Purification and characterization of formate oxidase from a

FEMS Microbiology Letters 214 (2002) 137^142
www.fems-microbiology.org
Puri¢cation and characterization of formate oxidase from
a formaldehyde-resistant fungus
Tetsuya Kondo
b
a;
, Yutaka Morikawa b , Naohiro Hayashi a , Noriyuki Kitamoto
a
a
Food Research Institute, Aichi Prefectural Government, 2-1-1 Shinpukuji-cho, Nishi-ku, Nagoya 451-0083, Japan
Aichi Industrial Technology Institute, Aichi Prefectural Government, Nishishinwari, Hitotsugi-cho, Kariya 448-0003, Japan
Received 25 May 2002 ; received in revised form 10 July 2002; accepted 13 July 2002
First published online 1 August 2002
Abstract
A formate oxidase activity was found in the crude extract of a formaldehyde-resistant fungus isolated from soil. The fungus was
classified and designated as Aspergillus nomius IRI013, which could grow on a medium containing up to 0.45% formaldehyde and
consumed formaldehyde completely. The specific activity of formate oxidase in the extract of the fungus grown on formaldehyde was
found to be considerably higher than that in the extracts of the fungus grown on formate and methanol. Formate oxidase from the fungus
grown on formaldehyde was purified to homogeneity. The enzyme had a relative molecular mass of 100 000 and was composed of two
apparently identical subunits that had a relative molecular mass of 59 000. The enzyme showed the highest activity using formate as
substrate. Hydrogen peroxide was formed during the oxidation of formate. The Michaelis constant for formate was 15.9 mM; highest
enzyme activity was found at pH 4.5^5.0. The enzyme activity was strongly inhibited by NaN3 , p-chloromercuribenzoate and
HgCl2 . 4 2002 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
Keywords : Formate oxidase; Formaldehyde; Formate; Aspergillus nomius
1. Introduction
While formate occurs ubiquitously in nature in microbial degradation of organic compounds, a diversity of microorganisms consume formate in a variety of metabolic
pathways. All of these microorganisms have enzymes capable of oxidizing formate. The enzymes have been classi¢ed into two types of formate dehydrogenases (FDH ;
for a review see [1,2]). One is a type of NAD-dependent
FDH. This enzyme has been highly puri¢ed from methylotrophic yeasts such as Candida boidinii [3,4], Hansenula
polymorpha [5] and Pichia pastoris [6]. In methylotrophic
yeasts, formate oxidation catalyzed by NAD-dependent
FDH is the last step of methanol oxidation [5,7]. This
dehydrogenase has also been found in aerobic bacteria
[8,9] and higher plants [10,11]. The majority of NAD-dependent FDHs display relatively high Michaelis constant
(Km ) values for formate and do not have any metals or
prosthetic groups. The second type is dye-linked FDH,
* Corresponding author. Tel. : +81 (52) 521 9316;
Fax : +81 (52) 532 5791.
E-mail address : [email protected] (T. Kondo).
which has been found in anaerobic bacteria [12,13]. This
enzyme is distinct from NAD-dependent FDHs. Dyelinked FDHs are extremely sensitive to oxygen and contain a diversity of iron^sulfur metal-containing units and
low Km values for formate. Thus FDHs have been found
in microorganisms from di¡erent taxons and higher plants
and well characterized both structurally and biochemically. On the other hand, little is known about the isolation of enzymes that catalyze formate oxidation followed
by an electron reduction of molecular oxygen to hydrogen
peroxide. NAD-dependent FDH from Pseudomonas oxalaticus [8] is the only known example; it is distinct from
the other NAD-dependent FDHs in higher molecular
mass, a complex quaternary structure containing iron, sulfur and FMN and a low Km value for formate. This enzyme uses oxygen, NADþ , FMN, FAD, ribo£avin and
various dyes as electron acceptors.
In the course of screening formaldehyde-degrading microorganisms, we isolated a formaldehyde-resistant fungus
from soil. This fungus was found to exhibit the formateoxidizing activity followed by formation of hydrogen peroxide in the crude extract. We tentatively named the enzyme ‘formate oxidase’ (FOX). In this paper, we describe
the isolation and some properties of the FOX.
0378-1097 / 02 / $22.00 4 2002 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
PII : S 0 3 7 8 - 1 0 9 7 ( 0 2 ) 0 0 8 5 6 - X
FEMSLE 10600 19-8-02
138
T. Kondo et al. / FEMS Microbiology Letters 214 (2002) 137^142
the enzyme solution. The reaction was done at 30‡C, and
the increase in absorbance at 505 nm was monitored.
2. Materials and methods
2.1. Chemicals
2.5. Measurement of NAD-dependent FDH activity
Paraformaldehyde, formic acid and sodium formate were
purchased from Wako (Osaka, Japan). Horseradish peroxidase (grade III) was purchased from Toyobo (Osaka,
Japan). DEAE^Sepharose CL-6B, Q^Sepharose FF, Sephacryl S-200 and Phenyl^Sepharose HP were from Pharmacia Biotech (Uppsala, Sweden). All other reagents were
of the highest grade available from commercial sources.
2.2. Media
The basal medium contained (all w/v) 1.0% glucose,
0.2% NaNO3 , 0.1% K2 HPO4 , 0.05% MgSO4 W7H2 O,
0.05% KCl, 0.001% FeSO4 W7H2 O and 0.01% yeast extract.
The pH of the medium was adjusted to 6.0. The formaldehyde medium used for production of FOX was the
basal medium supplemented with 0.1% (w/v) formaldehyde.
2.3. Fungal isolation, identi¢cation and cultivation
One hundred and seventy samples of soil and wastewater from Aichi Prefecture were spread on plates of the
formaldehyde medium containing 1.5% (w/v) agar and incubated at 30‡C for 5 days. An isolate, strain IRI013,
formed a white colony, the reverse color being orange
and the colony surface being £occose on potato^dextrose
agar. Under microscopic observation, it exhibited typical
characteristics of the Aspergillus genus [14]. In addition,
on the basis of 18S rDNA analysis [15], the strain was
taxonomically identi¢ed as Aspergillus nomius. A. nomius
IRI013 was maintained on a slant of the formaldehyde
medium. A loopful of the mycelium from a slant culture
of the fungus was transferred into the formaldehyde medium. The cultivation was carried out on a rotary shaker
at 30‡C for 7 days.
2.4. Measurement of FOX activity
A standard assay of FOX activity was performed by
determining the rate of oxygen consumption in a Clark
oxygen electrode (Central Kagaku, Japan) at 30‡C in the
standard assay mixture (1.5 ml) containing McIlvaine
bu¡er (pH 4.5), 30 Wmol of sodium formate/formic acid
(pH 4.5) and an appropriate amount of enzyme. One unit
of the enzyme activity was de¢ned as the amount of enzyme consuming 1 Wmol of O2 per minute. The enzyme
assay based on the measurement of hydrogen peroxide
formed in the enzyme reaction [16] was also used to detect
the enzyme during its puri¢cation. The reaction mixture
(3.0 ml) contained 250 Wmol of acetate bu¡er (pH 5.0),
60 Wmol of sodium formate, 1.5 Wmol of 4-aminoantipyrine, 6 Wmol of phenol, 30 U of horseradish peroxidase and
NAD-dependent FDH activity was assayed by measuring the increase in absorbance at 340 nm due to the formation of NADH (O = 6220 M31 cm31 ). The reaction mixture (3.0 ml) contained 250 Wmol of sodium phosphate
bu¡er (pH 8.0), 5 Wmol of NADþ , 300 Wmol of sodium
formate and the enzyme solution. The reaction was done
at 30‡C. One unit of the enzyme activity was de¢ned as the
amount that catalyzed the formation of 1 Wmol of NADH
per minute.
2.6. Puri¢cation of FOX
All operations were done below 4‡C, unless stated otherwise. Mycelia (165 g wet weight) obtained from 10.8 l of
culture broth were crushed with sea sand on a mortar and
suspended in 165 ml of 10 mM sodium phosphate bu¡er
(pH 6.0, bu¡er A). The suspension was centrifuged at
10 000Ug for 15 min. The supernatant obtained by centrifugation was retained as a crude extract. The crude extract was fractionated with ammonium sulfate, and then
proteins precipitating in the range 40^90% saturation were
collected by centrifugation. The precipitate was dissolved
in bu¡er A and dialyzed thoroughly against bu¡er A. The
enzyme solution was put on a DEAE^Sepharose CL-6B
column (2.5U23 cm) equilibrated with bu¡er A. After
washing the column with the same bu¡er, the enzyme
was eluted with a linear 0^0.5 M NaCl gradient in bu¡er
A. The active fractions were pooled (60 ml) and dialyzed
thoroughly against bu¡er A. The enzyme solution was put
on a Q^Sepharose FF column (2.5U19 cm) equilibrated
with bu¡er A. After washing the column with the same
bu¡er, the enzyme was eluted with a linear 0^0.5 M NaCl
gradient in bu¡er A. The active fractions were pooled
(32 ml) and concentrated by ultra¢ltration using a PL-30
membrane (Amicon). The concentrated enzyme solution
was put on a Sephacryl S-200 column (2.5U90 cm) equilibrated with 10 mM acetate bu¡er (pH 4.6, bu¡er B) and
eluted with the same bu¡er. The active fractions were
pooled (33.8 ml) and made 30% saturation with ammonium sulfate. The enzyme was put on a Phenyl^Sepharose
HP column (1.5U16 cm) equilibrated with bu¡er B containing ammonium sulfate at 30% saturation. After washing the column with the same bu¡er, the enzyme was
eluted with a linear 30^0% ammonium sulfate gradient
in bu¡er B. The active fractions were pooled (27 ml)
and concentrated by ultra¢ltration to 2.1 ml. The enzyme
was stored in a freezer at 360‡C.
2.7. Gel electrophoresis
Polyacrylamide gel electrophoresis of native protein (na-
FEMSLE 10600 19-8-02
T. Kondo et al. / FEMS Microbiology Letters 214 (2002) 137^142
139
Table 1
Enzyme activities of A. nomius IRI013 grown on several carbon sources in the presence of glucose
Basal mediuma plus
None
0.1% Formaldehyde
0.1% Formic acid
0.1% Methanol
3% Methanol
Speci¢c activityb
Total proteinc (mg)
FOX (U mg31 )
NAD-dependent FDH (U mg31 )
0.039
2.73
0.32
0.06
0.18
0.002
0.025
0.010
0.002
0
14.3
23.6
18.6
17.5
14.2
a
Containing 1% glucose.
The assays were performed in the presence of 100 mM formate.
c
The protein in mycelium obtained from 100 ml of culture broth.
b
tive-PAGE) in the absence of sodium dodecyl sulfate
(SDS) was done with a slab of 7.5% polyacrylamide gel
and 0.25 M Tris^1.92 M glycine bu¡er (pH 8.3) as a mobile phase. SDS^PAGE was done with a slab gel (10%
polyacrylamide gel) by the method of Laemmli [17]. After
electrophoresis, the proteins were stained with Coomassie
brilliant blue R-250.
2.8. Measurement of molecular mass
The molecular mass of the native enzyme was estimated
by gel ¢ltration on a Superdex 200 HR (10/30) column
(Pharmacia Biotech) equilibrated with 50 mM sodium
phosphate bu¡er containing 0.15 M NaCl (pH 7.0). As
molecular mass standards, rabbit aldolase (Mr = 158 000),
bovine serum albumin (Mr = 67 000), egg albumin
(Mr = 45 000), chymotrypsinogen A (Mr = 25 000), myoglobin (Mr = 17 800) and cytochrome c (Mr = 12 400) were
used.
2.9. Analytical methods
Protein concentration was measured by the method of
Lowry et al. [18] with bovine serum albumin as a standard. Absorption spectrophotometry was performed with
a Jasco V-550 spectrophotometer.
3. Results
3.1. Formation of FOX
A. nomius IRI013 was isolated from soil as a formalde-
hyde-resistant fungus. This fungus grew on the basal medium with 0.45% formaldehyde and consumed formaldehyde completely. In addition, FOX activity was detected
in the crude extract. The fungus did not grow on the formaldehyde medium without glucose. To investigate the effect of a carbon source in the presence of glucose on FOX
activity, A. nomius IRI013 was cultivated for 5 days in the
basal medium with formaldehyde, formic acid or methanol. On the basal medium, A. nomius grew well, but FOX
activity was considerably low in the crude extract. As
shown in Table 1, the addition of formaldehyde, formic
acid or methanol enhanced FOX activity. The maximal
formation of the enzyme activity was observed when the
fungus grew on formaldehyde. The speci¢c activity in the
fungus grown on formaldehyde was 70 times higher than
that in the fungus grown on the basal medium. NAD-dependent FDH activity was also increased by the addition
of formaldehyde or formic acid to the basal medium,
although the dehydrogenase activity was signi¢cantly lower than FOX activity. Alcohol oxidase could not be detected in the extract of the fungus grown on any of the
carbon sources tested.
3.2. Puri¢cation of FOX
The enzyme was puri¢ed 39-fold with a yield of 13%
from the crude extract of A. nomius IRI013 grown on
formaldehyde (Table 2). The speci¢c activity of the puri¢ed enzyme was 44 U mg31 at 30‡C and E0:1%
1cm at 280 nm
was determined as 1.4. The puri¢ed enzyme showed a
single protein band on native-PAGE and SDS^PAGE
(Fig. 1). When the enzyme was stored at 4‡C in 10 mM
acetate bu¡er (pH 4.6), 33% of the initial activity remained
Table 2
Puri¢cation of the FOX from A. nomius IRI013
Puri¢cation step
Protein (mg)
Total activity (U)
Speci¢c activity (U mg31 )
Yield (%)
Puri¢cation (fold)
Crude extract
Ammonium sulfate
DEAE^Sepharose CL-6B
Q^Sepharose FF
Sephacryl S-200
Phenyl^Sepharose HP
285
181
13.2
5.73
2.84
0.96
321
310
179
112
99.4
42.2
1.13
1.71
13.6
19.5
35.0
44.0
100
96
56
35
31
13
1
2
12
17
31
39
FEMSLE 10600 19-8-02
140
T. Kondo et al. / FEMS Microbiology Letters 214 (2002) 137^142
Fig. 2. Absorption spectrum of puri¢ed FOX. The concentration of the
enzyme protein was 0.53 mg ml31 in 10 mM acetate bu¡er (pH 4.6).
The same bu¡er was used as a blank. The absorption spectrum was
measured at room temperature.
100 000 by gel ¢ltration chromatography on Superdex
200 HR. These results show that the enzyme was composed of two identical subunits.
3.4. Absorption spectrum
Fig. 1. Native-PAGE and SDS^PAGE of the puri¢ed FOX. A: NativePAGE of FOX, B: SDS^PAGE of FOX: lanes 1, 3, puri¢ed enzyme;
lane 2, molecular mass markers containing phosphorylase b (97 400), bovine serum albumin (66 200), egg white ovalbumin (45 000) and bovine
carbonic anhydrase (31 000). In both cases, the protein band was stained
with Coomassie brilliant blue R-250.
after 10 days. The enzyme activity was maintained at
360‡C for several months. However, when the enzyme
was thawed, the activity decreased gradually and was
lost completely in 1 day at 4‡C.
3.3. Molecular mass and subunit structure
The relative molecular mass of the denatured enzyme
was found to be 59 000 on SDS^PAGE, while the molecular mass of the native enzyme was estimated to be
The puri¢ed enzyme had a yellow color. An absorption
spectrum was taken for a solution of the native enzyme.
The enzyme exhibited absorption maxima at 336, 372, 434,
478, 582, and 684 (Fig. 2).
3.5. E¡ects of pH and temperature on enzyme activity and
stability
The enzyme showed maximum activity at pH between
4.5 and 5.0 (Fig. 3A). The enzyme was most stable between pH 5.0 and 6.0 in McIlvaine bu¡er when it was
incubated at various pHs at 30‡C for 60 min (Fig. 3B).
The enzyme was stable below 30‡C in 60 min incubation
at pH 4.5. The activity dropped quickly during incubation
at higher temperatures. After 30 min incubation at 35‡C,
47% of the activity was left, and the enzyme was almost
inactivated at 45‡C.
Fig. 3. E¡ects of pH on activity and stability of FOX. A: The enzyme activity was measured under the standard assay conditions in McIlvaine bu¡er
at various pHs. B: The enzyme was incubated at 30‡C for 60 min in the following bu¡er: McIlvaine bu¡er (F), sodium phosphate bu¡er (b) and boric
acid^NaOH bu¡er (R). The remaining activity was measured under the standard assay conditions.
FEMSLE 10600 19-8-02
T. Kondo et al. / FEMS Microbiology Letters 214 (2002) 137^142
141
Table 3
E¡ects of various compounds on FOX
Compound
Concentration (mM)
Remaining activity (%)
p-chloromercuribenzoate
HgCl2
Iodoacetamide
Monoiodoacetate
Semicarbazide
Hydroxylamine
Hydrazine
EDTA
NaN3
o-phenanthroline
8-Hydroxyquinoline
CuSO4
0.1
1
1
1
1
1
1
1
1
1
1
1
1
0
99
90
90
76
70
93
1
59
38
81
The enzyme was preincubated with a test compound in McIlvaine bu¡er (pH 4.5) at 30‡C for 10 min, and then its activity was measured under the
standard conditions.
3.6. Substrate speci¢city and kinetic parameter
Formate was the best substrate for FOX. Glyoxylic acid
was also oxidized at 18% of the rate for formate. The Km
for formate was estimated to be 15.9 mM. The enzyme
activity was considerably low (1^2% of formate oxidation)
for acetate, malate, methanol, formaldehyde and glyoxal.
The enzyme was inert toward ethanol and acetaldehyde.
3.7. Electron acceptors
The availability of electron acceptors for FOX activity
was examined. To test the activity with various dyes, the
change of absorbance was monitored at 340 nm for
NADþ , 420 nm for ferricyanide and 600 nm for 2,6-dichlorophenolindophenol with or without phenazine methosulfate. Only oxygen served as an electron acceptor
for the enzyme reaction.
3.8. E¡ects of various compounds on activity of FOX
The e¡ects of various compounds on the enzyme activity were examined. As shown in Table 3, p-chloromercuribenzoate, HgCl2 and NaN3 completely inhibited the enzyme activity. Chelating reagents such as o-phenanthroline
and 8-hydroxyquinoline showed some e¡ects on inhibition
of the activity, but EDTA did not inhibit the enzyme.
Carbonyl reagents such as semicarbazide, hydroxylamine
and hydrazine did not a¡ect the enzyme activity.
4. Discussion
In methanol-utilizing microorganisms, methanol is successively oxidized to formaldehyde, formate and carbon
dioxide. In methylotrophic yeasts, NAD-dependent FDH
catalyzes the last step of methanol oxidation and involves
in the energy-generating system [2,3]. Table 1 showed that
a formaldehyde-resistant fungus, A. nomius IRI013, con-
tained both FOX and NAD-dependent FDH. FOX was
inducibly formed in cells grown on methanol, formaldehyde and formic acid, while NAD-dependent FDH was
not induced by methanol in contrast with that of methylotrophic yeasts [3]. Formaldehyde was most e¡ective for
the formation of FOX. The speci¢c activity of FOX
was signi¢cantly higher than that of NAD-dependent
FDH. These results suggests that FOX rather than
NAD-dependent FDH may contribute to the detoxi¢cation of formaldehyde, although it appears that formate
oxidation by FOX is not coupled to an energy-generating
system.
In the present investigation, FOX was puri¢ed and
characterized from A. nomius IRI013. This is the ¢rst report of the isolation of FOX from a fungus. The properties of FOX were similar to those of NAD-dependent
FDHs from methylotrophs. The FOX was composed of
two apparently identical subunits. Formate was the best
substrate for the enzyme. The Km value was 15.9 mM,
comparable to those obtained for NAD-dependent
FDHs from methylotrophic yeasts [3^6] and aerobic bacteria [9]. HgCl2 , p-chloromercuribenzoate and NaN3 signi¢cantly inhibited the enzyme activity (Table 3), suggesting some importance of a sulfhydryl group for the
expression of the enzyme activity.
On the other hand, FOX from A. nomius di¡ers from
other formate-oxidizing enzymes in several aspects. FOX
had an optimum pH at acidic pH, while FDHs have been
reported to have optimum pH within the neutral or alkali
region [2^6,9,13]. FOX was unstable in the alkali region in
contrast with FDHs. Finally, the enzyme required no additive cofactors except oxygen for its activity.
Acknowledgements
This work was performed through Special Coordination
Funds for Promoting Science and Technology (Leading
Research Utilizing Potential of Regional Science and
FEMSLE 10600 19-8-02
142
T. Kondo et al. / FEMS Microbiology Letters 214 (2002) 137^142
Technology) of the Science and Technology Agency of the
Japanese Government.
[10]
References
[11]
[1] Ferry, J.G. (1990) Formate dehydrogenase. FEMS Microbiol. Rev.
87, 377^382.
[2] Popov, V.O. and Lamzin, V.S. (1994) NADþ -dependent formate dehydrogenase. Biochem. J. 301, 625^643.
[3] Kato, N., Kano, M., Tani, Y. and Ogata, K. (1974) Puri¢cation and
characterization of formate dehydrogenase in a methanol-utilizing
yeast, Kloeckera sp. No. 2201. Agric. Biol. Chem. 38, 111^116.
[4] Schu«tte, H., Flossdorf, J., Sahm, H. and Kula, M.-R. (1976) Puri¢cation and properties of formaldehyde dehydrogenase and formate
dehydrogenase from Candida boidinii. Eur. J. Biochem. 62, 151^160.
[5] van Dijken, J.P., Oostra-Demkes, G.J., Otto, R. and Harder, W.
(1976) S-formylgluthathione : the substrate for formate dehydrogenase in methanol-utilizing yeasts. Arch. Microbiol. 111, 77^83.
[6] Hou, C.T., Patel, R.N., Laskin, A.I. and Barnabe, N. (1982) NADlinked formate dehydrogenase from methanol-grown Pichia pastoris
NRRL-Y-7556. Arch. Biochem. Biophys. 216, 296^305.
[7] Fujii, T. and Tonomura, K. (1972) Oxidation of methanol, formaldehyde and formate by a Candida species. Agric. Biol. Chem. 36,
2297^2306.
[8] Ho«pner, T., Ruschig, U., Mu«ller, U. and Willnow, P. (1982) Formate
dehydrogenase from Pseudomonas oxalaticus. Methods Enzymol. 89,
531^537.
[9] Egorov, A.M., Avilova, T.A., Dikov, M.M., Popov, V.M., Rodio-
[12]
[13]
[14]
[15]
[16]
[17]
[18]
nov, Y.V. and Berezin, I.V. (1979) NAD-dependent formate dehydrogenase from methylotrophic bacterium, strain 1. Puri¢cation and
characterization. Eur. J. Biochem. 99, 569^576.
Ohyama, T. and Yamazaki, I. (1974) Puri¢cation and some properties of formate dehydrogenase. J. Biochem. (Tokyo) 75, 1257^1263.
Uotila, L. and Koivusalo, M. (1979) Puri¢cation of formaldehyde
and formate dehydrogenases from pea seeds by a⁄nity chromatography and S-formylglutathione as the intermediate of formaldehyde
metabolism. Arch. Biochem. Biophys. 196, 33^45.
Schauer, N.L. and Ferry, J.G. (1986) Composition of the coenzyme
F420-dependent formate dehydrogenase from Methanobacterium formicium. J. Bacteriol. 165, 405^411.
Sebban, C., Blanchard, L., Bruschi, M. and Guerlesquin, F. (1995)
Puri¢cation and characterization of the formate dehydrogenase from
Desulfovibrio vulgaris Hildenborough. FEMS Microbiol. Lett. 133,
143^149.
Udagawa, S. (1978) Aspergillus genus In: Kinrui Zukan [in Japanese]
(Udagawa, S., Tsubaki, K., Horie, Y., Miura, K., Minoura, K., Yamazaki, M., Yokoyama, T. and Watanabe, S., Eds.), Vol. 2, 3rd edn.,
pp. 1006^1008. Kodansha, Tokyo.
Borneman, J. and Hartin, R.J. (2000) PCR primer that a⁄nity fungal
rRNA genes from environmental samples. Appl. Environ. Microbiol.
66, 4356^4360.
Yamada, H., Isobe, K. and Tani, Y. (1980) Oxidation of polyamines
by fungal enzymes. Agric. Biol. Chem. 44, 2469^2476.
Laemmli, U.K. (1970) Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature 227, 680^685.
Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951)
Protein measurement with the Folin phenol reagent. J. Biol. Chem.
193, 265^275.
FEMSLE 10600 19-8-02