Undetoxified corn stover hydrolysate fermentation by B. coagulans
B. coagulans as a potential industrial lactic acid producing strain has been received widespread attention. It could effectively convert various monosaccharides (pentose and hexose) derived from lignocellulose into high optical purity L-lactic acid through homolactic fermentation under thermophilic conditions [22]. To exploit the effective utilization of lignocellulosic feedstocks by B. coagulans, the undetoxified dilute acid pretreatment corn stover hydrolysate consisting of glucose (22.2 g/L), xylose (122 g/L), arabinose (~ 16 g/L), five carbohydrate degradation products, ten phenolic compounds and other unidentified inhibitors (Table 1) was used in this work. The hydrolysate was diluted to different concentrations (20%, 30% and 40% (v/v)) and then directly employed for lactic acid fermentation by B. coagulans NL01.
Table 1
Composition of sugars, weak acids, furans, and phenolic compounds present in concentrated undetoxified corn stover dilute-acid pretreatment hydrolysate.
| products | concentration |
Sugars (g/L) | Glucose | 22.36 ± 0.24 |
| Xylose | 121 ± 1.08 |
| Arabinose | 15.20 ± 0.28 |
Weak acids (g/L) | Formic acid | 1.35 ± 0.24 |
| Acetic acid | 14.50 ± 0.25 |
| Levulinic acid | 2.96 ± 0.05 |
Furans (g/L) | Furfural | 0.4 ± 16 |
| HMF | 0.99 ± 0.04 |
Phenolic compounds (mg/L) | 3’4-Dihydroxybenzoic acid | 34.25 ± 6.02 |
| 4-hydroxybenzoic acid | 41.34 ± 0.99 |
| Vanillic acid | 66.35 ± 5.33 |
| Syringic acid | 24.36 ± 1.60 |
| 4-hydroxybenzaldehyde | 59.35 ± 1.01 |
| Vanillin | 86.65 ± 2.03 |
| P-coumaric acid | 133.90 ± 2.64 |
| Syringaldehyde | 55.29 ± 2.21 |
| Ferulic acid | 214.20 ± 0.99 |
| Cinnamic acid | 0 |
| Total phenolics | 5.33 |
The growth and lactic acid fermentation capacity of B. coagulans in fermentation medium containing different concentrations of pretreatment hydrolysate were evaluated. As illustrated in Fig. 2a and 2b, the growth and xylose utilization of B. coagulans decreased as the concentration of the hydrolysate was elevated. At a concentration of 20% (v/v) hydrolysate with 31.93 g/L monomeric sugars, the bacteria showed obvious growth after an initial 12 h lag phase. Which can completely uptake the primary monosaccharides composed of xylose, glucose and arabinose within 72 h, and achieve the maximum OD600 6.6. Whereas, the growth and lactic acid fermentation of the B. coagulans were completely ceased at a higher hydrolysate concentration (30% (v/v)). Even after prolonged fermentation time to 96 h, the growth of B. coagulans was still seriously hindered. The same results were observed with 40% (v/v) hydrolysate.
These results demonstrated that B. coagulans had a certain tolerance to hydrolysate, but the tolerance ability of strain was not high. It was merely suited for hydrolysate with low inhibitors and low corresponding sugar concentration. It is well known that substrate titer is one of vital factors impacting the final product concentration [23]. However, the concentration of inhibitors in the hydrolysate would enhance with the increase of the concentration of sugar. Furthermore, synergistic effects of between inhibitors under high contents render of lignocellulosic hydrolysates more toxic. The poor fermentation performance indicated that the tolerance of B. coagulans to higher concentration of hydrolysate was very inferior. These features limit the potential of using it as a platform strain to handle the carbohydrate products of lignocellulosic waste [24]. To promote the effective conversion of lignocellulosic sugars in hydrolysate, the introduction of an outstanding detoxified strain capable of transforming and degrading inhibitors while preserving the sugars is highly desirable for removing the inhibitory and toxic compounds to improve the fermentation of B. coagulans.
Metabolic engineering of P. putida to block sugar metabolism and growth characterization of knockout strains
P. putida KT2440 has emerged as a promising candidate for a future lignocellulose-based biotechnology process due to its outstanding toxicity tolerance to numerous by-products of biomass hydrolysis [25–27]. This was confirmed here by incubating of wild-type P. putida in LB medium supplemented with major inhibitors found in lignocellulosic hydrolysate (Fig. S1). Besides, it is a robust microorganism capable of natively catabolizing glucose other than toxins. Unfortunately, the excessive utilization of sugar during the detoxification will adversely affect subsequent fermentation and reduce process efficiency. With the intent of developing a prominent detoxified strain capable of removing inhibitors in hydrolysate and retaining monomeric sugars, we engineered P. putida to delete its sugar metabolism. There are three glucose catabolism pathways in P. putida KT2440 (Fig. 3). The first route is that glucose is transported to cytoplasm from periplasm by means of an ATP dependent ABC transporter (PP_1015-PP_1018), encoded by the gtsABCD operon and is then phosphorylated by glucokinase to glucose-6-phosphate and subsequently converted by glucose-6-phosphate dehydrogenase to 6-phosphogluconate. The second is a periplasmic oxidation pathway mediated by PQQ-dependent glucose dehydrogenase gcd (PP_1444). Glucose is oxidized to gluconate by gcd in the periplasm, which is then transported to the cytoplasm and subsequently phosphorylated, later reduced to 6-phosphogluconate. Besides, previous studies revealed that xylose is also oxidized to xylonate by gcd [28]. The third route is the 2-ketogluconate loop, which involves the oxidation of gluconate to 2-ketogluconate. The resulting 2-ketogluconate is imported into the cytoplasm via the outer membrane and subsequently phosphorylated, in which the oxidation pathway and the direct phosphorylation route converge at the node of 6-phosphogluconate and that function simultaneously [28, 29]. Accordingly, the gcd and gtsABCD were selectively removed.
The growth and the consumption of fermentable sugars of P. putida and its knockout strains were characterized. As a reference, the cultivation of P. putida KT2440 in M9 minimal medium using glucose or xylose as a carbon source was performed. As shown in Fig. 4a, about 5 g/L of glucose can be used up by P. putida KT2440 in 6 h, and reach a remarkable maximum cell OD600 of 4.5. However, when the carbon source of P. putida KT2440 was switched to xylose, although the xylose concentration was significantly reduced and dropped to 0 at 30 h, the growth of the strain was not detected (Fig. 4c). The consumed xylose was oxidized to xylonate instead of being used as a carbon source for strain growth. The results obtained are consistent with previous study [30]. The growth performance of the P. putida pK18MS-Δgcd was evaluated under the same carbon source conditions. When incubating with glucose, P. putida KT2440 pK18MS-Δgcd showed a significantly slower growth than wild type and the consumption rate of glucose was remarkably decreased, which was not completely consumed until 36 h (Fig. 4b). The ability of the P. putida KT2440 pK18MS-Δgcd to utilize xylose was also tested (Fig. 4d). No growth of strain and reduced sugar concentration can be examined during the entire cultivation. The deletion of gcd could greatly retard the glucose metabolism and completely hinder the consumption of xylose. These results suggested that gcd is of paramount importance for glucose and xylose uptake in P. putida.
The successful knockout strain P. putida KT2440 pK18MS-Δgcd-ΔgtsABCD was verified using glucose or xylose as a substrate. As predicted, the strain did not illustrate any detectable growth and consumption of sugar during 4 days of the cultivation on glucose or xylose (Fig. 4e & Fig. 4f). Removal of gcd and gtsABCD from P. putida KT2440’s genome rendered it unable to utilize glucose and xylose. Biological detoxification of lignocellulosic hydrolysate has been proved to be a prospective approach for inhibitors removal and many microorganisms that either degrade or convert inhibitors have been identified. However, most of detoxified microorganisms consume main fermentable sugars, which causing a substantial loss of the sugar content in the lignocellulose hydrolysate. Efficient biological detoxification required microorganism that metabolized all inhibitors while preserving the sugar fraction intactly [31].
Inhibitors tolerance analysis of engineered P. putida KT2440 pK18MS-Δgcd-ΔgtsABCD
The ability of engineered P. putida KT2440 pK18MS-Δgcd-ΔgtsABCD to metabolize and convert common inhibitors encountered in lignocellulose hydrolysates was investigated. Firstly, the effect of acetate, which was formed as a major unwanted byproduct from the breakdown of hemicellulose, was tested by culturing cells in M9 minimal medium comprising different titers of sodium acetate ranging from 0–10 g/L. As presented in Fig. 5a, the knockout strain could still grow well with different concentrations of sodium acetate as a carbon source. The supplementation of high concentrations sodium acetate (5 g/L and 10 g/L) only slightly delayed the growth of cells in the early stage, but did not affect their complete catabolism by the strain. Besides, the maximum achieved biomass of the strain was positively correlated with the concentration of the initial substrate provided. With 1 g/L sodium acetate, the maximum OD600 0.77 was detected. Using 5 g/L sodium acetate, higher growth rate was measured. The cells reached a maximum OD600 of 3.1 at the highest applied 10 g/L of sodium acetate. During cultivation, the consumption of sodium acetate at different concentrations was also indicated in Fig. 5b. Up to 10 g/L sodium acetate could be totally depleted by engineered P. putida in 48 h. Although acetate itself is not a fierce inhibitor, its elimination could lessen the whole toxicity of hydrolysate [20]. Similarly, the utilization of levulinic acid by the knockout strain was assessed. The knockout strain did not exhibit lag phase at the tested concentrations (Fig. 5c). The strain could grow rapidly with levulinic acid. A maximum OD600 of 2.5 could be achieved starting at 5 g/L levulinic acid. Correspondingly, the consumption of levulinic acid can also be seen from Fig. 5d. Levulinic acid was rapidly consumed in a short period. Even though previous studies have reported that weak acids can inhibit cell growth and metabolism due to uncoupling and intracellular anion accumulation, we found that engineered P. putida grew well in the presence of carbohydrate derived acids [32]. FAL and HMF that commonly generated from high-temperature processing of lignocellulose are considered major inhibitors in microbial conversion processes [33]. The conversion performance of the engineered P. putida KT2440 pK18MS-Δgcd-ΔgtsABCD towards furan aldehydes was examined. As shown in Fig. 5e and 5f, the engineered strain could convert the representative concentrations of HMF and FAL into the corresponding carboxylic acids, 5‑hydroxymethyl‑2‑furancarboxylic acid (HMFCA) and furoic acid (FA), which are less toxic. In addition, a different conversion profile by engineered P. putida was noticed. Compared with FAL, the conversion rate of HMF were faster within 12 h (0.847 mM/h for HMF and 0.557 mM/h for FAL). The complete conversion of HMF was observed at 12 h of cultivation, whereas full conversion of FAL was observed at 24 h.
These results demonstrated that the genetically modified P. putida remained robust to the elimination and transformation of toxic inhibitors. It has extraordinary potential for the removal of inhibitors in the biomass pretreatment hydrolysate.
Production of lactic acid from undetoxified corn stover hydrolysate by using synthetic consortia of engineered P. putida and B. coagulans
Based on the above results, a co-cultivation reaction was performed for L-lactic acid production from undetoxified 30% (v/v) real hydrolysate that has the same composition as described above. During co-culture process, to gain detailed insights into the behavior of engineered P. putida KT2440 pK18MS-Δgcd-ΔgtsABCD on the complex mixtures of the hydrolysate, the detoxification reaction of the hydrolysate was set up using 1 g/L P. putida KT2440 pK18MS-Δgcd-ΔgtsABCD and no detoxified strain was supplemented as control. The concentration changes of the substances (fermentable sugars, organic acids, furans and phenol compounds) before and after the reaction were detected. After 24 h reaction, the concentration of major sugars such as glucose, xylose and arabinose in lignocellulosic hydrolysate remained nearly constant under both conditions (Table S1). However, there are considerable difference in the degradation and conversion of inhibitors. In the presence of detoxified strain, 4.3 g/L of acetic acid and 0.9 g/L of levulinic acid could be completely metabolized in 12 h. Besides, the removal rate of formic acid and HMF reached 73% and 100%, respectively (Fig. 6a). Furthermore, the concentration changes of total phenol and several typical monophenol inhibitors were also quantified. As shown in Table S2 and Fig. 6c, we observed that the concentration of total phenol and monophenol inhibitors have enormous degrees of reduction. The detoxification for 24 h by P. putida KT2440 pK18MS-Δgcd-ΔgtsABCD, 25% of total phenolic derivatives was found to be removed. This result is comparable to the electrochemical detoxification reported in the literature [23]. Additionally, most of the representative monophenol compounds especially ferulic acid, p-coumaric acid, vanillin, 4-hydroxybenzaldehyde and syringaldehyde could be rapidly degraded to very low concentration. A control experiment confirmed that the concentration of the organic acids, furan aldehydes (Fig. 6b), ten representative phenolic compounds (Fig. 6d) as well as total phenolic derivatives (Table S2) did not vary significantly in the absence of detoxified strain pK18MS-Δgcd-ΔgtsABCD under the same investigated conditions. These results indicated that P. putida KT2440 pK18MS-Δgcd-ΔgtsABCD is a superior biological detoxification strain, which can effectively eliminate the toxic compounds in depolymerized lignocellulose streams while retaining the hexose and pentose sugars in the hydrolysate for subsequent lactic acid fermentation.
The feasibility of detoxified hydrolysate for lactic acid fermentation by B. coagulans was investigated. As shown in Fig. 7, the hydrolysate treated by P. putida KT2440 pK18MS-Δgcd-ΔgtsABCD containing glucose (6.1 g/L), xylose (about 35 g/L) as well as arabinose (4.5 g/L) could be effectively consumed by fermentation of strain B. coagulans within 96 h, resulting in a lactic acid production of 35.8 g/L and a corresponding yield of 90%. However, the hydrolysate of the identical concentration that has not been detoxified by P. putida KT2440 pK18MS-Δgcd-ΔgtsABCD has neither the consumption of sugar nor the production of lactic acid was observed (Fig. 2). As anticipated, the degradation and transformation of diverse inhibitors by P. putida in hydrolysate greatly decrease the inhibitors concentration and contribute to the broth less toxic for fermentation. The results suggested that the detoxification of P. putida was extremely beneficial for improving the utilization of sugar and lactic acid fermentation by B. coagulans in the hydrolysate with high content of complex inhibitors. Additionally, which also suggested that the co-culture system was more advantageous than mono-culture fermentation for the production of biochemicals from highly toxic lignocellulosic hydrolysate.
Comparison of lactic acid production from lignocellulosic biomass with other studies
To date, extensive investigations have been conducted to realize the production of lactic acid from lignocellulosic biomass (Table 2). The production of lactic acid from the pretreated wheat straw hydrolysate by an adapted B. coagulans was reported. 35.5 g of lactic acid was obtained from 80 g wheat straw with a yield of 70.9% [34]. Besides, to reduce processing costs as well as simplification of the process, one genetically engineered Bacillus subtilis capable of both degrading the cellulose and producing lactic acid was developed. Nevertheless, the recombinant cellulolytic B. subtilis suffered from low titers of lactic acid [35]. Low efficiency was main barriers for lactic acid production from lignocellulosic biomass. Currently, most of studies on lactic acid production from lignocellulosic biomass commonly use the simultaneous saccharification and fermentation (SSF) method. Cui et al. reported a mixed cultures of Lactobacillus rhamnosus and Lactobacillus brevis for bioconversion corn stover into lactic acid through SSF process [36]. Although relatively high productivity of lactic acid was obtained, which require the addition of high-cost cellulase. Recently, consolidated bioprocesses (CBP) targeting lactic acid production was exploited by using a synthetic microbial consortia of Trichoderma reesei and Lactobacillus pentosus. Ultimately, 34.7 g/L and 19.8 g/L of lactic acid can be obtained from 5% (w/w) microcrystalline cellulose and pretreated beech wood, respectively [37]. Despite intensive research efforts spanning years, as an immature but promising technology, there are limit reports on the conversion of lignocellulose by microbial co-cultivation especially in lignocellulosic hydrolysates made up of numerous inhibitors and high concentration. In this work, we developed synthetic microbial consortia of P. putida and B. coagulans could achieve effective production of lactic acid from lignocellulosic hydrolysate with high inhibitors contents. The titer and yield of lactic acid attained by the coculture system were comparable to those previously reported. Although the results obtained by this approach are encouraging, the process regulation and optimization of microbial cocultivation are still essential in the future work, which probably promote the concentration and productivity of lactic acid to satisfy the requirement of practical lignocellulose biorefining.
Table 2
Comparison of lactic acid production from lignocellulosic biomass with other studies.
Strain | Strategy | Substrate | CLA, max (g L− 1) | Yield (%)/(g/g) | Productivity (g L− 1 h− 1) | References |
B. coagulans CC17A | adaptive evolution, SSF | wheat straw | 35.5 | 70.9 | n.a. | 34 |
B. subtilis | over-expression of endoglucanase and knocked out alpha-acetolactate synthase | cellulose | 4.1 | 63 | 0.03 | 35 |
L. rhamnosus/L. brevis | co-culture, SSF | Corn stover | 20.95 | 0.70 | 0.58 | 36 |
T. reesei/L. pentosus | CBP | microcrystalline cellulose | 34.7 | 62.4 | 0.16 | 37 |
T. reesei/L. pentosus | CBP | Beech wood | 19.8 | 85.2 | 0.10 | 37 |
P. putida/B. coagulans | Sequential co-culture | corn stover hydrolysate | 35.8 | 0.90 | 0.37 | this study |