Development of an Escherichia coli-based biocatalytic system for the efficient synthesis of N-acetyl-D-neuraminic acid

Highlights

• Altering the balance of transcription improved Neu5Ac synthesis.

• Structure-guided protein engineering increased Neu5Ac productivity significantly.

• System optimization yielded an efficient E. coli biocatalyst for Neu5Ac synthesis.

Abstract

N-acetyl-d-neuraminic acid (Neu5Ac) is a valuable resource that has seen increasing demand in both medicine and biotechnology. Although enzymatic systems and whole-cell biocatalysts have been developed for the synthesis of Neu5Ac, low yield and productivity still hamper the use of these methods on larger scales. We report the creation of an Escherichia coli biocatalyst for the efficient synthesis of Neu5Ac using a metabolic and protein engineering strategy. Expression of the two enzymes, N-acetyl-D-glucosamine 2-epimerase (AGE) and Neu5Ac lyase (NAL), was balanced using promoter engineering. Genes encoding competing pathways and GlcNAc catabolism were deleted, and then a structure-guided process was used to identify a more efficient NAL and an AGE mutant with a higher rate of Neu5Ac synthesis. The resulting biocatalyst produced 351.8 mM Neu5Ac with a yield of 58.6% from GlcNAc. This work exemplifies the use of rational design and protein engineering to construct a complex bacterial biocatalyst that can serve as a platform for the large-scale synthesis of a useful biological material.

Introduction

Sialic acids, derivatives of the nine-carbon monosaccharide neuraminic acid, are distributed widely in animal tissues and in other species ranging from plants and fungi to some microorganisms. N-acetyl-D-neuraminic acid (Neu5Ac), the main sialic acid in biological systems, plays important roles in biological, pathological, and immunological recognition processes, such as cell adhesion, cell signaling, glycoprotein stability, bacterial virulence, and tumor metastasis (Chen and Varki, 2010, Maru et al., 2002, Schauer, 2000). As a potential raw material for the synthesis of the antiviral medication Zanamivir, Neu5Ac can be used to prevent influenza type A and B infections (Tao et al., 2010). Neu5Ac is also an important additive in dairy products that can strengthen the immunity of infants (Bode, 2006). However, the high cost of Neu5Ac, which is caused by the conventional methods of its production, has impeded its industrial application.

Neu5Ac has been prepared by extraction from natural sources, such as egg yolk, and by hydrolysis of colominic acid (Maru et al., 2002). These methods are limited by low yield and unsatisfactory stereoselectivity. The chemical synthesis of Neu5Ac has been unsuitable for large-scale production because of its tedious protection and deprotection steps (Maru et al., 2002, Xu et al., 2007). In addition to the processes mentioned above, a two-enzyme, sequential system comprising N-acetylglucosamine 2-epimerase (AGE, EC 5.1.3.8) and N-acetylneuraminic acid lyase (NAL, EC 4.1.3.3) has been developed to produce Neu5Ac from N-acetylglucosamine (GlcNAc) (Gao et al., 2011, Lee et al., 2004, Schauer, 2000, Soundararajan et al., 2009, Traving and Schauer, 1998). The original process was hampered by a lengthy reaction time, low yield and low productivity. To improve synthesis performance, AGE and NAL were purified, immobilized, and coupled within one reactor to produce Neu5Ac from GlcNAc (Hu et al., 2010). The conversion rate from GlcNAc to Neu5Ac improved to 73% within 24 h. Moreover, a chemoenzymatic synthesis of Neu5Ac from GlcNAc using E. coli NAL (EcNAL) displayed on the surface of Bacillus subtilis WB600 spores was able to produce Neu5Ac with a concentration of 53.9 g L⁻¹ (Gao et al., 2011). However, these enzymatic processes are complicated and costly because the enzymes must be partially or completely purified, and ATP is required to activate AGE.

Compared with the isolation and purification of enzymes, whole-cell biocatalysts can be more readily and inexpensively prepared and are particularly useful for multi-enzyme reactions or those that require cofactor regeneration (Both et al., 2016). Two different E. coli systems have been developed for Neu5Ac synthesis. The first system couples two separate engineered strains to produce Neu5Ac; one overexpressing AGE and the other EcNAL. In the system produced in our laboratory, which used an AGE from Synechocystis sp. and EcNAL, the overall process generated 260.0 mM Neu5Ac (80.4 g/L) with a conversion yield of 43.3% from GlcNAc (Zhou et al., 2016). Lee et al. (2007a, b) used recombinant E. coli BL21(DE3) strains expressing Anabaena sp. CH1 AGE, and EcNAL to produce 0.4 M Neu5Ac with a yield of 33.3% from GlcNAc (Lee et al., 2007a, b). Zhang et al. (2010) constructed two recombinant E. coli strains expressing the AGE from Synechocystis sp. PCC6803 and EcNAL using a temperature-responsive expression system. Their system produced 61.9 mM Neu5Ac from 200 mM GlcNAc. The second system uses a single strain expressing both AGE and EcNAL to synthesize Neu5Ac. Lin et al. developed a single-cell biocatalytic process that achieved a maximum titer of 240 mM Neu5Ac (74.2 g/L) with a productivity of 6.2 g/L·h⁻¹ and a conversion yield of 40% from GlcNAc.(Lin et al., 2013) In similar processes reported by Tao et al. (2011) and Sun et al. (2013), the maximal Neu5Ac concentrations were 191.5 and 198.4 mM, respectively.

However, low conversion yields hamper biotechnological Neu5Ac production at larger scales. Because of the reversibility of AGE and NAL in the two-enzyme Neu5Ac synthesis cascade (Yu et al., 2004) (Fig. 1), systematic process improvement should receive more attention. These efforts should include screening for novel enzymes and achieving an elegant balance of enzyme expression. Here, we report the creation of a biocatalyst for the efficient synthesis of Neu5Ac using a strategy that combines rational metabolic engineering with protein engineering in E. coli BL21(DE3). Using the indicated strategy, which can facilitate the comprehensive evaluation of biocatalysts involving multiple genes, we developed a high-yield, high-productivity platform for Neu5Ac biosynthesis suitable for use in a 5-L biocatalytic reaction system (Fig. 1).