Enhancing fructosylated chondroitin production in Escherichia coli K4 by balancing the UDP-precursors

Highlights

• Chondroitin is widely used in medical, veterinary, drug, and cosmetic products.

• Balancing UDP-precursors maximizes fructosylated chondroitin in E. coli K4.

• Significant improvements were gained over current microbial production methods.

• The approach offers an alternative to extraction from animal tissues.

Abstract

Microbial production of chondroitin and chondroitin-like polysaccharides from renewable feedstock is a promising and sustainable alternative to extraction from animal tissues. In this study, we attempted to improve production of fructosylated chondroitin in Escherichia coli K4 by balancing intracellular levels of the precursors UDP-GalNAc and UDP-GlcA. To this end, we deleted pfkA to favor the production of Fru-6-P. Then, we identified rate-limiting enzymes in the synthesis of UDP-precursors. Third, UDP-GalNAc synthesis, UDP-GlcA synthesis, and chondroitin polymerization were combinatorially optimized by altering the expression of relevant enzymes. The ratio of intracellular UDP-GalNAc to UDP-GlcA increased from 0.17 in the wild-type strain to 1.05 in a 30-L fed-batch culture of the engineered strain. Titer and productivity of fructosylated chondroitin also increased to 8.43 g/L and 227.84 mg/L/h; the latter represented the highest productivity level achieved to date.

Introduction

Sulfated chondroitin is a type of glycosaminoglycan with a backbone consisting of a repeating disaccharide unit of D-glucuronic acid (GlcA) and N-acetyl-D-galactosamine (GalNAc) linked via β1 → 3 or β1 → 4 and substituted at various positions with O-sulfo groups (Cimini et al., 2010a, He et al., 2015). Sulfated chondroitin is widely used in medical, veterinary, pharmaceutical, and cosmetic products because of its outstanding biophysical properties (He et al., 2017, Suflita et al., 2015, Wang et al., 2016). Chondroitin and chondroitin-like polysaccharides can be produced by microbial fermentation (DeAngelis et al., 2002, He et al., 2015, Jin et al., 2016b); however, titer, yield, and productivity are much lower than those achieved by extraction from animal tissues. Of note, the capsular polysaccharide of E. coli K4 was reported to consist of a chondroitin backbone with fructose side branches at the 3-position in GlcA (Rodriguez et al., 1988). Accordingly, the production of chondroitin and chondroitin-like polysaccharides from E. coli K4 has been attempted by biochemical (Cimini et al., 2010b, Restaino et al., 2011, Restaino et al., 2013) and metabolic engineering (Badri et al., 2018) through gene overexpression (Cimini et al., 2015, Cimini et al., 2013, Cimini et al., 2010a, Cimini et al., 2014), gene deletion (Liu et al., 2014a), and transcriptional engineering (Wu et al., 2013). Although the titers of chondroitin and chondroitin-like polysaccharides increased significantly as a result, the metabolic flux to the precursors UDP-GalNAc and UDP-GlcA remained extremely imbalanced (Fig. 1) (Cimini et al., 2013, Tlapak-Simmons et al., 1999). Therefore, it may be necessary to balance this flux to maximize production.

In metabolic engineering, microbial metabolism is altered to boost the production of native metabolites or even produce nonnative metabolites (Chae et al., 2017, Nielsen and Keasling, 2016). A usual strategy is to increase the endogenous supply of precursors or to improve pathway efficiency. For example, efforts to increase the production of terpene from the precursors isopentenyl diphosphate and dimethylallyl diphosphate included protein engineering, in which site-saturation mutagenesis and error-prone PCR were used to mutate and enhance the catalytic activity of levopimaradiene synthase and geranylgeranyl synthase, resulting in a 2,600-fold increase in levopimaradiene production (Leonard et al., 2010). In addition, multivariate modular pathway engineering has been attempted, in which the taxadiene metabolic pathway was partitioned into an upstream methylerythritol-phosphate module and a downstream terpenoid-forming module; these were then simultaneously optimized (Ajikumar et al., 2010). Transport engineering, in which a pump from Alcanivorax borkumensis was expressed in limonene-producing E. coli to increase tolerance to exogenous biofuel and thereby improve limonene yield by 1.6-fold, has additionally been investigated (Dunlop et al., 2011). Finally, promoter engineering has been explored, in which synthetic constitutive promoters and redesigned 5′-untranslated regions were used to tune the expression of phosphoenolpyruvate synthetase A and glyceraldehyde 3-phosphate dehydrogenase, thereby balancing the metabolic flux to precursors and increasing the production of isoprenoids (Jung et al., 2016).

In the present study, we attempted to enhance the production of fructosylated chondroitin in a similar manner. To this end, we first identified rate-limiting steps in this pathway. We then knocked out pfkA and overexpressed pathway enzymes to enhance the synthesis of UDP-GalNAc and UDP-GlcA. Next, we attempted modular pathway engineering to simultaneously optimize UDP-GalNAc synthesis, UDP-GlcA synthesis, and chondroitin polymerization. We found that, under controlled culture conditions, the ratio of intracellular UDP-GalNAc to UDP-GlcA increased from 0.17 in the wild-type strain to 1.05 in the engineered strain ZQ25, which produced 8.43 g/L fructosylated chondroitin at the highest productivity (227.84 mg/L/h) level achieved to date. The data highlight the importance of balancing the metabolic flux to precursors to efficiently produce glycosaminoglycans in microbial factories.