Effect of initial 1,3-PDO concentration in batch cultures
First, the ability of Acetobacter sp. CIP 58.66 to perform growth-coupled 1,3-PDO bioconversion was investigated. Growth was tested in shake flasks on a complex medium containing either 0, 5, 10 or 20 g L-1 of 1,3-PDO. To some extent, the basal medium without 1,3-PDO was sufficient to support the growth of the strain. Indeed, in absence of 1,3-PDO, growth was observed, with a final cell density of 0.18 gDW L-1 (Table 1). When 1,3-PDO was added to the basal medium, changes in growth patterns were observed. Maximal growth rate µmax was found significantly higher with 1,3-PDO than without (p-value = 0.01), but it was independent from the initial concentration. With 5 and 10 g L-1 initial 1,3-PDO, final cell densities were also higher compared to the control, whereas they were lower for 20 g L-1 initial 1,3-PDO (Table 1). These results show that Acetobacter sp. CIP 58.66 is able to use 1,3-PDO during its growth, either as carbon or energy source. Yet, the highest cell density was achieved with the lowest initial 1,3-PDO concentration, meaning that higher substrate levels cause inhibition either directly (substrate inhibition) or indirectly (product inhibition). Previous studies on 1,3-PDO bioconversion into 3-HP using AAB focused on resting cells only, and biomass evolution was not measured over time (Dishisha et al., 2015; Zhao et al., 2015). Therefore, this is the first report of 1,3-PDO affecting the growth of an AAB.
Table 1
Comparison of growth and bioconversion for different initial 1,3-PDO concentrations in shake flask cultures. Finals values are calculated at 30 h of culture.
|
Initial 1,3-PDO concentration (g L-1)
|
|
0
|
5
|
10
|
20
|
Final DW (g L-1)
|
0.18 ± 0.01
|
0.33 ± 0.02
|
0.22 ± 0.03
|
0.12 ± 0.02
|
Number of generations (Ng)
|
2.21 ± 0.06
|
3.17 ± 0.03
|
2.58 ± 0.09
|
1.72 ± 0.02
|
Final 1,3-PDO consumption (%)a
|
ND b
|
99.4 ± 0.7
|
53.7 ± 1.3
|
14.4 ± 1.4
|
Final 3-HP titre (g L-1)
|
ND b
|
6.02 ± 0.06
|
3.77 ± 0.61
|
2.14 ± 0.06
|
Final 3-HP yield (mol mol-1)
|
ND b
|
1.07 ± 0.03
|
0.59 ± 0.10
|
0.59 ± 0.01
|
Final 3-HPA titre (g L-1)
|
ND b
|
0.16 ± 0.00
|
1.06 ± 0.06
|
0.55 ± 0.03
|
Final 3-HPA yield (mol mol-1)
|
ND b
|
0.04 ± 0.00
|
0.21 ± 0.01
|
0.19 ± 0.01
|
Final pH
|
6.7 ± 0.08
|
4.29 ± 0.05
|
4.44 ± 0.13
|
5.11 ± 0.17
|
µmax (h-1)
|
0.21 ± 0.01
|
0.28 ± 0.004
|
0.29 ± 0.01
|
0.27 ± 0.01
|
Normalised RMSE
|
2%
|
1%
|
2%
|
3%
|
a Calculated as the percentage of 1,3-PDO initially supplied
b Was not calculated because no 1,3-PDO was supplied to the medium
|
The conditions with the best growth performances (5 g L-1 initial 1,3-PDO) was also the one displaying the best bioconversion performances: only in that condition was 1,3-PDO fully depleted. For 10 and 20 g L-1 initial concentrations, consumption dropped to 53.7 % and 14.4 % respectively. This trend is consistent with the observations of Dishisha et al. (2015) on another AAB: in their study, resting cells of Gluconobacter oxydans consumed 95.3 %, 86.1 % and 31.3 % of 5, 10 and 20 g L-1 of initial 1,3-PDO. Furthermore, full substrate depletion was also associated with the highest 3-HP yield and titre, as well as with the lowest level of 3-HPA production (Table 1). Contrarily, for 10 and 20 g L-1 initial 1,3-PDO, final 3-HP yields were lower: this was attributed to higher HPA accumulation.
These results show that AAB growth and 3-HP production is best achieved with a 5 g L-1 initial 1,3-PDO concentration. In this case, 1,3-PDO was quasi-quantitatively converted to 3-HP; substrate and 3-HPA inhibition were thus prevented. The almost quantitative production of 3-HP may suggest that 1,3-PDO is not used as carbon source for growth, but rather as energy source. In that case, the strain may have used another carbon source from the complex medium.
Impact of initial pH on the growth on glycerol
Previous results showed that despite 1,3-PDO having a positive impact on the growth of Acetobacter sp. CIP 58.66, it could not ensure high biomass density. Growth was thus investigated on a separate substrate, namely glycerol. It was already used as growth substrate in other studies on 3-HP production with AAB (Dishisha et al., 2015; Li et al., 2016; Pyo et al., 2012; Zhao et al., 2015), but only few details were generally given on this step. In the present paper, four different initial pH were tested in shake flask, batch cultures of Acetobacter sp. CIP 58.66 on 10 g L-1 of glycerol (Table 2).
Table 2
Growth comparison of Acetobacter sp. CIP 58.66 with glycerol as substrate, at different initial pH.
|
Shake flask cultures a
|
Biocatalyst production b
|
Initial pH
|
4.0
|
4.5
|
5.0
|
6.5
|
5.0
|
DW (gDW L-1)
|
0.14 ± 0.01
|
0.33 ± 0.04
|
1.14 ± 0.04
|
0.43 ± 0.02
|
0.88 ± 0.05
|
Glycerol consumption (%)c
|
1.0 ± 0.8
|
4.6 ± 1.9
|
16.9 ± 0.7
|
5.8 ± 0.6
|
18.6 ± 1.7
|
Biomass yield ( )
|
ND d
|
0.83 ± 0.28
|
0.69 ± 0.05
|
0.76 ± 0.06
|
0.40 ± 0.05
|
Final pH
|
4.10 ± 0.02
|
4.58 ± 0.08
|
5.38 ± 0.03
|
6.55 ± 0.03
|
6.94 ± 0.13
|
µmax,1 (h-1)
|
0.10 ± 0.01
|
0.15 ± 0.004
|
0.18 ± 0.004
|
0.11 ± 0.004
|
0.24 ± 0.02
|
R²
|
0.94
|
0.98
|
0.99
|
0.98
|
0.88
|
µmax,2 (h-1)
|
0.03 ± 0.002
|
0.06 ± 0.002
|
0.09 ± 0.002
|
0.07 ± 0.001
|
0.18 ± 0.01
|
R²
|
0.92
|
0.97
|
0.99
|
0.99
|
0.96
|
a Comparison at 30.5 h of culture
b Values at 31 h of culture in bioreactor
c Calculated as the percentage of glycerol initially supplied
d Could not be calculated because no significant glycerol consumption occurred
|
When the initial pH was 4.0, glycerol remained unconsumed and biomass density remained low (0.14 gDW L-1). This condition was thus considered too stressful and could not ensure high biocatalyst concentrations. In other conditions (i.e. initial pH of 4.5, 5.0 and 6.5), glycerol consumption occurred and higher biomass densities were reached (Table 2). The highest concentration (1.14 gDW L-1) was obtained with an initial pH of 5.0. Even though glycerol consumption remained low at the time of comparison (30.5 h), full glycerol depletion was observed for all conditions after 115 h. In all instances, two distinct exponential growth phases were identified: a slope break appeared between 4 and 8 hours of culture. The highest growth rates were observed with an initial pH of 5.0, which was also the only condition reaching late exponential phase by 30.5 h of culture. Since it led to the best growth performance (highest cell density, growth rates and biomass yield), pH 5.0 was the selected condition for the production of biocatalyst in a bioreactor, during the sequential process (i.e. biomass production directly followed by bioconversion).
Biocatalyst production as part of a sequential strategy
First, we tested whether Acetobacter sp. CIP 58.66 could grow on 10 g L-1 glycerol in a bioreactor controlled at pH 5.0, while simultaneously converting 1,3-PDO into 3-HP. In this case, both growth on glycerol and 1,3-PDO conversion remained very limited (data not shown). Therefore, a sequential strategy was designed: biomass was first produced on glycerol in batch mode; once the late exponential phase was reached, 1,3-PDO was supplied in fed-batch mode. This section presents the first step (i.e. biocatalyst production) that was common to all sequential experiments.
Biomass was produced on the basal medium containing 10 g L-1 glycerol in a 3.6 L bioreactor. Initially, pH was set to 5.0 and it was then left uncontrolled. Results are shown in Table 2 and Figure 3. Except for one of the six replicates, bacterial growth started without any latency. In all instances, late exponential phase was reached after 31 h of culture, with a 0.88 g L-1 DW concentration. Similarly to previous shake flask experiments, two distinct exponential growth phases could be detected. Furthermore, by 31 h of culture, glycerol consumption remained low (18.6 % of initial glycerol). Since no buffer solution was added to the medium, pH rose up to 6.94. This rise in pH may be the cause of the slowing down of growth before full substrate depletion.
Growth on glycerol showed a satisfactory reproducibility: the coefficient of variation for final DW was 6 %. It was thus considered that these six growth replicates constitute a consistent basis to compare the subsequent bioconversions at different pH.
pH-based fed-batch bioconversion
Once produced as described in section 3.2.2, biomass of Acetobacter sp. CIP 58.66 was used as biocatalyst for 1,3-PDO selective oxidation into 3-HP in presence of the residual glycerol. After growth on glycerol, pH was adjusted to the desired value with H2SO4 and the fed-batch bioconversion was triggered by adding 5 mL of 1,3-PDO as starter, and by plugging the feeding solution to the bioreactor. The latter solution was a 1:1 molar mix of ammonium hydroxide and 1,3-PDO in order to ensure both pH control and substrate feeding. In the first instance, bioconversion was tested at pH 5.0. A secondary exponential growth phase was observed, without any latency (Fig. 3). DW concentration further increased to 2.08 ± 0.04 g L-1 at a maximal growth rate of 0.16 ± 0.06 h-1. No significant difference in growth rates was evidenced when comparing with previous results for growth on glycerol alone or on 1,3-PDO alone. Glycerol was further consumed during 1,3-PDO oxidation (Fig. 3): its consumption increased from 18.6 % at the beginning of bioconversion to 42.1 % at 25 h of bioconversion. So it remains unclear whether this secondary growth can be attributed to the uptake of glycerol or 1,3-PDO, or both of them. Consumption of 1,3-PDO began as soon as it was added to the medium, without any latency (Fig. 4.A). At 25 h of bioconversion at pH 5.0, a total of 51.45 ± 6.63 g of 1,3-PDO was consumed and converted quantitatively into 3-HP (Table 3; Fig. 4.B). Within the first three hours, a transient accumulation of 3-HPA occurred, that peaked at 0.10 ± 0.01 g L-1. In one of the replicates, 3-HPA was later accumulated again, reaching a final concentration of 0.78 g L-1 (Fig. 4.C).
Table 3
1,3-PDO bioconversion characteristics of Acetobacter sp. CIP 58.66, after its growth on glycerol.
|
pH of bioconversion
|
|
5.0
|
4.5
|
4.0
|
Final DW (g L-1)
|
2.08 ± 0.04
|
1.58 ± 0.06
|
1.51 ± 0.04
|
Final Ng a
|
1.21 ± 0.14
|
0.70 ± 0.03
|
0.54 ± 0.09
|
Final glycerol consumption (%)b
|
42.1 ± 0.6
|
36.0 ± 2.1
|
32.3 ± 1.8
|
Overall 1,3-PDO consumption (g)
|
51.45 ± 6.63
|
25.53 ± 1.07
|
7.15 ± 0.14
|
Final 3-HP titre (g L-1)
|
69.76 ± 6.00
|
36.48 ± 0.80
|
10.89 ± 0.65
|
Final 3-HP yield (mol mol-1)
|
1.02 ± 0.04
|
1.06 ± 0.05
|
1.11 ± 0.02
|
µmax (h-1)
|
0.16 ± 0.06
|
0.13 ± 0.02
|
0.10 ± 0.03
|
Normalised RMSE for ln(DWt/DW0)
|
11%
|
8%
|
13%
|
r3-HP,max (g3-HP L-1 h-1)
|
3.9
|
2.7
|
2.4
|
q3‑HP,max ( )
|
2.0
|
2.0
|
2.3
|
Normalised RMSE for DW production
|
9%
|
6%
|
17%
|
Normalised RMSE for 3-HP production
|
5%
|
2%
|
3%
|
a Number of generations of the secondary growth phase
b Calculated as the percentage of glycerol initially supplied
|
The pH was successfully controlled at 5.0 during these bioconversions. However, 1,3-PDO was not kept at 5 g L-1 as intended: a slight decrease happened and the final concentration was 2.96 ± 0.79 g L-1 (Fig. 4.C). In addition to 3-HPA accumulation in one replicate, two likely limiting factors for cells were identified: (i) high 3-HP concentrations, and (ii) oxygen limitation. Indeed, from 4.4 h of bioconversion, a drop in pO2 below the set point was observed, despite maximal stirring and air flow rates. In one instance, pO2 even dropped to 4% for 4 h before increasing again to the set point. Overall, the bioconversion process presented here achieved promising results, reaching the highest 3-HP titre with AAB. Also it is the first time growing AAB are used for 3-HP production instead of resting AAB. By doing so, the process could be performed with lower cell densities (from 0.88 to 2.08 gDW L-1) as compared with the previous literature, where cell densities vary from 3.25 gDW L-1 (Dishisha et al., 2015) to 10 gDW L-1 (Zhao et al., 2015).
Given the promising results obtained at pH 5.0, the process was further tested for lower pH values during the bioconversion stage, in order to try to take advantage of the great resistance of AAB to acidic conditions. The selected pH values were 4.5 and 4.0: by doing so, the bioconversion performance can be compared for a pH above the pKa of 3-HP (pH 5.0), equal to it (pH 4.5), or lower to it (pH 4.0). Again, a secondary growth phase happened (Fig. 3). Yet, the lower the pH was, the lower were the growth rates, the final DW concentration, and the final glycerol consumption (Table 3). This is consistent with the results of shake flask cultures on glycerol, during which growth was more limited with initial pH values of 4.0 and 4.5 compared to 5.0. Concomitantly to this secondary growth phase, 1,3-PDO was consumed in all pH conditions and was quasi-quantitatively converted to 3-HP (Table 3). Similarly to bioconversions at pH 5.0, a slight 3-HPA accumulation occurred and peaked around 2 h of bioconversion, at sub-inhibitory levels, but no later 3-HPA accumulation was observed (Fig. 4). Overall, bioconversion performances were lower with lower pH: less 1,3-PDO was consumed and therefore less 3-HP was produced.