STUDYING AND IMPROVING XYLOSE UPTAKE AND UTILIZATION IN ESCHERICHIA COLI
Open Access
- Author:
- Khankal, Reza
- Graduate Program:
- Chemical Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- June 03, 2009
- Committee Members:
- Andrew Michael Cirino, Dissertation Advisor/Co-Advisor
Patrick C Cirino, Committee Chair/Co-Chair
Costas D Maranas, Committee Member
William O Hancock, Committee Member
Andrew Zydney, Committee Member - Keywords:
- XYLOSE UPTAKE
ESCHERICHIA COLI
XYLITOL PRODUCTION
XYLE
XYLFGH
POLYOLS - Abstract:
- ABSTRACT Efficient microbial conversion of biomass into renewable fuels and value-added chemicals remains an important goal in biotechnology. Substrate cost is still a major factor in bioproduction of fuel and chemicals. Utilization of lower-valued substrates such as lignocellulose can reduce the cost of bioprocesses thereby enabling them to compete with traditional chemical synthesis. Xylose, which is the second most abundant sugar in nature and a major constituent of hemicellulose in lignocellulosic biomass and wastes, can play an important role in this contest. Escherichia coli is capable of utilizing a wide range of sugars as carbon and energy sources and producing native metabolites and non-native compounds. In E. coli, xylose uptake occurs primarily through a high-affinity, ATP-binding cassette transporter (XylFGH), although a second, low-affinity proton symporter (XylE) is also present. The efficiency of xylose utilization in this organism is therefore suboptimal due to energetic requirements for xylose uptake. Escherichia coli W3110 was previously engineered to co-utilize glucose and xylose by replacing the wild-type crp gene with a crp* mutant encoding a cAMP-independent CRP variant. Subsequent deletion of the xylB gene (encoding xylulokinase) and expression of xylose reductase from Candida boidinii (CbXR) resulted in a strain which produces xylitol from glucose-xylose mixtures. Our results show that although xylose is negligibly metabolized by wild-type E. coli W3110 in the presence of glucose (classic diauxic growth), xylose transport and xylitol production do occur in the presence of glucose in wild-type E. coli expressing XR, and xylitol production is significantly improved in a mutant strain expressing cAMP-independent CRP (CRP*). These results indicate that either the native xylose transporters are not tightly controlled by CRP or additional transport mechanisms exist. Finally, growth on xylose by E. coli W3110 strains with deletions in both native xylose transporters is recovered to nearly that of wild-type using high concentrations of xylose (~100 mM), demonstrating that xylose is transported by at least one other, low-affinity system. Accurate modeling of xylose metabolism in E. coli in the presence of high concentrations of xylose requires a more thorough understanding of xylose uptake mechanisms and their energy requirements. Conversion of xylose to xylitol (which is secreted) provides a measure of xylose transport without requiring xylose metabolism. This system allows studies of how different native and heterologous transporters (e.g., deleted or overexpressed) influence xylose uptake under various conditions (e.g. in the presence of glucose) and not coupled to growth or expression of genes specific to xylose metabolism. We have examined the contributions of the native E. coli xylose transporters (the D-xylose/proton symporter XylE and the D-xylose ABC transporter XylFGH) and CRP* to xylitol production in the presence of glucose and xylose. The final batch xylitol titer with strain PC09 (ƒ´xylB and crp*) is reduced by 40% upon deletion of xylG and by 60% upon deletion of both xyl transporters. This shows that, up to 40% of xylose transport occurs via secondary transport. This uptake is apparently not affected by individual deletions in other transporters known to be promiscuous. Xylitol production by the wild-type strain (W3110) expressing CbXR is not reduced when xylE and xylG are deleted, demonstrating tight regulation of the xylose transporters by CRP and revealing significant secondary xylose transport. Finally, plasmid expression of XylE or XylFGH with CbXR in PC07 (ƒ´xylB and wild-type crp) growing on glucose results in xylitol titers similar to that achieved with PC09 and provides an alternative strategy to the use of CRP*. Studies that correlate anaerobic growth of engineered strains on xylose to energy yield indicate that secondary xylose transport is largely energy dependent (not diffusion). E. coli K-12 strains W3110 and MG1655 and wild-type E. coli B were compared as platforms for xylitol production from glucose-xylose mixtures using either plasmid-based expression of a xylose transporter (XylE or XylFGH) or replacing the native crp gene with a mutant (crp*). The engineered strains were compared in fed-batch fermentations and as non-growing resting cells. Expression of CRP* in the E. coli B strains tested was unable to enhance xylose uptake in the presence of glucose. Xylitol production was similar for the (crp*, ƒ´xylB)-derivatives of W3110 and MG1655 expressing CbXR (average specific productivities of ~0.43 g xylitol g cdw-1 h-1 in fed-batch fermentation). In contrast, results varied substantially between different ƒ´xylB-derivative strains coexpressing either XylE or XylFGH. Thus, the differences in genetic background between these strains can profoundly influence metabolic engineering strategies. While previous studies have examined the effects of expressing CRP* mutants on the expression of specific catabolic genes, little is known about the global transcriptional effects of CRP* expression. We also compare the transcriptome of E. coli W3110 (expressing wild-type CRP) to that of mutant strain PC05 (expressing CRP*) in the presence and absence of glucose. While the simplest model of CRP*-mediated gene expression assumes insensitivity to glucose (or cAMP), our results show that gene expression in the context of CRP* is very different from that of wild-type in the absence of glucose, and is influenced by the presence of glucose. 238 genes were found to respond differently to glucose in PC05 compared to W3110. Genes whose expression is significantly altered by glucose in strain W3110 are generally not altered to the same extent in strain PCO5 and the glucose effect is significantly suppressed in this strain. We present a detailed transcription analysis and relate these results to phenotypic differences between strains expressing wild-type CRP and CRP*. Notably, CRP* expression in the presence of glucose, results in an elevated intracellular NADPH concentration and reduced NADH concentration relative to wild-type CRP. Meanwhile, a more drastic decrease in the NADPH/NADP+ ratio is observed for the case of CRP* expression in strains engineered to reduce xylose to xylitol via NADPH-dependent xylose reductase during glucose metabolism.