Removal of aromatic inhibitors produced from lignocellulosic hydrolysates by Acinetobacter baylyi ADP1 with formation of ethanol by Kluyveromyces marxianus

Abstract

Background

Lignocellulosic biomass is an attractive, inexpensive source of potentially fermentable sugars. However, hydrolysis of lignocellulose results in a complex mixture containing microbial inhibitors at variable composition. A single microbial species is unable to detoxify or even tolerate these non-sugar components while converting the sugar mixtures effectively to a product of interest. Often multiple substrates are metabolized sequentially because of microbial regulatory mechanisms. To overcome these problems, we engineered strains of Acinetobacter baylyi ADP1 to comprise a consortium able to degrade benzoate and 4-hydroxybenzoate simultaneously under batch and continuous conditions in the presence of sugars. We furthermore used a thermotolerant yeast, Kluyveromyces marxianus, to convert the glucose remaining after detoxification to ethanol.

Results

The two engineered strains, one unable to metabolize benzoate and another unable to metabolize 4-hydroxybenzoate, when grown together removed these two inhibitors simultaneously under batch conditions. Under continuous conditions, a single strain with a deletion in the gcd gene metabolized both inhibitors in the presence of sugars. After this batch detoxification using ADP1-derived mutants, K. marxianus generated 36.6 g/L ethanol.

Conclusions

We demonstrated approaches for the simultaneous removal of two aromatic inhibitors from a simulated lignocellulosic hydrolysate. A two-stage batch process converted the residual sugar into a non-growth-associated product, ethanol. Such a two-stage process with bacteria (A. baylyi) and yeast (K. marxianus) is advantageous, because the yeast fermentation occurs at a higher temperature which prevents growth and ethanol consumption of A. baylyi. Conceptually, the process can be extended to other inhibitors or sugars found in real hydrolysates. That is, additional strains which degrade components of lignocellulosic hydrolysates could be made substrate-selective and targeted for use with specific complex mixtures found in a hydrolysate.