Metabolic engineering of the oleaginous yeast Yarrowia lipolytica PO1f for production of erythritol from glycerol

Background Sugar alcohols are widely used as low-calorie sweeteners in the food and pharmaceutical industries. They can also be transformed into platform chemicals. Yarrowia lipolytica, an oleaginous yeast, is a promising host for producing many sugar alcohols. In this work, we tested whether heterologous expression of a recently identified sugar alcohol phosphatase (PYP) from Saccharomyces cerevisiae would increase sugar alcohol production in Y. lipolytica. Results Y. lipolytica was found natively to produce erythritol, mannitol, and arabitol during growth on glucose, fructose, mannose, and glycerol. Osmotic stress is known to increase sugar alcohol production, and was found to significantly increase erythritol production during growth on glycerol. To better understand erythritol production from glycerol, since it was the most promising sugar alcohol, we measured the expression of key genes and intracellular metabolites. Osmotic stress increased the expression of several key genes in the glycerol catabolic pathway and the pentose phosphate pathway. Analysis of intracellular metabolites revealed that amino acids, sugar alcohols, and polyamines are produced at higher levels in response to osmotic stress. Heterologous overexpression of the sugar alcohol phosphatase increased erythritol production and glycerol utilization in Y. lipolytica. We further increased erythritol production by increasing the expression of native glycerol kinase (GK), and transketolase (TKL). This strain was able to produce 27.5 ± 0.7 g/L erythritol from glycerol during batch growth and 58.8 ± 1.68 g/L erythritol during fed-batch growth in shake-flasks experiments. In addition, the glycerol utilization was increased by 2.5-fold. We were also able to demonstrate that this strain efficiently produces erythritol from crude glycerol, a major byproduct of the biodiesel production. Conclusions We demonstrated the application of a promising enzyme for increasing erythritol production in Y. lipolytica. We were further able to boost production by combining the expression of this enzyme with other approaches known to increase erythritol production in Y. lipolytica. This suggest that this new enzyme provides an orthogonal route for boosting production and can be stacked with existing designs known to increase sugar alcohol production in yeast such as Y. lipolytica. Collectively, this work establishes a new route for increasing sugar alcohol production and further develops Y. lipolytica as a promising host for erythritol production from cheap substrates such as glycerol.

Application of modified Logistic and Monod models in a single equations system framework to improve cell culture growth modeling and estimation

In modeling cell culture growth using unstructured model, two types of equations are normally used: logistic and Monod. However, these two equations are known for their limitations to model death phase of cell culture growth and to account for dead cells accumulation data. In this paper, we present a modeling framework whereby both Logistic and Monod equations can be used in a single set of equations system to overcome these limitations. First, it can be shown that the increase of total cell population that consists of viable and dead cells follows a logistic growth pattern with its own intrinsic growth rate and total carrying capacity. Furthermore, a hybrid Logistic-Monod equation with first-order decay kinetics can be used to model viable cell growth data with decline phase effectively. With this paradigm, a pseudo-rate equation can be written to account for dead cells accumulation data using population balancing with a simple understanding that dead cell population is simply the difference between total and viable cells. These equations can be adjoined with substrate consumption and product generation rate equations to depict complete batch growth data that covers exponential growth and death phases. This modeling framework has been fitted successfully to fit batch growth data of two cell lines from published literature with complete depictions of dead cell accumulation and cell viability profiles. The implication of this modeling framework for chemostat culture performance analysis is further investigated.