Discussion and Outlook
The field of synthetic biology has grown remarkably in the past two decades, and it has made substantial progress in engineering gene regulatory and metabolic networks in microorganisms. In recent years the field has put a spotlight on engineering microbial consortia (Brenner et al., 2008), with a focus again on their composition and genetic make up. Notwithstanding the many successful examples we cite above, and the many others we have surely missed, there is much room to be explored in engineering the environment where microorganisms grow. Among the most promising directions we would like to highlight is the development of evolutionary engineering approaches. While the examples we discuss above were executed in parallel, evolutionary engineering of the environment can also be implemented in time, where multiple environmental changes are introduced serially and the best environment is retained. The development of highly controllable milifluidic devices for continuous culture of microorganisms offer a promising platform for identifying optimal culture conditions for organisms and communities (Mancuso et al., 2021; Wong et al., 2018).
After a slow start, the field of artificial community-level selection is experiencing growing attention (Blouin et al., 2015; Chang et al., 2021; Doulcier et al., 2020, 2020; Panke-Buisse et al., 2015; Swenson et al., 2000; Vessman et al., 2023; Williams and Lenton, 2007; Xie et al., 2019; Xie and Shou, 2021). There exist obvious parallels between the process of optimizing the composition of an inoculum in a fixed environment, and that of optimizing an environment for a fixed inoculant. These may lead to a dialog between both fields, allowing the transfer of successful methodologies from one to another. Encouragingly, a wealth of novel evolutionary algorithms are being developed for the directed evolution of microbial communities (Chang et al., 2021; Vessman et al., 2023; Xie et al., 2019). These could be fruitful when applied to finding optimal combinations of environmental factors.
In a bottom-up manner, genome-scale metabolic models have been used for some time to rationally manipulate microbial interactions and to predict environments where microorganisms may coexist (Harcombe et al., 2014; Klitgord and Segrè, 2010), and they have been used to find culture conditions that optimize the growth of single strains or the production of target molecular products (e.g. (Swayambhu et al., 2020)). Constraint based models are being developed to quantitatively predict the behavior of microbial consortia (e.g. see (Heinken et al., 2021) and references therein), paving the way for their application to finding optimal environments.
Rather than being two parallel pursuits, bottom-up and top-down approaches can actually benefit from one another. By gaining a deeper understanding of the topology of the map (i.e. the response surface) between environment and function, researchers will be better equipped to design evolutionary algorithms that are capable of efficiently navigating those maps and finding optimal culture conditions. We hope that our review will stimulate efforts on this front.