Overexpression of membrane-bound gluconate-2-dehydrogenase to enhance the production of 2-keto-d-gluconic acid by Gluconobacter oxydans

Overexpression of ga2dh in G. oxydans DSM2003

Efficient expression of ga2dh (gox1230-1232) was essential for enhancing the 2KGA production in G. oxydans DSM2003. A well-characterized promoter is a prerequisite for the overexpression of an enzyme.
To achieve the optimum recombinant strains, three different promoters (G. oxydans_tufB, G. oxydans_ gHp0169 and the native promoter of the ga2dh gene) were introduced into the broad-host-range vector pBBR1MCS5 for expression of
the ga2dh gene. G. oxydans_tufB and G. oxydans_ gHp0169 were confirmed to be strong promoters for cloning and expression of homologous and
heterologous genes in G. oxydans18]. The three recombinant strains (G. oxydans_tufB_ga2dh G. oxydans_g2adh_ga2dh, and G. oxydans_gHp0169_ga2dh) and the control strain G. oxydans_pBBR1MCS5 were cultured in shaking flasks, and the activities of the obtained resting
cells toward GA for 2KGA production were compared. Growth behaviors of the recombinant
strains were similar to that of G. oxydans_pBBR1MCS5, but the biomass at the late-log phase was slightly lower than that of the
control strain (Table 1). During the biocatalysis of GA by resting cells, all of the ga2dh-overexpressing strains produced concentrations of 2KGA higher than that of the control
strain. Amongst these strains, G. oxydans_tufB_ga2dh and G. oxydans_gHp0169_ga2dh exhibited the highest specific productivities of 2KGA (0.83 and 0.85 g/g/h), about
100 % higher than that of G. oxydans_pBBR1MCS5 (0.42 g/g/h).

Table 1. Cell growth of different G. oxydans strains and their specific2KGA productivities

Concerning the biomass and specific productivity of 2KGA production, the optimal strain
G. oxydans_tufB_ga2dh was selected and batch bioconversion of GA by resting cells was conducted in a 7-L
fermenter with a GA concentration increased to 320 g/L. During the bioconversion,
the agitation speed and aeration rate were controlled at 600 rpm and 8 L/min, respectively.
pH was controlled at 5.8 using 4 mol/L NaOH solution. As shown in Fig. 1, almost all the GA was converted to 2KGA by 30 g
_{(wet wt)}
/L cells of the engineered strain in about 25 h, generating 319 g/L 2KGA at a productivity
of 12.76 g/L/h. In contrast, the control strain G. oxydans_pBBR1MCS5 required 51 h to complete this reaction,generating 307 g/L 2KGA at a productivity
of 6.02 g/L/h. The results show that enhanced expression of the ga2dh gene in G. oxydans under the control of tufB promoter efficiently improved the production yield of 2KGA from GA.

Fig. 1. Comparison of 2KGA production by G. oxydans_pBBR1MCS5 and G.oxydans_tufB_ga2dh. The biotransformations were carried out in 7-L fermenterat 30 °C, pH 5.8, 600 rpm,
aeration rate 8 L/min and cell concentration 30 g
_{(wet wt)}
/L. Gluconobacter oxydans_pBBR1MCS5 (blank), G. oxydans_tufB_ga2dh (filled)

As expected, the transcriptional levels of the ga2dh gene in the engineered strain (G. oxydans_tufB_ga2dh) were significantly enhanced (Fig. 2). The ga2dh expression level obtained was normalized in the control strain G. oxydans_pBBR1MCS5. The transcriptional abundance of the three subunits (gox1230, 1231 and 1232) of the ga2dh gene in G. oxydans_tufB_ga2dh were 180, 35 and 60-fold higher than those of the control strain G. oxydans_pBBR1MCS5 respectively. But the transcriptional abundance of the gdh gene (gox0265) in G. oxydans_tufB_ga2dh was about 85 % of that in G. oxydans_pBBR1MCS5.

Fig. 2. Relative transcriptional abundance of the ga2dh and gdh gene G. oxydans_pBBR1MCS5 (blank) and G. oxydans_tufB_ga2dh (shadow)

Optimization of the biocatalysis conditions by resting G. oxydans_tufB_ga2dh cells

To explore the potential of G. oxydans_tufB_ga2dh in 2KGA production and achieve a high production titer, the biocatalysis conditions
for 2KGA production from GA were optimized. In a 7-L fermenter, pre-experiments had
proven that the optimum reaction temperature and pH were 30 °C and 5.8, respectively.

A suitable amount of cell content is necessary for high 2KGA production and the economic
feasibility of the bioprocess. To determine the effect of the cell content on 2KGA
production, various concentrations of resting G. oxydans_tufB_ga2dh cells were used to catalyze 320 g/L GA. As shown in Fig. 3, 2KGA production and the reaction rate increased with the cell concentration increasing
to 30 g
_{(wet wt)}
/L, and then were nearly constant as cell concentration continued to increase. In
the presence of 30 g/L resting cells, 2KGA accumulation linearly increased and reached
a maximum after 25 h, at which time all GA had been converted to 2KGA, resulting in
the highest productivity of 12.76 g/L/h.

Fig. 3. Effect of cell concentration on 2KGA production. The biotransformations were conducted
by 10, 20, 30, 40 and 60 g/L G. oxydans_tufB_ga2dh resting cells, respectively

The effect of initial GA concentration on 2KGA production was also investigated. Reactions
with four different GA concentrations (320, 380, 440 and 480 g/L) were conducted with
30 g
_{(wet wt)}
/L resting cells at pH 5.8 and 30 °C. Almost all the GA at different concentrations
were converted to 2KGA with yields close to 100 %, but the productivity decreased
with increasing GA concentration, because of the extension of reaction time when the
substrate concentration was increased (Fig. 4a). The 2KGA productivities were 12.76, 9.04, 5.93 and 4.93 g/L/h, at initial GA concentrations
of 320, 380, 440 and 480 g/L, respectively. At a high initial GA concentration of
480 g/L, an enhanced concentration of resting cells (60 g
_{(wet wt)}
/L) had no effect on the productivity for 2KGA production (Fig. 4b). Both GA conversion rate and 2KGA production rate with 60 g
_{(wet wt)}
/L resting cells were identical with those at 30 g
_{(wet wt)}
/L resting cells.

Fig. 4. Effect of initial GA concentration on 2KGA production. a The initial GA concentration were 320, 380, 420 and 480 g/L, respectively. Cell concentration
was 30 g
_{(wet wt)}
/L. GA (open); 2KGA (filled), b Initial GA concentration 480 g/L; cell concentration 60 g
_{(wet wt)}
/L. GA (open); 2KGA (filled). c DO profile during the batch bioconversion, cell concentration 30 g
_{(wet wt)}
/L

Given that the formation of 2KGA from GA requires oxygen as the final acceptor of
electrons formed during the oxidation of GA, oxygen conditions were maintained via
a high agitation speed (600 rpm) and air flow rate (8 L/min) throughout the batch
bioconversion period. However, as shown in Fig. 4c, it was observed that when the reaction started, dissolved oxygen (DO) sharply decreased
to 0 % air saturation, and then remained constant (less than 0 %) until the latter
stages of the reaction (about 20 % of GA remaining). This may imply that DO is an
important factor influencing the conversion of GA to 2KGA.

The main factors controlling DO concentration during biotransformation are the degree
of agitation, gas flow rate and oxygen partial pressure in the supplied gas. Because
of the limitations of the fermenter design, the agitation speed and gas flow rate
could not be increased further. Thus, to increase DO levels, oxygen instead of air
was supplied continuously to support the oxidation of 480 g/L GA by 30 g
_{(wet wt)}
/L resting cells. Under this condition, the agitation speed and oxygen flow rate were
controlled at 600 rpm and 1 L/min, respectively. The DO level in the reaction mixture
was maintained above 100 % throughout the process. The time course for GA consumption
and 2KGA production are shown in Fig. 5a. During the first 24 h of batch bioconversion, approximately 50 % of GA was linearly
reduced and produced 199.42 ± 20.34 g/L 2KGA. Both 2KGA production and the conversion
rate of GA to 2KGA were higher than those in bioconversion experiments under continuous
air supply, in which a 2KGA titer of 129.42 ± 3.43 g/L and conversion rate of 27.4 %
were obtained at 24 h (Fig. 4a). After 24 h, a continuous rise in 2KGA titer was accompanied by a gradual decrease
in GA levels, as observed by declines in the product formation and substrate consumption
rates. By 108 h, all GA was completely transformed into 2KGA at a level of 461.09 g/L
with a productivity of 4.27 g/L/h. To determine the reasons for this behavior, samples
of resting cells in the reaction mixture were taken at different reaction times and
the relative activities toward GA were determined. The catalytic activities of the
samples at the beginning of the reactions from air supply experiments or oxygen supply
experiments were set at 100 %. As shown in Fig. 5b, the catalytic activities of the resting cells in these two experiments decreased
with the extension of reaction time; however, there was a more marked loss of activity
in the oxygen supply experiment than in the air supply experiment. Catalytic activities
of resting cells in the oxygen supply experiment decreased by 40 % in 24 h, and showed
33 % activity at the end of bioconversion. The results clearly reveal that high oxygen
levels suppressed the oxidative activity of resting cells toward GA, and an excess
of oxygen during bioconversion may result in decreasing productivity. Therefore, an
optimal oxygen level is important for high 2KGA productivity.

Fig. 5. Effect of DO control strategy on bioconversion of GA to 2KGA. a Time course of GA consumption and 2KGA production with oxygen supply b Relative activities of resting cells during bioconversion

In the case of enhanced oxygen levels via continuous supply of oxygen (Fig. 6), cell content in the reaction mixture was also increased to 60 g
_{(wet wt)}
/L to make up for the loss of activities of cells. As expected, the conversion time
was considerably shortened. By 45 h, 480 g/L GA was completely exhausted, and the
2KGA titer reached about 453.3 g/L, generating a productivity of 10.07 g/L/h, which
is 135.8 % higher than that achieved using 30 g/L resting cells.

Fig. 6. Bioconversion of GA to 2KGA with oxygen supply and enhanced cell content (60 g
_{(wet wt)}
/L)

Bioconversion of glucose to 2KGA by G. oxydans_tufB_ga2dh

Overexpression of the ga2dh gene in G. oxydans could significantly improve the productivity and 2KGA production from GA, and enhance
the product formation rate from glucose. As shown in Fig. 7, the profiles for 2KGA production from glucose by G. oxydans_tufB_ga2dh and the control strain G. oxydans_pBBR1MCS5 were similar, but the glucose conversion rates and 2KGA formation rates
were evidently different. As shown in Fig. 7a, 200 g/L glucose was consumed rapidly by 60 g/L resting G. oxydans_tufB_ga2dh cells, which was accompanied by an increase in 2KGA and GA accumulation. After all
glucose was fully depleted at 12 h, GA accumulation reached a maximum of 102.42 g/L,
with a glucose conversion rate of 16.67 g/L/h. However, using the same amount of resting
cells of the control strain G. oxydans_pBBR1MCS5, the glucose conversion time was extended to 24 h, corresponding to a lower
glucose conversion rate of 7.96 g/L/h (Fig. 7b). During the second period of GA conversion to 2KGA, all GA produced was further
converted to 2KGA by G. oxydans_tufB_ga2dh within 12–21 h (i.e. elapsed 9 h) with a GA conversion rate of 11.38 g/L/h, which
was about fivefold higher than that obtained using G. oxydans_pBBR1MCS5. 150.0 g/L GA produced was gradually decreased by G. oxydans_pBBR1MCS5 cells and was fully consumed over a long period of time (24–102 h, elapsed
78 h), resulting in a low GA conversion rate of 1.92 g/L/h. It was also demonstrated
that overexpression of the ga2dh gene improves the conversion of GA to 2KGA. Unexpectedly, overexpression of the ga2dh gene significantly enhanced the conversion of glucose to GA. Overall, the final 2KGA
titer reached 234.6 g/L at 21 h during the batch bioconversion of glucose by the engineered
strain, corresponding to a productivity of 11.17 g/L/h, which amounted to a 407 %
increase compared with that obtained using the control strain G. oxydans_pBBR1MCS5.

Fig. 7. Comparison of 2KGA production from glucose. The biotransformations were carried out
in 7-L fermenter at 30 °C, pH 5.8, 600 rpm, aeration rate 8 L/min, initial Glu concentration
200 g/L and cell concentration 60 g
_{(wet wt)}
/L. aG. oxydans_tufB_ga2dhbG. oxydans_ pBBR1MCS5

When the glucose concentration was increased to 270 g/L, full conversion of glucose
was observed at 15 h by 60 g/L resting cells of G. oxydans_tufB_ga2dh, and the 2KGA titer reached a maximum of 318 g/L at 48 h, giving a productivity of
6.63 g/L/h (Fig. 8a). During the overall reaction process, we also found that DO levels in the fermenter
remained below 0 %. Because sufficient oxygen supply could enhance 2KGA productivity
during GA conversion, oxygen instead of air was supplied continuously to support the
oxidation of 270 g/L glucose at the same cell mass (Fig. 8b). As expected, the reaction time under oxygen supply was significantly decreased
compared with that when air was used as the electron acceptor. The glucose conversion
rate during the first period of GA formation and the GA conversion rate in the second
period of GA conversion to 2KGA were increased by 400 and 268.8 %, respectively. All
of the supplied glucose was converted to 321 g/L 2KGA over 18 h by the constructed
strain G. oxydans_tufB_ga2dh, corresponding to a productivity of 17.83 g/L/h. Both the glucose concentration during
batch biotransformation and the 2KGA productivity in this study were relatively high
compared with those achieved by Pseudomonas fluorescens5], 19], 20].

Fig. 8. Comparison of 2KGA production from glucose by G. oxydans_tufB_ga2dh. Initial Glu concentration 270 g/L; Cell concentration 60 g
_{(wet wt)}
/L. a Air b O
₂