REVERSBBLE DISSDCMTION 0F YEAST
GLYCERALDEHYDE 3L PHOSPHATE
DEHYDROGENASE [N THE PRESENCE OF
ADENOSLNE TRIPHOSPHATE

Thesis fer the Degree f0 Ph. D‘
MECHIGAN STATE UNIVERSITY
GEORGE MICHAEL STANCEL
1970

 

~,.h.‘r>1~“ LIBRARY

Michigan State
' Uni'v‘eI‘SIty

 

This is to certify that the
thesis entitled

REVERSIBLE DISSOCIATION OF YEAST GLYCERALDEHYDE-B-PlDSPHATE
DEHYDROGENASE IN THE PRESENCE OF ATP

presented by

George Michael Stancel

has been accepted towards fulfillment
of the requirements for

 

Pho D. degree in Biochemistry

7/fl?2‘(‘(a.n\ (g [‘(Z’JJL

Major professor

Date December 18, 1910

0-7639

 

‘~.--“'—.“ RV»
-' ‘1‘ l1 - \
.-I_..~-"’ H ‘L
—————— ‘A”—\" i
7—. d I. . ~\
uguodoVJULHH “' '

‘. w
e bra"!
AQ‘F ‘
‘_ I ‘I I
«Viv-.338 J. GE:

Inch measure tr
:adenosine S"
..: self-maximal.
ussociatior
5308 '
. subunlt

’i-
s. \
‘v

, a 1.6 S 3‘“
'37 day incubat
Structural Chang
if: electrophoret

native enzyrr

 

 

 

ABSTRACT

REVERSIBLE DISSOCIATION OF YEAST GLYCERALDEHYDE-3-PHOSPHATE
DEHYDROGENASE IN THE PRESENCE OF ADENOSINE TRIPHOSPHATE

BY

George Michael Stancel

Tetrameric yeast glyceraldehyde-3-phosphate dehy-
drogenase undergoes a time-dependent inactivation at 0° in
the presence of adenosine 5'-triphosphate as a result of
dissociation into subunits. The dissociation has been
studied at 0° in 0.2 §_Tris buffer, pH 8.0, with varying
amounts of adenosine 5'-triphosphate. Inactivation studies
which measure the loss of catalytic activity indicate that
an adenosine 5'-triphosphate concentration of 0.5 mg produces
the half-maximal effect. An ultracentrifugal analysis of
the dissociation reveals in addition to the native tetramer
a 3.0 S subunit (produced by incubation times less than 12

hrS). a 1.6 S subunit (24 hr incubation) and a 0.9 S species

(3-7 day incubation). No dimer is observed in the analysis.

Structural changes are also indicated by the appearance of
an electrOphoretic band with a mobility less than that of

the native enzyme. The 3.0 S subunits can be reassociated

to the native 7.5 S tetramer.

 

   

of a;

'I-uvtn V
.

~4v.‘.ddol

:Wzbi an in»

I. up».

‘ I
‘54 “an A
‘ - H“ b “H
"‘ ‘Hve-U “.\.
I

' .
...;v~‘“‘F_n ‘“ _‘
V N I. F
"-~v.nJ_.)
.

A. ‘: ~ ‘
-‘dt 5‘ -
.o J‘DDCCLQ
. 'g:

 

u‘e :(‘y- ‘ .
‘4‘ +n‘
ks.‘S e,
‘
\..;S ‘i I
«1“
I
h‘.“‘ A: c
V . ‘
‘ ~01qu r:

George Michael Stancel

Tetrameric yeast glyceraldehyde-3-phosphate dehy-
drogenase undergoes a time-dependent inactivation in the
presence of adenosine 5'-triphosphate as a result of dis-
sociation into monomeric subunits. Optimal conditions,
which yield complete inactivation in 5 hours are: (l) l-2
‘mg adenosine 5'-triphosphate, (2) 0°, (3) protein concentra—
tions of 0.03-0.l mg per ml, (4) pH 9.0, and (5) 0.113
B-mercaptoethanol. Transition points (half-maximal loss
of activity in 5 hours) are: (1) 0.5 m§_adenosine 5'-tri-
phosphate, (2) 12°, (3) 0.5 mg per ml and (4) pH 8.6.
Adenosine 5'-mon0phosphate and adenosine 3',5'-monophosphate
do not dissociate the enzyme. They, and all the substrates
of the reaction, partially protect the enzyme from dissocia-
tion. Dissociation and inactivation are completely reversed
by warming to 17°. Reassembly is greatly stimulated by
adenosine 5'-triphosphate and by 10% sucrose. Optimal
reassembly conditions are: (l) 0.04 mg per ml protein,

(2) pH 7.0, (3) 1-2 mfl_adenosine 5'-triphosphate, (4) 17°,

(5) 10% sucrose and (6) 0.1 g_B-mercaptoethanol. Inactiva-
tion and dissociation apparently result from electrostatic
.repulsion. The results are discussed in terms of a possible
Inole for this enzyme in the regulation of glycolysis. Since
tflxis dissociation produces fairly compact subunits, associa-
tixmn of folded monomers to tetramers may be studied inde-
pendently of the polypeptide folding.

A detailed kinetic analysis of the dissociation of

tetrameric yeast GAPD to monomers at 0° in ATP has been

 

performed by
;E 8.0 in OJ
process is b
by a 3:! se
The
egulibri‘m T
2;;ilibrimn
iissociation
The
:15 a half-p
3551:ng that
'3 monomers.
fixation on
The dimer in
33:10“ beth
3'0: binding
iSSGCiatiOn
Strength is
325 ex'POSed

k
«D magmers .

George Michael Stancel

performed by following the loss of catalytic activity. At
pH 8.0 in 0.02 §_Tris and 0.075 §_B-mercaptoethanol the
process is biphasic. There is a £355 initial step followed
by a glgw_second step.

The fast step can be described by a rapid reversible
equilibrium between tetramers and dimers. The calculated
equilibrium constant (Keq = 10.2 x 10-8 g) for this initial
dissociation yields a value of 8.9 kcal/mole for A G° at 0°.

The slow step exhibits psuedo-first order kinetics
and a half-power dependence on protein concentration, sug-
gesting that this step reflects the dissociation of dimers
to monomers. The dependence of the dimer to monomer dis-
sociation on ATP concentration indicates that ATP binds to
the dimer in successive, reversible steps with no inter-
action between ATP binding sites. The average constant
for binding of ATP to the dimer is 0.75 mg, The rate of
dissociation of dimers to monomers is decreased as ionic
strength is increased, suggesting that hydrophobic residues
are exposed to the solvent medium as the dimer dissociates
to monomers.

A kinetic analysis of the reversal of dissociation
at 23° in the presence of 10% sucrose reveals that the
reversal process is second order with respect to protein
at low protein concentrations and first order at high
concentrations. These results suggest that the final step

in the reassembly of monomers to tetramers is a first order

 

gracess involv;
fr! the reasser
16.0 min at O.<
Exemin
gesezce of a g
:i yeast glycez'

sherirents ind

I

r)
u

c f
(3
(D

H

FJ
(I)

O

r 3

 

Lemar weigh.
:eterznined. Ma

ate: the inhibi

 

SCE’GIOgenase V

Effie“ .
. “31 Physme

o

Nan

: Curie-X betwe

353 7
Slbsequent

r“...
"“rthfless .

 

 

George Michael Stancel

process involving a conformational change. The half—life
for the reassembly process is 8.5 min at 0.20 mg/ml and
16.0 min at 0.010 mg/ml.

Examination of yeast cell extracts indicated the
presence of a protein which caused an apparent inhibition
of yeast glyceraldehyde—3-phosphate dehydrogenase. Initial
experiments indicated that this inhibitor protein was not
a protease, and that the inhibitor was present in intact
yeast cells. The inhibitor protein was purified and its
molecular weight (58,000) and amino acid composition were
determined. Maximal inhibition was then shown to occur
when the inhibitor protein and glyceraldehyde-3-phosphate
dehydrogenase were mixed in a 1:1 mglag ratio, however,
several physical tests failed to reveal the formation of
a complex between the two proteins. The inhibitor protein
was subsequently identified as triosephosphate isomerase.
Nevertheless, this identification of the inhibitor protein
as triosephosphate isomerase is not sufficient to clearly
explain the observed inhibition, since (1) the inhibition
is not completely overcome by glyceraldehyde-3-phosphate,
(2) the inhibition depends on the total protein concentra-
tion as well as the ratio of the two proteins, and (3)
triosephosphate isomerase seems to activate glyceraldehyde-

3-phosphate dehydrogenase under certain conditions.

 

,. u." H ‘-

-~°.*F?'\? W D“ "
.-I-.u.ddu ‘

v.0... A .. .,..
.rw Dz'r‘-‘ s

,1: H
Houndsxvyg.|s bk

 

REVERSIBLE DISSOCIATION OF YEAST GLYCERALDEHYDE-3-PHOSPHATE

DEHYDROGENASE IN THE PRESENCE OF ADENOSINE TRIPHOSPHATE

BY

George Michael Stancel

A THESIS

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

DOCTOR OF PHILOSOPHY
Department of Biochemistry

1970

 

The au‘
3!. W. C. Deal‘
r»

the course

fishers of his

 

 

ACKNOWLEDGEMENTS

The author wishes to eXpress his appreciation to
Dr. W. C. Deal, Jr., for his guidance and assistance through—
out the course of this work. The author also thanks the
members of his guidance committee, Drs. Wells, Kindel,
Karabatsos and Tulinsky, for their time and efforts.

The helpful discussions, criticisms and encourage-
ment of Drs. Suelter and Kuczenski are also greatly appreci-
ated.

The author would also like to thank Dr. William
Hare, Mr. Gregory Cook, Miss Shirley Ann Williams and
Miss Mary Kuhn for their assistance in preparing various

portions of this thesis.

ii

 

I. GEQTER O

A.

3.

Organ

Appro.

Liter.
Dehyd:

 

 

 

II.

TABLE OF CONTENTS

CHAPTER 0NE--INTRODUCTION . . . . . . . . .
A. Organization of the Thesis . . . . . . .
B. Approach to the Research . . . . . . . .

C. Literature Review: Glyceraldehyde-3-Phosphate
Dehydrogenase . . . . . . . . . . .

1. Introduction . . . . . . . . . .

2. Reaction Mechanism of the Oxidation of
Glyceraldehyde-3—Phosphate . . . . . .

3. Other Reactions Catalyzed by Glyceralde-
hyde-3-Phosphate Dehydrogenase . . . .

4. Substrate Binding . . . . . . . . .
a. NAD Binding . . . . . . . . .
b. Glyceraldehyde—3-Phosphate Binding . .

5. Subunit Structure of Native Glyceralde-
hyde-B-Phosphate Dehydrogenase . . . .

D. Literature Review: Subunit Association . . .
E. Literature Review: Cold Lability . . . . .
CHAPTER TWO-~STRUCTURAL STUDIES OF THE DISSOCIATION
AND REASSOCIATION OF GLYCERALDEHYDE-3-PHOSPHATE

DEHYDROGENASE IN THE PRESENCE OF ATP . . . . .
A. Abstract . . . . . . . . . . . . .
B. Introduction . . . . . . . . . . . .
C. Results . . . . . . . . . . . . .

1. Loss of Activity Upon Incubation with ATP

and Production of a Slower-Moving Component
in Polyacrylamide Disc Electrophoresis . .

iii

11

11
16
17

21
22
23
24

24

III.

Page

2. Production of a 3.0 and a 1.6 S
subunit 0 C O Q C O O O O O O 3 0

3. Sedimentation Equilibrium Studies of
Yeast GAPD in the Presence of ATP . . . 33

4. Production of a 0.9 S Subunit . . . . 36

5. Reversal of Dissociation: Production of
a 7.5 S Tetramer from a 3.0 S Monomer . 41

D. Discussion . . . . . . . . . . . . 47

1. Production of a 3.0 S and a 1.6 S
subunit . O O O O O D O O O O 47

2. Mechanism of Dissociation and the Overall
Control of GAPD . . . . . . . . . 49

E. Materials and Methods . . . . . . . . 50
l. Ultracentrifugal Analysis . . . . . 50
a. Sedimentation Velocity Analysis . . 50
b. Sedimentation Equilibrium Analysis . 51
2. Electrophoresis . . . . . . . . . 51
3. Enzyme Prepration and Assay . . . . . 51
CHAPTER THREE--SYSTEMATIC STUDY OF THE VARIABLES
AFFECTING DISSOCIATION AND REASSOCIATION OF
GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE AT 0°
IN THE PRESENCE OF ATP . . . . . . . . . 52
A. Mstract . O O O O O O O O O O O 53
B. Introduction . . . . . . . . . . . 55
C. Results . . . . . . . . . . . . . 57
1. Effect of ATP Concentration on Inactiva-
tion and Dissociation of Yeast
Glyceraldehyde-3-Phosphate Dehydrogenase 57

a. Effect of Protein Concentration . . 60

b. Effect of Temperature . . . . . 60

iv

(3

(7

 

IV.

0. Effect of pH . . . . . . .

d. Specificity of the Dissociation
Process for ATP . . . . . .

e. Protective Effects of 3',5'-AMP, AMP,

ADP, and Substrates . . . . .
2. Reversal of the Inactivation and
Dissociation of Yeast Glyceraldehyde-B-
Phosphate Dehydrogenase . . . . .
a. Effect of pH on Reassembly . . .
b. Effect of Temperature on Reassembly

c. Effect of Protein Concentration on
Reassembly . . . . . . . .

d. Effect of ATP Concentration on the
Reassembly . . . . . . . .

Do DiSCUSSiOD o o o o o o o o o o ' o
1. Mode of Action of ATP . . . . . .

2. Biological Significance of the Dis-
sociation Process for Metabolic Control

B. Materials and Methods . . . . . . .
1. Enzyme Preparation and Assay . ; .
2. Reagents . . . . . . . . . .
3. Sucrose Density Centrifugation . . .
4. Dissociation and Reassembly . . . .
CHAPTER FOUR--KINETICS OF THE DISSOCIATION AND
REASSOCIATION OF GLYCERALDEHYDE-3-PHOSPHATE
DEHYDROGENASE IN THE PRESENCE OF ATP . . .
A. Abstract . . . . . . . . . . .
B. Introduction . . . . . . . . . .

l. Subunit Structure . . . . . . .

2. Subunit Interactions . . . . . .

V

Page

60

67

67

72
79
79

79

84
87
87

94
95
95
96
96
97

98
99
101
101
102

Resu

 

 

3.

In Vivo and In_Vitro Folding and
Association . . . . . . . . . .

Results . . . . . . . . . . . . .

1.

2.

Description of the Model and Summary
of Evidence . . . . . . . . . .

Analysis of the Slow Process (Dissocia-
tion of Dimers) . . . . . . . . .

Separation of the Fast Process From the
Slow Process . . . . . . . . . .

Analysis of the Fast Process (Tetramer-
Dimer Equilibrium) . . . . . . . .

Effect of ATP on Dissociation . . . .
ATP Binding Constants . . . . . . .

Tests for Interaction Between ATP
Binding Sites . . . . . . . . .

Nature of Subunit Contact Sites: Effect
of Ionic Strength on the Dissociation
Process 0 O O O O O O O C O O

Reversal of Dissociation . . . . . . .

l.

2.

3.

First Order Dependence of the Reassembly
Process at Higher Protein Concentration .

Second Order Dependence of the Reassembly
Process at Low Protein Concentration . .

Change in Order with Change in Protein
Concentration . . . . . . . . .

Discussion . . . . . . . . . . .

1.
2.

Mechanism of Dissociation . . . . .

Absence of a Dimer Peak in Ultracentri-
fugal Analysis of Dissociation of Yeast
GAP D by ATP 0 I O O O O O O O O

Strengths of the Dimer-Dimer Attraction:

Energetics of the Tetramer-Dimer
EqUilibri‘m O O O O O O O O O 0

vi

Page

103
104

105

109

115

118
121
123

126

127

136

136

139

139

146

146

150

152

CHAPTER ’
AN IT'EIEL
GLYCERAL
IDENTIP;

TRIOSEPE

 

Page

Chemical Nature of the Subunit Contact
Sites 0 O O I O O I O O O O O 155

Reassociation of GAPD Monomers to
Tetramers . . . . . . . . . . . 159

Comparison of the Proposed Mechanism for
Reassociation of GAPD Subunits to Other
Enzymes . O O I O O O O C O O 161

F. Materials and Methods . . . . . . . . 162

l.

2.

Dissociation . . . . . . . . . . 162

Reassociation . . . . . . . . . 163

CHAPTER FIVE--ISOLATION AND CHARACTERIZATION OF

AN INHIBITOR PROTEIN FROM YEAST WHICH INHIBITS
GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE AND
IDENTIFICATION OF THE INHIBITOR PROTEIN AS
TRIOSEPHOSPHATE ISOMERASE . . . . . . . . 165

A. Abstract . . . . . . . . . . . . 166

B. Introduction . . . . . . . . . . . 167

C. Results . . . . . . . . . . . . . 168

1.

2.

Preparation of a Protease-Containing
Fraction . . . . . . . . . . . 168

Test for Inactivation of GAPD and Tests
For Protease Activity Against Hemoglobin,
GAPD, and Aldolase . . . . . . . . 169

Evidence that GAPD activity Loss was not
Due to a Protease . . . . . . . . 173

Possible Explanations of the Effect . . 177
Purification of the Inhibitor Protein . 184
a. Definition of Units . . . . . . 189

b. ‘Cell Lysis and Treatment of the
Crude Extract . . . . . . . . 190

c. 47-63% Ethanol Fractionation . . . 190

vii

D. M016
and
Inhi

 

Page

d. Protamine Sulfate and Ammonium
Sulfate Treatment . . . . . . . 192

e. Ethanol Fractionation . . . . . 193
f. Isoelectric Focusing . . . . . 193

Proof that the Inhibitor Protein was not
an Altered or Incomplete Subpolymer of
Glyceraldehyde-3-Phosphate Dehydrogenase 197

a. Tests for Cross Reaction Between the
Inhibitor Protein and Anti-GAPD and
Between GAPD and Anti-Inhibitor
Protein . . . . . . . . . . 198

b. Amino Acid Analysis of the Inhibitor
Protein and GAPD . . . . . . . 201

Molecular Weight of the Inhibitor Protein
and Molar Stoichiometry of the GAPD-
Inhibitor Protein Interaction . . . . . 204

1.

2.

Molecular Weight of the Inhibitor
Protein . . . . . . . . . . . 204

Stoichiometry of the GAPD-Inhibitor
Protein Interaction . . . . . . . 205

Tests for the Formation of an Inhibitor
Protein-GAPD Complex . . . . . . . 208

a. High-Speed Sedimentation Equilibrium 208

b. Sucrose Density Gradient
Centrifugation . . . . . . . . 215

c. Fluorometric Titration of (GAPD-NADH)
With the Inhibitor Protein . . . . 222

d. Immunochemical Titration of a GAPD-
Inhibitor Protein Mixture with Anti-
Inhibitor Protein . . . . . . . 225

Identification of the Inhibitor Protein
as Triosephosphate Isomerase . . . . 228

Mechanism of Inhibition of GAPD by
Triosephosphate Isomerase . . . . . 233

viii

Disc

Meti

 

‘71

6. Effect of Triosephosphate Isomerase on
the GAPD Catalyzed Hydrolysis of
pritrophenyl Acetate . . . . . . .

7. Effect of Glyceraldehyde-3-Phosphate

Concentration on the Inhibition of
GAPD by Triosephosphate Isomerase . . .

E. Discussion . . . . . . . . . . . .
1. Effect of Glyceraldehyde-3-Phosphate
Concentration on the Inhibition of
GAPD by Triosephosphate Isomerase . . .
2. Mechanism of the Inhibition of GAPD by
Triosephosphate Isomerase . . . . .
3. Hydrolysis of pritrophenyl Acetate by
Triosephosphate Isomerase . . . . .
F. MethOdS C O O O O O O O O O O O O
l. Assay for Inhibition of Glyceraldehyde-
3-Phosphate Dehydrogenase . . . . .
2. Reagents . . . . . . . . . . .
3. Isoelectric Focusing . . . . . . .
4. Antibody Preparation and Immunodiffusion
Tests 0 O O C O C O O O O O O
5. Fluorescence Measurements . . . . .
6. Amino Acid Analysis . . . . . . .
7. Assay for Triosephosphate Isomerase
Activity . . . . . . . . . . .
8. Assays for Esterase Activity Using
pritrophenyl Acetate . . . . . . .
9. Turbidimetric Measurement of Protein

Concentration . . . . . . . . .

CHAPTER SIX--KEY FINDINGS OF THE RESEARCH AND

SUGGESTIONS FOR FUTURE WORK ' . . . . . . .

A. Dissociation of Yeast GAPD in ATP

ix

Page

234

236

241

241

243

245
246

246
247
247

248
249

250

250

251

252

253
254

B. Raver
22. CRAPTER E

A. Apps:

 

VII.

Page
B. Reversal of Dissociation . . . . . . . 259
CHAPTER SEVEN--APPENDICES AND LIST OF REFERENCES 262

A. Appendix l--Determination of Reaction Order
by the Half-Life Method . . . . . . . 263

B. Appendix 2--Derivation of the Integrated
Rate Equation for a Tetramer-Dimer
Equilibrium . . . . . . . . . . . 254

C. List of References . . . . . . . . . 253

5.!

k4

 

 

 

DHAP
EDTA
FDP
GAP

GAPD

LDH

PLP
TCA
{PMV

{PPI

LIST OF ABBREVIATIONS

dihydroxyacetone phosphate
(ethylenedinitrilo) tetraacetic acid
fructose-1,6-diphosphate
Qfglyceraldehyde-3-phosphate
Qfglyceraldehyde-3-phosphate dehydrogenase:
NAD oxidoreductase (phosphorylating)

(EC 1.2.1.12)

L-lactate dehydrogenase: NAD oxidoreductase
TEC 1.1.1.27)

pyridoxa1-5'-phosphate
trichloroacetic acid
tobacco mosaic virus

triosephosphate isomerase (D-glyceraldehyde-B-
phosphate ketol isomerase) TEC 5.3.1.1)

xi

h? ,

6“.

We,
l.“.

Summary

q
hm
A3 I a

u‘

Report
Glyce

Enzymes

 

Table

II.

III.

IV. '

VI.

VII.

VIII.

IX.

LIST OF TABLES

Summary of Activites of 3-Phosphog1ycera1dehyde
Dehydrogenase . . . . . . . . . . .

Reported Molecular Weights for Rabbit Muscle
Glyceraldehyde-3-Phosphate Dehydrogenase . .

Enzymes which undergo Cold-Inactivation . . .

Calculated Equilibrium Constants for a
Tetramer—Dimer Equilibrium . . . . . . .

Dependence of Half-Lives for Reassociation on
Protein Concentration . . . . . . . .

Loss of Glyceraldehyde—3-Phosphate Dehydrogenase
Activity in the Presence of Inhibitor Protein

Summary of Purification of the Inhibitor
Protein 0 O O O O O O I O O O O 0

Amino Acid Composition of Glyceraldehyde-B-
Phosphate Dehydrogenase and the Inhibitor
Protein . . . . . . . . . . . . .

Hydrolysis of EfNitrophenyl Acetate by

Glyceraldehy —3-Phosphate Dehydrogenase and
Triosephosphate Isomerase . . . . . . .

xii

Page

13
19

122

140

170

191

202

235

 

Activity
\ ‘ .
Denvo:

. , ‘
Vltn n

Electro;
3-Phos

1n us;

‘ e
SESIZexfi
Glycer

with A

MOIECUIC
3‘Phos
ATP

‘ :
SECLmen-

ClYcerl

IRCuLe

SUCrQSe
GlYCer
DiSSOC
Cehtir

SuCIOSe

of ..

.
ad:

PhOSph

 

Figure

1.

:10

LIST OF FIGURES

Activity of Yeast Glyceraldehyde-3-Ph05phate
Dehydrogenase after 4 1/2 hr Incubation
Wi th ATP 0 O O O O C O O O O O O O

Electrophoretic Patterns of Yeast Glyceraldehyde-
3-Phosphate Dehydrogenase after 4 1/2 hr
incubation with ATP . . . . . . . . .

Sedimentation Velocity Patterns of Yeast
Glyceraldehyde-3-Phosphate Dehydrogenase
Wi th ATP 0 . O O O C O O O O O O 0

Molecular Weight Analysis of Yeast Glyceraldehyde-
3-Phosphate Dehydrogenase in the Presence of
ATP 0 C O O O O O O O O I O O O O

Sedimentation Velocity Pattern of Yeast
Glyceraldehyde-3-Phosphate Dehydrogenase After
Incubation with ATP for 7 Days . . . . . .

Sucrose Density Gradient Centrifugation of
Glyceraldehyde-3-Phosphate Dehydrogenase
Dissociated in ATP and Reassociated before
Centirfugation . . . . . . . . . . .

Sucrose Density Gradient Centrifugation Pattern
of Native Tetrameric Glyceraldehyde-B-
Phosphate Dehydrogenase . . . . . . . .

Activity of Glyceraldehyde-3-Phosphate
Dehydrogenase after Incubation for 5 hr at
0° With ATP 0 Q C O O O O O O O O O

Sucrose Density Gradient Centrifugation of
Glyceraldehyde—3vPhosphate Dehydrogenase
After 9 hr Incubation at 0° with 5 mM ATP . .

Effect of Protein Concentration on Inactivation

of Glyceraldehyde-3-Phosphate Dehydrogenase
by 4 m-M- ATP 0 O O O O O O I O O I O

xiii

Page

25

28

31

34

37

43

45

58

61

63

in! ~"'
”,“e

H.

Effect c
Glyce:

Activit;
Dehyd:

5 hr 'a

Effect c

 

 

Figure Page

11. Effect of Temperature on Inactivation of
Glyceraldehyde-3-Phosphate Dehydrogenase
by 4 In-M— ATP 0 O O O O O C O O O O O 6 5

12. Effect of pH of Tris Buffer on Inactivation of
Glyceraldehyde-3-Phosphate Dehydrogenase by
4mMATP.............68

13. Activity of G1yceraldehyde-3-Phosphate
Dehydrogenase after Incubation at 0° for r
5 hr with 1 mM_Adenine Nucleotides . . . . 7O

14. Protective Effects of Adenine Nucleotides and
Substrates against Inactivation of
Glyceraldehyde-3-Phosphate Dehydrogenase
by ATP . . . . . . . . . . . . . . 73

15. Effect of ATP on Reactivation of Inactive 2.8 S
Monomers . . . . . . . . . . . . . 76

16. Effect of pH on Reassociation of Glyceraldehyde-
3-Phosphate Dehydrogenase Subunits . . . . 80

17. Effect of Temperature on Reassociation of
Glyceraldehyde-3-Phosphate Dehydrogenase
subunits O O O O O O O O I O O O O 82

18. Effect of Protein Concentration on Reassociation
of Glyceraldehyde-3-Phosphate Dehydrogenase
subunits O C O O O O O O C O O O O 85

19. Effect of ATP Concentration on Reassociation of
Glyceraldehyde-3-Phosphate Dehydrogenase
subunits O C O O O O O O O O I I O 8 8

20. Time-Dependent Loss of Glyceraldehyde-3-Phosphate
Dehydrogenase Activity at 0° in ATP . . . . 107

21. Effect of Protein Concentration on the Rate of
Dissociation of Glyceraldehyde-3-Phosphate
Dehydrogenase in ATP . . . . . . . . . 110

22. Determination of the Order of the Slow Process
by the Half-Life Method . . . . . . . . 113

23. Synthetic Example of the Treatment of Kinetic

Data Used to Separate the Fast Process From
the Slow Process . . . . . . . . . . 116

xiv

I
”V"?
:.a“e
O

N. Kinetic ,
Fast P
for a
Tetrar

23. Determir;
the 81

M. Test for
Involv

N. Inhibit:
Ionic

M. The Per
Mixtu:

5- First 0:
at Hif

w' variatin
Protei
Conce:

“0 Change i
with c

Digestic
Imub;

Effect C
Inhibil
Dehydr

Effect C
0f GILI
by In?

Effect C
tion q
Phosph

 

 

Figure

24.

25.
26.
27.
28.
29.

30.

31.
32.

33.
34.
35.
36.

37.

Kinetic Fit of Corrected Data Points for the
Fast PrOcess to the Integrated Rate Equation
for a Rapid Equilibrium between Active
Tetramers and Inactive Dimers . . . . . .

Determination of the ATP Binding Constant for
the Slow Process . . . . . . . . . .

Test for Interaction between ATP Binding Sites
Involved in the Slow Process . . . . . .

Inhibition of the Slow Process by Increasing
Ionic Strength . . . . . . . . . . .

The Per Cent Tetramers of an Equilibrium
Mixture of Tetramers and Dimers . . . . .

First Order Dependence of the Reassembly Process
at Higher Protein Concentrations . . . . .

Variation of the Half-Lives of Reassembly with
Protein Concentration at Low Protein
Concentration . . . . . . . . . . .

Change in the Order of the Reassembly Process
with Change in the Protein Concentration . .

Digestion of Denatured Hemoglobin by the
Inhibitor Protein Preparation . . . . . .

Effect of the Time of Incubation on the
Inhibition of Glyceraldehyde-3-Ph03phate
Dehydrogenase by the Inhibitor Protein . . .

Effect of NAD Concentration on the Inhibition
of Glyceraldehyde-3-Phosphate Dehydrogenase
by Inhibitor Protein 0 O C O O O O O 0

Effect of GlyceraldehydeFB—Phosphate Concentra-
tion on the Inhibition of Glyceraldehyde-3-
Phosphate Dehydrogenase by Inhibitor Protein .

Inhibition of Glyceraldehyde-3-Phosphate
Dehydrogenase by Different Inhibitor Protein
Preparations . . . . . . . . . . . .

Effect of Inhibitor Protein Concentration on the
Inhibition of Glyceraldehyde-3-Phosphate
Dehydrogenase . . . . . . . . . . .

Page

119

124

128

131

134

137

141

147

171

175

179

181

185

187

JO

.0?

n".

J'-
K A)
o

u“
u!
o

.

Immunoc;
Q I

9305::

Prote;

‘S‘ew‘ w
va .An‘..
Q 'I ~
.nr~ n~
dunnLu‘

‘\ ‘ ~
W”V!fl V‘
i'“ In.
J

'U
'1
I ' (D
[f
L). (T) 'L’
| !

1. 9‘

'U L)
*1

O

(f ‘

o

I";
HILQP‘
V"
\f‘HYo O
nan B
De-

 

Figure Page

38. Purification of the Inhibitor Protein by
Isoelectric Focusing . . . . . . . . . 195

39. Immunodiffusion Patterns of Glyceraldehyde-B-
Phosphate Dehydrogenase and Inhibitor
Protein I O O O O I O O I O O O O 199

40. Determination of the Molar Stoichiometry of the
Inhibitor Protein-Glyceraldehyde-B-Phosphate
Dehydrogenase Interaction . . . . . . . 206

41. High-Speed Sedimentation Equilibrium of Inhibitor
Protein, Glyceraldehyde-3—Phosphate
Dehydrogenase and a Mixture of the Two
Proteins . . . . . . . . . . . . . 211

42. High-Speed Sedimentation Equilibrium of Inhibitor
Protein, Lactic Dehydrogenase and a Mixture of
the Two Proteins . . . . . . . . . . 213

43. Sucrose Density Gradient Centrifugation of the
Inhibitor Protein and Glyceraldehyde-3-
Phosphate Dehydrogenase . . . . . . . . 217

44. Sucrose Density Gradient Centrifugation of
Mixtures of the Inhibitor Protein and
Glyceraldehyde-3-Phosphate Dehydrogenase
with and without NAD . . . . . . . . . 220

45. Effect of the Inhibitor on the Fluorescence of
NADH Bound to Glyceraldehyde-3-Phosphate
Dehydrogenase . . . . . . . . . . . 223

46. Immunochemical Tetration of a Mixture of the
Inhibitor Protein and Glyceraldehyde-3-
Phosphate Dehydrogenase with Antibodies to
the Inhibitor Protein . . . . . . . .. . 226

47. Determination of the Km of Triosephosphate
Isomerase for DfGlyceraldehyde-3-Phosphate . . 231

48. Effect of Glyceraldehyde-3-Phosphate
Concentration on the Inhibition of Glyceralde-
hyde-3-Phosphate Dehydrogenase by Triosephos-
phate Isomerase . . . . . . . . . . . 239

xvi

 

CHAPTER ONE

INTRODUCTION

A “ .~ . . v
on ~:.E ‘iyarlOJf

.u
our 5.;
in! w.»

 

 

A .
«AMCI trke t \ ‘
5+4 m.

t -&C L‘121'7mes
‘

I'v'n. 1
“"‘J=pnate DORA
4...‘

{.1226}. and P"

c
p
F)
L11
'__J
’L}

””‘Pflate De.
$2,?

“5e M. Ste
N‘s-
. 3' A p:

r: ,

44 “Sieri ‘
an 5

\“

~Qf‘n

ORGANIZATION OF THE THESIS

The main areas of research are covered individually
in the various chapters of this thesis. For convenience
to the reader each chapter is presented as an independent
entity in the format of a scientific paper, with its own

Abstract, Materials and Methods, Introduction, Results,

 

 

and Discussion sections. The only deviation from this

 

format is that the references for all the chapters of the
thesis are combined at the end of the thesis.

Portions of Chapter Two have already been published
under the title "Metabolic Control and structure of Glyco-
lytic Enzymes. V. Dissociation of Yeast Glyceraldehyde-B-
Phosphate Dehydrogenase into Subunits by ATP," by George M.
Stancel and William C. Deal, Jr., (1968), Biochemical and

Biophysical Research Communications 31 398.

 

Chapter Three has already been published under the
title "Reversible Dissociation of Yeast Glyceraldehyde-B-
Phosphate Dehydrogenase by Adenosine Triphosphate," by
George M. Stancel and W. C. Deal, Jr. (1969), Biochemistry
8 4005. A preliminary report of the data in Chapter Three
had previously been presented at a meeting of the Federation
of American Societies for Experimental Biology (George M.

Stancel (1969), Federation Proceedings 28 490).

Chapte

 
   

Eise..e:ristr‘

   

results in Cha

: t. ‘ .
.::.art:1er s

 

 

Chapter Four is being prepared for submission to
Biochemistry, and it is hoped that some of the unpublished
results in Chapters Two and Five will provide a stimulus

for further study, and hopefully publication.

APPROACH TO THE RESEARCH

This research began with the observation that
glyceraldehyde-3-phosphate dehydrogenase exhibited a time
dependent loss of enzymatic activity when incubated with
ATP at 0°.

The time dependence of the loss of activity sug-
gested that structural changes might be occurring. The
ultracentrifugal analysis of the enzyme at 0° in the pres-
ence of ATP was therefore undertaken (Chapter Two). This
study revealed that the tetrameric enzyme was dissociated
into subunits at 0° (3.0 S and 1.6 S) in the presence of
ATP, the subunit species obtained depending on the length
of the incubation period. Further studies revealed that
the 3.0 S monomer could be reassembled to the native
tetrameric 7.5 S species.

Later structural studies revealed that very long
incubations at 0° in ATP (3-7 days) led to the appearance
of a 0.9 S species with an apparent molecular weight lower
than the expected value of 35,000 (based on the reported
tetramer molecular weight of 145,000, see Jaenicke, 23 El"

1968).

 

After
van: a syster’
fine dissociat‘
"szter Thre.

. 'r-

 

re
0:
wi
wilarly a C
,ieli informat

4: rom inac
izxrtant fO
process, sincx
net 83% and
«moms. 1'
lfitiuolite, E“
Abiding of \
«as ciat

After the observation of this reversible dissocia—
tion, a systematic study of the variables influencing both
the dissociation and reassociation processes was undertaken
(Chapter Three). There were many incentives for character-
izing both these processes. The study of the dissociation-
could yield valuable information concerning:

(1) the nature and strengths of the chemical forces

involved in subunit interactions,

(2) the binding of nucleotides to the enzyme,

(3) the stability and possible activity of indi-

' vidual subpolymers of a polymeric enzyme, and

(4) the possible biochemical significance with

respect to metabolic control of the interaction

of glyceraldehyde-3-phosphate dehydrogenase

with ATP .
Similarly, a study of the reassociation was expected to
yield information concerning the exact sequence of events
leading to the formation of the biologically active tetra-
mers from inactive monomers. This information should be
importantfor considerations of the in yiyg assembly
process, since the polypeptide chains of polymeric enzymes

must fold and associate after they are synthesized on the

 

ribosomes. This was especially interesting since another
metabolite, NAD, has been shown to be required for the
refolding of yeast glyceraldehyde-3-phosphate dehydrogenase

(dissociated in 8 M_urea), and it has been postulated that

 

2713 might con
centrolling a
Once ‘

‘

:ecissociat

3y
0) t
‘Eiiabil'
lZat'
10
A1 .
7 might b
wprot€01

5

NAD might control the synthesis of this enzyme in 2122 by
controlling a folding reaction (Deal, 1969).

Once the systematic study of the variables affecting
the dissociation and reassociation processes was completed,
a detailed kinetic analysis of both processes was under-
taken (Chapter Four). The kinetic study was performed in
order to obtain a more thorough understanding of the mecha-
nisms of dissociation and reassociation and to provide an
explanation for these processes on theoretical grounds.

The results in Chapters Two, Three and Four did not
suggest any physiological significance for the dissociation
of glyceraldehyde-3-phosphate dehydrogenase into subunits
by ATP, since the major effects occurred at low temperatures
(0°) and low protein concentrations, conditions not likely
to be encountered in yiyg, This work did provide a stimulus,
however, for other research in our laboratory. Thus,

Dr. S. T. Yang, a member of our research group, discovered
that very low concentrations of ATP (less than 1 mg)

greatly increased the susceptibility of yeast glyceraldehyde-
3-phosphate dehydrogenase to chymotrypic digestion at 23°
(Yang and Deal, 1969b). This result prompted us to investi-
gate our earlier suggestion (Stancel and Deal, 1968--see
Chapter Two) that the physiological significance of the
destabilization of glyceraldehyde-3-phosphate dehydrogenase
by ATP might be an increased susceptibility of the enzyme

to proteolysis, rather than a complete dissociation into

subunits.

 

O
\E Q“ '
“ Zine
A.
0.)? 0c

 

 

The experiments reported in Chapter Five were
initially undertaken to determine if yeast cells contained
a protease(s) which would degrade yeast glyceraldehyde-B-
phosphate dehydrogenase. If such a protease were found,
the effect of ATP on the £2E2.°f proteolysis could then
be examined. While the studies in Chapter Five did not
fulfill this original intent, they provided a number of
interesting observations which will hopefully form a basis
for further research in Dr. Deal's laboratory.

LITERATURE REVIEW: GLYCERALDEHYDE-B-PHOSPHATE
DEHYDROGENASE

 

INTRODUCTION. Since the crystallization of yeast GAPD in
1939 (Warburg and Christian, 1939), the properties of the
enzyme have been extensively studied because of an intrinsic
interest in the enzyme, because of the availability of
large quantities of the enzyme and because of the nature

of the catalytic reaction. A homogeneous preparation of
the enzyme, which is present in yeast in large amounts

[20% of the total soluble protein (Krebs et_al., 1953)],

is readily obtained in crystalline from (Krebs, 1955), thus
providing ample material for extensive investigations.
Furthermore, the enzyme catalyzes the only oxidation step
in glycolysis. This catalytic reaction is a substrate
level oxidative phosphorylation, and numerous mechanistic
studies have been performed because it was hoped this

reaction would provide a useful model for mitochondrial

oxidative pm
of studies cc
Fxf'ne, 1963
PERCTIOY NSC}?
E-phosphate t
inerganic pho
3f Phosphate

itd this SPOn'

 

The 1‘35
tu’e formatioy
Siilfhl’dr‘x'l err

31:;
“fly shown

31am
e aCYl‘en

cu .
...Z;"Yle With ah
Wu.
«4»;ka and P

 

 

oxidative phosphorylation (Racker, 1965). Several reviews
of studies concerning GAPD are available (Velick and
Furfine, 1963; Racker, 1965; Krebs, 1955).
REACTION MECHANISM OF THE OXIDATION OF GLYCERALDEHYDE-B-
PHOSPHATE. GAPD catalyzes the conversion of glyceraldehyde-
3-phosphate to 1,3-diphosphoglycerate in the presence of
inorganic phosphate and NAD. If arsenate is used instead
of phosphate unstable arsenOphosphoglycerate is produced
and this spontaneously decomposes to 3-phosphoglycerate,
and this is commonly used in assaying the enzyme.

The mechanism of the reaction involves the oxida-
tive formation of an S-acyl-enzyme intermediate between a
sulfhydryl group at the enzyme active site and the aldehyde
function of the substrate (Racker, 1965). This was most
clearly shown by the isolation and crystallization of a
stable acyl-enzyme formed by acylation of the rabbit muscle
enzyme with acetyl phosphate or 1,3-diphosphoglycerate
(Krimsky and Racker, 1955). It has also been shown that
the active-site cysteine is more reactive toward sulfhydryl
reagents than other cysteine residues in the GAPD molecule,
and that this reactive cysteine is situated in the same
octapeptide in the yeast, rabbit muscle and pig muscle
enzymes (Harris et,al., 1963; Perham and Harris, 1963).
The oxidative step is then followed by the phosphoryl

transfer step to yield 1,3-diphosphog1ycerate.

 

“Gav-

 

 

v--..+_

.:5.;ons is

 

 

OTHER REACTIONS CATALYZED BY GLYCERALDEHYDE-3-PHOSPHATE

 

DEHYDROGENASE. In addition to the oxidation of glyceral-

 

dehyde-B-phosphate, GAPD catalyzes a number of other
reactions (Taylor gt $1., 1963). A list of these other
reactions is presented in Table I. The involvement of
NAD and the requirement for free sulfhydryl groups of
these reactions is also noted. The transferase activity
listed refers to the transfer of the acetyl group of
acetyl phosphate to a number of receptors (Taylor et_§l.,
1963), while the diaphorase activity refers to the transfer
of electrons from NADH to 2,6-dichlorophenolindophenol
(Rafter and Colowick, 1957). The other activities are
self-explanatory. The significance of these various
reactions has been discussed by Park (1966). The variety
of reactions catalyzed by the GAPD may suggest that the
enzyme has multiple in_yiyg functions.

SUBSTRATE BINDING: NAD. Numerous studies concerning the

 

binding of the cofactor, NAD, to GAPD from both rabbit
muscle and yeast have been performed (Velick and Furfine,
1963; Racker, 1965), and it is clear that NAD is bound
very tightly. Yang and Deal (1969a) recently determined
a Km for the yeast enzyme of 0.18 m! NAD. Other studies
have measured dissociation constants much smaller, and
the reasons for such discrepancies have been previously

discussed (Velick and Furfine, 1963).

 

«I

Illilfflfllllfl-
mftpmniolfl AZ:>N~SQ rrcwuufuor .Lfl mfifl>Hn¥2EH
r_no "11*..nu‘~mv __A..a ema~zxva Lnnu AC/Nsa

.ln

    
 

EOHU
ITH>HO£QmO£Q UCG COdUU

ll .‘nlllfl L'vf .ul III.- nll ‘.i.||ll.'1l|.l||.l..|v all!

 

81"... .UII, .

1| £31-, i'flflpl‘l
M.—mh<zu.—n ~A.VV~1~>_~H.~A1~

uz~n_>:...~a.;<3—...~.Q>.~nva:~..:u.av-.~l.M. eiav «Mn...~fi.~.x~>}~.v~...J< .ANFV

II ll.|l."lfl'flln'l.|l, I" 8. II

\fVNlNzZK‘MJI I h .-.-~q5<.flh

.Amemfl ..mm_mm “cause some cmxmav

.muflnaana n<z+

.oaom oaumomopOH mo uommwo oz
#

 

 

+ + mDH + moé I ASH + mfifi
ommamnocmmosmmcmua .n
+ + mmMHOSQMHa .m
+ + xumomz n moaz .m
+ +0 oumumom
Hmsoomouuflslm mo mammaonmwm .v
*0 + oumnm
umosm Hauoom mo mwmaaouoam .m
+ + mammaosmmum Ho
.omcmnoxm mmm .mmmuwmmcmue .N
+ + GOAH
Imamnonmmonm 0cm GOwumpHxO .H
omuwsvmm mahusm coauommm ow oo>ao>cH
co mmsouo mm: modz H0 ndz

 

mm<zm00mnwmma MOMEMQQ¢MMUWAUOImmommIm ho mmHBH>HBU¢.hO Nm<223m .H mqmmfi

 

Pyrii

 

is name (E

have shown t‘:

 

 

vs :‘SC ‘
repanc
D‘:.;‘.’;I
‘ 1’91: r:C)l

.3:
aY and
‘87A
‘4"
J) enzyrtgn
f:- .p‘
“”9116: .
1‘
1‘30" “
‘lQ(1l 1:;‘h
.5: th
8 b
1nd:

10

Pyridine-3—a1dehyde NAD is a potent inhibitor of
the enzyme (Kaplan et_a1,, 1956) and more recent studies
have shown that adenine nucleotides are also inhibitors
(Yang and Deal, 1969a). It is the adenine protein of the
NAD molecule which appears essential for binding, and the
6-amino group and 2'-hydroxy1 group of the adenine moeity
appear to be critical (Yang and Deal, 1969a). Stockell
(1959) had earlier suggested the importance of the 6-amino
group, and Windmueller and Kaplan (1961) further suggested
the importance of the 1 position of adenine.

Data has also been presented (Racker, 1965) that
indicates that the 4 position of the pyridine ring may be
important for binding, but there are other possible explana-
tions for these observations (see Yang and Deal, 1969a).

As previously mentioned there have been a number
of discrepancies regarding the number of moles of NAD
bound per mole of enzyme as well as the magnitude of the
binding constants (Velick and Furfine, 1963). Quite
recently, it has been suggested for both the rabbit muscle
(Conway and Koshland, 1968) and yeast (Cook and Koshland,
1970) enzymes that the binding of NAD is a cooperative
phenomena. Kirschener 32 31., (1966) obtained weakly
sigmoidal NAD binding curves at 40° C for the yeast enzyme,
but the binding curves were hyperbolic at20° C. StOpped-
flow experiments have further revealed that the binding of

the first mole of NAD to the rabbit muscle enzyme is faster

 

 

than the hi:
and Slater,

It s
grocess whic
33: exasple,
of the enzy:

~' U. '
‘34 .ue anger

‘.
33 u: (Fens

: v
C- CC'A'IIOtr'f‘;
‘ - a.

the effect 1

Z‘acetamide_

(I)

%

 

 

 

11

than the binding of subsequent moles of NAD (De Vijider
and Slater, 1968).

It seems clear that the binding of NAD is a complex
process which involves conformational changes in the enzyme.
For example, NAD enhances the rate of carboxymethylation
of the enzyme by iodoacetate, while ATP decreases the rate
in the absence of NAD, but has no effect in the presence
of NAD (Fenselau, 1970). Similarly, ATP enhances the rate
of chymotryptic digestion of yeast GAPD, and NAD prevents
the effect (Yang and Deal, 1969b). Kirtley and Koshland
(1970) have also shown that the addition of NAD or ATP to
rabbit muscle GAPD, labelled at the active site with
Z-acetamide-4-nitrophenol, produces large difference spectra.
SUBSTRATE BINDING: GLYCERALDEHYDE-3-PHOSPHATE. GAPD
utilizes only the free aldehyde form of the triosephosphate
substrate glyceraldehyde-3~phosphate in the catalytic
reaction (Trentham gt 31., 1969). Glyceraldehyde-B-
phosphate, however, exists in two forms in aqueous solu-
tion, since the aldehyde may be hydrated to yield a diol
(Trentham gt al., 1969). At equilibrium, 96% of the
glyceraldehyde-3-phosphate exists as the non-utilizable
diol (Trentham 33 31., 1969). The free aldehyde is also
the form of g1ycera1dehyde-3-phosphate utilized by aldolase
and triosephosphate isomerase (Trentham e; 31,, 1969).
SUBUNIT STRUCTURE OF NATIVE GLYCERALDEHYDE-B-PHOSPHATE

DEHYDROGENASE. While the molecular weight of Yeast and

 

 

 

rabbit muscl:
there does :1:
zolecular we;
;:e'.'iO'asly pf

have rePOrte:
also reportef
156,303.

There
:13ch (Deal
1‘24 yeast (3;
3‘," a VariEty
35'000“38 .07
(Earl-is and II

“‘3 native

 

12

rabbit muscle GAPD has been examined in a number of studies,
there does not seem to be any general agreement on the
molecular weight of the native enzymes. This has been
previously pointed out in a review of the data by Elias
gt $1., (1960). Table I; summarizes the molecular weight
values obtained in a number of studies on the rabbit
muscle enzyme. Similarly, for the yeast enzyme Jaenicke
etflal., (1968) have reported values of 143,000--148,000,
while Deal and Holleman (1964) and Taylor and Lowry (1956)
have reported 116,000. Harrington and Karr (1965) have
also reported values for the pig muscle enzyme of 139,000-—
156,000.

There is agreement, however, that both rabbit
muscle (Deal and Holleman, 1964; Harrington and Karr, 1965)
and yeast (Deal and Holleman, 1964) GAPD subunits, produced
by a variety of treatments, have a molecular weight of
35,000--38,000. It is clear from these and other studies
(Harris and Perham, 1965; Meighen and Schachman, 1970) that
the native enzyme is composed of four polypeptide chains,
which are thought to be identical. Harris and Perham
(1965) have shown with the rabbit muscle enzyme that the
amino acid analysis, end-group analysis and peptide maps
obtained after tryptic digestion are consistent with a
tetramer of four identical subunits. Meighen and Schachman
(1970) have further shown that hybridization of succinylated

and native rabbit muscle GAPD produced five hybrids which

 

 

 

 

 

mi: 2:. .1; .JL._..7:..__ cco ms; t.....M.,-.“.41c finanl
mmma .xom Us... uexfiacccc ooo.mvalooo.mma mcwMOOumom 0:qu
GUCCHOCCZ OUSHQ> UGUHOQOE OSUflC£UQE HMUCQEHHQQXQ
CL/Knv 3...,vazz .fiHT—m~$~ 137% mpfiuzuhuéz -<HHDMUT~AHAOZ firefifivfioua~v~v~ cHH. .‘..~5HM~4\.H

13

.Eouuon Hamo on» an msam>

mcfluflfiwa may mm3 ooo.¢va mo msam> on» new .Aonmn mcflnomoummm soflumuucmocoov

msomflsoa on» on cowumaommuuxm an pocfimpno mm3 ooo.~h mo msHm> one

A

.cowunaflp muflcwmcw «0 uHEflH may 0» soap
Imaommnuxm on» com: pocflmuoo ooam> can ma many .mmflpspm may no umOE ch

 

mmma sHmHHmB pom oomammom

mama .Hnmm com coumcflunmm
mmma .xmm.mm OMOHcmwb
meme ..mm mm mxoucmmn
mmma ..mm mm uoamms

omaa suofimme

omma .meflaocmo pom xom
ammo ..mm_mm mmHHm

mmmH ..mm_mm 0x0Hcmob
mmma .Heon

coma ..mm.mm mafiam
mmmH ..wm.mm oxOHsomo
mmmH .xom pom meflaocmo

ooo.~h

s
nooo «Va

ooo.wvatooo.vma
ooo.mva
ooo.~va
ooo.o-
ooo.mHH
ooo.nm~
coo.0NH
ooo.mvH
ooo.mva

ooo.mHH
ooo.mva
ooo.wvalooo.mma

Aeoaumz managemev
sseuneaeswm emmdm seam

Aeoauoz manganese
asflunflaaswm emmmm seam

MHmeOEmO

Esannflaflswm cowumucmeflomm
nonm555fla\coflumuameflemm
GOHmSMMHQ\soMumosmEHUom
seemsmmaa\aoflumpcmsnemm
conSmmwQ\coflumvcoEapmm
:oflm5mmw0\sofluoucoaflpom
cofim5mmwn\cowpm»smsflpom

“camoflzohdv Esfluoflaflsvm
cowumucoeflvmm Op somoumm¢

mcflumuumom names
oceumuumom nemeq

 

mocmnommm

mosam> omuuommm

msvflccome Hopsofiwummxm

 

Dmfiw mqumDZ BHmmdm mom mBEUHMZ qusvmqoz QMBmOmmm .HH mqmda

 

could be eleC‘
iezsnstrates <
:c-zposed' of f<

The m
;r. the molecu
enzgfites under
ghenonena. T'
:gscle enzyme
eiider dimers

M

r; . '
..-._.erimenta l

 

(“cal and H01
W10 36,000 p—I
a Soluthrl

 

14

could be electrophoretically distinguished. This result
demonstrates quite clearly that the native enzyme is
composed of four polypeptide chains.

The most reasonable explanation for the discrepancies
in the molecular weight data is that the yeast and muscle
enzymes undergo some type of association-dissociation
phenomena. This has been clearly shown for the rabbit
muscle enzyme, which can be reversibly dissociated to
either dimers or monomers (depending on the choice of
experimental conditions) in dilute salt solutions at low
temperatures (Constantinides and Deal, 1970). Similar
behavior has also been suggested from other studies
(Hoagland and Teller, 1969; Meighen and Schachman, 1970).
The rabbit muscle enzyme is also reversibly dissociated
in the presence of ATP to dimers, or monomers, the Species
obtained depending on experimental conditions such as
protein concentration (Constantinides and Deal, 1969).

In addition to the dissociation produced by low
temperatures and dilute salts or ATP, the GAPD from yeast
(Deal and Holleman, 1964; Deal, 1969) and rabbit muscle
(Deal and Holleman, 1964) can be reversibly dissociated
into 36,000 molecular weight subunits by exposure to 8 M
urea solutions. In both cases, the cofactor, NAD, is
required for the reversal of dissociation. The implications
of these findings have been discussed in detail (Deal,

1969).

The 1
521385 is V1

as: muscle e

3341025 15 q
... .(C'

V 3
r.rnien tne

1:5“.- muscle

.9 spinach

it)

gross stz
Sizilar. 170.
“PM (1964)

Mix .
....1n tne ra

 

 

 

15

The subunit structure of GAPD from a number of
sources is very likely quite similar to that of the yeast
and muscle enzymes, since comparative studies have indicated
that the structure of the native enzyme from a number of
sources is quite similar. Allison and Kaplan (1964) have
purified the native GAPD's from rabbit muscle, beef muscle,
human muscle, chicken muscle, turkey muscle, pheasant
muscle, halibut muscle, sturgeon muscle, lobster muscle,
yeast and E, 9911. The corn, wheat, oat and barley
enzymes have also been studied (Hageman and Waygood, 1959).
The spinach leaf enzyme is different in that it uses NADP
and has a higher molecular weight (Yonuschot 32 31., 1970).
The gross structure of most of these enzymes seems quite
similar. For example, the enzymes studied by Allison and
Kaplan (1964) all have sedimentation coefficients, 3°20'w,
within the range 7.44 i 0.15 S. Because of this structural
similarity, it is eXpected that the studies reported in
this work with the yeast enzyme may have application to
other systems.

It is also interesting to note that the subunits
of the lobster GAPD, which is quite similar to both the
yeast and rabbit muscle enzymes (Allison and Kaplan, 1964),
appear to be geometrically arranged at the vertices of a
tetrahedron, since this arrangement is most consistent

with the D dihedral symmetry observed by X-ray diffraction

2
(Watson and Banaszak, 1964). At least four other tetrameric

 

I“- . f
SIC-281.55: 15”
A

Q ‘ I ‘ ‘
.aCtic denyflr

1970). Kirsc
Stufiies indi’:
tetramer res».

:cur equiva 16?

In a
139 different
is indicativa
cf enzymes.

Silggested rel

regulatorv t. *

L963; Koshla:

p: CESSeS ha.
E‘Jamate the
i (‘3

 

16

proteins, isocitrate 1yase, aldolase, hemoglobin and

lactic dehydrogenase possess similar symmetry (Klotz gt 31.,
1970). Kirschener (1968) has pointed out that the X-ray
studies indicate that the over-all structure of the native
tetramer resembles a dimer of dimers more closely than

four equivalent monomers.

LITERATURE REVIEW: SUBUNIT ASSOCIATION

 

In a recent review, Klotz at 31., (1970) have listed
109 different proteins which are composed of subunits. This
is indicative of the great interest in the subunit structure
of enzymes. In large part this interest stems from the
suggested relationships between subunit interactions and
regulatory properties of polymeric enzymes (Monod et 31.,
1963; Koshland and Neet, 1968; Atkinson, 1966).

A large number of polymeric enzymes undergo
reversible association phenomena of the type nP = Pn' For
the systems in which the equilibrium constants for such
processes have been obtained, it has been possible to
evaluate the free energy change per monomer, A G°m, where
A G°m is defined as A G°/n, with n being the stoichiometric
coefficient of the reactant in the reaction nP = Pn (Klotz
et 31., 1970; Langerman and Klotz, 1969). For all such
cases, values of A G°m between -2.1 and -6.3 kcal/mole

subunit have been obtained (Klotz et 21., 1970).

HistC
enr-mlogy tc

:errritted by

£550.35! 9' g
13195 and pr
resorts have

tEZd to 105 8

One C

19.:
11111119 en zyme

N‘AA
a P

....tly there
strated that
iissociation
texperature i
behavior migh
non-polar an:

Harri

4“

maline-trea

er'
ated the C.

L.
Phenome non .

wtained as a

par“

LlCle at 2|

1: 0°

 

the Self.

 

17

LITERATURE REVIEW: COLD LABILITY

 

Historically, it has been a common practice in
enzymology to perform studies at or near 0° C whenever
permitted by the experimental approach being used. This
has often been a successful procedure, since many enzymes
lose activity at higher temperatures for a variety of
reasons, e. g., heat denaturation, oxidation of sulfhydryl
groups and proteolytic degradation. A number of recent
reports have nevertheless demonstrated that some enzymes
tend to lose activity more readily at 0° C than at room
temperatures.

One of the earliest discoveries of such a cold-
labile enzyme was Jack bean urease (Hofstee, 1949).

Shortly thereafter, von Hippel and Waugh (1955) demon-
strated that casein undergoes a temperature-dependent
dissociation to monomers of molecular weight 15,000 as the
temperature is lowered to 1.5° C. They suggested that this
behavior might arise from the unusually large number of
non-polar amino acid residues in the casein molecule.

Harrington and Schachman (1956), working with
alkaline-treated tobacco mosaic virus (TMV) protein demon-
strated the dramatic effect of temperature on aggregation
phenomenon. Thus, a 133 S TMV protein particle, originally
obtained as an alkaline degradation product of a 194 S
particle at 25°, dimerized to a 170 S species at 25°, while

at 0° the same 133 S particle dissociated to a 94 S species.

I
'~--“pr (196:

fignav

. ‘ .
:e:er.oent po.
0

ever, if thi .-_

Q l . I
9- vnu- r~ fip~ .—
vag'u-e.‘za L..- -
l d

‘
n
:1“ l
’vA~ \' .
‘ v n .
..4,. .Ie J-C:J':
u.

 

 

 

 

18

Lauffer (1962) has also investigated the temperature-
dependent polymerization of TMV protein, and shown that
the pH of unbuffered solutions of TMV protein shows a
temperature dependent pH change. It is not clear, how-
ever, if this change is the cause of, or result of, the
polymerization.

An interesting study of glutamine dehydrogenase

from Neurospora crassa was reported by Fincham (1957).

 

In this study a mutant enzyme was shown to undergo a
reversible temperature dependent inactivation while the
wild-type enzyme did not exhibit this behavior. It thus
seemed that minor changes in primary structure, possibly
produced by a single point mutation, could affect the
temperature sensitivity of the enzyme.

More recently, Kuczenski (1970) and Jarabek et_al.,
(1966) have tabulated a number of proteins which are
inactivated at 0°. A partial list of these enzymes is
presented in Table III. It is readily seen that a variety
of enzymes from many different sources undergo cold
inactivation. In many of these cases, inactivation is
affected by metabolites.

Most of these enzymes are made up of subunits,
and for the examples marked with an asterisk in the
preceeding table, the cold inactivation is known to be
accompanied by dissociation. Kuczenski (1970) has further

pointed out that in most cases inactivation (measured as a

1-

 

 

 

>-
_.
K
‘s
.v.
1‘ .
I

 

‘
4.
“1"
.‘
“A.
v
a“;

 

TABLE III.

i 3.1 -__.

 

Enzyme

_.“=——‘-;_:: —-—.—. .:___._-——-. : :......---=r t——d—«"H":—_._._—' “ “I: ._- -

1.9

ENZYMES WHICH UNDERGO COLD-INACTIVATION

 

 

Source

-.._....._.1- ..

Effect of Metabolites
on Inactivation

 

. a *
Pyruvate kinase '

*
FDPaseb'

Carbamoylphosphate
c,*

synthase

*
ATPased'

l7-5-hydroxysteroid
dehydrogenasee

Carbamoylphosphate

synthetasef'

Glucose-6-phosphate
*
dehydrogenaseg’

h
Pyruvate carboxylase ’
i
Glycogen phosphorylase a
. . j,*
Arginnosucc1nase

. . k
Nz-f1x1ng enzyme

Glutamic acid decarboxylase
B-hydroxybutyric acid

m
dehydrogenase

Glyceraldehyde-3—ph05phate

n
dehydrogenase ’

ATPaseo

*
Fatty acid synthetasep’

UDP-galactose-A-epimerasecI

I

*

Baker's yeast

Candida utilis

 

Rat liver

Beef heart
mitochondria

Human placenta

Frog liver

Human erythrocytes

Chicken liver

Rabbit muscle
Steer liver
Clostridium
Pasteurianum

E. coli

 

Rhodospirillum
rubrum

 

Rabbit muscle

Yeast
Chicken liver

Bovine mammary
gland

FDP enhances
AMP required, FDP retards

ATP retards, g—acetyl-L—
glutamate enhances

NADH retards, l7-E—estradiol

retards

Acetvlqlutamate required

NADPH and AMP retard

Acetyl CoA, pvruvate, ATP,

nco3 retard

Glycogen, PLP, and AMP retard

Arginine and argininosuccinate
retard

PLP retards

NAD retards

ATP enhances, NAD prevents

 

*
Prolyl ERNA synthetaser’ E. coli - -
Threonine deaminaseS 3. subtilis PLP enhances
aKuczenski and Suelter, 1970. bRosen t al., 1967.

CGuthohrlein and Knappe, 1968.

eJarabak 33 31., 1966.

gKirkman and Hendrickson, 1962.

1Graves 31 31., 1965.

kDua and Burris,

mShuster and Doudoroff, 1962.

1963.

ORacker 93 31., 1963.

quai 33 al., 1970.

.—

dPullman et a1.,

1960;

Penefskyflgnd—Warner, 1965.

fRaijman and Grisola, 1961.

h

Scrutton and Utter,

1965.

jHavir 33 31., 1965.

lShukuya and Schwert,

1960.

nConstantinides and Deal, 1969.

pHsu and Yun, 1970.

rLee and Muench, 1970.

sHatfield and Umbarger, 1970.

* I Q o o O g Q I
Inactivation 15 accompanied by dissoc1ation into subunits.

 

 

"action of t
which the sei
me, in man;

fiissociation

In a
fat decreasr.
:yirophobic :

in their nati

 

20

function of time) appears to be a bi-phasic process, in
which the second step is frequently irreversible. Further-
more, in many cases the cold inactivation and accompanying
dissociation are reversed by raising the temperature above
20°.

In a classic paper, Kauzmann (1959) pointed out
that decreased temperatures could lessen the strength of
hydrOphobic interactions involved in maintaining proteins

in their native configuration.

 

"m ’m~
adlCILRAL

OF GL1

 

CHAPTER TWO

STRUCTURAL STUDIES OF THE DISSOCIATION AND REASSOCIATION
OF GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE

IN THE PRESENCE OF ATP

21

 

- 1
'1 a”; .'
,1": '1“: N

». thQ‘L -1 .a}.

L
f I ._ 1.. I
— — —_ —

 

ABSTRACT

 

Tetrameric yeast glyceraldehyde-3-phosphate dehy-
drogenase undergoes a time-dependent inactivation at 0° in
the presence of adenosine 5'-triphosphate as a result of
dissociation into subunits. The dissociation has been
studied at 0° in 0.2 M_Tris buffer, pH 8.0, with varying
amounts of adenosine 5'-triphosphate. Inactivation studies
which measure the loss of catalytic activity indicate that
an adenosine 5'-triphosphate concentration of 0.5 mM_produces
the half-maximal effect. An ultracentrifugal analysis of
the dissociation reveals in addition to the native tetramer
a 3.0 S subunit (produced by incubation times less than
12 hrs), a 1.6 S subunit (24 hr incubation) and a 0.9 S
species (3—7 day incubation). No dimer is observed in
the analysis. Structural changes are also indicated by
the appearance of an electrophoretic band with a mobility
less than that of the native enzyme. The 3.0 S subunits

can be reassociated to the native 7.5 S tetramer.

22

 

 

Name:
acleotides c
25:9 been re;
1365, 1966; El
azphasis has
35"} and 3' ,S'
PiCSphofructc
Itential abi
rate 0f leCO
Vinsour and M
Williamson, 1
:1 1963)

"5.35 lesser i

 

“3&5 on Othe

we have. her

23

INTRODUCTION

Numerous examples of the effects of adenine
nucleotides on the regulation of carbohydrate metabolism
have been reported (for recent reviews, see Atkinson,
1965, 1966; Stadtman, 1966; and Wood, 1966). Great
emphasis has been placed on the influence of ATP, ADP,
AMP and 3',5'-AMP on the activity of phosphorylase,
phosphofructokinase and fructose diphosphatase and their
potential ability to produce major fluctuations in the
rate of glycolysis (Krebs and Fischer, 1962; Cori, 1942,
Mansour and Mansour, 1962, Passonneau and Lowry, 1962,
Williamson, 1965; Taketa and Pogell, 1963, Bonsignore
31 31., 1963). The possibility of similar (although per-
haps lesser in magnitude) effects of these adenine nucleo-
tides on other glycolytic enzymes has largely been ignored.
We have, therefore, undertaken a study of the effects of
metabolites on the control of several important glycolytic
enzymes, including the glyceraldehyde-3-phosphate dehydro-
genases (GAPD) from yeast, rabbit muscle, and liver. Our
initial studies have shown that ATP produces two separate,
major effects with yeast GAPD, both of which affect its
activity. One effect, which is due to slow structural
change and which leads to 1333 3; activity, is described
in this communication. The other effect, which is due to

ATP competition with NAD for binding sites, and which leads

 

to instantaneous inhibition 31 activity is to be described

elsewhere (Yang and Deal, 1969a).

 

The In
ghenornenon of

'3'; in vitro 1

 

12:01.6 8 so

tion and pres

‘:;~r~4 .
*~:-Guatlon IT;

:53 or ACTIv
N

 

“ith ATP 51109:

SCmII‘ing Wit}
initially a s
comicted wit;
Of 4‘1/2 hOUrg

5‘ o.
.0 1110.2 1

\
I»
1;,“

 

 

 

24

The main purpose of this paper is to report the
phenomenon of destabilization of yeast GAPD, as evidenced
by 13.y1133 inactivation and dissociation of the enzyme
into 1.6 S subunits. One strong incentive for character-
izing the process was the possibility that the destabiliza-
tion and presumed increased susceptibility to proteolytic

degradation might be a mechanism for control 13 vivo.

RESULTS
LOSS OF ACTIVITY UPON INCUBATION WITH ATP AND PRODUCTION
OF A SLOWER-MOVING COMPONENT IN POLYACRYLAMIDE DISC
ELECTROPHORES13. The initial tests for inhibition of
yeast GAPD by ATP revealed that after incubation with ATP,
the enzyme showed a marked loss of activity (assayed in
the absence of ATP). The fact that the loss of activity
of GAPD increased significantly with time of incubation
with ATP suggested that marked structural changes were
occurring with time. To define the structural change,
initially a study of the effect of ATP concentration was
conducted with an arbitrarily selected incubation time
of 4-1/2 hours. Protein samples (0.2 mg/ml) were incubated
at 0° in 0.2 M tris (HCl) buffer, pH 8.0, with varying
amounts of ATP. After 4-1/2 hours aliquots were removed
and assayed in the standard assay. Since all but a
negligible residual amount of ATP was diluted out, any
activity loss should have been due to structural changes

which were not immediately reversible. As seen in Figure 1,

25

 

 

.“XTB mmm .mfid Sfifl»? COHfiMQDOQH HS N\H w

noumm ommcomoucmgoc opmnmmonmnmnopwnooamnoowam ammo» mo mpfl>flpom

.m onsmflm

 

26

 

 

 

 

o. .. m 4 n a _ o
q q q .4 1 q q 1 1 o
JON
0" :I i
o I. now
/.
/O
(100

1.4:
8

2.: «x. 4 .oo

1
O
2

Om<0 bm<u> “.0 20....<>_._.U<Z.

 

 

(.-9w-.-NIw-sa1owm mnv 'DadS

 

at ATP cone
of activity
5.

. *3 23Xm131.13?

tions of l-;

(V

9, I
m

Qrevented, t

he appeal-a;

This
at

'48an the
.15 enzyme.

ATP might ca:

Si: , .

Submit
8 by 1

JEal

I 1969) .
tie SlOWer‘nr

t
'19 native e»

 

27

at ATP concentrations of 2 mM or higher, about 50% loss

of activity occurred after 4-1/2 hours incubation at 0°.
The maximum ATP effect appears to be exerted by concentra-
tions of 1-2 mM; above this the curve levels off.

After activity measurements were performed as
above, aliquots were removed from the tubes and subjected
to polyacrylamide (6-1/2%) disc electrOphoresis at pH 8.3.
The sample without ATP showed only a single native band
(Figure 2). All other samples showed an additional,
slower-moving band whose concentration apparently was
greater at higher concentrations of ATP (Figure 2). Again
the maximum change appeared to be exerted by all concentra~
tions greater than 2 mM ATP. In contrast, incubation of
GAPD with an NAD concentration equal to that of the ATP
prevented, to a large extent, the loss of activity and
the appearance of the more slowly moving band upon electro-
phoresis.

This set of eXperiments provided the first direct
evidence that ATP produced a marked structural change in
the enzyme. A tempting explanation of the data was that-
ATP might cause dissociation of the enzyme into subunits,
since this enzyme has been shown to be dissociated into
subunits by urea (Deal, 1963; Deal and Holleman, 1964;
Deal, 1969). However, the alternative possibility that
the slower-moving species was an aggregated product of

the native enzyme could not be ruled out.

 

28

 

 

 

.uxms mom
.mad mo mucsoEm mooflnm> spas coflumnsozfl H: «\H v Hmwmm mmmcmmoupmsmp

momsmmonmumImpwsmpamumomam ummm> mo mcumpumm oaumnocmouuooam .M musmwm

29

a... min
Ru. 5 o... ad. 3 fid .3 36 o

.

 

----‘A"‘.

I-

-‘t--t"

. 1'... I II...

‘& .III
4 l

+

“A .u m P‘v
vavaOIJO‘
"ose DOSS.

were perfor

.P
“““““

0? 24 hours
3'31 and O.C;

3‘ a Simila:

 

 

 

30

PRODUCTION OF-A 3.0 AND A 1.6 S SUBUNIT. To decide between
these possibilities, sedimentation velocity experiments
were performed. It was desirable for the sedimentation
experiments to increase the protein concentration to

2.5 mg/ml. The concentration of the dissociating agent,
ATP, was made higher to help compensate for the unfavorable
effect of increased protein concentration on dissociation.
Samples of GAPD (2.5 mg/ml) were dialyzed at 0° for 4-1/2
or 24 hours against a solvent containing 10 mM ATP, 0.15 M
KCl and 0.02 M imidazole, pH 7.5. Native enzyme, treated
in a similar manner, but without ATP, served as a control.
As seen in Figure 3A and BB, the ATP—treated samples showed,
in addition to the usual 7.5 S native peak, a slower-moving
peak, suggesting a dissociation into subunits. It should
be noted that the sedimentation coefficients of the slower
moving peaks were 520,w = 3.0 S and 1.6 S, respectively,
for the samples incubated 4-1/2 hours and 24 hours.

To determine whether this might be a physiologically
significant process, two conditions used in the previous
experiments were changed to more closely approximate in
vivg conditions. The ATP concentration was lowered to
2 mM and the experiment was run at 15-16°. As seen in
Figure 3C, incubation of the enzyme with ATP under these
conditions for about 48 hours produced about 20-25% dis-

sociation. Although, the dissociation at 15° with 2 mM

 

 

' S.","

‘ '3:-

 

31

.mcoflpmnsosw
on» mo mmonu ow omoHo mousumummEmu um meHOMHmm onw3 moon madmanpcmo map

paw .mHQEMm 50mm How HE\mE m.m mmS coflumupcmocoo samponm one .me4 as m CH omH
um mflmflflu E 2. no sumpumm .92 ms 3 fi 00 pm 333% .3 em 5 883mm
.me< we OH 2H 00 um mflmMHmflp u: v um cumuumm .mcofluwosoo mooaum> Hops: mam

rhea coflpmnoosfl umumm ammo pmmmm mo mcuouumm wufloon> coepmucmaflpmm .m.musmflm

 

32

 

 

 

‘

 

I
ATP was not

ATP, there
tion.

SE33.;ET<TAT:
eezszr—cce 03
indicated t
it 03' it W
change by a
directly. '

5t 2° in tin

3" bpeed c.
The
* «tial CC

 

 

33

ATP was not as rapid or extensive as that at 0° with 10 mM
ATP, there clearly was a major destabilization and dissocia-
tion.
SEDIMENTATION EQUILIBRIUM STUDIES OF YEAST GAPD IN THE
PRESENCE OF ATP. While the sedimentation velocity studies
indicated that GAPD was dissociating into subunits in ATP
at 0°, it was also desirable to investigate this structural
change by a technique which yielded molecular weight values
directly. Therefore, the molecular weight of GAPD samples
at 2° in the presence of ATP was determined by conventional
low speed sedimentation equilibrium.

The molecular weight values of four GAPD samples
(initial concentrations of 1.0 mg/ml to 2.5 mg/ml) at 2°
in 5 mg ATP are illustrated in Figure 4, which is a plot
of the apparent weight-average molecular weight vs r2, the
square of the radial distance. The most distinctive
feature of the plot is the observed decrease in molecular
weight in going from the bottom of each cell (highest
protein concentration) toward the meniscus (lowest protein
concentration). This observed decrease in molecular
weight with decreasing protein concentration is consistent
with a dissociation of the native GAPD tetramer into sub-
units, since dissociation would be favored by lowering
the protein concentration.

The most.striking feature of this data, however,

was that the molecular weights observed at the meniscus

 

 

{YUP “—v w' - T—V‘r'f.

 

 

.o.m mm .93 m mood "Em Hofimnfioflfiflc m «0.0

34

.mflue.$ No.0 mos ucm>aom “2mm vam mo pmmmm m can om um meHOMHmm mmz

sou Esflunflaflowm coflumucmaflpmm one .mad mo mocmmmum on» cw mmmcwmompwcmp

oumzmmosmnmsopmzmpamumowam ummom mo mflmwamcm usmwms Hmasomaoz

.w musmam

 

35

 

..333 a d.
:30! on O
dixoi 3 o

aiin o.— O

 

8
“w

0!)
N

 

 

In
N

8
gmumaavu v)

“'9
N
—

Om—

 

 

{20,000 to ‘
:clecular
since the T
(Jaenicke [
tility that
screehow be:

1: ATP.

EXperiment a

 

36

(20,000 to 35,000) were considerably lower than the minimum
molecular weights expected for the monomer of GAPD (35,000),
since the molecular weight of the native tetramer is 145,000
(Jaenicke gt 31., 1968). These results raised the possi-
bility that the subunits of 35,000 molecular weight were
somehow being converted to a lower molecular weight species
in ATP.

To test this possibility, a sedimentation velocity

 

experiment was performed with the stock ATP-GAPD solution
(10 mg/ml), aliquots of which had been diluted to prepare

the samples for the sedimentation equilibrium experiments.

 

Since this stock solution had remained in the cold room
during the equilibrium run, and while the equilibrium
calculations were being performed, the GAPD solution

(10 mg/ml) had remained at 2° in 5 mg ATP for 7 days prior
to the sedimentation velocity experiment described in the
next section.

PRODUCTION OF A 0.9 S SUBUNIT. A sedimentation velocity

 

experiment using the GAPD sample which had been incubated
for 7 days with ATP at 0° revealed only two peaks, which

had sedimentation coefficients (3 ) of 7.7 S and 0.85 S

20,w
(Figure 5). The value of 7.7 S for the faster sedimenting
peak is the expected value for the native enzyme, but the
value of 0.85 S is considerably lower than the value of

1.6 S observed in 8 g_urea for subunits of 35,000 molecular

weight (Deal and Holleman, 1964). One explanation for

 

 

I 'u‘wx r“

37

.maflmump umanSM

How uxmw mom .m n.h mo ucmfloflmwmoo coflumucmfiflpmm m mm: “med mscfisv cumppmm
umsoa mnp cw xmmm was .mao>fluommmmu mxmmm mcflaflmuu paw mcflpmma may Hem m mm.o
pcm m h.> mum A3.0mmv mucwHOHmmooo coapmucmefipmm may Adam £uw3v cumpumm “odds
map cH .om um meM0mHmm spas can m@SMHHusmo paw coflumnsoofl mnp cpom .me¢
we m Acmmuumm HmBOHV usonufl3 Ho Acumuumm nmmmov cuflB mmmp b MOM coHumnoocH

umpmm AHE\mE OHV ammo ummmm mo mcumupmm mufloon> :oflumucmEHpmm .m mnsmflm

 

 

38

 

39

this observation and for the results from the preceding
molecular weight analysis was that in the presence of ATP
the 35,000 molecular weight subunits were being converted
to a lower molecular weight species by some unknown
mechanism. Alternatively, the sedimentation coefficient
of 0.85 S could suggest a more unfolded conformation

than that observed in 8 g_urea, but this would not explain
molecular weight values of 20,000 obtained in the previous
sedimentation equilibrium experiments.

At this point, however, there was another possible
explanation for the observation of the 0.85 S protein
species, based on the conditions under which the preceding
experiments had been performed. While ultra-centrifugal
studies are routinely performed with solvents containing
0.10 g_to 0.15 g_sa1t (usually KCl or NaCl) to "swamp out"
any electrostatic effects, 32 additional salt had been
added to the buffered solvent used in the preceding
sedimentation equilibrium and sedimentation velocity
experiments. Salt had been purposely omitted in these
experiments because other studies, which are fully described
in Chapter Four of this thesis, had indicated that increas-
ing ionic strength inhibits the over-all dissociation of

tetramers to monomers.1 Therefore, it was possible that

 

1Since this inhibition of dissociation by ionic
strength was not discovered until after the initial sedimenta-
tion velocity studies, i.e., the studies which demonstrated
the production of the 3.0 S and 1.6 S species, 0.15 g_K01
had been present in those experiments. Therefore, the
observation of the 3.0 S and 1.6 S species is not subject
to this criticism.

 

40

the observed sedimentation coefficient of 0.85 S arose from
a slowing of the sedimentation of 35,000 molecular weight
subunits due to the formation of an electrostatic gradient
in the centrifuge cell as the protein molecules sedimented.
This criticism also applied to the sedimentation equilbrium
studies, which were also performed in the absence of salt.
To test this possibility, sedimentation velocity experi-
ments were performed in the presence of 0.1 g KCl.

When further sedimentation velocity experiments
were performed with GAPD samples which had been incubated
at 0° in 5 mg ATP and 0.1 M.KC1 for 3-7 days, a species
with a sedimentation coefficient ($20,w) of 0.95 S was
observed, in addition to the native enzyme. This result
indicated that the observed sedimentation coefficient of
approximately 0.9 S was not an artifact resulting from
electrostatic effects.

Control samples of native GAPD minus ATP, which
were incubated at 0° for 3-7 days, sedimentated predominantly
as the native tetramer but also contained a small amount of
a more slowly sedimenting species. The concentration of the
slowly sedimenting species in these control samples (minus
ATP) was much lower than in the GAPD samples with ATP. The
concentration was so small that the sedimentation coefficient
of the slowly moving species in these controls (minus ATP)
could not be accurately measured.r However, experiments in

which two GAPD samples (with and without ATP) were run in

 

. ‘5. l a .P\ ,4 u \ V A c A 1.:

Y. ",4. nu a v. «Q n v .G 3..
a -7. .(A .l 0. .nn 0.. 1 i T. UV...
“.4. «Nb «3 r ab .12 d. a ,V.
.C .C at 5 ad .o. . A: A. e . . . ~ . q M
n: D; ,‘L I . ‘3 w». an a: \J. a...

nNy

 

~truct
"file I

n
s

 

41

separate cells in the same rotor suggested that the slowly
sedimenting peak in the controls corresponded to the 0.9 S
peak in the ATP sample. Therefore, it is clear that ATP
is in some way responsible for the production (or at least
for an increased rate of production) of the 0.9 S material.
At present, the exact nature of the 0.9 S species and the
mechanism of its formation are not clear, but further
molecular weight studies may provide a more precise
explanation.

REVERSAL OF DISSOCIATION: PRODUCTION OF A 7.5 S TETRAMER

 

FROM A 3.0 S MONOMER. One major question was whether the

 

structural changes, i.e., dissociation into subunits by
incubation of GAPD at 0°, produced in the presence of ATP
were reversible. Initial attempts to reverse the dissocia-
tion by removal of the ATP or by raising the temperature

to 23° were unsuccessful, since catalytic activity was not
recovered. Further studies revealed, however, that 3.0 S
subunits could be reassembled to yield a fully active
tetrameric enzyme, under the proper conditions. However,

both dissociation and reassociation processes are discussed

 

 

in great detail in the following chapters of this thesis,
and therefore only a minimum amount of data necessary to
demonstrate that 3.0 S subunits can be reassembled to
7.5 S tetramers is presented in this section.

While the sedimentation velocity studies reported

in the preceeding sections were performed at protein

42

concentrations above 2.0 mg/ml, reassembly experiments were
performed at concentrations an order of magnitude lower.
Lower concentrations were used for these studies because
Deal (1969) demonstrated that reversal of dissociation of
GAPD, from 8 §_urea, decreases at protein concentrations
above 0.04 mg/ml. Therefore, a sample of GAPD (0.2 mg/ml)
was completely dissociated2 by incubation at 0° in 5 mfl_
ATP (0.2 g Tris, pH 8.5 and 0.1 §_B-mercaptoethanol) for
6 hrs.

When a solution of GAPD dissociated at 0° by this

3 (by the addition

procedure was (1) made 7.5% in sucrose
of solid sucrose at 0°), (2) warmed to 23° for 2 hrs, (3)
diluted 1:1 with the above buffer, and (4) layered on a
sucrose gradient, the GAPD activity sedimented as a 7.5 S
species. This is illustrated in Figure 6. It is seen in
Figure 6 that the GAPD activity coincided with the 7.5 S
activity peak of the lactic dehydrogenase (LDH) marker
(Kaplan, 1964). For comparison, the sedimentation profile
of a GAPD control sample treated in a similar manner, but

minus ATP, is shown in Figure 7. This experiment proved

that inactive 3.0 S monomers could be reassembled to yield

 

2This sample had lost all activity after 6 hrs and
the dissociation was therefore considered complete.

3Sucrose was found to be required for reassembly.
The effect of sucrose on reassembly is described more
fully in Chapter Three, and the theoretical implications
of this effect are discussed in Chapter Four.

 

43

ll- ll1l\ > UC q‘C-h<~UOWW<W~‘

.uxme mom .coapmmSMflpusmo mHOMmQ omouoom mo mocmmmnm may cw omm
ou msafiums mp meMHoommmmu cam med cw pmpmeoommflc ommcmmoupmnmp mumnmmonmlm

m mmouoom .m ossmfim
Imcmnmpamumomam mo cumupmm coaummSMHHpcoo ucmflpmnm whflwc p

 

 

44

.—Zw.0<~_0 mmOgunm Z. om4m><¢~ EU

 

 

9m {N «A o
_ _ _ _
o m o. .02 man» ow mm o
d Ed 1 — a 1 q «wo— 1 q q q — n um: - — q q q d d d 14 d
. .7 7.. .
.1 ..
I .o \.... Ind
a ...
. a... -
... y» 17o
a: z. .
r ofi<_uomm<m¢\w\d .
I 920 .. 0...... 10.0
as...
I «3:22 :3\\V... I no
3x
I J o.—
-_»_.».._.»FL—bppp_..bs—.u.p

 

 

 

...—4‘ Dm<0l> ”.0 ZOZ.<_UOmm<m¢

(1W/Sian) - All/\llDV

45

.uxma mom

oauoEmnumu m>flumc mo cumupmm coflummSMHHpcmo yaowpmnm mpflmsop omouocm .5 on: am

 

.ommcmmoupwnmp mwmzmmogmlmumpwnmpamumowam

46

~2m_o<¢0 wwauam Z. omam><¢h <<U

 

 

 

 

 

em 3. S o
_ a. q _
m 0— 02 won... on ma
JSd-ddqdq—d.dqdqq.q—11«d—Jddu
r ..O 1

«3:22 :3\ L.....

I. 0. l
w.

,. : ..
O O

I 920 u>:<z\‘< I

T C 1
mos

vb.rpppP.—.pb._.kp._~.ppppppr

 

 

 

0m<0|> w>_.—<Z

ad

to

o.—

a;

('IW/Slan) — All/\llDV

 

.
p-‘ n .
‘p‘

HV 0‘ c

-
“TA’V‘ .
“I‘- 0L4 .A\
.

A: "

U. Cl.
F‘: A"!
w. Du.
1'“n:w
“ft/v.44

fin -
cue r:
n:
V o
P?“ I'm

47

an active 7.5 S tetramer. To date, it has not been possible
to reassemble either the 1.6 S or 0.9 S species obtained
in ATP to the native tetramer. Deal (1969), however, has
been able to reassemble subunits of GAPD (presumably 1.6 S)
produced by dissociation in 8 g urea.

Other experiments demonstrated that after warming
of dissociated GAPD subunits (3.0 S) to 23° in the presence
of sucrose, the slower—moving electrOphoretic band dis-
appeared and only the native electrophoretic band was
observed (see Figure 2). These experiments also illustrated
the reversal of the structural changes produced by ATP at

0°.

DISCUSSION
PRODUCTION OF A 3.0 S AND A 1.6 S SUBUNIT. One explanation
for the observation of both the 3.0 S and 1.6 S subunits
is that the enzyme (1) first undergoes only a limited
unfolding, (2) then dissociates, and (3) then unfolds
rather extensively to yield the 1.6 S subunit. It is
interesting that the sedimentation coefficient of 3.0 S
is consistent with that expected for a folded monomeric
subunit of molecular weight about 35,000 and the value of
1.6 S is identical to that found for unfolded subunits in
urea (Deal, 1963; Deal and Holleman, 1964; Deal, 1969).
For this reason, the 3.0 S subunits will be referred to

as folded or globular subunits throughout this thesis.

 

 

48

One interesting aspect of these structural studies
is that a dimer intermediate is not observed in ultra-
centrifugal analysis of the dissociation process. This
aspect of the dissociation process is discussed more
fully in Chapter Four of this thesis.

REVERSAL 0F DISSOCIATION. The discovery that 3.0 S monomers

 

could be reassembled to 7.5 S tetramers was exciting because
it provided a system ideally suited to investigate the

assembly of folded or globular subunits to native enzyme.

 

This is in contrast to most dissociation systems (e. g.,

urea or acid treatment) which produce extensive unfolding

 

of the individual polypeptide chains as well as dissocia-
tion into subunits. The discovery that folded GAPD monomers
could be reassembled to tetrameric enzyme was especially

exciting, since Deal (1969) had reassembled highly unfolded

 

subunits (produced in 8 g_urea) of GAPD to the native
tetramer. Thus, a comparison of the reassembly processes
for the two types of subunits should allow one to separate

folding and association effects for the overall assembly

 

of tetrameric enzyme from individual, unfolded polypeptide
chains. Since the individual polypeptide chains must fold
and assemble in 1129 after they are synthesized on poly-
somes, it is hoped that the studies of the reassociation
of folded subunits presented in this thesis will provide
an insight into the overall process of protein synthesis,

folding and association which occurs in_vivo.

49

MECHANISM OF DISSOCIATION AND THE OVERALL CONTROL OF GAPD.
Because of structural similarities it seems reasonable to
suspect that ATP is binding at an NAD site, and that the
additional phosphate and additional negative charge on the
terminal phosphate of ATP may be responsible for the
destabilization.

The conditions of ATP concentration (Betz and
Moore, 1967), salt, temperature, pH and protein concentra-
tion eXpected i2_yi!g in yeast are resonably close to those
described in this work. Furthermore, since only a small
structural change may be necessary to increase GAPD suscep-
tibility to yeast proteolytic enzymes, the destabilization
by ATP could be a physiologically significant process.
Studies concerning the role that this process might play
in the metabolism of yeast have led to results presented
in Chapter Five of this thesis.

For comparison, it is interesting to note that
rabbit muscle GAPD shows dissociation by ATP (Constantinides
and Deal, 1969), by urea (Deal and Holleman, 1964; Deal,
1969), and by low temperatures at low protein concentrations
(Constantinides and Deal, 1967; 1970) and all of these
dissociation processes are inhibited by NAD.

The stabilization by NAD of yeast GAPD against
ATP-induced destabilization and dissociation provides a
mechanism involving NAD in control of degradation of this
enzyme. These effects, together with the results (Deal,

1967; 1969; Constantinides and Deal, 1967) suggesting a

 

fl..—

C)"

50

role for NAD in accelerating the rate of synthesis of the
enzyme, provide mechanisms for a logical, systematic, and
effective control of both synthesis and degradation of
GAPD by NAD, with the degradative control being shared

antagonistically with ATP.

MATERIALS AND METHODS

 

ULTRACENTRIFUGAL ANALYSIS. All experiments were performed

 

in a Spinco Model B ultracentrifuge equipped with either
phase-plate schlieren optics or Rayleigh interference
Optics. The exact temperatures were obtained with a
calibrated rotor and temperature indicator control (RTIC)
unit. Photographic plates were measured with a Bausch and
Lomb microcomparator.

SEDIMENTATION VELOCITY ANALYSIS. Sedimentation velocity

 

experiments were performed at 56,000 RPM in an An-D Rotor
using wedge and plane single-sector capillary synthetic-
boundary cells. The sample volumes were approximately
0.35 ml, and approximately 0.15 ml were used in the
solvent chamber. After the boundary formation, the total
volume was therefore approximately 0.50 ml. Sedimentation

coefficients were calculated from the expression:

1 rp (t)

 

51

where S is the sedimentation coefficient in Svedberg units,

rp is the peak maximum in the refractive index gradient

curve of the picture taken at time to (first picture taken),
or at time t (later pictures), and w is the rotor velocity

in radians/sec.

SEDIMENTATION EQUILIBRIUM ANALYSIS. Rayleigh double sector
interference cells were used for these studies with an An-G

or an An-D rotor. The molecular weight values were calculated

using the equation:

2 RT 2.303 (d log c)

 

 

(1- ‘70) w2 (dr2>

where T is the absolute temperature, R is the gas constant
of 8.314 x 107 ergs/mole/degree, v is the partial specific
volume of the protein, p is the density of the solution,

w is the angular velocity of the rotor in radians/sec, and
d log c/d (r2) is the slope of a plot of concentration as

a function or radial distance, r, squared (semi-log plot).

ELECTROPHORESIS. Electrophoresis was performed in 7%

 

polyacrylamide at pH 8.3 (Tris/glycine buffer). A 5%
polyacrylamide stacking gel was also used. Protein samples
were applied in 10% sucrose. Approximately 50 pg of protein
was applied per gel. The protein bands were stained with
0.5% amido-schwartz and de-stained with 7.5% acetic acid.

ENZYME PREPARATION AND ASSAY. The enzyme assay and prepara-

 

tion is completely described in Methods and Materials,

Chapter Three of this thesis.

 

 

CHAPTER THREE

SYSTEMATIC STUDY OF THE VARIABLES AFFECTING DISSOCIATION
AND REASSOCIATION OF GLYCERALDEHYDE-3-PHOSPHATE

DEHYDROGENASE AT 0° IN THE PRESENCE OF ATP

52

ABSTRACT

 

Tetrameric yeast glyceraldehyde—3-ph05phate dehy-
drogenase undergoes a time-dependent inactivation in the
presence of adenosine 5'—triphosphate as a result of dis-
sociation into monomeric subunits. Optimal conditions,
which yield complete inactivation in 5 hours are: (1) 1-2
mg adenosine 5'-triphosphate, (2) 0°, (3) protein concentra-
tions of 0.03-O.l mg per ml, (4) pH 9.0, and (5) 0.1.M
B-mercaptoethanol. Transition points (half-maximal loss
of activity in 5 hours) are: (l) 0.5 mM adenosine 5'-
triphosphate, (2) 12°, (3) 0.5 mg per m1 and (4) pH 8.6.
Adenosine 5'-monophosphate and adenosine 3',5'-monophosphate
do not dissociate the enzyme. They, and all the substrates
of the reaction, partially protect the enzyme from dis-
sociation. Dissociation and inactivation are completely
reversed by warming to 17°. Reassembly is greatly stimu-
lated by adenosine 5'-triphosphate and by 10% sucrose.
Optimal reassembly conditions are: (1) 0.04 mg per ml
protein, (2) pH 7.0, (3) 1-2 liadenosine 5'-triphosphate,
(4) 17°, (5) 10% sucrose and (6) 0.1 M_B-mercaptoethanol.
Inactivation and dissociation apparently result from
electrostatic repulsion. The results are discussed in

terms of a possible role for this enzyme in the regulation

53

54

of glycolysis. Since this dissociation produces fairly
compact subunits, association of folded monomers to tetramers

may be studied independently of the polypeptide folding.

55

INTRODUCTION

 

Numerous examples of the effects of adenine
nucleotides on the structure and activity of various
enzymes of carbohydrate metabolism have previously been
reported (Atkinson, 1965, 1966; Wood, 1966; Stadtman,
1966; Scutton and Utter, 1968). Recent investigations
in this laboratory have demonstrated that ATP and other
adenine-containing compounds have pronounced effects upon
the structure and catalytic activity of yeast glyceraldehyde-
3-ph05phate dehydrogenase (GAPD). Glyceraldehyde-3-phosphate

dehydrogenase is inactivated by ATP as a result of dis-

 

sociation of the native tetrameric enzyme into monomeric
subunits (Stancel and Deal, 1968; Stancel, 1969). ATP
and other adenine nucleotides also produce an instantaneous

inhibition of the catalytic activity of yeast glyceraldehyde-

 

3-phosphate dehydrogenase (Yang and Deal, 1969a), which is
measured before any appreciable dissociation occurs.

This paper reports the reversal of the dissociation
of yeast glyceraldehyde—B-phOSphate dehydrogenase by ATP
(Stancel and Deal, 1968) and a detailed characterization
of the dissociation and reassembly processes (Stancel,
1969). Other papers (Constantinides and Deal, 1969) from
this laboratory show that rabbit muscle glyceraldehyde-B-
EHuDSphate dehydrogenase is also reversibly dissociated into
subunits by ATP and that the characteristics of that

systxmnare greatly different from the yeast system.

56

There were several strong incentives for character-
izing the dissociation of yeast glyceraldehyde-3-phosphate
dehydrogenase by ATP. First, the process might be involved
in metabolic control; other work in this laboratory has
shown that incubation of glyceraldehyde-3-ph08phate dehy-
drogenase with ATP greatly increases the susceptibility
of the enzyme to digestion and inactivation by chymotrypsin
(Yang and Deal, 1969b). Also, Williamson (1965, 1967) has
shown that in rat liver and heart, the glyceraldehyde-3-
phosphate dehydrogenase reaction is far from equilibrium
and pointed out that it is a potential control point for
glycolysis. If, as expected, the reaction is also not in
equilibrium in yeast and rabbit muscle, then a remarkable
situation exists, because glyceraldehyde—3-phosphate
dehydrogenase constitutes 20% of the total soluble protein
in yeast (Krebs 33 31., 1953), and about 10% in rabbit
muscle (Cori 33 31., 1948).

In contrast to most dissociating agents, which
extensively unfold the polypeptide chains of yeast
glyceraldehyde—3—phosphate dehydrogenase, ATP produces
a fairly compact subunit. This system seems ideal to
investigate several important biochemical questions:

(1) Can individual subunits exhibit catalytic activity?
(2) Do subunits associate after synthesis or during
synthesis? (3) Do metabolites affect the association

of subunits?

 

57

RESULTS

Except for the particular variable under study,
all experiments in this section used the following pro-
cedures and conditions. The enzyme (0.2 mg per ml) was
inactivated and dissociated at 0° for 5 hr in solutions
containing 0.1 M Tris (HCl), pH 8.5, 0.1 M_8-mercapto—
ethanol, and 4 mM_ATP. Control enzyme samples were
treated identically, except ATP was omitted.
EFFECT OF ATP CONCENTRATION ON INACTIVATION AND DISSOCIA-
TION OF YEAST GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE.
Yeast glyceraldehyde—3—phosphate dehydrogenase showed a
time—dependent loss of activity upon incubation with ATP.
As seen in Figure 8, at ATP concentrations of 2 mM or
higher, about 75% of the activity was lost after 5 hr.
At higher concentrations the curve leveled off. Half-
maximal loss of activity occurred at an ATP concentration
of 0.4-0.5 mM, a value which may be taken as the approxi-
lnate dissociation constant for the enzyme—ATP complex.
The residual ATP added into the assay with the enzyme was
txxo low to inhibit the enzyme (Yang and Deal, 1969a);
'therefore, the activity loss was due to structural changes
VflIiCh were not immediately reversible.

The inactivation by ATP was accompanied by dis-
sociation of the enzyme into a 2.8 S species, while the
enzyme without ATP sedimented as the tetrameric 7.6 S

species. This is shown by the sucrose density gradient

 

 

 

 

58

 

.m94 mo mucoosm

mcflmum> rues mmHmEMm wo mmfluom m once pouoaflp mmB dawned on» can .Hocmsum
noummohmelm m H.o paw .m.m mm .mflue.m H.o as do cmxmu mm3 umaamm may .pmmom
Ifluucoo mmz mEmNcm may mo coflmcmmmsm enamaom EsflcoEEm cm .Hocmnumoummonmelm
.m H.o cam .m.m mm .mflua.w H.o ca mam sufl3 00 am n: m How coepmncocfl Hmumm

AHE\mE m.ov mmmcmmomowzoo mpmnmmosmnmImpxnmpamumomHm mo wufl>flpo¢ .w on: Am

 

59

 

 

 

 

 

 

o. : m 4 m a c

I 4 4 : I . I q . 4 . q 0 S
o . . d
./ [ON .fiJ
V
/ . n

O 0/ 10V .
. )

1 ..fl
.. w
.l
a
S

Om .
m
L N
2.: Tue .09 W
. 9.
b p p p b b p p b p p p (

 

 

 

o ‘.. "-4_-_- ~

 

60

pattern given in Figure 9. It should be noted that since
the 2.8 S species was enzymatically inactive, a reassembly
process (see later section) was used to detect the other-
wise inactive enzyme in the gradient; the fractions col-
lected after centrifugation were warmed to 23°.

EFFECT OF PROTEIN CONCENTRATION. Dissociating-associating
systems are expected to exhibit a strong dependence on
protein concentration. The results shown in Figure 10
indicate that the inactivation of yeast g1ycera1dehyde-3-
phosphate dehydrogenase by ATP increased as the protein
concentration was lowered. Inactivation was complete at
protein concentrations lower then 0.1 mg per ml, and
inversely related to protein concentration in the con-
centration range 0.1 to 1.0 mg per ml. In contrast, the
control enzyme with no ATP was inactivated only slightly
at lower protein concentrations. These data provided
further evidence that the inactivation involved dissocia-
tion into subunits.

EFFECT OF TEMPERATURE. The dissociation showed a marked
«dependence on temperature as illustrated in Figure 11; the
specific activity of the control samples remained essentially
cxmnstant over the temperature range studied. Although no
aurtivity was lost at 23°, a conformational change occurred
as indicated by a difference spectrum.

EFFECT OF pH. Enzyme samples were incubated for 5 hr at

various pH values in 0.1 M Tris buffers. As shown in

 

 

, .,_._ __._ _.__-

61

.maflmymp umnuusw How uxmp mom .meuma m mm
poms mmB cmmcmmonpmnmp OHDOMH Dunc: mmmm .pommmmm pom Mg m How omm on
meHm3 .00 um pmnomaaoo mumz msoflpocuw one .coflpflmomEoo mEmm may cues
mucmflcmum mmouosm mno ca 00 no u: mm Mom pomSMHMpsmo cos“ mumz mmHmEMm

one .Am mucmflmv Eoflpme cofluma00wmflo pHMGQMDm may cw ma«_mfi m usonufla
no sufl3 00 um coaumnsosfl HA m umumm AHE\mE mm.ov mmmcmmoupmnmp mpmnmmonm

Imumownmpamumomam mo coflummSMHnucmo ucmfipmum wpflmsmc mmouosm .N musmflm

 

62

th_o<¢0 mmOgUDm Z. Owdm><~= EU

0d ta N.—
H q
o m o. .02 was» .ou

 

1%4-dwqq..—-«M-\qq—q4idfi.

t<\«

I. 2......

I

r. o.
w 34.

.:<\.........

no.5

I

 

o
_

kpprbrp.~bLP-Frpp.—pppp—ppbp

l

.1

 

 

.:< B 2.3..» .6 20.2.0035 »

o.—

m.—

ON

0N

(“N/SUN“) - All/\llDV

 

63

.mHHouoo Honpuom MOM uxop oom .mfid me v >2 omocomoupmnop opmnmmocdlm

Iopmaopamuoomam wo coeum>fluoocfl co cofluonucoocoo geopoum mo poomwm .dw onomam

 

Se<\0<5 ZO_.—<~:. ZwUZOU 23.—Om.—

 

 

 

 

o._ mud omd mud

- .. ov

r o o cacam“ 00

. x o 0

305200
- o .. 8
u :00—
m a: Too

 

 

o 20.2.6033

- p b . ~ . ~ . — _ — _ — — — _ ‘
a: 2 9:3.» a

 

'DJdS

'1 3V

(1-9W1-le 53 10W")

 

65

.m ousmflm mo pcomoa pom pxoy oom

.mB< we v an omocomoupmnoc opondmozmlmlopws

lopaouoomam mo coflum>fluoosfl co coflponsosfl mo ousumnomEoo mo poomwm .fll onomam

Col 3:25.33

 

on 8 - o.

1 a « s 4 . q . u 4 q 1 1

66

 

macahZOU

. 2:360

 

T
C
.\
O
N

'13V 'DSdS

J
0
q.

1
O
'0

 

 

45;

1
O
O

.. OO—

 

 

dp< >m Om<0|> “.0 20345035

(rewrmw snowd)

 

—
F

.\~

 

.‘-

T13

‘94

.\

I

§

.\
NA

H

C.
nuv.

67

Figure 12, inactivation by ATP was favored at high pH;
the control samples exhibited constant activity over the
pH range examined.

The pH-dependent transition is centered at about
pH 8.6 (Figure 12). The pH dependence was not a result
of titration of the Tris buffer, since samples in Tris
buffers of fixed pH (8.5), but varying concentrations
(0.02-0.2 M), gave similar results, as did other buffers.
SPECIFICITY OF THE DISSOCIATION PROCESS FOR ATP. Several
competitive inhibitors (Yang and Deal, 1969a), including
ATP, ADP, AMP and 3',5'-AMP were examined for dissociating
ability. Samples of glyceraldehyde-3-phosphate dehydro-
genase (0.2 mg per ml) were incubated at 0° for 5 hr in
0.1 M_Tris buffer, pH 8.1, and 0.1 M_B-mercaptoethanol
with 1 mM concentrations of either ATP, ADP, AMP, or 3',5'-
AMP. The control was treated similarly but no nucleotide
was added. As shown in Figure 13, AMP and 3',5‘-AMP had
no appreciable effect. The ADP and ATP samples lost 20%
and 60%, respectively, of the control activity. These
results indicated that the additional phosphoryl groups
on the ATP molecule were responsible for the dissociation.
PROTECTIVE EFFECTS OF 3',5'-AMP1 AMP, ADP, AND SUBSTRATES.
In light of the findings in the preceding section and the
«discovery that ATP and other adenine nucleotides are com-
;petitive inhibitors with respect to NAD (Yang and Deal,

1969a), it was of interest to determine whether the

68

.m onomflm wo pcomoa poo oxou oom .me¢_mE v >3 omocooonpmnop ouogmmonmlm

noomsooaomoowam mo coaum>flpomcfl so mommon mane mo mm mo poommm

 

.Mm

 

ouomam

In

 

 

 

m6 0.0 md Qm mN
4' u 1 1 4 u 1 q < q i q j a . _ e o

m. 0 O/

.. \ O .

u a: $.32. / .. cm

0 0i i

I C 4

.. .. O?

.l o 1 00

x u .......... o o

. 39:on

.. J 8

1 .. OO—

. 3:36 A .

L p p p P r b p b p P F P p b — r

 

 

 

ER >0 Om<0|> ”—0 ZO_.—<.U0mm_o

'DadS

'1 3V

(row l-NIW snow“)

 

7O

.HocoSDooumooHoEIm.a H.o poo .H.m mm
.mHHB w H.o cw moofluooaosc ocflcopm_mfi H cues H: m How 00 no coauonsocfl Houwm

AHE\mE m.ov omocomonownop ouonmmonmlmloomsopamuoo>am mo mufl>flwod alw onomwm

 

71

 

 

 

 

mo<

 

 

 

If... as? F L
anon #
3.53622
02

3.533: as;
mazm-oo

O

O
N

O .
V

O
0

O
Q

00—

 

 

n:< >m 0m<0|> ”.0 20345035

'13V'33d5

(row [..le ssnow")

72

substrates and above compounds protected glyceraldehyde-B-
phosphate dehydrogenase from dissociation by ATP. As in
the previous experiment, samples of glyceraldehyde-3-
phOSphate dehydrogenase (0.2 mg per ml) were incubated
with or without added nucleotides. One control contained
22_ATP; a second control contained only 1 mM_ATP. The
test samples all contained 1 mM_ATP plug 1 mM_additional
nucleotide (or substrate), except for the 10 mM_phosphate
sample. All these compounds protected the enzyme from
inactivation (Figure 14). The order of effectiveness

was 3',5'-AMP > NAD > AMP > Pi > ADP > g1ycera1dehyde-3-
phosphate.

REVERSAL OF THE INACTIVATION AND DISSOCIATION OF YEAST
fiLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE. The temperature
dependence of the dissociation process (Figure 11) sug-
gested a temperature-dependent equilibrium existed and
that activity might be regained by warming the dissociated
enzyme. The initial reassembly studies yielded only about
30% of the control value upon warming the samples to 23°.
With 10% sucrose, there was essentially complete recovery.
The sucrose effect is nonspecific, since maltose, fructose,
or glucose also increased recovery of activity. Studies
with.different concentrations of sucrose indicated the
Inaximal effect was achieved with about 7.5-10% sucrose.

.A sucrose concentration of 10% was used to provide a safe

Jnargin of error. Other experiments indicated that the

73

 

W

 

.MH ousmflm mo pcomoa poo

uxop oom .med wn omocomonpmzop ouwnmmondlmloomnopfimuoomam mo cofluo>flpoocfl

umcflmmo monouumnom poo mopfluooaooc ocflsopo mo mpoowmo o>flpoououm .wm ousmwm

74

 

 

 

 

 

 

 

.L

 

3.3. 3: E as. . q

mar—030:2
02

O

 

 

n34 >m Om<OI> “.0 20_h<.00mm.o

S
.d _
a
om .3
v
3
1.
ow
.m,
w
8 o
...
a
S
8
m
m
8. ...
W
9.

75

reassembled enzyme had the same Km' sedimentation coefficient,
and electrophoretic mobility on polyacrylamide disc electro-
phoresis as the native enzyme.

This surprising observation that the dissociation
was completely reversed, even in the presence of the dis-
sociating agent, raised the question of what role, if any,
ATP played in the reassembly process.

To answer this question a sample of glyceraldehyde-
3-ph05phate dehydrogenase (0.25 mg per ml) was completely
dissociated by a 9 hr incubation at 0° in 5 mM_ATP, and 3
aliquots were withdrawn. One aliquot was layered on a
sucrose gradient without ATP. If ATP were not required
for reassembly, the collected fractions from this aliquot
should regain enzymatic activity if warmed to 23° before
being assayed. This was not observed; only a trace of
activity was recoverable in the monomer position, as
indicated by the lower solid curve in Figure 15.

A second aliquot of the original dissociated enzyme
was layered on a gradient that contained 5 mM_ATP. When
the fractions collected from this sample were warmed and
assayed, substantial activity was recovered (Figure 15,
upper solid curve).

The third aliquot was layered on a gradient without
ATP, but concentrated ATP was added to the collected frac-

tions (to a final concentration of 5 mM) before they were

76

.maflonop Horunom “Om pxou oom .omm no
u: m Low ooumnsocfl corp cam 00 pm couooHHoo ouoz occapomum one .popmoflocfl
ouonS pohuflEo moz med pmooxo .coflwflmOQEoo oEmm osy cues mucoflponm omOHUSm
me 00 no H: mm How pomswfluucoo ouo3 hose .mefl we m npflz #52 Am ouomflmv
Esflpoe cofluofloommflp Unopooum osu ca u: m Low 00 no popmfl00mmflp ouo3 monEom

one .muoEocoE m m.m o>wuoocw mo coeuo>fluooon so med 00 poommm .mM ouomwm

 

77

him—043.0 mmOmUDm Z. omam><zh EU

0.”
_

o

ta u.—
_

_
m o_ .02 we“: on

 

I

I

r.

fl

1

 

dqqq—ddfid—quq—qufi

..2 oz\

.3...
$3 633 .=<\

ER 02

.:< \o

 

70¢: ...Zw.a<¢0

  

I

mQN

bbP-Pbppbbphrb—P%-—~P--b-P

       

o
_

mm

L

C5

In

N.

O
IAIIDV

and I.
A

who )
n

g
(1w/51IN

mu.—

 

 

mh< 2. Om<0l> “.0 20_.—<_00mm<m¢

78

warmed to 23°. The dashed line (Figure 15) shows that
ATP greatly increased activity recovery over that in its
absence.

Thus ATP did not inhibit the reassembly and both
ATP and 10% sucrose had to be present in the reassembly
solution to achieve maximal recovery of activity.

For all the following reassembly studies, the
standard dissociation procedure was changed to a 12 to
14 hr incubation at 0° in 10 mM ATP, 0.02 M Tris (pH 8.5),
0.1 M B—mercaptoethanol with 0.4 mg per ml protein. The
dissociated enzyme samples were then diluted into a reas-
sembly mixture at 0°. The higher concentrations of ATP
and protein in the dissociation mixture compensated for
the dilution, so that the final concentrations under which
reassembly occurred were the same as those for the previ-
ously described dissociation experiments. The standard
final reassembly conditions were: 0.2 mg per m1 glycer-
aldehyde-B-phosphate dehydrogenase, 0.1 M_B-mercaptoethanol,
5 mM.ATP, 10% sucrose and 0.2 M_Tris, pH 7.0.

The activity recovery in the following experiments
was only 80-85% of that previously observed (Figure 11),
presumably owing to the longer incubation times (12-14 hr),
since the subunits appeared to undergo a slow, irreversible
side reaction (half-time approximately 24 hr) at 0°. The
longem'incubation times were required to obtain complete

(lissociation at the higher protein concentrations used.

 

be

s.~\

79

EFFECT OquH ON REASSEMBLY. After incubation at 0°, aliquots
of the inactive monomers were diluted 1:1 into a series of
reversal mixtures at 0° containing 0.4 M Tris buffers of
varying pH. The samples were warmed at 23° for 2 hr and
assayed. As shown in Figure 16, there was a slight depend-
ence of the reassembly on pH in the range of 7.0-8.5 and at
higher pH values the activity recovery decreased even more.
Other eXperiments showed that activity recovery was less

in imidazole than in Tris buffers at given values of pH.
EFFECT OF TEMPERATURE 0N REASSEMBLY. Inactive monomers
were diluted into the reassembly mixture at 0° and warmed
to various temperatures for 1 hr. There was a sharp
dependence of activity recovery on temperature (Figure 17).
These samples were also assayed after 3 and 4 hr at the
indicated temperatures and equilibrium appeared to be
established within 2-3 hr in the temperature range of
0-12°. Presumably tetramers, dimers and monomers are all
present in this equilibrium mixture in this temperature
range. Since incubation at 17° appeared to yield maximum
reassembly under these conditions (Figure 17) it was used
as the optimal reassembly temperature. This is very near
the Optimal temperature for reassembly of the urea-
dissociated subunits (Deal, 1969).

IEFFECT OF PROTEIN CONCENTRATION ON REASSEMBLY. Aliquots
were removed from a stock solution of inactive monomers

«3.4 mg per ml) and diluted with various amounts of

80

.mcorpoz EH confluomop mo .omouosm moH coo
Hoamfiooudmoumsum m To .93. we. m wmfificoo 03.... $353 93. ...E m Low.
omm ou mcHEHma wn AHE\mE N.ov mnemonsm ommcomonomnop oumnmmonmlmnoomgopao

uuoowam mo :oflpofloOmmoou co Amuowmsn mHHB.$ m.ov mm Mo poommm .mw ousmflm

 

81

 

0.0 m m o.» m n 0.x
4 1. _ q 4 a . q a J 1 A .
2< as n .
:z\o¢< «.o
2: a- 2 ..
O
O
/ .
O
J! o
. lb!
0! L! 1
p p t L t b p F p p p P V F r bu

 

 

 

9

ON

on

ov

om

 

 

 

 

ER 2. am<O-> “.0 20:<_00mm<w¢

'33d5

'1 3V

(l-ow [..le snow“)

82

.ma onsmfim mo pcomoa oom .HMOHpcopH oHoB monp oocflm .mQOHponoocH H: v

pom m ocp Mom gamut ma o>noo oco maco

.muHESQSm omocomoupwnoc oponmmonm

Imlophnopaomoomam mo coauoHOOmmooH so onsnouomaou mo uoomwm .Mfl ouomflm

 

83

Co. 2.25.223
mu cu m. o—

 

I

 

P

d 4 q d d I d l1 d 1 1- ‘ fi 4

a? $2 m

.232 «.o o

 

8...— 7/0
3.... 4 62¢ a III... /4/ o\ o\4..
/o\\\0

p F b b p P p b

 

 

 

“:4 2. 9.40.9 m0 20_»<_00mm<m¢

'33dS

'1 3V

(l-ow rmw snow")

84

dissociation medium at 0° before the 1:1 dilution into the
reversal mixture. This kept all parameters except protein
concentration constant. After the dilution into the reas-
sembly mixture, the samples were placed at 17° and assayed
after 1 hr and 2 1/2 hr. The results indicated a require-
ment for a minimum protein concentration, 0.04 mg per ml,
for maximal reassembly (Figure 18). The same value was
obtained for the reassembly of yeast glyceraldehyde-B-
phosphate dehydrogenase dissociated by 8 M_urea (Deal,
1969). This probably reflects the concentration depend-
ence of a dissociation-reassociation equilibrium.

EFFECT OF ATP CONCENTRATION ON THE REASSEMBLY. For this
experiment, the ATP concentration in the dissociation
medium was lowered, since we wished to study reassembly
at much lower ATP concentrations in order to obtain larger
differences in reassembly. Enzyme samples were incubated
at 0° with 2 mM_ATP for 14 hr. Aliquots were withdrawn
and first diluted five-fold at 0° into a series of solu-
tions which were similar to the dissociation medium, but
contained varying amounts of ATP. A second dilution (1:1)
of these solutions at 0° with the reassembly mixture then
yielded a series of similar solutions containing 0.04 mg
per m1 glyceraldehyde-3-phosphate dehydrogenase and various
.amounts of ATP. The samples were then incubated at 17°.
.Maximal activity was recovered with 1-2 mM ATP; half-

Inaximal activity was recovered with 0.3 to 0.4 mM_ATP

85

.mpozpoz paw wH ouomflm mo pcomoa oom .mpflssnom omosomonpmzop ouosdmOSQIm

Iopmnooaonoomam mo coHOMHOOmmooH so coaumuucoocoo :Hopoum mo noommm

 

..mlm 353m

86

4<<\0<< P3305...—

 

 

 

 

o~.o m_.o : o_.o cod no.0
1 q 4 q d —‘ 4 q 4 . d1 d d d d J d
.:< 35....

a:
Silly
arms
lbw :llldllllo o
6. o _.
..1 .Q .1

p P b b p p p b h h p b P P b

0—

ON

on

0?

On

 

 

...—R Z. 0A<Ou> “.0 20:<.00mm<m¢

'lDV 'DJdS

(l-ow rmw snow")

87

(Figure 19). These values are in excellent agreement with
those obtained for the dependence of the dissociation

process on ATP concentration (Figure 8).

DISCUSSION

 

MODE OF ACTION OF ATP. Native g1ycera1dehyde-3-phosphate

 

_dehydrogenase does not undergo any appreciable dissociation
when incubated at 0° under the conditions employed in the
work reported here. Likewise, at 23° ATP does not signi-
ficantly dissociate yeast glyceraldehyde-3-phosphate
dehydrogenase. However, it does cause a conformational
change since an ultraviolet difference spectrum is
observed and the rate of degradation of the enzyme by
chymotrypsin is greatly increased (Yang and Deal, 1969b).
Thus, neither ATP alone nor low temperature alone can
shift the equilibrium of yeast glyceraldehyde-3-phosphate
dehydrogenase toward dissociation, but together they do,
under proper conditions.

Two major questions are: (1) How does the binding
of ATP by the enzyme alter the folding of the polypeptide
chains? (2) What types of forces involved in subunit
bonding would be sufficiently decreased by lowering the
temperature to lead to dissociation?

Electrostatic repulsion between the negatively
charged phosphate portion of ATP and negatively charged
groups at the active center of the g1ycera1dehyde-3-

phosphate dehydrogenase molecule appear to be responsible

88

.mpozuoz poo ma ousmflm mo pcomoa oom .muflcsnom omocomoupenop opmsmmonmlm

Iocwnopaohoomam mo cowpofloommoou co coflumnucoocoo med mo uoommm

 

 

..mm 6.33....

89

as T2...

 

 

 

 

 

(l-ow rmw ssnowd)

 

 

Nd QN in o.N o.—
q a a 1 u 1 q 1 -
._¢<\0(< v0.0
0h—
0
. ‘
— p p b b h n b p p b p p p -
n:.< Z. nun—<0; ”—0 .20:<.00mm<m¢

13" 03:15 .

90

for conformational changes produced by ATP as indicated
by the following: (1) reassembly experiments with urea
dissociated yeast glyceraldehyde—3-ph03phate dehydrogenase
indicate that it contains a region of electrostatic charge
near its active center which prevents proper refolding
unless neutralized by NAD or high salt (Deal, 1969).
(2) The order of the strength of binding of the adenine
nucleotides to glyceraldehyde-3-phosphate dehydrogenase
(Yang and Deal, 1969a) is 3',5'-AMP > AMP > ADP ~ ATP,
but their net charge and dissociation ability is exactly
opposite. (3) Mg2+ ion, which chelates with the negatively
charged phosphate portion of ATP, prevents inactivation of
yeast glyceraldehyde-3-phosphate dehydrogenase by ATP in
yeast g1ycera1dehyde-3wphosphate dehydrogenase solutions
containing both ATP and chymotrypsin (Yang and Deal, 1969b).
(4) Dissociation is favored at higher pH, with a transition
point at pH 8.6.

This transition point suggests the involvement of
a cysteine residue in the dissociation process. A sulf-
hydryl group has been reported to be involved in the binding
of NAD (Racker, 1965), and ATP is a competitive inhibitor
with respect to this substrate (Yang and Deal, 1969a).
Also, a different sulfhydryl group forms a covalent acyl
thioester with the substrate in the oxidation reaction

(Racker, 1965).

91

It appears that the loss of critical hydrophobic
bonds at 0° is the factor which, coupled with the ATP-
induced electrostatic repulsion, causes dissociation of
the enzyme into subunits. Kauzmann (1959) has discussed
the reasons for expecting hydrophobic bonds to be more
stable at room temperature than at 0°. A less likely
explanation for the temperature-dependence is that one
of the ionizable groups in the enzyme undergoes a tem—
perature-dependent change in pK value.

Thus, dissociation appears to proceed as follows:
(1) conformational changes occur in the individual subunit
polypeptide chains as a result of electrostatic repulsions
when ATP is bound. (2) This change in the folding of the
individual polypeptide chains alters the contact sites
between subunits in such a manner that hydrophobic inter-
actions assume a more critical responsibility for main-
taining the native tetrameric structure. (3) Lowering
the temperature to 0° is then sufficient to weaken hydro-
phobic interactions at the contact site between subunits
to such an extent that dissociation occurs.

It is also interesting that the temperature-
dependence of reasSociation (Figure 17) does not exactly
parallel that of dissociation (Figure 11). It thus appears
that the presence of 10% sucrose in the reassociation
Inixture alters the temperature—dependence of the equilibrium.

'Fhis may result from an effect of the sucrose on the water

92

structure, since alterations in water structure would be
expected to affect hydrophobic interactions.

Throughout this discussion we have implicitly
assumed the interaction of ATP and glyceraldehyde-3-
phosphate dehydrogenase to involve binding, followed by
dissociation. Tanford (1964) has pointed out, however,
that in some cases the process may be reversed; i.e.,
dissociation may occur first, followed by binding of
the denaturing agent. In such a case, the denaturing
agent binds preferentially to the dissociated product,
thereby shifting the equilibrium of the dissociation
process by removing the product.

It seems unlikely that this occurs in this system,
since ATP seems to bind to the native enzyme as tightly
as it does to the subunits. This is based on the mid-
points of graphs as a function of ATP concentration for
(l) chymotryptic inactivation at pH 7.0 and 25° (Yang and
Deal, 1969b), (2) dissociation at pH 8.5 and 25° and
(3) reassembly at pH 7.0 and 17°.

It is difficult at present to calculate an equi-
librium constant for this process, since dimers are almost
certainly involved in the conversion of the inactive monomers
to active tetramers and it is not known whether the dimers
are catalytically active or not. However, if the dimer is
assumed to be inactive, the equilibrium constants at 5 1/2°

(and the conditions of Figure 17) for a dimer-tetramer and

93

a monomer-tetramer equilibrium are 0.1 and 0.001 mg per ml
respectively.

It is interesting that ATP serves as a catalyst in
the reassembly reaction at 17° and perhaps in the dissocia-
tion reaction at 0°; this is in addition'to whatever
effect it may have on the equilibrium of the dissociation.
Further work is in progress on the mechanism by which ATP
stimulates the reactivation and reassociation processes.

As discussed in a previous paper (Stancel and
Deal, 1968), the 2.8-3.0 S dissociation product appears
to be a fairly compact monomer, since that sedimentation
coefficient is appropriate for the molecular weight of
38,000 obtained from subunit analysis using urea (Deal,
1963; Deal and Holleman, 1964; Deal, W. C., Jr., in
preparation). The sedimentation coefficient seems too
low for a dimer, since it would have to be extremely
unfolded to exhibit such a small sedimentation coefficient.
It is interesting that a dimer has not been detected in
the dissociation. This suggests that the monomer-monomer
interaction is somewhat dependent on the dimer-dimer
interaction for stability and is immediately broken when
the dimer-dimer interaction is broken.

The effect of ATP on yeast glyceraldehyde-3-phosphate
dehydrogenase appears to be similar.to the effect of oxygen
on hemoglobin.. The subunits of tetrameric hemoglobin

undergo a marked re-orientation with respect to each other

94

when oxygen is bound (Muirhead and Perutz, 1963), apparently
due to conformational changes in individual subunits upon
oxygenation (Edsall, 1968). The binding of oxygen also
increases the susceptibility of hemoglobin to proteolytic
degradation (Zito et_al., 1963).

BIOLOGICAL SIGNIFICANCE OF THE DISSOCIATION PROCESS FOR

METABOLIC CONTROL. Since the in_vivo ATP concentration

 

of yeast is 1—2 mM_(Betz and Moore, 1967), it seems likely
that the interaction of ATP and glyceraldehyde-3-phosphate
dehydrogenase may be physiologically significant. However,
the dissociation of yeast glyceraldehyde-3—phosphate
dehydrogenase into subunits by ATP occurs at low temper-
atures and occurs most readily at higher pH values. Also,
rapid but subtle conformational changes occur when ATP is
bound, but dissociation occurs much more slowly. The

rapid subtle conformational changes increase the rate of
inactivation of yeast glyceraldehyde-3—phosphate dehydro-
genase by chymotrypsin at 23° and neutral pH (Yang and
Deal, 1969b). This effect seems to have the greatest
potential as a control mechanism, and offers an additional
mechanism other than induction-repression or activation-
inhibition mechanisms to control metabolic processes. It
is also very significant that 3',5'-AMP and 5'—AMP alone

do not dissociate yeast glyceraldehyde-3-phosphate dehydro-
genase and, in fact, protect it from dissociation by ATP.

The possible physiological significance of these and other

95

effects is discussed in more detail elsewhere (Yang and
Deal, 1969b).

If we assume that the 2.8 S component is a folded
monomer, then the results from the reassociation experiments
suggest that folded monomers can associate to a tetrameric
enzyme. This would be consistent with, but would not prove,
that i§_yiyg synthesis of this polymeric enzyme may involve
two processes: the assembly and folding of the individual
polypeptide chains followed by a separate step, the
association of folded monomers to a Catalytically active
polymeric enzyme. A more thorough discussion of this
topic has recently been presented elsewhere (Deal, 1969).

There are very pronounced differences between the
effect of ATP on yeast glyceraldehyde-3—phosphate dehydro-
genase. The rabbit muscle enzyme is dissociated to a much
greater extent at high protein concentrations (10 mg per
ml) and is much more specific for ATP (Constantinides and

Deal, 1969).

MATERIALS AND METHODS

 

ENZYME PREPARATION AND ASSAY. The materials and methods

 

for preparation and assay of yeast glyceraldehyde-3-
phosphate dehydrogenase were those described previously
(Deal, 1969), with the following minor modifications. A
stock assay solution was prepared by mixing the following
(final assay concentrations in parentheses): (1) 12 m1

of 0.2 M Tris, pH 8.0 (100 mM); (2) 1.2 ml of 0.2 M_sodium

 

96

arsenate (10 mM); (3) 2.4 m1 of 0.01 M_NAD (1 mM); (4) 3.6
ml of 0.06 M cysteine hydrochloride (9 mM); (5) 1.2 ml of
3.0 M KCl (150 mM) and (6) 1.2 m1 of H20. The final pH
was 7.8. For assay, 0.36 was added to a cuvette, then
0.01 ml of enzyme solution (0.2 mg per m1) and the contents
were mixed. The enzyme reaction was then initiated by
addition of 0.03 ml of a stock glyceraldehyde-3-phosphate
solution (0.02 M), to give a final volume of 0.4 ml.
REAGENTS. Reagent grade B-mercaptoethanol was obtained
from Eastman. ATP (sodium salt) was obtained from Sigma
and stored dessicated under a vacuum at -20°. Stock
solutions were prepared at 0.04-0.05 M concentrations,
adjusted to pH 7.5 to 8.5, and stored frozen until used.
Normally this storage period was not longer than 30-60
days. ADP, AMP and 3',5'-AMP were all obtained from
Sigma (sodium salts) and were prepared as was the ATP.
All other reagents used were the highest quality reagent
grades commercially available.

SUCROSE DENSITY CENTRIFUGATION. Sucrose density gradient
centrifugation was performed using the method of Martin
and Ames (1961) with slight modifications which are
described elsewhere (Constantinides and Deal, 1969).
Samples were centrifuged at 40,000 rpm in a Beckman Model
iL preparative ultracentrifuge with either a SW 39 or a

SW 50 rotor for 22 hr at 0 to 2° C.

 

97

DISSOCIATION AND REASSEMBLY. For the dissociation experi-

 

ments, the enzyme (0.2 mg per ml) was incubated for 5 hr
at 0° in 4 mM ATP, 0.1 M Tris buffer, pH 8.5, and 0.1 M
B-mercaptoethanol. The samples were completely inactivated
by this treatment.

Reassembly was accomplished by diluting 0.5 ml of
the inactive monomers with 0.5 ml of reversal mixture at
0° and warming to 23° or 17° for 2 hr. The reversal
mixture contained 0.4 M Tris buffer, pH 6.2, 20% sucrose
and 0.1 M B—mercaptoethanol. After dilution, the resulting
mixture contained 0.2 mg per m1 g1ycera1dehyde-3-phosphate
dehydrogenase, 0.1 M_B-mercaptoethanol, 5 mM ATP, 0.21 M

Tris, 10% sucrose, and had a pH of 7.0.

CHAPTER FOUR

KINETICS OF THE DISSOCIATION AND REASSOCIATION OF
GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE

IN THE PRESENCE OF ATP

98

ABSTRACT

 

A detailed kinetic analysis of the dissociation of
tetrameric yeast GAPD to monomers at 0° in ATP has been
performed by following the loss of catalytic activity. At
pH 8.0 in 0.02 M Tris and 0.075 M_B-mercaptoethanol the
process is biphasic. There is a fag; initial step followed
by a slgw_second step.

The fast step can be described by a rapid reversible
equilibrium between tetramers and dimers. The calculated

8 £2.) for this initial

equilibrium constant (Keq = 10.2 x 10-
dissociation yields a value of 8.9 kcal/mole for A G° at 0°.
The slow step exhibits psuedo-first order kinetics
and a half—power dependence on protein concentration,
suggesting that this step reflects the dissociation of
dimers to monomers. The dependence of the dimer to monomer
dissociation on ATP concentration indicates that ATP binds
to the dimer in successive, reversible steps with no inter-
action between ATP binding sites. The average constant
for binding of ATP to the dimer is 0.75 mM. The rate of
dissociation of dimers to monomers is decreased as ionic
strength is increased, suggesting that hydrophobic residues

are exposed to the solvent medium as the dimer dissociates

to monomers.

99

 

100

A kinetic analysis of the reversal of dissociation
at 23° in the presence of 10% sucrose reveals that the
reversal process is second order with respect to protein
at low protein concentrations and first order at high
concentrations. These results suggest that the final step
in the reassembly of monomers to tetramers is a first order
process involving a conformational change. The half-life
for the reassembly process is 8.5 min at 0.20 mg/ml and

16.0 min at 0.010 mg/ml.

101

INTRODUCTION

 

In previous papers, we have reported the reversible
dissociation of yeast and rabbit muscle g1ycera1dehydro-

genases into unfolded subunits in urea (Deal and Holleman,

 

1964; Deal, 1969) and into folded subunits in ATP at 0°

 

(Stancel and Deal, 1968, 1969; Constantinides and Deal,
1969). These provide ideal models for intensive investi-
gations of the following properties: (1) subunit struc-
ture, (2) nature, types, and strength of subunit inter-
actions, (3) folding and association reactions leading to
formation of the polymeric enzymes, (4) the kinetic and
thermodynamic properties of these reactions i2_yiyg from
the initial stages of protein synthesis until the final
product, and (5) the significance of these properties
with respect to control of activity and synthesis of the
enzyme.

SUBUNIT STRUCTURE. The subunit structures of the two

 

enzymes are similar in that both yeast and rabbit muscle
glyceraldehydrogenases are tetramers of molecular weight
145,000 and are dissociated into unfolded subunits of

molecular weight 36,000 in urea (Dealand Holleman, 1964)
and guanidine hydrochloride solutions (Deal and Holleman,
1964; Harrington and Karr, 1965). The subunit structures
are different in that the yeast enzyme subunits are dif-
ferent from the muscle enzyme subunits; however, both

enzymes appear to be made up of identical subunits, one

102

type of subunit being found in yeast and another type
found in rabbit muscle.

SUBUNIT INTERACTIONS. The nature, types, and strength of

 

subunit interactions are also different for the two enzymes,
and one purpose of this paper is to provide further infor-
mation on this subject by defining the detailed reaction
mechanisms and thermodynamic and kinetic properties of

the reversible dissociation of yeast glyceraldehyde-3-
phOSphate dehydrogenase.

Both tetrameric yeast and rabbit muscle glyceralde-
hyde phosphate dehydrogenase are reversibly dissociated
into monomers by ATP at 0° under proper conditions, but
stable dimers are observed in sedimentation velocity
experiments only with the rabbit muscle enzyme under
proper conditions. With yeast glyceraldehyde phosphate
dehydrogenase no dimers are observed over the entire range
of protein concentration; only the native 7.4 S enzyme
and 2.8-3.0 S monomers are observed (Stancel and Deal,
1968, 1969). At long times (several days), the folded
monomer undergoes an irreversible change to a more slowly
sedimenting 0.9 S form (Stancel and Deal, 1968, 1969).

In contrast, rabbit muscle glyceraldehyde phosphate
dehydrogenase is dissociated into 4.6 S dimers at pro-
tein concentrations as high as 10 mg/ml and the enzyme
yields 4.6 S dimers in increasing relative amounts as
the protein concentration is decreased to 1 mg/ml, where

dimers alone exist. Below 1 mg/ml, the enzyme begins to

103

show dissociation into 2.8-3.2 S monomers and below 0.1
mg/ml, monomers alone are observed (Constantinides and
Deal, 1969). Since the dissociation of yeast glyceralde-
hyde phosphate dehydrogenase by ATP at 0° must also produce
a dimer (or trimer) as an intermediate step in the dis-
sociation to monomers, one of the specific questions to

be answered in this study was why no dimers are observed

 

in the ultracentrifugal analysis of the dissociation
products.

IN VIVO AND IN VITRO FOLDING AND ASSOCIATION. Another

 

purpose of this paper was to obtain information related
to the following question: "During, and after, its
polypeptide chains are synthesized on the polysomes,
what are the series of folding and association steps
which lead to the native tetrameric glyceraldehyde phos-
phate dehydrogenase?" The two dissociation systems for
the two glyceraldehyde-3-phosphate dehydrogenases have
certain features which make them rather uniquely suited
to help answer the question. With these two dissociation
systems, one of which yields dissociation with unfolding
(the urea system) while the other yields dissociation
essentially without unfolding (the ATP dissociation
system), we have the potential of studying not only the
overall process but also of studying independently the

two phases, folding and association, of the process ig

 

vitro.

 

104

In the previous study of the overall process of
reassembly (folding 33g association) from unfolded sub-
units (the urea system), it was discovered that NAD was
absolutely required for the process (Deal, 1969). A
translational control model has been proposed (Deal,
1969) in which NAD may control the rate of glyceraldehyde
phosphate dehydrogenase synthesis by dictating whether,
or how fast, the nascent polypeptide chain folds cor-
.rectly, which is postulated to be necessary to allow
synthesis to proceed to completion.

In the systematic study of conditions required
for the overall association process from folded subunits
(the ATP-0° system), it was discovered that ATP and
sucrose were necessary for the association processes.

It was of interest to know which steps of the reassembly
process were involved in these effects. The approach
taken was to figs; define the reaction steps by which the
native enzyme is produced from the "folded" subunits (the
ATP—0° system), as well as the kinetic and thermodynamic
parameters of these steps. This paper presents the

results of these studies.

RESULTS
Precise kinetic measurements of the ATP-induced
inactivation of yeast GAPD at low temperatures were
initially difficult to perform because after dilution from

stock solutions of enzyme stock ammonium sulfate suspensions

105

(Yang and Deal, 1969a) enzyme controls first showed an
increase in activity with time followed by a decrease in
activity. It was found that this problem could be
avoided if the enzyme was first diluted into buffered
solutions with small amounts of B-mercaptoethanol and
allowed to incubate for 4-12 hrs at either 0° or 23°
before the addition of ATP. Using this preincubation
technique the kinetics of the ATP-induced dissociation

at 0° were examined, and a model of the dissociation
mechanism was proposed.

DESCRIPTION OF THE MODEL AND SUMMARY OF EVIDENCE. Because
the kinetics of yeast GAPD inactivation in ATP are quite
complex we will attempt to make the concepts easier to
grasp by first presenting the resultant model, together
with a summary of the kinetic evidence on which it is
based. Then the kinetic data will be examined in detail.

The model for the sequence of reactions is given below:

RAPID EQUILIBRIUM

ATP k42 kZl
TETRAMER » [TETRAMER —> DIMERS] '4 MONOMERS (l)
(—
k24

FAST PROCESS SLOW PROCESS

The basic idea is that the overall process involves two
processes, one fast and one slow. The first step is

postulated to be the fast step and to involve a change

106

from stable active tetramers to an equilibrium mixture of
active tetramers and inactive dimers. The second step is
postulated to be slow (rate-limiting) and to involve the
dissociation of the inactive dimers into inactive monomers,
practically at a constant rate because the dimer concentra-
tion is practically constant. This is because dimer
removed is rapidly replaced by dimer produced by the very
fast tetramer-dimer dissociation, the relatively constant
amount of dimer being determined by the equilibrium con-
stant for the tetramer-to—dimer dissociation. The estab-
lishment of an equilibrium is presumed to result from the
fact that this dissociation into dimers is Very fast but
dissociation of dimers is so slow that it is rate-limiting.
This model is supported by the following observations,
which are amplified in succeeding sections.

(1) The process(es) (see Figure 20) cannot be
described by a single rate law.

(2) The overall process (see Figure 20) can be
separated into two phases, a fag: initial
process and a slgy process.

(3) Both processes are inversely dependent on
protein concentration, suggesting both pro-
cesses involve two dissociation steps.

(4) The slgw_process exhibits pseudo—first order
kinetics and a half—power dependence on tetramer
concentration, suggesting a dissociation of

dimers.

107

.poflhod oEHu meg“ mcfluop mmoa mnfl>fluoo on pozonm Auommsn mo

HE OH.O moan osemco popoQTUCflnon mo HE m.ov oadsmm Houwcoo ode .oEHp mo
GOHBUCSm o mo cozoaaow moz mmoH >OH>Hpoo odp poo .coflpsaom ofimmco pouonsocfl
loud or... ....o as To 8. some... 9.3 8.77 mm on @8332 92 as 8 do as 3.0
.cofluonoocfl moped .Hocmnpoowdoouoalm.w meo.o poo .o.m mm .mHHe m.H.o :e

00 no opflcuo>o poponcocfl mm3 AHE\mE vo.ov oaemco ode .med CH 00 pm >DH>Hvoo

omocomonoeaoc oposmmozdlmlopmgopHMHooeam mo mmoH poopcomoploEee .IM onomfle

 

4‘

108

5.52.2. ”2:.

 

 

00— ON— on O? o
...... .4 +4 .... ...... .4 I- 40
0’0, .
.l 1
.I7./ ON
o/o l
/
o/ Low 7..
co .2 fl... ... 9.0.x . W.
.23: 68.6 o .00 u
x. M
‘/ .. u
loo
,..

 

 

109

(5) The EEEE process can be kinetically described
by a rapid, reversible dissociation of
tetramers to dimers.

(6) The equilibrium constant calculated for the
proposed tetramer—dimer equilibrium fits the
data over a wide range of protein concentration.

This model assumes there is no appreciable unfolding and
thus, that we are dealing only with dissociation-association
reactions. The assumption is supported by the fact that

the model fits the observed data.

ANALYSIS OF THE SLOW PROCESS (DISSOCIATION OF DIMERS). Since
there clearly are two processes involved, we must attempt

to separate the two in order to analyze the data. Thus,

the logical starting point is with the slow process, since
for all practical purposes, the fast reaction has already
gone to completion, leaving only the slow reaction to
produce the results. Figure 21 shows a semi-log plot of

the loss of GAPD activity with time, for three different
initial protein concentrations. It is clearly seen that
the overall process is biphasic, that the slow phase does
not extrapolate back to 100% activity (also indicating

two processes), that both phases are inversely dependent

on protein concentration (suggesting two dissociation
steps), and that the slow phase is pseudo first-order.

The slow process, which is pseudo first-order, was

treated by the half-life method to obtain the order of

llO

.om mussflm

wo oComoH pCo onw oom .med CH ommComOHpedop opoCmmOlemloomsopHoHooeHm

wo CoHHoHOOmme mo opmu oCu Co CoHpoHpCoOCoo CHoHOHQ mo Hoomwm

minim

}

.HN musmflm

111

3.7.52.2. mi:
2.: 8. o: as 8. 8 -8 i 9.

on

 

.1 .JuW . .
h/ six at «8.6V. .
. /// . /
< /
‘/

d2\ 02 02.0 IV II

ooooooooo‘ll' d2\ 0: oncoo

 

b b n n P b -

 

0:

 

 

Om<0 “.0 20=<.00mm_0 “.0 «abuzz;

v.—

°9 N. *0. '0
(Au/\uav‘aa) 901

0.
N

112

this process with respect to protein concentration (see
appendix for details). As shown in Figure 22 a graph of
log (Tl/2) vs log (initial protein concentration) yields
a line with a slow process slope of 0.46, which yields an
order of 0.54 with respect to protein (see appendix).

Assuming that only the tetramer is active, an
interpretation of this half—power dependence is that the
slow process represents dimer dissociation, preceded by
the establishment of a tetramer—dimer equilibrium mixture.
The following derivation shows that this interpretation
is consistent with the results. From (1), the rate of

monomer formation, dM/dt, is given by eqn (2).

+ dM/dt = k21 [D] (2)
and since the dissociation equilibrium constant, Keq' is
given by eqn (3):

_ _ 2
Keg - k42/k24 — [D] /[T] (3)

solving for D in eqn (3) and substituting into eqn (2) yields

+ dM/dt = k21 Keql/2 [Tll/2 = k'ET11/2

‘Where M, D, and T are the monomer, dimer, and tetramer
concentrations, respectively; k42 and k24 are the rate
constants for dissociation of tetramers to dimers and

association of dimers to tetramers, respectively; k21 is

113

.pxop oom .HE\mE CH

COHHOHuCoOCOO CHoHOHm oCu OCo moHCCHE CH commonxo ono mo>HH|wHoC one
.Aflomon po>Homnogl H OH Hoovo mH Hopuo oCHV vm.o mo HoOHO Co OCHOHon
monp .wv.o mH oCHH oCu mo odOHm oCe .UOCpoE owHHIMHMC on en ACBOOxooHQ

HoEHO poEsmoHQV mwoooud BOHm on mo Hopho on mo COHHoCHEHopoQ .IM oHCmHm

114

_Hzapowau co. ... «

 

 

d

1 d I

3.0 ...-... was“

fi

0.—

7” l: 901

 

 

115

the rate constant for dissociation of dimers to monomers,

K 1/2. We consider this explanation to be

, .
and k 18 k2l eq

the correct one because we are unable to find another feasible
possible explanation for this dependence.

SEPARATION OF THE FAST PROCESS FROM THE SLOW PROCESS. In

 

order to elucidate the nature of the fast process, a method
was devised to separate the rates of the slow process and
fast process. As seen in Figure 21, the data points at
short times lay above the extrapolated rate of the slow
process, so the data for the slow process do not extra-
polate back to 100% activity but to a point A0. Now, A0

is the activity which would be observed at equilibrium if

 

only the first process, tetramer to dimer, were occurring
(i.e., at ts).

The points which lie above the slow process were
corrected for the rate of the slow process as illustrated
in Figure 23. After the extrapolation of the slow process
to a zero time value, A0, the difference in activity
between A0 (point 2) and a given point (point 1) on the
extrapolated line was Edged to the value of the experi-
mentally measured activity (point 3) at the same time
(Figure 23). With respect to the fast reaction, this
amount added was equal to, and therefore corrected for,
the excess loss of activity due to excess dimer formation
over that dictated by the equilibrium constant for the

tetramer to dimer equilibrium. A11 points between time

 

116

.mHHmpop HoCuHsm Low uxou oom .ACBOUxooHQ HoEHO UoEomoCmv mmoooum BOHm

on Eonm AESHHQHHHCUo HoEHOIHoEOCHoH OoECmoumv

mmooond Home on opouodom

Op poms ouop OHHoCHx wo 0CoEHooHp oCu mo onono OHuonCmm .IM oHCmHm

117

3332:): 22:

 

 

 

 

 

 

 

7 . . . q q a.» e
o/
1 o
/0
/O
.l /.
/o
t
.. /v0
0 :—
.. 33400.32:va od o: o d
N
. . l
On
.0
9:8 1020.27.08.20 10 3:32;.

v.—

m.—

0
F

«a N.
(AllAl13V96) 001

0:
p

0.
N

118

zero and the time at which the activity points coincide with
the slow process line (point 4, time ts) were corrected in
this way. Note that the first experimental point (4) on

the slow process line would have a corrected value of A0.

ANALYSIS OF THE FAST PROCESS (TETRAMER-DIMER EQUILIBRIUM).

 

The corrected points could not be fit to a psuedo first-
order rate law,1 but did fit an integrated rate equation
expected for a rapid equilibrium between active tetramers
and inactive dimers. This is shown in Figure 24, which is
a plot of log x vs time for the 0.036 mg/ml sample depicted

in Figure 2, where x represents the quantity

(—A°_AC)-(AO-AC) (T)
(5)

 

(’Ao-AC)+(Ao-AC)(T)

In this term A0 is the extrapolated activity at time zero
and AC is the corrected activity at a given time (see
Figure 23). For a given protein concentration T is a
constant which depends on the initial and equilibrium
activity values (see appendix for a complete derivation).
Similar fits were obtained at other protein and ATP con-
centrations. This behavior is consistent with a tetramer-
dimer equilibrium.

It is also possible to test for a tetramer-dimer

equilibrium by calculating equilibrium constants at

 

1The data also could not be fitted to a lst order
plot when differences between experimental points and the
extrapolated lines were plotted directly.

119

.mHflMme nonpo How paw m mpflpcmsq mgp mo coflpflcflmmo m MOM pxop
mom .mHoEflU 0>Hpomcfl ocm mumamupmb m>fluom cmprmQ Esflnnflaflsqm pflmmu 0
How cofipmsqm mwmn Umpmnmmnsfl mnn 0p AEDHHQHHHSUm HmEflplumfimupmp woesmmumv

mmmooum ummm mcu How mpcflom mymp Umpomuuoo map m0 new Oflposflm .wM musmflm

 

120

On

3222:): us:
an

 

 

..5 fl... ... .oo
:2on 086

 

 

X 901

121

different protein concentrations. If an equilibrium is in
fact being established, then the value A0 (obtained by
extrapolation of the slow rate to time zero) should cor-
reSpond to the activity of a tetramer-dimer mixture dictated
by the total protein concentration. Equilibrium constants
for each initial protein concentration can then be calcu-

lated as follows:

[T]. g

( [T] fl_) (% Act.,

initial’ extrapolated)
(6)

[D]. 92 (m.

initial’

9’1 " [T]: I!) (2)
K = [912 / [T3
42

Constants calculated from experiments with different
initial protein concentrations are listed in Table IV.
Considering the extrapolation procedure used and the
square term in the calculation, the K's seem to be constant
over a wide range of protein concentration. Since the
overall dissociation process is known to be a conversion
of tetramers t0 monomers (Stancel and Deal, 1968, 1969),
a mechanism consistent with all the data is the establish-
ment of a rapid equilibrium between tetramers (active) and
dimers (inactive) followed by the breakdown of dimers to
monomers.

EFFECT OF ATP ON DISSOCIATION. Although the previous

 

experiments involve dissociation in the presence of ATP,

 

we want information about the dissociation without ATP.

122

 

II .Nvm

 

 

z muoH x m.oH u io>ac m
h.m mO0.0
m.HH wmo.o
m.vH NB0.0
m.m OmH.o
l .vm
moa x z A.0Hmov M AHE\mEV .ocou camponm

 

ZDHmmHAHDOM mmEHQImmS<MBMB 4 mom m824BmZOU EDHmmHQHDOM DMBfiQDUAdU

.>H mqmdfi

123

In order to obtain this, we require knowledge of the
processes involving binding of ATP to the enzyme. Speci-
fically we must know the values of the ATP-to—enzyme
binding constants, the number of ATP molecules bound per
molecule of enzyme and whether there is interaction
between these ATP binding sites.

ATP BINDING CONSTANTS. To relate the equilibrium constants

 

calculated in the preceding section (obtained in the
presence of ATP) to the enzyme in the absence of ATP, the
ATP concentration was varied at fixed protein concentra-
tions. When eXperiments were performed at 0.036 mg/ml
yeast GAPD at various concentrations of ATP, biphasic
plots were obtained. When the data was treated as in
Figure 21, the extrapolated intercepts of the slow process
at time zero decreased as the ATP concentration was raised,
but the changes were not sufficiently large to be treated
in a quantitative manner. The rates of the slow process
also increased as the ATP concentration was raised. As
shown in Figure 25 these rate constants (slow process)
'were then plotted as l/k vs l/ATP to obtain the binding
constant for ATP and kmax'

From the intercept on the horizontal axis (Figure
25) an average binding constant for ATP of 0.80 mg_was
obtained, as well as a value of kmax' The fact that a

:plot of the form l/k vs l/ATPn is linear for n=1 indicates

that.the process appears first order with respect to ATP.

124

.umEHo wry 0p mam m0 mcflccflfl wry whomammu pOHm mflcu Eouw Umcflfiumuoc

accumcoo 0:0 .cBooxwoun ngflp on 0p cmESmmHm mH mmmooum 30am 0:0 mocflm

.mmmooum 30am mcu mom pampmcoo moapcfln 054 030 m0 soapmcflanmpmo

.IM muomflm

.73.: m.&\_
v

 

125

m m N p T
u q q q - —
X<S \/
x /7
l . N .I. 1
SE 0.0 n J.
. .\ .
.. :23: 2.6... o\ v -
0\ l
I 0 / ..
X
r m -
0
.. \ 2 -
h p h p P _

 

 

 

 

126

Since these are the psuedo-first order rate constants from
the presumed breakdown of dimers, this value of 0.80 mg
reflects the binding of ATP to the GAPD dimer.

Because a knowledge of the binding constants for
ATP was important, these experiments were repeated at a
second protein concentration of 0.0125 mg/ml. When treated
in a similar manner an average binding constant of 0.67 mg
was obtained, as well as a value for km .

51%
TESTS FOR INTERACTION BETWEEN ATP BINDING SITES. The rate

 

constants obtained at different ATP concentrations can
also be treated to test for any interaction between ATP
binding sites. If the binding of ATP is represented by

successive, reversible equilibria of the type

Enz(ATP)n_ + ATI>.<_.“> Enz(ATP)n

l

[Enz(ATP)n_1] [ATP]

 

[Enz(ATP)n]

then

0.
l-d

PKi

*where a is defined as the per cent saturation with ATP,
and n is the interaction constant for binding of successive
Jnoles of ligand (see Suelter, 1966). In this case n is
.analogous to the interaction constant obtained from the
:510pe of a Hill plot, i.e., a value of n=l indicates no

iJiteraction between ligand binding sites.

127

For the dissociation of GAPD in ATP, the preceding
relationship can be alternatively expressed in a form more
usable for out purposes as

d

- p( [ATP] ) = log ———

K
n(p avg 1-0

a values may be calculated by assuming that the rate of

inactivation is proportional to the degree of saturation

of enzyme with ATP. Therefore,

 

a (kATP , 1,)
ATP,i =
(kmax)
with k - the experimentally determined psuedo-first

ATP,i

order rate constant for the slow process at a given ATP
concentration and kmax determined from a plot of l/k vs
l/ATP (Figure 25).

To determine if there is any interaction between
ATP binding sites, one can then assume various values of
n and draw theoretical curves of 0 vs log ( [ATP] ) for
that value of n. Figure 26 is such a plot with the
theoretical line drawn for a value on n=1 and Kavg = 0.74
mg (the average of 0.80 and 0.67). The agreement between
the theoretical line for n=1 and the eXperimental points
suggests that there is no appreciable interaction between

ATP binding sites.

NATURE OF SUBUNIT CONTACT SITES: EFFECT OF IONIC STRENGTH

 

‘ON THE DISSOCIATION PROCESS. We have previously suggested

 

128

.Amopflm mcflocfla med cmozwmn SOHwomuopcfl 0s mmum0H©CH H mo 05Hm> mv H 0p

Hmsqm .s .wcmumcoo coflpUMHTNSH cm cam EE w>.0 mo wcmpmcoo COHNMHOOmmflp omnnm>m
so MOM czmnp TCHH Havaumuomgp map we mafia mgp mam .mCOHNMHNQTOCOU cflmponm
oopmoflpcfl mgu pm omcflmuho mucflom Hmycmfiflmmmxm map mum mwcflom 0:9 .mmmooum 30am

Exp CH ©m>Ho>CH mmuflm @cflpcfln me¢ cmmemQ soflpoMHTNCH How pmma .mm onomflm

 

129

3... 9350.3

 

 

ad ad m.— o.—
. 1 I W - . q q -
... .....- _
\“ L 0
\\\\
‘\
\\ C 1 N
.335. 3.0.0 -o as.
.
25: 030 -o .... . n.
n...
. t
o m,
\
\sssO . m. M
ssms .. 0. m
\\ l‘
\\
o \\\\ p u 8 . No
.... i... t .0 n J.
\\\\\\ 1 Q.
\\° .
. o

 

 

 

130

that the dissociation of yeast GAPD resulted from decreased
hydrophobic interactions between subunits at low tempera-
tures in ATP (Stancel and Deal, 1969). Physically this
implies that the contact sites between subunits are
hydrophobic in nature, since the contact sites are pre-
sumably the major structures broken by ATP induced dis- .3
sociation without unfolding r‘
The dissociation was therefore examined as a
function of Na SO4 concentration (Figure 27), since in—

2
creasing ionic strength should inhibit a process which

 

exposes hydrophobic residues to the solvent medium. It
is clearly seen that there is a marked inhibition of the
§l2w_process. Tests with 0.33 g NaCl also gave a notice-
able inhibition (Figure 27), so the inhibition was not
specific for sulfate.

The effect of Na2504 on the Slow process (Figure 27)

could be described by an expression of the type

log (ko/k) = 8 CS

where k0 is the rate constant measured with no added salt,
k is the rate constant measured at a given salt concentra-
tion, CS (g), and S is a constant. This observation sup-
ports the hypothesis that the contact sites between subunits
in the dimer involves considerable hydrophobic interactions

(Tanford, 1964; von Hippel, 1969).

 

131

.maflmpmp HozuH5m How axon mom .sumcmpum UAQOH mcflmmmuocfi

an Aczooxmmun HNEHG UmEDmmumv mmmoonm 30am 0:0 mo cofluflaflch

 

.mm mhsmem

132

It“
i
1

2.3.52.5: m2:

 

 

 

 

02 on. on on o .
q q q 11G] 1 q u e V—
. /.
. m.—
. o.—
o - 2
S.
1 w.—
«a.
2...
.:< $2 m .oo fl $03.2. 1 0..
..2\ 02 30.0
1 ON

 

 

0&40 m0 ZO:<_UOmm_o “.0 mu_»UmZ_v.

(All/\IIDV 96) 901

 

 

133

To insure that the effect of the salt on the dis-
sociation was not to alter the ATP binding constant or the
mechanism of the dissociation, the binding constant for
ATP and the order of the slow step were determined in the
presence of 0.33 g Na2504. Using procedures similar to
those in Figure 25 a binding constant of 0.5 mfl_was ‘Efih
obtained in the presence of salt, and using the half-life '

method (Figure 22) an order of 0.4 was determined for the

slow process in the presence of salt. Thus it seemed é

 

that the salt had no major effects on the ATP binding or
the mechanism of dissociation.

The intercepts obtained by extrapolation of the
rate of the slow process to time zero in the presence of

0.33 g'Na SO4 were also used to calculate equilibrium

2
constants for a tetramer-dimer equilibrium. The average
8

of 5 values in 0.33 b_d_|NaZSO4 was 4.8 x 10- g, The con-
stancy of the values obtained at the different protein
concentrations is illustrated in Figure 28, which is a
plot of % tetramers (active) as a function of protein
concentration. The line is a theoretical line drawn with
the experimentally determined constant of 4.8 X 10-8 g.
The points are the values of % tetramers calculated from
the extrapolation of the rate of the slow process to time
zero. The agreement between the points and the line

suggests that the fast process may also be described as

a tetramer-dimer equilibrium in the presence of salt.

 

134

.mHHMDmp “TEDHSM mom pxmu mom

.m mica x m.w How pcmumcoo Esflmnflaflsqm mmmnm>m so How nacho mafia HmoflDmHomzp
may we mafia mnu pom coflpmuwsoocoo campoum £000 pm mpcmumsoo Edennflaflsqm
UTQHEHmpmo mfiampcmEHHmmxw oz» Eonm oopmfiooamo who museom 0:8 .Amflmmn HNHOE
o co oommmumxmv coaumnpcoocoo cflououm HmuOD m0 coauocsm m mm mHmEHU pcm

muoamupmp mo ououxflfi Eganhflaflsqm so m0 mHmEMMpmp usmo Hog 0:9 .MM onsmflm

 

135

 

 

 

H252...”— 001 + o

 

 

 

 

9m mfi 9N m.— 0.. o

. u q . q _ . . . . . . _ q ,
u I now
. .:< 222 .oo

c u I \ .
. 0“ oz .2 and o . ov
r o -00
. 0 \
0
T \ 100
r. I oo—
Om<0 bm<m> “—0 202.5835
p - p . b u - p p . b b . _

 

saawvzuai %

136

REVERSAL OF DISSOCIATION
Previous work had indicated that the dissociation
«Add be reversed by warming the monomers to temperatures

above 15° in the presence of 10% sucrose and small amounts

cf B-mercaptoethanol (Stancel and Deal, 1969). It was

therefore of great interest to study the kinetics of the
reversal process.
FIRST ORDER DEPENDENCE OF THE REASSEMBLY PROCESS AT HIGHER

PROTEIN CONCENTRATION. When a completely dissociated

sample of GAPD was subjected to the reversal conditions
and warmed to 23°, the recovery of activity gave a very
good fit to an empirical first-order rate law (Figure 29).

This suggested that a first order process was the rate

limiting step in the reassociation. To further test this

possibility the points were replotted as the final activity
recoverable (Af) minus the activity at a given time (At)

vs time. Tangents were drawn to the curve at various

points and the logs of the slopes of these tangents were
plotted.vs the logs of (Af-At) at the point the tangent

was drawn. These points fell on a straight line with a

sltha of 0.96, thus indicating that the rate-limiting step
‘was le fact first-order with respect to protein concentra-
ti£H1 under the conditions of the experiment.

Another test of a first order process is that the

half—times should be independent of initial protein con-

centration. The dependence of the half-times for reversal

 

137

 

.mHHMDTU HTSDHSm How mcozwmz pcm uxmp 00m .HOQMQDOODQMOHOEIQ

m. 3.0.0 can 9E as m .8868 AS .og me .32.. m Rd E... 68308.6...th

mma AHE\@E H.0v mHmEmm QHTDOHQ 0:9 .moonpoz ca ponflnommo mm omm 0p mDHCDQSm

Dmdw ummmm mQHEHMB com: ©m>ummno mum>oomm >ue>flpom mo DOHQ mOHIHEOm

 

.NM musmwm

138

 

 

 

 

 

‘ .
o
\
\0
I .\.
.\
.\
r \
0000
..
...
_rL. ... ... h. .0 m ... ...
m I. .L .L .l .I. O 0
_H.<\.<L<M_ 00.

80 100 I20

60
TIME (MINUTES)

40

20

 

 

139

on initial protein concentrations was thus determined

under two sets of conditions, the standard condition and

the standard condition plus 0.3 I\_’I_ KCl. This was done since

a previous study of the reversal of urea-dissociated sub-
units employed this concentration of KCl (Deal, 1969) .
Results illustrated in Table V clearly indicate that the

half-times in this concentration range are independent of

protein concentration .

SECOND ORDER DEPENDENCE OF THE REASSEMBLY PROCESS AT LOW

To further insure, however, that

PROTE IN CONCENTRAT ION .
the half-lives did not depend on concentration, these

experiments were repeated at protein concentrations an
order of magnitude lower than those used for the data of

Table V. The results are given in semi-log form in

Figure 30 (A-D), and indicate that the half-times at
these lower concentrations do vary with initial protein

concentration. These results suggested that at the high

protein concentrations the rate-limiting step of the
reversal was a first order process, but that as the pro-

tein concentration was lowered, the overall order of the

reaction was increasing.

CHANGE IN ORDER WITH CHANGE IN PROTEIN CONCENTRATION. TO

more accurately define this apparent change in order with

changes in protein concentrations, a direct differential

method was employed. Various concentrations of monomers

produced by dissociation at 0° were warmed at 23° in 10%

 

If Q. I“.

 

 

 

 

m.b m.m ON.O

 

 

 

O m.n h.m OH.o
4
l
w.h o.m mm0.0
A.cHEVN\H H AHIQHEVN\H e
Hum.m m.o uamw cocoa oz AHE\mEV .osou sflmwoum

 

ZOHBdMBZmUZOU ZHMBomm ZO ZOHBdHUOmmdmm mom mm>HAImQ¢m m0 muzmazmmma .> mqmde

141

Honuuom How uxow mom

aflououg QDHB QOHDMAOOmmmog m0 mo>flalwam£ any m0 COHDMHH0> .mm mHsm.m

.maflmumo

.mcoflumnucmocoo afloaonm 30H um .coflumuucmocoo

142

 

,0

3952.5: 35..

— 3 N. o— a o

10‘
O

 

 

-

- I q q

d

u
2.2 «.2 ... 3
:30: 208

mm

b h p b

b

./

 

 

9.40 m0 ZO.»<.UOmn<w¢ “.0

mUZU

~23.

v.—

In
.—°

“2 ". ‘0.
(1:1-‘11)) 001

o;
(N/

0.
N

143

$202.5: «2:

 

 

p

b h b P b p E

V.—

In
.—°

N. o.
’v» 001

“2
(V-

o;
(511/

 

o— 3 «q— 0— m m w m o
T l
N
2.2 «.2 ... 2»
1/0 ..2\02 ONOd 1
.. /0 ..
0
O/
t on“ ./ I.
0/

 

9.40 50 ZO;<.UOmm<m¢ “.0 “mu—bug‘s.

0.
N

 

144

 

3952.2. mi..—

 

 

 

-

P n b P h b

.....1lm._..m . w ..
2.: ...: aux:
...:th n86 .
C 1
C
on« /0/ 1
C

 

 

Om<0 “.0 ZO_»<_UOmm<w¢ “.0 mU.»UwZ.¥

c.—

On
.—°

N. 0.
-‘V)) 001

a
(V

0;
(*V/

Q
N

145

 

3952:): m2:

 

 

b

2 3 N- o— o 0 V

d d -

q u

u
2.: of u 2»
:30: 29°

b p n b p p

q

p

Q.—

:'~. «3. J:
V» 901

«1:
(VJ

2.,
(*V/

 

0.
N

 

am<0 “.0 ZOZSUOmm<u¢ uO mUZUmZC.

146

sucrose and the rate of increase of activity was measured.
The data was then plotted directly as activity (umoles
substrate/min.) vs time, and an extrapolation was made
to time zero to obtain the initial rate at each protein
concentration. The logs of the initial rates obtained
in this manner were then plotted vs the logs of the initial
protein concentrations (Figure 31). The slope of such a
plot yields the order of the reaction directly.

As seen in Figure 31 there is a break in the
plot, indicating that the apparent order is dependent
upon protein concentration. The slopes at the two ends
of the plot yielded orders of 2.1 and 1.3 for the low and
high protein concentrations, respectively. This eXperi-
ment was repeated four times with essentially identical
results. The values of the reaction order ranged from
1.9 to 2.1 at low concentrations, and from 1.2 to 1.5 at
the high concentrations. Thus, the only type of mechanism
consistent with this data and the previous half—times
lnust involve a second order rate-limiting step at low
jprotein concentrations and a first order rate-limiting
step at higher protein concentrations. The interpretation

of these observations is considered further in the dis-

cussion.

DISCUSSION
IMECHANISM 0F DISSOCIATION. The simplest mechanism of dis-

:sociation most consistent with our kinetic evidence is as

follows:

 

147

.mfiuo0nflo sfl0poum 0D po0mm0u
:uflz H0©Ho 0:: 0H0H> m0sHH 0:: :0 m0m0Hm 0:8 .A.sflE\0D0Hum:00 m0HOEnv
mpfl>fluom mm wauo0nfio o0DDOHQ AOM0N 08H: 0: coflpmaom0wa0 Eouw @0sflmp:ov
coflywuus0osoo :000 :0 0:0“ Hmfluflsfl 0:: ma :o\xo .soflamups0osoo cfl0poum

0:: CH 0mcm:o :Dfl3 mm0ooum soHDMHoommm0H 0:: mo H0©Ho ca 00:0:U «NM 0H0mflm

 

148

o.—

8 9.205500. . o
3 .

Q— u.— G—

 

 

 

d

0 an 3
N F
('P/xp) 901

N
N

Q
N

 

 

 

149

RAPID EQUILIBRIUM

+ ATP
TETRAMER —" [(Tetramer) .ATPn : 2 (Dimers) .
j—ATP 1

.ATPn ]-—-->4(Monomers).ATPn
2 3

This mechanism is supported by the bi—phasic nature of the
activity loss with time, by the inverse relationship
between protein concentration and rates of both steps of
the observed biphasic process, by the fit of the first
step (short times) to the expected integrated rate equation
for a reversible dissociation and the fit of the second
step (long times) to a pseudo-first order rate law. The
mechanism is also supported by the agreement of equilibrium
constants calculated from the data for the first step at
various protein concentrations under two different sets of
conditions (Table IV and Figure 28). The half-power
«dependence of the second phase of the process on protein
(myncentration is also consistent with the mechanism (see
p. 106) .

Although the mechanism presented above is the

simplest mechanism consistent with our kinetic data the

 

trwne mechanism may involve an additional tetramer species.
TTnis possibility is discussed more fully in a following

section (see Chapter Six for discussion).

150

ABSENCE OF A DIMER PEAK IN ULTRACENTRIFUGAL ANALYSIS OF

DISSOCIATION OF YEAST GAPD BY ATP. In the proposed

 

mechanism for the dissociation of yeast GAPD the amount

of dimer should be relatively small (eSpecially at high
protein concentrations) due to the magnitude of the
calculated equilibrium constant, K42, for the dissociation
of tetramer to dimer while the amount of monomer is still
able to reach sizeable proportions due to the essential
irreversibility of dimer breakdown at 0°. Consequently,
physical measurements, such as sedimentation velocity,
which require protein concentrations several orders of
magnitude above the calculated equilibrium constant, should
reveal tetramers and monomers. (At a total protein con-
centration [on a mg/m1 basis] equal to K42, one would
observe equal concentrations of dimer and tetramer, but

at concentrations much greater than K42 tetramer would be
the predominant species.) However, the concentration of
dimers (controlled by K42) should be too low to observe,
especially since the relative amount of dimer decreases
1nith increasing concentration. This seems to be the case,
since sedimentation velocity experiments of GAPD incubated
1nith ATP reveal monomer and tetramer peaks,3 but do not

eexhibit a detectable dimer peak (Stancel and Deal, 1968).

 

3What we have called the tetramer peak in the noted
:reference may actually be a peak representing a rapidly
«equilibrating mixture of tetramers and dimers. The sedi-
Inentation coefficient observed for such a mixture would be

'the weighted average, Sobs. (cTsT + chD)/(cT + CD).

151

Rabbit muscle GAPD also seems to be dissociated
by ATP at low temperatures via a similar mechanism. In
this case, however, a dimer peak is observed in ultra-
centrifugal analysis of the dissociation process
(Constantinides and Deal, 1969), indicating that the
equilibrium constant for the tetramer-dimer dissociation ["1
of the rabbit muscle enzyme must be considerably larger
than that for the yeast system. This appears to be the

case, since rough calculations, similar to those in this

 

work (Table IV), based on the extrapolated activities
(A0) obtained for a tetramer-dimer equilibrium, indicate
that the K42 for the tetramer—dimer equilibrium is of

the order of 10'6

g for the muscle enzyme. Calculation

of the equilibrium constant for the rabbit muscle tetramer-
dimer equilibrium based directly on the distribution
between the tetramer and dimer peaks observed in ultra-
centrifugal analysis (Constantinides and Deal, 1969)

yields an equilibrium constant of 3 x 10-5 g, The calcu-
lated values of K42 for the muscle enzyme, obtained by two
different methods are at least two orders of magnitude

8

larger than the value of 10.2 X 10- M calculated for the

yeast enzyme. This indicates that the dimer-dimer attrac-
tion in the yeast enzyme is very much stronger than that

in the muscle enzyme.

 

jHowever, based on the equilibrium constant calculated in
this work, CD would be very small, and the observed sedi-
Inentation coefficient would be expected to very closely
approach the value for the tetramer alone.

 

152

STRENGTHS OF THE DIMER-DIMER ATTRACTION: ENERGETICS OF
THE TETRAMER-DIMER EQUILIBRIUM. Since energy changes
between different states are independent of the pathway
between the states, the free energy change upon dissocia-
tion of tetramers to dimers in the presence of ATP may be
related to the dissociation in the absence of ATP.

Using the average equilibrium constant calculated
for the tetramer-dimer equilibrium (Table IV) a free

energy change of 8.9 kcal/mole was obtained from

 

AG° = - RT In K
eq

Likewise, a value of 9.3 kcal/mole is calculated from the
average constant obtained in the presence of 0.33 g

80.

Na2 4

These constants are obtained from experiments
performed at 5-10 mg ATP and varying protein concentra-
tions. At these ATP concentrations the tetramer may be
.regarded as completely saturated with ATP (Yang and Deal,
.l969a; 1969b), and the dimer is also seen to be essentially
suaturated (Figure 25). Consequently, the calculated energy

cfluanges correspond to the process

T(ATP) __.< 2 D(ATP)

If, however, the ATP binding constants for tetramers
axnd dimers are equal, the calculated energy changes also
agnply to the dissociation in the absence of ATP. This is

i l lustrated below .

153

 

AGE
T(ATP) <————- 2 D(ATP)
n -—> n
1 2
2
T 1 1 3 1 I
1 3
n1 (ATP) n2 (ATP)
+ + [I]
4
T 5 D
>
O
AG4

 

For such a cycle

(AG: - AGg) + (AGZ - AGE) = 0

Therefore, if the binding constants of ATP to the tetramer
and dimer are equal (represented by steps 1 and 3 above),
then AGZ = AGE.

The association constant for the binding of ATP to
yeast GAPD dimer was determined in this work to be 0.75 mM,
as illustrated in Figure 25. This value agrees well with
binding constant estimated for the overall dissociation
process (Stancel and Deal, 1969) and with the values for
'the reassembly4 (Stancel and Deal, 1969). The value of

(3.75 mg is also close to the values obtained for the binding

(of ATP to the tetramer (Yang and Deal, 1969a; 1969b) by

 

4We have previously shown (Stancel and Deal, 1969)
-that ATP is required for reassociation of the subunits, as
1vell as for dissociation.

154

two separate measurements. Thus it seems reasonable that

lie

AG°, in which case AGZ 5 AGE.

We must stress that this calculation is only a

AG°

rough approximation of AG° in the absence of ATP, because
the binding constant of ATP to the yeast GAPD tetramer has
not been precisely determined under conditions identical

to those used to produce dissociation. Nevertheless, the
calculation is useful and appropriate because it allows ;

us to compare the strengths of yeast GAPD subunit inter-

 

Ht,

actions to the strengths of subunit interactions in other
proteins (Klotz gt al., 1970) and because it provides us
with a rough reference point for discussing the chemical
nature of the subunit contact sites.

By analogy with Langerman and Klotz (1969) for the
dissociation of hemerythrin we may define a quantity AGE,

such that
AG; = AG°/n

‘where n refers to the number of subunit contacts disrupted
in the dissociation process. For yeast GAPD, AG; becomes
approximately 4;§_kca1/mole at 0°, since two subunit
contacts are disrupted in the conversion of tetramers to
dimers. By way of comparison, Langerman and Klotz (1969)
determined a AG; for hemerythrin of 5.8 kcal/mole at 5°

in Tris/cacodylate buffer, pH 7.0, conditions very similar

to those employed in this work.

155

Klotz et 31. (1970) have also tabulated values of
AG; for a number of proteins involved in association-
dissociation reactions, and in all cases the values are
between 2 kcal/mole and 8 kcal/mole.

CHEMICAL NATURE OF THE SUBUNIT CONTACT SITES. There are
two major experimental findings which suggest that hydro-
phobic interactions between GAPD subunit contact sites
play a key role in maintaining the tetrameric structure of
the native enzyme. The first is that in the presence of
ATP the yeast GAPD tetramer dissociates at 0°, but does
not dissociate at 23° (Stancel and Deal, 1969). The
second is that the breakdown of dimers to monomers is
drastically inhibited by increasing the ionic strength
(Figure 27).

The susceptibility of polymeric enzymes to dis-
sociation at low temperatures indicates that hydrophobic
interactions, i.e., non-covalent interactions between
apolar residues, are important in maintaining the struc-
'tural integrity of the enzyme (Kauzman, 1959; Scheraga
SE. _a__]._., 1962). Studies with model systems suggest that
:such cold-lability of proteins may arise from unfavorable
eantropy losses for the interaction of apolar groups as
the temperature is lowered (Kauzman, 1959). Kauzman
(1959) has thoroughly discussed the phenomenon of
Irydrophobic interactions in a classic review. This
txopic has also been previously discussed for the yeast

GAPD-ATP system (Stancel and Deal, 1969) .

 

N’Vfi -

 

 

156

The breakdown of yeast GAPD dimers to monomers is
drastically inhibited by increasing the ionic strength
(Figure 27), which also suggests that the dissociation
involves the exposure of hydrophobic residues to the
solvent medium (Tanford, 1964; von Hippel, 1969). The
inhibition of the dissociation of rabbit muscle GAPD at {“7
0° in the presence of ATP is also inhibited by increasing
ionic strength (Constantinides and Deal, 1969). Further-

more, a similar behavior is observed for the dissociation

 

of yeast pyruvate kinase at 0° (Kuczenski and Suelter, 1970). kw
These observations suggest that in all three cases, hydro-

phobic interactions play an important role at the subunit

contact sites.

The requirement of sucrose (or other polyhydroxy
compounds) for complete reassociation (Stancel and Deal,
1969) is also consistent with the proposed role of hydro-
phobic interactions. If hydrophobic residues (involved
in subunit contacts) become exposed to solvent
as a result of dissociation at 0°, "iceberg" structures
(Frank and Evans, 1945) of water would probably form
aaround the exposed residues to form complexes similar to
(slatharate structures (Kauzmann, 1959). Reassociation of
saubunits (at 23°) would then require that water structures
“fliich formed around exposed hydrophobic groups "melt"
CBrandts, 1969) before the pr0per subunit contacts are

fiarmed, since water would be excluded from the hydrophobic

157

subunit contact sites in the tetrameric structure.
Therefore, polyhydroxy compounds such as sucrose, which
disrupt ordered water structures, should aid the reas-
sociation process. Consequently, the requirement of the
reassociation process for sucrose (Stancel and Deal, 1969)
is consistent with the idea that the subunit contacts
involve hydr0phobic interactions.

A more quantitative explanation for the effect
of salt on the rate of breakdown of dimers, can be obtained

using transition-state theory. Thus, from transition-state

theory we may write

D e 5 D* -——> M

The rate constant for the process at a given salt con-

centration, ks' is given by (Gould, 1959)

k8 = [kT/h][(D*)/(D)J = [kT/hJEKgq] (YD)
(YD*)

 

1Mhere k and h and the Boltzmann and Planck constants,
.nespectively, K* is the equilibrium constant for the
isomerization of dimers in the ground state (D) to the
transition state (D*) in the standard state (taken here
tx3 be the standard dissociation medium, minus salt), and
y luas the usual significance of activity coefficient.

Since the activity coefficients in the standard

state are unity, we may write

(.7

 

158
ko = [kT/h][K*]
and it is seen that the previous relationship reduces to

(YD)

S 0 (YD*)

 

Thus it is seen that

‘b‘t ' ."

log kO - log k5 = log YD* - log YD

 

Since k0 is clearly greater than ks (Figure 26), YD* must m

also be greater than YD’ This is the expected result for

a process in which hydrophobic residues become exposed to

the solvent medium (aqueous) in the transition state

complex. It also explains the description of the salt

effect on the rate by the relationship

log (kc/ks) = S C (page 130)

sixume the activity coefficients will depend on the ionic
strength of the aqueous medium in which dissociation
occurs.

Dissociation and reassociation studies provide
valuable information about subunit contacts, but a precise
determination of the chemical nature of subunit contacts

in a polymeric enzyme can only be obtained from detailed

x-ray analysis. At present, hemoglobin is the only polymeric

protein for which X-ray analysis has reached a resolution

159

down to 2.8 X (Perutz SE 31., 1968a; Perutz et 31., 1968b).
For the hemoglobin molecule contacts between unlike sub-
units (a and 8) involve mainly apolar residues.
REASSOCIATION OF GAPD MONOMERS TO TETRAMERS. The kinetics
of reassociation indicated that the rate limiting step was
first order at high protein concentrations, but second

order at low protein concentrations. A mechanism consistent

with these observations is given below

4 MONOMERS ——> 2 DIMERS ———-> TETRAMER*—-—-> TETRAMER

(inactive) 1' (inactive) 2 (inactive) 3_ (active)

The kinetic analysis does not allow us to determine which
of the pr0posed second order steps is limiting at low con-
centrations. Similarly, it is not possible to demonstrate
that the tetramer formed in step 2 is completely inactive,
but only that its activity is much less than that of the
tetramer depicted as active in the proposed scheme.

As pointed out earlier (p. 90) steps 2 and 3 of
the proposed mechanism should probably be written as
reversible, since recombination of globular subunits
produced at 0° with ATP (p. 79) and recombination of
«completely disordered subunits produced by exposure to
8 g urea (Deal, 1969) both Show decreasing recovery of
eactivity below protein concentrations of 0.04 mg/ml, thus
asuggesting that the recombination in both cases involves

a: concentration dependent equilibrium step.

.1.“ '3'!"

 

 

160

The most interesting aspect of the proposed
mechanism for reassociation is the proposed equilibrium

between active and inactive tetramers (step 3). In

 

addition to the kinetic evidence reported in this paper,
two independent studies (Yang and Deal, 1969b; Kirschner,
1968) have presented data supporting an equilibrium Fm;
between two different tetramers of yeast GAPD.

Yang and Deal (1969b) showed that ATP greatly

increased the susceptibility of yeast GAPD to chymotryptic

 

1‘] “Kli'v . ."

digestion. Since these experiments were performed under
conditions which did not yield any detectable dissocia-
tion, it is clear that the ATP caused a conformational
change in the tetramer sufficient to greatly increase the

susceptibility of the enzyme to digestion (Yang and Deal,

1969b).

Kirschner (1968) has also shown that the tetramer
of yeast GAPD undergoes an isomerization reaction at pH
8.5 and 40°. He has shown with st0pped-flow techniques
that.the thermodynamically stable form of the tetramer
.is catalytically active while the second tetrameric
cxynformation observed is practically devoid of catalytic
austivity (Kirschner, 1968). Interestingly, the inactive
iflarm.of the tetramer can bind all of the Substrates (NAD,
gfiLyceraldehyde-3-phosphate and arsenate), even though it
ceunnot catalyze the required hydride transfer (Kirschner,

.19E58). These results and the results presented in this

161

work for the reassociation of GAPD subunits to the
tetrameric enzyme compliment each other well in that they
provide independent evidence for the existence of two
forms of the GAPD tetramer which have greatly different
catalytic properties.

COMPARISON OF THE PROPOSED MECHANISM FOR REASSOCIATION rm-
0F GAPD SUBUNITS To OTHER ENZYMES. Steps 2 and 3 (dimer ‘
association and isomerization of an inactive to an active

enzyme respectively) of the proposed mechanism for the

 

reassociation of GAPD subunits are identical to the L
mechanism of recombination of aldolase dimers produced

by acid treatment of the native aldolase tetramer (Blatti,
1968). Furthermore, like the recombination of GAPD sub-
units, recombination of aldolase monomers produced by
acid dissociation of the native tetramer, also follows
«good.first order kinetics for the over-all process under
certain conditions (Deal et_al,, 1963). Thus it seems
that in.both cases, the formation of the native, active
enizymes involves a conformational change as the final
satep which produces fully active enzyme.

The kinetics of reassociation of inactive argin—
nosuccinase monomers to an active dimer (Schulze e5 £11.,
1£T70) are similar to the kinetics of reassociation of
GAPD and aldolase subpolymers in that the final step of
true reassociation process appears to be the isomerization
of an inactive dimer to an active dimer. For arginnosuc-

ciJmase, which also undergoes cold dissociation (see p. 19),

162

the kinetics of reassociation are second order for reas-
sociation in imidazole buffer and first order for reas-
sociation in phosphate buffer (Schulze $2.31)! 1970).
Thus, for all three enzymes it appears that the final step
in the reassociation process is a first order isomeriza-

tion step.

MATERIALS AND METHODS

DISSOCIATION. Prior to the addition of ATP, enzyme

 

samples were always pre-incubated for 4-12 hrs. at 0° at

 

71].“1‘VL I _ f
.‘

a protein concentration 1.11 times greater than that
incubated in the text and legends for the individual
experiments. After the incubation, 0.1 m1 of concentrated
ATP (10 times the final concentration for dissociation)
previously adjusted to pH 8.0, was added EE_Q: to 0.9 ml
of the pre—incubated enzyme solution. The time of addi-
tion of ATP was taken as time zero for the kinetic analysis.
For example, to obtain the data in Figure 20, GAPD (0.04
:mg/ml in 0.02 M Tris, pH 8.0 and 0.075 M_B-mercaptoethanol)
‘was incubated overnite at 0°. After this incubation, 0.1
Inl of 50 mM_ATP (0°) was added to 0.9 ml of the enzyme
solution at 0°. Therefore, the final concentrations of
<1APD and ATP were 0.036 mg/ml and 5 mg respectively.
.Aliquots of this solution, which were kept at 0°, were
'then withdrawn at various times and assayed.

Assays of catalytic activity were performed as

fkallows. A cuvette containing 0.36 ml of the assay mix

163

(see Chapter Three—-Methods) was placed in the sample
holder of the Gilford spectrophotomer (the temperature
of the sample holder was maintained at 25°). Exactly 30
sec after the cuvette was placed in the chamber, a 0.04
ml aliquot of glyceraldehyde-3—phosphate was added with F”
a 50 ul Hamilton syringe and the mixture stirred for
exactly 5 sec. with the tip of the syringe. The chart

drive was turned at this time. Exactly 30 sec. after

 

the glyceraldehyde-3-phosphate addition, an aliquot P~
(10-25 ul) of the GAPD-ATP mixture was added with a

second 50 ul Hamilton syringe and gently stirred for
exactly 5 sec. with the tip of the syringe, and the
monitoring of the absorbance at 340 mu was begun as
quickly as possible.

REASSOCIATION. GAPD (0.4 mg/ml in 0.02 M Tris, pH 8.0

and 0.075 M B-mercaptoethanol) was dissociated at 0° by

a 12-14 hr. incubation with 10 mN_ATP. The enzyme was
then diluted (at 0°) with the dissociation solvent
(including 10 mM_ATP) to a protein concentration exactly
two times that at which the reassociation kinetics were

to be followed. This diluted enzyme sample was then
further diluted (at 0°) with an equal volume of a solution
containing 0.4 M_Tris, pH 6.9, 0.075 M B-mercaptoethanol
and.20% sucrose. The final conditions for reassembly
vwere therefore 0.21 M_Tris, pH 7.0, 10% sucrose, 5 mM_ATP,

.and.0.075 M MSH. (A similar procedure in described in the

164

Methods of Chapter Three.) Finally, the sample in 10%

sucrose (normally 1 ml) was removed from ice and placed
in a large volume of water (at least 2 l) at 23° or 15°
as indicated for individual experiments--this was taken
as time zero in the reassembly studies. Aliquots were

then removed and assayed exactly as above for the dis—

sociation studies.

All other methods and materials have been described

in Chapter Three.

 

.2.“ ‘11 J

 

CHAPTER FIVE

ISOLATION AND CHARACTERIZATION OF AN INHIBITOR PROTEIN
FROM YEAST WHICH INHIBITS GLYCERALDEHYDE-3-PHOSPHATE
DEHYDROGENASE AND IDENTIFICATION OF THE INHIBITOR

PROTEIN AS.TRIOSEPHOSPHATE ISOMERASE

165

ABSTRACT

Examination of yeast cell extracts indicated the
presence of a protein which caused an apparent inhibition rm}
of yeast glyceraldehyde-3-phosphate dehydrogenase. Initial -
experiments indicated that this inhibitor protein was not

a protease, and that the inhibitor was present in intact

 

yeast cells. The inhibitor protein was purified and its 5
molecular weight (58,000) and amino acid composition were ,
determined. Maximal inhibition was then shown to occur

when the inhibitor protein and glyceraldehyde-3-phosphate
dehydrogenase were mixed in a 1:1 £9133 ratio, however,

several physical tests failed to reveal the formation of

a complex between the two proteins. The inhibitor protein

‘was subsequently identified as triosephosphate isomerase.
iNevertheless, this identification of the inhibitor protein

as triosephosphate isomerase is not sufficient to clearly

explain the observed inhibition, since (1) the inhibition

:is not completely overcome by glyceraldehyde-3-phosphate,

(2) the inhibition depends on the total protein concentra-

‘tion as well as the ratio of the two proteins, and

(3) triosephosphate isomerase seems to activate glycer-
aildehyde-B-phosphate dehydrogenase under certain condi-

tions .

166

 

167

INTRODUCTION

 

One major question concerning the dissociation of
yeast GAPD into subunits in ATP was whether this process
had any i2_yiyg significance. This seemed rather unlikely,
Since the major effects were produced only at low temper-
atures (0°) and low protein concentrations, conditions not [fig
likely to be encountered in_yiyg.

Another possibility was that the effects of ATP 2

at higher temperatures and higher protein concentrations

 

might be to produce minor conformational changes in the
native enzyme which would render the enzyme more sus-
ceptible to proteolytic degradation. This suggestion
seemed plausible for the yeast GAPD system because:

(1) Yeast contain numerous proteases with different
specificities (Hata SE Ei" 1967; Lenny and
Dalbec, 1967: Felix and Brouillet, 1966).

(2) ATP greatly enhances the digestion of native
GAPD by purified chymotrypsin (Yang and Deal,
1969b).

(3) GAPD comprises 20% of the total soluble protein
of baker's yeast (Krebs‘et_al., 1953), thus
raising the possibility that GAPD, or an
altered form of GAPD serves someig.yiyg_
function other than the catalysis of the
oxidative phosphorylation of glyceraldehyde-3-

phosphate.

168

Consequently, it was decided to examine yeast extracts to
determine if a protease was present which would degrade
GAPD.

Since GAPD had to be used as the substrate to

assay the protease, the presence of high levels of GAPD

 

in yeast extracts complicated the problem. Therefore, F“
the first approach was to obtain a broad activity protease .
fraction from yeast using procedures reported in the E
literature. It was hoped that such a fraction would be ‘
relatively free of GAPD activity, and could then be ‘ !.

tested for protease activity against purified GAPD. We
shall now designate this protein fraction as either

protease fraction or inhibitor protein in order to give

 

 

it consistent nomenclature throughout this thesis.

RESULTS

PREPARATION OF A PROTEASE-CONTAINING FRACTION. Initial

 

attempts at obtaining such a protease containing fraction
from yeast by utilizing the procedure of Lenny and Dalbec
(1967) were quite successful. One pound of baker's yeast
was allowed to autolyze in 80 ml of CHC13, 550 m1 of water
‘was added, the pH was adjusted to 7.3 and the mixture
centrifuged. The supernatant was adjusted to pH 4.8,

and the precipatate obtained between 47 and 63% saturation
‘with ethanol was taken up in citrate buffer, pH 5.5. The
absorbance of 2.5 at 280 mu of this preparation was taken
as a crude measure of the protein content; a value of

2.5 mg/ml was estimated.

169

TEST FOR INACTIVATION OF GAPD AND TESTS FOR PROTEASE
ACTIVITY AGAINST HEMOGLOBIN, GAPD, AND ALDOLASE. When
solutions of purified yeast GAPD were incubated with
aliquots of the inhibitor protein, there was a rapid loss
of GAPD activity. The agent responsible for this loss of
activity was a protein, since it was heat-labile, TCA-
precipitable and non-dialyzable. These experiments are
summarized in Table VI.

It was initially thought that this loss of activity
was probably due to a protease in the inhibitor protein
preparation. The inhibitor protein preparation was there-
fore examined for proteolytic activity using 1% hemoglobin
as a substrate (Figure 32). The inhibitor protein prepara-
tion released considerable amounts of TCA soluble ninhydrin
positive materials from hemoglobin and this was inhibited
by 1,mg_phenylmethylsulfonylfluoride, an inhibitor of
;proteases with serine residues at the active site
(Steinman and Jakoby, 1967). This seemed to provide
good evidence that the inhibitor protein preparation
contained a protease which could digest GAPD.

Surprisingly, treatment of the inhibitor protein
preparation with 1 mb_d_ phenylmethylsulfonylfluoride did 293;
prevent the loss of GAPD activity upon incubation with the
treated inhibitor protein preparation. This unexpected
result suggested that the observed loss of GAPD activity

(Table VI) was not the result of proteases present in the

 

170

TABLE VI. LOSS OF GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE
ACTIVITY IN THE PRESENCE OF INHIBITOR PROTEIN

 

Treatment of Inhibitor % GAPD Activity Remaining
Protein after 5 min. Incubation*

NONE 36 [Ta
1 g

Supernatant from

 

 

 

0.34 N TCA. b
preCIpitation ' 74
100°, 5 min.a 91 5
30°, 2 1/2 hrs. 25 L
20 hr. dialysis 29

 

*

Aliquots of the inhibitor protein preparation
(0.1 ml) were added to 0.9 m1 samples of GAPD (0.2 mg/ml)
in 0.1 N_Tris, pH 8.0 containing 0.1 g_8-mercaptoethanol.
After a 5 min. incubation at 23°, the GAPD activity was
measured as described in Methods. % GAPD activity was
calculated on the basis of the activity of a GAPD sample
treated in a similar manner, minus the addition of the

inhibitor protein.

aPrecipitate which formed was discarded by
centrifugation.

bAfter centrifugation, the supernatant was adjusted
to neutral pH with bicarbonate.

171

 

.QOHDDHom cHnonoe0: 0:: op coHuHoo0 o: HOHHQ

mmzm.m mloH :uH3 .CHE m How 0000:000H 003 CH0ponm HoanH:sH 0:: 00:0 Dm0ox0
10sHH oHH00IH0HHEH0 003 30000 Ammzmv 0oHHosHmecostmHm::0EHms0:m 0:9

.000“ 003 :8 0mm :0 00:0

1000000 0:: .GHE OH H0pm< .00000 003 Au0u03 :uH3 HHH ©000HHUV us0m00n sHHom 00

N02 0H 0m.mv momz_m m.o 00 HS m.~ coHumuuHHm H0000

HE ms.o 20:0 .00000 003 Amoo
.005 m vm.0 wo HE mN.H 0: @0000 0:0 ©0>OE0H 0003 mposcHH0 HE H 00EHu s0>Hm

may he .22 091000185 me 53.1.. He To :88 m 0.0 to 2.320980 3 00

HE 0.0 on 00000 003 Ao.m mm .0HHB CH HE\mE 00.0 00 HE o.mv 0H0uoum HODHOH:0H

.coHu0u0d0um 0H0uoum nouHQH:cH >0 cHnonoE0: 000000000 m0 00H000mHo .MM 0HsmHm

 

 

172

 

 

$052.2. ms:

 

 

 

.8 2 00 on 9 on .2 o. 00
< u 41 d14l 1 ‘ q . ¢< <4J<< 4 q ‘# 4.|.|¢I.l¢
.... - ....
.o..
2.39... «93.12. \ .. L «.0
3qu 3.5.. ......\
2.2.9... 3:232. .... . av.
323:2: _/ - 00 %
5‘0“... 8
... . w
.. a. N
.........o on« 0.. :a . D
.. 2.35020: o\o _. .. 3 m“
. O
m
.. QN W
Z-flOaOOEmI mo ZOP-hMO-O A

 

 

 

.- w-"F’f\_ ‘

 

 

 

173

preparation. It was therefore decided to examine the
protein specificity of this inhibition by the inhibitor

protein fraction. This was done by analyzing the effect

 

of the inhibitor protein preparation on the activity of
rabbit muscle aldolase, since this test would resemble
the experiments with GAPD more closely than the experiments rflfi
with the denatured hemoglobin. ;

When aldolase samples were incubated with the A

inhibitor protein preparation at 23° all the aldolase l

 

activity was lost within 30 min. This loss of activity
could not be reversed by the addition of phenylmethylsul-
fonylfluoride to samples of aldolase inactivated by this
treatment. If however, the inhibitor protein preparation
was treated with l mg_phenylmethylsulfonylfluoride prior
to incubation with the aldolase samples, the aldolase
retained 75% of its original activity after a 30 min.
incubation at 23°, indicating that a protease caused the
loss of the aldolase activity.

EVIDENCE THAT GAPD ACTIVITY LOSS WAS NOT DUE TO A PROTEASE.
tflue results with GAPD, hemoglobin and aldolase suggested
tflurt the inhibitor protein preparation contained a protein
‘whirfli caused the loss of GAPD activity by some mechanism
other than proteolytic degradation, and also contained a
protease(s) which degraded aldolase and hemoglobin. It
is ruyt surprising that GAPD was not degraded by the

protease(s) present since Halsey and Neurath (1962) showed

174

that native GAPD was not degraded by carboxypeptidase,
except in the presence of high concentrations of urea and
Yang and Deal (1969b) showed that chymotrypsin had little
effect on the enzyme in a 10 min. incubation. To further
test this possibility, samples of GAPD were incubated with
different amounts of the inhibitor protein preparation, K
and the GAPD activity was measured as a function of time.
If the loss of GAPD activity were due to proteolysis, the

total GAPD activity should decrease with time. This was

 

' {In a: “I...
i

not observed. As illustrated by the data of Figure 33 the
loss of GAPD activity did not increase with time after the
addition of the inhibitor protein aliquots to the GAPD
samples. These results and the failure of phenylmethy-
lsulfonylfluoride to prevent the loss of GAPD activity in
the presence of the inhibitor protein preparation clearly
indicated that the inactivation of GAPD was not a result
of proteolysis.

If the loss of GAPD activity was not due to
proteolytic degradation, it seemed likely that the activity
loss might be reversible. To test this possibility a 0.5
Inl sample of GAPD, specific activity of 104, was prepared
(0.4 mg/ml in 0.1 ngris, pH 8.0, and 0.1 fl.in B-
rmercaptoethanol). An aliquot of the inhibitor protein
lyreparation (50 ml) was then added and the mixture incubated
art 23° for 5 min. This treatment resulted in a decrease of

srxecific activity from 104 to 30. When this inactivated

175

.008:: 00:0o:©:: 0::

:0 0090000 0:03 0::0:onm 03: 0:: mo 0:::x:E 0:: mo 0:0:wflaa .HE O.H 003 0E:Ho>
H0:09 .Ho:0::0o:m0o:0E|m.w H.O 0:0 .o.m mm .0:H9 w H.O :: AHE\mE m.ov 000:0m
Ionpm:00 0:0:Q00:m|m:0©%:0pa0n0owfim o: @0000 0:03 AHE\mE m.mv :H0:onm Ho:::::::
:0 00ESHo> 00:0o:©:: 0:9 .::0:oum Ho::::::: 0:: %: 000:0mo:0>:0© 0:0:m00:m|m

I0Um:0ta0u0omaw :0 :o:::::::: 0:: :o :o::0::o:: mo 06:: 0:: mo :o0mmm aMI 0::mflm

”1|”:

176

 

Ann—.322): us..—

 

 

 

 

 

 

 

00 o: 3 on. 9. 00 ca 0. o
q a q q q a q 4 O
8. :0 o ...... m. .o F
a . ...S
on 4 nullinnnlluiltlillnuflz .
322
2.05.... «9.222. .1
00—
920 55> mo zo:<>:u<z.

 

 

(..9w-,-NIw-sa1owm NOV 0345

177

sample was then diluted ten-fold with buffer, the specific
activity increased from 30 to 108, the original value.
This experiment proved conclusively that the inactivation
of GAPD was not a result of proteolytic digestion.

POSSIBLE EXPLANATIONS OF THE EFFECT. There were basically

 

three other explanations for the observed loss of GAPD
activity in the presence of the inhibitor protein prepara-
tion.

(1) The inhibitor protein preparation could contain
a protein which was converting one of the sub-
strates of the GAPD reaction to some other
product.

(2) The activity loss of GAPD might be an artifact
of the procedure used to prepare the inhibitor
protein. For example, the protease fraction
might contain GAPD-like protein fragments,
arising from proteolytic degradation of GAPD
itself or other proteins during the preparation
of the inhibitor protein, which could bind to
the purified GAPD in the activity test and
cause activity loss.

(3) The inhibition of GAPD could result from the
reversible binding of a specific protein
inhibitor present in intact yeast cells.

The first possibility, that another enzyme in the

inhibitor protein preparation was rapidly removing the

 

 

178

substrates of the GAPD reaction, was tested by measuring
GAPD activity with and without added inhibitor protein as
a function of substrate concentration. Two samples of
GAPD were prepared, and an aliquot of the inhibitor
protein preparation was added to one of the samples.

The other sample (minus the inhibitor protein) served as

-,-"F

the control. Aliquots of both samples were withdrawn and
assayed with a fixed concentration of glyceraldehyde-3-

phosphate and varying concentrations of NAD; then the

 

samples were assayed at a fixed NAD concentration and varying 3~
concentrations of glyceraldehyde-3—phosphate. The results
are illustrated in Figures 34 and 35 respectively. In both
cases the activity loss was not overcome by the addition
of excess substrate above a concentration of about 1.5 mg.
Thus, it did not seem that the inactivation of GAPD resulted
from decreased substrate concentrations. (However, later
results disproved this conclusion--see p. 228.)

The second possibility, that the loss of activity
of GAPD might be caused by the presence of GAPD-like protein
fragments produced by proteolytic degradation of a protein
present in the original yeast extract, was evaluated by two
separate experiments. The first experiment was to see
whether preparation of the protease fraction in the
presence of the protease inhibitor, phenylmethylsulfonyl-
fluoride, would result in the fraction possessing decreased

ability to cause GAPD activity loss. The second experiment

179

.0HH0:00 :0: :x0: 00m .HE\mE H.o 003

:H0:oum Ho::::::: :0 :o::0:::0o:oo “HE\mE m.o 003 000:0mon0>:00 0:0:m0o:m
amn00m:00a0u0owam mo :o::0:::0o:oo .00o::02 :: 00::00000 00 0090000
A::000Hm ::0:o:m Ho::::::: :50::H3 0:0 :::3v 000:0mou0m:00 0:0:m00:mlm
n00>:00H0H0o%H0 am 010: x m.n 003 090000 0:: :: 0:0:m00:m|m1009:00H0H0omam
:0 :o::0:::0o:oo 0:9 .::0:o:m Ho::::::: m: 000:0wou0m:00 0:0:m00:mlm

1009:00H0H0omaw :0 :o:::::::: 0:: :o :o::0u::0o:oo :42 mo :o0wwm .wm 0::mHm

 

 

180

 

d

 

a... a4...

a

2.2.0.... 8020.22.
.... .
Om<0

Qm<0 hm<m> .0

   

20.95.12.

"“04 O
O

O
N

  

o o 6
Q 0 V

(.-ow-.-NIw-sa1owm mnv ‘DidS

 

 

 

 

 

 

 

 

l8l

.:x0: 0:0 em 0Hsmflm mo
0:0m0a 00m .$8 H 003 090000 0:: :: :42 mo :o::0:::0o:oo 0:9 .::0:onm
Ho::::::: 9: 000:0mom0>:00 0:0:m00:m|ml00>:00a0n0owam :0 :o::::::::

0:: :o :o::0m::0o:oo 0:0:m0o:m|ml00>:00fi0s0omam mo :o0mmm .IM 0usmflm

 

182

 

 

ji-

 

35 0.3.

Q

...... ...... .\

Choc 0\

 

 

00000000000000

0.30 54w» “.0 20.20.12.

d n a m u m . %o

l

8 9 2
(.-9w-.-le- snow") - mnv '3365

O
a

3

 

 

183

attempted to test for the production in the protease frac-
tion of GAPD-like fragments by a protease not inhibited by
phenylmethylsulfonylfluoride.

In the first experiment the inhibitor protein was
prepared by two different methods. The first preparation
method was as previously described. The second preparation
method was similar to the first except that all aqueous i
solutions used contained 1 mg phenylmethylsulfonylfluoride.

The protein concentrations of both the preparations were 0

 

the same, and both preparations produced the same amount
of activity loss when incubated with purified GAPD.

In the second experiment, a sample of baker's yeast
was autolyzed in CHC13, distilled water was added, and the
mixture allowed to stand at room temperature as previously
described. Aliquots were then removed at various times
and quickly placed in ice, since any proteolytic activity
should be substantially reduced by lowering the temperature
to 0°. After the last aliquot was removed and chilled in
ice (23 hrs), each of the 10 aliquots were used to prepare
a sample of the inhibitor protein as previously described.
If the inhibitor protein was not present in intact yeast
cells, but was produced as a result of proteolytic degrada-
‘tion during the preparation procedure, then the samples
vflnich remained at room temperature the longest should
(xyntain the highest levels of the inhibitor protein and

sruauld produce the greatest loss of GAPD activity. This

184

was not observed, as illustrated in Figure 36. The samples
prepared from the different aliquots (which had remained
at room temperature for various times) all produced the
same loss of GAPD activity, and the samples all had the
same protein concentration. Thus the inhibitor protein
responsible for the loss of GAPD activity was present in
the intact yeast cells, and was not produced by proteolysis
during the preparation procedure.

If the third possibility was correct, i.e., that
a specific GAPD inhibitor protein was causing the loss of
GAPD activity, then a plot of the loss of GAPD activity
(expressed as % inhibition) at a fixed GAPD concentration
vs increasing amounts of the inhibitor protein should
increase linearly when the GAPD was in excess and then
level off to a plateau value when the inhibitor was in
excess. The results of such an experiment (Figure 37)
<qualitatively seem to resemble such a titration curve,
suggesting that the GAPD is being "titrated" with a
pnnatein present in the inhibitor protein preparation,
ssince the curve (Figure 37) levels off to a plateau
region at less than 100% inhibition.
EEIRIFICATION OF THE INHIBITOR PROTEIN. Since the preceeding
results were quite exciting and seemed to offer great
prxxnise with respect to control of glycolysis, it was
decided to attempt a further purification of the protein(s)

responsible for the loss of GAPD activity. It was not

 

 

0 .

 

..\

185

.0HH0:00 H0::::: :0: 000::02 00m .Ho:0::0o:m0ou0film m H.o :::3 ~o.m mm

.0:H9 w H.o :: .HE\mE N.o. 000:0mon0>:00 0:0:m00:m|m100m:00a0u0owam mo HE m.o

o: :o::0n0m0um ::0:onm :o::::::: :0>:m 0:: mo :osvHH0 a: om 0 0::000 m:
00H5000E 003 000:0mou0>:00 0:0:m0o:mlmn00>:00a0u0omHm mo :o:::::::H

.maflmum0 00: 0x00 000 .0E:u :0 000020: 0:000:000 :00 0mm

:0 00::080: 00: :o::3 0:00H:x0 HH0o 00500 :0 0:osv:H0 Eon: 00::0::o 0:03

0:0::0H0m0nm ::0:osm Ho:H::::H .0:o::0u0m0um ::0:0Hm Ho::::::: ::0H0mw:0

m: 000:0mou0m:00 0:0:m00:mlmI00>:00H0H0owam mo :o:::::::H .lm 0nsm m

 

186

 

 

 

 

2 _. 3...: o2 2 22:
0— o 0 V a 0
o _1[- 1 1 q q a O
— u .0—
H 2
mm. .00
0
8 .
2
E n .. Om
C
m 20:32.2. o\o *0
R v: s ......0.\.. 0*
w Izloluolsm . .
B ../. “as...” ................ 0\0
A m 0H... .00
—— ......... o on
oeu<
0 .00

 

 

NOILIOIHNI "/0

187

.00o::0z :: 00::H0000 00 0090000 0:0 :30:0:::3 003

0:::x:E 0:: mo :oswfia0 :< .HE 0.: mo 0EsHo> H0:o: 0 :: .Ho:0::0o:m0ou0film
.2 To a»? .0... m: .32. .0 To 5 Se}: To. 00080809000 32:09.:
Iml009:00a0u0o9am 0: 00000 0:03 AHE\mE m.m. :o::0H000Hm ::0:o:m Ho:::::::
0:: mo 0:::050 0:0:u0> .000:0mon09:00 0:0:m0o:mlm:009:00H0H009H0 :0

:o:::::::: 0:: :o :o::0n::0o:oo :H0:onm uo::::::: wo :o0mmm .MM 0H5mwm

 

188

..2\.2.20¢.. «0:222. ....

 

 

1 q ‘ a [-i I

an 0.9 ...a \ l

920 .202 u .0 \ 1
..

(0... 0:. 00. -0» 00 9. 0a 0

8 9
NOlllfllHNl °/.

8

8

 

 

189

possible to attempt a purification directly from crude

cell extracts because of the very high amounts of GAPD
present in such extracts. The purification of the inhibitor
protein was therefore begun with the previously described
protease fraction obtained by the method of Lenny and
Dalbec (1967).

The following section describes the purification
of the inhibitor protein. The initial steps involve cell
lysis, acetic acid treatment and the 47-63% alcohol frac-
tionation and are identical to the steps in the preceeding
section. Assays for the presence of the inhibitor protein,
using purified GAPD to measure GAPD activity loss, were
begun with the 47-63% alcohol fraction, since this was
the first fraction which was relatively free of endogenous
GAPD activity. Assays could not be performed at earlier
steps because of the very high endogenous GAPD levels.

DEFINITION OF UNITS. One unit of activity is arbitrarily

 

defined as the amount of inhibitor protein which produces
10% loss of GAPD activity when added to a solution con-
taining 0.2 mg/ml of purified GAPD at pH 8.0 in 0.1 g

Tris and 0.1 g B-mercaptoethanol, in a total volume of

0.5 ml. The amount of inhibitor protein added was adjusted
so that the loss of activity was approximately 25-50%,
since the relationship between the amount of inhibitor
protein and activity loss was fairly linear in this range
(see Figure 37, p. 188). Total protein was by assuming

that an absorbance of 1.0 at 280 mu corresponded to a

190

protein concentration of 1.0 mg/ml. A summary of the
over-all purification is given in Table VII.

CELL LYSIS AND TREATMENT OF THE CRUDE EXTRACT. Two pounds
of fresh Red Star Baker's yeast were crumbled into small
pieces, and 160 ml of chloroform was added. The mixture
was kept at room temperature and stirred occassionally

for 60-90 min. until the suspension was thoroughly
liquified. At this point, 1100 ml of distilled de-ionized
water was added. The mixture was stirred briefly (ap-
proximately 5 min.) and the pH adjusted to 7.3 with l N
NaOH. The mixture was then left standing at room tempera-
ture overnite.

After standing overnite, the solution was centri-
fuged at 1300 x g for 20 min. The precipitate was discarded
.and.the supernatant (1475 ml) was adjusted to pH 4.8 with
3 91 HOAc. As the HOAc was added, a white precipitate
formed. To this mixture was added 2 ml of chloroform.
The mixture was then left stirring gently overnite. If
after stirring overnite, the precipitate had not completely
re-dissolved, it was removed by centrifugation at 1300 x g
for 10 min.

52—63% ETHANOL FRACTIONATION. The resulting solution

(1475 ml) was placed in an ice bath until the temperature

reached approximately 5°. At this point 1475 ml (l volume)

of 95% ethanol (-10°) was added slowly with stirring. Five

minutes after the addition of ethanol was complete, the

 

 

191

 

.::0E:000: :000 :0 000000000 0:: mo 0::0:00 000 :20: 000

 

 

 

000.02 000 02\02 0.0 . . 02 00 02002000
0:0:00:000H
000.02 000 02\02 2.0 :2 mm 2000020000000
0020202
000.00 00 :s\02 0.00 :2 20 020200000 0000020
20::0280 0:0
0:00:00 0::200000
000.00 00 02\02 0.0 02 00: 2000020000000
0020200 000-00
:2 0000 020200000 0202
020 00000 0:00
0:::0 9::>::om :o::00::00:00
:0009 0:0:0000 ::0:00m 0E::o> 0:08:0009
2020022 200020220 220.20 200020020222 220.20 0222220 .00> 22220

192

mixture was centrifuged at 1300 x g for 5 min. and the
precipitate was discarded. A second 1475 ml volume of
cold ethanol was added slowly with stirring. After
standing for 5 min., the precipitate was collected by
centrifugation at 1300 x g for 5 min. The precipitate
was taken up in 150 ml of 0.01 H_citrate buffer (Na),
pH 5.5. After standing at 0° overnite, any insoluble
material was removed by centrifugation. The total volume
obtained at this point was 158 ml.
PROTAMINE SULFATE AND AMMONIUM SULFATE TREATMENT. To the
fraction from the ethanol step (158 ml) 7.70 ml of protamine
sulfate (20 mg/ml) was added dropwise at 0°. The dark brown
precipitate which formed was removed by centrifugation.

The supernatant (162 ml) was brought to 80%
saturation with ammonium sulfate by the addition of 91 g.
of solid ammonium sulfate at 0°. The solution was stirred
gently for 5 min. after addition was completed, and the
precipitate was collected by centrifugation and re-
dissolved in 60 ml of 0.01 fl_citrate, pH 5.5. The resulting
solution (66 ml) was then brought to 70% saturation with
ammonium sulfate by the addition of 31.2 g. of solid
ammonium sulfate, and stirred gently for 5 min. after the
addition was complete. The precipitate was collected by

centrifugation and taken up in 50 ml of 0.01 g citrate,

pH 5.5.

193

ETHANOL FRACTIONATION. At this point 90 ml (1.75 volumes)
of cold (0°) ethanol (95%) was added. The precipitate
which formed was suspended in 50 ml of 0.01 g citrate
buffer, pH 5.5, and left at 0° overnite. Insoluble
material was then removed by centrifugation.

The resulting solution (52 ml) was then dialyzed
for 24 hrs. at room temperature against 2 l. of 0.01 g_
citrate buffer, pH 5.5, containing 0.07 5 KCl. After
dialysis the preparation was kept at 0° prior to the
succeeding steps.

When this preparation was examined by disc gel
electrophoresis one major band was observed along with
6-7 minor bands. Since the bands were quite widely
separated further purification of the inhibitor protein
by isoelectric focusing was attempted.

ISOELECTRIC FOCUSING. The approach was to first apply a
small amount of the dialyzed protein preparation to an
analytical size isoelectric focusing column utilizing a
broad pH gradient (3—10) in order to determine the approxi-
Imite isoelectric point of the inhibitor protein. A larger
anmdunt of protein from the dialysis step could then be
applied to a preparative size isoelectric focusing column
usiJng a much narrower pH gradient, ranging about the

isoelectric point; this should afford the maximum separa-

tion.

194

A small fraction of the solution from the dialysis
step (25 mg of protein) was therefore applied to an isoe-
lectric focusing gradient, pH 3-10, using a small (110 ml)
LKB column. After 72 hrs. at 0° fractions were collected.
The inhibitor protein banded at a pH of approximately 5.0.

Once the approximate isoelectric point of the

inhibitor protein was known, a larger fraction of protein

’
r .1
.1 _

(100 mg) from the dialysis step was applied to a larger

LKB isoelectric focusing column (440 ml) utilizing a much 1

 

narrower pH gradient of pH 4-6. After focusing for 72 hrs. EM
at 0°, fractions were collected and analyzed for both

protein content and inhibitor protein activity. The

results are shown in Figure 38.

Small aliquots of fractions 48-54 (which contained
the inhibitor protein activity) were analyzed by disc gel
electrophoresis. Fractions 49-54 showed only a single
band, and were therefore pooled. Fraction 48 showed a
minor contaminant and was therefore discarded.

The pooled fractions (40 ml) were then brought to
90% saturation with ammonium sulfate by the addition of
solid ammonium sulfate at 0°. After the addition was
(complete the solution was allowed to stand at 0° for 1 hr.
{the precipitate was then collected by centrifugation and
(iissolved in 10 ml of 0.01 fl_citrate, pH 5.5. The resulting

solution was stored at 0°.

195

.mconpmz CH confluommp mm cmusmmmfi mm3.>pfi>flpom mmmcomonphnmc

mpmzmmonmlmnmcmzocamumoham mo coapflnflncfl mpg paw cmcflfiuouop mmz 1E 0mm um
mocMQHOmnm may .cmnsmmms mmz mm mnu cmuooaaoo mums m20fluowum can “ovum

.munpoooum mcflmsoow mo maflmump now mponpwz mom .om um

Hg.mn now vchDUOM umumm cfisaoo mxn HE Ovv m Eoum cmuomaaoo mnoz mcoHuomum

.mcflmDOOM oanuomamomfl >3 cflmuoum Houflnflncfl msu mo GOAumoaMMHSQ .mm.mnsmwm

196

 

 

 

O
I ..m m.» 10—
— gm
A , . \i .
mu .. m4 Tau 20:352.“ .0“
m 5.
... a. ..i .8
C m
N m
M v: ."umu 10v
m Twila»?
m. . m
A m1 10m
0: 100

 

 

 

NOIlISIHNI "/9

 

197

PROOF THAT THE INHIBITOR PROTEIN WAS NOT AN ALTERED OR

INCOMPLETE SUBPOLYMER OF GLYCERALDEHYDE-3-PHOSPHATE

DEHYDROGENASE. As previously mentioned, one possible

explanation for the observed loss of activity of GAPD in

the presence of the inhibitor protein was that the inhibitor

protein was an altered or incomplete subunit (or dimer) of Fe:

GAPD with little or no GAPD activity. Consequently,

“-‘d :4“ ' . .

experiments had been designed to test this possibility

using impure inhibitor protein (p. 177). Rigorous tests

 

of this possibility, however, were not possible until im
the inhibitor protein was obtained in pure form. U
A further reason for rigorously evaluating this
possibility was the observation that the subsequently
purified inhibitor protein showed very low levels of GAPD
activity. The GAPD specific activity measured in the
purified inhibitor protein preparation was approximately
1% of the specific activity of pure, native GAPD. There
were two likely explanations for this observation.
The first explanation for the trace amounts of
GEEK) activity in the purified inhibitor solutions was that
the preparation was simply contaminated with GAPD. This
was a distinct possibility because of the large amount of
GAPD present in yeast; GAPD comprises (la-20% of the total
soluble protein of yeast (Krebs gt EH 1953) .
The second explanation was that the inhibitor

protein was an altered or incomplete subpolymer of GAPD.

198

Such an altered subpolymer could possibly bind to the
tetramer of yeast GAPD or could undergo an exchange reac-
tion with a subpolymer of the native enzyme to cause the
observed loss of GAPD activity. This possibility was also

likely, since the native tetramer is almost certainly

1.9!.
'Q
.7
,~

involved in equilibria with its subpolymers (see Literature

Review).

Tests for immunological cross reaction between the

1». _ .‘

two proteins provided a way to decide between these possi-

 

bilities. Therefore, antibodies to both GAPD and the
inhibitor protein were prepared. A second approach which
could indicate whether the inhibitor protein was an
altered form of a GAPD subpolymer, possibly produced by
proteolytic degradation, was to determine the amino acid
composition of the purified inhibitor protein and the
composition of purified GAPD. Therefore, the amino acid
composition of both proteins was also determined.

TESTS FOR CROSS REACTION BETWEEN THE INHIBITOR PROTEIN AND
ANTI-GAPD AND BETWEEN GAPD AND ANTI-INHIBITOR PROTEIN.
Antibodies to GAPD and the inhibitor protein (prepared as
described in the Methods section) showed no cross reaction
wiifli purified inhibitor protein and GAPD respectively.
This is clearly seen by the double diffusion patterns
illustrated in Figure 39. Starting with the tOp well in

each pattern and proceeding clockwise, the wells contain

AbGAPD' inhibitor protein, Abinhibitor protein and GAPD.

199

.mawmuoc

nosun5m How uxmp mom .omm um mpflcum>o coflmswwflp “comm po>ummno mums

cam .Hmoflpcmpfl mum omHSDOflm mcnmuumm nuom .AHE\mE.h.ov omdw can .cflmuoum
nouflhflccfinflucm .AHE\mE o.Hv cflmuonm HouflnflnCfl .omdonflucm cwmpcoo mHHmB
msu .mmHonoao mcflpooooum cam cuwuumm comm cw HHmB moo or» nufl3 mCfluumum

.cflmuoum Houflhflscw cam ammo ammo» mo mcumubmm coflmSMMHpocsEEH .mm on: em

200

 

....k.

...l'l' [’V’IL’ ‘7 I 'h‘
.41.!!NI‘u3-‘u‘I ‘11;

 

201

Both patterns depicted are identical and the concentrations
of GAPD and inhibitor protein are 0.7 and 1.0 mg/ml respec-
tively. Four other eXperiments were also performed in

duplicate using GAPD concentrations between 0.35 and 0.035

mg/ml and inhibitor protein concentrations between 0.4 and

 

0.04 mg/ml. All the patterns observed were similar to if?“
those illustrated in Figure 39. These experiments provided é
virtually conclusive evidence that the inhibitor protein :
was not some type of altered GAPD subpolymer.

AMINO ACID ANALYSIS OF THE INHIBITOR PROTEIN AND GAPD. Lm

 

The amino acid compositions of GAPD and the inhibitor
protein are given in Table VIII as residues per 69,000
molecular weight. This value was chosen since it is the
molecular weight of the GAPD dimer (Jaenicke 32 al., 1968),
and since preliminary studies had indicated that the
inhibitor protein had a molecular weight close to that

of the GAPD dimer (see following section). The value of
69,000 is also conveniently obtained for both proteins

by integral multiplication of the limiting value in both
cases, based on one tyrosine per limiting unit.

The values listed represent the average of two runs
for each protein, a 24 and a 48 hr. hydrolysis. The results
from the two runs were averaged for each protein, since the
two runs gave essentially identical results.

To check on the accuracy of the technique the amino

acid composition of GAPD as residues per 69,000 molecular

202

TABLE VIII. AMINO ACID COMPOSITION OF GLYCERALDEHYDE-3-
PHOSPHATE DEHYDROGENASE AND THE INHIBITOR PROTEINa

 

GLYCERALDEHYDE-3-PHOSPHATE INHIBITOR PROTEIN

DEHYDROGENASE (Residues per

 

 

 

(Residues per 68,712) 69,480)
AMINO ACID Literature ExPerimental Experimental
RESIDUE Value Value Value
b

Asp 69.2 70.6 69.1
Thr 44.3 39.9 28.3
Ser 47.1 43.7 39.4
Glub 35.9 35.8 64.7
Pro 23.3 22.3 16.6
Gly 50.3 53.1 62.1
Ala 66.4 66.5 70.7
Val 61.0 66.3 68.4
.Meth 12.9 11.5 NONE
Ileu 32.1 32.2 33.2
Leu 40.8 43.0 50.1
Tyr 21.6 24.0 18.0
Phe 19.6 19.9 29.9
Lys 40.1 47.8 48.8
Hist 10.8 13.2 6.4
.Arg 16.6 22.8 25.9

 

 

aValues based on one tyrosine per limiting unit.

bAssuming all residues are in free acid form and

not amidated .

203

weight was also calculated from the literature values
for the amino acid composition of GAPD (Velick and Furfine,
1963). The values are also listed in Table VIII, and they
agree quite well with the values for GAPD determined in
this work.

If the inhibitor protein was an altered GAPD Fen

subpolymer produced, for example, by proteolytic degrada-

— fl...“ .

tion, then the amino acid composition might be eXpected

to be similar to that of the GAPD. Examination of the

 

data in Table VIII, however, indicates that there are four
residues (underlined) that clearly occur in the inhibitor
protein more frequently than in GAPD. Of these, the
greatest discrepancy is noted for glutamic acid, which
occurs twice as frequently in the inhibitor protein as

in the GAPD. This composition seems highly unlikely if

the inhibitor protein arises from digestion of a GAPD
subpolymer, unless the hypothetical proteolysis selectively

jyielded a glutamate rich fraction.

Examination of Table VIII also reveals that methi-
onine is absent in the inhibitor protein, while the GAPD
contains approximately 12 methionine residues per 69,000
molecular weight. It thus seems inconsistent that the
iJfliibitor protein could be a degradation product of GAPD.

The results presented in this section for the
antibody experiments and the amino. acid (compositions

proved conclusively that the inhibitor protein was not

204

some type of slightly altered GAPD subpolymer still pos-

sessing the ability and structure to associate with GAPD.

MOLECULAR WEIGHT OF THE INHIBITOR PROTEIN AND
MOLAR STOICHIOMETRY OF THE GAPD-INHIBITOR
PROTEIN INTERACTION
The most likely explanation for the inhibition of
GAPD by the inhibitor protein seemed to be the formation
of some type of complex formation between the two proteins.

Such an explanation would be consistent with the data E

 

presented thus far (although later discoveries seemed to
rule out this interpretation). Furthermore, there seemed
to be some precedent for this type of complex formation,
since an examination of the literature revealed several

reports of inhibitor proteins for other enzymes (Bechet

and Wiame, 1965; Mikola and Suolinna, 1969).

The obvious approach to the problem at this stage
was to attempt to isolate a GAPD-inhibitor protein complex,
or at least to demonstrate that there was a physical
interaction between the two proteins. Such an attempt,
however, would be greatly facilitated by a knowledge of
the nature of the interaction, the molar stoichiometry
involved and the molecular weight of the inhibitor protein.
For these reasons the molecular weight of the inhibitor
protein was determined.

MOLECULAR WEIGHT OF THE INHIBITOR PROTEIN. The technique

used to determine the molecular weight of the inhibitor

205

protein was high-speed sedimentation equilibrium (thantis,
1964), often referred to as the meniscus depletion method
or thantis method. This technique was chosen because it
provides a sensitive test for the presence of low molecular
weight impurities in the sample and a test for dissociation
of the sample at low protein concentrations, as well as
yielding a molecular weight value.

In three separate experiments, at 17,980 RPM at

pH 8.0, and 20°, with an initial protein concentration of

 

0.5 mg/ml, the molecular weight of the inhibitor protein Li
determined by this technique was 56,000, 58,000 and 60,000.
These values were obtained from plots of the log of the
protein concentration (in fringe displacement) vs r2, the
distance from the center of rotation in cm2. In all cases
these plots were linear, indicating that no low molecular
weight impurities were present and that the inhibitor
protein itself was not involved in any association-
dissociation reactions.

STOICHIOMETRY OF THE GAPD-INHIBITOR PROTEIN INTERACTION.
The stoichiometry of the apparent interaction between the
inhibitor protein and GAPD was investigated by measuring
the loss of GAPD activity (at a fixed GAPD concentration)
as a function of the molar concentration of the inhibitor
protein. The results of such an experiment are illustrated

in Figure 40. The GAPD concentration was 3.6 x 10.8 M,

 

206

 

.moonuoz

a“ pmnflnommp mm omHSmmmE mmz coflufinflncw w map was .cflmuoum Hoaflnflncfl may

mo mDQSOEm mDOHHm> rues pmxHE mmz mmmcwmouphsmp mpmnmmonmumImphsmpacumoaam mo
coflumuacmocoo pmxflw m .cofluownmbcfl owmcomoucmnmp mumzmmonmnmlmpwropamumomam

Icflmuoum Houflnflncfl orb mo muuoEOH£0H0pm HmaoE may mo coflumcflfiumuma .dw musmwm

 

207

n23“: _HonE 3:212.”—

- rd «N 0.0 mfi Y? m.» «N —.— O

 

 

S

—.(- q - q 4 -.

 

ON
. .. a.
9.0. x o ... .. W28”— . \....... V o/o
M
\. m
. o m
\ ...... . u
‘ o\o .... O
...I.I.I.I.IP.L\.\- .... ..... .rOa N
3 u 0:5. .30: \...... .

 

208

and the inhibitor protein concentration was varied from

8 Mto 23 x 10'8 n4.

2 x 10-
The extrapolated lines from both ends of the

hyperbolic curve (Figure 40) intersect at an inhibitor

protein/GAPD molar ratio of 1.2. This data suggested

that the inhibitor protein might be interacting with GAPD fi‘“

to form a 1:1 complex, resulting in the observed inhibi-

tion.

At this point experiments were undertaken to physi-

 

__.'- ”I

cally test for an interaction between the inhibitor protein
and GAPD.

TESTS FOR THE FORMATION OF AN INHIBITOR PROTEIN-GAPD
COMPLEX. There are many possible physical techniques one
might use to test directly for the interactions or formation
of a complex between the two proteins. The techniques used
for the inhibitor protein-GAPD system were sedimentation
equilibrium, sucrose density gradient centrifugation,
fluorometric titration of (GAPD-NADH) mixtures with the
inhibitor protein and an immunochemical titration.
HIGH-SPEED SEDIMENTATION EQUILIBRIUM. If there is an
interaction between two proteins leading to complex forma-
'tion, then the observed mass distribution of a mixture of
‘the two proteins (with initial concentrations c1 and c2)
le a centrifugal field should differ from the sum of the
xmass distributions of the two proteins observed indi-

vidually at identical concentrations, c1 and c2, in an

209

identical centrifugal field. Consequently, the mass
distributions of the inhibitor protein and GAPD samples
were determined as a function radial position in two
separate centrifuge cells, and then the mass distribution
of a mixture of the two (in a third cell) was determined.

Experiments with GAPD and the inhibitor protein
were performed as follows, using Rayleigh interference
cells with wedge centerpieces. One cell was loaded with
0.06 ml of a dialyzed GAPD solution (1.0 mg/ml) and 0.06
ml of dialyzsate. A second cell was loaded with 0.06 ml
of a dialyzed inhibitor protein solution (1.0 mg/ml) plus
0.06 ml of dialyzsate. The third cell was then loaded
with 0.06 ml of the GAPD solution plus 0.06 ml of the
inhibitor protein solution, using the same protein solu-
tions which were placed in the first two cells. These
three cells and an interference counterbalance were then
placed in a 6-hole AnG rotor.

To test for non-specific protein-protein inter-

actions, a fourth cell was loaded with beef heart lactic

dehydrogenase (LDH) and a fifth cell with an LDH-inhibitor

protein mixture. The final concentration of LDH was

identical to that of the GAPD, and the final concentrations

of the LDH and inhibitor protein.in the mixture were
identical to the respective concentrations of the GAPD
eand the inhibitor protein in the GAPD—inhibitor protein

Inixture. The cells containing LDH were then placed in

 

210

the remaining two spaces in the rotor and a high-speed
sedimentation equilibrium experiment was performed.

Since the cells were all in the same rotor and
contained the same volumes of enzyme, minor variations
normally incurred between different centrifuge runs
(such as speed, temperature, vibrations and rotor expan—
sion) may be safely ignored. Special care was also taken
to insure that the radial distances measured for one cell
would be comparable to those in other cells. The plate
was carefully aligned and the patterns representing the
GAPD, inhibitor protein and the GAPD-inhibitor protein
mixture were read using the same fixed value for the
reference, i.e., the plate was not re-aligned between
the readings of the three different patterns, which lie
in rows above one another. The plate was then re-aligned
and the patterns representing the LDH, inhibitor protein
and LDH-inhibitor protein mixture were read using a second
fixed value for the reference. This procedure was used
since the maximum vertical displacement of the comparator
used to make the readings was only sufficient to read
four patterns without re-aligning the plate.

The results for the inhibitor protein, GAPD and
'the GAPD-inhibitor protein mixture are illustrated in
liigure 41, and the results for the LDH samples in Figure
442. The data is plotted as the log of the fringe dis-

}placement, (which is directly proportional to the log of

 

 

 

211

 

.maflmuop Hmrpnsm How pxou mom
.pommm omenomou Houwm n: em coxmu one: msmmnmouorm .Hocmnpmoummouoelm.w H.O
suflz .o.m mm .mflue.m «0.0 Ca 00m pom 2mm 0mm.na um poEHOmem mm? ucmEHHmmxm

one .HE\mE Om.o mmB samuonm some mo coflbmupcmosoo cflmuoum HMHDHCH are
.mcflmuonm 03h msu mo ousuxfle m cam ommcomompmcop mumgmmonmamImpmsopamnmozam

.cflopoum nouflhflscfl mo Esflnhflaflsom cofiwmucoeflpmm pmmmmlsmflm .ww musmflm

212

 

 

 

0.5 5m ..Om 0.0
. 1 . . . a . . . 4 q . 4 .ss . .
...-
\\
.0 ..
255:: 233.3 9 \\
35:2: 22:33: 4 us.
235.... 3552. o .1». 1
9:3 o ... ..
.... 3...?)
\‘ Ra Rama“.
CON °.Q In Cs k\\‘h
.0. 5.0.0.0. ..4..e.e.o 1
..V .Qoé HAM}.
\CHObb‘O t“‘\% 1
ab 4. h...
RV \« \0
\\\\.‘\ «Mme
\\ \s \«M& ..
s .1
. t .
ps » p‘MJ . b b

I o
P

'9
p

P.
N

 

 

(mawaavusno 39mm) 901M:

 

 

 

 

213

 

 

.mHHmpmp Hmsuusm How pxmu paw av whomem mo

pcmmma mom .mcwmuonm 03» opp mo musuxflfi M van AmDAV mmmcmmoupmnmp UflDOMH

.cflmponm HODHhflncH wo Esauhflaflsqm coflumucmEflpmm pmmmmlgmflm

...mw 853m

214

ANEUV Nu

 

«.3 NOm «On Nov «.0?
. a d u q . . . 4 1 . . d . T\j3
23.22 82:30 a .... .
.. £5.52: 28:88.: 4 .....e. .
. 2.22: «2352. o ........
1 :3 o .u.\o. o . O.—

 

 

 

 

(1N3W33V1d5l0 aonm) 901+:

215

the protein concentration in mg/ml), as a function of the
square of the radial distance from the center of rotation.
The theoretical curves for the two mixtures were obtained
by summation of the concentrations of the individual com-
ponents at a given radial position. i
The data of Figure 41 illustrates a small but
apparently significant difference between the theoretical
and observed curves for the GAPD—inhibitor protein mixture.
Furthermore, the mass distribution toward the bottom of
the cell should be higher for the observed mixture than
for the theoretical mixture if there is an interaction
between the two proteins. For the GAPD—inhibitor protein
mixture this is in fact the observed direction of devia-
tion. In contrast, the theoretical and observed curves
for the LDH-inhibitor protein mixture were practically
coincident. Although this data is consistent with complex
formation, the magnitude of the differences (Figure 41)
seems too small to offer strong evidence for complex
formation. Therefore, other procedures were used to test
for complex formation.
SUCROSE DENSITY GRADIENT CENTRIFUGATION. The resolving
jpower in a sucrose density gradient experiment is not as
great as the resolving power in the Model B, and a slowly
(aquilibrating system might be resolved in the Model E,
laut remain in equilibrium in sucrose density gradient

<axperiments. Thus, sucrose density gradient centrifugation

216

eXperiments are frequently useful for analyzing complex

formation of the type illustrated below
A+B:AB

Depending on the magnitude of the rate constants for

complex formation and breakdown, moving boundary eXperiments
for this type of system may reveal a single boundary or two
boundaries which do not separate if the equilibrium is
relatively rapid compared to the rate of migration in the
gradient. This technique thus offered an opportunity to
test the GAPD-inhibitor protein system for possible

complex formation.

Before the sedimentation behavior of GAPD and
inhibitor protein mixtures was interpreted, it was necessary
to examine the sedimentation behavior of the two proteins
individually. Figure 43 illustrates the individual sedi-
mentation profiles of the inhibitor protein and the GAPD.
These profiles were obtained by layering the inhibitor
protein on one gradient and GAPD on a second gradient,
and then running both gradients in the same rotor.l Under
the conditions of the experiment, the GAPD and inhibitor
;protein peaks (run on different gradients) separated

commdetely. Therefore, if the peaks did not separate

 

1The sedimentation coefficient of the inhibitor
protein obtained from this experiment is 4.1 S (based on
time value of the native GAPD), which agrees well with the
rmol. wt. of 58,000 which was previously obtained.

 

217

 

.mponuoz :H pmnfiuommp mm pflom oflccmw cpfl3 coaumuflmflomnm

omm<v maamofluuoaflpflhns» pmuommme mmz cflmuoum use

Hmuww A

.zmm 000.0v um mu: ea HOw mp3 coepmmDMHHusmo .cofluflmomfioo

mEmm orb mo mucmflpmnm co pmumwma mums .HOCMSpmoudmoumE|m_m mo.o paw
.Hox w mH.o .o.n mm .mumsmmosd w mo.o ca AHE\mE h.o mo HE Oa.ov mmHmEmm
campoum one .mmmcomompwswp mumbmmosmlmImphsmpamumowam ppm aflouonm

nouflhflnafl may wo coflpmmomeupcoo ucoflpmnm wuflmcop mmouosm .MI musmflm

 

218

 

..Zm.o<¢0 3082.3 2. Omam><¢h Eu

0.»

 

 

IN N.— O
— _
arOZquHON 0N
[due-.6 fl.’_qq-.—.-\-_...._qsq o
... .0\ o \ J
l/ \ / 0 IN. W
s
I O I o
a
.. \ \ .. w
0 N
I .. 3
a \ ..
I o 0:... no”
.. a: IQ. o2 .. 0
.w
I IN”
I m—.' ..
l IO.—
_pbb...._.-.h....~b..-_..L-

 

 

 

219

completely when a mixture of the two proteins was run on
a gradient together under these conditions, this would
suggest complex formation.

Other gradients had been layered with mixtures of
the inhibitor protein and GAPD and run in the same rotor
as those shown in Figure 43. One of these gradients was
identical to those used to obtain the profiles in Figure
43, and the other was identical except that the entire
gradient contained 0.5 mM NAD.

The sedimentation profiles of GAPD and inhibitor
protein mixtures are illustrated in Figure 44. In the
gradient without NAD there was no detectable complex
formation, since the two peaks separated completely. With
the NAD, the two peaks did not separate completely,2 which
seemed to indicate that some complex formation had occurred
in the presence of NAD. This eXperiment, however, was not
reproducible. Furthermore, in other experiments in which
NADH, ATP and mixtures of NADH and ATP and NAD and ATP
‘were present in the gradients, the two peaks separated
cummpletely, indicating no complex formation.

These experiments provided no evidence that the
.thibitor protein and GAPD formed a complex. However,
they'did not exclude this possibility, since the high

sucrose concentrations in the gradients may have prevented

 

2NAD at this concentration (0.5 mM) did not affect
tine sedimentation of either protein when they were run alone

in separate gradients.

 

 

220

.maflmpmp Hmnuusm MOM
pxmp cam me mmsmflm wo pcmmma mom .mpcmflpmnm mnu ca Ddz pmppm DUOSpHB
paw suHB mmmcmmoupwnmp wumnmmonmlmImpmsmpamnmowam paw :flmuoum Hoaflhflncfl

may mo mmHDuXHE mo coeummsmnflucmo ucmflpmum wuflmcmp mmouosm .wl.ousmflm

221

Jr... PI. JrIurEB

h2m3<¢0 mmOguam Z. Omaw><~= (<0

ed

-

m

{N . N.— O
. .

0— .02 was»

ON mN

 

'11.—
n
.

 

I-Nu—«qq-u—NN
...m \

~N—-q.—.u.d o

l a o I
......" \

..o o... . 61......l 32 Sand ..
/. \. o.(.........\ .6

_II M .«u II —.o
6 ..

.l O ......Ilo<z.l I.

I m... o 0.x 2.. IN.

- M ... ..

I ...u ... 3 ..q. m a... I

 

bh-_--—bb-—--—--—-~.

 

 

JDNVGUOSGV

”w 059

222

an interaction between the inhibitor protein and the GAPD.
Consequently, several other techniques were used to test
for an interaction between the two proteins.

FLUOROMETRIC TITRATION OF (GAPD-NADH) WITH THE INHIBITOR

 

PROTEIN. The fluorescence of NADH bound to GAPD is sensi-
tive to the environment provided by the enzyme and hence r‘“
to a number of factors which effect the conformation of

the enzyme (Velick, 1959). Thus it might provide a useful
probe to test for interactions between the inhibitor pro-

tein and GAPD. The fluorescence of bound NADH is especially

 

11—u—'_“ V? .".. 5
!
|

useful since the characteristic absorption and emission
bands of the nucleotide are far removed from any of the
absorption bands of the proteins.

The effect of GAPD alone on NADH fluorescence is
depicted by the open circles in Figure 45. As aliquots
of a stock solution of GAPD are added to a solution of

10'5

M NADH, there is an enhancement of the fluorescence
intensity which levels off to a plateau value. Extrapolation
of the two ends of this curve (Figure 45) yields an equiva—
lence point which corresponds to 3.5 moles of NADH bound
per mole of GAPD.

After the fluorescence reached the plateau level,
aliquots of the inhibitor protein were added directly to
the same cuvette (containing NADH and GAPD) to determine

the effect of the inhibitor protein on the fluorescence

of the GAPD-NADH complex. As illustrated in Figure 45

 

223

 

 

 

.maflmwop nonhusm How mponumz paw pxmu mom
.mmHOGMNHu 636 mg poumuumSHHfl we macaw

$042 mo cosmomouooam on“ so :Hmwoum Howwhwscfl or» NO pommmm one
.Ammaouflo
pomoHov ommcmmompmgmp oumnmwonmlmlophnmpamumomam was 3042 Mo mmsuxflfi why 06
wauoouflp poops corp mm3 aflououm Houflnflscw “Ho>ma smoumam m pmsommn coflmmflsm
cocoomouooam may Hausa Ammaouwo cmmov mmmcmmoupmsop mumnmmosmumlophsmp
lawnmowam sues pmpmuuflu m63 mo<z .mmmcmmouphsmp mumsmmosmumImpwnmpamumomam

on wagon maaz no moamommuosau was so nonannnaa man no humane .me mg: as

224

 

omoo< H.085 6.12. 02
VN. 3. N—. 00.

_ F 83¢ 9:6 62

 

 

q ‘ . . u 1

4 -

 

{ ‘

20¢... O...
#0: .222. on(

A§ZIOQ(OV O.— .
...Oum .212. OO<

l. O

m

Olull.

..o. u H.222“—

'

2

:32 o» \1

05(0 00( O

 

 

 

O 0.0 ‘0 ‘ NO 0

a.
on
All SNilNl 3 ON 33538001!

 

 

In
N

8

2

225

(closed circles) the addition of the inhibitor protein

did not produce any observable changes in the fluorescence
of the bound nucleotide. The lower curve in Figure 45
(triangles) illustrates that the inhibitor protein alone
does not alter the fluorescence of NADH. Thus, in the
region of the GAPD molecule where the NADH is bound, the
inhibitor protein did not sufficiently alter the struc-
ture of GAPD to produce any observable change in
fluorescence.

IMMUNOCHEMICAL TITRATION OF A GAPD-INHIBITOR PROTEIN

 

MIXTURE WITH ANTI-INHIBITOR PROTEIN. Another method to

 

test for possible complex formation between the inhibitor
protein and GAPD is to determine if antibodies to the
inhibitor protein could precipitate GAPD activity when
added to a mixture of GAPD and the inhibitor protein. If
the inhibitor protein and GAPD formed a complex as illus-

trated below
GAPD + INHIBITOR PROTEIN e——-(GAPD)-(INHIBITOR_PROTEIN)

antibodies prepared to the inhibitor protein might be able
to precipitate the complex as well as the free inhibitor
protein.

When added to a mixture of the inhibitor protein
and GAPD, antibodies to the inhibitor protein did not
precipitate GAPD activity. This is illustrated in Figure

46. GAPD and the inhibitor protein were mixed in a 1:1

 

226

 

 

.mflemump Morpnsm now mpospoz paw axon mom

.HE o.H mm3 oEsHo> HMHpHcH .Hosmspmonmmouoalm.m mo.o

rues .o.m mm .mflne.m m.o ca meHowumm mpcmeflummxm .cflmuoum HODflQflscfl
on» on mmflponflucm spflz ommcmmonpmnmp momsmmonmumImpwsmpamumomam paw

cflmuoum Hoganflcce may mo oughxfle 6 mo coHumnuflp HmowsosoocsEEH .lw musmflm

 

227

 

639.. 2.20.... ..oze_:z....=z< .0 .s.

 

 

 

.1 0.. 6.0 «.0 00
2.65.... «0:332. 02 «.0
...
920 02 ....0 ...-.......o ............ O—
/ ......3........o..m
on... ......6......
O.Q IQ ............. 1 ON
6....
...... 920 0.2 m .0 1 On
.....a . / .
........ |.I.|.|III|II..|_\‘
........° . . Ill . \. l 9
.......II.0..I 0
.. Om
nun—<0 bm<w> “.0 203—a .12.
00

 

 

(.-0w-.-NIw-sa1own) mnv 'DidS

228

molar ratio and the GAPD activity was measured (lower curve).

An aliquot of Ab.

inhibitor protein was added: the mixture

briefly centrifuged and the GAPD activity re-measured. As

the amount of the Ab.

inhibitor protein increased, the GAPD

activity increased to the level of the control, which was

a sample of pure GAPD to which Ab.

inhibitor protein was ‘zr‘

added. As seen in Figure 46 the addition of antibodies
to the inhibitor protein did not affect the activity of

the control sample of GAPD alone. Thus the results of

‘3! 7.-

 

this test for complex formation were also negative. E
IDENTIFICATION OF THE INHIBITOR PROTEIN AS TRIOSEPHOSPHATE

ISOMERASE. At this point there was no acceptable explana-

 

tion for the inhibition of GAPD by the inhibitor protein.
All the tests for complex formation in the previous section
had been either negative or inconclusive, but the kinetic
(inhibition) studies could best be explained by the forma—
tion of an enzyme-inhibitor protein complex. It was
therefore decided to review all the data and attempt to
formulate alternate hypotheses as to the origin of the
observed inhibition.

One possible explanation which had previously been
proposed but apparently ruled out was that the inhibitor

protein was irreversibly removing one of the substrates

 

of the GAPD catalyzed reaction. This possibility had
previously been rejected because the inhibition did not

seem to be overcome by increasing the substrate concentration

229

(p. 182) and because the tgtal absorbance change (due to
the conversion of NAD to NADH) observed at 340 mu in the
course of the GAPD assays was not noticeably affected by
the addition of the inhibitor protein to the assays.
Another possibility, however, was that the inhibitor

protein was reversibly converting one of the substrates of -:+

 

the GAPD reaction to a form that could not be directly
utilized by the GAPD, that is, it might lower the effective

glycera1dehyde-3-phosphate concentration at any given time 5

 

 

but as g1yceraldehyde-3-phosphate was used up, it would be L-
replenished by reversal of the original extraneous glycer-
aldehyde-3-phosphate removal reaction. The most obvious

example of this possibility, was that the inhibitor protein

was an enzyme which could rapidly and reversibly convert

 

one of the GAPD substrates to another chemical species.
When this possibility was considered, the first enzyme
which came to mind was triosephosphate isomerase.

When the inhibitor protein was assayed for triose-
phosphate isomerase activity, it rapidly catalyzed the
conversion of g1yceraldehyde-3-phosphate to dihydroxyacetone
phosphate in an assay coupled with a-glycerophosphate
dehydrogenase. Further examination revealed that the
inhibitor protein was in fact triosephosphate isomerase.
The data which support this conclusion is as follows:

(1) When assayed for triosephosphate activity,

the inhibitor protein had a specific activity

230

of 2000 umoles min.-l mg. protein-1, which
agrees well with the literature value for this
enzyme (Krietsch at 31., 1970).

(2) The Km for Qfglyceraldehyde-3-phosphate of 0.6
mM (Figure 47) agrees with the reported liter-
ature value for yeast triosephosphate isomerase -:“‘
(Krietsch at 31., 1970).

(3) The molecular weight of 58,000 (p. 204) agrees

with the reported values of 52,000 to 60,000

 

F-“‘ ‘r.-.~ IL. '. ‘ '
.
:'
.

for yeast triosephosphate isomerase (Krietsch
£3 31., 1970).

(4) The amino acid composition (p. 201) is in
agreement with the composition reported for
yeast triosephosphate isomerase (Krietsch 33 31,,
1970).

To insure that the inhibition of GAPD was definitely
due to the conversion of glycera1dehyde-3-phosphate to
dihydroxyacetone phosphate, an inhibitor protein preparation
was carried out as previously described (p. 186) and the
fractions obtained at the various steps were assayed for
inhibitor protein activity (as previously described) and
for triosephosphate isomerase activity. This preparation
was carried out up to the point of the electrofocusing
step. The ratio of the specific activities of the inhibitor
protein to triosephosphate isomerase remained constant
through these steps, indicating that the triosephosphate

isomerase activity was the cause of the inhibition.

231

.mpOCnmz CH ponflpommp mm mwMCmmompmnmp mCMCQmOCmoumowamIS CuHB
wmmmm pmadooo m CH pmusmmoE mmB >6fl>fluom .mumsmmonm-mumpmsmpamumomamLm

00m ommnoeomfl macromonmmmoflmw mo EM 036 wo COHDMCHEHmpoQ .lw mnsmwm

 

 

 

232

.14.... ..<0\.
a a

 

 

. 1 H q ‘1

 

‘

 

 

233

MECHANISM OF INHIBITION OF GAPD BY TRIOSEPHOSPHATE ISOMERASE.
Initially, it seemed that the identification of the inhibitor
protein as triosephosphate isomerase provided a simple
explanation for the inhibition of GAPD, but a careful
review of all data revealed that the following observa—
tions could not be satisfactorily explained by the triose-
phosphate isomerase catalyzed conversion of glyceraldehyde-
3—phosphate to dihydroxyacetone phosphate:

(1) increasing the glyceraldehyde-3-phosphate
concentration did not completely overcome the
inhibition; above 1.5 mg glyceraldehyde-3-
phosphate the GAPD activity showed a plateau
with increasing substrate (p 182);

(2) the inhibition of GAPD in the presence of
triosephosphate isomerase was reversible by
dilution (p. 173), e.g., samples diluted from
a stock solution mixture of GAPD and inhibitor
protein showed a decreasing % inhibition as
the total concentration of the mixture was
decreased;

(3) at a definite molar stoichiometry of 1:1
(GAPD:triosephosphate isomerase, see p. 204)

a maximal inhibition was observed; this suggests
an interaction between enzymes rather than a
competition for substrate, since a coincidental

ratio of 1:1 seems highly unlikely in View of

 

234

the fact that the specific activity of the

triosephosphate isomerase is 10 to 20 times

that of GAPD.
Therefore, further experimentation was undertaken to
determine if the observed inhibition could be explained
solely by the triosephosphate isomerase catalyzed conversion
of g1ycera1dehyde-3-phosphate to dihydroxyacetone phOSphate.

EFFECT OF TRIOSEPHOSPHATE ISOMERASE ON THE GAPD CATALYZED

 

HYDROLYSIS OF p-NITROPHENYL ACETATE. If the inhibition of

 

GAPD by triosephosphate isomerase resulted only from the
conversion of glyceraldehyde-3-phosphate to dihydroxyacetone
phosphate, then the triosephosphate isomerase should have
no effect on other reactions that GAPD catalyzes. Besides
the oxidation of g1yceraldehyde-3-phosphate, GAPD also
catalyzes the hydrolysis of penitrophenyl acetate (Taylor
gt 31., 1963). The effect of triosephosphate isomerase on
this GAPD-catalyzed reaction was thus investigated.

The results of these eXperiments are presented in
Table Ix, and it is readily seen that the triosephosphate
isomerase has no measurable effect on the hydrolysis of
penitrophenyl acetate by GAPD under several different
conditions. These eXperiments were performed by measuring
the esterase activity of GAPD and triosephosphate isomerase
separately and comparing the sum of these rates to the rate
observed when the two proteins were mixed and assayed in

the same cuvette. Comparison of the rates, which were

235

TABLE IX. HYDROLYSIS OF para-NITROPHENYL ACETATE BY
GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE AND
TRIOSEPHOSPHATE ISOMERASE

 

 

 

SOLVENT (1) (2) (3) (4)
*Rate with Rate with Rate with Rate (1)
GAPDa TPIb GAPD + +
TPI Rate (2)
0.1 M_Tris,
pH 8.0 0.052 0.060 0.104 0.112
0.1 M T is, c
+ 1.25 mM GAP 0.032 0.036 0.068 0.068
0.1 M Tris.
+ 1.25 mM GAP
+ 1.25 mg NAD 0.032 0.036 0.068 0.068
Complete GAPD
Assay Mix,
- cysteined 0.032 0.044 0.075 0.076

 

*
All rates expressed as change in absorbance at 450
mu per minute, after correction for spontaneous hydrolysis
of the penitrophenyl acetate.
aGlyceraldehyde-3-phosphate dehydrogenase.
bTriosephosphate isomerase.
cGlyceraldehyde-3—phosphate.

dSee Methods for exact composition of GAPD assay mix.

See Methods for further details.

236

corrected for spontaneous hydrolysis of the pfnitrophenyl
acetate, indicates no effect of triosephosphate isomerase
on GAPD under these conditions as judged by measuring the
esterase activity.

Surprisingly, these experiments revealed that the
triosephosphate isomerase alone catalyzed the hydrolysis

of the ester model substrate, pfnitrophenyl acetate. This

 

seems to be an active site catalysis since the triosephos-

phate isomerase substrate, g1yceraldehyde-3-phosphate,

 

inhibited the hydrolysis. Also, the catalysis by triose-
phosphate isomerase is comparable to the catalysis by

GAPD, which is considered to be quire active as an esterase
(Tayler §t_al., 1963). For example, yeast GAPD catalyzes
the hydrolysis of pfnitrophenyl acetate at a rate 1.5

times greater than chymotrypsin when the rates of
hydrolysis are expressed as umoles substrate min:1 mg.
protein"1 (Taylor gt al., 1963).

As noted in Table IX one set of conditions used
to measure the esterase activity was the complete GAPD
assay mix (see Methods) minus cysteine (normally present
at a concentration of 0.05 M.in the GAPD assay mix).
Cysteine had to be omitted from these experiments because
this amino acid alone catalyzed the hydrolysis of the
pfnitrophenyl acetate at a rate several times that of
either of the proteins.

EFFECT OF GLYCERALDEHYDE-3-PHOSPHATE CONCENTRATION ON THE

INHIBITION OF GAPD BY TRIOSEPHOSPHATE ISOMERASE. After

237

the inhibitor protein was identified as triosephosphate
isomerase, it became apparent that some of the previous
tests for the effects of triosephosphate isomerase on

GAPD could not be properly interpreted since the effective

 

concentration of the substrate g1yceraldehyde-3-phosphate

in the presence and absence of the triosephosphate isomerase
was drastically altered. It was thus not possible to
strictly interpret some earlier experiments to decide if

the triosephosphate isomerase had any effect p§£.§g on

the GAPD. In order to do this it is necessary to assay

the GAPD, with and without triosephosphate isomerase
present, in such a way that the concentration of glycera1de-
hyde-3-phosphate remains constant in both cases.

In order to determine what effect, if any the triose-
phosphate isomerase had on the GAPD reaction other than
lowering the effect g1yceraldehyde-3-phosphate concentration,
the following eXperiment was performed. A ggggg solution

of glyceraldehyde-3-phosphate (10"2 M) was diluted 1:30 with

 

water, and a 0.04 ml aliquot of this diluted solution was

added to a cuvette containing the complete GAPD assay mix. M2.
triosephosphate isomerase was added to this cuvette. The
cuvette was then allowed to stand at 23° for several minutes,
after which an aliquot of GAPD was added to initiate the reac-
tion. A second cuvette was prepared in a similar manner except
that a 0.04 ml aliquot of the sgggg.glyceraldehyde-B-phos-

2

phate solution (10- M) was used instead of an aliquot

of glyceraldehyde-3-phosphate diluted 1:30. After the

-...E'fiu‘ ~. if
I

“.1-

 

238

addition of glyceraldehyde-3—phosphate to the second cuvette
a large excess of triosephosphate isomerase (relative to

the GAPD) was added and the cuvette allowed to stand at

23° for several minutes before initiating the reaction

with GAPD. In the second cuvette an equilibrium was
established between g1yceraldehyde-3-phosphate and dihy-

droxyacetone phosphate before the addition of GAPD. The

W ...... 1

equilibrium of the triosephosphate isomerase reaction is

such that the dihydroxyacetone phosphate: glyceraldehyde-

 

:1f
1

3-ph03phate ratio is 30:1 at equilibrium (Trentham et a1},
1969). Therefore, both cuvettes contained the same con-
centration of glyceraldehyde-3—phosphate, and the only
difference was the presence of triosephosphate isomerase
in the second cuvette. Similar experiments were performed
with two series of cuvettes; comparable cuvettes in each
series contained identical glyceraldehyde-3-phosphate
concentrations and differed only in the presence or
absence of triosephosphate isomerase.

The completely unexpected result of this eXperiment

was that the triosephosphate isomerase seemed to activate

 

the GAPD rather than inhibit (Figure 48). The magnitude
of the activation depended on the glyceraldehyde-3-ph03phate
concentration and varied from 3 to 7-fold under the condi-

tions of the eXperiment.

 

239

.maflmwop nmnunsm How 6x06 00m .000H060mfl mumsmmonmmmoflnu poops

usonpe3 0C6 CuHB 050m 0C6 mm: CoflumupCooCoo opMCQmOCmIMImprmpamuoomHm
0C6 60C» CoCCmE m Conn Ce poEuomumm 0H03 muC0§HH0Qx0 0C9 .0m0H060mfl
0pmzmmonm0moeuu ma 0m0Comonpmsmp ouMCQmOCmum-mpwgmcamuoowam mo CoeuflhHCCH

0Cp Co CoflumupCooCoo oumnmmonmlm-mpNCmpHmnmoham mo poommm .MI.0H5mHh

 

240

”r I...

 

8 x as T3.

 

 

mu; m.— mN.— 0.. ms.

fl W

...h 02
0?- 03:20 m<0 O

r: .
... .6..
0.005 .20 .0...

..6

‘0

0.20 bm<m> “.0 2035.12.

 

‘ 2

 

 

In
F

0
N

In
N

('-9W'l-NIW- SilOWfl) 'AllDV 'DidS

241

DISCUSSION

 

The GAPD inhibitor protein discussed in this chapter
of the thesis has clearly been shown to be tirosephosphate
isomerase. The evidence for the identification of the
inhibitor protein as triosephosphate isomerase is that the
values of Km' specific activity, molecular weight and amino {flu
acid composition of the protein isolated in this work (p.
228) are very close to the literature values of these

quantities for yeast triosephosphate isomerase (Krietsch

 

et_al., 1970). E

It has also been clearly shown that the inhibition
of yeast GAPD is due to the trisoephosphate isomerase pg£_§g,
since the ratio of the specific activity of the triosephos-
phate isomerase to the Specific activity of the inhibitor
protein remains constant during purification of the inhibitor
protein from yeast extracts (p. 231).

The only question which remains is "What is the
precise mechanism of the inhibition of GAPD by triosephos-
phate isomerase?" In order to answer this question it is
first necessary to consider in detail the rather complex
effect of triosephosphate isomerase on the g1yceraldehyde-
3-ph05phate concentration in the GAPD assay.

EFFECT OF GLYCERALDEHYDE-3-PHOSPHATE CONCENTRATION ON THE

 

INHIBITION OF GAPD BY TRIOSEPHOSPHATE ISOMERASE. The

 

problem of the inhibition of GAPD by triosephosphate

isomerase is complicated by the chemical state of the

242

substrate glyceraldehyde-3-phosphate, which exists in
solution in two forms, the aldehyde and the hydrated diol

(Trentham at al., 1969). ‘This is illustrated below.

CHO CH(OH)2
;
H-C-OH

v

 

H-C-OH f“

H-C-O- <:> -H20 H-C-O- (:> p

I L

H H .
ALDEHYDE DIOL i

 

In aqueous solutions the interconversion of the two forms
is very rapid; the first order rate constants for diol
formation and for diol breakdown are 2.5 sec."1 and 8.7 x
10"2 sec.“1 respectively (Trenthan gg‘al., 1969).

In aqueous solution (pH 7.3-8.6, 20°) the equilibrium
ratio of DIOL:ALDEHYDE is 29:1 (Trentham et_al., 1969).
Furthermore, the aldehyde is the only form utilized by GAPD,
triosephosphate isomerase and aldolase; also, the product
of the aldolase and triosephosphate isomerase reactions is
the aldehyde (Trentham eg‘al., 1969).

In the presence of triosephosphate isomerase alone,

there are thus three forms of triosephosphate in solution:

 

 

triosephosphate
GAPaldehyde 4 D
T L isomerase
GA

Pdiol

243

In the presence of triosephoSphate isomerase, the equilibrium

ratios of GAP :DHAP are 1:29:660.

aldehyde:GAPdiol
Finally, in the presence of both GAPD and triose-
phosphate isomerase there are multiple processes occurring,

and these are illustrated below:

 

 

 

GAPD triosephosphate EA
PHOSPHO- < GAP - A’ DHAP
GLYCERATE +NAD aldehyde ‘isomerase i
+Arsenate 11\
GAPdiol

 

Thus, one of the effects of adding triosephosphate isomerase
to the GAPD assay system must be to lower the concentration

of GAP , which is the substrate for GAPD. Along with

aldehyde

this effect, the concentration of GAPDd 1 must also be

io
lowered.
MECHANISM OF THE INHIBITION OF GAPD BY TRIOSEPHOSPHATE

ISOMERASE. The lowering of the glyceraldehyde-3-phosphate

 

concentration should certainly cause an apparent inhibition
of GAPD. Thus, this accounts for the bulk of the inhibition
of GAPD by triosephOSphate isomerase. Furthermore, a
number of physical tests (p. 208 to p. 225) failed to
demonstrate any physical interaction between GAPD and
triosephOSphate isomerase.

These observations suggested that the only effect
of triosephosphate isomerase was to lower the glyceralde-

hyde-3-phosphate concentration. 0n the other hand there

244

are a number of observations that could not be explained
by the triosephosphate isomerase catalyzed conversion of
glyceraldehyde-3-phosphate to dihydroxyacetone phosphate.
These observations are as follows:
(1) Increasing the glyceraldehyde—3—phosphate con-
centration did not completely overcome the

inhibition; above 1.5 mM glyceraldehyde-3-

r——ra .-n- I'. (F !

phosphate the GAPD activity showed a plateau

with increasing substrate (p. 185).

 

(2) The inhibition of GAPD in the presence of t-.
triosephosphate isomerase was reversible by
dilution (p. 173); e.g., samples diluted from
a stock solution mixture of GAPD and triose-
phosphate isomerase showed a decreasing %
inhibition as the total concentration of the
mixture was decreased.

(3) At a definite molar stoichiometry of 1:1
(GAPD:triosephosphate isomerase, see p. 204)
a maximal inhibition was observed; this sug-
gests an interaction between enzymes rather
than a competition for substrate, since a
coincidental ratio of 1:1 seems highly un-
likely in View of the fact that the specific
activity of the triosephosphate isomerase is

10 to 20 times that of GAPD.

245

(4) When GAPD is assayed, with and without triose-
phosphate isomerase present, in such a manner
that the concentration of glyceraldehyde-3-
phosphate is the same in both cases (p. 239),
the triosephosphate isomerase seems to activate
the GAPD . IF

These observations suggest that the triosephosphate isomerase H
physically interacts with the GAPD. I

Since there is no single satisfactory explanation

 

for all the data, further experiments are needed to con- L-
clusively prove or disprove that the triosephOSphate isomerase

physically interacts with the GAPD.

HYDROLYSIS OF p-NITROPHENYL ACETATE BY TRIOSEPHOSPHATE

ISOMERASE. The observation in this work that triosephosphate

 

isomerase has considerable esterase activity with the model
substrate penitrophenyl acetate (p. 235) is itself inter-
esting, and was rather unexpected. The specific esterase
activity (umoles min."l mg. protein-l) of the triosephos-
phate isomerase is the same as that of GAPD, which is
considered to be quite active as an esterase with pf
nitrophenyl acetate as a substrate (Taylor gt a;., 1963).
For example, yeast GAPD is 1.5 times as active as chymo-
trypsin (Taylor gt al., 1963).

The esterase activity of the triosephosphate
isomerase also seems to be a Specific effect, since the
substrate glyceraldehyde-3—phosphate inhibits the hydrolysis

of the penitrophenyl acetate.. Glyceraldehyde-3-phosphate

246

also inhibits the esterase activity of GAPD (Taylor EE.E£°'
1963). It thus seems likely that a study of the esterase
activity of the triosephOSphate isomerase might provide
valuable information about the active site of this enzyme,
just as studies of the esterase activity of GAPD have
provided valuable information (Taylor gt 31,, 1963). FT-
At present no biochemical significance can be
assigned to this unusual catalytic function of triose-

phosphate isomerase.

 

METHODS

ASSAY FOR INHIBITION OF GLYCERALDEHYDE‘3-PHOSPHATE DEHY-

 

DROGENASE. Inhibition was measured by assaying an aliquot

 

of a solution of GAPD (usually 0.2-0.4 mg/ml in a volume

of 0.5-1.0 ml) to determine the 100% activity value,

adding a small amount of inhibitor protein (25-125 ug) to
the GAPD solution, and then assaying an aliquot of the
mixture for GAPD activity. Assays were performed as
previously described (p. 95). The solvent in which the

two proteins were mixed was 0.1 M Tris, pH 8.0, containing
0.1 M B-mercaptoethanol. The inhibitor protein was added
and mixed gently at room temperature. Identical values

for the % inhibition were obtained if the GAPD and inhibitor
protein were added separately to the GAPD assay cuvette,
instead of pre-mixing. Since pre-mixing of the two proteins
was more convenient, this technique was routinely used to

measure the inhibition.

247

REAGENTS. The pfnitrophenyl acetate used for the esterase

 

assays was obtained from Sigma (St. Louis, Mo.), and the
ampholytes used for the isoelectric focusing were obtained
from LKB (LKB Produkter AB, Fack, s-l6l 25, Bromma 1,
Sweden). The reagents used in the GAPD assays have previ-
ously been described (p. 95). All other reagents were
the best commercial grade available.

ISOELECTRIC FOCUSING. Isoelectric focusing was performed

 

with either an LKB 110 m1 apparatus or an LKB 440 ml
apparatus and a Savant high voltage power supply unit.

The anode solution was prepared by mixing 0.8 ml
of sulfuric acid, 56.0 ml of water and 48 g. of sucrose.
The cathode solution was 0.8 m1 of ethylendiamine and
40.0 ml of water.

For the narrow pH gradient (4-6), the ampholyte-
sucrose column (440 ml) was prepared by layering a series
of 46 fractions (9 ml each) of decreasing sucrose density
on 75 ml of anode solution. Fractions were layered with
tubing attached to a 10 ml funnel. The fractions contained
various proportions of a SEEEE sucrose-ampholyte solution
and a ligME aqueous ampholyte solution. The dense solu-
tion (40% sucrose) contained 7.6 ml of ampholytes, 142 ml
of water and 100 g. of sucrose. The light solution
contained 2.4 ml of ampholytes and 213 ml of water.
Starting with pure dense solution (lst 9 ml fraction)
the 9 ml fractions were prepared by successively decreasing

the amount of dense solution by 0.2 ml and increasing the

 

248

amount of light solution by 0.2 ml. Thus, the second 9 m1
fraction contained 8.8 m1 of dense solution and 0.2 ml
of the light solution. The 46th fraction was pure light
solution. Cathode solution was then added to a level 1 cm
above the platinum electrode. The procedure was analogous
for the broad pH range (3-10), except that 2.25 ml frac-
tions were used with the 110 ml column.

The protein, 25 mg (in 5 ml) for the 110 ml column

and 100 mg (in 20 ml) for the 440 m1 column, was applied

 

by adding small aliquots (each aliquot was 1 ml) of the
protein samples to the 20 ampholyte-sucrose fractions
used to form the middle of the column.

A potential difference of 600 v (110 ml column)
or 1000 v (440 ml column) was used. The temperature was
maintained at 2° with a circulating water bath. After
focusing for 72 hrs., the fractions were collected with a
Gilson fraction collector.

ANTIBODY PREPARATION AND IMMUNODIFFUSION TESTS. For the
primary immunization, GAPD and the inhibitor protein were

dialyzed overnite against 0.15 M NaCl and 10-4

M EDTA at
23°. Emulsions were prepared by mixing equal volumes of
dialyzed protein and Freund's adjuvant. Male-Wistar
rabbits were then given 3 ml of the emulsion (containing

7 mg of protein) in six injections (2 foot pads, 2 sides
of the neck, and 2 sides of the torso anterior to the hind

leg), i.e., 0.5 ml per injection site.

249

Three weeks after the primary immunization the
rabbits were given a booster (IV) in the ear vein. The
booster was 2 mg of protein (previously dialyzed as noted
above) in a volume of 0.5 m1.

Eight days after the booster, 15—20 ml of blood
was collected by cardiac puncture. The blood was left er:
overnite in the cold, the clot was removed by centrifuga- ;
tion and the serum was collected. 2

For the preparation of the double diffusion plates,

one gram of agar was mixed with 25 m1 of buffer (see below)

 

and 75 ml of water. The mixture was warmed until the agar
completely dissolved, and 1 m1 of 1% merthiolate was
added. The agar was stored in the refrigerator until
used.

The buffer used in the agar solutions consisted
of 0.05 M sodium diethylbarbiturate, 0.01 M_diethylbarbi-
turic acid and 0.05 M sodium acetate. The pH of this
buffer solution was 8.6 and the ionic strength was 0.1.

FLUORESCENCE MEASUREMENTS. Fluorescence measurements were

 

made at 20° with an Aminco—Bowman spectrophotofluorometer
equipped with an X—Y Recorder (American Instruments Co.,
Silver Springs, Md.). Fluorescence was measured at right
angles to the illumination. The NADH was excited at a
wavelength of 340 mu and the fluorescence emission was
measured at a wavelength of 465 mu. The temperature was

thermostat controlled.

250

Titrations of fluorescence change were begun with
a 1.0 ml sample of NADH in 0.1 M Tris, pH 8.0, which
contained 0.1 M B-mercaptoethanol, in a fused quartz cell.
Small aliquots of the enzyme titrant (GAPD or the inhibitor
protein) in the same solvent as the NADH were added and
mixed with a magnetic stirrer. The fluorescence intensity rt“
after each addition was then read and corrected for
dilution.

The initial concentration of NADH in these eXperi-

ments was either 1 x 10.5 M_or 2.5 x 10-5 M. E

 

AMINO ACID ANALYSIS. The protein samples (0.2 ml of 1.0

 

mg/ml) were de-salted by passage through a 1.6 x 12 cm
column packed with Sephadex G-25, which had been equilibrated
with distilled de—ionized water. The samples were 1yophil-
lized in hydrolysis tubes (200 ug/tube), 6 M HCl was added
and the tubes were placed in an acetone/dry ice bath. The
tubes were then evacuated and sealed. After hydrolysis at
110° for 24 or 48 hrs., 25 pg samples of protein were
analyzed.

The analysis was performed on an amino acid
analyzer constructed by Dr. D. Robertson and Dr. W. A.
Wood of this department.

ASSAY FOR TRIOSEPHOSPHATE ISOMERASE ACTIVITY.A Triosephos-

 

phate isomerase activity was routinely measured in a
coupled assay using a-glycerOphosphate dehydrogenase. The

assay contained the following components (the final

251

concentration of each component in the assay is given in
parenthesis) which are listed in the order of addition
to the assay cuvettes: Tris buffer (0.1 M), pH 8.0; rabbit
muscle a-glycerophosphate dehydrogenase (25 ug/ml); NADH
(1 mM); triosephosphate isomerase (0.1 -l.0 ug/ml);
glyceraldehyde-3—ph05phate (1 mM). The total volume was
0.4 ml.

The decrease in absorbance at 340 mu was followed

with a Gilford Model 2000 Spectrophotometer equipped with

 

temperature control and multiple sample absorbance recorder.
Assays were performed at 25°.

ASSAYS FOR ESTERASE ACTIVITY USING p-NITROPHENYL ACETATE.
The esterase activity of GAPD and triosephosphate isomerase
was measured with the model substrate pfnitrophenyl acetate
as described by Taylor EE.§l-r (1963). The assay contained
the following components (final concentrations in the assay
are given in parenthesis) which are listed in the order of
addition to the assay cuvettes: Tris buffer (0.1 M), pH
8.0; GAPD or triosephosphate isomerase (0.1-0.5 mg);

pfnitrophenyl acetate (2.17 x 10-3

M). Total volume was
0.4 ml.

Since the penitrophenyl acetate is unstable in
aqueous solutions, a stock solution (6.3 mg/ml, 3.5 x

10’2

M) was prepared in absolute methanol and stored at
0° until used. Stock solutions were never used more than

3 days after they were prepared.

252

The reaction was initiated with penitrophenyl
acetate and the increase in absorbance at 400 mu (due to
the release of pfnitrophenol) was followed with a Gilford
Model 2000 Spectrophotometer equipped with temperature
control and a multiple sample absorbance recorder. Assays
were performed at 25°. Blank rates, obtained from cuvettes
minus protein samples, were used to correct all reported
rates for spontaneous hydrolysis of the pfnitrophenyl
acetate.

Assays could not be performed in the presence of
cysteine, because of the high blank rates observed (without
the addition of either protein) when cysteine was present
in the assay mixture. This high blank rate apparently
arose from the catalysis of the hydrolysis of the p:
nitrophenyl acetate by the cysteine itself.

TURBIDIMETRIC MEASUREMENT OF PROTEIN CONCENTRATION. The
procedure used to locate the protein peaks on the sucrose
density gradients (Figures 43 and 44) was exactly as

described by Mejbaum-Katzenellenbogen and Dobryszycka

(1959).

 

CHAPTER SIX

KEY FINDINGS OF THE RESEARCH AND SUGGESTIONS

FOR FUTURE WORK

253

 

254

DISSOCIATION OF YEAST GAPD IN ATP
(1) Tetrameric yeast GAPD (7.5 S) dissociates into
subunits at 0° in the presence of ATP. Ultracentrifugal
analysis reveals that the first observable product of the
dissociation is a 3.0 S subunit (incubation time less than

12 hrs). This is the eXpected sedimentation coefficient

 

fit“
for a globular or folded subunit of molecular weight 36,000

(Holleman, 1966). After longer incubation times (24 hrs)

a 1.6 S subunit is observed. This is the expected sedimenta- '
tion coefficient for an extensively unfolded subunit, since E~

 

GAPD subunits in 8 M_urea have a sedimentation coefficient
of 1.5 S (Deal and Holleman, 1964). After a very long
incubation time (3-7 days) at 0° in ATP, a 0.9 S particle
is observed. A sedimentation coefficient of 0.9 S cor-
responds to a particle approximately one-half the molecular
weight of the GAPD subunit of 36,000.

The yeast GAPD-ATP dissociation is rather unique
since selective treatment can be used to obtain a folded

or unfolded subunit. This is in contrast to most dissocia-

 

tion systems which also produce extensive unfolding of the
individual polypeptide chains. This system is thus ideally
suited to answer a number of questions concerning subunit
interactions in polymeric enzymes.

Several questions raised by the observation of
the 0.9 S species are "What is the exact nature of the

0.9 S species?", "Does the production of the 0.9 8 species

255

involve the rupture of peptide bonds?" and "Is the produc-
tion of the 0.9 S species a spontaneous process?"

The first question, pertaining to the nature of
the 0.9 S species, can be answered by careful molecular
weight measurements. The technique best suited for this
measurement is probably electrophoresis in sodium dodecyl
sulfate (Weber and Osborn, 1969). Conventional sedimenta-
tion equilibrium studies would check the molecular weight
of the 0.9 S Species obtained by this technique.

The second question, whether or not peptide bonds
are ruptured in the formation of the 0.9 S species, can
also be answered by available experimental techniques.
This can be done by incubating GAPD at 0° in ATP solutions
prepared with Ola-enriched water, and then determining

whether 018

is incorporated into the terminal carboxyl
groups of the 0.9 S species.

Finally, experiments should be performed to determine
whether the production of the 0.9 S species is a spontaneous
process requiring only the incubation of GAPD with ATP at
0°, or whether any other protein, such as a protease
contaminant in the GAPD preparation, is required. This
question can be answered by observing the rate of appearance
of the 0.9 S species from native GAPD samples pre-treated
in several ways. The first sample would be GAPD prepared

as described in Methods, Chapter Three. The second sample

would be GAPD prepared as above and then passed over a

 

1“ “l’. ‘5'. 'i C
I

256

Sephadex G—200 column. The third sample would be GAPD
prepared as above and then "further purified" by isoelectric
focusing. Production of the 0.9 S species at the same rate
from all the GAPD samples so pre-treated would be strong
evidence that the 0.9 8 species arose solely from the
incubation of native GAPS with ATP at 0°.

(2) A systematic study of the variables affecting
the dissociation of GAPD at 0° in ATP reveals that electro-
static and hydrophobic interactions are both important in
the dissociation process. Thus dissociation involves:

(a) conformational changes in the individual subunit poly-
peptide chains produced as a result of electrostatic
repulsions between the negatively charged phosphate por-
tion of ATP and the negatively charged groups on the
protein when ATP is bound. (b) This conformational change
alters the subunit contact sites so that hydrophobic
interactions assume a more critical role in maintaining
the native tetrameric structure. (c) Lowering the temper-
ature to 0° then weakens hydrophobic interactions between
subunits enough to cause dissociation.

This interpretation is also consistent with the
results of studies of the assembly of yeast GAPD subunits
produced in 8 M_urea (Deal, 1969), with the results of
studies of the binding of adenine nucleotides (Yang and
Deal, l969a;b) and with the results of the kinetic analysis

of dissociation of yeast GAPD at 0° in ATP. Therefore, the

.J‘N". .-Np'Tl
|

 

257

nature of the forces involved in the dissociation process
seems well established.

(3) A kinetic analysis of the dissociation reveals
that the overall process is bi-phasic, involving a fa§t_
initial step followed by a slgg_second step.

The EEEE step can be described by a rapid reversible
equilibrium between tetramers and dimers. The kinetic data
can be treated to obtain an equilibrium constant for this
initial process. The calculated equilibrium constant

(K = 10.2 x 10"8

eq
a value of 8.9 kcal/mole for AG° at 0°.

M) for this initial dissociation yields

 

a“ ".' ‘1'.-

The §igw step can be described by the dissociation
of dimers to monomers, which is essentially irreversible
at 0°. The average constant for binding of ATP to the
dimer is 0.75 mM, and there is no measureable interaction
between ATP binding sites.

The methods devised in this thesis for obtaining
Keq (at 0°) in the presence of ATP for reversible dissocia-
tion of tetramers to dimers can now be used to obtain a
complete thermodynamic description of this initial dissocia-
tion as follows: (a) Keq can be measured at different

temperatures between 0° and 23°, and AG° as a function of

temperature may then be calculated from

AG° = -RT 1n K
eq

258

(b) The slope of a plot of 1n Keq vs 1/T at any temperature

will then yield - AH°/R at that temperature, since

_ 0
d 1n (Keq) AH

d (l/T) R

 

 

(c) The values of AH° and AG° at any temperature can then

be used to obtain AS° from

(AG° - AH°)
-AS°=

 

T

If a method can be devised to measure the constant

for binding of ATP to the tetramer of GAPD, the first Law

 

of Thermodynamics can be used to obtain values for the
thermodynamic parameters of the tetramer-dimer dissociation
in the absence of ATP. This possibility is discussed in
Chapter Four of this thesis.

A kinetic analysis of the dissociation process at
different temperatures can also be used to obtain a rate
constant, k, for dimer dissociation at different tempera-
tures. The activation energy, Ea' for this process can

then be calculated from the Arrhenius equation

where A is the constant obtained upon integration of the

differential form of the equation.

 

259

REVERSAL OF DISSOCIATION
(4) Globular subunits (3.0 S), produced by incuba-
tion of native GAPD with ATP at 0°, can be reassembled to
yield native 7.5 S tetramers by raising the temperature
(to 15° or above) in the presence of ATP, 10% sucrose and

reducing agent. This discovery is especially exciting

...-w- ‘E
L
1

since the reassembly of folded monomers (this work) may

now be compared to reassembly of unfolded subunits produced

 

in 8 M urea (Deal, 1969).

 

The most striking similarity between these two
reassembly systems is the nucleotide requirement for
i2_yitgg_reassembly. Reassembly of folded monomers requires
ATP and reassembly of unfolded monomers requires NAD.
Contrary to widespread views, these results clearly show
that the amino acid sequence alone is 22E adequate to
determine either the folding or association of yeast GAPD
subunits, thus raising the possibility that either folding
or association steps, or both, are subject to $2.!i22
control by metabolite levels (see Deal, 1969).

There are many other similarities between the two
reassembly systems: (a) the pH profiles of reassembly
show no significant dependence upon pH in the range of
pH 6.8. (b) Both systems require reducing agent for
complete activity recovery upon reassembly. (c) The protein
concentration which yields maximum activity recovery for

both reassembly systems is 0.04 mg/ml, suggesting a

260

concentration dependent equilibrium step in the reassembly
process. (d) Both reassembly systems exhibit a very
similar dependence on temperature for activity recovery.
(e) The half-time for activity recovery of unfolded sub-
units is 7.0 min (95 ug/ml, pH 6.9, 0.3 M_KCl, 16°), and
7.5 min for folded subunits (100 ug/ml, pH 7.0, 0.3.M

KCl, 23°, 10% sucrose), (f) The half-times of activity
recovery for both systems increase with decreasing protein
concentration. These results suggest that the mechanism

of assembly is very similar for both reassembly systems.

 

The only two major differences between the dif-

ferent reassembly processes seem to be the previously noted

nucleotide specificity and the effect of salt on reassembly.

KCl (0.3 M) yields a 5-fold increase in activity recovery
upon reassembly of unfolded subunits (Deal, 1969), but
does not affect the total activity recovery upon reassembly
of folded subunits. Thus it is clear that the effect of
KCl on the reassembly of unfolded subunits is on a folding
step and not on an association step (Deal, 1969). Further-
more, since KCl substitutes for NAD in the reassembly of
unfolded subunits, it is also clear that the effect of NAD
upon reassembly of unfolded subunits is an effect on the
folding of the polypeptide chains (Deal, 1969).

Taken together, the studies reported in this thesis
and the studies of Deal (1969) on the assembly of yeast

GAPD subunits to native enzyme provided one of the most

”'5

 

 

261

informative and comprehensive analyses available of the
assembly of a polymeric enzyme from its subpolymers.

(5) A kinetic study of the reassociation of globular
subunits (Chapter Four) reveals that the over—all order of
the reassociation process changes from second order at low
protein concentrations to first order at higher protein
concentrations. This is a significant observation because
it indicates that association of folded subunits is not
simply a sequence of bimolecular associations, but rather
the process may involve a first order isomerization reaction
as the final step in reassembly of the native enzyme. This
study should serve as a valuable model for the association
steps in a complete ig,yi££g synthesis of polymeric enzymes,

a goal which biochemists are rapidly approaching.

 

 

 

CHAPTER SEVEN

APPENDICES AND LIST OF REFERENCES

262

 

263

APPENDIX 1

DETERMINATION OF REACTION ORDER
BY THE HALFvLIFE METHOD

 

For any rate expression of the type
dx/dt = k (a - x)n (1)

the half-life may be defined for all values of a, the order

of the reaction, as

t = f (mm/4’“l (2)

1/2

where £,is some function of the order of the reaction, 2! and
the rate constant, M, and 3.15 defined as the initial con-
centration of reactant (Frost and Pearson, 1961). Expression

(2) may be placed in logarithmic form to yield (3)

log (tl/Z) = log (fEn.kl) - (n-l> log a (3)

Therefore a plot of the half-time of a given process vs the
initial concentration of reactant, a, should yield a straight

line with a slope of (l-n).

 

264
APPENDIX 2

DERIVATION OF THE INTEGRATED RATE EQUATION FOR
A TETRAMER-DIMER EQUILIBRIUM

The tetramer-dimer equilibrium is represented by

the process given below

k
A +___. 2B K = ——————
k eq [A]
24

If we start at time zero with A and 22.3! and let E equal
to the amount of A which is converted to B at any time,
the concentrations of A and B at time zero, time E_and at

equilibrium (te) will be those given below

 

 

 

k42
A: A’B
k24
TIME
t a 0
o o
t (ao-x) (2x)
te ae=(aO-x)e (2x)e

We may then write an eXpression for the rate of change of

an

dx/dt = k (ao-x) - k24 (2x)2

42

.5 .- ..’

 

265

I

At equilibrium, dx/dt 0, and from (1) we may therefore

write

0 = dx/dt = k42 (aO-x)e - k24 (2x)2 (2)

This may then be rearranged to give

k24 = k42 (ae/ (2x):) (3)

Substitution of this value of k24 into (1), yields

dx/dt = k42 (aO-x) - (k ae/ 4x2) (2x)2 (4)

42
We now define a constant M, such that

b = ae/ x: (5)

After substitution of p, (4) reduces to

dx/dt = k (aO - x - bxz) (6)

42
Separation of variables yields

dx/(-bx2 - x + a0) = k dt (7)

42

Integration of this relationship yields

kx-Zaofl - IxDl/ZI
t = log (8)

L_|(x-2ao)| + IxDl/ZLJ

1/2

 

42

 

 

 

266

where Qbis a constant given by

D = 1 + 4aob (9)

Now, M_was originally defined as the amount of A
(tetramer) which is converted to B (dimer) at any given
time. Therefore, since the tetramer is active and the
dimer is inactive, a measure of the quantity M'at any
time is given by the difference in enzymatic activity at

time zero and encymatic activity at any time, t.

x = A - A (10)

If we now let Ao be the measured enzymatic of a solution of
active tetramers (at a fixed protein concentration, 3), and
At the enzymatic activity of a mixture of active tetramers

and inactive dimers at a total concentration equal to 9, we

may re-write (8) in terms of enzymatic activity as

 

 

 

P _ _ _ _ 1/21
1/2 HA0 At 2A0” HA0 At” D
D k t = log (11)
42 RA - A - 2A I + kA - A n 01/2
._ o t o o t M
The term 2 in (11) must also be expressed in terms of
enzymatic activity, and this is given below
4A A
o e
D=1+ (12)

 

(A0 - Ae)2

 

267

where A0 is the activity at time zero and Ae is the

activity at equilibrium for the tetramer-dimer equilibrium.
This value, Ae, is thus seem to be the value of the activity
obtained by extrapolation of the slow process (dimer break-
down) of dissociation in ATP to time zero, since this
extrapolation provides a measure of the activity one would

observe at a given protein concentration for an equilibrium

 

mixture of tetramers and dimers if the dimer did not break-

down further to yield monomers.

 

Expression (11) may now be further rearranged to I

its final usable form

._
V|( -AO-At )I — |( Ao-At )(T)l
(T) k42 t = log

 

 

LR 4.0-2.t )1 + k 2.0-...: )(TI J.

where M_is simply the square root of Q'as given in (12).

268

LIST OF REFERENCES

 

Allison, W. S., and Kaplan, N.O. (1964), 1. Biol. Chem.
239,2140.

Atkinson, D. E. (1965), Science 150,851.

 

Atkinson, D. E. (1966), Ann. Rev. Biochem. 2§,85.

 

Bechet, J., and Wiame, J. M. (1965), Biochem. Biophys.
Res. Commun. 21,226.

Betz, A., and Moore, C. (1967), Arch. Biochem. Biophys.
120,268.

 

 

Blatti, S. P. (1968), Ph. D. Thesis, Michigan State

 

University, East Lansing, Michigan. ij.

Bonsignore, A., Mangiarotti, G., Mangiarotti, M. A.,
DeFlora, A., and Pontremoli, S. (1963), q.
Biol. Chem. 238,3151.

 

Brandts, J. F. (1969), in Structure and Stability of
Eiological Macromolecules, Timasheff, S. N., and
Fasman, G. D. (Eds.)7’Marcel Dekker, Inc., N. Y.

 

 

Constantinides, S. M., and Deal, W. C., Jr. (1967), 154th
Meeting American Chemical Society, Chicago,
Illinois, Abstract C-198.

Constantinides, S. M., and Deal, W. C., Jr. (1969), g,
Biol. Chem. 244,5695.

 

Constantinides, S. M., and Deal, W. C., Jr. (1970), 1.
Biol. Chem. 245,246.

 

Conway, A., and Koshland, D. E. (1968), Biochemistry Z,
4011.

Cook, R. A., and Koshland, D. E. (1970), Biochemistry g,
3337.

Cori, C. F. (1942), in "Symposium on Respiratory Enzymes,"
Univ. of Wisconsin Press, Madison, Wisc., p. 175.

Cori, G. T., Selin, M. W., and Cori, C. F. (1948), 1.
Biol. Chem. 173,605.

269

Dandliker, W. B., and Fox, J. B. (1955), 1. Biol. Chem.
214,275.

Deal, W. C., Jr. (1963), Fed. Proc. 23, 290.

Deal, W. C., Jr. (1967), 7th International Congress of
Biochemistry, Tokyo, Japan, Abstract G-144, p. 872.

Deal, W. C., Jr. (1969), Biochemistry 8, 2795.

 

Deal, W. C., Jr., and Constantinides, S. M. (1967), Fed. FF
Proc. Mg, 348. 1

Deal, W. C., Jr., and Holleman, W. H. (1964), Fed. Proc.
E3, 264.

Deal, W. C., Jr., Rutter, W. J., and Van Holde, K. E.
(1963), Biochemistryg, 246.

 

 

Dua, R. D., and Burris, R. H. (1963), Proc. Nat. Acad. 5""
Sci. 59, 169.

 

Edsall, J. T. (1968), in Rich, A. and Davidson, N., (Eds.):
Structural Chemistry and Molecular Biology, W. H.
Freeman and Co., San Francisco, p. 88.

 

Slias, H. G., Garbee, A., and Lanprecht, W. (1960), g,
Physiol. Chem. 319, 22.

 

Elodi, P. (1958), Acta Physiol. Acad. Sci. Hung, 13, 199.

 

 

Felix, F., and Brouillet, N. (1966), Biochem. Biophys.
Acta 122, 127.

 

 

Fenselau, A. (1970), 2. Biol. Chem. 245, 1239.

 

 

Fincham, J. R. S. (1957), Biochem. g. 65, 721.

Fox, J. B., and Dandliker, W. B. (1956),‘g, Biol.'Chem.
218, 53.

 

Frank, H. S., and Evans, M. W. (1945), J, Chem.‘Phys.'£3,
507.

Frost, A. A., and Pearson, R. G. (1961), Kinetics and
Mechanism, Wiley, N. Y., p. 41.

 

 

Gould, E. S. (1959), Mechanism and Structure £2 Organic
Chemist , Holt, Rhinehart and Winston, 959, p.
185.

270
Graves, D. J., Sealock, R. W., and Wang, J. H. (1965),
Biochemistry 4, 290.

Guthohrlein, G., and Knappe, J. (1968), Eur. g, Biochem.
7, 119.

Hageman, R. H., and Waygood, E. R. (1959), Plant Physiol.
34, 396.

 

Harrington, W. F., and Karr, G. M. (1965), 1. Mol. Biol.
13. 885.

 

Harrington, W. F., and Schachman, H. K. (1956), Arch.
Biochem. Biophys. 65, 278.

Harris, C. E., Kobes, R. D., Teller, D. C., and Rutter,
W. J. (1969), Biochemistry 8, 2442.

 

Harris, J. I., and Perham, R. N. (1965), g. Mol. Biol.
M1, 876.

 

 

 

Hata, T., Hayashi, R., and Doi, E. (1967), Agr. Biol.
Chem. 2;, 150.

 

Hatfield, G. W., and Umbarger, H. E. (1970), 9. Biol.
Chem. 245, 1736.

Havir, E. A., Tamir, H., Ratner, S., and Warner, R. C.
(1965), 1. Biol. Chem. 240, 3079.

 

von Hippel, P. H. (1969), Acc. Chem. Res. 3, 257.

 

 

von Hippel, P. H., and Waugh, D. F. (1955), g, 5M, Chem.
Soc. 11, 4311.

 

Hoagland, V. D., and Teller, D. C. (1969), Biochemistry 8,
594.

 

Hofstee, B. H. J. (1949), g. Gen. Physiol. 2, 339.

 

Hsu, R. Y., and Yun, S. L. (1970), Biochemistry g, 239.

 

Jaenicke, R., Schmid, D., and Knof, S. (1968), Biochemistry
7, 919.

Jarabak, J., Seeds, A. E., Jr., and Talalay, P. (1966),
Biochemistry g, 1269.

Kaplan, N. 0. (1964), Brookhaven SyMp. Biol. M1, 131.

 

Kaplan, N. 0., Ciotti, M. M., and Stolzenback, F. E.
(1956), g,Biol. Chem. 221, 833.

 

 

271

Kauzmann, W. (1954), 13 The Mechanism of Enzyme Action,
McElroy, W. D., and Glass, B., Eds., Johns Hopkins
Press, Baltimore, Md.

Kauzmann, W. (1959), Adv. Prot. Chem. 11, l.

 

Kirkman, H. N., and Hendrickson, E. M. (1962), 1. Biol.
Chem. 237, 2371.

 

Kirschner, K. (1968), in Kalckar, H. M., Klenow, H.,
Munch-Petersen, A., Ottsen, M., and Thaysen, J.
H. (Eds.), in The Role 9£_Nucleotides for the
Function and Conformation gngnzyMes, Academic
Press, N. Y., p. 158.

 

 

Kirschner, K., Eigen, M., Bittman, R., and Vioght, B.
(1966), Proc. Natl. Acad. Sci. (U.S.) 55, 1661.

 

 

Kirtley, M., and Koshland, D. E. (1970), 1. Biol. Chem. 245,
276.

Klotz, I. M., Langerman, N. R., and Darnall, D. W. (1970),
Ann. Rev. Biochem. 12, 25. .

Koshland, D. E., and Neet, K. E. (1968), Ann. Rev. Biochem.
37, 359.

Krebs, E. G. (1955), Meth. 1M Enzymol. 1, 407.

Krebs, E. G., Rafter, G. W., and Junge, J. M. (1953), 1.
Biol. Chem. 200, 479.

 

Krietsch, W. K. G., Pentchev, P. G., Klingenburg, H.,
Hofstatter, T., and Bucher, T. (1970), Eur. 1.
Biochem. 13, 289.

Krimsky, I., and Racker, E. (1955), Science 122, 319.

Kuczenski, R. T. (1970), Ph. D. Thesis, Michigan State
University, East Lansing, Michigan.

 

Kuczenski, R. T., and Suelter, C. H. (1970), Biochemistry
'2, 2043.

Langerman, N. R., and Klotz, I. M. (1969), Biochemistry 8,
4746.

 

Lauffer, M. A. (1962), in The Molecular Basis g£_Neoplasia,
Univ. of Texas Press, p. 180.

 

 

Lee, M., and Muench, K. H. (1969), 1. Biol. Chem. 2 4, 223.

 

 

272

Lenny, J. F., and Dalbec, J. M. (1967), Arch. Biochem.
Biophys. 120, 42.

Mansour, T. E., and Mansour, J. M. (1962), 1, Biol. Chem.
237, 629.

 

Martin, B. N., and Ames, R. G. (1961), 1, Biol. Chem.,
236, 1372.

Meighen, E. A., and Schachman, H. K. (1970), Biochemistry
9, 1177.

 

Mejbaum-Katzenellenbogen, W., and Dobryszycka, W. (1959),
Clin. Chim. Acta 1, 515.

. Mikola, J., and Suolinna, E. M. (1969), Eur. 1, Biochem.
9, 555.

Monod, J., Changeaux, J. P., and Jacob, F. (1963), g, ;
Mol. Biol. 6, 306. 4

 

 

Muirhead, H., and Perutz, M. F. (1963), Nature 199, 633.

 

Park, J. H. (1966) in Kennedy, B., (Ed.), Current Aspects
1E_Biochemical Engergetics, Academic Press, N. Y.,
1966.

 

Passonneau, J. V., and Lowry, 0. H. (1962), Biochem.
Biophys. Res. Commun. 1, 10.

Penefsky, H. S., and Warner, R. C. (1965), 1, Biol. Chem.
240, 4694.

Perutz, M. F., Muirhead, H., Cox, J. M., Goaman, L. C. G.,
Mathews, F. S., McGandy, E. L., and Webb, L. E.
(1968a), Nature 219, 29.

 

Perutz, M. F., Muirhead, H., Cox, J. M., and Goaman, L.
C. G. (1968b), Nature 219, 313.

Perham, R. N., and Harris, J. I. (1963), J, Mol. Biol. 1,
316.

Pullman, M. B., Penefsky, H. S., Datta, A., and Racker, E.
(1960), 1, Biol. Chem. 235, 3322.

Racker, E. (1965), Mechanisms 1g Bioenergetics, Academic
Press, N. Y., Ciio I-go

 

Racker, E., Pullman, M. E., Penefsky, H. S., and Silverman,
M. (1963), Proc. 5th Inter. Congr. Biochem. 1,
MacMillan, N. Y.

273

Rafter, G. W., and Colowick, S. P. (1957), Arch. Biochem.
Biophys. 88, 190.

Raijman, L., and Grisolia, S. (1961), Biochem. Biophys.
Res. Commun. 1, 262.

Rosen, O. M., Copeland, P. L., and Rosen, S. M. (1967),
1. Biol. Chem. 242, 2760.

 

 

Scheraga, H. A., Nemethy, G., and Steinberg, I. Z. (1962),
1, Biol. Chem. 237, 2506.

 

 

Schulze, I. T., Lusty, C. J., and Ratner, S. (1970), 1,
Biol. Chem. 245, 4534.

 

 

Scrutton, M. C., and Utter, M. F. (1965), 1, Biol. Chem.
240, 1.

 

Scrutton, M. C., and Utter, M. F. (1968), Ann. Rev. Biochem.
37, 249.

Shukuya, R., and Schwert, G. W. (1960), 1, Biol. Chem.
235, 1658.

Shuster, C. W., and Doudoroff, M. (1962), 1, Biol. Chem.
237, 603.

Stadtman, E. R. (1966), Adv. 1E’Enzymol. 88, 41.

Stancel, G. M. (1969), Fed. Proc. 28, 490.

Stancel, G. M., and Deal, W. C., Jr. (1968), Biochem.
Biophys. Res. Commun. 81, 398.

 

Stancel, G. M., and Deal, W. C., Jr. (1969), Biochemistry
8, 4005.

 

Steinman, C. R., and Jakoby, W. B. (1967), 1, Biol. Chem.
242, 5019.

 

Stockell, A. (1959), 1, Biol. Chem. 234, 1286.

 

Taketa, K., and Pogell, B. M. (1963), Biochem. Biophys.
Res. Commun. 18, 229.

Tanford, C. (1964), Brookhaven 82MB, Biol. 11, 154.

Taylor, E. L., Meriwether, B. P., and Park, J. H. (1963),
g, Biol. Chem. 238, 734.

Taylor, J. F. (1950), Fed. Proc. 8, 237.

 

274
Taylor, J. F., and Lowry, J. (1956), Biochem. Bithys.
Acta 19, 109.

Trentham, D. R., McMurray, C. H., and Pogson, C. I. (1969),
Biochem. 1, 114, 19.

Tsai, C. M., Holmberg, N., and Ebner, K. E. (1970), Arch.
Biochem. Biophys. 136, 233.

Velick, S. J. (1958), 1. Biol. Chem. 233, 1455.

 

 

Velick, S. F., and Furfine, C. (1963), The Enzymes Z, 243.

 

DeVijider, J. J. M., and Slater, E. C. (1968), Biochem.
Bithys. Acta 167, 23.

Watson, H. C., and Banaszak, L. J. (1964), Nature 204, 918.

 

Warburg, 0., and Christian, W. (1939), Biochem. 1, 303, 40.

Weber, K., and Osborn, M. (1969), 1. Biol. Chem. 244, 4406.

 

Williamson, J. R. (1965), 1, Biol. Chem. 240, 2308.

Williamson, J. R. (1967), 7th Intern. Congr. Biochem.
Abstracts (Tokyo) 1, 863.

 

 

Windmueller, H. G., and Kaplan, N. 0. (1961), 1, Biol.
Chem. 236, 2716.

 

Wood, W. A. (1966), Ann. Rev. Biochem. 11, 521.

Yang, S. T., and Deal, W. C., Jr. (1969a), Biochemistry 8,
2806.

 

Yang, S. T., and Deal, W. C., Jr. (l969b), Biochemistry g,
2814.

Yonuschot, G. R., Ortwerth, B. J., and Koeppe, 0. J.
(1970), 1, Biol. Chem. 245, 4193.

 

thantis, D. A. (1964), Biochemistry 1, 297.

 

Zito, R., Antonini, B., and Wyman, J. (1964), 1, Biol.
Chem. 239, 1804.

 

 

TF'L‘ r ..
.
'l