Microbial Characterization Facility

Principal Investigator: Adam Arkin, Lawrence Berkeley National Laboratory and UC Berkeley
Co-PIs: Terry Hazen, Inna Dubchak
Researchers: Adam Deutschbauer, Jeffrey Skerker, John Prausnitz, Michael Samoilov
Postdocs: Cindy Wu, Anna Gerasimova, Sharon Aviran
Student Research Associates: Jordan Mar, Kelly Wetmore
Faculty Collaborator: Doug Clark
Visiting Scholar: Chris Roberge
Administration: Gwyneth Terry

Bacteria are an attractive alternative to yeast for industrial level fuel ethanol production and perhaps the production of other types of fuels such as lipids and diesel. While, currently, few bacteria are close to the efficiency of fungi in the production of ethanol, recent findings show that rational engineering can greatly increase output. Bacteria like Zymomonas mobilis, a successful model microbe for ethanol production, have been moderately engineered and selected for industrial output of ethanol. However, there have been no systematic genomic or high-throughput genetic studies to elucidate how other metabolic and stress pathways affect the process in bacteria. Rational engineering of bacteria to optimize ethanol production or any fuel molecule requires a thorough, systems-level understanding of the how bacterial metabolism, gene regulation, and stress response influence this process. Our project will develop a facility for these studies and apply them to key biofuel relevant organisms starting with Zymomonas mobilis.

Building on our long expertise in the systems biology of microbes, we are building an experimental and computational comparative microbial systems biology infrastructure that we can apply widely to identify, understand and engineer the central pathways in bacteria and fungi that are important to biomass production and fermentation into biofuels. The approach combines functional genomics in the form of gene expression and metabolic analysis, high throughput genetics, and quantitative dynamic analysis of pathway function using highly controlled and reproducibly perturbed microbial cultures for elucidating the interactions between individual genes, pathways, and the environment. The core experimental technologies involve extensive use of custom high-density tiling microarrays for genome annotation and inference of regulatory elements, custom oligonucleotide microarrays for gene expression, metabolite analysis, and the parallel, quantitative analysis of thousands of sequence defined mutants using a molecular barcoding strategy. In addition to providing genetic resources for any microorganism, our proposed facility will enable automated sample preparation as well as automated, high-throughput microbial experiments (using liquid handling robotics and microplate readers). These experimental efforts will be supported by the development of a sophisticated computational infrastructure whose beginnings can be found in the MicrobesOnline (http://microbesonline.org) and RegTransBase (http//regtransbase.lbl.gov). For this project, these tools will be deepened with more genomes, better metabolic and stress gene annotation, regulatory network curation and inference, and functional genomic analysis tools, and tools for metabolic inference and engineering. Together, the facility will generate a quantitative, systems-level view of microbial physiology that will serve as the blueprint for the rational engineering of these organisms to meet bioenergy challenges.

Our first target is to use this framework to improve ethanol production in Zymomonas mobilis. Until very recently, the only commercial fermentations have relied on yeast and other super-fungi that have been identified as excellent producers of ethanol from C5 sugars and other carbon sources. However, for some time it has been recognized that the facultative anaerobe Zymomonas mobilis can have a higher yield and faster specific rate of ethanol production when compared to yeast. In addition, yeast has a higher aeration cost, a high biomass production, and low temperature and ethanol tolerance compared to Zymomonas. Its yield of ethanol from sugar, derived from the Entner-Doudoroff pathway, is around 96%, the same as yeast. Because of its industrial potential, DuPont and others have dedicated themselves to scaling up Zymomonas for commercial fermentation of lignocellulosic residues such as corn stover. Zymomonas has already, for over a decade, been a target of metabolic engineering for utilization of diverse C5 sugars such as xylose and arabinose that are present in lignocellulosic hydrolysates.

A number of components in hydrolysates can inhibit the growth and ethanol production of bacteria and yeast. While in Z. mobilis the major inhibitor is usually acetate, other products such as vanillin, syringaldehyde, hydroxymethyl-furfural and furfural all show significant inhibitory effects. These inhibitors, and ethanol itself, have differential effects on growth and on ethanol production and have different behaviors depending on the sugars being utilized. In fact, the osmotic stress of sugars themselves can have a strong affect on the productivity of fermentation and different sugars have different uptake rates and saturation points. Further, the growth mode of Z. mobilis can have a strong effect on its productivity. Immobilized and suspended cultures have very different productivities and the use of flocculent cultures can increase volumetric productivity by as much as ten-fold. Efforts to optimize Z. mobilis for production are limited because a detailed understanding of all the physiological factors affecting metabolism including stress responses and inessential secondary metabolic pathways are not currently available. Identifying the genetic factors that contribute to these effects is a main aim of this proposal.

Despite availability of both the genome and a developed genetic toolbox with plasmids, a promoter system, and conjugal shuttle vectors capable of delivering transposons from E. coli, there has not yet been a large-scale genomic analysis of Zymomonas physiology. Consequently, the factors affecting key metabolisms and behaviors for ethanol production and tolerances to inhibitors in the feedstock and in its products are unknown. In doing so, we hope to elucidate engineering principles to improve ethanol production from lignocellulosic sources by Z. mobilis and perhaps other species.

2009 Program Update:
This program is focused on the application of high-throughput functional genomics and phenotypic analysis of microbes relevant to bioenergy, and on the development of web-based software tools for data management, retrieval and analysis, served via the MicrobesOnline comparative genomics workbench (http://microbesonline.org). Using the ethanol-producing bacteria Zymomonas mobilis as a testbed, we have completed the construction of a "bar-coded" transposon library and carried out a high-throughput 96-well growth screen to determine the inhibitory dose range of purified hydrolysate inhibitors and fuel molecules. By mapping 14,009 transposon insertions by arbitrary PCR, a pooled transposon library that contains insertions in 1,695 different genes was created. This provides coverage of most non-essential genes in the Zymomonas genome. To complement mutant studies, high-density Nimblegen arrays for both gene expression and transcript mapping were developed. As an alternative to tiling arrays, we are exploring the use of next-generation sequencing to map transcript boundaries.

To facilitate downstream pathway and metabolic engineering of Zymomonas, we have developed better genetic tools for gene replacement by homologous recombination and plasmids for complementation and over-expression. In 2010 we expect to carry out pooled fitness experiments using the full transposon library and to obtain transcription profiles of wild type Zymomonas grown in the presence of key hydrolysate inhibitors and other fuel compounds. Integrating our fitness data with gene expression results will facilitate our understanding of tolerance in Zymomonas and should lead to hypotheses for subsequent engineering of improved strains. We also plan to initiate a comparative genomics project by sequencing related Zymomonas species and other closely related alpha-proteobacteria using next-generation sequencing and de novo genome assembly to create better annotation of the key metabolic and regulatory processes that underlie tolerance and production.