Engineering microbial consortia: a new frontier in synthetic biology Katie Brenner 1 , Lingchong You 2 and Frances H. Arnold 1 1 Division of Chemistry and Chemical Engineering, California Institute of Technology 210-41, Pasadena, CA 91125, USA 2 Department of Biomedical Engineering and the Institute for Genome Sciences and Policy (IGSP), Duke University, Center for Interdisciplinary Engineering, Medicine & Applied Sciences (CIEMAS) 2345, Durham, NC 27708, USA Microbial consortia are ubiquitous in nature and are implicated in processes of great importance to humans, from environmental remediation and wastewater treat- ment to assistance in food digestion. Synthetic biol- ogists are honing their ability to program the behavior of individual microbial populations, forcing the microbes to focus on specific applications, such as the production of drugs and fuels. Given that microbial consortia can perform even more complicated tasks and endure more changeable environments than monocultures can, they represent an important new frontier for synthetic biology. Here, we review recent efforts to engineer syn- thetic microbial consortia, and we suggest future appli- cations. Benefits and features of microbial consortia Synthetic biology [1–5] has generated many examples of what microbes can do and what we can learn from them [6–11] when they are creatively engineered in the labora- tory environment. From the synthesis of an anti-malarial drug [12] to the study of microbial genetic competency [13], engineered microbes have advanced technology while providing insight into the workings of the cell. Interest has recently emerged in engineering microbial consortia – multiple interacting microbial populations – because consortia can perform complicated functions that individual populations cannot and because consortia can be more robust to environmental fluctuations (Figure 1). These attractive traits rely on two organizing features. First, members of the consortium communicate with one another. Whether by trading metabolites or by exchan- ging dedicated molecular signals, each population or individual detects and responds to the presence of others in the consortium [14]. This communication enables the second important feature, which is the division of labor; the overall output of the consortium rests on a combi- nation of tasks performed by constituent individuals or sub-populations. Here, we briefly examine the complex functions that mixed populations perform, and the evi- dence for their robustness to environmental fluctuation. We then explore how engineers have employed communi- cation and differentiation of function in designing syn- thetic consortia, and we comment on their future applications. Mixed populations can perform complex tasks Mixed populations can perform functions that are difficult or even impossible for individual strains or species. Bal- ancing two or more tasks so that they are efficiently completed within one organism can pose insuperable challenges in some situations. For example, it is difficult to engineer efficient, metabolically independent path- ways within a single cell to enable it to consume the five- and six-carbon sugars produced by lignocellulose degra- dation; asynchrony in degradation, caused by glucose preference, lowers productivity [15]. These functions, however, can be separated into different, individually optimized populations. By compartmentalizing the mol- ecular components of each pathway, transcriptional regulators and chemical intermediates in each can be modulated separately without regard for potential inter- actions. For example, two strains of Escherichia coli have been engineered so that one metabolizes only glucose and the other only xylose, and can be tuned so that they consume their substrates at similar rates. When grown in co-culture, the two strains ferment the sugars more efficiently than would any single engineered cell perform- ing both functions [16]. Another important feature of microbial consortia is their ability to perform functions requiring multiple steps. Such tasks are possible when different steps are completed by dedicated cell-types. For example, cellulolytic microbes make and excrete several different protein components (e.g. scaffolding proteins and enzymes) that assemble into an extracellular cellulosome that is capable of cellulose degradation. Various organisms in nature can secrete all of the necessary cellulase components, but these organisms are often difficult to manipulate genetically [17]. Attempts to engineer more genetically tractable organisms to secrete all of the cellulase components heterologously have not yet been successful. This might be because the heavy meta- bolic burden associated with expression of the cellulase- associated proteins inhibits cell growth, or because intra- cellular assembly of the cellulosomal complexes interferes with their excretion. However, two engineered strains of Bacillus subtilis – one secreting the scaffold and the other secreting either an endoglucanase or a xylanase that binds to the scaffold to become active – exhibit the predicted enzymatic activity in co-culture [18]. In each of these examples, a combination of populations was used to achieve a desired outcome that is currently difficult to engineer in a single population. Review Corresponding author: Arnold, F.H. (frances@cheme.caltech.edu). TIBTEC-642; No of Pages 7 0167-7799/$ – see front matter ß 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.tibtech.2008.05.004 Available online xxxxxx 1