See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/319202863

Improvement of fermentation performance of Gluconobacter oxydans by

combination of enhanced oxygen mass transfer in compressed-oxygen-
supplied sealed system and cell-recycle techni...

Article  in  Bioresource Technology · August 2017

DOI: 10.1016/j.biortech.2017.08.107

CITATIONS READS

4 73

3 authors:

Xin Zhou Zhou Xuelian

Nanjing Forestry University Nanjing Forestry University
20 PUBLICATIONS   76 CITATIONS    11 PUBLICATIONS   27 CITATIONS   

SEE PROFILE SEE PROFILE

Yong xu
Nanjing Forestry University
61 PUBLICATIONS   1,045 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Xylonic acid View project

Whole cell catalysis of enzymatic hydrolysate to aldonic acids by Gluconobacter oxydans View project

All content following this page was uploaded by Xin Zhou on 18 October 2017.

The user has requested enhancement of the downloaded file.

 Bioresource Technology 244 (2017) 1137–1141

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Improvement of fermentation performance of Gluconobacter oxydans by MARK

combination of enhanced oxygen mass transfer in compressed-oxygen-
supplied sealed system and cell-recycle technique
⁎
Xin Zhou, Xuelian Zhou, Yong Xu
Jiangsu Co-Innovation Center of Efficient Processing and Utilization of Forest Resources, Nanjing Forestry University, Nanjing 210037, People’s Republic of China
College of Chemical Engineering, Nanjing Forestry University, Nanjing 210037, People’s Republic of China
Jiangsu Province Key Laboratory of Green Biomass-based Fuels and Chemicals, Nanjing 210037, People’s Republic of China

A R T I C L E I N F O A B S T R A C T

Keywords: Oxygen supply for microbial cultures is often identified as a limiting factor for aerobic fermentation. Through
Oxygen transfer rate (OTR) implementation of an integrated oxygen control strategy, the high oxygen mass transfer rate satisfied cellular
Compressed oxygen-supplied sealed stirred metabolic demands. Gluconobacter oxydans NL71 fermentation of xylose to xylonic acid was improved re-
tank reactor (COS-SSTR) markably. Finally, the productivity of xylonic acid from xylose by biooxidation was markedly increased to
Cell-recycle
32.5 ± 3.1 g/L/h compared to production levels using conventional laboratory-scale bioreactors. By improving
Gluconobacter oxydans
microbial fermentative vitality, we successfully bio-converted 1800 g xylose to 1813 ± 36 g xylonic acid by
Xylonic acid
combination of a fed-batch addition of xylose substrate as well as a cell-recycling strategy. Bioconversion results
demonstrated a highly efficient fermentation model that performs continuous bioreaction, assisting the effort to
industrialize microbial xylonic acid production.

1. Introduction microorganisms. Thus, the enhanced approaches for different micro-

The aerobic microorganisms used in industrial fermentation pro-
cesses generally demand a large quantity of oxygen in order to re-oxi-
dize nicotinamide adenine dinucleotide phosphate (NADPH) and/or
flavin adenine dinucleotide (FADH) to generate metabolic adenosine
triphosphate (ATP) along with by-product synthesis (Huang et al.,
2006; Xu et al., 2009). Problematically, the use of conventionally-aer-
ated bioreactors can lead to microorganism oxygen deprivation in the
fermentation medium due to oxygen’s low extent of solubility in broth,
rapid oxygen consumption rate, and low oxygen transfer rate (OTR)
(Yuan et al., 2016). When dissolved oxygen levels dip below critical
concentration, the metabolism of bacterial cells are significantly re-
stricted (Pollard et al., 2007). This occurrence eventually reduces the
catalytic performance. Thus, oxygen availability is recognized as one of
the most significant factors for aerobic microbial production and fer-
mentation processes at the industrial scale. To date, several techniques
have been employed in bioreactors to enhance the OTR including in-
creased broth agitation, elevated partial pressures of oxygen supplies,
the use of pure oxygen, or outright reconstruction of the fermentation
vessel. However, the turbulence and shear that are associated with high
rates of agitated aeration are often harmful to the relatively fragile
organisms require more targeted researches for improving the oxygen
supply level.
Gluconobacter oxydans (G. oxydans) is a representative obligate
aerobic bacterium, known for rapidly but incompletely oxidizing sugars
and sugar-alcohols. G. oxydans has wide industrial utilization as bio-
catalyst for production of 1,3-dihydroxyacetone, gluconic acid, vitamin
C, and vinegar (Deppenmeier et al., 2002; Gupta et al., 2001). G. oxy-
dans has also been found as an effective biocatalyst for converting xy-
lose to xylonic acid (Buchert et al., 1988; Zhang et al., 2017). Xylonic
acid is considered a potentially valuable chemical compound, due to
the several applications for which it can function. Examples of xylonic
acid utilization include usage as chelator, concrete additive, 1,2,4-bu-
tanetriol precursors, and oil well drilling fluid retarder (Chun et al.,
2006; Deppenmeier et al., 2002).
Metabolically speaking, G. oxydans biocatalysis of xylose is a close-
coupling bio-oxidation reaction of dehydrogenation and cellular re-
spiration that depends heavily on oxygen supply (Holscher et al., 2009).
Hence, oxygen is a major driving force for the serial oxidation reactions
occurring during biocatalytic conversion of xylose to xylonic acid.
Therefore, improvements in oxygen mass transfer should substantially
benefit G. oxydans bioconversion of xylose to xylonic acid. Zhou et al.,

⁎
Corresponding author at: College of Chemical Engineering, Nanjing Forestry University, No. 159 Longpan Road, Nanjing 210037, China.
E-mail address: xuyong@njfu.edu.cn (Y. Xu).

http://dx.doi.org/10.1016/j.biortech.2017.08.107
Received 22 July 2017; Received in revised form 16 August 2017; Accepted 17 August 2017
Available online 20 August 2017
0960-8524/ © 2017 Elsevier Ltd. All rights reserved.
X. Zhou et al.
have successfully demonstrated a method of delivering oxygen to G.
oxydans via compressed oxygen supply, resulting in a xylonic acid
production of 586 g from 600 g xylose in 125 h process (Zhou et al.,
2015). Although absolute production was very high, the volumetric
productivity was only 4.69 g/L/h. This productivity is relatively un-
satisfactory, especially when considering industrial implementation of
G. oxydans biocatalysis. Importantly, there is not any peer-reviewed
literature available to consult concerning the influence of oxygen de-
livery upon xylonic acid biocatalysis by G. oxydans. For this reason,
investigation of the effects of oxygen delivery are scientifically neces-
sary. In order to avoid the great cost associated with providing high
dissolved oxygen levels for fermentation, it is necessary to establish a
proper oxygen deliver system that can improve the productivity of
xylonic acid biotransformation.
In the present study, we have quantitatively demonstrated the sig-
nificance of oxygen delivery to G. oxydans for biotransformation of
xylose to xylonic acid. Oxygen delivery strategies were assessed
through adjustment of agitation, aeration, and oxygen partial pressure.
In addition, to address the feedback inhibition effects of high quantities
of product or substrate, a fed-batch bioprocess based on a cell-recycling
technique was developed. The information obtained provides a feasible
strategy for improving volumetric productivity of G. oxydans in in-
dustrial applications.
2. Materials and methods
2.1. Media and culture conditions
G. oxydans NL71 inoculum were prepared in a 3 L bioreactor (New
Brunswick GelliGen 115), containing 2 L medium (sorbitol 50 g/L,
yeast extract 5 g/L) and cultured for 24 h at 500 rpm and 30 °C. The cell
pellet was then harvested by centrifugation at 6000 rpm for 10 min and
the different cell contents were loaded into the medium (Miao et al.,
2015). The medium for bioconversion (batch or fed-batch) had the
following composition (g/L): yeast extract 5.0, MgSO4 0.5, KH2PO4 1.0,
K2HPO4 2.0, (NH4)2SO4 5.0, and glucose 1.0 per 100 g of xylose. Glu-
cose served as a xylose-cofactor for both batch and fed batch opera-
tions. pH was stabilized at 5.8 at the fermentative stage by automatic
addition of 50% NaOH (w/w) until 5.8 was reached (Buchert et al.,
1988; Zhou et al., 2015).
Bioconversion was conducted in three different reactors: an air-
aerated stirred tank reactor (A-ASTR), a pure oxygen-aerated stirred
tank reactor (O-ASTR), and a compressed oxygen supplied sealed
stirred tank reactor (COS-SSTR) (Zhou et al., 2015). For the A-ASTR,
the airflow/agitation rate was adjusted manually in the range of
1–5 vvm/300–1100 rpm. Oxygen from an oxygen cylinder
(purity ≥ 99.9%) was used to supply oxygen to O-ASTR (0.5 vvm/
500 rpm) and COS-SSTR (gas inlet pressure of 0.02–0.03 MPa).
Foaming was controlled by on-line addition of polyethylene as an an-
tifoam agent in A-ASTR, but not in COS-SSTR. During fermentation,
dissolved oxygen levels are conventionally considered to be in steady-
state, where the oxygen consumption rate is equal to the rate of oxygen
transfer. Once the substrate (xylose) is depleted, dissolved oxygen levels
begin to rebound. To account for this, xylose feed was manually con-
trolled according to changes of dissolved oxygen level in culture.
2.2. Repeated fed-batch culture with cell-recycle
Repeated fed-batch fermentation was carried out under the same
conditions as the in COS-SSTR system. This fermentation process in-
volves removal of the fermentation broth and recycling of G. oxydans
cells. Fermentation broth and G. oxydans cells were separated by cen-
trifugation at 6000 rpm for 10 min. Following centrifugation, the cells
were reloaded into the new medium containing 100 g/L xylose. The
total amount of xylose loaded over the course of repeated fed-batch
fermentation was 300 g/L. The recycled cells were re-implemented a
1138
Bioresource Technology 244 (2017) 1137–1141
total of six times.
2.3. Determination of dissolved oxygen levels and oxygen mass transfer rate
(OTR)
The dissolved oxygen levels were monitored using a dissolved
oxygen electrode (605-ISM, Mettler Toledo, USA) installed in the
bottom half of the reactor. The dissolved oxygen probe was calibrated
and tested before vessel insertion. The dissolved oxygen level was set to
100% in pure-water at 25 °C. A dynamic method was used to measure
the volumetric oxygen mass transfer coefficient (kLa) (Garcia-Ochoa
and Gomez, 2009). Off-gas analysis was used for determination of
oxygen uptake rate, obtained from the slope of the correlation between
dissolved oxygen concentration and time. Dissolved oxygen levels were
only measured when aeration was ceased. When aeration was restarted,
dissolved oxygen concentration increased until it reached steady-state.
Using the estimated oxygen uptake rate, kLa was formulated from the
profile of dissolved oxygen concentrations during biotransformation.
Finally, OTR (mmol O2/h/L) was estimated by the following Eq. (1),
OTR = kL a(C∗−CL) (1)
where C∗ is the concentration of dissolved oxygen (mmol/L) at the
bubble surface and CL is the dissolved oxygen concentration in the so-
lution bulk. The term kLa (h−1) is referred to as the overall volumetric
oxygen transfer coefficient, and was used to characterize the perfor-
mance of the different aeration devices.
2.4. Estimation of biomass concentration
Cell optical density (OD) was determined at 600 nm using a UV–Vis
spectrophotometer (Ultrospec 2100, Amersham Biosciences Corp.,
USA) in glass cuvettes. The path length was 5 mm. The samples were
diluted with double-deionized water to the linear response range
(0.3–0.7 absorbance units). Double-deionized water was used as a blank
control for measurements.
2.5. Analytical methods
Xylose and xylonic acid were simultaneously determined using high
performance anion-exchange chromatography coupled with pulsed
amperometric detection (Thermo ICS-5000). The separation column
used was CarboPac™ PA10 (Wang et al., 2014). Xylonic acid yield (%)
was calculated from the concentration of xylonic acid divided by the
total xylose concentration and multiplied by the constant 0.904. Vo-
lumetric productivity (g/L/h) of xylonic acid generation was calculated
from the concentration of xylonic acid divided by reaction time.
Two parallel assays were performed for each experiment.
3. Results and discussion
3.1. Effect of aeration and agitation rate on xylonic acid production
It has been previously shown that in a shaker flask culture method,
the available oxygen is the only source for surface aeration and this
could be a limiting factor for oxygen-dependent G. oxydans (Ricky
et al., 1996). Thus, the desired amount of xylonic acid production was
difficult to achieve using said system. Zhou et al. have already reported
that A-ASTR could significantly enhances bioconversion by G. oxydans,
due to improved oxygen transfer capacity provided by vigorous stirring
of air into small bubbles (Zhou et al., 2015; Galaction et al., 2010).
From this we have assumed that oxygen supply is a key factor for xy-
lonic acid production by G. oxydans. However, few studies exist fo-
cusing upon xylonic acid production, let alone how aeration affects it.
Thus, the influences of airflow rate and agitation rate upon xylonic acid
production were tested using an air-aerated stirred tank reactor.
Shown in Fig. 1 a, coupled with increases in aeration rate was also
X. Zhou et al.
a that we still observed a decline in the specific rate of oxygen con-
80
xylonic acid
kLa
OTR
60
Xylonic acid (g/L)
40
20
0 10
1 vvm 2 vvm 3 vvm 4 vvm
b 100
xylonic acid
kLa
80 OTR
Xylonic acid (g/L)
60
40
20
0 0
300 rpm 500 rpm 700 rpm 900 rpm
Fig. 1. Xylonic acid production of 10 h, OTR and kLa in A-ASTR, the experiments were
conducted at 300 rpm, 2 OD cell density and xylose concentration (100 g/L). (a) Different
aeration rate in A-ASTR. (b) Different agitation rate in A-ASTR.
increases in xylonic acid production, OTR, and kLa. The results showed
that kLa and OTR linearly rose from 83 ± 6 to 120 ± 8 h−1 and
16.3 ± 1.2 to 26.5 ± 3.4 mmol/L/h, when aeration rate was between
1 and 3 vvm. From these results it can be concluded that proper aera-
tion rate significantly enhances G. oxydans productivity. Greater aera-
tion rates could be providing better gas holdup, the residence time of
dissolved oxygen in the liquid, and higher area for gas-liquid interac-
tions (Kang et al., 2000). However, kLa nor OTR could not be improved
efficiently when aeration rate surpassed 3 vvm. We suggest that this is
due to aeration rate exceeding the flooding point, where air bubbles
were no longer dispersed. In these cases (> 3 vvm), kLa could not be
improved, therefore OTR failed to increase. To counter, an increase in
agitation rate might improve kLa by increasing air dispersion.
When the agitation speed varied between 300 and 1100 rpm, en-
hancements of kLa were observed from the Fig. 1b. The accumulation of
xylonic acid increased by 25%, compared to production (57.8 ± 2.6 g/
L) at 3 vvm and 300 rpm. Although higher agitation rate could sig-
nificantly improve bioconversion kinetics, the productivity of xylonic
acid generation did not increase when agitation speed exceed 500 rpm.
No improvement in OTR was observed at said agitation rate, indicating
that oxygen supply was not the sole limiting factor for xylose conver-
sion. We speculated that 2 OD G. oxydans could not energetically afford
to convert any more xylose. This speculation is supported by the dis-
solved oxygen concentration (CL) in solution, which increased when
agitation speeds were above 500 rpm (data was not shown). These re-
sults indicate that the dissolved oxygen in broth could not be consumed
at a suitable pace. Thus, the productivity of xylonic acid generation
would no longer increase despite expectations. It is important to note
Bioresource Technology 244 (2017) 1137–1141
150 sumption (data was not shown). Therefore, excessive agitation speed
exerted negative effects upon the productivity of xylonic acid produc-
tion by G. oxydans. A possible explanation for this observation could be
kLa (h-1) and OTR (mmol/h/L)
120
that the higher agitation induced frictional forces which inevitably lead
to microbial damage. Taking the above observations into consideration,
90
we believed that xylonic acid productivity could be improved through
increases in aeration rate and agitation rate.
30 Importantly, G. oxydans tends to not metabolize xylonic acid. This
signifies that there is no further catabolism of the xylonic acid molecule
or release of carbon dioxide. In addition, the use of sodium hydroxide
20 for adjustment of pH also needs to be considered. In conclusion, only
H2O was formed under the optimum cell biocatalysis conditions. In a
word, no exhausted gas give off. Thus, bioconversion of xylose to xy-
lonic acid can be summarized by the following chemical Eq. (2):
5 vvm
C5 H10 O5 (xylose) + 1/2O2 + NaOH = C5 H9 O6 ·Na(xylonate sodium)
+ H2 O (2)
250
That’s to say, almost all oxygen uptake was metabolically used for
oxidizing the xylose to xylonic acid. OTR was equal to half of the xy-
200 kLa (h-1) and OTR (mmol/h/L) lonic acid molar productivity, caused from increased levels of xylose
oxidation. Thus, further improvement of OTR should be studied to
150 improve the production efficiency of xylonic acid by G. oxydans.
50 3.2. Enhancement of OTR with different oxygen control strategies
Based on previous discussion, achievement of further-improved
25 xylonic acid productivity depends on achieving a sufficient OTR value.
The improvement in the aeration rate and the agitation rate enhanced
the kLa and OTR. However, OTR could not be increased continuously
even the agitation kept increasing. According to our deduced Eq. (1), it
1100 rpm can be concluded that OTR is also influenced by ΔC (the difference
between C∗ and CL). C∗ could be increased by the control of atmospheric
partial pressure or the usage of pure oxygen. Moreover, higher cell
content could effectively increase the consumption of oxygen and de-
crease CL (Garcia-Ochoa and Gomez, 2009). In attempt of further im-
provement, batch fermentation of xylonic acid was conducted in A-
ASTR, O-ASTR and COS-SSTR. In addition, 2 OD and 6 OD G. oxydans
cells were separately loaded into the medium for investigation the ef-
fect of the CL.
As shown in Fig. 2, the productivity of xylonic acid was predictably
enhanced, especially with O-ASTR (agitation rate of 500 rpm and pure
oxygen rate of 0.5 vvm) and COS-SSTR (agitation rate of 500 rpm). In
the A-ASTR system (agitation rate of 500 rpm and air rate of 3 vvm),
the productivity of 6 OD cell loading rose, but only by 20%. Fig. 2a
shows that in A-ASTR, dissolved oxygen levels dropped to values below
10% when cell loading loaded was 6 OD. This indicates that the OTR in
this system might be insufficiently supplying dissolved oxygen for 6 OD
of G. oxydans cells. Different for the O-ASTR and COS-SSTR systems,
xylonic acid productivity increases were almost proportional to the
increase in initial cellular inoculum. Fig. 2b and c clarify that the dis-
solved oxygen levels were all higher than 50%. In fact, when using the 6
OD inoculum, 100 g/L of xylose was consumed in ∼4 h, and more than
100 g/L xylonic acid was produced. As expected, oxygen limitations
were avoided through implementation of O-ASTR and COS-SSTR.
Although O-ASTR could effectively improve OTR and shorten pro-
duction time, the expense of pure-oxygen supply in aeration fermen-
tation would be considerable for large-scale industrial production.
Using COS-SSTR, no further catabolism is performed by G. oxydans,
hence the supplied oxygen was used exclusively for oxidation of xylose
to xylonic acid (with H2O as the sole by-product). In the COS-SSTR
system, oxygen supply is automatically equalized with no gas exhaus-
tion, avoiding oxygen waste. Therefore, the design of COS-SSTR is a
more financially feasible mode of xylonic acid production. In the
aerobic stirred bioreactor, the adsorption of cells to gas-liquid interface,
1139
X. Zhou et al. Bioresource Technology 244 (2017) 1137–1141

a 120
2-OD-xylose
6-OD-xylose
2-OD-xylonic acid
6-OD-xylonic acid 100
100
Production
Productivity
35

2-OD-dissolved oxygen 6-OD-dissolved oxygen 80 28

100

Productivity(g/L/h)
80

Production (g/L)
Xylose and Xylonic acid (g/L)

Dissolved oxygen level (%)

60 21
80
60
40 14
60
40 20 7
40

20 0 0
20 2-OD 4-OD 6-OD 8-OD 10-OD12-OD

Fig. 3. Xylonic acid production and productivity (in 2 h) with different cell content in
0 0 COS-SSTR.
0 2 4 6 8 10
Time (h) productivity (32.5 ± 3.1 g/L/h) increased slightly (Fig. 3). This sug-
gested that the oxygen supply in COS-SSTR was not required for more
b 120
2-OD-xylose
6-OD-xylose
2-OD-xylonic acid
6-OD-xylonic acid
250
than 8 OD cell content. In other words, the oxygen supply in COS-SSTR
2-OD-dissolved oxygen 6-OD-dissolved oxygen was not required for supporting greater cellular abundance. Thus, cel-
lular loading of 8 OD was adopted for the next study.
100
200
Xylose and Xylonic acid (g/L)

Dissolved oxygen level (%)

80 3.3. Oxydans fed-batch fermentation in COS-SSTR

150
60 Because high xylose quantities exert inhibitory effects on its bio-
conversion by G. oxydans (Zhou et al., 2015), a fed-batch strategy for
100
xylose was adopted. The feeding time was dependent on changes in
40
dissolved oxygen levels. Rebounding dissolved oxygen levels indicate
50 that the oxygen consumption rates have begun to decline, signaling a
20 shortage of substrate for microbial fermentation. Based on this, the fed-
batch was implemented in COS-SSTR (agitation rate of 500 rpm and 8
0 0 OD cell content) to improve cell bioconversion efficiency and achieve
0 2 4 6 8 10 high xylonic acid production. Results showed in Fig. 4, it showed that
Time (h) 406.3 ± 7.2 g/L of xylonic acid was produced from 376 g/L of xylose,
with a corresponding yield of 97.3 ± 1.7%. In addition, overall volu-
c 120
2-OD-xylose
6-OD-xylose
2-OD-xylonic acid
6-OD-xylonic acid 600
metric productivity of 16.9 ± 0.28 g/L/h was achieved. While the
final quantities of xylonic acid produced was inspiring, the desired
2-OD-dissolved oxygen 6-OD-dissolved oxygen
productivity levels could still not be reached. The results do indicate
100 that 300 g/L of xylose was almost fully converted within 15 h. How-
Xylose and Xylonic acid (g/L)

500 ever, for the remaining 100 g/L xylose, ∼10 more hours were needed
Dissolved oxygen level (%)

80 to achieve total conversion. It is evident from the time course of fed-

batch that the productivity suffered a dramatic decline when xylose
60 400 loading surpassed 300 g/L. As can be observed from Fig. 4, the solution
viscosity continually increased with increases in xylose loading. When
the loaded xylose levels reached over 300 g/L, liquid viscosity was over
40
5.5 mPa·s, and the dissolved oxygen level fell below 10%. Dissolved
300
oxygen in highly viscous broth exerted negative effects upon oxygen
20
450 9
Xylonic acid
0 200 Xylose
0 2 4 6 8 10 Vixcosity
Xylose and Xylonic acid (g/L)

360
Apparent viscosity (mPa·s)

Dissloved oxygen
Dissloved oxygen level (%)

Time (h)
6
Fig. 2. Fermentation profiles of different G. oxydans cell content at various bioreactor. (a) 270
A-ASTR (500 rpm, 3 vvm-air). (b) COS-SSTR (500 rpm, 0.02 MPa-compressed oxygen).
(c) O-ASTR (500 rpm, 0.5 vvm-oxygen).
180
3
with the increased apparent viscosity of broths, which is superior to the
90
water, boosts the bubbles coalescence, thus resulting in large bubble
formation and decrease of interfacial area or dissolved oxygen (Ozbek
and Gayik, 2001). In COS-SSTR, the issue of foam formation is elimi- 0 0
0 6 12 18 24
nated due to pressure control in a sealed system, which would also
Time (h)
avoid necessity of anti-foaming agents. To further improve the pro-
duction performance of COS-SSTR, the effect of higher cell loading was Fig. 4. The bioconversion of xylose to xylonic acid in COS-SSTR and fed-batch operation
examined. When cell loading exceeded 8 OD, the xylonic acid with four-batch addition of 100 g xylose at 0 h, 2.5 h, 7 h and 11.5 h.

1140
X. Zhou et al.
350
280
Dissolved oxygen level (%)
Xylonic acid (g/L)
210
140
70
0 0
0 12 24 36 48 60 72
Time (h)
Xylonic acid Loaded-xylose
Fig. 5. The profile of xylose to xylonic acid in COS-SSTR with fed-batch operation and
cell-recycle technique (6 times).
transfer. The accumulated viscosity of the broth could lead to greater
film resistance to oxygen transfer, resulting in limitations to xylonic
acid production or productivity (Garcia-Ochoa and Gomez, 2009). All
evidences suggested that the OTR was repressed when the total xylose
concentration surpassed 300 g/L, and that the diminished OTR directly
contributed to the observed decline in xylonic acid productivity.
3.4. Repeated fed-batch fermentation with cell-recycle technique
The overarching objective of this study was to convert the maximum
amount of xylose to xylonic acid through improvements to G. oxydans
bioconversion efficiency. Results from the previous sections demon-
strate that solutions containing higher concentrations of xylose or xy-
lonic acid can inhibit xylose oxidation by G. oxydans, attributed to a
sharp decline in OTR. To address this issue, a cell-recycle technique was
employed to lift oxygen transfer restrictions. As is well known, it is
difficult to incite G. oxydans proliferations (Gupta et al., 2001). Forcing
proliferation of G. oxydans cells will likely be costly. Therefore, effec-
tive utilization of G. oxydans cells is essential for successful industrial-
scale xylonic acid production. To improve cell utilization, we con-
ducted fed-batch fermentation using recycled cells within the COS-
SSTR system. This fermentation process involves removal of the fer-
mentation broth and recycling G. oxydans cells.
The fermentation profile obtained following six cycles of cell re-
cycling is shown in Fig. 5. It was observed that the activity of G. oxydans
cells was continuous and steady. For each experiment, a total of 300 g
xylose was loaded in the broth and allowed to ferment for 12–24 h.
After time, 1813 ± 36 g xylonic acid was produced in 1 L broth. When
the cells were reloaded into the new medium, productivity of xylonic
acid generation was recovered, similar to the first fed-batch fermenta-
tion. This once again indicated that the performance of G. oxydans was
restricted by diminished oxygen transfer to highly concentrated and
viscous fermentation broths. The results also reveal that overall volu-
metric productivity from cell-recycle fermentation decreases followed
repeated fermentation. Since the G. oxydans cells do not use xylose as
carbon source for cellular growth during fermentation, damage and/or
death of the remaining G. oxydans cells was inevitable with prolonged
time. Thus, fermentation time can also be considered as a significant
restriction to efficient xylonic acid production. These observations
suggest that cell-recycle is an efficient model for improving G. oxydans
utilization rate.
View publication stats
Bioresource Technology 244 (2017) 1137–1141
2000 4. Conclusions
1600 In this study, continuous xylonic acid fermentation was markedly
improved by an obligate aerobic bacterial reactor modification and
Loaded-xylose (g/L)
implementation of a cell-recycle system which was an effective tool to
1200
obtain the high xylonic acid production. We have confirmed that
oxygen supply play a critical role in production of xylonic acid. The
800 strategy of oxygen control was conducted through adjustments to agi-
tation, aeration, and oxygen partial pressure. All adjustments resulted
400 in a COS-SSTR system to significantly improve xylonic acid output and
enhance cellular productivity.
84 96 108 Acknowledgements
Dissolved oxygen The research was supported by the Key Research and Development
Program of Jiangsu (BE2015758), and the National Natural Science
Foundation of China (31370573). Also, the authors acknowledge fi-
nancial support from the Doctorate Fellowship Foundation of Nanjing
Forestry University for supporting the work presented in this paper.
References
Buchert, J., Puls, J., Poutanen, K., 1988. Comparison of Pseudomonas fraoi and
Gluconobacter oxydans for production of xylonic acid from hemicellulose hydro-
lyzates. Appl. Biochem. Biotechnol. 28, 367–372.
Chun, B.W., Dair, B., Macuch, P.J., Wiebe, D., Porteneuve, C., Jeknavorian, A., 2006. The
development of cement and concrete additive: based on xylonic acid derived via
bioconversion of xylose. Appl. Biochem. Biotechnol. 131, 645–658.
Deppenmeier, U., Hoffmeister, M., Prust, C., 2002. Biochemistry and biotechnological
applications of Gluconobacter strains. Appl. Biochem. Biotechnol. 60, 233–242.
Galaction, A.I., Cascaval, D., Camarut, S., 2010. Analysis of distribution of oxygen
transfer rate in stirred bioreactors for yeasts broths. Rom. Biotechnol. Lett. 15,
5423–5435.
Garcia-Ochoa, F., Gomez, E., 2009. Bioreactor scale-up and oxygen transfer rate in mi-
crobial processes: an overview. Biotechnol. Adv. 27, 153–176.
Gupta, A., Singh, V.K., Qazi, G.N., Kumar, A., 2001. Gluconobacter oxydans: its bio-
technological applications. J. Mol. Microbiol. Biotechnol. 3 (3), 445–456.
Hölscher, T., Schleyer, U., Merfort, M., Bringer-Meyer, S., Gorisch, H., Sahm, H., 2009.
Glucose oxidation and PQQ-dependent dehydrogenases in Gluconobacter oxydans. J.
Mol. Microbiol. Biotechnol. 16, 6–13.
Huang, W.C., Chen, S.J., Chen, T.L., 2006. The role of dissolved oxygen and function of
agitation in hyaluronic acid fermentation. Biochem. Eng. J. 32, 239–243.
Kang, X., Wang, Y., Harvey, L.M., Mcneil, B., 2000. Effect of air flow rate on scleroglucan
synthesis by Sclerotium glucanicum in an airlift bioreactor with an internal loop.
Bioprocess Eng. 23, 69–74.
Miao, Y., Zhou, X., Xu, Y., Yu, S., 2015. Draft genome sequence of Gluconobacter oxydans
NL71, a strain that efficiently biocatalyzes xylose to xylonic acid at a high con-
centration. Genome Announc. 3, 15–16.
Ozbek, B., Gayik, S., 2001. The studies on the oxygen mass transfer coefficient in a
bioreactor. Process Biochem. 36, 729–741.
Pollard, D.J., Kirschner, T.F., Hunt, G.R., Tong, I.T., Stieber, R., Salmons, P.M., 2007.
Scale up of a viscous fungal fermentation: Application of scale-up criteria with regime
analysis and operating boundary conditions. Biotechnol. Bioeng. 96, 307–317.
Ricky, L., Denis, P., Ronald, J.N., 1996. Enhancement of oxygen transfer rate using mi-
croencapsulated silicone oils as oxygen carriers. J. Chem. Technol. Biotechnol. 68,
37–46.
Wang, X., Xu, Y., Lian, Z.N., Yong, Q., Yu, S.Y., 2014. A one-step method for the si-
multaneous determination of five wood monosaccharides and the corresponding al-
donic acids in fermentation broth using high-performance anion-exchange chroma-
tography coupled with a pulsed amperometric detector. J. Wood Chem. Technol. 34,
67–76.
Xu, H., Dou, W., Xu, H., Zhang, X., Rao, Z., Shi, Z., Xu, Z., 2009. A two-stage oxygen
supply strategy for enhanced l-arginine production by Corynebacterium crenatum
based on metabolic fluxes analysis. Biochem. Eng. J. 43, 41–51.
Yuan, J., Wu, M., Lin, J., Yang, L., 2016. Enhancement of 5-keto-d-gluconate production
by a recombinant Gluconobacter oxydans using a dissolved oxygen control strategy. J.
Biosci. Bioeng. 122, 10–16.
Zhang, H., Han, X., Wei, C., Bao, J., 2017. Oxidative production of xylonic acid using
xylose in distillation stillage of cellulosic ethanol fermentation broth by Gluconobacter
oxydans. Bioresour. Technol. 224, 573–580.
Zhou, X., Lu, S., Xu, Y., Mo, Y., Yu, S., 2015. Improving the performance of cell bioca-
talysis and the productivity of xylonic acid using a compressed oxygen supply.
Biochem. Eng. J. 93, 196–199.
1141