Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/wiley/engineering-escherichia-coli-to-utilize-erythritol-as-sole-carbon-MaiNXuN7Za
- Publisher
- Wiley
- Copyright
- © 2023 Wiley‐VCH GmbH
- eISSN
- 2198-3844
- DOI
- 10.1002/advs.202207008
- Publisher site
- See Article on Publisher Site
Abstract
IntroductionThe global obesity epidemic and correlative diseases like diabetes, hypertension, and hyperlipidemia encourage people to change their dietary style by limiting excessive intake of carbohydrates. Yet, many people like sweet food. In this context, food industry is seeking to produce more carbohydrate/sugar‐reduced foods by using alternative sweeteners such as sugar alcohols (also called polyols that contain multiple hydroxyl groups).[1,2] Erythritol, a C4 sugar alcohol, naturally exists in some vegetables, fruits, mushrooms, and fermented foods.[3,4] It also has been found in human tissues such as semen, lens, cerebrospinal fluid, and serum.[1,4,5] Currently, erythritol is primarily produced by fermentation using engineered lactic acid bacteria and yeast.[6–10] Because erythritol is stable in a wide range of temperature and pH conditions, is not reactive in Maillard reactions, and is a non‐hygroscopic compound,[11] it has been recognized as a stable, cheap, and safe sugar alcohol as well as a natural sweetener for food industry. Since human bodies are not able to catabolize erythritol, most erythritol is absorbed in the small intestine and then directly excreted in urine within 24 h.[5,11–14] In addition, the human gut microbiota also lacks microbes to metabolize erythritol. Therefore, erythritol by food intake cannot be utilized by human bodies as a nutrient/energy source.[15,16]So far, two distinct erythritol degradation pathways have been identified from different species of microorganisms. The first pathway was discovered in several Hyphomicrobiales species, including the environmental bacterium Ochrobactrum,[17] the pathogenic bacterium Brucella,[18–23] and the nitrogen‐fixing plant endosymbionts Rhizobium[22] and Sinorhizobium.[24] These microorganisms usually uptake erythritol as an initial nutrient and yield erythrose‐4‐phosphate via a five‐step catalysis.[17] The second pathway was characterized in Mycolicibacterium smegmatis.[25] This pathway contains a short three‐step catalysis and the final product is also erythrose‐4‐phosphate. In both erythritol degradation pathways, erythrose‐4‐phosphate can serve as an intermediate product of the pentose‐phosphate pathway, entering into other primary metabolic pathways (e.g., glycolysis) as a C4 carbon source. Without such degradation pathways, most other microorganisms are not able to metabolize erythritol. To the best of our knowledge, no attempts have been made reprogramming a model chassis microbe to live solely on erythritol as a carbon source. However, this is important because erythritol is a safe compound not used by human metabolism, but it can work as a unique molecule to support the growth of engineered microbes/probiotics, which then will be potentially used as living therapeutics for disease treatment in human gut microbiota or other focal tissues (e.g., tumors).In metabolic engineering and synthetic biology, Escherichia coli (E. coli) has been engineered for a broad range of applications with a main focus on the production of valuable products.[26–28] However, the cultivation of E. coli in laboratory is predominantly based on easy‐to‐use carbon sources, especially, glucose (C6). Of note, a few studies recently tried to engineer E. coli to use C1‐based carbon sources including CO2[29,30] and methanol.[31,32] Yet, the use of C4‐based carbon sources (e.g., erythritol) for E. coli cultivation has not been reported so far.To address this opportunity, here we report two engineered E. coli strains (MG1655[33] and Nissle 1917[34]) that can utilize erythritol as a carbon nutrient. The gene cluster of a five‐step erythritol metabolic pathway is obtained from Ochrobactrum spp., which is isolated from the outdoor aerosol. To use the metabolic pathway, we initially characterize the erythritol‐responding genetic repressor eryD and its DNA‐binding site. Then, we perform mRNA transcriptional analysis to evaluate the metabolism of erythritol in E. coli MG1655. On the basis of the transcriptome data, we design genetic circuits to facilitate erythritol catabolism and significantly increase the growth of E. coli in a modified M9 medium (glucose is replaced with erythritol). Using the engineered E. coli MG1655, soft drinks that contain erythritol (as a sugar‐free sweetener) can be distinguished from those without erythritol according to the cell growth. Moreover, the probiotic E. coli Nissle 1917 (EcN) is also reprogrammed with the erythritol metabolic pathway, allowing EcN to grow in simulated intestinal fluid (SIF) containing erythritol as sole carbon source. Taken together, our study shows the successful transformation of a natural metabolism of erythritol into surrogate E. coli hosts for defined applications. Importantly, we fill the gap that E. coli can utilize not only C6 (e.g., glucose) and C1‐based carbon sources (CO2 and methanol), but also C4‐based erythritol in this work. Looking forward, we anticipate that the erythritol catabolic E. coli strains will provide new opportunities for compelling research in different fields, for example, carbon cycle, synthetic biology, metabolic engineering, biomedical engineering, and living therapeutics. Moreover, with the newly characterized erythritol‐responding repressor‐operator system, it has the potential to be used as a genetic circuit part installed in different other microbes/cells to specifically respond erythritol. In this context, the whole erythritol‐responding system can regulate cell's metabolisms (native or artificial pathways) to produce valuable products such as chemicals for industrial application or therapeutics for clinic treatment.ResultsEngineering Escherichia coli to Utilize ErythritolTo obtain the gene cluster for erythritol metabolism, we initially designed an erythritol‐based selective medium derived from the M9 minimal medium, in which glucose is completely replaced with erythritol (0.4%), to isolate microbes from the outdoor aerosol that can grow in the modified medium. By doing this, we successfully separated, identified, and characterized an environmental bacterium Ochrobactrum spp. from the aerosol (Figures S1 and S2, Supporting Information).[35] Then, ten erythritol catabolism‐associated genes were polymerase chain reaction (PCR) amplified from the genome of Ochrobactrum spp. followed by sequencing. Both DNA sequence and protein alignments suggested that all ten genes/proteins were almost the same as those of the reported Brucella abortus 2308 strain[36,37] (Figures S3–S5, Supporting Information), including five erythritol catabolic genes (eryA, eryB, eryC, eryH, and eryI), two putative erythritol‐binding transcriptional repressor (eryD and eryR), and three erythritol ABC‐transporter genes (eryE, eryF, and eryG) (Figure 1a). Catalytic functions of the five‐step erythritol catabolism proteins have been demonstrated previously in vitro.[17] At the end of the pathway, erythritol is eventually converted to D‐erythrose 4‐phosphate, which serves as one intermediate in the pentose‐phosphate pathway (Figure 1b). This C4 phosphate‐substrate will then enter into carbohydrate utilization via glycolysis, nucleotide synthesis, and the shikimate pathway for aromatic amino acids biosynthesis (Figure 1b).1FigureReconstitution of the erythritol metabolic pathway in E. coli to support cell growth on erythritol as sole carbon source. a) The erythritol metabolism gene cluster isolated from Ochrobactrum spp. and reconstituted in E. coli to utilize erythritol. b) Catalytic pathway for the conversion of erythritol to D‐erythrose 4‐phosphate, which then enters into the downstream metabolic pathways. c) The growth of E. coli MG1655 harboring plasmid pFB147 in liquid M9‐erythritol medium. Each OD600 value (mean ± s.d.) was measured with three biological replicates. d) Cell growth curve and the profile of erythritol utilization in M9‐erythritol medium. Erythritol concentrations were measured by HPLC with three biological replicates. The error bars represent the standard deviation (s.d.).Next, we aimed to reconstitute the gene cluster of erythritol metabolism (i.e., eryA, eryB, eryC, eryH, and eryI) in E. coli to see if cells can grow on erythritol as sole carbon source. To this end, we constructed a plasmid containing the above five genes and transformed it into E. coli MG1655. Then, the cells were cultivated in liquid M9‐erythritol medium with shaking for 72 h (Figure 1c). Clearly, E. coli cells were able to grow by using erythritol and the final OD600 value reached ≈0.35. To determine the profile of erythritol consumption, filtered medium supernatants were analyzed by HPLC. The results indicated that along with cell growth erythritol was gradually consumed by E. coli with 0.28% erythritol left in the medium (Figure 1d).Since the gene cluster worked to support E. coli growth, we were curious about the expression level of each protein in vivo as well as the strength of their regulatory parts (e.g., promoter and ribosomal binding site [RBS]). First, all in vivo expressed proteins except two membrane proteins (eryE and eryG) were analyzed by SDS‐PAGE and western‐blot (Figure S6, Supporting Information). The results showed that all proteins could be well expressed with a high percentage of soluble fractions. Second, the strength of the native promoters and RBSs of these genes (except eryR) were measured and compared with the iGEM standard regulatory parts (Figures S7 and S8, Supporting Information).[38–40] Overall, their strength was low as compared to the iGEM standard elements. In addition, three erythritol ABC‐transporter proteins (eryE, eryF, and eryG) were predicted and, in particular, the expression, transmembrane phenotype, and localization in E. coli cells of the two membrane proteins (eryE and eryG) were demonstrated (Figures S9 and S10, Supporting Information). However, coexpression of the three transporter proteins did not further enhance cell growth (Figure S11, Supporting Information). By contrast, E. coli MG1655 could transport erythritol without this ABC‐transporter system (Figure 1a,c), suggesting that E. coli’s native transporter systems can also be responsible for erythritol transportation.eryD Is an Erythritol‐Binding Transcriptional RepressorSangari et al. reported that erythritol might act as an inducer and probably bind to eryD in Brucella abortus 2308 to inhibit the repressor activity of eryD.[19] However, a detailed biochemical characterization of eryD is still lacking. Interestingly, our isolated erythritol cluster of Ochrobactrum spp. contains an eryD gene, whose DNA sequence is totally aligned with eryD in Brucella abortus 2308 (Figure S4, Supporting Information). In addition, DNA sequence analysis indicated that eryR, another putative DNA‐binding transcriptional repressor, might also act in a similar way as eryD. Hence, characterization of these two proteins will help understand their regulation mechanism and expand erythritol‐associated synthetic biology applications. To this end, we sought to determine the putative DNA‐binding sites within two regions. One is the 732 bp gap between eryE and eryA and the other region is ≈400 bp gap between eryD and eryR (Figure S3, Supporting Information). Here, we designed an in vivo characterization strategy to precisely identify the DNA‐binding sites. In principle, the repressor's DNA‐binding site was supposed to be located after a promoter. We individually evaluated different putative DNA sequences by using a reporter plasmid, which is derived from the iGEM standard backbone of pSB1C3 containing the reporter sfGFP. First, both directions of the 732 bp gap between eryE and eryA were presumed as two promoters called pEryF‐732 (F, forward) and pEryR‐732 (R, reverse), respectively (Figures S12 and S13, Supporting Information). Meanwhile, several compatible plasmids were constructed to express eryD or eryR using a gradient strength of constitutive promoters (Figure S14, Supporting Information). After coexpression of eryD or eryR with sfGFP, the results indicated that only eryD had a DNA‐binding site located in pEryF‐732 and the eryD expression level could regulate the sfGFP expression level, whereas eryR did not show an inhibition of sfGFP expression (Figure S14, Supporting Information). With this pEryF‐732 region, we then performed truncation experiments to precisely identify both 5′‐ and 3′‐site of eryD DNA‐binding site (we call it as an erythritol operator, eryO) (Figures S15–S20, Supporting Information). Finally, eryO was successfully identified as a 29 bp sequence located between the −35 and −10 regions of the promoter pEryF‐732 (Figure 2a). Moreover, we carried out in vitro electrophoretic gel mobility shift assay (EMSA) by incubation of the FAM‐labeled DNA fragment (pEryF, 732 bp) with different amounts of purified eryD. The EMSA results suggested that eryD was able to effectively bind to the pEryF‐732 promoter (Figure 2b) and a further investigation showed that eryD could precisely bind to the 29 bp region of eryO as well (Figure S21, Supporting Information). Using the same truncation strategy, however, we did not observe any DNA‐binding sites for eryD or eryR in the 400 bp region between eryD and eryR (Figure S22, Supporting Information).2FigureCharacterization of the transcriptional repressor eryD. a) DNA sequence of eryD binding site (erythritol operator, eryO, orange highlighted), promoter pEryF‐732 sequence (−35, −10, and +1 regions were annotated as red, blue, and underlined, respectively), and their genetic location in the erythritol catabolism cluster. b) EMSA analysis of the interaction between eryD and pEryF‐732 DNA fragment (5′ FAM‐labeled). c) Schematic shows two compatible plasmids pFB261 and pFB186 co‐existed in E. coli Mach1‐T1. sfGFP fluorescence indicates the transcription strength of pEryF‐732, which is repressed by eryD. d) Fluorescent image of E. coli Mach1‐T1 cell pellets harboring two plasmids in (c). Pellets were collected from 1.5 mL overnight LB culture (37 °C and 250 rpm for 16 h) with supplemented erythritol. The image is shown as one representative result from three biological replicates. e) Western‐blot analysis of sfGFP expression (with C‐terminal 6× His tag, 27.6 kDa) in (d). f) sfGFP fluorescence of E. coli Mach1‐T1 in (d). The samples were measured as normalized fluorescence with three biological replicates. g,h) Schematic diagram of used strains and plasmids for pT7‐eryO operon characterization. pFB276 and pFB261 co‐existed in E. coli BL21(DE3) and cell pellets were collected from 1.5 mL overnight LB culture (37 °C and 250 rpm for 16 h) with supplemented erythritol. The pellets were measured as normalized fluorescence with three biological replicates. i,j) Schematic diagram of used strains and plasmids for J231xx‐eryO operons characterization. pFB261 and pFB270–pFB275 co‐existed in E. coli Mach1‐T1 and cell pellets were collected from 1.5 mL overnight LB culture (37 °C and 250 rpm for 16 h) with supplemented erythritol. The pellets were measured as normalized fluorescence with three biological replicates. k,l) eryD‐erythritol orthogonality test was performed by the same strains and procedure as described in (c). sfGFP expression was analyzed by normalized fluorescence fold change and western‐blot. All the error bars represent the standard deviation (s.d.).Next, we aimed to further characterize eryD. First, structure prediction by both AlphaFold and Robetta indicated that eryD monomer has an N‐terminal helix‐turn‐helix motif responsible for DNA interaction and the C‐terminal domain for erythritol regulation. SWISS‐MODEL prediction showed that eryD is assembled as a homotetramer (Figure S23, Supporting Information), which is also confirmed by size exclusion chromatography (SEC) and a native‐PAGE gel analysis (the calculated molecular weight of eryD homotetramer is 138 kDa; Figure S24, Supporting Information). Then, we evaluated the effect of erythritol concentration on the regulation of eryD and the resulting reporter gene (sfGFP) expression (Figure 2c). The data indicated that eryD could be dynamically regulated by a gradient concentration of erythritol (Figure 2d–f and Figure S25, Supporting Information). To develop eryD as a potential regulatory element for genetic circuit design in synthetic biology, we further constructed two erythritol induction systems by using the T7 RNA polymerase‐based promoter (Figure 2g) and the E. coli σ70 factor‐based native promoters (Figure 2i), respectively. Briefly, the eryD‐binding site (29 bp) was inserted between each promoter and the RBS to see if the transcription and translation of the downstream sfGFP gene is regulated by eryD and erythritol (Figure 2g,i). Overall, their regulatory effect on sfGFP expression was observed with different constructs, albeit the expression levels were various (Figure 2h,j and Figure S26, Supporting Information). For instance, when strong promoters (e.g., T7 and J23100) were used in front of the eryD‐binding site, the expression level of the downstream gene (sfGFP) was notably higher than the use of weak promoters. In addition, if a weak promoter was used, the expression of sfGFP was relatively low and not obviously impacted by the concentration of erythritol, which is added to regulate eryD (Figure 2j). Finally, the orthogonality of erythritol–eryD interaction was investigated with a series of carbohydrates and polyols. We found that eryD showed the highest orthogonality with erythritol, whereas the interactions of eryD with other tested compounds were very low (Figure 2k,l).mRNA Transcriptional Analysis of the Erythritol Catabolism in Engineered Escherichia coliTo determine the erythritol catabolism in E. coli, we performed mRNA transcriptional analysis by comparing the exponential growth phase cultures of E. coli MG1655 between M9‐erythritol medium and M9‐glucose medium (Figure S27, Supporting Information). In total, 1298 genes were significantly different between these two cultivations, including 759 genes upregulated and 539 genes downregulated (Figure S28, Supporting Information). COG annotations showed that the significant differences were found in genes related to the metabolisms of carbohydrates, amino acids, energy, and inorganic ions, as well as the transcription and ribosome‐associated translation (Figure S29a, Supporting Information). In particular, most of the genes related to carbohydrate metabolism were overexpressed. Of special note, nearly all ribosome‐associated genes were downregulated, which indicates that the engineered strain decreased its overall translation level to adapt to a low‐carbon‐source environment. KEGG and GO annotations showed a similar trend/situation of gene expression as the COG analysis (Figure S29b,c, Supporting Information).Using transcriptomics data, we reconstructed the carbohydrate metabolism by connecting the heterologous erythritol degradation pathway to the native pentose‐phosphate pathway, glycolysis, and TCA cycle (Figure 3). In the hybrid metabolic pathways, erythrose‐4‐phosphate, the end product of erythritol catabolism, enters into pentose‐phosphate pathway and glycolysis through the catalysis of two sets of enzymes, which are transaldolase (talA and talB) and transketolase (tktA and tktB). Transcriptome analysis showed that talA, talB, tktA, and tktB were overexpressed by 5.2, 2.0, 1.3, and 5.9 times, respectively, compared to those when cells grew in glucose‐based M9 medium. In addition, several other carbohydrate degradation pathways were also upregulated, including d‐xylose, l‐arabinose, lactose, l‐rhamnose, and glycerol. By contrast, only galactose and d‐mannose degradation pathways were downregulated. In general, the gene expression levels in glycolysis and TCA cycle were not significantly impacted in the two cultivation conditions.3FigureReconstruction of the major carbohydrate metabolic pathways mapped with mRNA transcriptional analysis. Reconstructed major carbohydrate metabolic pathways of erythritol catabolic E. coli MG1655 (harboring pFB147) were based on the relative differential gene expression profiles between two cultures (M9‐erythritol as the experimental group and M9‐glucose as the control group). The genes with a significant (p < 0.05) differential expression of ≥1 log2 (fold change) are indicated with arrows (pathway) in green (upregulated in M9‐erythritol medium) or red (downregulated in M9‐erythritol medium). The other genes without significant differential expression are indicated with black arrows (pathway). The details of mRNA transcriptional analysis are shown in the Supporting Information.By screening the gene expression level, some notably upregulated gene clusters were summarized (Figure S30, Supporting Information). For example, the genes of glycolate utilization operon glcCDEFGBA were all significantly overexpressed (Figure S30b, Supporting Information). However, no glycolate or analogs were in the M9‐erythritol medium. We thus considered whether there is a transcriptional repressor in the glc cluster that might respond to erythritol catabolite(s). Previous work has reported that glcC is a DNA‐binding transcriptional repressor that can respond to glycolate and acetate.[41] Here, our experimental data suggested that glcC could also respond to erythrose‐4‐phosphate or its downstream metabolite(s) (Figure S31, Supporting Information). In addition, another gene cluster (i.e., phn operon[42]) involved in the phosphonate uptake and utilization was nearly all upregulated as well (Figure S30e, Supporting Information). However, we observed that the transcriptional repressor phnP did not respond to erythritol catabolism in our experiments (data not shown). This is likely due to the requirement of more phosphonate to support erythritol phosphorylation, leading to overexpression of the phn operon.Metabolic Engineering to Improve Cell Growth Based on Erythritol MetabolismWhile we have demonstrated that engineered E. coli can use erythritol as sole carbon source, the utilization efficiency of erythritol was low (only 30% of added erythritol was consumed) and cells could not grow to a high density (OD600 reached 0.35 in 72 h) (Figure 1d). Next, we set out to increase cell growth relying on erythritol by metabolic engineering. To begin, we compared the effect of five constitutive promoters with a gradient strength and the native promoter (pEryF‐732) on expression of the gene cluster and cell growth (Figure 4a). We found that two medium strength promoters (J23106 and J23105) and pEryF‐732 performed similar and the final OD600 values were comparable (Figure 4b,c). Yet, the strongest promoter J23100 just could support cell growth to a medium level, which is lower than pEryF‐732. Using the weakest promoter J23109, no cells could grow due to the low expression of the genes (Figure 4b,c). Overall, our results suggested that a suitable expression level of the gene cluster was sufficient for erythritol catabolism and cell growth. Since the tested promoters did not show a significant increase on cell growth, we finally chose the native promoter (pEryF‐732) to express the genes in our following experiments.4FigureImprovement of erythritol catabolism in E. coli by metabolic engineering. a) Schematic plasmids of pFB148–pFB152. The native promoter pEryF‐732 was replaced by a series of constitutive promoters with a gradient strength for erythritol metabolism optimization. b) E. coli MG1655 harboring the above each plasmid grew on M9‐erythritol agar plates after 72 h incubation at 37 °C. c) Growth curves of E. coli MG1655 harboring the above each plasmid in M9‐erythritol liquid medium (5 mL, 37 °C, 250 rpm). d) Schematic shows E. coli MG1655 harboring the erythritol degradation plasmid pFB147 and compatible plasmids (pFB293–pFB297) for the coexpression of talA+tktB. e) Cell growth curves of (d). f) Schematic shows E. coli MG1655 harboring the erythritol degradation plasmid pFB147 and compatible plasmids (pFB298–pFB301) for the coexpression of talB. g) Cell growth curves of (f). h) Schematic shows E. coli MG1655 harboring the erythritol degradation plasmid pFB147 and compatible plasmids (pFB302–pFB305) for the coexpression of tktA. i) Cell growth curves of (h). All the measurements were performed with three biological replicates. The error bars represent the standard deviation (s.d.).On the basis of the transcriptome analysis, the expression of transaldolase (talA/talB) and transketolase (tktA/tktB) were upregulated to convert erythrose‐4‐phosphate into other metabolic pathways. We then hypothesized that overexpression of these genes might help increase the consumption of erythrose‐4‐phosphate to accelerate carbon flux and cell growth. To verify this, we constructed three compatible plasmids that were able to constitutively express talA+tktB, talB, and tktA, respectively (Figure 4d,f,h). Note that the two genes of talA and tktB locate next to each other in the genome, thus we only constructed one plasmid to coexpress talA and tktB rather than two separate plasmids. By coexpression of the erythritol gene cluster and each of the above constructed plasmids, we obtained the profiles of cell growth after 108 h cultivation (Figure 4e,g,i). In general, overexpression of talA+tktB, talB, or tktA could obviously increase cell growth with the final highest OD600 value of ≈1.0 in each group, which is nearly threefold higher than that of the cultivation without overexpression (see Figure 4c for the OD600 of 0.35).Characterization of the Erythritol Transport System in Escherichia coliHaving demonstrated the utilization of erythritol in the engineered E. coli MG1655, we were curious which transport system(s) are responsible for erythritol transportation without expression of the erythritol ABC‐transporter system (i.e., eryE, eryF, and eryG[22]). Since eryE and eryG are membrane proteins, we first demonstrated their transmembrane phenotype in E. coli Mach1‐T1 cells (Figure S10, Supporting Information). However, when the three proteins were coexpressed in different E. coli strains (MG1655, Mach1‐T1, and Nissle 1917), we found that this ABC‐transporter could change cell's morphology from normal rods to elongated fibers (Figure S32, Supporting Information). Further experiments by individual expression of the three proteins indicated that eryF was the key component that caused the change of cell morphology (Figure S33, Supporting Information). On the other hand, coexpression of the transporter did not enhance cell growth and density (Figure S11, Supporting Information). Therefore, we then only focused on the potential native genes in E. coli MG1655 that assist erythritol transportation.First, we searched homologous genes of the erythritol ABC‐transporter, including all carbohydrate or polyol transporter genes in E. coli MG1655 (Figure S34, Supporting Information). After further protein–protein alignments by BLAST,[43,44] seven permease genes that are homologous to eryG were screened out (Figure S34c, Supporting Information), including the glycerol facilitator glpF, which has been reported to enable E. coli to transport erythritol.[45] Then, we individually knocked out the seven genes (Figure S34d, Supporting Information) and expressed the erythritol metabolic genes in each of the seven knock‐out strains. The results indicated that six transporter systems might facilitate erythritol transportation to support cell growth (Figure S35, Supporting Information). In addition, we found 34 putative carbohydrate transport genes upregulated according to the transcriptome data, constructed 34 knock‐out strains, and performed the same cultivation experiments. However, cell growth was not obviously impacted in each of the 34 knock‐out strains, suggesting that these putative systems are not responsible for erythritol transportation (Figure S36, Supporting Information). Taken together, while our preliminary data indicate that some endogenous carbohydrate transporter systems in E. coli might be able to transport erythritol, more characterization experiments should be performed to get more insights in the future study.Escherichia coli as a Living Detector to Distinguish Soda DrinksIn many soft drinks, some artificial and/or natural sweeteners are used as additives to reduce or replace the use of high‐fructose corn syrup (including fructose and glucose) and sugar (sucrose). For instance, erythritol has been used as a natural sweetener in soda and other foods. Detection of such compound in food often relies on expensive instruments such as HPLC and GC‐MS.[46] With the above engineered E. coli in hand, here we aim to use this strain as a living detector to detect erythritol from soda drinks. In principle, wild‐type E. coli strains can utilize fructose and glucose as carbon sources, but not erythritol (Figure S37, Supporting Information). Hence, the erythritol detection process is easy according to whether the cells can grow or not in the soda drinks‐modified M9 medium. As shown in Figure 5a, the E. coli strain (erythritol detector, ErD) containing the erythritol metabolic pathway is used for the detection. Meanwhile, a negative control (NC) strain is used for comparison, which harbors the same pathway but under the control of a weak promoter (J23109) that is not able to support cell growth on erythritol (Figure 4c).5FigureEngineered E. coli as a living detector to distinguish soda drinks. a) Schematic of two E. coli MG1655 strains shows the negative control (NC, with pFB151 that cannot utilize erythritol) and the erythritol detector (ErD, with pFB147 that can utilize erythritol). b,c) Liquid incubations were performed to distinguish synthetic M9‐based carbohydrate media and soda drinks. d,e) Solid incubations were performed for distinguishing synthetic M9‐based carbohydrate media (the same as [b]) and soda drinks (the same as [c]). All samples were performed with three biological replicates showing similar results.To prove the concept, we initially cultivated NC and ErD strains in M9 and M9‐modified media, in which glucose (0.4%) is completely replaced with the same amount of ddH2O, fructose, or erythritol. In both groups (Figure 5b), while NC and ErD cells were able to grow in fructose and glucose‐based M9 media, no cell growth was observed without carbon source (ddH2O). Clearly, only ErD cells could grow in the erythritol‐based cultivation, suggesting that our concept is feasible for detecting erythritol in soda drinks. Next, we obtained three commercial soda products (soda 1 is sugar‐free and contains two artificial sweeteners: acesulfame potassium and sucralose; soda 2 contains fructose and glucose; soda 3 is sugar‐free and contains erythritol and sucralose). Similarly, we modified the M9 medium by adding 10× diluted soda liquid to replace glucose. After cultivation for 36 h, we observed that the growth of ErD cells completely depended on the carbon sources from the selected soda drinks (Figure 5c). There was no cell growth in soda 1 without any carbon source; the growth in soda 2 (containing fructose and glucose) was the fastest (the liquid became turbid in 12 h); by contrast, ErD could grow in soda 3 that contains erythritol but with a slow growth rate (36 h). As a result, the cell growth profile could basically tell if the soda drinks contain erythritol or not by using our ErD strain. Besides, we also cultivated cells on solid agar plates and obtained similar results (Figure 5d,e). Of note, ErD cells/colonies became visible on the plates after 24 h incubation, a bit faster than the cultivation in liquid medium (36 h). Taken together, our living detector for erythritol is a potential method to distinguish erythritol‐containing soda drinks or other foods with several properties of rapid operation, cheap detection, and easy visualization.Erythritol Facilitates the Growth of Engineered Escherichia coli Nissle 1917 in Simulated Intestinal FluidNext, we aimed to bring the erythritol metabolic pathway to other E. coli strains for potential applications. In particular, we chose EcN, which is a probiotic (generally recognized as safe)[34] and has been widely engineered and applied for diagnosis and therapy.[47–50] After transferring the erythritol gene cluster into EcN, we then sought to evaluate if the cells can grow in the SIF supplemented with erythritol. The incubation was carried out at 37 °C for 4 days and at different time points the liquid cultures were spread to lysogeny broth (LB)‐agar plates for cell growth to count colony forming units (CFU) (Figure 6a). First, we mixed SIF solution (0.4% erythritol) with different inoculated cells (102, 104, 106, and 108) and evaluated their growth. Overall, cells could grow by utilizing erythritol and the CFUs reached the highest after 2‐day cultivation with the inoculation of 102, 104, and 106 cells (Figure 6b and Figure S38a, Supporting Information). Yet, the highest inoculation of 108 cells led to a rapid increase of CFU in just 2‐h cultivation, followed by a significant reduction of CFU within 1 day, which is likely due to the rapid exhaustion of nutrients by over‐inoculated cells. Since the growth profiles between 104 and 106 were similar (Figure S38a, Supporting Information) and their highest cell densities were comparable after 2‐day cultivation (Figure 6b), we then decided to use a lower density of 104 EcN cells to investigate the effect of erythritol concentration (0%, 0.01%, 0.1%, 1%, 4%, and 16%) on cell growth. During the 4‐day cultivation, CFU values in each group were counted at different time points (Figure S38b, Supporting Information). Then, we specifically compared cell growth after 1‐day cultivation because cells in most groups reached the highest numbers. As shown in Figure 6c, lower erythritol concentrations (0.01–4%) in general were beneficial to facilitate the growth of engineered EcN cells compared to the wild‐type cells. The growth of cells with the highest erythritol concentration (16%) was not detected at day 1 probably due to a high osmosis pressure in the culture. Unexpectedly, SIF without erythritol supplementation (0%) could also support cell growth (Figure 6c and Figure S38b, Supporting Information). Previous studies have shown that erythritol might exist in the animal bile flow.[51] We thus hypothesized that the bile salts we used to prepare SIF solutions might also contain residual erythritol. To test this hypothesis, we performed HPLC analysis and indeed detected erythritol in the bile salts, leading to the final concentration of the background erythritol in SIF at about 0.023%. As a result, we performed more control experiments by cultivating cells in modified SIF solutions without bile salts. The results showed that the growth profiles between engineered and wild‐type EcN cells were similar without significant difference and cell numbers in both cultivations were notably low (Figure 6c and Figure S38b, Supporting Information). This also suggests that erythritol is beneficial to support the growth of engineered EcN. Taken together, engineered EcN cells are able to grow in SIF and supplemented erythritol at low concentrations can help significantly increase cell density in this study. Yet, the inoculated cells and erythritol concentration are two factors subjected to be optimized in future potential applications.6FigureErythritol facilitates the growth of engineered probiotic E. coli Nissle 1917 in SIF. a) Schematic workflow of sample preparation, incubation, and measurement. The mixtures were incubated at 37 °C without shaking. b) Effect of inoculation on cell growth. The SIF solution was inoculated with 102, 104, 106, and 108 cells, respectively. Additional erythritol was added as 0.4% in SIF. The CFU/mL values were calculated after 2‐day cultivation. ND, not detected. Each value (mean ± s.d.) is calculated with three biological replicates and the error bar represents the standard deviation (s.d.). Student's t‐tests are used for statistical analysis, and p <0.05 indicates statistical significance (**p < 0.01; ns, p > 0.05). c) Effect of additional erythritol on cell growth. Note that the concentration of the background erythritol derived from bile salts in SIF was about 0.023%. The initial inoculated cells were the same with 104 in each group. The CFU/mL values were calculated after 1‐day cultivation. ND, not detected. Each value (mean ± s.d.) is calculated with three biological replicates and the error bar represents the standard deviation (s.d.). Student's t‐tests are used for statistical analysis, and p <0.05 indicates statistical significance (*p < 0.05, **p < 0.01, ***p < 0.001; ns, p > 0.05).DiscussionErythritol, a natural sweetener found/used in foods, cannot be metabolized by human bodies or the microbes in human gut,[13,15] which is excreted by human as a carbon waste. Without an erythritol metabolism, most other microorganisms are not able to utilize erythritol either. In this study, we engineer E. coli strains by reconstitution of the erythritol metabolic pathway in a special effort to expand the carbon source scope that can be utilized by the E. coli chassis. The erythritol gene cluster(s) have been identified from several microbes.[17,19,23,25] While in vitro characterization of the catalytic enzymes (eryA, eryB, eryC, eryH, and eryI) has been reported previously,[17] other regulatory proteins (eryD and eryR) and transporter proteins (eryE, eryF, and eryG) have not been investigated in detail with experimental evidence to our knowledge. Thus, we particularly investigate these proteins in this work in order to well employ the whole gene cluster for erythritol metabolism in a heterologous host (E. coli). A major finding of our study is the identification of the so far unknown DNA‐binding site (29 bp sequence) of the transcriptional repressor eryD (Figure 2a). This pair of eryD and DNA sequence (29 bp) can be further developed as an erythritol‐responding repressor‐operator system for metabolic engineering and synthetic biology applications. However, recombinant expression of the ABC‐transporter proteins in E. coli makes cell morphology changed (Figure S32, Supporting Information). But interestingly, E. coli strains are able to grow in pure erythritol‐based M9 medium without coexpression of the erythritol transporter system, which suggests that E. coli’s native transporter(s) can work for the transportation. To figure out the potential transporter(s) in E. coli, we construct a total of 41 knock‐out strains, yet only find six transporters showing positive effect on cell growth (Figures S35 and S36, Supporting Information).Importantly, different types of E. coli strains can be engineered to grow on erythritol as sole carbon source. From the engineered cells, we find that the expression of more than 750 genes is upregulated and most of the genes are related to carbohydrate transport and metabolism (Figure S29a, Supporting Information). According to these overexpressed genes, we particularly select transaldolase (talA/talB) and transketolase (tktA/tktB), which can convert erythrose‐4‐phosphate into other pathways, for overexpression, leading to a significant improvement of cell growth (Figure 4). This suggests that metabolic engineering strategies can further increase the efficiency of erythritol utilization and metabolism. With these engineered E. coli strains, we apply them to potential applications. First, E. coli MG1655 is used as a living ErD for distinguishing soda drinks whether they contain erythritol or not. This method is cheap and easily visible compared to the instrument‐based detection. Second, the probiotic EcN with erythritol metabolism is able to grow in the SIF supplemented with erythritol. This is promising and exciting because EcN might be developed as living therapeutics, which can respond to the human‐safety compound erythritol, to help regulate/cure human gut microbiota‐related diseases or even other focal tissues like tumors. While here two applications are showcased, we expect erythritol and its associated gene cluster to be used in more fields such as high cell‐density cultivation (Figure S39, Supporting Information) and recombinant protein expression (Figure S40, Supporting Information), which are just two straightforward examples of using erythritol as carbon source.Overall, we reconstitute the natural erythritol metabolic pathway in different E. coli hosts, which then show the capability to grow by using erythritol as sole carbon source. Our work has several key features. First, the scope of carbon nutrients that can be utilized by E. coli is expanded to C4‐based erythritol. This will provide a new strategy for strain engineering to produce high‐value products, because in engineered cells erythritol can be directly converted to D‐erythrose 4‐phosphate via the erythritol degradation pathway, which then serves as an important precursor for the synthesis of downstream products such as aromatic amino acids and pharmaceutical compounds.[52,53] Second, the DNA‐binding site of the transcriptional repressor eryD has been identified in this work (Figure 2a). Thus, erythritol can be developed as a regulator/inducer to regulate gene expression, providing a new repressor‐operator system for metabolic engineering and synthetic biology applications. Third, erythritol is a human‐safe compound with the potential for clinic applications. For example, erythritol supports the growth of engineered E. coli or other microbes/probiotics, which can be used as living therapeutics for the treatment of human diseases.[54–56] By contrast, E. coli has also been engineered to utilize C1‐based carbon sources like methanol,[31,32] however, such microbe cannot be used in human bodies due to the toxicity of methanol. Taken together, our work demonstrates the success of installing the erythritol metabolic pathway to E. coli, allowing cells to respond and use erythritol as a carbon nutrient. Looking forward, this success might be expanded to other heterologous hosts for more applications, if an erythritol‐responding and/or utilizing system is needed, for instance in the areas of metabolic engineering, synthetic biology, and biomedical engineering.Experimental SectionStrains, Vectors, Plasmids, Primers, and ReagentsThe details of E. coli strains, vectors, plasmids, and primers used in this study are listed in Tables S1 and S4–S6, Supporting Information. Some vectors and plasmids were derived from the authors’ previous work.[57] The full sequences of all plasmids are listed in Supporting Information (Excel Sheet Data) and their correctness was verified by Sanger sequencing (GENEWIZ) unless otherwise noted. Q5 High‐Fidelity DNA Polymerase (New England Biolabs), Phanta Super‐Fidelity DNA Polymerase (Vazyme), FastPure Gel DNA Extraction Mini Kit (Vazyme), and ClonExpress Ultra One Step Cloning Kit (Vazyme) were used for molecular cloning. DreamTaq Green PCR Master Mix (Thermo Scientific) was used for colony PCR. LB liquid medium contained 10 g tryptone, 5 g yeast extract, and 10 g sodium chloride in 1 L ddH2O. LB‐agar plates were prepared by adding 15 g L−1 LB. M9 minimal medium (also M9‐glucose medium) contains 200 mL of 5× M9 salts (15 g L−1 KH2PO4, 2.5 g L−1 NaCl, 33.9 g L−1 Na2HPO4, and 5 g L−1 NH4Cl), 20 mL of glucose (20% w/v), 2 mL of 1 m MgSO4, and 100 µL of 1 m CaCl2 in 1 L ddH2O. M9‐erythritol medium contained 200 mL of 5× M9 salts, 20 mL of erythritol (20% w/v), 2 mL of 1 m MgSO4, and 100 µL of 1 m CaCl2 in 1 L ddH2O. Antibiotic stocks (1000×) were 100 mg mL−1 ampicillin, 50 mg mL−1 kanamycin, and 34 mg mL−1 chloramphenicol.Genetic Parts and Gene ClustersAll genetic parts (promoter, RBS, promoter‐RBS, terminator, CDS, and protein binding site) and gene clusters used in this study are listed in Tables S2 and S3, Supporting Information. Genetic parts shorter than 100 bp were synthesized within the oligonucleotides (GENEWIZ) and inserted into PCR fragments during molecular cloning. Gene clusters were amplified from the isolated Ochrobactrum spp. genome and the referenced genome for DNA‐DNA alignments was Brucella anthropi strain T16R‐87 chromosome 2 (GenBank: CP044971.1). Gene clusters were checked for correctness with Sanger sequencing (GENEWIZ), which then were used as gene templates for PCR amplification.Plasmid ConstructionAll plasmids were constructed by Gibson Assembly. In brief, all linear PCR products (Q5 high‐fidelity DNA polymerase or Phanta super‐fidelity DNA polymerase) were extracted by gel extraction and then assembled by using ClonExpress ultra one step cloning kit (Vazyme). After assembly, the reaction mixture was added to 50 µL competent E. coli Mach1‐T1 cells for transformation, followed by incubation overnight on LB‐agar plate. Then, DreamTaq green PCR master mix (Thermo Scientific) was used for colony PCR. The PCR products were sequenced by GENEWIZ.Western‐Blot AnalysisE. coli pellets were collected in 1.5 mL centrifuge tubes at 5000 × g and 4 °C for 10 min. The supernatant was discarded and the pellets were resuspended with 1 mL 1× phosphate‐buffered saline (pH 7.4) and lysed by sonication (Q125 sonicator, Qsonica, 10 s on/off, 50% of amplitude, input energy ≈600 joules). The lysate was then centrifuged at 12 000 × g and 4 °C for 10 min. The total (cell lysate), soluble (supernatant), and pellet fractions were separated by SDS‐PAGE or Native‐PAGE (Omni‐Easy one‐step PAGE gel fast preparation kit, EpiZyme), followed by wet transferring to PVDF membrane (Bio‐Rad) with 1× transfer buffer (25 mm tris‐HCl, 192 mm glycine, and 20% v/v methanol in 1 L ddH2O, pH 8.3). Then, the PVDF membrane was blocked (protein free rapid blocking buffer, Epizyme) for 1 h at room temperature. After washing thrice with TBST for each 5 min, 1:10 000 (TBST buffer‐based) diluted His‐Tag Mouse monoclonal antibody (Proteintech) solution was added to the membrane and incubated for 1 h at room temperature. After washing thrice with TBST for each 5 min, 1:10 000 (TBST buffer‐based) diluted HRP‐Goat anti‐Mouse IgG (H+L) Antibody (Proteintech) solution was added to the membrane and incubated for another 1 h at room temperature. After the last washing with TBST thrice for each 5 min, the membrane was visualized using Omni ECL reagent (EpiZyme) under UVP ChemStudio (analytikjena).Ochrobactrum spp. Separation, Identification, and CharacterizationM9‐erythritol medium (25 mL) was added to a 50 mL centrifuge tube and the lid was replaced by gauze to block aerosol dust and enable microorganisms to drop into the medium. This device was stationary‐placed outdoors at ≈25 °C for about 7 days. On the 7th day, the medium became turbid and then the mixture was spread onto an M9‐erythritol agar plate for picking single colonies at 25 °C. After 3 days, some white colonies were formed and picked up for identification. 16s rRNA was sequenced by primer pair 27F/1492R. Cell morphology was characterized by scanning electron microscope (JSM‐7800F Prime) and transmission electron microscope (JEM‐2100 Plus). Ochrobactrum spp. genome sequence was referred to the Brucella anthropi strain T16R‐87 chromosomes 1 and 2 (GenBank: CP044970.1 and CP044971.1).Incubation of Erythritol Catabolic Escherichia coli in M9‐Erythritol Liquid MediumAll engineered E. coli strains that could utilize erythritol as sole carbon source were cultured and followed the same procedure. Cell pellets from overnight LB culture were collected in 1.5 mL centrifuge tubes at 5000 × g and 4 °C for 10 min. The supernatant was discarded and the pellets were washed thrice by M9‐erythritol medium, then the pellets were resuspended with M9‐erythritol and diluted to OD600 = 1.0 for standardization. Then, the resuspended mixture was inoculated into new M9‐erythritol medium (100 mL medium in a 250 mL conical flask or 1 L medium in a 2 L conical flask) as 1:500 v/v for the following incubation. All antibiotic concentrations in M9‐erythritol medium were diluted by five times to avoid growth inhibition in the minimal medium as far as possible.HPLC Analysis of Erythritol1 mL M9‐erythritol medium (with E. coli) cell cultures were collected in 1.5 mL centrifuge tubes at 5000 × g and 4 °C for 10 min. The supernatants were then passed through a 0.22 µm filter to remove cell pellets and other precipitation. The resulting samples were used for HPLC analysis (see Figure S41, Supporting Information, for a standard curve of erythritol).The 1260 Infinity II Prime HPLC System (Aligent) with a Hi‐Plex H column (Aligent) was used for erythritol analysis. The experiment parameters were set as follows: column temperature: 60 °C, mobile phase: 5 mm H2SO4 solution, flow rate: 0.6 mL min−1, detector: refractive index detector at 40 °C, sample loading volume: 2 µL, testing time: 20 min for each sample.Characterization of eryD Binding Site In VivoeryD binding site (eryO) was characterized and located in pEryF‐732, which was a 732 bp gap between eryE (5′ start) and eryA (3′ end). In general, the characterization was performed as an in vivo incremental truncation process in two steps.First, to determine the 5′ site, a series of 100 bp‐truncated variant plasmids (pFB186 to pFB191, and pFB210) were constructed and characterized that eryO was located between pEryF‐232 to pEryF‐132. Second, a series of 10 bp‐truncated variant plasmids (pFB192 to pFB198, pFB208, and pFB209) were constructed and characterized that eryO was located between pEryF‐162 to pEryF‐152. Third, a series of 1 bp‐truncated variant plasmids (pFB202–pFB207) were constructed and characterized that eryO 5′ site was located in pEryF‐159.To determine the 3′ site, the truncation process was the same as described above. The constructed plasmids (pFB214–pFB238) were characterized step‐by‐step. Finally, eryO 3′ site was located in pEryF‐131. The exact eryO site was characterized as a 29 bp DNA sequence of 5′‐gaaaaaaaatgcgccatctagaaaatttt‐3′.Electrophoretic Mobility Shift AssayDNA fragment containing the 732 bp eryE‐eryA intergenic region (pEryF‐732) was amplified with the 5′ FAM‐labeled primers from the plasmid pFB147 and then gel purified. The DNA fragment (200 nm) was incubated with different concentrations of purified eryD in the binding buffer (10 mm tris‐HCl pH 7.5, 1 mm EDTA, 100 mm NaCl, 0.1 mm dithiothreitol [DTT], 10 µg mL−1 BSA, 5% glycerol). The 20 µL mixtures were incubated for 30 min at room temperature and then mixed with 5 µL of the same buffer supplemented with 50% glycerol and bromophenol blue. Free DNA and eryD‐DNA complexes were separated on 7.5% polyacrylamide gels in 1× TAE buffer (4.844 g L−1 tris‐base, 1.21 mL L−1 acetate acid, 0.372 g L−1 EDTA disodium salt dihydrate, pH 8.3). The fluorescence of FAM‐labeled bands was visualized under UVP ChemStudio (analytikjena).Characterization of the eryD OrthogonalityE. coli Mach1‐T1 with plasmids pFB186 and pFB261 was used for the characterization. Inducers (erythritol, glycerol, l‐arabinose, d‐xylose, d‐glucose, d‐mannose, d‐galactose, d‐fructose, l‐sorbose, l‐rhamnose, d‐sucrose, lactose, maltodextrin, d‐sorbitol, and d‐mannitol) were dissolved in ddH2O at different concentrations (5%, 1%, 0.1%, and 0.01% w/v), followed by filtration through a 0.22 µm filter.E. coli Mach1‐T1 was initially transformed with the relevant plasmids. Starter cultures (LB containing 50 µg mL−1 ampicillin and 17 µg mL−1 chloramphenicol) were inoculated from a single colony and grew overnight at 37 °C for 16 h. The next day, 10 µL starter culture was used to inoculate 5 mL LB medium containing 50 µg mL−1 ampicillin and 17 µg mL−1 chloramphenicol in test tubes. The cultures were incubated at 37 °C with shaking (250 rpm) until OD600 reached 0.6–0.8. The cultures were then added with different inducer solutions and incubated for another 16 h at 37 °C and 250 rpm.Then, cell pellets were collected in 1.5 mL centrifuge tubes at 5000 × g and 4 °C for 10 min. The supernatant was discarded and the pellets were resuspended with 1 mL 1× phosphate‐buffered saline (pH 7.4). The sfGFP fluorescence of resuspended mixture was measured by microplate reader (Synergy H1, BioTek) and performed with excitation and emission wavelengths at 485 and 528 nm, respectively. All measurements were performed at least in triplicate. In addition, cell pellets were lysed by sonication (Q125 sonicator, Qsonica, 10 s on/off, 50% of amplitude, input energy ≈600 joules). The lysate was then centrifuged at 12 000 × g and 4 °C for 10 min. The total fraction was separated and analyzed by western‐blot.eryD Structure ModelingOchrobactrum spp. eryD amino acid sequence was the same as eryD (Uniprot ID: Q2YIQ4) from Brucella abortus (strain 2308). AlphaFold, Robetta, and SWISS‐MODEL were used for structure modeling.Size Exclusion ChromatographyTo determine the native molecular weight and assembly model of eryD, SEC was performed using a Superdex 200 increase 10/300 GL column (GE Healthcare) and two reference proteins including Vlm2 (284 kDa)[58] and T7 RNA polymerase (99 kDa) were used for comparison.Characterization of Transmembrane Phenotype of the Erythritol ABC‐Transporter SystemTMHMM‐2.0 was used for transmembrane prediction. eryE was predicted as a single transmembrane protein and eryG was predicted as a ten‐times transmembrane protein. To determine the transmembrane phenotypes, pFB162 and pFB164 were constructed for eryE‐sfGFP and eryG‐sfGFP expression. After transformation into E. coli Mach1‐T1, starter cultures (LB containing 34 µg mL−1 chloramphenicol) were inoculated from a single colony and grew overnight at 37 °C for 16 h. The next day, 10 µL starter culture was used to inoculate 5 mL LB medium containing 34 µg mL−1 chloramphenicol in test tubes. The cultures were incubated at 37 °C with shaking (250 rpm) until OD600 reached 0.6–0.8. The cultures were then added 1% arabinose (w/v) and incubated for another 6 h at 30 °C and 250 rpm. Afterward, cell pellets from 1 mL cell culture were collected in 1.5 mL centrifuge tubes at 5000 × g and 4 °C for 10 min. The supernatant was discarded and the pellets were resuspended with 1 mL 1× phosphate‐buffered saline (pH 7.4). Then, the mixture was imaged under laser scanning confocal microscopy (FV3000, Olympus).Protein PurificationPlasmids pFB161 and pFB167 were used to express eryD and eryR, respectively. The two proteins were expressed and purified with the same protocol. E. coli BL21(DE3) was transformed with the above plasmids, respectively. Starter cultures (LB containing 100 µg mL−1 ampicillin) were inoculated from a single colony and grew overnight at 37 °C for 16 h. 5 mL starter culture was used to inoculate 1 L LB medium containing 100 µg mL−1 ampicillin in a 2 L conical flask. The cultures were incubated at 37 °C with shaking (250 rpm) until OD600 reached 0.6–0.8. The cultures were then quickly cooled down on ice rapidly to 20 °C, and isopropyl‐β‐D‐thiogalactopyranoside (IPTG) was added for induction at a final concentration of 0.5 mm. The cultures then grew for another 16 h at 20 °C and 220 rpm. Afterward, cells were harvested by centrifugation (Avanti JXN‐26 high‐speed centrifuge, Beckman Coulter) at 5000 × g and 4 °C for 10 min. The pellets were resuspended in Buffer A, containing 20 mm sodium phosphate (pH 7.4), 1 m sodium chloride, 1 mm DTT, and 50 mm imidazole. The suspension was cooled on ice and then lysed thrice at 1500 bar by ultra‐high‐pressure homogenization (JNBIO). The lysate was then centrifuged at 4 °C and 20 000 × g for 30 min. The supernatant was collected and passed through a 0.22 µm filter. The filtered solution was then purified by Ni2+ affinity chromatography using a 1 mL HisTrap FF column (GE Healthcare). The column was equilibrated with 25 mL Buffer A at a constant flow rate of 1 mL min−1. After equilibration, the filtered protein solution was loaded, followed by washing with 25 mL Buffer A. Then, bounded proteins were eluted with Buffer B (Buffer A, but with 500 mm imidazole) and collected into 1.5 mL centrifuge tubes with 0.5 mL elution. Each 0.5 mL elution was analyzed by SDS‐PAGE. Eluted protein samples were combined in one tube and desalinated using an ultrafiltration tube (Amicon Ultra 3 kDa molecular weight cut‐off, Merck/Millipore) with Buffer C (25 mm tris‐HCl (pH 7.5), 1 mm DTT, and 1 m sodium chloride). The desalinated and concentrated protein solution was then mixed with 40% glycerol (v/v in water) with a volume ratio of 1:1. The concentration of the final protein solution was measured at 280 nm after molar attenuation coefficient correction. Purified proteins were flash frozen by liquid nitrogen and stored at −80 °C until further use.Protein Expression and Solubility CharacterizationPlasmids pFB158–pFB167 were used to express ten erythritol catabolism‐associated genes. E. coli strains were transformed with the relevant plasmids. Starter cultures (LB containing 100 µg mL−1 ampicillin or 34 µg mL−1 chloramphenicol) were inoculated from a single colony and grew overnight at 37 °C for 16 h. The next day, 10 µL starter culture was used to inoculate 5 mL LB medium containing 100 µg mL−1 ampicillin or 34 µg mL−1 chloramphenicol in test tubes. The cultures were incubated at 37 °C with shaking (250 rpm) until OD600 reached 0.6–0.8. The cultures were then added with inducers (1% arabinose [w/v] or 0.5 mm IPTG) and incubated continuously. Afterward, cell pellets were collected in 1.5 mL centrifuge tubes at 5000 × g and 4 °C for 10 min. The supernatant was discarded and the pellets were resuspended with 1 mL 1× phosphate‐buffered saline (pH 7.4) and lysed by sonication (Q125 sonicator, Qsonica, 10 s on/off, 50% of amplitude, input energy ≈600 joules). The lysate was then centrifuged at 12 000 × g and 4 °C for 10 min. The total (cell lysate), soluble (supernatant), and pellet fractions were separated and analyzed by SDS‐PAGE and western‐blot.Fluorescence Measurement StandardizationsfGFP (C‐terminal 6× His) was used as a fluorescence reporter. pFB286–pFB291 were performed as standardized sfGFP expression plasmids. These plasmids contained four parts within iGEM standardized vector pSB1C3: E. coli constitutive promoter library (J23100, J23106, J23105, J23114, J23109, and J23113, from strong to weak), strong ribosome binding site B0034, sfGFP, and a strong terminator B0015. Other plasmids containing sfGFP were compared to these six standardized plasmids to determine the sfGFP expression level. sfGFP fluorescence was measured by microplate reader (Synergy H1, BioTek) and performed with excitation and emission wavelengths at 485 and 528 nm, respectively. All measurements were performed at least in triplicate.Knocking Out Escherichia coli GenesGene knock‐out strategy was derived from Lambda‐Red recombination. In brief, plasmids pFB278–pFB284 and pFB306–pFB339 were constructed to obtain linear PCR products. The descriptions “homoL” and “homoR” of these plasmids represented “homologous left arm” and “homologous right arm” between the target knock‐out gene or gene cluster. These two DNA sequences were designed approximately longer than 300 bp for a higher recombination efficiency. The linear PCR products (homoL–FRT–KanR–FRT–homoR) were obtained and purified for use.Wild‐type E. coli MG1655 was first transformed with pKD46. On the next day, a single colony was picked up and inoculated in LB medium (with 1% arabinose, w/v) and incubated at 30 °C and 250 rpm for ≈4–6 h. When OD600 reached 0.6–0.8, the cell pellets were collected in 1.5 mL centrifuge tubes at 5000 × g and 4 °C for 10 min. The supernatant was discarded and the pellets were washed thrice with 4 °C ddH2O. Then, 1 µg purified linear PCR products were added to the cell mixture and proceeded electro‐transformation (Gene Pulser Xcell system [Bio‐Rad] and Gene Pulser/Micropulser electroporation cuvettes, 0.1 cm gap [Bio‐Rad] were used). Subsequently, colony PCR (primer pair: “homoL” forward primer/“homoR” reverse primer) was performed to test whether the gene or gene cluster deleted. Finally, the expected knock‐out E. coli MG1655 strains were continuously incubated for days at 37 °C for pKD46 (temperature‐sensing plasmid) elimination.mRNA Transcriptome AnalysisTo process transcriptome analysis, biological triplicate of E. coli MG1655 (with plasmid pFB147) in both M9‐glucose medium (control group) and M9‐erythritol medium (experimental group) were performed.Cell pellets from overnight LB culture (37 °C, 250 rpm for 16 h) were collected in 1.5 mL centrifuge tubes at 5000 × g and 4 °C for 10 min. The supernatant was discarded and the pellets were washed thrice by M9‐glucose or M9‐erythritol medium. Then, the pellets were resuspended and diluted to OD600 = 1.0 for standardization. Afterward, the suspensions were inoculated into new M9‐glucose or M9‐erythritol medium (1 L medium in a 2 L conical flask) as 1:500 v/v for the following incubation.For M9‐glucose cultivation, when OD600 reached 0.6 (mid‐log phase, ≈6 h after inoculation), cell pellets were collected. For M9‐erythritol cultivation, when OD600 reached 0.15 (mid‐log phase, ≈36 h after inoculation), cell pellets were collected. Then, the cell pellets were sent to GENEWIZ for further mRNA transcriptome analysis.Distinguishing Soda DrinksE. coli MG1655 harboring pFB151 and pFB147 were used as NC and living ErD, respectively. Three kinds of sweet soda drinks were bought from the market. The ingredients of soda 1 contained water, carbon dioxide, citric acid, potassium citrate, acesulfame potassium, sodium benzoate, sucralose, and sodium gluconate. Soda 2 contained water, high‐fructose corn syrup (including fructose and glucose), sugar (sucrose), carbon dioxide, citric acid, potassium citrate, sodium benzoate, sucralose, and acesulfame potassium. Soda 3 contained water, erythritol, carbon dioxide, citric acid, potassium citrate, sodium bicarbonate, and sucralose. All sodas were passed through a 0.22 µm filter to remove impurities and then shaken to eliminate carbon dioxide.For liquid incubation, the sodas were diluted by ten times in ddH2O to decrease the ingredient concentration, and the diluted sodas were used as different liquid mediums. NC and ErD cell pellets from overnight LB culture (37 °C, 250 rpm for 16 h) were collected in 1.5 mL centrifuge tubes at 5000 × g and 4 °C for 10 min. The supernatant was discarded and the pellets were washed thrice by ddH2O. Then, the pellets were resuspended and diluted to OD600 = 1.0 for standardization. Then, 10 µL mixtures were inoculated into 5 mL diluted sodas and incubated at 37 °C and 250 rpm for 12, 24, and 36 h, respectively. The solutions were imaged at each time point.For solid incubation, M9 agar plates were prepared (10 mL mixture in 10 cm plastic plates) for cell cultivation. 1 mL of filtered sodas (without dilution) were added and spread on the solid medium and waited for drying. Then, 10 µL NC or ErD cell mixtures were inoculated onto the solid medium and incubated at 37 °C for 24 and 48 h, respectively. The plates were imaged at each time point.Engineered Escherichia coli Nissle 1917 Grows in Simulated Intestinal FluidThe probiotic EcN (Mutaflor) was transformed with plasmid pFB147. Then, EcN cells were diluted in ddH2O with a gradient density (2 × 102, 2 × 104, 2 × 106, and 2 × 108) for the following experiments. 2× SIF was prepared according to a standard protocol with additional erythritol (0%, 0.02%, 0.2%, 2%, 8%, and 32% w/v).[59,60] The SIF solutions were passed through a 0.22 µm filter. Then, 7.5 mL EcN culture and 7.5 mL 2× SIF were mixed as 1:1 volume to obtain a 15 mL culture. Then, the cultures were stationary‐placed for incubation for days at 37 °C. At each time point, ≈100 µL samples were taken and spread (diluted if necessary) on LB‐agar plates for counting CFU. All experiments were performed at least in triplicate.AcknowledgementsThis work was supported by grants from the National Natural Science Foundation of China (Nos. 31971348 and 32171427) and the Double First‐Class Initiative Fund of ShanghaiTech University (No. SYLDX0292022).Conflict of InterestThe authors declare no conflict of interest.Author ContributionsJ.L. and F.B. designed the experiments. F.B. performed all experiments. X.J. helped perform cell cultivation. S.H. and W.‐Q.L. performed HPLC analysis. Y.Z. helped perform molecular cloning. F.B. analyzed the data and drafted the manuscript. J.L., S.L., and Y.L. revised and edited the manuscript. J.L. conceived and supervised the study. All authors read and approved the final manuscript.Data Availability StatementThe data that support the findings of this study are available from the corresponding author upon reasonable request.M. Grembecka, Eur. Food Res. Technol. 2015, 241, 1.P. Shankar, S. Ahuja, K. Sriram, Nutrition 2013, 29, 1293.W. O. Bernt, J. F. Borzelleca, G. Flamm, I. C. Munro, Regul. Toxicol. Pharmacol. 1996, 24, S191.I. C. Munro, W. O. Berndt, J. F. Borzelleca, G. Flamm, B. S. Lynch, E. Kennepohl, E. A. Bär, J. Modderman, Food Chem. Toxicol. 1998, 36, 1139.T. Oku, M. Okazaki, Nutr. Res. 1996, 16, 577.M. Veiga‐da‐Cunha, H. Santos, E. Van Schaftingen, J. Bacteriol. 1993, 175, 3941.H. Richter, D. Vlad, G. Unden, Arch. Microbiol. 2001, 175, 26.A. Mirończuk, A. Dobrowolski, M. Rakicka, A. Rywińska, W. Rymowicz, Process Biochem. 2015, 50, 61.W. Rymowicz, A. Rywińska, M. Marcinkiewicz, Biotechnol. Lett. 2009, 31, 377.X. Qiu, P. Xu, X. Zhao, G. Du, J. Zhang, J. Li, Metab. Eng. 2020, 60, 66.P. de Cock, in Sweeteners and Sugar Alternatives in Food Technology (Eds: K. O'Donnell, M. W. Kearsley), Wiley‐Blackwell, Hoboken, USA 2012, Ch. 10.G. Livesey, Nutr. Res. Rev. 2013, 16, 163.E. Arrigoni, F. Brouns, R. Amadò, Br. J. Nutr. 2005, 94, 643.G. J. den Hartog, A. W. Boots, A. Adam‐Perrot, F. Brouns, I. W. Verkooijen, A. R. Weseler, G. R. Haenen, A. Bast, Nutrition 2010, 26, 449.F. J. Ruiz‐Ojeda, J. Plaza‐Díaz, M. J. Sáez‐Lara, A. Gil, Adv. Nutr. 2019, 10, S31.M. Carocho, P. Morales, I. Ferreira, Food Chem. Toxicol. 2017, 107, 302.T. Barbier, F. Collard, A. Zúñiga‐Ripa, I. Moriyón, T. Godard, J. Becker, C. Wittmann, E. Van Schaftingen, J.‐J. Letesson, Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 17815.A. M. Lillo, C. N. Tetzlaff, F. J. Sangari, D. E. Cane, Bioorg. Med. Chem. Lett. 2003, 13, 737.F. J. Sangari, J. Agüero, J. M. Garcia‐Lobo, Microbiology 2000, 146, 487.J. F. Sperry, D. C. Robertson, J. Bacteriol. 1975, 121, 619.M. C. Rodríguez, C. Viadas, A. Seoane, F. J. Sangari, I. López‐Goñi, J. M. García‐Lobo, PLoS One 2012, 7, e50876.B. A. Geddes, G. Hausner, I. J. Oresnik, BMC Microbiol. 2013, 13, 46.T. Barbier, A. Zúñiga‐Ripa, S. Moussa, H. Plovier, J. F. Sternon, L. Lázaro‐Antón, R. Conde‐Álvarez, X. De Bolle, M. Iriarte, I. Moriyón, J. J. Letesson, Crit. Rev. Microbiol. 2018, 44, 182.B. A. Geddes, I. J. Oresnik, Microbiology 2012, 158, 2180.H. Huang, M. S. Carter, M. W. Vetting, N. Al‐Obaidi, Y. Patskovsky, S. C. Almo, J. A. Gerlt, J. Am. Chem. Soc. 2015, 137, 14570.S. Y. Park, H. Eun, M. H. Lee, S. Y. Lee, Nat. Catal. 2022, 5, 726.R. S. Ayikpoe, C. Shi, A. J. Battiste, S. M. Eslami, S. Ramesh, M. A. Simon, I. R. Bothwell, H. Lee, A. J. Rice, H. Ren, Q. Tian, L. A. Harris, R. Sarksian, L. Zhu, A. M. Frerk, T. W. Precord, W. A. van der Donk, D. A. Mitchell, H. Zhao, Nat. Commun. 2022, 13, 6135.J. Li, P. Neubauer, New Biotechnol. 2014, 31, 579.J. Bang, S. Y. Lee, Proc. Natl. Acad. Sci. U. S. A. 2018, 115, E9271.S. Gleizer, R. Ben‐Nissan, Y. M. Bar‐On, N. Antonovsky, E. Noor, Y. Zohar, G. Jona, E. Krieger, M. Shamshoum, A. Bar‐Even, R. Milo, Cell 2019, 179, 1255.F. Y.‐H. Chen, H.‐W. Jung, C.‐Y. Tsuei, J. C. Liao, Cell 2020, 182, 933.P. Keller, M. A. Reiter, P. Kiefer, T. Gassler, L. Hemmerle, P. Christen, E. Noor, J. A. Vorholt, Nat. Commun. 2022, 13, 5243.B. L. Adams, ACS Synth. Biol. 2016, 5, 1328.M. Reister, K. Hoffmeier, N. Krezdorn, B. Rotter, C. Liang, S. Rund, T. Dandekar, U. Sonnenborn, T. A. Oelschlaeger, J. Biotechnol. 2014, 187, 106.S. Bathe, W. Achouak, A. Hartmann, T. Heulin, M. Schloter, M. Lebuhn, FEMS Microbiol. Ecol. 2006, 56, 272.P. S. Chain, D. J. Comerci, M. E. Tolmasky, F. W. Larimer, S. A. Malfatti, L. M. Vergez, F. Aguero, M. L. Land, R. A. Ugalde, E. Garcia, Infect. Immun. 2005, 73, 8353.J. Lamontagne, M. Béland, A. Forest, A. Côté‐Martin, N. Nassif, F. Tomaki, I. Moriyón, E. Moreno, E. Paramithiotis, BMC Genomics 2010, 11, 300.J. R. Kelly, A. J. Rubin, J. H. Davis, C. M. Ajo‐Franklin, J. Cumbers, M. J. Czar, K. de Mora, A. L. Glieberman, D. D. Monie, D. Endy, J. Biol. Eng. 2009, 3, 4.J. Beal, T. Haddock‐Angelli, G. Baldwin, M. Gershater, A. Dwijayanti, M. Storch, K. de Mora, M. Lizarazo, R. Rettberg, iGEM Interlab Study Contributors, PLoS One 2018, 13, e0199432.J. Beal, G. S. Baldwin, N. G. Farny, M. Gershater, T. Haddock‐Angelli, R. Buckley‐Taylor, A. Dwijayanti, D. Kiga, M. Lizarazo, J. Marken, K. de Mora, R. Rettberg, V. Sanchania, V. Selvarajah, A. Sison, M. Storch, C. T. Workman, iGEM Interlab Study Contributors, PLoS One 2021, 16, e0252263.M. T. Pellicer, C. Fernandez, J. Badía, J. Aguilar, E. C. Lin, L. Baldom, J. Biol. Chem. 1999, 274, 1745.P. Seweryn, L. B. Van, M. Kjeldgaard, C. J. Russo, L. A. Passmore, B. Hove‐Jensen, B. Jochimsen, D. E. Brodersen, Nature 2015, 525, 68.S. F. Altschul, T. L. Madden, A. A. Schäffer, J. Zhang, Z. Zhang, W. Miller, D. J. Lipman, Nucleic Acids Res. 1997, 25, 3389.M. Johnson, I. Zaretskaya, Y. Raytselis, Y. Merezhuk, S. McGinnis, T. L. Madden, Nucleic Acids Res. 2008, 36, W5.K. B. Heller, E. C. Lin, T. H. Wilson, J. Bacteriol. 1980, 144, 274.T. Shindou, Y. Sasaki, H. Miki, T. Euguchi, K. Hagiwara, T. Ichikawa, J. Agric. Food Chem. 1989, 37, 1474.U. Sonnenborn, FEMS Microbiol. Lett. 2016, 363, fnw212.D. B. Pedrolli, N. V. Ribeiro, P. N. Squizato, V. N. de Jesus, D. A. Cozetto, Team AQA Unesp at iGEM 2017, Trends Biotechnol. 2019, 37, 100.V. W. Kelly, B. K. Liang, S. J. Sirk, ACS Synth. Biol. 2020, 9, 3184.A. Cubillos‐Ruiz, T. Guo, A. Sokolovska, P. F. Miller, J. J. Collins, T. K. Lu, J. M. Lora, Nat. Rev. Drug Discovery 2021, 20, 941.E. L. Forker, J. Clin. Invest. 1967, 46, 1189.A. Rodriguez, J. A. Martínez, N. Flores, A. Escalante, G. Gosset, F. Bolivar, Microb. Cell Fact. 2014, 13, 126.M. Jiang, H. Zhang, Curr. Opin. Biotechnol. 2016, 42, 1.Z.‐P. Zou, Y. Du, T.‐T. Fang, Y. Zhou, B.‐C. Ye, Cell Host Microbe 2023, 31, 199.V. M. Isabella, B. N. Ha, M. J. Castillo, D. J. Lubkowicz, S. E. Rowe, Y. A. Millet, C. L. Anderson, N. Li, A. B. Fisher, K. A. West, P. J. Reeder, M. M. Momin, C. G. Bergeron, S. E. Guilmain, P. F. Miller, C. B. Kurtz, D. Falb, Nat. Biotechnol. 2018, 36, 857.X. Yu, C. Lin, J. Yu, Q. Qi, Q. Wang, Microb. Biotechnol. 2020, 13, 629.F. Ba, Y. Liu, W.‐Q. Liu, X. Tian, J. Li, Nucleic Acids Res. 2022, 50, 2973.L. Zhuang, S. Huang, W.‐Q. Liu, A. S. Karim, M. C. Jewett, J. Li, Metab. Eng. 2020, 60, 37.A. Brodkorb, L. Egger, M. Alminger, P. Alvito, R. Assunção, S. Ballance, T. Bohn, C. Bourlieu‐Lacanal, R. Boutrou, F. Carrière, A. Clemente, M. Corredig, D. Dupont, C. Dufour, C. Edwards, M. Golding, S. Karakaya, B. Kirkhus, S. L. Feunteun, U. Lesmes, A. Macierzanka, A. R. Mackie, C. Martins, S. Marze, D. J. McClements, O. Ménard, M. Minekus, R. Portmann, C. N. Santos, I. Souchon, et al., Nat. Protoc. 2019, 14, 991.M. Minekus, M. Alminger, P. Alvito, S. Ballance, T. Bohn, C. Bourlieu, F. Carrière, R. Boutrou, M. Corredig, D. Dupont, C. Dufour, L. Egger, M. Golding, S. Karakaya, B. Kirkhus, S. L. Feunteun, U. Lesmes, A. Macierzanka, A. Mackie, S. Marze, D. J. McClements, O. Ménard, I. Recio, C. N. Santos, R. P. Singh, G. E. Vegarud, M. S. J. Wickham, W. Weitschies, A. Brodkorb, Food Funct. 2014, 5, 1113.
Journal
Advanced Science – Wiley
Published: May 1, 2023
Keywords: carbon source; erythritol; Escherichia coli; metabolic engineering; synthetic biology