Conferring cellulose-degrading ability to Yarrowia lipolytica to facilitate a consolidated bioprocessing approach

Zhong-peng Guo, Sophie Duquesne, Sophie Bozonnet, Gianluca Cioci, Jean-Marc Nicaud, Alain Marty, Michael Joseph O’Donohue

Yarrowia lipolytica, one of the most widely studied "nonconventional" oleaginous yeast species, is unable to grow on cellulose. Recently, we identified and overexpressed two endogenous β-glucosidases in Y. lipolytica, thus enabling this yeast to use cello-oligosaccharides as a carbon source for growth. Using this engineered yeast platform, we have now gone further toward building a fully cellulolytic Y. lipolytica for use in consolidated bioprocessing of cellulose. Initially, different essential enzyme components of a cellulase cocktail (i.e,. cellobiohydrolases and endoglucanases) were individually expressed in Y. lipolytica in order to ascertain the viability of the strategy. Accordingly, the Trichoderma reesei endoglucanase I (TrEG I) and II (TrEG II) were secreted as active proteins in Y. lipolytica, with the secretion yield of EG II being twice that of EG I. Characterization of the purified His-tagged recombinant EG proteins (rhTrEGs) revealed that rhTrEG I displayed higher specific activity than rhTrEG II on both cellotriose and insoluble cellulosic substrates, such as Avicel, β-1, 3 glucan, β-1, 4 glucan, and PASC. Similarly, cellobiohydrolases, such as T. reesei CBH I and II (TrCBH I and II), and the CBH I from Neurospora crassa (NcCBH I) were successfully expressed in Y. lipolytica. However, the yield of the expressed TrCBH I was low, so work on this was not pursued. Contrastingly, rhNcCBH I was not only well expressed, but also highly active on PASC and more active on Avicel (0.11 U/mg) than wild-type TrCBH I (0.065 U/mg). Therefore, work was pursued using a combination of NcCBH I and TrCBH II. The quantification of enzyme levels in culture supernatants revealed that the use of a hybrid promoter instead of the primarily used TEF promoter procured four and eight times more NcCBH I and TrCBH II expressions, respectively. Finally, the coexpression of the previously described Y. lipolytica β-glucosidases, the CBH II, and EG I and II from T. reesei, and the N. crassa CBH I procured an engineered Y. lipolytica strain that was able to grow both on model cellulose substrates, such as highly crystalline Avicel, and on industrial cellulose pulp, such as that obtained using an organosolv process. A Y. lipolytica strain coexpressing six cellulolytic enzyme components has been successfully developed. In addition, the results presented show how the recombinant strain can be optimized, for example, using artificial promoters to tailor expression levels. Most significantly, this study has provided a demonstration of how the strain can grow on a sample of industrial cellulose as sole carbon source, thus revealing the feasibility of Yarrowia-based consolidated bioprocess for the production of fuel and chemical precursors. Further, enzyme and strain optimization, coupled to appropriate process design, will undoubtedly lead to much better performances in the future.