Biotechnology Letters 26: 623–627, 2004.

© 2004 Kluwer Academic Publishers. Printed in the Netherlands.

623

Increased xylitol production rate during long-term cell recycle

fermentation of Candida tropicalis

Teak-Bum Kim1 , Yong-Joo Lee1 , Pil Kim2 , Chang Sup Kim1 & Deok-Kun Oh1,∗
1 Department of Bioscience and Biotechnology, Sejong University, Seoul 143-747, Korea
2 Division of Biotechnology, The Catholic University of Korea, Gyeonggi-do 420-743, Korea
∗ Author for correspondence (Fax: +82-2-3408-3911; E-mail: deokkun@sejong.ac.kr)

Received 9 January 2004; Revisions requested 23 January 2004; Revisions received 10 February 2004; Accepted 11 February 2004

Key words: Candida tropicalis, cell recycle fermentation, chemically defined medium, volumetric productivity,
xylitol

Abstract
Long-term cell recycle fermentations of Candida tropicalis were performed over 14 rounds of fermentation. The
average xylitol concentrations, fermentation times, volumetric productivities and product yields for 14 rounds were
105 g l−1 , 333 h, 4.4 g l−1 h−1 and 78%, respectively, in complex medium; and 110 g l−1 , 284 h, 5.4 g l−1 h−1 and
81%, respectively, in a chemically defined medium. These productivities were 1.7 and 2.4 times those with batch
fermentation in the complex and chemically defined media, respectively. The xylitol yield from xylose with cell
recycle fermentation using the chemically defined medium was 81% (w/w), which was 7% greater than the xylitol
yield with batch fermentation (74%); both modes of fermentation gave the same yield using the complex medium.
These results suggest that the chemically defined medium is more suitable for production of xylitol than complex
medium.

Introduction 120 h. The concentration and yield of xylitol were 5%

Xylitol is a natural food sweetener that prevents dental
caries, otitis, osteoporosis and inflammatory pro-
cesses; it also has an insulin-independent metabolism
(Makinen 1979, Makinen et al. 1995, Peldyak &
Makinen 2002). Due to these characteristics, xylitol
has attracted the attention of the food and pharma-
ceutical industries (Masalin 1992). The conversion of
xylose to xylitol by yeast has economic potential be-
cause it is a very specific process and requires little
energy. The microbial production of xylitol has been
extensively studied and, currently, xylitol is being
produced by Candida tropicalis KFCC10960 on an
industrial scale (Oh & Kim 1998). We recently re-
ported that a chemically defined medium containing
urea and vitamins, instead of yeast extract, as a ni-
trogen source could decrease xylitol production costs
(Kim & Oh 2003). In fed-batch cultures using the che-
mically defined medium, 237 g xylitol l−1 with a volu-
metric productivity of 2 g l−1 h−1 was obtained after
and 4% higher, respectively, than those obtained using
the complex medium, indicating that xylitol can be
effectively produced industrially using the chemically
defined medium.
Increases in volumetric productivity and yield are
important factors for industrial xylitol production. Cell
recycle fermentation has been reported elsewhere to
increase productivity and yield (von Weymarn et al.
2002). Although the chemically defined medium de-
scribed above showed reasonable results for xylitol
production in fed-batch cultures, the medium has not
been confirmed as beneficial for long-term operations,
such as cell recycle repeated batch culture.
In this paper, we describe the increases in pro-
ductivity and yield of xylitol using long-term cell
recycle fermentation and evaluate the use of the che-
mically defined and complex media in long-term fer-
mentation.
624
Materials and methods
Microorganism and media
Candida tropicalis KCTC 7221 (ATCC 13803) was
used grown on complex medium consisting of 150 g
xylose l−1 , 10 g yeast extract l−1 , 5 g KH2 PO4 l−1 ,
0.2 g MgSO4 ·7H2 O l−1 and minerals. The chemically
defined medium contained 150 g xylose l−1 , 5 g l−1
urea, 5 g KH2 PO4 l−1 , 0.2 g MgSO4 · 7H2 O l−1 , vit-
amins and minerals. The minerals in the chemically
defined and the complex media were as follows: 7 mg
MnSO4 · 4H2 O l−1 , 4 mg CoCl2 · 6H2 O l−1 , 2 mg
NaMoO4 l−1 , 2 mg ZnSO4 · 7H2 O l−1 , 1 mg AlCl3 ·
6H2 O l−1 , 2 mg CuCl2 · H2 O l−1 , 0.5 mg H3 BO3 l−1 ,
40 mg FeSO4 · 7H2 O l−1 . The vitamins in the chem-
ically defined medium were 5 mg ascorbic acid l−1 ,
10 mg biotin l−1 , 25 mg choline l−1 , 50 mg folic
acid l−1 , 10 mg inositol l−1 , 25 mg nicotinic acid l−1 ,
1 mg p-aminobenzoic acid l−1 , 5 mg pantothenic
acid l−1 , 1 mg pyridoxine l−1 , 100 mg riboflavin l−1 ,
and 5 mg thiamin l−1 . The minerals and the vitamins
were separately sterilized by filtration.
Culture conditions of batch and cell recycle
fermentations
Actively growing cells of C. tropicalis KCTC 7221
(3 ml) were pre-cultured in a 500 ml baffled flask con-
taining 100 ml growing-medium consisting of 20 g
glucose l−1 and 10 g yeast extract l−1 at 30 ◦ C with
agitation at 230 rpm for 12 h. This seed culture was
then inoculated into a 7 l jar fermenter (Biotron Co.,
Buchenon, Korea) containing 2 l of the complex or the
chemically defined medium. The initial pH was set to
be 6.5 and pH was uncontrolled during the fermenta-
tion. Temperature was maintained at 30 ◦ C. Agitation
speed was manually adjusted in the range of 350 to
400 rpm with aeration of 1 vvm. After finishing the
batch culture, the cells were recovered by centrifu-
gation (5000 × g, 4 ◦ C, 5 min) and resuspended in
fresh medium for the recycling of cells. The medium
was changed with fresh one whenever the remain-
ing xylose concentration in the fermenter fell below
10 g l−1 .
Analysis and enzyme activity measurement
The cell concentration was estimated using a standard
curve, which correlated absorbancy at 600 nm and dry
cell weight (1 OD660nm = 482 mg DCW l−1 ). Xylose
and xylitol were analyzed by HPLC equipped with a
Fig. 1. Batch fermentations for xylitol production in the complex
and chemically defined media using Candida tropicalis. Closed
symbols and solid lines are used for the complex medium; cell mass
(
), xylose (
), xylitol (). Open symbols and dotted lines are
used for the chemically defined medium; cell mass (), xylose (),
xylitol ().
RI detector using a BP-100 carbohydrate Ca2+ column
(Benson polymeric, Reno, NV) at 80 ◦ C with water as
an eluent at 1 ml min−1 .
Cells were disrupted by vortexing vigorously with
glass beads (SiLibeads d = 0.25–0.5 mm, Sigmund
Linder, Germany) at ice bath for 20 min (2.5 g beads
per 100 mg cells) after washing three times with
50 mM phosphate buffer (pH 8). The disrupted solu-
tion was centrifuged and the supernatant was used for
enzyme assay. Xylose reductase activity was estimated
by measuring a decrease of NADPH at 340 nm. The
reaction mixture (1 ml) consisted of 0.3 mM NADPH,
0.5 mM xylose, 100 µl cell extract in a 50 m M phos-
phate buffer (pH 8). The extinction coefficient for
NADPH was 6.22 cm−1 mM−1 at 340 nm. One unit
of enzyme activity corresponded to decomposition of
1 µmole xylose per 1 min (Rodrigues et al. 2002).
Protein concentrations were determined using a pro-
tein assay kit (BioRad, Richmond, CA) using BSA as
the standard.
Results and discussion
Batch fermentations in the complex and chemically
defined media
Batch culturing was performed in a 7-l jar fermenter
with 2 l of either the complex or chemically defined
medium containing 150 g xylose l−1 (Figure 1). A cell
mass of 24 g l−1 and a xylitol yield of 116 g l−1 were
obtained after 45-h batch culture in the complex me-

Fig. 2. Cell recycle fermentations for xylitol production using Can-
dida tropicalis. (A) Complex medium. (B) Chemically defined
medium. Cell recycles were performed 14 times. Cell mass (closed
bar), xylitol (open bar), volumetric productivity (
).
dium. Product yield and volumetric productivity were
78% and 2.6 g l−1 h−1 , respectively. Batch culture
for 48 h in the chemically defined medium gave a cell
mass of 28 g l−1 and a xylitol yield of 109 g l−1 , rep-
resenting a yield and volumetric productivity of 74%
and 2.3 g l−1 h−1 , respectively. These results indicate
that the fermentation parameters (i.e., product yield,
volumetric productivity, cell growth rate, xylose con-
sumption rate, and xylitol production rate) were higher
in the complex medium than in the chemically defined
medium in short-term batch fermentations.
Cell recycle fermentations in the complex and
chemically defined media
Long-term xylitol production was investigated with
the recycling of cells derived from batch cultures in
the complex and chemically defined media (Figure 2).
When the xylose concentration in the fermenter fell
below 10 g l−1 , the whole culture broth was centri-
fuged, and the cells were resuspended in fresh medium
for the next recycle batch. The remaining xylitol
625
Fig. 3. Xylose reductase activities in the complex () and chemic-
ally defined (
) media during cell recycle fermentations.
and xylose concentrations in each recycle batch were
included in the fermentation parameter calculation.
Fourteen rounds of cell recycle fermentation in the
complex medium produced a total of 2938 g xylitol
from 3767 g xylose in a 2-l reactor volume in 333 h,
which corresponds to an average xylitol concentration
of 105 g l−1 for each recycle batch (Figure 2A). The
total cell mass in the 2-l reactor volume was 122 g, and
the average culture time of each batch was 24 h. The
volumetric productivity and xylitol yield from xylose
were 4.4 g l−1 h−1 and 78%, respectively.
Fourteen rounds of cell recycle fermentation in the
chemically defined medium gave a total of 3078 g
xylitol in the 2-l reactor volume within 284 h (Fig-
ure 2B). An average xylitol concentration of 110 g l−1
was achieved in each recycle batch. In the chemic-
ally defined medium, the average culture time of each
batch was 20 h; the total cell mass was 100 g; the volu-
metric productivity was 5.4 g l−1 h−1 ; and the xylitol
yield from xylose was 81%.
The xylose consumption time remained almost
constant during cell recycle fermentation in the chem-
ically defined medium, whereas the xylose consump-
tion time in the complex medium increased as the
recycle batch number increased. This suggests that
the state of the cells, such as the specific production
rate, became worse over time in the complex medium.
These findings indicate that the chemically defined
medium is more valuable for economical xylitol pro-
duction by long-term fermentation not only because
of the lower cost of the medium but also because of
the more desirable state of the cells (less growth, more
xylitol conversion) in the chemically defined medium.
626
Table 1. Parameters for xylitol production in batch and cell recycle fermentations.

Medium Cell Total xylitol Volumetric Xylitol yield Time Time/

concentration productiona productivity from xylose (h) batch
(g l−1 ) (g) (g l−1 h−1 ) (%) (h)

Batch fermentation
Complex 24 232 2.6 78 45 45
Chemically defined 28 218 2.3 74 48 48

Cell recycle fermentation

Complex 61 2939 4.4 78 333 24
Chemically defined 50 3076 5.4 81 284 20
a Two-l reactor volume.

The intracellular xylose reductase activity of cells
in the complex medium was maintained at a level
similar to that of cells in the chemically defined me-
dium, even though the complex medium supported a
greater cell increase than did the chemically defined
medium (Figure 3). No significant difference mor-
phology was found between the cells from complex
and the chemically defined medium, though aggreg-
ation of cells was observed after 7th batches in both
cases (data not shown). Possible difference in cell
states between the two media would be available en-
ergy level, higher energy in the complex medium and
lower in the chemically defined medium. The higher
energy might drive the whole cellular metabolism to-
ward cell growth, which resulted in more biomass
increase and less xylitol conversion. These results im-
ply that more xylose was converted into cell mass
instead of xylitol in the complex medium.
Comparisons of batch and cell recycle fermentations
As expected, higher volumetric productivity was
achieved with cell recycle fermentation than with
simple batch fermentation (Table 1). Because cells
were re-used for the inoculation of the next recycle
batch, a higher initial cell concentration was main-
tained in recycle fermentation than in batch fermen-
tation. The time required for the bioconversion of a
given amount of xylose (initially, about 150 g l−1 in
each batch) to xylitol was shorter in the recycle fer-
mentation mode than in the batch fermentation mode
(24 vs. 45 h in complex medium, 20 vs. 48 h in chem-
ically defined medium) due to the increasingly greater
initial biomass in recycle fermentation. Correspond-
ingly, the volumetric productivity was greater with
recycle fermentation (4.4 vs. 2.6 g l−1 h−1 in com-
plex medium and 5.4 vs. 2.3 g l−1 h−1 in chemically
defined medium for recycle vs. batch fermentation,
respectively).
The xylitol yield from xylose in recycle fermenta-
tion using the complex medium (78%) was not differ-
ent from that in batch fermentation, while the xylitol
yield using the chemically defined medium was 7%
greater with recycle fermentation (81%) than with
batch fermentation (74%). Because the total cell mass
in recycle fermentation using the complex medium
(74 g) was greater than that using the chemically
defined medium (44 g), one would expect that more
xylose was directed toward increasing cell mass in
the complex medium, which would result in the lower
xylitol production.
Long-term cell recycle fermentation for the pro-
duction of xylitol showed a higher volumetric pro-
ductivity and yield than did simple batch fermenta-
tion, because shorter fermentation times and higher
substrate consumption rates were obtained in recycle
fermentation due to the higher cell concentration asso-
ciated with this method. In addition, the chemically
defined medium is more suitable for an economi-
cally valuable process than is the complex medium,
because it results in a lower cost of medium, desir-
able cell state, and better production rates. The cell
recycle fermentation process used in this study, how-
ever, is not considered appropriate for a large scale
operation, because the required batch-type centrifuga-
tion step is laborious and time-consuming. The use of
aseptic continuous centrifugation during commercial-
scale production could alleviate this potential problem,
although this technique is not currently available for
use in a lab-scale study.

Acknowledgement
This work was supported by a grant of the project
of cleaner production (Grant No. 10001990), Korea
Institute of Industrial Technology, Republic of Korea.
References
Kim TB, Oh DK (2003) Xylitol production by Candida tropicalis in
a chemically defined medium. Biotechnol. Lett. 25: 2085–2088.
Makinen KK (1979) Xylitol and oral health. Adv. Food Res. 25:
137–158.
Makinen KK, Pemberton D, Cole J, Makinen PL, Chen CY, Lam-
bert P (1995) Saliva stimulants and the oral health of geriatric
patients. Adv. Dent. Res. 9: 125–126.
627
Masalin K (1992) Caries-risk-reducing effects of xylitol-containing
chewing gum and tablets in confectionery workers in Finland.
Community Dent. Hlth. 9: 3–10.
Oh DK, Kim SY (1998) Increase of xylitol yield by feeding xylose
and glucose in Candida tropicalis. Appl. Microbiol. Biotechnol.
50: 419–425.
Peldyak J, Makinen KK (2002) Xylitol for caries prevention. J.
Dent. Hyg. 76: 276–285.
Rodrigues DC, Da Silva SS, Almeida ESJB, Vitolo M (2002)
Xylose reductase activity of Candida guilliermondii during xy-
litol production by fed-batch fermentation: selection of process
variables. Appl. Biochem. Biotechnol. 98–100: 875–883.
von Weymarn N, Kiviharju K, Leisola M (2002) High-level produc-
tion of D-mannitol with membrane cell-recycle bioreactor. J. Ind.
Microbiol. Biotechnol. 29: 44–49.