Engineering xylose metabolism in triacylglycerol-producing Rhodococcus opacus for lignocellulosic fuel production

Abstract

Background: There has been a great deal of interest in fuel productions from lignocellulosic biomass to minimize the conflict between food and fuel use. The bioconversion of xylose, which is the second most abundant sugar present after glucose in lignocellulosic biomass, is important for the development of cost effective bioprocesses to fuels. Rhodococcus opacus PD630, an oleaginous bacterium, accumulates large amounts of triacylglycerols (TAGs), which can be processed into advanced liquid fuels. However, R. opacus PD630 does not metabolize xylose.

Results: We generated DNA libraries from a Streptomyces bacterium capable of utilizing xylose and introduced them into R. opacus PD630. Xsp8, one of the engineered strains, was capable of growing on up to 180 g L-1 of xylose. Xsp8 grown in batch-cultures derived from unbleached kraft hardwood pulp hydrolysate containing 70 g L-1 total sugars was able to completely and simultaneously utilize xylose and glucose present in the lignocellulosic feedstock, and yielded 11.0 g L-1 of TAGs as fatty acids, corresponding to 45.8% of the cell dry weight. The yield of total fatty acids per gram of sugars consumed was 0.178 g, which consisted primarily of palmitic acid and oleic acid. The engineered strain Xsp8 was introduced with two heterologous genes from Streptomyces: xylA, encoding xylose isomerase, and xylB, encoding xylulokinase. We further demonstrated that in addition to the introduction and the concomitant expression of heterologous xylA and xylB genes, there is another molecular target in the R. opacus genome which fully enables the functionality of xylA and xylB genes to generate the robust xylose-fermenting strain capable of efficiently producing TAGs at high xylose concentrations.

Conclusion: We successfully engineered a R. opacus strain that is capable of completely utilizing high concentrations of xylose or mixed xylose/glucose simultaneously, and substantiated its suitability for TAG production. This study demonstrates that the engineered strain possesses a key trait of converters for lipid-based fuels production from lignocellulosic biomass.

Figure 1

Growth kinetics of R. opacus Xsp8 on high xylose concentrations in flask cultures. Xylose concentrations of modified defined media with 1.4 g L^-1 (NH₄)₂SO₄ and 10 mg L^-1 gentamicin were 40, 120, 160, 180 and 200 g L^-1. Initial inoculum densities were adjusted photometrically to obtain an OD₆₆₀ of 1.0. Values and error bars represent the mean and s.d. of triplicate experiments.

Figure 2

Optimization of TAG production from xylose by R. opacus Xsp8 in batch-culture fermentations. (a) Central composite experimental design matrix defining xylose and (NH₄)₂SO₄ concentrations. X₁, xylose concentration (g L^-1); X₂, (NH₄)₂SO₄ concentration (g L^-1). The strain was inoculated in the modified defined medium supplemented with gentamicin in Sixfors-bioreactors. The data for TAG production as fatty acids represent the maximum values during 10 days of cultivation. (b) Response surface plot of the effect of xylose and (NH₄)₂SO₄ concentrations on fatty acid production. Curves and points represent predicted values and experimental data, respectively.

Figure 3

TAG production from xylose and/or glucose by R. opacus Xsp8. (a-c) Time course kinetics of TAG production as fatty acids. The strain was inoculated in modified defined medium containing 16 g L^-1 xylose (a), a mixture of 8 g L^-1 xylose and 8 g L^-1 glucose (b), or 16 g L^-1 glucose (c) with 1 g L^-1 (NH₄)₂SO₄ and gentamicin in flasks. Values and error bars represent the mean and s.d. of triplicate experiments. (d) Fatty acids composition profile as % of total fatty acids (g/g) of R. opacus growing in the defined medium containing xylose (a), xylose and glucose mix (b), or glucose (c) for 4 days. Data are results of triplicate experiments, ±s.d.

Figure 4

TAG production from unbleached kraft hardwood pulp hydrolysate by R. opacus Xsp8 in batch-culture fermentations. (a) Saccharification of unbleached kraft hardwood pulp using commercially available enzymes. Eighty grams of dried pulp were suspended in 1 liter deionized water and adjusted to a pH 5.0 with 0.1N HCl. Four enzymes, 2 g of Pentopan and 10 ml each of Viscozyme, Celluclast and Novozyme 188, were added into the suspension. After 4, 8 and 12 h of duration, additional 80 g of the pulp, 2 g of Pentopan and 10 ml each of the three enzymes were mixed into the suspension three times. Values and error bars represent the mean and s.d. of triplicate experiments. (b) Time course kinetics of TAG production as fatty acids on the pulp hydrolysate in bioreactor cultivations. Xsp8 was inoculated in the saccharified solution containing 70 g L^-1 of total sugars composed of 47 g L^-1 glucose, 17 g L^-1 xylose and 6 g L^-1 other sugars supplemented with 3.47 g L^-1 (NH₄)₂SO₄, mineral components of defined medium and gentamicin, at an initial OD₆₆₀ of 0.5. Values and error bars represent the mean and s.d. of triplicate experiments. (c) Fatty acids composition profile as % of total fatty acids (g g^-1) of TAGs from Xsp8 cells growing in the saccharified solution for 7 days. Data are results of triplicate experiments, ±s.d.

Figure 5

Identification of the genes derived from S. padanus involved in improving xylose metabolism in R. opacus PD630. (a) Physical map of the 3603-bp insert on pXsp8. The numbers and arrows in the ORF boxes indicate those of the amino acid residue and the locations of ORFs, respectively. (b-d) Pairwise identities of amino acid sequences using the CLUSTALW program. (b) The ORF 1 (388 aa) and xylose isomerases of S. hygroscopicus accession no. AEY87893 (388 aa), S. avermitilis accession no. NP_828358 (388 aa), S. chartreusis accession no. ZP_09957234 (388 aa) and S. coelicolor accession no. NP_625460 (387 aa). (c) Xylulose kinase domain (XKD, 410 aa) of the ORF 2 and xylulose kinases of S. hygroscopicus accession no. AEY87894 (481 aa), S. avermitilis accession no. NP_828357 (481 aa), S. chartreusis accession no. ZP_09957235 (481 aa) and S. coelicolor accession no. NP_733525 (481 aa). (d) Cellulose-binding domain (CBD, 235 aa) of the ORF 2 and cellulose-binding proteins of S. hygroscopicus accession no. AEY91743 (311 aa), S. avermitilis accession no. NP_824035 (312 aa), S. chartreusis accession no. ZP_09956458 (311 aa) and S. coelicolor accession no. NP_629535 (310 aa).

Figure 6

Elucidation of the molecular targets involved in improving xylose metabolism in R. opacus PD630. (a,b) Colonies arising from transformation of various plasmids (pX0 which carries no gene, pX1 which carries xylA alone, pX2 which carries xylB alone, pX3 which carries xylA and xylB, and pX4 which carried xylA and xylB without cellulose-binding domain) into R. opacus competent cells. Each plasmid was electroporated into the competent cells of wt PD630 (a) or pXsp8-plasmid cured Xsp8C (b). The pulsed cells were plated onto defined medium containing 16 g L^-1 xylose, 1 g L^-1 (NH₄)₂SO₄ and spectinomycin, and incubated for 14 days. (c-f) Fermentation kinetics of R. opacus derivatives. Each strain was inoculated in defined medium containing 16 g L^-1 xylose, 1 g L^-1 (NH₄)₂SO₄ and spectinomycin at an initial OD₆₆₀ of 0.1 in flask cultures. (c,d) Growth kinetics of PD630 derivatives (c) and Xsp8 derivatives (d). (e,f) TAG production of PD630 derivatives (e) and Xsp8 derivatives (f) after 6 days of cultivation. Data represent the mean and s.d. of the six strains (n=1).