Current Challenges

Increasingly we are recognizing the vital role microbiomes play in essentially all aspects of life on Earth. Environmental microbial communities largely govern global nutrient cycling, including the carbon and nitrogen cycles, and microbiomes associated with soils and plants are vital to agriculture. In recent years, investment from Federal agencies, including the Department of Energy, the US Department of Agriculture, and the National Science Foundation, and from agricultural companies has been targeted at understanding how microbial communities associated with crop plants can promote low-input, high-productivity agriculture. It has been known for some time that up to 50% of carbon fixed by plants is exuded by the roots. Likely a major portion of this is to attract and support beneficial microbes. Yet, significant research has been conducted on only a few key plant-microbe mutualisms, such as legume-rhizobia or mycorrhizal symbioses, which enable plants to access otherwise unavailable nitrogen and phosphorus sources. However, critical details are missing about even these common forms of plant-microbe symbiosis and about a wide range of other growth promoting microbial interactions within the rhizosphere. Microbiome work to understand these interactions holds great promise to reduce the amount of fertilizer and water required for agriculture, while displacing petroleum-based products and improving the sustainability and yield of current practices.

 

Soil bacteria localized around plant root hairs within EcoFAB 1.0

The need to better understand environmental microbiomes has been highlighted not only in scientific literature, but also by the U.S. Department of Energy and other federal research sponsors. DOE’s Office of Science has highlighted its Grand Challenges, one of which is complex systems science across scales.[1] Many subsequent reports have highlighted understanding of microbial community dynamics in addressing key challenges in bioenergy development and climate. [2],[3]

Beyond DOE, the U.S. Department of Agriculture has recognized the importance of microbiomes to enable low-input agriculture and to mitigate the effects of the changing environment. A recent Dear Colleague letter explains that novel research into the role of microbiomes in increasing food security and carbon sequestration is a necessary element to achieving the goals set forth in the USDA strategic plan.[4],[5] The National Academies have noted the need for better understanding of microbial communities in recent reports, especially for communities found in the environment.[6] These efforts require detailed contextual information and model communities to be made accessible to the research community at large.

Tremendous advances in DNA sequencing now enable inexpensive sequencing bacterial genomes in high-throughput. As a result, we have accumulated and continue to accumulate vast amounts of sequence information from all types of organisms and environments. This, in principle, could provide the information necessary to make accurate predictions of microbial metabolism and fitness. However, we lack the necessary information on gene functions to make accurate predictions from environmental DNA sequence and we poorly understand the biology of soil organisms. Even for model organisms that have been studied for several decades, such as E.coli, 20% of genes still have no predicted function, and for another 10-20% we have only a very crude and general protein family information but lacking a precise enzymatic activity or function. The correlative research that has dominated the field may not be representative of the complex interactions happening in natural systems. The sequence-function gap is widely recognized within the research community as one of the grand challenges in microbiology (and indeed, biology in general), and it is now well established and attempts to understand biological phenomena based on incomplete data that would certainly lead to erroneous conclusions. Moving beyond correlation to causality and design will require greatly improved understanding of gene functions and this will require moving from isolate cultures (where many genes have no apparent function) to more realistic environments and greatly improved understanding of the biology of soil organisms. These systems will enable hypothesis testing, for example to test the role of a microbe or biosynthetic pathway within a community, and also enable hypothesis generation for testing in field studies.

Despite the investment and promise of plant-associated microbiome research, we lack vital understanding of the plant-microbe-soil interactions that govern these communities. Most approaches aimed at improving our understanding of soil microbial communities are focused on examination of individual isolates or field studies of complex native communities. One promising research direction is constructing laboratory consortia, since these have the advantage that the constituent isolates can (in most cases) be characterized independently and even genetically manipulated to determine causal mechanisms. While these consortia systems allow researchers to test hypotheses about community interactions, the validity of extrapolating consortia-based findings to authentic ‘field’ communities has not been determined. Conversely, approaches for studying field microbial communities are challenging because they are so unconstrained and complex, and they often show irreproducible results such that definitive links between specific taxa and effects on plant growth or ecosystem function cannot be identified.

One solution to this challenge is to develop model soil ecosystems to allow for controlled, replicated laboratory experiments that can be validated in the field. Broad scientific community acceptance of a few of these model ecosystems would no doubt exponentially increase our understanding of microbial communities as a whole by focusing diverse expertise and capabilities on the same systems. By analogy, model organisms (e.g. mice) have been instrumental in determining the molecular and cellular biology of multicellular organisms. Indeed, human diseases are often studied using organisms that dramatically differ from humans (e.g. zebrafish) because they provide reproducible systems that can be manipulated in carefully controlled experiments in labs around the world.

 

New technologies are urgently needed to construct these model ecosystems with the capability of controlling the “microbial microenvironment”—the sum of all chemical and physical interactions impressed upon a cell by its biotic and abiotic environment. Studying these microenvironments by engineering and manipulating them and measuring their phenotypic outputs builds strong bottom-up understanding of the foundational chemical and genetic factors structuring microbial communities. While this may sound futuristic, advances in 2D and 3D fabrication of biomaterials could rapidly enable the construction of microbiomes with carefully controlled microenvironments and targeted cellular interactions. Critically, these synthetic microbiomes can be at a range of scales (e.g. aggregate scale, plant-scale) and will allow use of extant microbial and host genetics tools to test the roles of individual taxa and combinations of genes and microbes in their microenvironment and interaction contexts, thus establishing causal connections. Central to this approach will be unlocking the enormous treasure trove of genetic diversity currently beyond the reach of laboratory microbiologists by using microenvironment control to enable high-throughput culturing of the “unculturable”.

The EcoFAB workshop defined these potential model soil ecosystems known as EcoFABs and determined the types of biotic and abiotic components integrated with the measurements and models required to accurately predict microbiome activities. An important consideration in the development of EcoFABs is that it be relevant to and used by the larger scientific community, ideally including laboratory, field and computational scientists. Thus, assemblies of communities of scientists around specific EcoFABs, analogous to the existing Brachypodium and Arabidopsis communities is essential. The discussions and outputs of the workshop are discussed in the report-outs from each breakout session.

[1] Biological and Environmental Research Grand Challenge Workshop Report DOE/SC-0135 (2010)

[2] BER Research for Sustainable Bioenergy DOE/SC-0167 (2013)

[3] BER Building Virtual Ecosystems Report DOE/SC-0171 (2014)

[4] http://www.nsf.gov/pubs/2016/nsf16058/nsf16058.jsp

[5] USDA Strategic Plan FY14-FY18 (2013)

[6] The Science and Applications of Microbial Genomics: Workshop, National Academies Press (2013)

Relevant model ecosystems, where detailed -omics measurements can be meaningfully extrapolated to the real world,  have not been achieved.  In our view, there are four key reasons why:

  1. Observability: to measure biogeochemical processes with unprecedented resolution on nested Eco_Fab1 copyscales.
    This would enable a theoretical modeling science known as mesoscale systems biology.
  2. Reproducibility: to recapitulate the system and its dynamics at will. Because this is a biological system, there will be variation, but this variation must be manageable as it is in model organism developmental biology. A reproducible system is required for testing predictions and evaluating models.
  3. Controllability: to make defined, targeted perturbations or manipulations to the cohort of species, their spatiotemporal community structures or dynamics, and their genetic complements – all in biochemical and geophysical context. The ability to manipulate our system is a prerequisite for hypothesis testing.
  4. Ecosimilarity: to assess the extent to which our bench-top model reproduces key behavior observed in open-field, real-world systems. Ultimately, the lessons we learn must be transferable to extant ecologies to enable interventions promoting ecoremediation and a healthier planet.

Overcoming these challenges using Berkeley Lab’s engineering, bioscience, and modeling capabilities will provide access to knowledge of biological and ecosystem function that has been inaccessible with current technologies.

chip_eco