have minor effects on this pKa

(Table 3). It is interesting that this increase is caused by

the replacement of a hydrophobic residue by another hydro-

phobic residue. This increase in pKa is very difficult to

explain since it could arise, for example, from rather small

conformational changes that affect the alignment or distance

of the hydrogen bonds to zinc-bound water and that stabilize

the zinc-bound water over zinc-bound hydroxide (see for

example Scheiner and Hillenbrand, 1985). It is known from

the structures of isozymes II and III that the zinc-bound

hydroxide (and zinc-bound water in the low pH form) is

hydrogen bonded with the side chain of Thr 199 and with two

water molecules in the active site forming part of a

partially ordered network of at least nine hydrogen-bonded

waters (Eriksson et al., 1988). One of these waters forms a

hydrogen bond with the t-electron cloud of the phenyl ring of

Phe 198 in HCA III (Eriksson, 1988) suggesting that the Phe

198 -4 Leu mutation may affect this water structure.










Proton Transfer


The replacement Phe 198 -4 Leu caused up to 10-fold

increases in kcat for hydration (Table 3) and RH20, the proton

transfer-dependent rate of release of water from the active

site (Figure 14) without introducing a residue capable of

proton transfer. For the single mutant F198L and the double

mutant R67N-F198L HCA III the values of kcat are the same and

generally independent of pH at 2 x 104 s-1, except at pH > 8.5

for the double mutant which may be due to the ionization of

Lys 64. The values of RH20/[E] are also the same for this

single and double mutant at about 1 x 104 s-1, again

independent of pH. These properties of kcat and RH20/[E] are

similar to those observed for H64A HCA II, a mutant of the

efficient isozyme in which the internal proton transfer group

of His 64 is replaced with alanine, a side chain which cannot

transfer protons. For H64A HCA II, kcat is near 1 x 104 s-1

(pH 7.2 with a noninteracting buffer such as Mops; Tu et al.,

1989), and RH20/[E] is independent of pH near 1 x 104 s-1.

The solvent hydrogen isotope effects of Table 5 are

consistent with a rate-contributing proton transfer in RH20

catalyzed by two mutants of HCA III containing Leu 198. A

reasonable hypothesis is proton transfer between the zinc-

water and water in the active site.

Both kcat and RH20/[E] were further increased for the

triple mutant; kcat increased to a value near 1 x 106 s-1 and

RH20/[E] to 1 x 105 s-1 (Table 3, Figure 14). This triple










mutant contains His 64 which increases these parameters

because of the introduction of an internal proton shuttle,

the imidazole of the side chain, to facilitate proton

transfer between zinc-water and solution. This is supported

not only by the increased values of kcat and RH20/[E] in Table

3 and Figure 14 but also roughly by the pH dependence of

these parameters (Steiner et al., 1975; Tu et al., 1989;

Jewell et al., 1991). The solvent hydrogen isotope effect in

RH20 for the triple mutant, Table 5, indicates a considerable

contribution of proton transfer to the overall rate of RH20-

We give, in Figure 15, the changes in energy barriers
for RH20/[E] upon mutations in HCA III. Each of the enzymes

in the first column contains Lys 64 and each in the second

column contains His 64; therefore, the energy changes

associated with the horizontal arrows of Figure 15 pertain to

the introduction of the internal proton shuttle. This

decrease in the energy barrier of 1.3 kcal/mole is

independent of the replacements we made at positions 67 and

198. The intramolecular proton transfer between His 64 and

the zinc-bound hydroxide is believed to occur through

intervening hydrogen-bonded water (Venkatasubban and

Silverman, 1980), and our results suggest that the

replacements Arg 67 -4 Asn and Phe 198 -4 Leu do not influence

this process. The measurements of RH20 were made in the

absence of added buffers so that the additional pathway for

proton transfer involving buffer in solution (Tu et al.,

1989) is avoided in these measurements.









Both kcat and RH20 for isozyme III are believed to be

dominated by proton transfer to water (Kararli and Silverman

1985; Rowlett et al. 1991). Proton transfer energetic

involving water are sensitive to changes in the distance of

the hydrogen bond and deformations of the 0-H ...O angle.

For example, increasing the 0 ... 0 distance by 0.1 A can

increase the energy barrier for the proton transfer (H20-H-

H20)+ by as much as 5 kcal/mole (Hillenbrand and Scheiner

1986). One other consideration is the frequent observation

that the energy barrier for proton transfer is directly

related to the difference in pKa of the proton donor and

acceptor groups. On this basis alone, we would predict that

the mutants containing Leu 198 would have kcat for hydration

less than for wild type because of the greater proton

affinity of the zinc-bound water in the mutants (pKa = 6.5)

compared with the wild type (pKa < 6.0). However, the

opposite was observed (Table 3). Consequently, changes in

the structure of hydrogen-bonded water caused by the

different side chains in the mutants may be more important

than the initial and final energy levels in explaining the

proton transfer rate between zinc-bound water and water in

the active site.

Another effect of the Phe 198 -4 Leu replacement in HCA

III is to enhance the apparent second-order rate constant

kH2OB/KeffB (see Tu et al., 1990) for the transfer of protons

from imidazole in solution to the enzyme. For F198L HCA III

this constant is 5 x 105 M-1s-1, obtained from Figure 16, and









for the double mutant R67N-F198L HCA III it is 9 x 105 M-1.
The constant kH20B/KeffB for the triple mutant is difficult to

estimate because of experimental scatter, but is quite low,

presumably because this enzyme has the internal proton
transfer group of His 64. The value of kH2OB/KeffB for wild

type HCA III is 1 x 105 M-1s-1 (Tu et al., 1990). The

increase in kH20B/KeffB caused by the replacement Phe 198 -4

Leu could be du'e to enhanced accessibility of the small

buffer imidazole to the active site caused by the removal of

the steric hindrance of Phe 198 or to some more subtle

structural change. It is possible that proton transfer

between imidazole and zinc-water occurs through the inter-

vening water structure just as proposed for His 64 (Silverman

and Lindskog, 1988).


Inhibition


Table 4 presents the apparent inhibition constants KI at

pH 7.2 for two sulfonamides and two anions determined by

inhibition of 180 exchange catalyzed by HCA II, HCA III, and

two mutants. It is known that anions such as SCN- and

halides bind directly to the metal in carbonic anhydrase,

displacing water or forming a fifth ligand (Lindskog, 1982;

Eriksson et al., 1988). The binding of anions to the zinc-

hydroxide form of isozyme II is very weak compared with that

of the zinc-water form (approximately 100-fold weaker; Tibell

et al., 1984). This has not been shown specifically for









isozyme III but is almost certainly the case. The pH

dependence of the binding of acetazolamide indicates that a

similar situation pertains (Lindskog and Thorslund, 1968).

We can convert the apparent values of KI in Table 4 to

pH-independent binding constants for the zinc-water form of

the enzymes by dividing each value of the apparent KI by (1 +

Ka/[H+]) using the values of pKa determined from the

hydrolysis of 4-nitrophenyl acetate (Table 3). These

estimates of pH-independent KI are given in parentheses in

Table 4. The replacement Phe 198 -- Leu caused a considerable

increase in tightness of binding for both acetazolamide and

ethoxzolamide, resulting in values of KI closer to that for

HCA II. Eriksson et al. (1988) have shown that the thia-

diazole ring of acetazolamide bound to HCA II is in van der

Waals' contact with the side chain of Leu 198. It is most

likely an unfavorable steric interaction between the

thiadiazole ring and Phe 198 that contributes to the weak

binding of acetazolamide to HCA III. On the other hand, the

Phe 198 -* Leu replacement in HCA III caused a weaker binding

of OCN- and I-, consistent with the known weaker binding of

anions to HCA II compared with HCA III (Table 4; Sanyal et

al., 1982).










Conclusion


It was previously suggested (Kararli and Silverman,

1985) that Lys 64 and Arg 67 were responsible for the unique

catalytic and inhibition properties of carbonic anhydrase

III, specifically its low activity in CO2 hydration and ester

hydrolysis, low pKa for zinc-bound water, and weak binding of

sulfonamides. The present work complements Jewell et al.

(1991) in showing that Lys 64 and Arg 67 have relatively

minor roles in determining these properties, but that Phe 198

can be identified as a major contributor to these unique

features of carbonic anhydrase III.














CHAPTER 4
EFFECTS OF THE MOLECULAR PROPERTIES OF RESIDUE 198 ON
CATALYSIS AND INHIBITION OF HCA III




Introduction


The previous chapter illustrated the important role Phe-

198 plays in the catalytic hydration of C02, hydrolysis of 4-

nitrophenyl acetate, and the binding of inhibitors by HCA

III. This chapter focuses on how the molecular properties of

residue 198 affect catalysis and inhibition, and what

consequences this has on the two-stage catalysis.

Six replacement residues (Ala, Asn, Asp, Leu, Tyr, and

Val) were introduced at position 198 which differ in size,

charge, hydrogen bonding capability, and hydrophobicity.

These residues were selected in part based on the surrounding

active-site cavity environment. Unique to this active site

is its water structure (Figure 2). For example, water 335

has been shown to hydrogen bond to the X-electron system of

the benzene ring of Phe 198. In turn, water 335 is involved

in a hydrogen bonding network through the hydroxyl group of

Thr 200 and water 318 to the zinc-bound water of the active

site (Eriksson, 1988). Since no analogous pattern is pos-

sible for HCA II, this unique aspect of the water structure

could contribute to the catalytic properties of HCA III.










Indeed we have found that all of the replacements caused

an increase in CO2 and HC03- interconversion. Non-hydrogen

bonding residues (Ala, Val, Leu) had kcat for hydration 10-

fold greater than hydrogen bonding residues (Asn, Asp, Phe,

Tyr). In addition, a correlation between hydrophobicity and

the turnover number was found. These data suggest an altered

proton transfer pathway in these mutants.


Results


The steady-state rate constant, kcat/Km for the hydration

of C02 catalyzed by HCA III and six mutants at position 198

was measured by stopped-flow spectrophotometry and 180

exchange between CO2 and water. It is an interesting

observation that every replacement made at this position

resulted in a mutant with kcat/Km for C02 hydration greater

than that of wild-type HCA III (Table 6). The largest

changes were observed with the replacement Phe 198 ->Asp

which resulted in a 130-fold enhancement of kcat/Km compared

to wild-type HCA III (Table 6). The replacement Phe 198 -4

Tyr caused the least effect. Two of the mutants, F198Y and

F198V HCA III, were similar to wild-type in that no pH

dependence of kcat/Km for hydration was observed in the range

of our measurements (pH 6 to 9). However, the pH dependence

of kcat/Km for the mutants F198A, F198D, F198L, and F198N HCA

III could be described by a single ionization with a maximum

at high pH (Table 6). Here the largest change was for F198D

HCA III which had the apparent pKa for catalysis increased to













Table 6. Maximal (pH-independent) steady-state constants and
values of the apparent pKa for the hydration of C02 and rate
of release of H2180 catalyzed by mutant and wild-type carbonic
anhydrasesa


Enzyme kcat/Km kcatb RH20/[E]
(x 10-6 M-1 s-1) (x 10-4 s-1) pKa (x 10-4 s-1)


wild-type HCA IIIc 0.3 0.2 <6.0 0.3

F198Y HCA III 0.5 0.2 <6.0 0.2

F198V HCA III 1.6 1.0 <6.0 0.3

F198N HCA III 6.2 0.2 7.1 d

F198A HCA III 6.5 2.5 6.5 1.5

F198L HCA III 7.4 2.2 6.9 1.0

F198D HCA III 43 0.2 9.2 d

wild-type HCA IIe 150 140 7.0 60


a Values of apparent pKa for catalysis were determined from
kcat/Km for the hydration of C02 with standard errors in the
pKa of 0.2. Values of kcat had standard errors less than 20%
and kcat/Km less than 12%.

b These values omit an increase at pH > 8 which has been
attributed to the ionization of Lys 64 (Jewell et al., 1991).

c From Jewell et al.(1991).

d pH-dependence not described by a single ionization. See
Figure 19.


e From Khalifah (1971).