Motif-guided identification of a glycoside hydrolase family 1 α-L-arabinofuranosidase in bifidobacterium adolescentis

Article abstract of DOI:10.1271/bbb.130279

Members of glycoside hydrolase family 1 (GH1) cleave glycosidic linkages with a variety of physiological roles. Here we report a unique GH1 member encoded in the genome of Bifidobacterium adolescentis ATCC 15703. This enzyme, BAD0156, was identified from over 2,000 GH1 sequences accumulated in a database by a genome mining approach based on a motif sequence. A recombinant BAD0156 protein was characterized to confirm that this enzyme alone specifically hydrolyzes p-nitrophenyl-α-L-arabinofuranoside among the 24 pnitrophenyl- glycosides examined. Among natural glycosides, α-1,5-linked arabino-oligosaccharides served as substrates, but arabinan, debranched arabinan, arabinoxylan, and arabinogalactan did not. A time course analysis of arabino-oligosaccharide hydrolysis indicated that BAD0156 is an exo-acting enzyme. These results suggest that BAD0156 is an α-L-arabinofuranosidase. This is the first report of a GH1 enzyme that acts specifically on arabinosides, providing information on GH1 substrate specificity.

Full text of DOI:10.1271/bbb.130279

Biosci. Biotechnol. Biochem., 77 (8), 1709–1714, 2013
Motif-Guided Identification of a Glycoside Hydrolase Family 1
ꢀ-L-Arabinofuranosidase in Bifidobacterium adolescentis
y
2
Hirokazu SUZUKI,1; Ayano MURAKAMI,2 and Ken-ichi YOSHIDA
1Organization of Advanced Science and Technology, Kobe University, Hyogo 657-8501, Japan
2Department of Agrobioscience, Graduate School of Agricultural Science, Kobe University, Hyogo 657-8501, Japan
Received April 2, 2013; Accepted April 23, 2013; Online Publication, August 7, 2013
[doi:10.1271/bbb.130279]
Members of glycoside hydrolase family 1 (GH1)
cleave glycosidic linkages with a variety of physiological
roles. Here we report a unique GH1 member encoded in
the genome of Bifidobacterium adolescentis ATCC
15703. This enzyme, BAD0156, was identified from over
2,000 GH1 sequences accumulated in a database by a
genome mining approach based on a motif sequence. A
recombinant BAD0156 protein was characterized to
confirm that this enzyme alone specifically hydrolyzes
p-nitrophenyl-ꢀ-L-arabinofuranoside among the 24 p-
nitrophenyl-glycosides examined. Among natural glyco-
sides, ꢀ-1,5-linked arabino-oligosaccharides served as
substrates, but arabinan, debranched arabinan, arabi-
noxylan, and arabinogalactan did not. A time course
analysis of arabino-oligosaccharide hydrolysis indicated
that BAD0156 is an exo-acting enzyme. These results
suggest that BAD0156 is an ꢀ-L-arabinofuranosidase.
This is the first report of a GH1 enzyme that acts
specifically on arabinosides, providing information on
GH1 substrate specificity.
contains an Arg residue (Fig. 1 and Supplemental
Fig. 1; see Biosci. Biotechnol. Biochem. Web site). We
also found that the DG motif sequences of characterized
GH1 members can be categorized into three main types
based on their glycone specificities: (i) DG motifs of
ꢀ-glycosidases consisting of 14 amino acids that include
Glu as a ligand-interacting residue (Supplemental
Fig. 1A); (ii) DG motifs of 6-phospho-ꢀ-galactosidases
(also having 6-phospho-ꢀ-glucosidase activity) consist-
ing of 14 amino acids that include Ser, Lys, and Tyr
residues as ligand-interacting residues (Supplemental
Fig. 1B); and (iii) DG motifs of 6-phospho-ꢀ-glucosi-
dases (with high specificity) consisting of 15 amino
acids that include Ser, Trp, Lys, and Tyr residues. These
findings suggest that DG motif sequences can be used to
predict the glycone specificities of uncharacterized GH1
members.
In this study, we analyzed the DG motif sequences of
GH1 members described in the CAZy database and
identified several GH1 members with unusual DG
motifs. Here we report that one GH1 member, Bifido-
bacterium adolescentis BAD0156, is an ꢁ-L-arabinofur-
anosidase that acts on arabino-oligosaccharides. This is
the first example of a GH1 arabinofuranosidase.
Key words: arabinofuranosidase; Bifidobacterium ado-
lescentis; arabino-oligosaccharide; glyco-
side hydrolase family 1 (GH1); genome
mining
Materials and Methods
Members of glycoside hydrolase family 1 (GH1)
catalyze the hydrolytic cleavage of glycosidic linkages
in a wide variety of glycosides, including disaccharides,
phosphorylated disaccharides, oligosaccharides, and
sugar–aromatic conjugates. The crystal structures of
GH1 members have been elucidated, revealing shared
ðꢀ=ꢁÞ8 barrel structures and interactions between amino
acid residues and ligand components within the active
site.1–8) In addition to the many GH1 enzymes that have
been identified and characterized, recent developments
in DNA sequencing technologies have revealed numer-
ous GH1 genes in various organisms. The carbohydrate-
active enzymes (CAZy) database (http://www.cazy.org)
currently describes more than 4,000 GH1 members,
although most of these are uncharacterized, and their
catalytic properties remain to be established.
Bacterial strains and culture conditions. B. adolescentis ATCC
15703 was purchased from the American Type Culture Collection
(Rockvilk, MD). Escherichia coli JM109 (Takara Bio, Otsu, Japan)
was used for DNA manipulation. E. coli BL21(DE3) harboring pG-
KJE8 (Takara Bio), which encodes protein chaperones, was used to
produce recombinant BAD0156 proteins. E. coli strains were cultured
aerobically at 37 ꢀC in Luria–Bertani (LB) medium. Ampicillin
(50 mg/L) and chloramphenicol (13 mg/L) were added as necessary.
L-Arabinose (0.5 g/L) and tetracycline (2 mg/L) were added to express
chaperone genes encoded by the pG-KJE8 plasmid.
Collection of DG motif sequences. All amino acid sequences of
GH1 members were obtained from the CAZy database (as of October
2010). Duplicate or partial GH1 sequences and sequences of putative
dTDP-4-dehydrorhamnose reductases were not included in subsequent
analysis. The sequences collected (the archaeal GH1, 61; bacterial
GH1, 1,725; and eukaryotic GH1, 346 sequences) were aligned using
the ClustalW algorithm (http://clustalw.ddbj.nig.ac.jp). Amino acid
sequences corresponding to those between Asp429 and Gly442 of a
6-phospho-ꢀ-galactosidase of Geobacillus kaustophilus HTA426
(GenBank ID, YP 149067)9) were extracted as DG motifs.
In a previous report,9) we identified a motif sequence
that governs GH1 glycone specificity. This motif, the
DG motif, is a loop comprised of a short linear amino
acid sequence flanked by Asp and Gly residues that
y
To whom correspondence should be addressed. Present address: Department of Bioscience and Biotechnology, Faculty of Agriculture, Graduate
School, Kyushu University, Fukuoka 812-8581, Japan; Fax: +81-92-642-3053; E-mail: hirokap@xpost.plala.or.jp
Abbreviations: CAZy, carbohydrate-active enzymes; GH1, glycoside hydrolase family 1; LB, Luria–Bertani; pNP, p-nitrophenyl; pNP-Araf, p-
nitrophenyl-ꢁ-L-arabinofuranoside; TLC, thin-layer chromatography

1710
H. SUZUKI et al.
Preparation of recombinant BAD0156 protein. The BAD0156 gene
(GenBank ID, YP 909019) was amplified from the B. adolescentis
chromosome using primers 50-GCCCTCGAGGTTAATCAAAAA-
GTCTCGCC-30 (XhoI site underlined and BAD0156 second codon
in italics) and 50-GCCGGATCCTTACTTACGATGGGCTTCCAC-30
(BamHI site underlined and BAD0156 stop codon in italics). Amplified
DNA was trimmed with XhoI and BamHI and cloned between the
XhoI and BamHI sites of pET-16b (Merck, Darmstadt, Germany). The
resulting plasmid was introduced into E. coli BL21(DE3) cells that
harbored pG-KJE8.
The transformants were cultured in 250 mL of LB medium
(supplemented with ampicillin, chloramphenicol, L-arabinose, and
tetracycline) until an optical density of 1.0 at 600 nm was achieved.
After the addition of isopropyl ꢀ-D-thiogalactopyranoside to a final
concentration of 0.1 mM, the culture was further incubated for 4 h.
Cells were harvested by centrifugation, suspended in 2.5 mL of buffer
A (50 mM sodium phosphate pH 7.0, 0.5 M NaCl, and 10% v=v
glycerol), and sonicated. The resulting lysates were clarified by
centrifugation and applied to a Talon-Co2þ column (0.5 mL; Clontech,
Mountain View, CA) equilibrated with buffer A. The column was
washed thoroughly with buffer
A containing 10 mM imidazole,
followed by 50 mM imidazole. Recombinant proteins were then eluted
with buffer A containing 250 mM imidazole. The buffer of the eluate
was changed to buffer B (10 mM sodium phosphate, pH 7.0), and the
preparation was applied to HiTrap Heparin (1 mL; GE Healthcare,
Little Chalfont, UK) equilibrated with buffer B at a flow rate of
¨
1 mL/min with the AKTAprime system (GE Healthcare). The column
was washed with 10 mL of buffer B, followed by 20 mL of buffer B
containing 0.2 M NaCl. Proteins were then eluted with a 0.2–1.0 M
linear gradient of NaCl in buffer B over 10 mL and by isocratic elution
with buffer B containing 1.0 M NaCl. Fractions containing pure
BAD0156 proteins were pooled and concentrated by ultrafiltration
along with buffer exchange to 10 mM sodium phosphate (pH 6.0). The
proteins were quantified by the Bradford method with a protein assay
kit (Bio-Rad, Hercules, CA) with bovine serum albumin as standard.
Fig. 1. Sequence Comparisons of DG Motifs of BAD0156 and the
Other GH1 Members Characterized (indicated by GenBank ID).
These included (i) ꢀ-glycosidases from Paenibacillus polymyxa
(AAA22263
and
AAA22264),42)
Pyrococcus
furiosus
(AAL80566),43–45) Micrococcus antarcticus (ACM66669),46) Fervi-
dobacterium islandicum (CCA60742),47) Sphingomonas paucimobi-
lis (AAG59862),48) Reticulitermes flavipes (ADK12988),49) Paeni-
bacillus sp. HC1 (BAE48718),50) Sphingomonas sp. 2F2
(ADY18331),51) Bacillus sp. GL1 (BAA36160),52) Thermoanaero-
bacter
(AAA72843),13) Thermus thermophilus (ABW87307),54) Spodoptera
frugiperda Dalbergia cochinchinensis
(AAC06038),55)
brockii
(CAA91220),53)
Sulfolobus
solfataricus
Enzyme assay for pNP-glycoside hydrolysis. The reaction mixtures
(100 mL) contained 50 mM sodium phosphate (pH 6.0), 2 mM pNP-
glycoside, and 2 mg/L of BAD0156 protein. The reactions were run at
30 ꢀC for 20 min and were terminated by adding 100 mL of 2 M sodium
carbonate. The amount of p-nitrophenol released was determined by
measuring the absorbance at 405 nm with reference to an empirically-
derived standard curve. One unit was defined as the amount of enzyme
that releases 1 nmol of p-nitrophenol in 1 min. Kinetic parameters were
determined from the initial reaction rates at substrate concentrations of
0.5–8.0 mM by Lineweaver–Burk analysis.
(AAF04007),56) Pyrococcus horikoshii (BAA29589),44) and Pyro-
coccus furiosus (AAL81332);45,57) (ii) 6-phospho-ꢀ-galactosidases
from Lactococcus lactis (AAA25183),58,59) Staphylococcus aureus
(AAA26650),59) Lactobacillus casei (AAD15134),59) and G. kaus-
tophilus (BAD77499);9) and (iii) 6-phospho-ꢀ-glucosidases from
E. coli (AAC75939 and AAC76744).59,60) Most DG motifs shared
Asp, Arg, and Gly residues (solid triangles). In addition, a Glu
residue was shared by ꢀ-glycosidases, whereas the Ser residue was
shared by 6-phospho-ꢀ-glalactosidases and 6-phospho-ꢀ-glucosides
(solid circle). 6-Phospho-ꢀ-glalactosidases and 6-phospho-ꢀ-gluco-
sides also shared Lys and Tyr residues (hollow circles).
Enzymatic characterization. The temperature profile of the pNP-
Araf hydrolyzing activity was determined by incubating the reaction
mixture at various temperatures. The pH profile was determined with
50 mM Britton–Robinson buffer10) at various pH values. The effects of
various chemicals on pNP-Araf hydrolyzing activity were evaluated by
adding a chemical to the reaction mixture. The activity without
additive chemicals was taken to be 100%. To determine temperature
and pH stabilities, the BAD0156 protein was treated for 20 min in
50 mM sodium phosphate (pH 6.0) at various temperatures and for 1 h
at 37 ꢀC in 50 mM Britton–Robinson buffer at various pH values.
Residual activity was determined with pNP-Araf as substrate.
Glycoside substrates. p-Nitrophenyl (pNP)-ꢀ-D-galactopyrano-
side-6-phosphate and pNP-ꢀ-D-glucopyranoside-6-phosphate were
synthesized as described previously.9) pNP-ꢁ-L-arabinofuranoside
(pNP-Araf), pNP-ꢁ-L-arabinopyranoside, pNP-ꢀ-L-arabinopyranoside,
pNP-N-acetyl-ꢀ-D-galactosaminide, pNP-N-acetyl-ꢀ-D-glucosaminide,
pNP-ꢀ-D-cellobioside, pNP-ꢁ-L-fucopyranoside, pNP-ꢀ-D-fucopyrano-
side, pNP-ꢀ-L-fucopyranoside, pNP-ꢁ-D-galactopyranoside, pNP-ꢀ-D-
galactopyranoside, pNP-ꢀ-D-galacturonide, pNP-ꢁ-D-glucopyranoside,
pNP-ꢀ-D-glucopyranoside, pNP-ꢀ-D-glucuronide, pNP-ꢀ-D-lactopyra-
noside, pNP-ꢀ-D-maltoside, pNP-ꢁ-D-mannopyranoside, pNP-ꢀ-D-
mannopyranoside, pNP-ꢁ-L-rhamnopyranoside, pNP-ꢁ-D-xylopyrano-
side, and pNP-ꢀ-D-xylopyranoside were purchased from Sigma
Aldrich (St. Louis, Mo). Sugar beet arabinan, debranched arabinan
prepared from sugar beet arabinan, wheat flour arabinoxylan, and
ꢁ-1,5-linked arabino-oligosaccharides (arabinobiose, arabinotriose,
arabinotetraose, and arabinopentaose) were from Megazyme Interna-
tional (Wicklow, Ireland). Larch wood arabinogalactan was from
Tokyo Chemical Industry (Tokyo, Japan). Nistose, 1-kestose, and
inulin were from Wako Pure Chemical Industries (Osaka, Japan).
Raffinose, sucrose, melibiose, and lactose were from Nakalai Tesque
(Kyoto, Japan).
Qualitative enzyme assay for natural saccharide hydrolysis. The
reaction mixtures (100 mL) contained 50 mM sodium phosphate
(pH 6.0), 1% (w=v) saccharide, and 79 mg/L of the BAD0156 protein.
The reactions were run at 30 ꢀC for 24 h, and were terminated by
heating at 100 ꢀC for 10 min. The reaction products were developed on
silica gel 60 thin-layer chromatography (TLC) F254 plates (Merck) with
1-butanol/2-propanol/water (5:3:2 by volume). The glycosides on
the plates were visualized with a diphenylamine/aniline/phosphate
reagent,11) followed by heating at 120 ꢀC for 10 min.
Quantitative enzyme assay for natural saccharide hydrolysis. The
reaction mixtures (100 mL) contained 50 mM sodium phosphate
(pH 6.0), 1% (w=v) arabino-oligosaccharide, and 180 mg/L of the
BAD0156 protein. The reactions were run at 30 ꢀC for 6 h, and were

Arabinofuranosidase of Glycoside Hydrolase Family 1
terminated by adding 100 mL of 0.8 M sodium carbonate (pH 10.5). The
1711
pNP-glycoside hydrolyzing activity
amount of glycosides released was determined with 2,20-bicinchoni-
nate.12) One unit of arabino-oligosaccharide hydrolyzing activity was
defined as the amount of enzyme that released 1 nmol of reducing
terminals in 1 min. Kinetic parameters were determined from the
initial reaction rates at substrate concentrations of 0.2–2.0 mM by
Lineweaver–Burk analysis.
The catalytic activities of the BAD0156 protein were
first analyzed using 24 pNP-glycosides as substrates.
Among these, BAD0156 hydrolyzed only pNP-Araf.
The specific activity and kinetic parameters are summa-
rized in Table 1. Other pNP-glycosides (see ‘‘Materials
and Methods’’ above), including pNP-ꢁ-L-arabinopyra-
noside and pNP-ꢀ-D-xylopyranoside, did not serve as
substrates (specific activity, <1 U/mg protein).
Results
DG motif analysis identified BAD0156 as a unique
GH1 candidate
Enzymatic characterization
To find GH1 members with unique glycone specific-
ities, we corrected all the amino acid sequences of the
GH1 members described in the CAZy database and
extracted their DG motif sequences. Among 2,132 DG
motif sequences, 1,745 sequences (82%) were catego-
rized into three main types (see the introduction). The
number of DG motif sequences flanked by Asp and Gly
residues was 2,111 (99%), but 21 sequences (1%),
including the DG motif of B. adolescents BAD0156,
were not flanked by Asp and Gly residues (Supplemental
Fig. 2). Some others comprised 13 amino acids. In view
of availability of microbe resources, we focused on
B. adolescents BAD0156 among these enzymes, and
characterized it.
The optimum temperature for pNP-Araf hydrolyzing
activity was 35 ꢀC (Fig. 3A). The enzyme was active
over a pH range of 3.5–7.5, with maximum activity at
pH 5.5 (Fig. 3B). BAD0156 completely retained its
activity after incubation for 20 min at pH 6.0 at <40 ꢀC
(Fig. 3C) and after incubation for 1 h at 37 ꢀC in a pH
range of 4.0–6.5 (Fig. 3D). Activity was partially
inhibited by the addition of 1 mM CuSO4 (relative
activity, 27%), 1 mM ZnCl2 (77%), 1 mM o-phenanthro-
line (69%), and 10% (v=v) of an alcohol (methanol,
57%; ethanol, 34%; 2-propanol, 28%; and 1-butanol,
18%). Negligible effects (90–130%) were observed for
the following chemicals (1 mM): Fe(NH4)2(SO4)2,
FeCl3, MnCl2, MgCl2, NiSO4, CaCl2, CoCl2, nitrilotri-
acetic acid, catechol-3,5-disulfonic acid, EDTA, NADþ,
NADH, NADPþ, NADPH, FAD, and dithiothreitol.
Sequence futures of BAD0156
The DG motif sequence of BAD0156 included an Asp
residue (Asp370), but no Gly or Arg residues conserved
in the usual DG motifs (Fig. 1). The entire sequence
included two Glu residues (Glu151 and Glu329) as
possible catalytically important residues13) and several
other residues that were conserved among GH1 mem-
bers as possible ligand-interacting residues (e.g.,
His108, Asn150, Asn337, Tyr339, Trp366, and
Trp374), but it did not exhibit significant similarity to
any GH1 member in the CAZy database (identity <
50%) or ꢁ-L-arabinofuranosidases that belong to other
GH families (about 20%). It appeared that the BAD0156
gene (BAD 0156) was part of an operon along with three
other genes (BAD 0153, BAD 0154, and BAD 0155),
which encode for components of a putative sugar ABC
transporter.
Natural saccharide hydrolysis
Among natural saccharides, BAD0156 hydrolyzed
arabino-oligosaccharides (arabinobiose, arabinotriose,
Preparation of the recombinant BAD0156 protein
The BAD0156 protein was produced in E. coli
BL21(DE3) cells that harbored pG-KJE8, and was
purified to homogeneity (Fig. 2). The yield was 0.7 mg
of purified protein per 250 mL of culture. When E. coli
cells without pG-KJE8 were used as host, BAD0156 was
produced in an insoluble form.
Fig. 2. Preparation of Recombinant BAD0156 Protein.
BAD0156 preparations were analyzed by SDS-poly acrylamide
gel electrophoresis after extraction (lane 1), Talon-Co2þ column
chromatography (lane 2), and HiTrap Heparin column chromatog-
raphy (lane 3). Lane M, molecular standards. Proteins were stained
with Coomassie Brilliant Blue R-250.
Table 1. Arabinoside Hydrolyzing Activities of BAD0156
Specific activitya
(U/mg protein)
ꢁ1
)
Substrate
pNP-Araf
Arabinobiose
kcat (sꢁ1
)
Km (mM)
kcat=Km (sꢁ1
M
1:6 ꢂ 103
0.6
1.7
3:8 ꢃ 0:2
NDb
ND
3:3 ꢃ 0:3
ND
ND
1:1 ꢂ 103 ꢃ 0:0 ꢂ 103
ND
Arabinotriose
Arabinotetraose
Arabinopentaose
1:3 ꢃ 0:2
3.4
1.8
7:9 ꢂ 10ꢁ3 ꢃ 2:2 ꢂ 10ꢁ3
4:2 ꢃ 1:2
1:9 ꢃ 0:1
5:8 ꢂ 10ꢁ3 ꢃ 3:9 ꢂ 10ꢁ3
5:2 ꢃ 3:7
1:1 ꢃ 0:1
a1 U was defined as the amount of enzyme that releases 1 nmol p-nitrophenol (for pNP-Araf) or 1 nmol reducing terminals (for arabino-oligosaccharides) in 1 min.
bValues could not be determined due to low activity (for arabinobiose) or high Km values (for arabinotriose).

1712
H. SUZUKI et al.
Fig. 5. Time Course Analysis of Arabino-Oligosaccharide Hydrol-
ysis.
Arabinotetraose (upper panel) and arabinopentaose (lower panel)
were incubated with BAD0156 at 30 ꢀC for the indicated durations.
Reaction products were analyzed by TLC with authentic arabino-
sides (A1, L-arabinose; A2, arabinobiose; A3, arabinotriose; A4,
arabinotetraose; and A5, arabinopentaose).
Fig. 3. Temperature (A) and pH (B) Profiles of pNP-Araf Hydro-
lyzing Activity and Temperature (C) and pH (D) Stabilities of the
BAD0156 Protein.
Discussion
Reaction mixtures were incubated at variosu temperatures (A) or
in 50 mM Britton–Robinson buffer10) at various pH values (B). The
BAD0156 protein was treated for 20 min in 50 mM sodium
phosphate (pH 6.0) at various temperatures (C) or for 1 h at 37 ꢀC
in 50 mM Britton–Robinson buffer at various pH values (D).
Residual activity was determined using pNP-Araf as substrate.
Values are means of two independent experiments. For each
analysis, the highest value was taken to be 100%.
Glycoside hydrolases are currently classified into
more than 113 families based on amino acid sequence
similarities.14) Among these families, glycoside hydro-
lases that act on ꢁ-L-arabinosides have been found in six
(GH3, 43, 51, 54, 62, and 93), and are classified into two
main types. One type, arabinan endo-1,5-ꢁ-L-arabina-
nase (EC 3.2.1.99), catalyzes the endo-hydrolysis of
(1!5)-ꢁ-arabinofuranosidic linkages in arabinans and
arabino-oligosaccharides.15–17) Enzymes of this type
generally show no activity toward pNP-Araf, as exem-
plified by Caldicellulorsiruptor saccharolyticus arabi-
nanase15) and Bacillus subtilis Abn2.16) The other type,
ꢁ-L-arabinofuranosidase (EC 3.2.1.55), catalyzes the
exo-hydrolysis of terminal non-reducing ꢁ-L-arabinofur-
anosidic linkages.18–41) Substrates of this type include
the terminal residues of 1,5-ꢁ-L-arabinan and 1,5-ꢁ-L-
arabino-oligosaccharides and branched arabinofurano-
side residues of arabinoxylan and arabinan. ꢁ-L-Arabi-
nofuranosidases generally act on pNP-Araf, although
there are exceptions.18,19,39,40)
In the present study, the DG motif sequences of all the
GH1 members described in the CAZy database were
analyzed, and we found that the motif is conserved in
almost GH1 members, while several GH1 members have
unique DG motifs. To clarify the catalytic activity of the
unique GH1 members, we characterized a unique
member, BAD0156, encoded in the B. adolescentis
genome. Among 24 pNP-glycosides and 16 natural
saccharides, the recombinant BAD0156 enzyme specif-
ically hydrolyzed pNP-Araf and ꢁ-1,5-linked arabino-
oligosaccharides. A time course analysis of arabino-
oligosaccharide hydrolysis indicated that BAD0156 was
an exo-acting enzyme. These results indicate that
BAD0156 is an ꢁ-L-arabinofuranosidase. Among the
GH1 members that have been characterized,9,13,42–60)
Sphingomonas BglSp hydrolyzes pNP-Araf as a minor
activity,51) but there is no member that acts specifically
on arabinosides. Thus BAD0156 is the first example of a
GH1 ꢁ-L-arabinofuranosidase.
Fig. 4. Natural Arabinoside Hydrolysis by BAD0156.
Arabinosides were incubated at 30 ꢀC for 24 h with (þ) and
without (ꢁ) the BAD0156 enzyme. The reaction products were
analyzed by TLC with authentic saccharides. A1, L-arabinose; A2,
arabinobiose; A3, arabinotriose; A4, arabinotetraose; A5, arabino-
pentaose; DA, debranched arabinan; AN, sugar beet arabinan; AX,
wheat flour arabinoxylan; AG, larchwood arabinogalactan.
arabinotetraose, and arabinopentaose), but did not
hydrolyze arabinan, debranched arabinan, arabinoxylan,
or arabinogalactan (Fig. 4). Kinetic analysis indicated
that arabinotetraose was the preferred substrate
(Table 1). Melibiose, lactose, and sugars containing a
D-fructofuranoside moiety (raffinose, sucrose, 1-kestose,
nistose, and inulin) did not serve as substrates (specific
activity, <0.5 U/mg protein; Supplemental Fig. 3). A
time course analysis indicated that arabinotetraose
hydrolysis produced arabinotriose first and then arabi-
nobiose, although L-arabinose production was obscure
due to a signal overlap of L-arabinose and arabinote-
traose on TLC analysis. Similarly, arabinopentaose
hydrolysis preferentially produced arabinotetraose and
then arabinotriose (Fig. 5). These results indicate that
BAD0156 acted on the terminal residues in the arabino-
oligosaccharides.
Bifidobacterium species are enteric bacteria that
degrade persistent oligosaccharides, such as arabinan,
arabinoxylan, arabinogalactan, and arabino-oligosaccha-
rides. In several bacteria, including Bifidobacterium

Arabinofuranosidase of Glycoside Hydrolase Family 1
1713
6) Gloster TM, Macdonald JM, Tarling CA, Stick RV, Withers
SG, and Davies GJ, J. Biol. Chem., 279, 49236–49242 (2004).
7) Seshadri S, Akiyama T, Opassiri R, Kuaprasert B, and Cairns
JK, Plant Physiol., 151, 47–58 (2009).
species, ꢁ-L-arabinofuranosidases belonging to other GH
families have been identified and characterized (see
above). These ꢁ-L-arabinofuranosidases, which belong to
other GH families, exhibit pH and temperature profiles
similar to those of BAD0156 (pNP-Araf hydrolyz-
ing activities: optimum pH, 4.5–6.0; optimum temper-
ature, 30–80 ꢀC). Catalytic efficiency toward pNP-Araf is
also comparable. Moreover, BAD0156 hydrolyzes
arabino-oligosaccharides, although the catalytic effi-
ciency is lower than those of other ꢁ-L-arabinofur-
anosidases.28,37,38) These observations suggest that
BAD0156 plays a role in utilizing arabino-oligosaccha-
rides and/or other arabino-containing substrates (e.g.,
partially-degraded arabinoxylan and arabinogalactan) by
B. adolescentis ATCC 15703 in combination with the
sugar ABC transporter, encoded by adjacent genes
BAD 0153–BAD 0155.
8) Nijikken Y, Tsukada T, Igarashi K, Samejima M, Wakagi T,
Shoun H, and Fushinobu S, FEBS Lett., 581, 1514–1520 (2007).
9) Suzuki H, Okazaki F, Kondo A, and Yoshida K, Appl.
Microbiol. Biotechnol., 97, 2929–2938 (2013).
10) Britton HTS and Robinson RA, J. Chem. Soc., 1456–1462
(1931).
11) Bailey RW and Bourne EJ, J. Chromatogr., 4, 206–213 (1960).
12) Zhang YH and Lynd LR, Biomacromolecules, 6, 1510–1515
(2005).
13) Moracci M, Capalbo L, Ciaramella M, and Rossi M, Protein
Eng., 9, 1191–1195 (1996).
14) Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V,
and Henrissat B, Nucleic Acids Res., 37, D233–D238 (2009).
15) Hong M, Park C, and Oh D, Biotechnol. Lett., 31, 1439–1443
(2009).
´
´
16) Inacio JM and de Sa-Nogueira I, J. Bacteriol., 190, 4272–4280
Except for the DG motif sequence and substrate
specificity, BAD0156 shares main features with other
GH1 members. BAD0156 activity was unaffected by the
addition of metal ions, vitamins, or chelating agents,
suggesting that BAD0156 requires no cofactors, sim-
ilarly to most of the other GH1 members that have been
characterized.9,13,42–45,47–60) The BAD0156 sequence
shares two glutamic acid residues (Glu151 and
Glu329) with other GH1 sequences. These might be
possible general acid/base and nucleophile residues. In
addition, the BAD0156 sequence contains several other
residues that are conserved in GH1 members. Because
the amino acid sequence similarity between BAD0156
and the other GH1 members characterized was not high
(identity < 50%), we cannot rule out the possibility that
various factors, other than DG motif structure, are
involved in BAD0156 substrate specificity, but our
results suggest that GH1 members with unusual DG
motifs have unique activities, and that a genome mining
approach based on DG motif sequences is practical to
find unique GH1 members.
(2008).
17) Sakamoto T, Inui M, Yasui K, Tokuda S, Akiyoshi M, Kobori
Y, Nakaniwa T, and Tada T, Appl. Microbiol. Biotechnol., 93,
1087–1096 (2012).
18) Lee D, Wi S, Lee Y, Cho E, Chung B, and Bae H, Mol.
Biotechnol., 49, 229–239 (2011).
19) Hashimoto K, Yoshida M, and Hasumi K, Biosci. Biotechnol.
Biochem., 75, 342–345 (2011).
20) Ichinose H, Yoshida M, Fujimoto Z, and Kaneko S, Appl.
Microbiol. Biotechnol., 80, 399–408 (2008).
21) Canakci S, Kacagan M, Inan K, Belduz AO, and Saha BC, Appl.
Microbiol. Biotechnol., 81, 61–68 (2008).
22) Ravanal MC, Callegari E, and Eyzaguirre J, Appl. Environ.
Microbiol., 76, 5247–5253 (2010).
23) Lee J, Hyun Y, and Kim D, J. Appl. Microbiol., 111, 1097–1107
(2011).
24) Raweesri P, Riangrungrojana P, and Pinphanichakarn P, Bio-
resour. Technol., 99, 8981–8986 (2008).
25) Guais O, Tourrasse O, Dourdoigne M, Parrou JL, and Francois
JM, Appl. Microbiol. Biotechnol., 87, 1007–1021 (2010).
26) Morana A, Paris O, Maurelli L, Rossi M, and Cannio R,
Extremophiles, 11, 123–132 (2007).
27) Miyanaga A, Koseki T, Miwa Y, Mese Y, Nakamura S, Kuno
A, Hirabayashi J, Matsuzawa H, Wakagi T, Shoun H, and
Fushinobu S, Biochem. J., 399, 503–511 (2006).
28) Cartmell A, McKee LS, Pen˜a MJ, Larsbrink J, Brumer H,
Kaneko S, Ichinose H, Lewis RJ, Viksø-Nielsen A, Gilbert HJ,
and Marles-Wright J, J. Biol. Chem., 286, 15483–15495 (2011).
29) Sørensen HR, Jørgensen CT, Hansen CH, Jørgensen CI,
Pedersen S, and Meyer AS, Appl. Microbiol. Biotechnol., 73,
850–861 (2006).
Acknowledgments
We thank Dr. Atsushi Iguchi for technical assistance
with bifidobacteria culture, and we thank Dr. Fumiyoshi
Okazaki and Ms. Nanami Nakashima for help with TLC
analysis. This work was supported by the program
Special Coordination Funds for Promoting Science and
Technology under the project Creation of Innovation
Centers for Advanced Interdisciplinary Research Areas
(Innovative Bioproduction Kobe) of Ministry of Educa-
tion, Culture, Sports, Science and Technology, Japan.
30) Shi P, Li N, Yang P, Wang Y, Luo H, Bai Y, and Yao B, Appl.
Biochem. Biotechnol., 162, 707–718 (2010).
31) Pei J and Shao W, Appl. Microbiol. Biotechnol., 78, 115–121
(2008).
32) Birgisson H, Fridjonsson O, Bahrani-Mougeot FK,
Hreggvidsson GO, Kristjansson JK, and Mattiasson B, Bio-
technol. Lett., 26, 1347–1351 (2004).
´
33) Margolles A and de los Reyes-Gavilan CG, Appl. Environ.
Microbiol., 69, 5096–5103 (2003).
References
34) Sakamoto T, Ogura A, Inui M, Tokuda S, Hosokawa S, Ihara H,
and Kasai N, Appl. Microbiol. Biotechnol., 90, 137–146 (2011).
35) Shin H, Park S, Sung JH, and Kim D, Appl. Environ. Microbiol.,
69, 7116–7123 (2003).
1) Wiesmann C, Hengstenberg W, and Schulz GE, J. Mol. Biol.,
269, 851–860 (1997).
2) Isorna P, Polaina J, Latorre-Garcıa L, Can˜ada FJ, Gonzalez B,
and Sanz-Aparicio J, J. Mol. Biol., 371, 1204–1218 (2007).
´
´
36) van den Broek LAM, Lloyd RM, Beldman G, Verdoes JC,
McCleary BV, and Voragen AGJ, Appl. Microbiol. Biotechnol.,
67, 641–647 (2005).
´
3) Sanz-Aparicio J, Hermoso JA, Martınez-Ripoll M, Lequerica
JL, and Polaina J, J. Mol. Biol., 275, 491–502 (1998).
4) Zechel DL, Boraston AB, Gloster T, Boraston CM, Macdonald
JM, Tilbrook DMG, Stick RV, and Davies GJ, J. Am. Chem.
Soc., 125, 14313–14323 (2003).
37) Lim Y, Yoon R, Seo E, Kim Y, Park C, and Oh D, J. Appl.
Microbiol., 109, 1188–1197 (2010).
´
38) Inacio JM, Correia IL, and de Sa-Nogueiral I, Microbiology,
´
5) Jeng W, Wang N, Lin M, Lin C, Liaw Y, Chang W,
Liu C, Liang P, and Wang AH, J. Struct. Biol., 173, 46–56
(2011).
154, 2719–2729 (2008).
39) Sakamoto T and Thibault J, Appl. Environ. Microbiol., 67,
3319–3321 (2001).

1714
H. SUZUKI et al.
40) Wong DWS, Chan VJ, and Batt SB, Appl. Microbiol.
Biotechnol., 79, 941–949 (2008).
50) Harada KM, Tanaka K, Fukuda Y, Hashimoto W, and Murata
K, Arch. Microbiol., 184, 215–224 (2005).
41) Numan MT and Bhosle NB, J. Ind. Microbiol. Biotechnol., 33,
247–260 (2006).
51) Wang L, Liu Q, Sung B, An D, Lee H, Kim S, Kim S, Lee S,
and Im W, J. Biotechnol., 156, 125–133 (2011).
52) Hashimoto W, Miki H, Nankai H, Sato N, Kawai S, and Murata
K, Arch. Biochem. Biophys., 360, 1–9 (1998).
´
42) Gonzalez-Candelas L, Aristoy MC, Polaina J, and Flors A,
Appl. Environ. Microbiol., 55, 3173–3177 (1989).
43) Kaper T, Lebbink JHG, Pouwels J, Kopp J, Schulz GE,
van der Oost J, and de Vos WM, Biochemistry, 39, 4963–4970
(2000).
53) Breves R, Bronnenmeier K, Wild N, Lottspeich F, Staudenbauer
WL, and Hoffmeister J, Appl. Environ. Microbiol., 63, 3902–
3910 (1997).
44) Kaper T, van Heusden HH, van Loo B, Vasella A, van der Oost
J, and de Vos WM, Biochemistry, 41, 4147–4155 (2002).
45) Bauer MW, Bylina EJ, Swanson RV, and Kelly RM, J. Biol.
Chem., 271, 23749–23755 (1996).
54) Gu N, Kim J, Kim H, You D, Kim H, and Jeon S, J. Biosci.
Bioeng., 107, 21–26 (2009).
55) Marana SR, Jacobs-Lorena M, Terra WR, and Ferreira C,
Biochim. Biophys. Acta, 1545, 41–52 (2001).
46) Fan H, Miao L, Liu Y, Liu H, and Liu Z, Enzyme Microb.
Technol., 49, 94–99 (2011).
56) Cairns JRK, Champattanachai V, Srisomsap C, Wittman-Liebold
B, Thiede B, and Svasti J, J. Biochem., 128, 999–1008 (2000).
57) Park SH, Park KH, Oh BC, Alli I, and Lee BH, N. Biotechnol.,
28, 639–648 (2011).
47) Jabbour D, Klippel B, and Antranikian G, Appl. Microbiol.
Biotechnol., 93, 1947–1956 (2012).
48) Marques AR, Coutinho PM, Videira P, Fialho AM, and
58) Schulte D and Hengstenberg W, Protein Eng., 13, 515–518
(2000).
´
Sa-Correia I, Biochem. J., 370, 793–804 (2003).
49) Scharf ME, Kovaleva ES, Jadhao S, Campbell JH, Buchman
GW, and Boucias DG, Insect Biochem. Mol. Biol., 40, 611–620
(2010).
59) Witt E, Frank R, and Hengstenberg W, Protein Eng., 6, 913–
920 (1993).
60) Wilson G and Fox CF, J. Biol. Chem., 249, 5586–5598 (1974).