The N-terminal regions of β and γ subunits lower the solubility of adenosylcobalamin-dependent diol dehydratase

Article abstract of DOI:10.1271/bbb.69.455

Adenosylcobalamin-dependent diol dehydratase is one of essential components of carboxysome-like polyhedral bodies. It exists as a heterohexamer (αβγ)₂, and its activity is recovered in a precipitant fraction of Klebsiella oxytoca and overexpressing Escherichia coli cells. Limited proteolysis of the enzyme with trypsin converted the enzyme into a highly soluble form without loss of enzyme activity. The N-terminal amino acid sequencing of the enzyme thus solubilized indicated that the N-terminal 20 and 16 amino acid residues had been removed from the β and γ subunits, respectively. Mutant enzymes with the same N-terminal truncations of either or both of the β and γ subunits were expressed on a high level in E. coli cells. All the mutant enzymes obtained were expressed in a soluble, active form. These results indicate that the N-terminal regions of the β and γ subunits lower the solubility of diol dehydratase. The mutant enzyme with the N-terminal truncations of both β and γ subunits was essentially indistinguishable in catalytic properties from recombinant wild-type enzyme or the enzyme purified from K. oxytoca in a soluble form.

Full text of DOI:10.1271/bbb.69.455

Biosci. Biotechnol. Biochem., 69 (3), 455–462, 2005
The N-Terminal Regions of ꢀ and ꢁ Subunits Lower the Solubility
of Adenosylcobalamin-Dependent Diol Dehydratase
y
Takamasa TOBIMATSU, Masahiro KAWATA, and Tetsuo TORAYA
Department of Bioscience and Biotechnology, Faculty of Engineering, Okayama University,
Tsushima-naka, Okayama 700-8530, Japan
Received June 14, 2004; Accepted December 27, 2004
Adenosylcobalamin-dependent diol dehydratase is
one of essential components of carboxysome-like poly-
organelles consist of at least 15 proteins, including
proteins having sequence similarity to the shell proteins
1
0)
hedral bodies. It exists as a heterohexamer ðꢂꢀꢁÞ , and
of carboxysome,
diol dehydratase (PduCDE), its
2
its activity is recovered in a precipitant fraction of
Klebsiella oxytoca and overexpressing Escherichia coli
cells. Limited proteolysis of the enzyme with trypsin
converted the enzyme into a highly soluble form without
loss of enzyme activity. The N-terminal amino acid
sequencing of the enzyme thus solubilized indicated that
the N-terminal 20 and 16 amino acid residues had been
removed from the ꢀ and ꢁ subunits, respectively.
Mutant enzymes with the same N-terminal truncations
of either or both of the ꢀ and ꢁ subunits were expressed
on a high level in E. coli cells. All the mutant enzymes
obtained were expressed in a soluble, active form. These
results indicate that the N-terminal regions of the ꢀ and
putative reactivating factor (PduGH), aldehyde dehy-
drogenase (PduP), and ATP:cob(I)alamin adenosyltrans-
ferase (PduO), and proposed that the organelles function
to minimize aldehyde toxity during growth on 1,2-
1
1)
propanediol. Similar bodies were observed in Kleb-
siella oxytoca grown on 1,2-propanediol as well (T. A.
Bobik and T. Toraya, unpublished results). But the
structure and protein assemblies of the bodies are still
far from understood. Diol dehydratase is a low-solubility
1
2,13)
enzyme by itself,
highly soluble enzyme,
their three-dimensional structures and catalyze essen-
whereas glycerol dehydratase is a
1
4,15)
although they are similar in
1
5–17)
tially the same reactions.
In order to promote the
ꢁ
subunits lower the solubility of diol dehydratase. The
mutant enzyme with the N-terminal truncations of both
and ꢁ subunits was essentially indistinguishable in
understanding of these enigmas, we attempted as the
first step to identify which region(s) of diol dehydratase
are responsible for its low solubility.
ꢀ
catalytic properties from recombinant wild-type enzyme
or the enzyme purified from K. oxytoca in a soluble
form.
The enzyme was first purified as a soluble protein
from Klebsiella oxytoca (formerly Klebsiella pneumo-
1
)
niae and Aerobacter aerogenes) ATCC8724, and was
shown to consist of subunits with Mrs of 60,000, 25,500,
1
8)
12)
Key words: diol dehydratase; glycerol dehydratase;
solubility of enzyme; adenosylcobalamin;
coenzyme B12
23,000, 15,500, and 14,000. Later, McGee et al.
reported that the enzyme purified after solubilization
with Triton X-100 consists of four subunits with Mrs of
60,000, 51,000, 29,000, and 19,000. They concluded
Diol dehydratase (DL-1,2-propanediol hydro-lyase,
EC 4.2.1.28) is an enzyme that catalyzes the adenosyl-
cobalamin (coenzyme B )-dependent conversion of
that the enzyme is membrane-associated, because the
majority of it was recovered in the precipitant fraction.
1
9)
Tanizawa et al. showed that the subunit heterogeneity
of the soluble enzyme might be due to proteolytic
cleavage during enzyme purification. This discrepancy
was finally solved by our cloning of the enzyme
12
1
,2)
1
is induced in some genera of Enterobacteriaceae, such
,2-diols to the corresponding aldehydes. The enzyme
3
,4)
5)
as Klebsiella and Citrobacter,
and other bacteria
2
0)
when they are grown anaerobically on 1,2-propanediol.
Glycerol is also effective for induction, but to a much
genes. We identified and overexpressed the pddABC
genes encoding the ꢀ (Mr 60,348), ꢁ (Mr 24,113), and ꢂ
(Mr 19,173) subunits of diol dehydratase in Escherichia
3
)
lesser degree. Diol dehydratase plays an essential role
in producing electron acceptors for the fermentation of
coli. The M of the ꢁ subunit determined by SDS–PAGE
r
was 30,000, close to that of the M 29,000 subunit of the
r
4
–8)
1
,2-diols and glycerol by these bacteria.
Recently,
Bobik and coworkers reported that Salmonella enterica
forms polyhedral organelles during B12-dependent
growth on 1,2-propanediol. They found that these
detergent-extracted enzyme from the precipitant frac-
tion. It turned out that the protein with Mr of 51,000 is
not the subunit of diol dehydratase but a CoA-acylating
9
)
y
To whom correspondence should be addressed. Fax: +81-86-251-8264; E-mail: toraya@cc.okayama-u.ac.jp
Abbreviations: PAGE, polyacrylamide gel electrophoresis; SDS, sodium dodecyl sulfate

456
T. TOBIMATSU et al.
propionaldehyde dehydrogenase (PduP).21,22) Recombi-
nant diol dehydratase expressed in E. coli was also
distributed mainly in the precipitant fraction and highly
pUSI2E(DD)20) was inserted into the BamHI–BglII
region of pUSI2ENd to produce pUSI2ENd(DD).
DNA segments encoding truncated ꢁ and ꢂ subunits of
diol dehydratase with deletions of N-terminal 20 and
16 amino acid residues, respectively, were amplified
by PCR using Taq DNA polymerase (Invitrogen,
1
3)
purified by solubilization with Brij 35. The X-ray
structure of the enzyme revealed that substrate 1,2-
þ
propanediol and essential K ion are bound inside the
1
6)
0
TIM barrel in the ꢀ subunit. The N-terminal 45 and 36
amino acid residues of the ꢁ and ꢂ subunits, respec-
tively, were missing during crystallization. These re-
gions appear not to be essential for catalytic function,
because the specific activities reported with different
enzyme preparations are not very much different.
In contrast, glycerol dehydratase, an isofunctional
Carlsbad, CA) and the following pairs of primers: 5 -
0
0
GCGAGCATATGGGCAGCGATAAACCG-3 and 5 -
0
TCAGATCTTAAAGCGCCACGCGCAG-3 for the
0
truncated ꢁ subunit gene, and 5 -TTGAGCCATA-
0
0
TGAACAGCCTGC-3 and 5 -AGAGATCTTAATCG-
0
TCGCCTTTGAG-3 for the truncated ꢂ subunit gene
(the NdeI and BglII sites are underlined). The 640-bp
and 500-bp DNA fragments amplified were cloned into
pCR2.1 vector using a TA cloning kit (Invitrogen, CA),
digested with NdeI and BglII, and ligated with the 5.0-kb
NdeI–BglII fragment from pUSI2ENd(DD) to produce
1
4,15,23)
enzyme, is highly soluble.
amino acid sequences of the corresponding subunits
A comparison of the
2
0,23–27)
between diol and glycerol dehydratases
high identities between them, except that the N-terminal
2 and 37 amino acid residues of the ꢁ and ꢂ subunits,
showed
0
plasmids pUSI2ENd(ꢁ ) and pUSI2ENd (ꢂ ), respec-
D
0
3
D
1
3)
respectively, of diol dehydratase are lacking in the
corresponding subunits of glycerol dehydratase. Hence it
was suggested that these regions might determine the
solubility of these enzymes, although the hydropathy
plots of diol dehydratase do not show any prominent
hydrophobic peaks in these missing regions.
In this paper, we report evidence that these regions of
diol dehydratase actually lower the solubility of the
enzyme. By protein engineering techniques, we obtained
highly soluble, truncated diol dehydratase that was
expressed in E. coli.
tively. Plasmid pUSI2E(ꢀD) was digested with BglII
and ligated with the 0.7-kb BamHI–BglII fragment
0
obtained from pUSI2ENd(ꢁ ) to yield plasmid
D
0
0
pUSI2E(ꢀDꢁ ). Plasmid pUSI2E(ꢀDꢁ ) was digested
D
D
with BglII and ligated with the 0.5-kb BamHI–BglII
0
fragment obtained from pUSI2ENd(ꢂ ) and
D
1
3)
0
0
pUSI2E(ꢂD)
and pUSI2E(ꢀ ꢁ ꢂ ),
pUSI2E(ꢀ ꢁ ) was digested with BglII and ligated
to yield plasmids pUSI2E(ꢀ ꢁ ꢂ )
D
D D
0
respectively.
Plasmid
D
D
D
1
3)
D
D
with the 0.5-kb BamHI–BglII fragment obtained from
0
0
pUSI2ENd(ꢂ ) to yield plasmid pUSI2E(ꢀDꢁDꢂ ). It
D
D
was confirmed by sequencing that no undesired muta-
tions occurred during the PCR amplifications.
Materials and Methods
Transformation of E. coli with an expression plasmid
and cultivation of the transformants were carried out as
described previously.
Materials. Crystalline adenosylcobalamin (coenzyme
B ) was a gift from Eisai, Tokyo. Trypsin was
2
0)
1
2
purchased from Sigma, St. Louis, MO. Other chemicals
were analytical grade reagents. The crude membrane
fraction of E. coli carrying recombinant diol dehydra-
tase was obtained as described previously.13)
Enzyme and protein assays. Diol dehydratase activity
was measured by the 3-methyl-2-benzothiazolinone
28)
hydrazone method using 1,2-propanediol as substrate.
One unit of diol dehydratase is defined as the amount of
enzyme activity that catalyzes the formation of 1 mmol
of propionaldehyde/min. Since the solubility of wild-
Trypsin treatment of recombinant diol dehydratase.
Recombinant E. coli cells expressing diol dehydratase
genes were disrupted by sonication in 50 mM potassium
phosphate buffer (pH 8.0) containing 2% 1,2-propane-
diol. The precipitate was washed three times with the
same buffer. The crude membrane fraction thus obtained
was re-suspended at a concentration of 5 mg of protein/
ml in the same buffer containing various amounts of
1
2,13)
type diol dehydratase is very low,
the homogenates,
extracts, and enzyme solution to be assayed were diluted
with 50 mM potassium phosphate buffer (pH 8.0) con-
taining 2% 1,2-propanediol and 0.5–1% Brij 35.
Protein was assayed by the method of Lowry et al.29)
with crystalline bovine serum albumin as a standard.
Specific activity was expressed as units/mg of protein.
ꢀ
trypsin. After incubation at 15 C for 12 h, the reaction
was terminated by the addition of excess soybean trypsin
inhibitor (400 mg/ml). The resulting suspension was
centrifuged at 16;000 ꢁ g for 30 min to separate soluble
and precipitant fractions.
PAGE. PAGE of cell-free extracts was performed
under non-denaturing conditions as described by
3
0)
20)
Davis in the presence of 0.1 M 1,2-propanediol, or
under denaturing conditions as described by Laemmli.3
Protein staining was carried out with Coomassie
Brilliant Blue R-250. Western blot analysis was per-
1)
Construction of expression plasmids. pUSI2E20) was
partially digested with NdeI and treated with a Klenow
fragment of DNA polymerase I in the presence of four
dNTPs, and then self-ligated to produce plasmid
pUSI2ENd. A 3.0-kb BamHI–BglII DNA fragment from
2
0)
formed as described previously.
Purification of mutant diol dehydratase with the N-

Highly Soluble Diol Dehydratase
457
Fig. 1. Effects of Tryptic Digestion on the Activity and Solubility of Diol Dehydratase.
ꢀ
A precipitant fraction containing 60 units/ml of diol dehydratase was incubated with the indicated concentrations of trypsin at 15 C for 12 h.
After centrifugation, enzyme activities in the supernatant (open bars) and precipitant (grey bars) fractions were measured (A). Proteins in the
supernatant and precipitant fractions after tryptic digestion were analyzed by SDS–PAGE on 11% gels (B). Fragmentation patterns (C) in the N-
terminal regions of the ꢀ, ꢁ, and ꢂ subunits upon limited proteolysis of diol dehydratase with trypsin.
terminal truncations of ꢁ and ꢂ subunits. The cell-free
extract was prepared from 3.3 g of E. coli cells carrying
the same buffer containing 50 mM KCl and then 100 mM
KCl. Purified enzyme with specific activity more than
35 units/mg was used to characterize the enzyme.
0
0
pUSI2E (ꢀDꢁ ꢂ ) by sonication. The precipitate
D
D
obtained by ammonium sulfate fractionation (40–50%)
of the cell-free extract was dissolved in Buffer A
containing 2% 1,2-propanediol and loaded onto a
Sepharose CL 6B column (bed volume, 350 ml).
Fractions containing the enzyme with specific activity
of more than 4 units/mg were collected. The enzyme
solution was concentrated with a Centricon (Millipore,
Billerica, MA) and mixed with 4 volumes of 2% 1,2-
propanediol. The solution was loaded onto a DEAE
cellulose (Serva) column (bed volume, 1 ml) previously
equilibrated with 10 mM KPB (pH 8) containing 2% 1,2-
propanediol. The column was washed with 5 volumes of
the same buffer, and then developed with 5 volumes of
Results and Discussion
Limited proteolysis of diol dehydratase with trypsin
The effect of limited proteolysis with trypsin on the
solubility of diol dehydratase was investigated by
incubating the suspension of a precipitant fraction of
ꢀ
E. coli overexpressing diol dehydratase at 15 C for 12 h
with various concentrations of trypsin. After centrifugal
separation of supernatant and precipitant fractions, the
diol dehydratase activity in each fraction was measured
(Fig. 1A). Almost all the activity was recovered in a
soluble fraction when digested with 1.0 mg/ml of trypsin

458
T. TOBIMATSU et al.
without loss of enzyme activity. This indicates that diol
dehydratase retains almost full activity after solubiliza-
tion by partial tryptic digestion. Higher concentrations
of trypsin, however, caused a slight loss (10–20%) of
enzyme activity.
former purifications, as well as the sizes of the subunits
1
8)
reported earlier.
Expression of mutant diol dehydratases with the N-
terminal truncations of the ꢁ and/or ꢂ subunits
As judged from the SDS–PAGE analysis (Fig. 1B),
the ꢁ subunit with Mr of 30,000 and the ꢂ subunit with
Mr of 19,000 decreased, and instead Mr 27,000 and Mr
The above-mentioned tryptic digestion experiment
suggests that the N-terminal region(s) of the ꢁ and ꢂ
subunit(s) lower the solubility of diol dehydratase. To
examine whether the removal of the N-terminal 20
amino acid residues from the ꢁ subunit and/or the N-
terminal 16 residues from the ꢂ subunit is sufficient to
make the enzyme highly soluble, we constructed
expression plasmids for the mutant diol dehydratases
with the N-terminal truncations of the ꢁ and/or ꢂ
1
tide appeared to be then digested to the M 23,000
7,000 polypeptides appeared. The M 27,000 polypep-
r
r
polypeptide with 1.0–100 mg/ml of trypsin, and further
to the Mr 21,000 polypeptide with ꢂ200 mg/ml of
trypsin. To confirm the digestion pattern, the digested
subunit polypeptides were separated by SDS–PAGE,
transferred to a PVDF membrane, and subjected to
Edman sequencing. The N-terminal amino acid se-
0
0
subunits. pUSI2E(ꢀDꢁ ꢂD) and pUSI2E(ꢀDꢁDꢂ )
D
D
were for the expression of mutant enzymes that lack
the N-terminal 20 residues of the ꢁ subunit and the N-
terminal 16 residues of the ꢂ subunit, respectively.
quences of the M 27,000 and 23,000 polypeptides were
r
determined to be QIIED and GSDKP, respectively.
Based on the amino acid sequence of the ꢁ subunit, these
polypeptides were identified as the ꢁ subunits that lack
the N-terminal 9 and 20 amino acid residues, respec-
tively. On the other hand, the Mr 17,000 polypeptide was
then digested to fragments with Mrs less than 15,000.
Similarly, the N-terminal amino acid sequences of the
0
0
pUSI2E(ꢀ ꢁ ꢂ ) was constructed for the expression
D
D
D
of a mutant enzyme in which both the ꢁ and ꢂ subunits
have these N-terminal truncations.
The expression and distribution of the mutant subunits
were analyzed by SDS–PAGE (Fig. 2). The density of
the ꢀ subunit band of all the mutant strains was almost
comparable to that of the wild-type strain. As reported
before, each subunit mutually affects the folding of the
M 17,000 and 15,000 polypeptides determined were
r
MNSLQG and SARVS, indicating that they are the ꢂ
subunits that lack the N-terminal 16 and 36 amino acid
residues, respectively. The reductions in the Mrs of these
ꢂ subunit-derived polypeptides corresponded to the
calculated molecular weights of the N-terminal polypep-
tides removed. But the reductions in the Mrs of the ꢁ
subunit-derived polypeptides were larger than expected.
One possibility is that the enzyme underwent partial
digestion at the C-terminal region as well. Another
possibility is that this discrepancy might be due to the
abnormal shape of the N-terminal region of the ꢁ
subunit. Such an abnormality was observed with the ꢁ
subunit itself—that is, Mr of 30,000 for the ꢁ subunit is
significantly larger than its molecular weight (24,113)
1
3)
others in diol dehydratase. Since the ꢀ subunit is the
product of the gene that is located just downstream of
the tac promoter, an excessive amount of the ꢀ subunit
is not folded correctly and thus accumulates as inclusion
bodies in the precipitant fraction. The bands of truncated
0
0
subunits ꢁ and ꢂ in the mutant strains were not as thick
as the corresponding ꢁ and ꢂ bands in the wild-type
strain. In the supernatant fractions, however, the bands
0
0
of ꢁ and ꢂ in the mutant strains were thicker than the
2
0)
calculated from the deduced amino acid sequence. In
contrast to the ꢁ and ꢂ subunits, the apparent size of the
ꢀ
subunit appeared not to be changed, irrespective of the
amount of trypsin used. The N-terminal amino acid
sequence of the ꢀ subunit of the solubilized enzyme was
MRSKR, which agreed with the deduced N-terminal
amino acid sequence. These results indicate that the ꢀ
subunit did not undergo digestion upon treatment with
trypsin under the conditions.
In summary, the enzyme was completely solubilized
ꢀ
without loss of activity by treatment at 15 C for 12 h
with 1.0 mg/ml of trypsin. The major subunit species of
the solubilized enzyme were Mr 60,000 (the ꢀ subunit
itself), Mr 23,000 (from the ꢁ subunit), and Mr 17,000
Fig. 2. Expression and Distribution of Mutant Diol Dehydratases
with the N-Terminal Truncations of ꢁ and/or ꢂ Subunits in E. coli.
Proteins in cell homogenates (homo) and supernatant (sup) and
precipitant (ppt) fractions of recombinant E. coli cells expressing
each mutant enzyme were subjected to SDS–PAGE on an 11% gel,
followed by protein staining. Experimental details are described in
the text. Molecular weight markers, SDS-7 (Sigma). Positions of the
(
from the ꢂ subunit) polypeptides. Figure 1C summa-
rizes their fragmentation patterns in the N-terminal
regions upon limited proteolysis with trypsin. These
results explain why the enzyme was obtained in a
0
0
ꢀ, ꢁ, ꢁ , ꢂ, and ꢂ subunits are indicated with arrowheads on the
right.
soluble form1 with subunit heterogeneity
)
18)
in the

Highly Soluble Diol Dehydratase
459
0
0
Fig. 3. Comparison of the ꢀꢁ ꢂ Mutant Enzyme with Trypsin-Digested Diol Dehydratase.
PAGE on a 5% gel under non-denaturing conditions, followed by protein staining (A) or activity staining (B). Positions of wild-type and
trypsin-digested diol dehydratases and glycerol dehydratase are indicated with arrowheads on the right. The protein bands that showed enzyme
activity were excised and analyzed by SDS–PAGE on an 11% gel, followed by protein staining (C). Positions of the subunits are indicated with
an arrowhead on the right. Experimental details are described in the text. Molecular weight markers, SDS-7 (Sigma).
untruncated subunit bands of the wild-type strain.
Western blot analysis using anti-diol dehydratase anti-
serum3 showed that the M 23,000 and M 17,000
had been removed, as well as with those of the trypsin-
digested enzyme. It was thus confirmed that the
engineered diol dehydratase expressed in E. coli cells
is catalytically active and consists of the Mr 60,000 (ꢀ),
23,000 (ꢁ : N-terminal truncated ꢁ), and 17,000 (ꢂ : N-
terminal truncated ꢂ) subunits.
2)
r
r
polypeptides in the supernatant fractions of mutant
0
0
0
0
extracts were actually the truncated subunits ꢁ and ꢂ of
diol dehydratase. Their sizes coincided well with those
of the Mr 23,000 and Mr 17,000 polypeptides, respec-
tively, formed by tryptic digestion of the enzyme.
Distribution and purification of diol dehydratases
with the N-terminal truncations of ꢁ and/or ꢂ subunits
When E. coli BL21(DE3) was used as a host strain,
three-fold higher activity was obtained than that with
E. coli JM109 (data not shown). Hence, the former
strain was used for the high level expression of mutant
enzymes. E. coli BL21(DE3) carrying each plasmid was
induced by the addition of 1 mM isopropyl-1-thio-ꢁ-D-
galactopyranoside, and analyzed for the distribution of
enzyme activity after sonic disruption of the cells. As
shown in Table 1, the percentages of diol dehydratase
activity recovered in the soluble fraction were more than
96% with either mutant enzyme, whereas the percentage
was only 2% with the wild-type enzyme. Thus it is
evident that diol dehydratase becomes highly soluble
when either of the N-terminal 20 residues of the ꢁ
subunit or 16 residues of the ꢂ subunit are removed. This
is consistent with the result of limited proteolysis with
trypsin. Thus it can be concluded that both of the N-
terminal regions of the ꢁ and ꢂ subunits are necessary
for precipitation or aggregation of this enzyme. It should
also be noted that the total activity of mutant enzymes
was lower than that of the wild-type enzyme. Since the
enzyme with N-terminal truncations of the ꢁ and ꢂ
subunits was almost fully active (Fig. 1A), perhaps the
mutant enzymes expressed in soluble forms are more
susceptible to proteolysis than the wild-type enzyme.
0
Excessive amounts of the ꢀ and ꢁ polypeptides ex-
0
pressed were found in the precipitant fraction, but the ꢂ
polypeptide was not. Smaller-sized bands that reacted
with anti-diol dehydratase aniserum were also observed
in the supernatant fractions upon Western blot analysis.
0
0
To compare the ꢀꢁ ꢂ mutant enzyme with the wild-
type (ꢀꢁꢂ) diol dehydratase and the enzyme treated with
ꢀ
1
E. coli cells carrying pUSI2E(ꢀ ꢁ ꢂ ) was subjected
.0 mg/ml of trypsin at 15 C for 12 h, cell-free extract of
0
0
D
D
D
to PAGE under non-denaturing conditions, followed by
2
0)
protein staining and activity staining
(Fig. 3). As
shown in Fig. 3A and B, only one band showing diol
dehydratase activity was observed with the mutant
extract. This band co-migrated with the trypsin-treated
enzyme and migrated slightly faster than the wild-type
enzyme and slower than glycerol dehydratase. When the
active band of mutant extract was cut out from the gel
and subjected to SDS–PAGE analysis, all three polypep-
tides co-migrated with the M 60,000, 23,000, and
r
1
(
7,000 polypeptides of the trypsin-treated enzyme
Fig. 3C). The N-terminal amino acid sequences of the
Mr 23,000 and 17,000 polypeptides in the mutant
enzyme were determined to be GSDKP and MNSLQG,
respectively. These coincided with the predicted N-
0
0
terminal amino acid sequences of the ꢁ and ꢂ subunits,
0
except that the N-terminal Met residue in the ꢁ subunit

460
T. TOBIMATSU et al.
Table 1. Distribution of Diol Dehydratase Activity of Mutant
Enzymes Expressed in E. coli Cells
Diol dehydratase
activity (units)
Percentage of activity
in supernatant (%)
Mutant
Supernatant Precipitant
ꢀ
ꢀ
ꢀ
ꢀ
ꢁꢂ (wild-type)
0
43
493
283
181
2,800
2
96
97
99
ꢁ ꢂ
0
23
9
ꢁꢂ
0
0
ꢁ ꢂ
1
1
0
Recombinant E. coli BL21 (DE3) (9:2{9:8 ꢁ 10 cells) harboring each
expression plasmid was washed, re-suspended in 2.6 ml of 50 mM potassium
phosphate buffer (pH 8.0) containing 2% 1,2-propanediol, and disrupted by
sonication. The resulting cell homogenates were centrifuged at 17;500 ꢁ g
for 30 min to separate supernatant and precipitant fractions. The precipitate
was washed with 2.6 ml of the same buffer and re-suspended in 10 mM
potassium phosphate buffer (pH 8.0) containing 2% 1,2-propanediol and 1%
Brij 35 to the same volume. Diol dehydratase activity was determined as
described in the text after appropriate dilution.
Fig. 4. Geometry of the N-Terminal Regions of ꢁ and ꢂ Subunits in
the Crystal Structure of Diol Dehydratase.
(A) Wireframe model of overall structure. Subunit ꢀ, light-grey;
, dark grey; ꢂ, black. (B) Cꢀs of the N-terminal residues of ꢁ and ꢂ
ꢁ
subunits that are located on an electron density map. Gꢁ46, dark-
grey ball; Sꢂ37, black ball.
This speculation is consistent with the fact that three-
fold higher activity was obtained when the wild-type
enzyme was expressed with E. coli BL21(DE3) as a
host, because this strain lacks lon and ompT proteas-
hydropathy profiles. Hence this point appears crucial for
our understanding of the way of protein assembles
around diol dehydratase in the polyhedral bodies in
bacterial cells. We have reported the X-ray structures of
3
3,34)
es.
Partially purified ꢀꢁ ꢂ enzyme was obtained from
the cell-free extract of E. coli cells carrying
0
0
0
0
16)
pUSI2E(ꢀDꢁ ꢂ ) by ammonium sulfate fractionation,
diol dehydratase in complexes with cyanocobalamin
D
D
3
5)
Sepharose CL-6B column chromatography, and DEAE-
cellulose column chromatography. The specific activity
and adeninylpentylcobalamin. The structures of the N-
terminal regions of the ꢁ and ꢂ subunits remain unclear,
because this enzyme becomes crystallized only after
cleavage of the N-terminal 45 residues of the ꢁ subunit
and the 36 residues of the ꢂ subunit. These regions are
considered to extend to the solvent and are believed to
16)
be inhibitory for compact filling up in crystallization.
0
0
of this preparation of ꢀꢁ ꢂ was 47 units/mg. Densito-
metric analysis after PAGE showed that the purity of the
0
0
16)
ꢀ
specific activity of the homogeneous preparation of
ꢁ ꢂ band was 55%. These results suggest that the
0
0
ꢀ
that of the enzyme purified from K. oxytoca in a soluble
ꢁ ꢂ enzyme is 85 units/mg, which is comparable to
When the Cꢀ atoms of the N-terminal amino acids
(Gꢁ46 and Sꢂ37) of the ꢁ and ꢂ subunits located on an
electron density map are indicated by grey and black
balls, respectively, in the X-ray structure of diol
dehydratase (Fig. 4A), it is evident that the N-terminal
regions of both subunits are located roughly at the tops
of tetrahedral structure. That is, the two tops are
occupied by the N-terminals of the two ꢁ subunits and
the other two by those of the two ꢂ subunits (Fig. 4B).
Such steric positions might be important for the
formation of large protein assemblies by intermolecular
interactions within the enzymes and/or of the enzyme
with other protein components in the polyhedral bodies.
1
8)
form (91.5 units/mg), and of the purified recombinant
wild-type enzyme (106 units/mg).13) The Km values for
0
0
a substrate (DL-1,2-propanediol) obtained with ꢀꢁ ꢂ and
wild-type enzymes were 0.11 and 0.10 mM, respectively,
both of which values are reasonably consistent with the
value obtained with the purified soluble enzyme from
2
)
K. oxytoca (0.18 mM). From the data described above,
it was concluded that the mutant enzyme with the
N-terminal truncations of both ꢁ and ꢂ subunits is
essentially indistinguishable in catalytic properties from
recombinant wild-type enzyme or the enzyme purified
from K. oxytoca in a soluble form. However, the K
0
m
0
value for AdoCbl obtained with the ꢀꢁ ꢂ enzyme was
.28 mM, smaller than that obtained with wild-type
0
Acknowledgments
enzyme (0.83 mM). The Km value for the coenzyme with
the enzyme purified from K. oxytoca in a soluble form is
This work was supported in part by Grants-in-Aid for
Scientific Research ((B) No. 13480195 and (C)
No. 14580627, and Priority Areas No. 753) from the
Ministry of Education, Culture, Sports, Science and
Technology of Japan. We thank Professor Hidenori
Yamada for N-terminal sequencing and Ms. Yukiko
Kurimoto for her assistance in manuscript preparation.
28)
0
.80 mM. The reason for this difference is not clear.
At present, it is not clear why both of the N-terminal
regions of the ꢁ and ꢂ subunits are required for making
this enzyme almost insoluble. This is an enigma because
no prominent hydrophobic peaks were found in these N-
terminal regions of either subunits as judged from their

Highly Soluble Diol Dehydratase
461
References
Allosteric interactions in glycerol dehydratase. Purifica-
tion of enzyme and effects of positive and negative
cooperativity for glycerol. Arch. Biochem. Biophys., 162,
321–330 (1974).
1)
2)
3)
Lee, H. A., Jr., and Abeles, R. H., Purification and
properties of dioldehydrase, an enzyme requiring a
cobamide coenzyme. J. Biol. Chem., 238, 2367–2373
15) Yamanishi, M., Yunoki, M., Tobimatsu, T., Sato, H.,
Matsui, J., Dokiya, A., Iuchi, Y., Oe, K., Suto, K.,
Shibata, N., Morimoto, Y., Yasuoka, N., and Toraya, T.,
The crystal structure of coenzyme B12-dependent glyc-
erol dehydratase in complex with cobalamin and
propane-1,2-diol. Eur. J. Biochem., 269, 4484–4494
(2002).
16) Shibata, N., Masuda, J., Tobimatsu, T., Toraya, T., Suto,
K., Morimoto, Y., and Yasuoka, N., A new mode of B12
binding and the direct participation of a potassium ion in
enzyme catalysis: X-ray structure of diol dehydratase.
Structure Fold Des., 7, 997–1008 (1999).
(
1963).
Toraya, T., Shirakashi, T., Kosuga, T., and Fukui, S.,
Coenzyme B12-dependent diol dehydrase: Glycerol as
both a good substrate and a potent inactivator. Biochem.
Biophys. Res. Commun., 69, 475–480 (1976).
Toraya, T., Honda, S., Kuno, S., and Fukui, S.,
Coenzyme B12-dependent diol dehydratase: Regulation
of apoenzyme synthesis in Klebsiella pneumoniae
(Aerobacter aerogenes) ATCC8724. J. Bacteriol., 135,
726–729 (1978).
4)
Toraya, T., Kuno, S., and Fukui, S., Distribution of
coenzyme B12-dependent diol dehydratase and glycerol
dehydratase in selected genera of Enterobacteriaceae
and Propionibacteriaceae. J. Bacteriol., 141, 1439–1442
17) Toraya, T., Radical catalysis in coenzyme B12-dependent
isomerization (eliminating) reactions. Chem. Rev., 103,
2095–2127 (2003).
(
1980).
18) Poznanskaja, A. A., Tanizawa, K., Soda, K., Toraya, T.,
and Fukui, S., Coenzyme B12-dependent diol dehydrase:
Purification, subunit heterogeneity, and reversible asso-
ciation. Arch. Biochem. Biophys., 194, 379–386 (1979).
19) Tanizawa, K., Nakajima, N., Toraya, T., Tanaka, H., and
Soda, K., Re-investigation of the protein structure of
coenzyme B12-dependent diol dehydrase. Z. Natur-
forsch., 42c, 353–359 (1987).
20) Tobimatsu, T., Hara, T., Sakaguchi, M., Kishimoto, Y.,
Wada, Y., Isoda, M., Sakai, T., and Toraya, T., Mo-
lecular cloning, sequencing and expression of the genes
encoding adenosylcobalamin-dependent diol dehydrase
of Klebsiella oxytoca. J. Biol. Chem., 270, 7142–7148
(1995).
5)
6)
7)
8)
9)
Toraya, T., Diol dehydratase and glycerol dehydratase.
In ‘‘Chemistry and Biochemistry of B12’’, ed. Banerjee,
R., John Wiley and Sons, New York, pp. 783–809
(
1999).
Toraya, T., Honda, S., and Fukui, S., Fermentation of
,2-propanediol with 1,2-ethanediol by some genera of
1
Enterobacteriaceae, involving coenzyme B12-dependent
diol dehydratase. J. Bacteriol., 139, 39–47 (1979).
Forage, R. G., and Foster, M. A., Resolution of the
coenzyme B12-dependent dehydratases of Klebsiella sp.
and Citrobacter freundii. Biochem. Biophys. Acta, 569,
2
49–258 (1979).
Forage, R. G., and Foster, M. A., Glycerol fermentation
in Klebsiella pneumoniae: Functions of the coenzyme
B12-dependent glycerol and diol dehydratases. J. Bac-
teriol., 149, 413–419 (1982).
Bobik, T. A., Havemann, G. D., Busch, R. J., Williams,
D. S., and Aldrich, H. C., The propanediol utilization
21) Sauvageot, N., Pichereau, V., Louarme, L., Hartke, A.,
Auffray, Y., and Laplace, J. M., Purification, character-
ization and subunits identification of the diol dehydratase
of Lactobacillus collinoides. Eur. J. Biochem., 269,
5731–5737 (2002).
(
pdu) operon of Salmonella enterica serovar Typhimu-
22) Leal, N. A., Havemann, G. D., and Bobik, T. A., PduP is
a coenzyme-a-acylating propionaldehyde dehydrogenase
associated with the polyhedral bodies involved in B12-
dependent 1,2-propanediol degradation by Salmonella
enterica serovar Typhimurium LT2. Arch. Microbiol.,
180, 353–361 (2003).
23) Tobimatsu, T., Azuma, M., Matsubara, H., Takatori, H.,
Niida, T., Nishimoto, K., Satoh, H., Hayashi, R., and
Toraya, T., Cloning, sequencing and high-level expres-
sion of the genes encoding adenosylcobalamin-depend-
ent glycerol dehydrase of Klebsiella pneumoniae. J.
Biol. Chem., 271, 22352–22357 (1996).
24) Seyfried, M., Daniel, R., and Gottschalk, G., Cloning,
sequencing, and overexpression of the genes encoding
coenzyme B12-dependent glycerol dehydratase of Cit-
robacter freundii. J. Bacteriol., 178, 5793–5796 (1996).
25) Bobik, T. A., Xu, Y., Jeter, R. M., Otto, K. E., and Roth,
J. R., Propanediol utilization genes (pdu) of Salmonella
typhimurium: Three genes for the propanediol dehydra-
tase. J. Bacteriol., 179, 6633–6639 (1997).
26) Tobimatsu, T., Azuma, M., Hayashi, S., Nishimoto, K.,
and Toraya, T., Molecular cloning, sequencing and
characterization of the genes for adenosylcobalamin-
dependent diol dehydratase of Klebsiella pneumoniae.
Biosci. Biotechnol. Biochem., 62, 1774–1777 (1998).
rium LT2 includes genes necessary for formation of
polyhedral organelles involved in coenzyme B12-
dependent 1,2-propanediol degradation. J. Bacteriol.,
1
81, 5967–5975 (1999).
1
1
0) Havemann, G. D., and Bobik, T. A., Protein content of
polyhedral organelles involved in coenzyme B12-de-
pendent degradation of 1,2-propanediol in Salmonella
enterica serovar Typhimurium LT2. J. Bacteriol., 185,
5
086–5095 (2003).
1) Havemann, G. D., Sampson, E. M., and Bobik, T. A.,
PduA is a shell protein of polyhedral organelles involved
in coenzyme B12-dependent degradation of 1,2-propane-
diol in Salmonella enterica serovar Typhimurium LT2.
J. Bacteriol., 184, 1253–1261 (2002).
2) McGee, D. E., and Richards, J. H., Purification and
subunit characterization of propanediol dehydratase, a
membrane-associated enzyme. Biochemistry, 20, 4293–
1
4
298 (1981).
1
3) Tobimatsu, T., Sakai, T., Hashida, Y., Mizoguchi, N.,
Miyoshi, S., and Toraya, T., Heterologous expression,
purification, and properties of diol dehydratase, an
adenosylcobalamin-dependent enzyme of Klebsiella
oxytoca. Arch. Biochem. Biophys., 347, 132–140 (1997).
4) Stroinski, A., Pawelkiewicz, J., and Johnson, B. C.,
1

462
T. TOBIMATSU et al.
2
7) Macis, L., Daniel, R., and Gottschalk, G., Properties and
the assembly of the head of bacteriophage T4. Nature,
227, 680–685 (1970).
32) Toraya, T., and Fukui, S., Immunochemical evidence for
the difference bteween coenzyme-B12-dependent diol
dehydratase and glycerol dehydratase. Eur. J. Biochem.,
76, 285–289 (1977).
33) Gottesman, S., Genetics of proteolysis in Escherichia
coli. Ann. Rev. Genet., 23, 163–198 (1989).
34) Gottesman, S., Proteases and their targets in Escherichia
coli. Ann. Rev. Genet., 30, 465–506 (1989).
35) Masuda, J., Shibata, N., Morimoto, Y., Toraya, T., and
Yasuoka, N., How a protein generates a catalytic radical
from coenzyme B12: X-ray structure of a diol-dehydra-
tase-adeninylpentylcobalamin complex. Structure, 8,
775–788 (2000).
sequence of the coenzyme B12-dependent glycerol
dehydratase of Clostridium pasteurianum. FEMS Micro-
biol. Lett., 164, 21–28 (1998).
2
8) Toraya, T., Ushio, K., Fukui, S., and Hogenkamp, P. C.,
Studies on the mechanism of the adenosylcobalamin-
dependent diol dehydrase reaction by the use of analogs
of the coenzyme. J. Biol. Chem., 252, 963–970 (1977).
9) Lowry, O. H., Rosebrough, N. J., Farr, A. L., and
Randall, R. J., Protein measurement with the Folin
phenol reagent. J. Biol. Chem., 193, 265–275 (1951).
0) Davis, B. J., Disc electrophoresis. II. Method and
application to human serum proteins. Ann. N.Y. Acad.
Sci., 121, 404–427 (1964).
2
3
3
1) Laemmli, U. K., Cleavage of structural proteins during