- Open Access
Protein-based biorefining driven by nitrogen-responsive transcriptional machinery
Biotechnology for Biofuels volume 13, Article number: 29 (2020)
Protein-based bioconversion has been demonstrated as a sustainable approach to produce higher alcohols and ammonia fertilizers. However, owing to the switchover from transcription mediated by the bacterial RNA polymerase σ70 to that mediated by alternative σ factors, the biofuel production driven by σ70-dependent promoters declines rapidly once cells enter the stationary phase or encounter stresses. To enhance biofuel production, in this study the growth phase-independent and nitrogen-responsive transcriptional machinery mediated by the σ54 is exploited to drive robust protein-to-fuel conversion.
We demonstrated that disrupting the Escherichia coli ammonia assimilation pathways driven by glutamate dehydrogenase and glutamine synthetase could sustain the activity of σ54-mediated transcription under ammonia-accumulating conditions. In addition, two σ54-dependent promoters, argTp and glnAp2, were identified as suitable candidates for driving pathway expression. Using these promoters, biofuel production from proteins was shown to persist to the stationary phase, with the net production in the stationary phase being 1.7-fold higher than that derived from the optimal reported σ70-dependent promoter PLlacO1. Biofuel production reaching levels 1.3- to 3.4-fold higher than those of the σ70-dependent promoters was also achieved by argTp and glnAp2 under stressed conditions. Moreover, the σ54-dependent promoters realized more rapid and stable production than that of σ70-dependent promoters during fed-batch fermentation, producing up to 4.78 g L − 1 of total biofuels.
These results suggested that the nitrogen-responsive transcriptional machinery offers the potential to decouple production from growth, highlighting this system as a novel candidate to realize growth phase-independent and stress-resistant biofuel production.
Protein-based biorefining for the production of biofuels and ammonia fertilizer constitutes a promising technology to simultaneously reclaim the carbon and nitrogen from waste proteins [1,2,3,4,5]. To realize protein-to-fuel conversions, the amino acids must be forced to degrade through artificial transamination and deamination cycles in the host organism in order to release the carbon skeletons for biofuel synthesis (Fig. 1a) . However, although currently utilized conversion pathways have been precisely designed, the resultant biofuel productivity remains far below the theoretical level [3, 6]. A major hurdle preventing efficient conversion of proteins into value-added chemicals is the growth-dependency of the production process . As high productivity can only persist in periods of cell growth, but rapidly ceases once the cells enter the stationary phase or encounter stresses, a considerable proportion of the protein source is reallocated to cell growth and maintenance rather than biofuel production. Therefore, to enhance production, we hypothesized that the protein-to-fuel flux could be engineered to resist shifts in the growth phase and the stresses accompanying the fermentation process.
The protein-to-fuel flux is sustained by robust expression of the biofuel synthetic pathway. In general, the engineered biosynthetic pathways in bacteria for the production of value-added chemicals are mostly governed by σ70-dependent promoters , the transcription of which is determined by the number of the RNA polymerase (RNAP) carrying the σ70 subunit (Eσ70). Owing to its dominant abundance, this subunit can easily outcompete alternative σ factors for the finite core RNAP . However, the relative advantage of σ70 over other σ factors in recruiting the core enzyme is highly compromised once the cells enter the stationary phase or encounter stresses. This derives in part from the sharp increase in the number of alternative σ factors (e.g., σ38 and σ24) in response to both intra- and extracellular disturbances. In addition, regulatory molecules such as Hofmeister salts, regulator of σD, and guanosine pentaphosphate or tetraphosphate [9, 10] simultaneously accumulate, whereas chromosomal DNA supercoiling decreases . Together, these physiological shifts suppress the association between the core RNAP and σ70, facilitating core RNAP interaction with alternative σ factors at the expense of Eσ70 . As a result, the protein conversion pathway ceases to function once the fermentation persists to the late stage. To overcome the innate drawbacks of σ70-mediated transcription, we posited that metabolic engineering could transform the intrinsic transcriptional regulation process into a driving force for the robust biorefining of waste protein.
In particular, the nitrogen-responsive transcriptional machinery offers a possible solution to achieve growth phase-independent and stress-resistant protein-to-fuel conversion. This machinery consists of the RNAP-σ54 holoenzyme (Eσ54), the nitrogen regulatory proteins, and the corresponding σ54-dependent promoters. For Escherichia coli, σ54 constitutes one of the most abundant σ factors following σ70, and its intracellular concentration tends to remain stable throughout the whole growing stages . In addition, σ54 appears to exhibit the highest affinity to the core RNAP among all the alternative σ factors  and as the regulatory molecules accumulate, the association between σ54 and the core RNAP is further strengthened upon entering into the stationary phase or encountering stresses. In combination, these determinants would be expected to facilitate the formation and maintenance of sufficient Eσ54 to support biofuel production under a wide range of physiological conditions.
Moreover, the transcription of σ54-dependent promoters is mostly related to the nitrogen status. A limited supply of ammonia can activate σ54-mediated transcription, whereas it is inhibited by an increase in the ammonia concentration. This nitrogen-specific response of the σ54-dependent promoter is achieved through a cascade of phosphorylation of the nitrogen regulatory proteins such as nitrogen regulatory protein C (NtrC). Under nitrogen starvation conditions, NtrC receives the phosphoryl group from phosphorylated nitrogen regulatory protein B (NtrB) and hydrolyzes ATP, which enables Eσ54 to denature the double-stranded DNA and initiate transcription (Fig. 1b). Notably, compared with ammonia, the amino acids support slower growth and are considered to be poor nitrogen sources . The use of amino acids from protein hydrolysate as the sole nitrogen supply can thus cause nitrogen starvation and induce σ54-mediated transcription [14, 15]. Therefore, the nitrogen-responsive transcriptional machinery holds promise to resist transcriptional switchover during the stationary phase and under stressed conditions, and should remain active throughout the whole process of protein conversion (Fig. 1c).
Accordingly, in this study we aimed to regulate the expression of the protein conversion pathway to realize robust production of biofuels from protein biomass. Toward this end, the ammonia assimilation pathway in E. coli was first engineered to sustain the activities of σ54-dependent promoters. To exploit σ54-mediated transcription, a series of NtrC-dependent promoters were characterized using a fluorescence reporting system under nitrogen starvation conditions and the promoters exhibiting high activities were identified. Effects of the candidate promoters in maintaining robust protein-to-fuel conversion were evaluated under both optimal and stressed conditions with various protein sources. We postulated that the nitrogen-responsive transcriptional machinery could realize stable pathway expression throughout exponential growth to the stressed stationary phase, rendering the σ54-mediated transcriptional machinery a novel candidate to drive robust chemical production from waste proteins.
Influence of ammonia assimilation on sustaining σ54-mediated transcription
In general, to maintain active σ54-mediated transcription, the E. coli cells must be maintained under nitrogen starvation conditions. However, the deamination of amino acids will release NH3, which would be reused by the cells as a preferred nitrogen source; as NH3 accumulates, the σ54-dependent promoters would cease to function. Therefore, disrupting the intrinsic ammonia assimilation pathway was flagged as the key to achieving continuous transcription from σ54-dependent promoters. Genes involved in the glutamate dehydrogenase (GDH) and glutamine synthetase–glutamate synthase (GS–GOGAT) pathways  were selectively deactivated (Fig. 2a) and the activity of σ54-mediated transcription was measured using a green fluorescence protein (GFP)-based reporting system driven by the typical σ54-dependent promoter glnAp2 in an amino acid-rich environment. As expected, suppressing ammonia assimilation led to sharp increase in the fluorescence intensity (GFP/OD600) from nearly 0 (strain LM10) to 2800 (LM13) (Table 1 and Fig. 2b). Disrupting the GDH and GOGAT pathways by knocking out gdhA together with gltB or gltD led to similar fluorescence intensity levels of around 1100, as shown for LM11 and LM12, respectively. In comparison, disruption of the GS instead of the GOGAT pathway in a GDH-deficient background achieved a 2.4-fold increase in the fluorescence intensity (LM13).
To further enhance the transcriptional activity, the quorum sensing (QS) pathway, which has the potential to increase strain robustness under stressed conditions , was blocked by deleting lsrA, which encodes the transporter for autoinducer-2. This GDH-, GS-, and QS-deficient strain (LM14) exhibited 1.8-fold increase in glnAp2-mediated transcription compared with that of LM13 (Fig. 2b). When amino acids were supplied as the sole nitrogen source, the maximum OD600 of LM14 was half that of LM10. The growth rate of LM14 at the exponential phase was 15 to 47% that of LM10 (Additional File 1: Figure S1). As a result, more resources in LM14 were allocated to biofuel production than that of LM10 (Additional File 1: Figure S2).
Mining of σ54-dependent promoters
To identify σ54-dependent promoters that are able to maintain high transcriptional activity under nitrogen-limited conditions, a total of 12 σ54-dependent promoters (Additional file 1: Table S1), of which 10 were regulated by the nitrogen regulator NtrC, were each inserted upstream of gfp and characterized based on fluorescence intensity. When yeast extract was used as the sole nitrogen source, strains harboring different gfp expression cassettes exhibited fluorescence intensities ranging from 538 to 29,836 in the exponential phase. The gfp expression driven by argTp showed the highest intensity, which was 1.7-fold that of glnAp2 and 17–56 times higher than that of the remaining candidates (Fig. 3a). The transcriptional activities for argTp and glnAp2 were strengthened when cells entered the stationary phase (Fig. 3b). Compared with the exponential phase, 1.8- and 1.3-fold increases in the fluorescence intensities for argTp and glnAp2 were observed, respectively. The superior activity of argTp was confirmed by repeated measurement (Additional file 1: Figure S3), for which the fluorescence intensity generated from the argTp-gfp construct was 2.7–7.5 times higher than that of the remaining candidates. However, the fluorescence intensity produced by the glnAp2-gfp construct was comparable to that of the remaining promoters upon assay repetition. Nevertheless, to cover as many potential candidates as possible, both argTp and glnAp2 were tested for their performance with regard to driving protein-to-fuel conversion.
Sequence alignment showed that seven promoters including argTp and glnAp2 contained − 12/− 24 elements identical to the previously defined − 12/− 24 consensus sequences  (Fig. 3c) and three promoters including astCp2, glnKp, and yhdWp exhibited perfect matches with the consensus, whereas puuPp and pabBp6 displayed 1–3 mismatches in both their − 12 and − 24 elements. Mismatches in the extended − 12/− 24 regions (− 15 to − 8 and − 31 to − 20) in comparison to the consensus sequences are also presented in Fig. 3c.
Biofuel production driven by the σ54-dependent promoters
The effect of σ54-dependent promoters in driving protein-to-fuel conversion was investigated in comparison with that of three σ70-dependent promoters including rrnBp1, J23100, and PLlacO1. rrnBp1 constitutes a typical σ70-dependent promoter that drives the bulk transcription of ribosomal DNA . J23100 is a strong synthetic promoter (iGEM Part: BBa J23100) and PLlacO1 is a commonly used promoter for biofuel production [3, 17]. Isobutanol (C4) and methylbutanols (2-methyl-1-butanol and 3-methyl-1-butanol, C5) were produced through the synthetic pathway. For all promoters, the total biofuel titers increased sharply in the first 48 h when cells were in the exponential phase (Fig. 4a–e). Upon entering into the stationary phase, the biofuel production from rrnBp1 and J23100 ceased rapidly (Fig. 4d, e). In comparison, the biofuel titer for strain LM19 with promoter argTp continued to increase at a steady rate during the stationary phase with a total of 0.84 g L− 1 biofuel being produced from 48 to 120 h (Fig. 4a), which was 1.7-fold of the corresponding net biofuel production for strain LM15 with promoter PLlacO1 (Fig. 4c). argTp, glnAp2, and PLlacO1 promoters produced similar amounts of the C4 and C5 alcohols in the final products, whereas the C5 alcohol dominated the produced biofuels for J23100 and rrnBp1. The σ70-dependent promoters achieved final titers of 0.17 to 1.03 g L− 1 for C4 alcohol and 0.63 to 1.13 g L− 1 for C5 alcohol, whereas higher titers of 0.79 to 1.14 and 0.97 to 1.36 g L− 1 were achieved for C4 and C5 alcohols, respectively, by the σ54-dependent promoters. In total, the pathway driven by argTp produced 2.50 g L− 1 biofuels, which was 16% higher than that of PLlacO1. glnAp2 also showed advantage over the σ70-dependent promoters by achieving a final titer 1.3- to 2.2-fold higher than that of J23100 and rrnBp1.
The performance of σ54-mediated biosynthesis was also evaluated under stressed conditions. Compared with optimal conditions, osmotic stress induced by 400 mM NaCl suppressed biofuel production of the σ70-dependent promoters by 41% to 49% (Fig. 4f). In comparison, the σ54-dependent promoter sustained its advantage in biofuel production. argTp achieved the highest biofuel production of 1.61 g L− 1, which was 26% higher than that of PLlacO1. A titer equivalent to 1.6- and 2.4-fold that of J23100 and rrnBp1, respectively, was also achieved by glnAp2. Moreover, stress related to sharply decreased pH may also arise consequent to the pretreatment of protein biomass. Compared with the σ70-dependent promoters, argTp retained its advantage in driving pathway overexpression under an initially acidified (pH 5.0) condition, leading to one of the highest biofuel titers of 1.77 g L− 1 (Fig. 4g). The biofuel production driven by glnAp2 also exceeded that of J23100 and rrnBp1 by 2.1- and 1.5-fold, respectively.
The biofuel synthetic pathway driven by argTp produced more pathway enzymes than that driven by the σ70-dependent promoters in both the optimal and the stressed conditions. When fermentation proceeded to the stationary phase under optimal conditions, the activity of the acetolactate synthase (AlsS) enzyme expressed via the argTp promoter was 42% higher than that from PLlacO1 (Additional file 1: Figure S4a). Under osmotic (Additional file 1: Figure S4b) or acid stresses (Additional file 1: Figure S4c), the activity of AlsS expressed via argTp remained 34% and 93% higher than that from PLlacO1, respectively, and was 1.5- to 3.5-fold that of the remaining σ70-dependent promoters.
Batch fermentation driven by the σ54-dependent promoters
To mimic the industrial fermentation process, a fed-batch fermentation was carried out in the presence of oleyl alcohol to extract the produced biofuel from the aqueous phase (Additional file 1: Figure S5). The σ54-dependent promoters realized a rapid and stable production of biofuels, with a longer producing period and a higher final titer than those of σ70-dependent promoters. In the exponential phase, the σ54-dependent promoters synthesized the biofuels at a rate of 47.81 to 49.81 mg L− 1 h− 1 with the titer reaching 2.29 to 2.39 g L− 1. After entering the stationary phase, these promoters continued to produce at a rate of 10.47 to 14.22 mg L− 1 h− 1 for 168 h, with the final titer reaching 4.05 to 4.78 g L− 1. In contrast, the σ70-dependent promoters produced biofuels at a rate of 35.69 to 42.26 mg L− 1 h− 1 during the exponential phase. However, the production rate dropped sharply by 66 to 77% upon entering the stationary phase and then decreased to zero after 72 h. The final titer achieved by σ70-dependent promoters was 75 to 88% that of the σ54-dependent promoters.
Biofuel production from waste protein biomass
To examine the performances of the σ54-dependent promoters in driving waste protein conversion, microbial proteins from E. coli and Corynebacterium glutamicum cells, and plant proteins from soybean meal were used as feedstock for biofuel production (Fig. 5a). A total of 0.32 to 1.55 g L− 1 higher alcohols were produced using these protein biomasses. When E. coli biomass was supplied, 62 to 81% of the produced biofuel comprised C5 alcohol, whereas for C. glutamicum biomass, the main portion of the biofuel changed to C4 alcohol. Plant proteins produced lower amounts of biofuels ranging from 0.32 to 0.54 g L− 1, of which the majority constituted C4 alcohol, which accounted for 78 to 82% of the total biofuel (Fig. 5b). The highest biofuel production was achieved when yeast protein was used as the feedstock, with C5 alcohol representing 50 to 77% of the titer. In general, pathways driven by the σ54-dependent promoters produced more biofuel than those driven by the σ70-dependent promoters. argTp demonstrated advantage over the other promoters for fermentation with all feedstocks. On average, the biofuel production driven by argTp was 1.41 g L− 1, which was up to 2.53-fold that the production driven by σ70-dependent promoters. Pathways driven by the σ54-dependent promoters also produced higher proportions of C4 alcohol in the total biofuel than those from the σ70-dependent promoters.
In this study, we demonstrate that the nitrogen-responsive transcriptional machinery in E. coli is able to drive robust biofuel production from protein biomass. Compared with the commonly used σ70-dependent promoters, σ54-mediated biofuel production could retain high productivity in the stationary phase and achieve higher biofuel titers under both optimal and stressed conditions. With simple modifications of the ammonium assimilation pathways and the use of suitable promoters, the σ54-mediated scheme therefore offers a promising alternative to the conventional σ70-mediated approach for the biorefining of waste proteins.
The nitrogen-responsive transcriptional machinery provides multiple advantages over the conventional σ70-mediated scheme in driving protein conversion [3, 17]. The first is extension of the period supporting biosynthetic activity, as the pathway driven by argTp retained a relatively high biofuel productivity from early to late stationary phases, whereas that driven by σ70-dependent promoters basically ceased to function upon completion of cell growth. In particular, although the σ70-dependent promoters could only retain activity for 48–72 h after entering the stationary phase, the σ54-dependent promoters could function for more than 168 h (Additional file 1: Figure S5). Notably, maintaining production in the stationary phase is of marked significance for the industrial production of chemicals. For example, the cell factories face trade-offs between growth and production. When cells enter the stationary phase, the resources dedicated for growth are minimized and could be rechanneled for production; thus, this represents an optimal stage to boost overall yield. For this purpose, the concept of decoupling growth and production has been proposed and considered as a grand challenge for metabolic engineering [21, 22]. In the present study, glnAp2 and argTp reached the highest yield when using microbial protein biomass as the feedstock, achieving 17% of the theoretical level (Additional file 1: Figure S6), and was up to 2.93-fold that of the σ70-dependent promoters. Therefore, improving productivity in the stationary phase appears to be beneficial for industrial processes. Moreover, most industrial production consists of continuous batch fermentation, in which the majority of production occurs during the stationary phase . In batch fermentation, the strain carrying the σ70-dependent promoter could only maintain productivity up to 11.13 mg L− 1 h− 1 after entering the stationary phase, whereas the σ54-dependent promoter retained a productivity 28% higher throughout the whole stationary phase (Additional file 1: Figure S5).
The nitrogen-responsive transcriptional machinery also avoids the utilization of inducers, which are essential for the biosynthetic pathways governed by inducible σ70-dependent promoters such as the isopropyl β-d-1-thiogalactopyranoside (IPTG)-induced PLlacO1. The exclusion of chemical inducers reduces the fermentation cost and is more feasible for industrial scale-up of the protein-based biorefinery. Additionally, as the NtrC-mediated promoters mainly respond to nitrogen starvation, their transcription can be strengthened through depletion of the nitrogen source, as fermentation persists to the late stages. This intrinsic driving force can compensate for the reduced pathway expression resulting from the stationary phase and accompanying stresses, endowing argTp and glnAp2 equivalent or even superior performance than that of the strong inducible promoter PLlacO1.
The resistance to general stresses such as high osmolarity and low pH constitutes another advantage, as these can significantly inhibit biofuel production driven by σ70-dependent promoters [24, 25]. It should be noted that for protein-based biosynthesis, the pH of the medium is less likely to decrease during the fermentation due to the continuous release of NH3 from amino acid deamination. However, acid stress can still stem from pretreatment processes, such as purification and hydrolysis of the protein biomass . Therefore, the stress resistance endowed by σ54-mediated biosynthesis may further promote robust biofuel production from waste proteins. The biofuel titers achieved in this study did not exceed those previously reported . This may be largely attributed to differences in the yeast extract used for feeding, for which the amino acid concentration was only half that of the yeast extract used in other studies [3, 6].
Maintaining nitrogen starvation constitutes an important prerequisite to induce stable transcription of the nitrogen-responsive promoters. When using amino acids as the feedstock, the knockout of both the GDH and GS pathways could create a nitrogen-limited intracellular status under an ammonia-rich environment, thereby achieving high activities of the NtrC-mediated promoters. This specific phenotype could be inferred from the regulatory cascade in nitrogen assimilation (Additional file 1: Figure S7). For many bacteria, the nitrogen status is sensed by the bifunctional uridylyltransferase/uridylyl-removing enzyme (GlnD) and PII signal transduction systems. In E. coli, GlnD responds to intracellular glutamine and PII is regulated by 2-oxoglutarate . The accumulation of glutamine leads to the dephosphorylation of phosphorylated NtrC (NtrC-P) and suppresses NtrC-mediated transcription. In contrast, excessive 2-oxoglutarate promotes the release of free NtrB and subsequently increases the level of NtrC-P and activates the targeted transcription. In general, a high 2-oxoglutarate to glutamine ratio indicates nitrogen starvation [16, 27], which signals the cell to activate nitrogen assimilation pathways controlled by the NtrC. In this case, the conversion of 2-oxoglutarate to glutamate is prevented owing to the deficiency of GDH, and the production of glutamine from glutamate is also blocked after knocking out the GS. This results in the accumulation of 2-oxoglutarate and consumption of glutamine, leading to an increased ratio between these two effectors and thus deceiving the cells into maintaining high transcriptional activities of the NtrC-mediated promoters. Upon deletion of the ammonia assimilation pathway, the engineered strain can no longer reuptake the newly produced NH3, which limits its growth. However, when equipped with a σ54-dependent promoter, the LM19 strain could reallocate more resources dedicated to growth instead to production, achieving 3.16-fold higher biofuel production than that of the LM10 strain equipped with biofuel synthetic pathways (Additional file 1: Figure S2).
Disruption of QS further enhanced protein expression driven by the σ54-dependent promoters. When cells enter the stationary phase or encounter stresses, the QS signaling molecule autoinducer-2 (AI-2) accumulates , causing DNA damage and metabolic shifts [29, 30]. Therefore, blocking AI-2 uptake by deleting lsrA, which encodes the ATP-binding component of the AI-2 transporter, could prevent the cell population from collectively switching their metabolic state from biosynthesis to maintenance, thereby leading to robust chemical production throughout the whole growth phase.
Analysis of the σ70-dependent promoters suggested that inclusion of the σ binding sites most resembling the consensus sequence would facilitate promoter opening and the formation of an open complex, thereby increasing the transcriptional strength of the promoters . However, this appeared not to be the case for the 12 σ54-dependent promoters evaluated in the present study, as promoters (e.g., astCp2, glnKp, and yhdWp) exhibiting perfect matches to the − 12 and − 24 consensus sequences basically showed the lowest transcriptional activities throughout the exponential to the stationary phases (Fig. 3). In comparison, although containing mismatches in the − 12 or − 24 element to the consensus sequence, both the argTp and glnAp2 promoters exhibited dominant transcriptional activities in both the exponential and stationary phases. Therefore, the resemblance of the core promoter elements to their consensus sequences might not constitute a reliable criterion for identifying strong σ54-dependent promoters. The lack of a close relationship between the transcriptional strength and the degree of sequence conservation for the σ54-dependent promoters may be attributed to the involvement of activators, which play additional roles in initiating σ54-dependent transcription. Taking this into consideration, sequences in the upstream region of the core promoter elements may be crucial for activator binding, its orientation with regard to Eσ54 , and the DNA looping that leads to the formation of the activator–Eσ54 complex . In the present case, the NtrC binding sites on argTp and glnAp2 might provide the proper binding strength and favored relative positioning of NtrC-P toward Eσ54. In addition, the sequences in between the NtrC binding sites and the core promoter region may also produce the right angle of DNA bending to facilitate the NtrC–Eσ54 interaction. Therefore, the outstanding performance of argTp and glnAp2 could likely be attributed to the combined effects of the core promoter regions, NtrC binding sites, and intervening bending regions. Moreover, transcription from argTp has been shown to increase with procession of the stationary phase , which supports the use of argTp to drive robust pathway expression.
Our results demonstrated that the composition of produced alcohol is affected by two factors, the amino acid composition of the protein biomass and the dominance of the biofuel synthetic pathway. As the precursors of higher alcohols, branched-chain amino acids (BCAAs) including valine, leucine, and isoleucine can be directly converted into the corresponding alcohols. Thus, the relative abundance of BCAAs in the raw material could directly affect the composition of the produced biofuels. In addition, driven by the biofuel synthetic pathway, the central metabolite pyruvate would be mainly converted to C4 alcohol . Therefore, the more pyruvate is converted from amino acids other than BCAAs, the more C4 alcohol would be produced. However, as pyruvate might be channeled to various pathways, the metabolic flux to biofuels relies on the strength of the biofuel synthetic pathway. A strong and stable promoter would drive the continuous expression of the biofuel synthetic pathway, converting more pyruvate to C4 isobutanol. This may be the reason that the proportion of the produced C4 alcohol in the total biofuel driven by argTp and glnAp2 exceeded that of the σ70-dependent promoters. However, it should be noted that both the C4 and C5 alcohols are ideal alternatives to traditional gasoline . Therefore, the composition of the final products has little effect on the quality of the produced biofuels.
By targeting protein bioconversions, this study demonstrated that the nitrogen-responsive transcriptional machinery can be employed for chemical production and has the potential to realize growth phase-independent and stress-resistant overexpression of the biosynthetic pathways. Constructing robust microbial cell factories is essential for achieving higher productivity and represents an important challenge for metabolic engineering . σ54-mediated transcriptional regulation offers a convenient solution as it requires only the substitution of a single promoter and small modification of the ammonia assimilation pathway. Subsequent engineering of the host strain, reconstruction of the synthetic pathway, and optimization of the fermentation process might also be performed to further enhance the production. Considering the wide distribution of σ54-dependent promoters in prokaryotes [36, 37], the σ54-mediated regulation scheme might be further exploited by mining the natural promoter libraries, as for E. coli in particular, nearly one hundred σ54-dependent promoters have already been experimentally identified or predicted according to the latest collections in RegulonDB . To expand the transcriptional strength and dynamic range of σ54-mediated transcription, synthetic σ54-dependent promoters might also be generated by engineering the naturally occurring − 12/− 24 elements, spacer region, activator binding sites, or the sequence responsible for DNA looping. These endeavors would facilitate the fine-tuning of the σ54-mediated biosynthetic pathways.
The waste proteins as feedstock can be supplied in various forms including microbial, plant, and animal biomass for the σ54-mediated biorefineries. In addition to the abundant sources of feedstock, another advantage is that the σ54-mediated biosynthetic scheme could theoretically be applied for the production of a myriad of value-added amino acid derivatives (Additional file 1: Figure S8), such as carbocyclic aromatic compounds derived from the shikimate pathway for aromatic amino acid biosynthesis. However, it should be noted that efficient protein-based biorefinement relies upon rewiring the transamination network that channels different amino acids into specific amino acid precursors. Therefore, the design of the transamination network should consider the amino acid compositions of the source materials (Additional file 1: Figure S9) and the corresponding amino acid synthetic pathways for the targeted products. Combined with the σ54-dependent promoters, the engineered transamination and the biosynthetic pathway would be expected to promote efficient recycling of both the ammonia and carbon skeletons from waste proteins.
Theoretically, the σ54-mediated biosynthetic scheme is not limited to protein conversion. When the nitrogen supply is tightly controlled, σ54-mediated transcription can remain active throughout the whole growth stage. Therefore, when materials other than proteins are used as the feedstock, the amino acids may serve as the poor nitrogen source rather than ammonia. As a result, the low availability of the amino acids would trigger the NtrC-mediated promoters and achieve the robust conversion of renewable sources into value-added chemicals.
Overall, our findings demonstrated that biosynthesis mediated by the nitrogen-responsive transcriptional machinery outcompetes the conventional σ70-mediated scheme by enhancing the robustness and productivity of the biosynthetic pathways, rendering it a novel solution to realize growth phase-independent and stress-resistant protein-to-fuel conversion.
Strains and plasmids
A previously reported E. coli strain with enhanced amino acid utilization was used for the conversion of proteins into biofuels (Table 1). The derivative strains with gene deletions including gdhA, gltB, gltD, or lsrA were created by using P1 transduction or λ phage recombination. Two adjacent promoters were located upstream of glnA: the σ54-dependent glnAp2 and the σ70-dependent glnAp1. The latter was located between the two NtrC binding sites at the 5′ end of glnAp2. To eliminate the interference from glnAp1 on σ54-mediated transcription, only the − 1 to − 99 region that encompasses the core glnAp2 promoter and its first three NtrC binding sites from the 5′ end was cloned from E. coli MG1655 genomic DNA. Other σ54-dependent promoters were also cloned from the genomic DNA and inserted into the 5′ end of the two gene cassettes comprising the biofuel biosynthetic pathway  using Gibson assembly (Additional file 1: Tables S1 and S2). Cloning was carried out using E. coli strain XL10-Gold (Agilent Technologies, Santa Clara, CA, USA). The E. coli strains were routinely cultured in Luria–Bertani (LB) broth or LB agar supplemented with 50 μg mL− 1 kanamycin or 100 μg mL− 1 ampicillin.
Medium and fermentation
Amino acid medium used for biofuel production comprised 40 g L− 1 yeast extract (amino acid profile shown in Fig. 5a) with M9 salt containing 6.0 g L− 1 Na2HPO4, 3.0 g L− 1 KH2PO4, 0.5 g L− 1 NaCl, 0.12 g L− 1 MgSO4, 11 mg L− 1 CaCl2, and 1.0 mg L− 1 vitamin B1. Ampicillin (100 μg mL− 1) and kanamycin (50 μg mL− 1) were added when required. The E. coli and C. glutamicum grown in LB medium were used as microbial protein sources and soybean meal was used as the plant protein source. The bacterial cells were treated by ultrasonication at 500 W for 40 min, whereas the soybean meal was first autoclaved at 121 °C for 20 min and then disrupted ultrasonically. Released proteins were measured using the Bradford assay and hydrolyzed overnight by protease (120 U mg− 1) at a concentration of 4 g kg− 1 (dry weight). The produced free amine groups were then quantified using the ninhydrin reaction. For all protein sources, the total amount of peptides and amino acids used for fermentation was adjusted to 12.8 g L− 1, which is equivalent to the protein concentration in 40 g L− 1 yeast extract (Angel Yeast Co., Ltd., Yichang, Hubei, China). For biofuel fermentation, the overnight seed culture was prepared in 5 mL LB medium at 37 °C in a shaker at 250 rpm. The culture was inoculated at 1% into 20 mL of amino acid medium in a 250 mL screw-cap conical flask. Fermentation was performed in triplicate at 30 °C in a shaker at 250 rpm. The cell OD and the concentrations of higher alcohols were measured at defined time intervals. For continuous fermentation, an equal volume of oleyl alcohol was added to the flask prior to incubation for extraction of the produced isobutanol. Following inoculation, IPTG was added to a final concentration of 0.1 mM. Fermentation was performed in triplicate at 30 °C in a shaker (250 rpm). The aqueous and organic phases were sampled (5 mL each) at defined time intervals. After sampling, the fermentation broth was replenished with fresh medium and oleyl alcohol. All amino acids mentioned in this study were l-amino acids except glycine, which has no chirality.
Assay of promoter strength
Cells were cultured in 5 mL LB medium at 37 °C in a shaker at 250 rpm. The overnight culture was inoculated at 1% into 200 μL amino acid medium in a 96-well plate, which was sealed with breathable film and incubated at 37 °C. At defined time points, the fluorescence was measured using a Cytation Hybrid Multi-Mode Reader (BioTek, Winooski, VT, USA) with the excitation and detection wavelengths set at 400 and 508 nm, respectively; the OD600 was also measured. The ratio of fluorescence to OD600 (GFP/OD600) was used to represent the promoter strength. The background fluorescence was measured using a strain harboring a promoterless plasmid that carries the gfp gene.
The activity of AlsS was measured according to Atsumi . Briefly, 50 μL crude cell extract was mixed with 150 μL MOPS buffer (pH 7.0) containing 100 mM MOPS, 20 mM sodium pyruvate, 0.1 mM thiamine pyrophosphate, and 1 mM MgCl2. The enzyme reaction was carried out at 37 °C for 30 min and terminated by adding 20 μL of 50% H2SO4. The produced 2-acetolactate then underwent acid hydrolysis to form acetoin, which could be quantified using the Voges–Proskauer assay by measuring the absorbance of the red mixture at 535 nm . The enzyme activity was represented as the amount of 2-acetolactate produced by 1 mg of total protein in 1 min.
Gas chromatography (GC) detection of higher alcohols
Higher alcohols were quantified using an Agilent 6890 GC with flame ionization detector. A DB-FFAP capillary column (30 m × 0.32 mm × 0.25 μm; Agilent Technologies) was used to separate C5 and C4 alcohols with n-pentanol as the internal standard. For the analysis of higher alcohols in aqueous phase, the GC oven temperature was first held at 80 °C for 3 min, increased to 230 °C at 115 °C min− 1, and held for 1 min. The alcohols in organic phase were separated by first holding the oven temperature at 90 °C for 0.5 min followed by heating at a rate of 20 °C min−1 to 110 °C and holding for 0.5 min. The temperature was then increased to 235 °C at 120 °C min−1 and held for 2 min. Samples were injected at a split ratio of 1:50 and detected at 280 °C.
High-performance liquid chromatography detection of amino acids
The concentrations of amino acids were measured using an Agilent 1290 Infinity™ II liquid chromatography system equipped with a Durashell C18(L) column (Bonna-Agela Technologies, Torrance, CA, USA) using the phenylisothiocyanate derivatization method . A solution containing 0.1 M sodium acetate (pH 6.5) and acetonitrile in volumetric ratio of 99.3:0.7, and a solution containing 80% acetonitrile were used as mobile phases . The derived amino acids were detected at 254 nm using a diode array detector.
Availability of data and materials
All data generated or analyzed during this study are included in this published article.
- Eσ70 :
- Eσ54 :
Nitrogen regulatory protein B
Nitrogen regulatory protein C
Green fluorescent protein
Branched-chain amino acid
Li S-Y, Ng I-S, Chen PT, Chiang C-J, Chao Y-P. Biorefining of protein waste for production of sustainable fuels and chemicals. Biotechnol Biofuels. 2018;11:256.
El-Dalatony MM, Saha S, Govindwar SP, Abou-Shanab RAI, Jeon B-H. Biological conversion of amino acids to higher alcohols. Trends Biotechnol. 2019;37:855–69.
Huo YX, Cho KM, Rivera JG, Monte E, Shen CR, Yan Y, Liao JC. Conversion of proteins into biofuels by engineering nitrogen flux. Nat Biotechnol. 2011;29:346–51.
Wernick DG, Liao JC. Protein-based biorefining: metabolic engineering for production of chemicals and fuel with regeneration of nitrogen fertilizers. Appl Microbiol Biotechnol. 2013;97:1397–406.
Huo Y-X, Wernick DG, Liao JC. Toward nitrogen neutral biofuel production. Curr Opin Biotechnol. 2012;23:406–13.
Choi K-Y, Wernick DG, Tat CA, Liao JC. Consolidated conversion of protein waste into biofuels and ammonia using Bacillus subtilis. Metab Eng. 2014;23:53–61.
Jin L-Q, Jin W-R, Ma Z-C, Shen Q, Cai X, Liu Z-Q, Zheng Y-G. Promoter engineering strategies for the overproduction of valuable metabolites in microbes. Appl Microbiol Biotechnol. 2019;103:8725–36.
Österberg S, Peso-Santos TD, Shingler V. Regulation of alternative sigma factor use. Annu Rev Microbiol. 2011;65:37–55.
Sharma UK, Chatterji D. Transcriptional switching in Escherichia coli during stress and starvation by modulation of σ70 activity. FEMS Microbiol Rev. 2010;34:646–57.
Gralla JD, Huo Y-X. Remodeling and activation of Escherichia coli RNA polymerase by osmolytes. Biochemistry. 2008;47:13189–96.
Kusano S, Ding Q, Fujita N, Ishihama A. Promoter selectivity of Escherichia coli RNA polymerase Eσ70 and Eσ38 holoenzymes: effect of DNA supercoiling. J Biol Chem. 1996;271:1998–2004.
Jishage M, Iwata A, Ueda S, Ishihama A. Regulation of RNA polymerase sigma subunit synthesis in Escherichia coli: intracellular levels of four species of sigma subunit under various growth conditions. J Bacteriol. 1996;178:5447.
Maeda H, Fujita N, Ishihama A. Competition among seven Escherichia coli σ subunits: relative binding affinities to the core RNA polymerase. Nucleic Acids Res. 2000;28:3497–503.
Wang J, Yan D, Dixon R, Wang Y-P. Deciphering the principles of bacterial nitrogen dietary preferences: a strategy for nutrient containment. mBio. 2016;7:e00792-00716.
Maheswaran M, Forchhammer K. Carbon-source-dependent nitrogen regulation in Escherichia coli is mediated through glutamine-dependent GlnB signalling. Microbiology. 2003;149:2163–72.
Huergo LF, Dixon R. The emergence of 2-oxoglutarate as a master regulator metabolite. Microbiol Mol Biol Rev. 2015;79:419.
Atsumi S, Hanai T, Liao JC. Non-fermentative pathways for synthesis of branched-chain higher alcohols as biofuels. Nature. 2008;451:86.
Atsumi S, Cann AF, Connor MR, Shen CR, Smith KM, Brynildsen MP, Chou KJ, Hanai T, Liao JC. Metabolic engineering of Escherichia coli for 1-butanol production. Metab Eng. 2008;10:305–11.
Barrios H, Valderrama B, Morett E. Compilation and analysis of σ54-dependent promoter sequences. Nucleic Acids Res. 1999;27:4305–13.
Rosenow C, Saxena RM, Durst M, Gingeras TR. Prokaryotic RNA preparation methods useful for high density array analysis: comparison of two approaches. Nucleic Acids Res. 2001;29:e112–e112.
Chubukov V, Mukhopadhyay A, Petzold CJ, Keasling JD, Martín HG. Synthetic and systems biology for microbial production of commodity chemicals. NPJ Syst Biol Appl. 2016;2:16009.
Venturelli OS, Egbert RG, Arkin AP. Towards engineering biological systems in a broader context. J Mol Biol. 2016;428:928–44.
Venayak N, Anesiadis N, Cluett WR, Mahadevan R. Engineering metabolism through dynamic control. Curr Opin Biotechnol. 2015;34:142–52.
Huo Y-X, Guo L, Ma X. Biofuel production with a stress-resistant and growth phase-independent promoter: mechanism revealed by in vitro transcription assays. Appl Microbiol Biotechnol. 2018;102:2929–40.
Ma R, Zhang Y, Hong H, Lu W, Lin M, Chen M, Zhang W. Improved osmotic tolerance and ethanol production of ethanologenic Escherichia coli by IrrE, a global regulator of radiation-resistance of Deinococcus radiodurans. Curr Microbiol. 2011;62:659–64.
Huergo LF, Chandra G, Merrick M. PII signal transduction proteins: nitrogen regulation and beyond. FEMS Microbiol Rev. 2013;37:251–83.
van Heeswijk WC, Westerhoff HV, Boogerd FC. Nitrogen assimilation in Escherichia coli: putting molecular data into a systems perspective. Microbiol Mol Biol Rev. 2013;77:628.
Moslehi-Jenabian S, Gori K, Jespersen L. AI-2 signalling is induced by acidic shock in probiotic strains of Lactobacillus spp. Int J Food Microbiol. 2009;135:295–302.
Winzer K, Hardie KR, Burgess N, Doherty N, Kirke D, Holden MTG, Linforth R, Cornell KA, Taylor AJ, Hill PJ, Williams P. LuxS: its role in central metabolism and the in vitro synthesis of 4-hydroxy-5-methyl-3(2H)-furanone. Microbiology. 2002;148:909–22.
Delisa MP, Valdes JJ, Bentley WE. Mapping stress-induced changes in autoinducer AI-2 production in chemostat-cultivated Escherichia coli K-12. J Bacteriol. 2001;183:2918–28.
Shimada T, Yamazaki Y, Tanaka K, Ishihama A. The whole set of constitutive promoters recognized by RNA polymerase RpoD holoenzyme of Escherichia coli. PLoS ONE. 2014;9:e90447.
Huo Y-X, Tian Z-X, Rappas M, Wen J, Chen Y-C, You C-H, Zhang X, Buck M, Wang Y-P, Kolb A. Protein-induced DNA bending clarifies the architectural organization of the σ54-dependent glnAp2 promoter. Mol Microbiol. 2006;59:168–80.
Huo Y-X, Zhang Y-T, Xiao Y, Zhang X, Buck M, Kolb A, Wang Y-P. IHF-binding sites inhibit DNA loop formation and transcription initiation. Nucleic Acids Res. 2009;37:3878–86.
Weichart D, Querfurth N, Dreger M, Hengge-Aronis R. Global role for ClpP-containing proteases in stationary-phase adaptation of Escherichia coli. J Bacteriol. 2003;185:115–25.
Gong Z, Nielsen J, Zhou YJ. Engineering robustness of microbial cell factories. Biotechnol J. 2017;12:1700014.
Zhang N, Buck M. A perspective on the enhancer dependent bacterial RNA polymerase. Biomolecules. 2015;5:1012–9.
Buck M, Gallegos M-T, Studholme DJ, Guo Y, Gralla JD. The bacterial enhancer-dependent σ54 (σN) transcription factor. J Bacteriol. 2000;182:4129–36.
Santos-Zavaleta A, Salgado H, Gama-Castro S, Sánchez-Pérez M, Gómez-Romero L, Ledezma-Tejeida D, García-Sotelo JS, Alquicira-Hernández K, Muñiz-Rascado LJ, Peña-Loredo P. RegulonDB v 10.5: tackling challenges to unify classic and high throughput knowledge of gene regulation in E. coli K-12. Nucleic Acids Res. 2018;47:D212–20.
Atsumi S, Li Z, Liao JC. Acetolactate synthase from Bacillus subtilis serves as a 2-ketoisovalerate decarboxylase for isobutanol biosynthesis in Escherichia coli. Appl Environ Microbiol. 2009;75:6306.
Westerfeld WW. A colorimetric determination of blood acetoin. J Biol Chem. 1945;161:495–502.
Cohen SA, Bidlingmeyer BA, Tarvin TL. PITC derivatives in amino acid analysis. Nature. 1986;320:769–70.
Huo Y-X, Zheng B, Wang N, Yang Y, Liang X, Ma X. Identifying amino acid overproducers using rare-codon-rich markers. J Vis Exp. 2019. https://doi.org/10.3791/59331.
This work was jointly supported by the National Key R&D Program of China (Grant No. 2019YFA0906500), the National Nature Science Foundation of China (Grant No. 91951113) and the Innovation Team Project of Colleges and Universities in Jinan (Grant No. 2019GXRC033).
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Additional file 1: Table S1.
σ54-dependent promoters characterized in this study. Table S2. List of primers. Figure S1. The growth curves of the parent strain and the ammonia-assimilation-pathway-deleted strain in defined media. Figure S2. The biofuel titer of the LM19 and LM10 with biofuel synthetic pathway in M9 medium with 40 g L− 1 yeast extract as the carbon and nitrogen sources. Figure S3. The fluorescence intensities for GFP expressed from different σ54-dependent promoters. Figure S4. Activities of AlsS in the biofuel synthetic pathway driven by different promoters. Figure S5. Biofuel production from pathway driven by aglnAp2, bargTp, cPLlacO1, d J23100 and errnBp1 in batch fermentation. Figure S6. The percentage of theoretical yield (g of product per g of consumed raw material) for biofuel produced from pathway driven by different promoters. Figure S7. The regulatory cascade for nitrogen assimilation in E. coli. Figure S8. Precursors of the 20 amino acids and their value-added derivatives. Figure S9. Amino acid composition of different protein sources.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
About this article
Cite this article
Ma, L., Guo, L., Yang, Y. et al. Protein-based biorefining driven by nitrogen-responsive transcriptional machinery. Biotechnol Biofuels 13, 29 (2020). https://doi.org/10.1186/s13068-020-1667-5
- Amino acid
- Higher alcohol
- Transcription regulation
- Stationary phase