Mechanistic insights into the 1,3-xylanases: Useful enzymes for manipulation of algal biomass

Article abstract of DOI:10.1021/ja211836t

Xylanases capable of degrading the crystalline microfibrils of 1,3-xylan that reinforce the cell walls of some red and siphonous green algae have not been well studied, yet they could prove to be of great utility in algaculture for the production of food and renewable chemical feedstocks. To gain a better mechanistic understanding of these enzymes, a suite of reagents was synthesized and evaluated as substrates and inhibitors of an endo-1,3-xylanase. With these reagents, a retaining mechanism was confirmed for the xylanase, its catalytic nucleophile identified, and the existence of -3 to +2 substrate-binding subsites demonstrated. Protein crystal X-ray diffraction methods provided a high resolution structure of a trapped covalent glycosyl-enzyme intermediate, indicating that the 1,3-xylanases likely utilize the ¹S₃ → ⁴H₃ → ⁴C₁ conformational itinerary to effect catalysis.

Full text of DOI:10.1021/ja211836t

Article
pubs.acs.org/JACS
Mechanistic Insights into the 1,3-Xylanases: Useful Enzymes for
Manipulation of Algal Biomass
Ethan D. Goddard-Borger,† Keishi Sakaguchi,‡ Stephan Reitinger,†,∥ Nobuhisa Watanabe,§ Makoto Ito,‡
and Stephen G. Withers*,†
†Department of Chemistry, University of British Columbia, 2036 Main Mall, Vancouver, British Columbia V6T 1Z1, Canada
‡Department of Bioscience and Biotechnology, Graduate School of Bioresource and Bioenvironmental Sciences, Kyushu University,
Fukuoka 812-8581, Japan
§Department of Biotechnology, Graduate School of Engineering, Nagoya University, Nagoya 464-8603, Japan
S
* Supporting Information
ABSTRACT: Xylanases capable of degrading the crystalline microfibrils of 1,3-xylan that
reinforce the cell walls of some red and siphonous green algae have not been well studied,
yet they could prove to be of great utility in algaculture for the production of food and
renewable chemical feedstocks. To gain a better mechanistic understanding of these
enzymes, a suite of reagents was synthesized and evaluated as substrates and inhibitors of
an endo-1,3-xylanase. With these reagents, a retaining mechanism was confirmed for the
xylanase, its catalytic nucleophile identified, and the existence of −3 to +2 substrate-
binding subsites demonstrated. Protein crystal X-ray diffraction methods provided a high
resolution structure of a trapped covalent glycosyl−enzyme intermediate, indicating that
1
4
4
the 1,3-xylanases likely utilize the S3 → H3 → C1 conformational itinerary to effect catalysis.
note: Caulerpa lentillifera (cultivated for human consumption)7
INTRODUCTION
■
and Caulerpa taxifolia (a highly invasive pest).8
Collectively, the algae produce a considerable number of
unusual polysaccharides that are not found in higher plants.
This is perhaps a consequence of their broad taxonomic
diversity, simple body plan, and the unique challenges and
opportunities presented by their aquatic environments. The
composition of algal cell walls can vary radically between phyla,
in contrast to the higher plants, which for the most part utilize a
common system of cell wall polysaccharides and are invariably
reliant on cellulose for cell wall strength. A number of algae
produce cell walls that are completely devoid of cellulose, a
curious observation first noted over a century ago1 and
unequivocally confirmed some six decades later.2 Many of these
cellulose-free algae rely on other polysaccharides to endow their
cell walls with the mechanical properties necessary to survive.
One such polymer is 1,3-xylan, a homopolysaccharide of β-1,3-
linked D-xylopyranose units, which is found in some red and
siphonous green algae. Within these organisms, molecules of
1,3-xylan bundle into right-handed triple helices, which pack
hexagonally to form crystalline microfibrils.3,4 These fibrils
envelope the algal cell in transverse, longitudinal, and/or
random orientations, just as cellulose microfibrils do in higher
plants.5
Understanding how Nature goes about depolymerizing 1,3-
xylan goes beyond a general interest in the ecology of these
algae. Enzymes capable of degrading 1,3-xylan are currently
used to degrade the cell wall of living algal cells in order to
generate protoplasts, which enables cell fusion and genetic
manipulation to create new cultivars with characteristics that
are desirable for algaculture.9 It has also been suggested that
1,3-xylan-containing algal biomass could serve as a feedstock for
the production of renewable chemical commodities.10 If such
technology were developed, the 1,3-xylanases (EC 3.2.1.32)
would no doubt play a central role in the conversion of that
biomass into useful products.
Enzymes capable of degrading 1,3-xylan were first alluded to
by Iriki and co-workers in the 1960s, when it was noted that
several fungal species appeared to secrete enzymes capable of
hydrolyzing the 1,3-xylan that they had only recently isolated
from siphonous green algae.11 Some time later, several fungal
proteins with 1,3-xylanase activity were isolated from the
culture medium of an Aspergillus terreus strain that originated
from soil collected near a seaweed-processing factory.12 Since
then a number of secreted 1,3-xylanases induced by the
presence of 1,3-xylan have been identified in bacteria cultured
from marine sediments.13−15 The genes encoding several of
these bacterial 1,3-xylanases (from various Alcaligenes, Vibrio,
and Pseudomonas species) have since been cloned and
A number of algae possessing 1,3-xylan-based cell walls are of
significant relevance to human activities. Many red algae
cultivated for human consumption, such as those harvested as
Japanese “nori”, Korean “gim”, Anglo-American “dulse”, and
“laver”, are Porphyra species that possess cell walls reinforced by
microfibrils of 1,3-xylan.6 Of the many siphonous green algae
that produce 1,3-xylan, two Caulerpa species are of particular
Received: December 19, 2011
Published: January 30, 2012
© 2012 American Chemical Society
3895
dx.doi.org/10.1021/ja211836t | J. Am. Chem. Soc. 2012, 134, 3895−3902

Journal of the American Chemical Society
Article
recombinantly expressed in E. coli, and to date, they are the
only functionally characterized 1,3-xylanase genes.16−18
Bacterial 1,3-xylanases, like many cellulases, hemicellulases,
and chitinases, have a modular architecture: they consist of a
catalytic domain, glycine rich linker(s), and one or more
carbohydrate binding modules (CBMs).17,18 The unique CBMs
of the 1,3-xylanases belong to CBM family 31 of the
Carbohydrate Active enZyme (CAZy) database (http://www.
cazy.org),19 and they are characterized by an immunoglobulin-
like fold with high affinity only for insoluble, crystalline 1,3-
xylan microfibrils.18,20,21
Scheme 1. Synthesis of Disaccharide Substrates 5 and 6
The catalytic modules of the known 1,3-xylanases all belong
to glycoside hydrolase (GH) family 26 of the CAZy database.
This sequence-based classification suggests that these xylanases
are retaining GHs, although this detail has yet to be empirically
confirmed for this subclass of enzymes. GH family 26 also
harbors β-mannanases (EC 3.2.1.78)22 and lichenases (EC
3.2.1.73).23 These two enzymes act on very different substrates
and invoke different pyranose conformational itineraries for the
electrophilic migration events that occur during catalysis.24 The
mannanases appear to operate with a 1S5 → B2,5
→
OS2
itinerary,25 while the lichenases utilize a S3 → H3 → C1
mechanism.26 The conformational itinerary of the 1,3-xylanases
remains unknown, although, based on studies of the 1,4-
xylanases, only two possibilities are likely: the 1S3 → 4H3 → 4C1
1
4
4
disaccharide 6, which would likely be a poor glycosyl acceptor
for 1,3-xylanases and thus minimize any potential competitive
transglycosylation reactions.
24
5
2
and S1 → 2,5B → SO itineraries.
From the extensive body of literature describing the 1,4-
xylanases, it is evident that the development of a range of
unnatural substrates and inhibitors has been invaluable in
delineating the details of the 1,4-xylanase mechanism.27−32
Similar molecular tools are required to foster more detailed
investigations of 1,3-xylanases; to date, no inhibitors or
chromogenic/fluorogenic substrates have been described for
these enzymes. The present work describes the development
and evaluation of such compounds, as well as their use in
addressing some of the mechanistic questions that remain
unanswered for the 1,3-xylanases.
Many endo-acting GHs prefer or require oligosaccharide
rather than disaccharide substrates, although kinetic analyses
can sometimes be complicated by the hydrolysis of several
different glycosidic bonds within the substrate. Regardless, such
substrates can be useful for exploring the role of more distant
substrate binding subsites in catalysis. A [2 + 1] glycosylation
strategy was employed to assemble the β-1,3;β-1,3-linked
trisaccharide 7. To prepare a suitable acceptor, triacetate 2 was
chloroacetylated at O-3 and then converted into the known
glycosyl bromide 8 (Scheme 2). Koenigs−Knorr glycosidation
Scheme 2. Synthesis of Trisaccharide Substrate 7
RESULTS AND DISCUSSION
■
Substrates. A number of fluorogenic oligosaccharide
substrates were initially synthesized to simplify the kinetic
analysis of 1,3-xylanases and to provide some insight into their
substrate preferences, which would then inform the design of
inhibitors. 1,3-Xylan is not currently commercially available,
which means that the preparation of 1,3-linked xylo-
oligosaccharides from natural sources is somewhat impractical.
As a consequence, the substrates were instead assembled from
monosaccharide synthons.
The known trichloroacetimidate 133 and alcohol 234 were
prepared from D-xylose and used to synthesize the β-1,3-linked
disaccharide 3 (Scheme 1). This product was treated with
hydrogen bromide in acetic acid, presumably providing the
corresponding glycosyl bromide, which was used to perform a
Koenigs−Knorr glycosidation of 4-(trifluoromethyl)-
umbelliferone (CF3UB) to return disaccharide 4. This product
of 4-(trifluoromethyl)umbelliferone and base-catalyzed trans-
esterification of the chloroacetate ester provided alcohol 9. The
β-1,3-linked disaccharide 3 was converted into the trichlor-
oacetimidate 10 and then used to glycosylate alcohol 9,
providing trisaccharide 11 in good yield. As before, a base-
catalyzed transesterification was used to effect deacetylation,
providing the desired product 7.
The glycosides 5−7 were evaluated as substrates for Xyl4, the
endo-1,3-xylanase from Vibrio sp. strain AX-4 (Table 1).17 The
enzyme was capable of processing both disaccharide substrates
5 and 6, but it exhibited markedly greater activity on the
was subjected to a Zemplen
desired disaccharide 5.
́
transesterification to provide the
Some polysaccharide-processing GHs are capable of perform-
ing transglycosylation at comparable rates to hydrolysis, which
complicates kinetic characterization. In this event, substrates
that cannot act as glycosyl acceptors can be very useful for
studying the kinetics of hydrolysis.35 For this reason, a similar
sequence of reactions was used to synthesize the 3′-O-methyl
3896
dx.doi.org/10.1021/ja211836t | J. Am. Chem. Soc. 2012, 134, 3895−3902

Journal of the American Chemical Society
Article
competitive and mechanism-based inhibitors was considered.
The classical mechanism of retaining GHs invokes transition
states with considerable oxocarbenium ion characterplacing
partial positive charge on C-1 and O-5 of the pyranose ring.
The potency of iminosugars as competitive inhibitors of GHs
can be rationalized in terms of their mimicry of this partial
positive charge developed at the transition state, although it is
also reasonable to suggest that fortuitous Columbic interactions
between a protonated iminosugar and the catalytic carboxylates
of the enzyme contribute significantly to inhibitor potency.
Potent competitive inhibitors of retaining 1,4-xylanases have
been prepared by glycosylating (3R,4R)-3,4-dihydroxypiper-
idine, an iminosugar that is the D-xylose equivalent of the
potent β-glucosidase inhibitor isofagomine, to produce 1,4-
xylan-like oligosaccharides.38,39 A similar approach to produce
1,3-xylan-like oligosaccharides 13 and 14 was expected to yield
good 1,3-xylanase inhibitors.
Table 1. Kinetic Parameters of Each Substrate for Xyl4
substrate
KM (μM)
kcat (s−1
)
kcat/KM (mM−1·s−1
)
5
(1.7 0.1) × 103
(7.1 0.2) × 102
3.8 0.1
1.9
1.1
6
2.1
3.0
7
0.89
n.d.
2.3 × 102
n.d.
12
nd
trisaccharide substrate 7. Indeed, the kcat/KM value of
trisaccharide 7 was some 210 and 80 times greater than those
of the disaccharides 5 and 6, respectively. This significant
difference appears to be driven by tighter binding of the
trisaccharide to Xyl4, as alluded to by the KM values determined
in these experiments. It is also possible that the lower KM value
could be a result of greater accumulation of the glycosyl−
enzyme intermediate for the xylotrioside substrate. Equivalent
enhanced accumulation of the intermediate was seen previously
for the GH family 10 β-1,4-xylanase Cex36 and was shown to
result from an increase in the glycosylation rate constant due to
addition of the extra sugar, with no effects on the
deglycosylation rate constant. This explanation could also
account for the considerably lower KM values observed for the
aryl glycoside substrates 5−7 relative to those reported for 1,3-
xylotetraose and 1,3-xylopentaose.17
Using established procedures, alcohol 15 was prepared and
converted into pseudodisaccharide 13 (Scheme 3).38,39
Scheme 3. Synthesis of Competitive Inhibitors 13 and 14
Thin layer chromatography and mass spectrometry revealed
that the only carbohydrate product of the Xyl4-catalyzed
hydrolysis of trisaccharide 7 (at ca. 50% completion) was 1,3-
linked xylotriose, indicating that only the reducing-end
glycosidic bond of this substrate is initially cleaved by the
enzyme. It follows that Xyl4 possesses a −3 subsite that must
be occupied for optimal activity, which is commensurate with
the apparent endo-activity of Xyl4 on 1,3-xylan itself.17 While
the simple Michaelis−Menten kinetics observed for substrates
5 and 7 suggest that transglycosylation does not occur at a rate
that is competitive with hydrolysis for this particular enzyme, it
is interesting that the methylated disaccharide 6 was in fact a
better substrate than the unmethylated disaccharide 5. This
difference appears to be driven by ground state binding
interactions, as evidenced by differing KM but similar kcat values
for the substrates. This phenomenon is typical of many endo-
GHs and is likely a consequence of the fact that the natural
polymeric substrates do not possess a free hydroxyl group at
the position of methylation but instead have an “ether-like”
oxygen capable only of accepting hydrogen bonds.35 Finally, it
should also be noted that Xyl4 had no detectable activity on 4-
methylumbelliferyl β-D-xylopyranosyl-(1→4)-β-D-xyloside 12,
suggesting that a β-1,4-link between the −1 and −2 subsites of
the enzyme is not tolerated.
Disaccharide imidate 10 was used to glycosylate alcohol 15
to provide the pseudotrisaccharide 16. This product was
subjected to transesterification and hydrogenolysis to return the
desired product 14.
Iminosugars 13 and 14 were found to competitively inhibit
Xyl4 with inhibition constants of 3.1 0.3 mM and 2.0 0.1
μM, respectively (Figure S2 of the Supporting Information).
The almost 1000-fold difference between these Ki values
illustrates once again that the −3 subsite of Xyl4 makes a
significant contribution to ligand recognition. A comparison of
the Ki values of the inhibitors to the KM values determined for
the corresponding disaccharide 5 (1.7
0.1 mM) and
trisaccharide 7 (3.8 0.1 μM) substrates indicates that these
iminosugars are in fact relatively poor inhibitors of Xyl4. A
similar phenomenon has been observed for analogous β-1,4-
linked iminosugars with a GH family 11 1,4-xylanase from
Bacillus circulans (Bcx).38 In that case it was proposed that the
poor inhibition of Bcx was the result of the iminosugars failing
to mimic the conformation of the pyranose ring in the −1
subsite at the enzyme’s transition state. This was a reasonable
Stereochemical Outcome of Xyl4-Catalyzed Hydrol-
ysis. Since the 1,3-xylan substrate of Xyl4 is quite different
from those of other enzymes from GH family 26, and given that
at least one GH family has been shown to contain enzymes of
opposing stereospecificities,37 it was important to confirm that
Xyl4 is indeed a retaining glycosidase, since this mechanistic
feature is of particular relevance to the design of inhibitors.
Accordingly, the hydrolysis of disaccharide substrate 5 was
5
explanation, since Bcx utilizes a somewhat unorthodox S1 →
2,5B → 2SO conformational itinerary,24,40 which is very different
1
4
4
from the S3 → H3 → C1 itinerary of the GH family 10
xylanases that were potently inhibited by the same
iminosugars.41 This raises the interesting possibility that
iminosugars 13 and 14 are poor inhibitors of Xyl4 because
this enzyme utilizes a similar conformational itinerary to that of
the 1,4-xylanases of GH family 11. If so, then GH family 26
would be the first GH family demonstrated to utilize at least
three different conformational itineraries to effect catalysis. This
intriguing idea provided an added incentive to develop a novel
mechanism-based inhibitor of 1,3-xylanases that functions via
1
monitored directly by H NMR (Figure S1 of the Supporting
Information), revealing the initial, exclusive formation of the β-
configured hemiacetal product. This confirmed that Xyl4 is
indeed a retaining GH.
Competitive Inhibition. Having confirmed that Xyl4
utilizes a retaining mechanism and has a preference for β-1,3-
linked trisaccharides over disaccharides, the design of
3897
dx.doi.org/10.1021/ja211836t | J. Am. Chem. Soc. 2012, 134, 3895−3902

Journal of the American Chemical Society
Article
the accumulation of the glycosyl−enzyme intermediate. Not
only could such a molecule be a more effective inhibitor of
these enzymes, but it would also provide a way to address the
question of the enzyme’s conformational itinerary using X-ray
diffraction methods.
Mechanism-Based Inhibition by Accumulation of a
Covalent Intermediate. Activated 2-deoxy-2-fluoroglycosides
are frequently used as mechanism-based inhibitors of retaining
glycosidases, with potent inhibition being achieved through the
accumulation of a covalent glycosyl−enzyme intermediate.42
When the hydroxyl group at C-2 of a glycoside is replaced with
fluorine, the oxocarbenium ion-like transition states for
formation and turnover of the glycosyl−enzyme intermediate
are destabilized both through inductive effects and through the
loss of transition state-stabilizing interactions between that
hydroxyl group and the enzyme. Installation of a good
nucleofuge at C-1 of the fluorosugar inhibitor accelerates the
first step, enabling formation of the glycosyl−enzyme
intermediate. However, hydrolysis of this glycosyl−enzyme
intermediate remains slow, so the intermediate accumulates and
enzyme activity is lost. Because Xyl4 has a strong preference for
trisaccharide substrates, synthetic efforts were directed at
preparing an activated 2-deoxy-2-fluoro trisaccharide to inhibit
1,3-xylanases. To this end, the alcohol 17 was prepared from D-
xylose according to established methods27 and then glycosy-
lated using the disaccharide imidate 10 to provide the
trisaccharide 18. This product was deprotected by the action
of methanolic hydrogen chloride to provide trisaccharide 19
(Scheme 4).
Figure 1. Time dependent inactivation of Xyl4 by 19. (A) Plot of Xyl4
activity with respect to time when incubated at 37 °C with 19 at the
●
○
following concentrations: 0 nM (∗); 25 nM ( ); 30 nM ( ); 40 nM
□
◇ ◆
▲ ▼
); 45 nM ( ); 60 nM ( ); 75 nM ( ); 90 nM ( ); and 120 nM
(
( × ). Each curve represents a nonlinear fit of each data set to a single
exponential decay equation, except for the data set collected in the
absence of 19, which was fit to a linear regression model. (B) Plot of
the observed rate constants as a function of inactivator 19
concentration. This data was best fit by a linear regression model,
the gradient of which provides an estimate of the second-order rate
constant of inactivation ki/Ki.
Scheme 4. Synthesis of a Mechanism-Based Inhibitor 19
The trisaccharide fluorosugar 19 rapidly inactivated Xyl4 in a
time-dependent manner with a second-order rate constant (ki/
Ki) of 129 1 mM−1·s−1 at 37 °C; the rapidity of the reaction
precluded an estimation of the individual parameters ki and Ki
(Figure 1). Indeed, this is one of the most efficient mechanism-
based inactivators of any glycosidase to date. To provide some
context, inactivations of 1,4-xylanases from GH family 10 and
11 with analogous activated fluorosugars are reported to
proceed with ki/Ki rate constants in the range 0.007−0.3
mM−1·s−1.27,29 As a consequence, complete inactivation of Xyl4
can be achieved within an hour at low nanomolar
concentrations of 19. Proof that the inactivation by 19 is
indeed active site-directed was provided by the protection
against inactivation afforded by competitive inhibitor 14
concentration of the 1,3-xylobioside 20, a molecule very similar
to the reducing end fragment of a natural substrate, increased
the rate of reactivation by a factor of just 3. Very little, if any,
increase in the rate of reactivation was observed for 1,4-
xylobioside 2143 or xyloside 22.39 The greater rate of
reactivation observed in the presence of 1,3-xylobioside 20
than either 1,4-xylobioside 21 or xyloside 22 alludes to the
presence of a +2 subsite and a preference of the +1/+2 subsites
for β-1,3-linked oligosaccharides. This observation is once again
commensurate with the enzyme’s endo-activity on natural 1,3-
xylan.
Identification of the Catalytic Nucleophile. Sequence
alignment of Xyl4 with the mannanases of GH family 26
suggests that the catalytic nucleophile should be E234.
Inactivator 19 provided a convenient tool to obtain direct
empirical evidence to confirm this assignment. Thus, samples of
Xyl4 were incubated with or without inactivator 19 for several
hours, thermally denatured, and digested with pepsin. Both
peptidic digests (inactivated and a control) were analyzed by
reverse-phase high performance liquid chromatography mass
spectrometry (HPLC-MS), which enabled the identification of
two glycosylated peptides that are unique to the labeled sample
(m/z = 960.0 and 1016.6, z = 2) and their unlabeled
counterparts (m/z = 761.0 and 817.5, z = 2) in the control
sample. These peptides were sequenced by tandem mass
(Figure S3). A 2.0 μM concentration of 14 (Ki = 2.0
0.1
μM) decreased the observed rate of inactivation by 60 nM 19
from 8.4 0.4 × 10−3 s−1 to 4.3 0.1 × 10−3 s−1.
The 2-fluoroxylotriosyl-enzyme species was shown to be
catalytically competent, since it reactivated by spontaneous
hydrolysis. However, hydrolytic turnover was determined to be
relatively slow, with a rate constant for hydrolysis of just 1.1
0.1 × 10−5 s−1, corresponding to a half-life of 17.5 h (Table 2,
Figure S4). Consistent with the catalytic relevance of the
trapped intermediate, reactivation was accelerated by trans-
glycosylation onto various D-xyloside acceptors. However, again
this reactivation was not fast: incubation with a 20 mM
3898
dx.doi.org/10.1021/ja211836t | J. Am. Chem. Soc. 2012, 134, 3895−3902

Journal of the American Chemical Society
Article
Table 2. Reactivation of 2-Fluoroxylotriosyl-Xyl4 through Hydrolysis and Transglycosylation
acceptor (20 mM)
reactivation rate constant, k (s−1
)
half-life for reactivation, t1/2 (h)
(1.1 0.1) × 10−5
(3.4 0.2) × 10−5
(1.4 0.1) × 10−5
(1.3 0.1) × 10−5
17.5
5.7
20
21
22
13.8
14.8
spectrometry (MS-MS), which revealed that the putative
nucleophile E234 was indeed glycosylated (Figure 2).
Figure 3. Covalent glycosyl−enzyme intermediate obtained by
inactivation of Xyl4 with 19. (A) The electron density shown is a
2Fobs − Fcalcd map contoured at 1σ. (B) Illustration of the interactions
made by the covalently bound trisaccharide (orange) with Xyl4
(green) and the surrounding ordered water molecules.
Figure 2. Identification of the catalytic nucleophile of Xyl4 by mass
spectrometry using 19.
Structure of the Trapped Glycosyl−Enzyme Inter-
mediate. To address the question of what conformational
itinerary the 1,3-xylanases might use, the structure of the
catalytic domain of Xyl4 after inactivation with 19 was
determined using X-ray diffraction methods to a resolution of
1.2 Å. Well-defined electron density was observed for the
complete ligand, clearly revealing a covalent bond between C-1
of the proximal 2-deoxy-2-fluoro-D-xylose unit and the side
chain carboxylate of E234 in the −1 subsite (Figure 3A). The
interactions of the pyranose units in the −1 and −2 subsites
with Xyl4 are typical for a glycosidase (Figure 3B). The D-
xylose unit in the −3 subsite, which is very important for
enzyme activity, makes few direct interactions with the protein
but is an integral part of a large hydrogen-bonding network that
facilitates many water-mediated interactions with Xyl4. Clearly
these water-mediated interactions are important for catalysis.
The covalently bound 2-deoxy-2-fluoroxylopyranosyl moiety
adopts a 4C1 conformation, which indicates that the 1,3-
xylanases do not operate via an intermediate with the 2,5B
conformation, as had been thought probable, but rather utilize
1
4
4
the S3 → H3 → C1 conformational itinerary. In this regard,
they are therefore similar to the lichenases of GH family 2626
and the 1,4-xylanases of GH family 10.44 This means that
conformational considerations cannot explain the poor
inhibitory activity of the iminosugars 13 and 14. Further
kinetic, thermodynamic, and structural investigations will be
necessary to address that question.
3899
dx.doi.org/10.1021/ja211836t | J. Am. Chem. Soc. 2012, 134, 3895−3902

Journal of the American Chemical Society
Article
incubated at 37 °C. Aliquots of the reactions were removed at
appropriate time points and diluted into cuvettes containing substrate
7 in buffer. The initial rates of these reactions were determined as
described above and represent a measure of the residual xylanase
activity in the inactivation mixture at that time point. The initial rates
obtained at each time point for a given concentration of 19 were fit to
a single exponential decay equation using nonlinear regression with the
program Grafit 5.0.13 to provide the pseudo-first-order rate constant
of inactivation at that concentration of 19. A plot of these pseudo-first-
order rate constants against the concentration of 19 (Figure 1)
suggested that the greatest concentration of 19 used in these
experiments (120 nM) was in fact much less than Ki, and so the
data was fit to a linear regression model with kobs ≈ ki[19]/Ki, to
provide the second-order rate constant (ki/Ki) of inactivation. To
determine the rate of hydrolysis of the glycosyl−enzyme intermediate
generated by the inactivation of Xyl4 by 19, or its rate of
transglycosylation onto acceptors 20−22 (at a concentration of 20
mM), a sample of Xyl4 was first incubated with 19 (90 nM) in buffer
at 37 °C for 2 h. This sample was concentrated to 2% of its original
volume at 4 °C using a centrifugal filter tube (Millipore, Biomax 30K
NMWL membrane) and diluted to its original volume with a
NaH2PO4/Na2HPO4 buffer (50 mM, pH 7.5) containing 150 mM
NaCl. This process was repeated three times, and then samples were
diluted further into the same buffer with or without one of the
acceptors 20−22 (final concentration 20 mM). The mixtures were
incubated at 37 °C, and aliquots were removed at appropriate time
points and diluted into cuvettes containing substrate 7 in buffer. The
initial rates of these reactions were determined as described above and
represent a measure of the residual xylanase activity in the reactivation
mixture at that time point. The initial rate data obtained for each
reactivation experiment were modeled to first-order reaction kinetics
using nonlinear regression by the program Grafit 5.0.13 to provide a
pseudo-first-order rate constant of reactivation.
Identification of the Catalytic Nucelophile of Xyl4 by Mass
Spectrometry. Xyl4 (1.0 mg·mL−1) was inactivated by 19 (0.1 mM)
as described above, except in the absence of BSA. This inactivated
sample and a control sample of untreated Xyl4 (1.0 mg·mL−1) were
concentrated to 2% of their original volume at 4 °C using a centrifugal
filter tube (Millipore, Biomax 30K NMWL membrane) and diluted
with NaH2PO4/Na2HPO4 buffer (50 mM, pH 2.0) to their original
volume. These samples were thermally denatured (90 °C, 2 min),
cooled to 25 °C, and subjected to proteolysis by the addition of pepsin
(final concentration 0.1 mg·mL−1). After 3 h at 25 °C, digestion was
halted by freezing the samples at −80 °C. The proteolysis products
were analyzed by HPLC-MS immediately after thawing. Peptides were
separated by reverse phase HPLC interfaced directly with a mass
spectrometer. Mass spectra were obtained with a triple quadrupole
mass spectrometer equipped with an ionspray ion source. Each
proteolytic digest was loaded onto a C-18 column equilibrated with
0.05% CF3CO2H and 2.0% MeCN in H2O (v/v, solvent A). The
peptides were eluted with a gradient of 0−60% solvent B [0.045%
CF3CO2H and 80% MeCN in H2O (v/v)] over 60 min, followed by
100% solvent B for 2 min. Solvents were pumped at a constant flow
rate of 50 μL·min−1. Spectra were obtained in the single-quadrupole
scan mode (HPLC-MS) or in the tandem MS-product scan mode
(MS-MS). In the HPLC-MS mode, the quadrupole mass analyzer was
scanned over a mass to charge ratio (m/z) range of 300−2200 amu
with a step size of 0.5 amu and a dwell-time of 1.5 ms per step. The ion
source voltage was set at 5.5 kV, and the orifice energy was 45 V. In
the MS-MS mode, the spectra were obtained by selectively introducing
the precursor peptide with the m/z value of interest from the first
quadrupole (Q1) into the collision cell (Q2) and observing the
fragment ions in the third quadrupole (Q3). The Q3 scan range was
50−1900 atomic mass units with the same step size and dwell time as
above.
CONCLUSION
■
A collection of 1,3-xylanase substrates has been synthesized and
validated on the 1,3-xylanase from Vibrio sp. strain AX-4. This
enzyme was confirmed to be a retaining glycosidase and
demonstrated to have a marked preference for trisaccharide
substrates, which is commensurate with its reported endo-
activity. Two isofagomine-like oligosaccharides were prepared
and shown to be relatively poor inhibitors of the xylanase, with
Ki values that are comparable to the KM value of the analogous
substrates. A 2,4-dinitrophenyl 2-deoxy-2-fluoroxylotrioside was
also synthesized and found to be an excellent mechanism-based
inhibitor of the xylanase that functions through the formation
of a stable, but catalytically competent, glycosyl−enzyme
intermediate. Using this inactivator, the catalytic nucleophile
of the enzyme was identified unequivocally by liquid
chromatography−mass spectrometric analysis of proteolytic
digests, and through reactivation experiments, evidence was
obtained that the +1/+2 subsites have a preference for β-1,3-
linked xylooligosaccharides. Finally, X-ray diffraction methods
were used to determine the structure of the inactivated
xylanase, providing evidence that these enzymes likely utilize a
4
4
1S3 → H3 → C1 conformational itinerary to effect catalysis,
much like the lichenases that also reside in GH family 26.
Together, the compounds prepared and tested here represent
an excellent set of molecular tools that set the stage for more
detailed kinetic and structural studies of this biotechnologically
relevant class of enzymes.
EXPERIMENTAL SECTION
■
General Materials and Methods. All chemicals were obtained
from the Sigma-Aldrich Chemical Co. (USA) or SynChem Inc. (USA)
and were of reagent grade. Thin layer chromatography (tlc) was
performed on Merck silica gel 60 F254 plates. Flash chromatography
was performed with Merck silica gel 60 (230−400 mesh). 1H and 13
C
NMR spectra (referenced using the residual solvent peak or an internal
MeOH standard in the case of 13C for D2O) were recorded on a
Bruker AV-300 or AV-400 instrument. High-resolution mass spectra of
all compounds were obtained in the mass spectrometry laboratory of
the Chemistry Department at U.B.C.
Enzymology. Xyl4 was recombinantly expressed in E. coli, as
described previously.17 All experiments with Xyl4 were conducted at
37 °C using a NaH2PO4/Na2HPO4 buffer (50 mM, pH 7.5)
containing 150 mM NaCl and 0.1% w/v bovine serum albumin
(BSA) unless otherwise stated. Reactions were initiated by the
addition of Xyl4 to a solution of substrate (and inhibitor) in the
aforementioned buffer. The increase in absorbance (400 nm) or
fluorescence (λex = 380 nm, λem = 500 nm) of the reaction mixture
over time gave the initial reaction rate. Initial rates were constant over
the 3 min for which each reaction was monitored. Michaelis
parameters (vmax and Km) for substrates 5−7 with Xyl4 were
determined by a nonlinear fit of the initial rates at different substrate
concentrations to the Michaelis−Menten equation using the GraFit
5.0.13 program. To determine the Ki value of inhibitors 13 and 14 for
Xyl4, a series of initial reaction rates were measured at four
concentrations of substrate 5 and a range of inhibitor concentrations
(typically bracketing the Ki value ultimately determined such that 0.2Ki
< [inhibitor] < 5Ki). These data were fit to a competitive inhibition
model using nonlinear regression analysis, as performed by the GraFit
5.0.13 program, to provide the Ki value. Dixon and Lineweaver−Burke
plots of each data set validated the use of a competitive inhibition
model. To determine the second-order rate constant of inactivation
X-ray Structure Analysis. The recombinant Xyl4 catalytic module
(rXYL4-CM) was prepared and crystallized as described previously.45
Crystals were transferred into the reservoir solution with 110%
precipitant containing 10 mM 19 and then incubated for 2 weeks. X-
ray diffraction data up to 1.2 Å were collected using the synchrotron
(ki/Ki) of Xyl4 by 19, pseudo-first-order rate constants, where kobs
=
ki[19]/(Ki + [19]), were measured for the decay in 1,3-xylanase
activity at various concentrations of 19 (25, 30, 40, 45, 60, 75, 90, and
120 nM, as well as a control of 0 nM). Inactivation reactions were
initiated by the addition of Xyl4 to a solution of 19 in buffer and
3900
dx.doi.org/10.1021/ja211836t | J. Am. Chem. Soc. 2012, 134, 3895−3902

Journal of the American Chemical Society
Article
radiation source at the beamline BL-17A of the Photon Factory
(Ibaraki, Japan). The data were processed using HKL2000.46 The
structure of rXYL4-CM in complex with β-D-xylopyranosyl-(1→3)-β-
D-xylopyranosyl-(1→3)-2-deoxy-2-fluoro-α-D-xylopyranosyl was deter-
mined by molecular replacement using the native rXYL4-CM structure
(Protein Data Bank code 2DDX) with the program Molrep.47 Manual
model fitting and the structure refinement were performed with the
programs Coot48 and Refmac5.49
(15) Yamaura, I.; Matsumoto, T.; Funatsu, M.; Mukai, E. Agric. Biol.
Chem. 1990, 54, 921−926.
(16) Araki, T.; Hashikawa, S.; Morishita, T. Appl. Environ. Microbiol.
2000, 66, 1741−1743.
(17) Kiyohara, M.; Sakaguchi, K.; Yamaguchi, K.; Araki, T.;
Nakamura, T.; Ito, M. Biochem. J. 2005, 388, 949−957.
(18) Okazaki, F.; Tamaru, Y.; Hashikawa, S.; Li, Y.-T.; Araki, T. J.
Bacteriol. 2002, 184, 2399−2403.
(19) Cantarel, B. L.; Coutinho, P. M.; Rancurel, C.; Bernard, T.;
Lombard, V.; Henrissat, B. Nucleic Acids Res. 2009, 37, D233−D238.
(20) Hashimoto, H.; Tamai, Y.; Okazaki, F.; Tamaru, Y.; Shimizu, T.;
Araki, T.; Sato, M. FEBS Lett. 2005, 579, 4324−4328.
(21) Kiyohara, M.; Sakaguchi, K.; Yamaguchi, K.; Araki, T.; Ito, M. J.
Biochem. 2009, 146, 633−641.
(22) Bolam, D. N.; Hughes, N.; Virden, R.; Lakey, J. H.; Hazlewood,
G. P.; Henrissat, B.; Braithwaite, K. L.; Gilbert, H. J. Biochemistry 1996,
35, 16195−16204.
(23) Taylor, E. J.; Goyal, A.; Guerreiro, C. I. P. D.; Prates, J. A. M.;
Money, V. A.; Ferry, N.; Morland, C.; Planas, A.; Macdonald, J. A.;
Stick, R. V.; Gilbert, H. J.; Fontes, C. M. G. A.; Davies, G. J. J. Biol.
Chem. 2005, 280, 32761−32767.
(24) Davies, G. J.; Ducros, V. M.-A.; Varrot, A.; Zechel, D. L.
Biochem. Soc. Trans. 2003, 31, 523−527.
(25) Ducros, V. M.-A.; Zechel, D. L.; Murshudov, G. N.; Gilbert, H.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental information for the synthesis of novel com-
pounds, demonstration of the retaining mechanism of Xyl4
using 1H NMR spectroscopy, Dixon plots for inhibition of Xyl4
by 13 and 14, reactivation experiments on inactivated Xyl4, and
1H and 13C NMR data for all novel compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
withers@chem.ubc.ca
■
Present Address
∥Institute for Biomedical Aging Research, Austrian Academy of
Sciences, Innsbruck A-6020, Austria.
́
J.; Szabo, L.; Stoll, D.; Withers, S. G.; Davies, G. J. Angew. Chem., Int.
Ed. 2002, 41, 2824−2827.
(26) Money, V. A.; Smith, N. L.; Scaffidi, A.; Stick, R. V.; Gilbert, H.
J.; Davies, G. J. Angew. Chem., Int. Ed. 2006, 45, 5136−5140.
(27) Ziser, L.; Setyawati, I.; Withers, S. G. Carbohydr. Res. 1995, 274,
137−153.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
E.D.G.B thanks the Canadian Institutes of Health Research for
a postdoctoral fellowship. S.R. thanks the Austrian Science
■
(28) McIntosh, L. P.; Hand, G.; Johnson, P. E.; Joshi, M. D.; Korner,
̈
M.; Plesniak, L. A.; Ziser, L.; Wakarchuk, W. W.; Withers, S. G.
Biochemistry 1996, 35, 9958−9966.
Fund for an Erwin Schrodinger Fellowship. S.G.W. acknowl-
̈
(29) Miao, S.; Ziser, L.; Aebersold, R.; Withers, S. G. Biochemistry
1994, 33, 7027−7032.
edges support from the Natural Sciences and Engineering
Research Council of Canada, the Canada Research Chairs
Program, the Canadian Foundation for Innovation, and the
B.C. Knowledge Development Fund. We thank Shouming He
of the Proteomics Core Facility at UBC for mass spectrometric
analyses.
(30) Notenboom, V.; Williams, S. J.; Hoos, R.; Withers, S. G.; Rose,
D. R. Biochemistry 2000, 39, 11553−11563.
(31) Wicki, J.; Schloegl, J.; Tarling, C. A.; Withers, S. G. Biochemistry
2007, 46, 6996−7005.
(32) Ludwiczek, M. L.; Heller, M.; Kantner, T.; McIntosh, L. P. J.
Mol. Biol. 2007, 373, 337−354.
(33) Mori, M.; Ito, Y.; Ogawa, T. Carbohydr. Res. 1990, 195, 199−
224.
REFERENCES
■
(1) Correns, C. Ber. Dtsch. Bot. Ges. 1894, 12, 355.
(2) Nicolai, E.; Preston, R. D. Proc. R. Soc. B 1952, 140, 244−274.
(3) Veluraja, K.; Atkins, E. D. T. Carbohydr. Polym. 1987, 7, 133−
141.
(34) Utille, J.-P.; Gagnaire, D. Carbohydr. Res. 1982, 106, 43−57.
(35) Damager, I.; Numao, S.; Chen, H. M.; Brayer, G. D.; Withers, S.
G. Carbohydr. Res. 2004, 339, 1727−1737.
(36) Tull, D.; Withers, S. G. Biochemistry 1994, 33, 6363−6370.
(37) Gloster, T. M.; Turkenburg, J. P.; Potts, J. R.; Henrissat, B.;
Davies, G. J. Chem. Biol. 2008, 15, 1058−1067.
(38) Williams, S. J.; Hoos, R.; Withers, S. G. J. Am. Chem. Soc. 2000,
122, 2223−2235.
(4) Atkins, E. D. T.; Parker, K. D.; Preston, R. D. Proc. R. Soc. B 1969,
173, 209−221.
(5) Frei, E.; Preston, R. D. Proc. R. Soc. London B 1964, 160, 293−
313.
(6) MacArtain, P.; Gill, C. I. R.; Brooks, M.; Campbell, R.; Rowland,
I. R. Nutr. Rev. 2007, 65, 535−543.
(39) Goddard-Borger, E. D.; Fiege, B.; Kwan, E. M.; Withers, S. G.
ChemBioChem 2011, 12, 1703−1711.
(7) Nguyen, V. T.; Ueng, J.-P.; Tsai, G.-J. J. Food. Sci. 2011, 76,
C950−C958.
(40) Sidhu, G.; Withers, S. G.; Nguyen, N. T.; McIntosh, L. P.; Ziser,
L.; Brayer, G. D. Biochemistry 1999, 38, 5346−5354.
(41) White, A.; Tull, D.; Johns, K.; Withers, S. G.; Rose, D. R. Nat.
Struct. Biol. 1996, 3, 149−154.
(8) Meusnier, I.; Olsen, J. L.; Stam, W. T.; Destombe, C.; Valero, M.
Mol. Ecol. 2001, 10, 931−946.
(9) Araki, T.; Hayakawa, M.; Tamaru, Y.; Yoshimatu, K.; Morishita,
T. J. Phycol. 1994, 30, 1040−1046.
(42) Withers, S. G.; Street, L. P.; Bird, P.; Dolphin, D. H. J. Am.
Chem. Soc. 1987, 109, 7530−7531.
(10) Umemoto, Y.; Shibata, T.; Araki, T. Mar. Biotechnol. 2012, 14,
10−20.
(43) Lop
745.
́ ́
ez, R.; Fernadez-Mayoralas, A. J. Org. Chem. 1994, 59, 737−
(11) Iriki, Y.; Suzuki, T.; Nisizawa, K.; Miwa, T. Nature 1960, 187,
82−83.
(44) Notenboom, V.; Birsan, C.; Warren, R. A. J.; Withers, S. G.;
Rose, D. R. Biochemistry 1998, 37, 4751−4758.
(45) Sakaguchi, K.; Kiyohara, M.; Watanabe, N.; Yamaguchi, K.; Ito,
M.; Kawamura, T.; Tanaka, I. Acta Crystallogr., D: Biol. Crystallogr.
2004, 60, 1470−1472.
(12) Chen, W. P.; Matsuo, M.; Yasui, T. Agric. Biol. Chem. 1986, 50,
1183−1194.
(13) Araki, T.; Inoue, N.; Morishita, T. J. Gen. Appl. Microbiol. 1998,
44, 269−274.
(46) Otwinowski, Z.; Minor, W. Methods Enzymol. 1997, 276, 307−
326.
(14) Araki, T.; Tani, S.; Maeda, K.; Hashikawa, S.; Nakagawa, H.;
Morishita, T. Biosci. Biotechnol. Biochem. 1999, 63, 2017−2019.
3901
dx.doi.org/10.1021/ja211836t | J. Am. Chem. Soc. 2012, 134, 3895−3902

Journal of the American Chemical Society
Article
(47) Vagin, A.; Teplyakov, A. Acta Crystallogr., D: Biol. Crystallogr.
2000, 56, 1622−1624.
(48) Emslcy, P.; Cowtan, K. Acta Crystallogr., D: Biol. Crystallogr.
2004, 60, 2126−2132.
(49) Murshudov, G. N.; Vagin, A. A.; Dodson, E. J. Acta Crystallogr.,
D: Biol. Crystallogr. 1997, 53, 240−255.
3902
dx.doi.org/10.1021/ja211836t | J. Am. Chem. Soc. 2012, 134, 3895−3902