THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 241, No. 21. Issue of November 10, PP. 4966-4976, 1966

Printed in U.S.A.

Glucose 6-Phosphate Dehydrogenase of Human Erythrocytes

I. PURIFICATION AND CHARACTERIZATION OF NORMAL (B+) ENZYME*

(Received for publication, April 22, 1966)

AKIRA YOSHIDA

From the Division of Medical Genetics, Department of Medicine, University of Washington School of Medicine, Seattle, Washington 98105

SUMMARY

Human erythrocyte glucose 6-phosphate dehydrogenase (o-glucose 6-phosphate :nicotinamide adenine dinucleotide phosphate oxidoreductase, EC 1.1.1.49) was purified by column chromatography with diethylaminoethyl cellulose, calcium phosphate gel, carboxymethyl cellulose, diethyl- aminoethyl Sephadex, and carboxymethyl Sephadex. A

homogeneous preparation was obtained in over-all yield of about 50%. The specific activity (750) and the yield were significantly greater than those so far reported for the en- zyme.

The sedimentation patterns of analytical ultracentrifuga- tion and interference pattern of sedimentation equilibrium in- dicated a homogeneous preparation. The sedimentation constant (s%J was 10.0 S at a protein concentration of 0.3 %. The molecular weight was estimated as 240,000.

The molecular weight was about 43,000 in 4 N guanidine- HCl, indicating that the enzyme consisted of six similar size subunits.

Removal of nicotinamide adenine dinucleotide phosphate from the enzyme caused dissociation into inactive (or weakly active) subunits with a molecular weight of about one-half of the native protein.

The enzyme was partially inactivated by dilution without dissociating into subnuits.

Optimal pH and Michaelis constants for the primary sub- strates were determined. Analogues of n-glucose 6-phos- phate accepted as substrates were 2-deoxy-D-glucose 6-phos- phate and D-galactose 6-phosphate. Nicotinamide adenine dinucleotide was a weak substrate.

Amino acid composition of the enzyme was determined.

Because of the existence of many genetic variants in man, glucose 6-phosphate dehydrogenase (n-glucose 6-phosphate: nicotinamide adenine dinucleotide phosphate oxidoreductase, EC 1 .l. 1.49) from erythrocytes is of particular interest in studies of human biochemical genetics. About 20 types of glucose 6-phosphate dehydrogenase, which are distinguishable

* This research was supported by Research Grant H 3901 from the National Institutes of Health and Institutional Cancer Grant 11-3786.

by electrophoretic mobility in starch gel, or by enzymic character- istics, or by both methods, have been reported (l-3). Thus, about 20% of American Negroes have an electrophoretically rapid variant of enzyme with normal enzyme activity, while another 10 to 15y0 have an electrophoretically similar variant with deficient activity (1). The deficient glucose 6-phosphate dehydrogenase variant found in many Mediterranean populations has normal electrophoretic migration and a high relative rate of utilization of substrate analogues (i.e. galactose 6-phosphate or 2-deoxyglucose 6-phosphate) (4). Deficiencies of the enzyme have been related to drug- and food-induced anemia and to chronic hemolytic diseases (2, 5, 6).

In spite of the efforts of several investigators (7-9), the enzyme has not yet been sufficiently purified to elucidate genetic altera- tions on the basis of molecular structure. In an attempt to work out hereditary alterations and defects of the enzyme on the molecular level and to understand the nature of alterations of enzyme specificity introduced by mutational change, glucose 6- phosphate dehydrogenases from blood of normal and mutant human individuals is being purified and characterized with re- spect to chemical, serological, and enzymic properties. This paper provides data on isolation and purification of glucose 6- phosphate dehydrogensse from normal human blood. Molecular weight, amino acid composition, and some enzymic character- istics of a homogeneous preparation are described. It has been known that the activity of glucose B-phosphate dehydrogenase from human erythrocytes is markedly reduced when bound NADP is removed (l&13). This reversible inactivation has been attributed to dissociation of the enzyme into subunits. Contradictory results regarding molecular size of the active enzyme and its inactive subunits have been presented with the use of partially purified preparations (12, 13). In this paper, the molecular weight of the enzyme has been determined by a sedi- mentation equilibrium method under various conditions, and the relationships between enzyme activity and structural changes are discussed.

MATERIALS AND METHODS

Human blood was obtained from the King County Blood Bank (Seattle, Washington). The blood (clotting was prevented by acid-citrate-dextrose, ACD Formula A) had been stored at 4” for 2 to 4 weeks. Several milliliters of each blood specimen were hemolyzed separately and the enzymic activity and electro- phoretic mobility in starch gel electrophoresis (14) were meas-

4966

by guest on February 4, 2020http://w

ww

.jbc.org/D

ownloaded from

ured. Only blood containing normal type (B+) of the enzyme bath. All the buffer solutions used for the preparation contained was pooled and used for these investigations. 10m3 M mercaptoethanol, lop3 M EDTA, and 2 X 10m5 M NADP,

Ion &changer-DEAE-cellulose (0.78 and 0.9 meq per g) and unless otherwise stated. CM-cellulose1 (0.6 meq per g) were purchased from Bio-Rad. DEAE-Sephadex (A-50, 3.5 meq per g, 40- to 120~M particle size) Hemolysis of Red Cells

and CM-Sephadex (C-50, 4.5 meq per g, 40- to 120-p particle About 2 liters of blood were centrifuged at 15,000 x g for 20 size) were from Pharmacia, Uppsala. min, and supernatant plasma was removed by a pipette attached

Culcium Phosphate Gel-Hydroxylapatite was prepared by the to an aspirator bottle. The red cells were washed twice by cen- method of Tiselius, HjertCn, and Levin (15). trifugation with 2 volumes of 0.15 M NaCl containing 10V3 M

BuJer-Phosphate buffer was Ka2HP04-KH2P04; Tris buffer EDTA, and were lysed by mixing with an equal volume of water was Tris-HCl; acetate buffer was acetic acid-sodium acetate; containing 1O-3 M EDTA, 1O-3 M mercaptoethanol, and 50 ml of glycine buffer was glycine-NaOH. toluene. After vigorous shaking for a few minutes, they were

ChemicalsNAD, NADP, n-glucose g-phosphate, n-galactose centrifuged (15,000 X g for 30 min), and the hemolysate was 6+hosphate, 2-deoxy-n-glucose 6-phosphate, 6-phospho-n glu- collected by a pipette attached to an aspirator bottle. The pre- conate, n-glucose l-phosphate, n-fructose l-phosphate, D- cipitated debris was suspended in about 2 volumes of water and fructose 1,6-diphosphate, and glucosamine 6-phosphate were centrifuged. The extract was combined with the first hemoly- obtained from Sigma, Calbiochem, Nutritional Biochemicals, sates. Starting from 2 liters of whole blood, 2 to 2.5 liters of he- and Mann. Guanidine hydrochloride and p-dimethylamino- molysates were obtained. benzaldehyde were recrystallized twice.

Glucose &Phosphate Dehydrogenuse dssuy-The dehydrogenase DEAE-cellulose Column

activity was measured by the initial rate of reduction of NADP Two to 2.5 liters of the hemolysates were placed on a DEAE- at 25”. The increase of absorbance at 340 rnp was recorded on a cellulose column (in a funnel with porous glass filter 10 cm in Gilford automatic recording spectrophotometer. The reaction diameter) prepared from 100 g of DEAE-cellulose powder mixture used for routine assay contained: 7 x 10m5 M NADP, buffered with 0.005 M phosphate buffer (pH 6.4 to 6.5). The 7 x 1OW M n-glucose 6-phosphate, 7 X 1O-3 M MgC12, 0.05 M column was washed with 2 liters of 0.005 M phosphate buffer Tris buffer (pH 8.0), and enzyme. In crude hemolpsate, which (pH 6.4 to 6.5). This procedure removed most of the hemoglo- contains bot.h glucose 6-phosphate dehydrogenase and 6-phos- bin and 6-phosphogluconate dehydrogenase, while glucose 6- phogluconate dehydrogenase (6-phospho-r-gluconate :NADP phosphate dehydrogenase remained on the column. The en- oxidoreductase decarboxylating, EC 1.1.1.44)) the difference of zyme was eluted with 2 liters of 0.1 M phosphate buffer (pH 5.8) the reduction rate of NADP between the reaction mixture containing 0.5 M NaCl. In order to prevent hydrolysis of the containing 7 X 1OW M n-glucose g-phosphate and 7 X lop4 M enzyme by contaminating proteolytic enzymes, diisopropylfluoro- 6-phospho-n-gluconate and the reaction mixture containing only phosphate (final concentration, 4 X 10e5 M) or e-amino-N-caproic 6-phospho-n-gluconate was used to calculate glucose g-phosphate acid (final concentration, lop3 M) was added to the effluents. dehydrogenase activity (16). The compositions of the reaction After adjusting the pH to 6.2 with 0.5 M Na2HP04, solid am- mixtures used for determination of pH dependence, Michaelis monium sulfate (350 g per liter) was added to the effluents. The constant, and substrate specificity are specified below. process from the beginning of washing of the red cells to the

Since the specific activity of the enzyme rapidly declined to a precipitation of the enzyme by ammonium sulfate was completed lower value when the enzyme was diluted below 1 mg of purified within 12 hours. After being kept in the cold overnight, the enzyme per ml (see “Inact.ivation of Glucose 6-Phosphate De- precipitates were collected by centrifugation (20,000 x g for hydrogenase by Dilution”), it was practically impossible to 30 min). examine all of the enzymic properties with fully active enzyme. Most assays were made with the enzyme that had been diluted Curboxymethyl Cellulose Column Chromatography

to 10 to 20 pg per ml in 0.01 M phosphate buffer (pH 7.2) or The precipitate obtained in the previous step (from a total 0.05 M acetate buffer (pH 6.0) containing 5 X 10e5 M NADP, about 11 liters of blood) was suspended in about 250 ml of 0.005

10-3 M EDTA, and 10m3 M mercaptoethanol, and was kept at M phosphate buffer (pH 5.8), dialyzed against the same buffer, O-4” for several hours. After this incubation, the activity of the and centrifuged (37,000 X g for 15 min). The supernatant purified enzyme remained unchanged at least for 2 weeks at solution (total, about 350 ml) was placed on a CM-cellulose

04” and was about 25% as active as fully active enzyme (see column (5 x 35 cm) buffered with the same buffer as above. below). In studies on dilution effect and turnover rate, the The enzyme was eluted with increasing concentrations of NaCl

assays were made immediately (within 5 see) after dilution of a from 0 to 0.3 M and increasing pH. The gradient was produced concentrated fully act.ive enzyme solution ( >5 mg per ml). by adding 0.05 M phosphate buffer (pH 6.4) containing 0.3 M

Enzyme activity is defined as micromoles of NADP reduced NaCl into a mixing chamber which contained 1 liter of 0.005 M

per min at 25”. Specific activity is the number of enzyme units phosphate buffer (pH 5.8) (fixed initial buffer volume). The per mg of protein. Protein was assayed by Lowry’s method flow rate was 1 ml per min. An elution pattern is shown in Fig. (17), with crystalline bovine serum albumin as a standard. 1. The major enzyme fraction was treated with ammonium

sulfate (360 g per liter). EXPERIMENTS AND RESULTS

Isolation of Glucose B-Phosphate Dehydrogenuse First Calcium Phosphate Gel Column Chromatography

All the procedures were carried out in the cold or in an ice The precipitate obtained in the previous step was suspended

in 0.01 M phosphate buffer (pH 6.8) and dialyzed against thesame ‘The abbreviation used is: CM, carboxymethyl. buffer. The dialyzed solution (75 ml) was placed on a calcium

Irsue of November 10, 1966 A. Yoshida 4967

by guest on February 4, 2020http://w

ww

.jbc.org/D

ownloaded from

4968 Glucose &Phosphate Dehydrogenase from Human Erythrocytes Vol. 241, No. 21

EFFLUENT ML

c 3

i : : ‘.

I _ I -.-- . 100 200 300 400 500 600

EFFLUENT ML

100 200 300 400 500 600 700 800 EFFLUENT ML

6

100 200 300 400 500 600 700 EFFLUENT ML

40

30

!O

IO

FIG. 1. Elution pattern from a CM-cellulose column. Partially purified glucose 6-phosphate dehydrogenase (obtained through a DEAE-cellulose column) was placed on a CM-cellulose column (5 X 35 cm) and eluted with increasing pH and concentration of N&l. O---O, absorbance; A- --A, dehydrogenase activity.

FIN. 2. Elution pattern from a calcium phosphate gel column. Partially purified glucose g-phosphate dehydrogenase (peak from a CM-cellulose column) was placed on a calcium phosphate gel column (2.5 X 30 cm) and eluted with phosphate buffer (pH 6.8) of in- creasing concentration. O-O, absorbance; A- --A, dehydrogenase activity.

FIG. 3. Elution pattern from a DEAE-Sephadex column. Partially purified glucose 6-phosphate dehydrogenase (peak from first calcium phosphate gel column) was placed on a DEAE-Sephadex column (1.5 X 35 cm) and eluted with increasing concentration of NaCl. O-O, absorbance; A- --A, dehydrogenase activity.

FIG. 4. Elution pattern from a calcium phosphate gel column. Partially purified glucose 6-phosphate dehydrogenase (peak from first DEAE-Sephadex column) was placed on a calcium phosphate gel column (1.5 X 30 cm) and eluted with phosphate buffer (pH 6.8) of increasing concentration. O--O, absorbance; A- - -A, dehydrogenase activity.

phosphate gel column (2.5 x 30 cm) buffered with 0.01 M phos- DEAE-Sephadex column (1.5 X 35 cm) buffered with 0.02 M

phate buffer (pH 6.9, and was eluted with phosphate buffer phosphate buffer (pH 6.4) and was eluted with increasing con- (pH 6.8), the concentration of which was gradually increased centrations of NaCI. The gradient was produced by adding from 0.01 M to 0.2 M. The gradient was produced by adding 0.02 M phosphate buffer (pH 6.4) containing 0.25 M NaCl to a 0.2 M buffer to a mixing chamber which contained 500 ml of mixing chamber which contained 400 ml of 0.02 M phosphate 0.01 M buffer (fixed initial buffer volume). The flow rate was buffer (pH 6.4) containing 0.05 M NaCl (fixed initial buffer 30 ml per hour. An elution pattern is shown in Fig. 2. Glu- volume). case 6-phosphate dehydrogenase was eluted at phosphate buffer An elution pattern is shown in Fig. 3. This procedure removed concentration ranging from 0.065 M to 0.12 M. The bulk of the the bulk of yellow material and the enzyme fraction became al- enzyme fraction was precipitated with ammonium sulfate (350 g most colorless. The dehydrogenase fraction was precipitated per liter). with ammonium sulfate (350 g per liter).

First DEAE-Sephadex Column CM-Xephadex Column Chromatography

The precipitate obtained in the previous step was suspended The enzyme precipitate was dissolved in 0.005 M phosphate in 0.02 M phosphate buffer (pH 6.4) and dialyzed against the buffer (pH 5.8) and dialyzed against the same buffer. The same buffer. The dialyzed solution (35 ml) was placed on a dialyzed solution (12 ml) was placed on a CM-Sephadex column

by guest on February 4, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of November 10, 1966 A. Yoshida

(1 X 30 cm) buffered with 0.005 M phosphate buffer (pH 5.8), and was eluted with increasing concentrations of N&l and rising pH. The gradient was produced by adding 0.02 M phos- phate buffer (pH 6.4) containing 0.15 M NaCl to a mixing cham- ber which contained 400 ml of 0.005 M phosphate buffer (pH 5.8) (fixed initial buffer volume). Flow rate was 15 ml per hour. The enzyme was eluted between 100 and 400 ml of effluent. This fraction was collected and treated with ammonium sulfate (350 g per liter).

Second Calcium Phosphate Gel Column Chromatography

The precipitate was dissolved in 0.01 M phosphate buffer (pH 6.8) and dialyzed against the same buffer. The dialyzed solution (about 15 ml) was placed on a calcium phosphate gel column (1.5 x 30 cm) buffered with 0.01 M phosphate buffer (pH 6.8), and was eluted by increasing concentrations of buffer. The gradient was produced by adding 0.2 M phosphate buffer (pH 6.8) to a mixing chamber which contained 500 ml of 0.01 M phosphate buffer (pH 6.8) (fixed initial buffer volume). Flow rate was 20 ml per hour. An elution pattern is shown in Fig. 4. The enzyme fraction was treated with ammonium sulfate (350 g per liter). After this stage of purification, the protein was easily denatured on the surface of glassware and in foam. All glassware, including the chromatographic column, was siliconized by treatment with dimethyldichlorosilane (Applied Science Lab- oratories). The enzyme solution was treated extremely gently, avoiding foam as much as possible.

Fractionation with Ammonium Sulfate

The precipitate obtained in the previous stage was dissolved in the smallest feasible amount of 0.05 M phosphate buffer (pH 6.4), and a small amount of insoluble material was removed by centrifugation. Saturated ammonium sulfate solution was added drop by drop to the supernatant solution until the solution became slightly turbid. After keeping overnight, the suspension was centrifuged. Most of the enzyme activity remained in the supernatant, and a small amount of inactive precipitate was discarded again by centrifugation.

Saturated ammonium sulfate was added to the supernatant solution until it became slightly turbid, and it was kept at room temperature for a few hours. The precipitate, which contained about 70% of the activity, was collected by centrifugation at room temperature (first precipitate). Saturated ammonium sulfate solution was added to the supernatant solution until it became slightly turbid at 0”. The enzyme was collected by centrifugation after keeping the solution at room temperature for a few hours (second precipitation). Both precipitates were combined and used for further purification.

The specific activity of the enzyme at this stage of purification (specific activity of fully active enzyme was more than 570, and specific activity after partial inactivation by dilution was 127 to 146) was much higher than that of the most highly purified preparation so far reported (113 units), which was assumed to be about 80% pure (9). The enzyme preparation showed a single protein band on disc electrophoresis (Canalco model 12) in polyacrylamide gel. Homogeneity of the enzyme preparation was examined by analytical centrifugation in a Spinco model E centrifuge. The sedimentation pattern indicated that the preparation contained a major high molecular weight fraction together with a minor lower molecular size fraction.

I

1 /J--J r; * :L I -, 100 200 300 400

4969

00 1

80 ;

60 2

40 2

20

EFFLUENT ML

FIG. 5. Elution pattern from a DEAE-Sephadex column. Par- tially purified glucose g-phosphate dehydrogenaae (peak from second calcium phosphate gel column) was placed on a DEAE- Sephadex column (1 X 30 cm) and eluted with increasing concen- tration of NaCl. O--O, absorbance; A- - A, dehydrogenase activity.

Second DEAE-Sephadex Column Chromatography

The precipitate obtained in the previous step was dissolved in 0.02 M Tris buffer (pH 8.0), and the solution was dialyzed against the same buffer and placed on a DEAE-Sephadex column (1 x 30 cm) buffered with 0.02 M Tris buffer (pH 8.0). The enzyme was eluted by a gradient obtained by adding the buffer containing 0.5 M NaCl to a mixing chamber which contained 400 ml of the buffer (fixed initial buffer volume).

An elution pattern is shown in Fig. 5. The major dehydro- genase fraction (260 to 325 ml of the effluent) was precipitated by ammonium sulfate (350 g per liter).

Crystallization and Precipitation by Ammonium Sulfate

The precipitate was dissolved in about 1 ml of 0.05 M acetate buffer (pH 6.0) and a small amount of insoluble material was discarded after centrifugation. Saturated ammonium sulfate solution was added to the supernatant until the solution became turbid. The suspension was kept at room temperature for a few hours, and centrifuged at room temperature. The supernatant, which contained less than 2% of the total activity, was discarded. The precipitate was dissolved in the smallest feasible amount of 0.05 M acetate buffer (pH 6.0) and ice-cold saturated ammonium sulfate solution was added drop by drop until the solution be- came slightly turbid. The enzyme suspension was kept in an ice bath for a few days. A silky sheen appeared and fairly homogeneous tiny needle crystals were observed (Fig. 6).

However, more than 70% of the enzyme activity still remained in the supernatant. I f more ammonium sulfate solution was added or if the supernatant was kept at room temperature for several hours in order to precipitate all the enzyme, the precipi- tates were no longer homogeneous crystals, but appeared as a mixture of somewhat bigger rod type crystals and cubic type particles which apparently were not crystals. Specific activities of the needle crystals and noncrystalline precipitate were identical (170 to 180 at the stationary state after dilution inactivation). As is discussed below, the enzyme has more than one form, de- pending on conditions, so that the crystals and the amorphous precipitate must be essentially the identical protein.

Yield and activity of the enzyme at various steps of purification are shown in Table I. It should be mentioned that the dehydro-

by guest on February 4, 2020http://w

ww

.jbc.org/D

ownloaded from

4970 Glucose 6-Ph.osphate Dehydrogenase from Human Erythrocytes Vol. 241, No. 21

genase showed a greater total activity when it was more concen- of purified glucose 6-phosphate dehydrogenase reprecipitated trated. Starting from about 11 liters of blood which contained (recrystallized) twice. 5.75 x 103 units of the enzyme, about 16 mg of the purified preparation (total activity, 11 X lo3 units in full active state, Measurement of Sedimentation Constant 2.7 X 103 units after dilution inactivation; see “Inactivation of Glucose 6-Phosphate Dehydrogenase by Dilution”) have been

Ultracentrifugation experiments were carried out in a Spinco

obtained. model E centrifuge. The protein was dissolved in 0.05 M acetate

All of the following experiments were carried out with the use buffer (pH 6.0) containing 10-s M EDTA, 10-S M mercaptoeth- anal, and 2 X lop5 M NADP at a concentration of about 0.3%. Schlieren patterns of the enzyme showed single and symmetrical sedimentation boundaries (Fig. 7).

The sedimentation constant (szo,,) was calculated as 10.0 S.

Measurement of Molecular Weight

The molecular weight of the enzyme and subunit molecular size were determined by the sedimentation equilibrium method (18) with the use of Rayleigh interference optics in a Spinco model E centrifuge.

The protein solutions used for the determination were as follows.

Solution I-The enzyme was dissolved in 0.05 M acetate buffer (pH 6.0) containing 10m3 M EDTA, 10-s M mercaptoethanol, and 2 x lOti5 M NADP, and dialyzed against the same buffer. The concentration of the protein was 0.015% and 0.008%, and its specific activity was 170 to 180.

Solution II-The enzyme solution (about 8 mg per ml) was mixed with 10 volumes of ice-cold 60% saturated ammonium sulfate containing 0.05 M H2S04. After keeping for 10 min at O”,

FIG. 6. Micrograph of glucose 6-phosphate dehydrogenase the mixture was centrifuged. The precipitate was suspended in crystals (see text). X ~650. the same ammonium sulfate solution as above, kept for 10 min at

Steps of purification Total volume Total protein

Hemolysate DEAE-cellulose column ef-

fluent Ppt with (NHa)$04 and resus-

pended in buffer CM-cellulose column effluent Ppt with (NH1)804 First calcium phosphate gel

column effluent First DEAE-Sephadex column

effluent CM-Sephadex column effluent Ppt with (NHI)#O~ and dis-

solved in buffer Second calcium phosphate gel

column effluent (NH&S04 fractionation, dis-

solved in buffer Second DEAE-Sephadex col-

umn effluent Crystallization and precipita-

tion

ml

- 15 x 103 -1.3 x 103

300

+w

2,030 X lOa 24.6 X lo3

65

13

280

3.5

65

1.7

13.7 x lo-3

5.35 x 103 2.8 X 1O8 5.5 x 102

1.85 x 102

55 50

37

28

18

15.6

-

-

-

TABLE I

-

Total activity

Glucose-&P dehydrogenase

6.Phosphoglu- conate dehy-

drogenase

units

5.75 x 108 5.5 x 108

5.5 x 103

4 x 103 3.9 x 108 3.7 x 103

3.5 x 103

3.2 X lOa 7.5 x 103

(3.7 x 103)” 3.4 x 103

16 X lo3 (3.5-4.1 x 103)b

3 x 103

11.5 x 103 (2.8 x 108)b I -

2.1 x 103 <loo

<lO

<O.l

-

-

Specific activity of glucose-6-P dehydrogenase

2.83 X lG+ 2.33 X 19-l

4.02 X 10-l

7.5 x lo-’ 14 x 10-1

6.72

19

58 150 (74P 92

570 (127-146)b

165

730 (18O)b

1

.-

-

142

265 495

2,380

6,700

20,600 53,000

(26,169)” 32,300

2O1,oclo (45,999-51,590)b

54,500

258,000 (63,500)”

a Ratio of specific activity of the preparations taking that of hemolysate as 1.00. b Activity measured after dilution inactivation (see text). The concentrated enzyme solution was diluted (lOO- or %%fold) in a

buffer and kept at 9-4” for several hours before the activity measurement.

by guest on February 4, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of November 10, 1966 A. Yoshida 4971

0”, and centrifuged. This procedure eliminates tightly bound NADP from the enzyme and reduces enzyme activity (13). In this particular instance, the specific activity was 70 (60yo in- activation) at a protein concentration of 0.03yo and 35 (80% inactivation) at a protein concentration of 0.003% after keeping the enzyme solution for 1 hour at room temperature. The precipitate w&s dissolved in 0.05 M acetate buffer (pH 6.0) and dialyzed against the ssme buffer. The protein concentrations used for the molecular weight measurement were about 0.020/, and 0.008%.

Solution III-The protein was dissolved in 4 M guanidine hydrochloride solution containing 0.05 M acetate buffer (pH S.O), 10-3 M EDTA, and 10m2 M mercaptoethanol, and dialyzed against the same solvent. In this solvent, the enzyme had no enzymic activity. The protein concentration used for the measurement was about 0.02%

Molecular weight was calculated by the equation (18)

2K = 2.363 (d log Ay/dr) mRT rav WP (1 - ia

where R = gas constant, T = absolute temperature, m = mag- nification factor, TaV = radius to middle point of fringes, @ = partial specific volume of protein, p = density of solvent, w = angular velocity, and d log Ay/dr = slope of logarithm of fringe displacement with respect to radius.

If a given protein solution is homogeneous, plots of log Ay against radius should be linear. If protein molecules reversibly dissociate into small subunits at a range of concentration used for measurement, the plots should be curvilinear. Higher d log Ay/dr values should be obtained at larger rather than at smaller r values. Because the protein solution is more concentrated near the bottom of the cell, a greater fraction of the molecules should associate. By contrast, in the upper layer of the cell, the protein solution is more diluted and a greater part of the molecule should be in the dissociated form.

Some examples of the relationships between the logarithm of the fringe displacement and the radius obtained from the inter- ference patterns are shown in Fig. 8, A to C. The interference patterns obtained in the solution of the native enzyme (Solution I described above) and the solution of the denatured enzyme (Solution III described above) showed no concentration depend- ence and indicated a high degree of homogeneity of protein size.

In protein Solution I, the molecular weight was estimated as 240,000 f 8,000, based on a partial specific volume value of 0.731 ml per g, which was calculated for the amino acid composition of the enzyme according to the method of Cohn and Edsall (19).

In protein Solution III, the molecular weight was estimated as approximately 43,000 f 3,000, correcting change of partial specific volume value due to denaturation and unfolding in guanidine-HCl (0.724 ml per g with correction) and amount of guanidine-HCI bound with protein (about 5%) (20). This result, together with the molecular weight value of undenatured active enzyme (240,000), indicates that the enzyme consists of six similar size subunits.

The interference patterns of the partially inactivated enzyme treated by acidic ammonium sulfate solution showed concentra- tion dependence (Fig. 8B). The molecular weight was estimated as about 180,000 i 10,000 near the bottom of the cell and as about 123,000 i 10,000 in the upper layer of the cell. Thus the

results indicate that removal of the bound NADP induces dis- sociation of the enzyme accompanied by inactivation.

Amino Acid Composition

The purified enzyme was dialyzed against water and lyophi- lized. The protein was hydrolyzed for 20 hours at 110” in hydrochloric acid of constant boiling point in an evacuated, sealed tube. The hydrolysates were evaporated to dryness under reduced pressure in the presence of NaOH.

In order to estimate the cystine (or cysteine) content, a sample of oxidized protein was subjected to amino acid analysis. The procedure used for the oxidation of the protein with performic acid was essentially that described by Schram, Moore, and Big- wood (21). After oxidation, the reaction mixture was diluted with 10 volumes of cold water, and the solution was immediately lyophiliied. The samples were dissolved in a small amount of water, and the solution was again lyophilized and hydrolyzed as described above.

The amino acid analysis was performed with a Technicon amino acid analyzer.

Tryptophan was estimated by p-dimethylaminobenzaldehyde method (22,23).

The amino acid composition of the enzyme is presented in Table II. The number of amino acid residues in 1 enzyme mole- cule (mol wt 240,000) and the number of amino acid residues per subunit, assuming tentatively that the enzyme consists of six identical subunits, are also given in the table.

FIG. 7. Schlieren pattern of purified glucose 6-phosphate de- hydrogenase in the centrifuge at 56,740 rpm 32 min after reaching final speed at 20”. The concentration of the protein was about 0.3% in 0.05 M acetate buffer (pH 6.0) containing IO-3 M EDTA, 10-a M mercaptoethanol, and 2 X lU+ M NADP.

by guest on February 4, 2020http://w

ww

.jbc.org/D

ownloaded from

2. a-

:

Some of the enzymic properties of glucose 6-phosphate de-

$ 2.6- hydrogenase from human red cells have been examined previously 2.4-

2

by several investigators (4, 25, 26) with the use of partially 2.2-

t purified enzyme preparations. Although most of the reported

8 2.D properties are probably genuine characteristics of the enzyme, 2 1.8- Q

some fundamental properties, i.e. optimal pH, Michaelis con-

$ 1.6- 1.4- stant, and substrate specificity, as well as unique phenomena of

$ 1.2- dilution inactivation of the enzyme, were examined in more

~ l.O- detail with the purified homogeneous enzyme preparation.

2 .8- Optimal pH-The effect of pH was examined with acetate

.6- buffer (pH 5.0 to 6.0), Tris buffer (pH 7.0 to 9.0), and glycine

I buffer (pH 9 to 10). Phosphate buffer was not used since it .4 6 0 1.0 1.2 1.4 1.6 inhibited the reaction. As Fig. 9 shows, the optimal pH for the

RADIUS COORDINATE, mm enzymic reaction was pH 8.0.

2.8- B E$ect of Substrate Concentration-The rate of the enzymic ,.

$ 2.6- reaction was examined as a function of the concentration of

5 2.4- NADP and glucose 6-phosphate. The usual Michaelis-Menten

!$ 2.2- relationship was observed, since a Lineweaver-Burk plot (27) iy 2.0- gave a straight line (Fig. 10). The apparent Michaelis con- 3 1.8-

b 1.6- TABLE II p 1.4-

$ Amino acid composition of glucose 6-phosphate dehydrogenase

1.2-

5 l.O- .8- Percentage of Assumed no. of Assumed no. of

Amino acid amino acid residues per residuesa IllOlEXUl~~

residues per subunitC

.6-

.4 Aspartic acid. 10.85 f 0.10 227 38

0.2 0.4 0.6 0.8 1.0 1.2 1.4 Threonined. 3.97 zk 0.17 94 16

1.6 1.8 RADIUS COORDINATE, MM Serinee. 4.07 f 0.20 114 19

Glutamic acid. 13.79 zt 0.16 260 43 Proline 5.82 f 0.08 144 24

$ 2.2

1

Glycine 3.89 zt 0.05 164 27 Alanine......... 3.88 f 0.01 132 22

2 2.0 Half-cystinef 1.68 40 7

i$ 1.8

/-

Valine . . . 5.21 f 0.01 126 21

3 1.6 Methionine 2.86 52 9

2 1.4 Isoleucine 4.37 f 0.01 93 17

3 1.2 Leucine. 9.07 zt 0.23 193 32 Tyrosine. 5.45 80 13 Phenylalanine. 6.46 f 0.26 105 18 Lysine . 6.22 f 0.09 117 19 Histidine . 3.06 f 0.05 54 9

.5 I.0 1.5 2.0 2.5 3.0 3.5 Arginine 7.02 zk 0.17 108 18 RADIUS COORDINATE, mm Tryptophano.. 2.29 30 5

FIG. 8. The logarithm of the fringe displacement as a function A mmoniah 1.56 f 0.03

of the distance from the center of the rotor. The data were obtained from the interference pattern of sedimentation equilib- Total . . . 100 rium by comparator reading. Magnification factor (m) = 2.229; vertical axis, log (millimeters of fringe displacement X 100); axis of abscissa, millimeters of comparator coordinate.

a Mean value and deviations of analysis of native and performic A, initial

concentration of the enzyme was 0.015% in 0.05 M acetate buffer acid-oxidized protein. The contents of methionine sulfone and

(pH 6.0) containing lO+ M EDTA, lo+ M mercaptoethanol, and tyrosine of performic acid-oxidized protein have not been included

2 X lO+ M NADP. After 18 hours at 23,150 rpm and at lO.l”, in the mean value because of uncertainty of their recovery.

average radius was 6.6 cm. B, initial concentration of the enzyme Total of individual amino acid residues is taken as 100.

treated nith acidic ammonium sulfate was 0.02% in 0.05 M acetate b The nearest integral number to the calculated number of

buffer (pH 6.0) containing lO+ M EDTA, lo+ M mercaptoethanol, amino acid residues per molecule based on molecular weight of and 2 X 1Om6 M NADP. After 18 hours at 23,150 rpm and at 12.5”, 240,000. average radius was 6.5 cm. C, initial concentration of the enzyme c The nearest integrated number to the calculated number of was 0.02% in 4 M guanidine-HCl containing 0.05 M acetate buffer (pH 6.0), low3 M EDTA, and lo+’ M mercaptoethanol. After 20

amino acid residues per subunit, assuming tentatively that the

hours at 27,690 rpm at 14.3”, average radius was 6.5 cm. enzyme consists of six identical subunits (see text).

d The recovery after hydrolysis is assumed to be 95% (24).

Enzymic Properties 6 The recovery after hydrolysis is assumed to be 90% (24). J Calculated from the cysteic acid content in performic acid-

Glucose 6-phosphate dehydrogenase catalyzes Reaction 1. oxidized protein. The recovery of cysteic acid is assumed to be

n-Glucose B-phosphate + NADP+ $ 95% (21).

(1) Q Estimated by p-dimethylaminobenzaldehyde method. 6-phosphogluconolactone + NADPH + H+ h Not included in the total.

4972 Glucose B-Phosphate Dehydrogenase from Human Erythrocytes Vol. 241, No. 21

by guest on February 4, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of November 10, 1966 A. Yoshida 4973

$ 40-

*O- / I

6 7 8 9 IO PH

FIG. 9. Effect. of pH on glucose 6-phosphate dehydrogenase activity. Composition of the reaction mixture was 7 X 1OW M

glucose B-phosphate, 7 X W6 M NADP, 7 X 1OW M MgCl, enzyme, and either 0.1 M acetate buffer (pH 6.0), Tris buffer (pH 7.0 to 9.0), or glycine buffer (pH 9 to 10).

stants (K,) for NADP and glucose 6-phosphate are shown in Table III.

Specijic Activity and Rate of Turnover-The specific activity of the fully active enzyme (see “Inactivation of Glucose 6-Phos- phate Dehydrogenase Dilution”) measured in the reaction mixture containing 7 X lop4 M D-glucose g-phosphate, 7 X 10-S M NADP, 7 x lop3 M MgC12, and 0.1 M Tris buffer (pH 8.0) was 750. Based on this value, together with minor correction because of substrate concentration dependence described above, the ap- proximate maximum specific activity that would obtain at optimal pH and under conditions of saturation of the enzyme with substrates was estimated as 900 at 25”. Since the molecular weight of the enzyme is about 240,000, the maximal turnover number is about 2.2 X lo5 moles of substrate per min per mole of enzyme.

Specificity of Substrate AnaloguesAmong a number of carbo- hydrate phosphates examined, only D-galactose 6-phosphate and 2-deoxy-n-glucose g-phosphate were oxidized by the enzyme. The Michaelis constants and relative maximum reaction rates estimated from Lineweaver-Burk plots for these compounds are shown in Table III. Both n-galactose d-phosphate and 2- deoxy-n-glucose 6-phosphate were fairly good substrates at higher concentrations since their maximum reaction rates were 45% and 12% of n-glucose g-phosphate. However, their Michaelis constants were much higher (8.0 x 1OV M and 6.0 X 10-a M) than that for D-glUCOSe 6-phosphate .(3.9 X lop5 M).

Thus, at lower concentrations, their reaction rates were very low. For example, relative rates were 2.4% for D-galactose 6-phos- phate and 3.1yo for 2-deoxy-D-glucose 6-phosphate at a concen- tration of 7 X 10e4 M, taking the rate of D-glucose g-phosphate at the same concentration as 100%.

None of the following compounds were used as substrates: 6-phosphogluconate, D-glucose l-phosphate, n-fructose l-phos- phate: n-fructose 1,6-diphosphate? D-glucosamine 6-phosphate.

NAD was a weak substrate for the dehydrogenase. In con- trast to NADP, the reaction rate for NAD was reduced about

2 Commercially available o-fructose l-phosphate and n-fruc- tose 1,6-diphosphate were oxidized by the enzyme. However, maximum reaction rates for them were identical with that. for D-glucose 6-phosphate itself. It was conclrtded that the apparent reduction of NADP should be attributed to D-glucose 6-phosphate contaminating in the samples.

40% in the presence of 7 X lop3 M MgC12. The Michaelis constant and relative maximum reaction rate for NAD are shown in Table III.

Inactivation of Glucose C-Phosphate Dehydrogenase by Dilution

As has been described in the section of purification procedures, the enzyme showed higher specific activity (and total activity) when it was more concentrated. This phenomenon was ex- amined in more detail by the use of purified homogeneous enzyme preparation.

When the enzyme was diluted in a buffer (pH 6.0 to 8.0, acetate or Tris buffer) from a concentrated solution (more than 5 mg of enzyme per ml) to give a concentration less than 0.1 mg per ml, the specific activity decreased. The decrease WBS more rapid and slightly more severe at 0” than at 25” (Fig. 11). After a certain period of time in dilute solution, some of the lost activity

I/ IGlucose-6-PI (x10-3)

4 8 12 I6 20

400.

FIG. 10. Lineweaver-Burk plot of the effect of substrate con- centration of glucose B-phosphate dehydrogenase. A, efect of NADP concentration. The reaction mixture contained 7 X lo+ M glucose 6-phosphate, 0.1 M Tris buffer (pH 8.0), enzyme, and various concentrations of NADP (e-0). B, effect of glucose B-phosphate concentration. The reaction mixture contained 7 X 10V5 M NADP, 0.1 M Tris bufler (pH 8.0), enzyme, and various concentrations of glucose 6-phosphate (A--&.

TABLE III Michaelis constants and relative reaction rate for primary substrates

and analogues

Substrate Michaelis

constant 6%)

M

D-Glucose 6-phosphatea.. . 3.9 X IO-” D-Galactose 6-phosphatea. 8.0 X W3 2-Deoxy-n-glucose 6-phos-

phate....................6.0 X 1(r3 NADPb....................4.4 X 1O-6 NADc _....____........__.. 4X10-3

Vm,, (relative to glucose 6-

vm,x

phosphate) (relative to

NADP)

100 45

12 loo

4.7

a The reaction mixture contained 0.1 M Tris buffer (pH 8.0), 7 X 1V M MgC12, 7 X 10m5 M NADP, enzyme, and various con- centrations of glucose g-phosphate or its analogues.

b The reaction mixture contained 0.1 M Tris buffer (pH 8.0), 7 X 10d3 MgC12, 7 X lo+ M D-glucose 6-phosphate, enzyme, and various concentrations of NADP.

c The reaction mixture contained 0.1 M Tris buffer (pH 8.0), 7 X 1OW M glucose A-phosphate, enzyme, and various concentra- tions of NAD (without MgC12).

by guest on February 4, 2020http://w

ww

.jbc.org/D

ownloaded from

Glucose 6-Phosphate Dehydrogenase from Human Erythrocytes Vol. 241, No. 21

INCUBATION TIME AFTER DILUTION

FIG. 11. Time course of inactivation of glucose 6-phosphate dehvdrogenase bv dilution. The concentrated solution of glu- cose 6-phosphate” dehydrogenase (7 mg per ml) was diluted in 0.05 M acetate buffer (oH 6.0) containing 10-3 M EDTA, lO+ M

mercaptoethanol, and 2 X 10- M NADP to give a final concen- ‘5 tration of 0.07 mg per ml. O---O, kept at O”, A---& kept at 25”. t , part of the diluted enzyme solution kept at 0” was transferred at 25”, and the activity (0) was measured after keep- ing at 25” for 1 hour.

100

1 I

-2.2 I I I I I I I

-1.8 -1.4 -1.0 -0.6 -0.2 0.2 0.6 I.0 PROTEIN CONCENTRATION LOG (Mg PROTEIN/ ML)

FIG. 12. Specific activity of glucose 6-phosphate dehydrogenase at different concentrations. The enzyme was dissolved in 0.05 M

acetate buffer (pH 6.0) containing W3 M EDTA, 10e3 M mercap- toethanol, and 2 X 1C5 M NADP to give a final concentration shown in the figure. After holding for more than 30 min at 25”, the activity was assayed.

was partially recovered, and reached a stationary state. The recovery was more rapid at 25’ than at 0”, but the specific activities at the stationary state were same in both cases (about 170 to 180 at a concentration less than 0.3 mg per ml).

The relationships between the protein concentration in a buffer (pH 6.0) containing 2 X lo+ M NADP and the stationary specific activity are shown in Fig. 12. The enzyme was fully active (700 to 750 units per mg) at concentrations higher than 2

mg per ml, and it was about 25% active (170 to 180 units per mg) at concentrations ranging from 0.0007 to 0.3 mg per ml.

The presence or absence of NADP (2 X 10m5 M) and bovine serum albumin (5 mg per ml) in the buffer used for dilution showed no effect on the stationary specific activity. The station- ary specific activity is not significantly different at pH 7.0, 8.0 (Tris buffer), and 6.0 (acetate buffer).

The partial inactivation by dilution was reversible, since precipitation of the diluted enzyme by ammonium sulfate and redissolving of the enzyme at higher concentration restored the activity, as repeatedly shown in the process of enzyme purifica- tion (Table I).

DISCUSSION

Since the discovery of glucose B-phosphate dehydrogenase (Zwischenferment) by Warburg and Christian in 1931 (28), the enzyme has been purified to varying degrees from a variety of sources. The enzyme has been crystallized from yeast (specific activity, 680) (29) and bovine udder (specific activity, 68) (30). However, the latter was not homogeneous judging from analytical ultracentrifugation. Homogeneity and molecular size (NADP- free enzyme, mol wt 102,400 f 2,400 at infinite dilution) of the yeast crystalline enzyme were reported recently (31).

Kirkman (7) and Chung and Langdon (9) attempted to purify the enzyme from human red cells. The best preparation so far reported had a specific activity of 113, but the yield was only several per cent (9).

In this work, particular attention was paid to avoid hydrolysis and inactivation of the enzyme by proteolytic enzymes (the major proteolytic enzyme in hemolysates is plasmin) by diisopropyl- fluorophosphate and e-amino-N-caproic acid in the process of purification and to improve the yield of the enzyme. Thus, after more than 20 steps of purification, a homogeneous enzyme preparation (specific activity, 750 in fully active state, 170 to 180 after dilution inactivation) was obtained in an over-all yield of about 50 %.

Controversial results have been reported regarding the molec- ular size of glucose 6-phosphate dehydrogenase of human red cells. Kirkman and Hendrickson (12), using fairly crude en- zyme preparations (0.3 to 0.6 unit per mg of protein), estimated the molecular size from the relative rate of sedimentation in sucrose gradient centrifugation and relative rate of diffusion through sintered glass, together with an arbitrary assumed value of partial specific volume of the enzyme: the molecular weight of the active NADP-bound enzyme was estimated as 105,000 and that for inactive (or less active) subunit as about one-half. Tsutsui and Marks (11) also carried out sucrose gradient centrif- ugation of the enzyme in the presence and absence of NADP. Their results were similar to those of Kirkman and Hendrickson (12).

On the other hand, Chung and Langdon (9, 13), using a more purified enzyme preparation (up to 113 units per mg of protein), estimated the molecular weight as 190,000 from a rough estima- tion of the sedimentation constant by ultracentrifugation and the diffusion constant by the porous diaphragm method. They observed that removal of NADP resulted in dissociation of the enzyme into inactive subunits with a molecular weight of about one-half of the native protein.

This discrepancy could be settled in the present work by more accurate measurement of molecular size of the enzyme under various conditions. Dilution of human red cell glucose 6-phos-

by guest on February 4, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of November 10, 1966 A. Yoshida

phate dehydrogenase below certain concentrations caused partial inactivation of the enzyme (75% inactivation of concentrations less than 0.3 mg per ml) (Fig. 12). This inactivation was en- tirely different from the inactivation of the enzyme due to removal of NADP (lo-13), since NADP had no effect on the former inactivation, while the latter was prevented and reversed by NADP.

The molecular weight of the enzyme measured by the sedi- mentation equilibrium method at concentrations of 0.2 mg and 0.08 mg per ml was about 240,000. The interference patterns showed no concentration dependence. This finding, and the sed- imentation constant (820, ,) of fully active enzyme (measured at a concentration of about 3 mg per ml) of 10.0 S, indicated a value of 240,000 for the molecular weight of the fully active enzyme as well as of the partially inactive enzyme. It is unlikely that some metal contamination caused partial inactivation, since EDTA, reducing agents, and various buffers showed no effect.

The partial inact,ivation by dilution which is NADP-independ- ent may be attributed to structural change of the protein without dissociation into subunits. It should be mentioned that a similar phenomenon, i.e. partial inactivation by dilution without dis- sociation into subunits, has been observed in Bacillus subtilis L-alanine dehydrogenase (32) and L-lactate dehydrogenase (33).

I f one assumes the shape of the protein to be a prolate ellipsoid, one can calculate its major dimensions from the molecular weight (240,000), the partial specific volume (0.731 ml per g), and the sedimentation constant (10.0 S). The major axes are then 30 and 4.3 mp, respectively.

The inactivation by removal of bound NADP can be attributed to reversible dissociation of the protein into subunits, since the molecular weight was estimated BS 123,000 f 10,000 in the upper layer of the cell and as 180,000 Z!Z 10,000 near the bottom of the cell after removal of NADP from the enyme.

The molecular weight value observed in guanidine-HCl (43,000 =I= 3,000) indicates that the enzyme consists of six similar size subunits. It is not yet known whether or not these subunits are identical.3 This problem may be elucidated by peptide mapping of tryptic hydrolysates of the enzyme, which is under study in conjunction with investigations of genetically altered enzyme (A+). The number of tryptic peptides and the fre- quencies of lysine and arginine in the molecule (Table II) will give further information.

Considering the results of this work and the results reported by previous investigators, the following relationship between activity and configuration of the enzyme may be suggested.

The fully active enzyme (Hexamer I, mol wt 240,000), which has probably six tightly bound NADP per molecule,4 consists of

3 Chung and Langdon (9) reported that the enzyme has at least, two peptide chains, one with tyrosine and the other with alanine as amino-terminals. Their enzyme preparatiov,. which they assumed to be about 80% pure, contained impurltles of smaller molecular weight (9). Comparing the activity of their prepara- tion (113 units per mg) and our preparation (750 units per mg in fully active form, 170 to 180 units per mg in partially inactive form), their sample may be about 65% pure, since presumably they measured the enzyme activity in partially inactive form. The observed amino-terminals, therefore, may not, be attributed to glucose 6-phosphate dehydrogenase itself.

4 The content of bound NADP in glucose 6-phosphate dehydro- genase was estimated from the increment of 340 rnp absorbance (13) or fluorescence emission (7) after adding excess n-glucose 6- phosphate to the enzyme. Chung and Langdon reported that 2.09 mg of enzyme, which had a specific activity of 95, contained

NNN N N N N Hexamer -i Tri mer-1 Hexamer-n

-NADP t NADP -NADP tNADP

II

Trimer-II

dilution

Hexamer-El

FIG. 13. Schematic presentation of the relationship of various configurations and activities of glucose g-phosphate dehydrogen- ase. Hexamer I (mol wt 240.000). fullv active (snecific activitv. 750)) exists at higder conce&ati&s (>0.3 mg p&-ml). Hexam& II (mol wt 240,000)) partially active (specific activity, 170 to 180)) exists at concentrations of 0.007 to 0.3 mg per ml. Hexamer III (mol wt 240,000), partially active (specific activity, <170). exists at concentrations-higher-than 0.1 mg per ml in-the absence of NADP. Trimer I (mol wt 120.000). nartiallv active (snecific activity, <150), exists in extremely’&&ed s&ion (<i&-l mg per ml) and appears during the transition between Hexamer I and Hexamer II. Trimer II (mol wt 120,000), inactive (or very weakly active), exists at concentrations lower than 0.1 mg per ml in the absence of NADP.

six similar size subunits. By dilution, Hexamer I dissociates into two partially active (specific activity, less than 20%) sub- units with bound NADP (Trimer I), which eventually reassociate to form a new partially active (specific activity, 25%) hexamer (Hexamer II). At higher concentrations ( >3 mg per ml), Hexamer I is predominant, while at lower concentrations (0.007 to 0.3 mg per ml) most of the enzyme molecules are partially active Hexamer II. Trimer I can be detected only in extremely diluted solution (<10m4 mg per ml), such as that used by Kirk- man and Hendrickson (12) and Tsutsui and Marks (11) for sucrose gradient centrifugation.

Removal of bound NADP from the enzyme by acidic am- monium sulfate treatment and dilution to concentrations of <O.l mg per ml induce dissociation of the enzyme into inactive (or very weakly active) trimers without bound NADP (Trimer II). At higher concentrations (>0.2 mg per ml), a partially active ( <170 units per mg) hexamer without bound NADP (Hexamer III) is predominant (Fig. 8B). Higher temperature (25”) rather than low temperature (0’) favors this formation of partially active hexamers from trimers. When Trimer II is further diluted, it may dissociate into smaller subunits (12). These relations are summarized schematically in Fig. 13.

about 20 mpmoles of NADP (13). From these values, one can estimate that 1 enzyme molecule (mol wt 240,000; stationary specific activity, 170 to 180) contains 4 to 5 NADP molecules. According to Kirkman (7), 0.13 to 0.15 X 100 units of enzyme con- tained 1 mole of NADP. This indicates that 1 enzvme molecule I has about, 6 NADP molecules.

by guest on February 4, 2020http://w

ww

.jbc.org/D

ownloaded from

4976 Glucose G-Phosphate Dehydrogenase from Human Erythrocytes Vol. 241, No. 21

Molecular size of glucose 6-phosphate dehydrogenase from other mammalian species seems to be similar to that of human enzyme, since Nevuldine and Levy (34), using gel filtration methods, estimated the molecular weight of rat mammary en- zyme as about 130,000 in the diluted state and as about a quarter million at higher concentrations ( >1.9 mg of protein per ml) in the presence of NADP.

It should be mentioned that human glucose 6.phosphate dehydrogenase of various tissues including erythrocytes is under control of a single structural gene linked to the sex chromosome (35) ; accordingly, glucose 6-phosphate dehydrogenase from various tissues should be structurally identical.

Michaelis constants of major substrates and relative rate of utilization of substrate analogues measured for the homogeneous glucose 6-phosphate dehydrogenase were not significantly differ- ent from values reported for fairly crude preparations (25). The Michaelis constants of substrate analogues (NAD, n-galac- tose g-phosphate, 2-deoxyn-glucose 6-phosphate) were much higher than those of primary substrates. Mg++, which stimu- lated the NADP-linked reaction (8), inhibited the activity for NAD, as has been observed in glucose 6-phosphate dehydro- genase from other mammalian sources (36).

The specific activity of fully active human glucose 6-phosphate dehydrogenase (750) was similar to that of the crystalline yeast enzyme (680) (29). These values are considerably higher than that of the crystalline enzyme from bovine mammary gland, which had a specific activity of 68 (30). Although the bovine enzyme has been obtained in crystalline form, it was not homog- eneous in analytical centrifugation (30). At the same time, since the activity of the enzyme is so sensitive to variations in its environment, the activity of bovine enzyme might have been measured in a partially inactivated state. True specific activity of bovine enzyme may be of the same order as that of human enzyme.

AcknowledgmentsThe author is most grateful to Professor A. G. Motulsky for his support and advice in this investigation. The author is also grateful to Mrs. L. Steinmann, Miss P. Harbert, and Mrs. A. Morrow for assistance, and to King County Blood Bank, Seattle, Washington, for providing human blood.

REFERENCES

1. BOYER, S. H., PORTER, I. H., AND WEILBACHER, R. G., Proc. Natl. Acad. Sci. U. S., 48, 1868 (1962).

2. KIRKMBN, H. N., ROSENTHAL, I. M., SIMON, E. R., CARSON, P. E., AND BRISON, A. G., J. Lab. Clin. Med., 63, 715 (1964).

3. KIRKMAN, H. N., MCCURDY, P. R., AND NAIMAN, J. L., Cold Spring Harbor Symp. &ant. Biol., 29, 391 (1964).

4. KIRKMAN, H. N., SCHNETTINI, F., AND PICKARD, B. M., J. Lab. Clin. Med., 63,726 (1964).

5.

6.

7. 8.

9.

10.

11.

12.

13.

14.

15.

16. 17.

18. 19.

20.

21.

22.

23. 24. 25. 26.

27.

28.

29.

30.

31.

32.

33. 34.

35.

36. LEVY, H. R., J. Biol. Chem., 238, 775 (1963).

CARSON, P. E., FLANAGAN, C. L., ICKES, C. E., AND ALVING, A. S., Science, 124, 489 (1956).

TARLOV, A. R., BREWER, G. J., CARSON, P. E., AND ALVING. A. S., Arch. Internal Med., 109, 209 (1962).

KIRKMAN, H. N., J. Biol. Chem., 237, 2364 (1962). BALINSKY, D., AND BERNSTEIN, R. E. Biochim. Biophys.

Acta, 67, 313 (1963). CHUNG, A. E., AND LANGDON, R. G., J. Biol. Chem., 238, 2309

(1963). MARKS, P. A., SZEINBERG, A., AND BANKS, J., J. Biol. Chem.,

236, 10 (1961). TSUTSUI, E. A., AND MARKS, P. A., Biochem. Biophys. Res.

Commun., 8, 338 (1962). KIRKMAN, H. N., AND HENDRICKSON, E. M., J. Biol. Chem.,

237, 2371 (1962). CHUNG, A. E., AND LANGDON, R. G., J. Biol. Chem., 238, 2317

(1963). PORTER, I. H., BOYER, S. H., WATSON-WILLIAMS, E. J.,

ADAM, A., SZEINBERG, A., AND SINISCALCO, M., Lancet, 1, 895 (1964).

TISELIUS, A., HJERT~N, S., AND LEVIN, D., Arch. Biochem. Biophys., 66, 132 (1956).

GLOCK, G. E., AND MCLEAN, P., Biochem. J., 66,400 (1953). LOWRY, 0. H., ROSEBROUGH, N. J., FARR, A. L., AND RANDALL,

R. J., J. Biol. Chem., 193, 265 (1951). YPHANTIS, D. A., J. Biochem., 3, 297 (1964). COHN, E. J., AND EDSALL, J. T., Proteins, amino acids and

peptides, Reinhold Publishing Coporation, New York, 1943, p. 370.

KIELLEY, W. W., AND HARRINGTON, W. F., Biochim. Biophys. Acta, 41, 402 (1960).

SCHRAM, E., MOORE, S., .~ND BIGWOOD, E. J., Biochem. J., 67, 33 (1954).

SPIES, J. R., AND CHAMBERS, D. C., Anal. Chem., 22, 1209 (1950).

SPIES, J. R., Anal. Chem., 22, 1447 (1950). REES, M. W., Biochem. J., 164, 151 (1946). KIRKMAN, H. N., Nature, 184, 1291 (1959). BALINSKY, D., AND BERNSTEIN, R. E., Biochim. Biophys. Acta,

67, 313 (1963). LINEWEAVER, H., AND BURK, D., J. Am. Chem. Sot., 66, 658

(1934). WARBURG, O., AND CHRISTIAN, W., Biochem. Z., 238, 131

(1931) ; 242, 206 (1931). NOLTMANN, E. A., GUBLER, C. J., AND KUBY, S. A., J. Biol.

Chem., 236, 1225 (1961). JULIAN, G. R., WOLTE, R. G., AND REITHEL, F. J., J. Biol.

Chem., 236, 754 (1961). YUE, R. H., ROY, R. N., NOLTMANN, E. A., AND KUBY, S. A.,

Federation Proc., 25, 712 (1966). YOSHIDA, A., AND FREESE, E., Biochim. Biophys. Acta, 96,

248 (1965). YOSHIDA, A., Biochim. Biophys. Acta, 99, 66 (1965). NEVULDINE, B. H., AND LEVY, H. R., Biochem. Biophys. Res.

Commun., 21, 28 (1965). LINDER, D., AND GARTLER, S. M., Am. J. Human Genet., 17,

212 (1965).

by guest on February 4, 2020http://w

ww

.jbc.org/D

ownloaded from

Akira YoshidaENZYME

PURIFICATION AND CHARACTERIZATION OF NORMAL (B+) Glucose 6-Phosphate Dehydrogenase of Human Erythrocytes: I.

1966, 241:4966-4976.J. Biol. Chem. 

  http://www.jbc.org/content/241/21/4966Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/241/21/4966.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on February 4, 2020http://w

ww

.jbc.org/D

ownloaded from