Production of Rainbow Colorants by Metabolically Engineered Escherichia coli

Abstract

There has been much interest in producing natural colorants to replace synthetic colorants of health concerns. Escherichia coli has been employed to produce natural colorants including carotenoids, indigo, anthocyanins, and violacein. However, production of natural green and navy colorants has not been reported. Many natural products are hydrophobic, which are accumulated inside or on the cell membrane. This causes cell growth limitation and consequently reduces production of target chemicals. Here, integrated membrane engineering strategies are reported for the enhanced production of rainbow colorants-three carotenoids and four violacein derivatives-as representative hydrophobic natural products in E. coli. By integration of systems metabolic engineering, cell morphology engineering, inner- and outer-membrane vesicle formation, and fermentation optimization, production of rainbow colorants are significantly enhanced to 322 mg L^-1 of astaxanthin (red), 343 mg L^-1 of β-carotene (orange), 218 mg L^-1 of zeaxanthin (yellow), 1.42 g L^-1 of proviolacein (green), 0.844 g L^-1 of prodeoxyviolacein (blue), 6.19 g L^-1 of violacein (navy), and 11.26 g L^-1 of deoxyviolacein (purple). The membrane engineering strategies reported here are generally applicable to microbial production of a broader range of hydrophobic natural products, contributing to food, cosmetic, chemical, and pharmaceutical industries.

Keywords: membrane engineering; metabolic engineering; natural products; rainbow colorants; vesicle.

Figure 1

Overview of the metabolic engineering and membrane structure expansion strategies for the enhanced production of rainbow colorants (red, astaxanthin; orange, β‐carotene; yellow, zeaxanthin; green, proviolacein; blue, prodeoxyviolacein; navy, violacein; purple, deoxyviolacein). Morphology engineering was performed by knocking down the genes involved in cell division or cell wall metabolism. Inner‐membrane vesicles (IMVs) were formed by introducing the cav1 gene encoding human caveolin‐1. Outer‐membrane vesicles (OMVs) were formed by knocking down the genes involved in OMV formation. The synthetic sRNA technology was employed to knockdown the expression levels of target genes by blocking translation. Bent arrow and T‐shape represent promoter and terminator, respectively. Solid and dotted lines represent single and multiple reactions, respectively. Abbreviations: G3P, glyceraldehyde 3‐phosphate; E4P, erythrose 4‐phosphate; PEP, phosphoenolpyruvate; PYR, pyruvate; DXP, 1‐deoxy‐D‐xylulose 5‐phosphate; SKM, shikimate; FPP, farnesyl diphosphate; GGPP, geranylgeranyl pyrophosphate; TRP, L‐tryptophan; IPA, indole pyruvate; Sp., spontaneous.

Figure 2

Morphology engineering and IMV formation for the enhanced production of rainbow colorants. A) Schematic representation of morphology engineering to expand the space in which rainbow colorants can be accumulated. Knockdown of genes involved in cell division would lead to elongated cells whereas knockdown of genes involved in cell wall synthesis or maintenance would lead to shorter cells with spherical and irregular shapes. The synthetic sRNA technology was employed to knockdown the expression levels of target genes by blocking translation. B) β‐Carotene production in the engineered strains introduced with the sRNAs targeting genes related to cellular morphology. C) Deoxyviolacein production in the engineered strains introduced with the sRNAs targeting genes related to cellular morphology. D) Schematic representation of employing IMVs (caveolae). IMVs were formed by introducing the cav1 gene encoding human caveolin‐1. E) β‐Carotene production by employing IMVs. F) Deoxyviolacein production by employing IMVs. TEM (upper panels) and SEM (lower panels) images of G) the control β‐carotene producer BTC1, H) BTC1 expressing cav1, I) the control deoxyviolacein producer deoxyviolacein, and J) deoxyviolacein expressing cav1 are shown. For panels (H and I), red arrows represent IMVs. Error bars are mean ± SD (standard deviation; n = 3). B,C) *P < 0.05, determined by two‐tailed Student's t‐test. E,F) *P < 0.01, **P < 0.002, determined by two‐tailed Student's t‐test. P‐value thresholds were adjusted using Bonferroni correction (corrected significance levels represented as α/m; α, original significance level; m, number of hypotheses). NS, not significant.

Figure 3

Formation of OMVs for the enhanced production of rainbow colorants. A) Schematic representation of OMVs formation. OMVs were formed by repression of genes encoding outer‐membrane proteins, repression of genes related to outer‐membrane or peptidoglycan integrity, or activation of the σ^E factor. The synthetic sRNA technology was employed to knockdown the expression levels of target genes by blocking translation. Abbreviations are OMP, outer‐membrane protein; LPS, lipopolysaccharide. B) β‐Carotene production by employing OMVs. *P < 0.0083, **P < 0.0017, ***P < 0.00017, determined by two‐tailed Student's t‐test. C) Deoxyviolacein production by employing IMVs. *P < 0.0125, **P < 0.0025, ***P < 0.00025, determined by two‐tailed Student's t‐test. TEM (upper panels) and SEM (lower panels) images of D) BTC1 introduced with anti‐rffD sRNA and E) DVIO introduced with anti‐rfaI sRNA. F) SEM images of purified OMVS from BTC1 harboring anti‐rffD sRNA (upper panel) and DVIO harboring anti‐rfaI sRNA (lower panel). TEM (upper panels) and SEM (lower panels) images of I) BTC1 introduced with anti‐rffD and anti‐rfaD sRNAs and cav1‐plsBC and J) DVIO introduced with anti‐rfaI sRNA and cav1. G) β‐Carotene production by testing synergistic effects of employing the strategies of forming IMVs and OMVs. H) Deoxyviolacein production by testing synergistic effects of employing the strategies of morphology engineering, IMVs formation, and OMVs formation. Abbreviations in (G,H) are Tot, total titer; Sec, the titer obtained from the extracellular medium; Morph, morphology engineering. For panels (D–J), Red arrows represent IMVs or OMVs. Error bars are mean ± SD (n = 3). P‐value thresholds were adjusted using Bonferroni correction (P < α/m). NS, not significant.

Figure 4

Time profiles of the fed‐batch fermentation of engineered strains producing seven rainbow colorants. Fed‐batch fermentation profiles of A) ATX68 (pWAS‐anti‐rffDrfaD) producing astaxanthin, B) BTC1 (pWAS‐anti‐rffDrfaD) producing β‐carotene, C) ZEA20 (pWAS‐anti‐rffDrfaD) producing zeaxanthin, D) PVIO (pWAS‐anti‐rfaI‐cav1) producing proviolacein, E) PDVIO (pWAS‐anti‐rfaI) producing prodeoxyviolacein, F) VIO (pWAS‐anti‐rfaI‐cav1) producing violacein, and G) DVIO (pWAS‐anti‐rfaI‐cav1) producing deoxyviolacein are shown. (H) Picture of rainbow colorants produced by the engineered strains. Each colorant was appropriately diluted in chloroform (β‐carotene), acetone (zeaxanthin), or DMSO (astaxanthin and violacein derivatives). The concentrations of the colorants shown in the picture are: red, 43.3 mg L^–1 astaxanthin; orange, 80.5 mg L^–1 β‐carotene; yellow, 14.9 mg L^–1 zeaxanthin; green, 6.56 mg L^–1 proviolacein; blue, 5.51 mg L^–1 prodeoxyviolacein; navy, mixture of 59.7 mg L^–1 violacein and 3.8 mg L^–1 deoxyviolacein; purple, 46.9 mg L^–1 deoxyviolacein. Summary of the titers of I) β‐carotene and J) deoxyviolacein obtained by flask culture of the strains applied with the major strategies in this study. Abbreviations are Tot, total titer; Sec, the titer obtained from extracellular medium; DCW, dry cell weight; KD, knockdown; OE, overexpression.