What is industrial microbiology ? Microorgansims and other cell types, such as plant and animal cells, see widespread use as cellular factories, or biocatalysts, that are critical for making a wide variety of products. These include low-margin chemicals such biofuels, biorenewable chemicals and animal feed to high-margin commodities such as pharmaceuticals and cosmetics. Well known cellular factories include the bacterium, Escherichia coli, and the brewer’s yeast, Saccharomyces cerevisiae.

What is the robustness problem ? As with any production process, yield and productivity are critical components of economic success when using biocatalysts to make products. Improving the yield and productivity of cellular factories is an important endeavor and a variety of tools can be brought to bear on the problem. However, the central problem facing those who wish to improve cellular factories is what we call the “robustness problem”. At the heart of this problem is the fact that microorganisms have spent millions of years evolving to perform certain tasks in specific ecological niches and forcing them to make a desired end product under the often-stressful conditions required for economically viable industrial processes causes them to perform poorly. Industrial microbiology has long sought a rapid and vigorous method to improve the performance of cellular factories under real-world industrial conditions.

What is evolutionary optimization ? Agricultural operations have long relied on the breeding—which is really just a targeted method of evolution—to improve the characteristics of domesticated plants and animals. Without a doubt, this has been wildly successful. To see this, one need only look at the tremendous variety of domesticated dog breeds that were derived from the gray wolf. Not surprisingly, industrial microbiologists have also long sought a robust method for improving yield and productivity of cellular factories that is akin to breeding and myriad methods have been developed over the years. However, for much of human history the only method that existed for improving cellular factories was evolutionary optimization via continuous culture, although it is clear that early industrial microbiologists had no idea they were using the method. In evolutionary optimization via continuous culture, cellular factories are continuously grown under relevant industrial conditions, which means that cultures are periodically, grown, diluted and re-grown. This is repeated for thousands of generations and over time, such repetitive cultures facilitate the selection of naturally occurring genetic variants that have become adapted to and optimized for the culture conditions. In this way, microbes with new properties emerge and the key is to make sure that these new properties are compatible with the intended industrial application. Indeed, this is precisely how the brewer’s yeast, Saccharomyces cerevisiae, became so good at fermenting sugar to ethanol — through thousands of years of selection for more productive strains by brewers. Of course, early brewers had little knowledge of microbiology, let alone the process of evolution. More importantly, they lacked the technology to facilitate rapid and robust evolutionary optimization, which is why evolutionary optimization has been widely viewed as being slow and ineffective.

What is required for evolutionary optimization ? The key to evolving cellular factories is the ability to culture them continuously. While establishing a continuous culture of cells is simple, it is practically quite difficult to utilize these cultures to direct rapid and robust evolution. This is because successful evolutionary optimization requires that the method of continuous culture meet the following criteria :

• First, the culture must have a sufficiently large population size. More cells means more genetic diversity upon which to act.

• Second the cultures must be growing at their maximal growth rate to ensure more generations per unit time.

• Third, the ability to monitor population dynamics in real time is crucial in order to know when to increase selective pressure to accelerate evolution or decrease selective pressure as needed or prevent collapse of the population.

• Fourth, evolution acts on genetic diversity within a population. Each cell contains unique mutations that are spontaneously accumulated through natural errors in the machinery that replicates DNA. Thus, evolutionary optimization can be accelerated by increasing the accumulation of these mutations and the ideal method of continuous culture should incorporate a simple method of achieving this.

Evolutionary optimization via continuous culture is an old technology. And, considering that all of the tremendous diversity in nature arose through evolution, why isn’t this powerful technology more widely used ? Clearly, microbiologists and brewers have been unknowingly using evolutionary optimization to produce better cellular factories for millennia. Moreover, following the development of the theory of evolution by Darwin and Wallace in the mid-1800s, microbiologists quickly began applying to theory for the alteration of microbial properties. The earliest published example of experimental evolution via serial batch culture was published by Dallinger in 1878. Since then, interest in the use of serial batch culture to experimentally evolve microorganisms grew. However, it was recognized early that serial batch culture as a means of maintaining a continuous culture could not meet the criteria needed for facilitating rapid and robust evolution. Two main problems plague this methodology. The first is that it is difficult and often impossible to monitor population dynamics in real time, meaning that selective pressure cannot be consistently applied when it is needed most nor can it be rapidly dialed back when too much pressure is applied. The second is that this methodology is ineffective for evolving cellular factories that don’t grow evenly in suspension or on ‘dirty’ real-world substrates that are mixtures of solubles and particulates. Beginning around 1930, researchers began to develop the concept of true continuous culture as a means for replacing serial batch culture, culminating in the disputed invention of the chemostat by competing researchers. In continuous culture, exemplified by the chemostat, fresh substrate is continuously fed into a fermenter while spent medium saturated with cells is removed. The in-flow rate of fresh substrate is matched to the out-flow rate of saturated medium and both are matched to the growth rate of the cellular factory. In theory, continuous culture can maintain cultures with high population sizes for indefinite periods of time. Since the earliest publication of a chemostat-like device in 1930, many improvements and iterations of the concept have been developed. However, they all function on the same premise. Unfortunately, continuous culture never reached its potential and has failed to deliver effective microbes for industrial applications. As with serial batch culture, the reasons for this failure are twofold. First, the methodology is ineffective for evolving cellular factories that don’t grow evenly in suspension or on ‘dirty’ real-world substrates that are mixtures of solubles and particulates. Second, since the out-flow of a chemostat-like device is the road to extinction, cells rapidly acquire mutations that allow them to stick to the surfaces inside the chemostat. This problem, called wall growth or biofilm formation, reduces the ability of the user to monitor population dynamics in real time and biofilms can even shield cells from selective pressure. Unfortunately, this problem can be attenuated but never eliminated. The result is that, despite their potential, there are very few examples of the use of chemostats or related devices for improving cellular factories. The inherent technical difficulties with maintaining serial batch cultures and chemostat-like devices result in the ability to apply weak selective pressure at best. And, with reduced selective pressure, evolution proceeds at a snail’s pace. Still, even with these limitations, evolutionary optimization via serial batch culture or chemostats remained the only tools in the strain developer tool kit until the advent of genetic engineering in the 1970s. With genetic engineering, developers realized they could accelerate ‘evolution’ by artificially moving DNA around within an organism (cisgenesis) and from one organism to another (transgenesis). Each new iteration of genetic engineering, given trendy names like ‘synthetic biology’ and ‘metabolic engineering’, demonstrate the tremendous potential of genetic engineering. And the use of engineering largely eclipsed interest in evolutionary optimization until recently when strain developers began to realize that, despite the enormous potential, genetic engineering cannot address the robustness problem. In fact, genetic engineering is actually a frequent cause of robustness issues. This has led strain developers back to evolutionary optimization. Evolugate has developed new continuous culture technology, called the Evolugator™ that circumvents the traditional problems associated with serial batch culture and chemostats. The Evolugator™ allows for real time monitoring of population dynamics, counter-selects against biofilm formation and can support the culturing of any type of cellular factory (from bacteria to mammalian cells) on any type of substrate. Moreover, the technology can incorporate methods for increasing the rate at which mutations are introduced into a population and is highly compatible with genetic engineering, both of which can greatly accelerate the rate of evolutionary optimization. The Evolugator™ represents the first real breakthrough in the use of continuous culture for evolutionary optimization in decades. Improving natural micro-organisms through experimental evolution is now practical and our technology has reopened a wide field of applications for green chemistry using cellular factories.

How do you position your technology relative to genetic engineering ? There is little doubt as to the power of genetic engineering to accelerate the number and magnitude of genetic changes upon which natural selection can act. However, like all technologies, genetic engineering has its limitations.

• First, microorganisms have spent millions of years evolving exquisitely complex metabolic pathways, making it exceedingly difficult to properly re-engineer new functionality. We simply do not know enough about most traits to be able to predict what types of changes need to be engineered. Moreover, even high-throughput recombinant methods of genetic engineering have difficulty altering complex traits, which may require the acquisition of many low probability mutagenic events. Evolutionary optimization is blind to these concerns. With enough selective pressure and time, experimental evolution can easily access the genetic diversity needed to alter complex and poorly understood traits.

• Second, genetic engineering requires the ability to insert or remove genetic material. For most microbes, the tools required to make this work simply do not yet exist, requiring significant investment in the development of the appropriate molecular biology methods.

• Third, tinkering with one trait via genetic engineering often comes at the expense of other traits, such as growth rate and even successful examples of genetic engineering almost always tend to produce strains that are less robust than the parent strain. To make an analogy, it is a bit like adding or deleting some lines in a piece of software when you did not write the program and are not a master of all the code. Thus, engineered microbes may be able to perform a certain industrial task, but do so too slowly to be an effective biocatalyst from an economic standpoint. Strains produced by experimental evolution are simultaneously optimized for both the desired trait as well as growth rate.

• Fourth, the use of genetic engineering results in strains that are genetically modified organisms (GMOs). Right or wrong, the use of GMOs for industrial purposes is often complicated by resistance from the general public, and this resistance is growing. This is particularly true when the end-use of the cellular factory is the in the production of byproducts that are sold into food and feed markets or when there is the possibility that the GMO will escape into the biosphere. On the other hand, strains produced by evolutionary optimization are naturally occurring genetic variants of the original parent strain—no foreign DNA has been inserted and no endogenous DNA has been deliberately removed. As such, these strains are not legally considered “genetically-modified organisms” (GMOs). This said, our technique can be highly complementary with genetic engineering, which can give evolutionary optimization a much need jumpstart.

Will the strains you evolve revert over time ? Clearly, microbes have spent millions of years adapting to be the way they are. If selective pressure to change is removed from a microbe we have adapted for a particular set of conditions and the microbe is returned to its original environment, it will experience pressure to revert back to its original state. However, long term experimental evolution results in the accumulation of many different genetic changes and their rapid reversion is highly improbable. Thus, it will take time for our microbes to alter phenotype once they are removed from the environment to which we have adapted them. That said, it is important for us to continuously apply selective pressure on our microbes, otherwise we run the risk of undoing the good work we have done.

When you adapt a microbe to a different optimal growth temperature, does it lose its ability to grow at its initial optimal growth temperature ? Does it narrow its range of temperature ? In evolutionary biology, there is a concept called antagonistic pleiotropy which postulates that mutations that are adaptive for a certain set of conditions, may result in less robustness under other conditions. So the short answer is yes, if the difference between starting and final growth temperature is large enough and if the microbe spends enough time learning to be a specialist at a particular temperature, it is possible that a microbe will show reduced ability to grow at the initial temperature. We have not pushed the limits of thermal adaptation and the strains we have produced are still capable of growing at both temperatures, although we do see a drop in robustness at the initial temperature.

Is it possible to predict the time it takes to achieve some specific goal ? Unlike some traditional engineering, IT, or building developments, the accumulation of biological mutations is a stochastic process. There is no way to predict when an adaptive mutation will occur, nor is it possible to predict what type of adaptive mutation might occur. Evolution is generally characterized by a series of high-probability mutational events that usually have only modest adaptive effects. However, it is our experience that the adaptation process also occasionally hits “metabolic walls” where it takes a significantly longer time for an adaptive variant to sweep through the population. At these “walls” it is likely that low-probability mutagenic events are required to continue the adaptation process. It is impossible to predict where these “walls” might occur or how long it might take to break through them. Progressing in the field of experimental evolution is like exploring a wild country beyond a new frontier as no one in the past of the industry has ever been able to pursue thoroughly this route.

How is your technology integrated in the industry ? Our purpose is to produce variants of industrially important microbes that can grow robustly under the conditions that prevail in industry. To achieve this result, our growth conditions are designed to be as close as possible to the actual conditions a microbe would experience in the real world. As a result, we are not making microbes that work in a lab, we will be able to more easily incorporate them into a functioning biorefinery.

When you evolve microbes, does it produce GMOs ? Strains produced by evolutionary optimization are naturally occurring genetic variants of the original parent strain—no foreign DNA has been inserted and no endogenous DNA has been deliberately removed. As such, these strains are not considered “genetically-modified organisms” (GMOs).

Is your technology harmless for the environment ? Many of the microbes we develop are destined for use in confined fermenters for bioproduct production at industrial scale, thus the risk to the environment is minimal. However, when the industry requires that a particular microbial biocatalyst be released into the biosphere, there is always a risk, however small, that that microbe can have a deleterious effect on the environment. There are two ways in which our technology minimizes this risk and makes our microbes the greenest solution to this problem. First, through evolutionary optimization, microbes become hyper-specialists at thriving under the conditions to which we adapted them. They do so at the expense of being able to thrive when those conditions change. Thus, once they are out in the wild they are at a distinct disadvantage once the conditions change from what they were designed for. This means that the most likely fate for our microbes after they “escape” is extinction. When they have nothing left to eat they just starve and die, resulting in microbial biomass that is highly and quickly biodegradable.

What is done with spent culture ? Spent cultures of cellular factories are destroyed according to the appropriate regulations, which change depending on the jurisdiction and the type of material.