Side-chain structure of cell surface polysaccharide, mannan, affects hypocholesterolemic activity of yeast

Article abstract of DOI:10.1021/jf900347q

We previously reported that Kluyveromyces marxianus YIT 8292 exhibited more potent hypocholesterolemic activity than other yeasts containing Saccharomyces cerevisiae. To clarify the reason for the higher hypocholesterolemic activity, we examined the side-

Full text of DOI:10.1021/jf900347q

J. Agric. Food Chem. 2009, 57, 8003–8009 8003
DOI:10.1021/jf900347q
Side-Chain Structure of Cell Surface Polysaccharide, Mannan,
Affects Hypocholesterolemic Activity of Yeast
Y
K
ASUTO
Y
OSHIDA,* EIICHIRO
IMURA, WAKAE
ASAHIKO
N
Y
I
AITO, HARUMI
OKOI, TADASHI
TO AND ARUJI
M
S
IZUKOSHI, YOKO
ATO, TAKEKAZU
AWADA
W
ATANABE
,
AZUMASA
K
O
KUMURA
,
M
,
H
S
Yakult Central Institute for Microbiological Research, Yaho 1796, Kunitachi-shi, Tokyo 186-8650, Japan
We previously reported that Kluyveromyces marxianus YIT 8292 exhibited more potent hypocho-
lesterolemic activity than other yeasts containing Saccharomyces cerevisiae. To clarify the reason
for the higher hypocholesterolemic activity, we examined the side-chain structure of cell surface
polysaccharide, mannan, of K. marxianus YIT 8292. The result shows that K. marxianus YIT 8292
had shorter R-(1,2)-linked oligomannosyl side chains and lower phosphate content in mannan than
S. cerevisiae. The association between its structural features and hypocholesterolemic activity was
investigated by comparing the hypocholesterolemic activities of S. cerevisiae mannan mutants in
rats fed a high-cholesterol diet. S. cerevisiae mnn5 mutant with deficiencies in the phosphorylation
and elongation of mannan side chains showed higher hypocholesterolemic activity than the wild-
type strain. These results show that the side-chain length and phosphate contents of mannan affect
hypocholesterolemic activity.
KEYWORDS: Kluyveromyces marxianus; mannan; side chain; cell wall; cholesterol
INTRODUCTION
LDL-cholesterol levels in hypercholesterolemic subjects at doses
of 3.0 and 4.0 g/day (10).
Yeasts have diverse fermentation activities and therefore have
been used for the production of many kinds of fermented foods
such as bread, beer, wine, and fermented milk (1). Dried yeasts
(usually heat inactivated) have also been used as a nutritional
supplementbecause wholeyeast cells are richinvitamin B, dietary
fiber, and protein (1). Three species of yeast, Saccharomyces
cerevisiae, Candida utiliz, and Kluyveromyces marxianus, are
widely used as dried yeasts (1). Recently, the beneficial effects
of dried yeasts on health have been investigated (2-11). In our
previous study comparing the hypocholesterolemic activities of
81 yeast strains (8), we found that the hypocholesterolemic
activities of the yeasts varied remarkably between strains, with
K. marxianus YIT 8292 exhibiting more potent hypocholestero-
lemic activity than the other yeasts containing S. cerevisiae
(baker’s yeast and brewer’s yeast). We have also found that cell
wall polysaccharides are the major active components of K.
marxianus YIT 8292 (9). Rats fed cell wall polysaccharide-
enriched fractions had increased fecal sterol excretion and cecal
concentration of short-chain fatty acids (SCFA) (9). Thus, the cell
wall polysaccharides of K. marxianus YIT 8292 can exert hypo-
cholesterolemic activity through the suppression of intestinal
cholesterol absorption, the interruption of the enterohepatic
circulation of bile acids, and/or the production of SCFA. Our
previous paper has shown that daily intake of K. marxianus YIT
8292 crude cell wall fraction decreases serum total cholesterol and
The cell wall of yeast is mainly composed of dietary fiber such
as β-glucan and mannan (12). Mannoprotein, a mannan-rich
glycoprotein, surrounds the inner layer consisting of β-(1,3)
glucan, β-(1,6) glucan, and a small amount of chitin. It has been
reported that the proportions of β-glucan and mannan within the
cell wall and the structures of those polysaccharides differ
according to the strain (13-15). We therefore suggest the possi-
bility that the structural and/or compositional differences of cell
wall polysaccharides among yeast strains are involved in the
variation of hypocholesterolemic activities.
The structural differences in the outer chain of N-linked
mannan among other cell wall polysaccharides are well charac-
terized and utilized for the serological classification of yeasts. The
mannan structures of several yeasts such as S. cerevisiae,
S. italicus, C. albicans, C. kefyr, and K. lactis have been identi-
fied (13, 14). A long branched R-(1,6)-linked mannose backbone
structure has been found in a wide variety of yeast species,
suggesting it is a general feature of most yeasts. These chains
are usually branched with R-(1,2)-linked mannoses, which are
then extended with a variety of genus- or species-specific linkages.
Mannan of K. lactis is closely analogous to that of S. cerevisiae,
except that it lacks phosphorylated side chains and instead
possesses N-acetylglucosamine substituents on some of the man-
nnotetrarose units (13,16). Ballou documented that K. marxianus
and K. dobzhanskii have R-mannan structurally similar to that of
K. lactis (13). On the other hand, Kobayasi showed that C. kefyr
(the imperfect state of K. marxianus), strain IFO 0586, has a
comblike structure consisting of an R-(l,6)-linked backbone and
*Author to whom correspondence should be addressed (telephone
þ 81 42 577 8968; fax þ 81 42 577 3020; e-mail yasuto-yoshida@
yakult.co.jp).
© 2009 American Chemical Society
Published on Web 08/11/2009
pubs.acs.org/JAFC

8004 J. Agric. Food Chem., Vol. 57, No. 17, 2009
Yoshida et al.
short R-(l,2) branches (17,18). Its side-chain structure differs from
that of K. lactis and S. cerevisiae by a lack of phosphate and
terminal R-(1,3)-linked mannose units.
Table 1. Composition of Crude Cell Wall Fractions Isolated from S. cerevisiae
X2180-1A, Its mnn Mutants, and K. marxianus YIT 8292a
S. cerevisiae
Ballou and colleagues isolated and characterized several man-
nan mutants of S. cerevisiae X2180-1A to investigate the mechan-
ism of mannan synthesis (13,14,19). The mnn4 and mnn1 mutants
lack the mannosyl-phosphate side chain and the terminal R-(1,3)-
mannose residues of the outer chain portion, respectively. The
mnn5 mutant is defective in the addition of the second R-(l,2)-
linked mannose. These mutants can be used to study the effects of
the side-chain lengths and phosphate groups of mannan on
function.
In the present study, we analyzed the mannan structure of
K. marxianus YIT 8292. The association between its structural
features and hypocholesterolemic activity was investigated by
comparing the hypocholesterolemic activities of S. cerevisiae
mannan mutants in rats fed a high-cholesterol diet.
wild typeb mnn4 mnn1 mnn4 mnn5
KM
dietary fiber (%)
crude protein (%)
crude fat (%)
44.1
49.5
41.4
52.0
43.7
49.7
41.1
52.5
46.6
44.2
2.82
2.89
2.90
2.82
4.04
ash (%)
other componentsc (%)
1.69
1.88
1.73
1.93
1.74
1.94
1.69
1.88
3.09
2.13
a Each result is expressed as the mean value of two replicates. b Wild type,
S. cerevisiae X2180-1A; KM, K. marxianus YIT 8292. c Components excluding
dietary fiber, protein, fat, and ash.
and lard, respectively. Their moisture, ash, and other components replaced
sucrose.
The rats were given a high-cholesterol diet with or without 3.5% crude
cell wall fraction isolated from S. cerevisiae, the mutants of S. cerevisiae
mnn1, mnn1 mnn4, mnn5, or K. marxianus YIT 8292 ad libitum for 14 days.
Food intake and body weight were measured three times or once a week,
respectively. The rats were then anesthetized with an intraperitoneal
injection of sodium pentobarbital (Nembutal, Abbot Laboratories,
Chicago, IL) at 10 a.m. on the last day of the experimental period, and
nonfasting blood samples were collected from the abdominal aorta. The
blood from each rat was put into a plastic tube containing heparin, and the
plasma was separated by centrifugation. Each liver was perfused with
0.9% NaCl, removed, and weighed. Liver lipids were extracted according
to the method of Folch et al. (21).
MATERIALS AND METHODS
Yeast and Cultivation. S. cerevisiaeX2180-1A (MATR SUC2mal mel
gal2 CUP1) and its mannan mutants mnn4 (LB6-5D MATR mnn4-1 SUC2
mal CUP1), mnn1 mnn4 (LB 97-3C MATR mnn1 mnn4), and mnn5 (LB65-
5D MATR mnn5) were obtained from the American Type Culture
Collection (ATCC). The other strains were obtained from the culture
collection of the Yakult Central Institute for Microbiological Research
(Tokyo, Japan).
All strains were grown in liquid culture medium containing 0.5% yeast
extract, 1% polypeptone, 3% glucose, 0.2% KH2PO4, 0.5% K2HPO4, and
General Analytical Procedures. The total carbohydrate con-
tent of mannan was measured by using the phenol-sulfuric acid meth-
od (22). Sugar compositions of mannan and crude cell wall fraction
were analyzed by high-performance liquid chromatography (HPLC)
after 1 h of 4 M trifluoroacetic acid (TFA) hydrolysis at 80 °C and
methanolysis (5% HCl-methanol, 100 °C for 2 h) combined with 4 M
TFA hydrolysis (80 °C, 1 h), respectively, followed by labeling with
1-phenyl-3-methyl-5-pyrazolone (PMP) (23). Total phosphate was deter-
mined according to the method of Ames and Dubin with KH2PO4 as the
standard.
Approximately 50 mg of mannan was acetolyzed as described pre-
viously (24). A 10:10:1 (v/v/v) mixture of (CH3CO)2O, CH3COOH, and
H2SO4 was used for preferential cleavage of the R-(1,6) linkage. The
O-acetylated mannooligosaccharide mixture was extracted from the reac-
tion mixture with CHCl3 and deacetylated in 0.2 M sodium methoxide in
methanol. One-tenth of the resultant mannooligosaccharide mixture was
pyridylaminated and then analyzed by HPLC with a size exclusion
chromatographic column (TSKgel R-2500, Toso, Tokyo, Japan) (25).
The other part of the acetolysis product was applied to a column packed
with Bio-Gel P2 (Bio-Rad). The biose and triose fractions were isolated
and then identified by NMR analysis.
1H and 13C NMR (DEPT 135) spectra were recorded with a FT-NMR
spectrometer (ECA-500, JEOL, Tokyo, Japan) for solutions in D2O at
30 °C with acetone as the internal standard. 2D-NMR (HSQC, HMQC,
HMBC, COSY) spectra were measured with a FT-NMR spectrometer
(ECA-500, JEOL) for solutions in D2O at 60 °C with acetone as the
internal standard.
The methylation of mannan was performed according to the method
of Ciucanu and Kerek (26). The methylated product was hydrolyzed with
4 M TFA at 121 °C for 1 h, reduced with NaBH, and acetylated. The alditol
acetates of methylated sugars were identified by GC-MS, performed on a
Hewlett-Packard 6890A gas chromatograph connected to a JMS-700 V
0.05% MgSO4 7H2O (8). Cultures were grown aerobically at 30 °C for
3
24 h and harvested by centrifugation.
Crude Cell Wall Fraction and Mannan Preparation. Crude cell
wall fraction was prepared as described previously (9). Mannan was
extracted by autoclaving at 120 °C for 90 min in 0.1 M citrate buffer at
pH 7.0 and was further purified by selective precipitation of mannoprotein
with borate and hexadecyltrimethylammonium bromide (Cetavlon) (14).
Cell Surface Hydrophobicity (CSH) Assay. CSH of crude cell wall
was determined by hydrophobic interaction chromatography as described
previously (20). Yeast crude cell wall was washed twice with 100 mM
sodium acetate buffer (pH 4.2) and resuspended in the same buffer to a
final concentration of 15 mg/mL. Phenyl Sepharose 6 Fast Flow gel (GE
Healthcare, Piscataway, NJ) was packed in a disposable chromatography
column (0.6 cm ꢀ 3.7 cm) (Biospin empty column; Bio-Rad, Richmond,
CA) to a volume of 1.0 mL and equilibrated with 100 mM sodium acetate
buffer (pH 5.0) containing 125 mM NaCl. Yeast cell suspensions (50 μL)
were then applied to the column and eluted by gravity flow with 3 mL
of buffer containing NaCl. The OD660 of the eluent was measured,
and CSH was determined by using the following equation: CSH (%) =
100(Aapplied - Aeluent)/Aapplied, where Aapplied is the OD660 of 0.1 mL
of the cell suspension diluted with 3 mL of elution buffer and Aeluent is the
OD660 of the eluent.
Animals and Diets. Five-week-old male Wistar rats were obtained
from Clea Japan, Inc. (Tokyo, Japan). All animal experiments were
performed in accordance with the guidelines of the Ethical Committee
for Animal Experiments of Yakult Central Institute for Microbiological
Research (Tokyo, Japan). The rats were housed individually in stainless
steel cages and kept in an isolated room at a controlled temperature (24 (
1 °C) and humidity (60 ( 5%). Lighting was maintained on a 12 h light-
dark cycle (lights on from 8 a.m. to 8 p.m.). The rats were fed a commercial
solid diet (MF; Oriental Yeast Co. Ltd., Tokyo, Japan) during the
acclimatization period of 7 days. After this period, they were randomly
divided into groups (n = 8/group) and fed the powdered diets described
next. All treatment groups started with similar mean body weights ranging
from 176 to 179 g.
The high-cholesterol diet consisted of (g/kg) casein, 223; sucrose, 578;
mineral mix, 40; vitamin mix, 10; soy oil, 10; lard, 100; choline bitartrate,
1.5; cholesterol, 5.0; sodium cholate, 2.5; and cellulose, 30 (9). Each cell
wall fraction was added to the high-cholesterol diet at a concentration of
3.5% (w/w). To equalize the contents of fiber, protein, and fat in each diet,
the constituents of the cell wall fraction (Table 1) replaced cellulose, casein,
mass spectrometer (JEOL). Gas chromatography of O-methyl-O-acetyl-D-
mannitols was performed using a SE52 glass capillary column (30 m ꢀ
0.25 mm i.d.) at 200 °C using He as the carrier gas at a flow rate of
0.7 mL/min. The MS detection conditions were as follows: interface
temperature, 270 °C; ionization mode, EIþ; electron energy, 70 eV.
Total cholesterol, triglycerides, and phospholipids in the plasma and
liver were enzymatically measured on two replicates with commercial
kits (Determiner TC555, Kyowa Medics, Tokyo, Japan; Triglyceride
E test, Wako Pure Chemical Industries, Tokyo, Japan; and Phospholipid
B test, Wako Pure Chemical Industries). HDL-cholesterol was measured

Article
J. Agric. Food Chem., Vol. 57, No. 17, 2009 8005
RESULTS AND DISCUSSION
with a commercial kit (Determiner HDL, Kyowa Medics, Tokyo,
Japan). Dietary fiber, crude protein, crude fat, and ash were measured
according to the methods of the Association of Official Analytical
Chemists (27).
Analysis of Mannan Structure Isolated from K. marxianus YIT
8292. The results of compositional analysis of K. marxianus YIT
8292 mannan are given in Table 2. It contained a large amount of
carbohydrate, 98% of which was mannose. The phosphate
content was 50 μmol/g, which is lower than that of S. cerevisiae
X2180-1A (230 μmol/g) or baker’s yeast (S. cerevisiae YIT 10025)
(200 μmol/g).
Statistical Analysis. Data are presented as mean ( SD. Statistical
significance was determined using the Tukey test, and p < 0.05 was
considered to be statistically significant.
Table 2. Composition of K. marxianus YIT 8292 Mannana
K. marxianus YIT 8292 mannan
1H NMR, 13C NMR DEPT 135, HSQC, HMBC, and COSY
NMR spectra of intact K. marxianus YIT 8292 mannan are
shown in Figure 1. Two major signals were detected in the
chemical composition
1
total carbohydrate (g/g)
total phosphate (μmol/g)
sugar composition
mannose (%)
0.73 ( 0.02
50 ( 2
anomeric region at approximately 5.09 and 5.01 ppm in the H
NMR and were labeled H1A and H1B, respectively. Two major
anomeric carbon signals, 102.4 and 98.5 ppm, were assigned as
C1B and C1A from cross peaks at 5.01/102.4 and 5.09/98.5 ppm,
respectively, in the HSQC spectrum. The chemical shifts of H2A
and C2A were identified from the presence of cross peaks at
5.09/4.01 ppm in the COSY spectrum and at 4.01/78.9 ppm in the
98 ( 1
1 ( 0
0.5 ( 0.0
glucose (%)
glucosamine (%)
a Each result is expressed as the mean value of three replicates.
Figure 1. 1H NMR (A), 13C NMR DEPT 135 (B), HSQC (C), HMBC (D), and COSY (E) spectra of intact K. marxianus YIT 8292 mannan.

8006 J. Agric. Food Chem., Vol. 57, No. 17, 2009
Yoshida et al.
Table 3. 1H and13C NMR Chemical Shifts of Intact Mannan Isolated from K. marxianus YIT 8292 and Its Acetolysis Products
sugar residue
B
chemical shift (ppm)
chemical shift (ppm)
oligosaccharide
C
A
proton
C
B
A
carbon
C
B
A
acetolysis-released triose
ManR1f2ManR1f2ManR
H1
H2
H1
H2
4.99
4.01
5.25
4.06
5.01
4.05
5.32
3.88
5.09
4.01
C1
C2
C1
C2
C6
102.3
70.5
100.7
78.6
92.6
79.4
98.5
78.9
65.9
intact mannan
V6
ManR1f2ManR
V1
102.4
70.3
61.4
Table 4. Methylation Analysis of K. marxianus YIT 8292 Mannan
partially methylated mannitol acetate
linkage
molar ratio
2,3,4,6-tetra-O-methyl
2,4,6-tri-O-methyl
3,4,6-tri-O-methyl
2,3,4-tri-O-methyl
3,4-di-O-methyl
4-O-methyl
Man 1f
f3 Man 1f
f2 Man 1f
f6 Man 1f
f2,6 Man 1f
f2,3,6 Man 1f
1.00
0.36
0.36
0.07
0.71
0.07
HSQC spectrum. H2B and C2B were assigned as 4.05 and
70.3 ppm, respectively, from cross peaks at 5.01/4.05 ppm in
the COSY spectrum and at 4.05/70.3 ppm in the HSQC spectrum.
In the HMBC spectrum, it was suggested that an R-(1,2) linkage
was present between residues A and B from a cross peak between
H1B (5.01 ppm) and C2A (78.9 ppm). The DEPT 135 spectrum
shows two inverted methylene signals (61.4 and 65.9 ppm). In the
HMBC, the methylene signal (65.9 ppm), which is a downfield
shift, correlated with the signal at 5.09 ppm (H1A), indicating the
presence of an R-(1,6)-linked backbone. Therefore, it was sug-
gested that K. marxianus YIT 8292 mannan consists mainly of the
partial structure shown in Table 3. In the methylation analysis of
K. marxianus YIT 8292 mannan, the major methylation products
2,3,4,6-tetra-O-methylmannitol and 3,4-di-O-methylmannitol
made up approximately 70% (mol/mol) of the total, which is
consistent with the above-mentioned structure (Table 4).
K. marxianus YIT 8292 mannan was subjected to acetolysis,
which selectively cleaves backbone R-(1,6) linkages, to obtain the
oligosaccharides corresponding to the side chains. The propor-
tions of biose (M2), triose (M3), tetraose (M4), and pentaose (M5)
in the resultant total oligosaccharides were 76.5, 11.8, 8.8, and
2.9%, respectively, on a molar basis (Figure 2). The composition
of mainly short side chains is similar to that of S. cerevisiae mnn5
mutant and C. kefir IFO 0586 (17-19), but different from that of
S. cerevisiae and K. lactis (16).
Figure 2. HPLC chromatogram of pyridylamino derivatives of acetolysis
products of K. marxianus YIT 8292 mannan (A) and baker’s yeast
(S. cerevisiae YIT 10265) mannan (B).
possesses N-acetylglucosamine substituents on some of the man-
nnotetrarose units (15). Ballou described that K. marxianus has
R-mannan structurally similar to that of K. lactis (13, 16). The
results of our structural analysis show that the cell surface
polysaccharide, mannan, of K. marxianus YIT 8292 consists of
a backbone of R-(1,6)-linked residues with short R-(1,2)-linked
branches, but lacks terminal R-(1,3)-mannose residues. It has a
comblike structure with shorter side-chain lengths and a lower
number of phosphate groups than that of S. cerevisiae. These
structural characteristics are quite similar to those reported for
C. kefyr IFO 0586 (17, 18), which is the imperfect state of
K. marxianus, and mannan mutants mnn1 mnn4 and mnn5 of
S. cerevisiae. The discrepancy concerning the mannan structure of
K. marxianus may be attributed to the existence of intraspecific
variations in mannan structure.
Comparison of Hypocholesterolemic Activities in Rats and Cell
Surface Hydrophobicity of S. cerevisiae X2180-1A and Its Mannan
Mutants. In our previous study, which compared the hypo-
cholesterolemic activities of 81 yeast strains, we found that
K. marxianus YIT 8292 exhibits greater hypocholesterolemic
activity compared with other yeasts containing S. cerevisiae (8).
We also found that cell wall polysaccharides are the major
active components of K. marxianus YIT 8292 (9). On the basis of
these previous findings, we suggest that the structural features of
mannan side chains of K. marxianus YIT 8292 may be involved in
its potent hypocholesterolemic activity. This possibility was
confirmed by comparing the hypocholesterolemic activities of
mannan mutants derived from S. cerevisiae X2180-1A in rats fed
a high-cholesterol diet. Schematic structures of N-linked mannan
outer chain of S. cerevisiae mannan mutants are shown in
Figure 1A on the bases of previous papers (13,14). The nutritional
composition of each crude cell wall is shown in Table 1. Their
dietary fiber contents were very similar. The hydrolysates of these
cell walls had an approximate composition of glucose and
mannose in the ratio of 1:1. These data suggest that no important
difference in β-glucan and R-mannan contents exists.
M2 and M3 were isolated and then identified by NMR analysis.
1
The H NMR and 13C NMR data of M2 are identical to those
of authentic Man-R-(1,2)-Man. The major component of M3 was
identified as Man-R-(1,2)-Man-R-(1,2)-Man by HMQC, HMBC,
and COSY NMR spectra (Figure 3; Table 3). The chemical shifts
are consistent with those in previous papers (17, 18).
The presence of M3 in the acetolysis products and 3,4,6-tri-
O-methylmannitol in the methylation products indicates that K.
marxianus YIT 8292 mannan also contains the internal R-(1,2)-
linked mannose unit. The weak cross peak observed at 5.26/
100.8ppm inthe HSQC spectrumof intactK. marxianus YIT 8292
mannan (Figure 1) was deduced to correspond to the internal
mannose unit from the NMR data of C. kefir IFO 0586 (17, 18).
The cross peaks in the HSQC and HMBC spectra of K. marxianus
YIT 8292 mannan were identical to those observed in S. cerevisiae
mnn5 mutant (spectrum not shown). These findings show that K.
marxianus YIT 8292 possesses a comblike structure comprising an
R-(1,6)-linked backbone substituted with a large number of short
side chains composed of R-(1,2)-linked mannose.
Mannan of K. lactis is closely analogous to that of S. cerevisiae
Food intake and body weight gain did not differ among groups
except that it lacks phosphorylated side chains and instead
(Table 5). Plasma triglycerides, phospholipids, and HDL-cholesterol

Article
J. Agric. Food Chem., Vol. 57, No. 17, 2009 8007
Figure 3. 1H NMR (A), 13C NMR (B), HMQC (C), HMBC (D), and COSY (E) spectra of triose obtained from acetolysis product of K. marxianus YIT 8292
mannan.
Table 5. Effects of Crude Cell Wall Fractions Isolated from S. cerevisiae X2180-1A, Its mnn Mutants, and K. marxianus YIT 8292 on Body Weight Gain, Food Intake,
Plasma Lipids, and Liver Lipids in Rats Fed a High-Cholesterol Dieta
S. cerevisiae
control
wild type
mnn4
mnn1 mnn4
mnn5
KM
body weight gain (g/14 days)
food intake (g/14 days)
plasma lipids (mg/dL)
HDL-cholesterol
94.5 ( 9.2 a
254 ( 22 a
90.1 ( 8.2 a
252 ( 15 a
86.2 ( 9.4 a
236 ( 20 a
91.5 ( 7.2 a
247 ( 12 a
85.0 ( 6.6 a
233 ( 10 a
87.0 ( 6.6 a
238 ( 15 a
11.4 ( 1.3 a
136 ( 44 a
156 ( 24 a
5.9 ( 0.4 a
16.7 ( 2.3 a
191 ( 75 a
159 ( 17 a
5.6 ( 0.3 ab
17.6 ( 2.1 a
228 ( 81 a
164 ( 24 a
5.3 ( 0.6bc
18.3 ( 2.3 a
150 ( 89 a
137 ( 31 a
5.4 ( 0.3 bc
17.4 ( 2 a
153 ( 51 a
148 ( 16 a
5.1 ( 0.3 bc
16.7 ( 1.9 a
162 ( 83 a
145 ( 28 a
5.0 ( 0.4 c
triglycerides
phospholipids
liver weight (dry g)
liver lipids (mg/g)
triglycerides
160 ( 17 a
126 ( 4 a
169 ( 17 a
124 ( 14 a
163 ( 17 a
118 ( 13 a
158 ( 20 a
123 ( 13 a
153 ( 15 a
117 ( 4.5 a
150 ( 13 a
114 ( 11 a
phospholipids
a The rats were given a high-cholesterol diet with or without 3.5% crude cell wall fraction isolated from S. cerevisiae, mnn1, mnn1 mnn4, mnn5, or K. marxianus YIT 8292 for
14 days. Each value is the mean ( SD (n = 8). Mean values within a row not sharing a common letter are significantly different (p < 0.05, Tukey test). Wild type, S. cerevisiae
X2180-1A; KM, K. marxianus YIT 8292.
levels did not change significantly (Table 5). Liver dry weight
was decreased by feeding the crude cell wall of mnn4, mnn1 mnn4,
and mnn5, but not by wild type of S. cerevisiae (Table 5). Feeding
crude cell wall isolated from mnn4 mutant, which possesses
phosphate-deficient mannan (13, 14), significantly decreased the
total cholesterol levels in the plasma and liver, whereas the crude
cell wall of the wild-type strain did not (Figure 4). In a comparison
of mannan mutants with different side-chain lengths, the hypo-
cholesterolemic activity showed a tendency to decrease as the side-
chain lengths shortened (Figure 4). The hypocholesterolemic activity
of mnn5 mutant, lacking the addition of the second R-(1,2)-linked
mannosyl residue to the mannan side chains (13,14,19), was higher
than that of the wild-type strain and approximately equal to that of
K. marxianus YIT 8292. These results indicate that the side-chain

8008 J. Agric. Food Chem., Vol. 57, No. 17, 2009
Yoshida et al.
Figure 4. Effects of crude cell wall fractions isolated from S. cerevisiae X2180-1A, its mannan mutants, and K. marxianus YIT 8292 on total cholesterol levels
in plasma (B) and liver (C). Each value is the mean ( SD (n = 8). Values not sharing a common letter are significantly different (p < 0.05, Tukey test).
Wild type, S. cerevisiae X2180-1; KM, K. marxianus YIT 8292.
structure of cell surface mannan has an impact on the hypo-
cholesterolemic activity of yeast. The results also suggest that
the structural characteristics of mannan side chains of K. marxianus
YIT 8292, low levels of phosphate, and short side chains are
responsible, at least in part, for its potent hypocholesterolemic
activity.
In the present paper, the effect of mannan side-chain structures
on CSH was also investigated. CSH was determined by measur-
ing the affinity of crude cell wall to the hydrophobic resin as
described previously (20). As shown in Figure 5, CSH of all
mannan mutants, which possesses phosphate-deficient mannan,
was higher than that of wild-type strain. CSH of mnn5 mutant,
lacking the addition of the second R-(1,2)-linked mannosyl
residue to the mannan side chains (12, 13, 18), was not signifi-
cantly different from that of the mnn4 mutant (p = 0.051), but
significantly higher than those of the wild-type strain and mnn1
mnn4 mutant. This result shows that deficiencies in the phos-
phorylation and the addition of the second R-(1,2)-linked manno-
syl residue increase CSH. The increase of the CSH may explain
part of the reason for the potent hypocholesterolemic activities of
mnn5 mutants.
Figure 5. Cell surface hydrophobicity of S. cerevisiae X2180-1A and its
mannan mutants.
At present, it is not clear why the side-chain length and
phosphate contents of mannan affect hypocholesterolemic
activities. It is well-known that some dietary fibers exhibit

Article
J. Agric. Food Chem., Vol. 57, No. 17, 2009 8009
hypocholesterolemic activity. The mechanisms by which dietary
fibers elicit their hypocholesterolemic effect have been pro-
posed (28). One of the primary actions is to reduce cholesterol
uptake and dietary fat in the intestine. It has been proposed that
the viscosity associated with soluble fibers interferes with key
physiological events in the cholesterol absorptive process. These
interference mechanisms include direct binding of sterols within
the intestinal lumen, interference with the diffusion of cholesterol
toward the epithelial cell surface, and/or antagonization of the
cholesterol emulsification process. Dietary fibers also interfere
with the enterohepatic circulation of bile acids, resulting in an
increase in the conversion of cholesterol to bile acids in the
liver (28). Another mechanism is the effect of short-chain fatty
acids (SCFA) on cholesterol metabolism. SCFA are products of
the colonic bacterial fermentation of dietary fiber. Several studies
have shown that the suppressive effect of certain dietary fibers on
the plasma cholesterol level was at least partly due to the
inhibition of cholesterol biosynthesis caused by SCFA (28). We
previously reported that feeding cell wall fraction isolated from
K. marxianus YIT 8292 increased the fecal excretion of acidic and
neutral sterols and the cecal SCFA concentration (9). These
observations suggest that cell wall polysaccharides of K. marxi-
anus YIT 8292 exerted a cholesterol-lowering effect through the
several mechanisms described above. Thus, we speculate that the
side-chain length and phosphate contents of mannan may affect
the interaction with cholesterol and/or bile acids and the ferment-
ability of cell wall polysaccharides. The association of CSH with
hypocholesterolemic activity among S. cerevisiae mannan mu-
tants may imply partial involvement of CSH in the cholesterol
lowering (e.g., through direct binding of cholesterol). Further
studies are necessary to demonstrate these possibilities.
marxianus YIT 8292 in rats fed a high cholesterol diet. Biosci.,
Biotechnol., Biochem. 2004, 68, 1185–1192.
(9) Yoshida, Y.; Yokoi, W.; Ohishi, K.; Ito, M.; Naito, E.; Sawada, H.
Effects of the cell wall of Kluyveromyces marxianus YIT 8292 on the
plasma cholesterol and fecal sterol excretion in rats fed on a high-
cholesterol diet. Biosci., Biotechnol., Biochem. 2005, 69, 714–723.
(10) Yoshida, Y.; Naito, E.; Ohishi, K.; Okumura, T.; Ito, M.; Sato, T.;
Sawada, H. Effect of Kluyveromyces marxianus YIT 8292 crude cell
wall fraction on serum lipids in normocholesterolemic and mildly
hypercholesterolemic subjects. Biosci., Biotechnol., Biochem. 2007,
71, 900–905.
(11) Mifuchi, I.; Hosoi, M.; Yokokura, T. Promotion of resistance to
Ehrlich ascites carcinoma cells by injection of tumor cells mixed with
yeast cell wall. Gann 1969, 60, 655–656.
(12) Fleet, G. H. Cell walls. In The Yeasts, 2nd ed.; Rose, A. H., Harrison, J.
S., Eds.; Academic Press: London, U.K., 1991; Vol. 4, pp 199-277.
(13) Ballou, C. E. Structure and biosynthesis of the mannan component
of the yeast cell envelope. Adv. Microb. Physiol. 1976, 14, 93–158.
(14) Ballou, C. E. Isolation, characterization, and properties of Sac-
charomyces cerevisiae mnn mutants with nonconditional protein
glycosylation defects. Methods Enzymol. 1990, 185, 440–470.
(15) Nguyen, T. H.; Fleet, G. H.; Rogers, P. L. Composition of the cell
walls of several yeast species. Appl. Microbiol. Biotechnol. 1998, 50,
206–212.
(16) Smith, W. L.; Nakajima, T.; Ballou, C. E. Biosynthesis of yeast
mannan: Isolation of Kluyveromyces lactis mannan mutants and a
study of the incorporation of N-acetyl-D-glucosamine into the
polysaccharide side chains. J. Biol. Chem. 1975, 250, 3426–3435.
(17) Kobayashi, H.; Komido, M.; Watanabe, M.; Matsuda, K.; Suzuki,
M.; Ikeda, T.; Oyamada, H.; Shibata, N.; Suzuki, S. Structure of cell
wall mannan of Candida kefyr IFO 0586. Infect. Immun. 1994, 62,
4425–4431.
(18) Kobayashi, H.; Watanabe, M.; Komido, M.; Matsuda, K.; Ikeda-
Hasebe, T.; Suzuki, M.; Shibata, N.; Hisamichi, K.; Suzuki, S.
Assignment of 1H and 13C NMR chemical shifts of R-mannan
composed of R-(1,2) and R-(1,6) linkages obtained from Candida
kefyr IFO 0586 strain. Carbohydr. Res. 1995, 267, 299–306.
(19) Cohen, R. E.; Ballou, L.; Ballou, C. E. Saccharomyces cerevisiae
mannoprotein mutants: isolation of the mnn5 mutant and compar-
ison with the mnn3 strain. J. Biol. Chem. 1980, 255, 7700–7707.
(20) Akiyama-Jibiki, M.; Ishibiki, T.; Yamashita, H; Eto, M. A rapid and
simple assay to measure flocculation in brewer’s yeast. Master
Brewers’ Assoc. Am. Tech. Q. 1997, 34, 278–281.
In conclusion, mannan of K. marxianus YIT 8292 has a
comblike structure with short side chains of R-(1,2)-linked
mannopyranose residues and low phosphate content. Our results
show that these structural characteristics have an impact on
hypocholesterolemic activity.
ACKNOWLEDGMENT
We thank Dr. Kouji Miyazaki and Dr. Fumiyasu Ishikawa
for providing us with valuable advice. We thank members of
the Animal Experiment Research Laboratory at our institute
for their maintenance of the rats and Yuko Nakamura for tech-
nical help.
(21) Folch, J.; Lees, M.; Sloane-Stanley, G. H. A simple method for the
isolation and purification of total lipides from animal tissue. J. Biol.
Chem. 1957, 226, 497–509.
(22) Dubois, M.; Gilles, K. A.; Hamilton, J. K.; Rebers, P. A.; Smith, F.
Colorimetric method for determination of sugars and related
substances. Anal. Chem. 1956, 28, 350–356.
LITERATURE CITED
(23) Honda, S.; Akao, E.; Suzuki, S.; Okuda, M.; Kakehi, K.; Nakamura,
L. High-performance liquid chromatography of reducing carbohy-
drates as strongly ultraviolet-absorbing and electrochemically sensi-
tive 1-phenyl-3-methyl-5-pyrazolone derivatives. Anal. Biochem.
1989, 180, 351–357.
(1) Dziekak, J. D. Yeast and yeast derivatives; definitions, character-
istics and processing. Food Technol. 1987, 2, 104–120.
(2) Robbins, E. A.; Seeley, R. D. Cholesterol lowering effect of dietary
yeast and yeast fractions. J. Food Sci. 1977, 42, 694–698.
(3) Ebihara, K.; Kiriyama, S. Plasma cholesterol-normalizing activity of
brewer’s yeast and its various components in rats. Nutr. Rep. Int.
1983, 28, 861–871.
(4) Offenbacher, E. G.; Pi-Sunyer, F. X. Beneficial effect of chromium-
rich yeast on glucose tolerance and blood lipids in elderly subjects.
Diabetes 1980, 29, 919–925.
(24) Kocourek, J.; Ballou, C. E. Method for fingerprinting yeast cell wall
mannans. J. Bacteriol. 1969, 100, 1175–1181.
(25) Kimura, K.; Matsumoto, K.; Ishihara, C.; Harada, K.; Miyagi, A.
Structure determination of galacto-oligosaccharides by pyridylami-
nation and NMR spectroscopy. Carbohydr. Res. 1995, 270, 33–42.
(26) Ciucanu, I.; Kerek, F. A simple and rapid method for the permethyl-
ation of carbohydrates. Carbohydr. Res. 1984, 131, 209–217.
(27) AOAC. Official Methods of Analysis, 15th ed.; Association of Official
Analytical Chemists: Arlington, VA, 1990.
(28) Rideout, T. C.; Harding, S. V.; Jones, P. J.; Fan, M. Z. Guar gum
and similar soluble fibers in the regulation of cholesterol metabolism:
current understandings and future research priorities. Vasc. Health
Risk Manag. 2008, 4, 1023–1033.
(5) Elwood, J. C.; Nash, D. T.; Streeten, D. H. Effect of high-chromium
brewer’s yeast on human serum lipids. J. Am. Coll. Nutr. 1982, 1,
263–274.
(6) Nakamura, T.; Agata, K.; Mizutani, M.; Iino, H. Effects of brewer’s
yeast cell wall on constipation and defecation in experimentally
constipated rats. Biosci., Biotechnol., Biochem. 2001, 65, 774–780.
(7) Nicolosi, R.; Bell, S. J.; Bistrian, B. R.; Greenberg, I.; Forse, R. A.;
Blackburn, G. L. Plasma lipid changes after supplementation with
β-glucan fiber from yeast. Am. J. Clin. Nutr. 1999, 70, 208–212.
(8) Yoshida, Y.; Yokoi, W.; Wada, Y.; Ohishi, K.; Ito, M.; Sawada, H.
Potent hypocholesterolemic activity of the yeast Kluyveromyces
Received January 31, 2009. Revised manuscript received May 1, 2009.
Accepted July 11, 2009.