21'

H

Za]

Ho
{:C.‘
.\.,

ABSTRACT
AN X-RAY CRYSTALLOGRAPHIC EXAMINATION OF THE INTERACTIONS
OF a-CHYMOTRYPSIN NITH IRREVERSIBLE INHIBITORS ANO
POTENTIAL TRANSITION-STATE ANALOGS
By

Michael Norman Liebman

The three-dimensional structure of a-chymotrypsin
(a-CHT), a representative enzyme of the serine protease
family, as determined by Tulinsky _t _l., provided an
ideal system in which to examine the structure-function
relationship. The existence of two independent molecules
of a-CHT within the crystalline dimer permitted an exten-
sive probe of specificities within the binding and
catalytic regions of the active site using two inhibitor
classes, irreversible, covalently-linked inhibitors and
potential transition-state analogs, while allowing the
observance of structural variability.

Irreversible inhibition of a-CHT by phenyl alkyl
sulfonyl fluorides has been established in solution
studies. The compounds studied, p-toluene sulfonyl
fluoride (TOS) and phenyl methyl sulfonyl fluoride (PMS)
both form covalent sulfonyl ester bonds with SER l95.
Examination of the structure of the complexes allowed

observation of definitive cOvalent linkages with SER l95

while also allowing the study of those interactions

Michael Norman Liebman

responsible for specificity. Structural analysis
emphasized the similar but independent molecules within
the crystalline dimer while recognizing two unique but
consistent binding orientations across the two-fold axis.
Observation of the SER l95 interaction allowed evaluation
to be made of the applicability of the potential transi-
tion-state analogs.

As a-CHT is expected to involve a tetrahedral inter-
mediate in its catalytic mechanism, it has been proposed
that phenyl alkyl boronic acid derivatives would be iso-
steric with normal substrates and allow observation of
the bound state by crystallographic methods. Variation
of the phenyl alkyl group, phenyl ethane boronic acid
(PEBA), phenyl propyl boronic acid (PPBA) and phenyl
butyl boronic acid (PBBA), as well as pH with PEBA,
allowed for the observation that at pH 5.4 and pH 7.3,
the interaction more closely resembled the covalent
linkage of TOS, but formed a weaker complex at pH 3.6.
Further studies involving PPBA and PBBA established the
limiting features of the specificity pocket of a-CHT
and correlations with previous kinetic studies.

The direction of the research has been to establish
the basis of enzyme recognition and specificity, with
emphasis on molecular organization and conformational

adaptability. Towards these goals, the method of

ral

f!)
"

n.

Michael Norman Liebman

diagonal plots was expanded to incorporate difference
Fourier analysis and in this manner the domain structure

of a-CHT was further studied.

ll 1-21! '-‘

V l

a
!
y I J b

in Dar

AN X-RAY CRYSTALLOGRAPHIC EXAMINATION OF THE INTERACTIONS
OF a-CHYMOTRYPSIN WITH IRREVERSIBLE INHIBITORS AND
POTENTIAL TRANSITION-STATE ANALOGS

By

Michael Norman Liebman

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1977

elc‘lc‘fso

In Memory of My Father

ii

"n
..e a
a::'e:ia:‘l

;-.":‘a'.:e a

V R
'0 er

u pl
‘fl I
l‘ n
, M t.
i "I
«.m w
J 5-H
333p
l ,
ii“.
I Vna1
3;

ACKNOWLEDGMENTS

The author wishes to express his sincerest
appreciation to Dr. Alexander Tulinsky for his constant
guidance and support, enthusiasm and friendship.

To Dr. S. N. Timasheff and Dr. R. B. Beard, the
author would like to extend his appreciation for their
early support and encouragement. To Dr. Richard L.
Vandlen, the author extends particular appreciation for
his interactions and encouragement in the early part of
this work. The author would like to express apprecia-
tion for their many helpful discussions to Dr. Irene
Moustakali, Dr. Aristides Mavridis, Mr. Lawrence Weber
and Mr. Lyndon Hibbard.

The author would like to thank Mr. Paul Hunt,

Mr. Tom Smith, for their technical assistance during
various phases of this study and Ms. Janet L. Sayers
for her technical assistance and helpful discussions
during the final stages of this work.

Support of the Molecular Biology Section of the
National Science Foundation during the course of this
study is gratefully acknowledged as is the support of
the Lower Merion Township Scholarship Fund for support

of the author's undergraduate education.

-gfl‘
.144

IN

r0

lull,

TABLE OF CONTENTS

I. INTRODUCTION.
A. General.
B. a-Chymotrypsin (a-CHT): Chemistry .
C. a-Chymotrypsin: Crystallography .
D. Transition-State Analogs and a-CHT .
II. EXPERIMENTAL .
A. Crystallization of a-CHT .
B. Data Collection.
C. Data Processing.
III. METHODS OF ANALYSIS .
A. Emphasis of Analysis

B. Examination of the Substitution in the
Active Site. . . .

C. Examination of Conformational Adaptation
Accompanying Inhibitor Binding .

D. Inhibitor Orientation in the Independent
Molecules. . . . . . . .

E. Inhibitor Orientation Definitions.
IV. RESULTS AND DISCUSSION
A. Irreversible Inhibitors.

l. Substitution characteristics Of
TOS and PMS

2. Orientation of inhibitor molecule
and interaction with a-CHT.

iv

18
29
36
36
41
49
64
64

66

74

82
82
87
87

87

92

3.

TABLE OF CONTENTS

Comparison of TOS (MSU) and
T05 (MRC) . .

B. Competitive Reversible Inhibitors:
Transition—State Analogs

l.

7.

Substitution characteristics of
PEBA as a function of pH.

Orientation of PEBA molecule and
interaction with a-CHT as a
function of pH.

Evaluation of PEBA as a transition-
state analog.

Comparison with enzymezinhibitor
(natural) complex structures.

Phenyl alkyl boronic acid orienta-
tion as a function of alkyl chain

length. . . . . . . . . .

Orientation and interaction of

phenyl alkyl boronic acids in the
active site .

Substrate specificity of o-CHT.

V. DIAGONAL PLOT ANALYSIS.

A. Analysis of the Distance Diagonal Plot
of a-CHT (Native, pH 3.6). . . . .

3. Comparison of the Structures of a-CHT
with Pancreatic Trypsin Inhibitor by
Use of Distance Diagonal Plots

C. Perturbations to the Tertiary Structure
Accompanying Active Site Binding

REFERENCES.

Page

105

110

110

118

130

T38

T42

144
150
T6]

161

I74

182
200

=::*.::x

”ESZIX

”EVZIX

V

APPENDIX A.

APPENDIX 8.

APPENDIX C.

TABLE OF CONTENTS

Coordinates of Inhibitors Determined
in This Study (a = 49.2 A, b =
67.2 , c = 65.95 A, 6 = lOl.8°).

Calculated van der Naals Contacts

of TOS, PMS, TOS (MRC), PEBA pH 3.6,
PEBA pH 5.4, PEBA pH 7.3, PPBA and
PBBA. . . . . . . . . . . . . . . .
Coordinates of Difference Fourier
Peaks Used in Calculation of
Difference Diagonfil Plots (a =

49.2 A, b = 67.2 , c = 65.96 A,

B = lOl.8). . . . . . . . . . . .

vi

208

218

226

Table

10

ll

l2a

12b

LIST OF TABLES

Specificity of a-CHT with Respect to
Methyl Esters of Some N- .Acetyl- L-
Amino Acids . . .

Sequence Homology in the Vicinity of
Active Site Residues Among Several

Serine Proteases.

Classes of Derivatives Studied in a-CHT .

Comparison of Lattice Parameters for
a-Chymotrypsin with Addition of
Dioxane . . . . . . .

Possible Configuration of Tetrahedral
Centers for Interaction with SER 195.

Lattice Parameters of the a-CHT
Derivative Crystals Used in This Study.

Comparison of Derivative Crystal
Characteristics in This Study

Comparison of Derivative Crystal |AF|
Distribution Statistics . . . .

Comparison of Cell Parameters of
Irreversibly Inhibited a-CHT.

Comparison of the Centroid Positions of
Substitution Density in the Active Sites
of Molecules l and l' (TOS and PMS)

Orientation Angles Determined for the
Inhibitor-Enzyme Structures

Summary of van der Naals Contacts (< 3.5 A)
Between Native a-CHT and the Inhibitor
Molecules . . . . . . . . . . . .

Summary of Inhibitor Contacts with Specific
Residues of Native a-CHT (< 3.5 A . .

Page

14

21

23

32

48

50

55

87

91

93

95

96

a.

H u
:4

PFV

or;
n/ys

n/5

Q/\

«(J

.l/S

l‘
‘15

Table
l3

14

15

16

17

18
19

20

21

22

23

24
25

LIST OF TABLES

Comparison of Unit Cell Parameters
of TOS (MSU) and T05 (MRC).

Comparison of Contacts Nithin van der Waals
Distances Between Tosyl and a-CHT in MSU
and MRC Studies . . . . .

Comparison of Cell Parameters of PEBA
at pH 3.6, 5.4 and 7.3.

Comparison of Peak Heights of Difference
Electron Density Regions in the Active
Site of PEBA pH 3.6, 5.4 and 7.3.

Comparison of Centroid Positions of

Substitution Density in the Active Sites
of Molecules l and l' (PEBA pH 3.6, PEBA
pH 5.4 and PEBA pH 7.3) . . . . . . . .

Cell Parameters for PPBA and PBBA .
Comparison of Peak Heights of Difference
Electron Density in the Active Sites of
the PPBA and PBBA Derivatives

Comparison of Centroid Positions of
Substitution Density in the Active Sites
of Molecules l and l' (PPBA and PBBA)
Comparison of Homologous Series of
Inhibitors, Substrates and Potential
Transition-State Analogs of a-CHT .

Structural Features of a-CHT Diagonal
Plot. . . . . . . . . . .

Intersection of Mirror Planes with
Diagonal of Distance Plot of a-CHT.

Intrachain Loop Palindromy.
Residues in Proximity of Localized

Sulfate Ions.

viii

Page

106

111

112

113

117
142

143

147

153

165

167
169

173

P

3:39'd‘ix AI
fiird‘ix A9

1»-

I‘vu

In.

 

.51GC

 

LIST OF TABLES

Table

26 Sequence Comparison of PTI and a-CHT
Domain A Based on Alignment of Distance
Diagonal Plots.

27 Individual Analysis of Difference
Electron Density Distance Diagonal
Plots . . . . . . . . . . .

28 Correlation (%) of Difference Electron

Density Peaks (Other Than Substitution)
Within the Active Site Region of a-CHT.

Appendix A-l

Appendix A-2

Appendix B-l

Appendix B-2

Appendix C

Coordinates of Inhibitor Determined
in This Study (a = 49. 2 b =
67.2 c = 65. 95 A., e= lOl. 8° )

Bond Distances and Bond Angles Of
Inhibitors Determined in This
Study .

Calculated van der Naals Contacts
Between Native a-CHT and Irreversible
Inhibitors (TOS and PMS) Inhibitor
Coordinates from Appendix .

Calculated van der Naals Contacts
Between a-CHT (MRC) and Tosyl
Inhibitor (MRC Results) (49).

Coordinates of Difference Fourier
Peaks Used in Calculation of
Diffe ence Diagonal Plots (a =
49.2 , b = 67. 2 A, c = 65.95 A,

B = 101.8°) . . . . . .

ix

179

192

198

209

216

219

225

227

a
354 “I,

11'

n I
a a

(sauna Pu

..

0nd

”(U

5U
9.1

a. U AH T. i...

Figure

10
11

12

13
l4

155

LIST OF FIGURES

Amino Acid Sequence of a-Chymotrypsinogen.

Activation and Autolytic Reactions of
a-Chymotrypsinogen A (4)

pH-Activity Profile of a-CHT Catalyzed
Hydrolysis of a) N-acetyl-L-tryptophan
amide, and b) N-acetyl-L-phenylalanine
amide in water at 25°C (10). . . . . .

Early Ideas of Serine l95 Participation
in Anomalous Activity. . . . .

Chagge- Relay Mechanism of Blow _t _1.
3l . . . . . . . . .

3D Derivatives of a-CHT at 2. 8 A
Resolution . . . . . .

Schematic Packing Diagram of a-CHT Viewed
Along a*-Direction. Non-crystallo raphic
Dimer = I and I' or II and II' (4l?

Variation of Unit Cell Parameters of
a-CHT with pH (43) . . . .

Distance Diagonal PlotA of Native a-CHT
(45), Contoured at l5 A(Solid Regions).

Potential Transition-State Analogs

Preparation of Crystals of Native
a-Chymotrypsin . . . . . . .

Procedure for Preparation and Verification
of Derivative a-CHT Crystals . . .

Inhibitor Molecules Used in This Study

Crystal Selection and Data Collection
Procedures . .

Distribution of Diffracted Intensities
Along Principal Axes of a-CHT Complexes
Used in This Study .

Page

10

17

19

20

22

24

27
34

37

39
40

43

45

r\
m

I\I
r\>

6‘? (J

‘n

(‘1 TI:

 

Figure
16a

16b

17

18

19

20

2la

21b

22

23

24

LIST OF FIGURES

Oat? Trocessing--Reduction of Intensities
to F .

Data Processing--Conversion of IFI'S to
F's. . . . . . . . . . . .

Histograms of Derivative Crystal |AF|
Distribution Statistics. . . . .

Absorption Correction Curves-~PEBA pH 3. 6,
Crystal 6- 67- SI- 3. . . . . .

Schematic of y-z Projections of the
Difference Electron Density in the Active
Site Region for TOS a-CHT. .

Comparison of Composite Projections of

the TOS Difference Electron Density in the
Active Si es of the a-CHT Dimer; Contoured
at 0.15 e -3, Positions of TOS l and T05
l' are Arbitrary in z. . . . . . . .

Comparison of Composite Projections of
the Substitution Electron Density of TOS
and PMS in the Active Site of Molecule 1;
Contoured at 0.15 eA 3 . .

Comparison of Composite Projections of
the Substitution Electron Density of TOS
and PMS in the Active Site of Molecule l';
Contoured at 0. l5 eA-3 . . .

Model of TOS as Fitted to Neighted
Centroid Projections of the Difference
Electron Density

Distance Diagonak Plot, r1; , of a-CHT,
Contoured at 15 only (45) and Exhibiting
Dimer. . . . . . . . .

Schematic Representation of a-CHT
Domains. . . . .

xi

56

59

67

68

69

70

73

76

78

"I
QTIV

Figure
25

26

27a

27b

28

29

30

31

32
33a

33b

34

LIST OF FIGURES

Intermolecular Domain Contacts in a-CHT
Crystalline Dimer--a) Schematic of
Cross-Distance Diagonal Plot, and b)
Linear Packing of Dimers and Domains of
a-CHT.

Definitions for Diagonal Distance Plot
and for Difference Distance Plots.

Difference Distance Plot of Conformational
Changes on Increasing the Native, pH = 3. 6
a-CHT Crystal to pH = 7. 3 in the Presence
of PEBA. . . . . . . . . . . . .
Substitution Distance Plot of Substitution

of PEBA Observed on Increasing the Native,
pH = 3.6 a-CHT Crystal to pH = 7.3 .

Definition of Orientation Angles for
Comparison of Inhibitor Orientation.

Variability Between Irreversible Inhibitozs

Viewed Down y,z-Axis, Contoured at O.l5 e -

with Approximate Two-Fold Representation

TOS and PMS Orientation in x-z Projection
of Active Site .

TOS (MRC) Orientation in x- 2 Projection
of Active Site (MRC) .

TOS (MSU) vs. TOS (MRC) Position .

PEBA in Region of Active Site, Molecule
1, Viewed Down y,z-Axis, Contoured at
0.15eA-3

PEBA in Region of Active Site, Molecule
1', Viewed Down y,z-Axis, Contoured at
O.TSeA-3

PEBA pH 3. 6 Orientation in x- 2 Projection
of Active Site . . . .

xii

£9.92

79

81

83

84

85

90

98

108

109

115

116

119

”Vi

'Hv
A)"

:95

51‘

Figure
35

36

37

38a

38b

39
40
41
42
43

44

45
46

LIST OF FIGURES

PEBA pH 5. 4 Orientation in x- 2 Projection
of Active Site . . .

PEBA pH 7. 3 Orientation in x- z Projection
of Active Site . . .

Energy Diagram of Proposed Enzymatic
Mechanism Involving a Transition-State
for Catalysis (96) .

PPBA in Region of Active Site Viewed Down
y,z-Axis, Contoured at 0.15 eA-3 3, with
Approximate Two-Fold Representation.

PBBA in Region of Active Site Viewed Down
y,z-Axis, Contoured at 0.15 eA3 3, with
Approximate Two-Fold Representation.

PPBA and PBBA Orientation in x- z Projec-
tion of Active Site. .

Substrate/Inhibitor Side- Chain Specificity
in a- CHT . . . .

Distance Diagonal Plot Definitions

Distance Diagonal Plot of a-CHT Dimer De-

picting Representative Structural Features,

Contoured at 15 A (Solid Regions).
Domain Configuration of a-CHT Dimer.

Comparison of Distance Diagonal Plots of
a-CHT and PTI. . . . . . . .

Difference Diagonal Plots.

Identification of Diagonal Plot Features

of Native a-CHT to be Compared in Difference

Diagonal Plots

xiii

£999.

120

121

132

145

146
149
157
162
163
171
177

185

191

S

9': '3

.V

g
D

T...
a.

bin

II WHI 0‘1 :0..- “NM
:0. C PC a) n-
F V n n AF»
. 3a a. 1: I P:- n x.
“a ri .14 r . r
A \- ‘ a 1.. as :-
F . P .w . IA: 1 P
0 out A 1! AUX .\ I t n
3 I I I. 1' D t R d

I. INTRODUCTION

A. General

The chemical processes of living systems are almost
entirely dependent on the catalytic activity of enzymes.
Two fundamental questions in biochemistry are: What are
enzymes and how do they operate? One answer that has
emerged is that enzymic activity depends on the structure
of the molecule and particularly on the chemical nature
of that part of the enzyme which comes into contact with
the substrate and is generally referred to as the active
site. The purpose of the present work is to utilize
X-ray diffraction in studying two specific structural
aspects of a-chymotrypsin (a-CHT): the interaction at
the active site upon binding of substrate/inhibitor, and

the source of the specificity of the enzyme.

B. a-Chymotrypsin (a-CHT): Chemistry
a-CHT is a proteolytic enzyme which is found in
pancreatic tissue and pancreatic juice. Its empirical
formula, C1113N3000349H1752512, indicates a molecular
weight of 25,300 amu and corresponds to a total polypep-
tide chain length of 241 amino acid residues. As with
othe1~ enzymes, a-CHT is secreted in an inactive form, or
ZymosJen, in this case, alpha-chymotrypsinogen (a-CTN).

The anmino acid sequence of this zymogen is used as a

.;.;. ,
'31'396'15

:An. 00 .

'4‘ ~-
.3QI'II-
’1 i. ‘-

“r3 5" V'-
' VI 5
‘4’ .-
a.vil 1n¥._
137‘ I:
'w 14 36:
Q 1'. ”'1.

 

Numbering basis for a-CHT (Figure l). The sequence

determination was performed by Hartley £3 31. (1), and
verified by Meloun e_t_ 11. (2), with a final minor
revision based on examination of the X-ray results of

Blow at g. (3).
Al though the activation scheme for a-CTN has been

establ ‘i shed (Figure 2) (4), the exact function of each

form of the enzyme is not yet known. The activation to

a-CHT involves the removal of two dipeptides, SER 14-

ARG 15 and THR 147-ASN 148, by tryptic and chymotryptic

hYdr‘o‘lysis of a-CTN. The final structure consists of

three polypeptide chains, A, B, and C, containing 13,

131. and 97 residues, respectively, connected by five
disu'l fide bridges and various polar and non-polar inter-
aetiohs.

Kunitz and Northrop (5) first isolated a-CHT in

1935 and crystals of the enzyme were among the first to

be examined by X-ray diffraction (6). Because of the

ease Of obtaining the alpha form and its relatively high
puri ty when prepared from bovine sources, a-CHT has
become one of the most extensively studiP-d enzymes.
”Never, it is beyond the scope of this work to allude
to 3“ 1 of the observations reported and the intention is
only to deal with some of the results which have been

ob
SeY‘ved consistently and with the theories which have

be
e“ developed to explain them.

z:< 2.; zx< 2.. 2.. :.u >_: >_: r>c .I; r_: 3:; >3: ::_ r>4 ar< 2.: 3:3 zen ~<> 2.: ale :1; 12R 2.x

>1; 31.. -.<> <.< 2 1.. 2 .2 >4... Z...< _<> 1. :«::<«I__.w 22.; >3. 1.... 3-... ..<> 3:: 2-x; .41: <4< Ell .:\> >42 7.x...»

.... .... ... «u- .u.

camocwmaxgpos»;u-a to aucmacom cwu< 6:45< .4 443446

 

.TI.II.II.II.:wm;o-u
ommume .Nmpuwmp .Pomummp .mmum¢ .NNFIF

.IIIIIII :wmgonm m lllllll cwmcuu< onh<szmmo zH<Iu

 

mmuammm mon4=mHo
.cmmazggosxsoia op conm>Ppom newest pmop mw mzuwmmg mwzp A44
mew u mmsuwmmg mo Lucas: Punch

mm ¢ m mm mm m o N ep mp o— N mm m op op m ¢— ¢ mm

4<> m>p amp mzh mum om; mza Hm: m>4 =m4 m4H mH: >4w 24c 24w m>o am< zm< wm< <4<

 

onhHmoazoo

m>4 mzh >4w nah m>h m>4 m>4 m>u zm< mzh zm< mum :u4 =m4 can :M4 mum <4< 24w 24w am; wm< am< can «I»

 

zm< <4<«zm<¥mzh m>h wm< azh :m4 >4o nah >4w mzh mzh 4<> m>o azh «IF >49 <4< <4< mzm mm< mm< mum <4<

 

mum om; =m4 m>u 4<> <4< mum 4<> mzp 24w mum mza mum <4< <4< mzh mum =m4 m>4 =m4 3mg «1* m4H mm< zm<
zm< u4H mzk :m4 mum zm< m>h m>4 mum zm< w>4 mzm 4<> m>4 <4< m4H m>4 =m4 m>4 24w m4H m>4 24c mmm mmm

 

 

a E 244 444 4:4 34 :4 54 444 444 4g 2 E E 44: 4g 2 E =44 :4 s 44 4444 s 4: 444:

zm< =49 zm

Pom
cup
Fm—
omp
pop
mm

TIIIIIIIT.< 444 :44 «mm >44 >4u m>u 41a m4: ”In >44 4:» m>4 am< 24¢ :44 «mm 4<> 249 644 cam axe mum mm

:m4 >4@ «mm 3m; 4<> can 24w m4H <4< cum 4<> >4w m>u
op m

>44 oma 4<> <4< :46 :44 >44 zm< 4<> 444 wm< mum
mN ON imP

~

1

1
1r\

“15va CIT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A9615“ LlLE‘G
j .
L J1 L 11
] 1
f [
M. ‘1 «I J
--——-4>»
J
I_1 L—IL——'J r n
k 1
}— *
4 an] A .. v-cut
coir 1 .
€013 ULE
LJ 6' 'I‘TT
I [1 ' L [I
J j
E .(l
001 L] . k j
"R“G‘ ‘ALA j
' I II [VIII
,1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-- Moo-CTN 4-“?
O I O
,Issuu uu:16_] dim, unit—J
[' 1F L 11
1 [)Li [ T‘T
. _ 1 ——-b . k . j
,. 9 <—-—-—-
R146; .ALAuj R146; :ALAMO.)
L 11
" I-CHT " r-cut

FIGURE 2. Activation and Autolytic Reactions of a-Chymotrypsinogen A
4

U

.--.;4
6".
'1’1
n'ou“

Au.
.5.)-

0"

.IJ

..
91A
“'14

all

a-CHT catalyzes the hydrolysis of proteins and
their derivatives, polypeptides and amides, as an endo—
peptidase. .Ig.vitgg, a-CHT has also been noted to
catalyze the hydrolysis of esters. Bergmann and Fruton
(1936) (7) first observed the a-CHT catalyzed hydrolysis
of carbobenzyloxyglycyl-L-tyrosy1 glycinamide, proving
that synthetic peptides could be used as substrates.
The specificity of a-CHT for aromatic peptide side-chains
has been observed and associated with the presence of
an aromatic binding pocket in the active site; the order

of specificity is summarized in [1] below.

0
Site of g_
proteolysis-—\\ ' [1]
~CH-C-NH-—CH
1" 1
R2 0 R]

R2: Aromatic > Cycloalkyl > Alkyl

The effect of this specificity is seen in Table l (8),
from which it can also be seen that other large, hydro-
phobic peptides are susceptible to chymotryptic attack,
although at a greatly reduced catalytic rate. The
characteristic rate enhancement observed in enzyme
catalysis is seen in N-acetyl-L-tyrosine ethyl ester

1

hydrolysis, where the rate constant is 0.45 min" at

25°C for standard base catalysis (NaOH), and 0.12 x 105

»~.\ ..,._c—-.< C:—_:<11_1~>a.eu<12 OEOW to

n........x.. .XZQSZ 3.. 0.431132 24.3 5.201». 6‘3 k.upU~&‘Ua..~1W

- u:-n~<~

4.o.4=oav 4 64454

 

 

chmpmpxcmga

. . Iwzu -ogczgmxm:

N o co? x o m -4-_»4du<-z

. . N Pamogxp

N o 404 x mm m 1 Io‘ 0: -4-r»4ad<-z
z

Fxcngouaxgg

4.0 m04 x «.4 -4-4»466<-z
.WIU

425v 24 44-64-2V 24\4564 atsud=s4m ewaeo-du4m agate 446<

 

AQV mowo< ocps<-4-4»4au<-z meow 46
6264mm 44:46: 64 podamdm 444: 4:4-5 to xuwuecwuaam

_ m4m<h

I ...u
21:; Ev. A—iw.IZV .v.\«.. y. OLJHUELuW :‘ccoicUfim

A.u.u:ouv F wpnmp

 

 

m
:u chowcpms
_-oF x _.N -NIU-NIU-+mnmm -4-_ngas
:u -mipzwmo<-z

m

TN 1 \\:u
. :o :o/@ quamp
_ N No, x mm.~ :u -4-_zpmo<-z
. N N m Pacorspms
No, x m N i :o. :oim- Io -4npapmu<iz
. . Iwzu Fxcmpmpxcmsa
N F ¢o_ x N a -4-_xpmo<-z
E

425v 4 44-64-2v s4\4a64 at=4u=t4m =4aeo-664m asosm Fsu<

 

4.6.4EOOV _ 6446p

:2; Ex 4.1.6.}:

5. AV U
xxu x oxzoozcum :46:015o.m QJOLo NAOQ

Nah». s~hnl~m~V \ L.U~A§l.‘§

E

pcmpmcou mcwucwn mwpmmguwz u x

 

 

pcmwmcou mam; ovpzpmpmu n pmux mews:
mm m-o4 x w.m .: 446445
-4-_»4dd<-z
so Fxsmuapm
N-o_ x m.m -N:4-N=4-0Am - -4-_»4au<-z
o
425V 24 4_-mp-zv 54\4ad atsudstpm =45;U-du4m asstw _»d<

 

4.6.4eouv 4 644a»

‘1‘

P:;.-‘
uh". ,

(a
“It:

Abs
in

321219

min.1 at 25°C for a-CHT catalysis (9). Other solution

studies have yielded information concerning the composi-
tion of the active site of a-CHT, which has been verified
by the crystallographic results.

Examination of the activity profile of o-CHT vs. pH
Shows a bell-shaped behavior with a maximum occurring
between pH 7.5 and pH 8.5 (Figure 3) (10), with two
separate ionizing groups apparently controlling this
behavior. When a carboxylic acid is placed in a solution

18

near optimal pH containing 0 -enriched water, the

reaction, shown by [2] (11), takes place.

18-

R-COO- + H20 --- R-C00 + H20 [2]

The positioning of amino acids in the active site must
account for these observations, and locating their
positions can only be accomplished by careful chemical
experimentation.

A unique serine residue, SER 195, was located in the
active site as a result of inhibition studies with

diisopropyl fluorophosphate (DFP) [3].

CD
(3
I

\ [3]

10

Ao-v uamN an noun: a? uuvsa mcwcmpapacmga
-4upxumuauz An new .muPEm cmsaounxyau4ipaumum-z Am
46 mvmapatcaz ua~4454au Azu-a.co 6444644 444>446<-=a .m 443444

 

Nd . mic fwmiMummimL
m4 4 mx

 

 

 

 

11

Jansen _t‘_l. (1949) (12) first observed that a-CHT was
stoichiometrically inactivated in a (1:1) manner by
reaction with DFP. Later work by Shaffer gt 11. (1953)
(13) revealed that this was due to the acylation of

SER l95. Oosterbaan and von Adrichem (1958) (14) showed
that the acylation step of the reaction with substrate
was analogous to the reaction with DFP when comparison of
the acetyl peptide and phosphoryl peptide segments proved
them to be identical. These results were particularly
interesting since under ordinary conditions, serine is
not susceptible to acylation or phosphorylation, while

in a-CHT, one serine of twenty-eight is reactive (15).
Chemical conversion of SER 195 to ALA in the production
of anhydro-chymotrypsin eliminates catalytic activity but
not the substrate binding capability. SER 195 has thus
been implicated as a catalytic residue of a-CHT but not
directly involved in the specificity of binding.

HIS 57 was first implicated in catalytic function
upon the observation that photo-oxidation of this residue
eliminates enzyme activity (16). Due to the unique pK
(6.0) for histidine residues, HIS 57 has been assigned
as the group responsible for the control of the low pH
leg of the activity profile. The strongest evidence
for the definition of its role in catalysis was the work
of Schoellman and Shaw (17), who alkylated HIS 57 in

a-CHT and also in trypsin (T), a related enzyme, using

shylatirg

ti’liiC‘it;

'9."
.31:
1,": .63
owz ~|-
'-
.1.
Ta
IV
A :1
Y

(‘8 “—
vhf-191113.."
N'CETOIE;

:EEA carri

 

12

alkylating agents which take advantage of enzyme

Specificity and are shown in [4] below.

 

a-CHT 0 ’T 0
¢-CH2-EH-C-CH2-Cl CHZ-CHz-CHZ-CHZ-fH-C-CHZCl [4]
HH NH2 TH
Tosyl Tosyl

¢ = @ Tosyl = CH3 9 502-

Recent X-ray studies of a-CHT in the presence of the

Schoellman—Shaw reagent have established some new ideas

Concerning this work (18). Methylation of HIS 57 has also

been carried out but without total loss of enzymatic

acti‘vity (19,20). Solution studies of HIS 57-methylated

GPCPTT reveal a slight reduction of the acylation rate
cOHS'tant but no reduction in the binding constant, thus
1|“P'l'icating HIS 57 in a catalytic role similar to SER l95,
rattler than participation in substrate binding (20).

ASP l02 was misinterpreted to be asparagine (ASN) in

the original amino acid sequencing work (3). Comparison

of flhe homologous enzymes trypsinogen, thrombin, and
Porszine elastase, coupled with crystallographic results
13" ‘to a reassessment of the observations associated with
this residue and its reassignment to aspartic acid.

ASpv 102 has also been assigned a catalytic role in the

charge relay system of Blow t _a_l_. (3), but its role is

Sti'11 subject to debate.

”a;

L. “‘
v..yc

'Fiancrea

”a: a can

 

13

It should be noted that several enzymes, some of
them isolable from pancreatic juices, including chymo-
trypsinogen A and B, chymotrypsin A and B, trypsinogen,
trypsin and elastase, have homologous sequences in the
vicinity of SER 195, H15 57, and ASP l02 (Table 2) (2l).
The enzymes indicated in Table 2, subtilisin, thrombin
and several other proteases, constitute a class denoted
as "serine proteases," all of which exhibit proteolytic
activity centered around the serine, but with varying
side chain specificities (22). It has been suggested
that the variation in specificity of the proteases found
“I pancreatic juice is an example of divergent evolution
from a common precursor, while the other proteases,
Par“ticularly those of bacterial origin, are examples of
the! convergent evolution of similar enzymes (23).

Three other amino acid residues of a-CHT have been
in\Iestigated for possible involvement in the catalytic
activity of a-CHT although they are not necessarily
homologous with other serine proteases. MET l92 has been
1""‘lb'licated in enzymic activity on the basis of selective
°Xi dation experiments. Oxidation of MET 192 by either
ph(Ito-oxidation or chemical means affects substrate
hydrolysis without alteration of MET 180, the only other
methionine in a-CHT. Koshland _ei 1. (24) showed that

the maximal velocity of hydrolysis of the substrate

89:;5

nnnnnn

: CYS t
1 HS t
.f, CYS (
f: US l
:j CYS l
2" CYS
CYS
CYS l
TYR .

.wgti, .
\JS.

' Deletic
:5 . Reilon
; ' 80‘!an
' POY‘Clrg

 

 

 

I; . 7mm
{i ' BC'J‘Ene
3’: ' 303/in
:" 10min:

.' . Elasta:
' i‘Lyt :
‘ wan”.-

 

Sequence Homology in the Vicinity of Active Site

14

TABLE 2

Residues Among Several Serine Proteases (21)

 

 

 

 

W

('3

)-
Illllllllll

Deletion

Region unclear

Bovine Trypsin

Porcine Trypsin
Thrombin

Bovine Chymotrypsin A
Bovine Chymotrypsin B
Porcine Chymotrypsin A

Elastase

a-Lytic Protease

Subtilisin

Protease 55 56 57 58 59

81 ALA ALA HIS CYS TYR

PT ALA ALA HIS CYS TYR

TH ALA ALA HIS CYS -

BCA ALA ALA HIS CYS GLY

BCB ALA ALA HIS CYS GLY

PCA ALA ALA HIS CYS GLY

E ALA ALA HIS CYS VAL

LP ALA GLY HIS CYS GLY

S ASN SER HIS GLY THR
Protease 191 192 193 194 195 196 197 198
BT CYS GLN GLY ASP SER GLY GLY PRO
PT CYS GLN GLY ASP SER GLY GLY PRO
TH CYS GLU GLY ASP SER GLY GLY PRO
BCA CYS MET GLY ASP SER GLY GLY PRO
BCB CYS MET GLY ASP SER GLY GLY PRO
PCA CYS - GLY ASP SER GLY GLY PRO
E CYS GLN GLY ASP SER GLY GLY PRO
LP CYS MET GLY ASP SER GLY GLY PRO
S TYR ASN GLY THR SER MET ALA SER
Abbreviations:

100 101 102

103

104

 

ASN ASN ASP
ASP ASP ASP
ASN ASP ASP
ARG ASP ASP
GLY ASP ASP
GLY ASP ASP
GLN ASP ASP

213 214 215

VAL SER TRP
ALA - -

VAL SER TRP
VAL SER TRP
VAL SER TRP
THR SER PHE
GLY ASN VAL

ILE
ILE
ILE
ILE
ILE
ARG
ASN

216
GLY
GLY
GLY
GLY

VAL
GLN

MET
ALA
THR
THR
ALA
ALA
SER

217 218
SER X

GLU X

SER SER
SER SER
SER LEU
SER GLY

 

15

remains unchanged although substrate binding is increased by
afactor of five upon oxidation with H202. Scramm and Lawson
(25) observed normal a-CHT inhibition by DFP although MET 192
had been alkylated by benzyl bromide. More recent work by
Taylor _t _l. (26) using C13CSOZCl, a specific MET l92
oxhfizing agent, renewed interest in the role of MET 192.
lheir observations indicate that the oxidized enzyme retains
nmximum catalytic activity over a broader pH range.

GLY 216 occurs in a region of the enzyme which shows
variability in sequence among the serine proteases (Table 2).
As this residue has its alpha carbon atoms positioned at the
entrance of what appears to be the specificity pocket of the
enzwne, the observed variation suggests a source for the
varied specificities. If a larger side-chain, such as
valine in elastase, is present at this position, the specifi-
city is altered from that of an aromatic group in a-CHT to a
small hydrophobic group in elastase (e.g., alanine) (27).

TYR 146 is the exposed carboxylic acid terminal
residue of the B-chain which is formed during the activation
process from a-CTN. It is not involved in the catalytic
mechanism since its removal by carboxypeptidase has no
effect on the catalytic rate (28). However, TYR 146 is
apparently critical for dimerization of a-CHT which pre-
dominates in solution at low pH, and which is observed in

the crystal structure at pH 3.6; the removal of TYR 146

prevents dimerization in solution.

.\C
.p.
v.

a“.

nr"

9“

ii:

IpQA

id

'v‘a.

16

The first proposed mechanisms of a-CHT analysis
attempted to deal with the anomalous reactivity of SER 195.
It was believed that the unexpected reactivity was the
result of the incorporation of the serine hydroxyl group
into intermediate covalent ring structures. Rydon (29)
proposed the formation of an oxazoline ring (Figure 4a).
Cohen t al. (30) as well as Schneider (30) and Smith gt

al. (30) had predicted that an internal ester link between
the serine hydroxyl and the carbonyl of ASP 194 was
responsible for its reactivity (Figure 4b). A fused ring
structure with the serine hydroxy1 protruding was proposed
by Bernhard gt_al. (Figure 4c) (31). In all cases, the ring
structures were presumed to open before the catalytic
process was complete since isolation of the acylated or
phosphorylated serine peptide sequence never revealed these
ring structures:

The observation that imidazole catalyzes ester hydroly-
sis in solution by general base catalysis suggested the
manner in which HIS 57 might participate in a-CHT catalysis.
Gutfreund and Sturtevant (32) suggested that HIS 57 was a
cationic group which was not transiently acylated, but
rather acted as a base catalyst for the acylation and de-
acylation of the SER 195 hydroxyl group.

In the work of Blow et_al. (3), which re-assigned

residue 102 as aspartic acid, a charge relay mechanism

17

 

(‘2021-1
T42 fflcg (o)29
\' CH—c’o
HN—CH
\N/
8—0 “”30
HzT/ \THz
max—(_a—
H H H H
\C'O H-N/

FIGURE 4. Early Ideas of Serine 195 Participation in

Anomalous Activity 1

9"
O:
u w I

35 n Ale
r- .1 .«d i..

”1 ti -\ nih
.‘ U r ‘yl "I HI
A I {F5 ‘49 ero P v

‘ I U -4 ~

‘ . ’I l i -
oi. .. .. o. I .V. v. u

18

involving SER l95, HIS 57, and ASP 102 evolved. The charge
relay system (Figure 5) involves the hydrogen bonding of

ASP 102 to HIS 57 which is further hydrogen bonded to SER 195.
The strong nucleophilic character of SER 195 would then be

a result of the stabilization of hydrogen transfer through

the histidine to the negatively charged aspartic acid

(pK 1.5) (33). This interpretation resulted from observa-
tions of an inhibited derivative of a-CHT, tosyl-o-CHT (TOS),
and is subject to re-evaluation based on the structure of

native-a-CHT (NAT), as determined by Tulinsky gt 31, (34).

C. a-Chymotrypsin: Crystallography

 

The crystallographic determination of the structure of
a-CHT has been the result of two efforts, both independent
and differing fundamentally in their approaches to the
problem. 0. M. Blow's group at the MRC (35) described the
structure initially, as well as some of the early ideas
concerning the structure-function relationship of a-CHT
(36,37,38,39). A. Tulinsky's group at MSU (40) has been
primarily concerned with the details of the structure and
how they relate to several classes of chemical and/or
biochemical behavior (Table 3, Figure 6) (18,41,42,43).

In the MSU approach, individual members of the classes
prove to be only as important as their contribution to the

overall understanding of the behavior of that class.

\‘f/

‘6

102

102

FIGURE 5 .

19

102

Charge-Relay Mechanism of Blow g §_1__.

(31)

 

0:: nLuserM

 

Uruc O'COLCD ccozufiueza .p \N Z“

6.: I; . p. N I :C a 9.-.: I 2.. o .6.5 I 2.. .:.u:.: I>.¢CZ. v N4” 1 IA .v €..N l I3

0
2

. mzatou<ssmno=uz any

gauosomucacmnavv

aga-aua-tm-az-zaav .N
ath-aua-.u-z .F
Fu~omm_uu

\e,
NmF pa:

:0 a x
Nap pa: co.ua_ȴ_< we a. o

tron chogon memsuuumnm .N

tron O’COLon mamcuwlmcm .F

n

a.w Ia +.m.h . =a_+ ¢.m . =a.+ mm.” a In .P=Q-a a>.pan_+ ~.m . :a +_m.~ . za

czogu Noam .m

coruapommm < m.~ on hxuua mo ma>*uo>*guo an .o mmawuu
v'ua uvconapam ocmapou
Law-wsauoua-z
Emc:_uugm.m
mvwsuyamop

unopux-uz-aghtouatz . Fxmnva + Lahmm

v

vmuc uwcogonuocmuaauona .m waua
kum opoucu
uvcogoa-ucoaogn-mgm .N
aghtapuumumuz
Am.“ .a.m .o.m Iav uvua

uvcogonnmcusum-mga .F o:a-»~o-muutz

 

[xflpuoopv .N aah-eaoe-z
I
A, my .p a:a.Etoa-z
3o 35 \/
ouapm
co?uwmccgh
mmcmguxo
a - a
“team nomv aaouaa.gcfi
opnwmgm>mm
u>vuwpunsou

Amll aghtpxsgomnz .p

 

 

 

 

 

 

 

 

. . . . . .
«a «a ‘e- to \O r~ tn :5 c: r-
l- '—

.Amzmv pxcoypzm
panama p»:mca

guaa.<uu<-z
xULaH
pxmapn

O
'—

FAmou
\l.

 

aaaa.avgcunntut_
1| opfiflgotu

sh

§
.

21

TABLE 3

Classes of Derivatives Studied in a-CHT

 

1. pH Conformers

Irreversible Inhibitors
Competitive Reversible Inhibitors
Transition-State Analogs

SO4 -Se04= Exchange
Oxidation of MET 192

Noam-poem

Alkylation of MET 192

 

Both groups have established the same general features
of the a-CHT structure. a-CHT crystallizes in the monoclinic
crystal system, space group P2], with four molecules per unit
cell and two molecules per asymmetric unit. A local two-fold
axis of non-crystallographic origin exists within the
asymmetric unit between the two independent molecules. The
two molecules probably resemble the dimer of o-CHT which is
observed to form in solution at lower pH values. A schematic
of the dimer is shown in Figure 7; however, it does not
possess exact two-fold symmetry (42). This view of the unit
cell, which is approximately parallel to the a*-axis, is
also parallel to the local two—fold axis. The interstitial
space between molecules is occupied by solvent (about 36%

by weight).

22

B x
1! l
P x l O

’ ,
5‘
\ II x
Y TY ‘ T
L f‘yx l, .

A,B=non-crystallographic dyods
it active sites
U =uranyl binding sites

 

 

FIGURE 7. Schematic Packing Diagram of a-CHT Viewed Along
a*-Direction. Non-crystallographic Dimer = I
and I' or II and II' (41)

23

A comparison of the native a-CHT structures as
determined by the MSU and MRC groups has been summarized
by Tulinsky (34,42). A summary of the changes in the
cell parameters as a function of pH (Figure 8), and a
comparison with earlier reported cell dimensions of
a-CHT, is given by Mavridis 33 al. (43). The cell
dimensions for the NAT crystals as used in this study

are given in Table 4. From Figure 7, the molecule

TABLE 4
Comparison of Lattice Parameters for
a-Chymotrypsin with Addition

of Dioxane

 

MSU, Native,

 

MSU, Native, MRC, Native, pH 3.6, addition
pH 3.6, no grown from of 3% (v/v)
dioxane dioxane (34) dioxane
(3| 49.24(7)3 49.1(1)R 48.9(1)A
|E| 67.20(10)A 67.4(1)E 63.3(1)K
|E| 65.94(9)A 65.9(1)A 65.8(1)A
B 101.79(8)° 101.7(1)° 101.8(1)°

v 213,600(1000)A3 214,000(1100)K3 215,ooo(1000)33

 

0

resembles a prolate ellipsoid of about 25 x 28 x 33 A,
with the ellipsoid axes corresponding closely to the a*,

b and c crystallographic axes.

O
'a" '3'-
‘vqt

8.

 

 

 

 

 

24

 

 

 

: s E a ‘ s ! a s 2 a
1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 100
pH

FIGURE 8. Vargatgon of Unit Cell Parameters of a-CHT with
pH 43 ~

5::r3xi'
'75 :once
:arted a

iistorte

:"C '1
Pa‘,
~ 9C1.)

25

An examination of the structure of a-CHT reveals
approximately 7.5% helix and 34% beta sheet. The helix
is concentrated into two regions--10 residues of undis-
torted alpha helix (236-245) and about 10 residues of
distorted helix (164-173). The larger component of
beta sheet is observed in 13 segments of varying length
and degrees of distortion in the anti-parallel mode.
Deviation in both a-CHT and other proteins from the
proposed models of tertiary structures for helix and
sheet has been a major problem in the correct prediction
of structure by spectroscopic techniques. The ILE 16-
ASP 194 internal salt bridge and ASP 102 are three major
ionic groups which are found in the interior of the
molecule, and all are important in that they are involved
in the formation of the active site region. The majority
of the interior is comprised of non-polar groups with
the exception of 18 non-polar groups whose side-chains
appear on the surface of the molecule in the company of
polar residues. Except for +NH3-ILE 16, the amino-
terminal and carboxylic-terminal groups of the three
polypeptide chains formed during CTN-activation appear
on the surface. The intricate folding of the polypeptide
chain has the A-B chain and the C chain forming individ-
ual domains with little overlap. Two areas of
particular interest occur at the interface of these

regions--the active site and a secondary binding site

“f ',H§A§I'-
i'rJr'J"

szidies 1'
tie pcssi

O

.u: sew

26

(tryptophan cluster) (42). Detailed conformational
studies involving the folding of the backbone indicate
the possibility of an evolutionary relationship between
two separate domains within the a-CHT molecule. This

is evidenced by the presence of an apparent mirror plane
in the distance diagonal plot (44) which represents
interactions between all the alpha carbons within the
molecule (Figure 9). Further analysis of this relation—
ship in light of changes which occur upon binding will
be discussed in the present work.

Several other crystallographic studies are of
interest in relation to the a-CHT structure. Several
inhibitor:a-CHT structures have been examined by the MRC
group [DFP (39), Tosyl (39,46,49), N-formyl-L-tryptophan
(47) and indoleacryloylimidazole (48)]. In addition,
Freer £5 31. (50) have described the structure of a-CTN
and have made some observations concerning the process
of activation. Wright (51) has pursued this work
further in a detailed comparison of the structural
differences between a-CHT and a-CTN, and an interpreta-
tion of the process of activation. He has concluded
that although the catalytic triad is intact (SER l95,
HIS 57, and ASP 102), four major enzyme-substrate inter-
actions do not occur in a-CTN:

1) binding within an ill-formed specificity
pocket;

' MP}: 9
‘

27

   
   

MIRROR
1C).

RESIDUE'i
150:

 

 

/ ' RESIDUE')

FIGURE 9. Distance Diagonal Plot of Native a-CHT (45),
Contoured at 15 A (Solid Regions)

n-CHT} r
it revea‘.
residues

are the ‘
211-216 ~'
frat the‘
1"}Suorn.
has not

Siler {F5

1'39 Sat

 

28

2) the GLY (193)-NH-to-substrate hydrogen bond;
3) hydrogen bonds from substrate to GLY (216);

4) the inability for interaction of a polypeptide
substrate leaving group with MET (192).

The structure of the monomeric gamma-chymotrypsin
(v-CHT) has been determined by Segal gt al. (52), and
it reveals little difference in the positioning of
residues in the active site. Of particular interest
are the indications of the importance of the segment
214-216 in forming the specificity pocket which results
from their inhibitor studies. It can be concluded from
this work that the dimerization process in the crystal
does not greatly affect the active site configuration
other than blocking the active site from interacting with
large substrates.

Subsequent X-ray crystallographic studies have
determined the structures of other serine proteases,
subtilisin (SUB) (53), elastase (E) (54), and trypsin
(T) (55). Although SUB is of bacterial origin, it seems
to have converged evolutionarily to the catalytic triad
of other serine proteases SER-HIS-ASP. The divergence
in specificity remains the primary distinction among
these enzymes. However, since the pH activity profiles
also differ, this suggests an influence from some other
phenomenon such as conformational adaptability to

external (i.e., pH) change. The understanding of the

o

“3211“

r\a
3..

.’.I

29

requirement for this particular geometric orientation
as the only one used in the proteolytic activity of
these enzymes remains the goal of the current research

on serine proteases.

D. Transition-State Analogs and a-CHT

As early as 1946, Pauling (56) had expressed the
idea that the greater efficiency in enzyme catalysis
resulted from the complementarity between the transition
state of the reaction catalyzed and the active site of
the enzyme. This suggestion was the foundation for
Nolfenden's application of transition-state analog theory
to enzyme catalysis (57).

Of particular interest to enzyme chemists has been

the structure of ES+ and EA+ [5]:

E + s r-—-=-Es us Es+ --> EA + P -_. EA+_ --> E + P2 [5]

1
where, E = native enzyme,
S = substrate,
ES = initial enzyme-substrate complex
+ (Michaelis-Menten)
ES = transition-state complex,
EA = acylated enzyme,
P] = acylation product, and
P2 = hydrolysis product.

In 1951, Bender (58) had postulated the existence of a
tetrahedral intermediate in ester hydrolysis. The
mechanism for the reaction [6] offered an explanation

for the isotopic exchange he observed when an alkyl ester

30

 

 

_ .1+
1‘8 - + 1
OH or H
R-C-DR + H2016 ........ > €18 3
R--E--0-R -- R-C-OH [6]
o16 _ + o + R-DH
" 18 OH or H |
R-C-DR + H20 ; ------- ‘> L_ H

 

18-enriched water was catalytica11y

18

of benzoic acid in 0
hydrolyzed by OH', and 0 was incorporated into the

unhydrolyzed ester [7].

 

- 1
1 o
18 - "
0‘8H

 

The same process has been observed with a-CHT. The
possibility of observing the suggested short-lived
intermediate has been discarded on the basis of life-
time. However, if the mechanism is correct, the tetra-
hedral intermediate should also appear during catalysis

by a-CHT. To approximate the tetrahedral configuration,
aldehydes and ketones with susceptible carbonyl groups had
been studied with elastase (59), but they yielded com-
plexes with a-CHT which were too weak to permit examina-
tion of their interaction with SER 195 by crystallographic
methods. The tosylated enzyme had a tetrahedral arrange-
ment around the sulfur which is covalently linked to 0y-

SER 195, but the tetrahedral sulfur is not dimensionally

E'IZTIE

3
Hbviat

1".

A
u

a 'n ..

-'e Si

31

isosteric with a tetrahedral carbon (Table 5) (60). It
is apparent that the more closely a model approaches the
correct configuration, the more susceptible it becomes to
enzyme catalysis, and, the more closely the active site
interactions should approximate those which occur in
actual catalysis.

Bell (61) had observed that boronic acids equilibrate

a hydroxyl group according to [8]:
R-B(0H)2 + OH- --‘- R-B(0H)g [8]

The structure produced in this manner has been observed
to be dimensionally isosteric with that proposed for the
tetrahedral carbon (Table 5). As boronic acids are easily
prepared with varying alkyl- and phenyl-alkyl groups,
they appeared to be ideal as transition-state analogs and
to be adaptable to meet the specificity requirements of
the enzyme. Exceptionally large inhibition constants,

as predicted for transition-state analogs, were observed
with phenyl ethane boronic acid and both a-CHT (62) and
SUB (63). A pH dependence for this inhibition was noted
in both studies which reflected the pK of the tetrahedral
transformation of R-B(0H)2. This pH-inhibition profile
also appears to be a function of the pH-activity profile
of the enzyme. Antonov gt 1. (64,65) studied the

effects of pH and variation of the R group in both the

32

TABLE 5
Possible Configuration of Tetrahedral Centers

for Interaction with SER l95

 

 

Enzyme Inhibitor/
Complex Bond Distances (60) Substrate
0
' O
CH2——O-C-<R rR-C = l 54 A Phenylalanine
. O
//H ----- x rc_o - 1 43 A
N+
9H
CHZ—oth-R rR_B = 1.57 A Phenyl Alkyl
o Boronic Acids
H ----- 6H r8 0 = 1.43 A
/ -
.N+
I o
CH2——OD-s-—‘R rR-S = l 83 A p-Toluene and
o Phenyl Methyl
. ----- 5 rs_o = 1.47 A Sulfonyl
Fluorides

 

:J
AH4
‘11
!.1\

n\..v

. "3'5

5'

0 .-
‘Fu

‘,U

A
U

n‘a
.uil

£21116

’1!
.n V

33

alkyl boronic acid and phenyl alkyl boronic acid series.
The side-chain variation allowed for examination of the
interactions within the specificity pocket as well as the
SER l95 interaction. Although a comparison of the experi-
mental results reveals similar dependencies on pH and a
comparable inhibition constant for phenyl ethane boronic
acid, Koehler and Lienhard (1971) (66) have predicted

that SER 195 is involved in the boronate binding, while
Antonov has suggested HIS 57. Phenyl ethane boronic

acid closely resembles a phenylalanine peptide (Figure
10), a typical a-CHT substrate. Thus, if PEBA were a
transition-state analog, it would be of great interest

to determine the details of the interactions within the
active site. The X-ray results of Matthews gt g1, (1975)
(67) on the structure of the complexes of benzene

boronic acid and phenyl ethane boronic acid with sub-
tilisin BPN' (Novo) provide a further source of comparison
of the serine protease class and the specific modes of
binding both to SER 195 and within the specificity

pocket.

The emphasis of the present study has been to
examine the substitution and related changes in enzyme
configuration upon the binding of boronic acid deriva-
tives. Because of the specificity of o-CHT for aromatic
side-chains, the phenyl alkyl boronic acid series was

studied with variation in the alkyl side chain length,

34

POTENTIAL TRANSITION-STATE ANALOGS

 

®
7'95 Q 7I95 3
ENZYME-O-H + B —’ ENZYME -O —-B--OH
//"Ԥ E
on 1 0" ow

n--cw2-cwz.

PHENYL ETHANE BORONIC ACID @ "pH 3.6, pH 5.4, pHT. 3

(PEBA) ' (PEBA3.6)(PEBA5.4)(PEBAZS)
n- -cwz-cwz-cwz~. R- -cw2 -CH2-CH2 -CH2.
PHENYL PROPYL eonomc ACID PHENYL BUTYL. eonomc ACID

(PPBA) ”’33“

FIGURE 10. Potential Transition-State Analogs

35

n = 2,3,4 (Figure 10). The effects of pH on the binding
of the phenyl ethane boronic acid molecule have also

been investigated at three pH values, 3.6, 5.4, and 7.3.
To aid in the interpretation of the SER l95—HIS 57 inter-
actions with these inhibitors, the structures of a-CHT
covalently linked with a tosyl group and with a phenyl
methane sulfonyl group have also been determined, and

the molecules were compared. The differences between

the interactions in the independent subunits were
evaluated. The approach pursued throughout this work
involved interpreting the independent interactions within

each molecule and then comparing these results.

n».
-v

-+
b

.1-

II. EXPERIMENTAL

A. Crystallization of a-CHT

 

The crystals of native a-CHT used in this study were
prepared by the method outlined in Figure 11. This is a
modification of the original procedure developed by
Kunitz and Northrup (5). a-CHT was obtained from
Worthington Biochemical Corporation (68) as three-times
recrystallized, salt-free, lyophilized a-CHT and was used
without further characterization or purification. The
procedure, which involves the salting-out of the enzyme,
was carried out at pH 3.6, close to the optimum pH for
dimerization (69), and corresponds to the pH 3.6 conformer
of Mavridis gt gt. (43). The diamond-shaped platelet
crystals reached a size of 1.5 x 1.0 x 0.5 mm within one
to two weeks of growth at room temperature. The density
of the a-CHT crystals grown in this manner is approxi-

mately 1.25 g-cm3

which corresponds to a solution
containing 37% mother liquor, or an enzyme concentration
of ~0.1 M within the crystal. The presence of the mother
liquor enables the diffusion of inhibitors, substrates,
heavy metal ions and hydrogen ions into crystals and
permits the assessment of the catalytic activity of the
crystalline enzyme.

Crystals of a-CHT were modified by addition of
inhibitor to the mother liquor, by alteration of the pH

36

37

Boil'w500 ml of distilled
[H20 and cool to room temp.l

Adjust m300 ml to pH 3.6
(«.1 drop of 18 M H2504)

 

L_‘

Prepare «400 ml of 75% Adjust to pH 3.6
saturated (v/v) (NH4)2504 ""““*’ («J drop of 18 M H2304)
+ 3.07 M, pH 4.2

[Weigh out 0.570 gm of a-CHT--dissolve in 26.5 ml of pH 3.6 H20 by

 

 

layering the o-CHT on top of the solvent (in 150 ml beaker);
Stir gently (otherwise denaturation might occur) -* 0.85 mM a-CHT

 

 

i

i

 

 

 

Titrate with 75% sat. (NH4)2504 Back titrate (few drops

to first cloudiness which a only) to clearness with

persists (m60-70 m1) pH 3.6 H20, ~53% (NH )2504;
2.17 M (NH4)2$04, 0.2% mM
a-CHT

 

 

 

 

 

l

 

Place m2 m1 of final solution into each of m40 test tubes and
cover with double layer of parafilm; store in plastic test tube
racks at room temperature--do not disturb (growing solution is

 

 

 

f m7 mg/ml)

 

Stop crystal growth upon reaching average crystal size of 1.5 x
1.0 x 0.5 mm, by replacing the growing solution with 75% sat.
(NH4)ZSO solution at pH 3.6--store in water bath at 19°C +
3.07 M (AH4)2504

 

 

 

7

FIGURE 11. Preparation of Crystals of Native a-Chymotrypsin

38

of the mother liquor, or both (Figure 12). The inhibitors
used in this study (Figure 13) were all of non-commercial
sources except for the para-toluene sulfonyl fluoride
(70). Phenyl methane sulfonyl fluoride (PMS) was the

gift of Dr. Lawrence J. Berliner (71); phenyl ethane
boronic acid (PEBA) was donated by Dr. Gustav E. Lienhard
(72); and Dr. V. K. Antonov (73) donated phenyl propyl
boronic acid (PPBA) and phenyl butyl boronic acid (PBBA).
All pH measurements were performed using a Leeds-Northrup,
Model 7411, general purpose pH meter with temperature
compensation and equipped with a miniature Ag/AgCl element
combination electrode, Model 117202. Although the manu-
facturer's specifications state an accuracy of t 0.05 pH
units with t 0.02 pH units reproducibility (74), a more
conservative estimate for the protein work would be a
reproducibility of t 0.05 pH units. The buffers used to
standardize the pH meter were Mallinckrodt AR Standard
Buffer Solutions, pH 4.01 (#0029) for the range pH = 3.0
to pH = 5.5, and pH 7.00 (#0031) for the range pH = 5.5

to pH = 8.5. The pH conformers discussed in the present
work are those described by Mavridis _t gt. (43). In
derivatives where both the pH of the mother liquor was
modified and an inhibitor was added, the change in pH

and subsequent verification of equilibrium preceded the

inhibitor addition.

mpmumxgu H1015 w>wum>wsmo $0 cowvouc._.sm> Ucw covaLmawLm SOL. mLzmeOLm .N_. magma...

In. 8:52. on LBEES 4..----

 

 

 

 

In amigos no 3.58:8 A1111 233.8 2.58: mo-
. mccmuumq ucmucmum guwz :ommgmaeou
m m In an Louwnwsca am11111. an cowumELoE m>wpm>mgwu xwwgm>
mgmsgomcou In .332 A, W
mo:mnumwwmwcmmwwmmwM.1.1 1..1 1 1 1 1. . mmxm Pmpmacu Fmawucwga
gospowcou.:g quLm> . mo cgwppma cowpumcwcmu m>gmmno

 

 

 

 

 

 

 

 

. 11
m 1 ‘1 flow 39. 5 .5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

    

 

cowuapom mew
mmxm qumxgu Lopwnwscw :5 one an PE e_um
Fmawucwca 0o :gmuumg . nugopwnwccw to mmmoxm upom Loaawp gmspos mo Noop mum—amm....1
3508:? 9:38 omfip 53 In cmtmmn pm 1 1 c + _I vm. J
o , com A 1.: .an “E 83o: m - H _ m u
. . - _

9 I _v + C H C —
3 mmowumm _ W ”1| 1 1o11 m 1 1 5.538 933m 5: 5.; Ar 1 1L
p H . Lozcwp gmgpoe co Rom muapamm

. Lona?” gmguoe, zop cop to .
o6 In 338: 1 .V :3; oB Ti . o u c
1 , _
i a . a
. e Nd .
corp cm A Izv pom amu guwz Louvawncv
uspom mcwxoOm 3m: . P5 m cu mcwcn cucu11mpmgpw: :5 oma an F5 op1muucouwnmccw
sow; Lassa. canoes . -ooooa Fe P.o cw Loowawgcw to mmooxo EWOL o~-o_ gov: o.m 2a
Lo weep oompamm . m>Pommwu .xgmmmouw: mH um commfi :zv .amm mm“ mcwawcm
_
misc: em r1111”? *1

 

ova:

 

 

- aommz to zoazz
coves—om mcwxmom sup: Losowp gmsuoe pcocgsu . sosawp
3m: guy: Lozqu zzopmn Lo m>onu maps: m.o mo :8 canvas :? vommfiezzv umpugspmm amu
toaoos to “on ooapaom oa oomNAezzv .oAm amu oaaaooa .8.» 1a .Azo-a to upcomxgo Lo onah

 

 

 

 

 

 

40

IRREVERSIBLE INHIBITORS

 

CNZYIC‘Oy'N ‘I’ F —T_° -. ENZYME-Or —:—.

.. _@.. .._.,.,_@

p TOLUENE SULFONYL PHENYL METHYL SULFONTL

FLUORIDE FLUORIDE

(T03, (PMS)

POTENTIAL TRANSITION-STATE ANALOGS

C
7" - 7a: 1
ENZYME-04" O —-DENZYME -O -—-—o--o11
. on 3..

11- -cwz-cw,-@

PHENYL ETHANE aonomc ACID @ '9" 3.11, 9115.4, [1117. 3

 

1 PEBA 1 (Peasaneeeasmeeeum
'11- -cwz-cw,-cwt.@ n- -c11a -c1-1,«cn2 -cwz©
91mm. mom aonomc ac1o 19115va sum. aonomc ACID
("on 1 roan

FIGURE 13. Inhibitor Molecules Used in This Study

41

The modified a-CHT crystals were allowed to equili-
brate for a minimum period of one to two weeks before
X-ray examination. Single crystals were mounted in glass
capillaries according to the method of King (75), and the
seal was reinforced by the addition of paraffin wax. The
inclusion of a drop of mother liquor both above and below
the crystal maintained a constant humidity level and
helped prevent drying the crystal. Further positional
stabilization of the crystals mounted in this manner was
obtained by allowing the crystals, fixed to the capillary
wall by a drop of mother liquor, to stand inverted for a
period of several hours. Examination of the crystal
included the observation of the diffraction pattern along
the principal axes, and the determination of the lattice
parameters. Subsequent use of the crystal in intensity
data collection came only after allowing further posi-
tional stabilization over a 24 hour period. To minimize
the effects of unanticipated corrections, the crystals
used for data collection were always mounted with the
monoclinic g-axis coincident with the phi-axis of the

diffractometer.

B. Data Collection
All X-ray work was performed using a Picker Nuclear
FACS—l automated, four-circle diffractometer modified at

MSU to replace the slow-operating filter selection wheel

5%
LL
— I.
a,
.4.
n

91131111.
The co
the 01
131111
“e 11'

,
- 1T.

CIOV

42

with a simpler filter-solenoid system. Additional
modifications, made during the course of this work,
included the repositioning of the balanced filters from
the diffracted beam to the incident beam side of the
crystal. The X-radiation incident on the crystal now
consists of little more than the characteristic wavelength
of the copper Ka line (1.5418 A). As a result, the life-
time of a crystal with respect to X-ray decay has been
extended without loss of intensity of the diffracted beam.
An additional modification involved the placement of a

60 cm, helium-filled tube between the detector and the
crystal (43,76). Increasing the distance between the
detector and the crystal improves the resolving power of
the diffractometer and allows in the present case the
examination of crystals with unit cells of up to 180 A.
The constant flow of helium eliminates air absorption of
the diffracted beam. The net effect of these two
modifications was a slight increase in the intensity of
the diffraction pattern peak-to-noise ratio, a marked
improvement in resolution and an extension of a crystal's
lifetime during X-ray exposure.

X-ray crystallographic examination of the crystals
and determination of their lattice parameters initially
involved inspecting the relative intensities of the
diffraction pattern along the crystal axes (Figure 14).

Preliminary evaluation of the quality of a derivative

43

 

 

 

Mount crystal in glass Y w
capillary and seal, rein-
‘forcing with parafilm

Examine diffraction pattern; , Allow to soak with 1
along crystal axes; deter- small or additional inhibitor
mine lattice parameters; no changes for one week

compare to native standard
crystals

 

 

 

 

 

 

 

 

significant

changes
Observe reproducibility of ,
diffraction pattern to variable
verify equilibrium pattern
’ ' repeat after
stable one week
pattern

 

 

 

 

 

 

 

Examine crystal characteristics, beyond acceptable Select new;
twin size, absorption, background limits crystal

 

 

 

 

 

within
acceptable
limits

 

_di

 

Collect data

1. le st squares orientation matrix

2. 6 (hOl) projection (129 reflec-
tions) before data

3. 3-dimensional data collection 'n72 hours of .
(6 00 reflections, m+ > 0.7) X-ray exposure

4. 6 (hOl) projection (129 reflec-
tions) after_ggta

 

 

 

 

Re-examine diffraction pattern along
principal crystal axes to determine if
any significant changes have occurred
as a result of X-ray exposure

[Transport data to computer center
for further processing

+m = figure-of-merit

FIGURE 14. Crystal Selection and Data Collection Procedures

44

involved comparison of the axial diffraction pattern and
lattice parameters with those of the native enzyme at the
same pH. Changes in the distribution of axial intensities
(Figure 15) were considered more significant than changes
in the lattice parameters alone. The lattice parameters
listed in Table 6 were obtained by the least squares
analysis of the positions of twelve reflections, except
for the native enzyme at pH = 4.6, where only a three
reflection matrix was used, and at pH = 3.6, which
represents the average of several crystals. In those
experiments involving both pH change and inhibitor
binding, attainment of the desired pH conformer was
verified by comparison to the native enzyme before the
inhibitor was added to the soaking solution. It should be
noted that the diffraction pattern of the native enzyme
has proved to be reproducible over a period of years;
the same probably applies to the other pH conformers (43).
Before initiating a full, three-dimensional intensity
data collection (Figure 14), each crystal was further
evaluated with respect to its ability to diffract X-rays
(intensity of diffraction pattern in general, and at
higher diffraction angles), size of the crystal and its
twin, X-radiation absorption characteristics and peak
shape of the reflections. Optimum diffracting ability
would allow operating the high intensity X-ray tube at a

F1

45

1 «I .
‘ TOS _
"F . .
-1- " "v .1».
~ __ .. - 11. 'w.
II‘" _“I..
W '

.—~

 

 

 

 

 

 

 

 

 

l l
I T
Q
E“:
6.3-.
L;‘-‘——
ck— _
L.
z...“
CI.
:‘
I l I
1'
D
N
h
’ C
I
j I I
0 U :

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-L‘ I-
__ PMS _ _
-.- . 1 ii ! I x «D
' a . F F *1 9 II, 1““ - I.“
a ”I . ]_LIII'LLIL1J|I' «b o : )vv -I
1 I I“ III “I” ._ ..
T‘U I ‘1 I “1 g L ‘L g 1 I 1 I I 1
S "a“ l 0 a a “at. 0 0
(1100) (002) (OkO)

FIGURE 15. Distribution of Diffracted Intensities Along
Principal Axes of a—CHT Complexes Used in This
Study

 

 

Ii.

.'7 1:1
I,"‘
VVJ

4'6

. pH7.3 __ .-:oo
o " ~
- - -u
I-” 1mm

 

 

 

 

 

 

 

 

 

 

 

 

I I I I
a
8
n
a
r
u
.539
“a :.-:.—-~°
_ ._ ..-—_—-.-—..4
‘2 - -— .I
_r‘- ' .: .1 _—,
1_-— '
flier-”F. 3
J 1 1 1
1 l l
T .
u
. .
C
1 I I ‘n
a o g

 

 

'
L I l
I l
.
7
-:——L—
J I
m: :
J “I I
3
'E 5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.LL . .1-
.. PEBA ' .1-
. pH36 _ II " _m .
- .3 II I‘ : a 9 Ln 1%”
o 13 1 ‘
.. 8 a I~ III IJ‘I‘I‘ .,. . I Has
3
‘I‘ z" 93:! “ II -- d.
f M: ”I
-- Ufl'yiiii‘" ._ "
1 1”1 1 I 1 181 1 . 1 181 i .

(hOO) (002.) (OkO)

 

 

 

.47

 

 

 

 

 

 

 

 

 

L
P PBBA ‘II‘
‘ .“ . . ‘I'
‘ IAL—J . ‘ J . .. ‘n “- fl
, c 9 . .. 3 ° I -. 10%,,
ULJU J" 'a: '. I? j JII‘I‘ ‘~ ‘ ‘
MIMIWJMI -. |
1. r 2 o y
9 :

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

f Lay? 1 =2 {I}

(hoo)
(009,)
(Oko)

Pv>vdax>vthp~flb Fvfiwlh. mw—‘F hfip~ .JILP-‘ai-:GSFC ‘11-..14-

as Ms 1. 5< ~.

48

mcowumcwsgmpmu xwgums cowpumpwmg mmgzp $o mppammm +

mpmumxgu Fmgm>mm mo mmmgm>< *

 

 

 

Aooevoom.mpm Amvmo.mcp Acvmm.mm Amvmm.~m Amvem.me Am_vm.~ m>wpmz
AOONVOOm.¢_N Amvmm._o_ A~V_m.mm Auvmm.km Amvm_.me Am_ve.m m>wumz
Aooovooo.m_~ A_Vo.mo. A_vm.mm A_Vm.~o Apvm.m¢ +o.¢ m>wpmz
Aooo_voom.m_m onmg._o_ Aavea.mm Ao_vom.~m Auvem.m¢ *m.m m>vpmz
Aoomvoom.m_m Aevmo.NoF Amvmm.mm Auvmo.~m vamm.a¢ M.“ <mm¢
Aoomvoo~.mpm Amvmm._o_ Amvm~.mm A¢vmm.~o Amvhm.m¢ ¢.m <mm¢
Aoomvoom.m_m Acvmm._o_ Anymo.mm Amvm¢.~m Amvem.m¢ o.¢ <mm¢
Aooovoom.m_m Amvmm.Po_ Asvoo.oo Amvmm.uo Aevem.me m.¢ <ma¢
Aoomvooa.¢_m A¢vmm.FoF onmm.mm A¢Vm¢.~o Aevmm.m¢ m.m «mum
Aooovooe.wpm Amvw~._oF onmm.mo Amvm~.~o Aevom.m¢ o.m m2;
Acouvooe.m_w Amvmn._o_ Amvmm.mo onmo.~o Amvmp.m¢ o.m WOF
Ammv we:_o> flavmgmn Amvu Anya Amvm In w>wum>wgma

 

m>wpu>wgwo hxuua ms» we msmpmsagmm muwppmA

m u4m<h

xcapm mpsp cw umms mpmgmxgo

uhu Au
0 (y-

a...“
Tu

C-

49

power setting of 240-480 watts (six to twelve milliamps

at 40 kilovolts). The previously detailed instrumental
modifications allowed most measurements to be collected

at a power setting of 300 watts, further reducing X-ray
exposure. The acceptable twin ratio was arbitrarily set
at 2:l (crystal:twin) with a larger ratio being more
desirable (Table 7). The absorption characteristics were
determined by measuring the intensity of reflections along
the b-axis, aligned at X = 90°, as the phi axis of the
diffractometer was rotated, which was also coincident

with the b-crystallographic axis.

C. Data Processing

 

The methods used here for data reduction and processing
have been described in detail elsewhere (34). A flow chart
showing the path from data collection to the comparison of
results with a Kendrew model of the enzyme is shown in

Figure l6 and is summarized below.

(A) (B)
I ------------- > |F| ------- > |F|
obs-der rel-der abs-der
(diffractometer (relative (absolute
measurements) derivative derivative
structure structure

amplitudes) amplitudes)

where (A) = data reduction:

 

 

= (lobs-der)(decaY)(ab50rPt10n)(twin)(scale)
rel-der

 

IFI
(Lorentz factor)(polarization)

>._.3uw m—zb Ev mu.n~vaOaUUL€£U pcumxfiLU C>uuc>vkwa .LO EOmsLmvlzasu

\ m—»—:<.

 

 

cm as
.. WE . . NE A
copumm mcapmcmasmp owqocpomw

genome mFmom FFmLo>o x

HANK\®N=PmmN-VaxaH m

N

9299:”:

cowuznwgumwu m>vpmc op mcwppwe co_p=nmgpmwu meumc chowmcmewuum Eon; vmcvmpno mcmuwsmcma

mapsmmc cowwumwoca Apocv < m Eocm cmcwmuno mpmum x

_mmbPCP _acwc Fawpwcp

AHV\A .Ava n xdowo e

mwxmum mo mpmcm e n we mews: .mAe+omPvH\mAevH u xgumsE>m< cowpngomn< a

Av ..

:PEAeVH\meAevH u Ezewxms zowpaeomn< «

cwzp Ac c we co suvmemacH
m>wpm= Ao C 0v co apwmcach u owpmm cwzp +

 

 

50

 

om.o w.~- ca._ op.o mo._ hm.F _HN.¢ <mma
No.o o.m- Fo._ mm.o mm.o mm.F P o.m «mam
mm.o o.m- ou.~ m~.o mo._ mm., _"e.m m.~ In <mma
No.F «.m- mo.” m~.o _o.F mF.~ s e.m In <mma
mm.o m.~ mm.o FN.o mm.o oo.~ _"m.~ o.m In «mum
mo._ N._ mm.F mp.o o_._ oe.F F P.~ mza
No._ ~.e mN.F mp.o No._ NN.F _"N.N woe
*x mm fl33m excumo amaszm< . *.xmz +owamm m>wpm>wgwo
goma< agomn< :ch

 

scant Pacovm
-caeGA so:

 

zuzpm mwgh cw mowumwcmpuusmsu Foungu m>Pum>wcmc mo comwgmaaoo

m m4m<p

3531 z... 3.5.5 0.. ,....._u_.....z..c— .23 Lee 33,22 to 3::o 3.933.»: a: «3.: .332 2.25
p.....—,,..q.u—.b..u._. L...a—:3:. U.....J:3.. 3...: -Ufi< .:O..s fist... ....:C.,-.u: .71).;ZJ 23591.53125 1L~2TP~§

Fun—fl cq- —.~ 1.....h < w“ A a F...‘ ahnu_ G .uahuhehfflrh Pvrh‘ Jvudpwlhhyrha— A-Vau m&\ \ .w\~a\.\\-.v N If.'V—\ \ -=~o§y.\

 

 

 

 

51

 

 

cowuumw

m._m_ op mmvamcmacH do coppuzvmmuumcwmmmuocm mung

w

.mmp mmszm

Ago, weaned
m.u op m._u
yo cowmgw>cou

 

 

Ndumu .coauom wmem «cow»

uncomnm .mucmpmn mo Jump .owpmc cwzu
.cowumNPmeoq .Nucmgog com mcwpowggou

.pmm cyan Pacowmcwepvum mozumm

 

 

-oca mucmgme

-cwu A_ogv
13.8.. AL

< o mumcmcmu

 

mpmu m m com co?»
m

 

-umccoo omv :mopno

mm

 

comuumnocn

xpwmcwo cocpumpm
33630 :05 Illul.
=memm= < m mumcmcmw

 

 

 

 

 

 

 

cowuumn

1 -

cowpqcomno .mucmpma
mo xump .owpmc :wzp .covp
-mNPcmFoa .Nucwcog com pumcgooun

 

”but APOEV =zmpca= m m eczema

 

 

-oca mucmcmm
-cau APO;V
=mco$ma=

< m muogmcmw

 

cowuomnoca mg» seem
qungu m>wpm>wgmu com
accumcou mFmUm :Fopao

:owuumwocn

a.

 

 

apwmcmu coguumpm
umimmno :9: I'll.
=mgommn= < o wumgmcwm

 

 

 

 

 

 

L) ‘

 

cowuapomms

m.~ um coppumgcou

Acumv gmwpa pmm op mmvupmcmucw

copuumpymg

gouvcos vogucaa mm:

 

covuauomnm .mocmpmn
we xomF .owaog cwzu .cowa
-m~wgmpon .Nucmco4 cow pumccoo--

 

moan Azoev =azocmn= m m museum

 

 

Fumo<m scum

 

psapao on mooumn seem
pmpprcomuumpemg mgungou

 

i)

 

 

wmwmmmuoga come

one cow caoumm op muou
HHum< seem upon muoumn

paaa=o _-mu<a seem
.mucmpma mo xumfi .o_umg cpzu
.mm>L=u cowuacomnm «gunmen

 

 

 

 

 

~\

~§=LU£J
J up!» a nonhnuqv

C

.5;
2:6
3— 1 a...

.51.. LET... C;
:3. d U3..L3,v
1:. .:L...¢ .6:

 

:2. 432...... I“?
ll. .1) 9 a C: .s s ‘2. .4....l::.,.u

~ 1‘ t 1L

‘1‘33L ~ £21: $2.2:‘tlx. 22‘s.?“

 

«..L.\:.L:m m.) .4
:3: ~3Q..L.~r:. .

\.i
it

\

~.-\

\

52

m.u op m__u_ mo copmgo>oouuum=pmmooogo opoo

.amp mmame

 

 

paces opaepmn=m\.oppnpgcp

po mopocpogooo opsopo ocomooz

 

ocppppp _mooe coppspppmnam oz.
mpmxpooo Low xoo m.ogo;opm op
xppmcoo cospoopo o>ppoc opco
ooe Lopgzom oucogopmpo ouopo

 

 

 

 

 

 

.Am F u so my one copgzom
o cogommpo opoomuzmcocox
pooopmcospoum opogocow

 

 

 

Am F u Eu m.Qev

coppozpsoxo xgocpeppoco

cow one gopsoom oooogopmpo
-pops pocopmcospoum opococow

 

 

coppanp.pwpo ~_._ m>ppac
op ooFoom apmoop>oco m._m_
op mopmco omogo o>ppoc cmpmm<

 

 

o r;
..3......-.:. .._.. m; 2...;
om Loposocoo
Foggocp m oco coppoocgoo

poopmcoo opoom ocpEgopoo

 

 

 

coppoopcpmpo popooc
o>ppo= opp: ogoosoo

 

Am<\o :pm m> «_e—
opp apzpmpo «_u
Forum; pocopmcospoum oopsgopoo

 

 

(I)

a-‘

.'.'

Q)

53

O

with scaling and 6 A resolution decay corrections
derived from "before" and "after" (hOT) electron
density projections;

(B) = scaling of the radial distribution of [F12;

2 _ 2 _ 2 2 . 2 2 ,
IFIabs-nat - lFIabs-der - k IFIrel-der exp(-2351n e/A )’

(see last two columns, Table 7)

(C)
[FISBEIEEL"" m(lFlabs-der ' lFIabs-nat) exPhO‘nat)
(absolute
derivative
structure
amplitudes)

and (C) = assignment of the native phases to the derivative
structure amplitudes and generation of "best"
difference Fourier coefficients:

Fabs-der = IFIabs-der exp(io‘nat)

"best" phase angle,

Q
I

nat -
A = mHFIabs-der ' IFlabs-nat) exP(mnat)

where A is the "best" difference Fourier coefficient
and m is the figure of merit.
At the time native phase angles are assigned to the

derivative data, the distribution of IIFlabs-der ‘ IFlabs-nat

is also determined. The distributions of the derivative

(a

V6

54

difference coefficients used in this study are listed
in Table 8, and in the histograms of Figure l7. From
Table 8, it can be seen that the differences are

generally observable (IFI = 33 electrons/unit cell)

unobs
and that excessive changes in intensity are minimal,
suggesting good isomorphism.

A series of typical absorption curves from a PEBA
pH 3.6 crystal is shown in Figure l8. A comparison of
the relative absorption maxima and asymmetry of the deri-
vative crystals used in this study is made in columns
2 and 3 of Table 7. Evaluations of the peak shape of
the reflections were based on the appearance of the axial
diffraction pattern and reflection centering character-
istics. Crystals were rejected if peaks appeared too
broad or split. Most omega-spreads of derivative
crystals (0.3-0.4°) were slightly larger than those in
native crystals (0.2-0.3°), but this did not seem to
impair the diffraction pattern in any way (4). Severely
cracked crystals were also not used. As experiments
were performed on tubes containing ten or more crystals,
these evaluations could be repeated until an acceptable
crystal was found.

The intensity data collection procedures (Figure
l4), including the wandering count-six-drop-two omega
step scan, have been extensively discussed by Vandlen

and Tulinsky (34,78). The data collection system used

 

 

 

 

 

mappmppepm :CPuJQprmpa +~L<~ ptmeLq w>+po>pan k0 :OmpLflQEOU

I.

«~ J :,..—< ~

 

.pxop op oommoompo

 

55

 

 

mo nos xppmcoo :ocpoopo mocogowmpo op Locgo ogoocopmt _pocimoo_m_ . LooL....oo_h___ n _m<_ +
. - - - . . . . . . mu o A...
mpo o p o m o m m m mm m pm o p. wepo m.p In o>ppmz
. - - - . . . . . . op_o ape.
mpo o p o m o p p m mp p mN o mm “ammo ¢.m In m>ppmz
mpo.o - - - - N.o m._ N.3_ m._m m.m¢ .mmpm o.e :3 «mm.
cope
. - - - - . . . . . owoo .
mpo o m o p N m mp o _m e p. .mmmm o o In <mma
opo.o - - - _.o o.o p.m N..N m.m~ o.N¢ .wwwm m.p In <mm.
NNo.o _.o m.o m.o w.o N.N m.p N.om N.p~ p.Nm .mmwm e.m :3 <mm.
Npo.o - - - - N.o m._ p.mp m.om o.Nm .Nmpw o.m In <mm.
mepo
mpo.o - _.o ~.o m.o “.0 o.N N.o_ p.0m m.om .wm¢w m2.
Npo.o - - - - ~.o o._ m... m.pm p.Nm .wm¢m mop
Apv ARV ARV Apv ARV Apv ARV Apv flay
Am- avpomeaqvo om. co. 3mm pom ooN «mp pm mm ooeamaos . o>ppa>peao

 

ago; m usovcopm com upoe immm imam upom immp iwm tom :0 uw>gwmno *

 

Appmo p.==\m=oepumpov mcoppzapcpmpo +_.<_

 

moppmppopm coppoopgpmpo +_m4_ popmxgu o>ppo>pgoo mo oompgoosou

m u4m<h

56
we

°°‘ TOS

40<

'Io

 

30-1 315$

 

145$

 

 

 

 

 

001

 

5°‘ 5°“ PMS

201

 

 

 

 

 

won on. on.

o- 34- se- issjzoi- '269.‘ ans-ram- Imoo
a: 57 :34 zoo 267 334 400 499
ofmronVunéi can

FIGURE l7. Histograms of Derivative Crystal IAFI Distribution
Statistics

 

57

. 383303.59...
00v 00. van «3 oo~ on. so an

83.533559.- ..obe .3

.‘H

.0

 

I... loo 3.:
2d (a

 

5N

 

 

3.. 2,32. .
”Kin (mum

.a -

 

 

 

 

 

 

Das—

 

INN"

 

 

 

 

 

 

2.-. 02‘... v

1610 (mum . H

 

 

 

 

5p

 

 

 

 

92.339.

 

 

 

58

3" PPBA

4249.

40-

“I.

 

30‘ no;

20d

 

183$

10.-

 

 

 

 

 

 

5%,... PBBA

 

30‘ 31.31.

 

 

 

 

 

1.8-1.1 “2'“

 

o- 34- aa- 135'20i- '550-‘335-‘401-3500
:3 oz 134 zoo 267 334 400 499

matte-my wit «u

59

. m-pm-po-o pmpmazo
.o.m za <mua--mo>.=o coppuoetoo coppaLOmn< .op mmaop.

 

 

 

 

O.N

o3 HXmEuzu?

60

in this study, which was developed at MSU, permitted the
monitoring of the crystal alignment during intensity data
collection and the redetermination of the crystal
orientation matrix if the crystal apparently moved. In
addition, the programs measured a reduced reflection data
set which only considered reflections which had a figure
of merit greater than 0.7 in the native data (79). The
hardware and software developments described here and
elsewhere made it possible to obtain a complete set of
observations from a single crystal, not the five or more
crystals required in previous studies. The reduced
reflection set included approximately 6300 reflections.
The total number of independent reflections is approxi-
mately 10,500 at 2.8 A resolution. The electron density
maps calculated with the reduced set of coefficients,
whose phase angles were more accurately known, compared
well with maps calculated from the native data set
containing all reflections with figure of merit greater
than 0.3 (approximately 8500 reflections). Data
collections were completed within a three to four day
period, averaging 70 to 80 hours of total X-ray exposure
to the crystal. In all the derivatives studied, the
decay as determined by fall-off in intensity of monitored
reflections was 35% or less (Table 7).

For each derivative crystal, the intensity of the

129 reflections of the centrosymmetric (h0l) projection

61
at 6 A resolution was measured both before and after the
three-dimensional intensity data collection. The "before"
(th) projected electron density permitted scaling to the
parent native crystal data, normalizing the crystal size
and beam intensity. The "after" (h0l) projected density
allowed the determination of intensity fall-off due to
X-ray damage in the 6 A resolution range. These projec-
tions also permitted the evaluation of possible changes
in inhibitor binding which might have accompanied the
extended X-ray exposure. The derivative crystals used
in this study did not show the latter changes.

A method of evaluating the error of the difference
electron density has been developed by Henderson and
Moffat (80), the mean-square-error of the difference
density being given by

<Ap2> = (2/v2)zzz[A|F|§_n(hk1) + A|F|§_n(hkl )(l-m2(hkl))

+ 52mm
Where, AIFld_n = lFlder-nat = llFlder ' |F|natI
(52 = 02(AIFld-n)
m = figure of merit

V unit cell volume

The first term, A|F|§_n, represents the error caused by

assuming that the phase angles of the parent native

62

crystals are unchanged in the derivative crystals.
Although the structural changes observed in this study
are generally small, Luzzati (8l) has shown that the
substitution could be 50% higher due to this type of
phase angle difference. The second term, (A|F|§_n)
(l-m2(hkl)), corrects for the errors made in the original
determination of the native phase angles by using the
figure of merit; the latter is related to the error of

2, represents

the phase angle. The third term, 6
experimental observational errors in the difference
coefficient. The root-mean-square (rms) errors of the
derivative difference electron density maps are included

in Table 8 (the 62

term was not included in the calcula-
tion). This error estimation of the maps was used to
establish levels of significance in the maps. Although
the method has become a standard means of estimating the
error in difference electron density maps, it should be
noted that the calculation has been shown to be inaccurate
by a factor of 2 (82). In the present work, a more
conservative approach was employed based on an estimate

of background noise in the Fourier maps. Earlier pH
studies had indicated that an appropriate random background
noise level in the difference map was about i 0.05 eA'3

(l2), and confidence was placed in those peaks which

occurred at about three times this value. The same

m
3..
r1)
(‘5
1+
"1

featur
even 3

 

 

 

 

63

procedure was used in the present work, with adjustments
being made on the basis of the random background noise

for each derivative crystal studied. This screening
process, along with the ability to correlate the difference
electron density peaks with intelligent structural

features of the native structure, will be shown to give
even more validity to the interpretation of the difference

maps. The significance levels established in this manner

were usually about l0 x 0(Ap).

III. METHODS OF ANALYSIS

A. Emphasis of Analysis

 

The two classes of inhibitors used in this study
(sulphonyl halides and boronic acids) and the selection
of the compounds within these classes permitted a
detailed comparison of the a-CHTzinhibitor complexes
with respect to the following important points:

l. The orientation of each inhibitor in the
independent molecules of the a-CHT dimer and
the interactions of the inhibitor in the
active site upon binding.

2. The perturbation of the native conformation
which occurs upon binding in regions removed
from the active site, and its relationship to
the binding in the active site.

3. The effect of varying the side-chain of the
inhibitor upon:

a. The binding orientation and inter-
actions within the specificity site,

b. The perturbation of the native con-
formation removed from the active
site region.

4. The structural and functional variability among
the different pH conformers observed in the

crystal upon inhibitor binding.

64

use 0‘
inabi'
atomic
here.
tions

CESCr

 

 

 

65

5. The evaluation of the suitability of phenyl
alkyl boronic acids as potential transition-
state analogs for a-CHT catalyzed hydrolysis.

The present analysis has been based primarily on the
use of the difference Fourier method. However, the
inability to examine the three-dimensional structure at
atomic resolution limits the correlative studies described
here. On the other hand, several refinements and innova-
tions to present techniques have been made and will be
described below.

The areas of major concern in this study have been
the comparison of the inhibitor binding interactions and
orientations within the active site and their external
structural perturbations which might indicate a form of
cooperativity upon active site binding. A detailed
comparison of the inhibitor molecule binding orientations
has utilized the examination of composite sections of
the difference electron density of the substitution and
the fit of a model representing the substitution to
electron density centroid positions in planes of the
difference Fourier electron density map. The examination
of the changes which occur in the native a-CHT conforma-
tion upon binding has been carried out by the use of a
difference diagonal plot method which was developed during
the course of this work and which is a variation of the

diagonal plot originally proposed by Phillips (83).

66

B. Examination of the Substitution in the Active Site

Initial examination of inhibitor substitution involved
the superposition of the "best" difference electron
density map onto the native electron density map in the
MSU Richard's box (84). The orientation of the MSU-
Kendrew model of a-CHT presented problems in placement of
a model of the inhibitor within the electron density. The
Fourier map, calculated along the a-direction and stacked
along the gf-direction, revealed a more or less "end-on"
view of each of the inhibitors used in this work (Figure
19). This made a direct comparison of the various inhi-
bitor substitutions difficult and the same applied to the
exact placement of the inhibitor model in the active site.
Therefore, the difference electron density was calculated
along both the g- and g-directions, and composite projec-
tions were constructed as shown in Figure 20. These
composite projections permitted a simple comparison of
the substitution density in each of the independent
molecules of the dimer and, also, among the various
derivatives (Figure 2l). The lowest contour level in
the projections is drawn at about three times the noise
level previously discussed (O.l5 eA-3), and only these
contours are drawn in each section.

A detailed comparison of the various substitutions
Was possible after the Kendrew model of the inhibitor was

fit to the difference density and the coordinates for the

67

 

Z —>
-Y TWO-FOLD AXIS

/V \I \
I I I
1’ 1 [_‘j,/’ 1’ I
z' I I
I ' 7 I

\ n 1

\ II I

\ . l‘ /

\‘— l/ \ Ill
SUBSTITUTION SITE

 

 

 

 

@/]\\

@

 

 

 

 

 

FIGURE l9. Schematic of y-z Projections of the Difference
Electron Density in the Active Site Region for
TOS a-CHT

68

 

 

 

 

FIGURE 20. Comparison of Composite Projections of the TOS
*Difference Electron Density in the Actixe‘Sites
of the a-CHT Dimer; Contoured at O.l5 e '3,
Positions of TOS l and T05 l' are Arbitrary in z

 

XI

 

‘ h
‘ H ====
' \

69

 

 

 

 

FIGURE 2la. Comparison of Composite Projections of the
Substitution Electron Density of TOS and PMS
in the Active Site of Molecule 1; Contoured
at O.l5 eA'3 . -

 

 

 

.YI

III
cl:

 

70

 

 

 

 

 

FIGURE 21b. Comparison of Composite Projections of the
Substitution Electron Density of TOS and PMS
"in the Ac ige Site of Molecule l'; Contoured
at O.l5 '

7l

fitted model were obtained. The positioning of the model

in the difference electron density used a method of

plotting the projections of the calculated difference

electron density centroids. The method involved the

following scheme:

1.

For each Fourier section calculated along the
g-direction, line sections are constructed
through the peak density position in the

region of the substitution in both the y- and
z-directions.

The centroid peak position (maximum of the above
plot) for each Fourier section is determined to
an accuracy of i 0.25 Fourier grid units (0.l6 A
in y- and 0.l3 A in z-directions).

These centroid positions are plotted on a scale
of 2 cm = l A in both the x—z and x-y projec-
tions. For the centroid position in each
section, a boundary occurring at i 3 x 0(Ap)

is also plotted, establishing a new weighted
composite density projection, which is weighted
by the difference electron density.

A Kendrew model of the inhibitor is "fitted" to
the three projections and the atom coordinates

are measured.

72

5. Bond distances and angles are calculated and
the atomic positions are then adjusted to
correct any major disparities from accepted
geometries (85). The proper orientation is
maintained within the difference electron
density centroid projections.

The inhibitor bond distances determined in this manner
approach accepted values to within : 0.05 A (A-l, A-2).
The fit of tosyl to the centroid density of p(TDS-a-CHT)
- p(a-CHT) is shown in Figure 22.

The method approximates results obtained from a
computer analysis of substitution density which uses
visual interactive graphics display techniques (86). A
goal of both this method and that of the visual display
system is to manipulate inhibitor orientations within
the difference electron density without disturbing the
native model and to obtain an initial set of atomic
coordinates for the substitution. Although not quite as
precise as the automated techniques, the method is
easily applied and has allowed the detailed comparison
of the orientation of the substitutions for the inhibitors
used in this work.

A detailed analysis of the inhibitor orientation in
the active site was accomplished by the use of the native

enzyme coordinates measured using the model of a-CHT at

r»
‘0

73

<—-X

 

 

 

 

é"“X

 

 

 

 

FIGURE 22. Model of TDS as Fitted to weighted Centroid
Projections of the Difference Electron Density

genag
Und-

be re

74

MSU (79), and refined by Diamond's program, "Model Build"
(87). A comparison of the calculated distances between
the inhibitor molecules and the amino acid residues
forming the specificity pocket and the catalytic site of

the native configuration is made in the Results section.

C. Examination of Conformational Adaptation Accompanying

Inhibitor Binding

 

Kirkwood (88) and Lumry (89), among others, have
suggested that enzymes might exhibit a cooperative struc-
tural relationship upon the binding of a substrate or
inhibitor in the active site of the molecule. The igtgr-
molecular cooperativity observed in the hemoglobin
system (sigmoidal kinetics) has raised the question of
the generality of possible igtrgmolecular cooperativity
and molecular recognition. An examination of the confor-
mational changes which might occur upon the interaction
of a-CHT with individual members of two classes of
inhibitors could provide insight concerning such questions
as why large biopolymers, such as the serine proteases,
are required for the maintenance of a small active site
region. In this respect, particular emphasis is placed
upon the domain structure of proteins, as first uncovered
for the nucleotide binding domains of lactate dehydro-
genase (90) and observed later with other nucleotide
binding enzymes. These bio-architectural features can

be revealed by a diagonal distance plot (l).

75

The diagonal distance plot involves the calculation
and representation of the distances separating the alpha-
carbons (Ca) in a protein (polypeptide) molecule. The
distance map, shown in Figure 23, where rij
the distance between residues i and j along the a-CHT

represents

chain, was generated by Nishikawa _t_gl. (44), and was
contoured at intervals of 15 A. This original map used
the MRC coordinates of a-CHT and is virtually the same
as the map generated using the MSU coordinates. This is
in agreement with the comparison of the MSU and MRC
results which were within the 15 A resolution limit of
the distance map. A striking feature of the a-CHT
distance plot is the appearance of two similar structural
domains, A and 8, beginning at residues 1 and 123
(Figure 23) (44). The presence of separate but similar
structural features is not restricted to a-CHT and has
been observed with elastase (91) and lactate dehydro-
genase (92). This structural repetition in the construc-
tion of a protein molecule has been attributed to gene
duplication with evolutionary differences occurring in
the related regions (49).

The potential cooperativity which might occur
between domain regions as a result of binding of an

inhibitor or substrate, chemical modification or change

in the pH or salt content of the mother liquor, has been

dddddddd

molecu|e1

FIGURE 23. Distance Diagonal Plot, r

c‘ l‘

V
‘5”
.Ja
J

76

molecule 1'

'I

150’ 200’

._ O$;::;E1::

 

 

of a-CHT, Contoured

1
at 15 A only (45) and Exhigiting Dimer

Sui

di-

di-

77

examined in this work. A schematic representation of
the domains of the distance diagonal plot of Figure 23
is shown in Figure 24. Both the individual domain
regions (A and B) and the inter-domain regions (A-B) of
a-CHT are depicted. It is also possible to examine the
inter-domain interactions in the dimer of a-CHT by
construction of the cross-distance diagonal plot shown
in Figure 25a. This method plots rij
alpha carbons (Ca) from molecules 1 and 1', respectively.

where i and j are

The packing along the z-direction of the dimers and the
domains of a-CHT is shown in Figure 25b. Although the
inter-domain regions have been observed in the distance
diagonal plots of other enzymes (92,93), no detailed
description of them has been reported.

In examining the structural changes accompanying
substitution, some innovations to the original distance
diagonal plot have been made. One of these readily
displays the variability of conformational changes
between the two independent molecules of the a-CHT dimer
(42). The method makes use of the inherent symmetry in

the diagonal plot (e.g., ). Since the distance

"ij "ji
diagonal plot exhibits structural features arising from
the interaction of individual Cu atoms, the peaks of a
difference Fourier map, which represent changes in the

structure of the native enzyme, can be examined for their

78

Residue# Molecule 1

 

 

 

:5
Residue#
°Af
l¥-B’
A1
0)
'3
8
3 a
z:
AA"B
8'
Ir

 

 

 

 

 

FIGURE 24. Schematic Representation of a-CHT Domains

79

 

 

 

Molecule V
A'—A
. A-B'
A-A’

Q
'5
8
'3 B-I
z:

B-A'

B - 3'
u

 

 

 

 

 

 

FIGURE 25.

 

 

 

 

 

Molecule 1‘ Molecule 1

Intermolecular Domain Contacts in a-CHT
Crystalline Dimer--a) Schematic of Cross-Distance

Diagonal Plot, and b) Linear Packing of Dimers
and Domains of a-CHT

a»

tn

or

80

proximity to these atoms. A substitution can also be
examined in a similar manner. We have termed these as
difference distance plots and difference substitution
plots, respectively. The definition of the quantity
plotted, Rij’ is presented in Figure 26. In the
difference distance plot, the Ca atoms, i and j, and
the difference Fourier peaks are not simply considered
as points. Instead, a prolate ellipsoid surface is
generated as an ellipsoid of revolution about the line
between i and j. By allowing the sum of the distances
from the peak to i and to j to equal the separation,

O

rij’ increased by 2 A as appears below
ij rij + 2 A i eri + erI

R

the equation for the resulting ellipsoid of revolution
is

xZ/(R-j/z)2 + yz/HR-j/Z)2 - (r-j/2)2)

+ zZ/HR-sz)2 - (R-j/2)2) = 1.
07‘

x2/1r-j/4 + r.. + 1) + yZ/(rij + 1) + 22mij + 1) g 1.

Thus, if a peak occurs within this surface, its value,

R is plotted. Moreover, the Rij of the two independent

ij’
molecules of a-CHT dimer are plotted on either side of the
diagonal of the distance plot. Thus, deviations from

diagonal symmetry represents asymmetrical behavior. An

81

 

DISTANCE DIAGONAL PLOT DIFFERENCE DISTANCE DIAGONAL
COORDINATES‘ PLOT COORDINATES

FIGURE 26. Definitions for Diagonal Distance Plot and for
Difference Distance Plots '

for

moi
5.09

Val"

IT!

82

example of the difference distance plot calculated from
the difference Fourier map obtained from changing the

pH of a-CHT from pH 3.6 to pH 7.3 in the presence of PEBA
is shown in Figure 27a. The substitution distance plot
for PEBA pH 7.3 using the peaks of the substitution
itself are shown in Figure 27b. Appendix C contains a
list of all difference Fourier peaks used in calculating
the difference distance plots of this work along with
their peak heights. The substitution peaks are listed in

Table A-1.

D. Inhibitor Orientation in the Independent Molecules

 

The detailed comparison of the orientation of each
inhibitor molecule used in this study in the independent
molecules of a-CHT dimer probes both the source of the
specificity of a-CHT and the effects of the dimeric

variability in structure.

E. Inhibitor Orientation Definitions

 

In comparing the inhibitor orientations, three para-
meters have been defined and examined, an x-z and x-y
inclination angle and an aromatic ring rotation angle.

The inclination angles are defined as the angles formed
between a line from the C-1 to C-4 positions of the phenyl
ring and the x-z or x-y planes, which are parallel to the

non-crystallographic two-fold axis (Figure 28). Inhibitors

83

 

 

 

 

 

 

1 l 1 7i I T I T
n W
1L ' db; {‘1 U
o >r" n I 1 I“) q» ("n l {I} R“ (‘1‘; V? .V,‘ (I I» f“
L.(P I?” H Lu . -w ‘1“‘ ~’
I) . ‘ L1
I) 1) 9. a
o E 5.; I i. ["013 $.11
ix'fl‘f’-’ 7' ' ' I. 1‘
"‘ fi .h'. Ink ,1 - f“.
n “go/ht Lg:- (, ,1 gm 3
er \ 11 11.1 1 .~
‘— .- V ' x-" ,. L J (A) C: - r
50 I~PJ [‘ H [\J .
5I : a n : «T! f“. "‘
351- 1‘ U1 31
— .r-L n
. .14. O L
' 2;} a. ; . . {.9 y 1. a. --
__ " c ; ’2'1'. :32 \ (is: [4.11.77 :31“.
c :5 "EL '_---1‘t].1
#9
. L T' ,‘Urfi, ~ 7"
f6 r f‘fi/Jl '17:: O S‘- T...’
- 3" L‘ 3 .1 r1. :3 3 A
100 {U} o ' ' .J D 6 <1.
. \1 ,. " V .. _.--
L. $1.. a J D - “‘v’
33 t.
L - 1;
°F
1.1 ' ‘ 0
f 1 L.) 1 .3 .- In 0
U :1
'50” “so 11%) 0
J3". 2. .LJ a
C ' Liz
I]
{i -. .-
0 a u 3 '1 '
° “I? U I} I) & c311
200—11> h ° ' 5'
I’ 11 v

 

 

 

 

FIGURE 276.

J

Difference Distance Plot of Conformational

Changes on Increasing the Native, pH = 3.6

a-CHT Crystal to pH = 7.3 in the Presence of

PEBA

 

34

 

 

J
5.0 100 15:0. zoo
' l I 1
I I
3;. .3 .-
50' :. .' q
I! .
IOO' ' '
I
150‘, ’ ' 11 "i“.
“ £3 .3 m o‘ “a": 41-w-
200" .
b d! a :35 a .
‘ ' ' .3 . ' 83%? .19.

 

 

 

 

FIGURE 27b. Substitution Distance Plot of Substitution of
PEBA Observed on Increasing the Native, pH =
3.6 a-CHT Crystal to pH = 7.3

85

 

molecuh 1 molecule f

 

 

 

x-y INCLINATION ANGLE

 

 

x-z INCLINATION ANGLE

 

 

CF! CP3 c9)
' Cks a
on ) ( c4
Cé-\ 2
moteculef meleculo i

AROMATIC ROTATION ANGLE

FIGURE 28. Definition of Orientation Angles for Comparison
of Inhibitor Orientation

86

which are oriented identically in the specificity sites
of the two independent molecules would have inclination
angles which are identical in magnitude but opposite in
sign. Values of 0° in the inclination angles are
observed when the Cl-C4 line is parallel to the two-fold
axis in the x-z or x-y plane.

The aromatic ring rotation angle defines the
orientation of the phenyl ring of the inhibitor by rota-
tion about the line joining C1 to C4 (also, C1 to CA or
C1 or S). It is taken to be 0° when the ring lies in the
x-z plane if the x-y inclination angle is set equal to 0°.
The positive rotation direction is taken to be a clockwise
rotation around the Cl-CA or Cl-S bond when viewed from
the Cl side. The rotation angle for two identically
oriented inhibitors would also be identical in magnitude

but opposite in sign.

In

IV. RESULTS AND DISCUSSION

A. Irreversible Inhibitors

l. Substitution characteristics of TOS and PMS

Initial examination of the diffraction pattern along
the principal axes of the TOS and PMS derivatives revealed
substantial but similar intensity changes when compared to
native a-CHT (Figure 23). Moreover, the changes are
similar in magnitude to those observed with isomorphous
heavy atom derivatives. The lattice parameters of the
unit cells show little difference in dimensions beyond the

calculated error (Table 9).

TABLE 9
Comparison of Cell Parameters of

Irreversibly Inhibited a-CHT

 

a(A) b(A) C(R) e(°) Volume (R3)

 

TOS pH 3.6 49.15(5) 67.02(6) 65.88(9) lOl.78(3) 212,400(700)
PMS pH 3.6 49.36(4) 67.25(5) 65.98(6) lOl.78(6) 214,400(600)
NAT pH 3.6 49.24(7) 67.20(IO) 65.94(9) 101.79(6) 213,600(IOOO)

 

The cell volume of TOS appears to shrink slightly while
that of PMS expands slightly; however, both differences
approach the limit of error in the native cell volume.
The larger standard deviation indicated for the native

structure results from the combination of the seven

87

SE

111'

th

V

der

OCE

by

5y:

I.
h-

511'

88

separate native crystals used in the data collection (34)
while the errors for the derivative structures represent
the parameters for the one single crystal used for three-
dimensional data collection.

A preliminary examination of the difference electron
density maps showed that the substitution approaches 100%
occupancy, as expected from solution studies of inhibition
by aromatic sulfonyl fluorides, with excellent two-fold
symmetry with respect to occupancy. The active site

3

substitution of TOS in molecule 1 reaches 0.65 eA' and

3

0.62 eA’ in molecule 1'; while in PMS, the corresponding

3 and 0.65 eA'3

occupancies are 0.62 eA' , respectively.
Further study of the substitution density of TOS and PMS
showed differences in orientation reminiscent of the
variability and asymmetry in the native structure (42).
In addition to the substitution density, gradient peaks
were observed, indicative of small rearrangements of the
native enzyme structure to accommodate the inhibitors.
Examination of peaks greater than 3 x C(Ap) in the
difference maps showed more extensive changes in the
structure with the PMS derivative than with the TOS
derivative.

Construction of x-z and x-y composite difference
electron density projections of the substitution, as

described in the Methods section, revealed significant

and consistent differences in the inhibitor binding

89

orientation both between the two independent molecules
of the a-CHT dimer, and between the TOS and PMS deriva-
tives (Figure 29). A detailed comparison of the centroid
distribution of the inhibitor density is summarized in
Table 10. The comparison within the dimer was accomplished
by the rotation of the electron density of the substitution
in molecule 1' about the two-fold axis followed by the
calculation of the differences in centroid positions in
the specificity region (x 5 49/76) and in the catalytic
region (x 3 50/76). These regions were selected on the
basis of inhibitor interactions as discussed later. The
estimated error in these comparisons is about 0.4 A; thus,
only the differences in the y-direction are significant.
The systematic difference across the two-fold axis (the
rotated density of molecule 1' is displaced in the (-)y-
direction) is compatible with the variability in structure
observed in the dimer and will be discussed more fully
below.

Kendrew models of the inhibitor molecules were
fitted to the composite bounded difference electron
density and atomic coordinates were measured as discussed
in the Methods section. The atomic coordinates determined
in this manner for all the inhibitors used in this study
are listed in Table A-1 and these were used in all dis-

tance calculations and comparisons. The bond lengths of

1g Z9. .. Rx...

9O

. coppopcomogoom upon
1oz» opoepxogoo< sppz m- m_.o po omeoopcou .mpx<-~.a
ozoo oozop> mcoppnpscm op pmco>oggp coozpom appppoopco>

AN. .mze ANV mza

 

.¢N masonu

 

IIJ‘ fl" u-‘Wt‘fuc ll‘".-.¢~.‘vllr&-U “c Inf-(T‘Thco TYC‘Afi-XGL ext, 1*: :CUmroLONC—tcu

C p mu 1. max;-

91

 

.A~<v

.o ... ....

pm.o- a. poxuoa

 

 

 

 

 

 

 

 

 

 

 

 

 

oppxpopoo ogp op .p oco F mopooopoe :p mo» po moopppmoo o oo ogp com: on ooppop>oo one "oposoxm
Ao~\om.m xv .mop .mop mop mop
pmxoog < mo.o u we. < a..o- n ma. .mza < m_.o- u we < o~.o- n ma. m2.
oppapopoo o o o o
Aoe\me.w_xv
poxooa < mp.o n ma < mo.o- n ma .WZQ e m_.o- n ma < oo.o n ma m2.
appupwpuoom o o o o
Loewe p:o-..a=p_pepmxgo
9.3 B5 mwpzumpoz ucwucwnmug ms» *0 zoom 5. cowuzpwpmnam 259.525 mpnwmgm>mggm mo :oflLmanu .n
.1 .m2a .m2a .mop .mop
Ae~\om A x.
poxuoa < _e.o n ma < o~.o- n ma mze < mm.o n ma < _m.o- n ma mop
oppxpopoo o o o o
Aoe\me.w xv
poxuoa < 3N.o n ma < mN._- n ma m:. < 4N.o n ma < m~._- n ma mop
xppopppooom o o o o

 

 

 

 

 

 

 

 

F opooo—oz sogm ooppop>oo ogp mcpmoco><
xo oozoppom .._ opooopoz op coppopppmozm ogp po mpx< opomioz» pooo< coppopom mo compsooeou .o

.p opzoopos mopocoo A.v
pose we. mop. .. use _ mapsumpoz co moppm a>ppu<
ogp op appmcoo ooppopppmoom mo mcopppmoo opoepcou ozp po compgooEoo
op m4m<h

tt

Ta

(Tl

Ill

92

the individual inhibitor molecules are also reported in

Table A-2.

2. Orientation of inhibitor molecule and inter-

 

action with a-CHT

 

The orientation angles, as defined in the Methods
section, for TOS and PMS and the other inhibitors used
in this study are presented in Table ll. It is apparent
from this table that the two-fold related pairs, TOS-TOS'
and PMS-PMS', exhibit good two-fold symmetry with respect
to the x-y and x-z inclination angles, but are asymmetric
with respect to the aromatic rotation angle. The orienta-
tions of TOS and PMS in each of the independent a-CHT
molecules differ in both their x-y inclinations and
aromatic rotation angles. The differences between the
TOS and PMS derivatives can be related to the specificity
requirements of a-CHT, while the two-fold asymmetry in
both derivatives can be related to the structural varia-
bility of the dimer. A more detailed comparison of the
individual inhibitor molecular orientations and inter-
actions with the enzyme, and a full comparison of the TOS
derivative results with those reported by the MRC (49),
are given below.

The effect of the variability of a-CHT structure on
inhibitor orientation can be seen by examining the inter-

actions which occur between a-CHT and the inhibitors. A

93

TABLE ll
Orientation Angles Determined for the

Inhibitor-Enzyme Structures

 

 

Inclination Inclination Aromatic

Angle (x-y) Angle (x-z) Rotation
Derivative (:_5°) (:_5°) Angle (:_l0°)
TOS MSU +44 -12 +10
TOS' MSU -42 + 9 -30
TOS MRC] +40 -12 +14
PMS +65 -12 + 0
PMS' -65 + 8 -45
PEBA pH 3.6 +48 -12 +60
PEBA' pH 3.6 -39 + 8 -75
PEBA pH 5.4 +48 + 5 +45
PEBA' pH 5.4 -39 + 8 -45
PEBA pH 7.3 +45 - 4 +30
PEBA' pH 7.3 -32 + 8 -20
PPBA +4] + 5 +70
PPBA' -44 - 2 ?
PBBA +4] + 5 +70
PBBA' -42 - 2 ?

 

l

Calculated from coordinates reported by Birktoft and Blow,

J. Mol. Biol. §§, 187 (1972).

 

94

summary of proteinzinhibitor contacts which are 3.5 A or
less (van der Naals) of all the inhibitors used in this
study is given in Table 12. The individual protein:
inhibitor contacts are listed in Table B-l. This table
was compiled using the inhibitor coordinates and a set
of a-CHT coordinates which resulted from one cycle of
Diamond's model-build procedure. Only the coordinates of
molecule l were used and the analysis was based on the
inhibitor interactions in molecule l and the known
variability in both inhibitor orientation and enzyme
conformation (42) as determined from the difference
electron density maps.

From Table 12, it is apparent that the majority of
van der Waals contacts between the aromatic group of the
inhibitor and the protein molecule involve residues
TRP 2l5-GLY 2l6. The structure of the a-CHT dimer in this
region of the specificity pocket (i.e., residues TRP 215-
SER 2l8) shows significant deviations from local two-fold
symmetry, due either to repulsive or attractive inter-
actions (42). The orientations of TOS and PMS in x-z
projection are shown in Figure 30, which also details
some interactions in the specificity pocket. The shortest
contacts occur with GLY 2l6 Ca, thus establishing its im-
portance in defining a boundary to the specificity

pocket. This is in agreement with observations in

.Amm\om M.xv muvm uwpapmpmu cw mcowpomgwucw nausea :mmogu»: Puppcmaoa .m
.Am5\om.m xv muwm uwaxpmumo cw mcowuomgmucw amazon :mmoguxgucoz .N
.Amnmpv mmp «mm ._owm .Poz .n .3opm new pmoaxgwm an umpcoamg mmpmcwugoou soc» umpmpaupmu .F

 

95

 

 

 

Nm PN mp N N 0 mp - « o.m
No om mN o, o_ o_ 0N - q m.m <mma
mm m_ NF m N F NF - q o.m
we am on N_ m_ m cm - n m.m <maa
oN a_ N_ N m N N_ - q o.m
we MN mN a N m mN - n m.m m.N :a «mum
0N N_ N a o N m - q o.m
Nm 0N oN __ op m NN - « m.m ¢.m Ia <mNa
m. N N e m P N - q o.m
NN 4F m_ m m 4 my - q m.m m.m In <mma
oN NF 4 N N o 4 - u o.m
Fe NN 4? _F m_ F a_ - a m.m mza
e. N, N m m - N o a o.m
m4 NN _N o_ N_ - o_ m « m.m um: mop
4N m, N m N - a q q o.m
we mN NP P_ e_ - N, m a m.m 2m: mop
_mpo» paxuoa omxuoa m N cwmgu macaw _sgpmz m>wpm>waao
uwpx_~pmu spvuwcwomam _»¥_< uwpmsog< -F»=_op
muompcoo muumucou
Peach Fnyoh cowmmm panacea opaumpoz gopwnwch

 

mopsumpoz sopwnwch mzu ucm hzuua m>wumz cmmzumm A< m.m wq muumucou mpmmz Lou cm> mo xngE=m
0

amp m4m<k

96

A.u.p=ouv .nN_ afinm»

 

INN

P

I l'—

I I
Ir— Ir—l—r-
I I
Il-‘l—‘I
Il—l—'

I I

I
II—MNNVN

F

NNQ’MFLDKOLD

INN Ir—
FI—F-I—r- I
I—d'r—I II

I I II

I I I I

r— I“) I

mm LOMKONQ'QO‘O
N

IOLDIDN

r—LO

<Nma

«mag

m.N Ia «mum
e.m In <mma
m.m Ia <NNN
mza

_omz mop
am: mop

<mma

<maa

m.N Ia (mug
¢.m Ia «mum
o.m In <Nma
mza

_umz mo»
an: mo»

Fuzz mob
3m: woe

cwmco
szp<

azogu
upmeoL<

Pseums
-F»=_0h

 

oNN

Rpm

e_N mFN e_N m_N mm, «a, mm_ Nap
mauwmam uwu< o=_s<

_m_

omp

mm

Ne

m>wum>wgmo

 

A< m.m.wv N:o-5 m>wpmz co mm=u_mmx uwcwumam cow: mpuapcoo canvasscH co Nameszm

amp m4m<h

 

«.U.HCOU»

 

IGNI‘

t N Q t;

.328 M xv 3.; 0.5338 5 23323:? 32.8 53.62 33:38 .m
.Stom M xv 3.5 “9.5333 5 20308.35 32.2 swaggeucoz .N
.333 BF .wm .35 .3: .n .265 25 99852.5 .3 «.3289. 335208 so: “.32:qu .P

 

 

97

- - - - - <NNN
- <maa
m.N 2a <Nma
e.m In <mma
o.m Ia <mNN m
- mZN
FUN: mo»
- - _ - 3m: mop

IF-r—r—N
I—I—MF-N NI—
Ir—F- I I

r— I l
I I I
I Ir—r-F-F-
I I

F

- - - m <Nma
_ - - - <maa
N.N In <Nma
- - - - - ¢.m Ia <NNN
m.m Ia <mwa N
- - - mza
Pom: mop
- - - am: mop

F-I—
II—F-LON NNN

IMF-NCO
F

I
I
I
,—
I

I
I
I
I
I
I
sommmxotxoo hemwnumm
I<r
IF-

N
N
NNN
I—Q'N
I
I
I

 

CNN N_N e_N m_N e_N N_N mm, «mp mm. No. _m_ cm. Nm Ne m>Ppa>Ncma
wauwmmm uwu< ocps<

 

A.u.p=ouv .nNF mpnn»

\OG

98

apvm aswpu< to cowpumHOLa N-x =2 covpmpcavao mza was mop

mmwfiv
$5.va c o u
+ m _..0 "mam
m 2Q I Q mu
mOHI © m9.
00 2
cm mo
0 @I
U
mfiwiu Au
.0
@I
5th

.om mmnwam

 

99

elastase where substitution of this GLY with VAL modifies
the specificity by reducing the size but not the chemical
nature of the side-chains of acceptable elastase-specific
substrates (54). The observed displacement of the TRP 2l5'-
SER 2l8' chain by approximately 2 A in both the y- and x-
directions from the two-fold related position of TRP 215-
SER 218 (42) appears to be responsible for the differences
in orientation of the aromatic groups of TOS and PMS
(Table ll). The observed displacement in the y-direction
of the phenyl group (Table 10) and its difference in
rotation between molecule 1 and molecule l' reflects an
apparent attempt to maintain maximum contact between the
inhibitor phenyl ring and the specificity pocket of the
enzyme. There are approximately 19 contacts within a
distance of 3.5 A to the phenyl group of TOS and l4
contacts to PMS. The main difference in contacts between
TOS and PMS comes from six additional contacts of the

TOS methyl group; thus, approximately l3 contacts appear
to be responsible for maintaining the orientation of the
aromatic group of TOS which compares favorably with the
14 contacts observed with PMS. Although there might be

a slight difference in orientation between TOS and PMS

to accommodate the CA of PMS, the contact region for the
aromatic group remains constant. The contact region is
notable in that it principally involves main chain atoms.

It would therefore appear that the specificity site might

100

be influenced, or even possibly determined, by the
conformation of the main chain. The only polar groups

in this region are the carbonyls and amides of the main
chain with the remainder being hydrophobic in nature. The
side-chains of residues SER 214-SER 217 other than GLY 2l6
extend away from the pocket. A more detailed examination
of the specificity region will be discussed with the PEBA
results since this inhibitor more closely resembles a
typical substrate.

The difference electron density maps contain peaks
other than those attributable to substitution of an
inhibitor which occur at much lower electron density levels
are are probably associated with changes in the native
conformation which accompany the binding of an inhibitor
in the active site. These changes are indicated
schematically in the x-z projection of the specificity
region by circles for TOS and triangles for PMS (Figure
30), the enclosed number giving the peak height of the
change in eA'3. The difference peaks indicate a small
reorientation of the main chain from SER 2l4-SER 218 to
accommodate the inhibitor, and principally involves
sections of the polypeptide chain making close van der
Naals contacts (< 3.5 A) between the native enzyme and
inhibitor. The changes of molecule l do not have exact
two-fold correspondence in molecule l"due to variability

in the native structure between TRP ZlS-SER 2l8 (42).

101

The differences observed in molecule 1 are larger and more
extensive than those in molecule 1' and suggest a small
enlargement of the specificity pocket of molecule l to
approximate that present in molecule 1'. The asymmetric
inhibitor accommodation leads to a dimeric molecule
satisfying the two-fold symmetry more exactly. Although
the phenyl and the sulfonyl groups are relatively fixed in
their interactions in a-CHT, the orientation of the phenyl
groups in T05 and PMS must be slightly different to
accommodate the additional tetrahedral carbon atom of
PMS (Cl-CA-S angle is ~110°). The difference in orienta-
tion between the phenyl groups of TOS and PMS is achieved
by the increased flexibility in orientation of the phenyl
ring in PMS due to rotational freedom about the Cl-CA and
CA-S bonds. 4

The exact orientation of the sulfonyl group in the
catalytic site is not immediately discernible from the
difference electron density maps due to the limitations
imposed by 2.8 A resolution data, and to the presence of
a sulfate ion adjacent to SER 195 OY in native a-CHT
(94) which is displaced by both TOS and PMS. Although
the sulfur atom of TOS and PMS must maintain an approxi-
mately tetrahedral arrangement utilizing one carbon atom
and three oxygen atoms (one of which is SER 195 OY)’ this

tetrahedral array is not directly observable. To

102

establish a more accurate and detailed representation of
the interactions of the sulfonyl group with SER l95, it
was necessary to apply chemical knowledge. Although
actual distortions from a tetrahedral sulfur will be
somewhat obscured as a consequence, any such distortion
should not introduce significant misinterpretation.

Among the factors used in determining the orienta-
tion of the sulfonyl group were: 1) optimization of
potential interactions generated by rotation about the
Cl-S bond of TOS or Cl-CA and CA-S bonds of PMS; 2)
presence of hydrogen-bonding solvent molecules as evidenced
by positive difference electron density located adjacent
to the sulfonyl electron density; and 3) difference
density gradients involving reorientation of SER l95 0y.

A positive peak which occurs in the difference
electron density map in the immediate vicinity of the
sulfonyl group was assigned to a solvent molecule in both
the TOS and PMS derivatives and in both molecules 1 and
l'. It is labeled as H20 in the composite difference
electron density projections shown in Figure 29. Assign—
ment of such peaks to water or solvent molecules is common
in protein crystallography. In addition to the sulfate
ion assignments of MSU a-CHT, where sulfates have been
found in particularly ordered positions, counter ions
have also been suggested (94). Due to the covalent bond

between TOS and SER 195 OY with the resultant orientation

5M

103

of the sulfonyl group, the water molecule appears to be
highly ordered and hydrogen-bonded, thus defining the
position of one sulfonyl oxygen.

Further evidence for orienting the sulfonyl group
(which relates the TOS and PMS derivatives directly)
involves a density gradient at the SER l95 OY position;
in the derivatives the SER l95 hydroxyl moves approxi-
mately l.5 A by a rotation of approximately -60° (counter-
clockwise) about the ca-CB bond. The native SER l95 OY
position is 2.4 K from the sulfur in both derivatives,
much longer than the accepted S-O bond length (ml.4 A).

The difference peaks in the vicinity of the sulfonyl
in molecule l' differ from those in molecule 1 similarly
for both the TOS and PMS derivatives. This is probably
due to the slight difference in positioning of SER l95
and SER 195' (42). In molecule 1' the SER l95 OY lies
in an x-z plane which passes closer to the local two-fold
axis than the similar plane in molecule l. Once again,
it appears that the binding of both TOS and PMS tends to
reduce some of the variability of structure which occurs
in the dimer. Such effects have also been noted with the
pH 5.4 conformer by Vandlen and Tulinsky (4l).

The difference electron density map also shows that
the native active site conformation undergoes significant
changes upon binding TOS and PMS. The catalytic triad,
SER l95-HIS 57-ASP l02, appears to transmit the effects of

104

the covalent involvement of SER l95; HIS 57 rotates

counter-clockwise about its Ca-C bond away from the

8
new SER l95 OY orientation and ASP 102 retracts slightly
from its position near HIS 57. The peaks associated
with these changes appear somewhat larger in the PMS
derivative and probably reflect an additional perturba-
tion caused by the CA position which is only about 4 A
away from HIS 57 NE2. In TOS, the ASP l02 movement is
smaller than that observed in PMS and is just above the
observable limit. The movements involving HIS 57 and
ASP lOZ occur similarly in both molecules.

Other significant changes in the active site are
associated with residues involved in the dimer inter-
action. The B-chain carboxyl terminus, TYR 146', is
positioned in the active site region of molecule l (and
vice versa) and interacts via a hydrogen bond from OEE
to a sulfate ion, which in turn is hydrogen bonded to
SER l95 Oy; in addition, the carboxylic acid group of
TYR 146' interacts with HIS 57. The sulfate is lost on
inhibition, expelled by the sulfonyl group of TOS and
PMS, and subsequent reorientation of HIS 57 produces
changes in TYR l46'. This reorientation is resolved by
a rotation of approximately -40° about the Ca'CB bond

of HIS 57. The change in unit cell volume accompanying

the introduction of the methylene group in the PMS

105

derivative is probably a result of increased interaction
with TYR 146 of the two-fold related molecule, which in
turn, can affect the dimer interface.

MET l92, which shows an asymmetric orientation about
the local two-fold axis in the native dimer to minimize
the non-covalent SY-Sy. contact, exhibits a complex
movement on inhibition of the enzyme. The close approach
of MET 192 to MET 192' (m3 3) (42) makes a detailed
analysis of this region difficult but it would appear
that a rotation occurs about the CB-SY bond.

Another region displaying electron density changes
which occur in both the TOS and PMS derivatives involves
the disulfide bridge of CYS l9l-CYS 220. Similar obser-
vations have been made in other derivative studies in
this laboratory (18). Such features might be due to the
higher electron density associated with the sulfur atoms
in the native structure so that small changes in the
position can easily produce observable difference density

peaks of 0.20 eA'3 or larger.

3. Comparison of TOS (MSU)and TOS (MRC)

 

It has already been noted that the structure of
TOS-a-CHT was also determined by the MRC group (49). The
TOS derivative was used as the parent compound by the MRC
group in phase determination. The approximations concomi-

tant with such an approach have already been detailed

106

in addition to others such as averaging the structure of
the two independent molecules of the dimer.
The unit cell parameters of the MRC and MSU struc-

tures are compared in Table 13.

TABLE 13
Comparison of Unit Cell Parameters of TOS (MSU) and TOS (MRC)

 

a(A) b(A) c(3) (°) Volume (X3)

 

'45—”

TOS-pH 3.6 49.15(5) 67.02(6) 65.88(9) lOl.78(3) 212,400(700)
75% saturated

Ammonium

Sulfate

MATSpH 3.6 49.24(7) 67.2(1) 65.94(9) 101.79(6) 213,600(1000)
34

 

fl

TOS-pH 4.2 49.3(l) 67.3(1) 65.9(l) lOl.8(l) 214,000(llOO)
(49), 65%

saturated

Ammonium

Sulfate,

0.1 M

citrate,

2% dioxane

H2;3pH 4.2 49.1(1) 67.4(1) 65.9(l) 101.7(1) 214,000(1100)

 

The apparent shrinkage in unit cell volume upon tosylation
observed in the present work is not indicated in the MRC
work. The changes in the MSU cell parameters, although

small, appear significant in comparison with changes

107

observed as a function of pH. The large standard
deviations of the MRC results could possibly arise
from the introduction of dioxane into the crystalliza-
tion solvent.

The TOS orientation, generated from the MRC
coordinates, is shown in xz projection in Figure 31.
Comparison with the results of this study (Figure 32)
shows small (< 0.5 A) differences in both the placement
of the tosyl group in the active site and the orienta-
tion of the chain defining the substrate specificity
pocket (SER 214-SER 218) (Figure 31). The gradients
indicated in Figure 30 suggest shifts which lead to a
structure approximating the tosylated a-CHT structure
of the MRC group.

A direct comparison of the tosyl orientations is
shown in Figure 32, which indicates that the orientation
derived from the averaged structure of the tosylated
dimer reported by the MRC group is slightly displaced
from both tosyl group orientations of the present study.
Both the inhibitor orientation in molecule 1 and in
molecule 1' after rotation about the local two-fold
axis are compared with the MRC results. The van der
Haals contacts in MRC TOS (with TOS-a-CHT [MRC]) have
been calculated and are fewer in number than those of
MSU TOS (with native-a-CHT) using a limit of 3.5 3

(Table B-l). To compensate for the concerted mo.5 A

'O

108

Aumzv muwm m>wuu< we cowuumnogm Nix cw copuopcmwso “umzv mob .Fm mazwmm

 

Eu moo

mm.
0 <0

. U

. . . _ O V

\—
o. SmEVmOH . u

. <0

0 2 U AFNZ 9N2 (U QNZ .

.flU . .wU

mU

oo o. .oo

 

 

 

 

 

 

 

FIGURE 32 .

109

   

-— MSU T051
---- MSU Tos1’(rotated)
__ MRC TOS (published)

- - - MRC TOS (corrected)
- toluyl-
methyl

1.0 A

 

TOS (MSU) vs. TOS (MRC) Position

110

shift in the GLY 216-SER 218 backbone observed in this
work upon tosylation, the limit was extended to 4.0 A
in examining the MRC results. It can be seen in Table
14 that the number of contacts is comparable in the two
studies when the difference in choice of parent enzyme

is taken into consideration.

B. Competitive Reversible Inhibitors:Transition-State

Analogs

l. Substitution characteristics of PEBA as a
function of pH

The a-CHTzPEBA complex was examined with respect to
‘three pH conformers of a-CHT, pH 3.6, pH 5.4, and pH
‘7.3. As described in the Experimental section, 30 mM
IDEBA was added to the mother liquor of the desired
(:rystalline pH conformer. The appropriate pH conformer
vvas verified initially by comparison with known axial
(diffraction patterns and unit cell dimensions. The
iaddition of PEBA produced further significant changes
'in both the diffraction pattern and unit cell parameters
(Table 15).

From Table 15, it can be noted that the addition
(If PEBA to the pH conformers increased the b-axis at
I3}! 3.6 but decreased it at pH 5.4 and pH 7.3, while the
it1teraxial angle remained unchanged with the substitution.
These two parameters had proven to be the most sensitive

12:) pH changes in earlier studies, although a pH study

.mau .Nau .mo .eu .mo .No .Po macaw
gee; eseee meeeeeeu « m.m esp co ANFV Nme .Nesem am: we» eH "e_esexm .F-< speak some

 

111

e_ mN me um:
4N ee - an:
«em «ma Nee
msoum _Pm so» _mpoh
ANmNV N_ ANNeV e_ m o.N
Aeemv PN ANNmV mN m m.m
gays - «o3

muumpcou Punch
“mo .No ._o .m macaw meeu=PU=Hv

ANe_v N ANNNV N m o.m
ANO_V e ANNeV me n m.m
Aemev FN - m o.e

meoepcoo Peach
IAmeu .Nao .mo .eo .mu .No .Pu agape meee=_ecev
mma Neeem weep

meeeeem UN: eee am: e? #10.5 eee Psmoe geezeem

mmuceumwo m—mmz emu cm> cwgpvz muumucou mo comwcmasou

¢_ uqnmh

112

TABLE 15
Comparison of Cell Parameters of PEBA at pH 3.6, 5.4 and 7.3

 

 

a(A) b(A) C(A) s(°) Volume (A3)
PEBA pH 3 6 49.36(4) 67.43(4) 65.99(6) 101.83(5) 215,000(600)
NAT pH 3 6 49.24(7) 67.20(10) 65.94(9) 101.79(6) 214,000(1000)
PEBA pH 5.4 49.37 3) 67.32(4) 65.73(5) 101.92(6) 213,700(500)
NAT pH 5 4 49.13 5) 67.83(7) 65.81(7) 101.92(6) 214,600(700)
PEBA pH 7 3 49.35(2) 67.68(2) 65.98(3) 102.03(4) 215,500(300)
NAT pH 7 3 49.24(2) 67.98(3) 65.85(4) 102.03(3) 215,600(400)

 

performed on a-CHT specifically oxidized at MET 192 shows
effects similar to the b-axis behavior observed here. It
should be noted further that among the boronic acid
derivatives studied, PEBA, PPBA, and PBBA, those deriva-
tives which do not appear to interact strongly at both
the specificity site and the catalytic site (PEBA pH 3.6,
PPBA and PBBA as discussed below) show an increase in
their respective unit cell volume, while those which
appear to interact more strongly at both sites (PEBA pH
5.4 and 7.3) show a decrease in unit cell volume or
remain constant. The TOS and PMS derivatives would
not necessarily be expected to follow this pattern as
they interact covalently with SER 195 and do not meet
the specificity requirements of a-CHT.

As with the irreVersibly inhibited derivatives,

fiaxamination of the difference electron density map

113

revealed small, localized regions of both positive and
negative electron density (> 0.15 eA'3) accompanying the
large substitution density of PEBA at pH 3.6, both in and
removed from the active site region. The difference maps
calculated from the derivative and native data at pH 5.4
and pH 7.3 revealed more extensive regions of perturbation
of the native structure., Due to the separation of the
phenyl ring from the boronate group by an ethyl group,
two maximum density regions occur. These regions corres-
pond to the phenyl group in the specificity pocket (x 1
49/76) and to the boronate group near the catalytic site
(x 3 50/76). A comparison of these peak heights is
representative of the occupancy and/or ordering of the

inhibitor in the two sites (Table 16). The independent

TABLE 16
Comparison of Peak Heights of Difference Electron

Density Regions in the Active Site

of PEBA pH 3.6, 5.4 and 7.3

 

 

Derivative Site Molecule 1 Molecule 1'
PEBA pH 3.6 catalytic 0.47 eA'3 0.38 eA'3
specificity 0.37 0.33
PEBA pH 5.4 catalytic 0.46 0.45
specificity 0.46 0.45
PEBA pH 7.3 catalytic 0.44 0.33
specificity 0.39 0.45

‘

114

molecules exhibit good two-fold symmetry in their
occupancies only at the higher pH values.

The x-y and x-z composite bounded projections of
the difference electron density for the substitution of
PEBA at the three pH values are shown in Figure 33,
from which it can be seen that there is a striking
similarity in the orientation of the phenyl group density
in the specificity pocket as a function of pH in the
independent a-CHT molecules.

A comparison of the centroids of the difference
electron density of PEBA pH conformers across the local
two-fold axis shows a significant variation in the y-
direction (Table 17) as previously noted with TOS and
PMS (Table 10). Again, the estimated error in coordinate
is about i 0.4 A. The above comparison indicates a strong
pH effect in the catalytic site; the results at pH 3.6
show a large deviation from local two-fold symmetry (in
the y-direction), while at pH 5.4 and 7.3, the deviation
is within the limits of error. The constant variation in
the specificity site for this series seems indicative of
the fact that the side-chain of PEBA approximates that
of phenylalanine, an a-CHT specific substrate. The atomic
coordinates for PEBA at pH 3.6, 5.4 and 7.3 were determined
in a manner similar to that used for TOS and PMS and are

given in Table A-1.

115

7% 36 an vatsoucoo £324.» :38

eezee> .P «paee_oz .eeem e>_ee< co seemem er <mma

A~v m.~ :a <mma

ANV ¢.m :n (mum

 

.mmm mmzuuu

ANV e.m 2a (owe

 

116

mine m_.o an vacaoucoo .mwx<-~.z axon
umzmp> ..F upsuopoz .muvm m>muu< mo cowmmm cm <mue

ANV N.N Ia .<Nma ANN e.m ze .<mNN

\

 

Aav m.N In .<NN¢ Aav e.m Ia .<mme

\r

 

.nmm mmsomu

ANV e.m xa .<mNe

Q.

\

k

4‘

Axv m.m In .<mNa

"

9)
J

117

.uoxuoa uruapouau
2: 5 as a 8.? as 33 < mm... 2. 3338 33:83 2: 5 as m 35 2; as as x
as c g

 

 

 

 

 

 

 

 

.p 232 . 53 .2 a. 58 2e 4.... ._e (3.. ea 22:3,. 282.8 3 e353 5: :2. 2: 3:.st
I 2 i .3”: . S .3 <9:
$28 A 5 ea .3 ed _a
39.98 < 2.? < Ni? 28.. < 3.? < .85 «a...
ovuapauau . o e o o
25. .v. .3 3 1e 3 In
a gene m NN.o m 8.? .52 m 35 m Rd «3..
:eEwe m .3 1e .3: .3 E «a: ..
$28 A 5 3” E 3. _a
3938 < 55 < 3.? < 9.5 < 3.0 .52 < 9...? < 3.? < B? < 8..- SE
UPQ%P0U~U O O O O O 0 O 0
323 .v. 5 3 a. em a
33.08 < NN.o < 8.? < 85 < NM? .23”: < 3.? < 2.0 < 3.? < 3.? 5m:
:PUrFrUUQm o o o O o o o

 

 

 

 

 

 

 

 

 

 

 

 

No me Ne mq N< an m< me

.855 :58
oceppoumxcu ex» ea mopsumpoz ucwucmamvcn on» mo comm cw covaauvumnam soupnvcca m>vupumneou mo comvganeou .n

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«.5 In M.“ In ¢.m In ¢.m :a o.m In w.m In
.<mum .<mwa .<mma. .<mma .<mmq .<mwm
. I me me me me me me
AON\Omqu_UMW ".5 In . ¢.m _..—a @.M $9
ovuhpuuau m p~.o m m¢.0I . <mmn M no.0 m m~.oI (mum M mm.o m mF.FI <mwm
823 .v. 3 2 a. 3 1a . 3” E
auxuoa < mp.o < Nm.on <mmm < ov.o < m~.oI (mum K h~.o K mo.pn (mum
auruvpvumam - o o o - o
~< m< . ~< . m< m< me

p opsuupoz sage cowuov>oo on» mcFuaLu><
an uozoppom ..p opsoopoz er coeuaupumnzm as» no umscomeue mrx< upouuozp uson< covumuoa mo conveoaeou .o

an.“ :a <mmm one e.m :q <mue .c.m :a («Nev .F can F mopsuupo: eo
moupm o>_uo< use :N,»u_mcun covusuNumaam mo mcoeupmoa uwogucou mo comvcuqeou
. 2 39:

118

2. Orientation of PEBA molecule and interaction

with a-CHT as a function of pH

The PEBA orientations at each pH value are shown in
Figures 34-36. The peaks observed in the difference
electron density maps of both the native pH conformer
and the inhibited conformer are indicated. The enzyme
inhibitor contacts for each of the PEBA derivatives have
been summarized in Table B-1.

It is apparent from examination of Tables 11, A-1,
and B-1 that, although the phenyl group extends into
approximately the same translational position at each of
the hydrogen ion concentrations studied, significant
differences in orientation occur, both among the pH
conformers and within the a-CHT dimer. The PEBA results
are of particular interest as the phenyl ethyl group
‘resembles the phenylalanine side-chain and the interactions
‘in the hydrophobic pocket should approximate those respon-
ssible for substrate specificity. The possible application

c>f PEBA as a transition-state analog for a-CHT-catalyzed

(asster hydrolysis is partially dependent on this isosterism.
Examination of the inclination angles of PEBA reveals

a similar orientation of the phenyl ring to that of the

13<>$yl ring in molecule 1 at all the pH's investigated.

I't will be seen below that an x-y inclination angle of

45° also occurs with the PPBA and PBBA derivatives in

119

eeem eseee<
co eoeeeeweea N-x =_ eeeeeeeeeeo e.m :a <Nme .em mmaoHe

820
o
o
o z . z 50
AMOOVXC mu . o 320 u o u 095
+9.6» mxm _ mu ._.v 2
N9? _ 9:0
8 . 2
N00

- moo

8 mo. . .-6 6w 8

a” o O
0 mFN. .
u
u o .
No
$6 6.0 . u ,v&<o z
+
o 055 998 @
@. z z 2
U mU
mu @+ a mu
0
0
00

120

we cowpumnogm NIx

mupm m>wpu<
e. coeeeeeeeeo e. m In <mma

 

.mm museum

mmsxu
Amoéx c o
.+mm r.AVI nmwxxmv .U FOnV
emcSU nu
vale 4mmeI© . 8 8, .0
mm In -4 A 3. . mes
, me 2_
N00
am. “no.
AV
6 e-
m=N<u_ Aw+
.U
. o
@I hm<m ® . @I
nut. No he
@+ 00 G- . ,

omcflu

NmXu

 

121

poem eseee<

mo corpumnosm NIx cw cowpmucmwgo m.n In <mma .om wastu
O
820
o .
$0.va c o w 50 o
. . n V
mm In <m me @ ee 3 _ 2
o < m.
muk.+tnu AN 2 e<o
8 N00
N8
om . - 5.0 mu
WU . o‘qgo Q...
. 0
meme.
0. U
0 6+ . Noe
9+ 0 fl. GO 220 z
th<U AV.
0 @- 9m 0 z
.. mo
29. /oo 4V.mvo
mu nU+
mu

 

122

Inolecule l, suggesting that this orientation optimizes
the interactions between the specificity pocket of o-CHT
and the aromatic group of an inhibitor or substrate.

The deviation observed in PMS (m65°) is consistent with

a tetrahedral configuration about the CA atom (S-CA-Cl
angle = ll0°), and the involvement of a side-chain too
long to possess complete orientational freedom and too
short to extend properly into the hydrophobic cleft.

This further explains the identical x-y inclination angle
for PMS in molecules 1 and l'. The variability exhibited
in the orientations in the dimer in all three PEBA deri-
vatives is explainable in terms of the conformational
variability of the dimer. The observed displacement of
the SER 214'-SER 217' chain in molecule 1' away from

SER 195 (along x) orients the aromatic ring with a
smaller x-y inclination angle to maximize the contacts
within the specificity pocket.

A more substantial variation in the aromatic
rotation angle of PEBA is observed with each of the pH
conformers. It should be apparent that the contacts
listed in Table 12 are responsible for the primary
interactions within the specificity pocket. Although
little structural change is noted in the SER 2l4-SER 217
region with pH, the movement which is observed (at pH
5.4) seems to decrease the asymmetry of the dimer, thus

possibly accounting for the similarity in aromatic

123

rotation angle. The anomalous values of the x-z
inclination angle for the pH 5.4 conformer is indicative
of the asymmetry remaining.

Overall, the introduction of the phenyl ethyl group
into the specificity pocket at the hydrogen ion concen-
trations studied produces movement of the SER 214-SER 217
chain to enlarge the pocket and optimize interactions
between enzyme and inhibitor. Although it is not
immediately obvious, distance calculations reveal that
those changes in orientation which occur as a function
of pH result in almost doubling the number of contacts
with the phenyl ring (Table 12a). At pH 3.6, the PEBA
phenyl ring (15 contacts) interacts at approximately
the same extent as the phenyl rings of PMS (14 contacts)
and TOS (l9 contacts). At pH 5.4 and 7.3, these contact
regions expand to include 26 and 25 contacts, respectively
(Table 12). As with TOS and PMS, the majority of the
contacts within the specificity pocket involve main
chain atoms of a-CHT, reinforcing the importance of the
main chain conformation. It should also be noted that
the pH 5.4 and pH 7.3 PEBA conformers show a substantial
increase in the number of contacts in the active site
region (29 at pH 3.6, 52 at pH 5.4 and 48 at pH 7.3;
Table l2).

The most significant difference in the orientation

of PEBA occurs in the catalytic region. Examination of

124

the composite difference electron density (Figure 33)
and the orientation diagrams (Figures 34-36) reveals
that as the pH is increased between pH 3.6 and pH 7.3,
the boronate group moves toward SER l95, shifting almost
1 A along the z-direction towards SER 195 07' The pH
5.4 structure is intermediate in this movement. The
o-CHT PEBA contacts (Table 12) further show that the
number of contacts in the catalytic site also increases
substantially upon raising the pH above 3.6 (l4 contacts
at pH 3.6, 26 at pH 5.4 and 23 at pH 7.3).

The orientation of the boronate functional group
is not clearly defined at 2.8 A resolution, similar to
the TOS and PMS derivatives. Unlike the tetrahedral
sulfonyl fluoride inhibitors, the boronic acids (R-B-
(0H)2) exhibit a tendency to accept a hydroxyl group
as a function of pH, converting a trigonal boron arrange-
ment to a tetrahedral boronate configuration. Examination
of the composite electron density of the PEBA substitution
(Figures 34-36) shows the orientational differences of
the —B(0H)2 group among the three pH conformers. This
variability is assumed to be a function of the interaction
of the various ionic equilibrium states (R-B(0H)2 +
0H'22 R-B(OH)§) (95) previously observed in solution
studies. This structural variability of the boronate
group complicates the interpretation of the detailed

interaction in the catalytic site because a covalent bond

125

to SER l95 is not obvious as with TOS and PMS. Evidence
for a tight interaction between the boronate group and
OY-(l95) is given by the inability to wash out the
inhibitor-soaked crystal as discussed below.

The boronate group occupies different orientations
in the catalytic site at each of the three pH conformers
of the a-CHT dimer studied. At pH 3.6, this group is
positioned furthest from SER l95 OY (and l95') as
evidenced by fewer contacts less than 3.5 A between a-CHT
and inhibitor. The difference electron density indicates
that a water molecule might be hydrogen bonded to 02 of
the boronate and that the separation between the boron
and SER l95 OY is about 3.5 A after applying a shift to
the position of SER 195 as indicated by a difference
electron density gradient. Thus, at pH 3.6 SER l95 0Y
is probably involved in a weak interaction with a tri-
gonal -B(0H)2-H20. The same applies to the interactions
in the catalytic site of molecule 1', except for the
difference in SER l95 0Y position already discussed for
TOS. The catalytic region generally undergoes some
small rearrangements upon inhibitor binding at pH 3.6,
probably a result of the displacement of the sulfate
ion bridging monomers of the dimer (94).

As with TOS and PMS and all boronic acid derivatives

used in this study, the sulfate ion, present in the active

126

site region of native a-CHT adjacent to SER 195 OY’ must
be displaced by the sulfonyl or boronate group. In
examining the inhibitor orientation by the difference
Fourier method, the analysis in the catalytic region of
the active site is complicated because the inhibitor
density occupies the region of the displaced sulfate

electron density

pprotein ' pprotein = pinhibitor ' psulfate

+ inhibitor + sulfate
Since the sulfate in native a-CHT links HIS 57, SER 195,
GLY 193 and TYR 146' by hydrogen bonding and charge
interactions, the presence of an inhibitor, whether coval-
ently bound or weakly coordinated to SER 195 0y, will cause
a reorientation of this cluster of residues.

The SER 195 changes associated with PEBA pH 3.6 are
both smaller and opposite in direction from those observed
in TOS and PMS. At pH 3.6, PEBA produces larger changes
in the vicinity of MET 192-GLY 193 than near SER 195
which results from the orientation of the boronate group.
HIS 57 appears to rotate about Ca'CB as in TOS and PMS
but the associated peak is somewhat smaller in PEBA pH
3.6, while TYR 146' retracts from its position in the
native structure by an apparent rotation about Ca-C of

8
about +45°.

127

The PEBA derivative at pH 5.4 is most notable because
of a shift in position of the boronate group of approxi-
mately 1 A which results from a rotation of 120° around
the inhibitor CA-CB bond from that of the pH 3.6 deriva-
tive. The effect of this reorientation is to double the
number of interactions within a range of 3.5 A between
a-CHT and PEBA (Table 12). The increased interaction at
pH 5.4 has already been noted in the specificity region
so that the number of van der Naals contacts and the
potential hydrogen bonds have increased markedly with the
pH 5.4 conformer.

Although PEBA at pH 3.6 assuredly contains a trigonal
-B(OH)2 group, the configuration in the pH 5.4 a-CHT
complex will be dependent upon the pH of its local micro-
environment. The pH of the conformers of a-CHT referred
to here represents the hydrogen ion concentration of the
mother liquor in which the crystals have been soaking.

The pH of the catalytic site is taken to be the same as

the soaking solution. The value of pH 5.4 is close to

the transition of the boronate group from its trigonal

to tetrahedral form. Due to the lack of atomic resolution,
determination of the exact boronate configuration is not
possible from the electron density. However, the differ-
ence electron density map suggests that the interaction

between SER 195 CY and the boronate forms a generally

128

more intimate complex than that at pH 3.6. The observed
boron-SER l95 0Y distance in the pH 5.4 complex is
about 2.5 A.

A change of the native a-CHT conformation upon PEBA
inhibition at pH 5.4 is suggested by difference electron
density gradients around the region CYS 191-SER 195.

The rotation of SER 195 0Y in PEBA pH 5.4 is both larger
and opposite in direction to that at pH 3.6 (a rotation

of 'b+90° about the Ca-C bond). The other shifts in this

8
region cannot be analyzed definitely but would appear to
indicate an increased interaction between the inhibitor
and GLY 193, probably through its amide group. The
remainder of the catalytic triad (HIS 57 and ASP 102)
remains intact, probably because the boronate occupies
approximately the same position vacated by the sulfate
ion and interacts in a similar manner (except for the
shift in SER 195 0Y' It has already been observed by
Vandlen and Tulinsky that the two active sites become
more similar at pH 5.4 and this is in good agreement
with the occupancies of PEBA pH 5.4 (Table 16). The
reorientation of SER 195 0Y could thus be a pH effect
which is either more pronounced because of the inter-
action with the inhibitor, or easier to observe in the
latter.

At pH 7.3, PEBA is oriented with its boronate group

in yet another position. The activity profile of a-CHT

129

would suggest even a greater interaction might occur
between the enzyme and a substrate. Moreover, at pH
7.3, it can be assumed that the boronate is tetrahedral
(pK = 9.2). Thus, in the pH 7.3 conformer the boron is
positioned about 1.8 A from SER 195 0Y' However, the
electron density distribution in this region at pH 7.3
neither proves nor disproves tetrahedral boronate.

It should be noted that the orientation of SER 195
is unique to each pH conformer and is greatly affected
by the presence of PEBA. There is no apparent reorienta-
tion of SER 195 in PEBA pH 7.3. The conformational
changes between CYS 191-SER l95 involve MET 192-GLY 193

and result in a rotation of the Ca-C bond of ASP 194,

B
which affects the crucial ASP 194-ILE 16 ion pair. The
rotation of ASP 194 appears to be a precursor to the
reorganization observed with the pH 8.6 conformer (43),
and from solution studies (10) is thought to be largely
responsible for the high pH fall-off of the a-CHT-pH
activity profile. The pH 8.6 conformer shows an
apparent difference in the pK of the ASP 194-ILE 16 ion
pair in the two molecules of the asymmetric dimer. The
reorganization of the active site of molecule 1 could
be responsible for the decrease in occupancy of PEBA

observed in molecule 1 on raising the pH from 5.4 to

7.3 (Tab1e 16).

130

The asymmetry of the orientation of PEBA in a-CHT
dimer can be seen from Tables 16, 17 and in Figure 33.
It would appear from the non-identical conformational
changes which occur as a function of pH that the orienta-
tional asymmetry results from an attempt to maximize the
interaction between PEBA and a-CHT. It is apparent that
both the x-y and x-z inclination angles are relatively
conserved with the largest deviation occurring in the
aromatic rotation angle. While the inclination angles
are responsible for the primary orientation of the
aromatic group of PEBA in the specificity site, it is
the aromatic rotation angle which produces the differences
in the interactions observed between TOS and PMS, and

between PEBA at pH 3.6, 5.4 and 7.3.

3. Evaluation of PEBA as a transition-state analog

The tetrahedral form of the boronate ion is expected
to approximate the tetrahedral intermediate suggested in
the mechanism of ester hydrolysis catalyzed by a-CHT.
This tetrahedral intermediate has been implicated in non-
enzymatic ester hydrolysis by isotope exchange experiments
so that it would be of great interest to understand the
mechanism of the rate-enhanced hydrolysis by enzyme
catalysis. The emphasis in current research is on those
features of enzymatic catalysis which increase the

catalytic efficiency: their cause and their effect. The

131

use of molecules which approximate the proposed transition-
state (structure of highest energy) and not the substrate
itself should enable the investigation of the transition-
state of the enzyme-catalyzed mechanism.

A comparison between enzymatic and non-enzymatic
catalysis is made in Figure 37 (96), if it is assumed
that these proceed by the mechanisms indicated. The com-
parisons also assume that the transition-state structure
of the substrate is identical in both the enzymatic and
non-enzymatic process but that the enzyme allows lower
free energy of activation to the transition-state (57).

It has been suggested that this results from chelating
effects within the active site (i.e., a high concentration
of weak interactions such as hydrogen bonding, dipolar
interactions, etc.) and within the specificity region of
the enzyme which distort the substrate molecule to best
accommodate these interactions.

The above ideas concerning enzymatic catalysis have
been expanded to consider the design of analogs to the
transition-state for use in probing the interactions
responsible for high catalytic efficiency. It was anti-
cipated that a molecule which resembled the transition-
state of the substrate would be bound more closely to
the enzyme than the substrate itself (56), thus permitting
the examination of a complex approximating a species

whose half-life is approximately 10'10 sec. Since these

132

 

 

 

 

 

 

20
'32
O
E u)
‘a
a?
LL
‘q
0
~10
R EACTION PATH WAY
E = Enzyme
S = Substrate
P = Product
ES = Enzyme Substrate Complex
ES* - Complex Transition-State
S1' = Substrate Transition-State
EP = Enzyme Product Complex
AFS° - Formation of ES from E and S
AFT = Formation of ES* from E and 5+
AFN = Formation of S+ from S
AFE+ = Formation of ES from ES

FIGURE 37. Energy Diagram of Proposed Enzymatic Mechanism
Involving a Transition-State for Catalysis (96)

133

complexes are stable, they can be studied by physical
techniques such as X-ray diffraction (96). It should be
noted that a perfect transition-state analog cannot be
synthesized because the bonds differ in length, charge
distribution, etc., and lead to minor differences in
free energies of solvation, all of which ultimately
disturb the transition-state equilibrium. Also, the
more closely the analog resembles the substrate transi-
tion-state, the greater is its susceptibility to
catalytic attack. The observed binding constants (inhi-

3 4

to 10' mM, can differ appreciably

8 ~14

bition constants), 10'

from the theoretical values (97), 10' to 10 mM and
both non-ideality and electronic interactions have been
held responsible. The major consideration in choosing

an analog is structural similarity in both the catalytic
and specificity regions of the inhibitor molecule.

The methyl ester of phenylalanine is a primary
substrate of a-CHT, its phenyl ethyl side-chain occupying
the specificity pocket. This substrate has as an analog,
phenyl ethane boronic acid, which is closely isosteric
to the former. The use of the boronic acid group to
approximate carbonyl compounds was pioneered by Antonov
gt _1, (65) by studying the interactions of both alkyl
and phenyl alkyl boronic acids with a-CHT. Philipp and

Bender (98) examined substituted aryl boronic acids as

inhibitors of both a-CHT and subtilisin. Koehler and

5E

134

Lienhard (66) investigated the PEBA system in comparison
to other competitive inhibitors, emphasizing the struc-
tural correspondence of PEBA to the proposed tetrahedral
transition-state of the phenylalanine substrate. A
companion study of the subtilisin-PEBA system was performed
by Lindquist and Terry (63). Kinetic studies of the
reaction of benzene boronic acid with both subtilisin and
a-CHT have been performed by Nakatani _t _l, (99) in an
effort to understand the mechanism of catalysis. All
these solution studies arrived at the same general con-
clusions concerning the applicability of transition-state
analog theory to the serine proteases and alkyl or phenyl
alkyl boronic acids.

Boronic acids, as a class, have been shown to inhibit
a-CHT and subtilisin by the formation of a tightly bound

2

complex; for PEBA a-CHT, KD = 4.7 x 10' mM, for subtili-

2

sin, KD = 2.8 x 10' mM. These values are larger than

those predicted by transition-state theory, where

KD = Kw x kn/(Ka x ke) (62)

where,
KD = transition-state dissociation constant
Kw = dissociation constant of water
Ka = dissociation constant of boronic acid
kn = non-enzymatic rate constant (base

(catalysis)

k = enzymatic rate constant

135

5 4

The predicted K for a-CHT is 9 x 10' mM and 7 x 10' mM

0
for subtilisin. These differences have been attributed
to non-exactness of the transition-state analogs and
overlap of pH transition ranges. In examining both the
a-CHT and subtilisin interactions with PEBA, it has been
observed that the inhibition constant is a function of
both pH and side-chain length. The pH-Ki (Ki = l/KD)
profile for the two systems is identical below pH 8 and
reflects the conformational transition of a-CHT which
occurs at higher pH values with PEBA, hydrocinnamide and
2-phenyl ethane sulfonate (66). Although more extensive
specificity studies have been performed on a-CHT, it is
apparent that the substrate specificity is identical to
that observed with the boronic acids.

A significant difference occurs in the interpretation
of the specifics of the boronate-enzyme interaction.
Hhile Antonov gt_al. (65) and Philipp and Bender (98)
report the occurrence of the boronic acidzenzyme complex
to be an interaction primarily with HIS 57, Koehler and
Lienhard (66) implicate SER 195 as the primary site of
interaction.

Hess 33 al. (100) have reported the existence of a
tetrahedral intermediate in the complex of a-CHT and PEBA
after examining the complex using laser Raman spectro-

scopic techniques. The appearance of a peak in the

spectrum of the complex at pH 7.0, but not at pH 5.5, is

136

highly suggestive of the presence of a tetrahedral PEBA,
but still requires a normal coordinate analysis of the
boronate structure. The observation that the PEBA
saturates the active site of a-CHT at pH 5.0 as well as
at pH 7.0 (100) agrees with the results of the present
study which indicate approximate full occupancy of PEBA
at pH 3.6, 5.4 and 7.3 in crystalline a-CHT. This result
does not conflict with the observations concerning the
variation of binding strength versus pH since it was
possible to wash out the PEBA from the pH 3.6 and pH
5.4 crystals with original PEBA-free mother liquor.
The diffraction patterns returned to those of the native
conformers at pH 3.6 and 5.4 after washing. At pH 7.3
the result was not definitive; changes were observed upon
attempts to wash out the PEBA but without attainment of
the pattern of the native pH 7.3 conformer. One possi-
bility could be the removal of the inhibitor from one
of the molecules of the dimer.

Matthews gt al. (67) have investigated the structure
of boronic acid adducts with subtilisin BPN (Novo) in
the crystalline state. Their 2.5 A resolution study
considered both benzene boronic acid (BBA) and PEBA, at
pH 7.5. The initial phasing for the subtilisin structure
was derived from the PMS-inhibited structure at pH 5.9.
The phases used to examine both pH differences and

inhibitor complexes were determined by a structure factor

137

calculation based on a Kendrew model constructed at
pH 5.9 and were extended to 2.0 A resolution. The inter-
actions in the catalytic region of subtilisin correspond
to those in a-CHTzPEBA (HIS 64, GLY 219, SER 221,
subtilisin; HIS 57, GLY l93, SER 195, o-CHT) and would
imply the same mode of binding of PEBA in subtilisin at
pH 7.5 as with a-CHT at pH 7.3. The previous examination
of the TOS (MSU) and TOS (MRC) structures lacked the
exacting comparison necessary to evaluate the full influ-
ence of the choice of parent compounds for protein MIR
phasing. Differences were observed between the native
structures in the two studies. At less than atomic
resolution the uncertainty caused by the weighting of
the parent phasing structure upon detailed structural
analysis, particularly when utilizing Ap maps, should be
examined at high resolution before attempting to make
reliable, absolute comparisons of such results.

In summary, PEBA binds more directly to SER 195
than HIS 57 in a-CHT by interacting through the boron
and SER 195 0y. At pH 7.3, it appears that the boron
could be in a tetrahedral arrangement utilizing the
0Y of SER 195 and the interaction of PEBA is consistent
with observations both in solution and with subtilisin.
However, lack of atomic resolution prevents the direct

observation of this geometry. The interaction in the

138

catalytic region of the active site appears only slightly
different at pH 5.4, but very different at pH 3.6. The
interpretation of both the substitution interaction and
the concomitant conformational changes which occur upon
binding are affected by the conformational variability
previously observed in a-CHT with change in pH. It

would appear that the high occupancy observed at all
three pH values results from a complementarity between
PEBA and the a-CHT specificity and catalytic sites, while
the degree of the interaction principally results from
SER 195 and secondarily from GLY 193 (hydrogen bonding).
This pH variability reflects a probable overlap of the

pH “activation" of the active site (probably resulting
from HIS 57 deprotonation) and the tetrahedral transfor-
mation of the boronic acid group. While the observed
binding and interaction of PEBA with a-CHT do not in

any way disprove the existence of a tetrahedral inter-
mediate, they also do not permit a definitive proof

that it occurs in this system.

4. Comparison with enzyme:inhibitor (natural)

 

complex structures

 

Another crystallographic approach to the enzyme:
substrate complex problem has utilized the inhibition of
trypsin (T) by the natural protein inhibitors, pancreatic

trypsin inhibitor (PTI) and soybean trypsin inhibitor

“L

 

139

(STI). The applicability of such a complex to the a-CHT
system would seem to be straightforward in view of the
similarity in the catalytic mechanism of a-CHT and T.
Such studies have been completed, on bovine PTI-T by
Ruhlmann t l. (101), and on porcine T-STI by Sweet

gt al. (102). These crystallographic results are of
particular interest since Blow and Huber had previously
collaborated in predicting the manner of association

of PTI with a-CHT and T based on model-building studies
utilizing the structures of PTI, a-CHT and T (103).
Huber's group had previously determined the structure
of PTI at 1.5 A resolution (104).

Crystals of the PTI-T complex yielded diffraction
data to 2.8 A resolution (101). Previous work had
determined that the interaction between PTI and T
involved the LYS lSI-ALA 161 link of PTI. The establish-
ment of a tetrahedral configuration about this amide
carbon came primarily from the position of SER l95 0y.
Comparison of the PTI-T active site with that of the
MRC a—CHT structure revealed the orientation of SER 195 0Y
to be in the "acylated" position observed in indoleacroyl-
(48) and tosyl-a-CHT (49). The absence of electron density
in the native orientation further suggested the involve-
ment of SER l95 0Y in an interaction other than a
Michaelis complex, particularly as the bond between

LYS lSI-ALA l6I appeared intact. Including the tetrahedral

140

configuration into a real space refinement produced a
final orientation in agreement with the starting model.
The stability of this complex (Ki > 1014) was thus
thought to depend on the strain present in the tetra-
hedral carbon bond angles observed in the PTI structure,
121° and 98° compared with 109° (ideal). It was also
concluded that the structural changes between the native
structures and those in the complex represented the
minimum changes necessary to accommodate PTI and T,

and that this additionally lowered the energy of the
transition-state. Further study suggested that the
intermediate might be somewhat between the acyl-enzyme
and the tetrahedral adduct due to a slightly extended
C(151)-0Y(195) distance.

Blow and his collaborators investigated the STI-T
complex, in which STI can be reversibly hydrolyzed, because
of its greater stability and homogeneity. Crystals of
the complex yielded diffraction data to 2.6 A resolution.
Solution studies had shown that the scissle bond of STI,
ARG 63I-ILE 641, was reversibly cleaved but bound to the
enzyme in an inhibitory way (102). The refinement
procedure used a modified set of bovine coordinates,
structure factor calculation and a combination with the
MIR phases. They attempted to examine the nature of the
scissle bond by deleting the constraint of the peptide

linkage and allowing the independent refinement of

141

SER l95, ARG 631 and ILE 641. A comparison was made of
these results (CB(195)-Ca(631) distance 3.7 A and
Ca(63I)-Ca(641) distance 3.9 K) with those established
for a Michaelis complex (5.2 A and 3.8 A), tetrahedral
intermediate (3.7 A and 3.8 A), and acyl-enzyme (3.7 A
and 5.0 A). The coordinates reflect a tetrahedral
configuration, even though Huber gt_al. (101) have
suggested that C(63I)-0Y(195) is longer than normal C-0
single bonds. Unlike PTI, $11 was not observed to have
strained angles about the scissle bond, and the complex

stability (Ki > 1011)

is attributed to the availability
of free energy from entropy driven effects. It is
further hypothesized that the tetrahedral form is
stabilized as a minimum of free energy of the total
system and that the function of trypsin-like enzyme is
to lower the activation energy of the overall reaction.
This is accomplished by the delocalization of the
negative charge on the tetrahedral group which also
strengthens potential hydrogen bonds between carbonyl
oxygen and NH (193) and NH (195) and improves alignment.
In accordance with the above, the tetrahedral
intermediate has been assumed as fact; although con-
ceivably it could be an artifact of the refinement

procedures. It would appear imperative therefore in

applications of this nature that the dependence of

142

refinement procedures upon the initial model be established
in order to evaluate the reliability of the details

resulting from these analytical approaches.

5. Phenyl alkyl boronic acid orientation as a
function of alkyl chain length

The structures of a-CHTzPPBA and a-CHTzPBBA were
studied to further probe the specificity requirement of
a-CHT. The study was carried out at pH 4.6 and the cell
parameters for these two derivative crystals proved to
be identical; however, they differ markedly along the a-
and b-axes from those observed in native a-CHT pH 4.6

crystals (Table l8). Both derivative crystals exhibit

TABLE 18
Cell Parameters for PPBA and PBBA

 

a(A) b(A) C(R) e(°) Volume (33)

 

PPBA pH 4.6 49.54(4) 67.39(6) 66.00(7) 101.92(5) 215,600(600)
PBBA pH 4.6 49.54(2) 67.43(2) 66.03(3) 101.97(4) 215,800(300;
NAT pH 4.6 49.3 (1) 67.8 (1) 65.9 (1) 102.0 (1) 215,000(600

 

cell volumes greater than that of the native pH 4.6 confor-
mer. Although the variation in the cell parameters of all
the boronic acid derivatives is probably indicative of
changes occurring in the crystal, there are too few obser-
vations to make more conclusive statements. The variation

in unit cell volume, although only slightly beyond the

143

estimated error for all the boronic acid derivatives and
for TOS and PMS, appears significant and is of the same
magnitude as the variation observed among the native pH
conformers.

An examination of the difference electron density
maps of PPBA and PBBA showed lower occupancies in the

active site region for both derivatives (Table 19). The

TABLE 19
Comparison of Peak Heights of Difference Electron
Density in the Active Sites of the
PPBA and PBBA Derivatives

 

 

Derivative Site Molecule 1 Molecule 1'

PPBA pH 4.6 catalytic 0.16 eA'3 0.12 eA'
specificity 0.31 0.20

PBBA pH 4.6 'catalytic 0.17 0.14
specificity 0.21 0.20

 

large difference in the occupancy between these two
derivatives and that of PEBA initially generated some
concern about the integrity of the inhibitors themselves,
since PPBA and PBBA were from the same source and PEBA
from a different source. However, a new sample of PPBA
gave essentially identical results with respect to
occupancy of the substitution and conformational changes

throughout the molecule. The differences in occupancy

144

observed upon the lengthening of the alkyl chain appear
to be due to a limitation imposed by the specificity
pocket.

The x-y and x-z composite bounded difference electron
density projections of PPBA and PBBA are shown in Figure
38. A comparison of these projections to those of the
other inhibitors indicates the generally lower occupancy
since all are drawn at approximately the same contour
levels; however, the comparison shows excellent agreement
with respect to orientation of the inhibitor both between
the two independent molecules and even between PPBA and
PBBA. This observation is further corroborated by the
comparison of the centroid positions of the substitution
densities as previously described (Table 20). The dis-
placement observed in the y-direction with the other
inhibitors upon rotation of molecule 1' about the two-fold
axis is also apparent with the PPBA and PBBA derivatives.
As with the other inhibitors, Kendrew models of PPBA and
PBBA were fitted to the composite bounded electron density
projections and the atomic coordinates were determined

and are listed in Table A-1.

6. Orientation and interaction of phenyl alkyl

boronic acids in the active site

 

It is apparent from examination of the composite

bounded electron densities (Figure 38) that the orientation

145

,‘ . .
(e : Q
‘34 ; ‘9
© 1
. . i
PPBA (y) _ PPBA' (y)

 

PPBA (z) PPBA' (2)

FIGURE 38a. PPBA in Region of Active Site Viewed Down
y,z-Axis, Contoured at 0.15 eA'3, with
Approximate Two-Fold Representation

146

©

 

PBBA(y) PBBA] (y)
PBBA (z) PBBA' (z)

FIGURE 38b. PBBA in Region of Active Site Viewed Down
y,z-Axis, Contoured at 0.15 eR-B, with
Approximate Two-Fold Representation

147

.888 8.5.233 as. E as 8.8.. 2; as $8 at

A.« mpzumpozv <mma new «P e_aumpozv <mm¢ mo mcowpwmoa uwogucm an» :mmzumn cow mw>mu ugh "mpasmxm

:58 M 5
pmxuoa
umuxpmpmu

328 w 5
pmxuoa
«growewumam

as» to

$28 .A. 5
pmxuoa
owpapmpmu
:58 w 5

pmxuoa
xuwumewomam

 

 

 

 

 

.88 .88 88 88
«8.8- u we «8.8- h .2 .88 «8.8 u we «$8 n me 88
«8.8 u .2 «2.8 n .2. .88 «8.8- u E «8.0 n 3 88

 

 

 

 

 

 

 

 

Loewe «10-8 mcwppmpmxgu
mmpzumpo: pcmccmawuc« mcp mo comm cw cowuspwumazm Lopwnwch m>wuwamqeou co comvgoasou .n

 

 

 

 

.88 .88 .88 .88
«8.8- u we «38 n me 88 «8.8 h we «8.8 n me 88
c .
«8.? h E «8.7 n ..3 88 «2.8 u E «8.8- n m< 88

 

 

 

 

 

 

 

 

F mpzuwpoz sage cowpmw>mo ms» mcwmmgm>< an vwzoP—om

..F mpaumpoz cw cowuzpwpmaam mgu co nmELomgmm mwx< vpomiozh uaon< coppmpom mo comwgmasou .m

A<mmm ncm <mmmv .P van P mmpzumpoz mo mmpwm

m>wuo< on» cw «armcmo coppzuwpmnam mo mcopupmom avogucmu mo commgmasou

om m4m<h

148

of PPBA and PBBA are essentially the same in the
specificity region of molecule 1. The occupancy in
molecule 1 is lower than that observed with PEBA, but

is higher than that observed in molecule 1' where only
the PPBA density appears above the three sigma level.

In the catalytic region of the active site, significant
density is only present for the substitution in

molecule 1. The net result is that the binding orienta-
tions and interactions between the independent molecules
cannot be accurately compared.

It is interesting to note the similarity of both
the orientation and interaction of PPBA and PBBA in the
specificity site (Figure 39). Although initial examina-
tion of the Kendrew model of native a-CHT indicated that
side-chains longer than phenyl ethyl could possibly be
accommodated by the specificity site, comparison of
the coordinates of the bound inhibitor molecules
indicates that there is an apparent length limitation in
this region. The same number of van der Haals contacts.
within 3.5 A, between a-CHT and the aromatic group of
the inhibitor occur with PPBA and PBBA as occur with
PEBA, but the number of closer contacts increases markedly
(within 3.0 A-PPBA (l7), PBBA (l6), PEBA pH 7.3 (12).
PEBA pH 5.4 (9) and PEBA pH 3.6 (7)). The interactions
in the catalytic site involve van der Naals contacts

between the inhibitor alkyl chain and a-CHT and the

 

fl—I‘T ‘.

 

 

149

aewm aswec<

ea eaveeafiaca ~-x cw eaveaeeapco <mma tea <maa

Au
30.va c
+m,__..Oum.<®
<mm due 8
<m&&i@ N9.
.00 2
sum Mao
no
u G-
mzmau. no .
nvxu .
@l
saws

mos

.mm 5.23...

 

150

boronic acid group is oriented in a "weak" binding
orientation near SER l95 (m3.5 A). The alkyl chain,
instead of extending into the vacant area of the active
site and thereby minimizing conformational strain,
attempts to maintain an interaction between the boron
and SER l95 0y. The gradient peaks along the 214-219
chain lining the rear of the specificity pocket resemble
the changes observed in PEBA pH 3.6.

It would appear that the lower occupancy of PPBA
and PBBA results from a strain imposed upon the I'bi-
dentate"-like binding in the active site. An apparent
cooperativity exists between the specificity region and
the catalytic region of the active site. The specificity
of a-CHT is determined by its ability to interact
correctly in both regions. It would also appear that
the importance of this cooperativity is reflected in the
weak binding of PPBA and the weaker affinity towards
PBBA. An increase of the alkyl chain appears sufficient
to strain the interactions between the specificity and
catalytic regions and prevents higher occupancy in the

active site.

7. Substrate specificity of a-CHT

 

In attempting to discuss the high efficiency of
enzyme catalysis, it is necessary to examine the sources

and types of interactions which determine enzymic

151

specificity. The initial binding (complexing) of a
substrate or inhibitor which is specific for a given
enzyme must rely on the enzyme's ability to interact
selectively with a limited segment of a multi-component

pool 1 vivo. The serine protease enzyme class contains

 

members of different species origin with similar but
varying structures and specificities. Although each
member of this class exhibits a unique response to pH,
temperature and other external perturbations, the
reaction and interaction between the enzymes and their
substrates provides evidence of those forces most
critical to natural processes. Probably similar forces
enable a polypeptide chain to conform to a pre-determined
tertiary structure as it is produced on the ribosome and
form the basis of all biochemical processes. The
inhibitors used in this study provide a means to examine
the specificity requirements of a-CHT and to compare
results obtained from solution studies.

As indicated in Table 10, a-CHT exhibits specificity
with respect to both size and stereoisomerism, but
enzymatic activity exists over a wide range of structures
with varying efficiency. The reactivity can be evaluated
in terms of two parameters, the relative binding constant,
K

and the catalytic rate constant, k Various

m’ cat“

attempts have been made to quantitize the interaction

152

specificity, most notably using reactivity parameters,
steric fit, hydrophobicity, and molecular size consid-
erations (105,106). Based on the phenyl alkyl boronic
acid results obtained in this study, it would appear

from the cooperativity implied between the catalytic

and specificity regions of the active site that molecular
size is of prime concern within a homologous series.

This idea does not ignore the effects of reactivity,
steric effect and hydrophobicity. However, the limited
number of orientations and interactions in the binding
region suggests a selectivity based on volume, surface
area, potential van der Naals contacts and packing within
the binding region, in combination or separately.

Solution studies of a-CHT specificity have also
been performed using homologous substrate and inhibitor
series. Some of these, N-acetyl-L-amino acids esters
(107), phenyl alkyl-para-nitro phenyl esters (108),
alkyl boronic acids (65) and phenyl alkyl boronic acids
(64), are discussed below in terms of side-chain volume
and surface area.

The van der Haals volumes and surface areas of
various side-chains have been determined using the
approximate method of Bondi (109) (Table 21). That
portion of the side-chain which extends into the specifi-
city region of the active site (beyond Cu) is compared

in Figure 40. Because of the diversity of experimental

153

«.8.88888 .88 8_888

 

 

88.8 88.8 88.8 8.8 88.8 8.88 8PI888
8._ 8.8_ 8._ 8._ 88.8 8.88 .8888:
88.8 88.8 88.8 8.8 «_.8 8.88 8:888
88.8 88.8 88.8 8.8, 88.8 _.88 8:888
88.8 88._ 88.8 8.88 88.8 8.88 8:88
88.8 88.. 8.8 888 8_.8 8.8_ 8:8
8.8 8.8.8 _8 8.8 8.8 I
Pacx «_-8888 88 Facx «:88 mg 88 > 8
:=«N:ovg

 

8:8-w-=8-:z-w-8=8

 

.«m888v w «Mm .uom .Emgu .sme< .8 .cmm.mm 88:88 .Amgmpmm «Anyway 88888 8:858-4-_88888-z

888-8 88 888_888 88888-88?»

mmpmcpmazm .mgouwawccm mo mmpgmm

«N u4m<k

«8:888 8888:8888 8:8

888888058: mo 8888888580

.—

154

8.8.88888 .88 88888

 

 

88.8 88.8 88.8 8.88 888888-88
8.8 88.8 88.8 8.88 888888-88
88.8 88.8 88.8 8.88 888888-88
888 88.8 8.88 888-88

88 88.8 8.88 888-8

88.8 88.8 88.8 8.88 888888888888
88.8 8.88 88.8 8.88 888888888888
8.8 8.88 88.8 8.88 888888888888
88.8 88.8 88.8 8.88 888888888
88.8 88.8 8.8 8.88 8:8
888 8.8 8.8 I

88888 «8.8888 888 8 m8 888 88 888 > .8

8:88-8828

.mninu

.«08mpv N «Hm..::EEoo .mmm .mazmowm .Emsuomm ..8 pm xwmazo .mxmumm «AcmcqroLuwcumgmm .N

 

 

8.8.88888 .88 88888

155

«.8.88888 .88 88888

 

 

 

 

 

 

88.8 .8888 88.8 8.88 888888888888
88.8 .8888 88.8 8.88 888888888888
8.8 .88888 88.8 8.88 «8888888888
88.8 .888 88.8 8.88 «88888
.88 88

.8
88888 88.88888 s8 88 > .8

8:8 8888 888888-8888 h .8 888888-88888v-8888

.888888 88 «mm 8888888 8888 ..«8.mm.>8888=< .88888 8888888 88888 888888 .8
88.8 888.8 88.8 8.88 8888888
88.8 88.8 88.8 8.88 88888888
8.8 88.8 8.8 8.88 888
888 8.8 8.8 8

8
88888 88.8888 888 x «8 88V 88 «88 > .8

8.8.88888 .88 88888

x8 3888 838:8

 

h 8888 u 88

 

156

 

8.8 .888 88.8 8.88 8888888
8.8 .888 88.8 8.88 8888888
8.8 .888 88.8 8.88 8888888
88.8 .888 88.8 8.88 888888:
88.8 .88 88.8 8.88 8A888v=
88.8 .88 88.8 8.88 8888888
88.8 .88 8.8 8.8 88
8888 «8.88888 ”8 88 > .8
8
888 8888 888888: n .8 888888-8888888

 

 

.888888 88 am 8888888 8888 . «8.88 >88888< .88888 8888888 88888 .8
8.8.88888 .88 88888

157

 

 

1.0
.9
.e .
.7
.6
.5 .
-‘4
.53 . e
3‘ .2 .
.1
.o \ ' b

 

 

10 2‘0, 3‘0 .4‘0 86 60 7o 86——
VOLUME(cm3 mole")

 

 

1.0 . 1
1 k 'V
.9 . d\ \
.8
.7 .
.6.
.5 .

_.4 ' a
‘83 e
‘5 .2

.1 .
b
.o

 

 

 

3 4 5 6 7 8 9 i)
SURFACE AREAth cm2 mole")

FIGURE 40. Substrate/Inhibitor Side-Chain Specificity in
a- CHT

158

parameters and procedures, it is of particular interest
to note the bell-shape of these steric relationships.

The Optimal surface area (9:120 AZ/molecule) and volume

(~92 A3/molecule) occur with the same structures, and if
a spherical approximation is used for the inhibitor side-

chain, the approximate van der Haals radii are 3.1 A and
It appears that as the side chain

0

2.8 A, respectively.
is increased in size, it passes through an Optimal size
for interaction in the specificity region. This has

been observed with the contacts in TOS, PMS, PEBA (3.6,

5.4, 7.3) and PPBA and PBBA. Further expansion of the

side-chain would require either reorientation of the

group within the specificity region or would strain the
interactions within the catalytic site. If the inter-

actions in the catalytic site are weak by nature (i.e.,
van der Naals contacts and hydrogen bonding only). it

is conceivable that the multiple contacts in the

specificity region would govern the orientation and
conformation of the inhibitor. This apparently occurs

with PPBA and PBBA, and is suggestive of the similar
observations made recently with alkyl isocyanates (110)
Although there is a slight displacement of the phenyl

ring by about 0.3 A along the x-direction away from

it is apparent from the available free volume

SER 195.
of this region (l93-312 A3), that the aromatic group

does not reposition itself, but maintains apparently

159

optimalinteractions as observed with PEBA within the

specificity region. However, the alkyl chain of PPBA

is forced to reorient in an attempt to maximize its

interaction in the catalytic region. Thus the boronate

group is forced to extend past SER 195 OY’ which weakens
their interaction and accounts for the low occupancy.

This interpretation also explains the smaller occupancy

observed with PBBA in the catalytic region. These

results are indicative of the cooperativity which exists
between the specificity and catalytic regions of the
active site and indicate that the specificity pocket
establishes the initial orientation of the substrate

for subsequent interaction and catalysis at SER 195.

The results further confirm the hydrophobic nature

of the specificity of a-CHT as did the comparisons made

between alkyl and phenyl alkyl side-chains. The elec-

tronic factors involved in substrate specificity were

not specifically detailed in this study although further

analysis of the data of Dupaix gt_al. (108) indicates

complexities with halo-alkyl side-chains. A further

point of interest concerns the behavior of branched

alkyl side-chains. a-CHT appears to exhibit restrictive

catalytic behavior if extensive branching occurs at Ca

witfli‘the restriction decreasing for C8 branching. This

is most probably due to strained interactions in the

catalytic region of the active site, particularly the

 

. )1-Hfl1llls£!

160

inability of the apparently important hydrogen bond
with GLY 193 NH to form.

. ~8__._-___.f .8 44

V. DIAGONAL PLOT ANALYSIS

A. Analysis of the Distance Diagonal Plot of a-CHT

(Native, pH 3.6)

Because inhibitor binding affects the conformation
of the active site of native a-CHT and produces conforma-
tional changes in regions of the molecule removed from
the catalytic site, the structure of a-CHT (Native, pH

3.6) will be discussed in terms of a distance plot

representation. The distance diagonal plot comprises

one diagonal half (i 3 j or i g j) of a square matrix of
order N, where N is the number of amino acids in a poly-
peptide chain with the point (ij) indicating the

separation (in A) of alpha carbon atoms i and j (Figure

41). The plot of the native a-CHT structure is shown

0

in Figure 42, with all regions < 15 A darkened. Using

the alpha carbon positions for each of the independent

molecules of the a-CHT dimer, both molecules can be

represented in the same diagonal plot, one above and the

other below the diagonal. This permits convenient

comparisons to be made across the diagonal between the
independent monomers.

The inherent organization of a distance plot provides
a natural progression between sequential relationships

and secondary and tertiary structure. Closed segment

features include nearest-neighbor interactions (Ii - .1120).

161

 

162

 

r- P r.
91 ' (NESZZESESU
DISTANCE DIAGONAL pLor ' DIFFERENCE DISTANCE DIAGONAL
COORDINATES PLOT COORDINATES

FIGURE 41. Distance Diagonal Plot Definitions

163

 

, moleculei'
j:
50' 100' 180’ 200’ N.
(as .o g ". - . ._-
P2. ‘4' - . O‘ W in. -
st? "2+ . " ?‘ Jfl—aD
5O \ ' ‘ : z
1397 i - t4 :1 ~
688* 33, f . . .. -
E‘B. 4 ,v ° ‘ . ’ 100
Too ,. \.\ ; I . ,
8, l .I
.8 . , f v “’1 .
O'H- «g ~150
150 , ‘17 ‘. . 5;;

 

g ‘ .‘8
_ 18 ;. x,.§'14/ '12" g .

jllTIIIITIIIII

\/ so 100 150 200
j

molecu|e1

FIGURE 42. Distance Diagonal Plot of a-CHT Dimer Depicting
Representative Structural Features, Contoured
at 15 A (Solid Regions)

164

which occur at or along the diagonal, mid-range distance

interactions (Ii - .ll 1 W4) are slightly offset from the
diagonal, while open segment features involve long-range

interactions (|i - j| >N/4) and are located distant from

the criagonal. Although a-helices (broadened segments

along the diagonal) are nearest-neighbor interactions,
anti-parallel B-strands (extending perpendicular to the
Iiiagonal) are mid-range or long-range depending on the

length of the polypeptide chain segment connecting the

strands of the B-loop (Table 22) (111). Parallel B-strands
(features parallel to but offset from the diagonal) are

typically long-range. Table 22 summarizes these several

features as well as others of tertiary structure observed

in a-CHT. These examples are idealized as distortion

from these occurs in actual diagonal maps of proteins.

The mid-range features are primarily those discussed by
Kuntz (112).

Further examination of the distance diagonal plot

ofchHT shows the presence of long-range features

related by mirror planes of symmetry, oriented about 45°
to thecfiagonal and located distantly from it. These
offcfiagonal long-range interactions of a-CHT displaying

IMrrorsymmetry have not heretofore been discussed. The

reghwm are designated in numbered blocks in Figure 42

mm wfll be called: crosses (l), anti-crosses (2), "v's"

(3)andinverted "v's'I (4) and they arise from the

 

165

TABLE 22

Structural Features of a-CHT

Diagonal Plot

 

 

 

 

 

 

Backbone Theoretical Feature
Structure Conformation Feature (Figure 42)
l. a-helix 9
”(3322’ \
2. B-loop
:fi‘RM" X l,2,3,4,5,6,7,8
' 10.11
i C
{j \
PARALLEL
Multi-strand
Loops 'fCES"
.5“ \
EL.
5‘
.83 8x W3
Loop-loop
1:25: i) ‘5
1:: x ‘5-‘7
5. Loop-hel i x
uqrvb’i /A\ 18
88 v w

 

166

structural features shown schematically in Table 22.

The B-strand-B-strand interaction gives rise to mm

symmetry in the diagonal plot and the particular strands

involved in the interaction can be identified from the
intersection of the symmetry elements with the diagonal.

'The Inasitions of the intersections define the sequential

positjcni of the bend of the B-strand. For instance, the

crosses of Figure 42 arise from B-strands with bends
near (37-148) and (49-204), respectively (Table 23).
The intersection of the mirror plane of the "v's" and
inverted "v's" can be interpreted in the same manner.
Thus, the "v" of Figure 42 arises from the interaction
Of the carboxyl terminal helix of a-CHT with a B-strand
possessing a bend near residue 49 while the inverted "v"
arises from a similar interaction but with a bend near
97 and with the helix running in an opposite sequential

sense. It should be noted that these interactions are

all of an interdomain type.

The long-range interactions which appear as symme-
trical features in the region i 3_123, j < 122 reveal
the humrdomain contacts to involve beta structures.

It Muwld be apparent that by examining these diagonal
(flotscm a finer contour scale and extending them to
uneeafimensional surfaces, Observation of distance
grmfients (sharpness, etc.) would allow the calculation

ofdinmrtions from the idealized interactions of

167

TABLE 23
Intersection of Mirror Planes with Diagonal

of Distance Plot of a-CHT

 

 

 

 

 

 

RESIDUE fl) INTERCHANGE (245-11)
ASN 18 —————— 227 VAL
THR 37 “—"—'1""“l ------ 208 THR
ASN 50 —-— ._ '1 ——————— 195 SER
SER 76—1- — — 1 —————— 169 LYS
LEU 97 A - - — —————— 149 ALA
TYR 146]--— ~— -——— — — -—- 99 LEU
TRP 172--4---i -— —— ————— 73 GLN
GLY 197—- - -— ————— 48 ASN
THR 208 ____._ ————— 37 THR
SER 221 ————— 24 PRO

 

 

 

Palindromy observed in a-CHT

--------------- Predicted by interchange (245-N)

 

168

Table 22. It would thus be possible to approximate
separation angles between sheets and helices, sheet
strand separation, etc.

Nishikawa and Ooi (45) have noted that the distance
plot of a-CHT (Figure 42) possesses an approximate mirror
plane of symmetry perpendicular to the diagonal, which
separates the molecule into two similar structural
domains (domain A, residues 1-122 and domain 8, 123-245).
This qualitative feature based on a comparison of the
plot contoured at a 15 A level is indicative of the presence
of two segments of chain of approximately equal length
(122 residues) and conformation. Matthews (38) originally
noted that the main chain hydrogen bonding scheme of
a-CHT implied the presence of two approximately cylindrical
surfaces of anti-parallel pleated sheet structure.
Birktoft and Blow (49) later reported the details of these
cylindrical surfaces in terms of six adjacent anti-
parallel chains and four disulfide bridges. Their exam-
ination of this intramolecular symmetry led them to
conclude that a-CHT contained two structurally similar
halves in one segmented polypeptide chain, and suggested
this was the result of gene duplication (49). This type
of folding is not unique since Shotton gt al. (91) have
made similar observations with sequentially homologous
elastase. However, the domain structure in the distance

diagonal plot contradicts the possibility of gene

169

duplication, since the cylinders show greatest similarity
in structure when the sequence in one of the domains is
reversed. The algorithm for generating the corresponding
residue in domain 8 from domain A simply involves sub-
tracting the sequence number of the residue from the
length (245) of the peptide chain termed "interchange."
This observation suggests that the implication of gene
duplication is incorrect since the latter does not give
rise to a palindromic or interchanged polypeptide sequence.
Although sequence homologies do not exist between the two
domains, various structural features are observed to be
present in both.

Five interchain loops occur in a-CHT, three of which
are closed by disulfide bridges and two of which are

approximately palindromic to two others (Table 24).

TABLE 24

Intrachain Loop Palindromy

 

 

Loop Name Residues (N) Interchange (245-N)
Histidine 42-58* 203-187
Methionine 168-182* 78-63
Serine 190-220* 55-25
Aspartic Acid 84-110 161-135
Autolysis 133-164 81-112

where (*) denotes lOOp closure by disulfide bridge

 

170

It can be further noted that part of the specificity
pocket of the active site TRP 215-SER 217 is sequentially
palindromic with part of a secondary binding site of
a-CHT formed by a cluster of three tryptophan residues
(TRP 27, TRP 29 and TRP 209) (42).

The segments with mirror symmetry reveal several
interesting points concerning the inter-domain inter-
actions and the dimeric structure. Locating the residues
involved at the beta-bends (intersection points on the
diagonal) shows that they are on the surface of the
molecule, within approximately 5 A from one of two non-
crystallographic two-fold axes which are approximately
parallel to the a*-axis (Figure 43). The construction of
line segments connecting the alpha carbons of each pair
of these palindromically related residues in the three-
dimensional structure gives a series of lines with
colinear mid-points. A new intra-molecular dyad axis can
now be constructed passing approximately through the mid-
points of these line segments. This dyad passes through
the active site region and is perpendicular to the non—
crystallographic two-fold axes discussed above (Figure
43). There is only one pair of residues which does not
show this relationship (TYR 146-LEU 97). If a correction
is made to the TYR 146 position based on the position of
this residue in the zymogen (51), the pair then satisfies

the new dyad. Thus the a-CHT dimer can be approximated

171

NON-CRYSTALLOGRAPHIC TNO- FOLD AXES

 

>1.

 

 

. . PALINDROMIC THO-FOLD AXIS
V _‘ 2 V

Molecule 1' Molecule 1 ' .
x = CATALYTIC BINDING POCKET
y = SPECIFICITY POCKET
Z = SECONDARY BINDING SITE

 

 

 

 

. - '. , .
. O . .
53’A\88_\£N ,,l2£f' _8YNI
Molecule 1' Molecule 1

,/9- LOOP PAIRS-J

 

FIGURE 43. Domain Configuration of a-CHT Dimer:

172

by a tetramer comprised of domains A, B, A', and 8'.
Such a tetrameric organization of quaternary structure
might be a basic format observed in nature in other
macromolecular systems. The new dyad might exist to
define the region of intra-domain contact; it is inter-
esting that both the active site and the secondary site
(tryptophan cluster) (42) are situated around this dyad
(Figure 43). In the crystal, the domains of a-CHT occur
as alternative rows of A and B type, extending parallel
to the crystallographic c-axis.

An additional point of interest related to the domain
structure but which is not yet fully understood, centers
around the location of the ordered sulfate molecules of
a-CHT (identified by sulfate-selenate exchange experiments)
(94). Each of the sulfates can be assigned to a specific
amino acid residue according to proximity. Except for
CYS 1, all residues so indicated (Table 25) are in domain
8. CYS l-122 constitutes both termini of domain A and is
located on the intramolecular dyad.

The foregoing construction arises from the fact that
the bio-architecture of a-CHT relies on the presence and
interaction of structurally similar domains. It is
therefore of particular interest to study the correlation
of both the domain sub-structure with other members of a
particular enzyme class, and the domain conformational

changes which occur upon inhibitor binding or chemical

173

TABLE 25

Residues in Proximity of Localized Sulfate Ions (94)

 

 

Sulfate (Residue #) Domain Location
l95/195' - B, B'
1/1' Terminus of A domain:

covalently linked to
122, A'

149/149' 8, 8'

154/154' 8, B'

177 8

192/192' B, 8'

217/217' 8, 8'

236/236' 8, B'

 

174

modification with respect to variability in protomer
structure. Potentially, this can form a basis for
molecular recognition. Since domain structures occur

in other proteins and enzymes, the basic concepts
operative in the a-CHT system might, in fact, be of more

general application.

8. Comparison of the Structures of a-CHT with Pancreatic

 

Trypsin Inhibitor by_Use of Distance Diagonal Plots

 

Due to the similarity in the active site regions of
enzymes belonging to the serine protease class, it is of
interest to examine their three-dimensional structures
in an attempt to recognize common structural features
which might provide a basis for better understanding the
structure-function relationship of this enzyme class. It
would appear that the distance plot affords an efficient
means by which to catalog these structures for detailed
comparisons.

The domain of a-CHT centers around an observed beta-
barrel. The representation of this region in the distance
plot has been discussed above as a "multi-loop super-
helix" in a-CHT, but Kuntz (112) also assigned this feature
to extended segments of anti-parallel B-sheet structure.
It would be of particular interest to further evaluate
these conformations, as described above, in terms of the

packing of ideal sheet and helical structures to establish

175

whether these configurations are possibly formed as
functional segments of the polypeptide chain, or as
simple minima on conformational energy surfaces. This
aspect appears even more significant upon the examination
of the crystallographic structure of PTI.

PTI (101) is a natural inhibitor of trypsin found
in bovine pancreas as well as other organs. It consists
of 58 amino acid residues, with a molecular weight of
6500 amu. It is anomalously stable towards standard
denaturants, chemical agents and heat, and resists
proteolytic digestion, while being capable of inhibiting
either a-CHT or T in a stoichiometric (1:1) complex.
The three-dimensional structure of PTI has been determined
by Huber's group (101).

The folding of PTI is based on a twisted double-
stranded anti-parallel B-sheet from ALA 16 to GLY 36.
The N-terminal segment is in an extended conformation and
is anti-parallel to the first strand of the B-sheet. The
remaining C-terminal segment consists of a short extended
region (39-43) and three turns of a-helix, both of which
are anti-parallel to the second strand of sheet and the
N-terminal segment. Thus, the folding consists of four
anti-parallel and two parallel structural features with
about 12 and 10 residues between turns.

The overall folding of PTI is similar to half of
that of cylinder 1 of a-CHT. The latter is based on six

176

B-sheet features arranged in a fairly distorted cylindrical
fashion with about 12, 10, 15 and 22 residues between
turns. The resemblance can be seen from Figure 44 which
compares the diagonal distance map of a-CHT and with that
of PTI translated by 21 residues from the N-terminal to
obtain the best congruence in folding with a-CHT. This
suggests that the folding of an a-CHT cylinder might be
based upon and related to that of PTI. The closest
structural similarity occurs between the double-stranded
anti-parallel sheet of PTI (CYS 14-39) and the double-
stranded sheet of the HIS loop of a-CHT (CYS 42-58). The
principal difference between the two occurs in the
vicinity of and beyond residue 60 in a a-CHT (37 in PTI):
from Figure 44, it can be seen that the C-terminal a-helix
of PTI leads to a broader interaction perpendicular to
the diagonal at residue 60. In the folding of PTI, the
a-helix corresponds closely to the position of the fourth
strand of the a-CHT cylinder so that the PTI folding
resembles part of the cylindrical well. Removal of the
helix and duplication of the remainder of the chain

could conceivably lead to a cylinder.

The distance plot of PTI is shown in Figure 44, along
with that of a 58 amino acid segment of a-CHT which bears
a structural resemblance to PTI. It is apparent that this
segment of a-CHT which occurs in domain A (residue 24-81)

is quite similar to PTI when comparing alpha-carbon

 

 

2'15
1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FIGURE 44. Comparison of Distance Diagonal Plots Of
a-CHT and PTI [contouring program courtesy
of S. Ernst (18)].

178

separation distances although the comparison of their
respective amino acid sequences does not yield the same
conclusion (Table 26). The distance plots show similar
placement of anti-parallel sheet structure originating
at residues 37(221), 49(204) and 75(172) in a-CHT, and
10, 23 and 48 in PTI, as well as similar "helical"
thickening along the diagonal near the segment termini.
The major differences between domain A(24-81) and PTI
interestingly involve a region which exhibits variability
between domains A and B in native a-CHT. A possible
interpretation is that this region of the molecular con-
formation is more susceptible to evolutionary change, and
this is further buoyed by the comparison of this domain
with elastase as described (45). This variability would
not produce significant changes in the overall structure
of a-CHT as the residues involved occur on the surface of
the molecule, frequently with exposed side-chains. The
three segments approximate a single loop (class C) of the
multi-loop super helix. a-CHT also exhibits other "class"
examples as previously indicated, most notably the
approximately three strand loop beta structure (class B)
which encompasses the region 75(172) to 97(146).
Concerning the possible structure-function relationship
present in PTI, it should be noted that there are three
disulfide bridges within the 58 residue chain. It is

observed in a-CHT that only one disulfide appears in

 

179

 

 

 

 

 

«888V 888888888 souc8gmmm.m n A8v mmm h ON

88 <8< mm ZOO Pm 88 ON: ON 8<> Nm
O >8O 8m OOH ON 8 >8O ON 888 >8
8 >8O mm m>8 88 N8 <8: 8N zm< O8
8 m>O mm OOO O8 O8 m>8 ON OOO 88
88 8:8 88 mum 88 N8 <8< mN zm< O8
8 O8< mm 8mm O8 8 zm< 8N OOH 88
N 882 Nm mum m8 8 8>8 ON 388 O8
8 m>O _m >8O 88 8 8:8 NN mum 88
88 8m< om 28w 88 8 «>8 8N >8O 88
88 OOO O8 8m< N8 8 O8< ON >8O O8
8 <8< O8 uza 88 O m:« 88 m>O N8
88 mum 88 OOO O8 88 OOH m8 m:8 >8
8 m>8 O8 >8O 88 ON O8< 88 m8: O8
8 mza 88 <8< NO 8 <8< 88 m:8 88
O zm< 88 8<> 8O 8 m>8 m8 >8O mm
N zm< m8 8<> OO 8 m>O 88 8:8 88
m Om< N8 8<> 88 m O88 88 m>8 om
O8 m>8 _8 8m< 88 >8 >8O N8 8m< mm
88 <8< O8 mum mm 88 8:8 88 ZOO 88
O O8< 88 8:8 NO O «>8 O8 2:: mm
8 m>O Om 8:8 88 88 O88 8 8mm Nm
8 >8O 8m 8<> OO O O88 O 8<> «8
ON >8O Om >8O mm 88 OOO 8 ZOO om
8 8>8 mm m>O mm 8 2:: O 888 ON
O 8<> 8m m«: 88 8 m>O m O88 ON
8 8:8 mm <8< mm 88 8:8 8 888 8N
N8 8:» N8 <8< mm N8 8m< m mum ON
88 ZOO 88 8:8 88 N8 O88 N >8O ON
88 m>O Om 8<> mm 8 O8< . 8 O88 8N
8888888 8 8888888 8888888 8 8888888 888888: 8 8888888 8888888 8 888888:
88: «88 >:O-8 888m 888 8:O-8

 

88888 88:8888O 88:8888O yo pcwscmw>< :8 88888 < :88EOO pzuia 8:8 «88 co c88888gsou mucmzcmm

ON m4m<h

 

Vi...

180

domain A (CYS 42-CYS 58), while three appear in domain B
(136-30l; 168-182; and l9l-220), and two of the latter
are linked externally to the 58 residue segment. It is
possible that the disulfide links, themselves, are
responsible for the relative chemical stability of the
inhibitor molecule as compared to the serine proteases,
in general. This conjecture is supported by the observa-
tion that upon the reduction of the disulfide bond CYS 14-
38, the LYS lS-ALA l6 peptide bond is cleaved by T (ll4).
Based on the above observations, it is interesting
to hypothesize on the evolution of the class of serine
proteases and their natural inhibitors. It is possible
that the domains of a-CHT represent conservative struc-
tures present in all serine proteases (cylinders have
been identified in elastase) (9l) resulting from a
common evolutionary gene, and that the different speci-
ficities and three-dimensional structures are representa-
tive of presently stable gene mutations. The observation
of domains in other protein classes suggests that
proteins consist of several building blocks (domains),
which might have specific conformational properties or
functions associated with them. It is interesting to
note that several of the bacterial serine proteases (e.g.,

Staphylococcus aureus) are of a molecular weight about

 

l2,000-l3,000 daltons (approximately lOO-llo residues in

 

 

 

18]

length), consistent with the size of the o-CHT domain.
PTI is one-half this size (58 residues) which might
imply the presence of still smaller domains or domain
cores. The comparison with other members of this enzyme
class and its inhibitors should prove most useful in
establishing domain conservation. The functional
importance with respect to a-CHT:inhibitor interactions
is discussed below, examining both the reversible and
irreversible inhibitors used in this study.

One last hypothesis incorporates the concept of a
master macromolecule comprised of both the zymogen
(inactive enzyme) and its natural inhibitor. This, in
effect, would provide a bio-feedback mechanism by which
a single molecule, consisting of sub-segments (domains),
could be employed to generate the independent zymogen,
the active enzyme and/or its natural inhibitor by appro-
priate cleavage. This concept is particularly of interest
when one considers that the activation of the serine
proteases from their zymogens appears to function with
cleavage by other serine proteases--principally tryptic
or chymotryptic attack. The idea of a single macro-
molecule structured as an internally complete system would
further explain the observation of the large, apparently
non-allosteric enzyme structures necessary to maintain a

small, concerted active site configuration.

 

f? ‘8
A
I!

 

 

 

 

 

182

C. Perturbations to the Tertiary Structure Accompanying

 

Active Site Binding

 

The intermolecular cooperativity exhibited by the
tetrameric hemoglobin molecule gave early indications
of the manner in which protein transmits conformational
changes through its three-dimensional structure. In
describing his theory of proton fluctuation, J. G.
Kirkwood (88) stated:

Relatively distant groups may influence
local interactions of complementary sites on two
protein molecules by serving as sources or sinks
of mobile charges supplied or withdrawn from the
interacting sites.

And more recently, R. w. Lumry (89) has proposed the
notion of mobile defects:

Proteins in solution are in a constant state
of restless conformational fluctuation due to the
presence of 'mobile defects' resulting from the
evolution of imperfect bonding solutions. In
addition to high mobility in most regions of
protein, the mobile defects provide a store of
potential energy necessary for specific function.

Both of these ideas emphasize the importance of mid- and
long-range tertiary interactions and the dynamic nature

of the three-dimensional structure of an enzyme,
suggesting the ability to recognize and accommodate a
substrate or inhibitor molecule and transmit this informa-
tion through conformational changes. The group of
structurally/functionally related inhibitors used in this
study provides an excellent basis for the comparison of

such structural adaptability. This will be accomplished

183

using a new technique developed in the course of this
study, a distance plot of the difference electron
density of a derivative which will be referred to as a
difference diagonal plot. The construction of these
difference diagonal plots has been described in the
Methods section.

The difference diagonal plot considers the distance
of a peak in a difference electron density map from the
alpha-carbon atoms of residues i and j. The quantity
plotted in the difference diagonal plot is the sum of
the magnitudes of these distances (Figure 41b). The
derivative difference distance plot shows structural
changes on two levels. The primary level identifies the
difference electron density peaks associated with individ-
ual alpha-carbon atoms (or residues) as contours along
the diagonal. These represent the close approach of a
peak in the difference electron density to a residue
although, presently, positive and negative difference peaks
are not distinguished. The peaks which appear off-
diagonal are representative of changes occurring in the

tertiary structure at both mid-range (Ca. i = j)

1 - Ca'

J.
and long-range (i >j). Since the diagonal plot is
symmetrical about the diagonal, the difference diagonal
plot of the two monomers of o-CHT can be plotted on
either side of the diagonal. Thus, any asymmetry about

the diagonal represents an asymmetry in response of the

184

dimer to a perturbation. Some maps displaying the
application of this new approach are shown in Figure
45(a-e), from which it is immediately apparent that,

in general, mirror symmetry about the diagonal is not
present, thus showing the asymmetrical response of the
dimeric molecule. The difference peaks used in

generating the plots of Figure 46(a-e) are listed in
Appendix C. The asymmetry simply reflects the variability
in structure observed in the dimer and between domains A
and B.

The optimum correlation between the difference
electron density distance plot and the structure of the
native enzyme comes only after the distance diagonal
plot of the enzyme itself has been elucidated. The
detailed interpretation of the features of the three-
dimensional structure which have been found in the
distance diagonal plot of a-CHT (Figure 42) suggests the
capability of this technique. It would be of further
interest to determine the variability in interaction
that could be caused by rotation, twisting or bending
of adjacent regions of helices, B-sheets, random coils
or combinations thereof. As this mode of examination of
tertiary interactions of macromolecular structure is
still under development, the interpretations and compari-

sons reported here may be subject to some revision.

185

828

muo88 88888888 80:8808888

 

.88.».
QMMW 88.0.8mauou.“ .
w»...

88.8/WV _...dw: ..

par... 88.9
a.

fine”. a...

.00“

 

 

 

 

 

 

on—
.88
I no .0
w 1
I kw . OO—
8 88
-on
.0
do I.
I. o
l. o

 

 

 

 

 

 

 

.88 882888
mO8
.. .u a a. L
.. a 9. 8
v C 9 u n o
8.. 2.8 .8... m s.
. o
o a .9...
06 3 § up
. I 88.1 fill-III“.
. .w w" . v98 -2:
0 0” mod
“I O c o -
n 8. Aw . =.Auv. 888 .x
c 8. $8 a...” .8. .r..... 5.8., 8 A.
.r. 84.8 .8. .. . 9.... . .
.?3. to _D 87 . .
. r. .. 5 rs, . .8 my
8« 3. . 8. 8 o

186

mK In gun—

¢.m =8 <88.

o.m 18 (mm.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8888 m.8 :8

8.88 8.8

 

 

187

61

v... .8“...
888.. "a

Cat.

 

 

a“ 8... ..

 

“a
F

 

.C'OCI“

'6

 

 

 

 

 

 

8..

00—

on

188

.<mma

 

 

 

8 8...... .. . .
a to u
at .. .ma . -8...
m8 .8. .
..R.v .w I . 1
O .
8 S. -2.
.8 8 . . ®@ .
o 0 Co
. o u» .n J
8; .88
L8—
.8. . .
8 .8 .8 88....
.. .a .3
a V I M. 'v
6 0.0 .. .6
o .uo 8. ®_@ 3%. -
L . . r . .8 .

 

 

 

(mam

 

 

 

 

 

 

a .a . .
8.. .
18"
31¢. .8 Wauww O a»? a
.2. 8. L
. .
. . . a...
8 8 . .
v... v “.88 a” .6 10“-
a... q . $6 .8
av
100.
my a 8.. .... .
.8 .. 8. .. .. 8.
. a
.3
:9 ."f 0 a .. .88». Q
._.. n... 8.. c. .2 9a
. . a ..
.8 m.— ..H.“ .0 o J
r 8. . om. om. - m. o

189

8888 8.8-8-5...-2

 

 

 

 

can

0»...

oo—

 

 

Q a
. 88:67 .
:v. 0
Q.— a 3 a
.... 88 .
.mwmu. X 2.0 fiv .. &
.8 r8: 8.
. w
8. . . ..
9 o O
D. o
0.8 . om. 9.2! on

Am C mcaugétgé

 

/88. .88.. 8.8;.

C v7.7 09

8.. 8 .. ..

 

 

 

 

 

can

on.

190

The features of the difference diagonal plots to
be discussed below are denoted as l-lO and l'-l0' in the
distance diagonal plot of native a-CHT shown in Figure
46 and detailed in Table 27. These correspond to the
B-barrel (cylinder), B-sheet loops and B-sheet stacking
which were discussed in the previous section. Several
observations can be made from this comparison which
relate to the relative adaptability of the structural
facets represented by these features.

l. Features (l,l') and the domain-symmetry related
features (4,4'), are regions of the structure
which show the greatest tendency to change,
with (4,4') being generally more susceptible.
(l,l') appears to be more sensitive to pH
change than (4,4') due to the presence of
ILE l6, l6' and H15 40, 40'. The local two-
fold agreement appears to improve with pH in
(l,l') and with duality of interaction within
the catalytic region and specificity region of
the active site for (4,4'). Comparison of
(4,4') with PMS points out the variability
in specificity site dimensions discussed above
for the two independent molecules and might be
indicative of differences in specificity.

2. Features (2,2') and the domain-symmetry related

features (3,3') exhibit a greater fluctuation

191

molecule 1'

 

 

 

 

 

1oo _ '
150 L g5 ,.
6

2.“ {7, .
200 .. ”‘6,” M. ‘ 4“

iii—ll

H Ti

 

 

molecule1 . -
FIGURE 46. Identification of Diagonal Plot Features of

Native a-CHT to be Compared in Difference
Diagonal Plots

1532

It} . lyil‘ll!l‘.lll

 

8.u.NcouV .NN o8nu8

 

¢o\~u\mm Loom . 88am Loom

Loom Loom 58am voou coco uoou Loom nooo u8om-oz»

 

 

 

+N N 8 N 8 m m N N m 8 m .e
+N 8 N 8 N N N N N m m m e
_.u\mu\om u soon .8008. Loom Loom swam coca swan. LE.”— Loom Loom upouuox...
+N 8 8 N N _ N m N m . N N N .m
N 8 8 . 8 8 N m N N m N 8 m
Loon 58o; Loom Loom Loom L8ou voou 88mm gmmu nooo L8Nu unou-ozp
338: >588: 388.: >588: >588: :98: D83: 3,388 38: E38: :98: 208838833
. . m :8N20a
8\ou\mm Loon. . a goon .8008. soon. .. soon I u U pom—-03...
+8 8 8 , N m N 8 8 8 8 .N
8 N 8 8 8 . 8 8 N 8 8 N
8u\¢8\¢a a8ea N8a8 e8a8 e8~8 vooa e8~8 Looa eooa coca u8oa-oz8
+N . N N m N N . N N 8 N .8
N N N . N N N N 8 N N 8
.39.. 2.0... 5.2..— Lmau— soon .30.... Loan— Loan— Loom Upon—no}?
one: ago: \sa8eaz \ss8emz \ss8uaz NNsaa: \ss8emz \sa8uo: \Ngm88 \ss8uaz \238emz Neo883888mnam
< 58.8
88meo>o aN8-N N:N-N «NNN <NNN N.N 2a N.8 ma ¢.m :n ¢.m In o.m ma m=a m08 «Laueaa\e8asoo
omega>< .seoc-z -eaoe-z <NNN <mNa «NNN

 

88 u ocoz\unm84 8N - 5:8uu: 8n n xyoo: "meowpuoeouc8v
muo8m pacomo8o ou=N8m8a au8mcmo coeuum8u mucmgmwm8o mo m8ma8mc< 8a=v8>8u=8

mm u4m<h

1533

 

 

 

-\NN\- - - - - L8N8 - e8NN - - - - u8oa-oz8.
8 8 8 8 8 N 8 N 8 8 8 8 .o8
8 8 8 8 8 8 8 8 8 8 8 8 N8
-88N\- - - - - - - - g8NN - - - u8oe-oz8
8 8 8 8 8 8 8 8 N 8 8 8 .N
8 8 8 8 8 8 8 8 8 8 8 8 N
8N\Na\oa - - - - e8N8 N8NN L8N8 L8NN .8Na e8Na NOON N8oa-oz8
8 8 8 8 N . 8 . 8 N N N N .N
8 8 8 8 N N 8 N 8 N N N
NNN888oa . - - N8NN N8Na uooo N8NN e8NN L8N8 NOON L8N8 L8NN u8oa-oz8
+8 8 8 8 8 N N 8 N N N N .8
+8 8 8 8 8 N N 8 N N N N 8
NN8N888N L8N8 .8N8 - - L8Na N8N8. coca .8NN Noam L8N8 NOON N8oa-oz8
8 8 8 8 N N 8 N N N N .N
8 8 8 8 N N 8 N. N . N N N
NNNN8\NN .008 a8ua .8N8 N8oa a8Na NOON .8NN L888 coco . NOON goo. N8oa-oz8
8 8 N N N N _ 8 N N N N .N
8 8 N N N N N N N N 8 N
.8008. .8008. .886... .30.... .30“— mem .83... LG..— mem swan— Lmam Upouaozh
\Nsm88 \N8N8N \88N8N \88N88 \ss8naz \538uaz \ss8umx \ss8uaz Ne=8emz 85:8noz 85:8umz 8:0883N88NN3N
NN =8Ne°a
88Nea>o 8N8-N N:8-N <NNN <NNN N.8 :8 N.8 Ia. e.N :8 N.N za N.N 2a N28 N08 NL=8NNNNN8NEON
amoeo>< .saoe-z .seoe-z <NNN «NNN «NNN

 

8.u.8=ouv .8N m8aN8

194

in their variability with only light occupancy
in (2,2') and slightly heavier occupancy in
(3,3'), indicative of respective conformational
adaptability. In (2,2'), of domain A, there
again appears a greater pH susceptibility, now
due to the presence of ASP 102, while in (3,3')
the pH fluctuation appears to function coopera-
tively with the side-chain specificity difference.
Neither feature exhibits extremely good or
consistent two-fold correspondence.

The features (5,5'), (6,6'), (7,7') and (8,8')
are the tertiary interactions of the B-strands
(loops) between the two domains and along with
(9,9') and (l0,lO') reveal the conformational
adaptability in the inter-domain interface.

The interactions of a-helix and B-loop, (9,9')
and (l0,l0') (Table 22), show only slight if

any change upon inhibitor binding. The princi-
pal change observed occurs only with (l0,lD')
upon pH variation and only in molecule l'.

This adaptability is tempered (eliminated) upon
incorporation of PEBA.

Feature (5,5') represents conformation fluctua-
tion which includes the segmented loop
containing TYR l46, l46'. Although this feature

is somewhat complex, involving two interactive

195

features, one with ILE l6 and one with HIS 40
(Figure 19), perturbations to this feature

are combined in this discussion. These features
exhibit the best two-fold symmetry of the more
heavily changing features. There is no

apparent correlation other than a greater

degree of fluctuation occurring at pH 7.3 (with/
without PEBA), or at pH 3.6 where the orienta-
tion of the boronate implies a more direct
interaction than simple conformational adapta-
bility.

Features (6,6'), (7,7') and (8,8') reflect the
interactions between the loop involving LYS 203
and ILE l6, HIS 40 and ASN 48, respectively.

In general, these features exhibit fairly good
two-fold agreement, indicative of their generally
non-dimer interface involvement. (6,6') reveals
a greater density of adaptation as would be
expected when considering that the 203 loop
includes MET 192, SER l95 and both CYS l9l and
CYS 201. Feature (7,7') exhibits a similar
degree of adaptability and the best two-fold
agreement of any of the features examined.
Feature (8,8') reveals fairly good two-fold

symmetry between molecule 1 and l', but

196

differences in the intensity of the adaptability,
with pH influencing molecule 1 more than mole-
cule l'.

The availability of the difference electron density
maps of two competitive reversible inhibitors, N-formyl-z-
tryptophan and N-formyl-z-phenylalanine (unpublished
results of this laboratory), permitted an interesting but
tentative generalization to be made from an examination
of their difference electron density distance plots
(Figure 45). The implication of the map is that changes
which result from the occupation of the specificity region
of the active site, as with these two inhibitors, are
confined to the B domain of the structure. It is of
interest to note that the two distance plots are very
similar, generally differing in magnitude. The two-fold
symmetry of the distance plots is best in the (4,4')
region. The (3,3') features shows less change than with
inhibitors which interacted more directly with both the
catalytic and specificity regions of the active site.

Thus it becomes apparent that TYR l46, l46' is essentially
not involved in the interaction and its importance lies

in the catalytic triad interaction in the dimer inter-
face. Among the off-diagonal features, (5,5') exhibits
the same changes in both derivatives, but only a fair
degree of two-fold symmetry and is very light in substi-

tution. The (6,6') feature exhibits greater substitution,

197

two-fold symmetry and similarity between the two deriva—
tives.

Another comparison of changes in structure
accompanying the introduction of inhibitors or pH change
has been carried out using difference electron density
peaks within the active site region. Only those peaks
which occur within 1 8 A of the center of the active site
in the x and y directions, but extending across the local
two-fold axis separating the active sites of the dimer
in the z direction, were considered. The correlation
among those peaks is presented in Table 28, where further
delineation is made between the catalytic and the
specificity regions of the active site. Two peaks were
assumed to be equivalent if both appear above the three
sigma (difference electron density) level, both had the
same sign and were located within 1 A of one another.

It is apparent from examination of Table 28 that,
in general, the difference electron density peaks other
than the substitution are fairly conserved and reproducible
in terms of observation of combined effects of inhibitor
and pH change. This conservation is in agreement with
concepts of molecular recognition and conformational
adaptability. One intention of this work was to introduce
and explore these concepts by X-ray crystallographic

methods and studies in order to establish the feasibility

of such approaches. Although this work cannot expound

198

.88Npouv Nxmma 8\8¢ENNV Nxmma 8 u 8

.Nm>88m>8882 ucwgwmw8u 5°88 2OOO288N
=m8N NEON 828 me 888 new 8mgpocm mco 8o < 8 282883 Lzuuo 8N2» 88 scams op Omgmn8Ncou mgm 8.88888.O
O

 

 

 

 

- OO8 ON ON ON OO ON ON O N <NN8
NN - 88 NN 88 ON NN 88 O N <N88
NN NN - NN OO8 NN O O 88 O N.8 28
ON NO 8N - NN NN 8N 8N N N8 N.8 28 ONN8
NO NO OO8 NO - NO O O8 O8 8 O.N 28
ON N8 ON NN ON - ON ON N8 88 O.N 28 ONN8
ON ON O ON O ON - ON ON O8 N.N 28 ONN8
N8 O O N8 O ON ON - NN N N28
88 O 88 NN 88 ON 8N OO8 - N NO8
<NN8 ON88 N.8 28 N.8 28 O.N 28 O.N 28 N.N 28 N28 NO8 NNNO8 8
ONN8 <NN8 ONN8 8N88ON.M xv OO8O88 88888OOOO
- OO8 O OO8 O NN NN NN O N <NN8
N8 - NN N8 ON ON O NN O O <N88
O O - N8 NN O O N8 O 88 N.8 28
ON ON NN - NN 8N NN NN OO N8 N.8 28 <NN8
ON O OO8 OO - OO O ON O N O.N 28
ON ON NN 8N 88 - 88 ON ON N8 O.N 28 <NN8
OO ON O8 ON O ON - ON ON O8 N.N 28 <NN8
O OO ON ON O ON ON - OO8 N N28
NN ON O N8 O ON N8 N8 - N NO8
<NN8 <N88 N.8 28 N.8 28 O.N 28 O.N 28 N.N 28 N28 NO8 NNNO8 8
ONN8 ONN8 <NN8 8N8800.w x8 OO8OON 888O888OO8N

 

82o88288umn=m cosh gunpov Nxmm8 8u8N=mo coggum8m mucmsmm88o 80 88v =o8um8mggoo

.82O-O 8O OO8OON OO8N O>888< 828 88288:

mm m4m<8

199

with finality of ultimate results, it has established
new methods and techniques with which to examine concepts
of bio-molecular architecture utilizing specific enzyme

inhibitors and potential transition-state analogs.

10.

11.

12.

13.

14.

15.

REFERENCES

Hartley, 8 5., Nature 201, 1284 (1964).

 

 

Hartley, B. S. and Kauffman, D. L., Biochem. J. 101,
229 (196 ).

6
Meloun, B., Kluh, I., Kostka, V., Moravek, L.,
Prusik, Z., Vavécék, J., Keil, B. and Sorm, F.,
Biochim. Bigphys. Acta 130, 543 (1966).

 

Blow, D. M., Birktoft, J. J. and Hartley, B. 3.,
Nature 221, 337 (1969).

 

Blow, D. M., The Enzymes, P. D. Boyer, editor,
V01. III, 3rd. Edition, Academic Press, Chapter 6.

 

Kunitz, M. and Northrop, J. H., J. Gen. Physiol. 18,
433 (1935).

 

Bernal, J. 0., Fankuchen, I. and Perutz, M. F.,
Nature 141, 523 (1938).

 

Bergmann, M. and Fruton, J. S., Science 88, 306 (1936).

Bender, M. and Brubacher, L., Catalysis and Enzyme

Action, McGraw-Hill, New York, 1973.

Zerner, B. and Bender, M. L., J. Am. Chem. Soc. 88,
357 (1963).

 

Bender, M. L., Clement, G. E., Kezdy, F. J. and
Heck, H. d'A, J. Am. Chem. Soc. 88, 3680 (1964).

 

Sprinson, D. B. and Rittenberg, 0., Nature 167,
484 (1951).

 

Jansen, E. F., Nutting, M. D. F., Jang, R. and Balls,
A. K., J. Biol. Chem. 179, 189 (1949).

 

Shaffer, N. K., May, S. C. and Summerson, w. H.,
J. Biol. Chem. 202, 67 (1953).

 

Oosterbaan, R. A. and van Adrichem, M. E., Biochim.
Biophys. Acta 81, 423 (1958).

 

Baker, B. R., Design of Active-Site-Directed
Irreversible Enzyme Inhibitors, Wiley, New York,
1967.

 

 

”L

16.

17.

18.
19.
20.

21.

22.

23.

24.

25.

26.

27.

28.

29.
30.

201

Ray, W. J., Ruscica, J. J. and Koshland, D. E.,
J. Am. Chem. Soc. 88, 4739 (1960).

 

Schoellmann, G. and Shaw, E., Biochemistry 8, 252
(1963).

 

Unpublished results of this laboratory.
Henderson, R., Biochem. J. 124, 13 (1971).

 

Wright, C. S., Hess, G. P. and Blow, D. M.,
J. Mol. Biol. 88, 295 (1972).

 

Dayhoff, M. 0., Atlas of Protein Sequence and Struc-
ture, National Biomedical Research Foundation,
Washington, D. 0., Vol. 5, 1972, p. D-107.

Goodwin, T. W., Harris, J. I. and Hartley, B. 5.,
Structure and Activity of Enzymes, FEBS Symposium
#1, March 24, 1964, Academic Press.

Wright, C. S., Alden, R. A. and Kraut, J., Nature
221, 235 (1969).

Koshland, D. E., Strumeyer, D. H. and Ray, W. J.
Brookhaven Symp. Biol. 18, 101 (1962).

 

Schramm, H. J. and Lawson, W. B., Hoppe-Seyl. Z.
332, 97 (1963).

Lawson, W. B. and Schramm, H. J., Biochemistry 8,
377 (1965).

 

Taylor, R. P., Vatz, J. B. and Lumry, R., Biochemistry

 

1_2_, 2933 (1973).

Hartley, B. 5., Phil. Trans. Roy. Soc. London 8257,
77 (1970).

 

Horbett, T. A. and Teller, D. C., Biochemistry 18,
1349 (1973).

 

Gladner, J. A. and Neurath, H., J. Biol. Chem. 206,
911 (1954).

 

 

 

Rydon, H. N., Nature 182, 928 (1958).

Cohen, J. A., Oosterbaan, R. A., Jansz, H. S. and
Berends, F., J. Cell. Comp. Physiol. 88, Suppl. 1,
231 (1959).

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

202

Schneider, F., Hoppe-Seyl. Z. 332, 38 (1963).

 

Smith, E. L., Light, A. and Kimmel, J. R.,
Biochemical Society gymposia No. 21, The Structure
and Biosynthesis of Macromolecules, A61, Cambridge

 

University Press, p. 88.

Bernhard, S. A., Berger, A., Carter, J. H.,
Katchalski, E., Sela, M. and Shalitin, Y.
J. Am. Chem. Soc. 88, 2421 (1962).

 

Gutfreund, H. and Sturtevant, J., Biochem. J. 88,
656 (1956).

 

Fersht, A. R. and Sperling, J., J. Mol. Biol. 88,
137 (1973).

 

Tulinsky, A., Mani, N. V., Morimoto, C. N. and
Vandlen, R. L., Acta Cryst. B29, 1309 (1973).

 

Medical Research Council, Laboratory of Molecular
Biology. Cambridge, England.

Blow, D. M., Rossmann, M. G. and Jeffrey, B. A.,
J. Mol. Biol. 8, 65 (1964).

 

Matthews, B. W., Sigler, P. B., Henderson, R. and
Blow, D. M., Nature 214, 652 (1967).

 

Birktoft, J. J., Matthews, B. M. and Blow, D. M.
Biochem. Biophys. Res. Commun. 88, 131 (1969).

 

Sigler, P. B., Jeffrey, B. A., Matthews, B. W. and
Blow, D. M., J. Mol. Biol. 15, 175 (1966).

 

Department of Chemistry, Michigan State University,
East Lansing, Michigan.

Vandlen, R. L. and Tulinsky, A., Biochemistry 18,
4193 (1973).

 

Tulinsky, A., Vandlen, R. L., Morimoto, C. N.,
Mani, N. V. and Wright, L. H., Biochemistry 18,
4185 (1973).

 

Mavridis, A., Tulinsky, A. and Liebman, M. N.
Biochemistry 88, 3661 (1974).

 

Nishikawa, K., Ooi, T., Isogai, Y. and Saito, N.,
J. Phys. Soc. Japan 88, 1331 (1972).

 

45.
46.
47.

48.
49.

50.

51.
52.

53.

54.

55.

Commun. 88, 337 (1971).

203

Nishikawa, K. and Ooi, T., J. Theor. Biol. 88, 351
(1974).

 

Sigler, P. B., Blow, D. M., Matthews, 8. M. and
Henderson, R., J. Mol. Biol. 88, 143 (1968).

 

Steitz, T. A., Henderson, R. and Blow, D. M.
J. Mol. Biol. 88, 337 (1969).

 

Henderson, R., J. Mol. Biol. 88, 341 (1970).

 

Birktoft, J. J. and Blow, D. M., J. Mol. Biol. 88,
187 (1972).

 

Freer, S. T., Kraut, J., Robertus, J. D., Wright,
H. T. and Xuong, NgH, Biochemistry 8, 1997 (1970).

 

Wright, H. T., J. M01. Biol. 18, 1 (1973).

 

Wright, H. T., J. M01. Bio1. 18, 13 (1973).

 

Segal, D. M., Powers, J. C., Cohen, G. H., Davies,
?. R.)and Wilcox, P. E., Biochemistry 88, 3728
1971 .

 

Wright, C. S., Alden, R. A. and Kraut, J., Nature
221, 235 (1969).

Alden, R. A., Birktoft, J. J., Kraut, J., Robertus,
J. D. and Wright, C. 5., Biochem. Biophys. Res.

 

Robertus, J. D., Kraut, J., Alden, R. A. and
Birktoft, J. J., Biochemistry 88, 4293 (1972).

 

Drenth, J., Hol, W. G., Jansonius, J. N. and
Koekoek, R., Cold Spring Harbor Symposium on
Quantitative Biology, Vol. XXXVI, Cold Spring

 

Harbor Laboratory, 1971, p. 107.

Shotton, D. M., Watson, H. C., Phil. Trans. ng.
Soc. London B257, 111 (1969). '

 

 

Watson, H. C., Shotton, D. M., Cox, J. M. and
Muirhead, H., Nature 225, 806 (1970).

 

Stroud, R. M., Kay, L. M. and Dickerson, R. E.

Cold Spring Harbor Symposium on Quantitative

Biolo , Vol. XXXVI, Cold Spring Harbor Laboratory,
1971, p. 125.

56.
57.
58.
59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

204

Stroud, R. M., Kay, L. M. and Dickerson, R. E.
J. Mol. Biol. 88, 185 (1974).

 

Pauling, L., Chem. Eng8 News 88, 1375 (1946).

 

Wolfenden, R., Acc. Chem. Res. 5, 10 (1972).

Bender, M. L., J. Am. Chem. Soc. 88, 1626 (1951).

 

 

Thompson, R. C. and Blout, E. R., Biochemistry 88,
44 (1973).

 

International Tables for X-Ray Crystallography,
Vol. III, p. 202, Kynoch Press, Birmingham, T962.

Bell, R. P., Edwards, J. 0. and Jones, R. B., The
Chemistry of Boron and Its Compounds, Muetterties,
E. L., Editor, Wiley, New York, Chapter 4.

 

Rawn, J. D. and Lienhard, G. E., Biochemistry 88,
3124 (1974).

 

Lindquist, R. N. and Terry, C., Arch. Biochem.
Biophys. 160, 135 (1974).

 

Antonov, V. K., Ivanina,12 V., Berezin, I. V. and
Martinek, K., FEBS Letters 8, 23 (1970).

 

Antonov, V. K., Ivanina, T. V., Ivanova, A. G.,
Berezin, I. V., Levashov, A. V. and Martinek, K.
FEBS Letters 88, 37 (1972).

 

Koehler, K. A. and Lienhard, G. E., Biochemistry
88, 13 (1971).

 

Matthews, 0. A., Alden, R. A., Birktoft, J. J.,
Freer, S. T. and Kraut, J., J. Biol. Chem. 250,
7120 (1975).

Worthington Biochemical Corporation, Freehold,
New Jersey.

Schwert, G. W. and Kaufman, S., J. Biol. Chem.
190, 807 (1951).

para-toluene sulfonyl fluoride purchased from
Aldrich Chemical Company, Milwaukee, Wisconsin,
#ll,745-5.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.
82.

83.

84.
85.

86.

205

Department of Chemistry, Ohio State University,
Columbus, Ohio.

Department of Chemistry, University of New
Hampshire, Durham, New Hampshire.

Institute for Chemistry of Natural Products, USSR,
Academy of Science, Moscow, USSR.

Operating Manual, Leeds-Northrup general purpose
pH meter, Model 4711.

King, M. V., Acta Cryst. 8, 601 (1954).

 

Personal communication, Bud Suddath, Department of
Chemistry, Massachusetts Institute of Technology,
Cambridge, Massachusetts.

Hunt, P., Senior Honors Report, Department of
Chemistry, Michigan State University, East Lansing,
Michigan.

 

Vandlen, R. L. and Tulinsky, A., Acta Cryst. 827,
437 (1971).

 

Vandlen, R. L., Ph.D. Thesis, Department of
Chemistry, Michigan State University, East Lansing,
Michigan, 1973.

Henderson, R. and Moffat, J. K., Acta Cryst. 27,
1414 (1971).

Luzzati, V., Acta Cryst. 8, 142 (1953).

 

 

Ford, L. 0., Johnson, L. N., Macin, P. A.,
Phillips, 0. C. and Tjian, R.,J. Mol. Biol. 88,
349 (1974).

 

Phillips, 0. C., British Biochemistry, Past and

 

Present, T. W. Goodwin, Editor, Academic Press,

London, 1970, pp. 11-28.
Richards, F. M., J. Mol. Biol. 81, 225 (1968).

 

The Chemical Society, Tables of Interatomic Distances
and Conformations in Molecules and Ions, Special

 

Publicatibn #10, Burlington House, W.l, 1958.

Hass, B. S., Willoughby, T. V., Morimoto, C. N.,
Cullen, D. L. and Meyer, E. F., Acta Cryst. B31,
1225 (1975).

 

87.
88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

206

Diamond, R., Acta Crygt. 88, 253 (1966).

 

Kirkwood, J. G., "The Nature of the Forces Between
Protein Molecules in Solution," The Mechanism of
Enzyme Action, Johns Hopkins Press, Baltimore,

 

 

(1954).

Lumry, R., "Fundamental Problems in the Physical
Chemistry of Protein Behavior," Electron and Coupled
Energy Transfer in Biological 8ystems, Vol. I, Part
A, King and Klingenberg, Editors, Marcel Dekker, New
York, 1971.

 

Adams, M. J., Buehner, M., Chandrasekhar, K., Ford,
G. C., Hackert, M. L., Liljas, A., Rossmann, M. G.,
Smiley, I. E., Allison, W. S., Everse, J., Kaplan,

N. 0. and Taylor, 5., Proc. Nat. Acad. Sci. USA 88,
1968 (1972).

Shotton, D. M. White, N. J. and Watson, H. C.,
Cold Spring Harbor Symposium on Quantitative Biology,
Vol. 36,187 (1971).

Liljas, A. and Rossmann, M. 6., Annual Review of
Biochemistry 88, 475 (1974).

 

 

M. G. Rossmann, Purdue University, West Lafayette,
Indiana, personal communication.

Tulinsky, A. and Wright, L. H., J. Mol. Biol. 88,
47 (1973).

 

Cotton, F. A. and Wilkinson, 6., Advances in Inorganic
Chemistry, 2nd Edition, Interscience Publishers,

 

New York.

Lienhard, 0., Science 180, 13 (1973).

 

Schray, K. and Klunman, J. P., Biochem. Biophyg.
Res. Commun. 88, 641 (1974).

 

USA Q, 478 (1971).

Philipp, M. and Bender, M. L., Proc. Nat. Acad. Sci.

Nakatani, H., Hanai, K., Uehara, I. and Hiromi, K.,
J. Biochem. 77, 905 (1975).

 

Hess, G. P., Sybert, D., Lewis, A., Spoonhower, J.
and Cookingham, R., Science 189, 384 (1975).

 

101.

102.

103.

104.

105.

106.
107.

108.

109.

110.
111.

112.
113.

114.

207

Ruhlmann, A., Kukla, D., Schwager, P., Bartels,
K. and Huber, R., J. Mol. Biol. 11, 417 (1973).

 

Sweet, R. M., Wright, H. T., Janin, J., Chiotia,
C. H. and Blow, D. M., Biochemistry 18, 4212 (1974).

 

Blow, D. M., Wright, C. S., Kukla, D., Ruhlmann, A.,
8teig§mann, W. and Huber, R., J. Mol. Biol. 88, 137
1972 .

 

Huber, R., Kukla, D., Ruhlmann, A., Epp. 0., and
Formanek, H., Naturwissenschaften 5 , 389 (1970).

 

Yoshimoto, M. and Hansch, C., J. Med. Chem. 18, 71
(1976).

Simon, Z., Angew. Chem. Int. Ed. 18, 719 (1974).

 

 

Jones, J. B., Kunitake, T., Niemann, C. and Hein,
G. C., J. Am. Chem. Soc. 81, 8 (1965).

 

Dupaix, A., Bechet, J. and Roucous, C., Biochem.
Biophys. Res. Commun. 88, 464 (1970).

 

Bondi, A., Physical Properties of Molecular Crystals,
Liquids and Glasses, Wiley, New York, 1968.

 

 

Brown, W., Biochemistry 18, 5079 (1975).

 

Rossmann, M. G. and Liljas, A., J. Mol. Biol. 88,
177 (1974).

Kuntz, I. D., J. Am. Chem. Soc. 81, 15 (1975).

 

 

Dayhoff, M. 0., Atlas of Protein Sequence and
Structure, 1972, Vol. 5, National Biomedical Research

 

 

Foundation, Silver Springs, Maryland, 1972.

Kress, L. F. and Laskowski, M., J. Biol. Chem. 242,
4925 (1967).

 

APPENDICES

APPENDIX A
Coordinates of Inhibitors Determined in This Study

(a = 49.2 K, b = 67.2 K, c = 65.95 K, s = 101.8°)

208

209

8.2.ucouv 81< x8ucm88<

 

 

 

mwmm.o Nmom.o momm.o N8mm.o N8m~.o 8mmm.o m o
O8mN.o 888N.o NNNN.o 888m.o mo8m.o OO8o.o N o
NNNN.o ommm.o mmmo.o mmoO.o O88m.o 8~mo.o 8 o
8N8N.o 8omN.o NN8N.o Omwm.o ~8om.o mo8o.o m
o8ON.o N8m~.o NNNN.o 8mmm.o N8w~.o Oomo.o N8o
NNNN.o O8om.o O8om.o mumm.o No8~.o mmom.o N80
8NON.o 80mm.o mmom.o mm8m.o omem.o mOmm.o O o
momm.o momm.o mmom.o mmmm.o m88~.o omom.o m o
No8m.o Nomm.o Nmmo.o mNoO.o Omm~.o ommo.o N o
momm.o o¢om.o NNON.o ommm.o m8m~.o oevo.o 8 u
8NmN.o Ommm.o 8mmm.o O88m.o mON~.o N8Nm.o N o
N 8 x N a x «so: Eo8<
.8 m8zum8oz 8 8838802
NO8

8.N.808 . N .m NO.NN . O .m N.8N . O .m N.OO u 88
8OOON N828 O8 OOO828888O N8O88O8OO8 88 NOOOO8O8OOO

814 xmozmma<

210

8.O.NOOOO 8-2 x8OOOO8<

 

 

 

NN8N.o mm8m.o O8NN.o 8NOO.o OO8N.o N8ON.o m o
88NN.o N8NN.o 8oo8.o N8o2.o 80om.o N888.o N o
8mmm.o NNON.o 8N08.o OO8m.o NowN.o o8o8.o 8 o
omNN.o NOON.o m8mo.o NNON.o NOON.o NOON.o m
8NNN.o NmNN.o NNNN.o Ommm.o oeom.o N8NN.o <0
ONNN.o mmom.o ON8N.o 8Nom.o NNNN.o NNNN.o N80
oONN.o NO8m.o comm.o oomm.o 8NNN.O NOom.o m8u
ONON.o mmmm.o 88ON.o 888m.o NNNN.o mmmm.o O o
m8NN.o wOmm.o ON8N.o momm.o NONNOO NNoN.o N u
N8Nm.o Om8m.o ONNN.o 8mmm.o 8NNN.o N8NN.o N u
NNNN.o mmom.o ammo.o NNNN.o 8ONN.o m8mm.o 8 u
N a x N 8 x 8582 sou<
.8 m8zum8oz 8 m82um8oz
mz8

8.O.OOOOV 8-2 x8OOO88O

211

8.0.00000 8-2 x8O2888<

 

 

 

000N.o 88NN.o NN80.0 0N80.o 8N8N.o 0N88.o N o
oNON.o NNNN.o ONO0.o NN8N.o NONN.o NNNO.o 8 o
08mm.o 8N8N.o N080.o OONN.o N88N.o 8NNO.o N
NNNN.o 08NN.o N8mm.o NNNN.0 8NNN.o NN80.o N0
88NN.o 8N8N.o 8NNN.o 8NNN.o NNON.o ONO0.o <0
NNNN.o 888N.o N800.o ome.o O08N.o NNNO.o N80
NNNN.o NNNN.o NNNN.o 8NON.o NONN.o 8N00.o N80
8NON.o omON.o mme.o O88N.o ONNN.o N8mm.o O 0
o8NN.o 08NN.0 N080.o 8NNN.0 NNNN.o wwmm.o N 0
OONN.o NNON.o NON0.o O8NN.0 8NNN.o N000.o N 0
O80m.o NNNN.o ONNO.o NowN.o 08NN.o NNNO.o 8 0
N O x N 8 x mENz sow<
.8 m8zom8oz 8 m8=om8oz
N.N 28 <NN8

8.2.82OOV 8-< x8ucmaa<

212

8.O.8OOOO 8-< x8OOO882

 

 

 

N000.0 0O0N.0 0080.0 NNON.0 N08N.0 0ON8.0 N 0
0O00.0 O80N.0 0808.0 ON80.0 000N.0 0008.0 N 0
0N00.0 0O0N.0 0000.0 0080.0 ONON.0 8008.0 8 0
NN80.0 O88N.0 0080.0 8NON.0 880N.0 N008.0 0
0080.0 O00N.0 0800.0 0800.0 0N8N.0 ON00.0 00
0000.0 0O0N.0 O000.0 000N.0 ONON.0 0000.0 <0
0NN0.0 0O8N.0 0O80.0 N80N.0 O00N.0 NNN0.0 N80
00N0.0 NONN.0 8000.0 0NON.0 00ON.0 0000.0 N80
8NO0.0 O0ON.0 0O00.0 NN8N.0 0NON.0 0000.0 O 0
0000.0 0NON.0 8000.0 O000.0 OO0N.0 0O00.0 N 0
O000.0 08NN.0 00N0.0 8800.0 880N.0 0080.0 N 0
ON00.0 8880.0 8NN0.0 080N.0 800N.0 0NN0.0 8 0
N 8 x N 8 x mENz sou<
.8 0828802 8 0820080:
0.0 :8 <008

8.O_OOO80 8-2 x8O8888<

213

8.O.OOO80 8-8 x8OOO888

 

 

 

0O00.0 008N.0 0000.0 0080.0 08N0.0 8N00.0 N 0
0080.0 0NON.0 0O80.0 0000.0 O000.0 8N88.0 N 0
ONO0.0 NO0N.0 N000.0 O800.0 80NN.0 NO80.0 8 0
0N00.0 N80N.0 8880.0 8000.0 0880.0 8N00.0 0
8N00.0 800N.0 00O0.0 0000.0 0NON.0 0O00.0 00
0000.0 0N00.0 0000.0 8000.0 880N.0 8000.0 <0
N800.0 08NN.0 N8N0.0 0000.0 0O0N.0 08N0.0 N80
ON00.0 O800.0 0000.0 0N00.0 O0ON.0 0000.0 080
8000.0 0000.0 0880.0 8080.0 00ON.0 OO00.0 O 0
0000.0 OONN.0 0N00.0 O000.0 000N.0 8800.0 0 0
ON00.0 0N80.0 0800.0 0800.0 088N.0 0000.0 N 0
0O00.0 0O8N.0 88N0.0 0080.0 008N.0 0080.0 8 0
N 8 x N 8 x msmz sou<
.8 083080: 8 083080:
0.8 :8 <008

8.O_OOO80 8-8 x8O2888<

Appendix A-1 (cont'd.)

214

PPBA

 

 

 

 

 

Molecule 1
Atom Name x y 2
C 1 0.6003 0.2711 0.3727
C 2 0.5725 0.2707 0.3621
C 3 0.5578 0.2527 0.3600
C 4 0.5700 0.2356 0.3699
CP3 0.5971 0.2364 0.3819
CP2 0.6126 0.2539 0.3824
CA 0.6161 0.2903 0.3750
CB 0.6467 0.2862 0.3745
CC 0.6623 0.3061 0.3745
8 0.6938 0.3019 0.3748
0 1 0.7010 0.3099 0.3557
0 2 0.7111 0.3117 0.3931
0 3 0.6988 0.2802 0.3759
PBBA
Mo1ecu1e 1
Atom Name x y z
' C 1 0.6017 0.2697 0.3691
C 2 0.5748 0.2672 0.3574
C 3 0.5610 0.2491 0.3580
C 4 0.5730 0.2341 0.3717
CP3 0.5990 0.2373 0.3847
CP2‘ 0.6139 0.2546 0.3826
CA 0.6166 0.2891 0.3685
c3 0.6479 0.2852 0.3705
CC 0.6644 0.3039 0.3789
co 0.6835 0.3100 0.3641
8 0.7142 0.3040 0.3739
0 1 0.7265 0.2935 0.3582
0 2 0.7145 0.2907 0.3918
0 3 0.7306 0.3221 0.3808

 

215

C5 ATOM LABELS

C4 (PBBA shown)

  

216

APPENDIX A-2
Bond Distances and Bond Ang1es of Inhibitors

Determined in This Study

As described in the text, the deviations in bond distances and
bond ang1es for the inhibitors determined in this study are very

simi1ar to idea1 va1ues, as summarized be1ow:

1) Carbon-Carbon Distances

 

14010053

 

 

 

 

 

<r>aromatic =

<r>pheny1~a1ky1 = 1.50 :_0.05 A

<r>alkyl = 1.54 :_o.05 A
2) Carbon-Su1fur Distances

<r>c_S = 1.83 :_o.05 A
3) Carbon-Boron Distances

<r>C_B = 1.57 :_0.05 A
4) Oxygen Distances

<r>s_0 = 1.47 :_o.05 A

<r>B_0 = 1.48 :_o.05 A
5) Aromatic Ring Bond Ang1es

<e> = 120 :_1°
6) A1ky1 Chain Bond Ang1es

<e>pheny1-a1ky1 120 i-1

= 109.45 :_1°

<e>a1ky1

217

7) Boronate/Su1fonate Bond Ang1es

 

<e>-B(OH)2 120 :1

3
8) Deviations from P1anarity of Pheny] Rings

 

<Ar> = :_0.02 A from 1east-squares p1ane

APPENDIX B
Ca1cu1ated van der Naa1s Contacts of T05, PMS, TOS (MRC),
PEBA pH 3.6, PEBA pH 5.4, PEBA pH 7.3, PPBA and PBBA

218

219

0.0.00000

 

 

00.0080 000 --- <0
--- --- _u
A8.mv o 800 Am.mv o ¢_N
00.00 080 080 Ao.m0 880 080
“0.00 z 008 Am.m0 z 00. N0
08.00 50 .Am.mv z 080 A~.mv z 080
Am.mv o .Ae.mv 0 080 Am.mv 0 0.0
Am.m0 o 800 A¢.m0 o 800 mu
--- 00.00 0 808 000
0,.00 z .Am.~0 50 08m AN.m0 z .Ao.mv 50 08m mag
00.00 2 .Am.Nv do 080 Am.mv u .00.N0 z .Am.Nv au e_N
Am.mv o .AN.mV 0 08m Am.m0 u 080 80
Ao.m0 z 0_~
00.00 0 .Am.~0 z .Am._0 50 0.0
--- 80.00 0 .A~.mv 0 080 mu
8< 000000800 Eo0< 0000000 0 0008000 8< 000000800 so0< 0000000 0 0008000 Eo0< 000808008

020

< x80=000< 0°00 mmpmcwugoou 000800000 8000 0:0 0080

008

0000808000 0808000>0000 000 8:010 0>80mz 0003000 00000000 08003 000 00> 0000800800

810 x002000<

 

1L

220

 

00.00 80 .A¢.mv z

 

000
00.00 2 808
a
8 08.00 0 .08.00 0 .08.~vaz 008
80.00 00 .Am.80 0 mm_ 00.00 0 .Ao.mv 0 008
00.00 2 mm8 00.00 0 800 m0
. 88
A0 00 0 080
._o
8 00.00 0 8
.Am.NV 00 .88.00 0 .8m.80 z 008 88.00 800 080
08.00 0 .80.00 0 80.00 m .Ao.80 0
6
.Am.00 0 .A_.mv 00 .80.00 2 808 .80.00 00 .Am.80 o .Am.80 z 008
d
80.00 0 .A¢.mv z 008 00.00 0 .00.00 0 .Ao.00 z 008 00
00.00 z 80_
00.00 2 mm_ 8
00.00 0 .Am.00 50 .Ao.mv z 008 00.00 00 .Am.~0 o 008
0
00.00 0 .Am.~v 0 _08 08.00 N z 80 80
d .6
00.00 80 00.00 80
.32 00 .800 z .223 o 02 .843 00 .300 z .300 o 02 m
8< 000000800 500< 0000000 0 0008000 8< 000000800 Eo0< 0000000 0 0008000 Eo0< 000808000

020

008

0.0.00000

2121

0.0.00000

 

 

00.00 2 008 00.00 2 000 -
00.00 z 008 00.00 0 800 00.00 80 .00.00 00 .00.00 z 008 00
00.00 0 080 - 00.00 0 080 00
00.00 2 080 00.00 0 808 00.00 0 080 80
- - 00.00 2. 000
00.00 60 .00.00 2 080 - 00.00 0 000
00.00 00 .00.00 0 .00.00 0 080 - 00.00 0 080
00.00 0 000 00.00 880 000 00.00 080 000 00
00.00 60 .00.00 2 000 08.00 50 .00.00 2 000 -
00.00 0 .00.00 0 080 00.00 0 000 00.00 50 .00.00 2 080
00.00 00 008 00.00 80 .00.00 00 008 00.00 0 .00.00 0 000 00
- 00.00 00 .00.00 00 .00.00 2 008 -
00.00 0 .00.00 0 800 00.00 60 .00.00 0 .00.00 0 008 - 000
- 00.00 80 000 - .
- 00.00 0 .00.00 2 000 -
- 00.00 50 080 -
08.00 00 080 00.00 2 .00.00 00 008 -
00.00 .00 .00.00 0 .00.00 0 808 00.00 60 .00.00 0 008 . 00 00 .00 000
00.00 0 .00.00 0 000 08.00 0 000 00.00 2 .00.00 00 080 000
. 00.00 .00 .00.00 2 080 00.00 50 000 00.00 2 000
00.00 00 .08.00 0 .00.00 0 008 00.00 00 .00.00 0 .00.00 0 008 00.00 00 .00.00 2 000 00
8 000000800 0 000000000 8 000000.800 0 00080000 2 000000.800 .00 0008000 03¢
a 00 0000000 0 00 0000:00 0000 0000000 , . 000000000
0.0 :0 0000 0.0 :0 0000 0.0 :0 <000

 

2122

 

00.00 00 .00.00 00

00.00 000
.00.00 00 .00.00 00 .00.00 0

 

000
.00.00 00 .00.00 0 .00.00 0 000 00.00 0 .00.00 0 000
00.00 0 .00.00 2 000 00.00 00 00.00 00 .00.00 00 000
00.00 0 000 .00.00 0 .00.00 0 .00.00 0 000 00.00 0 000 00
00.00 00 00.00 00
. 0 . . 0 . . . . 0 . . 0 . . . 0 . . 0 . . 0
.00 00 0 00 00 0 00 00 z 000 00 00 0 00 00 0 00 00 z 000 00 00 0 00 00 0 00 00 0 000
00.00 000 00 00.00 002 00 00.00 000 .00.00 000 00 00
00.00 000 000 00.00 00 000 - 00
r a m 3.3»:
00.00 0 .00.00 0 .00.00 0 .00.00 0 000
.00.00 00 .00.00 00 .00.00 2 000 00.00 2 000 00.00 00 .00.00 00 000 0
0 - 00.00 0000 .00.00 2 000 -
00.00 0 .00.00 0 000 00.00 0 000 -
0 300300000 0 05303. 0 3:30:00 0. 0:303. 0 3:30.50 .00 203030 .82
0 00 0000000 0 00 0000000 0 00 0000000 000000000
0.0 =0 0000 0.0 00 0000 0.0 :0 <000

 

0.0.00000

223

0.0.00000

 

 

00.00 2 000 00.00 2 000
00.00 0 000 00.00 0 000 00
d a a o a a c
00.00 0 00 00 z 000 00 00 0 00 00 z 000
0 00.00 0 000 0 00.00 0 000 00
00.00 0 .00.00 0 000 00.00 0 .00.00 2 .00.00 0 000
a o o 6 a a a o
00.00 0 00 00 0 000 00 00 0 00 0000 00 00 0 000 00
00
00.00 0 .00.00 2 .00.00 0 000 00.00 0 .00.00 2 000
00.00 0 .00.00 0 000 00.00 0 .00.00 0 000 00
.6
00.00 0 000 -
00.00 0 000 00.00 2 .00.00 0 .00.00 0 000
00.00 00 000 00.00 00 .00.00 0 000 000
6 .00
00.00 0 000 0 00.00 0 000
00.00 2 000 00.00 0 .00.00 2 000
00.00 00 00.00 00
d
.00.00 00 .00.00 0 .00.0000 000 .00.00 0 .00.00 0 .00.0000 000 000
00.00 0 000 00.00 0 000
00.00 00 00.00 00
.00.00 00 .00.00 0 .00.00 0 000 .00.00 00 .00.00 0 .00.00 0 000 00
< 00:0000o Eo0< uumucou 0 0300000 < mocmpmmo Eop< pompcoo 0 m=u0mmm sou< 000000200
00 0 0o 0

<mmm

<mmm

 

0 00000000 0000 00000000000 000000000 0000 000

(man— Ucm ._.IQI6 m>_.u.mz :mmzvmn mvumwcoo mpwmz Lmfi cw> kum 200—00

224

 

00.00 00 .00.00 00 .00.00 00

00.00 0000 .00.00 00 .00.00 2

 

000
.00.00 0 .00.00 0 .00.00 2 000 00.00 00 .00.00 0 .00.00 2 000
00.00 0 .00.00 0 .00.00 2 000 00.00 00 .00.00 0 .00.00 2 000
00.00 0 .00.00 0 000 00.00 0 000
m 00
00.00 0 .00.00 0 00 00.00 0 000 00
00.00 00 .00.00 00 000
00.00 00 00 00.00 00 .00.00 000 000 00
00.00 00 .00.00 2 000 00.00 >0 .00.00 00 .00.00 00
00.00 0 000 .00.00 0 .00.00 0 .00.00 2 000
00.00 00 .00.00 2 000 00.00 0 000
00.00 0 000 00.00 002 00 00
00.00 00
.00.00 00 .00.00 00 .00.00 2 000
00.00 00 00.00 0 .00 00 z 000
.00.00 00 .00.00 00 .00.00 2 000 00.00 2 000 0
00.00 0000
.00.00 00 .00.00 00 .00.00 2 000
00.00 002 00 - 00
00.00 00 .00.00 00 000 00.00 00 .00.00 00 .00.00 2 000 00
00.00 2 000 00.00 2 000 00
0m m000000av sop< 0000000 0 0300000 0< 00:0000o0 Eop< panacea 0 0300000 sou< 000000500

<mmm

<maa

 

0.0.00000

225

APPENDIX B-2
Ca1cu1ated van der Naais Contacts Between a-CHT (MRC)
and Tosy) Inhibitor (MRC Resuits) (49)

 

 

Inhibitor Atom Residue # Contact Atom (Distance 3)
cs 216 ca (3.6), N (3.9), c (4.0)
217 o (4.0), N (4.0)
C4 215 c (3.8)
216 N (3.5), ca (3.8)
CP3 216 N (4.0)
cpz 191 c (3.8), o (3 9)
192 N (4.0) .
195 oY (3.9)
c3 215 c (3.8)
oz 191 o (3.7)
195 oY (3.0)
213 CY] (3.7)
C] 191 0 (3.4), c (3.9)
195 (33 (3.9), oY (2.6)
s 191 o (3.8)
195 N (3.5), Ca (3.5), cB (2.5),
0Y (1.5)
051 191 c (3 2), o (2.7), N (3.3)
192 ca (3.2)
195 oY (2.8)
0E2 191 o (2.9)
192 ‘ ca (3.3), N (3.1)
193 N (3.5)
194 N (2.9)
195 ca (3.5), cB (2.9), oY (2.5)
oY 194 N (2.6)

195 ca (2.4), CB (1.8), 0Y (0.8)

 

APPENDIX C
Coordinates of Difference Fourier Peaks Used in

Calculation of Difference Diagonal Plots

(6 = 49.2 A, b = 67.2 A, c = 65.95 A, B = 101.8°)

226

227

APPENDIX C
Coordinates of Difference Fourier Peaks Used in

Calculation of Difference Diagonal Plots

(a = 49.2 K, b = 67.2 K, c = 65.95 K, s = 101.8°)

 

 

TOS
Peak Height* Derivative x y z
-9 TOS 0.197 0.716 0.000
-9 TOS 0.328 0.333 0.650
-9 TOS 0.552 0.333 0.500
-9 TOS 0.684 0.366 0.350
-9 TOS 0.697 0.266 0.600
9 TOS 0.802 0.216 0.000
12 TOS 0.815 0.366 -0.016
-8 TOS 0.302 -0.050 0.333
8 TOS 0.460 0.416 0.600
8 TOS 0.500 0.266 0.400
-8 TOS 0.526 0.316 0.433
-8 TOS 0.552 0.250 0.400
-8 TOS 0.578 0.250 0.333
-8 TOS 0.605 0.350 0.583
—8 TOS 0.697 0.450 0.666
-8 TOS 0.828 0.283 0.433
8 TOS 0.881 0.283 0.400
8 TOS 0.907 -0.033 0.266
-8 TOS 0.947 -0.033 0.433
8 TOS 1.000 0.500 -0.050

 

Appendix C (cont'd.)

Appendix C (cont'd.)

228

 

 

PMS
Peak Height* Derivative x y z
10 PMS 0.118 0.783 0.600
-10 PMS 0.157 0.783 0.566
10 PMS 0.250 0.800 0.616
10 PMS 0.276 0.616 0.366
10 PMS 0.289 0.816 0.650
-10 PMS 0.289 0.783 0.466
11 PMS 0.302 0.633 0.666
-10 PMS 0.315 -0.050 0.333
-14 PMS 0.447 0.750 0.600
10 PMS 0.473 0.666 0.916
10 PMS 0.526 0.166 0.083
-12 PMS 0.526 0.316 0.450
-14 PMS 0.552 0.250 0.400
10 PMS 0.684 0.366 0.350
11 PMS 0.697 0.133 0.333
-10 PMS 0.697 0.450 0.666
-10 PMS 0.710 0.316 0.350
-10 PMS 0.710 0.283 0.533
10 PMS 0.723 0.116 0.633
10 PMS 0.750 0.300 0.383
12 PMS 0.815 0.366 -0.016
-10 PMS 0.842 0.283 0.433
10 PMS 0.881 0.283 0.400

 

Appendix C (cont'd.)

Appendix C (cont'd.)

229

 

 

PEBA pH 3.6
Peak Height* Derivative x y z

9 PEBA 0.592 0.366 0.383
-11 PEBA 0.605 0.300 0.416
-9 PEBA 0.644 0.266 0.516
-9 PEBA 0.697 0.283 0.516
-9 PEBA 0.750 0.283 0.466
9 PEBA 0.763 0.283 0.550
9 PEBA 0.789 0.333 0.500
10 PEBA 0.802 0.216 0.000
10 PEBA 0.815 0.383 -0.016
8 PEBA 0.236 0.266 0.266
8 PEBA 0.355 0.050 0.433
8 PEBA 0.381 0.650 0.633
-8 PEBA 0.434 0.233 0.433
-8 PEBA 0.434 0.350 0.433
-8 PEBA 0.434 0.016 0.766
-8 PEBA 0.565 0.250 0.333
-8 PEBA 0.565 0.516 0.233
8 PEBA 0.618 0.150 0.366
-8 PEBA 0.631 0.300 0.416
8 PEBA 0.657 0.450 0.566
-8 PEBA 0.684 0.450 0.666
-8 PEBA 0.710 0.266 0.366
8 PEBA 0.723 0.316 0.416
-8 PEBA 0.736 0.300 0.500
-8 PEBA 0.802 0.116 0.266

 

Appendix C (cont'd.)

Appendix C (cont'd.)

230

 

 

PEBA pH 5.4
Peak Height* Derivative x y z
-13 PEBA 0.171 0.166 0.383
12 PEBA 0.171 -0.050 0.500
14 PEBA 0.315 0.433 0.566
12 PEBA 0.355 0.650 0.466
-13 PEBA 0.434 0.233 0.433
-12 PEBA 0.539 0.250 0.383
12 PEBA 0.644 0.150 0.533
-13 PEBA 0.684 0.166 0.500
-14 PEBA 0.684 0.183 0.533
-14 PEBA 0.684 0.416 0.416
-13 PEBA 0.802 0.233 0.616
12 PEBA 0.828 0.450 0.500
14 PEBA 0.855 0.333 0.616
-13 PEBA 0.881 0.350 0.783
-13 PEBA 0.894 0.400 -0.050
-13 PEBA 0.947 -0.016 0.433
21 PEBA 0.947 0.000 0.400
-12 PEBA 0.986 0.233 -0.116

 

Appendix C (cont'd.)

Appendix C (cont'd.)

231

 

 

pH 5.4
Peak Height* Derivative x y z

—11 pH 5.4 0.052 0.066 0.416
—11 pH 5.4 0.052 0.633 0.416
-9 pH 5.4 0.078 -0.033 0.550
-10 PH 5.4 0.289 0.066 0.366
9 pH 5.4 0.381 -0.066 0.600

9 pH 5.4 0.434 0.616 0.366

9 pH 5.4 0.565 0.116 0.633
11 pH 5.4 0.565 0.266 0.466
9 pH 5.4 0.631 0.433 0.400
-13 pH 5.4 0.684 0.166 0.533
-9 pH 5.4 0.684 0.416 0.416
-10 pH 5.4 0.723 0.566 0.633
-9 pH 5.4 0.763 0.333 0.483
-9 pH 5.4 0.789 0.466 0.383
9 pH 5.4 0.868 0.000 0.000
12 pH 5.4 0.881 0.316 0.600
-9 pH 5.4 0.921 0.333 0.583
-9 pH 5.4 0.934 0.400 0.450
-11 pH 5.4 0.947 0.133 0.583
-11 pH 5.4 0.947 0.566 0.583

 

Appendix C (cont'd.)

232

Appendix C (cont'd.)

 

 

PEBA pH 7.3
Peak Height* Derivative x y z
11 PEBA 0.118 -0.066 0.333
11 PEBA 0.315 0.416 0.566
-13 PEBA 0.434 0.250 0.433
-11 PEBA 0.500 0.300 0.466
-12 PEBA 0.526 0.250 0.483
12 PEBA 0.578 0.250 0.483
-12 PEBA 0.592 0.300 0.516
11 PEBA 0.605 0.400 0.600
11 PEBA 0.684 -0.083 0.433
-14 PEBA 0.684 0.183 0.533
-11 PEBA 0.763 0.300 0.483
11 PEBA 0.881 0.433 0.666
12 PEBA 0.947 0.000 0.400

 

Appendix C (cont'd.)

Appendix C (cont'd.)

233

 

 

pH 7.3
Peak Height* Derivative x y z
13 pH 7.3 0.052 0.500 0.600
-11 pH 7.3 0.092 0.516 0.633
-13 pH 7.3 0.276 0.066 0.366
-15 pH 7.3 0.315 -0.083 0.583
11 pH 7.3 0.368 -0.066 0.600
-11 pH 7.3 0.605 0.233 0.533
11 pH 7.3 0.631 0.433 0.400
-18 pH 7.3 0.671 0.183 0.533
-14 pH 7.3 0.671 0.416 0.416
-13 pH 7.3 0.723 0.566 0.633
14 pH 7.3 0.789 0.200 0.016
-12 pH 7.3 0.828 0.300 0.583
11 pH 7.3 0.881 0.316 0.600
-11 pH 7.3 0.907 0.016 0.366
-11 pH 7.3 0.921 0.333 0.583
13 pH 7.3 0.947 0.000 0.400

 

Appendix C (cont'd.)

Appendix C (cont'd.)

234

 

 

PPBA
Peak Height* Derivative x y z
-9 PPBA 0.171 0.283 0.466
-9 PPBA 0.233 0.266 0.166
10 PPBA 0.315 0.583 0.566
10 PPBA 0.328 0.100 0.250
-9 PPBA 0.368 -0.033 0.083
-10 PPBA 0.434 0.250 0.433
-9 PPBA 0.644 0.466 0.933
-9 PPBA 0.671 0.350 0.466
10 PPBA 0.671 0.600 0.750
10 PPBA 0.684 -0.066 0.433
12 PPBA 0.855 0.333 0.616
10 PPBA 0.868 0.550 0.616
9 PPBA 0.894 0.433 0.666
10 PPBA 0.947 0.000 0.400
-10 PPBA 0.947 0.283 0.433

 

Appendix C (cont'd.)

235

Appendix C (cont'd.)

 

 

PBBA
Peak Height* Derivative x y z

9 PBBA 0.052 0.350 0.200
11 PBBA 0.052 0.500 0.600
10 PBBA 0.315 0.433 0.566
9 PBBA 0.328 0.100 0.250
-9 PBBA 0.368 0.033 0.083

-19 PBBA 0.434 0.233 0.433
9 PBBA 0.486 0.333 0.500
9 PBBA 0.671 0.600 0.750 -
9 PBBA 0.789 0.333 0.016
12 PBBA 0.855 0.333 0.616
11 PBBA 0.868 0.333 0.616
-9 PBBA 0.947 0.283 0.433
11 PBBA 0.947 0.000 0.400
-9 PBBA 1.000 0.316 0.300

 

* °_
Electrons/A 3 = (Peak Height)/50.