Our current research falls into one of two main categories:

• Retrobiosynthesis for the production of organic compounds
• Optimization and understanding of gene dosage effects in metabolic engineering

Retrobiosynthesis for the production of organic compounds

We are interested in combining the traditions of metabolic engineering with the practices of biocatalysis to construct novel metabolic pathways for the production of non-natural compounds (small molecules) in microbial hosts. Metabolic engineering has produced many examples of successful and optimized production of biochemicals, whether such products come from existing pathways in the original organism, for example, improving lysine production in bacteria, or are "new," i.e., not normally produced in the host organism. Such products may be naturally-occurring in other organisms and only produced in the target organism following transfer of the metabolic pathway from the source to target (e.g., carotenoid production in E. coli), or the products themselves may be novel, structurally distinct from those found in nature. Of particular interest is the creation of novel pathways for new product formation. A novel pathway is defined as a sequence of conversions from a designated substrate to a designated product that is not natural in the host organism. In this definition, each individual conversion step need not be new for the sequence as a whole to be considered novel. Novel pathways have most often been assembled in two ways. First, a novel pathway may be constructed by combining two or more existing pathways from different organisms. This method produces a compound that is un-natural in the target host. Second, an existing pathway can be modified and/or extended through the introduction of new activities to produce compounds that are both un-natural and structurally distinct from those products of the original pathway. Our goal is to create biosynthetic pathways for un-natural (and possibly novel) production formation by combining a series of enzymatic reactions that are not known to exist in sequence in nature for the formation of a designated product.

To do so, we aim to build upon the principles employed in the field of biocatalysis. Generally speaking, production of any chemical compound through biological means can be considered biocatalysis, which would encompass nearly all aspects of metabolic engineering. However, the term is traditionally used in a more restrictive manner to describe enzymatic or whole cell single-reaction conversions of an externally-supplied substrate. It is especially useful for the synthesis of chiral compounds in the fine chemical and pharmaceutical industries. While a biocatalytic transformation is typically employed as an alternative to a single step of an organic chemical synthesis route; it is usually done in the context of the full synthesis, which relies on the organic chemistry principles of retrosynthesis. As such, biocatalysis is product-focused (though the starting substrate is usually pre-determined from the overall synthesis scheme), and the products formed are usually non-natural, produced from non-natural substrates. The choice of enzymes is based only on consideration of the required chemical conversion. Using a "retrobiosynthetic" approach and existing databases of known biochemical reactions, we propose pathways by analyzing the desired product to determine a series of enzyme-catalyzed steps that are likely to result in its formation. We wish to develop a methodology whereby a number of potential starting compounds could be utilized, with the objective of choosing reactants that exist as cellular metabolites or are commonly used microbial carbon sources (e.g., glucose) such that the ultimate pathway can be linked to basic metabolism. The final configuration of these production systems is envisioned as custom-designed "microbial factories" for organic chemical production.

back to top

Optimization and understanding of gene dosage effects in metabolic engineering

We are also interested in the optimal manner in which metabolic pathways should be constructed with respect to gene dosage levels and the choice of starting substrates. One approach towards metabolic engineering is to overlay traditional genetic engineering for the introduction of recombinant genes into a desired host organism with a systems-based methodology for the analysis of the resulting cellular phenotypes. This approach has proven successful in allowing one to understand and control the impact of changes of enzyme levels, and hence flux through the associated pathways, within a variety of recombinant systems. However, in using the framework of metabolic engineering to more fully describe the state of a recombinant cell, many researchers have neglected full consideration of the impact of the means by which the genetic modifications are made to arrive at the final cell producing the product of interest. The most common approach involves the use of readily available, high-copy-number plasmids for the introduction of new or enhancement of existing enzyme functions in the host cell. Yet the use of such plasmids is well-known to carry with it the additional burden of maintaining the plasmid DNA and expressing its encoded genes, a burden that may be reflected in a re-organization of the host cell's basic metabolism. Therefore, the optimal cell for production purposes may require that recombinant gene copy numbers be limited to balance enzyme levels and activity against gene levels. Conversely, if metabolites involved in basic host maintenance and propagation are required substrates for the final product of interest, the gene levels required to effectively divert such substrates towards alternative pathways may be quite high. Those scientists advocating a "minimal-genome" approach towards the construction of novel microorganisms with unique production properties may therefore neglect to consider such limitations in defaulting towards a single-copy system for the expression of heterologous genes.

Our interest is in characterizing "metabolically engineered" cells in order to define the recombinant pathway characteristics that necessarily require higher gene dosages for adequate expression and those that benefit from lower gene dosages for maximum productivity. This can lead to better predictability of the best strategies for optimization of recombinant, i.e., transferred, pathways, providing genetic "design rules" for metabolic pathway engineering. The knowledge gained will also provide useful insights into which substrates are more easily diverted from their intended pathways towards product formation (i.e., result in high productivity from a low gene dosage) and would likely be better suited for use as starting reagents in our novel pathways assembled from the principles of retrobiosynthesis. The choice of appropriate expression levels and the production of particular metabolites are also being considered in the context of industrially relevant production conditions (e.g., high cell density fermentations, growth in defined medium).

back to top