Malonyl-CoA decarboxylase from the mammary gland of lactating rat

187

Biochimica et Biophysics @ Elsevier/North-Holland

BBA

Acta, 531 Biomedical

(1978) Press

187-196

57256

MALONYL-CoA DECARBOXYLASE LACTATING RAT PURIFICATION,

YU SAM

KIM and

PROPERTIES

P.E.

KOLATTUKUDY

Department of Agricultural State University, Pullman, (Received (Revised

February manuscript

Chemistry, Wash. 99164

28th, 1978) received June

5th,

FROM THE MAMMARY GLAND OF

AND SUBCELLULAR

LOCALIZATION

* Program (U.S.A.)

in Biochemistry

and Biophysics,

Washington

(1978)

Summary Malonyl-CoA decarboxylase (EC 4.1.1.9) was purified 500-600-fold from the mammary gland extracts by (NH4)$04 precipitation, gel filtration with Sepharose 4B, anion-exchange chromatography with QAE-Sephadex, and chromatography with NADP-Agarose. This enzyme (spec. act. 200-300 nmol/ min per mg protein) had a molecular weight of approx. 170 000. It did not cross-react with rabbit antiserum prepared against either fatty acid synthetase from the mammary gland or malonyl-CoA decarboxylase from the uropygial gland of goose. The decarboxylase showed a pH optimum near 8.5-9.0 and a K, of 0.33 mM, decarboxylated neither malonic acid nor methylmalonyl-CoA and was inhibited by thiol directed reagents but not by avidin. Sucrose density gradient centrifugation of the gland homogenate showed that the major peak of decarboxylase activity coincided with that of cytochrome oxidase. Breakage of mitochondria released >80% of the decarboxylase activity into the 105 000 X g supernatant, suggesting that malonyl-CoA decarboxylase may be located in the mitochondrial matrix.

Introduction Malonyl-CoA decarboxylase activity organisms from both animals and plants

has been observed in a variety of [l--3]. However, very little is known

* To whom all correspondence should be addressed. This is scientific paper No. 5041. project 2001. College of Agriculture University, Pullman. Wash. 99164. U.S.A.

Research

Center,

Washington

State

188

about the properties and function of malonyl-CoA decarboxylase. Recently, it was found that malonyl-CoA decarboxylase causes the production of multimethyl-branched fatty acids in the uropygial gland of domestic goose [4 1. In this gland, a tissue specific and substrate specific malonyl-CoA decarboxylase is found together with an enzyme, which carboxylates both acetyl-CoA and propionyl-CoA. Therefore acetyl-CoA and methylmalonyl-CoA would be expected to be the substrates available to the fatty acid synthetase in this tissue. As a result, multimethyl-branched fatty acids are the major products generated in this gland. This example constitutes an extreme case of regulation of malonyl-CoA levels by malonyl-CoA decarboxylase. It is possible, however, that such an enzyme can regulate the ratio of acetyl-CoA to malonyl-CoA in other tissues and thus play a regulatory role in lipid metabolism. For example, mammary glands generate large amounts of short (
Lactating Sprague Dawley strain rats were purchased from the Small Animal Resources Center of this University. Sources of all reagents used and the antisera preparations were described previously [4,9]. Enzyme assays Malonyl-CoA

decarboxylase. Malonyl-CoA decarboxylase was assayed either by measuring the 14C02 released from [3-14C]malonyl-CoA or by measuring the rate of production of acetyl-CoA spectrophotometrically. The radiochemical assay was done as described before [ 41. Labeled acetylCoA was identified as a reaction product by paper chromatography of the hydroxamic acid derivative [lo]. The amount of acetyl-CoA generated was equal to the amount of CO, released. In the spectrophotometric assay the rate of production of acetyl-CoA was measured by coupling the malonyl-CoA decarboxylase reaction with the reactions catalyzed by malic dehydrogenase and citrate synthetase. First, the malic dehydrogenase equilibrium was set up and then the shift in this equilibrium caused by the removal of oxaloacetate by condensation with acetyl-CoA generated by the decarboxylase was measured by the increase in absorbance at 340 nm [8]. Protein was determined by the method of Lowry et al. [ 111 using bovine serum albumin as standard. The spectrophotometric assay was used for the determination of specific activity, the kinetic experiments and for monitoring enzymatic activity in fractions from the column chromatographic procedures. The radiochemical assay was used during the first two steps of purification and for characterization of the enzyme when the spectrophotometric coupling assay was not suitable.

189 Fatty acid synthetase. Fatty acid synthetase activity was measured spectrophotometrically; initial rates of NADPH oxidation at 30°C were obtained by measuring the absorbance decrease at 340 nm [lo]. Marker enzymes. Cytochrome oxidase activity was measured by the spectrophotometric method [12]. Cytochrome c reductase activity was measured spectrophotometrically [ 131. The glands were gently Localization of malonyl-CoA decarboxylase. homogenized in a chilled mortar with 0.32 M sucrose solution and the homogenate was centrifuged at 600 X g for 10 min and the resulting supernatant (3 ml) was centrifuged on a 15-6070 linear sucrose gradient as described elsewhere [ 141. Isolation of malonyl-CoA decarboxylase. Rats, which were lactating for 4-5 days, were killed, and the mammary glands excised. The glands from 5 rats (approx. 100 g) were cut into small pieces and homogenized at full speed in a Sorval mixer for 30 s with approx. 40 ml 100 mM phosphate buffer (pH 7.6)/ 250 mM sucrose/l mM MgClJ0.5 mM dithioerythritol. The resulting mixture was thoroughly homogenized with a Ten Broeck glass homogenizer. The homogenate was centrifuged at 15 000 Xg for 20 min, the floating fatty material was removed and the supernatant was centrifuged at 105 000 Xg for 90 min. The supematant, which contained about 40 mg/ml protein, was diluted 4-fold to adjust protein concentration to about 10 mg/ml with 100 mM phosphate buffer (pH 7.6)/0.5 mM dithioerythritol. The protein precipitated at 30-50% saturation with (NH4)2S04 was recovered, suspended in approx. 20 ml 100 mM phosphate buffer (pH 7.6)/0.5 mM dithioerythritol and dialyzed against the same buffer overnight. After removal of any insoluble material by centrifugation this enzyme solution (7-10 ml) was subjected to gel filtration with a Sepharose 4B column (3.2 X 90 cm), using the above buffer (flow rate 0.5 ml/ min, fraction size 7 ml). Fractions containing malonyl-CoA decarboxylase activity were pooled and concentrated using a PM-30 Amicon ultrafiltration membrane. The resulting enzyme solution was again subjected to Sepharose 4B gel filtration in a manner identical to that described above. The enzyme solution, resulting from the above step was dialyzed overnight against 20 mM Tris-HCl buffer (pH 7.6)/0.5 mM dithioerythritol and was applied to a QAE-Sephadex A-25 column (2.5 X 40 cm) which had been equilibrated with the same buffer, followed by elution of proteins from the column with a linear gradient of O-O.3 M NaCl in a total volume of 500 ml 20 mM Tris-HCl buffer/O.5 mM dithioerythritol (fraction size 6 ml). Fractions containing malonyl-CoA decarboxylase activity were pooled, concentrated as above, and dialyzed overnight against 20 mM phosphate buffer (pH 7.0)/0.5 mM dithioerythritol. This enzyme solution was applied to a NADP-Agarose column (bed volume 10 ml, containing 2.4 Dmol NADP per ml gel) equilibrated with the above buffer. After washing the column with the same buffer the absorbed proteins were eluted with a linear gradient of O-O.2 M KC1 in a total volume of 200 ml of the same buffer. Fractions (1.7 ml each) containing malonyl-CoA decarboxylase activity were combined, dialyzed overnight against 100 mM phosphate buffer (pH 7.6)/0.5 mM dithioerythritol and concentrated. Electrophoresis. Polyacrylamide disc gel electrophoresis was performed in an analytical electrophoresis apparatus from Hoefer Scientific Instruments,

190

according to the procedure of Ornstein and Davis [ 15 J. The gel dimensions and procedures used for electrophoresis, staining and destaining were the same as described elsewhere [ $1. Results and Discussion Purification

of malonyl-CoA

decarboxylase

Under the homogenization conditions used over 80% of the decarboxylase activity present in the extract was recovered in the 105 000 X g supernatant from the mammary gland and this preparation usually had a specific activity of approx. 0.5 nmol/min per mg (range 0.2---1.0). This specific activity was much lower than that found in the extracts of the uropygial glands of domestic geese f4]. Major part of the decarboxylase contained in this extract was precipitated between 30 and 50% saturation with (NH4)$04 resulting in a 52% recovery of the activity. Gel filtration with Sepharose 4B showed that malonyl-CoA decarboxylase emerged from the column immediately after fatty acid synthetase (Fig, 1A). A repetition of the gel filtration showed that the decarboxylase activity was at the shoulder of what appeared to be a fairly symmetrical protein peak (Fig. 1B). The two gel filtration steps resulted in a tripling of specific activity. At pH 7.6 the decarboxylase activity was absorbed by QAESephadex and the enzyme was eluted at around 0.09 M NaCl (Fig. 2) resulting in over 16-fold purification with nearly complete recovery of the enzyme at this step. The final step of purification involved chromatography with NADPAgarose. When the enzyme solution obtained from the QAE-Sephadex step was passed through NADP-Agarose bulk of the protein was not absorbed (Fig. 3), but this protein fraction did not catalyze malonyl-CoA dec~boxylatioll. The decarboxylase could be eluted from the column at around 0.07 M KC1 (Fig. 3)

I.0

A t

03

MALONYL-CoA

MALONYL-CoA

$

DECARBOXYLASE

n rn-

DECARBOXYLASE

05L PROTEIN &_-_____*-.__c______ ------_-0 50 100 FRACTION NUMBER

150

Fig. 1. A. Sepharose 4B gel filtration of the protein fraction obtained by ammonium precipitation from the 105 000 Xg supematant of the mammary gland of lactating repetition of the gel filtration with the malonyl-CoA decarboxyiase recovered from A.

sulfate (30. -50%) rat. B, represents

191

/ / 7= zi

04-

MALONYL-CoA

DECARBOXYLA?G

, 0

A\

______

’ .L___-

-20

FRACTION

r

VI \

/

/’

2

/ -015

_____/“/,; /’ _a

-_-

/

,___;

40 NUMBER

60

Fig. 2. QAE-Sephadex chromatography of the malonyl-CoA Sepharose 4B gel filtration step shown in Fig. 1.

s 2

-0.03

decarboxylase

; -0

preparation

obtained

from the

resulting in ELfold purification with a 33% recovery. The overall purification obtained by the combination of methods described above ranged from 500600-fold with a recovery of about 6% (Table I). Polyacrylamide disc gel electrophoresis of the enzyme preparation obtained by this final step of purification showed a major band and one or two very faint bands (Fig. 4). Measurement of the enzymatic activity of the gel slices showed that the major protein band, which represented about 90% of the total protein, contained all of the decarboxylase activity. Thus it appears that the enzyme has been highly purified by the methods employed. The specific activity of the purified enzyme ranged from 200 to 300 nmol/min per mg. This specific activity is much lower than that observed with malonyl-CoA decarboxylase purified from the uropygial glands [4]. If the activity of the purified enzyme accurately reflects in vivo activity the mammary gland has the ability to decarboxylate 15--40 nmol malonyl-CoA/min per g tissue. In contrast, the uropygial glands of domestic geese decarboxylate about 50 pmol malonyl-CoA/min per g tissue [ 41. Properties of malonyl-CoA decarboxylase Molecular weight. Molecular weight of malonyl-CoA

I I

T ts c! Yz

I (-J*_

[

;yN

I I I

!

I I I I

I I I I I I

0

was esti-

\

I I ;

2 8

decarboxylase

MALONYL-CoA

/’

/

/

DECARBOXYLA,sG’

/’

/’ 20

40 FRACTION

60

80

NUMBER

Fig. .‘j. NADp-Agarose chromatography of the malon~l-CoA decarboxylase QAE-Bephadex ion-exchange chromatography step shown in Fig. 2.

PrePaMion

obtained

from the

192

TABLE

I

PURIFICATION TATING

OF

MALONYL-CoA

Fraction

105

DECARBOXYLASE

000

x B supernatant

(NH4)2SO4

THE

MAMMARY

GLANDS

OF

LAC-

Protein

Enzyme

Specific

Recovery

Purifi-

(mg)

activity

activity

(%)

cation

(nmol/min)

(nmol/min 0.4

**

100

1533

0.6

**

52

1.5

842

0.9

28

2.2

(I)

936

Sepharose

4B

(II)

330

QAE-Sephadex

both

were tracer

*

from

239.5

and

by

coupling

assay

tracer with

glands assay.

respect

(10

The

1

20

4.5

18

74.0

6

557.0

229.6

183

g mammary the

29.6

532

0.8

measured

1.8

594

18

NADP-Agarose

These

(-fold)

2964

4B

* It was obtained

per mg)

2555

7410

(30--50%)

Sepharose

**

FROM

RATS

rats). rate

to time

and

of

malonyl-CoA

protein

decarboxylation

concentration

for each

was

linear

for

fraction.

mated to be 170 000 by gel filtration with a calibrated Sepharose 6B column. This molecular weight is quite similar to the value (186 000) obtained with malonyl-CoA decarboxylase isolated from the uropygial glands of domestic geese [4] but substantially lower than that (250 000) reported for the enzyme purified from beef liver [ 161. Cofactors. Even though NADP-Agarose retained the decarboxylase, NADP, NADPH, NAD and NADH (each at 1 mM) showed no effect on the decarboxylase activity. Metal ions such as Mg2+, Mn”, and Ca*’ at 1 mM concentration and chelators such as EDTA were without effect on the activity of the

1

0

Fig.

4.

Agarose the gel.

05 MOBILITY

Polyacrylamide chromatography

disc

IO

gel electrophoresis

shown

in Fig.

3 and

of the

malonyl-CoA distribution

decarboxylase of

malonyl-CoA

obtained

from

decarboxylase

the NADPactivity

in

193

enzyme. Avidin did not inhibit the decarboxylase suggesting that biotin is not involved in the reaction. Thus a component of acetyl-CoA carboxylase is probably not involved in this reaction. Specificity. Neither maIonic acid nor methyImalonyl-CoA was a good substrate for the purified decarboxylase. For example, under identical conditions with 0.3 mM of each substrate 220, 0.09, and 0 nmol 14C0, were released from [3-14C]malonyl-CoA, [1,3-“Clmalonic acid, and methyl[3-14C]malonyl-CoA, respectively. Thus the CoA thioester is an essential feature of the substrate and the enzyme cannot accommodate the bulky methyl group (in place of H) at C-2. These results are similar to those observed with the decarboxylase from other sources [ 4,16,17]. Effect of pH. Maximal rates of decarboxylation were found at about pH 8.5-9.0. This pH optimum is close to those previously observed with the crude mitochond~al malonyl-CoA decarboxylase from rat brain [ 17 J and the purified enzyme from the uropygial gland of goose [4] but considerably higher than those (6.0-7.5) reported for the crude decarboxylase from yeast [lS] and Pseudomonas fluorescence [l] as well as that (5.5) from Mycobacterium tuberculosis (unpublished results). It appears that the pH optimum for malonyl-CoA decarboxylase from higher animals is generally higher than that from microorganisms. Effects of time, protein, and substrate concentration. Under the present spectrophotometric assay conditions linear rates were observed at least up to lo-15 min and only the initial linear rates were used for assays. The amount of acetyl-CoA generated from malonyl-CoA was directly proportional to the amount of enzyme added at least up to about 25 pg/ml. Under the tracer assay conditions 14C02 release was linear up to 225 pg prote~n/ml and 1.5 h incubation time. The rate of formation of acetyl-CoA increased with increasing concentrations of malonyl-CoA and a typical Michaelis-Menten type substrate saturation pattern was observed. From linear double reciprocal plots a K, value of 0.33 mM was calculated. With the crude enzyme preparation from the mammary gland K, values of 0.4-0.7 mM were obtained, which is quite similar to that (0.5 mM) observed previously with a crude enzyme preparation from rat brain mitochondria [17]. On the other hand much lower K, values were reported for the crude decarboxylase preparation from rat liver [ 191, partially purified enzyme from beef liver [ 161, and the purified enzyme from goose uropygial gland f 41. inhibitors, Malonyl-CoA decarboxylase was strongly inhibited by the thiol directed reagent, p-hydroxymercuribenzoate (Table Ii) as previously observed with malonyl-CoA decarboxylase from rat liver mitochondria [ 191 and goose uropygial glands [ 41. Diisopropylfluorophosphate showed only a slight inhibition of the activity of the present enzyme. The reaction product, acetyl-CoA, a substrate analog, me~ylm~onyl-CoA, and a possible product analog, propionyl-CoA, inhibited malonyl-CoA decarboxylase and the sensitivity of the enzyme to these three compounds was similar, whereas free coenzyme A (up to 1 mM) showed no inhibition. Acetyl-CoA was previously found to be an inhibitor of malonyl-CoA decarboxylase preparations from other sources 14, l&-18]. Methylm~onyl-CoA was reported to be a competitive inhibitor to the

194

TABLE EFFECT

II OF

Enzymatic

INHIBITORS

activity

ON

MALONYL-CoA

was measured

by the tracer

Inhibitors

p-Hydroxymercuribenzoate

DECARBOXYLASE method

Relative

(mM)

(W of control)

0.5

Methylmalonyl-CoA

Acetyl-CoA

Coenzyme

A

Propionyl-CoA

in Experimental

Concentration

0.25

Diisopropylfluorophosphate

described

section.

rate

6 2

0.5

84

1.0

71

0.5

83

1.0

64

0.5

15

1.0

57

0.5

111

1.0

98

0.5

76

1.0

66

-~

decarboxylase from rat brain mitochondria [ 171 and a noncompetitive inhibitor to the enzyme from the goose [4]. Preliminary experiments with the purified decarboxylase from the rat mammary gland showed that both propionyl-CoA and methylmalonyl-CoA were noncompetitive inhibitors as previously observed with beef liver enzyme [ 161. Immunological comparison. Ouchterlony double diffusion experiments showed that malonyl-CoA decarboxylase isolated from rat mammary glands did not cross-react with rabbit antiserum prepared against the decarboxylase purified from the uropygial glands of domestic geese. In addition, the antiserum did not cause any inhibition of the enzymatic activity of the decarboxylase isolated from the mammary gland. Thus the decarboxylase from the avian source is immunologically yuite different from the mammalian enzyme.

Evidence that malonyl-CoA acid syn thetase

decarboxylase

is not a proteolysis

product

of fatty

In view of the observations that fatty acid synthetase can, under certain conditions, catalyze decarboxylation of malonyl-CoA it appeared possible that the decarboxylase we isolated might be modified fatty acid synthetase. Since the size of the decarboxylase (170 000) 1s . only somewhat smaller than that of the protomer (220 000) of the fatty acid synthetase of the mammary gland 1201, there was a possibility that the former might be derived from the latter as a result of proteolysis. However under conditions which maintained fatty acid synthetase in the dimeric form [21], the yield of malonyl-CoA decarboxylase from the Sepharose 4B step was unaffected and the decarboxylase was clearly resolved from fatty acid synthetase. Furthermore, dissociation of fatty acid synthetase (isolated from the above gel filtration) by extended dialysis (48 11) against low ionic strength buffer (1 mM Tris-HCl (pH 9.2)/35 mM glycine/l mM EDTA/0.5 mM dithioerythritol) did not result in any malonyl-CoA decarboxylase activity. Furthermore, rabbit antiserum prepared against fatty

195

acid synthetase isolated from mammary glands of lactating rat neither cross reacted with nor inhibited the purified malonyl-CoA decarboxylase. Since even a small tryptic fragment (32 000) derived from fatty acid synthetase from both rat mammary gland [ 221 and goose uropygial gland [23] is known to cross react with antifatty acid synthetase, it is highly unlikely that a protein as large (170 000) as the present decarboxylase would fail to cross react, if the decarboxylase were derived from fatty acid synthetase. Thus, it appears clear that the malonyl-CoA decarboxylase is not derived from fatty acid synthetase. The subcellular location of the decarboxylase described in the succeeding section also strongly supports this conclusion. Subcellular

localization

of malonyl-CoA

decarboxylase

In order to determine the subcellular location of the enzyme, gland homogenates prepared by mild techniques were fractionated using a sucrose density gradient centrifugation technique. The main peak of malonyl-CoA decarboxylase activity was coincident with cytochrome oxidase activity at a density of about 1.18 g/cm3 strongly suggesting that the enzyme was located mainly in the mitochondria. Malonyl-CoA decarboxylase was previously reported to be located in the mitochondria in liver, kidney, heart and brain of rat [ 1,19,24]. Since breakage of mitochondria by suspension in dilute buffer (10 mM phosphate) released substantial amounts (approx. 85%) of the decarboxylase it appears that the extracts used in the present enzyme purification contained a substantial portion of the mitochondrial enzyme. In fact, the high speed supernatant used in the purification of the enzyme contained 82% of the decarboxylase activity of the homogenate. Yet during all of the steps used in the purification procedure multiple forms of the decarboxylase could not be detected. Therefore it appears that the mammary gland contains only the mitochondrial decarboxylase. The function of malonyl-CoA decarboxylase in the mammalian tissues is obscure. If the in vivo activity of the enzyme is accurately shown by the in vitro measurements used in the present study, it would appear that decarboxylase activity might be too low to bring about a large enough increase in the [acetylCoA] : [malonyl-CoA] ratio to significantly alter the chain length of the fatty acids produced by the synthetase. The mitochondrial location of the enzyme and the reported inability of the carnitine system to transport malonyl-CoA into the mitochondria [16] also agree with this conclusion. A possible function for the decarboxylase in mammalian tissues might be to decarboxylate malonyl-CoA generated within tine mitochondria by propionyl-CoA carboxylase which is known to carboxylate acetyl-CoA at low rates [ 251. Acknowledgments We thank Angelika Boos and Linda Rogers for technical assistance. This work was supported in part by grant GM-18278 from the U.S. Public Health Service. References 1

Hayaishi.

2

Wolfe,

J.B.

0.

(1955) and

J. Biol.

Rittenburg.

Chem. S.C.

215.

(1954)

125-136 J. Biol.

Chem.

209,

885-892

196

3

Hatch,

4

Buckner,

M.D.

5

Carey,

6

Hanson,

H..

7

Sumper,

M.,

8

Kim.

and

J.S.. E.M.

Y.S.

Stumpf,

P.K.

Kolattukudy,

and

DiIs.

Carey.

R. (1970)

E.M.

and

Oesterhelt.

and

(1962)

P.E.

R.

P.E.

37,

A.J.

(1970)

(1978)

Arch.

Arch.

Acta.

Biochim. C. and

121-126

(1976)

Biophys.

Riepertinger,

Kolattukudy,

Physiol.

Poulose,

Biochim.

DiIs,

D..

Plant

and

210,

Biophys.

Lynen.

Biochem.

Buckner.

J.S.

and

Kolattukudy,

P.E.

(1976)

Biochemistry

15.

Buckner,

J.S’

and

Kolattukudy,

P.E.

(1975)

Biochemistry

14,

II

Lowry,

12

Wharton,

13

Phillips,

A.H.

and

Langdon,

14

Soliday,

C.L.

and

Kolattukudy,

15

Omstein,

16

Scholte.

17

Koeppen,

18

Lynen,

19

Landriscina,

D.C.

and

L. and H.R. F..

Tzagoloff,

David,

(1973)

A.H.,

N.J.,

C..

A.

G.F..

Gnoni.

J. Biol.

(1978)

Ann.

Goldmann,

G.V.

and

Chem.

Acta

309. A.A.

M. and

E. (1971)

S. and

Abraham.

S. (1970)

J. Biol.

Chem.

245,

S. and

Abraham,

S. (1971)

J. Biol.

Chem.

246.6428-6435

E..

C.J.,

Libertini,

24

Scholte,

H.R.

25

Lane,

D. and

L. and

Kolattukudy, (1969) Halenz,

Smith. P.E.

Biochim. D.R.

S. (1976)

and

Biophys.

(1962)

Biochem.

Rogers.

L. (1978)

Acta

Methods

Chem.

193.

265--275

188.

338-347

404-408

178.

Enzymol.

13,

Biochem.

Eur.

Smith, Agradi.

J. Biol.

245-250

Biochemistry

Smith, Bedord,

377-387

2652-2660

I. (1962)

20 22

10,

457-465

21 23

400-410

J. Biochem.

1768-1773

Biophys..

121,

(1974)

Kessel.

QuagIiarieIlo,

539-551

in the press

(1951)

237,

Sci.

177,

1948-1957

10,

Biochem.

Acad.

Ammoumi.

R.J.

Enzumol.

Arch.

N.Y.

Biophys. and

Randall.

(1962) P.E.

E.J.

and

Methods

R.G.

Biochim.

A.L.

(1967)

B. (1964)

Mitzen,

Domagk,

Farr,

Biophys.

210,

Eur.

Biophys..

9

Rosebrough.

Acta

F. (1969)

10

O.H.,

Biochem.

388-399

Z.

J. Biochem.

3589-3595 335. 19,

519-539 573-580

3209-3217 Biophys. Arch.

Res.

Biochem.

137-144 5. 576-581

Commun. Biophys.

68, 186.

894-900 139-151