ABSTRACT

REVERSIBLE DISSOCIATION AND CHARACTERIZATION
OF RABBIT MUSCLE erLYCEROPHOSPHATE DEHYDROGENASE

By William H. Holleman

Physical and chemical studies on native and dissociated
rabbit muscle dyglycerophosphate dehydrogenase ( apGDH) have
shown the native enzyme (s20,W = 4.868, Dgo,w = 6.20 x 10'7
cm2/sec, M$(s/D) = 74,400) to consist of two noncovalently
bound polypeptide chains (s20,W = 1.708, D30,w = 0.1 x 10"7
cm2/sec, M3(s/D) = 40,000). Each mole of native enzyme was
found to contain 2 moles of C-terminal methionine as deter-
mined by the carboxypeptidase technique, and 21 moles of
free sulfhydryl groups as determined by both carboxymethyla-
tion of reduced protein and by performic acid oxidation.
Fingerprinting of the tryptic peptides suggests the poly-
peptide chains are not grossly dissimiliar and may be iden-
tical. The partial specific volume ofci-GDH in 0.13 KCl was
determined, using density gradient techniques, to be 0.746
cc/g, in good agreement with the value calculated independently
from the amino acid composition. The enzyme may be dissociated
into stable subunits by either limited performic acid oxidation,
followed by dialysis into 8-Ofl guanidine-HCl. or in the
solvent system of 7.2g guanidine-HCl, 0.lfl_mercaptoethanol.
Dilution of the guanidine.HCl dissociated enzyme (s20.W = 1.708)

at optimum conditions reversed the dissociation process with

90% recovery of enzyme activity. The optimum conditions for

William H. Holleman
reversal were 0.li«_I Tris-H01, at pH 7.42, 0.001131 EDTA, 0.2;;
mercaptoethanol 5111 a final enzyme concentration of 0.025

mg/ml.

REVERSIBLE DISSOCIATION AND CHARACTERIZATION

OF RABBIT MUSCLECI-GLYCEROPHOSPHATE DEHYDROGENASE
By

William H. Holleman

A THESIS

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

DOCTOR OF PHILOSOPHY
Department of Biochemistry

1966

ACKNOWLEDGEMENTS

The author wishes to thank Dr. w. C. Deal for his
guidance and aid during the course of this research. The
technical help of Doris Bauer is also greatly appreciated.
The support of a National Institutes of Health predoctoral
fellowship is gratefully acknowledged.

ii

TO BETTY

Page
IlqrrLiODLTCr-E102Q’O00......OOOOOOOOOOOOOO0.OOOOOOOOOOOOIOOOOOO l
LITE-iAPBUL‘iE :QEVIET’q...0.........O.C....................... 2

Occurence of a-glycerophOSphate Dehydrogenase...... 2
Properties of nativeCX-glycerOphosphate Dehydro-
renase.................................o....... 3
Preparation and Purification of a-GDH......... 3
Stability Of OFGDHoooooooooooooo0000000000000. 3
PhySical Properties Of(X’GDHoooooooooooooooooo 4
Interaction of OkGDH with Aldolase............ 6
Prosthetic Groups of c»GDH.................... 7
Afllno A011 AnalySiS Of(l-GDHoooooooooooooooooo 10
SUlfhydryl Content Of lyGDHooooooooooooooooooo 10
N-terfiinal StUdieS Of x-GDHooooooooooooooooooo ll
Flnfierprlntln: Of aGDfi....................... 11
Properties Of Denatured(i-GDHoooooooooooooooooooooo 12
Effect of Acid and Alkali......................12
Effect of Other Dissociating Agents on a-GDH.. 12
Properties of the Catalytic Reaction............... l3
Specificity of the ckGDH Beaction............. 13
DH for Optimum ACtiVityo00000000000000.0000... 13
Turnover Number and Substrate Km's............ l3
Inhibitors and Stimulators of thecx-GDH
HeaCtionooooooooooooooooooooooooo0.00.000 14
Inhibition of a-GDH by(x-glycerophOSphate..... 15
Effect of Sulfhydryl Agents on a—GDH Activity. 15
3018 Of(1"GDH in Cell MetabOlismooooooooococo-coco. 16
{LglycerOl phosphate Shunt.................... 16

crglycerophosphate-pyruvate Dismutation....... 1?
Absence of a-GDH Activity in Malignant Tissue 18

NATERIALS AND METHODSOOOOOOOQOQQno...oncooooooooooooooooo 19

I’laterialso00.0000000.000000000000000...0000000000000 19

I'lethOdSOCOCCOOOOOOCCO.......00....OOOOOOOOOOOOOOOOOO 20

RESULTS.OOOOOOOOOOOOOOOOO0.0.0.0000...OOOOOOOOOOOOOOOOOOI 28

Physical Properties of Native a-glycerophosphate
Dehydrogenaseoo00000000000000.0000.000000000000 23
Preparation of "aggregate" free GLGDH.......... 28
Wolecular Weight of OhGDH as Determined by
D t Sedimentation Equiligrium...........3..... 29

e erm na ion of s D and M s D of
Native a-GDHO§9:Y:OO%9:?IOOOOOOYEO{CO.00.. 34

iv

TABLE OF CONTENTS - Continued Page

Partial SpGCifiC VOlumeO..OOOOOOOOOOOOOOOC0.0. 40
Chemical Properties Of a-GDHooooooooo00000000000000 49
Analysis for Disulfide Bonds and Sulfhydryl
Groups Of a-GDHOOOOCOOOOOOOOOOOOOOOOOOOO. 1+9
PHDIB TitrationOOOOOOOOOOOOOOOOOOOOOO.00.. Ll'9
Carboxymethylation of a—GDH.............. 51
Performic Acid Oxidation of a-GDH........ 51
Combination of Carboxymethylation and
Performic Acid Oxidation............ 51
Amino Acid Analysis of 0-GDH.................. 54
Carboxy-terminal Analysis of d—GDH............ 56
Fingerprinting Of aGGDHooooooooooooooooocoo-co 59
Subunit Structure of a-glycerophosphate Dehydro-
genaSeOOOOOOOOOOOOOOOO.OOOOOOOOOOOOOOOOOOOOOOO 63
Attempted Dissociation of a-GDH in urea....... 63
Effect of Guanidine-HCl Concentration on
the S W Of a-GDHooooooooooooooooooooooo 64
SedimentatIgn Equilibrium Studies of Q-GDH
subunitst.OOOOOOOOOOOOIOOOCOOOOOOOOOOO... 73
The Requirement for Mercaptoethanol........... 73
Performic Acid Oxidation of a—GDH............. 79
Molecular Weight of Unreduced, Carboxy-
methylated- a-GDH0.0000000000000000000000. 82
Summary of the Molecular Weight Analysis of
the G-GDH subunit-10......OOOOOOOOODOOOOOOO 86
Benaturation Of a-GDHOOOOOOOOOOOOOOOOOOOOOO00...... 86
Effect of pH on Renaturation.................. 88
Effect of Protein Concentration on Renatura-
tionOO0.00.000000000000000.00000000000000 88
Effect of Metabolites on Renaturation......... 91
Product Inhibition of a—GDH bycx-glycero-
phosphateOOOOOOOOOO0.00000000000000000... 96
Kinetics of Benaturation in the Presence of
NADH Or DHAPoo00000000000000.0000oooooooo 97
Renaturation in NerCaptoethanOlooo000.000.000.100

DISCUSSIONOOCCOCCCOOO0.....OOOOOOCCCOCCOCO.0...’........103
SUMMARYOOCOO0.000.000.0000...OOOOOOOOOOOOO000.00.000.000109
BIBLIOGRAPHY...ooooooooooooooooooooooooooooooooooooooooolll

APPENDIXOOOOOOOOOI0.0.0.000...0.0.0.000...0.0.00.0000000116

LIST OF TABLES
TABLE Page

I. Partial specific volume as calculated from
amino acid analysisOO0.0.00.0...0.00.00.00.00... 48

II. Titration of OLu-G-DH with PHMB.................... 50

III. Analysis for sulfhydryl groups and disulfide
bOndS Of OFGDHoooooooooooooooococoon.ooooooooooo 52

IV. Total amino acid analysis of d-GDH at varying
times Of hydr01ySiSoo0.0000000000.000.000.000... 55

V. Release of C—terminal amino acids with time..... 59
VI. Staining of peptides for specific amino acids... 60

VII. Summary of molecular weight analysis of the
G‘GDH Subunit.’COOOOOOOOOOOOOOOIOOOOOOOOOO.0000. 86

VIII. EffeCt OfCX-GP on &‘GDHooooooo0.0000000000000000 97

vi

FIGURE

1.

10.

ll.

12.

13.

LIST OF FIGURES

Extrapolation of the apparent weight average
(MW) molecular weights of nativecx-GDH to
zero protein concentration....................

Extrapolation of the apparent weight average
(NW) and z-average(M ) molecular weights to
zero protein concentration....................

Extrapolation of sedimentation coefficients of
nativecx-GDH to zero protein concentration....

Extrapolation of the diffusion coefficients
of native a-GDH to zero protein concentration.

Graphs of theoretical and empirical sedimen-
tation coefficients as a function of molecular
weights for glObular proteinS.................

Graphs of theoretical and empirical diffusion

coefficients as a function of molecular weights

for globular proteinS.........................

Protein concentration versus apparent partial
specific volume(v) ofcx-GDH...................

Destruction of serine and threonine during
801d. hydrOlySiS Of a-GDHOOOOOOOOOOOOOOOOOOOOOO

A reconstructed peptide map of a trypsin
digest of reduced carboxymethylatedc1-GDH.....

Effect of guanidine-HCI concentrations on the
sedimentation coefficient of a-GDH............

Extrapolation to zero protein concentration

of the sedimentation coefficients of a-GDH sub-

units in guanidine, mercaptoethanol...........

Extrapolation to zero protein concentration

of the diffusion coefficients of CLGDH subunits

in guanidine-HCl, mercaptoethanol.............

Extrapolation to zero protein concentration of

apparent weight average and z-average molecular

weights ofCI-GDH subunits in guanidine-HCl,
mercaptoethanol...............................

v1 i

Page

30

32

35

38

#1

43

45

57

61

65

69

71

7h

LIST OF FIGURES - Continued Page

Figure

14.

15.

16.

17.
l8.
19.
20.

21.

Extrapolation to zero protein concentration of
the apparent weight average (MW) molecular weights
of a—GDH in guanidine-HCl with no mercaptoethanol....77

Extrapolation to zero protein concentration of
z-average molecular weight (M2) of performic
aCid OXidized aPGDHOQQOOQOOOOO0.00000000000000000000080

Extrapolation to zero protein concentration of
the apparent weight average ( ) and z-average
(M ) molecular weights of a-GD simultaneously
cagboxymethylated and dissociated....................84

Effect of pH on renaturation of a—GDH................89
Effect of protein concentration upon renaturation....92
Effect of a-glycerophosphate upon renaturation.......9u

Effect of NADH or DHAP upon the renaturation

Of a-GDHO.0.0.0.000...0......OOOOOOOOOOOOOOOOOOOOOO0.98

Renaturation of GPGDH in the presence of various
concentrations of mercaptoethanol...................101

INTRODUCTION

One of the ultimate problems of biochemistry is to
determine the correlation between the structure and func-
tion of enzymes. It was hOped that by providing a detailed
analysis of the structure of an enzyme additional knowledge
concerning such a relationship would be gained. Preliminary
studies (Deal 32 31., 1963) onCX-glycerophOSphate dehydrogenase
from rabbit muscle had revealed a reversible denaturation.
Because of this and the interesting prOperties described
below this enzyme was chosen for further analysis. The mole-
cular weight of nativecx-GDH seemed too large for a single
polypeptide chain and therefore it was suSpected thatcx-GDH
was composed of two or more subunits. The original goal of
this investigation was to find conditions for dissociation
of the enzyme into stable subunits and to determine the size
and number of the subunits of rabbit muscle a-glycerOphosphate
dehydrogenase by both chemical and physical methods. A study
of reversal of the dissociation process was also planned
pending successful dissociation of the enzyme into subunits.
In the course of these investigations it became clear that
the physical properties of native a-GDH reported in the
literature were inconsistent. Since a reliable knowledge
of the structure of the native enzyme was a prerequisite to
a knowledge of the subunit structure, a detailed reinvestiga-
tion of the physical prOperties of native G-GDH was also

conducted.

LITERATURE REVIEW

Occurence of a-glycerophosphate Dehydrogenase

a-glycerOphOSphate dehydrogenase (a-GDH) is an enzyme
catalyzing the reduction of dihydroxyacetone phOSphate (DHAP) to
aeglycerOphOSphate (d-GP) accompanied by the oxidation of
NADH to NAD: It was originally discovered by Meyerhof in
1919. The subject of this study is the water soluble
cytOplasmic rabbit muscleCI-glycerOphOSphate dehydrogenase.
There is also a mitochondrial<1-g1ycerophosphate dehydro-
genase which differs from its cytOplasmic counterpart in
that the coenzyme for the reaction is FAD rather than NAU’
and the reaction greatly favors the oxidation ofCX-GP,
whereas the cytOplasmic enzyme greatly favors the reduction
of DHAP. For this reason the mitochondrial enzyme is known
asa -glycer0ph05phate oxidase.

(I-GDH activity has been demonstrated in insect muscles,
different rat organs (Young and Pace, 1958b), components of
the blood and is found to some extent in all the organs of
both vertebrates and invertebrates (Delbruck 23 al., 1959;
Zebe, 1960). With the notable exceptions of the Morris
hepatoma 5123 and the ascites Ehrlich-Lettre tumor of the
mouse (Morris gt a1., 1960), a-GDH activity is either lacking
or very low in most malignant tissues. The observations of
low levels ofcx-GDH activityine not due to the presence of
an inhibitor, since the addition of tumor extracts to

2

3
extracts from normal tissues did not inhibith-GDH activity

in the normal tissue extracts.

PrOperties of Native a-glycerophOSphate Dehydrogenase

 

Preparation and Purification Quill};

As in the original purification and crystallization
(Baranowski, 1949) the standard method of preparation
involves fractionation of the enzyme in the 42-60% saturated
ammonium sulfate range. Of several new procedures and modifi-
cations (Disteche, 1948; Beisenherz 32 al., 1953) two have
yielded significant increases in Specific activity. Van Eys
33 a1, (1959) obtained a three fold increase by adding a
heat step, and of secondary importance, DEAE chromatography
under conditions where all other proteins were absorbed. A
recent procedure (Telegdi, 1964) claimed an activity three
times that of any previous procedure. A rabbit muscle
CZ-glycerOphOSphate dehydrogenase with a higher molecular
weight and different crystalline form (Young and Pace, 1958a)
is probably a dimer of the "standard" enzyme produced by
either the isolation procedure or by the high concentration
of ammonium sulfate (0.5m) in the solvent used for their

molecular weight analysis.

Stability,g£‘g:ggfi

The thermal stability of ObGDH has Special interest
because of the heat step sometimes employed for purifica-
tion (Van Eys gt_al., 1959). Although the enzyme loses no
activity upon standing at 250 for 30 minutes or for 2 weeks

at O0 in 0.2g ammonium sulfate, the loss after 1 minute at

u

550 is 53% and the loss after 1 minute at 600 is 100% (Young
and Pace, 1958a). The only information on stability in the
pH range 5-9 has been obtained with the previously mentioned
"unusual" enzyme (Young and Pace, 1958a). Using catalytic
activity as the criterion for stability the enzyme is most
stable at pH 5.7-6.2, is least stable at pH 7.0 and shows
fair stability at pH 8.5. Beisenherz (1953) reported that,
even in the absence of salts, the use of redistilled water

prevented denaturation ofCI-GDH.

Physical Properties ofcx-GDH

The physical properties of nativeCX-GDH in both the
nucleotide-containing and nucleotide-free form have been
studied in several laboratories. In either 0.323 ammonium
sulfate or 0.1fl'malate, pH 6.28, both the native and the
nucleotide-free protein yielded values1 of Sgo,w = 4.98,
and Dgo’w = 5.1 x 10"7 cmZ/sec (Van Eys §§_a1., 1959) which
were only Slightly dependent on protein concentration. Using
the value1 of 0.70 cc/g found for the partial specific volume,
together with the sedimentation and diffusion data Van Eys

gt 1. (1959) calculated a molecular weight forCI-GDH of

 

1Our data indicate a value of 0.746 cc/g for the partial
Specific volume of GFGDH in contrast to the results in the
literature which are 0.70 cc/g (Van Eys E£.§l" 1959) and
0.75 cc/g (Young and Pace, 1958a). Inspection of the data
in Table 1 shows this value Should be 0.70 gc/gz Our data
also indicate a value of D30 = 6.20 x 10' cm /Sec for the
diffusion coefficient, in congrast tozthe result reported in
the literature which is 5.1 x 10‘7 cm /sec (Van Eys gt a1.,
1959). Van Eys g§_al. (1959) calculated a value of 1.44—for
the frictional ratio (f/fo) while the value obtained by us is
1.23. Further discussion of these discrepancies and their
effect on molecular weight calculations, may be found in the
text.

78,000. From these combined data the frictional ratio
(fO/f) was calculated to be 1.44, which is larger than values
usually obtained for globular proteins (Tanford, 1961).

Ankel 22 a1. (l960)using a solvent of 0.053 phosphate, 0.g§

0.8%
20,w

native and nucleotide free protein. Young and Pace (1958a),

NaCl, pH 6.8, found an S value of 4.948 for both the
although using conditions (0.5g ammonium sulfate) which
differed only slightly from the conditions of Ankel gt a},
(1960) and of Van Eys at 31. (1959), found a considerably
higher value of Sgo,w = 6.58. As previously suggested these
values may represent a dimer of the enzyme, or they may be
due to aggregation. Although Young and Pace (1958a) reported

a value1

of 0.75 cc/g for the partial Specific volume,
reevaluation of their data suggests that 0.70 cc/g is a better
value (see Table 1). Using their value of 7.2 oc/g for the
intrinsic viscosity Of<1rGDH in ammonium sulfate, Young and

Pace (1958a) calculated a molecular weight of 173,000, and

a frictional ratio of 1.4, in good agreement with the frictional
ratio of 1.44 found by Van Eys 23 El. (1959). These values

for the frictional ration and the intrinsic viscosity are

much higher than expected for globular proteins which have

frictional ratios which range from 1.1 to 1.2 and 3.0 to

4.0 respectively (Tanford, 1961).

Table 1. Apparenta specific volume ofcx-GDH, taken from
Young and Pace (1958a).

 

 

Determination Specific Volume (cc/gm)

 

0.70
0.70
0-95
0.70

0.20
Mean 0.75

\n-{I-‘KJONH

Standard deviation 0.11

 

aEnzyme (4 mg/ml) in In ammonium sulfate at 240

Jirgensons (1965) has measured the Optical rotary diSper-
sion Of(I-GDH and calculated the Moffitt constant, b0, of
2080, which correSponflfl.to anCX-helical content of 34%,
measured at a Naof 216mu. For this calculation polyglutamic
acid was used as a standard for 100%CX—helix and itwas assumed
that the only kind of helical structure presentvun the right-
handed<1~helix.

Interaction QQCX-GDH with Aldolase

 

The first data suggesting interaction between.a-GDH and
aldolase was the observation (Baranowski, 1939; 1949b) that
crystalline Myogen A containedCI-GDH activity as well as
aldolase activity. Several recrystallizations of the Myogen A
did not separate theCI-GDH activity from the aldolase activity.
This interaction has been substantiated by Gulyi (1959), but

they could not detect<1-GDH activity in Myogen A, unless the

Myogen A was treated with 0.53'urea for two hours.

7

Addition

of purified aldolase or Myogen A to pure d-GDH resulted in a

decrease of‘l-GDH activity and an increase in aldolase activity

(Litvinenko, 1963).

Sereda (1960) demonstrated that the incu-

bation of CZ-GDH with aldolase resulted in a 15% increase in

aldolase activity as well as protecting a-GDH from thermal

denaturation.

these phenomena.

Prosthetic Groups 22a -GDH

No physiological role has been assigned to

The most important prosthetic group ofcx-glycerophOSphate

dehydrogenase is NADH.

As seen in Table II, various attempts

to measure the amount of NADH bound to the enzyme have yielded

 

 

 

Table II. Binding of NADH by<I-GDH.
fiales of NADH Bound Method Reference
of
Determination
1.0 per 70,000g Fluorometric Ankel gt a1., (1960)
titration
1 per 65,0003 Sedimentation Ankel gt al., (1960)
(400,000 x g
1-3 Specifically Equilibrium Van Eys gt_a1., (1959)
per 78,000g dialysis
31 non specifically Equilibrium Van Eys 23 a1., (1959)
per 78,000g dialysis

22 per 78,000g

Sephadex gel
filtration

Pfleiderer and Aur-
ricchio (1964)

 

different results ranging from 1 to 3.

The fluorometric

titration (Ankel 2E.§l°9 1960) is based on the fact that when

8
free NADH binds tOCX-GDH, its absorbance maximum shifts from
463mu to 454mu and its fluorescence increases 7-10 fold.
The dissociation constants for the two classes (Van Eys gt a1.,
1959) of NADH binding found by equilibrium dialysis are:
(l) 6.0 x 10"LF for the nonspecific class and (2) 1.8 x 10-6
for the Specific class. While there is no explanation for
the values of 1 mole of NADH bound per mole of enzyme obtained
by Ankel gt El. (1960) the Sephadex gel filtration technique
(Pfleiderer and Auricchio, 1964) seems to have no experimental
pitfalls or deficiencies and thus seems to us the best value.
In connection with this data, it is of interest that Pfleiderer
and Auricchio (1964) found one mole of NADH bound per 30,000-
45,000 grams of protein for seven different dehydrogenases;
this protein mass presumably represents the individual poly-
peptide chains, which therefore would be predicted to possess
one active Site per polypeptide chain.

In contrast to the majority of the other common dehydro-
genases which catalyze reactions whose equilibrium favors
oxidation,C1-GDH catalyzes a reaction in the reductive direc-
tion. As a result of this, most of the other common dehydro-
genases bind NAD‘but not NADH. In contrast, a-GDH binds NADH
and not NAD: Extensive attempts (Van Eys §t_a1., 1958) to
demonstrate NAD+binding have not been successful.

A second prosthetic group of a-GDH is ADP-ribose (Van
Eys e£.al., 1964; Ankel gt al., 1960; Celliers gt al., 1963).
It is possible that the ADP-ribose is a degradation product

of bound NADH. The bound ADP-ribose is not an integral part

9
of the active site ofa -GDH for its removal (Ankel SE.§$°9
1960) does not affect the Specific activity, $20,W, or
turnover number of a-GDH; however its removal lowers the
electrophoretic mobility and exposes an additional sulfhydryl
group.

The ADP-ribose was first detected as a result of the
observation by Baranowski (1949a) and Beisenherz g§,_1. (1953)
thatCX-GDH crystallized from rabbit muscle exhibiUfl.an abnor-
mal ultraviolet Spectrum in the 260-280mu region. Van Eys
23 a1. (1959) was able to separate the bound ADP-ribose from
(l-GDH by treatment with charcoal, trichloroacetic acid, acid
alcohol, heat, or by dialysis against cysteine at a pH above
8. Quantitative analysis of this nucleotide (Ankel et a1.,
1960) gave 2 moles of ribose, 2 moles of phOSphate, and 0.9
moles of adenine per 70,000 grams of protein. The substance,
demonstrated by enzymatic means to be neither NAD+or NADH,
migrated identically in paper chromatography eXperimentS in
several different solvent systems with authentic ADP-5'-ribose.

The existence of a third prosthetic group has been
reported (Van Eys, 1960) but its identity is uncertain. This
compound, originally thought (Van Eys, 1960) to be 3-(4-methyl-
5-(B-hydroxethyl)-thiazoly1)-succinic acid, was given the
trivial name of thiamic acid. Celliers g: 31. (1963) have
questioned this structure and Van Eys gt 21. (1964) indicated
in a later report that the structure seemed less certain than
it had previously. Thiamic acid (Van Eys, 1960) reactivated

charcoal treated yeast or rabbit muscle a-GDH, suggesting

that it might have some function in the enzyme. As additional

10
evidence for the presence of a thiazole derivative as part of
the a-GDH molecule he showed that thiamine deficiency in rats

resulted in a significant lowering of theCZ-GDH content.

Amino Acid Analysis g§_g;ggg

Van Eys ggpgl. (1964) have calculated from the amino
acid analysis a minimal molecular weight of 38,400 for the
half molecule ofCI-GDH, based on 28 aSpartic acid residues
per half molecule ofCI-GDH. Since the molecular weight from
physical measurements (Van Eys g2_§l., 1959) was 78,000 a
molecular weight of 76,800 was indicated. Approximately 4
moles of tryptOphan were found per mole of a-GDH by both
microbiological and Spectrophotometric methods, after removal

of the prosthetic groups with trichloroacetic acid.

Sulfhydryl Content g£_g:ggfl

Ankel EE.§$* (1960) reported the presence of 15 moles of
cysteine per 70,000 grams of protein as determined by
p-mercuribenzoate titration in neutral phOSphate buffer.
Van Eys £3 al. (1964) also using p-mercuribenzoate, as well as
(di-3-carboxy-4-nitrophenyl) disulfide, found only 12 moles
of cysteine in the presence of 8E urea and values as low as
8 in the absence of urea. When the free sulfhydryl groups
were blocked by dinitrophenylation and then oxidized with
performic acid, 1.7 moles of cysteic acid were found (Van Eys
‘32 a1,, 1964). This suggested thatai-GDH had a maximum of 1
disulfide bond; these authors interpreted this as evidence

that no disulfide bonds were present.

11

N-terminal Studies girl-GDH

 

Using the fluorodinitrobenzene technique van Eys gt El.
(1964) found no free N-terminal amino acids, raising the poss-
ibility that the N-terminal amino acids might be N—acetyl
derivatives, as in enolase and several other proteins(winstead
and Wold, 1964). Using the colorimetric technique of LUdoweig
and Dorfman(l960) they calculated a value of 2 moles of
acetyl per 78,000 grams of protein; unfortunately, the accuracy
of this determinationvmm;limited by a large correction (50$)
for the blank, because of the presence of adenosine diphOSphate
ribose (sxaProsthetic Groups). The method involves deacetyla-
tion by hydrolysis of the N-acetyl bond in a 2g HCl methanol
mixture. The methyl acetate product is spectrOphotometrically
determined as the hydroxylamine-ferric complex by its absorbance
at 520 mu. This complex is formed in a reaction mixture con-
taining hydroxylamine, ferric chloride, perchloric acid and
HCl. Presumably ADP-ribose also reacts with this reagent to

give a colored complex. Nevertheless, this analysis suggested

the existence of 2 or 3 polypeptide chains in the native enzyme.

Fingerprinting g: E:g21

Further analysis of the number and identity of the poly-
peptide chains in d—GDH utilized the fingerprinting of tryp-
tic peptides; this revealed no more than 16 peptides (Van Eys
gt 31., 1964) despite repeated experiments. This is less than
half the number of tryptic peptides which would be

predicted if the native enzyme contained two identical

12

polypeptide chains.

Properties of Denatured a-GDH

 

 

Effect g£_Acid and Alkali

 

The reversible inactivation of rabbit musclecx—GDH in
acid has been studied by several workers (Deal and Van Holde,
1962; Deal g3 31., 1963; Van Eys, 1963; Van Eys gt gl., 1964).
Deal and co-workers reported an inactivation without dissocia-
tion into subunits in 0.013 citrate, pH 2.6; the Van Eys
group reported dissociation into subunits in 0.01fl glycine,
pH 2.6, in the presence of 0.2§_Na2304, as well as in 1.5M
NHQOH. Van Eys (1963) reported that alkali treated a-GDH
moved with the salt boundary on Sephadex G-25 in high ionic
strength solvent (0.35fl_ammonium sulfate); but in a low ionic
strength solvent the protein was excluded from the Sephadex
and came through the column with the solvent front. Both the
alkali and acid treated<1-GDH had sedimentation coefficients

Of SZO,W = 2063.

Effect of Other Dissociating Agents on gzggg

Deal and Holleman (1964) reported that treatment of
CLGDH with 8.53 guanidine-HCl in the presence of mercaptoethanol
resulUfl.in the Splitting of the enzyme into subunits which
have a molecular weight of 4 x 104 and a sedimentation
coefficient of 820,w = 1.68. This dissociation was not
observed if the mercaptoethanol was omitted. Chilson §£__1.

(1965) have reported the reversible denaturation of OFGDH

from guanidine-HC1. Sodium dodecyl sulfate, p-mercuribenzoate,

13
an inhibitor of OFGDH, urea or zinc ions did not dissociate

the protein into subunits (Van Eys gt_§1., 1964).

PrOperties g; the Catalytic Reaction

Specificity g: the a-GDH Reaction

Rabbit muscle a-glycerophOSphate dehydrogenase is Specific
for L-glycerol-l-phosphate and dihydroxacetone phosphate. The
reduction of 1,2-propanediol-l-phoSphate attributed to CLGDH
by Selinger and Miller (1949) has been shown by them to be due
to a distinct l,2-pr0panediol-l-phosphate dehydrogenase
(Selinger and Miller, 1958). Borreback gt a;. (1965) have
Shown that a-GDH oxidizes NADPH at a rate approximately 10%
the rate of NADH oxidation; values of the pH optimum Shifts
from pH 75 for NADH to pH 57 for NADPH. The reaction catalyzed
by rabbit muscle a-GDH is stereOSpecific for B-NALr(Levy and
Vennesland, 1957); the oxidation of NADH, deuterated in the
a-position proceeds with complete retention of the deuterium

in the NADZ

pa for Optimum Activity

Young and Pace (1958a) using the higher molecular weight
enzyme mentioned earlier, found pH Optimums for dihydroxy-
acetone phOSphate and a-glycerOphOSphate as substrates of
7.5 and 10.2 reSpectively. The rate of 0'--GP oxidation at
pH 10.2 is about one-twelfth the rate of DHAP reduction at

pH 7.5, using the same substrate concentrations.

Turnover Number and Substrate Km's

The equilibrium of the a-GDH reaction greatly favors the

formation of<1-glycer0phOSphate. Values of 4.6 x 10‘123

14
(after extrapolation of the apparent equilibrium constants to
zero ionic strength) and 5.7 x 10'12fl were found respectively
by Burton and Wilson (1953) and Bucher and Klingenberg (1960).
From equilibrium and free energy data (Burton and Wilson,
1953) for the oxidation of NADH a free energy change of—10.24
kcal. was calculated.

The turnover number for the reduction of DHAP (20°, pH 7,
ionic strength 0.15) was calculated by Baranowski (1949b) to
be 20,700 moles per minute per 78,000 grams of protein and by
Beisenherz gt gl., (l953),23,400 moles per minute per 78,000
grams of protein at 25°. Young and Pace (1958a) reported a
much higher value of 35,000 moles per minute per 78,000 grams
of protein at 220. Other kinetic data are scarce. The
approximate Michaelis constants at pH 7.0 have been reported

“g for NAD: 4.6 x 10-43

-4

as 1.1 x 10-43 for Q-GP, 3.8 x 10'
(Bacher and Klingenberg, 1960) and 3.63 x 10 fl (Blanchaer,

1965) for DHAP.

lphibitors and Stimulators gt the OLGDH Reaction

 

A large number of NAD analogs studied by Van Eys gt gt.
(1959) were found to be potent inhibitors of O‘-GDH. The only
exception was 4-(l-imidazoy1)NAD, which is neither a coenzyme
Hm an inhibitor. Ankel 22.21- (1960) showed that only 60%
maximal activity is obtained if EDTA is not included in the
reaction mixture. Emmelot and 808 (1962) reported a stimulatory
effect of potassium cyanate on crude a-GDH preparations, which
they attributed to a protective action of the cyanate anion

against inactivation of the enzyme by heavy metals. Other

15

labortories using crystalline apcDH (Boxer and Devlin, 1962;
Borst, 1962; Pette and Huge, 1963) could not reproduce this
effect and Pette and Huge Showed that this stimulatory effect
was due to the use of impure DHAP. KCN caused decomposition
of this impurity to pyruvate which was a substrate for lactic
dehydrogenase, present in the crude preparations of Emmelot
and B08 (1962). Sacktor and Dick (1965) reported that many
substituted cinnamic acids inhibited crystalline d-GDH, the most
potent of these being p-nitrocinnamic acid which gave 50%
inhibition at 3 x lO'ufl. This inhibition was not competitive
with DHAP and could be reversed by dilution of the enzyme.
Shonk (1962) reported that zinc ions at a concentration of

10'6fl strongly inhibited the enzyme in both directions.

Inhibition g£<x-GDH tz_a.glycerophos;hate

 

 

Blanchaer (1965) reported that W-GP inhibits d-GDH and
measured an inhibition constant of 5.6 x lO'SE. Blanchaer
(1965) also showed that there was a sharp decrease in DHAP
reduction (12.2222) when the concentration of d-GP rose above
1 umole per gram of muscle tissue. Consideration of the
estimated concentrations of DHAP and OFGP present in muscle
during contraction (680 and 60 mu, respectively) suggests
that apGP may influence the pathway of triose phosphate

utilization (Blanchaer, 1965).

 

Effect gt Sulfhydryl Agents on d—GDH Activity
Van Eys gt_gl, (1959) found that one mole of p-hydroxy-

mercuribenzoate (PHMB) per 87,000 grams of protein completely
-41»,

inhibited dsGDH as did a concentration of l x 10

16
N-ethylmaleimide. Iodoacetic acid (1 x lo-ufl) gave only a
23% decrease in activity. In contrast to these results,
Telegdi and Keleti (1964) reported that titration of 3
sulfhydryl groups per mole of a—GDH gave a 50% loss of
activity and it was necessary to titrate 9-11 sulfhydryl

groups in order to abolish all enzymatic activity.

Role f chDH 1 Cell Metabolism

 

 

(rglycerOphosphate Shunt

 

The role of<1-GDH in cell metabolism, except for the
production ofCI-GP for fat synthesis, remained largely
unexplored until Lehninger (1951) noted that extramitochon-
drial NADH is not a substrate for the electron tranSport chain.
A system for the transfer of hydrogen from the extramitrochon-
drial NADH to the respiratory chain was postulated by Zebe
gt g_. (1956) utilizing thecx-glycerophOSphate cycle illustra-

ted below.

4..

Qytoplasm NAD NADH

a-glycerOphOSphate ,X. J) Dihydroxyacetone-P

 

 

 

 

\
a -GDH /
Mitochondria FAD FADH2
w L f
(x-glycerOphOSphate \, Dihydroxyacetone-P
7

 

(I-glycerOphOSphate
oxidase

17

Support for this model come from the demonstration that
mitochondria from various sources contained a flavin linked
enzyme that oxidized cxglycerOphOSphate by way of the phos-
phorylating electron transport chain (Sachtor gt gt., 1959;
Klingenberg and Slenczka, 1959; Zebe gt_gt., 1957). Recently
an NAD—linked, mitochondrial(x—glycerophosphate dehydrogenase
has also been found (Tomita and Helling, 1965). The equilib-
rium of the a-glycerOphOSphate oxidase greatly favors the
formation of dihydroxacetone phOSphate as a result of the FAD
potential. The alternate extramitochondrial reduction of
DHAP and mitochondrial oxidation of O’mGP thus form an effective
system for shuttling reducing equivalents from extramitochon-
drial NADH to the intramitochondrial electron tranSport system.
It should be noted that any dehydrogenase which is located in
both the cytoplasm and mitochondria can provide a means of
transferring electrons into the mitochondria, if the substrates
are permeable to the mitochondria membrane. Two enzymes in
this category are lactic dehydrogenase and 63-hydroxybutyrate

dehydrogenase (Boxer and Devlin, 1961).

agglycergphosphategpyruvate Dismutation
a -GDH may also play a role in the so-called "a-glycero-
phOSphate-pyruvate dismutation" which is outlined in the

N
diagram below (Klingenberg and Bucher, 1960). The result of

 

 

lactate a -g1yCerophOSphate
+-
~———NAD
%NADH
\L

 

pyruvate dihydroxacetone-P

18
this cycle is the transfer of hydrogens from lactate to
DHAP with the formation of an extra pyruvate. The end result
is that lactate does not accumulate at the expense of pyruvate.
An extra production of pyruvate from a-GP could allow the
production of lactate without a corresponding lowering of the
lactate-pyruvate redox potential because of a build up of
lactate. This mechanism could provide a stabilization of the
redox potential of the extramitrochondrial ATP system under
inadequate oxygen supply. Such a system can also explain the
lack of a change in the lactate-pyruvate ratio when there is
a rise in the lactate level in Skeletal muscle, heart, brain
and liver of the rat (Hohorst gt_gt., 1959; Bucher and
Klingenberg, 1958) and during tetanic work of striated muscle
(Sachs and Morton, 1956).

Absence gtcx-GDH Activity tg Malignant Tissue

In most malignant tissues, a-GDH is either absent or
present in very low levels (Boxer and Devlin, 1961). Holzer
_t__l. (1958) first observed this characteristic in Ehrlich
tissues and Yoshida hepatoma cells. In a variety of normal
mammalian tissues, including regenerating liver and embryonic
tissue the ratios of lactic dehydrogenase (LDH) to a-GDH was
found (Boxer and Slonk, 1960) to vary between 0.531 and 7.031;
in a large series of tumors of rodents and human beings this
ratio ranged from 1031 to several hundredszl. The increase
in the a-GDE: LDH ratio was due to a large decrease in the
amount of a-GDH present in the tumors and not due to an increase

in the amount of LDH (Boxer and Devlin, 1961).

MATERIALS AND METHODS

Materials

 

NADH, NAD: dihydroxyacetone phoSphate-cyclohexylamine
salt-dimethyl ketal'HZO and a-glycerOphOSphate were obtained
from the Sigma Chemical Co. NADH was prepared daily. A stock
solution (2 x lO'ZE) of dihydroxyacetone phosphate was prepared
for use by acidification with Dowex-50 (H+) and heating for
4 hours at 400 in order to hydrolyze the ketal. This solution
was stored frozen until used. Guanidine carbonate, iodoacetic
acid and mercaptoethanol were products of the Eastman Kodak
Chemical Co. All of the reagents used in the course of these
investigations were reagent grade quality and solutions were
made up in distilled deionized water.

Guanidine°HCl was prepared from guanidine carbonate by a
Slight modification of the method of Anson (1941). The carbonate
salt was recrystallized twice from 95% alcohol, dried, and
concentrated HCl added to the guanidine carbonate to a pH of 2.
The guanidine‘HCl was evaporated to near dryness, filtered,
washed with cold (-20°) 95% alcohol and dried in a vacuum at
500 for several days. Urea was obtained from the J. T. Baker
Chemical Co. and recrystallized from hot (40°) analytical 95%
alcohol.

S-carboxymethyl cysteine was prepared by the method of
Dickens (1933); it had a melting point of 1850, which compared

well with the reported value of 1840 (Dickens, 1933).

19

20

S-carboxymethyl cysteine sulfone was prepared by adapt-
ing the procedure of Moore (1963) for the preparation of
cysteic acid. The carboxymethyl cysteine was oxidized with
performic acid for 4 hours at 0°. Performic acid was prepared
by mixing 1 ml 30% H202 with 9 ml 88% formic acid. The reaction
was stOpped by the addition of 0.3 m1 of 48% HBr and the
reactants removed with a rotary evaporator at 400.

Rabbit muscle a-glycerophosphate dehydrogenase was obtained
from the Calbiochemical Co. Lot numbers used were 54734,
63490, 63688, 63684, 45818, 52614, 54520, 52610. Trypsin
(bovine pancreas) two times recrystallized was obtained from
the Sigma Chemical Co. Carboxypeptidase-A treated with
diiSOpropyl fluorOphOSphate was obtained from Worthington

Corp.

Methods

Ultracentrifigal Analysis

Samples were prepared for sedimentation equilibrium and
sedimentation velocity analysis by centrifigation of the pro-
tein from an ammonium sulfate suSpension and dissolving the
protein in the test solvent. This stock solution was then
dialyzed against this solvent for 48 hours. For a series of
experiments done at different protein concentrations, samples
of the stock protein were diluted directly with the dialysate
to give the desired concentrations.

Two Spinco analytical Model E ultracentrifuges equipped
with phase-plate schlieren Optics were used for these studies.

The Short column sedimentation techniques of Van Holde and

21

Baldwin (1958) were utilized. Techniques were developed for
the use of multicell (2 or 3 cells) operation using the double
sector cells equipped with sapphire windows and filled epon
centerpieces. Equilibrium was Judged complete in 36 hours for
the guanidine‘HCl-protein solutions and in 24 hours for the
native protein. Average initial protein concentrations (CO)
were obtained from a series of synthetic boundary experiments
performed at different protein concentrations immediately after
completion of dialysis. The computations were performed on a
Control Data Corporation 3600 digital computer using a program
develOped and extensively evaluated in this laboratory. The
calculations include statistical analysis of the data and a
plot of l/r-dc/dr versuSAC.

Sedimentation velocity eXperiments were run at 59,780
rpm at low temperature; exact temperatures were obtained from
an BTIC unit. The diffusion coefficients and synthetic
boundary experiments were performed in a double sector synthetic
boundary cell at approximately 5900 rpm at a temperature of 5°.
Diffusion coefficients were calculated using height to area
analysis (Schachman, 1957). Densities of solvents were mea-
sured to three Significant figures with hydrometers and viscos-
ities were either obtained from the International Critical

Tables or were measured using an Oswald free-fall viscometer.

Preparation gt OFGDH Derivatives
Limited performic acid oxidation of a-GDH, in preparation
for molecular weight analysis, was carried out by the method

of HirS (1956). The ammonium sulfate pellet was taken up in

22
88% formic acid and sufficient performic acid (1 ml 30% H202
+ 9 ml 88% formic acid) added to give a 10 fold excess over
the half-cysteine content of a-GDH. After oxidation the
protein was dialyzed for 24 hours against 8.53 guanidine°H01,
0.2g NaCl, 0.04fliTris'HCl, 0.001M EDTA, pH 8.3 to remove the
oxidation reactants. Sedimentation equilibrium eXperiments
were performed on the dialyzed sample.

Reduction and carboxymethylation of a-GDH was done
according to the method of Crestfield gt gt. (1963) with the
exception that the 81'urea was replaced by 8.5g guanidine.HCl
to insure complete unfolding for reduction and alkylation
with iodoacetic acid. Reduction of the protein (5-10 mg) was
carried out with 0.1; mercaptoethanol in a 0.14fl'TrisofiCl,
pH 8.6 buffer, under a nitrogen atmOSphere. Iodoacetic acid
(0.2 ml of 0.2g) was added to the reduced protein and allowed
to react for 30 minutes. The reactants were removed from the
protein by gel filtration on Sephadex G-75 which was equilib-
rated with 50% acetic acid. The acetic acid was removed from
the protein by rotary evaporation at 400.

The method of Crestfield gt gt. (l963)was also used to
prepare carboxymethylated protein with the exception that the
protein was not reduced prior to alkylation with iodoacetic
acid.

Oxidation of<1-GDH for amino acid analysis was accom-
plished by the method of Moore (1963). To 5-10 mg of protein,
2 m1 of performic acid solution (1 ml of 30% H202 plus 9 ml
of 88% formic acid) were added, reacted at 00 for 4 hours

and the reaction stopped by adding 0.3 ml of 48% HBr.

23

Heactants were removed with a rotary evaporator at 400.

Amino Acid Determinations

 

Samples for all of the amino acid determinations were
desalted on Sephadex G—25 and lyophilized to dryness. For
each analysis 5 to 10 mg portions of the dried protein or
protein derivative were hydrolyzed in 6Q HCl, in sealed tubes
evacuated to 50 microns. Hydrolysis was carried out at 1100
for 24 hours or more as deSired. After hydrolysis the HCl
was rapidly removed at 400 in a rotary evaporator, and the
hydrolysate dissolved in 2 mls of a 0.23 citrate buffer,
pH 2.2. Quantitative determination of the amino acids‘was

Performed on a Beckman Spinco Model 120 amino acid analyzer

according to the method of Moore and Stein (1954).

Peptide Mapping

 

All protein used for peptide mapping was reduced and
carboxymethylated in 8.53 guanidine‘HCl according to the
method of Crestfield gt gt. (1963). Trypsin (3 mg/ml) was
added to 5 mg portions of the carboxymethylated QFGDH sus-
pended in water, pH 8.6. The ratio of trypsin to carboxy-
methylated a-GDH was 1:50. A Radiometer automatn: recording
titrator, Model TTT-l/SBHZ-SBUl/TTA31 was used to monitor
the reaction by maintaining the pH at 8.6 by the addition of
0.01fl NaOH. The reaction was 90% complete after 10 minutes
of digestion and was stopped after 2 hours by adjusting the
pH to 2.8 with 1Q’HC1. The reaction mixture was frozen,
lyophilized and dissolved in a minimum amount of water. The

mapping of the tryptic peptides was carried out by the technique

24
of Katz gt__t. (1959). Chromatography was performed with
butdnol :acetic acid:water (4:1:5) and electrophoresis was
carried out at pH 3.7 in pyridine:acetic acid:water (1:10:289)
for 90 minutes at 2000 volts. The peptides were located with

0.2% ninhydrin or with one of several specific amino acid

sprays (given in the text--see Table VI).

C-terminal Analysis

The carboxy-terminal amino acids of OLGDH were determined,
using carboxypeptidase, by the method of Koorajian and Zabin
(1965). This method involved the use of sodium dodecyl sulfate
as a denaturing agent in order to make the carboxy terminal
amino acids more susceptible to attack by carboxypeptidase.
a-GDH (5 mg) was dialyzed overnight against 0.13,Tris-acetate,

0.013’MgC1 0.001fl EDTA, 0.5% sodium dodecyl sulfate, pH 7.7,

2,
then heated at 1000 for 10 minutes. Carboxypeptidase-A was
added to the protein (ratio of 1:20) and allowed to react at
room temperature for the desired time. Protein concentrations
were determined by measurement of O.D. at 280mu using an
extinction coefficient of 0.62 ml/mg cm-1 for a-GDH and

1.94 ml/mg cm-1 for carboxypeptidase-A (Neurath gt gt., 1947).
After various intervals of time (5, 10 and 90 minutes) the
reaction was stOpped by the addition of 20% trichloroacetic

acid (TCA) to a final concentration of 5% TCA. The precipitated
protein was washed twice with water, and the supernatant and
washings were combined and lyophilized. The dried sample was

dissolved in 1 ml of 0.23 citrate buffer, pH 2.2 and the amino

acids were analyzed on a Spino Model 120 amino acid analyzer.

25

Henaturation

 

a-GDH was centrifuged from an ammonium sulfate suspen-
sion and dissolved in the dissociation medium 7°2fl guanidine-HCl,
0.1g Tris'HCl, 0.13 mercaptoethanol, 0.001m EDTA, pH 7.5, to
give a final protein concentration of 1 mg/ml. After 2 hours
at 0°, 25 ul of the denatured protein was diluted into 1.0 ml
of the renaturing media at room temperature to give a final
protein concentration of 0.025 mg/ml and a final guanidine-HCI
concentration of 0.23. The protein was allowed to renature
for exactly 15 minutes at room temperature after which 10 ul
of the protein was assayed for enzymatic activity as described

below.

Enzyme Assay

 

a-GDH was assayed by following the decrease in absorbance
at 340mm using dihydroxyacetone phosphate (1 x 10'3M) and

NADH (2.5 x 10‘4g)

as substrates with 0.1E'TriS°HCl, 0.0013
EDTA, pH 7.42 as the buffering agent. A Beckman DU Spectro-
photometer with a Gilford Multiple Sample Absorbance Recorder
attachment was used; the recorder Speed was 1 in. per minute
and the sensitivity of the detector was 1 O.D. full scale.
Temperature was maintained at 250 by a constant temperature
circulator. An enzyme unit is defined as 100 divided by the

time in seconds required for a change of 0.1 O.D. per mg of

protein per ml of assay mixture.

Partial Specific Volume
The partial Specific volume (V) ofd -GDH was calculated
from the amino acid composition by the method of Cohn et al.,

“—

(1934). The partial Specific volume was determined

26
experimentally by the falling drop method of Barbour and
Hamilton (1926). The apparatus consists of a 50 x 1.5 cm
jacketed column containing a mixture of bromobenzene and
o-xylene, the density of which was adjusted to give a drop
time of 30-60 seconds. The temperature of this column was
maintained at 200 i 0.005. A 5 ul drop of the enzyme, prepared
as described below, was applied to the tOp of the column and
the time for the drOp to fall between two marks approximately
25 cm apart timed to the hundrdfifliof a second. The drop fell
approximately 15 cm before reaching the first mark in order
to allow for complete temperature equilibration of the drOp.
A standard curve of density versus reciprocal fall time, was
prepared using KCl solutions of known density (0.998-l.004
gm/cc). The fall times of protein solutions of different
concentrations were measured and their densities determined
from the standard curve. The partial Specific volumes were

calculated using the following equation (Schachman, 1957).
l_J—_d"do)

V = do x ( do

where V partial Specific volume
d = density of the solution
do= density of solvent (0.10003_KCl)
x = grams of protein/ml
The protein samples were prepared by dialyzing against
either distilled water or 0.10003 KCl for 48 hours at 40. The
protein concentration of the test sample was determined on a

separate portion by evaporating a known volume of a-GDH to

dryness in an evacuated P205 dessicator and subsequently drying

27
to constant weight in a 1100 oven. The contribution due to
the K01 solvent was accounted for and subtracted from the

total weight.

Sulfhydryl Titration

Sulfhydryl groups were determined by the method of Boyer
(1954) using parahydroxymercuribenzoate (PHMB) and following
the increase in absorbance at 250 mu. In order to insure
complete access of the PHMB to the sulfhydryl groups, the
titrations were carried out in the presence of 8g_urea, pH 4.6.
Guanidine°H01 was not used, although preferred because of its
greater effectiveness as a denaturing agent, due to its absorb-
ance at 250mu. Glutathione (l x 10'3fl) was used to standarize
the PHMB (3 x 10‘53) immediately before each titration.
Experimental O.D. differences of 0.001 were routinely measured
reproducibly with a Beckman DU monochromator equipped with an
absorbance indicator. Proper controls were run to account for
the absorbance of glutatione and Ol-GDH at 250mu.

PiWB was purified (Boyer, 1954) for use by dissolving in
a aqueous basic solution and precipitating by acidifying with
concentrated HCl. This process was repeated, the crystals
washed twice with distilled water and dried.

The concentration of OFGDH was determined by the method
of Lowry 33 a1. (1951) using bovine serum albumin as a standard.
For the analysis ObGDH was dialyzed for 24 hours against Sfl

urea, pH 4.6.

RESULTS

Physical Properties 3: Native a—glycerophOSphate
DehydrOgenase

 

 

An accurate knowledge of the native enzyme is a pre-
requisite for determining the subunit structure of chDH.
Due to the limited nature of the eXperiments reported in
the literature and to an inability to reconcile our results
with some of the values in the literature, a complete phys-
ical characterization of native a-GDH was undertaken. This
included determination of 330,w’ D30,w’ molecular weight
and partial specific volume. A major problem in the deter-
mination of the molecular weight of native a—GDH by sedi-
mentation techniques was the presence of high molecular
weight aggregates in native a-GDH. The first phase of this
research was directed toward finding a solution to this pro-

blem.

Preparation of "aggregate free” orGDH

The formation of aggregates was considered likely to
be due to the interaction of reactive groups as a result of
a slight unfolding, or to the formation of random interchain
disulfide bonds. When high ionic strength (1.0fl NaCl), or
co-factor stabilization (0.1fl’NADE 0.05m NADH) did not

eliminate aggregates, a-GDH was chromatographed on Sephadex

29

G-lOO or G-200 in the presence of a reducing agent (either
0.13 mercaptoethanol or 0.1g dithiothreitol (DTT)). Chromatog-
raphy on Sephadex G-lOO in the presence of 0.1fl‘mercapto-
ethanol gave unsatisfactory results. Although chromatography
on Sephadex G-200 initially appeared promising, aggregates
formed upon standing,even in the presence of O.lm_mercapto-
ethanol. This suggested the need for a reducing agent
sufficiently strong to maintain the thiol groups in the
reduced form. a-GDH, chromatographed on Sephadex G-200 in
the presence of O.lfl_DTT,gave satisfactory results. Since
the concentration of the protein eluted from the Sephadex
columns was low (ca. 1 mg/ml), water was removed by layering
Sephadex G-200 over dialysis casing which contained the
deaggregatedCX-GDH. In approximately 10 hours, a-GDH was
concentrated tenfold. This was kept in DTT and used for the
following experiments.
Molecular Weight gfpgjggg aleetermined by Sedimentation
Equilibrium

The results of a series of sedimentation equilibrium
experiments conducted on a-GDH obtained in this manner are
shown in Figure 1. Least square analysis of the extrapola-

tion to zero protein concentration data yielded a weight

c

average molecular weight of MW = 83,900. The individual
sedimentation equilibrium schlieren patterns clearly showed
a small amount of aggregate at the bottom of the cell. Fur-

ther evidence for the existence of significant amounts of

aggregate is shown by inspection of Figure 2 where it is

30

.Hs\ms m see .m .m .NH

.ma mzfisampsoo moamsmm exp :0 cosmogamm mummefihmaxm hhmdQSon capmgpshm
Scam pmzamppo modam> exp wcawwao>m an omsampno mm; SoapSHOm Moonm

mg» no Aoov Soapmapnmocoo Hmauasa one .mpsmaahomxo SSHHQHHHSwm Coapmp
Izmaavom map on Modem pswashmbo on>H0m mHSp pmsammm commamao Son»

was Qamponm one .mmhp smmzxo Eopmmm 03p Qmox on mode @903 msoszmo
toga cam mms on HOHHQ QmmOHpHQ Spas cowhsm mangOHOSp ohm: poms mpsmb
IHOm HH< .psmeaom godpsao 0:9 mm m:.m mm .Hom.maha.ma.o .meom flaoo.o
.Hcez mm.o .HopoeeeSpoHneae 2H.o anew: oomuc ameesemm :0 eeeaeamoees
IOMSo mm: mDoJc .mmpmwmnwmm pgwamz amazooaoa swan obosmh cu Romeo SH
.N\nno+aov mm ompmdam>m who: mnofipmapnoosoo .om pm paw Ema Hem.m mo
commm m pm Edaanaaaswm soapmpnoaaoom an Um£HEHmpmo mama mpsmflos HmHSO
ImHoB psmsmaa< .Qoapmhpzmonoo Samponm chow on moolxco>apmn mo mpswamz

Hmadomaoa fizzy mwmnm>m unwamz unommmam exp mo Goapmaommhpxm .H mhsmflm

APPARENT MOLECULAR WEIGHTXIO-B

35:03: ZOrEKHZmGZOU

m m m . e o
_ i _ _ _ a _ _

 

ooT oommm... e 2 no.

to
00
I
o
\
S
i“

E

 

9

r0
LO
1
l

IQOlO m>_._.<z

 

 

_ _ _ c _ _ _ _

 

32

.mshwcm exp mo goapmwmhwmm

HaapsmmeSm mpmofidza H8\ms NH mbonm mnoapmHQSmoqoo Camposg pm ANS\H

mo mosam> Soav N2 mo mmSHm> swag one .maampmo HmemEHnmaxo mom H masmam

ca Gnowma exp mom “a oHSMHm Ca onsm omOSp mm mpsoaflammxm csmw mgp 809%

ohm mpmw one .mmmomadm somaamano mom mpsmfimz amazooaoa NS osm 3: Spon
mm©H>0Ha mammag mane .Qofipmnpsmocoo Sampoam 09mm on mpswamz HmadomaoE ANEV

mmmao>mnw Una Azzv mmmao>m pswamz psmammmm esp mo zofipmaoamnpxm .m madmam

33

3282 ,zofiqEzmozoo
m m

 

O
Q

[0
CO

APPARENT MOLECULAR WEIGHTX I63
[‘0
L0

[C

 

_ H) _ fl

N20
320

JOEMEIHQICQ Z. IDOLO m>:.<z

_ _ _ _

 

 

 

0..

W
QOIXT

34

seen that most of the z-average molecular weights at a given
concentration are much higher than the corresponding weight-
average molecular weights. As a result the weight-average
molecular weight of 83,900 must be considered a maximum.
Although there is a significant amount of aggregation in DTT,
it does not approach the amount of aggregation observed in
0.13 mercaptoethanol. In further contrast to the results
obtained in 0.1g'mercaptoethanol, additional aggregation did
not occur rapidly upon standing for several days. The
evidence from the experiments above supports the theory that
the aggregates in native N-GDH arise from the formation of
random interchain disulfide bonds. Since the presence of
aggregates prevented an unequivocal determination of the
molecular weight of native d-GDH using sedimentation equili-
ibrium techniques, we turned to an analysis using a combination
of Sgo,w and 330,w° This method is much less sensitive to
aggregatior since the sample is freed from aggregates in the
sedimentation velocity experiments. This is because the

aggregates sediment faster than the main protein boundary.

 

Determination 9;; £30“, 230””, and gwgsflgl 9;; native a-GDH
A series of sedimentation velocity eXperiments were carr-

ied out in parallel with the sedimentation equilibrium ex-

periments, using the same stock enzyme preparation which had

been deaggregated on Sephadex-G-ZOO in the presence of

0.13 dithiothreitol. For comparison, two values are included

which were obtained using 0.1Q'mercaptoethanol rather than

DTT as the reducing agent. As shown in Figure 3, there was

35

.om pm 899

omu.mm pm ooposmsoo ohms mpsosahomxo mpaooao> soapmpcoEHoom .pzmfinhobo
oonHmfio ohms .psomm msHoSooH 02p on coapaoom EH .m:.m mm .gemm Naoo.o
Joanie .36 .32 mama maafiooaoo .noaaenn fiom .Anoaofio Bait
HoQQSpoOpamcaoepmH.o wQHQHmpsoo uno>aom m SH oobaommao mmz oaaamm osooom
< .AmoHOHHo noaov HOpHmHSpoHSpHo wa.o mo monomohm map ma oomlo Mop
lasaom no poSQmHmOmeOHzo mos oaaamm oco .Qoapmhpsoosoo GaopOHQ oaoN 0p

mmolo O>Hpms mo mpsoaoammooo Coapmpsmaapom no Qoapmaoamapxm .m omswam

36

35:02 V 20_._.<m._.Zm0200

v

m

N _

 

 

_ _

1.0._._m_mI.r0_I.:0 5:0 0
AOZ<IPMOHQ<0¢ME 5:0 .-

Iooio M222 no 2.on 1

_ _

_

F

_

 

_

_ _

meme "ouom

8

0.
m

o.
q.
(M038) lNBlOIddBOO NOLLVlNBWICJBS

 

_

 

37

no significant difference between the two experiments.
Extrapolation of the apparent sedimentation coefficients
to zero protein concentration (see Figure 3) yields a sed-
imentation coefficient of Sgo,w = 4.863 which is identical
to that reported by Van Eys gt a1. (1959).

Since the determination of the molecular weight required
both sedimentation and diffusion data it was necessary
to determine the diffusion coefficient of native<1-GDH.
Extrapolation of the apparent2 diffusion coefficients to
zero protein concentration yielded a diffusion coefficient
(see Figure 4) of D30,w = 6.20 x 10-7cm2/sec, which is

independent of protein concentration. Using the value of

So 0
20,w 20,w

for the diffusion coefficient a molecular weight of 74,400

= 4.868 and a value of D = 6.20 x 10-7cm2/sec
is obtained. Using this molecular weight of 74,400 and a
partial specific volume of 0.746 cc/g, a frictional ratio

of 1.23 is calculated. It is obvious that this data is
internally consistent, in contrast to the data reported

in the literature. This will be shown later in the discu-
ssion. Furthermore, assuming the molecular weight of 74,400

the sedimentation coefficient and diffusion

 

2Apparent simply denotes that the quantity is measured
at a finite concentration and depends on concentration.
This is in contrast to the value obtained by extrapolation
to zero protein concentration, wnich is assumed to be an
intrinsic property and thus not a function of protein
concentration.

38

mx> m WVM . E “moapmavo

mmonowbm ocp scam oopmadoamo macs mpswflo; amazooaoa AQ\mv32 .sah oom.m
no woman a wsamd om hampMSHNOHan pm doahomhoa ohms muscafinoaxo ose

.ma.a ma .<Hom mHoo.o .Hooz mm.o .Hom.maaa_mH.o .Hoacnpoooaooaos.ma.o

pmsamwm masos 3N oonHmao mm: QHopOHQ ose .zoapmeSooQoo :HopOHQ chow

on moansoo>apmz mo wpnoaoammooo scamswmao on» mo Scapmaoamhpxm .3 oaswam

39

N_

33: 0 S: 20_._.<E.Zm0200

w

v

 

 

a

 

 

:86 H.222 do some

F

O

_

. a >>.ON
70mm cause. x 0mm - coo

_

a

'7

co

_

_ ,

J;
W.
q-

l l l
C 8 g 8
(M030) lNI—JIOdeEIOO NOISrL-HICI

I
LO. ,
no

 

 

40

coefficients both lie close to the curves expected for
globular proteins, which is indirect evidence for their

accuracy (see Figures 5 and 6).

Partial Specific Volume

 

As stated in the introduction, the literature value
of 0.70 cc/g for the partial specific volume, gave am-
biguous results for the subunit analysis ofcx-GDH. Using
this value a molecular weight of 26,000 was obtained for
the Q-GDH subunit, thus indicating that a-GDH was composed
of three polypeptide chains (see the subunit section).
This result does not agree with the 2 moles of C-terminal
(see Table I) or 2 moles of N-terminal (Van Eys gt al.,
1964) amino acids found per mole of OEGDH, which indicated
2 polypeptide chains per native enzyme unit. For this
reason and because a precise value for partial Specific
volumevms necessary for the calculation of reliable
molecular weights, the partial specific volume ofcx-GDH
was determined by two independent means: (1) a direct
eXperimental measurement using the falling drOp method
of Barbour and Hamilton (1926) and (2) an indirect, em-
pirical determination calculated from the amino acid
composition (Cohn and Edsall, 1943).

The initial determination of the partial specific
volume was conducted with OPGDH dialyzed in distilled
water. It did not yield satisfactory results at the lower

protein concentration (see Figure 7). Since protein-

41

.Amoosoaomoh

no pmHH How xaocomam oomv chapmhopaa mcp Scam mozfimpno who: mpgwamz
amazooaoa ozm mpsofloammooo cofipmpsmanom pom modam> AIIIV HmOHHHmam
.AmSOprHSOHmo Mom Nfiosoaaw oowv w\oo mm.o mo mEdHo> camaoomm Hmaphmm
m spas monosam pom ompmadoamo cams mpsoaofiwmooo coapmeoEwoom alllv
Hmoapopoose .msaopohm amasnoaw How mpswaoz HmHSooHoe mo :oHposdm w mm

mpCoHoammooo Soapmpsoaaoom HQOHHHQSo ocm HmoapohooSp mo wsmmmo .m ohswam

)1")

one” x Bonus €30.39
oooa cos pom- com com ooa E. on on om

Jfi 1 71.3.1}? . _ _ 111.41 a M 2%,:
l g _ _ m l . l w . . _ x . . , a. I .
a, _ . _ l m a
k l . , .. .

 

 

_
_

n

.
a

a

 

 

 

 

 

 

.L. 1---... ~.__.... _._. '
.

 

. Guacamoaufl O \\
00533.8»on magnoamououaamlfl . O

unnaocu

annexe}... \ \
. flag a 0d . 0x a
. L Hm. SHUHOWMHH . O\ _ . Hnuwuwgm AIIIV
\ auowuouoona Ally
(awash; -- 0 \ 1
. mg magnumm: 0V x i ,
4 0.33.6 W \\O L
and». 0
weaning“

_ .\
030g _\.\0 m i H .
Tnuvmznuaoam gnome nos 92
«I'VE—«Ema «on 95303300 dam

0H

              
 

S
5‘03
( 8) announce nomvmmmas

Ln
H

O
N

43

.Amoosohomoa mo pmaa Hog Napgommm oomv chapmmopfia oSp Boa“ oosamppo
who: mpsmaoz Hwadooaos cam mpmoaoamuooo scamdmmao mom mozamb AIIIV
HmOHHHQEm .AmCOHpdeoaco Mom Kaocoamm oomv w\oo mw.o mo mESHo> camfi
loogm Hmapmmg m Spa: moaosam Mom oopdafioamo ohms mpcoaoawhooo Goamzwmflo
Alllv Hmoapmhoose .manpoya Hmadfloam mow magmaoz Hmadooaoa mo noapossg m

mm mpsoaoammooo CofimdmmHo Hmoaaameo cam Hmoapohoosp mo mgmmnw .0 oasmfim

44

ha on

 

 

 

 

 

 

 

 

 

 

.1 1 1 -1 :4 -11 ., 1 1 1-11 WE 1 1 -
1 1 1 11 .............. . . 1 H
a 1 ..... ..... 1 i 1 -1-
: 1 1
1..... .. ,, 1 , . 311, ..... - u ....... 1 M 1 i . 1 n

 

1.0 1 0.3.1.8. 2.2.3.312 . .. . . .1 1. 1 1
I , 0 tag: 325...?— 1 .. . .
[Ly...1 . ‘UEN. 11‘? 1 r. .1111 . .. .. . . ... ..1”. . . . . *—

1311 ..,. , .
Io/T 3&0 833.2 1 1 M

.1 f 1 .. [11/5 -11-...0m1uufocolour

T3115-..

 

 

 

       
           

{5.095133 wa§uozmd~8-¢cd

 

 

 

1
1 file/1832i
. , 1 .1 I01. I [8303033 . . . p
_ gangs
1 33303336. m
. . . . . a . . ,1 .1; . I 41...! .1. . u Q
1 1 1 1 . . .71.-.. .1 .. . 41w? Boeing? 25.0
111.. 1.1 HIDMHH’H AIIIV . i 1. 1a. L 1 1 1 4., 5 am . . OH
1 Guam-33.1 1'1 1 H H . 1 . 1 l
1 .

 

. . . 1 . 1 . 1 .1 1 1 . _
1 . 1 L 1 . . 1 .LthLa 1-1-5111. 1 1
Tissue: «453.8 3.. a!
TIER-nun «on Soup—n8 £3

 

‘
(A “30) muomaoo uorsnum

.mmmShhv on

memmmm xOOpm mgp go Coaphom m we nofipmhomw>m an UmQHShopmv 03m:

45

moahmm Scam CH Coflpsaom XOOpm Hmsfimfiho ms» mo msoapmeSmozoo Sampopm
.mHSOS m: Mom me>Hom mpmHHQOMQQm asp pmsfimmm cmmmamfiw mwz seas;
soapsaom thQO xooum m Scum meme mum: mcoapdaflm .Ammaohao :mmov
Hom.mdooa.o 62w AmoHOHHo cmwoaov Mmpmz “mpcmbaom pCmHom%H© 03p 23
©m>aommflw SampOMQ so muwe mum: wpszmMSmMmz .mewld mo va mESHo>

oawaomam Hmaphmm pcmhmgam m5mhm> :oHpmesmozoo CampOHm .5 ohsmfim

1.1125121 2012115828

 

 

 

 

m v N
1 1 1 1 _ 1 1
O
Jomd
10x 02 o
Immd
181 121.0 0 o
0 [one
0
O O
0
HI 0 o o. o + o \mfifio
98 550 u. 05>
IQQLO m>_._.<Z
_ _ _ _ _ _ _

 

(ED/OOBWD'IOA OHIOBdS WVILHVd

47
protein interactions are known to often be a significant
factor in the absence of salt, the V measurement was
repeated using 0.10003 KCl as solvent. Under these
conditions precise measurements of V'were possible and an
average value of 0.746 oc/g was obtained for the partial
specific volume of apGDH (see Figure 7). It is also
seen from Figure 7 that V is essentially independent of
protein concentration.

McMeekin and Marshall (1952) have shown that the
partial specific volume of proteins calculated from the
amino acid composition agreed very closely with the
values obtained by direct experimental measurements.
Therefore the partial specific volume of a-GDH was cal-
culated from the amino acid analysis listed in Table I.
The partial specific volumes for the individual amino
acids were obtained from Cohn and Edsall (1943) and the
partial specific volume of cysteine was obtained from
McMeekin §t_§;. (1949). The value of 0.746 cc/g obtained
by this method (see Table I)agre&ivery closely with the
experimentally measured value of 0.747 cc/g and indi-
cated that the value of 0.70 cc/g indicated by the data
of Young and Pace (1958a) and the value reported by
Van Eys gt a; (1959) was in error. A partial specific
volume of 0.706 cc/g was also calculated from the amino
acid composition of dpGDH reported by Van Eys 22.2}!
(1964). The excellent agreement between the values for

V determined by these two widely different methods argues

48

Table I. Partial Specific volume as calculated from amino
acid analysis8

 

 

Number of (% by weight Specific (% by

 

Amino acid residues/ of residue) volume of volume of
residue molecule residue residue
Lysine 55 9.28 0.82 7.610
Histidine 18 3.25 0.67 2.178
Arginine 16 3.29 0.70 2.303
Aspartic acid 20 3.03 0.59 1.788
Asparag‘ineb 36 5.40 0.60 3.240
Threonine 28 3.73 0.70 2.611
Serine 21 2.41 0.63 1.158
Glutamic acid 46 7.81 0.66 5.015
Glutamineb 36 6.06 0.67 4.060
Proline 31 3.96 0.76 3.096
Glycine 77 5.79 0.64 3.706
Alanine 67 6.27 0.74 4.640
Valine 62 8.09 0.86 6.596
Methionine 15 2-59 0.75 1.943
Isoleucine 55 8.19 0.90 7.371
Leucine 58 8.64 0.90 7.776
Tryosine 10 2.15 0.71 1.526
Phenylalanine 32 6.20 0,77 4,770
Half-cystinec 21 2.85 0.63 1.796
Tryptophaned 4 0.98 0.74 0.725

Total 74.63

 

aSee text for details.

bThe 72 residues of ammonia were divided evenly between

glutanine and asparagine.

CThis value was determined separately, both as cysteic
acid and as S-carboxymethyl cysteine (See Table III).

d

‘Taken from the data of Van Eys gt al.(l964).

49
strongly for the accuracy of the value of 0.746 cc/g.

Analysis for Disulfide Bonds and Sulfhydryl Groups pf g;g2g

In order to evaluate the presence or absence of di-
sulfide bonds incx-GDH, the two quantities which are need-
ed are: (l) the total half-cystine content and (2) the
number of free sulfhydryl groups; the number of disulfide
bonds is then the difference between the two values.
These values were obtained by measuring the sulfhydryl
content of unreduced and reduced protein; the sulfhydryl
content in the former case corresponds to the number of
free sulfhydryl groups and in the later case to the
number of free sulfhydryl groups plus twice the number
of disulfide bonds. The sulfhydryl content values were
obtained by two methods namely, (1) titration with para-
hydroxymercuribenzoate and (2) by measurement of S-carb-
oxymethylcysteine formed by reaction of a-GDH with 10do-
acetic acid.

PHMB titration.- The PHMB experiments were con-
ducted in 33 urea, pH 4.6,to expose the sulfhydryl
groups to the PHMB. The results are shown in Table II.
If disulfide bonds were present in<x-GDH the reduced
protein should have a greater number of titratable sulf-
hydryl groups than the unreduced protein. Unfortunately,
the results were neither consistent nor reasonable; a
greater number of sulfhydryl groups wenafound in the
unreduced samples (experiments 1 and 2) than in the

reduced samples (experiments 3b and 4). However eXperi-

50

Table II. Titration ofchDH with PHMB

;
i

No. of —SH groups per

 

Treatment Exp. No. 76,000 g protein
None(ie., unreduced) 1 19.0
2 20.4
3a 15.2
Reduced and mercapto- 3b 15.1
ethanol dialyzed out 4 15.4
Carboxymethylation 5 4.8

followed by reduction

 

ments 3a and 3b, which were done in parallel using the same
stock solution, raised the possibility that there was no
appreciable difference between unreduced and reduced pro-
tein. A direct measurement of the disulfide content of
a-GDH was achieved by measuring with PHMB the sulfhydryl con-
tent of a-GDH carboxymethylated prior to reduction(Experiment
5). This suggested thatcx-GDH contained 1 or more disul-
fide bonds. The explanation for these results is not known;
however one possibility is that when the enzyme is unfolded,
as it is in urea, that very reactive sulfhydryl groups are
exposed which then react to form disulfide bonds. Since

the results obtained by PHMB titration were contradictory

and did not allow unequivocal conclusions concerning the
presence of disulfide bonds in native a-GDH, it was nec-
essary to resort to more refined and more reliable chemi-

cal techniques. Carboxymethylation with iodoacetic acid was

the technique chosen for additional experiments.

5l

Carboxymethylation of d»GDH.————-The aim in this
series of experiments was to determine values for the
total number of half-cystines and free sulfhydryl groups
in.0«GDH by measurement of carboxymethyl cysteine, formed
by reaction of reduced and unreduced a—GDH with iodoacetic
acid respectively. Values of 20.8 and 15.9 (see Table III)
moles of carboxymethyl cysteine per mole of protein were
found for the reduced and unreduced protein respectively.
This obviously suggested the presence of two disulfide
bonds in the sample under study. But again the variability
(12, 17, and 19 sulfhydryls per mole) in the experiments
on unreduced, carboxymethylated protein cast doubt on the
results. We thus turned to still another method, per-
formic acid oxidation, to provide an analysis for total
half-cystine, in an effort to firmly establish this
value.

Performic acid oxidation of d—GDH.-—-—-Analysis of
the cysteic acid produced by the performic acid oxidation,
yielded a value of 21.4 moles of half-cystine per mole
of protein, in very good agreement with the value of
20.8 obtained from the reduced and carboxymethylated
protein (see Table III). These data established firmly
and reliably the half-cystine content of d-GDH at 21
moles per mole of protein, but the possible existence
of disulfide bonds still remained1x>be answered.

Combination of carboxymethylation and performic

 

acid oxidation. The following experiment was de-

Table III.

52

bonds of a-GDH.

Analysis for sulfhydryl groups and disulfide

 

 

 

Derivative Number of

Treatment measured residues founda Ave.
performic acid cysteic acid 20.6 21.4 22.1 21.4
oxidized

reduced, car- carboxymethyl- 20.9 20.6 20.8
boxymethylated cysteine

carboxynethyl- carboxymethyl- 11.7 16.7 19.4 15.9
ated cysteine

carboxymethyl- cysteic acid 7.2 8.1 7.7

ated, performic
acid oxidized

8based on 56 aSpartic acid residues per 76,000 grams of
protein.
signed to provide a direct measurement of the disulfide
bonds present in a—GDH. The first step was to carboxy-
methylate the protein in 8.5g guanidine'HCl and the second
step was to oxidize the carboxymethylated protein with
performic acid. All free sulfhydryl groups should have been
carboxymethylated in the first step and all disulfide
bonds oxidized to cysteic acid in the second step. 0f
the two products obtained by this procedure one, cysteic
acid, should have provided a measurement of the disulfide
bonds present in.a-GDH, and the other, S-carboxymethyl
cysteine sulfone, should have provided a measurement of
the free sulfhydryl groups. The latter is produced by
the oxidation of the S-carboxymethyl cysteine. Following
such treatment (See Table III) a-GDH gave rise to a

considerable amount of cysteic acid (8 moles per 76,000

53

grams of protein) suggesting the presence of 4 disulfide
bonds. However examination of the structure of carboxy-
methyl cysteinesulfone suggested that some or all of the
cysteic acid might have been formed from carboxymethyl
cysteine sulfone during hydrolysis of the protein in

63 HCl, via the following reaction.

 

H00? 9 110° Hooq
HF-CHZ-S-CHZCOOH + H20 7 HQ-CHZSOBH + CHBCOOH
ZHN 0 6y. HCl ZHN
carboxymethyl cysteine cysteic acid
sulfone

To test this possibility, a mixture containing equi-
molar quantities of carboxymethyl cysteine and cysteine.
HCl was oxidized with performic acid. A portion of the
mixture was incubated in 6y HCl at 1100 for 24 hours,
while a second portion remained at room temperature.

Both samples were than analyzed for cysteic acid. The
sample which had been subjected to HCl hydrolysis had
only 1/3 as much sulfone as the sample which had not been
heated in 63 HCl, thus substantiating the theory that
hydrolysis in 63 HCl at 1100 for 24 hours results in a
breakdown of carboxymethylcysteine sulfone with the
possible formation of cysteic acid. As expected, some
increase in cysteic acid was also observed. The experi-
mental approach just described was thus unable to evaluate
the existence of disulfide bonds. With this experiment
this phase of the chemical analysis was terminated. Al-

though the collective data (See Tables II and III) indica-

54

ted that disulfide bonds did exist in the particular
samples analyzed, this did not prove that these disulfide
bonds were an integral part of the original native urGDH
molecule. For example, the disulfide bonds could have
been formed in the treating and handling of the enzyme. In
conclusion the chemical analysis had provided a reliable
value for the half-cystine content but was unable to
unequivocally evaluate the possible presence of diSulfide
bonds. A final answer to this question was later obtained,
however by molecular weight analysis of enzyme dissolved

in a carboxymethylating dissociation media. This is des-

cribed in the subunit section.

Amino Acid Analysis offzzggfl

As mentioned previously an amino acid analysis was
performed in order to calculate an empirical value for the
partial Specific volume of erDH and to obtain a detailed
analysis for the number of disulfide bonds and sulfhydryl
groups present in d-GDH. The results of a total amino
acid analysis are shown in Table IV. Because the values
obtained for aSpartic acid remained constant during
hydrolysis, the values for the other amino acids were
based on aSpartic acid, which was set equal to 1.000.
The absolute values for the the individual amino acids were
obtained by extrapolation of these ratios to zero time
of hydrolysis for amino acids destroyed during hydrolysis

and by appropriate extrapolation to the maximum value

55

A:

00Hv.mm um was 00> as emsasppo osass

D

Campoaa UmNfiCHNo UHom

oaeaomamm pom ompmeSposaonamo mo mamzamsw no woman “HHH mapme Scam coxmp mpHSmomm

 

: a

0m am
00 mm
m 0H
00 mm
00 00
0a ma
:0 N0
00 00
00 00
mm Hm
mm mm
am am
0m mm
00 00
ma 0H
m0 m0
ma 0H
00 mm

ll llAqmmHv sflmpOHm

.Hs pm mam ss> an a 000.00
Hog odeamom

oocfiwppo modam>

mom.o
m0a.o
mmo.a
:m0.o
dwm.o

HOH.H
m0a.a
:00.H
000.0

Hms.a
000.0
m©:.o
000.H
mmm.o
HON.H
mmm.0
m00.0

oSHmb

popmaoamapxm

 

 

 

 

l1

 

 

 

0mm.0
0ma.0
000.H
000.0
00m.0

mmH.H
moa.a
mmm.a
msm.0

mm:.a
smm.0
0m0.0
000.H

m0m.0
00m.a
0mm.0
000.0

mm

m0:.0
sma.0
000.H
000.0
H0N.0

00H.H
0mH.H
ems.a
0am.0

00:.H
H0m.0
Hms.0
000.H

00m.0
mam.a
mam.0
000.0

mm

0am.0
00H.0
000.H
000.0
mmm.0

m00.a
sma.a
0:0.H
Hmm.0

mm:.a
mam.0
000.0
000.H

H0m.0
00m.a
0mm.0
000.0

m:

mmm.0
00H.0
H00.H
000.0
000.0

mmo.a
mea.a
:00.H
mmm.0

00:.H
000.0
mes.0
000.H

00m.0
mmm.a
0mm.o
:m0.0

3N

Amazosvwamaaoapzs mo mafia

aye

Qmao

.m

osm
nae
Sod
omH
p02

Hs>
«ht—”Mm
saw
oam

Saw
mom

9:8
am¢

ma<
mfisofis<
mam
mad

sacs
oQHE<

 

 

 

1"

llll

 

Ii

mawhaohoms ho mmEHp msaaam> pm maolnvmo mawhamcm macs OQHEM proa

.>H mHnt

56

for those whichwere incompletely hydrolyzed after 24 hours.
The use of leucine or arginine as standards gave essent-
ially the same results as aSpartic acid. Correctionsfor
losses were made from observations of 24, 48, 72 and 96
hour hydrolysates ofcx-GDH. Losses during hydrolysis were
significant in the cases of serine, threonine,(Figure 8)
and tryosine. The molecular weight forCX-GDH calculated
on the basis of 56 residues of aspartic acid per mole of
protein is 76,000. The agreement is good between our
results and the results of Van Eys gt al.(1964), which

are shown in the extreme right column of Table IV.

Carboxy-terminal Analysis of a-GDH

 

As an independent means of determining the number of
polypeptide chains in the native molecule of a-GDH a
kinetic study of the appearance of C-terminal amino acids
with time of digestion with carboxypeptidase was under-
taken. Prior to digestion with DFP treated carboxypep-
tidase-A, a-GDH was thoroughly denatured by heating at
1000 for 10 minutes in 0.5} sodium dodecyl sulfate. The
digestion was stopped after the indicated time periods
by precipatating the proteins with trichloroacetic acid;
the amino acids in the supernatant were concentrated and
analyzed on an amino acid analyzer. After 5 or 90 minutes
of digestion (See Table V), methionine was the only
amino acid released in sifnificant amounts. These results

are consistent with a-GDH being composed of two poly-

57

.ofiom capamamm Hog ooo.H mo mzamb esp no woman was moHpma egg .OOHH pm Hum
mm msflsflmpsoo moDSp meadow copmSom>o CH pofiaomama mm: mamaaoapam .meIn

no mamaaopwhs Uflom moaadp osasooazp Una osaaom mo :oHpOSHpmoQ .w madmam

58

AmmDOIv m_m>JOmQ>I do 92:.

vm

 

00 ms 00
_ m

w

 

.O/

 

_ _ _

_

I 0 Ln.“ v.0
52801: /

Omd

mfio

Odd

Omd

 

 

(0001 =- GIOV OllHVdS-V)OLLV8

59

peptide chains both of which contain a methionine residue
on the carboxy-terminal end. The fact that no other

amino acid is released after extended digestion with

Table V. Release of C-terminal amino acids with time.

 

 

 

 

Minutes of Moles amino acid per 76,000
digestiona grams of protein

Amino

Acid 5 10 90
Serine trace 0.24 0.46
Glycine trace trace trace
Methionine 1.86 2.11 2.03
Leucine trace 0.23 trace

 

 

3ratio of carboxypeptidase-A to<1-GDH was 1:20.

carboxypeptidase-A suggests that the amino acid adjacent
to methionine is either lysine or arginine, for carboxy-
peptidase-A will not attack either carboxy-terminal lysine,
arginine, or proline and may be blocked by glutamate or

aspartate (Canfield and Anfinsen, 1963).

Fingerprinting g: ggggi

Amino acid analysis of a-GDH indicatecia total of 71
lysines and arginine residues per 76,000 grams of protein.
Assuming that trypsin is Specific for lysyl and arginyl
bonds, one would eXpect to find 72 peptides if chDH

is composed of subunits which have no repeating amino

60

acid sequences, either in the same or in different poly-
peptide chains. On the other hand if the protein is
composed of polypeptide chains whose amino acid sequences
are wholly or partly unique, the number of peptides ob-
tained will be some fraction of 71. The results of a
typical fingerprint, obtained as described in materials
and methods, is shown in Figure 9. A maximum number of
25 peptides was obtained with 22-23 being commonly found.
Other peptide patterns obtained in a similiar manner
were analyzed separately with specific spray reagents

for peptides containing tyrosine, histidine, and arginine.
As shown in Table VI, the number of peptides containing

each of these amino acids was in all cases approximately

Table VI. Staining of peptides for specific amino acids

 

 

 

Spray reagent No. of spots No. of amino acid
residues/ 76,000 g

of aPGDH

Ninhydrin 20-25 71 lysine + arginine

Diazotized sulfanilic 5-6 18 histidine

acid (Block, 1951)

a-Nitroso-b-naphthol 3 10 tyrosine

Acher and Crocker (1952)

8-Hydroxyquin011ne-sodium 5-6 16 arginine

hypobromite (Jepson 22 al.,

,1953)

 

one third the number of residues present in the native
molecule, which is consistent with the theory that d—GDH

is composed of at leaSt two and a maximum of 3 identical

.mcflsflmam

sasssoo 0H 0:0 0H .HH .0 .0 mmaapsms 020 “ssa0asmas mm 000 0H .0H .0H
.0 mmzamoaap Campsoo ma was m .3 mocflpaom .mopdSHS OOH Mom mpao>
000M pm Soapomnac ampsomfiao: osp CH poSHogama was mflmoaosooapomam
.psobaom mm Amnaudv Hops: «UHow 00pmom “HonpSQ mcfimd noflpoohflp Hwoap
lamb onp QH p50 coaahmo mmz 0SQMHwOmeoaso .mmwlnvuopwH0Spofihxophmo

pecanma mo pmomflp QHmQ0Hp w 00 Ads mpapaom pmpodhpmsooma m .m oaswam

62

 

L
\

AHdVHOOlVWOHHO

 

® / @mmmOIdOmFowmm ®

V

 

 

©»\Z_0_KO
@ M: Q
0. ®®
90 001 -. ,
Q Qm n.\ @300 0
® ® Q

Q
a Q

 

100-5 SE do mmocfimd .,s._..r_>cx_

 

63

polypeptide chains. It seems likely that extensive tests of
other chromatOgraphic solvent systems would have revealed con-
ditions for better separation and thus yielded more peptides.
Since it is likely that some peptides are very similiar and
were not separated by the conditions used in the fingerprint-
ing and due to the limited nature of the experiments, the num-
ber of peptides detected by this method will be a minimum.

Subunit Structure gfci-glycerophosphate
Dehydrogenase

 

 

 

As stated in the introduction, a principal objective of
this research was to determine the number of polypeptide
chains present in the native moleculae ofcl-GDH. This first
phase of the research involved finding a dissociation sys-
tem which produced stable subunits. When this was success-
fully accomplished, a complete physical characterization of
the d-GDH subunits was undertaken. This included determina-
tion of the sedimentation coefficients, the diffusion coeffi-
cient and the molecular weight of the dyGDH subunits by both
sedimentation equilibrium and by combination of the sedimenta-

tion and diffusion coefficients.

Attempted Dissociation ngi-GDH in Urea

 

The results obtained in the carboxy terminal and in the
fingerprinting analysis of d-GDH suggested that dnGDH was
composed of two, or at most three, subunits. The next ob-
jective was to find conditions which would dissociate the
enzyme into subunits. It had been shown previously by

Deal 23 a1. (1963) that the eXposure of ahGDH to acid

64

(0.0lfl citrate, pH2.6) was ineffective as a means of dis-
sociating the protein. Since urea is a convenient and
inexpensive reagent which is often sumnssful in dissociat-
ing proteins, its dissociating ability was tested on a-GDH.
a-GDH in Sl'urea aggregated even with the presence of

the reducing agent mercaptoethanol(0.1fl). This made
accurate conclusions from the sedimentation equilibrium
experiments impossible and necessitated consideration of

other dissociating agents.

Effect gfiGuanidine°HCI Concentration 93_pgg_§20,u_q§1gjggg
From the previous section the need for a denaturing
agent stronger than 81 urea was obvious. Since guanidine-
HCl was the strongest protein denaturing agent available,
it was tested for dissociation ability. Preliminary
experiments indicated that 7g guanidine’HCl dissociated
abGDH into stable subunits, if 0.1fi mercaptoethanol was
present. The experiments which follow were designed to
determine the minimum guanidine-H01 concentration at which
dissociation occured. For this experiment the sedimentation
coefficients of a-GDH in various concentrations of guanidine°
HCl was determined (see Figure 10); the solutions also
contained 0.13'mercaptoethanol. The protein concentration
was 6 mg/ml for these experiments; further experimental
details are given in the legend for Figure 10. Eval-
uation of the experiments at guanidine'HCl concentrations

of 0.5g and lfl'was impossible because the enzyme pre-

65

.HosmzmepamoHoE m:p one mpamw mgp on 630 0pamooma> Hmsoapapww ms»
pQSooom ouca MQmep Ammoav paoyzme 0cm mamsmzmm mo mums msp Sop“ pogae
lampop mama onHpSHow Hom.msapazmzm cs» mo moapamooma> .EQH ow0.©m
mm: mcfippmm uocdm pouch 0:9 .maso: NH mo ESFHQHE m pom comaamflp cams
moHQEmm 0:9 .HS\mE m was nofipmapcoosoo anpoaa exp and omlm pm mmop
who: mpsosfiaoaxm Hag .Hom.osawfismsm mo noapmapscosoo vendoapsfi exp
030 .004. m0 .Hossssmopssosss H.420 .30.. 04200.0 .802 0W0 082.3:

mfi.o mm: mpnmsfiamaxo Ham SH pom: uno>aom ose .mmwlo mo pnmflofimmmoo

GoapmpSmsHUom 03p :0 Soapmapsmocoo Hom.msfipasm5m mo pommmm .oa oaswam

66

>.._.. m 4402 .01 $2.924 D o

N m m w m N .

 

 

_ J _ _ a _ _ _ _ _ _ a, _ _

1
//O

IQOLO do ZO_._.<_oOmm_O l

<r M N
33) iNBIOIjleOD Nouvmswnoas

DU

(

 

__,_T_ _F_ _ _ _ _L_

 

M‘0
0/09'

67

cipitated from solution under these conditions. The

sedimentation coefficient of dpGDH in 23 guanidine-H01
0.6%

W88 820 ,W

corresponding value of s

= 3.68, which was a significant decrease from the

0.6%
20,w

The nature of the 3.68 species is not known but it may be

= 4.28 for the native enzyme.

due to a slightly unfolded native molecule or may represent
a rapid equilibrium mixture consisting of the native enzyme
and theenbunit. Dissociation of a—GDH occured at concentra-
tions of guanidine.HCl 3E and above. The slow decrease in
sedimentation coefficients in the guanidine.HCl range of 3 to
7m suggested that some unfolding of the enzyme continued to

occur as the guanidine.HCl concentration was increased.

0 0 HO
£20,w’ 220,w’ and LW(S{D) g£_the apGDH Subunit

The next phase of the research involved determination

 

of the molecular weightcfdpGDH by two independent
methods, namely (1) sedimentation equilibrium and (2)
combination of sedimentation and diffusion coefficient
data, using the Svedberg equation. The two series of ex-
periments were carried out simultaneously and used the
same stock subunit solution; the series of equilibrium
experiments is described in the next section. Although
a.GDH dissociated in BE guanidine-H01 the earlier ex-
perience with the pronounced tendency of the enzyme to aggre-
gate provided a convincing argument for the use of

higher concentrations of guanidine.HCl than 3m to avoid

aggregation. Therefore these two series of experiments

were performed in 7.23 guanidine'HCl with 0.lfl_mercapto-
ethanol also present. A series of sedimentation velocity
experiments at different protein concentrations was con-
ducted in the dissociating system described previously.

Extrap01ation of the apparent sedimentation coefficients

0

to zero protein concentration yielded a value of 820 w =
9

1.708(see Figure 11).

From a similiar series of experiments at different
protein concentrations the diffusion coefficient of dis-
sociated a-GDH was also measured so its molecular weight
could be determined by combination of sedimentation and
diffusion data. A value for the diffusion coefficient
of D30,W = 4.1(f0.2) x 10‘7cm2/sec was obtained upon
extrapolation of the apparent diffusion coefficients to
zero protein concentration(see Figure 12). The reasons
for the scatter of the diffusion coefficients were not
obvious; but as a result of the scatter, the value for the

diffusion coefficient was not very accurate. Using the

value of SEQ w = 1.703 for the sedimentation coefficient
3
and the value of D30 w = 4.1 x 10-7cm2/sec for the dif-
9

fusion coefficient a subunit molecular weight of Mg(s/D) =
+

40,000 2,000 was calculated. This value for the molec-
ular weight is approximately one-half the value of
M§(s/D) = 74,400 for native a-GDH. Thee data suggested
that a-GDH consisted of two polypeptide chains. The

sedimentation equilibrium molecular weight analysis of

the subunits described in the next section complements

.909 000.00
mo msflupom mmmam Hopoh 6 Spas om pm pofiaomacm mhmz mpsmsflhcaxm m£8

m” .00.0 m0 .aom.masa.ma.0 000 .«aom ma00.0 .Hosmspmopssosmscma.0 .H0sz

2N.o .Hom.osacasmsm mm.0 pmsfimfim mano: 3N pom pommamflo was Campoaa 03H

.Hoswgmepamoaos msapfismsm SH mpHQSQSm mawlacmo mpsmaoamwooo Cofipmpsms

Iapmm ozp go Scapmapsmosoo Campoam chow on moflpmaoamapxm .HH onsmam

70

3253: ZOEEHZMQZOU

 

 

 

m_ w v 0
. _ _ 1 _ _
is
Imd
/O/ I.
o . 3.00 JN._
O/O //M.@._
/
403100100085: 5:0 ._oI.mz_0_z<:0 20.0 -
_ 0 _ _ _ _

 

(M‘Ozsuwsolsaaoo NOIlVlNI-JWIClZ—IS

71

.mpsosaaoaxo mo mofihmm mpmamamm 03p on momma moaoaao

Como pcm comoao mSH .00 mo chapmhmafmp w 090 Ema oom.m mo woman m
00 00.0 ms 00000000000805 30 .0090 300.0 000.020 10.0 .802
mm.o .Hom.osacasmsm Wm.0 CH cospomaom who: mucosfiamoxo Scamsgmflp 0:9

.HosmzpoOmeOHoE .Hom.onfiwfism5m SH mpHQSQSm mmolo mo memHOHgmooo :onSM

1000 on» no cofipmapcoosoo :Hopoha oaom on Coapmaoamhpxm .NH masmflm

72

333%: ZO:.<m._.Zmozoo

 

 

 

 

N_ m V
_ m _ _ _ _
low
0 oz mxm 0 Tomm N200.0. 0. m0 .0 _.0 ”ammo
_ .OZ mxm O
[QM
O O
O
C
Di 0 0 04v
0
O O
o o Imam
JOZ<IFMOHQ<UKM2 _>__.O ._OI.m_Z_O_Z<DO _>_NN
_ _ _ _ _ _

 

(M030) lNBlOlsldEJOO NOISfl-HICJ

73

the results described above by providing another inde-

pendent measure of the subunit molecular weight.

Sedimentation Equilibrium Studies of 0%GDH Subunits

 

As mentioned earlier a series of sedimentation equili-
brium eXperiments in 73 guanidine'HCl, 0.1fl'mercapto-
ethanol was simultaneously conducted with the sedimentation
and diffusion experiments presented in the previous section.
The same stock enzyme subunit solution was used for both.
The results of the sedimentation equilibrium experiments
are shown in Figure 13. A molecular weight of 43,300 was
obtained by extrapolation of apparent weight average(MW)
and z-average molecular weights to zero protein concentra-
tion. This value is slightly greater than the value of
40,000 based on the sedimentation and diffusion coefficients;
the equilibrium technique is somewhat more sensitive to
aggregation than the other technique. However the indivi-
dual schlieren patterns gave no suggestion of aggregation,
suggesting that the aggregation was not pronounced. Fur-
ther proof that there was not a great deal of aggregation
wx;given by the fact that the z-average molecular weights
lay on the sace curve as the weight average molecular

weights(sce Figure 13).

The Heouirement for Nercaptoethanol

 

Having obtained a reliable value for the subunit
molecular weight, the next question was whether the
mercaptoethanol was necessary to obtain stable subunits

when the guanidine°H01 concentration was as great as 7E°

.mqupmHSono on» 90% cows mm: w\oo w:m.o mo madaob oagaoomm
Hwaphmm m .moaaemm mpdflhmpaw so omfihomamm mpsosahomxo mHGCQSOD capoSp
Imam anm vocampno modam> osu wmamwho>m an cosawpno was GoHpSHom moopm
msp mo AoovsoHpmesooQoo HmeHCH mge .m:.m mm .HosmswoOpQQOHos.mH.o
.Som $00.0 .Homhfla 35 .82 mmé .Hombfigsgm mm; pmfimmm

mfimzawsm FSHHQHHHSUo coapwpsosfioom on HOHHQ mason 2m pom wommawaw mm:

71+

:Hoponm age .NE mom A90 + How mm Ugo 3: Mom N\Apo + 80v mo ooprHm>m
ohms msoapwhpsoosoo .smh oma.mm mo snappom woman HOpOH w spas on ad
oopososoo mucosaaomxm Edfiapaaaswm Scapprosaoom Scam oozafihopoo who:
mpSMams amazooaos psoamaa< .HosmzmepQQOHoE .Hum.osaoasw5w CH mpass

IDSm magic go mpswaoz Hmadomaos Amzv ommao>wlm cso Azavowwmo>mlpzwams

pzmhmaaw msp ho soHpmezoozoo sampoam chow 0p QoprHommhoxm .mH madmam

75

-3
r0 0
r0 LO

LO
(\l

APPARENT MOLECULAR WEIGHT XIO
O
(\l

35:95: ZOF<mFZwUZOQ

m. g m. o o
_

 

t

 

_ _ m

_ i

N_>_o

.oqufioqumms _>_ _.o roxmzazqao saw
_ a _ _ _ a

  

 

 

q-

LO

N
90” l

76

To answer this question a series of eXperiments was con-
ducted in a manner identical to that just described, with
the exception that mercaptoethanol was omitted from the
dissociation solvent. With one single exception, the
samples were very aggregated and gave molecular weights
in the range of 100,000 to 250,000(see Figure 14). In
the one case where dissociation occured the "subunits"
were unstable and a repeat eXperiment conducted after
several days yielded a schlieren pattern comparable with
that of the other samples. At first appearance, this
might seem to indicate the presence of disulfide bonds
in the enzyme; but this does not explain the pronounced
aggregation which occured. The aggregation made it im-
possible to tell whether dissociation had occured and
was followed by aggregation, or whether the enzyme had
simply aggregated without dissociating. The existence
of the one "subunit" suggested that the former was the
case. If this be the case, a further conclusion is that
the sulfhydryl groups of a-GDH in an unfolded State,
such as they experience in 7.23 guanidine°HCl, are ex-
tremely reactive, and that the presence of a reducing
agent is necessary to prevent formation of disulfide
bonds.

In order to further evaluate the possiblity that
dissociation might occur in auanidine'HCl alone, sedi-

mentation velocity experiments were performed on two

77

.H9\me HH Ucm NH .MH .dH .ma msHQprsoo mmaaswm mgp no poshomhmm mpgmsahmmxm
zhmosdon cauoSpcmm 909M oosfimpno moSHm> mnp msfimwhmbm an omSHMon mmz
Ac

Qoszaom xOOQm on» do ovmsofipmhpsoosoo Hoapfisfi one .mapms agapwaoommfiw

ozp CH pcomohm no: mmz Hoswgmepawoamz .om.m ma .¢B_m flaoo.o .Hom.mHHH
HH.o .Homz Hm.o .Hom.osflwflsm5m Hm.m pmsHmNd mamzadgw Esflanflaflzwm
soflpwpzmsfiomm on Hoflnm mhzoz 3N How oomaawfib mm: CampOHQ mgw .N\

ADD + How mm UopdSHm>m ohms msoapwhpzoosoo .Ema oom.mH go wsfippom boomm
Hopoa m Spas find om w pmp05©soo mucosflhoaxm sSHHDHHHSGm Coapdpsosflomm
Foam omsashopmfl macs mpsmfioz HmH50oHos psohwaa¢ .HozmspmouQQOHmE

on Spa; Hom.msaoaswsm SH SH maelo mo mpsvfioz amazooaofi Assvmvwam>m pcwfioz psmamm

lam mgp mo SoapmapCoosoo sampopa oHoN op :oHpmHomepxm .ia mHSMHm

78

APPARENT MOLECULAR WEIGHT x 10’3

j_>_\o_>: ZOF<¢FZM 0200

O_ m w

v

 

o
“P

0
<1—
I
O

[‘0
r0
[

 

A _ q a 4 _

:_OZ<IHmO.E<omm_>_ OZ
_oI.mz_Q_Z<Do Emu.

_

Om

 

 

79

samples of QPGDH in 8.03 guanidine.HCl, 0.2fl NaCl, 0.04m
Tris-HC1, 0.001fi EDTA, pH 7.42 at an enzyme concentration

of 3 mg/ml. One of the samples contained 0.1flimercapto-
ethanol, while the other did not. There was no difference

in the sedimentation coefficients of the reduced and unreduced
samples; both had a value of 336?? = 1.578. These re-

sults suggested that dissociation occured in both cases

and that in the mercaptoethanol system the reactive sulf-

hydryl groups remained reduced, while in the absence of

mercaptoethanol they reacted to form random disulfide bonds.

Performic Acid Oxidation EQCI—GDH

 

 

Since (l-GDH dissociated in the presence of 7.2m
guanidine-H01, 0.lfl mercaptoethanol and aggregates oc-
curred if the mercaptoethanol was not included, an ex-
periment was designed where the formation of disulfide
bonds was not possible. This experiment involved mild ox-
idation of O-GDH with performic acid. This process ox-
idized both cystine and cysteine to cysteic acid and
oxidized methionine to methionine sulfone without cleavage
of peptide bonds (Hirs, 1959). After oxidation, the pro-
tein was prepared for sedimentation equilibrium experi-
ments by dialyzing against 7.55 guanidine.HCl, 0.2fi NaCl,
0.043 Tris.H01, 0.001fl EDTA, pH 8.3 for 24 hours. The
extrapolation of the z-average molecular weights to

zero protein concentration (see Figure 15) yields a NZ 2

80

.onapoHSOHmo

03p Mom cows was m\oo m:m.o mo mESHo> ofigaooam Hoapamm ¢ .omm

pm Ema omm.:m pm csp ohms mpcmfianmmxm .mpsmpomoh mQHNHoHNo oSp
o>osoh on mflmzamsm esfianfiaazmo sofipmpsmsfloom on Hofiam m.m mm .¢amm
H.285 48733 .9405 .32 mmé .3329me fig pmimmm meson em
Mom vowmamae nos» one wfiom oHEHOMHom Spa; Ummfioaxo mm: oszmso age

.maoio umNHUHNOIUHom anhomaom ANszpsmHoz amazomaoe ommao>mum

esp mo zofipmapcmosoo Qfiopoaa oaoN op Soapmaommapxm .ma mazmam

81

APPARENT MOLECULAR WEIGHTx 10'3

33:05: ZO_._.<m._.2m_ozoo
m m .

 

9

LC)
M

0
r0

If)
N

 

a q _ _ _ u

 

:oImZQZdjo 2mm O._.Z_ QwNZSQV
100-6 omN_Q_XO 904 BSEOnEMd

_ _ e _ _ _

0.?

 

 

82

40,100, which confirms the value of Mam/D) = 40.000

found with dissociated reduced d-GDH. This experiment

thus showed that, for subunit studies, the best procedure
for avoiding aggregatiOn was to use performic acid oxidized
enzyme. The question of whether disulfide bonds existed

in the native protein still remained unanswered. The fol-
lowing experiment was designed to resolve this dilema; it
conclusively demonstrated that the subunits in the native

enzyme are not linked to one another by disulfide bonds.

Molecular Weight of Unreduced,Carboxymethylatedti-GDH

 

If the aggregation of unreduced O—GDH in guanidine.HCl
was due to the oxidation of sulfhydryl groups to disulfide
bonds, it seemed possible that it might be prevented by
dissolving the protein in a carboxymethylating dissociation
solvent; i.e. a solvent containing the iodoacetic acid in
the dissociation solvent before the native enzyme is dissolv-
ed in it. This is in contrast to the usual procedure of
dissolving the enzyme in a solvent containing guanidine.HCl
only, and then adding the iodoacetic acid to this solution.
With this simultaneous carboxymethylation dissociation the
sulfhydryl groups would become carboxymethylated as soon
as they became exposed to the media and therefore they
could not form disulfide bonds. The subunit samples for
this series of sedimentation equilibrium experiments were
prepared by dissolving the ammonium sulfate pellet con-

taining native d—GDH in a dissociation carboxymethylating

solvent containing 7.5fl'guanidine-HCl and 0.2g iodoacetic
acid(for other details see legend of Figure 16). Ex-
trapolation of the weight-average and z-average molecular
weights to zero protein concentration gave molecular
weights of 46,200 and 52,600 respectively. This indicated
that dissociation had been acmeved with this enzyme de-
spite the fact that it was not reduced. Although slightly
higher molecular weight values were obtained for the sub-
units, thereizs no doubt that this molecular weight value
represenkfl suounits andvm£;slightly high because of ag-
gregation. The individual sedimentation equilibrium
schlieren patterns showed a small amount of aggregate at
the bottom of the cell. Additional evidence for the
existence of aggregated protein is shown by inspection

of Figure 16 where the z-average molecular weights are
greater than the weight average molecular weights at
corresponding protein concentrations. As a result the
weight average molecular weight of 46,200 is expected

to be high. Indeed based on previous experience with
aggregation, it was predicted that some aggregation would
occur.

It would be difficult to overestimate the importance
of the results from this series of experiments. These
results directly, clearly, and conclusively prove that
the polypeptide chain subunits of native a-glycerOphos-

phate dehydrogenase are not linked by disulfide bonds.

.H9\ms 0 Use .m .m .m .oa.mcficfimp

Isoo moamsmm no UmEHOMHmQ meoEaHomMo mam©CSOQ capoSpnzm Eon“ cmcflmpno
mosHm> mzp wsammho>m an comamppo mm: :oHpSHom XoOpm on» go AooVQOHpmap
usoosoo Hmapaza one .mhfios :m 90% om.w ma .¢gbm maoo.o .Hom.mHHB fla.o
.Homz mm.o .Hom.ocfioficmsw mm.m pmsamwm commamac mm: zfiopoam exp mopzcfis
m: empe< .n.m me .cfiom ofipmomoeofi mm.o .«gnm mmoo.o .Hom.mfiee.mea.o
.Hom.onfinasm3mrmm.m SH oopaommfio mm: sfiopoam mge .N5 90% ADD + Eov

mm Ugo 3: Mom N\AQQ + SUV mm oopmSHm>o ohms mzoapmhpCoonoo .EQH

L omm.mm mo wQprmm cmomm Hooch 6 Spa: 62m om pm copososoo mucosaammxm
ESHHpaaflsvo :oHpmpsoEHUom soap UmQHEHmpoU mam: wpswaoz amazooaos
pcmhmaa¢ .Umpmaoommao use @opmHmSmemeonamo mHmSoQOpHSEHm mmwio

wo mpsmaoz amazomHoE Asz omdho>dnm cam “asymmmhobm pgmaoz pcmhmm

lam mgp mo :oprHpsoocoo Campoma oaom 0p QoHpMHogmnpxm .mH masmam

85

APPARENT MOLECULAR WEIGHT x 6“

0
L0

pr)
PO

LO
(\J

O
(\l

ON

Ale/<05: ZO_._.<m._.ZmoZOo
Q m__ m
A

 

 

a

N2.

 

. u N
86$th

:00 - 0 3.934 >Ihm§>xomm<o
_ _ _

 

 

 

om

W
gleT

86

Summary of the Molecular weight Analysis g£_the aaGDH Subunit

 

 

Table VII gives a summary of the molecular weight

analysis of the subunits of a-GDH

Table VII. Summary of the molecular weight analysis of
the erDH subunit

 

 

Treatment Molecular Weight
I; 11‘: ( s /D) 11:
7.2g guanidine-H01 43,300 40,000 43,300
0.13’mercaptoethanol
Performic acid 40,100
oxidized
7.5g guanidine.HCl 40,000 to

250,000

carboxymethylation in 46,200
7.5fl,guanidine.HCI

 

Renaturation of a—GDH

 

Upon dissociation of d-GDH into stable subunits it
was of interest to see if active enzyme could be reformed
by removal of the dissociating agent. It was hoped that
the knowledge gained from these experiments would yield
information concerning the conditions necessary for the
folding of the newly synthesized polypeptide chains into
active enzyme molecules. Therefore a study was undertaken
to determine the conditions necessary for maximum re-
covery of enzyme activity. These conditions included

pH, protein concentration, the presence of the substrates

and products, and the effect of reducing agents.

Preliminary attempts to reverse the dissociation of
d-GDH by removing the dissociation agent were immediately
successful. When an aliquot of denatured d-GDH was di-
luted directly into the assay mixture there was a lag of
a few seconds followed by a dramatic increase in enzymatic
activity as the enzyme recovamfl.its catalytically active
conformation. The following experiments were then de-
signed to determine conditions necessary for a maximum
recovery of catalytic activity. The enzyme was prepared
for analysis by dissolving a pellet of native d-GDH,
centrifuged from an ammonium sulfate suSpention, into the
dissociating solvent which contained the following: 7.5m
guanidine0HCl, 0.13'Tris-HCI, 0.0013 EDTA, pH 7.46. The
final protein concentration was 1 mg/ml. This stock dissocia-
ted enzyme was allowed to sit in this dissociation mixture
for 2 hours. The denatured protein was diluted (1:40) from
this stock solution into tubes containing 1 m1 of the rever-
sal solvent. The optimum reversal conditions were: 0.1fl’Tris-
HCl, 0.2§ mercaptoethanol, 0.001; E TA, pH 7.42 and a final pro—
tein concentration of 0.025 mg.ml. It is important to note
that mercaptoethanol was not included in the reversal solvents
since its effect was discovered later. After 15 minutes
at room temperature 10 ul of the renatured enzyme was
assayed in a volmmaof 0.3 ml. For the experiments to

to follow, all variables (except mercaptoethanol) were at

88

optimum conditions except the one under study. In order
to accurately determine the % recovery, the activity of
the stock enzyme solution was determined with each series
of experiments. Although the Specific activity of the
native enzyme varied from one stock solution to another,
the relative percent did not. It was constant for a
given condition, regardless of the specific activity of
the stock enzyme. The native enzyme control was treated
exactly as the dissociated sample with the single ex-
ception that the dissociating agent was not present.

All dilutions of the denatured enzyme into the reversal

media were performed in duplicate.

ffect 2f. F

t1]
(5

[,

r—Hl

on Benaturation

 

Experiments were designed to examine the effect of
the pH of the renaturation solvent on the degree of re-
naturation. As shown in Figure 17 the amount of renatur-
ation is very dependent on pH; two maxima were obtained
at pn's 6.8 and 7.5. At pH 7.5 the activity recovery was
in the range of 55%. The amount of renaturation did not
appear to be dependent on the type of buffer used. For
example, at pH 7.4 the amount of recovery obtained with
a phOSphate buffer and that obtained with a Tris'HCl

buffer differed by only 2%.

Effect g: Protein Concentration 9g Renaturation

 

 

As previously mentioned it was suSpected that the

concentration of protein in the renaturation media might

.cpdcawao
psmaa mgp So cmppoam ma apmbooma & end opmsacao pmma mgp So moppoaa

o/ ma mpabapom camaomam .Amaampoc ampscsflaoaxo Mom pxop oomv mpfl>apom
03

oHpmsmNQc mom cmhmmmm mm: mshmso ogp mo QoHpaom m magpmammgmp Boos
pd mmpSQHE ma Hmpg< .o.m on o.m mos“ mcfimsma mm macaamb mo macmmsn
opQH cmpfiaac mmz mahmsm mne .om.n mg .geom.mHoo.o JonSmepQQOHmE

EH.o .Hom.maaaima.o .Hom.osacfismzm_flm.m SH mHSos N Mom UmMSpmscw mm:

AHB\mE av sfiopoam mge .mmwn mo sofipmazpmsmm So am mo pommmm .mH mHSmHm

% RECOVERY

Om

0v,

Om

Om

OO_

[03

L0
CPO

 

 

—.L

m._.<m._._o O
_oI.m_E. m

whqldmOId m

13010 SE do m<mmm>mm

F

_

_ h

 

 

All/\IIOV BWAZNB

91

affect the amount of recovery of enzymatic activity. The re-
sults of a series of experiments where the concentration

of protein in the reversal mixture was varied are shown

in Figure 18. This experiment was designed so that the
amount of guanidine-H01 in the renaturing mixture was

always constant. A maximum recovery of 55% was obtained

at 0.025 mg/ml. Renaturation at concentrations of 0.1

mg/ml and higher was not possible, for the protein pre-
cipitated immediately upon dilution into the renaturation

media.

Effect of Metabolites gn Renaturation

 

 

Since it has been shown by many investigators(Chilson

1., 1965; Hill 33 al., 1964; Beychok et 1., 1959)

gt
that coenzymes or substrates may have an effect on the
amount of renaturation, the effects of the O5-GDH sub—
strates and products were tested. With the exception of
a-glycerophosphate, none of the substrates or products
had any effect on the degree of renaturation in the

-6

concentration range tested (1 x 10 g to 1 x 10
The results of a series of experiments in which the
concentration ofCX-GP was varied from 1 x 10‘73 to 3 x

10‘2

g are shown in Figure 19. It is seen that at con-
centrations of a-GP greater than 5 x 10'53 the renatur-
ation of a-GDH is inhibited; for example, only 30% re-

covery of enzyme activity was observed at a11-GP con-

centration of 3 x 10-33. The possibility that lower

92

.opmgflcho pgmfih 05p So ecppoaa

ma mao>oooa R end mumsfieao pmma ogp So Umppoam ma mpfi>apom oflawo
loam .maspmaomamp 9009 pm coshomaom mmB goapmHSmemm .mmmmo Ham SH
ommm exp mms Soapmhpsmosoo Hom.msaefismsm Hmcam exp pmmp Hmeao SH
mamamm Hmmno>oa homo on cmeem macs Hom.osaefisdsm mo mpcdoem UcHSmmc?

.cozmmmm mm: mmhmao msp mo Soapaoa m mopSQHE ma Hmpw< .Qofipmapgmosoo
saopoaa Umaammo ozp cbam op m:.n mm .aBQm mHoo.o .Hom.mflam_ma.o

go He o.H ousfi @mpSHHG Emma mms posvflam gm .m:.m mm .HonSpcOpmmo
Lee H6 .58 waooé £21.35 mic .Homéfieasmsm Hm; ca 936:

m Mom chapmsme on mozoaam wmz AHE\mE omv mfizmso mo :oHpSHom xoOpm

m .Qoapdhdmeoa Com: Soapmapcmogoo Campoam go pommmm .wa cadmam

93

O
(\l

°/o R ECOVERY

Q
q-

Om

Ale/<03: ZOqu HzmoZOo

and

 

 

coo VQG
_ a
O O
\\o
1810 SE

_ _ .

do

_

J<mmm>mm

P

_

OO

 

 

All/\IIOV BWAZNB

94

.mpmsaeho pgmaa on» so ecppoag ma mambooma & cam mpmSHcao pmma mSp

no coppoam ma apa>apom cagaomam .dadma zmmmm mo Ha m.o cpCa dmpdaac
mm: mammcm dmASpMth mSp mo munama OH .oHSpmammamp Boom pm mmpzsda
ma ampwm .Ha\wa mmo.o mm; CoHpmeCmocoo Campoaa Hmsdm 0:9 .molo

mo Coapmppcmocoo ompmcwammc ms» chHMpsoo Scams N:.m ma .maam flaoo.o
.Hom.mdpe wa.o opmd UmpSHHU mmz mason 03» Mom dongpmomeoaoE MH.o
.maé me .Homéfie mic .fibm ”1.80.0 .Smtfieagism mmé 3 @8393

.momnav .coHpmHSpman coma mpmzmmosmoamomamlo ho pommwm .mH mnswfim

95

°/o RECOVERY

.9

>530: mzzamoxaommohoo
Nb. m.o_ Vo_ no 99 No.

 

ON

Ov

Owl

 

_ _ l a l a

10010 Em do n<mmm>mm
_ _ h _ _ _

 

 

All/\l 13V EWAZNE

96

recovery of catalytic activity might be due to product
inhibition of active enzyme in the assay had to be con-
sidered. This could occur for example as a result of

the contribution of a-GP from the reversal sample to the
total OPGP concentration in the assay. Since the dilu-
tion of the reversal mix ure into the assay mixture is
1:30, the assay mixture concentration of a-GP Will be
1/30 that of the reversal mixture. For ex mple, a con-
centration of 3 x 10'3;fiz-GP in the renaturation mixture
will yield a concentration of l x 10-4; in assay mixture.
Thus, from these experiments the only conclusion possible
was that chP was inhibiting; (1) either the refolding

of the dissociated enzyme in the reversal mixture or, (2)

the expression of activity of active enzyme in the assay.

1

Product Inhihition of a—?l€ 3y oezlyceronnoschate

 

 

In order to distinguish between the two possililites
presented by the previous results the following experi-
ment was performed. The effect of 1 x 10‘3; erP was
measured on two samples; (1) oeGDH was renatured in the
presence of a-SP, (2) native a—GWH activity was assayed
in the presence of (£31?. In the forder case the con-
centration of a—GP in the assay mixture was 3 x 10‘5g
while it is l x 10'3;_in the latter case. Examination of
Table VIII shows that inhibition was greater when a-GDH
was reversed in l x 10‘33 than when a—GDH was assayed

at the same a—GP concentration. Two conclusions could

97

Table VIII. Effect of a-GP on.a-GUH

Concentration of

 

G-GP in assay % inhibition
Benatureda 3 x 10‘53 40(inhibition
enzyme of reversal)
Ifative 1 x 10‘311 3o
enzyme

 

adenatured in the presence of l x 10‘3M a-GP. Therefore

a -GP concentration in the assay was 3‘: 10'5;.
be drawn from these experiments, (1) l x lO'BEcx-GP inhi-
bited the refolding of the dissociated enzyme and (2) l x 10-33

a-GP inhibited the native enzyme.

('1

Kinetics of Renaturation in the Presence of

m _

v7

IADH 93 3A?

 

 

 

 

Although the substrates and products of the a-GDH
reaction showed no affect on the extent of renaturation,
it was possible that they might effect the rate of renatur-
ation. To test for this effect, the amount of recovery of
enzymatic activity after various periods of time of incu-
bation in the reversal mixture was measured. OPGDH was
dissolved in 7.5g guanidine-331, 0.13’TrisoHCl, 0.0011
EDTA, 0.001E mercaptoethanol, pH 7.52 to give a protein con-
centration of 1 mg/ml. After two hours in this solvent 25 pl
was diluted into 1.0 ml of 0.lfl,TriS°HCl, 0.001fi EDTA, pH
7.42. Aliquots of loul were withdrawn after the indicated
time period and assayed for enzymatic activity. Examina-

tion of Figure 20 suggests that the presence of NADH or

98

.omeHcHo psmflh mgp So

coppoaa ma mambooma R can mumsfioso pmma mgp Go ooppoaa ma mpflbfipcw
cagaoomm .cmhwmmm mm: mamwzo cmHSpmsmH mzp ho soapaom w ambampsa
mean cmpwcacsa mg» smpwd .mambauooamma mmuoa N H was mined N m.m
pm psmmmaa macs moocmpmndm mmmnp .mwmm so mmdz mo cocomoam mSp

ma CoapmpSpmsoa mo mmaozpm 0:» Mom .N:.m ma .¢BQm_flHoo.o .Hom.mHaB
ma.o .Hom.ocH©H:msm mm.m CH maso: N How bandpmsmc was Campoam one

.mmoio mo soapQHSpmsmH oSp com: m<mQ Ho ma<z do muoohmm .om masmam

99

°/o RECOVERY

AmDZOommv ZO_._.<mD._.<Zm_m “.0 m_.>__._.

 

00m ooN 09
fl a _ _ - _ _ N\
cm]. 1
Q<IQ a
quz . ea 1
_o:.nms o
oq..
4
owl

 

 

IQOLO Em LC J<mmm>mm

_ _ _ _ _

 

 

All/\IIOV SWAZNEI

lOO

DHAP had no effect on the rate of renaturation. Maximum
recovery was in the range of 6C%. The half life of
renaturation, i.e., the time required for 50% recovery,

was in the vicinity of 70 seconds.

Renaturation in mercaptoethanol

It was of interest to test the effect of mercaptoethanol
on reversal for two reasons. First it was thought that the
presence of mercaptoethandiin the renaturing mixture might
prevent aggregation thus yielding a greater degree of recovery.
Secondly, if renaturation could be accomplished in the pre-
sence of mercaptoethanol, indirect proof would be obtained for
the absence of disulfide bonds in the catalytic unit. Fig-
ure 21 illustrates the dramatic effect that mercaptoethanol
had on the extent of Q-GDH renaturation. The extent of
renaturation increased with increasing mercaptoethanol
concentration, with an approximate reammry of 90% being
obtained at 0.2g_mercaptoethanol. This increase in acti-
vity was not due to an activation of the native protein,
for the enzymatic activity of native OLGDH did not increase
when incubated with 0.2fl.mercaptoethanl.

The ability of the enzyme to renature in the presence
of 0.2fl'mercaptoethanol provides further indirect evidence
that disulfdie bonds are not necessary for catalytic

activity of the enzyme.

101

.mprHpao pswfia mSp so dmppoaa ma hambooca

R ozm opdsficao whoa oSp no Umppoam ma mpfi>apom oflwflommm .mQSpmammEmp
Boos pd omenomaom was sofipmHSmeom .HocmgmepdmoaoE go mpszoam oopmo
lawns mSp Umsawpsoo moans m:.m mm .wemu maoo.o .Hom.maa9 mH.o do He o.H
cpsa UopSHHo was posvfiflm Q< .om.m mm pm HosmSpoOpgwoaos mH.o .weom maoo.o
.Hom.mHHB ma.o .Hom.msacficmzm Hm.m CH mummsaaogwm scapmadpmsma mnp 0p

Hofiam mason 03p 90% UmhdpmSmc was Qampoam mSB .HOQdSmepmmoama Mo msoHp

umapsoosoo msofiam> do mosmmoaa esp SH momma mo sofipmHSpmsom .Hm maswam

102

RECOVERY

°/o

>tm<n§2 JOZ<IHm0Fd<0mmE

000.0

0

 

0m0
_

0
q-
l

 

0..0
_

00.0 _0.0
_ _

100-5 SE d0 44mmw>mm

# _

 

 

 

Ail/\IIDV SWAZNH

DISCUSSION

In this investigation the values for the following

physical properties of native rabbit musclecl-glycerophos—

O
20,w

x 10-7cm2/sec, fO/f = 1.23, V = 0.746 cc/g and h;(s/D) =

phate dehydrogenase were obtained: 830 w = 4.863, D = 6.20
9

74,400. Van Eys gt El. (1959) also measured these quantities

o _ o _
and found the following values, 820,w - 4.98, D20,w - 5.1 x

10'7cm2/seo, fO/f = 1.44, V = 0.70 cc/g and N;(s/D) 78.000.
Young and Pace (1958a) found a value of 0.70 cc/g for V,and a
value for fO/f of 1.4. Thus the same values were obtained

by these groups for Sgo,w and M$(s/D), but different values
were obtained for fO/f, D30,w’ and V.

While the value for fo/f is somewhat secondary in in-
terest it is necessary to know the correct Dgo,w and V values,
since both affect the molecular weights.

The diffusion coefficient will be considered first.

If the diffusion coefficient values obtained in this research
are compared with typical results for other globular pro-
teins (see Figure 6) with known molecular weights and diff-
usion coefficients, it is seen that our value of D0 =

20,w
6.20 x 10'7cm2/sec lies on the empirical curve (—--) expected
for a protein of molecular weight ca. 75,000 while the value
of Van Eys g£_al. (1959) of D30,w = 5.1 x lO'7cm2/sec falls
far below the line. This suggests that the diffusion co-

efficient value of Van Eys gt El. (1959) may be in error. Pur-

104

suing this further, it is of interest that the value of s30 w
9

= 4.98 found by both groups does lie near the empirical line
(--—) for the sedimentation coefficients of a globular pro-
tein of molecular weight 75,000 (see Figure 5). Thus, our
values for both 330 w

for typical globular proteins while the Van Eys 33 a1. (1959)

and DO lie on the empirical line
20,w

value for the Sgo,w lies on the typical line but the D20,w
value does not.

A8 mentioned before, the value of fO/f measured by
Van Eys gt_§l. (1959) is greater by 20% than that expected
for most globular proteins (Tanford, 1961) thus raising the
possibility that this value is also in error. This 20% varia-
tion in fo/f is directly attributable to the low value of
the diffusion coefficient, for fO/f is directly related to

the diffusion coefficient as seen below.

2 =30
f Do f

D = ———- and

where the subscript, 0' refers to a sphere, k is Boltzmann's
constant, t is absolute temperature and f is the frictional
coefficient. Now, fO/f is also dependent on V but this de-
pendence is small since fO/f depends on the l/3 power of
V (see appendix). Thus the use of either of the two values
for V (0.70 and 0.746 cc/g) would give an fO/f differing
by less than 2%.

If we assume that the correct value of D20,w is 6.20
x 10-7cm2/sec and use the values of Sgo,w and V calculated

by Van Eys 23.2l- (1959) a value of ca. 64,000 is calculated

for the molecular weight of nativeIX-GDH.

105

The fact that the value of 0.70 cc/g found by Van Eys
g£,§l. (1959) and indicated by the data of Young and Pace
(1958a) is probably in error is suggested by the follow-
ing; (1) the value of 0.70 cc/g is lower than that eXpected
for typical globular proteins which have V values in the
range 0.73 to 0.74 cc/g (Tanford, 1961) and (2) using this
value for V and the correct value for D20,w a molecular weight
of 64,000 is obtained for<1-GDH which is much lower than
expected for a protein with a sedimentation coefficient of
4.93 (see Figure 5), and (3) the value of 64,000 in conjunc-
tion with the subunit molecular weight of 40,000 does not
yield a whole number for the the number of polypeptide
chains present in a-GDH.

There are three reasons for the confidence placed in
the value of 0.746 cc/g obtained in this work for the v of
d.GDH; (1) extensive direct measurement (falling drop method)
and empirical calculations (based on amino acid composition)
yielded identical results, (2) the value of 0.746 cc/g
is close to that expected for globular proteins and (3) the
use of 0.746 cc/g in molecular weight calculations gave
values for the native enzyme and for subunits which yield-
ed an integral number of subunits in the enzyme. There is
no obvious explanation for the discrepancy in our results
and those of Van Eys gt a1. (1959) and of Young and Pace (1958a).
One possible eXplanation is the limited scope of the exper-
iments of Van Eys §£_§l. (1959) and the unusual conditions

(1.0fl ammonium sulfate) used by Young and Pace (1958a).

106

The precise value used for the partial Specific volume
in the juanidine-Hil has a large effect on the molecular
weight results because of the high solvent density(71 guani-
dine-H01 = 1.195). There are two problems which occur
in the use of the correct value for the V for the enzyme, (1)
whether the V is the same for the folded(native) and un-
folded(subunits) protein molecule and (2) if there is pre-
ferential binding of one of the solutes(water or guanidine-HG )
to the protein molecule.

The results reported in the literature are conflicting
as to the effect of guanidine-301 on partial specific volume.
Kiellfirand Harrington {1960) reported that the V of myosin in
51 guanidine'HCl decreased by 0.01 cc/g relative to the
value obtained for the native protein in aqueous salt solu-
tion. izarier 213. 3;.(1960) showed. similiar effect for V—glo-
bulin. 0n the other hand, Reithel and Sakura(1963) have
found little change in V for a number of proteins. Cal-
culation of the V of a protein by summation of the partial
Specific volume contributions of the individual amino acids
(Cohn and Edsall, 1943) would seem to us to be the best
approximation of the V for an unfolded protein. The fact
that the V obtained from amino acid analysis is identical to
the experimentally determined value argues for the conclusion
that the partial specific volume of unfolded erDH is the
same as that for the native molecule.

Schachman and Edeletein(l966) have made calculations

for the preferental interaction of water with the denatured

107

protein and found a substantial contribution of water to
the partial Specific volume of aldolase.

It was assumed in the calculations used in this re-
port that the partial Specific volume was the same for the
native and unfolded protein and therefore no correctirns
were made for the effect of guanidine'HCl on the partial
specific volume of CkGDH.

The chemical and physical evidence presented here is
compatible with the theory that a-GDH is composed of two
polypeptide chains. This was supported by the following
experiments: (1) the molecular weight of the subunit ob-
tained from sedimentation and diffusion experiments is appro-
ximately one half that of the native enzyme(h;(s/I) = 74,400,
(2) digestion of erDH with carboxypeptidase-A released
2.0 moles of methionine per mole of protein.

The question of whether the polypeptide chainsofcx-GDH
were held tOgether by disulfide bonds was a very perplexing
problem which was solved only after great difficulty. The
presence of disulfide bonds in the native protein was sug-
gested in several ways: (1) the necessity of 0.l;_mercapto-
ethanol in the dissociating mixture in order to observe stable
subunits, (2) the production of stable subunits when the
protein was oxidized with performic acid, (3) failure to
obtain the maximum number of carboxymethyl cysteines when the
enzyme was carboxymethylated but not reduced. There are
two obvious explanations for these results: (1) disulfide
bonds are an integral part of the native QPGDH molecule or

(2) disulfide bonds are not an integral part of the native

108

molecule but are formed upon treatment and handling of the
enzyme.

Fortunately, evidence was obtained which conclusively
proved that the latter was the case and that disulfide bonds
were ngt an an integral part of the native molecule. This
experiment involved dissolving the unreduced protein in a
dissociating-carboxymethylating solvent, which prevented
sulfhydryl groups from being oxidized to disulfide bonds.
Under these conditions dissociation did occur. A second
proof for the absence of disulfide bonds in the catalytic
unit was the observaticn that the dissociation process could

be reversed in the presence of 0.2g mercaptoethanol.

S UICI-lABY

(1) It has been shown that native d-GDH has the

following properties.

a. D30 w = 6.20 x 10‘7cm2/sec (new value:of. Van Eys
I
23 g__1_.. 1959)
b. 830 w = 4.863 (confirming Van Eys 23.31): 1959)

c. fO/f = 1.23 (new value: of Van Eys gt al., 1959)
d, v = 0.746 oc/g (new value: of Van Eys 23 al., 1959:
Young and Pace, 1958a)

e. Ifi(s/D) = 74,400

(2) A partial Specific volume of 0.746 cc/g was found
for the native enzyme by two independent methods; (a) in-
directly from the amino acid composition and (b) by direct
measurement using the falling drop method.

(3) The values obtained for the Sgo,w and Dgo,w
fell on a line drawn for typical globular proteins of known
molecular weights.

(u) There are 21 moles of half-cystine per mole of
protein; determined as carboxymethyl cysteine and as cysteic
acid

(5) Two moles of C-terminal methionine were found per
mole of dFGDH, showing that arGDH is composed of two subunits.

(6) Fingerprinting of carboxymethylatedél-GDH suggests
that dsGDH is composed of a maximum of three identical poly-

peptide chains.

110

(7) The complete amino acid analysis has been performed
(confirming Van Eys 33 al., 1964)

(8) The enzyme dissociates into subunits in 7.22
guanidine-H01, O.lfl'mercaptoethanol. The subunits have the
following prOpertieS.

O a
a. 820,.W = 107OD

o
20,w

O
c. MW(s/D) = 40,000

b. D = 4.1 x 10'7cm2/Sec
d. N; (sedimentation equilibrium) = H3,300. This value
is high due to slight aggregation
(9) The enzyme aggregates if mercaptoethanol is not
included in the dissociating solvent.

(10) The inclusion of iodoacetic acid in the dissoc-
iating solvent results in the production of subunits thus
proving that the subunits are not linked by disulfide bonds.

(ll) Reversal of the dissociating process is accom-
plished upon dilution of the dissociating agent.

a. A maximum recovery of 90% is obtained.

b. The optimal conditions for renaturation are 0.1fl
Tris-HCl, pH 7.42, 0.23 mercaptoethanol, at 250 at
a final protein concentration of 0.025 mg/ml.

c. NADH, NAD: DHAP and erP do not affect renaturation

BIBLIOGRAPHY

Acher, 3., and Crocker, C. (1952), Biochim. BiOphys.
Acta 2, 704.

 

Ankel, H., Bflcher, Th., and Czok, R. (1960), Biochem.
Q! 332, 315.

Anfinsen, C. 3., and Canfield, H. E. (1963), The Proteins
2nd edition, Vol. 1, p 334. Academic Press, New York

Anson, H. L. (1941), Gen. Physiol. 24, 399.

 

Arendt, B., and Pfleiderer, G. (1953), ;. Naturforsch

i2; 555-
Baranowski, T. (1939), ;. Physiol. Chem. 260, 43.

 

 

Baranowski, T. (1949a), i. Biol. Chem. 180, 535.

 

 

Baranowski, T., and Niederland, T. H. (1949b), i. Biol.
Chem. 180, 543.

Barbour, H. G., and Hamilton, W. F. (1926), i. Biol.
Chem. _6_2, 625.

 

Beisenherz, G., Boltze, H. J., Bucher, Th., Czok, 3.,
Garbade, K. H., Meher-Arendt, and Pfleiderer, G. (1953),
;. Naturforsgh_813, 555.

 

Blanchaer, H. C. (1965), Can. 1. Biochem. 43, 17.

Block, N. J. (1951), Arch. Biochem. Biophys. 1, 266.

 

 

Borreback, B., Abraham, 8., and Chaikoff, I. L. (1965),
Biochim. Bioohys. Acta 26, 237.

 

 

Borst, P. (1962), Biochem. Biophys. Acta 513 270.

 

Boxer, G. E., and Shonk, L. E. (1960). Cancer Research
2.9.. 85.

 

Boxer, G. 3., and Devlin, T. M. (1961), Science 134,
1495.

 

Boyer, P. D. (1954), i. Am. Chem. Soc. 16, 4331.

 

Bucher, Th., and Klingenberg, M. (1958), Angew. Chem.

0, 552.

111

112

Bflcher, Th., and Klingenberg, M. (1960), Biochem.
g, 332, 328.

Burton,/K., and Wilson, T. H. (1953), Biochem. J.
L)“, 80.

Celliers, P. 3., Stock, A., and Pfleiderer, G. (1963),
Biochim. Biophys. Acta 71, 577.

Chilson, O. P., Kitto, G. B., and Kaplan, N. O. (1965),
Proc. R. A. E. 22, 1006.

Ciaccio, E. 1., Keller, D. L., and Boxer, G. E. (1960),
Biochim. BiOphyS. Acta 31, 191.

 

Ciaccio, E.K., and Keller, D. L. (1960), Fed. Proc.
.22, 34.

Cohn, E. J., McHeekin, T. L., Edsall, J. T., and
Blanchard, H. H. (1934), J, Am. Chem. Sec. 56, 784.

 

 

Cohn, E. J., and Edsall, J. T. (1943), Proteins, Arino
Acids, and Peptides as Ions and Dipolar Ions, Reinhold
Pulb. Co., New York, N. Y.

 

 

Crestfield, A. M., Moore, 3., and Stein, W. H. (1963),

 

Deal, W. C. Jr., and Van Holde, K. E. (1962), Fed, Proc.
21, 254.

 

Deal, w. C. Jr., Butter, w. J., Massey, V., and Van Holde,
K. E. (1963), Biochem. Bigphys. Hes Conn. 19, 49.

 

Deal, w. C., and Holleman, W. H. (1964), Fed. Proc.
Q, 254.

 

Delbrflck, A., Schimassek, A., Bartscg, K., and Bdcher,
The (1959), BiOChemo a. 22]., 2790

Dickens, L. (1933), Biochem. J, 21, 1144.

Disteche, A. (1948), Biochim. BiOphyS. Acta 2, 265.

 

Emmelot, P. M., and B03, C. J. (1962), Biochim.
Biophys. Acta 52, 495.

Gulyi, M. F., Litvenenko, L. T. (1959), Dvornikova,
Doklady Ahad. Naak. 8383. 129,451; Chem. Abstracts 59, 5394(1962)

 

Hirs, C. H. w. (1956), J. Biol. Chem. 219, 611.

 

 

113

Hohorst, H. J., Kreutz, F. A., and Bdcher, Th. (1959),
Biochem. g, 332, 18.

Holzer, P., Glogner, P., and Sedlmayr, G. (1958),
Bioohe‘no Z... 220, 59.

Jepson, J. B., and Smith, I. (1953). Nature 172. 1100-

 

Jirgensons, B. (1965), J. Biol. Chem. 240, 1069.

 

Kawahara, K., and Tanford, C. (1966), J. Biol. Chem.
241, 3228.

 

Kielley, w. W., and Harrington, w. F. (1960), Biochim.
BiophLTSo ACta 3;, 40]..

 

Klingenberg, M., and Slenczka, W. (1959), Biochem. g.

331, 254.

Klingenberg, K., and Bdcher, Th. (1960), Ann. Rev.
Biochem. 29, 669.

Koorajian, S., and Zabin, I. (1965), Biochem. Biophys.
Res. Comm. J8, 384.

 

Kusama, K. (1957), J. Biochem. Japan 44, 375.

Kutz, A. M., Dreyer, w. J. and Anfinsen, C. B. (1959),
J. Biol. Chem. 234, 2897.

Lehninger, A. L. (1951), J. Biol. Chem. 190, 345.

 

Levy, H. B., and Vennesland, B. (1957), J. Biol. Chem.
228, 85.

 

Litvinenko, L. T. (1963), Ukrai, Biokhim. §Q° 35, 410.

Lowry, O. H., Hosebrough, N. J., Farr, A. L., and
dandall, H. J. (1951), J. Biol. Chem. 193, 265.

 

Ludoweig, J., and Dorfman, A. (1960), Biochim. Biophys.
Acta 33, 212.

Harler, B., Nelson, C. A., and Tanford, C. (1964),
Biochemistry g, 279.

HcHeeKen, T. L., Groves, M. L., and Hipp, N. J. (1949),
J. Am. Chem. Soc. 1;, 3298.

 

McMeeken, T. L., and Marshall, K. (1952), Science
116 , 142 o

Meyerhof, O. (1919), Pflugers Arch. :es Physiol. 175, 20.

114

Moore, S. (1963), J. Biol. Chem. 238, 235.

Morris. H.P., Sidvansky, H.. Wagner, B.P., and Dryer,
H.M. (1960), Cancer Research 29, 1252.

Neurath, H.. Elkins, B., and Kaufman, S. (1947), J.
Biol. Chem. 170, 221.

 

Pette, D.. and Huge, G. (1963), Biochem. Biophy_.
Res. Comm. 23. 150

 

Pfleiderer, G., and Auricchio, P., (1964), Biochem.
Biophys. Res. Comm. J6, 53.

 

Reithel, F.J., and Saltura, J.D. (1963), J. Phys. Chem.
éz. 2497.

Sachs, J., and Morton, J.D. (1956), fig. J. Physiol.
186,221.

Sacktor, B. (1955), J. Biophys. Biochem. Cytol. J, 29.

Sacktor, B., and Dick, A.R. (1965), Biochim. Biophys.
Acta 99, 549.

Sacktor, B., Packer, L., and Estabrook, B.W. (1959), Arch.
Biochim. Biophys. §9, 69.

 

Schachman H.D. (1957), Methods lg_§nzymology 4, 2.

Schachman, H.K., and Edelstein, S.J. (1966),
Biochemistry_5, 2681.

 

Shonk, C.E. (1962), Federatign_Proc. 2;, 251.

Tanford, C. (1961), Physical Chemistry_g§ Macromolecules
John Wiley and Sons, Inc.. New York.

Telegdi, w. (1964),Acta Physiologica Acad. Sci. Hung,
25, 177; Chem.Abstracts 62, 8053(1965).

Telegdi, H.. and Keleti, T. (1964), Acta Physiologica Acad. Sci.
Hung. 25, 181; Chem.Abstracts 62 8054(1965).

Tomita, K., and Helling, S. (1965), Federation Proc.
22; 351.

Van Eys, J., Stolzenback, F.E.. Sherwood, L., and Kaplan,
N.O. (1958), Biochim. Biophys. Acta 2 , 63.

115

Van Eys, J., Nuenke, B. J., and Patterson, M. K. (1959),
J. Biol. Chem. 234, 2308.

 

Van Eys, J. (1960), Fed. Proc. 29, 26.

 

Van Eys, J. (1961), Nutrition 13, 403.

 

Van Eys, J. (1963), Fed. Proc. 22, 290.

Van Eys, J., Judd, J., Ford, J., and Womack, w. B. (1964),
Biochem. 3, 1755.

Van Holde, K. B., and Baldwin, H. L. (1958), J. ths.
Chem. 62, 734.

Winstead, J. A., and Wold, F. (1964), Biochemistry 3, 791.

 

Young, Ho Lee, and Pace N. (1958a), Arch. Biochem.
Biophys. 15, 125.

Young, Ho Lee, and Pace, N. (1958b), Arch. Biochem.
Biophys. o, 112.

Zebe, B., Lalbrflck, A., and Bucher, Th. (1957), Anwew.
Chem._59, 56; Ber gee Physiol. Bxptl. Pharmakol. l8 , 115.

 

 

Zebe, E. (1960), Biochem. 2. 332, 328.

APPENDIX

117

 

 

.000000000 .mm .0000 .0000000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.0000000 ..0.0 .0000 000 .0.0 .000000 0.0 0.00 000.000 0000000000000

I .Ammmav mmmsww

000 0000 ..0000 .0000 .0 ..0.0 .00000002 0.0 0.0 000.000 -0000000 000000
.000000000 .mm ..0000000

.0000000 .0000 ..0 .0000000 000 h.e .00000000 0.0 0.00 000.000 00000000

.0000000000 ..000 .0000 .0 ..0.0 .0000 0.0 0.00 000.000 0000000 00000000
:1: .0000000000 .0 .000000000

-000 ..0.3 .0000 000 ..0.: .000000000 00.0 00.00 000.000 000000 000>0000
I 0000000000 .mwm

.e000 .0000 .0 ..0.0 .0000 ..0 .0000000 00.0 00.0 000.000 00000000
.00 may 900000000 090 208200300

.0000000000 . .000000000000 ..0 .0000000 00.0 0.0 000.000 00000000
.0000000000 .0

.000000000000 ..0 .0003 000 ..<.0|»00000003 0.0 0.0 000.00 0000000
.000000000 .0 .0000000

mmoamsoo .pQH £00.000m ..m.m .00£0000 m0.m 00.0 000.00 wmmahswld

000:000000500

m0mo£p 00:0 wwm om.0 00.0 000.00 mpmSQmOSQOHmomflwld
-, .000000000 .00

“h0000000 .pmmem< h.3 .000000000 00.0 00.0 000.00 0000000000
0000000000 .00 .000 .0000 .00 .0 ..0 .0000020

000 ..0 .00000 ..0 .0000000 ..0 .0000000 00.0 00.0 000.00 000000000
.I. .00000000 .00

.0000000 .0 ..0 .000000 ..0 .00000 00.0 00.0 000.00 0000000000000-0
III. I. .000000000

.000 ..0000 .0000 .0 ..3.0 .000300m 0.0 00.0 000.00 0000000000000000
.000000000 .000 ..0000 .0000 .0

..0 .03000 000 ..0 .0mmm00 ..0 .00000 00.00 00.0 000.00 000000
I .000000000 .000 ..0000 .0000

.0 ..3 .00000 000 ..0 .00000 ..0 .00000 0.00 00.0 000.00 00000

. . 000003
002000000 3 om 3 0N0 000500002 Q0mponm
0 0cm 0

0005000 :0 000: mus0000wwmoo £00000s080dmm 0:0 :00030000 0:» 000 modam>

.0 00000

118
The theoretical values for the sedimentation coeffici-
ents of Spheres as a function of of molecular weight(see
Figure 5) were calculated in the following manner.
The frictional coefficient(f) for a Sphere is given

by Stokes .Law as,

f0 = {YURI‘O 1)
Y): viscosity of solvent
at 20°, in this case
water(0.0lO“‘ poise)

r = radius of a Sphere

subscribt
sphere

,0, refers to a

The radius of a Sphere is defined as,
r =(0g)1/3 2)
0 Tr
Q): volume of Sphere
The volume of a Sphere of a given molecular weight

is given as,

v = R71: 3)
‘ partial specific
volume

<|
n

M = molecular weight of
sphere

N = Avoqadro's number
(6.023 x 1023)

Combination of equations 1), 2) and 3) yields the

following for the frictional coefficient of a Sphere

f0 = 69WWl/3 14')

The sedimentation coefficient, S, is defined as
S : Hgl - vp) 5)
N f
f9: density of solvent

119

Substitution of the value for the frictional coeffi-
cient, equation 4% into equation 5) yields a value for

the sedimentation coefficient of a sphere.

 

z m - 391 1m 1/3 6)
SO (flTNn (3%)

Rearrangement of equation 6) zives,

so .._._____Iv12/3 __ (4 1/3 1 - W 7)
("N)2/3 6n '3' (W173

 

Substitution of the known constants into equation

7) yields the following relationship.

8 = (391)2/3 g1 - Eb)x 1.199 x 10‘15 8)
0 ml 3

For the calculations of the values used in Figure

5, V'was set equal to 0.73 cc/g.

120

The theoretical values for the diffusion coeffici-
ents of spheres as a function of molecular weight(see Fig-
ure 6) were calculated in the following manner

The diffusion coefficient is defined as follows

D = _L£. 1)

k = Boltzmann's gonstant
(1.38 x 10'1 )

t = absolute temperature

f = frictional coeffi-
cient

The frictional coefficient for a Sphere is given
by Stckes law as,
f0 = oflWLrO 2)
r = radius of Sphere
n = viscosity of solvent
in this case water at

200(0.01oo5 poise)

subscribt,o, refers to
a sphere

Substitution of of the value for fO in equation 2)

into equation 1) yields,

D = kt
O Efinro 3)
The radius of a Sphere is defined as

I‘0 = %&}/B a)

v = volume of Sphere
The volume of a Sphere of a given molecular weight

is defined as

121

_ 7H1
V _ N 5)

v = partial Specific
volume

H

molecular weight
of Sphere

N = Avogsdmfs nggber

Combination of equations 3), 4) and 5) yields an
equation which can be used to calculate the diffusion

coefficient of a Sphere of a given molecular weight.

D : Lrt (”W‘:18 __l__ /3 O)
O gfifi 3 $11

Substitution of the known constants into equation 6)

gives the following,

-4 '
D0 = 2.905 x 10 J(*;;$/3

v14

For the calculations of the theoretical diffusion
coefficients listed in Figure 6, 5 was set equal to 0.73

cc/g.