Corn steep liquor is a superior nitrogen source compared to yeast extract in enhancing methanol assimilation
In addition to carbon source, nitrogen source is another key nutrient that affects cell growth and product formation. In native methylotrophic P. pastoris, nitrogen source was reported capable of regulating the gene expression in methanol utilization pathway [21]. In synthetic methylotrophic E. coli, the addition of yeast extract could also enhance methanol assimilation [23]. Thus, we hypothesized that the types of nitrogen sources contributed differently to the methanol assimilation of B. methylotrophicum. To test it, different kinds of nitrogen sources including peptone, corn steep liquor, beef extract, ammonium chloride and ammonium sulphate were supplemented in the modified DSM 135 medium with 200 mM methanol to replace the yeast extract with the equal nitrogen content.
As shown in Fig. 1, when the yeast extract was replaced by peptone, beef extract or inorganic nitrogen source of ammonium chloride and ammonium sulphate, the cell growth and methanol consumption were significantly decreased. With ammonium sulphate as an example, the final biomass and methanol consumption rate were decreased by 43% and 32%, respectively (Fig. 1). On the contrary, the use of corn steep liquor increases the methanol utilization rate of B. methylotrophicum by 35.9% (Fig. 1b). The final biomass yield in corn steep liquor medium was increased by 56.5% with a yield of 0.248 gDCW/gMeOH (Fig. 1a). The butyric acid formation in different nitrogen source was consistent with the methanol consumption. In corn steep liquor medium, 2.56 g/L butyric acid was produced with a yield of 56.0%, a 1.07-fold higher than that in yeast extract medium, while 2.16 g/L acetic acid was accumulated (Fig. 1c, d). The results indicated that corn steep liquor was a superior nitrogen source for B. methylotrophicum to grow in methanol.
Lysine was identified to be responsible for enhanced methylotrophy in B. methylotrophicum
Free amino acid is the main content of organic nitrogen source. To elucidate the potential molecular mechanism by which the corn steep liquor (CSL) was benefit for methanol assimilation of B. methylotrophicum compared to yeast extract (YE), the amino acid composition in CSL medium and YE medium was comparatively investigated during the fermentation process. The culture broth was sampled at 0 h, 36 h (middle exponential phase), 60 h (later exponential phase), and 84 h (stationary phase), respectively. Through the GC/MS analysis, 15 kinds of amino acids were detected including Gly, Ala, Lys, Asn, Thr, Glu, Asp, Val, Ile, Ser, Tyr, Pro, and Phe. The levels of the 15 amino acids were all gradually decreased with the fermentation both in the CSL medium and YE medium (Additional file 1: Fig. S1).
Among them, the content of 11 amino acids in CLS medium was initially higher than those in YE medium, which are Gly, Lys, Glu, Asn, Asp, Ile, Thr, and Met respectively (Fig. 2a). These results indicated the advantage of CLS in supplying free amino acids. During the whole fermentation process, the contents of Glu, Asn, Asp, Ile, Thr and Met in CLS medium were always higher than those in YE medium, while the contents of Phe, Pro, Ser and Leu in CLS medium were always lower than those in YE medium (Fig. 2a). For the amino acids with the higher level in CLS medium, Gly, Lys, and Asn represented the most significant difference. In addition, the difference of Gly, Lys, Ala and Val in CLS medium and YE medium changed most significantly during the whole fermentation process (Fig. 2a). We speculated that these significantly different amino acids with higher content in CSL medium might be related to enhanced methanol assimilation in B. methylotrophicum.
To further identify the certain amino acids important for enhanced methanol metabolism in B. methylotrophicum, Gly, Lys, Ala, Val, and Asn were added into the medium to evaluate their effects, while the other 9 amino acids were also comparatively analyzed. From the results shown in Fig. 2b, when methanol was used as a sole carbon source, the specific growth rate and the final biomass of B. methylotrophicum were improved by 12.1% and 29.2% with the addition of Lys, while the addition of Gly, Val and Ala showed a moderately increase in final biomass. Based on these results, lysine was determined as an important factor for regulating methanol assimilation of B. methylotrophicum.
As high lysine level was supposed to be beneficial for improving methanol assimilation of B. methylotrophicum, the effect of different concentrations of lysine was further evaluated. The results are shown in Fig. 3. The lysine concentration ranged from 5 to 30 mM. When a small amount of 5 mM lysine was supplemented, the cell growth and methanol consumption of B. methylotrophicum were slightly affected (Fig. 3a, b). With a higher lysine concentration from 10 to 30 mM, the cell growth, methanol consumption and butyric acid production of B. methylotrophicum were significantly improved (Fig. 3a–c), further confirming our conclusion that increasing the availability of lysine level could enhance methanol assimilation of B. methylotrophicum. At the condition of 20 mM lysine, the cell growth, methanol consumption and butyric acid production of B. methylotrophicum reached the maximum level, which were increased by 38.44%, 21.52%, and 53.07%, respectively (Fig. 3a–c). In addition, we found that the addition of lysine could also improve the butyric acid yield from methanol (Fig. 3d), while the production of acetic acid was decreased (Additional file 1: Fig. S2). In acetogenic bacteria, several amino acids have been reported to be oxidized and degraded to support cell growth. For example, alanine could serve as growth substrate for Sporomusa aerivorans (S. aerivorans) and A. woodii [24, 25], and E. limosum was able to use isoleucine and valine as growth substrate [26]. We also evaluated whether lysine could be used as a growth substrate for B. methylotrophicum and therefore stimulated growth on methanol. As the result shows in Fig. 3a, in the cultivation medium containing 3 g/L yeast extract, methanol was removed and 20 mM lysine was supplemented to determine the effect on cell growth. The results showed that the cells of B. methylotrophicum could not grow when methanol was removed in presence of lysine, and none of lysine consumption was observed (Fig. 3a and Additional file 1: Fig. S3). Under the condition of 100 mM methanol supplemented with different lysine concentration, lysine was also barely consumed with methanol consumption and cell growth (Additional file 1: Fig. S3). These results indicated that lysine was not served as a growth co-substrate to improve methanol assimilation in B. methylotrophicum.
Transcriptional analysis revealed that up-regulated expression of ABC transporters was triggered by lysine addition
To elucidate the potential mechanism of lysine in regulating methanol assimilation of B. methylotrophicum, the transcriptional response to lysine addition was determined. Through the RNA-seq experiments, differentially expressed gene (DEG) analysis identified 58 up-regulated and 920 down-regulated genes (Fig. 4a). The significantly changed genes are illustrated in Additional file 1: Tables S2 and S3. KEGG pathway analysis was subsequently conducted to identify the pathways for these DEGs. With the corrected p-value < 0.05, the up-regulated genes were significantly enriched in 8 pathways, while the down-regulated genes were enriched in 7 pathways (Fig. 4b and Additional file 1: Fig. S4). In the enriched pathways for up-regulated genes, ABC transporters represented the most significant one (Fig. 4a). In the enriched pathways for down-regulated genes, the pathways involved in ribosome, fatty acid synthesis, phosphotransferase system, fructose and mannose metabolism and lysine biosynthesis changed most significantly (Additional file 1: Fig. S4).
Here, the genes in the up-regulated pathways attracted our interested. Eight significant up-regulated ABC transporters including NikA, NikB, NikC, NikD, NikE, FhuB, FhuC, and FhuD was clustered into two classes. The first cluster including genes of NikA, NikB, NikC, NikD and NikE was defined as the nickel transport system [27], while another gene cluster of FhuB, FhuC, and FhuD was defined as an iron complex transport system involved in the uptake of siderophores, heme and cobalamin (vitamin B12) [28]. In B. methylotrophicum, methanol was assimilated to acetyl-CoA with CO2 as an electron acceptor through the methyltransferase system along with the carbonyl branch of the WLP pathway [16]. In the methyltransferase system, the corrinoid-dependent methyltransferase (MtaB) first transfers the methyl group of methanol to a corrinoid protein (MtaC), and then methyltetrahydrofolate-methyltransferase (MtaA) transfers the methyl group from methyl-MtaC to tetrahydrofolate (THF), where cobalamin is an important cofactor for the activity of MtaB and MtaC [29], up-regulation of the FhuBCD transporter may enable MtaB and MtaC catalytic activity, thereby affecting methanol assimilation in B. methylotrophicum. In the carbonyl branch of the WLP pathway, both carbon monoxide dehydrogenase (CODH) and acetyl-CoA synthase (ACS) are nickel-containing enzymes [30], up-regulation of the NikABCDE transporter may enhance the uptake of nickel and affect the catalytic activity of CODH and ACS. The gene expression of NikA, NikB, NikC, NikD, and NikE was significantly up-regulated by lysine addition with fold changes of 7.9, 9.9, 8.8, 3.6, 10.8, respectively (Fig. 4c), and the expression of FhuB, FhuC, and FhuD was increased by 3.3-fold, 3.0-fold, and 2.7-fold, respectively (Fig. 4d). To further demonstrate whether lysine stimulated the up-regulation of NikABCDE and FhuBCD transporters, we carried out RT-qPCR experiments. Following the results shown in Additional file 1: Fig. S5, the relative expression level of NikABCDE and FhuBCD transporters was significantly up-regulated by 2.19 and 1.98 times, respectively, when 20 mM lysine was added into the methanol medium. We thus speculated that the up-regulated cofactor uptake system might be involved in the improved methanol metabolism triggered by lysine in B. methylotrophicum.
Overexpression of NikABCDE or FhuBCD improved methanol assimilation of B. methylotrophicum
To further identify whether the up-regulation of these two-transport system was responsible for the improved methanol utilization of B. methylotrophicum, NikABCDE or FhuBCD was engineered to evaluate their effects on methanol metabolism of B. methylotrophicum. As shown in Fig. 5a, the overexpression of FhuBCD or NikABCDE resulted in a significant improvement in the methylotrophic phenotype over the empty plasmid control. With methanol as the sole carbon source, the specific growth rate of NikABCDE overexpressing strain reached 0.01393 h−1, 1.4-fold higher than the empty plasmid control, and the overexpression of FhuBCD increased the specific growth rate to 0.01641 h−1, 1.2-fold higher than the control (Fig. 5a). Higher final biomass titer was also achieved by the overexpression of FhuBCD or NikABCDE (Fig. 5a). At the meantime, the methanol consumption rate of the recombinant B. methylotrophicum/pXY1-FhuBCD and B. methylotrophicum/pXY1-NikABCDE increased by 24.5% and 34.7%, respectively (Fig. 5b). The titer and yield of butyric acid from the methanol were also significantly improved when these two systems were overexpressed (Fig. 5c). The butyric acid titer in the recombinant B. methylotrophicum/pXY1-FhuBCD and B. methylotrophicum/pXY1-NikABCDE was increased by 38.9% and 52.5%, respectively, and the butyric acid yield in the recombinant B. methylotrophicum/pXY1-FhuBCD and B. methylotrophicum/pXY1-NikABCDE was increased by 9.9% and 16.1%, respectively. Based on these results, we confirmed that the overexpression of NikABCDE or FhuBCD could improve methylotrophy of B. methylotrophicum (Fig. 5c). We thus proposed a possible regulatory mechanism that increasing lysine level triggered the expression of NikABCDE and FhuBCD transport system responsible for improved methanol utilization in B. methylotrophicum (Fig. 5d).
The engineering of lysine synthetic pathway enhanced methanol utilization and butyric acid production in B. methylotrophicum
As exogenous lysine addition could improve methanol utilization of B. methylotrophicum, we hypothesized that increasing lysine synthesis through the genetic modification of the de novo lysine synthetic pathway was able to improve methanol assimilation. The lysine biosynthetic pathway in B. methylotrophicum was identified to be similar with that in E. coli, where L-aspartate was converted to generate L-lysine by the catalysis of lysC, asd, dapA, dapB, dapC, dapD, dapE, dapF, and lysA (Fig. 6a). The overexpression of lysA, dapA, or dapB has been confirmed to be an efficient approach for enhancing flux through the lysine synthetic pathway [31,32,33,34]. Here, two plasmids of pXY1-Pthl-dapA-dapB for the overexpression of dapA and dapB, and pXY1-Pthl-lysA for the overexpression of lysA were constructed and electro-transformed into B. methylotrophicum. As the results illustrated in Fig. 6b, both the overexpression of lysA or co-overexpression of dapA and dapB could significantly enhance the methylotrophic phenotype of B. methylotrophicum. The biomass concentration and methanol consumption of B. methylotrophicum/pXY1-Pthl-lysA was increased by 38.8%, and 23%, respectively (Fig. 6b, c). The recombinant strain B. methylotrophicum/pXY1-Pthl-dapA-dapB exhibited the best growth advantage in methanol medium, in which the final biomass concentration was increased by 51%, and the specific growth rate reached 0.0259 h−1 (Fig. 6b). The methanol consumption of B. methylotrophicum/pXY1-Pthl-dapA-dapB was increased by 63.2% (Fig. 6c). In addition, the enhancement of lysine could also improve butyric acid production titer and yield from methanol (Fig. 6d, e). The overexpression of lysA increased the butyric acid production by 33.8% with a final titer of 0.99 g/L, while the accumulation of acetic acid was significantly decreased (Additional file 1: Fig. S6). For the engineered B. methylotrophicum/pXY1-Pthl-dapA-dapB, a titer of 1.33 g/L butyric acid was achieved with a yield of 83.1%, which was increased by 79.7% and 10.4% compared to that in the control strain, respectively (Fig. 6c, d). These results confirmed that increasing the flux through the lysine biosynthetic pathway could efficiently improve methanol utilization and butyric acid production of B. methylotrophicum, further identifying that lysine is an important target for the regulation of methanol utilization in B. methylotrophicum.
The presence of CO2 as an electron acceptor could affect methanol consumption and product distribution in B. methylotrophicum. Different concentrations of bicarbonate were therefore supplemented and the cell growth on methanol was largely improved (Fig. 7a). The percentage of methanol consumption was increased to 97% in conditions of 20 mM or 40 mM bicarbonate from 55% in conditions of 0 mM bicarbonate (Fig. 7b). The titer of butyric acid reached 1.45 g/L when the methanol to bicarbonate ratio was 100 mM:20 mM (Fig. 7c). With the further increase of bicarbonate concentration to 40 mM, butyric acid titer was just increased to 1.6 g/L, and meanwhile large amount of acetic acid (2.0 g/L) was produced (Fig. 7d). To further improve the butyric acid production of the engineered B. methylotrophicum/pXY1-Pthl-dapA-dapB from methanol, the fermentation was performed in CSL medium. As shown in Fig. 8, under the condition of 200 mM methanol and 40 mM bicarbonate, the cell grew into the stationary phase after the fermentation 96 h. With the prolonged fermentation time to 158 h, the production of butyric acid reached 3.69 g/L, while 1.92 g/L acetic acid was accumulated. Finally, 5.7 g/L methanol and 15.97 mM bicarbonate were totally consumed, and the yield of butyric acid from methanol could reach 76.3%.