Stereoselective oxidation of protected inositol derivatives catalyzed by inositol dehydrogenase from Bacillus subtilis

Article abstract of DOI:10.1039/b417757f

Inositol dehydrogenase (EC 1.1.1.18) from Bacillus subtilis is shown to have a nonpolar cavity adjacent to the active site, allowing racemic protected inositol derivatives such as 4-O-benzyl-myo-inositol to be recognized with very high apparent stereoselectivity.

Full text of DOI:10.1039/b417757f

View Article Online / Journal Homepage / Table of Contents for this issue
C O M M U N I C A T I O N
Stereoselective oxidation of protected inositol derivatives catalyzed
by inositol dehydrogenase from Bacillus subtilis†
Richard Daniellou,a Christopher P. Phenix,a Pui Hang Tam,b Michael C. Lalibertea and
David R. J. Palmer*a,b
a Department of Chemistry, University of Saskatchewan, 110 Science Place, Saskatoon,
Saskatchewan, Canada. E-mail: palmer@sask.usask.ca; Fax: 306-966-4730;
Tel: 306-966-4662
b Department of Biochemistry, University of Saskatchewan, 110 Science Place, Saskatoon,
Saskatchewan, Canada
Received 23rd November 2004, Accepted 26th November 2004
First published as an Advance Article on the web 22nd December 2004
Table 1 Kinetic constants of IDH-catalyzed oxidationsa
Inositol dehydrogenase (EC 1.1.1.18) from Bacillus subtilis
is shown to have a nonpolar cavity adjacent to the active
site, allowing racemic protected inositol derivatives such
as 4-O-benzyl-myo-inositol to be recognized with very high
apparent stereoselectivity.
Compound
K
App/mM
V
App/lmol min−1 mg−1
(V/K)rel/%
myo-Inositol
18
8
7
4
3
44
13
—
57
—
2
1
1
1
1
5
2
8
1
100
26
30
51
58
8
12
—
8
1
0.8 0.1
0.8 0.1
0.9 0.1
0.8 0.1
1.6 0.1
0.7 0.1
Trace activity detected
2.0 0.5
Trace activity detected
2
myo-Inositol dehydrogenase (IDH), EC 1.1.1.18, is a bacterial
enzyme that catalyzes the selective oxidation of the (axial)
hydroxyl group attached to carbon-2 of myo-inositol to a ketone
with concomitant conversion of NAD+ to NADH, as shown in
Fig. 1. The reaction allows the organism to grow on inositol,
which is abundantly available in soil, as its sole carbon source.
The gene encoding this enzyme has been identified in many
environmental bacteria, but only in the case of IDH from
Bacillus subtilis has the catalyzed reaction been studied in detail.1
Among the findings of that work, it was reported that IDH
catalyzed the oxidation of a-D-glucopyranose and D-xylose to
the corresponding 1,5-lactones, but that scyllo-inositol and b-D-
glucopyranose (in which all the substituents are equatorial) were
not substrates. This indicates that an axial hydroxyl group is a
required feature of the substrate, but that other aspects of the
structure may vary. Many other common monosaccharides and
alditols were also excluded as substrates, suggesting that the IDH
active site does discriminate against most variations in structure.
For example, galactose and mannose, containing an additional
axial hydroxyl group, were not oxidized, nor was glucose-6-
phosphate a substrate. Our laboratory has chosen to probe the
active site of B. subtilis IDH further in order to learn more
about its catalytic properties, and to discover if such an enzyme
could be used as an efficient means of separating enantiomers of
myo-inositol derivatives, providing a source of optically active
cyclitols. Inositol derivatives are ubiquitous natural products,
and the enzymes that manipulate them are therapeutic targets
for the treatment of several illnesses.2 The efficient synthesis of
optically active inositols is by no means routine.3
3
4
5
6
7
Melibiose
Gentiobiose
6
—
a Conditions: 100 mM Tris–HCl, pH 9.0, 25 ◦C. [NAD+] = 0.5 mM.
sequence resulted in an amplified sequence which differed by
one nucleotide from the coding region reported. This was
true of two separate PCR experiments, performed on different
template preparations, by different individuals, suggesting that
the sequence of our template is accurately reflected in the clone.
The difference, a thymine at nucleotide 434 rather than the
reported adenine, results in an amino acid change at position
145, from asparagine to isoleucine. Inspection of apparently
homologous sequences, such as those found in the cluster of
orthologous groups COG06735 indicates that the amino acid
corresponding to position 145 of the iolG gene product is always
a branched, hydrophobic residue such as leucine, isoleucine
or valine. We therefore believe this sequence represents that
of wild-type IDH, and it has been deposited in GenBank
(accession no. AY676876). The gene was ligated into pET-28b
and expressed in E. coli BL21(DE3). Purification of IDH bearing
an N-terminal histidine tag by Ni2+-affinity using a gradient
of increasing imidazole concentration resulted in protein that
was determined to be homogeneous by SDS-PAGE analysis.
Electrospray mass spectrometry indicated a molecular mass
of 40382.0; the predicted exact mass based on our sequence
is 40382.76. The resulting IDH, assayed by observing the
appearance of NADH at 340 nm, was active without removing
the His tag. The kinetic constants for the oxidation of myo-
inositol in the presence of recombinant enzyme, shown in
Table 1, are in good agreement with those reported previously
for the native enzyme. Ramaley et al.1 showed the IDH-catalyzed
reaction to follow an ordered sequential Bi Bi mechanism, with
KNmAD = 0.23mM andanapparentMichaelis constant for inositol
KApp = 18 mM.‡
Fig.
1 Oxidation of myo-inositol to scyllo-inosose by inositol
dehydrogenase.
The sequence of the gene encoding IDH, iolG, has been
reported previously.4 PCR-amplification from B. subtilis 168
(trpC2) genomic DNA using primers based on the reported
The previous report that IDH can oxidize D-glucose1 led
us to propose that IDH might tolerate small substituents on
inositol at the position analogous to C-6 of D-glucose. For
example, 1L-4-O-methyl-myo-inositol is approximately isosteric
with a-D-glucopyranose, as indicated in Fig. 2. We confirmed
† Electronic Supplementary Information (ESI) available: experimen-
tal details: molecular biology, enzyme kinetics, HPLC conditions,
and synthesis and characterization of all synthetic substrates. See
http://www.rsc.org/suppdata/ob/b4/b417757f/
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 0 1 – 4 0 3
4 0 1
©

View Article Online
the previous results with D-glucose and D-xylose, and also
tested L-glucose and L-xylose, but could observe no activity,
suggesting a high degree of stereoselectivity, i.e. one ‘side’ of
the active site can tolerate variations in structure, but the other
‘side’ recognizes equatorial hydroxyl groups. The 4/6-position
is conveniently alkylated, sulfonylated or phosphorylated using
the 1,3,5-monoorthoformate route pioneered by Kishi.6 Briefly,
myo-inositol monoorthoformate is formed by an acid-catalyzed
reaction of triethyl orthoformate with inositol. This product is
alkylated or acylated with the appropriate electrophile, and the
protecting group(s) subsequently removed. While this process
results in good yields and can be run on a large scale, it does result
in racemic products. We have used these methods to synthesize
an array of substituted inositols, shown in Fig. 2.†
3 by published methods resulted in an overall yield of less than
2%.
The presence of a relatively nonpolar cavity adjacent to the
active site of IDH suggests that the enzyme may have a role
in the oxidation of other natural products in vivo. Alternatively
this may indicate that IDH has evolved from an enzyme that
recognizes more complex structures, and has retained some
of the characteristics of the progenitor. Speculation as to the
evolutionary origin of this active site led us to investigate
disaccharides as possible substrates. We chose gentiobiose [6-O-
(b-D-glucopyranosyl)-D-glucopyranose] and melibiose [6-O-(a-
D-galactopyranosyl)-D-glucopyranose] because they are roughly
isosteric with 1L-4-O-substituted inositols, and were readily
available. Neither ‘gentiobiose dehydrogenase’ nor ‘melibiose
dehydrogenase’ are reported in the literature, however melibiose
transporter and hydrolase genes are apparently present in B.
4b
subtilis. Both compounds were oxidized in the presence of
IDH, although the activity observed with gentiobiose was very
low. The apparent Michaelis constant of the melibiose reaction
was within experimental error of that previously reported for
a-D-glucopyranose. Initial experiments with isomaltose [[6-O-
(a-D-glucopyranosyl)-D-glucopyranose] show activity similar to
that observed with melibiose suggesting the a(1,6) linkage results
in a better fit in the IDH active site.
In summary, we have probed the active site of IDH, revealing
a hydrophobic cavity that allows the recognition and oxidation
of 1L-4-O-substituted myo-inositols bearing large hydrophobic
groups. This property allows resolution of racemic mixtures
of inositol derivatives, including those blocked with cleavable
protecting groups that may be used in the asymmetric synthesis
of more complex inositol derivatives. The relatively low substrate
selectivity and high stereoselectivity of IDH suggest it is well-
suited for development as a tool for synthetic applications. The
oxidation of disaccharides by IDH suggests the enzyme may
fulfil such a role in vivo, or it is homologous with an as-yet-
unidentified disaccharide dehydrogenase. Understanding the
relationship between enzyme structure and substrate recognition
will help in assigning function to putative oxidoreductases
whose metabolic role cannot be predicted based on our current
knowledge.
Fig.
2 Racemic 4-O-substituted myo-inositol derivatives (the
1L-enantiomer is drawn) and D-glucose derivatives oxidized by IDH.
4-O-Methyl-myo-inositol was a substrate of IDH, as we
had predicted. When we repeated our experiments with larger
substituents, we were surprised to find that solubility was the
primary limitation to suitability as a substrate. Even a very
bulky substituent, such as the (1S)-10-camphorsulfonyl group,
was tolerated by the enzyme. All showed normal hyperbolic
dependence of rate on substrate concentration. Disappointingly,
when inositol 4-phosphate (7) was the substrate only trace
activity could be detected, perhaps unsurprising given that no
activity could be observed with glucose-6-phosphate. However,
we were particularly encouraged that substituents commonly
used as protecting groups in multi-step syntheses, such as the
allyl and benzyl groups, could be incorporated while maintaining
enzymatic activity. The data in Table 1 show that a more
favourable Michaelis constant is observed for several substrates
relative to the natural substrate, indicating an enhanced enzyme–
substrate interaction in the ground state. The results using
compounds 4 and 5 show that while a negatively charged
substituent can be tolerated in the active site (when distant from
the inositol moiety), the Michaelis constant increases by an order
of magnitude.
The equilibrium of the IDH reaction favors the inositol rather
than the inosose form. To drive the reaction to completion, the
reaction was performed in the presence of lactate dehydrogenase
and an excess of pyruvate. This allowed a catalytic amount
of cofactor to be used, since it is continually recycled. HPLC
analysis of the reaction mixture using a reverse-phase column
indicated that the amount of substrate had decreased by 50%.
Using a cyclodextrin-based chiral HPLC column, we found that
racemic 3 could be resolved. After 72 h of enzymatic reaction,
one of the peaks was completely consumed. The remaining
substrate was isolated, and a specific rotation of +6 could
be measured. This is the value reported previously for 1D-4-
O-benzyl-myo-inositol.7 When this isomer was re-subjected to
the IDH assay, no activity could be observed. The optically
active 1L-3 synthesized in our laboratory7,8 co-elutes with the
reactive isomer when analyzed by HPLC, and is a substrate
for IDH. Note that we synthesized racemic 3 in 73% yield
starting from myo-inositol, and could obtain the 1D-isomer
quantitatively using IDH. Chemical synthesis of optically active
This work was supported by a Research Innovation Award
(R10532) from Research Corporation, and a Discovery Grant
from the Natural Sciences and Engineering Research Council of
Canada.
Notes and references
‡ Determined at [NAD+] = 0.5 mM. Reactions were fit to the equation
m = VApp[S]/(KApp + [S]) where S is the varied substrate and VApp
is the maximal velocity at the concentration of the fixed substrate.
In an ordered sequential Bi Bi mechanism, KApp = Kminositol(KNi AD
+ [NAD+])/(KmNAD + [NAD+]), where KNi AD is the product inhibi-
tion constant for the reverse process (binding of NAD+ to free
enzyme).9 Using our data, we calculate KiNAD = 5 mM, and Kminositol
=
1 mM.
1 R. Ramaley, Y. Fujita and E. Freese, J. Biol. Chem., 1979, 254, 7684–
7690.
2 (a) I. Schulz and E. Krause, Curr. Mol. Med., 2004, 4, 313–322;
(b) R. P. Ebstein, B. Lerer, E. R. Bennett, B. Shapira, S. Kindler,
Z. Shemesh and N. Gerstenhaber, Psychiatry Res., 1988, 24, 45–52;
(c) A. Crossman Jr., M. J. Paterson, M. A. J. Ferguson, T. K. Smith
and J. S. Brimacombe, Carbohydr. Res., 2002, 337, 2049–2059 and
references therein.
3 R. J. Kubiak and K. S. Bruzik, J. Org. Chem., 2003, 68, 960–
968.
4 (a) Y. Fujita, K. Shindo, Y. Miwa and K. Yoshida, Gene, 1991, 108,
121–125; (b) F. Kunst, N. Ogasawara, I. Moszer, A. M. Albertini, G.
Alloni, V. Azevedo, M. G. Bertero, P. Bessieres, A. Bolotin and S.
Borchert, et al. Nature, 1997, 390, 249–256.
5 R. L. Tatusov, N. D. Fedorova, J. D. Jackson, A. R. Jacobs, B. Kiryutin,
E. V. Koonin, D. M. Krylov, R. Mazumder, S. L. Mekhedov, A. N.
Nikolskaya, B. S. Rao, S. Smirnov, A. V. Sverdlov, S. Vasudevan,
4 0 2
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 0 1 – 4 0 3

View Article Online
Y. I. Wolf, J. J. Yin and D. A. Natale, BMC Bioinformatics, 2003, 4,
41.
6 (a) H. W. Lee and Y. Kishi, J. Org. Chem., 1985, 50, 4402–4404;
(b) D. C. Billington, R. Baker, J. J. Kulagowski, I. M. Mawer, J. P.
Vacca, S. J. deSolms and J. Huff, J. Chem. Soc., Perkin Trans. 1, 1989,
1423–1429.
7 S. M. Kheronsky and Y.-T. Chang, Carbohydr. Res., 2002, 337, 75–78.
8 (a) S.-K. Chung, B.-G. Shin, Y.-T. Chang, B.-C. Suh and K.-T. Kim,
Bioorg. Med. Chem. Lett., 1998, 8, 659–662; (b) S.-K. Chung and Y.
Ryu, Carbohydr. Res., 1994, 258, 145–167.
9 W. W. Cleland, Adv. Enzymol. Relat. Areas Mol. Biol., 1977, 45, 273–
387.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 0 1 – 4 0 3
4 0 3