Novel substrate specificities of two lacto-N-biosidases towards β-linked galacto-N-biose-containing oligosaccharides of globo H, Gb5, and GA1

Article abstract of DOI:10.1016/j.carres.2015.03.005

We describe the novel substrate specificities of two independently evolved lacto-N-biosidases (LnbX and LnbB) towards the sugar chains of globo- and ganglio-series glycosphingolipids. LnbX, a non-classified member of the glycoside hydrolase family, isolated from Bifidobacterium longum subsp. longum, was shown to liberate galacto-N-biose (GNB: Galβ1-3GalNAc) and 2′-fucosyl GNB (a type-4 trisaccharide) from Gb5 pentasaccharide and globo H hexasaccharide, respectively. LnbB, a member of the glycoside hydrolase family 20 isolated from Bifidobacterium bifidum, was shown to release GNB from Gb5 and GA1 oligosaccharides. This is the first report describing enzymatic release of β-linked GNB from natural substrates. These unique activities may play a role in modulating the microbial composition in the gut ecosystem, and may serve as new tools for elucidating the functions of sugar chains of glycosphingolipids.

Full text of DOI:10.1016/j.carres.2015.03.005

Carbohydrate Research 408 (2015) 18e24
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Novel substrate specificities of two lacto-N-biosidases towards
b-linked galacto-N-biose-containing oligosaccharides of globo H,
Gb5, and GA1
Aina Gotoh a, Toshihiko Katoh a, Yuta Sugiyama a, Shin Kurihara a, Yuji Honda a,
Haruko Sakurama a, Taiho Kambe b, Hisashi Ashida c, Motomitsu Kitaoka d,
, b, *
Kenji Yamamoto a, Takane Katayama a
a Ishikawa Prefectural University, Nonoichi, Ishikawa 921-8836, Japan
b Graduate School of Biostudies, Kyoto University, Sakyo-ku, Kyoto 606-8502, Japan
c Faculty of Biology-Oriented Science and Technology, Kinki University, Wakayama 649-6493, Japan
d National Food Research Institute, National Agriculture and Food Research Organization, Tsukuba, Ibaraki 305-8642, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
We describe the novel substrate specificities of two independently evolved lacto-N-biosidases (LnbX and
LnbB) towards the sugar chains of globo- and ganglio-series glycosphingolipids. LnbX, a non-classified
member of the glycoside hydrolase family, isolated from Bifidobacterium longum subsp. longum, was
Received 7 January 2015
Received in revised form
6 March 2015
Accepted 8 March 2015
Available online 14 March 2015
shown to liberate galacto-N-biose (GNB: Galb
1-3GalNAc) and 20-fucosyl GNB (a type-4 trisaccharide)
from Gb5 pentasaccharide and globo H hexasaccharide, respectively. LnbB, a member of the glycoside
hydrolase family 20 isolated from Bifidobacterium bifidum, was shown to release GNB from Gb5 and GA1
oligosaccharides. This is the first report describing enzymatic release of
b-linked GNB from natural
Keywords:
Lacto-N-biosidase
substrates. These unique activities may play a role in modulating the microbial composition in the gut
ecosystem, and may serve as new tools for elucidating the functions of sugar chains of
glycosphingolipids.
b-Linked galacto-N-biose
GA1
© 2015 Elsevier Ltd. All rights reserved.
Gb5
Globo H
1. Introduction
microbes in the gut.7 Meanwhile, recent studies have shown that
host-derived glycans are also a significant determinant in gut mi-
The consortium of gut microbes exerts a considerable influence
on host health. Short-chain fatty acids produced by microbes in the
intestines have been shown to affect body weight, energy balance,
and immune responses via GPR43 in mice,1,2 while butyric acid
excreted by clostridia is known to induce the differentiation of
mouse colonic regulatory T cells.3 In addition, some behavioral
abnormalities such as anxiety-like behavior and impaired cognitive
function have been linked to gut microbial metabolic activity in
mice.4e6 It is therefore essential for host health to understand what
shapes and influences the gut microbiota.
crobial composition as these serve as a nutrient source for certain
microbes.8 The genome of Bacteroides thetaiotaomicron, a dominant
symbiont in adult intestines, contains polysaccharide utilization
loci responsible for the utilization of mucin O-glycans and glycos-
aminoglycans.9 Infant gut-dominating species of the genus Bifido-
bacterium are equipped with enzymes dedicated to the metabolism
of human milk oligosaccharides (HMOs), the third-most abundant
solid component in human milk.9e13 It is therefore highly likely
that these commensal bacteria have evolved selected glycosidases
for degrading host-derived glycans to prosper in this ecological
niche and to establish a harmonious relationship with the host.
Hence, unraveling the unique functions of these glycosidases is
necessary for us to determine the principles underlying the for-
mation and maintenance of normal gut microbiota. This notion is
Dietary glycans (dietary fibers) are one of the known factors
affecting gut microbial composition, and these have gained
considerable attention in efforts to develop prebiotics (generally,
oligosaccharides), which stimulate the growth of beneficial
strengthened by the finding that FUT2-mediated H-antigen (Fuc
2Gal) expression on the gut epithelial cell surface (host side) and
1,2- -fucosidase-mediated release of Fuc from the antigens
(microbe side) are important for establishing beneficial symbiosis14
a1-
a-L
* Corresponding author. Tel.: þ81 76 227 7513; fax: þ81 227 7557.
E-mail address: takane@ishikawa-pu.ac.jp (T. Katayama).
http://dx.doi.org/10.1016/j.carres.2015.03.005
0008-6215/© 2015 Elsevier Ltd. All rights reserved.

A. Gotoh et al. / Carbohydrate Research 408 (2015) 18e24
19
and for decreasing bacterial virulence gene expression in the in-
testines.15,16 1,2-
-Fucosidase (EC 3.2.1.63) thus could be repre-
In this paper, we report a novel glycan-scavenging activity of
LnbX and LnbB, i.e., the activity of these enzymes towards the oli-
gosaccharides of globo- and ganglio-series sphingolipids. This is the
a-L
sentative of the selected, host glycan-specific microbial
glycosidases affecting the gut ecosystem.17
first report describing
b-linked galacto-N-biose (GNB)-releasing
Lacto-N-biosidase (EC 3.2.1.140), a key enzyme that degrades
activity on natural substrates with biological implications. These
enzymes will likely serve as a future tool for not only understanding
glycan-mediated host-microbe interaction but also for examining
the role of sugar chains of complex glycosphingolipids.
lacto-N-tetraose (LNT: Gal
b1-3GlcNAcb1-3Gal
b1-4Glc),
a main
component of HMOs, into lacto-N-biose I (LNB: Gal
b
1-3GlcNAc)
and lactose (Lac), also appears to be such an enzyme. Two different
lacto-N-biosidases have been isolated18e20; one belongs to the
glycoside hydrolase family (GH) 20 while the other is a non-
classified member listed in the CAZy database (http://www.cazy.
org).21 These enzymes are lacto-N-biosidase from Bifidobacterium
bifidum (LnbB), and the enzyme from Bifidobacterium longum subsp.
longum (LnbX), respectively (Fig. 1, see legend for detailed struc-
tural description). Our previous studies demonstrated that GH20
LnbB exhibits strict specificity towards the unmodified LNB struc-
2. Results
2.1. Inhibition of LnbX activity in the presence of GalNAc
Although the main function of LnbX must be the hydrolysis of
lacto-N-tetraose (LNT: Galb1-3GlcNAcb1-3Galb1-4Glc) based on
the catalytic efficiency for this substrate being relatively high (see
Table 2),19 the quite rare occurrence of LnbX homologues in nature
and the distant relationship of LnbX with other proteins in the
database prompted us to examine the heretofore unknown enzy-
matic activities of this enzyme. To this end, we examined the
inhibitory effects of monosaccharides that constitute the sugar
chains of host glycoproteins, glycolipids, and glycosaminoglycans,
i.e., Fuc, Gal, GalNAc, Glc, GlcA, GlcNAc, Man, Neu5Ac, and Xyl. 4-
ture, while LnbX is capable of liberating GalNAc
disaccharide structure found in O-mannosyl glycans of
glycan)22 and 20-fucosyl LNB (Fuc
1-2Gal 1-3GlcNAc) in addition
to LNB.18,19 LnbB has evolved from exo-
-N-acetylhexosaminidase,
the main member of GH20, by acquiring the ꢀ2 subsite to accom-
modate the
-(1/3)-linked Gal residue of LNB.23 GH20 lacto-N-
b
1-3GlcNAc (a
a-dystro-
a
b
b
b
biosidase is distributed only in the Bifidobacterium and Strepto-
myces genera; nevertheless, these two bacterial groups likely
evolved these enzymes differently because the amino acid
sequence homology between these two enzymes is 38% at
maximum. Although we do not yet have a clear answer as to what
the driving forces were for the evolution of this enzyme in Strep-
tomyces, in the case of the bifidobacterial enzyme, the host-derived
glycans, especially the HMOs, were likely to have been the main
factor. Homologues of LnbX have so far been identified in gut mi-
crobes belonging to only five genera (Bifidobacterium, Clostridium,
Ruminococcus, Roseburia, and Enterococcus) and these have not yet
been assigned to clusters of orthologous groups (COGs), indicating
a distant relationship to other proteins in the database. LnbX could
thus be an environmentally driven, selected enzyme. Considering
that lacto-N-biosidase itself is rarely found in nature (only three
enzymes have been characterized), the occurrence of two very
different lacto-N-biosidases in one genus, Bifidobacterium, is of in-
terest (Fig. 1), and this enzyme should therefore serve as an
attractive target for better understanding the molecular basis of
host-microbe symbiosis and co-evolution, especially in terms of the
glycan-foraging strategies of the gut microbes.
Nitrophenyl-b-lacto-N-bioside (LNB-b-pNP) was used as the sub-
strate. Only GalNAc was found to inhibit the activity (Fig. 2a). The
inhibition by GalNAc appeared to be competitive with the Ki value
of 6.4 mM (Fig. 2b). Neither Gal nor GlcNAc (LNB-constituting
monosaccharide) had a significant inhibitory effect. Considering
that the Km values of LnbX for LNB-
b
b-pNP and GalNAcb1-3GlcNAc-
-pNP were comparable,19 the results suggest that the enzyme may
recognize GalNAc at subsite ꢀ1 more tightly than GlcNAc. Note that
LnbX does not act on GalNAc-
vious paper.19
b-pNP, as was described in our pre-
2.2. Activity of LnbX and LnbB towards GNB-containing
oligosaccharides
We first examined the Galb1-3GalNAc (GNB)-b-pNP-hydrolyz-
ing activity of LnbX. The activity of GH20 LnbB was also examined
for comparison. The kinetic parameters of LnbX and LnbB obtained
for LNB-
reported in previous studies (Table 2).19,23 As expected, LnbX hy-
drolyzed GNB- -pNP, although the specific activity was 50-fold
b-pNP in this study were essentially the same as those
b
Fig. 1. Schematic representation of two lacto-N-biosidases LnbX and LnbB. (a) The active enzyme molecule consists of the gene product of lnbX only, while the lnbY gene product
acts as a designated chaperone for LnbX. The lnbX and lnbY genes constitute an operon. LnbX from B. longum subsp. longum contains a signal peptide (black bar), a right-handed
b-
helix region (b-helix, Pfam 13229) (light gray), an uncharacterized sugar-binding domain (FIVAR, Pfam 07554) (shaded box), and a membrane anchor (black bar) in this order from
the N-terminus.47 LnbY has a signal peptide (black bar). (b) LnbB from B. bifidum has a signal peptide (black bar), a GH20 catalytic domain (dotted box), a carbohydrate-binding
module 32 (CBM32 in CAZy database) (shaded box), a bacterial immunoglobulin-like 2 domain (Big_2, PF02368) (dark gray), followed by a membrane anchor (black bar). Both
LnbX and LnbB are cell surface-exposed proteins.10

20
A. Gotoh et al. / Carbohydrate Research 408 (2015) 18e24
(a)
(b)
120
100
80
60
40
20
0
120
100
80
60
40
20
0
-10
-5
0
5
10
15
1/[S]0 (/mM)
Fig. 2. Inhibitory effect of monosaccharides on the LNB-
-pNP, and 10 mM monosaccharide (each) in 200 mM sodium phosphate buffer (pH 6.0) was incubated for 15 min at 30 ꢁC. The amount of pNP released in the presence of each
sugar was compared with that released in the control experiment without monosaccharides. (b) The double reciprocal plot of LNB- -pNP hydrolysis in the presence or absence of
-pNP, and 10 mM GalNAc in 50 mM sodium phosphate buffer (pH 6.0) was
b-pNP-hydrolyzing activity of LnbX. (a) The reaction mixture (total volume of 50 mL) containing 0.23 mg/mL LnbX, 1 mM LNB-
b
b
GalNAc. The reaction mixture (total volume of 100
incubated for 15 min at 30 ꢁC.
mL) containing 0.092 mg/mL LnbX, 0e0.5 mM LNB-b
lower than for LNB-
ingly, reflecting the results obtained in the inhibition assay, LnbX
exhibited a lower Km value for GNB-
(<0.05 mM vs 0.12 mM). GH20 LnbB also hydrolyzed GNB-
with 3-fold lower efficiency than for LNB-
(2500 U/g vs 7700 U/g). Activity of LnbB towards GNB-
surprising because LnbB likely evolved from the GH20 exo-
acetylhexosaminidases, most of which equally act on both
acetylglucosaminides and
b
-pNP (510 U/g versus 25,000 U/g). Interest-
lower Km value of the enzyme for GNB-
-pNP (<0.05 mM vs 0.12 mM). The difference in kcat was 130-fold
between LNT and Gb5 pentasaccharide (110 sꢀ1 vs 0.84 sꢀ1), which
was 2-fold greater than the difference between LNB- -pNP and
GNB-
-pNP (60-fold: 83 sꢀ1 vs 1.5 sꢀ1). These results suggest that
b-pNP compared with LNB-
b
b
-pNP than for LNB-
b
b
-pNP
-pNP
b
b-pNP-hydrolysis
b
b
-pNP is not
steric impedance may occur at the þsubsite(s) of LnbX, as Gb5
b
b
-N-
-N-
oligosaccharide has an extension of one sugar ring compared with
LNT, and also has an
a
-(1/4)-linkage between þ1 Gal and þ2 Gal.
b
-N-acetylgalactosaminides. Neither of
Globo H hexasaccharide was hydrolyzed by LnbX with 40-fold
lower efficiency than Gb5 pentasaccharide at a substrate concen-
tration of 5 mM (5.1 U/g vs 210 U/mg). Significant decreases in
LnbX and LnbB acted on a-linked GNB, which is present in mucin-
type O-glycans (data not shown).24
We next examined whether LnbX and LnbB could hydrolyze
GNB-containing natural oligosaccharides derived from globosides
and gangliosides (see Table 1 for glycan structures). LNB-containing
sugars LNT and lacto-N-fucopentaose I (LNFP I) were also included
in the assay for reference. Thin-layer chromatography analysis
showed that LnbX acts on LNT, LNFP I, Gb5 pentasaccharide, and
globo H hexasaccharide, while LnbB acts on LNT, Gb5, and GA1
oligosaccharides (Fig. 3). LnbB did not liberate GNB from GM1
pentasaccharide (data not shown). The hydrolysis of the substrates
was confirmed by MS analysis, in which molecular ion peaks cor-
responding to the sodium adducts of the expected reaction prod-
ucts were detected (Table 1 and Fig. 4): for LNT hydrolysis, Lac
(calculated, 365.1; observed, 365.1) and LNB (calculated, 406.1;
observed, 406.1 or 406.2); for LNFP I, Lac (observed, 365.1) and 20-
fucosyl LNB (calculated, 552.2; observed, 552.3); for Gb5, GNB
activity in the presence of the a-(1/2)-linked terminal Fuc residue
was also observed when LNT was replaced with LNFP I (Table 2).
LnbX did not liberate GNB from GA1 tetrasaccharide, but it did act
on Gb5 pentasaccharide with considerable activity, suggesting that
this enzyme does not tolerate non-planarity of the substrate be-
tween the ꢀ1 and þ1 subsites.
Activity of GH20 LnbB towards Gb5 oligosaccharide was signif-
icantly low. The catalytic turnover differed by 170-fold between
GNB-b
-pNP and Gb5 (4.9 sꢀ1 vs 0.029 sꢀ1) and the Km value differed
by at least 60-fold (3.3 mM vs <0.05 mM). It is interesting to note
that LnbB has no apparent þsubsite and its catalytic pocket is open
to the solvent23; nevertheless, its Km value was significantly
affected by the structure of the leaving groups. LnbB did not act on
globo H oligosaccharide as the enzyme does not hydrolyze LNFP I.
Specific activity of LnbB for GA1 tetrasaccharide was 5-fold higher
than for Gb5 pentasaccharide (49 U/g vs 9.4 U/g). In this case too,
the structural difference in the leaving group affected the hydro-
lysis activity of LnbB.
(calculated, 406.1; observed, 406.1) and Gal
a
1-4Gal
b
1-4Glc
1-
(calculated, 527.2; observed, 527.3); for globo H, Gal
a
1-4Galb
4Glc (observed, 527.3) and 20-fucosyl GNB (calculated, 552.2;
observed, 552.3); and for GA1, Lac (observed, 365.1) and GNB
(observed, 406.1).
3. Discussion
Kinetic parameters of LnbX and LnbB for Gb5 pentasaccharide,
and specific activities of LnbX for Gb5 and globo H oligosaccharides
and of LnbB for Gb5 and GA1 oligosaccharides, were determined
(Table 2). LnbX showed a six-fold higher Km value for Gb5 penta-
saccharide than for LNT (2.5 mM vs 0.4 mM) in contrast with the
Glycosphingolipids are a class of glycolipids containing a long-
chain amino alcohol (sphingoid base), generally in amide linkage
to a fatty acid as the lipid moiety. In vertebrates, those carrying
glycan chains consisting of more than two sugars are categorized
Table 1
Substrates used for the specificity analyses of LnbX and LnbB
Substrate (oligosaccharide)
Structure
Gal 1-3GlcNAc
Hydrolysis products
LNT
LNFP I
Gb5
Globo H
GA1
b
b
1-3Gal
b
b
a
b
1-4Glc
1-3Gal
1-4Gal
1-3Gal
Gal
Fuc
Gal
Fuc
Gal
b
1-3GlcNAc/Gal
1-2Gal 1-3GlcNAc/Gal
1-3GalNAc/Gal 1-4Gal
1-2Gal 1-3GalNAc/Gal
1-3GalNAc/Gal 1-4Glc
b1-4Glc
Fuc 1-2Gal 1-3GlcNAc
Gal 1-3GalNAc 1-3Gal
Fuc 1-2Gal 1-3GalNAc
Gal 1-3GalNAc 1-4Gal
a
b
b
b
a
1-4Glc
1-4Glc
1-4Gal
a
b
b
b
1-4Glc
1-4Glc
b
b
b
a
a
b
b1-4Glc
a
b
a1-4Galb1-4Glc
b
b
b
1-4Glc
b
b

A. Gotoh et al. / Carbohydrate Research 408 (2015) 18e24
21
(a) 2 U/mL enzyme
LNT
Gb5
Globo H
GA1
LNFP I
pentasaccharide
hexasaccharide
tetrasaccharide
(b) 20 U/mL enzyme
Gb5
Globo H
GA1
LNFP I
pentasaccharide
hexasaccharide
tetrasaccharide
Fig. 3. Substrate specificities of LnbX and LnbB towards natural oligosaccharides containing LNB or GNB disaccharide structures. The reaction mixture (total volume of 50
mL)
containing 2 mM substrate in 50 mM sodium phosphate buffer (pH 6.0) was incubated at 30 ꢁC for 12 h in the presence of 2 U/mL (a) or 20 U/mL (b) of either enzyme (LnbX or
LnbB). Note that the unit (U) was determined using LNB-b-pNP as a substrate. The reaction products were analyzed by thin-layer chromatography using a solvent system of 1-
butanol/acetic acid/water (2/1/1 by volume). The sugars were visualized with diphenylamine-aniline-phosphoric acid.48 Control experiments were performed without the
enzyme. LNB or GNB was used as the standard (last lane in each image).
into four major groups (globo-, ganglio-, lacto-, and neolacto-series)
based on the core structures of the glycans. These glycolipids are
found in the outer leaflet of the cell membrane where they form the
so-called lipid rafts and play important roles in cellecell in-
teractions and signal transduction.25 A deficiency in the lysosomal
enzymes required to degrade glycosphingolipids causes sphingo-
lipidoses in humans, including Fabry disease and TayeSachs dis-
ease.26,27 In addition, the glycan moieties of glycosphingolipids are
exploited by several bacteria and their toxins as pedestals of
adhesion to the host cells.28,29 Therefore, finding unique enzymes
that modify these glycan structures is important for the future
analysis of the physiological and pathological functions of complex
glycosphingolipids.
from Gb5 and globo H oligosaccharides, respectively,33 but did not
act on GA1 tetrasaccharide, whereas LnbB liberated GNB from Gb5
and GA1 oligosaccharides. To our knowledge, this is the first report
describing enzymatic release of b-linked GNB from natural sub-
strates. This is also the first description of type-4 trisaccharide-
releasing activity. Considering that globosides and gangliosides are
present on the surface of gastrointestinal epithelial cells, and that
both LnbB and LnbX are cell wall-anchored proteins (Fig. 1), these
activities might give bifidobacteria a nutritional advantage in the
gut, as these organisms have a GNB/LNB assimilation pathway34
and/or may influence the gut microbial ecosystem.35 The latter is
due to the sugar chains of various glycolipids frequently acting as
receptors for many bacteria in the gut.
3.1. Physiological implications
3.2. Enzymatic tool
Degradation of the sugar chains of glycosphingolipids by gut
microbes was first mentioned by Larson et al. in 1988 and later by
Falk et al.30,31 They described several bifidobacteria and rumino-
cocci express enzymes that can degrade lacto- and neolacto-series
glycosphingolipids, the sugar moieties of which are essentially the
same as for HMOs. Degradation of the major milk gangliosides GM3
and GD3 by bifidobacteria was also recently reported.32 In the
present study, we found that LnbX and LnbB are able to act on the
sugar chains of GNB-containing globo-series (Gb5 and globo H) and
asialoganglio-series (GA1) glycosphingolipids, although these ac-
tivities were relatively low compared with those towards HMOs
(Table 2). LnbX liberated GNB and 20-fucosyl GNB (type-4 chain)
In recent years, the need for glycan engineering techniques has
increased for use in studies on embryology, immunotherapy, and
pathogenicity. Globo H is a carbohydrate marker that is highly
expressed in tumor cells, especially in cancers of the breast, pros-
tate, and lung, and is probably involved in metastasis.36 Gb5, the so-
called stage-specific embryonic antigens 3a (SSEA-3a), is expressed
on the cell surfaces of multilineage-differentiating stress-enduring
(muse) cells, bone marrow stroma, embryonic stem cells, and
induced pluripotent stem cells,37e41 although its role in these cells
remains elusive.42 GA1 acts as a receptor for Pseudomonas aerugi-
nosa and Pseudomonas cepacia.43 Unique activities of LnbX and
LnbB could therefore serve as useful tools for identifying the glycan

22
A. Gotoh et al. / Carbohydrate Research 408 (2015) 18e24
Fig. 4. MALDI-TOF-MS analysis of the reaction products catalyzed by LnbX and LnbB. The substrates used were LNT (a), LNFP I (b), Gb5 pentasaccharide (c), globo H hexasaccharide
(d), and GA1 tetrasaccharide (e). See Fig. 3 caption for assay conditions.
moieties required for these biological activities, together with the
specific exo-glycosidases and endoglycoceramidases.44,45 In addi-
tion, these enzymes may serve as tools for synthesizing GNB-
containing glycoconjugates where glycosynthase technology is
applicable, because we have already established a method for the
large-scale synthesis of GNB (~100 g).46 X-ray structural analysis of
LnbX is currently in progress in our laboratory.
important for deciphering how gut microbes have co-evolved with
their hosts, driven by the glycans available in their environment
and their influence on the gut microbial ecosystem. In addition,
these enzymes may prove useful for trimming or remodeling the
sugar chains of various glycoconjugates.
4. Materials and methods
3.3. Conclusions
4.1. Chemicals
Unexpectedly, novel glycan-modifying activities of two sepa-
rately evolved lacto-N-biosidases were revealed. These findings are
LNB-
b
-pNP and LNFP I were purchased from SigmaeAldrich
-pNP was purchased from Tokyo Chemical
(Missouri, USA). GNB-
b

A. Gotoh et al. / Carbohydrate Research 408 (2015) 18e24
23
Table 2
Activities of LnbX and LnbB towards GNB-containing sugars
LnbX (non-classified GH member)
LnbB (GH20)
Km (mM)
kcat (sꢀ1
)
kcat/Km (mMꢀ1 sꢀ1
)
Sp. act.a (U gꢀ1
)
Km (mM)
kcat (sꢀ1
)
kcat/Km (mMꢀ1 sꢀ1
)
Sp. act. (U gꢀ1
)
LNB-
GNB-
LNT
b
b
-pNP
-pNP
0.12
<0.05b
0.40c
15c
2.5
ND
d
88
730
d
280c
0.93c
0.34
ND
25,000
510
NDd
ND
210
5.1
<0.05b
<0.05b
0.63c
d
16
d
7700
2500
ND
1.5
110c
14c
0.84
ND
d
4.9
42c
d
d
67c
d
LNFP I
nde
9.4
Gb5f
3.3
d
0.029
d
0.0088
d
Globo Hf
GA1f
nd
49
d
nd
ND
ND
ND
a
Specific activity was determined at the substrate concentration of 5 mM.
The Km values were too low to be determined.
Data from Sakurama et al.19 (measured at 25 ꢁC).
ND, not determined.
nd, not detected.
Oligosaccharides were used as the substrate.
b
c
d
e
f
Industry Co. (Tokyo, Japan). Lacto-N-tetraose (LNT) was obtained
from Dextra Laboratories (Reading, UK). Globo H, Gb5 and GA1
oligosaccharides were from Elicityl (Crolles, France). Gb5 penta-
saccharide was further purified by semi-preparative high-perfor-
mance liquid chromatography (HPLC) equipped with an Asahipak
NH2P-90 20F column (Showa Denko, Tokyo, Japan). Other re-
agents of analytical grade were obtained from various commercial
sources.
4.4. Matrix-assisted laser desorption/ionization time-of-flight
(MALDI-TOF) mass spectrometry (MS)
Prior to the analysis, proteins were removed by acetone pre-
cipitation. The resulting supernatant was dried and dissolved in
50% methanol containing 20 mg/mL 2,5-dihydroxybenzoic acid as
the matrix (SigmaeAldrich). MALDI-TOF MS analysis was per-
formed with an UltrafleXtreme (Bruker Daltonics, MA, USA) in the
positive ion and reflector mode.
4.2. Enzyme preparation
Acknowledgments
LnbX and LnbB proteins were expressed in Escherichia coli BL21
A part of this work was conducted in JAIST, supported by the
Nanotechnology Platform Program (Molecule and Material Syn-
thesis) of the Ministry of Education, Culture, Sports, Science and
Technology (MEXT), Japan. This work was supported in part by a
Grant-in-Aid for Scientific Research (C) 24580119 from the Ministry
of Education, Culture, Sports, Science and Technology, Japan; a
Grant-in-Aid from the Asahi Glass Foundation; a Grant-in-Aid from
the Science and Technology Research Promotion Program for
Agriculture, Forestry, Fisheries and Food Industry; and a Grant-in-
Aid from the Institution for Fermentation, Osaka.
D
lacZ (pRARE2) as hexahistidine-tagged forms and purified as
described previously.18,19 Protein concentrations were determined
using a BCA protein assay kit (Thermo Scientific, MA, USA) with
bovine serum albumin as a standard.
4.3. Enzyme assay
The standard reaction mixture contained 50 mM sodium
phosphate buffer (pH 6.0). pNP-sugars and oligosaccharides were
used as the substrates. The reaction was initiated by addition of the
enzyme and the mixture was then incubated at 30 ꢁC for the
appropriate time, in which the linearity of the reaction rate was
References
observed. The concentration of LnbX was 0.09e0.3
-pNP), 4e6 g/mL (for GNB- -pNP), 50e60 g/mL (for Gb5 pen-
tasaccharide), or 970 g/mL (for globo H hexasaccharide). The
concentration of LnbB was 0.1e0.8 g/mL (for LNB- -pNP and GNB-
-pNP) or 1e2 mg/mL (for Gb5 and GA1 oligosaccharide). When the
kinetic parameters were determined, the substrate concentrations
varied from 0.3 to 2 times the respective Km values. The reaction
was stopped by adding Na2CO3 (for pNP-sugars) or heating at 95 ꢁC
for 5 min (for oligosaccharides). The amount of released p-nitro-
phenol was determined by measuring the absorbance at 405 nm.
The hydrolysis of oligosaccharides was monitored with an HPLC
system equipped with an Asahipak NH2P-50 4E column (Showa
Denko). Elution was performed at a flow rate of 1 mL/min with 73%
acetonitrile, and monitored using a charged aerosol detector
(Thermo Scientific). When globo H hexasaccharide was used as the
mg/mL (for LNB-
1. Kimura I, Ozawa K, Inoue D, Imamura T, Kimura K, Maeda T, et al. Nat Commun
2013;4:1829.
2. Maslowski KM, Vieira AT, Ng A, Kranich J, Sierro F, Yu D, et al. Nature 2009;461:
1282e6.
b
m
b
m
m
m
b
3. Furusawa Y, Obata Y, Fukuda S, Endo TA, Nakato G, Takahashi D, et al. Nature
2013;504:446e50.
b
€
4. Diaz Heijtz R, Wang S, Anuar F, Qian Y, Bjorkholm B, Samuelsson A, et al. Proc
Natl Acad Sci U S A 2011;108:3047e52.
5. Goehler LE, Park SM, Opitz N, Lyte M, Gaykema RPA. Brain Behav Immun
2008;22:354e66.
6. Hsiao EY, McBride SW, Hsien S, Sharon G, Hyde ER, McCue T, et al. Cell
2013;155:1451e63.
7. Brownawell AM, Caers W, Gibson GR, Kendall CWC, Lewis KD, Ringel Y, et al.
J Nutr 2012;142:962e74.
8. Koropatkin NM, Cameron EA, Martens EC. Nat Rev Microbiol 2012;10:323e35.
9. Martens EC, Chiang HC, Gordon JI. Cell Host Microbe 2008;4:447e57.
10. Asakuma S, Hatakeyama E, Urashima T, Yoshida E, Katayama T, Yamamoto K,
et al. J Biol Chem 2011;286:34583e92.
11. Kiyohara M, Tanigawa K, Chaiwangsri T, Katayama T, Ashida H, Yamamoto K.
Glycobiology 2011;21:437e47.
substrate, the reaction products were treated with 1,2-
dase17 prior to HPLC analysis to remove fucosyl residues from
Fuc 1-2Gal 1-3GalNAc for quantification. The standard curves
a-L-fucosi-
12. Miwa M, Horimoto T, Kiyohara M, Katayama T, Kitaoka M, Ashida H, et al.
Glycobiology 2010;20:1402e9.
13. Ashida H, Miyake A, Kiyohara M, Wada J, Yoshida E, Kumagai H, et al. Glyco-
biology 2009;19:1010e7101.
14. Bry L, Falk PG, Midtvedt T, Gordon JI. Science 1996;273:1380e3.
15. Pickard JM, Maurice CF, Kinnebrew MA, Abt MC, Schenten D, Golovkina TV,
et al. Nature 2014;514:638e41.
16. Goto Y, Obata T, Kunisawa J, Sato S, Ivanov II, Lamichhane A, et al. Science
2014;345:1254009.
a
b
were created with known concentrations of Lac and GNB. The ki-
netic parameters were calculated by curve-fitting the experimental
data to the MichaeliseMenten equation (incorporating competitive
inhibition where necessary), using GraFit software (Erithacus
Software, West Sussex, UK).

24
A. Gotoh et al. / Carbohydrate Research 408 (2015) 18e24
17. Katayama T, Sakuma A, Kimura T, Makimura Y, Hiratake J, Sakata K, et al.
J Bacteriol 2004;186:4885e93.
18. Wada J, Ando T, Kiyohara M, Ashida H, Kitaoka M, Yamaguchi M, et al. Appl
Environ Microbiol 2008;74:3996e4004.
19. Sakurama H, Kiyohara M, Wada J, Honda Y, Yamaguchi M, Fukiya S, et al. J Biol
Chem 2013;288:25194e206.
20. Sano M, Hayakawa K, Kato I. Proc Natl Acad Sci U S A 1992;89:8512e6.
21. Henrissat B, Davies G. Curr Opin Struct Biol 1997;7:637e44.
22. Yoshida-Moriguchi T, Yu L, Stalnaker SH, Davis S, Kunz S, Madson M, et al.
Science 2010;327:88e92.
33. Hakomori S. Biochim Biophys Acta 1999;1473:247e66.
34. Kitaoka M, Tian J, Nishimoto M. Appl Environ Microbiol 2005;71:3158e62.
35. Karlsson KA. Annu Rev Biochem 1989;58:309e50.
36. Heimburg-Molinaro J, Lum M, Vijay G, Jain M, Almogren A, Rittenhouse-
Olson K. Vaccine 2011;29:8802e26.
37. Shevinsky LH, Knowles BB, Damjanov I, Solter D. Cell 1982;30:697e705.
38. Wright AJ, Andrews PW. Stem Cell Res 2009;3:3e11.
39. Takahashi, Tanabe K, Ohnuki M, Narita M, Ichisaka T, Tomoda K, et al. Cell
2007;131:861e72.
40. Thomson JA. Science 1998;282:1145e7.
23. Ito T, Katayama T, Hattie M, Sakurama H, Wada J, Suzuki R, et al. J Biol Chem
2013;288:11795e806.
41. Kuroda Y, Kitada M, Wakao S, Nishikawa K, Tanimura Y, Makinoshima H, et al.
Proc Natl Acad Sci U S A 2010;107:8639e43.
24. Katayama T, Fujita K, Yamamoto K. J Biosci Bioeng 2005;99:457e65.
25. Schnaar RL, Suzuki A, Stanley P. In: Varki A, Cummings RD, Esko JD, Freeze HH,
Stanley P, Bertozzi CR, Hart GW, Etzler ME, editors. Essentials of Glycobiology. New
York, NY: Cold Spring Harbor Labortory Press, Cold Spring Harbor; 2009. p. 129e41.
26. Patterson MC. Handb Clin Neurol 2013;113:1707e8.
42. Brimble SN, Sherrer ES, Uhl EW, Wang E, Kelly S, Merrill AH, et al. Stem Cells
2007;25:54e62.
43. Krivan HC, Ginsburg V, Roberts DD. Arch Biochem Biophys 1988;260:493e6.
44. Sakurama H, Fushinobu S, Hidaka M, Yoshida E, Honda Y, Ashida H, et al. J Biol
Chem 2012;287:16709e19.
27. Thomas AS, Hughes DA. Pediatr Endocrinol Rev 2014;12(Suppl. 1):88e101.
28. Basu I, Mukhopadhyay C. Langmuir 2014;30:15244e52.
29. Krivan HC, Roberts DD, Ginsburg V. Proc Natl Acad Sci U S A 1988;85:6157e61.
30. Larson G, Falk P, Hoskins LC. J Biol Chem 1988;263:10790e1078.
31. Falk P, Hoskins LC, Larson G. J Biochem 1990;108:466e74.
32. Lee H, Garrido D, Mills DA, Barile D. Electrophoresis 2014;35:1742e50.
45. Ito M, Yamagata T. J Biol Chem 1986;261:14278e82.
46. Nishimoto M, Kitaoka M. Carbohydr Res 2009;344:2573e6.
47. Finn RD, Mistry J, Tate J, Coggill P, Heger A, Pollington JE, et al. Nucleic Acids Res
2010;38:D211e22.
48. Anderson K, Li SC, Li YT. Anal Biochem 2000;287:337e9.