JOURNAL OF BACrERIOLOGY, Nov., 1966 Vol. 92, No.

5
Copyright © 1966 American Society for Microbiology Printed in U.S.A.

Diaphorases from Aerobacter aerogenes'

CARL BERNOFSKY AND RUSSELL C. MILLS
Department of Biochemistry, University of Kansas Medical Center, Kansas City, Kansas
Received for publication 20 June 1966

ABSTRACT
BERNOFSKY, CARL (The University of Kansas, Kansas City), AND RUSSELL C. MILLS.
Diaphorases from Aerobacter aerogenes. J. Bacteriol. 92:1404-1414. 1966.-Five enzymes
which catalyze the reduction of 2, 6-dichlorophenol-indophenol by reduced nicotinamide

Downloaded from http://jb.asm.org/ on November 22, 2017 by guest

adenine dinucleotide (NADH2) have been separated from sonic extracts of Aerobacter
aerogenes B199 by diethylaminoethyl (DEAE) cellulose chromatography. Three major
chromatographic fractions (enzymes I, 11, and III) account for most of the activity in the
extract. Of the two minor fractions, one is associated with cytochrome b,. The other
is extremely labile, and was not studied further. The chromatographed diaphorases ap-
pear to have a specific requirement for flavin mononucleotide. They are also readily inac-
tivated by dilution; however, this can be prevented by a combination of phosphate buffer,
bovine serum albumin, and flavin mononucleotide. The different enzymes are clearly dis-
tinguishable by their activities with NADH2 and reduced nicotinamide adenine dinucleo-
tide phosphate (NADPH2) in the presence of various electron acceptors (2, 6-dichloro-
phenol-indophenol, ferricyanide, menadione, and cytochrome c), and by their responses
to inhibitors (amobarbital, antimycin A, Atabrine, p-chloromercuribenzenesulfonate,
dicumarol, and 2, 4-dinitrophenol). With 2, 6-dichlorophenol-indophenol as acceptor, en-
zymes I, II, and III have comparable activities with either NADH2 or NADPH2. With
menadione and ferricyanide as acceptors, enzymes II and III exhibit very high, NADH2-
specific activities. When cytochrome c is the acceptor, however, enzyme III shows greater
activity with NADPH2 as the electron donor. Ferricyanide is the most active acceptor for
the cytochrome b,-containing fraction. Coenzyme Q6 does not appear to serve as an ac-
ceptor. All the diaphorases, with the exception of that in the cytochrome b1-containing
fraction, are inhibited by p-chloromercuribenzenesulfonate. Amobarbital is relatively in-
effective and inhibits only the indophenol reductase activity of enzyme I. The menadione
reductase activity of enzymes I, and II, and the diaphorases in the cytochrome b&-contain-
ing fraction are strongly inhibited by antimycin A, 2,4-dinitrophenol, dicumarol, and
Atabrine. However, the menadione reductase activity of enzyme III is affected only by the
last three of these inhibitors. The diaphorases in sonic-treated extracts do not appear to be
associated with a particulate fraction.

A diaphorase may be defined as any flavopro- responsible for coupling NADH2 oxidation to the
tein which can catalyze the oxidation-reduction electron transport system. When isolated from its
reaction between a reduced pyridine nucleotide mitochondrial association, this dehydrogenase is
and a suitable acceptor. Such enzymes have a wide active with ferricyanide, but not with cytochrome
distribution in nature, and characteristically ex- c, as an electron acceptor (14). It is readily con-
hibit a broad specificity for electron acceptors. verted by mild treatment into a diaphorase that
Well-known examples of flavoproteins which can function with cytochrome c as an acceptor
were first recognized as diaphorases before their (26).
physiological functions were understood are lipoyl Lipoyl dehydrogenase is a component of the
dehydrogenase and the respiratory chain-linked pyruvate and cx-ketoglutarate dehydrogenase
reduced nicotinamide adenine dinucleotide complexes in which it serves to catalyze the oxida-
(NADH2) dehydrogenase. The latter enzyme is tion-reduction reaction between NADH2 and
1 This paper was taken from a thesis submitted by lipoic acid. This enzyme, first isolated and
Carl Bernofsky to the University of Kansas Graduate characterized as a diaphorase by Straub (21), was
School in partial fulfillment of the requirement for the established in its present role through the work of
Ph.D. degree. Massey (13) and Searls and Sanadi (19).
1404
VOL. 92, 1966 DIAPHORASES FROM AEROBACTER AEROGENES 1405
It is generally assumed that diaphorases are M potassium phosphate (pH 7.4-8.0) and were centri-
involved only in pathways concerned with the fuged at 15,000 X g for 15 min. The concentration and
regeneration of pyridine nucleotides from their pH were dependent on the intended subsequent treat-
reduced forms. In the oxidation of a-keto acids, ment. All operations after the harvesting procedure
however, lipoyl dehydrogenase normally func- were conducted at 0 to 5 C. The cells were resuspended
in 75 ml of buffer and were centrifuged at 30,000 X g
tions in the direction of nicotinamide adenine di- for 15 min. The usual yield of packed wet cells from 2
nucleotide (NAD) reduction by dihydrolipoic liters of culture medium was 30 g, 5% of which repre-
acid (17), thus showing that a diaphorase can be sented protein as measured by the method of Lowry
involved in a reversible, pyridine nucleotide- et al. (10).
linked reaction in which the isolated enzyme Extracts. The packed cells were suspended in an
shows little specificity for the non-nucleotide re- equal volume of buffer and subjected to sonic treat-

Downloaded from http://jb.asm.org/ on November 22, 2017 by guest

actant. ment for 10 min in a tap-water cooled Raytheon 9-kc
The lack of specificity for the non-nucleotide magnetostriction oscillator, at a plate voltage of 100.
component is a property which enables diaph- After initial centrifugation at 30,000 X g for 30 min,
the sonic-treated material was centrifuged at
orases to be readily detected. However, it is also 105,000 X g for 2 hr to obtain a clear yellow super-
a property which obscures their physiological natant fraction containing essentially all of the diaph-
roles. Although such nonspecific enzymes may orase activity. In preparation for chromatography, the
seem to be of questionable value to the cell, it is supernatant fractions were dialyzed overnight with
apparent from a consideration of lipoyl dehydro- 0.01 to 0.02 M potassium phosphate (pH 7.4-8.0), de-
genase that a flavoprotein which has little intrinsic pending upon the initial conditions of the chroma-
specificity of its own can have specificity conferred tography.
upon it by virtue of its association with a larger Assays. The standard assay for measuring indo-
complex in which its contact with substrates is phenol reductase activity in chromatographic fractions
contained: 300 ,moles of potassium phosphate (pH
governed by the specificities of other enzymes. 7.4), 6 mg of crystalline bovine serum albumin (BSA,
In the present study, extracts of Aerobacter Pentex Inc., Kankakee, Ill.), 0.4 ,umole of flavin mono-
aerogenes were examined for their content of nucleotide (FMN, Sigma Chemical Co., St. Louis,
diaphorases as a first step toward gaining an in- Mo.) 0.15 umole of 2,6-dichlorophenol-indophenol
sight into the processes by which reduced pyridine (2,6-DCPIP, Fisher Scientific Co., Pittsburgh, Pa.),
nucleotides are oxidized by this bacterium. The 0.48 Amole of NADH2 (Sigma Chemical Co.), and
assumption that these enzymes necessarily partici- enzyme, in a total volume of 3.0 ml. In practice, a stock
pate in oxidative pathways is, of course, subject solution containing the buffer, BSA, FMN, and 2,6-
DCPIP was dispensed into a cuvette, followed by
to the above reservation. addition of the reduced pyridine nucleotide and en-
It was found that sonic extracts of A. aerogenes zyme. This stock solution was stable for at least 2
contain five diaphorases which could be iso- weeks at 0 to 5 C in total darkness; the 2,6-DCPIP
lated by chromatography on diethylaminoethyl underwent an FMN-catalyzed bleaching in light.
(DEAE) cellulose. One of these, because of its Other acceptors used in specificity studies were:
instability, was not studied further. The specific- menadione (Mann Fine Chemicals, Inc., New York,
ities of the other diaphorases for NADH2 and N.Y.), equine cytochrome c (Sigma Chemical Co.),
reduced nicotinamide adenine dinucleotide phos- and potassium ferricyanide. When ferricyanide was
phate (NADPH2) have been determined with a used, the amount of reduced pyridine nucleotide was
doubled. NADPH2 was the product of Calbiochem.,
number of electron acceptors, and an attempt has Los Angeles, Calif.
been made to ascertain whether these enzymes are The inhibitors used in this study were: antimycin
associated with a particulate fraction. A (Sigma Chemical Co.), in ethyl alcohol solution;
Atabrine hydrochloride (Sterling-Winthrop, Rensse-
MATERIALS AND METHODS laer, N. Y.); amobarbital (Gane and Ingram, New
Culture. A. aerogenes B199, obtained from the De- York, N. Y.); p-chloromercuribenzene-sulfonic acid
partment of Bacteriology, University of Kansas, was (PCMS, Sigma Chemical Co.); dicumarol (Nutri-
cultured in a nutrient broth consisting of 0.5% NaCI, tional Biochemicals Corp., Cleveland, Ohio), and 2,4-
1% glucose, 1% Peptone (Difco), and 0.7% Beef Ex- dinitrophenol. The last four inhibitors were used as
tract (Difco), pH 7.4. The same medium, solidified their sodium salts, and all solutions were prepared
with 2% agar (Difco), was used for the maintenance with deionized distilled water.
of stock cultures. Spectrophotometry. The amounts of the acceptors
Fernbach flasks (2.8 liter) containing 500 ml of me- employed and the conditions used to measure the
dium were incubated at 37 C for 24 hr on a recipro- course of the various reactions are listed in Table 1.
cating shaker. Inoculation was with 7 ml of a subcul- The concentrations of acceptors were chosen so that
ture previously incubated in test tubes on the shaker the maximal change in absorption which could occur
for 24 hr. was 1.00.
After centrifugation, the cells from each 2 liters of Initial rates of all enzymatic reactions were meas-
culture medium were washed in 150 ml of 0.01 to 0.02 ured with a Bausch & Lomb Spectronic 505 recording
1406 BERNOFSKY AND MILLS J. BACTERIOL.

TABLE 1. Summary of reactions investigated

Acceptor Reaction measured Wave-

length Molar extinction
coefficient Amtlof
reactant Refer-
efe
per assay ec

mA ,umoles
Menadione .......... Disappearance of NADH2 or 340 6.22 X 103 0.482a (9)
NADPH2
2,6-DCPIP .......... Disappearance of oxidized 2,6- 600 20.0 X 103 0.150 (18)
DCPIP
Ferricyanide ........ Disappearance of ferricyanide 420 1.04 X 103 2.88 (8)

Downloaded from http://jb.asm.org/ on November 22, 2017 by guest

Cytochrome c ....... Formation of reduced cytochrome c 550 21.0 X 103b 0.143 (12)
a This figure is for NADH2 and NADPH2. The amount of menadione present per assay was 0.40
,Amole, added in 0.04 ml of 95% ethyl alcohol solution. Oxygen was the final acceptor in this system,
since the product of menadione reduction is rapidly autoxidized.
b Based on the difference of molar absorption coefficients between the oxidized and reduced forms of
cytochrome c at 550 mp&.
Spectrophotometer which was modified by removal of taining 0.25 M potassium phosphate (pH 8.0) which
tubes V-211 (6CM6) and V-210 (12AT7) from circuit fed into the first two of mixing chambers in series,
board no. 2, and by connection of the instrument to a each containing 250 ml of the initial buffer used. The
constant voltage transformer. This procedure discon- second mixing chamber discharged into the column.
nected the drum-braking mechanism and increased The flow rate was maintained at 15 to 20 ml/hr by
the uniformity of the drum speed. adjustment at the reservoir.
Cuvettes of 1.00-cm light path were used, and the
assays were conducted at room temperature. A cuvette RESULTS
containing the assay mixture, but no acceptor or en- Chromatographic separation. Chromatography
zyme, was placed in the reference beam of the spectro- of the supernatant fraction with DEAE-cellulose
photometer. After addition of enzyme to the sample by a stepwise procedure resulted in the elution of
cuvette, the change in absorption was recorded for 1
min. These rates were nearly always linear. The non- the diaphorase activity between 0.05 and 0.15 M
enzymatic reduction of 2,6-DCPIP by NADH2, which potassium phosphate (pH 8.0) with a recovery of
was AA600 = 0.02 per min, was subtracted from the 51%.
enzymatic assays. For separation of the various enzymes by
Units of activity. One unit of diaphorase (indo- gradient elution chromatography, 175 ml of a
phenol reductase) activity is defined as causing a supernatant fraction containing 1.32 g of protein
AA600 per min of 1.00 in 3.00 ml of solution of 1.00-cm and 1,313 units of indophenol reductase activity
light path, when 2,6-DCPIP is the electron acceptor. was dialyzed with 0.02 M potassium phosphate
Specific activity is the micromoles of acceptor reduced (pH 8.0) and was adsorbed onto an 18 by 170-mm
per min per milligram of enzyme protein, regardless of
whether the electron change is two per mole (menadi- column of DEAE-cellulose. The upper surface
one and 2,6-DCPIP), or one per mole (cytochrome c was occasionally stirred to improve the flow rate.
and ferricyanide). It should be noted that, since sub- After the extract, 35 ml of 0.02 M buffer was
strate and acceptor concentrations were arbitrarily passed through the column, and the gradient elu-
chosen, the latter on the basis of photometric con- tion was begun. The mixing chambers contained
venience, the activities reported are not necessarily the 0.04 M potassium phosphate (pH 8.0) and the
maximal rates obtainable with the diaphorases. gradient was measured by phosphate analysis of
DEAE-cellulose. Type 20 Selectacel-DEAE (Brown about every 10th tube, by use of the method de-
Co., Berlin, N. H.) was repeatedly washed with 1.0 M scribed by Taussky and Shorr (22).
neutral phosphate buffer and filtered on a Buchner
funnel until tan pigment was no longer extracted. The activity of the fractions, measured as
After being washed with deionized water, the fines indophenol reduction, is shown in Fig. 1 as open
were removed and the product was suspended in water circles. Fifty-five per cent of the activity was re-
and stored in the cold. Fast-flowing columns were pre- covered. The small peak labeled cytochrome bi
pared by pouring the DEAE-cellulose at room tem- was brown in color and had spectral characteris-
perature all at once and allowing it to settle by gravity. tics identical with- the cytochrome bi from A.
They were then equilibrated with the buffer used to aerogenes described by Smith (20). The difference
dialyze the extract to be chromatographed. spectrum (Fig. 2) was taken at room temperature
Chromatography. All chromatography was per- with a Zeiss PMQ II spectrophotometer by place-
formed in a cold room at 4 C by use of a Rinco frac-
tion collector with a volumetric siphon which delivered ment of the peak fraction into two microcuvettes
an average of 5.8 ml per fraction. The apparatus for and addition of a small pinch of dithionite to the
gradient elution consisted of an upper reservoir con- one in the sample position.
VOL. 92, 1966 DIAPHORASES FROM AEROBACTER AEROGENES 1407
reductase activity was measured directly by fol-
7 lowing NADH2 oxidation at 340 m,u.
It should be pointed out that, although the
mixed diaphorase fraction had no significant
-
0 6 0.2

>- J I--
zwz 0.1 cr +.20 430
Z 5
N _ 0
L -
- 4 0.0
2 z +. 6
cr X
0. Lu
2:
ZL 3
560
i +.12
op-
QLo W
2

Downloaded from http://jb.asm.org/ on November 22, 2017 by guest

>
o - w 530
0 4 0
4 J Z
+.08
a: I
co
0
C,)

I
c
<r +.04
40 60 80 100 120
FRACTION NUMBER
I
FIG. 1. Gradient-elution chromatography of a sonic .00
extract ofAerobacter aerogenes. Activity measured with
2,6-DCPIP. Symbol:)X = fluorescent emission at 517
mfiA upon activation at 444 m,u.
-.04
The three major fractions of diaphorase activity
shown in Fig. 1 are labeled I, II, and III, in order 400 500 600 700
of their emergence from the chromatographic
column. In addition, a fluorescent yellow fraction WAVELENGTH (mp)
(labeled flavoprotein) which possessed no di- FIG. 2. Difference spectrum of cytochrome b1 ob-
aphorase activity was eluted between the cyto- tained from chromatography illustrated in Fig. 1.
chrome bi and enzyme I peaks. When activated
with light at 444 m,u, this fraction fluoresced with 12
an emission peak at 517 m,u, as measured with an w

Aminco-Bowman Spectrophotofluorometer.
Flavin requirement. Although the dialyzed N
z 10
supernatant fraction did not show a dependence w

for added flavin coenzymes, the diaphorases ob-

tained by DEAE-cellulose chromatography ap- 8
peared to have an absolute requirement for FMN.
The flavin requirement was determined with a w
I-
preparation from a preliminary chromatography 6
z
which failed to separate the activities. Figure 3
shows the dependence on FMN of this enzyme 4
mixture. The apparent Km for FMN, as calculated
from a double reciprocal plot, was 1.9 x 10-5 M.
FAD was inactive. 2
(0
In this study, a coupled menadione-2, 6-DCPIP
assay was used in which 0.4 ,umole of menadione
in 0.04 ml of ethyl alcohol was added to the 44 0

standard indophenol reductase assay. This system, .1 .2 .3 .4 .5

which essentially measured menadione reduction, pMOLES FMN OR FAD PER ASSAY
was followed at 600 m,u and depended upon the FIG. 3. Flavin dependence of an unfractionated peak
rapid nonenzymatic reduction of 2,6-DCPIP by of diaphorase activity eluted from DEAE-cellulose.
reduced menadione. In later studies, menadione Activity measured with menadione and 2,6-DCPIP.
1408 BERNOFSKY AND MILLS J. BAcTERioL.

menadione reductase activity in the absence of TABLE 2. Stabilization of a mixed diaphorase

added FMN, the apparent Km obtained probably fraction toward dilution"
reflected the activity of the component with the Activity
smallest catalytic requirement for FMN. The Final concn in reaction mixture remaining 120
flavin dependencies of the individual diaphorases min after five-
fold dilution
were not determined.
Effect of dilution. Another characteristic of the
chromatographed diaphorases was their in- 0.01 M KPO4, pH 7.4 7
................

stability to dilution. This property was studied 0.05 M KPO4, pH 7.4 27

................

with the same diaphorase preparation described 0.05 M KPO4, pH 7.4 + 1.0% BSA. 46
above. In the experiment shown in Fig. 4A, the 0.05 M KPO4, pH 7.4 + 0.13 mM
mixed diaphorase fraction in 0.05 M potassium FMN.............................70
phosphate (pH 7.5) was diluted fivefold at 0 C 0. 10 M KPO4, pH 7 4 + 1 .0%
with the same buffer, and was assayed at varying BSA + 0.13 mM FMN ..... ....... 100

Downloaded from http://jb.asm.org/ on November 22, 2017 by guest

time intervals with the coupled menadione-2,6- a See text for experimental procedure.
DCPIP assay. There was initially a rapid inactiva-
tion followed by a period during which the ac-
tivity did not change. The undiluted enzyme assay of the diaphorases, 12-ml samples of the
preparation showed no loss of activity after 4 hr. supernatant fraction containing 200 mg of protein
Although the inactivation described here probably and 242 units of indophenol reductase activity
reflected the behavior of the most unstable com- were dialyzed with 0.01 M potassium phosphate
ponent with menadione reductase activity, more (pH 8.0) and were adsorbed onto each of two
than one diaphorase was found to be unstable to identical 11 by 120-mm columns of DEAE-cellu-
dilution as shown below. lose. These were then eluted in the usual manner
With use of the activity remaining 120 min by use of 0.01 M buffer in the niixing chambers.
after fivefold dilution as a quantitative measure The eluate from one column was collected into
of inactivation, it was found (Fig. 4B) that in- empty receiving tubes, whereas the eluate from
creasing concentrations of phosphate buffer could the other was collected into tubes containing 1.2
stabilize the preparation. Additional effects of ml of 1.0 M potassium phosphate (pH 7.4) and 5.84
BSA and FMN are given in Table 2. The final X 104 M FMN. As far as possible, all other
combination shown in Table 2 which resulted in conditions were kept the same.
satisfactory stabilization for fivefold dilution was The recovery from the stabilized fractions was
not as effective where dilutions of 20-fold were 83%, compared, with 40% for the unstabilized
required, and, in these cases, stabilization was fractions; this increase was not uniform among
achieved by substituting 0.5 M potassium phos- all the fractions (Fig. 5). From these data, it
phate for the 0.1 M buffer. appears that enzyme I was the most stable, that
Effect of stabilization. To assess the extent of enzyme III and the enzyme associated with cyto-
inactivation which occurred between elution and chrome b1 were the least stable, and that enzyme
II was of intermediate stability. Figure 5 also

100
shows the presence of a previously unobserved
fraction of diaphorase activity which passed
so 0 0~~~~~~~~~~~~~~~
directly through the cellulose exchanger. This
49 activity was extremely unstable, and no further
-J
4
60 characterization was attempted. This unstable di-
z 40
0 . pofassum phosphaM aphorase was not detected when chromatography
was carried out in the absence of stabilizing solu-
_^... / ^pH74 tions.
Z 20
Activities with various acceptors. The enzymes
CO
50 100 ISO 200 0
used for this study were the peak fractions of di-
- .1 .2 .3 .4 .5
MINUTES AFTER DILUTION MOLARITY
aphorase activity obtained from the DEAE-
cellulose chromatography illustrated in Fig. 1.
Fia. 4. Effect of dilution on an unfractionated peak Table 3 lists the specific activities of the
of diaphorase activity in 0.05 M phosphate (pH 7.4) diaphorase fractions with NADH2 and NADPH2,
eluted from DEAE-cellulose. Activity measured with and with the various acceptors. The activity of the
menadione and 2,6-DCPIP. (A) Effect offive-fold di-
lution with 0.05 M potassium phosphate (pH 7.4) on ac- cytochrome bi fraction was too smaU in some
tivity. (B) Activity remaining 120 min after fivefold di- cases to be measured accurately.
lution with various concentrations of potassium phos- It can be seen that for the general purpose of
phate (pH 7.4). Abscissa indicatesfinal molarity. assaying the chromatographic fractions, 2,6-
 VOL. 92, 1966 DIAPHORASES FROM AEROBACTER AEROGENES 1409
1.4 DCPIP was the acceptor of choice, since enzymes
I, II, and III had similar specific activities, in
addition to having comparable activities with
either pyridine nucleotide. It is interesting to note
w that the rates of menadione and ferricyanide re-
duction closely paralleled one another, and that
N
z enzyme II was highly specific for NADH2 with
w
these acceptors.
-J
With cytochrome c as an acceptor, the existence
of NADPH2-specific activity in the original ex-
w
a- tract was indicated by the NADH2-NADPH2
z ratio of 0.25. However, only enzyme III, with a
NADH2-NADPH2 ratio of 0.37, appeared to be
w
significantly NADPH2-specific. It is possible that
a- a pathway for NADPH2 oxidation (i.e., trans-

Downloaded from http://jb.asm.org/ on November 22, 2017 by guest

0
0 hydrogenation) which was present in the extract
was lost upon chromatography.
Inhibition studies. The enzymes for these experi-
ments were obtained from a number of DEAE-
cellulose fractionations. The analogous enzyme
peaks were pooled, precipitated by 90% sat-
uration with ammonium sulfate, and dissolved in
a small volume of stabilizing diluent (last line,
O 10 20 30 40 50 60 70 80 90 100 Table 2). Before use, the enzymes were diluted so
FRACTION NUMBER that 0.1 ml would produce a A A per min of 0.250
FIG. 5. Gradient-elution chromatography of a sonic to 0.400 in the absence of inhibitor.
extract of Aerobacter aerogenes. Activity was measured In the typical assay, the enzyme was put into a
with 2,6-DCPIP, and 242 units were applied to each small tube together with the assay medium, but
column. Symbols:O = fractions (5.8 ml) collected into without NADH2. After addition of the inhibitor
1.2 ml of 0.58 mm FMN; 1.0 potassium phosphate and incubation for 3 min at 37 C, NADH2 was
m

(pH 7.4). Corrections were made for dilution. A = added. The reaction mixture was then transferred
fractions collected into empty receiving tubes. See text to a cuvette, and the optical density was recorded.
for conditions of chromatography. The inhibitors were tested with NADH2 only,

TABLE 3. Specific activities of chromatographed diaphorases with various acceptorsa

Conditions Diaphorase fraction

Electron acceptors Coenzyme Before chro- Cytochrome Enzyme I Enzyme II Enzyme III

2,6-DCPIP ....... NADH2 0.16 0.084 0.71 0.56 1.00

NADPH2 0.12 0.041 0.54 O.60 0.82
NADH2-NADPH2 1.29 2.05 1.32 0.94 1.22
Cytochrome c. NADH2 0.015 b 0.11 0.18 0.015
NADPH2 0.061 0.007 0.10 0.18 0.040
NADH2-NADPH2 0.25 1.05 0.95 0.37
Menadione ....... NADH2 1.05 0.09 1.75 17.7 5.49
NADPH2 0.44 0b0.54 1.32 4.44
NADH2-NADPH2 2.40 3.27 13.4 1.24
Ferricyanide ...... NADH2 1.76 0.37 4.44 59.0 12.8
NADPH2 0.76 0.32 1.97 4.50 7.50
NADH2-NADPH2 2.32 1.17 2.25 13.1 1.71
a See text for experimental details.
b Activity too small to measure.
1410 BERNOFSKY AND MILLS J. BACTERIOL.

with menadione and 2, 6-DCPIP as acceptors. An column without being adsorbed (Fig. 6A). Imme-
attempt was made to determine those concentra- diately after this fraction, there appeared a small
tions which caused an inhibition of 50%. How- peak of diaphorase activity which was only
ever, where this was not achieved, the per cent of slightly retarded by the column. Addition of the
inhibition at the highest concentration of inhibitor 0.2 M buffer immediately eluted the remaining
used is reported (Table 4). diaphorases, together with a small amount of
Dicumarol generally caused initial precipita- particulate matter. The total activity recovered
tion which, in the more dilute cases, dissolved was 147% of that placed on the column.
during incubation. The photometric interference Chromatography of the high-speed supernatant
of persistent precipitates and the ultraviolet ab- fraction into the stabilizing medium is shown in
sorption of dicumarol were compensated for by Fig. 6B. It is apparent that, whereas the high-
positioning a cuvette in the reference chamber so speed centrifugation removed the material re-
that the ground-glass sides blocked part of the sponsible for the turbidity, it did not affect the
light beam. Erratic tracings were repeated over activity of the diaphorases. Upon addition of the

Downloaded from http://jb.asm.org/ on November 22, 2017 by guest

each other by use of the displacement control to 0.2 M buffer, a small amount of particulate ma-
reposition the pen over the previous tracing. terial was eluted. The total recovery of activity
Reproducible data were obtained. Amobarbital was 161% of that added.
also generally caused precipitation; however, Another portion of the high-speed supernatant
little or no inhibition by this barbiturate was ob- fraction, dialyzed with 0.005 M potassium phos-
served. phate (pH 8.0) was chromatographed as above,
The data (Table 4) clearly indicate that the except that the fractions were not collected into
enzymes in the various fractions are quite different the FMN-phosphate stabilizing medium. Stepwise
from each other. In general, indophenol reductase elution was carried out with 50 ml of 0.005 M, 50
activities are less affected by the inhibitors than ml of 0.02 M, and 75 ml of 0.2 M potassium phos-
are menadione reductase activities, PCMS inhibits phate (pH 8.0). The initial peak of activity previ-
all of the enzymes except the cytochrome bi frac- ously seen was essentially absent (Fig. 6C). Full
tion, and amobarbital does not appreciably in- activity, however, was retained in the peak after
hibit any of the diaphorases, with the exception of the addition of the 0.2 M buffer. It is assumed that
the indophenol reductase activity of enzyme I. Of the unadsorbed diaphorase decomposed rapidly
particular interest is the relatively strong inhibition in the absence of the stabilizing medium.
of the menadione reductase activities of enzymes These experiments failed to demonstrate the
I, 11, and the enzyme in the cytochrome bi fraction association of diaphorase activity with a particu-
by antimycin A, 2,4-dinitrophenol, dicumarol, late fraction not sedimented at 20,000 X g under
and Atabrine. The menadione reductase activity the conditions stated. The unexpected increases in
of enzyme III is not inhibited by antimycin A. activity after chromatography may have been due
Distribution study. An attempt was made to to the removal of some inhibitor, or to the dissoci-
demonstrate an active particulate fraction. Bac- ation of the natural flavin followed by recombina-
terial cells were subjected to sonic disintegration tion with FMN, with which the activity was
in 0.02 M potassium phosphate (pH 8.0), and the greater, or to some other cause. In contrast to
extract was centrifuged at 20,000 x g for 20 min, these high activities obtained when the enzymes
yielding a highly turbid supernatant fraction. were on the column only a short time, recovery
Two-thirds of this low-speed supernatant fraction was only 81% from the gradient elution collected
was centrifuged at 105,000 x g for 4 hr. After into stabilizing medium. It is apparent that con-
dialysis with the above buffer, the supernatant siderable losses of activity resulted when the
fraction from the low-speed centrifugation (10.0 diaphorases were on the column for extended
ml) and that from the high-speed centrifugation periods of time.
(11.5 ml), each containing 279 units of indophenol DIscussIoN
reductase activity, were applied to 11 by 120-mm
columns of DEAE-cellulose. A 50-ml amount of The present work furnishes another example of
0.02 M potassium phosphate (pH 8.0) was put the observation that microorganisms contain a
through the column, followed by 75 ml of 0.2 M number of enzymes which can catalyze the oxida-
buffer; 5.8-ml fractions were collected into 1.2 ml tion of reduced pyridine nucleotides. Robinson
of the stabilizing medium described earlier. Tur- and Mills (16) isolated three diaphorases from the
bidity was measured with a 10-mm round cuvette soluble fraction of Pasteurella tularensis, and a
by use of a Klett colorimeter with a 660-m,u filter. fourth one by deoxycholate extraction of the
Most of the material responsible for the tur- particulate fraction. These diaphorases were
bidity of the low-speed supernatant fraction characterized by their coenzyme specificities and
(cross-hatched areas) passed directly through the transhydrogenase activities. Dolin (4) reported
VOL. 92, 1966

E
(-1

4,))
*_
0
0.~

Q..
0
Ctn

Fl.-
4)
___

E~~~

u
I

=
N)0
._

a
fi

P-4
C;
p
xo
el

_
0

.o

0
0

ra

._
.Z
_

1.
.4=

4)
0

n
g
- __

N~~~
As
vo-
DIAPHORASES FROM AEROBACTER AEROGENES

<
o

v ~;4g=

4)0

¢E

p44
0)0

A".,

_N
0
-
-
x
0

X
0

o
-

X
-

x xx

41
o
0

0
-
n

t)
0
e

ew
X
o
O
o
CO
00~C
_

-: CD

~xz ox x

10^
X IC61X00

II

68
xx
-.

XX
-0

ttr
ON

-~
0)

00

XX
C
r
)

'-O U:

t
-
o
i

C)

r~'..O

--~C

00
sr
XX
oo
r

tIXX
'r0
n
-)C)C)

v cor
0
>X
tm
'-0N

ee:

-~~C)O
tYx
r

.~
o

,N
o o

xx
00 "
00

00
1r

_O
IX x
x
%0 0

xx
o
-

O O
xx
co

O
V)
on

00
X

x
o

r)'ft
00

XX

XX X
o o 4)

.
X

cm

0
c.
0
~~~~~~~~~~~~o
X
O
0

O
0

0
X
O

o~x

x
0

~-

If)
0

X
0\
00

C
f

0
C
x
-6
4)

o
._
.-

0
.)
0
._
4-)

-oQ
.0

cu.0
-=
4.4 (U
_.

<_ Z

e
._

20
1411

Downloaded from http://jb.asm.org/ on November 22, 2017 by guest

 1412 BERNOFSKY AND MILLS J. BACTERIOL.

also been described by Repaske and Lizotte (15).

Wosilait (27), in his studies of menadione re-
ductase from dog liver, found that this enzyme
was active with both NADH2 and NADPH2, and
that it was able to reduce vitamin K1. Marki and
30'55
tiu/t2
units
6 units
units Martius (11) described a similar vitamin K1 re-
3 °| 3D5 5 394 36i6 |
ductase which they were able to obtain in a very
0

highly purified form from bovine liver. Ernster

cz 2 50
w
et al. (5) reported on the occurrence of an abun-
~~~~~~~~~20
dant "DT diaphorase" from rat liver which was

Downloaded from http://jb.asm.org/ on November 22, 2017 by guest

active with both NADH2 and NADPH2. This
0 I.-
~ ~ ~ ~ ~
enzyme could utilize menadione as well as 2,6-
DCPIP and ferricyanide as electron acceptors,
to 20 20
10 l0 20 30
and was inhibited by Dicumarol.
FRACTION NUMBER In the present study, enzyme II is seen to be a
FIG. 6. Stepwise-elution chromatography of low- and very active menadione reductase, relatively spe-
high-speed supernatant fractions from a sonic-treated cific for NADH2. It is inhibited by Dicumarol,
extract. Symbols: 0 = units of diaphorase activity per 2,4-dinitrophenol, PCMS, Atabrine, and anti-
milliliter in each fraction; X = turbidity with 660-msu mycin A, but not by amobarbital. Enzyme III,
filter. (A) Low-speed supernatant fraction, collected also an active menadione reductase, functions
into FMN-phosphate medium. (B) High-speed super- almost equally with either NADH2 or NADPH2.
natant fraction, collected into FMN-phosphate medium. It is not inhibited by antimycin A. These enzymes
(C) High-speed supernatant fraction, not collected into from A. aerogenes differ from the menadione re-
FMN-phosphate medium. Corrections made for volume. ductase isolated from E. coli by Bragg (2), not
See text for conditions of chromatography.
only in their sensitivity to inhibitors, but also in
the relative ease with which they are obtained in
the presence of five NADH2-oxidizing enzymes in the soluble state. Menadione reductase from E.
Streptococcus faecalis. One of these, an FMN- coli must be extracted from a particulate fraction.
containing diaphorase which was purified 600- This difference between the two bacteria with
fold, was active with 2, 6-DCPIP, ferricyanide, regard to intracellular associations is not limited
and certain benzoquinones and naphthoquinones. to menadione reductase. Fujita et al. (6) found
The other four enzymes were characterized as that, in E. coli, cytochome bi is also localized in
being an oxidase, a peroxidase, a menadione the particulate fraction and requires vigorous con-
reductase, and a cytochrome c reductase, respec- ditions for its solubilization. Preparations have
tively. been obtained which contain this cytochrome in
Two fractions which specifically catalyze the association with either formate dehydrogenase or
oxidation of NADH2 have been isolated from formate dehydrogenase plus nitrate reductase.
Lactobacillus casei by Walker and Kilgour (24, In contrast, cytochrome bi is readily obtained in a
25). The first requires FMN for maximal activity soluble state from Micrococcus denitrificans and
and can utilize 2, 6-DCPIP, ferricyanide, or p- Pseudomonas denitrificans (23), as well as from A.
benzoquinone as electron acceptors. The other aerogenes.
fraction, which contains bound FAD, possesses From the data in the present study, one cannot
both peroxidase and oxidase activities which determine whether cytochrome bi is itself catalyti-
could not be separated. cally active, or whether it co-chromatographs with
Of special interest are the diaphorases which another diaphorase. In either case, the cyto-
can utilize as acceptors the simple benzoquinones chrome-containing fraction is relatively
and naphthoquinones that are models of com- a diaphorase under the conditions used. It is of
inactive as
pounds having a very broad distribution in nature. interest that cytochrome bi from E. coli is active as
Many of the organisms surveyed by Wosilait and an electron acceptor with purified menadione
Nason (28) were found to contain such activities, reductase from the same organism (3).
and these investigators isolated from Escherichia
coli both menadione reductase and a quinone re- The requirement for FMN and the relative ease
ductase. Menadione reductase from this source with which the native flavin is dissociated upon
has been further purified and studied by Bragg chromatography appear to be general properties
(2). In addition, Asano and Brodie (1) isolated of the diaphorase fractions prepared from A.
menadione reductases from extracts of Myco- aerogenes. Flavin is not lost appreciably upon
bacterium phlei. An NADH2-specific menadione overnight dialysis, and it is likely that the cellu-
reductase from Hydrogenomonas eutropha has lose exchanger is directly involved in separating
VOL. 92, 1966 DIAPHORASES FROM AEROBACTER AEROGENES 1413
the flavins from the enzymes. This situation re- 5. ERNSTER, L., L. DANIELSON, AND M. LJUNGGREN.
sembles that of the vitamin K1 reductase from dog 1962. DT diaphorase. I. Purification from the
liver (27), in which the enzyme does not exhibit a soluble fraction of rat-liver cytoplasm, and
requirement for flavin until after chromatog- properties. Biochim. Biophys. Acta 58:171-188.
raphy on DEAE-cellulose. In the latter case, the 6. FUJITA, T., E. ITAGAKI, AND R. SATO. 1963. Puri-
fication and properties of cytochrome b, from
enzyme is more strongly activated by FAD than Escherichia coli. J. Biochem. (Tokyo) 53:282-
by FMN. 290.
The yellow fluorescent band shown in Fig. 4 7. HUENNEKENS, F. M., S. P. FELTON, N. A. RAO,
may be free flavin, since the conditions under AND B. MACKLER. 1961. Flavin component of
which it appears are similar to those described by reduced diphosphopyridine nucleotide de-
Huennekens et al. (7) for the chromatography of hydrogenase. J. Biol. Chem. 236 :PCS7-PC59.
FMN on DEAE-cellulose. The sensitivity to FMN 8. IBERS, J. A., AND N. DAVIDSON. 1951. On the inter-
of the FMN-specific diaphorase system reported action between hexacyanatoferrate (III) ions
here is about 10-fold that of the NADPH2-cyto- and (a) hexacyanatoferrate (II) or (b) iron (III)
ions. J. Am. Chem. Soc. 73:476-478.
chrome c reductase of yeast (7); this may be useful 9. KORNBERG, A., AND W. E. PRICER, JR. 1953. Di-

Downloaded from http://jb.asm.org/ on November 22, 2017 by guest

as an analytical tool for flavin identification. phosphopyridine nucleotide, p. 20-24. In E. E.
The instability to dilution is another charac- Snell [ed.], Biochemical preparations, vol. 3.
teristic of diaphorases from A. aerogenes. As seen John Wiley & Sons, Inc., New York.
in Fig. 5, the presence of FMN and concentrated 10. LOWRY, 0. H., N. J. ROSEBROUGH, A. L. FARR,
phosphate buffer in the collecting tubes appears to AND R. J. RANDALL. 1951. Protein measure-
minimize the dilution-inactivation which probably ment with the Folin phenol reagent. J. Biol.
is responsible for the poor recoveries in the experi- Chem. 193:265-275.
11. MARKI, F., AND C. MARTIUS. 1961. Vitamin K-
ments illustrated in Fig. 1. The unstable diaph- Reduktasen aus Rinds- und Rattenleber. Bio-
orase (Fig. 5) is apparently an extreme example of chem. Z. 334:293-303.
this lability. 12. MASSEY, V. 1959. The microestimation of succi-
The physiological roles of the isolated diaph- nate and the extinction coefficient of cyto-
orases are, of course, still unknown, since what chrome c. Biochim. Biophys. Acta 34:255-256.
was measured is the ability of these enzymes to 13. MASSEY, V. 1960. The identity of diaphorase and
catalyzethe transfer of electrons from NADH2 and lipoyl dehydrogenase. Biochim. Biophys. Acta
NADPH2 to arbitrarily chosen electron acceptors. 37:314-322.
Elucidation of the pathways of reduced pyridine 14. MINAKAMI, S., T. CREMONA, R. L. RINGLER, AND
T. P. SINGER. 1963. Studies on the respiratory
nucleotide oxidation will require further studies chain-linked reduced nicotinamide adenine di-
both of the isolated components of the systems nucleotide dehydrogenase. III. Catalytic proper-
and of their interactions within the cell. ties of the enzyme from beef heart. J. Biol.
Chem. 238:1529-1537.
ACKNOWLEDGMENTS 15. REPASKE, R., AND C. L. Lizom. 1965. The elec-
This investigation was supported by Public Health tron transport system of Hydrogenomonas
Service grant E-806, and by National Science Founda- eutropha. II. Reduced nicotinamide adenine di-
tion grant G-8245. nucleotide-menadione reductase. J. Biol. Chem.
We are indebted to John Forman, Carolyn Huntley, 240:4774-4779.
and Roberta Jacobs for their assistance with this work. 16. ROBINSON, D. A., AND R. C. MILLS. 1961. Oxida-
tion of reduced pyridine nucleotides in Pas-
LITERATURE CITED teurella tularensis. II. Separation of diaphorases
1. ASANO, A., AND A. F. BRODIE. 1965. The proper- and transhydrogenases. Biochim. Biophys.
ties of the non-phosphorylative electron-trans- Acta 48:85-92.
port bypass enzymes of Mycobacterlum phlei. 17. SANADI, D. R., AND J. W. LITTLEFIELD. 1953.
Biochem. Biophys. Res. Commun. 19:121-126. Studies on a-ketoglutaric oxidase. III. Role of
2. BRAGG, P. D. 1965. Purification and properties of coenzyme A and diphosphopyridine nucleotide.
menadione reductase of Escherichia coli. Bio- J. Biol. Chem. 201:103-115.
chim. Biophys. Acta 96:263-271. 18. SAVAGE, N. 1957. Preparation and properties of a
3. BRAGG, P. D. 1965. Quinone-independent reduc- highly purified diaphorase. Biochem. J. 67:146-
tion of cytochrome bi by reduced nicotinamide 155.
adenine dinucleotide in a purified soluble sys- 19. SEARLS, R. L., AND D. R. SANADI. 1961. A com-
tem from Escherichia coli. J. Bacteriol. 90: parative study of dihydrolipoyl dehydrogenase
1498-1499. and diaphorase. J. Biol. Chem. 236:580-583.
4. DOLIN, M. I., AND N. P. WOOD. 1960. The Strep- 20. SMITH, L. 1954. Bacterial cytochromes. Bacteriol.
tococcus faecalis oxidases for reduced diphos- Rev. 18:106-130.
phopyridine nucleotide. V. A flavin mono- 21. STRAUB, F. B. 1939. Isolation and properties of a
nucleotide-containing diaphorase. J. Biol. flavoprotein from heart muscle tissue. Biochem.
Chem. 235:1809-1814. J. 33:787-792.
1414 BERNOFSKY AND MILLS J. BAcThRIoL.

22. TAUSSKY, H. H., AND E. SHORR. 1953. A micro- bacillus casei. II. Oxidase and peroxidase. Arch.
colorimetric method for the determination of Biochem. Biophys. 111:534-539.
inorganic phosphorous. J. Biol. Chem. 202:675- 26. WATARI, H., E. B. KEARNEY, T. P. SINGER, D.
685. BASINSKI, J. HAUBER, AND C. J. LusTY. 1962. A
23. VERNON, L. P. 1956. Bacterial cytochromes. I. comparison of reduced nicotinamide adenine
Cytochrome composition of Micrococcus dinucleotide-cytochrome c reductases from
denitrificans and Pseudomonas denitrificans. J. heart mitochondria and their formation from
Biol. Chem. 22:1035-1044. the respiratory chain-linked dehydrogenase. J.
24. WALKER, G. A., AND G. L. KIouR. 1965. Pyri- Biol. Chem. 237:PC1731-1734.
27. WOSILAIT, W. D. 1961. Role of vitamin K in
dine nucleotide oxidizing enzymes of Lacto- electron transport. Federation Proc. 20:1005-
bacillus casei. I. Diaphorase. Arch. Biochem. 1011.
Biophys. 111:529-533. 28. WOSILAIT, W. D., AND A. NASON. 1954. Pyridine
25. WALKER, G. A., AND G. L. KILGOUR. 1965. Pyri- nucleotide-menadione reductase from Escheri-
dine nucleotide oxidizing enzymes of Lacto- chia coli. J. Biol. Chem. 208:785-798.

Downloaded from http://jb.asm.org/ on November 22, 2017 by guest