Proc. Nati. Acad. Sci.
USA
Vol. 84, pp. 76-79, January 1987
Biochemistry
Solubilization of a membrane-bound diol dehydratase
with retention of EPR g = 2.02 signal by using 2-(N-
cyclohexylamino)ethanesulfonic acid buffer
(oxygen-sensitive enzyme/Clostridium glycolicum/radical scavengers/EPR radical signal)
MARIS G. N. HARTMANIS AND THRESSA C. STADTMAN
Laboratory of Biochemistry, National Heart, Lung, and Blood Institute, National Institutes of Health, Building 3, Room 103, Bethesda, MD 20892
Contributed by Thressa C. Stadtman, September 18, 1986
ABSTRACT A procedure for solubilization of the oxygen-
labile, membrane-bound diol dehydratase from Clostridium
glycolicum with retention of enzymatic activity is described.
The procedure involves sonication of crude membrane prepa-
rations anaerobically in 0.1 M 2-(N-cyclohexylamino)ethane-
sulfonic acid (CHES) buffer (pH 8.6-9.0) containing 2 mM
dithiothreitol. The addition of dimethylsulfoxide (30%) and
lysophosphatidylcholine (0.15 mg/ml) to the solubilization
buffer resulted in a 10-fold increase in recovery of solubilized
diol dehydratase activity. After ultracentrifugation, an overall
recovery of 50% of the activity initially present in the crude
membrane preparations was achieved. Active membrane prep-
arations and the solubilized enzyme exhibited an EPR signal at
g = 2.02. Both enzyme activity and EPR signal were sensitive
to oxygen and the radical scavengers, NH20H and hydroxy-
urea.
A diol dehydratase isolated from Clostridium glycolicum
converts 1,2-ethanediol to acetaldehyde (1) in a reaction that
is the first step in the fermentation pathway that leads to the
conversion of the diol to equimolar amounts of ethanol and
acetate (2). The diol dehydratase from C. glycolicum differs
from other bacterial diol dehydratases (3-6) in that it is
strongly associated with the cell membrane, it is extremely
oxygen sensitive, and it does not appear to utilize a cobamide
coenzyme as cofactor (1). Moreover, marked sensitivity of
the enzyme to radical scavengers such as hydroxylamine and
hydroxyurea and correlation of enzyme activity with an EPR
signal at g = 2.02 suggested that some other enzyme-bound
radical species is involved in the mechanism of the reaction
catalyzed by this clostridial diol dehydratase (1). In these
initial studies, detailed characterization of this interesting
enzyme was impossible because a wide variety of solubili-
zation procedures that were tried were ineffective in releasing
active enzyme from the membrane matrix. These procedures
included exposure to high ionic strength; treatment with
organic solvents; incubation with numerous ionic detergents
such as 3-[(3-cholamidopropyl)dimethylammonio]-1-pro-
panesulfonate (CHAPS), 3[(3-cholamidopropyl)dimethylam-
monio]-2-hydroxy-1-propanesulfonate (CHAPSO), and
many commonly used nonionic detergents; osmotic shock
followed by sonication in the presence of sucrose; treatment
with phospholipase A2; and limited proteolysis (1). Recently,
however, it was found that sonication of C. glycolicum
membrane preparations anaerobically in 2-(N-cyclohexyl-
amino)ethanesulfonic acid (CHES) buffer at alkaline pH
liberated active diol dehydratase. Details of this procedure
and results of preliminary EPR studies on the membrane-
bound and soluble forms of the diol dehydratase are reported
here.
The publication costs of this article were defrayed in part by page charge
payment. This article must therefore be hereby marked "advertisement"
in accordance with 18 U.S.C. §1734 solely to indicate this fact.
MATERIALS AND METHODS
Materials. The following products were obtained from
commercial sources: dimethylsulfoxide (Me2SO), Aldrich
Chem (Metuchen, NJ), dithiothreitol, Boehringer Mann-
heim; CHES, phosphatidylcholine, and lysophosphatidylcho-
line, Sigma; electrophoresis reagents, Bio-Rad. All other
reagents were of the highest grade available.
Growth of Bacteria and Preparation of Membranes. C.
glycolicum was grown anaerobically on 1,2-ethanediol in a
400-liter fermentor, and cells were harvested, frozen, and
stored as described (1). Frozen cell pastes were thawed
anaerobically in 100 mM Tris-HCl buffer, pH 8.0/2 mM
dithiothreitol (buffer A) and sonicated in a nitrogen atmo-
sphere. The disrupted cells were first centrifuged for 30 min
at 3000 x g, and the pellet, which contained all of the diol
dehydratase activity, was washed three or four times with
deoxygenated buffer A. The washing also removed unbroken
cells, since only the upper membrane layer was removed and
washed after each centrifugation. For maximum recovery of
diol dehydratase activity, cells were disrupted, and mem-
branes were prepared in the Anaerobic Laboratory (7) at the
National Institutes of Health.
Assays. Diol dehydratase activity was determined anaero-
bically by measuring aldehyde formation with the 2,4-
dinitrophenylhydrazine method (8) as modified (1). Protein
concentrations were estimated by the method of Bradford (9)
with bovine serum albumin as standard.
Electrophoresis. NaDodSO4/polyacrylamide gel electro-
phoresis in 12% acrylamide gels was carried out by the
method of Laemmli (10). The gels were washed overnight
with a solution containing 25% 2-propanol and 10% acetic
acid, stained with Coomassie brilliant blue G-250 as de-
scribed in (11) and destained with 7.5% acetic acid.
Solubilization. Solubilization experiments were carried out
by sonication (Branson LS-75 sonifier; 1-min bursts) of crude
membrane preparations anaerobically in the buffers indicat-
ed. Membrane preparations were sonicated in metal tubes
chilled in ice. After centrifugation at either 40,000 x g or
105,000 x g, supernatant solutions and resuspended pellets
were assayed for diol dehydratase activity. The inclusion of
2 mM dithiothreitol in the solubilization buffer was necessary
for optimal activity (1). All solubilization procedures and
centrifugation at 40,000 x g were carried out inside the
Anaerobic Laboratory. For the ultracentrifugation step,
tubes were filled and sealed inside the Anaerobic Laboratory.
EPR Spectroscopy. EPR experiments were carried out by
using a Varian E-109 spectrometer equipped with an E-102
microwave bridge. The temperature was kept at - 1960C
using liquid nitrogen. The g values were calibrated with
Abbreviations: CHES, 2-(N-cyclohexylamino)ethanesulfonic acid;
Me2SO, dimethylsulfoxide; CHAPS, 3-[(3-cholamidopropyl)-
dimethylammonio]-1-propanesulfonate; CHAPSO, 3-[(3-cholamido-
propyl)dimethylammonio]-2-hydroxy-1-propanesulfonate.
 Biochemistry: Hartmanis and Stadtman Proc. Natl. Acad. Sci. USA 84 (1987) 77

1,1-diphenyl-2-picrylhydrazyl powder (g = 2.003) (12). Other

details are described in the legend to Fig. 3.

RESULTS AND DISCUSSION -

mo -92
-- - 66
In contrast to all of the solubilization procedures tried
previously for liberation of diol dehydratase from membranes - -31
of C. glycolicum (1), sonication in CHES buffer released the
amp
diol dehydratase in a form that retained enzymic activity. The
membrane suspension was almost entirely solubilized by
sonication in 100 mM CHES, pH 8.6/2 mM dithiothreitol,
and only a minor pellet fraction of insoluble material re-
mained after centrifugation. The time course of the release of _ -2 1 FIG. 2. NaDodSO4/polyacrylamide gel elec-
diol dehydratase from its membrane matrix is shown in Fig. trophoresis of solubilized proteins from mem-
1. After sonication for 3 min, <10% of the dehydratase - 14 branes of C. glycolicum. The left-hand lane
activity remained in the pellet that was sedimented by contains 50 pyg of solubilized proteins. The
centrifugation at 40,000 x g for 30 min. In contrast, sonica- right-hand lane contains protein markers: phos-
phorylase B (92 kDa), bovine serum albumin (66
tion of a membrane preparation in 100 mM Tris-HCl, pH kDa), ovalbumin (45 kDa), carbonic anhydrase
8.6/2 mM dithiothreitol under the same conditions failed to (31 kDa), soybean trypsin inhibitor (21 kDa),
solubilize any diol dehydratase, and also little or no protein and lysozyme (14 kDa).
was released to the supernatant solution. Thus, the solubili-
zation by sonication in CHES was caused by the buffer itself -1960C, and subsequent thawing caused a total loss of
and was not an effect of alkaline pH alone. Apparently, activity of the membrane-bound enzyme (1), similar freezing
CHES has detergent-like properties that make it capable of followed by thawing 24 hr later of the solubilized enzyme
rupturing membrane structures, thereby releasing integral caused a 60% loss of activity. Neither type of enzyme
membrane proteins in a relatively mild way. CHES is an preparation withstood storage under anaerobic conditions at
amphiphilic compound exhibiting a hydrophobic hydrocar- 40C for more than a few hours without marked loss of activity.
bon moiety in the form of a cyclohexyl group and a hydro- In order to increase the extent of solubilization of the
philic part consisting of an amino group and a sulfonic acid membrane-bound diol dehydratase and perhaps improve its
group, making the molecule zwitterionic. The buffer has a stability, a large variety of substances were tested as addi-
PKa value of 9.27 in water at 25°C (13). Since CHES, unlike tives to the CHES/dithiothreitol buffer prior to sonication.
detergents such as Triton X-100, is compatible with many Supplementation with cell extracts or lipid-containing pellet
commonly used enzyme purification procedures, it need not fractions from C. glycolicum had no beneficial effect on
be removed and instead can be used as buffer in subsequent recovery of enzyme activity (Table 1). Inclusion of 20%
isolation steps.
In the initial experiments using CHES buffer as the (wt/vol) glycerol or 50 mM 1,2-ethanediol, the substrate,
solubilization agent, only about 5% of the diol dehydratase almost totally destroyed activity. Addition of catalase and
activity originally present in the crude membrane prepara- Table 1. Effect of various compounds on the recovery of
tions was recovered in an active soluble form. By varying the diol dehydratase after solubilization
concentration and the pH of the CHES buffer, the recovery
of active enzyme could be increased to 8-10o. This was Relative activity,t
achieved by using 0.1 M CHES, pH 9/2 mM dithiothreitol. A Addition Amount* %
typical NaDodSO4/polyacrylamide gel electrophoresis pro- No addition 100
file of the large number of membrane proteins solubilized by Supernatant 10% 93
this procedure is shown in Fig. 2. 35,000 x g pellet 10%0 94
Solubilization of the diol dehydratase, however, had little 35,000 x g pellet 20%o 84
effect on its stability properties. Whereas freezing, even at Supernatant 10%
+ 35,000 x g pellet 10% 97
Polyethylene glycol 400 10% 82
Polyethylene glycol 8000 6% 73
1,2-Ethanediol 50 mM 8
Glycerol 209% 3
Tween 80 1% 97
Dithionite 2 mM 71
Dithiothreitol 10 mM
+ FeSO4 5 mM 55
Catalase 600 units/ml 42
Superoxide dismutase 600 units/ml 85
Bovine serum albumin 15 mg/ml 32
Phosphatidylcholine 0.2 mg/ml 83
Lysophosphatidylcholine 1 mg/ml 69
Sonication, min Dimethylformamide 20% 88
Tetrahydrofuran 20% 79
FIG. 1. Solubilization of diol dehydratase by sonication of crude *Percentage additions were on a wt/vol basis except for supernatant
membrane preparations in 0.1 M CHES, pH 8.6/2 mM dithiothreitol. and resuspended pellet (vol/vol).
Supernatants and pellets obtained by centrifugation for 30 min at tActivity in the supernatant after solubilization and centrifugation for
40,000 x g were assayed for enzyme activity. o, Enzyme activity in 1 hr at 40,000 x g. The solubilization solution was 0.1 M CHES, pH
the resuspended pellet; *, enzyme activity released into the super- 9.0/2 mM dithiothreitol containing the indicated addition. An
natant. Maximum activity released equals 5% of the activity present activity of 100% is equivalent to -8% of the total membrane-bound
in the original membrane preparation. enzyme activity.
78 Biochemistry: Hartmanis and Stadtman
Table 2. Effect of Me2SO on the recovery of diol
dehydratase after solubilization
Me2SO, Relative activity,*
0 100
5 103
20 155
30 365
50 15
100 0 C
*Activity in supernatant after solubilization and centrifugation for 1
hr at 105,000 X g. The solubilization solution was 0.1 M CHES, pH
9.0/2 mM dithiothreitol containing the concentration of Me2SO
indicated. An activity of 100%o is equivalent to -8% of the total
membrane-bound enzyme activity. The reported pKa value for
CHES in 30%o Me2SO at 00C is 9.89 (13).
superoxide dismutase to decompose active oxygen species or
supplementation with additional reducing agents decreased
recovery of active enzyme. Similar negative effects were
observed with phosphoglycerides, dimethylformamide, and
tetrahydrofuran as additives.
In marked contrast to these results, sonication of mem-
brane preparations in the presence of Me2SO greatly in-
creased recovery of soluble active enzyme (Table 2). Addi-
tion of 30% Me2SO (vol/vol) gave a more than 3.5-fold
increase in activity recovered in the supernatant after cen-
trifugation. Activity remaining in the pellet was negligible.
The further addition of phosphoglycerides to the solubiliza-
tion system containing 30% Me2SO/0.1 M CHES, pH 9.0/2
mM dithiothreitol resulted in even greater improvement of
soluble enzyme yields (Table 3). The more than 6-fold
increase in recovery of soluble enzyme observed when both
30% Me2SO and lysophosphatidylcholine (0.15 mg/liter)
were added to the buffer system corresponds to an z50%
recovery of the total diol dehydratase initially present in the
crude membrane preparation. Addition of phosphoglycerides
alone decreased recovery of soluble diol dehydratase (Table
1), but in the presence of Me2SO (Table 3), they increased
recovery of the enzyme substantially. Unfortunately, stabil-
ity of the solubilized enzyme during storage, either at 40C or
at - 196°C, was not improved by the additions of Me2SO and
phosphoglycerides. Further studies to preserve or regenerate
the possible enzyme-bound radical species (see below) may
provide a solution to this problem.
As mentioned in a previous publication (1), membrane-
bound diol dehydratase preparations exhibit a characteristic
EPR signal that correlates with enzyme activity. Moreover,
a correlation was noted between diol dehydratase activity
and the amplitude of this EPR signal in inhibition studies with
radical scavengers such as hydroxyurea. Data documenting
these observations are shown in Fig. 3. Of particular impor-
tance is the difference between the EPR spectrum of mem-
Table 3. Effect of phosphoglycerides in the presence of Me2SO
on the recovery of diol dehydratase after solubilization
Amount, Relative activity,*
Phosphoglyceride added mg/ml % sources. In the cobamide coenzyme-independent ribonucle-
None 100
Phosphatidylcholine 1 107
Lysophosphatidylcholine 1 149
Lysophosphatidylcholine 0.35 152
Lysophosphatidylcholine 0.15 171
*Activity in supernatant after solubilization and centrifugation for 1
hr at 105,000 x g. Solubilization solution was 30% Me2SO/0.1 M
CHES, pH 9.0/2 mM dithiothreitol containing the concentration of
phosphoglyceride indicated. A relative activity of 171 represents
50% of the activity originally present in the membranes.
Proc. Natl. Acad. Sci. USA 84 (1987)
A D
B E
16 G
FIG. 3. EPR spectra of crude membranes and solubilized mem-
brane preparations of C. glycolicum. EPR conditions: microwave
power, 5 mW; frequency setting, 9.16 GHz; temperature, -196°C;
scan range, 160 G; modulation amplitude 3.2 G (spectra A, B, and C)
or 8 G (spectra D and E); modulation frequency, 100 kHz; time
constant, 0.5 s; gain, 2.5 x 104 (spectra A, B, and C) or 4 x 104
(spectra D and E). All spectra are centered around a g value of 2.02.
Spectra: A, crude membrane preparations from 1,2-ethanediol-
grown cells in buffer A; B, as for spectrum A, but after a 15-min
incubation with 10 mM hydroxyurea; C, crude membrane prepara-
tions from glucose-grown cells in buffer A; D, membrane preparation
after solubilization with 0.1 M CHES, pH 9.0/2 mM dithiothreitol
and centrifugation; E, as for spectrum D, but after incubation with
10 mM hydroxyurea. The intensities of spectra A, B, and C can be
compared directly on a relative basis; also intensities of spectra D
and E can be compared directly. However, the absolute intensities
of the two groups are on different scales.
brane preparations from cells grown on 1,2-ethanediol (Fig.
3, spectrum A) and that of membrane preparations of cells
grown on glucose (Fig. 3, spectrum C). The latter cells lack
diol dehydratase activity, which is induced by growth on
diols (1), and the characteristic EPR signal at a g value of 2.02
is virtually absent in membrane preparations of these cells.
Spectrum B shows the effect of a radical scavenger on the
amplitude of the EPR signal. Incubation of the membrane
preparations from diol-grown cells with 10 mM hydroxyurea
for 15 min decreased the amplitude of the g = 2.02 signal by
50% (compare spectra A and B). A comparable quenching of
the signal (spectrum E) was observed when hydroxyurea was
added to a solubilized diol dehydratase preparation. Addition
of 10 mM hydroxylamine to the membrane-bound enzyme
had a similar effect (data not shown). The solubilized enzyme
preparation in the absence of inhibitors exhibited spectrum
D. The decreases in signal amplitude observed in response to
the added radical scavengers was accompanied by virtually
complete loss of diol dehydratase activity in both types of
enzyme preparations. Sulfhydryl modifying agents, which
strongly inhibit diol dehydratase activity (1), also diminished
the radical signal. Treatment of the enzyme with 2 mM
p-chloromercuriphenylsulfonate for 15 min decreased the
amplitude of the signal to 25% of the control. The EPR signal
and the enzyme activity were also abolished by exposure to
oxygen (data not shown).
These correlative effects suggest that a radical species may
be an integral property of the diol dehydratase from C.
glycolicum and that this serves as a substitute for a cobamide
coenzyme, which functions in diol dehydratases from other
otide reductase of Escherichia coli, a tyrosyl free radical with
a g value of 2.0047, present in the protein B subunit of the
enzyme, has been shown to be involved in the reaction
mechanism (14). Ribonucleotide reductase activity and the
EPR signal of the enzyme show a similar sensitivity to the
radical scavenger hydroxyurea.
The authors are grateful to Dr. Hideo Kon for running the EPR
spectra. M.G.N.H. thanks the Swedish Natural Science Research
Council for their financial support.
 Biochemistry: Hartmanis and Stadtman
1. Hartmanis, M. G. N. & Stadtman, T. C. (1986) Arch. Bio-
chem. Biophys. 245, 144-152.
2. Gaston, L. W. & Stadtman, E. R. (1963) J. Bacteriol. 85,
356-362.
3. Lee, H. A. & Abeles, R. H. (1963) J. Biol. Chem. 238,
2367-2373.
4. Poznanskaya, A. A., Tanizawa, K., Soda, K., Toraya, T. &
Fukui, S. (1979) Arch. Biochem. Biophys. 194, 379-386.
5. McGee, D. E. & Richard, J. H. (1981) Biochemistry 20, 4293-
4298.
6. Toraya, T. & Fukui, S. (1982) in B,2, ed. Dolphin, D. (Wiley,
New York), Vol. 2, pp. 233-262.
7. Poston, J. M., Stadtman, T. C. & Stadtman, E. R. (1971)
Proc. Natl. Acad. Sci. USA 84 (1987) 79
Methods Enzymol. 22, 49-54.
8. Bohme, H. & Winkler, 0. (1954) Z. Anal. Chem. 142, 1-5.
9. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254.
10. Laemmli, U. K. (1970) Nature (London) 227, 680-685.
11. Reisner, A. H., Nemes, P. & Bucholtz, C. (1975) Anal. Bio-
chem. 64, 509-516.
12. Bolton, J. R., Borg, D. C. & Swartz, H. M. (1972) in Biolog-
ical Applications of Electron Spin Resonance, eds. Swartz,
H. M., Bolton, J. R. & Borg, D. C. (Wiley, New York), p.
100.
13. Taylor, M. J. & Pignat, Y. (1982) Cryobiology 19, 99-109.
14. Sjoberg, B.-M., Reichard, P., Graslund, A. & Ehrenberg, A.
(1977) J. Biol. Chem. 252, 536-541.