Isolation and characterization of a hyperbranched proteoglycan from Ganoderma Lucidum for anti-diabetes

Article abstract of DOI:10.1016/j.carbpol.2014.09.051

Presently, an efficient protein tyrosine phosphatase 1B (PTP1B) inhibitor, named FYGL-n, was isolatedfrom Ganoderma Lucidum and characterized for its structure and bioactivity. Structure and chain con-formation of FYGL-n based on both chemical and spectroscopic analysis showed that FYGL-n was ahyperbranched heteropolysaccharide bonded with protein via both serine and threonine residues by O-type glycoside, and showed a sphere observed by AFM. Specifically, monosaccharide compositionindicated that FYGL-n consisted of D-arabinose, D-galactose, L-rhamnose and D-glucose in a mole ratio of 0.08:0.21:0.24:0.47, with a molecular mass of 72.9 kDa. The analysis of amino acids in FYGL-n indicatedthat there were 16 common amino acids, among which aspartic acid, glycine, serine, alanine, glutamicacid and threonine were the dominant components. Also it was demonstrated that FYGL-n could inhibitthe PTP1B activity on a competitive mechanism in vitro.

Full text of DOI:10.1016/j.carbpol.2014.09.051

Carbohydrate Polymers 117 (2015) 106–114
Contents lists available at ScienceDirect
Carbohydrate Polymers
journal homepage: www.elsevier.com/locate/carbpol
Isolation and characterization of a hyperbranched proteoglycan from
Ganoderma Lucidum for anti-diabetes
Deng Pana,1, Linqiang Wangb,1, Congheng Chena, Bingwen Hub,∗, Ping Zhoua,∗
a State Key Laboratory of Molecular Engineering of Polymers, Department of Macromolecular Science, Fudan University, Shanghai 200433, PR China
b Shanghai Key Laboratory of Magnetic Resonance, Department of Physics, East China Normal University, Shanghai 200062, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 28 June 2013
Received in revised form 25 August 2014
Accepted 18 September 2014
Available online 28 September 2014
Presently, an efficient protein tyrosine phosphatase 1B (PTP1B) inhibitor, named FYGL-n, was isolated
from Ganoderma Lucidum and characterized for its structure and bioactivity. Structure and chain con-
formation of FYGL-n based on both chemical and spectroscopic analysis showed that FYGL-n was a
hyperbranched heteropolysaccharide bonded with protein via both serine and threonine residues by
O-type glycoside, and showed a sphere observed by AFM. Specifically, monosaccharide composition
indicated that FYGL-n consisted of d-arabinose, d-galactose, l-rhamnose and d-glucose in a mole ratio of
0.08:0.21:0.24:0.47, with a molecular mass of 72.9 kDa. The analysis of amino acids in FYGL-n indicated
that there were 16 common amino acids, among which aspartic acid, glycine, serine, alanine, glutamic
acid and threonine were the dominant components. Also it was demonstrated that FYGL-n could inhibit
the PTP1B activity on a competitive mechanism in vitro.
Keywords:
Ganoderma Lucidum
Structure characterization
Hyperbranched
Proteoglycan
Polysaccharide
© 2014 Elsevier Ltd. All rights reserved.
1. Introduction
the sensitivity of insulin (Kenner, Anyanwu, Olefsky, & Kusari,
1996). Such agents provide the promise as new therapies for the
Reversibility of protein phosphorylation is a predominant
strategy to control the activity of proteins in eukaryotic cells.
Approximately 30% proteins in the typical mammalian cells are
thought to be phosphorylated (Szczepankiewicz et al., 2003).
Tyrosine phosphorylation is considered to be a key reaction in
the initiation and propagation of insulin action. Insulin exerts its
physiological effects by binding to its receptor in insulin target
tissues such as adipose, liver, and muscle. Insulin binding to the
insulin receptor induces activation of the intrinsic tyrosine kinase
in ␤-subunit, which serves as the first step in insulin signaling, and
is followed by a cascade of intracellular events, such as glucose
uptake (Kahn, 1994). There are compelling evidences that protein
tyrosine phosphatase 1B (PTP1B) is primarily responsible for the
dephosphorylation of insulin receptor, consequently, leading to
the block of insulin signaling transduction pathway. Thus, an
idea PTP1B inhibitor would be expected capable of increasing the
half-life of the phosphorylated insulin receptor and enhancing
prevention and treatment of metabolic disorders, particularly for
the treatment of diabetes which is considered to be one of the
major leading causes of death in the world and currently affects
more than 346 million people worldwide, with almost 3.4 million
mortalities (Kim, Hyun, Choung, & Choung, 2006).
Nowadays the agents used for diabetes treatment mainly are
synthetic drugs as well as insulin. However, these drugs usu-
ally come with considerable side effects, such as hypoglycemia,
drug-resistance, dropsy and weight gain (Tahrani, Piya, Kennedy,
& Barnett, 2010). Apart from those currently available therapeu-
tic options for diabetes, herbal medicines have been recommended
by World Health Organization (WHO) (Gupta & Mal, 2005). Herbal
remedies are relatively effective, cheap and almost no side effects,
compared to synthetic agents. One of examples is Ganoderma
lucidum (G. lucidum), a white rot fungus, which has been widely
used as a tonic in promoting longevity and health for thousands
of years in China. Recently, a great deal of work has been car-
ried out on this fungus, and demonstrated that some ingredients
extracted from G. lucidum can promote the release of serum insulin
and decrease the plasma glucose concentration in vivo (Huang et al.,
2010; Trajkovic, Mijatovic, Maksimoviclvanic, & Stojanovic, 2009;
Weng et al., 2009). Previously, we successfully isolated an effi-
cient PTP1B inhibitor, named FYGL, from G. lucidum fruiting bodies
(Teng et al., 2011). FYGL was an ingredient capable of decreasing
∗
Corresponding authors at: East China Normal University, Shanghai Key Labo-
ratory of Magnetic Resonance, Department of Physics, Shanghai 200062, PR China.
Tel.: +86 21 55664038; fax: +86 21 55664038.
E-mail addresses: bwhu@phy.ecnu.edu.cn (B. Hu), pingzhou@fudan.edu.cn
(P. Zhou).
1
These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.carbpol.2014.09.051
0144-8617/© 2014 Elsevier Ltd. All rights reserved.

D. Pan et al. / Carbohydrate Polymers 117 (2015) 106–114
107
the plasma glucose level and enhancing the insulin sensitivity by
2.3. General analysis
inhibiting the activity and expression of PTP1B in vivo. We have
already investigated the dominant components, inhibition kinet-
ics, pharmacology and toxicity of the efficient extract in vivo, and
demonstrated that FYGL can serve as a drug candidate or a health-
care food for the diabetic therapy or protection with high safety
(Teng et al., 2011, 2012; Wang et al., 2012; Pan et al., 2013, 2014).
It is well known that the bioactivities of natural resources are
dependent on their chemical structures and chain conformations
(Liu et al., 2010; Tong et al., 2009; Sun, Liang, Zhang, Tong, &
Liu, 2009). However, the structure information of FYGL, including
molecular structure, residue linkages, branches, and morphology,
etc., are still unclear due to the complexity of the natural resource.
Therefore, in the present work, FYGL was further fractionated and
structurally characterized for clarifying the structure–bioactivity
relationship in detail. Herein, a neutral component, named FYGL-
n, was fractionated from FYGL. The chemical structure of FYGL-n
was investigated systematically by chemical and 2D NMR spec-
troscopic methods. The chain conformation of FYGL-n was also
studied by dynamic and static light scattering method, and the
structural parameters of FYGL-n were established based on the the-
ory of polymer solution. Furthermore, FYGL-n inhibition kinetics
on PTP1B activity was also evaluated in vitro. Hopefully, this work
could provide valuable information on further understanding of the
bioactive extracts from the natural resources.
Total contents of carbohydrate and uronic acid in FYGL-n were
determined by colorimetric methods of phenol–sulphuric acid and
m-hydroxydiphenol, using glucose and glucuronic acid as refer-
ence, respectively, (Blumencrantz & Hansen, 1973). Content of
protein was determined by Lowry’s methods with bovine serum
albumin (BSA) as a standard (Bensadoun & Weinstein, 1976). UV
absorption spectrum was recorded with a UV spectrophotometer
(Model Lambda-35, Perkin Elmer, Massachusetts, USA).
2.4. Monosaccharide and amino acid component analysis
For monosaccharide analysis, FYGL-n (5 mg) was hydrolyzed
with 2 M TFA (5 mL). The hydrolyzed product was reduced with
NaBH4, followed by acidification with 40% formic acid, and then
the reduced sugars of alditols were acetylated with pyridine–acetic
anhydride to produce the alditol acetates which were ana-
lyzed by gas chromatograph (GC, Tokyo, Japan). For absolute
configuration analysis, a solution of 250 ␮L 0.625 M HCl in -(+)-
2-butanol were added into the hydrolyzed product of FYGL-n,
and kept at 80 ◦C for 16 h. The trimethylsilylated (TMS)-(+)-2-
butyl glycoside-derivatives of monosaccharide were prepared with
N,O-bis(tri-methylsilyl) trifluroacetamide (BSTFA) which were also
analyzed by GC. For amino acids contents determination, the sam-
ple was vacuum-dried and placed in a hydrolysis vessel containing
18% HCl and 1% phenol. The vessel was purged with nitrogen gas
and sealed under vacuum. The sample in the vessel was hydrolyzed
at 110 ◦C for 24 h. After hydrolysis of the sample, the vessel was
cooled and vacuum-dried to remove the residual HCl. The dried
sample was dissolved in a citrate buffer of pH 2.2 and analyzed
using a HitachiL-8500 amino acid analyzer packed with a cation
exchange resin and eluted with a series of buffers ranging from
low (0.25 M sodium citrate, pH 3.05) to high (0.25 M sodium nitrate,
pH 9.5) pH values. Detection of amino acids was carried out using
post-column derivatization with o-phthalaldehyde, a fluorescent
reagent that reacts with all of the amino acids except proline. For
praline detection, the sample was treated with sodium hypochlo-
rite, and the same procedure as that used for other amino acid
analysis was carried out.
2. Materials and methods
2.1. Materials
All the dried fruiting bodies of G. lucidum, grown in northeast-
ern China, were purchased from Leiyunshang Pharmaceutical Co.
Ltd. (Shanghai, China). Standard monosaccharide, DEAE-52 cel-
lulose, Sephadex G-75, and Sephadex G-15 were bought from
Whatman Co. (Shanghai, China). PTP1B was donated by Shanghai
Viva Biotech Co. Ltd (Shanghai, China). p-Nitrophenyl phosphate
(p-NPP), dimethyl sulfoxide (DMSO), methyl iodide (CH3I) and
methylbenzene from Sinopharm Co. Ltd. (Sichuan, China) were
redistilled to remove trace water before use. Trifluoroacetic acid
(TFA), sodium periodate (NaIO4), sodium iodate (NaIO3) and other
necessary reagents from Sinopharm Co. Ltd. (Shanghai, China) were
of analytical grade without further purification.
2.5. Methylation, periodate oxidation and Smith degradation
In order to determine the glycosyl linkages, FYGL-n (5.0 mg)
was methylated four times, using distilled DMSO and finely pow-
dered dry NaOH. The vacuum-dried FYGL-n (5 mg) was dissolved in
DMSO (1.5 mL) and then methylated with a saturated NaOH/DMSO
solution (1.5 mL) and CH3I (1 mL). The reaction mixture was
extracted with chloroform, and the organic phase was washed
with double-distilled water. Complete methylation was confirmed
by the disappearance of OH band (3200–3700 cm−1) in FTIR
spectrum. The permethylated FYGL-n was hydrolyzed, reduced
and acetylated similarly as mentioned in Section 2.5. After those
procedures, the methylated alditol acetates were re-dissolved in
chloroform and analyzed using a gas chromatograph–mass spec-
trum (GC–MS, Shimadzu GC-14A, Tokyo, Japan).
Periodate oxidation and Smith degradation were performed to
confirm the glycosidic linkages deduced from methylation analysis.
FYGL-n was oxidized with 0.015 M sodium periodates at 4 ◦C in the
dark. After the oxidation was completed, the solution was reduced,
neutralized, dialyzed and lyophilized to give the polyalcohol which
was then hydrolyzed and acetylated for GC analysis.
2.2. Preparation of FYGL-n
FYGL preparation procedures were based on a previous work
(Teng et al., 2011). Briefly, after the dried fruiting bodies of G.
lucidum were milled, defatted with boiling ethanol and decocted
with boiling water, the solid residue was extracted with ammo-
nia aqueous solution, and then the supernatant was neutralized,
concentrated, dialyzed and lyophilized successively, and the crude
extract was collected. Subsequently, the crude extract was further
purified using Sephadex G-75 column (l.d. 80 × 26 cm) chromatog-
raphy with 0.5 M NaCl solution as the eluent. The eluted fractions
were monitored by the phenol–sulfuric acid method with ultravio-
let (UV) absorption at wavelength of 490 nm, and the main fraction,
named as FYGL, was obtained.
In the present work, FYGL was further fractionated using DEAE-
52 column (l.d. 60 × 40 cm) with 0–0.3 M NaCl as eluent in a
stepwise gradient manner. A neutral fraction, named as FYGL-n,
was isolated from FYGL, and treated by Sevage agent (Staub, 1965)
to remove free protein, and then dialyzed and lyophilized. The
isolation process was carried out several times to obtain enough
amounts for the analysis.
GC and GC–MS experiments were performed on a Shimadzu
GC-14A gas chromatograph equipped with a DB-5MS column
(30 m × 0.25 mm × 0.25 ␮m, Tokyo, Japan) and a flame ionization
detector, using high purity of nitrogen as the carrier gas at a flow

108
D. Pan et al. / Carbohydrate Polymers 117 (2015) 106–114
rate of 8 cm3/min. The temperature was heated up from 180 to
240 ◦C at 10 ◦C/min, and kept at 240 ◦C for 25 min. The vaporizer
temperature was 260 ◦C and the detector temperature was 240 ◦C.
A liquot (15 ␮L) was injected for each run.
where kh and kk are Huggins and Kraemer coefficients, ꢀsp is the
reduced specific viscosity, and ꢀr is the relative viscosity. The
intrinsic viscosity [ꢀ] is obtained from the intercept of both linear
functions of ꢀsp/c vs. c and ln ꢀr/c vs. c.
2.10. Determination of molecular structural parameters
2.6. Polysaccharide and protein linkages analysis
For determination of molecular parameters of FYGL-n, the multi-
angle static laser-light scattering measurements (MALLS) were
performed at 25 ◦C with the multi-angles at 30, 45, 60, 75, 90, 105,
120, 135, 150◦. Weight averaged molecular mass (Mw) and gyra-
tion radius (Rg) for the studied polymers in dilute solution were
calculated on the Zimm’s method based on Rayleigh–Gans–Debye
theory as Eq. (3)
The linkages between polysaccharide and protein in FYGL-n
were analyzed by the ␤-elimination reaction. FYGL-n were dis-
solved in 28% ammonium hydroxide aqueous solution saturated
with ammonium carbonate at 60 ◦C for 40 h, and subsequently
scanned by a UV spectrophotometer from 220 to 400 nm. Further-
more, 0.5 M boric acid was added into the reaction mixture which
was then incubated at 37 ◦C for 30 min. After those procedures,
the reaction product was desalted by Sephadex G-15 column (l.d.
5 cm × 1 cm), and subjected to the amino acid analysis. The linkage
bonds between polysaccharide and protein in FYGL-n were deter-
mined by comparison with the amino acid contents in FYGL-n and
the ␤-eliminated sample.
ꢀ
ꢁ ꢂꢃ
Kc
Rꢁ
1
Mw
16ꢂ2ꢀ2
3ꢃ2
ꢁ
=
1 +
Rg2 sin2
+ 2A2c
2
(3)
ꢄ
ꢅ
2
4ꢂ2n2 dn
K =
NAꢃ4
dc
where NA, n, ꢁ, c, and ꢃ are Avogadro number, solvent refractive
index, observed angle, polymer concentration, and light wave-
length in vacuum, respectively. Additionally, the refractive index
increment (dn/dc) of the FYGL-n in distilled water was measured
by a double-beam differential refractometer. A plot of Kc/Rꢁ→0 vs.
c and Kc/Rc→0 vs. sin2(ꢁ/2) can be used to determine the parameter
of Mw and Rg, respectively.
2.7. NMR characterization
All NMR experiments were recorded at 298 K on a Varian
NMR System 700 spectrometer using a 5 mm [1H, 13C, 15N]-
triple resonance probe, with 1H and 13C ␲/2 pulses of 9 and
12.0 ␮s, respectively. 1H NMR of FYGL-n was acquired by sup-
pressing H2O signal using WET (Water suppression Enhanced by
T1 effects) pulse sequence (Smallcombe, Patt, & Keifer, 1995).
13C NMR and 2D NMR spectra (COSY, TOCSY, NOESY, HSQC and
HMBC) of FYGL-n in D2O solution were recorded at 298 K using TSP
(3-(trimethylsilyl)-propionic acid-d4 sodium salt) as an external
standard. The relaxation delay for the NMR experiments was 1.5 s.
The NOESY spectrum was recorded with a mixing time of 100 ms
and TOCSY spectrum with a mixing time of 110 ms.
Hydrodynamic radiuses (Rh) were measured by a Zetasizer
nanoparticle analyzer (Malvern Instruments Ltd., Malvern, U.K.)
using dynamic light scattering (DLS) technique.
2.11. Morphology measurement
AFM images of FYGL-n were recorded on an atomic force micro-
scope (Digital Instruments, USA). A drop of 10 ␮L FYGL-n solution
at concentration of 1 ␮g/mL in distilled water was dropped onto
the freshly cleaved mica substrate and dried for 1.5 h. The freshly
prepared sample was mounted on AFM stage and imaged under
atmosphere (25 ◦C, relative humidity of 40–50%) using tapping
model. The image was recorded at 256 × 256 pixels.
2.8. Fractionation of FYGL-n on different molecular weight
To determine the molecular parameters, it is necessary to frac-
tionate FYGL-n on different molecular weight by the non-solvent
addition method. Briefly, FYGL-n was dissolved in distilled water,
and then methanol was slowly added into the solution drop-wise
with stirring at room temperature until the solution turned tur-
bid. The turbid solution was then heated to 50 ◦C and kept for
2 h, and the turbidity was somewhat decreased. After cooled down
to 25 ◦C and kept for 2 h, the turbid mixture was centrifuged at
10,000 rpm for 30 min and the precipitate was washed several
times by methanol, resulting in the first fraction. Similarly, the
supernatant was subjected to the next fractionation by adding more
volumes of methanol. Seven fractions, designated as F1, F2, F3, F4,
F5, F6 and F7 were thus obtained. Each fraction was dialyzed and
lyophilized.
2.12. PTP1B inhibition and its kinetics measurement
The inhibitory potency of FYGL-n on the PTP1B activity was
evaluated according to the method described in the literature
(Szczepankiewicz et al., 2003). Briefly, 50 ␮L 1 mM p-nitrophenyl
phosphates (pNPP) reacted with 50 ␮L 100 ␮g/mL PTP1B in a tris-
NaCl buffer solution (pH 8.0) with or without inhibitor (FYGL-n) at
37 ◦C for 30 min, and the enzyme reaction was terminated by 3 M
NaOH. The amount of produced p-nitrophenol was measured by
UV absorbance at wavelength of 405 nm. The IC50 values referred
to the concentration of the inhibitor required to decrease the initial
PTP1B activity by 50% were used to evaluate the inhibitory potency
of the inhibitor.
2.9. Measurement of intrinsic viscosity
For determining the inhibition kinetics, FYGL-n aqueous solution
at different concentrations of 0.0 (control), 50.0, 75.0 ␮g/mL were
individually added to the reaction mixtures of 50 ␮L, 100 ␮g/mL
PTP1B enzyme and 1 mM pNPP. The inhibition kinetics mechanism
was determined using the Lineweaver–Burk plot.
The intrinsic viscosities [ꢀ] of FYGL-n and its fractions F1 to F7
in distilled water and DMSO were measured at 25 ◦C by Ubbelohde
capillary viscometer. The effluent time of the solvent was longer
than 120 s, so that the kinetic energy correction was negligible.
Huggins and Kraemer equations (1) and (2) were used to estimate
the [ꢀ] value.
3. Results and discussion
ꢀsp
2c
(1)
(2)
3.1. General properties of FYGL-n
= ꢀ + k
ꢀ
[ ] h[ ]
c
ꢀr
c
FYGL-n was light yellow powder. The analysis of total sugar
and protein contents indicated that FYGL-n was composed of
ln
= ꢀ + k
[ ] k[ ]
ꢀ
2c

D. Pan et al. / Carbohydrate Polymers 117 (2015) 106–114
109
Table 1
Monosaccharide and amino acid components in FYGL-n .
Components
Monosaccharid
content (%)
Components
Amino acid
contents (%)
Components
Amino acid
contents (%)
Components
Amino acid
contents (%)
Glucose
47.0
8.0
24.0
21.0
Asp
Thr
Ser
Glu
Phe
Pro
14.40
8.98
10.20
9.04
3.58
5.84
Gly
Ala
Val
Ile
Leu
Tyr
12.80
9.88
7.81
5.00
7.54
0.43
Arg
Lys
His
Cys
Met
1.20
1.54
1.21
0.39
0.15
Arabinose
Rhamnose
Galactose
Asp, aspartic acid; Asn, asparagine; Gly, glycine; Glu, glutamic acid; Gln, glutamine; Ala, alanine; Ser, serine; Thr, threonine; Leu, lecine; Phe, phenylalanine; Val, valine; Pro,
proline; Ile, isoleucine;Tyr, tyrosine; Arg, arginine; Lys,lsine; His, histidine; Cys, cysteine; Met, methionine.
approximate 82 2% polysaccharide and 12 2% protein. More-
over, no uronic acid was detected in FYGL-n. Therefore, it was
suggested that FYGL-n was a proteoglycan, i.e. the proteins in
FYGL-n were covalently bound to the polysaccharide since the free
proteins were removed by the Sevage method.
(2.34%), 3,4-Me2-Rhap (4.96%), 2,3,4-Me3-Glcp (28.8%). Thus, the
degree of branching (DB) could be calculated on the following equa-
tion:
NT + NB
DB =
(4)
NT + NB + NL
For the monosaccharide analysis, FYGL-n was hydrolyzed,
reduced and acetylated into alditol acetate derivatives. Four peaks
are detected at 6.26, 7.25, 8.01 and 8.11 min in GC spectrum (Data
in Supporting information Fig. S1). By comparison with monosac-
charide references, it is concluded that FYGL-n is composed of
arabinose, galactose, rhamnose and glucose, with a ratio of 0.08:
0.21: 0.24: 0.47 (Table 1). For absolute configuration analysis,
FYGL-n was hydrolyzed and derivatizated into TMS-derivatives,
using -(+)-2-butanol and N,O-bis(tri-methylsilyl) trifluroacetamide
(BSTFA). By comparison with those TMS-derivatives of standard
monosaccharides, it can be concluded that the configurations of all
glycosyl residues in FYGL-n are in d-enantiomers except rhamnose
in l-enantiomer. The analysis of amino acids in FYGL-n indi-
cated that there were 16 natural amino acids including aspartic
acid/asparagine (Asx, 14.40%), glycine (Gly, 12.80%), serine (Ser,
10.20%), alanine (Ala, 9.88%), glutamic acid/glutamine (Glx, 9.04%),
threonine (Thr, 8.98%), valine (Val, 7.81%), leucine (Leu, 7.54%), pro-
line (Pro, 5.84%), isoleucine (Ile, 5.00%), phenylalanine (Phe, 3.58%),
lysine (Lys,1.54%), histidine (His, 1.21%), arginine (Arg, 1.20%), tyro-
sine (Tyr, 0.43%), cysteine (Cys, 0.39%) and methionine (Met, 0.15%).
The high contents of serine (10.2%) and threonine (8.98%) implied
the possible existence of O-glycosidic linkage between polysaccha-
ride and protein in FYGL-n (Chen, Xie, Nie, Li, & Wang, 2008).
where NT, NB, and NL are the total contents of the terminal residues
(15.6%), branched residues (45.3%), and backbone residues (39.1%),
respectively. The DB value of FYGL-n was thus calculated to be 0.61,
implying a highly branched biomacromolecule (Aulenta & Hayes,
2003).
Periodate oxidation and Smith degradation of FYGL-n were
performed to confirm the glycosidic linkages deduced from
the methylation analysis. The residue products from periodate
oxidation and Smith degradation of FYGL-n contained glycerol,
rhamnose, erythrose, glucose and galactose in the GC spectrum
(Supporting information Fig. S3). Large amount of the glycerol
production implied that there might exist many glycosyl linkages,
such as 1→, 1 → 2, 1 → 6 or 1 → 2,6, which could be attributed
to the residues of 2,3,4,6-Me4-Glcp, 2,3,4-Me3-Galp, 2,3,4,6-Me4-
Galp, 3,4-Me2-Rhap, and 2,3,5-Me3-Araf, as characterized by
methylation analysis (Liu & Sun, 2011). Similarly, the production
of erythrose indicated that there might exist glycosyl linkage
of 1 → 4 or 1 → 4,6 which could be assigned to 2,3,6-Me3-Glcp
(Askera & Shawky, 2010). The presences of glucose, galactose and
rhamnose suggested that there were part of those monosaccharide
linked in the form of 1 → 3, 1 → 2,3, 1 → 2,4, 1 → 3,4, 1 → 3,6 or
1 → 2,3,4, which might be attributed to 2,4,6-Me3-Glcp, 3,6-Me2-
Galp, 2,4-Me2-Galp or 6-Me-Glcp, respectively (Chandra, Ghosh,
Ojha, & Islam, 2009). The results of periodate oxidation and Smith
degradation well agreed with those of methylation analysis.
Linkage bonds between polysaccharide moiety and protein one
in the proteoglycan can be classified into two types on the basis
of their stability against alkali, one is O-glycosidic linkage, and
the other is N-glycosidic linkage (Tao, Zhang, & Zhang, 2009). The
alkali-sensitive O-glycosidic linkages (involving serine and threon-
ine residues) can be easily broken in relatively mild condition by
␤-elimination reaction resulting in the release of the carbohydrate
moiety, and the serine and threonine residues linked to hydroxyl of
3.2. Linkage features of FYGL-n
As demonstrated by methylation analysis, there were at least
eleven types of linkages present in FYGL-n (Table 2 and sup-
porting information Fig. S2). Specifically, the analytical results
revealed that FYGL-n contained terminal residues of 2,3,4,6-Me4-
Glcp (8.18%), 2,3,4,6-Me4-Galp (4.49%), 2,3,5-Me3-Araf (2.93%),
branched residues of 2,4-Me2-Galp (11.15%), 3,6-Me2-Glalp
(9.38%), 6-Me-Glcp (19.1%), 3-Me-Rhap (5.69%), and remaining
backbone residues of 2,3,6-Me3-Glcp (2.97%), 2,3,4-Me3-Galp
Table 2
GC–MS analysis of methylated FYGL-n.
Peak no. in GC
MS fragments (m/z)
Methylated monosaccharide
Glycosidic linkage
Mole ratio (%)
1
2
3
4
5
6
7
8
43,87,101,117,129,161
2,3,6-Me3-Glcp
2,3,4-Me3-Galp
2,3,4,6-Me4-Glcp
2,3,4,6-Me4-Galp
3,4-Me2-Rhap
2,4-Me2-Galp
3,6-Me2-Galp
2,3,5-Me3-Araf
6-Me-Glcp
→4)-Glcp-(1→
→6)-Galp-(1→
Glcp-1→
2.97
2.34
8.18
4.49
4.96
11.15
9.38
2.93
19.1
5.69
28.8
43,58,71,87,101,117,129,161,189,233
43,101,71,75,88,99,143,175,102,222,115
43,59,71,87,101,117,129,145,161,205
43,131,129,89,87,117,189,99,113
43,117,129,189,87,159,233
Galp-1→
→2)-Rhap-(1→
→3,6)-Galp-(1→
→2,4)-Galp-(1→
Araf-(1→
43,117,190,233
117,129,101,45,161,87,71,145
43,73,85,103,115,127,145,175,187,217
43,129,143,87,117,189,101,203
58,97,87,129,143,59,71,75,85,171,103,111,145,203
9
10
11
→2,3,4)-Glcp-(1→
→2,4)-Rhap-(1→
→6)-Glcp-(1
3-Me-Rhap
2,3,4-Me3-Glcp

110
D. Pan et al. / Carbohydrate Polymers 117 (2015) 106–114
respectively. Meanwhile the residue A had an anomeric chemical
3
shifts larger than ıH 5.0, and very small JH-1
,
value (<3 Hz),
H-2
suggesting the residue A being ␣-linked (Elizabeth, 1995). Based
on the 1H-13C cross peak at ıH/C 1.25/16.48 observed in HSQC spec-
trum (Fig. 2a), which was attributed to the methyl group (Marvelys,
Maritza, Lilian, & Julio, 2006), it was concluded that residue A was ␣-
l-rhamnose residues. Combined with proton signals from COSY and
TOCSY spectra, the carbon signals in residue A were easily assigned
from HSQC spectrum (Fig. 2a) as C-1 of ıC 106.85, C-2 of ıC 82.45,
C-3 of ıC 68.63, C-4 of ıC 70.43, C-5 of ıC 58.95 and C-6 of ıC 16.48.
The relatively downfield shifts of C-1 and C-2, referred to typical
monosaccharide, indicated that those carbon atoms were substi-
tuted. Additionally, both 13C chemical shifts of C-3 and C-5 smaller
than 80 ppm indicated that the residue A was pyranose instead of
furanose. Conclusively, residue A was suggested to be 1,2-linked
␣-l-rhamnose, identical to the result from chemical analysis. Sim-
ilarly, according to NMR characteristics for FYGL-n as well as the
standard monosaccharide, the residues from B to J were suggested
to be 1,2,4-linked ␣-l-rhamnose, 1,3-linked ␣-d-glucose, 1-linked
␣-d-glucose, 1,2,3,4-linked ␣-d-glucose, 1,6-linked ␤-d-galactose,
1,2,4-linked ␣-d-galactose, 1,3,6-linked ␤-d-galactose, 1-linked ␤-
d-galactose and 1,6-linked ␤-d-glucose, respectively. The complete
carbon and proton signals of residues B to J were assigned similarly
as that of residue A and summarized in Table 3. Furthermore, the
signals of ␣, ␤ and ␥ protons and carbons of the linked serine and
threonine residues were also assigned in Table 3, according to the
typical amino acid NMR.
Fig. 1. 1H NMR (a) and 13C NMR (b) spectra of FYGL-n in D2O solution; ıH 4.65 is
attributed to the trace H2O.
1H 13C HMBC NMR spectrum, which could provide valuable
information of 13C 1H three-bond coupling within intra-residue
and the connection between residues through the glycosidic link-
ages, was applied to figure out the complete sequence of the
the polysaccharide would be transformed to ␣-amino acrylic and ␣-
amino butenoic acid, respectively, which have strong absorbance
at 245 nm in UV spectrum. Furthermore, when the ␤-eliminated
products were reduced and hydrolyzed, the contents of serine and
threonine in the studied sample would be decreased while those of
alanine and ␣-amino butyric acid would increase. UV spectrum of
the alkali-treated FYGL-n (supporting information Fig. S4) showed
a strong absorbance at 245 nm, implying that the ␤-elimination
reaction took place and the ␣-amino acrylic or ␣-amino butenoic
acid were formed. Moreover, after the ␤-eliminated products were
reduced and hydrolyzed, the serine and threonine contents were
drastically decreased from 8.98% and 10.2% to 5.00% and 6.40%,
respectively, while the alanine content was increased from 9.88%
to 12.1% (Supporting information Fig. S5 and T1). The results con-
firmed that the polysaccharides in FYGL-n were covalently linked to
the protein in O-linkage type via both serine and threonine residues.
residues in FYGL-n. Amounts of 1
H
13C cross-peaks were observed
in HMBC spectrum of FYGL-n (Fig. 2d). For instance, 1H 13C
cross-peak at ıH/C 5.85/69.41, denoted as (1) in HMBC spectrum,
suggested the residue A of 1,2-linked ␣-l-rhamnose connect-
ing the residue F of 1,6-linked ␤-d-glucose via 1 → 6 glycosidic
linkage. Additionally, 1H 13C cross-peaks at ıH/C 4.37/75.88 and
3.96/68.701, denoted as (14) and (15) in HMBC spectrum (Fig. 2d),
respectively, suggested that the residues of threonine and ser-
ine were connected to the residues of 1,3-linked ␣-d-glucose and
1,6-linked ␤-d-glucose, identical to the conclusion in which the
polysaccharides in FYGL-n were covalently linked to the protein
in O-linkage type via both serine and threonine residues. Simi-
larly, according to 1H 13C cross-peaks in HMBC spectrum as well
as NOESY spectrum (Supporting information Fig. S6), the connec-
tions in the repeated units in FYGL-n were assigned systematically
and summarized in Table 4. Therefore, the deduced hyperbranched
chain sequence of FYGL-n was determined as Fig. 3 (detailed struc-
ture analysis in the Supporting information).
3.3. NMR characterization of FYGL-n
The chemical structure of FYGL-n was also characterized by one-
and two-dimensional NMR. As shown in 1H NMR (Fig. 1a), 1H sig-
nals at ıH 5.85, 5.75, 5.34, 5.09, 5.03, 4.99, 4.94, 4.65 and 4.39 ppm
in the anomeric area further confirmed the heteropolysaccharide
of FYGL-n. As shown in 13C NMR (Fig. 1b), seven 13C signals at ıC
106.92, 106.85, 106.13, 99.68, 98.30, 98.04 and 91.97 ppm were
observed in the anomeric area, also suggesting a heteropolysaccha-
ride of FYGL-n. 13C NMR signals at ıC 170–175 ppm were assigned
to the carbonyl groups, indicating the existence of protein (Chen
et al., 2008). 1H 13C cross-peaks at ıH/C 5.85/106.85, 4.99/99.68,
5.34/98.30, 5.09/91.79, 5.75/106.92, 4.39/102.50, 5.03/106.13,
4.94/98.04, 4.65/102.44 and 4.64/102.44 ppm, designated as A,
B, C, D, E, F, G, H, I and J residues, observed in 2D HSQC NMR
spectrum (Fig. 2a) indicated that the signals from carbons and
protons in anomeric region were corresponded in each group pair,
respectively.
3.4. Molecular structural parameters of FYGL-n
According to the chemical structure as in Fig. 3, FYGL-n was con-
sidered to be highly branched. Thus, the molecular parameters,
including intrinsic viscosity ([ꢀ]), gyration radius (Rg), hydro-
dynamic radius (Rh) and molecular mass (Mw) of FYGL-n in aqueous
solution were also investigated using viscosity and light scattering
methods. Good linear correlations of the plots of ꢀsp/c vs. c and
lnꢀr/c vs. c for FYGL-n in both distilled water and DMSO within
concentration of 3–8 mg/mL at 25 ◦C (data in Supporting informa-
tion Fig. S7a) suggested that FYGL-n was a neutral macromolecule
instead of polyelectrolyte (Esquenet, Terech, Boué, & Buhler, 2004).
By extrapolation of those two linear curves to the infinite dilution
concentration, [ꢀ] value of FYGL-n in distilled water and DMSO were
determined as 20.6 mL/g. It was noted that there was no significant
difference of [ꢀ] of FYGL-n in both DMSO and water, implying that
1H 1H cross-peaks at ıH/H 5.85/3.88, 3.88/4.08, 4.08/3.71,
3.71/4.23, 4.23/1.25 detected in 1H 1H COSY and 1H 1H TOCSY
spectra (Fig. 2b and c) indicated that the signals of ıH 5.85, 3.88,
4.08, 3.71, 4.23 and 1.25 corresponded to H-1 to H-6 of residue A,

D. Pan et al. / Carbohydrate Polymers 117 (2015) 106–114
111
Fig. 2. 2D NMR spectra of anomeric groups in FYGL-n in D2O. (a) 1
H
13C HSQC; (b) 1
H
13C HMBC; (c)1H 1H COSY and (d) 1
H
1H TOCSY NMR.
Table 3
Complete assignments (ppm) of 1H and 13C NMR of FYGL-n.
No.
Residues
ıH/C
ıH/C
ıH/C
ıH/C
ıH/C
ıH/C
A
B
C
D
E
F
G
H
I
J
→2)-␣-l-Rhap-(1→
→2,4)-␣-l-Rhap-(1→
→4)-␣-d-Glcp-(1→
␣-d-Glcp-1→
5.85/106.85
4.99/99.68
5.34/98.30
3.88/82.45
3.78/75.34
3.29/72.83
3.48/70.54
3.92/79.07
3.18/72.93
3.73/75.36
3.50/72.57
3.26/72.96
3.26/72.96
4.37/73.38 (␤)
3.96/69.80 (␤)
4.28/68.63
3.62/71.26
3.65/70.51
3.68/71.36
3.24/73.13
3.37/72.82
4.07/70.99
3.24/73.02
3.55/72.81
3.55/73.21
1.23/17.0 (␥)
/
3.71/70.43
3.08/82.21
3.97/75.88
3.97/70.51
3.55/80.64
3.62/73.01
3.34/75.73
3.37/71.62
3.77/73.05
3.77/71.61
/
4.23/58.95
4.23/88.95
3.61/70.85
3.60/70.85
3.31/72.50
3.51/70.67
3.62/72.97
3.62/72.97
3.34/69.49
3.73/70.64
/
1.25/16.48
1.26/16.48
3.78/60.66
3.77/60.66
3.60/60.59
3.24/69.42
3.37/61.08
3.34/75.73
3.58/62.64
3.52/68.70
/
5.09/91.79
→2,3,4)-␣-d-Glcp-(1→
→6)-␤-d-Galp-1→
→2,4)-␣-d-Galp-(1→
→3,6)-␤-d-Galp-(1→
␤-d-Galp-1→
5.75/106.92
4.39/102.50
5.03/106.13
4.94/98.04
4.65/102.44
4.65/102.44
4.32/61.88 (␣)
4.50/58.35 (␣)
→6)-␤-d-Glcp-(1→
Threonine
Serine
Thr
Ser
/
/
/
Table 4
13C 1H connections of FYGL-n from HMBC NMR.
No.
Residues
ıH/C
13C 1H Connections
A
B
C
D
E
F
→2)-␣-l-Rhap-(1→
→2,4)-␣-l-Rhap-(1→
→4)-␣-d-Glcp-(1→
␣-d-Glcp-(1→
(1) 5.85/69.41
(2) 4.99/79.08
(3) 5.34/76.08
(4) 5.09/75.38
(5) 5.75/69.34
(6) 4.39/68.6
(7) 4.39/80.63
(8) 4.39/69.42
(9) 5.03/82.23
(10) 5.03/68.72
(11) 4.94/73.13
(12) 4.65/75.27
(13) 4.65/82.45
(14) 4,37,75.88
(15) 3.96,68.70
→2)-␣-l-Rhap-(1→6)-␤-d-Galp-1→
→2,4)-␣-l-Rhap-(1 → 4)-␣-d-Glcp-(1,2,3→
→4)-␣-d-Glcp-(1 → 4)-␣-d-Galp-(1,2→
␣-d-Glcp-(1 → 2)-␣-d-Galp-(1,4→
→2,3,4)-␣-d-Glcp-(1 → 6)-␤-d-Galp-(1,3→
→6)-␤-d-Galp-(1 → 6)-␤-d-Glcp-(1→
→6)-␤-d-Galp-(1 → 4)-␣-d-Glcp-(1,2,3→
→6)-␤-d-Galp-(1 → 6)-␤-d-Galp-(1→
→2,4)-␣-d-Galp-(1 → 4)-␣-l-Rhap-(1,2→
→2,4)-␣-d-Galp-(1 → 6)-␤-d-Glcp-(1→
→3,6)-␤-d-Galp-(1 → 3)-␣-d-Glcp-(1,2,4→
␤-d-Galp-(1 → 3)-␤-d-Galp-(1,6→
→6)-␤-d-Glcp-(1 → 2)-␣-l-Rhap-(1→
Thr → 1)-␣-d-Glcp-(3→
→2,3,4)-␣-d-Glcp-(1→
→6)-␤-d-Galp-(1→
G
→2,4)-␣-d-Galp-(1→
H
I
J
Thr
Ser
→3,6)-␤-d-Galp-(1→
␤-d-Galp-(1→
→6)-␤-d-Glcp-(1→
Threonine
Serine
Ser → 6)-␤-d-Glcp-(1→

112
D. Pan et al. / Carbohydrate Polymers 117 (2015) 106–114
Fig. 3. Structure of FYGL-n based on chemical and NMR analysis. Rs represent the carbohydrate residues of →2,4)-␣-l-Rhap-(1→, →6)-␤-d-Galp-1→, Araf-(1→ or →3,6)-␤-
d-Galp-(1→.
where ˚0 is a constant of 2.87 × 1023 g/mL (Zimm and Kilb, 1959).
FYGL-n molecule was an isolated polymer in both DMSO and water
1/4
1/2
within such concentration range. Similarly, [ꢀ] values of F1 to F5
in distilled water were also obtained (Supporting information Figs.
S7b, S7c and T2).
The weight averaged molecular mass (Mw), the hydrodynamic
radius (Rh) and the gyration radius (Rg) of FYGL-n and its frac-
tions F1 to F7 were measured by dynamic and static laser-light
scattering technology, indicating that FYGL-n was a macromolecule
with weight averaged molecular mass (Mw) of 72.9 kDa (Supporting
information Fig. S8 and T2).
Thus, a plot of [ꢀ]/Mw
vs. Mw
(plot in Supporting information
Fig. S8b) was established on the following Zimm–Kilb equation:
ꢀ
[ ]
= 0.73 + 0.0024 Mw1/2
(8)
1
⁄4
Mw
The unperturbed dimension of A value for FYGL-n was thus cal-
´
˚
culated from Eq. (8) to be 1.36 0.15 A which was close to that
(1.48 A) of the hyperbranched glycoprotein extracted from sclero-
˚
tia of regium reported by Tao et al. (Tao, & Yan et al., 2007a; Tao, &
Zhang et al., 2007b). Usually, the unperturbed dimension of a poly-
mer is affected by the bond angle and the steric hindrance from both
backbone and side chain. In general, a hyperbranched polymer has
a Gaussian behavior with large unperturbed dimension because of
its segments crowding each other in the vicinity of branching point
(Tao & Xu, 2008). The unperturbed dimension value of FYGL-n much
larger than that of the reported linear polymers was identical to the
hyperbranched structure of FYGL-n, as shown in Fig. 3.
It is well-known that, for a given polymer solution, the ratio
of the gyration radius to hydrodynamic radius, ꢄ = Rg/Rh, depends
on the architecture and conformation of polymer chains. It has
been reported that the ꢄ value for an extended stiff chain or rod
like chain is higher than 2.0, such as the comb-branched polysac-
charide, named AF1, extracted from Auricularia auricula-judae by
Xu et al. with a ꢄ value of 2.3 in water (Xu, Xu, & Zhang, 2012).
The ꢄ value from 1.5 to 1.7 indicates a flexible linear polymer in
a good solvent; whereas the ꢄ value of 0.775 indicates a compact
sphere (Nichifor, Lopes, Carpov, & Melo, 1999). For a hyperbranched
structures, its theoretical ꢄ parameter was predicted as 1.22 on
the hydrodynamic interaction according to the Kirkwood–Riseman
pre-average approximation (Ioan, Aberle, & Burchard, 2000). The ꢄ
parameter of FYGL-n was calculated to be 1.27 0.06 by averaging ꢄ
values of F1 to F7 (Supporting information Fig. S10 and T2), similar
to that the reported hyperbranched polymers (Tao et al., 2007a,b;
Lyulin, Adolf, & Davies, 2001; Burchard, Schmidt, & Stockmayer,
1980), suggesting that FYGL-n exhibits a hyperbranched macro-
molecule, probably as extendedly spherical particles in the aqueous
solutions.
The intrinsic viscosity of a polymer in solution is related to Mw
on the Mark–Houwink equation (5):
ꢀ = kM˛
(5)
[ ]
w
The power value (˛) of Mw in Mark–Houwink equation is
dependent on shape and nature of the macromolecule in solution.
Generally, ˛ value of 0–0.5 suggests the shape of polymer being a
dense sphere, and the value from 0.6 to 0.8 being a flexible chain,
and the value greater than 1 being an elongated rod, while the value
of 0 being a typical hard sphere (Ma, Zhang, Nishiyama, Marais, &
Vignon, 2011). Therefore, the Mark–Houwink function was estab-
lished by fitting the curve of lg[ꢀ] vs. lg Mw (plot in Supporting
information Fig. S9a) as Eq. (6)
0.02
ꢀ = 0.178 0.12 M0.42
(6)
[ ]
(
)
w
The ˛ value of 0.42 0.02 suggests a sphere-like structure for
FYGL-n in aqueous solution, i.e., a highly branched structure, iden-
tical to the results deduced from degree of branching (DB) and NMR
analysis as above.
Moreover, the intrinsic viscosity can also be applied to evalu-
ate the unperturbed dimension A (A = (ꢀr2ꢁ0/Mw)1/2, where ꢀr2ꢁ0 is
mean-square end-to-end distance) of a polymer by the intercept of
plot of [ꢀ]/Mw1/4 vs. Mw1/2 on Zimm and Kilb theory as Eq. (7) which
is more applicable for the hyperbranched polymer characterization.
[ꢀ]
Mw1/4
= K0 + BM1/2
3.5. Molecular morphology
w
ꢁ
ꢂ
(7)
3/2
Molecular parameters of FYGL-n deduced from the polymer
solution theory indicated that FYGL-n exist as a spherical chain
conformation in aqueous solution. AFM is a powerful tool in
ꢀr2ꢁ0
K0 = ˚0
MW

D. Pan et al. / Carbohydrate Polymers 117 (2015) 106–114
113
bonding, consequently keeping the WPD loop in open status, there-
fore deactivating the PTP1B. The detailed interaction mechanism
of FYGL-n with PTP1B is still under investigation in our lab.
4. Conclusion
In summary, a highly water-soluble proteoglycan, FYGL-n, was
isolated from G. lucidum. The present results demonstrated that
FYGL-n was a hyperbranched heteropolysaccharide bonded with
protein via both serine and threonine residues in O-type glycoside,
and looked like a sphere in aqueous solution. The hyperbranched
chain structure of FYGL-n may play special roles for its bioactivities
of PTP1B inhibition and antihyperglycemic potency.
Acknowledgments
This work was supported by the NSFC (nos. 21074025,
21374022), Scientific Program of Shanghai Municipal Public Health
Bureau (no. 2010231), Open Program of Shanghai Key Laboratory of
Magnetic Resonance (no. 20101206), Science and Technology Inno-
vation Program of k cjx.f (no. 11DZ1971802), Innovation Program
of Shanghai Municipal Education Commission (no. 2012Z102460)
and The Soft Science Program of Jing-An district in Shanghai (no.
2008RZ002). We also thank Dr. Min Xiao in Tennessee University
for his assistance in article refinement
Fig. 4. AFM image of FYGL-n in distilled water at 25 ◦C.
investigating long chain-like molecule, such as polysaccharide, on
the nanometer scale. Thus, to provide direct evidence of the chain
conformation, AFM was used to observe the morphology in the
present work. As shown in Fig. 4, the AFM image exhibits obviously
globular particles in diameter within 150–200 nm, indicating that
the FYGL-n molecules are dispersed individually in the extremely
dilute solution. The globular particle sizes of FYGL-n in AFM image
were identical to the hydrodynamic diameter (Dh) distribution
measured by dynamic light scattering (Supporting information Fig.
S11), supporting the hyperbranched structure of FYGL-n based on
the polymer solution theory.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.carbpol.
2014.09.051.
References
3.6. Inhibitory potency and interaction of FYGL-n on PTP1B
Askera, M. M. S., & Shawky, B. T. (2010). Structural characterization and antioxidant
activity of an extracellular polysaccharide isolated from Brevibacterium otitidis
BTS 44. Food Chemistry, 123, 315–320.
Aulenta, F., & Hayes, W. (2003). Dendrimers: A new class of nanoscopic containers
and delivery devices. European Polymer Journal, 39, 1741–1771.
Barford, D., Flint, A. J., & Tonks, N. K. (1994). Crystal structure of human protein
tyrosine phosphatase 1B. Science, 263, 1397–1404.
The inhibitory potency of FYGL-n on PTP1B activity in vitro was
also investigated. IC50 value of FYGL-n on PTP1B was determined to
be 7.8 0.2 ␮g/mL. Meanwhile, the Lineweaver–Burk plots demon-
strated that the inhibition mechanism of FYGL-n on PTP1B was
typically competitive (Supporting information Fig. S12).
Bensadoun, A., & Weinstein, D. (1976). Assay of proteins in the presence of interfering
materials. Analytical Biochemistry, 70, 241–250.
PTP1B is a ca. 50 kDa protein with highly polar phosphatase
active sites surrounding protein surface, which is difficult to bind
the lipophilic compound. PTP1B displays a substrate preference for
phosphotyrosine flanked in the N-terminus by the acidic residues
which positive active sites are likely to attract negatively charged
substrates (Park, Bhattarai, Ham, & Cho, 2009). However, the
significant negative charge substrates or agents are not always cell
compatible in the drug development. Structure of PTP1B contains
highly conserved WPD (tryptophan, proline, aspartic acid residues)
loop and pTyr (tyrosine) loop (Barford, Flint, & Tonks, 1994). The
WPD loop is flexible and in an ‘open’ conformation in the absence
of its substrate, while closes once the pTyr loop binds the substrate.
The key step during closure required for catalysis is the formation of
hydrogen bond between the residue of Trp179 in the WPD loop and
the residue of Arg221 in the pTyr loop (Kamerlin, Rucker, & Boresch,
2007). The extensive inter- and intra-molecular hydrogen bond
donors and acceptors present in the hyperbranched biopolymer of
FYGL-n. Some of the amino acid residues locating on the surface of
FYGL-n might allow FYGL-n binding to the active sites in PTP1B com-
petitively through hydrogen bonding and electrostatic interaction
to prevent the interaction of PTP1B with its substrate. Meanwhile,
the hyperbranched structure could maintain and stabilize the
sphere-like conformation and its function (Seiler, Kohler, & Arlt,
2002). Thus, we preliminarily speculate that FYGL-n could block
the pTyr loop in PTP1B binding to its substrate through hydrogen
Blumencrantz, N., & Hansen, G. A. (1973). New method for quantitative determina-
tion of uronic acids. Analytical Biochemistry, 54, 484–489.
Burchard, W., Schmidt, M., & Stockmayer, W. H. (1980). Information on polydis-
persity and branching from combined quasi-elastic and intergrated scattering.
Macromolecules, 13, 1265–1272.
Chandra, K., Ghosh, K., Ojha, A. K., & Islam, S. S. (2009). Chemical analysis of a
polysaccharide of unripe (green) tomato (Lycopersicon esculentum). Carbohy-
drate Research, 344(16), 2188–3194.
Chen, Y., Xie, M. Y., Nie, S. P., Li, C., & Wang, Y. X. (2008). Purification, composition
analysis and antioxidant activity of a polysaccharide from the fruiting bodies of
Ganoderma atrum. Food Chemistry, 107, 231–241.
Elizabeth, F. H. (1995). 1H NMR in the structural and conformational analysis of
oligosaccharides and glycoconjugates. Progress in Nuclear Magnetic Resonance
Spectroscopy, 27, 445–474.
Esquenet, C., Terech, P., Boué, F., & Buhler, E. (2004). Structural and rheological
properties of hydrophobically modified polysaccharide associative networks.
Langmuir, 20, 3583–3592.
Gupta, S., & Mal, M. (2005). Evaluation of hyperglycemia potential of S. xanthocarpum
fruit in normal and streptozin induced diabetic rats. European Bulletin Drugs
Research, 13, 51–55.
Huang, C. Y., Chen, J. Y. F., Wu, J. E., Pu, Y. E., Liu, G. Y., & Pan, M. H. (2010). Ling-Zhi
polysaccharides potentiates cytotoxic effects of anti-cancer drugs against drug-
resistant urothelial carcinoma cells. Journal of Agricultural Food Chemistry, 58,
8798–8805.
Ioan, C. E., Aberle, T., & Burchard, W. (2000). Structure properties of dextran. 2. Dilute
solution. Macromolecules, 33, 5730–5739.
Kahn, C. (1994). Insulin action, diabetogenes, and the cause of type II diabetes.
Diabetes, 43, 1066–1084.
Kamerlin, S. C. L., Rucker, R., & Boresch, S. (2007). A molecular dynamics study of
WPD-loop flexibility in PTP1B. Biochemical and Biophysical Research Communi-
cations, 356, 1011–1016.

114
D. Pan et al. / Carbohydrate Polymers 117 (2015) 106–114
Kenner, K. A., Anyanwu, E., Olefsky, J. M., & Kusari, J. (1996). Protein–tyrosine
phosphatase 1B is a negative regulator of insulin- and insulin-like growth factor-
I-stimulated signaling. Journal of Biological Chemistry, 271, 19810–19816.
Kim, S. H., Hyun, S. H., Choung, S. Y., & Choung, S. Y. (2006). Anti-diabetic effect of
cinnamon extract on blood glucose in db/db mice. Journal of Ethnopharmacology,
104, 119–123.
Liu, C., Lu, J., Lu, L., Liu, Y., Wang, F., & Xiao, M. (2010). Isolation, structural characteri-
zation and immunological activity of anexopolysaccharide produced by Bacillus
licheniformis 8-37-0-1. Bioresource Technology, 101, 5528–5533.
Szczepankiewicz, B. G., Liu, G., Hajduk, P. J., Cele, A. Z., Pei, Z., Xin, Z., et al. (2003).
Discovery of a potent, selective protein tyrosine phosphatase 1B inhibitor
using a linked-fragment strategy. Journal of American Chemical Society, 125,
4087–4096.
Tahrani, A. A., Piya, M. K., Kennedy, A., & Barnett, A. H. (2010). Glycogenic control in
type 2 diabetes: Targets and new therapies. Pharmacology & Therapeutics, 125,
328–336.
Tao, Y., & Xu, W. (2008). Microwave-assisted solubilization and solution properties
of hyperbranched polysaccharide. Carbohydrate Research, 344, 3071–3078.
Tao, Y., Yan, F., & Zhang, L. (2007). Fractionation and characterization of a protein-
polysaccharide complex from Pleurotus tuberregium sclerotia. Journal of Polymer
Science: Polymer Physics, 45, 2546–2554.
Liu, J., & Sun, Y. (2011). Structural analysis of an alkali-extractable and water-soluble
polysaccharide (ABP-AW1) from the fruiting bodies of Agaricus blazei Murill.
Carbohydrate Polymers, 86, 429–432.
Lyulin, A. V., Adolf, D. B., & Davies, G. R. (2001). Computer simulations of hyper-
branched polymers in shear flows. Macromolecules, 34, 3783–3789.
Ma, Z. C., Zhang, L., Nishiyama, Y., Marais, M. F., & Vignon, M. (2011). The molecu-
lar structure and solution conformation of an acidic heteropolysaccharide from
Auricularia auricula-judae. Biopolymers, 95, 217–227.
Marvelys, L., Maritza, M., Lilian, S., & Julio, H. (2006). Structural elucidation of the
polysaccharide from Sterculia apetala gum by a combination of chemical meth-
ods and NMR spectroscopy. Food Hydrocolloids, 20, 908–913.
Tao, Y., Zhang, L., Yan, F., & Wu, X. (2007). Chain conformation of water-insoluble
hyperbranched polysaccharide from fungus. Biomacromolecules, 8, 2321–2328.
Tao, Y., Zhang, Y., & Zhang, L. (2009). Chemical modification and antitumor activ-
ities of two polysaccharide–protein complexes from Pleurotus tuber-regium.
International Journal of Biological Macromolecules, 45, 109–115.
Teng, B. S., Wang, C. D., Yang, H. J., Wu, J. S., Zhang, D., Zheng Fan, Z. H., et al. (2011).
A protein tyrosine phosphatase 1B activity inhibitor from the fruiting bodies of
Ganoderma lucidum (Fr.) Karst and Its hypoglycemic potency on streptozotocin-
induced type 2 diabetic mice. Journal of Agricultural and Food Chemistry, 59,
6492–6500.
Teng, B. S., Wang, C. D., Zhang, D., Wu, J. S., Pan, D., Pan, L., et al. (2012). Hypo-
glycemic effect and mechanism of a proteoglycan from Ganoderma Lucidum on
streptozotocin-induced type 2 diabetic rats. European Review for Medical and
Pharmacological Sciences, 16, 166–175.
Tong, H., Xia, F., Feng, K., Sun, G., Gao, X., Sun, L., et al. (2009). Structural characteriza-
tion and in vitro antitumor activity of a novel polysaccharide isolated from the
fruiting bodies of Pleurotus ostreatus. Bioresource Technology, 100, 1682–1686.
Trajkovic, L. M. H., Mijatovic, S. A., Maksimoviclvanic, D. D., & Stojanovic, M. B. (2009).
Tufegdzic, S. J. Anticancer properties of Ganoderma lucidum methanol extracts
in vitro and in vivo. Nutrition and Cancer, 61, 696–707.
Nichifor, M., Lopes, A., Carpov, A.,
& Melo, E. (1999). Aggregation in water
of dextran hydrophobically modified with bile acids. Macromolecules, 32,
7078–7085.
Pan, D., Zhang, D., Wu, J., Chen, C., Xu, Z., Yang, H., et al. (2014). A Novel proteoglycan
from Ganoderma Lucidum fruiting bodies protects kidney function and amelio-
rates diabetic nephropathy via its antioxidant activity in C57BL/6 db/db Mice.
Food and Chemical Toxicology, 63, 111–118.
Pan, D., Zhang, D., Wu, J., Chen, C., Xu, Z., Yang, H., et al. (2013). Antidiabetic, antihy-
perlipidemic and antioxidant activities of a novel proteoglycan from Ganoderma
Lucidum fruiting bodies on db/db mice and the possible mechanism. PLoS One,
8(7), e68332.
Park, H., Bhattarai, B. R., Ham, S. W., & Cho, H. (2009). Structure-based virtual
screening approach to identify novel classes of PTP1B inhibitors. The European
Journal of Medicinal Chemistry, 44, 3280–3284.
Seiler, M., Kohler, D., & Arlt, W. (2002). Hyperbranched polymers: new selective
solvents for extractive distillation and solvent extraction. Separation and Purifi-
cation Technology, 29, 245–263.
Smallcombe, S. H., Patt, S. L., & Keifer, P. A. (1995). WET solvent suppression and
its applications to LC NMR and high-resolution NMR spectroscopy. Journal of
Magnetic Resonance, 117A, 295–303.
Weng, C. J., Chau, C. F., Yen, G. C., Liao, J. W., Chen, D. H., & Chen, K. D. (2009). Inhibitory
effects of Ganoderma lucidum on tumorigenesis and metastasis of human hep-
atoma cells in cells and animal models. Journal of Agricultural and Food Chemistry,
57, 5049–5057.
Wang, C. D., Teng, B. S., He, Y. M., Wu, J. S., Pan, D., Pan, L., et al. (2012). Effect of a
novel proteoglycan PTP1B inhibitor from Ganoderma Lucidum on amelioration
of hyperglycemia and dyslipidemia in db/db mice. British Journal of Nutrition,
108, 2014–2025.
Staub, M. A. (1965). Removal of proteins: Sevag method. Methods in Carbohydrate
Chemistry, 5, 5–6.
Sun, Y., Liang, H., Zhang, X., Tong, H., & Liu, J. (2009). Structural elucidation and
immunological activity of apolysaccharide from the fruiting body of Armillaria
mellea. Bioresource Technology, 100, 1860–1863.
Xu, S., Xu, X., & Zhang, L. (2012). Branching structure and chain conformation of
water-soluble glucan extracted from Auricularia auricula-judae. Journal of Agri-
cultural and Food Chemistry, 60, 3498–3506.
Zimm, B. H., & Kilb, R. W. (1959). Dynamics of branched polymer molecules in dilute
solution. Journal of Polymer Science, 37, 19–42.