Enhanced Propionic Acid Fermentation

by Propionibacterium acidipropionici
Mutant Obtained by Adaptation in
a Fibrous-Bed Bioreactor

Supaporn Suwannakham, Shang-Tian Yang

Department of Chemical and Biomolecular Engineering, The Ohio State University,

140 West 19th Avenue, Columbus, Ohio 43210; telephone: 614-292-6611;
fax: 614-292-3769; e-mail: Yang.15@osu.edu
Received 15 August 2004; accepted 18 January 2005

Published online 23 June 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/bit.20473

Abstract: Fed-batch fermentations of glucose by P. acid- Keywords: Propionibacterium acidipropionici; propionic

ipropionici ATCC 4875 in free-cell suspension culture and
immobilized in a fibrous-bed bioreactor (FBB) were
studied. The latter produced a much higher propionic
acid concentration (71.8
0.8 g/L vs. 52.2
1.1 g/L),
indicating enhanced tolerance to propionic acid inhibition
by cells adapted in the FBB. Compared to the free-cell
fermentation, the FBB culture produced 20–59% more
propionate (0.40–0.65
0.02 g/g vs. 0.41
0.02 g/g), 17%
less acetate (0.10
0.01 g/g vs. 0.12
0.02 g/g), and 50%
less succinate (0.09
0.02 g/g vs. 0.18
0.03 g/g) from
glucose. The higher propionate production in the FBB was
attributed to mutations in two key enzymes, oxaloacetate
transcarboxylase and propionyl CoA: succinyl CoA trans-
ferase, leading to the production of propionic acid from
pyruvate. Both showed higher specific activity and lower
sensitivity to propionic acid inhibition in the mutant than
in the wild type. In contrast, the activity of PEP carbox-
ylase, which converts PEP directly to oxaloacetate and
leads to the production of succinate from glucose, was
generally lower in the mutant than in the wild type. For
phosphotransacetylase and acetate kinase in the acetate
formation pathway, however, there was no significant
difference between the mutant and the wild type. In
addition, the mutant had a striking change in its morphol-
ogy. With a threefold increase in its length and &24%
decrease in its diameter, the mutant cell had an &10%
higher specific surface area that should have made the
mutant more efficient in transporting substrates and
metabolites across the cell membrane. A slightly lower
membrane-bound ATPase activity found in the mutant
also indicated that the mutant might have a more efficient
proton pump to allow it to better tolerate propionic acid. In
addition, the mutant had more longer-chain saturated
fatty acids (C17:0) and less unsaturated fatty acids (C18:1),
both of which could decrease membrane fluidity and
might have contributed to the increased propionate
tolerance. The enhanced propionic acid production from
glucose by P. acidipropionici was thus attributed to both a
high viable cell density maintained in the reactor and
favorable mutations resulted from adaptation by cell
immobilization in the FBB. ß 2005 Wiley Periodicals, Inc.
Correspondence to: Shang-Tian Yang
Contract grant sponsors: U.S. Department of Agriculture; Consortium for
Plant Biotechnology Research, Inc.
Contract grant number: CSREES 99-35504-7800
acid fermentation; fibrous-bed bioreactor
INTRODUCTION
Propionibacteria are Gram-positive, nonspore-forming, non-
motile, anaerobic, rod-shaped bacteria. They are widely used
in probiotic and cheese industries (Playne, 1985). They also
can be used to produce vitamin B12, tetrapyrrole compounds,
and propionic acid. The latter is an important chemical used
in production of cellulose plastics, herbicides, and perfumes.
As a strong mold inhibitor, calcium, sodium, and potassium
salts of propionate are also widely used as food and feed
preservatives. Currently, almost all propionic acid is
produced by petrochemical processes, at an annual produc-
tion rate of &400 million pounds in the U.S. Although there
has been a high interest to produce propionic acid from
biomass via fermentation with propionibacteria, the rela-
tively low propionic acid concentration, yield, and produc-
tion rate from the fermentation have been the major barriers
for economical production of propionic acid from glucose
and other fermentable sugars.
The problems associated with conventional propionic acid
fermentation largely stem from the strong end product
inhibition caused by propionic acid even at a very low
concentration of 10 g/L (Gu et al., 1998; Hsu and Yang,
1991). The conventional batch propionic acid fermentation
usually takes more than 3 days to reach &20 g/L propionic
acid, with a product yield usually less than 0.45 g/g sugar
fermented. Attempts to improve propionic acid fermentation
in terms of its yield, final product concentration, and
production rate have resulted in the development of new
bioprocesses and mutant strains but with limited success
(Emde and Schink, 1990; Huang et al., 1998; Jin and Yang,
1998; Lewis and Yang, 1992; Paik and Glatz, 1994; Rickert
et al., 1998; Solichien et al., 1995; Woskow and Glatz, 1991).
Among all these previous efforts, a fibrous-bed immobilized-
cell bioreactor has shown to be a promising technology that
can significantly improve volumetric productivity, product
yield, and final product concentration in several organic acid

ß 2005 Wiley Periodicals, Inc.

fermentations (Huang and Yang, 1998; Lewis and Yang,
1992; Silva and Yang, 1995; Yang et al., 1994, 1995; Zhu and
Yang, 2003; Zhu et al., 2002).
The potential of using the fibrous bed bioreactor to produce
high-concentration propionic acid from glucose was evaluated
in this study. We found that a final concentration of more than
70 g/L of propionic acid could be produced from glucose in a
fed-batch fermentation. This propionic acid concentration was
not only much higher than that from a similar fed-batch
fermentation with free cells grown in a conventional stirred-
tank bioreactor, but also higher than the highest concentration
(57 g/L) previously reported for an acid-tolerant strain
immobilized in calcium alginate beads (Paik and Glatz,
1994). To understand the underlying mechanisms contributing
to the improved propionic acid production and tolerance by
cells immobilized in the FBB, the effects of propionic acid on
both the original (wild type) and adapted (from the FBB)
cultures were also studied and compared, with respect to their
specific growth rates, cellular activities of several key enzymes
in the dicarboxylic acid pathway, and membrane-bound
ATPase. In addition to significant differences in some of these
factors studied, it was also found that there were significant
changes in the membrane fatty acid composition and cell
morphology. Based on these results, how the adapted (mutant)
cells from the FBB acquired the ability to tolerate and produce
more propionic acid is also discussed in this article.
MATERIALS AND METHODS
Culture and Media
Propionibacterium acidipropionici ATCC 4875 was grown
in a synthetic medium containing (per liter): 10 g yeast
extract (Difco Laboratories, Detroit, MI), 5 g Trypticase
(BBL), 0.25 g K2HPO4, 0.05 g MnSO4 and 50–100 g glucose
as the substrate. The medium was sterilized by autoclaving at
1218C, 15 psig for 30 min. The stock culture was kept in
serum bottles under anaerobic conditions at 48C.
Free-Cell Fermentation
Unless otherwise noted, the fermentation was carried out in a
5-L fermentor (Marubishi MD-300) containing 2 L of the
synthetic medium, at 328C, pH 6.5, and 100 rpm for agitation.
Anaerobiosis was established by sparging the medium with
N2 for &30 min, and thereafter the fermentor headspace was
maintained under 5 psig N2. The fermentor was inoculated
with &100 mL of a fresh culture grown in a serum bottle
(OD600 & 2.0), and liquid samples were withdrawn at regular
time intervals. The glucose level in the fermentation broth
was replenished by pulse feeding a concentrated glucose
solution to allow the fermentation to continue until reaching
the maximum propionic acid concentration.
Immobilized-Cell Fermentation
The fermentation was carried out with cells immobilized in a
fibrous-bed bioreactor (FBB) connected to a 5-L fermentor
326 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 91, NO. 3, AUGUST 5, 2005
(Marubishi MD-300) through a recirculation loop (&1 m
long, tubing ID: 3.1 mm; Microflex Norprene 06402-16; Cole
Palmer, Chicago, IL) and operated under well-mixed
conditions with pH and temperature controls. The FBB
was constructed by packing a spiral wound cotton towel into
a glass column bioreactor, and had a working volume
of &690 mL. A detailed description of the reactor construc-
tion has been given elsewhere (Silva and Yang, 1995). After
inoculation with &100 mL of cell suspension (OD600 & 2.0)
into the fermentor, cells were grown for 3–4 days to reach an
optical density (OD600) of &3.5. The fermentation broth was
then circulated at a flow rate of &30 mL/min through the
FBB to allow cells to attach and be immobilized in the fibrous
matrix. The process continued for 60–72 h until most cells
had been immobilized in the FBB. The medium circulation
rate was then increased to &80 mL/min, and the fermenta-
tion was run at a repeated batch mode to obtain a high cell
density in the fibrous bed. Fed-batch fermentation with pulse
additions of concentrated glucose solution was then per-
formed to study the fermentation kinetics and to evaluate the
achievable maximum propionic acid concentration. Samples
were taken at regular time intervals throughout the
fermentation for analyses. At the end of the fed-batch
fermentation, adapted cells (mutants) in the FBB were
removed from the fibrous matrix by vortexing the matrix in
sterile distilled water. The cells were collected for viability
assay and subcultured in serum bottles for further analyses.
Effect of Propionic Acid on Cell Growth
To determine the inhibition effect of propionic acid on cell
growth, cells were grown under anaerobic conditions in
serum tubes containing 10 mL of the synthetic medium with
20 g/L glucose and varying amounts of propionic acid (0–
100 g/L). Cell growth was followed by measuring the optical
density (OD) at 600 nm. The experiment was done in
duplicate. The specific growth rates at various initial
propionic acid concentrations were then estimated from the
semilogarithmic plots of the OD vs. time.
Enzyme Assays
Cells grown in the synthetic medium (50 mL) at 328C to the
exponential phase (OD600 & 1.8) were harvested, washed,
and resuspended in 15 mL of 25 mM Tris/HCl (pH 7.4). The
cell suspension was then sonicated to break cell walls and
centrifuged at 10,000 rpm for 1 h to remove cell debris. The
cell extracts were kept cold on ice before they were used in
the enzyme activity assays. The protein content of the
extracts was determined in triplicate by Bradford protein
assay (Bio-Rad, Hercules, CA) with bovine serum albumin as
the standard.
Phosphotransacetylase was assayed by the method of
Andersch et al. (1983). The assay solution (final total volume:
200 mL) contained: 0.1M potassium phosphate buffer
(pH 7.4), 0.2 mM acetyl-CoA, 0.08 mM 5,50 -dithio-bis
(2-nitrobenzoate) (DTNB), and the cell extract (120 mL). Membrane-Bound ATPase Assay
The enzyme activity was determined at 328C by measuring
Cells were grown at 328C in the synthetic medium (50 mL)
the absorbance at 405 nm (SpectraMax 250) following the
containing 0.4% (w/v) glycine to the exponential phase
liberation of coenzyme A, which has a molar extinction
(OD600 & 1.8) or stationary phase (OD600 & 3.0), and
coefficient of 13.6 mM1cm1 (Ellman, 1959). One unit of
harvested and suspended in 2 mL of TE buffer (10 mM
activity is defined as the amount of enzyme converting 1 mmol
Tris-HCl, 1 mM EDTA, pH 8.0). Then, mutanolysin
of acetyl-CoA per minute.
(Sigma) at the final concentration of 100 mg/mL was added
Acetate kinase was assayed by the method of Allen et al.
to lyse the cells at 378C for 40 min. The crude membrane
(1964). The assay mixture (final total volume: 300 mL)
fraction was obtained by centrifugation at 48C for 30 min and
contained: 81 mM Tris/HCl buffer (pH 7.4), 4 mM ATP,
was suspended in 2.5 mL of 100 mM PIPES buffer (pH
4 mM MgCl2, 1.6 mM phosphoenolpyruvate, 81 mM
5.95). ATPase activity was determined based on the method
potassium acetate, 0.4 mM NADH, 0.4 U/mL pyruvate
of Gutierrez and Maddox (1992) by measuring the release of
kinase, 0.4 U/mL lactate dehydrogenase, and the cell extract
inorganic phosphate (Pi) from ATP in the reaction mixture
(160 mL). The reaction rate at 258C was followed by
(total volume: 500 mL) containing 5 mM ATP, 10 mM
measuring the absorbance at 340 nm, which decreased
MgCl2 and 50 mL of the crude membrane fraction. The
linearly with time. A control without acetate was used as the
mixture was incubated at 378C for 20 min and the reaction
blank to correct for any ADP that might be present in the ATP
was stopped by adding 1 mL of 15% (w/v) ice-cold
preparation.
trichloroacetic acid. The mixture was then centrifuged twice
Oxaloacetate transcarboxylase was assayed by the method
at 13,200 rpm for 5 min, and the inorganic phosphate Pi in the
of Wood et al. (1969). Briefly, the reaction mixture (final total
supernatant was determined spectrophotometrically at
volume: 300 mL) contained: 0.1 mM NADH, 0.2 mM
830 nm (Shimadzu UV-1601) following the molybdenum
methylmalonyl CoA, 10 mM sodium pyruvate, 350 mM
blue method (Vogel, 1978). One unit of activity is defined as
potassium phosphate (pH 6.8), 2 U/mL malate dehydro-
the amount of enzyme that releases 1 mmol of Pi per minute,
genase, and the cell extract (70 mL). The reaction rate at 258C
and the specific activity of ATPase is defined as the unit of
was followed by measuring the absorbance at 340 nm, which
activity per mg biomass.
decreased linearly with time, for at least 4 min. The reaction
mixture without methylmalonyl CoA was used as the blank.
Propionyl CoA:succinyl CoA transferase (CoA transfer- Cell Membrane Fatty Acid Analysis
ase) was assayed following the method of Schulman and
Cells were grown to the exponential phase and then analyzed
Wood (1975). The reaction mixture (250 mL) consisted of:
for their membrane fatty acid composition with fatty acid
100 mL of Mixture 1 (250 mM Tris/HCl buffer (pH 8.0), 1
methyl ester (FAME) analysis done by Microcheck, Inc.
mM sodium malate, 2.5 mM NAD), 10 mL of 1.5M sodium
(Northfield, VT).
acetate, 10 mL of Mixture 2 (220 units of malate
dehydrogenase, 35 units of citrate synthase, and 0.1M
potassium phosphate buffer (pH 6.8) to make up a 1-mL
Cell Viability Assay
volume), 10 mL of 15 mM succinyl CoA, and 60 mL of the
cell extract. The reaction rate at 258C was followed by Cell viability assay followed the method previously described
measuring the absorbance at 340 nm, which increased by Glenner (1977). Briefly, 2,3,5-triphenyl-2H-tetrazolium
linearly with time, for 3–5 minutes. chloride (final concentration: 1 g/L) was added to the cell
Phosphoenolpyruvate carboxylase (PEP carboxylase) was suspension and incubated at room temperature for 30 min. for
assayed by the method of Maeba and Sanwal (1969). The color development. After centrifugation at 10,000 rpm for
assay mixture, with a total volume of 300 mL, contained: 10 min, the cell pellets were collected and then suspended in
66.67 mM Tris-HCl buffer (pH 9.0), 66.67 mM NADH, 10 methanol to extract the pink color. The absorbance of the
mM MgCl2, 10 mM NaHCO3 (freshly prepared), 3.33 mM methanol extract at 485 nm, which is proportional to the
PEP, 6.25 U/mL malate dehydrogenase, and the cell extract viable cell number, was measured with a spectrophotometer
(180 mL). The reaction rate at 258C was followed by (Shimadzu, UV-1601). The cell viability (%) is reported with
measuring the absorbance at 340 nm. A blank control had no cells cultured in serum bottles and harvested in the
PEP in the reaction solution. exponential phase as the control with 100% viability.
Unless otherwise noted, the activities of above enzymes
were determined on the basis of the molar extinction
Scanning Electron Microscopy
coefficient of 6.2 mM1cm1 for NADH (Van der Werf
et al., 1997) and one unit of activity is defined as the amount Cells were laid by gentle filtration of cell suspension onto the
of the enzyme catalyzing the reaction at a rate of producing or filter paper (Fisherbrand Grade P5, Fisher Scientific,
consuming 1 mmole of NADH per minute. The specific Hampton, NH). The samples were fixed in 2.5% (w/v)
enzyme activity is defined as the unit of enzyme activity per glutaraldehyde for 15 h at 48C and rinsed with distilled water
mg of total protein. All enzyme activity assays were done in twice. The samples were processed through a progressive
duplicate. dehydration with 20–100% ethanol at 10% increment, dried

SUWANNAKHAM AND YANG: ENHANCED PROPIONIC ACID FERMENTATION BY P. ACIDIPROPIONICI MUTANT 327
with HMDS, and coated with gold/palladium. The samples
were scanned and photographed with a Philips XL 30
scanning electron microscope at 15 kV.
Analytical Methods
Fermentation samples were analyzed for the optical density
at 600 nm (OD600) using a spectrophotometer (Sequoia-
turner, Model 340). One unit of OD600 was equivalent to
0.435 g/L of cell dry weight. The concentrations of glucose
and acid products (mainly, propionic, acetic, and succinic
acids) were analyzed by high-performance liquid chromato-
graphy (HPLC) with a Bio-Rad HPX-87H organic acid
analysis column at 458C using 0.01 N H2SO4 at 0.6 mL/min
as the eluent. The HPLC system (Shimadzu) consisted of a
controller, a pump (LC-10Ai), an automatic injector (SIL-
10Ai), a column oven (CTO-10A), a refractive index detector
(RID-10A), and a computer with data analysis software
(Shimadzu version 4.2).
RESULTS AND DISCUSSION
Fermentation Kinetics
Figure 1 shows typical kinetics for fed-batch fermentations
with free cells in a stirred-tank fermentor and immobilized
cells in the fibrous bed bioreactor. These fed-batch
fermentations were allowed to continue until cells ceased
to consume glucose or produce propionic acid. The purpose
of the experiment was to allow cells to gradually adapt to the

Figure 1. Fed-batch fermentations of glucose by P. acidipropionici at pH 6.5, 328C. (A) Free-cell fermentation. (B) Immobilized-cell fermentation in the
FBB.

328 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 91, NO. 3, AUGUST 5, 2005

high-propionate concentration environment to test the
maximum propionic acid concentration that can be produced
by the bacteria. As seen in Figure 1, the immobilized-cell
fermentation in the FBB produced more propionate from
glucose and reached a higher final propionic acid concentra-
tion of 71.8 g/L, which was &40% higher than that from the
free-cell fermentation (52.2 g/L). It should be noted that 71.8
g/L is the highest propionic acid concentration ever produced
in propionic acid fermentation. Previously, the maximum
propionate concentration produced from glucose was 57 g/L
by a propionate-tolerant strain P. acidipropionici P200910
immobilized in calcium alginate beads (Paik and Glatz,
1994). A propionic acid concentration of 65 g/L was also
attained from fermentation with cells immobilized in the
FBB using de-lactose whey permeate as the substrate (Yang
et al., 1995). The higher final product concentration obtained
in the fermentation would facilitate product separation and
recovery, and significantly reduce the production costs (Van
Hoek et al., 2003).
The higher propionic acid concentration produced in the
immobilized-cell fermentation indicated that cells in the
FBB were possibly less sensitive to propionic acid inhibition.
As shown in Figure 2, at comparable propionic acid
concentrations, the volumetric productivity for propionic
acid was much higher in the immobilized-cell fermentation
than in the free-cell fermentation. The higher volumetric
productivity in the immobilized-cell fermentation could also
be attributed to the high cell density (>45 g/L reactor
volume) and cell viability in the FBB. At the end of the fed-
batch fermentation, the total amount of cells in the FBB
was &32 grams, of which more than 95% was immobilized
in the fibrous matrix, with an average cell viability of greater
than 70%. The volumetric productivities shown in Figure 2
were calculated based on the total volume of the liquid
medium (2 L) in the fermentation, instead of the actual
working volume of the FBB (0.69 L), which was about one
Figure 2. Effect of propionic acid on the volumetric productivity of
propionic acid in fed-batch fermentations with free cells and immobilized
cells.
SUWANNAKHAM AND YANG: ENHANCED PROPIONIC ACID FERMENTATION BY P. ACIDIPROPIONICI MUTANT
third of the total liquid volume. It should be noted that the
specific productivity in the immobilized-cell fermentation
was approximately 80% of that in the free-cell fermentation
because of the much higher cell density in the immobilized-
cell system (11.5 g viable cell/L vs. 2.4 g viable cell/L for the
free-cell fermentation). A lower specific productivity was
often found in immobilized-cell fermentations (Goncalves
et al., 1992; Krischke et al., 1991; Norton et al., 1994;
Senthuran et al., 1997) because of nutrient limitation and
reduced cell growth.
Compared to the free-cell fermentation, the immobilized-
cell fermentation in the FBB produced more propionic acid
and less succinic and acetic acids from glucose. Figure 3
shows cumulative consumption of glucose and production of
propionate, acetate, and succinate in the fed-batch fermenta-
tions. It is noted that the overall propionic acid yield from the
immobilized-cell fermentation was high, &0.65 g/g, for
propionic acid concentrations up to 45 g/L, but then
decreased to 0.4 g/g as the propionic acid concentration
increased to 71 g/L. In contrast, the average propionic acid
yield from the free-cell fermentation was 0.41 g/g. The lower
propionic acid yield in the free-cell fermentation can be
partially attributed to the fact that more carbon sources are
used for cell growth and energy production. In the FBB with
high cell density, cell growth was limited (Huang et al., 2002;
Yang et al., 1994; Zhu and Yang, 2003), as indicated by the
relatively low OD value throughout the fermentation (see
Fig. 1). Furthermore, more acetate and succinate were
produced in the free-cell fermentation than in the immobi-
lized-cell fermentation. As can be seen in Table I, the free-
cell fermentation produced 20% more acetate (0.12 g/g vs.
0.10 g/g) and 100% more succinate (0.18 g/g vs. 0.09 g/g)
than the immobilized-cell fermentation. The reduced cell
growth in the FBB would lower the ATP demand for biomass
formation and thus result in decreased acetate production as
the acetate formation pathway is the major pathway for ATP
production in propioinibacteria (Goswami and Srivastava,
2000; Rickert et al., 1998). The lower succinate production in
the immobilized-cell fermentation can be attributed to
changes in the activities of several key enzymes in the
dicarboxylic acid pathway, which will be discussed later in
this article.
The results from the fermentation experiments clearly
indicate that cells immobilized in the FBB acquired an ability
to tolerate and produce a higher propionate concentration that
could not be achieved by the original culture grown in free
suspension. Also, there was a significant shift in the
metabolic pathway, leading to a higher propionate yield
from glucose and decreased production of acetate and
succinate. Further experiments were thus conducted to study
the underlying changes or causes that allowed the adapted
culture in the FBB to produce more propionate at a higher
concentration than previously reported. To determine if there
were any phenotypic changes in the FBB culture, cells in the
FBB were removed and grown as suspension culture to test
their propionate tolerance. The activities of several key
enzymes in the acid-forming pathways and cell membrane
329

Figure 3. Comparison of product yields from glucose in fed-batch
fermentations with free cells and immobilized cells of P. acidipropionici.
(A) Cumulative propionic acid production vs. glucose consumption. (B)
Cumulative acetic acid production vs. glucose consumption. (C) Cumulative
succinic acid production vs. glucose consumption. The product yields can be
estimated from the slopes of these plots.
ATPase as well as membrane fatty acid composition were
also examined and compared with those of the wild type
culture used in seeding the bioreactor.
Propionic Acid Inhibition
Propionic acid is a strong growth inhibitor (Balamurugan
et al., 1999) and 1% (w/v) propionic acid in the medium could
reduce the specific growth rate of propionibacteria by more
than 50% (Jin and Yang, 1998). Figure 4 compares the specific
growth rates for the wild type and the adapted culture from the
330 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 91, NO. 3, AUGUST 5, 2005
FBB at various initial concentrations of propionic acid in the
growth media. Although the wild type had a slightly higher
growth rate in the absence of propionic acid, the adapted
culture had a significantly higher growth rate for the
propionate concentrations tested between 5 g/L and 60 g/L.
Similar results have been reported for a propionate-tolerant P.
acidipropionici strain, which also had a slightly lower
specific growth rate than the wild-type strain (0.185 h1 vs.
0.199 h1) in the absence of propionic acid but a higher
growth rate at 8% propionic acid (0.047 h1 vs. 0.033 h1)
(Woskow and Glatz, 1991). In this study, the specific growth
rate reduced by 70% for the wild type and 50% for the FBB
adapted culture, respectively, as the propionate concentration
increased to 10 g/L. There was minimal growth when the
propionate concentration was 80 g/L and higher.
The effect of propionic acid on cell growth can be
described by the following noncompetitive product inhibi-
tion model:
mmax Ki 1 1 1
m¼ or ¼ þ P
Ki þ P m mmax mmax Ki
where m is the specific growth rate (h1),
max is the
maximum specific growth rate (h1), K i is the inhibition rate
constant (g/L), and P is the propionic acid concentration
(g/L). The values of
max and Ki were determined from the
linear plot of 1/m vs. P shown in the inset of Figure 4, and are
given in Table II. Compared to the wild type, the adapted
culture had a slightly lower mmax (0.133
0.008 vs.
0.154
0.017 h1), but a much higher Ki (8.93
0.77 vs.
4.33
0.67 g/L). It is clear that the adapted culture is less
sensitive to propionic acid inhibition.
Acid-Forming Enzyme Activities
The cellular activities of several key enzymes in the
dicarboxylic acid pathway (see Fig. 5) that would have
major effects on the production of propionic acid, acetic acid,
and succinic acid were examined and compared between the
mutant strain and the wild type. As shown in Figure 6, the
activities of both oxaloacetate transcarboxylase and propio-
nyl CoA:succinyl CoA transferase (CoA transferase), two
key enzymes in the pathway leading to propionic acid
production, were significantly higher in the mutant than in the
wild type. Oxaloacetate transcarboxylase, a key enzyme in
the pathway for propionate synthesis, catalyzes a coupled
reaction of pyruvate to oxaloacetate and methylmalonyl CoA
to propionyl CoA. It is responsible for supplying the carbon
flux from the central carbon metabolism toward the end
product, propionate. CoA transferase also catalyzes a
coupled reaction of succinate to succinyl CoA and propionyl
CoA to propionate. It was found that CoA transferase was
less active but more sensitive to propionic acid inhibition
than was transcarboxylase for both the mutant and wild type.
Clearly, the reaction catalyzed by CoA transferase was the
rate-limiting step in the pathway leading to propionic acid
production in P. acidipropionici.
 Table I. Comparison of product yields and maximum propionic acid concentrations from fed-batch
fermentations with free cells and immobilized cells in the FBB.

Free-cell fermentations
Immobilized-cell
Wild type Mutanta fermentation

Max. propionic acid concentration (g/L) 52.2
1.1 51.5
0.9 71.8
0.8
Product yield (g/g)
Propionic acid 0.41
0.02 0.47
0.01 0.40  0.65
0.02b
Acetic acid 0.12
0.02 0.11
0.01 0.10
0.01
Succinic acid 0.18
0.03 0.09
0.01 0.09
0.02
Total viable cell concentration (g/L) 2.41 5.67 11.48c
Total carbon recovery (%)d 88.2
8.5 87.1
3.7 95.7
6.0

Note: Mean
standard deviation.

a
Cells from the FBB after adaptation to high propionic acid concentration in the fed-batch fermentation.
b
Yield was &0.65 for propionic acid concentrations up to &45 g/L and then decreased with increasing
propionic acid concentration to &0.40 at &71 g/L.
c
Including all viable cells in suspension and immobilized in the fibrous matrix; based on the total liquid volume
of 2 L.
d
Based on the assumption that 46.2% of the biomass was carbon. The amount of CO2 produced was estimated
based on the assumption that one mole of CO2 was coproduced with each mole of acetic acid in the fermentation.

On the other hand, the activities of phosphotransacetylase
(PTA) and acetate kinase (AK) in the mutant were either
about the same as or slightly lower than those of the wild type
(Fig. 7). PTA and AK are two key enzymes in the pathway
from pyruvate to acetate, which is the major supply route
for ATP. PTA catalyzes the reaction of acetyl CoA to acetyl
phosphate, and AK catalyzes the conversion of acetyl
phosphate to acetate. Both PTA and AK were highly sensitive
to propionic acid inhibition, but PTA had a much
lower activity and appeared to be the rate-limiting enzyme
in the acetate-formation pathway. However, it was not
clear why the activity of PTA increased with increasing
the propionic acid concentration from the minimum level at
60 g/L of propionic acid. The increased PTA activity at
high propionic acid concentrations could be an induced cell
response to direct more pyruvate towards the acetate-forming
pathway instead of the propionate-forming pathway.
Also, the increased PTA activity would increase the cellular
level of acetyl phosphate, which can function as a global
regulator in a metabolic network (McCleary et al., 1993).
Acetyl phosphate can be used in the phosphorylation of a
group of proteins regulating cell’s responses to environ-
mental stimulations. The global phosphorylation could lead
to acetate formation from acetyl phosphate (Wanner and
Wilmes-Riesenberg, 1992) without the activity of acetate
kinase.

Figure 4. Effects of propionic acid on specific growth rates of P. acidipropionici wild type and mutant from the FBB. Inset shows the determination of rate
constants in the noncompetitive inhibition of propionic acid. The curves from the model predictions simulate the data (symbols) well.

SUWANNAKHAM AND YANG: ENHANCED PROPIONIC ACID FERMENTATION BY P. ACIDIPROPIONICI MUTANT 331
Table II. Comparison of rate constants for specific growth rate and several key enzymes in P: acidipropionici wild type and mutant from the FBB.

Wild type Mutant from FBB

mmax (h1) or vmax (U/mg) Ki (g/L) mmax (h1) or vmax (U/mg) Ki (g/L)

Specific growth rate 0.154
0.017 4.33
0.67 0.133
0.008 8.93
0.77
OAA transcarboxylase 0.111
0.001 84.8
4.9 0.138
0.014 451.9
42.9
CoA transferase 0.027
0.001 10.1
0.2 0.059
0.004 13.5
0.5
Phosphotransacetylase 0.0012
0.0001 18.4
0.5 0.0011
0.0001 19.0
0.8
Acetate kinase 0.019
0.001 10.2
0.8 0.021
0.002 10.6
0.5

Note: Mean
standard deviation; n ¼ 2. Both wild type and mutant cells were grown in suspension cultures for specific growth rate determination and
enzyme activity assays.

PEP carboxylase is the enzyme that catalyzes the reaction
from PEP directly to oxaloacetate, thus leading to the
production of succinate from glucose. As shown in Figure 8,
the activity of PEP carboxylase was generally lower in the
mutant than in the wild type. Furthermore, unlike other acid-
forming enzymes, PEP carboxylase was relatively stable in
the presence of propionic acid up to &80 g/L. In fact, the
enzyme activity in the wild type seemed to increase
significantly when the propionic acid concentration increased
from 10 to 60 g/L. These characteristics could explain
why succinate production increased significantly later in the
fed-batch fermentations when the propionic acid concentra-
tion was higher (see Fig. 3C) and why the adapted culture or
mutant in the FBB had much lower succinic acid production
as compared to the wild type. This finding is also consistent
with the result obtained in an extractive fermentation
where propionic acid was continuously removed from
the fermentor and the production of succinic acid was
reduced to almost zero as a result of the low propionic acid
concentration maintained in the fermentation broth (Jin and
Yang, 1998).
In summary, the enzyme activity assay results were in good
accordance with the fermentation results that the mutant
produced more propionic acid and less acetic and succinic
acids than did the wild type. It is clear that changes in various
key enzyme activities in the dicarboxylic acid pathway led to
a metabolic shift to favor more propionic acid production
over acetic and succinic acids. These results also provide
insights on how one could metabolically engineer the
propionibacterium to further increase propionic acid produc-
tion and reduce byproduct formations.
Propionic acid is an inhibitor to oxaloacetate transcarbox-
ylase, CoA transferase, PTA, and AK (Figs. 6 and 7), and the
inhibition can be modeled by the following non-competitive
inhibition kinetics equation:
vmax Ki 1 1 1
v¼ or ¼ þ P
Ki þ P v vmax vmax Ki
where v is the specific activity of enzyme (U/mg), vmax is the
maximum specific activity (U/mg), Ki is the inhibition rate
constant (g/L), and P is the propionic acid concentration

Figure 5. The dicarboxylic acid pathway for the conversion of glucose to propionic, acetic, and succinic acids. The five key enzymes assayed in this study are
labeled in the pathway.

332 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 91, NO. 3, AUGUST 5, 2005

Figure 6. Effect of propionic acid on oxaloacetate transcarboxylase and
propionyl CoA:succinyl CoA transferase activities of P. acidipropionici wild
type and mutant from the FBB. The curves from the non-competitive
inhibition model simulate the data (symbols) well. (A) Oxaloacetate trans-
carboxylase; (B) CoA transferase.
Figure 7. Effect of propionic acid on phosphotransacetylase and acetate
kinase activities of P. acidipropionici wild type and mutant from the FBB.
The curves show the noncompetitive inhibition model predictions
that simulate the data (symbols) well for propionic acid concentrations up
to 60 g/L. (A) Phosphotransacetylase; (B) Acetate kinase.

propionic acid concentration and yield obtained in the fed-

batch fermentation by cells immobilized in the FBB.
(g/L). The best values of vmax and Ki for these enzymes were
determined from nonlinear regression of the data and are Membrane-Bound ATPase
listed in Table II. The higher vmax indicates a more active The uncoupling effect of propionic acid on oxidative
enzyme, whereas the higher Ki indicates that the enzyme is phosphorylation may interfere with the establishment and
less sensitive to propionic acid inhibition. Compared to the
wild type, oxaloacetate transcarboxylase and CoA transfer-
ase in the mutant not only were more active, but also less
sensitive to propionic acid inhibition, especially when the
propionic acid concentration was lower than 60 g/L. In fact,
the mutant’s transcarboxylase activity was almost unaffected
until the propionic acid concentration was higher than 60 g/L
(Fig. 6A). In contrast, both PTA and AK in the mutant and
wild type had almost the same vmax and Ki , indicating their
similar activity and sensitivity to propionic acid inhibition.
Compared to the wild type, the higher propionate tolerance of
the mutant was mainly attributed to the improvements in its
oxaloacetate transcarboxylase and CoA transferase that
could maintain relatively high activities and thus allow the
cells to survive even at elevated propionate concentrations.
The alteration in the sensitivity to propionic acid inhibition of Figure 8. Effect of propionic acid on PEP carboxylase activity of P.
these two enzymes also contributed to the increased final acidipropionici wild type and mutant from the FBB.

SUWANNAKHAM AND YANG: ENHANCED PROPIONIC ACID FERMENTATION BY P. ACIDIPROPIONICI MUTANT 333
maintenance of a functional pH-gradient across the cell
membrane for the transport of metabolites (Gutierrez and
Maddox, 1992). To maintain a proper pH gradient, the extra
protons must be pumped out at the cost of ATP via mem-
brane ATPase (Deckers-Hebestreit and Altendorf, 1996;
Kobayashi, 1987; O’Sullivan and Condon, 1999). Thus, an
active ATPase is essential to prevent the acidification of
cytoplasm by propionic acid. The effects of propionic acid
concentration on the membrane-ATPase activities of the
mutant and wild-type strains were studied with cells
harvested at exponential and stationary phases. As shown
in Figure 9, the activity of ATPase increased with increas-
ing propionic acid concentration in the range between 0 and
10 g/L, but then decreased rapidly to zero at 30 g/L.
Surprisingly, the ATPase regained its activity when the
propionic acid concentration was 60 g/L and increased with
increasing propionic acid concentration. Also, cells grown in
the presence of 10 g/L of propionic acid and harvested in the
exponential phase had &5% more ATPase activity as
compared with cells grown in the absence of propionic acid
(data not shown). These results are consistent with the fact
that the ATPase activity increases as the acid concentration
increases (O’Sullivan and Condon, 1999; Zhu and Yang,
2003) or as the extracelluar pH value decreases (Belli and
Marquis, 1991; Miyagi et al., 1994; O’Sullivan and Condon,
1999). In all cases studied, the effect of propionic acid on
ATPase was essentially the same for the mutant and wild-type
Figure 9. Effect of propionic acid on membrane-bound ATPase activity of
P. acidipropionici wild type and mutant from the FBB. (A) Cells in the
exponential phase. (B) Cells in the stationary phase.
334 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 91, NO. 3, AUGUST 5, 2005
strains, but the former appeared to have a consistently lower
ATPase activity level, an indication that the mutant was more
efficient in pumping out protons and could survive in a higher
propionate environment. The lower ATPase activity for cells
in the stationary phase was attributed to the lower ATP
requirement, as cells were not actively growing in the
stationary phase. On the other hand, for cells in the
exponential phase, a lower membrane-bound ATPase activity
can be an indication of a more efficient proton pump, as it has
been reported that the ATPase activity was lower in faster-
growing cells than in more slowly growing cells (O’Sullivan
and Condon, 1999).
Membrane Fatty Acid Composition
It has been reported that microorganisms such as Sacchar-
omyces cerevisiae (Casey and Ingledew, 1986) and Escher-
ichia coli (Ingram, 1976) could increase their tolerance to
organic solvents by regulating their membrane lipid compo-
sition in response to environmental stresses. Compared to the
wild type, the membrane fatty acid composition was
significantly changed in the mutant, which had more
longer-chain saturated fatty acids (C17:0) and less unsatu-
rated fatty acids (C18:1) (Table III), both of which could
decrease membrane fluidity and thus might have contributed
to the increased propionate tolerance.
Morphological Change in Mutant
Cell adaptation and mutation are often accompanied with a
significant change in cell morphology as a response to
environmental perturbations (Jan et al., 2001; Ye et al., 1999).
Propionibacterium acidipropionici has a rod shape; how-
ever, the mutant appeared to be much longer and thinner
than the wild type as observed under SEM (Fig. 10). The
mutant had an average length of 3.2 mm (vs. 1.1 mm for the
wild type) and an average diameter of 0.50 mm (vs. 0.66 mm
for the wild type). Compared to the wild type, the thinner and
longer appearance of the mutant had an &10% increase in its
specific surface area, which should contribute to proportional
increases in substrate uptake and metabolite excretion rates.
This could explain why the mutant might have a more
efficient proton pump and can tolerate a higher propionic acid
concentration. In the FBB high cell density environment with
relatively low glucose and high propionic acid concentra-
tions, the observed morphological change would be a natural
response for cells to adapt and survive. Interestingly, this
dramatic morphological change was permanent and the
Table III. Comparison of membrane fatty acid compositions of
P: acidipropionici wild type and mutant from the FBB.
Fatty Acid Wild type Mutant
C15:0 70.76% 68.5%
C16:0 2.3% 2.9%
C17:0 14.75% 20.06%
C18:1 1.37% 1.24%

Figure 10. Scanning electron micrographs of P. acidipropionici showing
morphological difference between the wild type and the mutant from the
FBB. (A) Wild-type cells with a short rod shape. (B) Mutant cells with an
elongated slimmer rod shape.
mutant preserved this morphology even after numerous
subculturing in normal growth media for one year. However,
this finding is not a surprise as similar observations have also
been reported. Propionibacterium freudenreichii lost its
original rod shape and became shorter under an extreme
acidic pH of 2.0 (Jan et al., 2001). Lactobacillus plantarum
immobilized in chitosan treated polypropylene matrix
experienced a morphological change from a normal rod
shape to a coccoid shape and a metabolic shift from
homofermentative to heterofermentative (Krishnan et al.,
2001). Lactobacillus rhamnosus immobilized on solid
support changed from isolated rods to thinner, chain-
linked consortia during continuous lactic acid fermentation
at high dilution rate (Goncalves et al., 1992).
Effects of Cell Immobilization in FBB
Clearly, adaptation of P. acidipropionici in the FBB
provided an efficient method to obtain a metabolically
robust mutant that can better produce propionic acid
from glucose with a higher concentration and yield. It
should be noted that free cell fermentations carried out
under similar fed-batch conditions could not achieve the
same adaptation effect and was not as effective in obtaining
SUWANNAKHAM AND YANG: ENHANCED PROPIONIC ACID FERMENTATION BY P. ACIDIPROPIONICI MUTANT
propionate tolerant mutants. It was the unique environment in
the FBB that facilitated rapid adaptation of cells and allowed
mutants with higher propionate tolerance to survive when
subjected to adverse conditions. The FBB has also been
successfully used for adapting and screening for acid-tolerant
strains of Clostridium formicoaceticum (Huang et al., 1998)
and Clostridium tyrobutyricum (Zhu and Yang, 2003). The
ability to obtain acid-tolerant mutants in the FBB can be
attributed to: (a) the high cell density (>45 g/L) and viability
(>70%) maintained in the FBB, and (b) distinct physiology
and survivability of immobilized cells resulting from their
direct contact with a solid surface and with each other. None
of these conditions existed in the free-cell fermentation. It
should be noted that the mutant was also tested in free-cell
fed-batch fermentation and gave similar higher propionate
yield and lower succinate yield from glucose, but could not
reach the same high level of propionate concentration as in
the FBB (see Table I). This indicated that besides the
observed mutations in cell morphology, membrane fatty acid
composition, and enzyme activity and sensitivity to propio-
nic acid, there were additional phenotypic changes for cells
immobilized in the FBB that could not be easily duplicated in
the suspension culture.
Immobilization mimics what occurs in nature when cells
adhere and grow on surfaces or within natural structures. The
high cell density in the immobilization system also provides
conditions conducive to cell-to-cell communication. It is
natural for cells to modify their pattern of growth and
replication because of direct contact with a solid surface or
with other cells. Immobilization also results in the change of
physicochemical properties of the microenvironment, includ-
ing the presence of ionic charges, reduced water activity,
altered osmotic pressure, modified surface tension, and cell
confinement. Growth under this situation can induce responses
at both biochemistry and genetic levels that cause fundamental
changes in the cells. As a cellular response to the environ-
mental stress, immobilization caused changes in cell meta-
bolic pathway and morphology (Krishnan et al., 2001). Cells in
an immobilization system are usually maintained in a non- or
low-growing but metabolically active state (Paik and Glatz,
1994), which may have allowed them to better survive adverse
conditions. When subjected to a high ethanol concentration,
immobilized cells were able to maintain a higher cytoplasmic
pH or a more efficient proton pump (Groboillot et al., 1994;
Loureiro-Dias and Santos, 1990). All these factors might have
contributed to the better adaptation and survivability of cells in
a high-density immobilized system such as the FBB.
In the free cell fermentation, both acetate and succinate
production increased with increasing cumulative glucose
consumption (see Fig. 3) or propionate concentration.
Apparently, free cells grown in suspension would need more
energy (ATP) when the propionic acid concentration was
higher to maintain their intracellular pH. Thus, there was a
clear metabolic pathway shift; however, which could not
continue to work at higher concentrations of propionic acid
and the fermentation ceased to produce propionic acid
at &52 g/L. On the other hand, for the FBB fermentation,
335
cells responded differently. Instead of shifting the metabolic
pathway, immobilized cells underwent some mutations,
including changes in morphology and membrane fatty acid
content to increase their tolerance to propionic acid. These
changes allowed the immobilized cells to tolerate and
produce more propioinic acid at a higher concentration.
CONCLUSIONS
Immobilization and adaptation of P. acidipropionici in the
fibrous-bed bioreactor provided an effective means to obtain
a metabolically advantageous mutant with the ability to
tolerate and produce more propionic acid at higher
concentrations and yield from glucose. The mutant had
significant changes in its morphology, membrane lipid
composition, and key enzymes in the pathways leading to
propionic acid and succinic acid. The higher propionate yield
was attributed to the higher activity levels of oxaloacetate
transcarboxylase and CoA transferase in the mutant,
while the lower succinate yield in the mutant resulted from
the lower activity of PEP carboxylase. The increased
propionate tolerance of the mutant is possibly because of
changes in its membrane fatty acid composition and its
slimmer morphology making the proton pump more efficient
as indicated by the lower ATPase activity. The decreased
sensitivity to propionic acid inhibition for the mutant’s
oxaloacetate transcarboxylase and CoA transferase also
allowed the cells to survive and continue to produce
propionate even when the propionate concentration was
high. The higher final product concentration, coupled with
the significantly reduced by-product and cell concentrations
in the fermentation broth should facilitate simple separations
for product purification and reduce the overall production
cost of propionic acid by the FBB fermentation process.
References
Allen SHG, Kellermeyer RW, Stjernholm RL, Wood HG. 1964. Purification
and properties of enzymes involved in the propionic acid fermentation. J
Bacteriol 87:171–187.
Andersch W, Bahl H, Gottschalk G. 1983. Level of enzymes involved in
acetate, butyrate, acetone and butanol formation by Clostridium
acetobutylicum. Eur J Appl Microbiol Biotechnol 18:327–332.
Balamurugan K, Venkata Dasu V, Panda T. 1999. Propionic acid production
by whole cells of Propionibacterium freudenreichii. Bioprocess Eng
20:109–116.
Belli WA, Marquis RE. 1991. Adaptation of Streptococcus mutans and
Enterococcus hirae to acid stress in continuous culture. Appl Environ
Microbiol 57:1134–1138.
Casey GP, Ingledew WE. 1986. Ethanol tolerance in yeasts. Crit Rev
Microbiol 13:219–280.
Deckers-Hebestreit G, Altendorf K. 1996. The F0F1-type ATP synthases of
bacteria: Structure and function of the F0 complex. Annu Rev Microbiol
50:791–824.
Ellman GL. 1959. Tissue sulfhydryl groups. Arch Biochem Biophys 82:
70–77.
Emde R, Schink B. 1990. Enhanced propionate formation by Propionibac-
terium freudenreichii subsp. freudenreichii in a three-electrode
amperometric culture system. Appl Environ Microbiol 56:2771–
2776.
336 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 91, NO. 3, AUGUST 5, 2005
Glenner GG. 1977. Formanzans and tetrazolium salts. In: Lillie RD, editor.
H.J. Conn’s biological strains: A handbook on the nature and uses of the
dyes employed in the biological laboratory. Baltimore, MD: Williams
and Wilkins. p 201–236.
Goncalves LMD, Barreto MTO, Xavier AMBR, Carrondo MJT, Klein J.
1992. Inert supports for lactic acid fermentation—A technical assess-
ment. Appl Microbiol Biotechnol 38:305–311.
Goswami V, Srivastava AK. 2000. Fed-batch propionic acid production by
Propionibacterium acidipropionici. Biochem Eng J 4:121–128.
Groboillot A, Boadi DK, Poncelet D, Neufeld RJ. 1994. Immobilization of
cells for application in the food industry. Crit Rev Biotechnol 14:75–107.
Gu Z, Glatz BA, Glatz CE. 1998. Effects of propionic acid on
propionibacteria fermentation. Enzyme Microb Technol 22:13–18.
Gutierrez NA, Maddox IS. 1992. Product inhibition in a nonmotile mutant of
Clostridium acetobutylicum. Enzyme Microb Technol 14:101–105.
Hsu ST, Yang S-T. 1991. Propionic acid fermentation of lactose by
Propionibacterium acidipropionici: Effects of pH. Biotechnol Bioeng
38:571–578.
Huang YL, Mann K, Novak JM, Yang S-T. 1998. Acetic acid production
from fructose by Clostridium formicoaceticum immobilized in a
fibrous-bed bioreactor. Biotechnol Prog 14:800–806.
Huang YL, Wu Z, Zhang L, Cheung CM, Yang S-T. 2002. Production of
carboxylic acids from hydrolyzed corn meal by immobilized cell
fermentation in a fibrous-bed bioreactor. Bioresource Technol 82:51–59.
Huang Y, Yang S-T. 1998. Acetate production from whey lactose using co-
immobilized cells of homolactic and homoacetic bacteria in a fibrous-
bed bioreactor. Biotechnol Bioeng 60:498–507.
Ingram LO. 1976. Adaptation of membrane lipids to alcohols. J Bacteriol
125:670–678.
Jan G, Leverrier P, Pichereau V, Boyaval P. 2001. Changes in protein
synthesis and morphology during acid adaptation of Propionibacterium
freudenreichii. Appl Environ Microbiol 67:2029–2036.
Jin Z, Yang S-T. 1998. Extractive fermentation for enhanced propionic acid
production from lactose by Propionibacterium acidipropionici. Bio-
technol Prog 14:457–465.
Kobayashi H. 1987. Regulation of cytoplasmic pH in streptococci. In: Reizer
J, Peterkofsky A, editors. Sugar transport and metabolism in Gram-
positive bacteria. London, UK: Ellis Harwood. p 255–269.
Krischke W, Schroder M, Trosch W. 1991. Continuous production of L-lactic
acid from whey permeate by immobilized Lactobacillus casei subsp.
casei. Appl Microbiol Biotechnol 34:573–578.
Krishnan Sudha, Gowda LR, Misra MC, Karanth NG. 2001. Physiological
and morphological changes in immobilized L. plantarum NCIM 2084
cells during repeated batch fermentation for production of lactic
acid. Food Biotechnol 15:193–202.
Lewis VP, Yang S-T. 1992. Continuous propionic acid fermentation by using
immobilized Propionibacterium acidipropionici in a novel packed-bed
bioreactor. Biotechnol Bioeng 40:465–474.
Loureiro-Dias MC, Santos H. 1990. Effects of ethanol on Saccharomyces
cerevisiae as monitored by in vivo 31P and 13C nuclear magnetic
resonance. Arch Microbiol 153:384–391.
Maeba P, Sanwal BD. 1969. Phosphoenolpyruvate carboxylase from
Salmonella typhimurium, strain LT 2. Methods Enzymol 13:283–288.
McCleary WR, Stock JB, Ninfa AJ. 1993. Is acetyl phosphate a global signal
in Escherichia coli? J Bacteriol 175:2793–2798.
Miyagi A, Ohta H, Kodama T, Fukui K, Kato K, Shimono T. 1994. Metabolic
and energetic aspects of the growth response of Streptococcus rattus to
environmental acidification in anaerobic continuous culture.
Microbiology 140:1945–1952.
Norton S, Lacroix C, Vuillemard J-C. 1994. Kinetic study of continuous
whey permeate fermentation by immobilized Lactobacillus helveticus
for lactic acid production. Enzyme Microb Technol 16:457–466.
O’Sullivan E, Condon S. 1999. Relationship between acid tolerance,
cytoplasmic pH and ATP and Hþ-ATPase levels in chemostat cultures of
Lactococcus lactis. Appl Environ Microbiol 65:2287–2293.
Paik H-D, Glatz BA. 1994. Propionic acid production by immobilized cells
of a propionate-tolerant strain of Propionibacterium acidipropionici.
Appl Microbiol Biotechnol 42:22–27.
Playne MJ. 1985. Propionic and butyric acids. In: Moo-Young M, editor.
Comprehensive Biotechnology, Vol. 3. New York: Pergamon. p 731–
759.
Rickert DA, Glatz CE, Glatz BA. 1998. Improved organic acid production by
calcium alginate-immobilized propionibacteria. Enzyme Microb Tech-
nol 22:409–414.
Schulman M, Wood HG. 1975. Succinyl-CoA:propionate CoA-transferase
from Propionibacterium shermanii. Methods Enzymol 35:235–242.
Senthuran A, Senthuran V, Mattiasson B, Kaul R. 1997. Lactic acid
fermentation in a recycle batch reactor using immobilized Lactobacillus
casei. Biotechnol Bioeng 55:841–853.
Silva EM, Yang S-T. 1995. Kinetics and stability of a fibrous-bed bioreactor
for continuous production of lactic acid from unsupplemented acid
whey. J Biotechnol 41:59–70.
Solichien MS, O’Brien D, Hammond EG, Glatz CE. 1995. Membrane-based
extractive fermentation to produce propionic and acetic acids: Toxicity
and mass transfer considerations. Enzyme Microb Technol 17:23–31.
Van der Werf MJ, Guettler MV, Jain MK, Zeikus JG. 1997. Environmental
and physiological factors affecting the succinate product ratio during
carbohydrate fermentation by Actinobacillus sp. 130Z. Arch Microbiol
167:332–342.
Van Hoek P, Aristidou A, Hahn JJ, Patist A. 2003. Fermentation goes large-
scale. CEP 99:37S–42S.
Vogel IA. 1978. Vogel’s textbook of quantitative inorganic analysis. London:
Longman Group Limited. 756 p.
SUWANNAKHAM AND YANG: ENHANCED PROPIONIC ACID FERMENTATION BY P. ACIDIPROPIONICI MUTANT
Wanner BL, Wilmes-Riesenberg MR. 1992. Involvement of phosphotransa-
cetylase, acetate kinase and acetyl phosphate synthesis in control of the
phosphate regulon in Escherichia coli. J Bacteriol 174:2124–2130.
Wood HG, Jacobson B, Gerwin BI, Northrop DB. 1969. Oxalacetate
transcarboxylase from Propionibacterium. Methods Enzymol 8:215–
230.
Woskow SA, Glatz BA. 1991. Propionic acid production by a propionic acid-
tolerant strain of Propionibacterium acidipropionici in batch and semi-
continuous fermentation. Appl Environ Microbiol 57:2821–2828.
Yang S-T, Huang Y, Hong G. 1995. A novel recycle batch immobilized cell
bioreactor for propionate production from whey lactose. Biotechnol
Bioeng 45:379–386.
Yang S-T, Zhu H, Li Y, Hong G. 1994. Continuous propionate production
from whey permeate using a novel fibrous bed bioreactor. Biotechnol
Bioeng 43:1124–1130.
Ye K, Shijo M, Miyano K, Shimizu K. 1999. Metabolic pathway of
Propionibacterium growing with oxygen: enzymes, 13C NMR
analysis, and its application for vitamin B12 production with
periodic fermentation. Biotechnol Prog 15:201–207.
Zhu Y, Wu Z, Yang S-T. 2002. Butyric acid production from acid hydrolysate
of corn fiber by Clostridium tyrobutyricum in a fibrous-bed bioreactor.
Proc Biochem 38:657–666.
Zhu Y, Yang S-T. 2003. Adaptation of Clostridium tyrobutyricum for
enhanced tolerance to butyric acid in a fibrous-bed bioreactor.
Biotechnol Prog 19:365–372.
337