LF306

SYNTHETIC BIOLOGY
Synthetic Microbial Communities - Engineering communities

PART 2 (of 2)

What we will cover on these two lectures?

Synthetic Communities as new frontier in SynBio.
Why engineered communities? How to engineer communities?
Rational engineering vs. natural community engineering.
Future directions.

Associated literature (in addition to those cited on slides)

Grosskopf, T, and Orkun S Soyer. "Synthetic Microbial Communities." Curr Opin Microbiol 18, (2014): doi:10.1016/
j.mib.2014.02.002

Mee, Michael T, and Harris H Wang. "Engineering Ecosystems and Synthetic Ecologies." Molecular bioSystems 8, no.
10 (2012): doi:10.1039/c2mb25133g.

Lindemann, S. R., Bernstein, H. C., Song, H.-S., Fredrickson, J. K., Fields, M. W., Shou, W., … Beliaev, A. S. (2016).
Engineering microbial consortia for controllable outputs. The ISME Journal, 10(9), 1–8. https://doi.org/10.1038/
ismej.2016.26

Widder, Stefanie, Rosalind J Allen, Thomas Pfeiffer, Thomas P Curtis, Carsten Wiuf, William T Sloan, Otto X Cordero,
and others. "Challenges in Microbial Ecology: Building Predictive Understanding of Community Function and
Dynamics." The ISME journal (2016).

Two main approaches to synthetic community

engineering

Rationally
~~ engineer
links
~~
Exploit
natural links
i.e. create to or
tendencies
study and apply
i.e. learn
from Nature
 ngs | PETROCHEMICALS

Microbial Communities. They are everywhere.

es of
tion
High industrial and medical Scienti c frontier with a
relevance
ds with multitude of open questions
nea@ Biofilms IKC
(water
INDUSTRIAL SECTORS
BIOTECH | AGRI-TECH How can we explain diversity in
tteries) WATER | MEDICAL
COATINGS
| PETROCHEMICALS microbial communities?
living coatings

What is the function (if any) in a

ers given microbial community?
novel routes of
ofilm bioproduction
sers Does community stability relate
INFORMATION to diversity?
biofilm colloids with TECHNOLOGIES |
high surface area (water
MANUFACTURING
BIOTECH | AGRI-TECH
ut/wound
treatment, batteries) WATER What are the key interactions in
ansplants microbial communities?

biofilm
Online
erasers
presentations on latest What is the relation between ecological
and evolutionary processes?
communities research TECHNOLOGIES
(2022):
INFORMATION
|
MANUFACTURING
gut/wound
ilm assays https://www.newton.ac.uk/event/umcw06/
ng, plant) transplants Widder S. et al, ISME J 10:11 (2016)
fi
 Natural vs. Synthetic Microbial Communities
Mono- Synthetic
cultures Communities
*Five species cellulose-
degrading community
7104 KATO ET AL.
Co-
Mono- Bio lm
culture FIG. 3. Network model of substrate flow and the putative roles of
each member in SF356. The solid lines indicate the flow of the sub-
culture strates. The thickness of each line represents the relative contribution
Complexity
of a particular pathway in SF356. The dotted lines indicate promoting
or inhibiting factors with respect to the efficiency of cellulose degra-
dation. The names of the substrates are presented in boxes. The des-
ignations of the bacterial strains are presented in ovals.
Tractability (computational and experimental)
istence of two bacterial strains in mutualistic or commensalistic
association (e.g., see references 10 and 31). In addition, some
Gene-circuits
groups have reported the construction of a defined mixed cul-
ture of more than two isolates in order to reveal metabolic
Microscopy sharing and synergistic interactions (2, 5, 6). For example,
qPCR
Plate Reader Dejonghe et al. (6) isolated bacteria corresponding to all of the
Fluorescence labeling
major bands detected by DGGE analysis from a stable micro-
flora that degrades the herbicide linuron, and they constructed
a defined mixed culture that degraded linuron in order to
reveal the relevant synergistic interactions. However, they did
not make note of the stable coexistence of the members of the
community. To the best of our knowledge, there appears to
have been no reports to date regarding the construction of a
defined mixed culture composed of more than two microor-
ganisms that was able to stably coexist for a long period of
* Kato, S et al. App. & Env Microbiol 71:11 (2005)
time. Hence, our defined mixed culture, SF356, is considered a
novel accomplishment, as it consists of dominant bacteria in an
original microflora, and all of the members of the culture were
fi
fi
Microbial
Communities
APPL. ENVIRON. MICROBIOL.
aerobic static conditions, and the degradation efficiency was
significantly low even under anaerobic conditions. These ob-
servations suggested that the contributions of aerobic noncel-
lulolytic bacteria (i.e., the consumption of oxygen, pH neutral-
ization, and stimulation of growth) (20) (Fig. 3) are essential
for effective cellulose degradation by the community. In addi-
tion, the growth of noncellulolytic isolates was promoted in the
cell-free culture filtrate of C. straminisolvens CSK1 (Table 1).
1000s of
Thus, the results are suggestive of a synergistic relationship
between C. straminisolvens CSK1 and the aerobic noncellulo-
lytic bacteria.
(ii) Clostridium sp. strain FG4. Clostridium sp. strain FG4 is
Granules
an anaerobic noncellulolytic isolate. It appears that Clostridium
sp. strain FG4 consumes saccharides derived from cellulose
species
degradation and produces a large amount of acetate and a
small amount of ethanol (Fig. 3), as based on the knockout
experiments (Fig. 2C to E). This prediction was also supported
by the results of the pure-culture experiments (Table 1).
Water-soluble oligosaccharides, especially cellobiose, are
known to repress cellulase expression (18, 33), to inhibit cel-
lulase activities (19, 23), and to inhibit cellulose degradation
(24) by cellulolytic clostridia. Hence, the consumption of sac-
charides by Clostridium sp. strain FG4 is assumed to have a
positive effect on achieving effective cellulose degradation.
However, although the accumulation of saccharides in !FG4
was higher than that in the case of SF356 (Fig. 2C), !FG4
Meta-omics
exhibited more effective degradation of filter paper than SF356
DNA
(Fig. 2A). This observation indicated that Clostridium sp. strain
FG4 has a negative effect on cellulose degradation, and this
effect exceeded the positive effects exerted by the saccharide
ngerprinting
scavenge. Although the pH value in !FG4 dropped once below
6 before day 3, it returned to an approximately neutral value,
and the acidic pH in SF356 (the mixed culture containing
Clostridium sp. strain FG4) was generally maintained during
the cultivation period (Fig. 2B). The production of acetic acid
by Clostridium sp. strain FG4 would be expected to maintain an
acidic pH. This could account for the negative effects on cel-
lulose degradation exerted by C. straminisolvens CSK1, which
exhibited only slight cellulose degradation under a pH value of
6.0 (21).
(iii) Pseudoxanthomonas sp. strain M1-3. Pseudoxanthomo-
nas sp. strain M1-3 is a facultative anaerobe and is considered
to greatly contribute to oxygen consumption, pH neutraliza-

Understanding Microbial Communities
Tamames et al. BMC Microbiology 2010, 10:85
http://www.biomedcentral.com/1471-2180/10/85
Sequence and thou shall understand…
specificity but does, undoubtedly, indicate a clear environ- diversity. To assess the relationship between environ-
mental preference. ments and taxa, we clustered the different environmen-
The families that are present in many environments, tal types according to the affinities of their different taxa
but not showing relevant affinity values for any of them, (Figure 3).
may be considered ubiquitous. Amongst these we The environments are separated into five different
mainly find Proteobacteria, such as Xanthomonadaceae, groups. The first one is associated with animal tissues
Comamonadaceae, Pseudomonadaceae and Burkholder- (oral, gut, vagina, other human tissues, samples from
iaceae, but also the archaeal Methanosarcinaceae and animal tissues and aerial specimens, the last mostly
Page 6 of 14
the Clostridia Peptococcaceae families. coming from air expired from human subjects). These
It is interesting to note that the Clostridia clade har- habitats clearly differ from the rest, and some of the
bors cosmopolitan families, such as Peptococcaceae, and prokaryotes living there do not thrive in other locations
environment-specific ones such as Lachnospiraceae or [28]. Thus, host association with animals emerges as the
Oscillospiraceae. This indicates that phylogenetically first discriminating factor in the composition of the pro-
close families can show strikingly
Vol 464 | 4 March 2010different environmen-
| doi:10.1038/nature08821 karyotic assemblages.
tal preferences and distribution patterns, which at least The second group to segregate is composed of ther-
for some cases, questions the validity of the proposed mal environments (geo- and hydrothermal), and also
relationship between phylogenetic distance and environ- ARTICLES
shows a clearly distinct taxonomic profile. Both environ-
mental preferences [26,27]. ments are separated by long distances in the dendro-
Taxonomic distributions can be used to explore the gram, which indicates significant differences between
A human gut microbial gene catalogue
characteristics of the environments themselves. Group-
ing environments according to similarity in their taxo-
them. The absence of oxygen and light in hydrothermal
locations accounts for the presence of some anaerobic
established by metagenomic sequencing
nomic profiles can help us to understand the main
environmental features at play in selecting prokaryotic
methanogenic archaea in hydrothermal, but not geother-
mal sources, or for some photosynthetic cyanobacterial
Junjie Qin1*, Ruiqiang Li1*, Jeroen Raes2,3, Manimozhiyan Arumugam2, Kristoffer Solvsten Burgdorf4,
Chaysavanh Manichanh5, Trine Nielsen4, Nicolas Pons6, Florence Levenez6, Takuji Yamada2, Daniel R. Mende2,
Junhua Li1,7, Junming Xu1, Shaochuan Li1, Dongfang Li1,8, Jianjun Cao1, Bo Wang1, Huiqing Liang1, Huisong Zheng1,
Yinlong Xie1,7, Julien Tap6, Patricia Lepage6, Marcelo Bertalan9, Jean-Michel Batto6, Torben Hansen4, Denis Le
Paslier10, Allan Linneberg11, H. Bjørn Nielsen9, Eric Pelletier10, Pierre Renault6, Thomas Sicheritz-Ponten9,
Keith Turner12, Hongmei Zhu1, Chang Yu1, Shengting Li1, Min Jian1, Yan Zhou1, Yingrui Li1, Xiuqing Zhang1,
Songgang Li1, Nan Qin1, Huanming Yang1, Jian Wang1, Søren Brunak9, Joel Doré6, Francisco Guarner5,
Karsten Kristiansen13, Oluf Pedersen4,14, Julian Parkhill12, Jean Weissenbach10, MetaHIT Consortium{, Peer Bork2,
S. Dusko Ehrlich6 & Jun Wang1,13
To understand the impact of gut microbes on human health and well-being it is crucial to assess their genetic potential. Here
we describe the Illumina-based metagenomic sequencing, assembly and characterization of 3.3 million non-redundant
microbial genes, derived from 576.7 gigabases of sequence, from faecal samples of 124 European individuals. The gene set,
,150 times larger than the human gene complement, contains an overwhelming majority of the prevalent (more frequent)
microbial genes of the cohort and probably includes a large proportion of the prevalent human intestinal microbial genes. The
genes are largely shared among individuals of the cohort. Over 99% of the genes are bacterial, indicating that the entire
cohort harbours between 1,000 and 1,150 prevalent bacterial species and each individual at least 160 such species, which are
also largely shared. We define and describe the minimal gut metagenome and the minimal gut bacterial genome in terms of
functions present in all individuals and most bacteria, respectively.
It has been estimated that the microbes in our bodies collectively individuals from the United States or Japan8,16,17. To get a broader
make up to 100 trillion cells, tenfold the number of human cells, overview of the human gut microbial genes we used the Illumina
and suggested that they encode 100-fold more unique genes than Genome Analyser (GA) technology to carry out deep sequencing of
our own genome1. The majority of microbes reside in the gut, have total DNA from faecal samples of 124 European adults. We generated
a profound influence on human physiology and nutrition, and are 576.7 Gb of sequence, almost 200 times more than in all previous
crucial for human life2,3. Furthermore, the gut microbes contribute to studies, assembled it into contigs and predicted 3.3 million unique
energy harvest from food, and changes of gut microbiome may be open reading frames (ORFs). This gene catalogue contains virtually
associated with bowel diseases or obesity4–8. all of the prevalent gut microbial genes in our cohort, provides a
To understand and exploit the impact of the gut microbes on broad view of the functions important for bacterial life in the gut
human health and well-being it is necessary to decipher the content, and indicates that many bacterial species are shared by different
diversity and functioning of the microbial gut community. 16S ribo- individuals. Our results also show that short-read metagenomic
somal RNA gene (rRNA) sequence-based methods9 revealed that two sequencing can be used for global characterization of the genetic
bacterial divisions, the Bacteroidetes and the Firmicutes, constitute potential of ecologically complex environments.
over 90% of the known phylogenetic categories and dominate the
distal gut microbiota10. Studies also showed substantial diversity of Metagenomic sequencing of gut microbiomes
the gut microbiome between healthy individuals4,8,10,11. Although this As part of the MetaHIT (Metagenomics of the Human Intestinal
difference is especially marked among infants12, later in life the gut Tract) project, we collected faecal specimens from 124 healthy, over-
microbiome converges to more similar phyla. weight and obese individual human adults, as well as inflammatory
Metagenomic sequencing represents a powerful alternative to bowel disease (IBD) patients, from Denmark and Spain (Supplemen-
rRNA sequencing for analysing complex microbial communities13–15. tary Table 1). Total DNA was extracted from the faecal specimens18
Applied to the human gut, such studies have already generated some and an average of 4.5 Gb (ranging between 2 and 7.3 Gb) of sequence
3 gigabases (Gb) of microbial sequence from faecal samples of 33 was generated for each sample, allowing us to capture most of the
Figure 3 Relations between environments, and between environments and taxonomic families. Heat-map of the posterior medians of the
affinities and the resultingResearch
dendrogram from the cluster analysis
Molecular of theLaboratory,
environment types, usingVIB—Vrije
log-affinities and euclidean distance. Purple and
1 2 3 4
BGI-Shenzhen, Shenzhen 518083, China. European Biology 69117 Heidelberg, Germany. Universiteit Brussel, 1050 Brussels, Belgium. Hagedorn
Institute, DK 2820 Copenhagen, Denmark. 5Hospital Universitari Val d’Hebron, Ciberehd, 08035 Barcelona, Spain. 6Institut National de la Recherche Agronomique, 78350
orange cells represent lowJouyand high
en Josas, affinity
France. 7
values,
School of Software respectively.
9
Engineering, South China University of Technology, Guangzhou 510641, China. 8Genome Research Institute, Shenzhen University Medical
10
School, Shenzhen 518000, China. Center for Biological Sequence Analysis, Technical University of Denmark, DK-2800 Kongens Lyngby, Denmark. Commissariat à l’Energie
…but, it seem that
Atomique, Genoscope, 91000 Evry, France. 11Research Center for Prevention and Health, DK-2600 Glostrup, Denmark. 12The Wellcome Trust Sanger Institute, Hinxton, Cambridge
CB10 1SA, UK. 13Department of Biology, University of Copenhagen, DK-2200 Copenhagen, Denmark. 14Institute of Biomedical Sciences, University of Copenhagen & Faculty of Health
Science, University of Aarhus, 8000 Aarhus, Denmark.
*These authors contributed equally to this work.
{Lists of authors and affiliations appear at the end of the paper.
59
©2010 Macmillan Publishers Limited. All rights reserved
everything is everywhere
Tamames, J et al BMC Microbiol 10 (2010)
RI 6 August 2008 14:55

Understanding - Metagenomics
Culture-based
identity

Escherichia coli
DH10B (Temporal) Metagenomics

(Temporal) Metatranscriptomics
Classic microbial
physiology tests
16S rRNA sequence +

Metagenomics
Species isolation
Who is
present? What is their
function? Stable Isotope Tracking

….
Microbial
community Comparative
assemblage genomics

Phylogenetics

Can we make How do Culture

predictions? they interact? bioassays

O O
O
[14c]- N
H H
O
Computational Stable isotope tracking
modeling

Figure 3 Little, AE et al Annual Rev of Microbiol 62 (2008)

 Understanding - Temporal Metagenomics
www.anaerodynamics.com
1e−02

●
●
●

● ● ●
● ●

●
●

1e−03 ●

Relative abundance
●
●

● ●
●
● ●

●
● ●
● ● ●
1e−04 ●
● ●
●

●
●

● ●

●
1e−05 ● ●

10 20 30 40 50
Week

• Weekly samples over a year

• Metagenomics
• Metadata on methane, pH, and
(in some cases) feed.

Raguideau, et al. unpublished results

 tents via exposure to an alternating-current electric field (corona Appendix, Fig. S6) with on-c

Understanding - Massively Parallel Cultivation

Fig. 1. kChip enables massively parallel construction of microbial communities. (A) To run a kChip screen, 1-nL d
contains a color code (a specific ratio of three fluorescent dyes) that maps to a corresponding input. After they hav
the kChip, where they randomly group into microwells (SI Appendix, Fig. S1). The microwells are designed to group
to identify the contents of each microwell from the droplet color codes. Droplets are then merged within their
alternating-current electric field, generating parallel synthetic communities. Community phenotypes can be tracke
protein expression and respiration-driven reduction of resazurin to the fluorescent product resorufin. (B) Example m
Downloaded by guest on November 11, 2020

of droplets for different microwell types, the designs for which are described in SI Appendix, Fig. S2. Microwell
microwell density varying inversely with size (k). A single microwell type can be arrayed across a kChip (“Full kChip”).
Figs. 3 and 4, we have generated a “k = {1:7;19} Chip” that includes different microwell types arranged in paralle

Kehe et al. PNAS | Jun

Kehe et al. PNAS 116:26 (2019)

 Exploiting Natural Microbial Communities GasHigh

2.0 Biogas
Anaerobic Digestion: Animal Guts, Biogas Country
Reactors, ReportTreatment
and Water GermanyPlants
2009
heterogeneous biofilms 2.1 introduction
and granules, enhances the metabolic interactions in a BMC10,11, and that
diversity and redundancy increase BMC stability and
Since the introduction
performance of 12 . a feed ofin these
tariff (FIT) for iselectricity from biogas plan
Organic Waste a continual development
convincingly proven.
Neither proposals
in biogas technology was taking place in the 1990i
It has been argued that synthetic design of
early 2000. Innovel thiscommunities
period mostly small
or manipulation companies
of existing ones can were developing the farm
biogas technology be utilisedwith the help of practical experiences
to address some of the
and to rationally engineer complex and relevant
above questions 2
from plant operator
after the introduction
functions4. The of an latterenergy crop abonus
view underpins new wave in of2004, which was paying fo
synthetic biology research, which has already resulted
production costs of energy crops, bigger companies developed the technolo
in the engineering of “toy communities” with specific
industrial standard.populationSpecial
dynamics6,7boni. Whilefor not the
drivenutilisation
by either anof heat and/or new techno
ecological or synthetic biology agenda, other studies
have given very good
defined incentives
several co- and for the development
tri-cultures capable of of a high technologica
in these fields. Through these measures and the long term development of
degradation of organic compounds into methane 13-17.

The most recent of these used metabolic models to

technology on a farm
study a specificscale level
co-culture Germany
18, and subsequently is showed
now the technology leader in t
that
wide. lactate into methane19. Currently ofit
biogas world converting this minimal synthetic community is capable
Methane
remains unknown how much the complexity of
Figure 1. SchematicBiogas such simple communities can be extended (e.g. by
degradation of organic material in
development
of the multistep
increasing in Germany
species numbers) until end 2008
to achieve an

installed electrical power [MW]

methanogenic ecosystems. Each increase in the range of substrates they can digest
step is carried out by a specific and their biomethane productivity.
functional group of micro-organisms
(also known as guilds). 2.Preliminary work leading up to this proposal
In recent published and unpublished studies, we have
been developing a better understanding of stability and
function in microbial communities and how these are affected by the evolutionary process. Here,
we summarise key findings and insights from these studies which are relevant for this proposal.

2.1. Species composition and diversity of BMCs can vary over time and with bioreactor
conditions. Different studies have shown that diversity and composition of species in a BMC can
be stable or temporally varying, with seemingly no effect on performance. For example, amplified
ribosomal DNA restriction analysis of species abundance identified temporal variation in both
archaea and bacteria populations in otherwise stable reactors20. Similar results were obtained from
a two-year monitoring of an anaerobic reactor fed with wine distillation waste21. However, other
studies on communities from reactors treating wastewater and dairy waste, indicated stable
communities despite changes in process parameters such as feed composition22,23. Process
parameters, in particular stirring, acidity and temperature, are all shown to alter composition and
diversity in a BMC24-27. In a recent study, we analysed the effects of a large number of different
operating parameters on the microbial diversity and performance of BMCs using laboratory-scale
activated sludge reactors. This analysis failed to identify clear patterns between changes in such
parameters, microbial diversity and performance28. It is possible that these disparate findings arise
 Mid Summary 1

Microbial communities are the norm in Nature

Natural microbial communities are complex in terms of

species numbers and interactions, and can display spatial
organisation

Community composition and interactions can be studied - and

interacting species pairs or groups can be picked up - using
metagenomics, temporal metagenomics and cultivation studies.
 Effect Ecological Metabolic References

0/+ Commensalism Food chain Freilich et al.

Metabolic interactions are prevalent in Nature

[22], Xu et al.
[34]

heterogeneous biofilms and granules, enhances the metabolic interactions in a BMC10,11, and that
Figure 1
diversity and redundancy increase BMC stability and
performance . Neither of these proposals is-/- Competition Substrate competition Foster & Bell
Metabolic interactions among microbes
12

convincingly proven. [19]

It has been argued that synthetic design of
novel communities or manipulation of existing ones can
be utilised to address some of the above questions2
and to rationally engineer complex and relevant Effect Ecological Metabolic References
functions4. The latter view underpins a new wave of
synthetic biology research, which has already resulted0/+ Commensalism Food chain Freilich et al.
in the engineering of “toy communities” with specific-/+ Predation Food chain with waste product inhibition Balagadde
population dynamics6,7. While not driven by either an [22], Xu et et
al.a
ecological or synthetic biology agenda, other studies [23]
[34]
Syntrophic Kerr et al. [24]

}
defined several co- and tri-cultures capable of
degradation of organic compounds into methane13-17.
The most recent of these used metabolic models to
interactions
study a specific co-culture18, and subsequently showed
that this minimal synthetic community is capable of
converting lactate into methane 19 . Currently it -/- Competition Substrate competition Foster & Bell
remains unknown how much the complexity of0/0 No interaction No common metabolites [19]
Figure 1. Schematic of the multistep such simple communities can be extended (e.g. by
Schink B Microbiol
degradation Mol Biolin Rev
of organic material increasing
61:2 (1997)species numbers) to achieve an
methanogenic ecosystems. Each increase in the range of substrates they can digest
step is carried out by a specific and their biomethane productivity.
functional group of micro-organisms
(also known as guilds). 2.Preliminary work leading up to this proposal
In recent published and unpublished studies, we have
-/+
+/+
been developing a better understanding of stability and Predation
Cooperation Food chain withSyntrophy
waste product inhibition Balagadde
Shou et al. et a
[25
function in microbial communities and how these are affected by the evolutionary process. Here, [23]
Kerner et al. [2
we summarise key findings and insights from these studies which are relevant for this proposal. Kerr et al. [24]
Hillesland et a
[28], Summers
Auxotrophic
2.1. Species composition and diversity of BMCs can vary over time and with bioreactor
conditions. Different studies have shown that diversity and composition of species in a BMC can
be stable or temporally varying, with seemingly no effect on performance. For example, amplified
al. [29],

interactions
ribosomal DNA restriction analysis of species abundance identified temporal variation in both
archaea and bacteria populations in otherwise stable reactors20. Similar results were obtained from 0/0 No interaction No common metabolites
a two-year monitoring of an anaerobic reactor fed with wine distillation waste21. However, other0/- Amensalism Waste product inhibition
studies on communities from reactors treating wastewater and dairy waste, indicated stable Grosskopf T & Soyer OS, Curr. Opin. in Microbiol. 18 (2014)
Networks of energetic and metabolic interactions
communities despite changes in process parameters such as feed composition22,23. Process
parameters, in particular stirring, acidity and temperature, are all shown to alter composition and
define dynamics in microbial communities
diversity in a BMC24-27. In a recent study, we analysed the effects of a large number of different
operating parameters on the microbial diversity and performance of BMCs using laboratory-scale
activated sludgeJoanne
Mallory Embree, reactors.
K. Liu,This analysis
Mahmoud failed to
M. Al-Bassam, andidentify
Karsten clear
Zenglerpatterns
1
between changes in such
parameters, microbial diversity and performance 28. It is possible that these disparate findings arise
Department of Bioengineering, University of California, San Diego, CA 92091
because composition and diversity in microbial communities is driven by neutral processes such as
+/+ Cooperation Syntrophy Shou et al. [25
Embree
Edited
stochasticby Mary E. M. Etdynamics
Lidstrom,
population al PNAS
University of Washington,
and 112:50
random(2015)
Seattle, WA, and approved October 21, 2015 (received for review March
immigration29 (e.g. species incorporation via feed). 26, 2015) Figure 1. The six basic motifs of microbial interactions. Blue and yellow circles denote
Kerner et al. [2
the individual
We do not know which selective pressures (i.e. operation conditions) can override such
Microorganisms form diverse communities that have a profound species within these communities, but bacteria different
and microbial strains respectively, while boxes represent metabolites. Stimulating
Hillesland etanda
impact on the environment and human health. Recent technological archaea all seem to collectively persist and thrive. The layering of
neutral processes to drive the adaptation
advances have enabled elucidation of community diversity at high
of BMCs, and how the resulting community would
multiple types of interspecies interactions, particularly those that inhibitory interactions mediated by species traits or metabolites are indicated with red and
[28], Summers
look like. Investigation
resolution. We will addressof microbial these open
communities questions
has revealed that in our
extend research
beyond basic carbonprogramme.
exchange, may explain the presence ofgreen arrows respectively. References are to studies in which synthetic microbial
al. [29],
nockout that lead to an auxotrophic amino acid use to maximize energetic efficienc
ssential aminoHow acidsdo(Materials and
metabolic produced
interactionsamino acids emerge? (e.g., S, G, and P) are
ix essential amino acids were left out quently than expensive ones (e.g., M, H, aroma
ither did not have single-gene targets previous computational approaches have sugge
Metabolic
uxotrophic or the interactions amongtionships
resulting mutants microbes (24),are thisprevalent
is the first in Nature meas
experimental
defects even in richly supplemented aware of regarding this important property. Beca
designate each auxotrophic strain by genotypes are prevalently observed in nature, ou
—for example, the methionine ΔmetA that many microbes may be subjected to this e
14 auxotrophs (C, F, G, H, I, K, L, M, zation and that amino acid exchange may be a
Cross feeding
confirmed to show no growth in M9- guiding theSyntrophy
evolution of emerging these microbial from ecosyst
emerging from trade-offs To moreenvironmental
deeply investigate limitations the onproperties
(and sympatric speciation) metabolite electron
exchangeacceptors of amino acids in microbi
we developed Schink B synthetic ecosystems
Microbiol Mol Biol Rev 61:2 (1997) using th
auxotrophic E. coli. Studying synthetic communi
offers the benefits of robust and fast-growing
genetics, and well-developed in silico models wh
a standardized and reproducible genetic backg
developed a two-member syntrophic communi
twoAuxotrophy
different species emergingfrom from our 14 characteriz
Eachoptimization
auxotroph for metabolite
is unable to growusage? in M9-glucos
tentially grow as a coculture when paired w
Akashi, H & Gojobori T PNAS 99:6 (2002)
auxotroph. We probed all 91 possible pairwise s
actions in M9-glucose minimal media. In agreem
 To determine whether increased amino acid
Results production also incurred fitness costs, the three
Metabolic auxotrophies are evolutionarily stable?
Construction of cross-feeding interactions
single-gene deletion mutants DnuoN, Dmdh, and
Dppc mutants were competed against WT. The
In order to construct a synthetic cooperative inter- results of this experiment indicated significant
action, we first generated strains that released
ing interactions
fitness costs of amino-acid overproduction of
Sincreased
Pande et al amounts of amino acids into the around 5–7% (Figure 2b).

se cases,
udied in Stability of obligate cross-feeding interactions
S Pande et al
h as the
5
cInerney Growth advantage of cross-feeding relative to
achment prototrophic WT
al., 2008) To verify whether the newly constructed strains
s specific could cross-feed each other, all possible combina-
tions between two amino-acid auxotrophs, over-
tunately, producers, and cross-feeders were mixed (1:1) in
rs makes minimal medium and their realised growth rate
eractions within 24 h (that is, the Malthusian parameter)
al., 2002; determined as a measure of fitness. In this experi-
Figure 1 Design strategy of synthetic cross-feeding interactions. Stability of obligate cross-feeding interactions
d Figure
Moran,2 Characterisation ment, genotypesand
of genes
mutants carrying the experiment.
same amino-acid
Overview over the deleted incoculture
E. coli wild type (WT) to(a) Amino
yield acid
S Pande et al(AA) production levels (that is, arginine, tryptophan,
leucine, and histidine) auxotrophy-causing
of WT, four auxotrophsmutation
(Aux), were
threenot paired, and (OV), eleven cross-feeders (CF) and the cross-feeding genotype
overproducers
questions one of four single-gene deletion mutants that are either
WT cells inoculated at a similar initial density
auxo- 5
DleuBDnuoNtrophic
lutionary (CF*) for
within 24
one amino
Growth h determined
acid
advantage (AA,
of thatasis,productivity
arginine,
cross-feeding relative of cocultured
tryptophan,
to
served as control. The results indicated that con-
AA auxotrophs (n ¼ 8 for every auxotroph-genotype
obligate Cross-feeding auxotrophs can
combination).leucine andprototrophic
Combinationshistidine)
deletion mutants
AA-overproducing mutants
orWT
with
sortia composed
overproduce
matching
(thatwhether
To verify
AAs, asacid
amino
of two auxotrophic
is, ‘cross-feeders’),
relative toless
significantly WTfit within
than 24
well as
in which the
the newly constructed
double
auxotrophies
strains were
h inSurprisingly,
WT.
two
strains
minimal medium. the
were excluded. (b) Competitive fitness of the three
Relative fitness is the ratio of Malthusian parameters and the
mutations causing AA auxotrophy and overproduction were
outcompete WT through
could
dashed line indicates equality cross-feed each other, all possible combina-
fitnessinoffitness
two out between
of threeWTpairs
and of
competitor. Asterisks indicate fitness values that were significantly different from 1
overproducing
ng inter-
(that is, WT combined intions
all possible combinations.
strainsbetween
fitness, one-sample astwo
t-test,
as well **Po0.01,
43 of
Coculturing
amino-acid
the 54 ***Po0.001,
possible
two of those
auxotrophs, n ¼over-
combinations10). (c) Fitness given as the Malthusian parameter of WT, pairs of
um: why cross-feedersproducers,
should result
and in reciprocal exchange
cross-feeders were mixed of essential
(1:1) in
abolite to
DleuBDnuoN metabolic “division of labour”
cocultured auxotrophs of cross-feeding
amino acids(Aux,
(inset).
minimal
levels24
(CF*, 9) within h (n ¼ 8
mutants overproducers
6 combinations),
for WT
significantly exceeded
medium and their realised growth rate
(Figure 2c; Supplementary
and everyFigure
pair
within 24 h (that is, the Malthusian parameter)
(OV, 3),WT
S2).
of
cross-feeders (CF, 45), and cross-feeding consortia involving strain
Of the
genotypes). Different letters indicate significant differences (LSD post hoc
ortest,
itself? 11 cocultures of cross-feeding mutants that did not
Po0.001). Boxplots:determined
median (horizontal
as a measure linesofin boxes),
fitness. this experi- range (boxes), 1.5 " -interquartile range (whiskers).
Ininterquartile
ecause of show this pattern, 9 involved the double deletion
ment, genotypes carrying the same amino-acid
mutant DleuBDnuoN (Figure 2c; Supplementary
e on the among two Pande, S etwhose
bacterial al. ISME
genotypes?
auxotrophy-causing
Figure S2), J 8:5(2)(2014)
mutation
amino-acid
wereArenotcooperative
paired, and
production levels
Journal cross-feeding
WT interactions
cells inoculatedvulnerable
at a similartoinitial
the exploi-
density
lutionary were also not sufficient to support growth of the
tation served as control. The results indicated that con-
of
le to a four focal auxotrophs (Figure 2a; Supplementary
sortia composed of two auxotrophic strains were
ism. For non-cooperating
Figure genotypes? Our results provide
S1a).
significantly less fit than WT. Surprisingly, the
compar- evidence fitness
for aof significant
two out of three fitness
pairs ofadvantage
overproducingof
y related obligate cross-feeding
strains as well as relative
43 of the 54topossible
metabolic auto-
combinations Figure 3 Competition of single-gene deletion mutants against
Benefit of cross-feeding is larger than the costs WT. (a) Fitness of amino-acid auxotrophic and overproducing
coopera- of
nomy andOne cross-feeding
suggest mutants
that for
thethe significantly
metabolic exceeded
cross-feeding WT
explanation observed synergistic growth single-gene deletion mutants relative to WT within 24 h in
levels (Figure
interactions 2c; Supplementary Figure S2). Of the
re rarely couldcan
be a stably
division coexist
of metabolic with other,
labour amongnon-
pairs minimal medium supplemented by either culture extract from
11 cocultures of cross-feeding mutants that did not the Dmdh mutant, or (b) a mixture of amino acids (150 mM) that
cooperating genotypes—even in the absence of
of cross-feeding mutants. For this mechanism to
 lity
hav
Metabolic auxotrophies are evolutionarily stable?
Figure 5 Competition of cross-feeding consortia against wild from
type. Representative pairs of cross-feeding mutants (Supplementary pro
Figure S2) were competed against WT for 24 h in liquid minimal
medium. Relative fitness is the ratio of Malthusian parameters tha
and the dashed line indicates equality in fitness between WT and rele
the cross-feeding consortia. All fitness values were significantly env
different from 1 (Pp0.03, n ¼ 10), except the ones labelled from
And can be maintained against possible ‘cheats’. with ‘ns’.
Stability of obligate cross-feeding interactions am
S Pande et al bin
7 Thi
Discussion ute
Cooperative cross-feeding interactions are very cos
common in bacterial populations and thus of critical com
importance to our understanding of microbial com-
munities. Methodological difficulties to cultivate am
and study naturally evolved interactions, however, the
have so far impeded a mechanistic understanding of tha
the costs and benefits associated with this type of lise
symbiotic lifestyle as well as its robustness in the
presence of non-cooperating genotypes. Here we con
used an approach of synthetic microbial ecology to mu
experimentally determine the fitness consequences
of an obligate and reciprocal exchange of metabo-
lites between two partners as well as the effects of
non-cooperating genotypes on the ecological stabi- Dis
lity of cooperative cross-feeding interactions. We
have shown that the loss of two metabolic genes
We
Figure 5 Competition of cross-feeding consortia against wild from a bacterial genome was sufficient to turn a gen
type. Representative pairs of cross-feeding mutants (Supplementary Figure bacterial
prototrophic 6 Reciprocal invasion-from-rare
cell into a cooperating strain experiments with four stab
Figure S2) were competed against WT for 24 h in liquid minimal cross-feeding consortia and the corresponding auxotrophs.
Pande, S et al. ISME J 8:5 (2014)
medium. Relative fitness is the ratio of Malthusian parameters that was unable to produce a certain amino acid, yet can
and the dashed line indicates equality in fitness between WT and
Competition
released increased experiments
amounts ofwith otherseachintoof the
the four representative
the cross-feeding consortia. All fitness values were significantly see
cross-feedingalso;
environment. The ecological interactions resulting S2) and one of the
consortia (Supplementary Figure eno
different from 1 (Pp0.03, n ¼ 10), except the ones labelled two cross-feeding
from the corresponding of Shou, Wwere
auxotrophs
essential et al. initiated
amino PNAS
acids104:6
at a(2007)
ratio of 1:100, a lo
with ‘ns’. 100:1 and 1:1 and the fitness of the cross-feeding populations
among two deletion mutants increased their com- cel
bined relative to auxotrophs
fitness relative within 24 h ancestor.
to their prototrophic determined. Shown is the
(Ko
percentage fitness
This unexpected of the advantage
cross-feedingcould consortia
be attrib-at the onset of the
uted toexperiment
a division(x-axis) and their
of metabolic fitnessThe
labour: relative to the focal auxotrophs
fitness Ho
costs of amino-acid
after overproduction
24 h (y-axis). Different were more correspond
symbols than to the eight liti
compensated
different by the advantage
combinations tested andresulting from is the mean of eight
every point con
amino-acid auxotrophies.
replicates. The solid Moreover,
line showscharacterising
the quadratic regression of fitness
2
stab
 Metabolic cross-feeding is evolutionarily stable?

Ecological and evolutionary dynamics of coexisting

lineages during a long-term experiment with 2008, generation 45K
Escherichia coli
Mickaël Le Gaca,b, Jessica Plucaina,b, Thomas Hindréa,b, Richard E. Lenskic,d,1, and Dominique Schneidera,b,1
a
Laboratoire Adaptation et Pathogénie des Microorganismes, Institut Jean Roget, Université Joseph Fourier, F-38041 Grenoble, France; bCentre National de la
Recherche Scientifique, Unité Mixte de Recherche 5163, F-38041 Grenoble, France; and cDepartment of Microbiology and Molecular Genetics, and dBEACON
Center for the Study of Evolution in Action, Michigan State University, East Lansing, MI 48824

Contributed by Richard E. Lenski, April 27, 2012 (sent for review January 17, 2012)

Closely related organisms usually occupy similar ecological niches,
leading to intense competition and even extinction. Such compe-
tition also can promote rapid phenotypic evolution and ecological
divergence. This process may end with the stable occupation of
distinct niches or, alternatively, may entail repeated bouts of
evolution. Here we examine two Escherichia coli lineages, called L
LF306,and
Lecture 11,coexisted
S, that 15more than 30,000 generations after di-
Slide: for
their ecological divergence and stabilizing their interaction.
Third, fitness may improve relative to previous generations both
within and between lineages, leading to more complex changes in
their interaction. For example, one lineage might encroach on
the ecological niche of the other, and that encroachment in turn
might cause the extinction of the affected lineage, or might lead to
further evolution of the affected lineage that enables its persistence.
Trade-off leads to evolution of specialisation
and cross-feeding

Grosskopf T et al. BMC Evol. Biol 17:163 (2017)

 species systems.
How does metabolic cross-feeding emerge?
SECTION B. Methodology
Workpackage A. Establishing the design principles of and experimental tools for c
controlling microbial metabolic interactions.
The overarching vision of this workpackage is to enable rational engineering at the
Respiratory Fermentative
species. Achieving such rational engineering will require modelling and predicting m
points based on their metabolism,
Metabolism and characterising and manipulating such interactions
Metabolism
will develop these tools and enable easier and systematic engineering of microbial metabo
Sugars Sugars
Project A1. Establishing a ther
modelling framework for metaboli
microbial organisms display oxida
they utilise under the presence of a
acceptors for complete oxidation
fermentative pathways that they ut
Ethanol, Acetate, Lactate,
that result
Butyrate, … in incomplete oxidation o
5). It is also possible that under ce
conditions, e.g. at high levels of su
Respiration both pathways are utilised to harves
Figure 5. A cartoon representation of metabolic energy from oxidative pathways an
flux among oxidative (black) and fermentative substrate into fermentation40,41,45,52. T
(red) pathways. In presence (A) of strong electron these
??? two modalities
Evolution of growth, or
of cross-
acceptors (blue), the oxidative pathways would can be understood
feeding from metabolicin the context
carry most of the flux, while in their absence (B), and resource allocation.
fermentative pathways would dominate. switching ?
We will develop a modell
microbial interactions that will tak
 amino acids, proteins, and DNA) require energetic supply
that is provided by ATP. It has been observed that if mi-
How does metabolic cross-feeding emerge?
crobes are limited by their energetic resource, the amount
of biomass formed per unit of ATP is approximately con-
stant and independent of the mode of ATP production
(Bauchop and Elsden 1960). This implies that with in- Pfeiffer T & Bonhoeffer S, Curr. Am. Nat. 163:6 (2004)
creasing rate of ATP production, the rate of biomass for-
mation and thus the growth rate of an organism increases.
More direct support comes from a study by Dykhuizen
and Dean (1990), which showed that in an E. coli pop-
ulation in a chemostat with lactose as limiting energy re-
source, fitness of an organism strongly correlates to the
rate of the lactose pathway. The concentration of enzymes
of a metabolic pathway should be minimized because en-
zymeTrade-offs in growth
synthesis involves costs in terms of ATP, carbon, and
other compounds and because the capacity of protein syn-
cterial Figure 1: A, Unbranched biochemical pathway with m intermediates and
thesis may be limited (Heinrich and Schuster 1996). Ev-
tewart
idence for evolutionary forces reducing the costs of enzyme mCostly intermediates
! 1 enzymes. andis denoted
The rate of the pathway enzymes by J S. B, Unbranched
ase of
ATP-producing pathway with nATP ATP-producing steps. The rate of ATP
production comes from molecular evolution studies dem- production is given by J ATP p nATPJ S . C, Extended pathway scheme. An
nd the
onstrating that a negative correlation exists between bio-
hange intermediate, Xk, can be excreted or imported by a reversible transporter,
synthetic costs of amino acids and their frequency of usage Em!2. The rate of ATP production is given by J ATP p n1J S ! n2J P.

(5)

nstant
bolism
trient.
e pro-
uation
Eqs. 3–
source Fig. 1. Possible trade-off curves ra ! g(rg) between maximal growth
Doebeli M, Pop. Ecology 44:2 (2002)
due to rate on glucose, rg, and maximal growth rate on acetate, ra. Shown are
the functions ra ! u0 " v0r 2g with u0 ! 3 and v0 ! 0.1, and ra !
 Testing effect of trade-offs on the evolution
of cross-feeding EvoFBA

Species A

FBA2
Inflow Outflow
medium medium

FBA1
Species B

FBA boundary conditions are optimised independently by each species

At each time interval ∆t

Growth rate is scaled by FBA and nutrients removed/added from the medium

Similar to: Harcombe W. et al.Curr. Cell Rep. 7:4 (2014)

Chiu et. al., PLCB 10:7 (2014)

Incorporating evolution and trade-offs:

EvoFBA

Uptake Reaction Bounds

Species A n

Species A ∑α i = const.
i=1
FBA2
n

Species B ∑α i = const.
i=1

FBA1 Mutation
n

Species B Species B’ ∑α i = const.

i=1

Ace
Glu

…..
Ser
Inspired from:
Continue ad infinitum Species …. Beg et al.PNAS 104:31 (2007)
Zhuang et. al., MSB 7:500 (2011)
Simulating evolution of E.coli using EvoFBA

1/100 1/100

Continue ad infinitum
24h 24h

in silico version of
Lenski,'R.'et'al.,'1991.#Long'term#experimental#evolu3on#in#Escherichia#coli.#I.#
Adapta3on#and#divergence#during#2,000#genera3ons.#American#Naturalist,#
138(6),#pp.1315–1341.#
Many clones emerge over in silico evolution

•  98678&&clones&
generated&
•  3978&&&(~4%)&clones&
survived&a&sub<
culturing&event&
•  235&+/<&30&clones&
present&each&day&
Two clones dominate…
Trade-off leads to evolution of specialization
and cross-feeding

Species A
Glucose specialist

“Fermenter”
FBA

FBA

Species B
Acetate specialist
Grosskopf T et al. BMC Evol. Biol 17:163 (2017)
“Respirer”
FBA
See also:
Pfeiffer T & Bonhoeffer S, Curr. Am. Nat. 163:6 (2004)
Doebeli M, Pop. Ecology 44:2 (2002)

 Workpackage A. Establishing the design principles of and experimental tools fo

Cellular trade offsmicrobial
controlling can shiftmetabolicmetabolism
interactions.
to fermentative pathways.
The overarching What
vision of this are isthe
workpackage to enable rational engineering at
species. Achieving such rational engineering will require modelling and predicting
consequences?
points based on their metabolism, and characterising and manipulating such interactio
will develop these tools and enable easier and systematic engineering of microbial met

Project A1. Establishing a t

modelling framework for meta
microbial organisms display ox
they utilise under the presence o
acceptors for complete oxidati
fermentative pathways that they
Ethanol, Acetate, Lactate,
that result
Butyrate, … in incomplete oxidatio
5). It is also possible that under
conditions, e.g. at high levels of
both pathways
CH4, CO2,are H2 utilised to har
energy from oxidative pathways
High energyFigure 5. A cartoon representation of metabolic
respiratory Low energy fermentative
substrate into fermentation40,41,45,
flux among oxidative (black) and fermentative
pathways (red) pathways. In presence (A) ofpathways
strong electron these two modalities of growth,
acceptors (blue), the oxidative pathways would can be understood in the conte
carry most of the flux, while in their absence (B), and resource allocation.
ΔG0 < −500kJ / mol
fermentative pathways would dominate. ΔG > −300kJ / mol
0 We will develop a mo
microbial interactions that will
point. The reason for this focus is that the interplay and choice between oxidative and
in each species would exert a direct effect on metabolic interactions among species tha
the end-products of fermentative pathways.
 Metabolic syntrophy is a necessity?

Certain degradations in certain environments266 SCHINK

only possible (thermodynamically) in syntrophy TABLE 2. Pure or defined cultures of bacteria catalyzing syntrophic substrate oxidations via interspecie
organized on the basis of the usual substrate
Isolate Substrate range Gram ty

Oxidation of primary alcohols !

Strain S Ethanol !
Desulfovibrio vulgaris Ethanol " sulfate !
Thermoanaerobium brockii Ethanol, sugars, etc. "
Pelobacter venetianus Ethanol, propanol
Pelobacter acetylenicus Ethanol, acetylene !
Pelobacter carbinolicus Ethanol, 2,3-butanediol

Oxidation of butyrate and higher homologs

Syntrophomonas wolfei C4–C8 !
Syntrophomonas sapovorans C4–C18 !
Syntrophospora bryantii C4–C11, 2-methylvalerate "
Strains SF-1, NSF-2 C4–C6 "
“Thermophilic coculture” C4 ?
“Thermophilic coculture” C4 ?

Oxidation of propionate
Syntrophobacter wolinii Propionate !
Syntrophobacter pfennigii Pyruvate !
Thermophilic coculture Propionate " fumarate !
Strain MPOB Propionate " fumarate !

Oxidation of acetate
Strain AOR Acetate, ethanol, ethylene glycol "
Clostridium ultunense Acetate, formate, cysteine "

Oxidation of isovalerate
Schink B Microbiol Mol Biol Rev 61:2 (1997) Strain GraIva1 Isovalerate only "

Oxidation of glycolate
Syntrophobotulus glycolicus Glycolate, glyoxylate "

Oxidation of aromatic compounds

Syntrophus buswellii Benzoate, crotonate !
Syntrophus gentianae Benzoate, gentisate, !
 Metabolic syntrophy is a necessity?

heterogeneous biofilms and granules, enhances the metabolic interactions in a BMC10,11, and that
diversity and redundancy increase BMC stability and
Hydrogen
12
performanceDesulfovibrio
. Neither of these Sul de proposals is
convincingly proven.
vulgaris
It has been argued that synthetic design of
novel communities or manipulation of existing ones can
be utilised toLactate
address some of theH2above questions2
and to rationally
Sulfate engineer complex and relevant
functions4. The latter view underpins a new wave of
synthetic biology research, which has already resulted
in the engineering of “toy communities” with specific

}
population dynamics6,7. While not driven by either an
ecological or synthetic biology agenda, other studies
defined several co- and tri-cultures capable
Desulfovibrio Methanococcus
of
degradation of organic
vulgaris compounds into maripaludis
methane 13-17.

The most recent of these used metabolic models to

study a specific co-culture18, and subsequently showed
Lactate CH4
that this minimal synthetic community H2 is capable of
converting lactate into methane 19 . Currently it
remains unknown how much the complexity of
such simple communities can be extended (e.g. by
SchinkFigure 1. Schematic
B Microbiol of the
Mol Biol Revmultistep
61:2 (1997)
increasing species numbers) to achieve an
degradation of organic material in
methanogenic ecosystems. Each increase in the range of substrates they can digest
step is carried out by a specific and their biomethane productivity.
functional group of micro-organisms
(also known as guilds). 2.Preliminary work leading up to this proposal
Lactate + 2Sulfate + 3H + In recent published and unpublished
3CO + 2HS + 3H O - studies, we have
ΔG0 = -259.09 kJ/mol
been developing a2better understanding of2stability and
function in microbial communities and how these are affected by the evolutionary process. Here,
we summarise key findings and insights from these studies which are relevant for this proposal.
Lactate + H2O Acetate + CO2 + 2H2 ΔG0 = -8.79 kJ/mol
2.1. Species composition and diversity of BMCs can vary over time and with bioreactor
conditions. Different studies have shown that diversity and composition of species in a BMC can
Lactate + H O
be stable or temporally
2 Acetate + 0.5CO + 0.5CH
varying, with seemingly no effect on performance.
2 For example,
4 amplified ΔG0 = -74.19 kJ/mol
ribosomal DNA restriction analysis of species abundance identified temporal variation in both
fi
 n mutualism, we isolated the species populations from These evolutionary improvements in coculture growth rate
culture and used them to produce cocultures of mixed could represent general adaptations to conditions that are the
We compared these cocultures with cocultures that had same in both treatments, such as the challenge of growing on
estral or only evolved populations.
Metabolic syntrophy is a necessity?
lactate without an electron acceptor. In that case, evolved
cocultures would perform similarly whether they were examined
in the heterogeneous environment or in the uniform environ-
ary Changes in Stability of Syntrophic Communities. The ment. Alternatively, the populations might have adapted to dif-

EVOLUTION
ge of mutualist evolution was characterized by erratic ferent ecological conditions in the two environments. In that
Fig. 1). Cocultures typically consumed all of theDesulfovibrio
resources vulgaris
case, (sulfate
cocultures reducing
that evolved bacteria)
in the uniform environment would
week and achieved stationary-phase densities between 0.25 perform poorly in the heterogeneous environment and vice versa.
OD600 nm. However, after the first four transfers, few We investigated
CHa3CHOHCOO - + H 2O this CH
possibility
3COO- by measuring
+ CO 2 + 2H2 the growth
∆G = rate
-8.8 of
kJ/mol
K
es did not show appreciable growth (0.0–0.06 OD600); all of the evolved cocultures in their alternate evolution envi-
eq

ere not transferred until they reached stationary-phase ronment (Fig. 2A and Table S1). The magnitude of improvement
Methanococcus
1 week or more later. Every coculture entered one of these maripaludis
in coculture (hydrogenotrophic
growth rate methanogen)
was not affected by environment (F1,18 =
wth phases at least once during the first 6 months of 0.41; P = .531) or by the interaction between the evolution and
ion. The final densities achieved by slow-growing COcocul- CH4 + H2O(F1,136 = 2.03; P = .157). Thus, it ∆G
2 + 4H2assay environment appears that kJ/mol
= -130.7
ied from 0.15 to 0.35 OD600. The timing, frequency, and K
the eq
populations have mostly adapted to general aspects of syn-
of slow growth varied considerably among the cocultures, trophic growth, not specifically to the heterogeneity of the
∆G = ∆G0’ + RTln(Γ)

0.5
At equilibrium; Keq = Γ
OD600 nm at time of transfer

0.4
kf [CO2][H2]4 = kr[CH4][H2O]
0.3

kf / kr =[CH4][H2O] / [CO2][H2]4
0.2

0.1

0
1 45(~300 doublings) 60 112
Transfer

A ‘natural’ syntrophy ‘evolving’ increased stability.

Fig. 1. Coculture density at each transfer during the evolution experiment.

and Stahl Hillesland, KL & Stahl, DA PNAS 107:5 (2010) PNAS | February 2, 2010 | vol. 107 | no. 5 | 2125

Engineering Metabolically Coupled Communities

Desulfovibrio
vulgaris Methanococcus
maripaludis

Lactate
CH4
Sulfate H2
0.50"

0.25"

0.6"
0.00"
0.5"
OD600%(in%Hungate)%

0" 5" 10" 15" 20"

0.4" Days%a:er%Inocula=on%

0.3"

0.2"

0.1"

0"
1" 3" 5" 7" 9" 11" 13" 15" 17" 19" 21" 23" 25" 27" 29" 31" 33" 35" 37" 39"
Replicate%number%(unrelated)%
Engineering Metabolically Coupled Communities
Hypothesis: Syntrophy enabling mutation allows energy investment to
overcome thermodynamic hurdle:
Lactate + H2O Acetate + CO2 + 2H2 ΔG0 = -8.79 kJ/mol
DVU2287
DVU2287 DVU3023
DVU2287 DVU3023
5798 WALKER ET AL.
Lactate Pyruvate + H2 ΔG0 = 43 kJ/mol
? Wild type
Na+ (out) Na+(in) Dv
Sytnrophic
Dv

Grosskopf T. et al. ISME J. (2016).

 Mid Summary 2
Metabolic interactions are common in microbial communities

Cross-Feeding
Syntrophy
Complementary
Auxotrophy

This might be because they increase tness (shown for some

interactions)

Metabolic interactions can emerge and evolve due to environmental and

biophysical conditions (e.g. availability of electron donors / acceptors,
thermodynamic limits)
fi
Engineering Metabolically Coupled Communities
dysfunctional
natural
community
‘Microbiome Engineering’

functional natural Enrichment

community

~~
Cross-Feeding
Metabolic Cycles

functional ‘synthetic’ Module based

‘design’
community
Syntrophy
Complementary
Auxotrophy
 not allow specific kinds of bacteria to be targeted [12]. How
constrained by a lack of appropriate tools (Box 2). Thus, new s
Microbiome Engineeringthe key mechanisms underlying microbiome-associated phe
ical roles of specific bacteria. These strategies may allow long
mented for pathological conditions to which resident bacte
dysfunctional
To broaden our understanding of microbiome-linked disease
natural
the in situ production of currently available or novel therape
community
alizable technologies to engineer commensal microbes. Her
Trends in Immunology
‘Microbiome of microbiome engineering needed for understanding host–
Engineering’
of the most pressing challenges in the field.
Review
Emerging Frontiers in Microbiome Engineering
Marı́a Eugenia Inda,1,2,5 Esther Broset,3,4,5 Timothy K. Lu,1,2,* and Cesar de la Fuente-Nunez3,*

The gut microbiome has a significant impact on health and disease and can actively contribute to
952 Trends in Immunology, October 2019, Vol. Highlights
40, No. 10
obesity, diabetes, inflammatory bowel disease, cardiovascular disease, and neurological disor- The microbiome plays a funda- https:/
ª 2019 Published by Elsevier Ltd.
ders. We do not yet have the necessary tools to fine-tune the microbial communities that consti-
tute the microbiome, though such tools could unlock extensive benefits to human health. Here,
mental role in our health and
disease.
we provide an overview of the current state of technological tools that may be used for micro-

Indeed functionally distinct ‘states’ seem to

Engineering the microbiome might
biome engineering. These tools can enable investigators to define the parameters of a healthy
enable studying the contribution of
microbiome and to determine how gut bacteria may contribute to the etiology of a variety of
individual microbes and generating
diseases. These tools may also allow us to explore the exciting prospect of developing targeted
potential therapies against meta-
therapies and personalized treatments for microbiome-linked diseases.

exist in complex communities and depend

bolic (e.g., phenylketonuria and
chronic kidney disease), inflamma-
tory, and immunological diseases,
Modulating the Microbiome: An Emerging Paradigm for Understanding and Treating among others.
Human Diseases

on their species composition.

Current methods for probing the
The human body is home to at least as many microbial cells as human cells [1]. However, the most
microbiome include fecal micro-
salient characteristic of the interaction between microbes and the human body is not the number biota transplantation and the use of
of cells involved, but their inextricable link with each other. The microbiome (i.e., the microbial com- antibiotics, probiotics, and pre-
munities living within our bodies) can affect metabolism, development, immunity, and other aspects biotics; these can produce broad
of human health. Imbalances in the microbiome may alter an individual’s health status. The ‘hygiene alterations in the microbiome and
hypothesis’ (see Glossary) suggests, for example, that loss of microbial diversity in the human micro- its reciprocal interactions with the
biome may underlie the recent increase in the incidence of asthma [2]. immune system.

Novel tools are needed to precisely

Bacteria exert their effects not only locally where they reside, but also at other sites of the human
reprogram microbial communities
body; they release molecules that travel to other tissues, such as the liver or brain, traditionally consid- in order to improve health and
ered sterile (Box 1). It is estimated that gut microbial metabolites represent 10% of the metabolites disease outcomes (among other
found in mammalian blood [3] and their effects on hosts are only starting to be identified, as reviewed applications).
elsewhere [4]. As a result of these novel insights, animals are no longer considered autonomous units
but, rather, ‘holobionts’, the ensemble of the host and its microorganisms [5,6].

Existing methods to modulate the microbiome can cause sweeping alterations (e.g., fecal microbiota
transplantation, antimicrobial drugs, and prebiotics). These methods are inadequate for understand-

Raguideau, et al. unpublished results

ing how the microbial community is structured and which particular species, strains, or secreted pro-
teins produce the observed effects. Existing methods are also not refined enough to fine-tune dys-
1Synthetic Biology Group, MIT Synthetic

Biology Center, Department of Biological

biosis. For example, fecal transplant to treat Clostridium difficile diarrhea is clinically successful in Engineering and Department of Electrical
Engineering and Computer Science,
humans [7]; nevertheless, it is difficult to standardize and such therapies can also perturb the micro-
https://www.biorxiv.org/content/10.1101/2021.07.05.451125v1
Massachusetts Institute of Technology,
biome quite broadly [8], as the FDA has also warned [9]. Antibiotics frequently used to treat infections Cambridge, MA 02139, USA
cause broad-spectrum changes to microbiome communities and their overuse can lead to resistance 2Research Laboratory of Electronics,

[10], reducing the proportion of beneficial microbes within the community [11]. Prebiotics and dietary Massachusetts Institute of Technology,
Cambridge, MA 02139, USA
interventions that induce the growth of beneficial bacteria may be useful for some indications, but do
3Machine Biology Group, Departments of
not allow specific kinds of bacteria to be targeted [12]. However, current microbiome engineering is Psychiatry and Microbiology, Perelman
constrained by a lack of appropriate tools (Box 2). Thus, new strategies are being sought to determine School of Medicine, and Department of
Bioengineering, University of

649
the key mechanisms underlying microbiome-associated phenotypes and to pinpoint the physiolog-
Pennsylvania, Philadelphia, PA, USA
ical roles of specific bacteria. These strategies may allow long-lasting, effective therapies to be imple-
4
 not allow specific kinds of bacteria to be targeted [12]. How
constrained by a lack of appropriate tools (Box 2). Thus, new s
Microbiome Engineering the key mechanisms underlying microbiome-associated phe
ical roles of specific bacteria. These strategies may allow long
mented for pathological conditions to which resident bacte
dysfunctional
To broaden our understanding of microbiome-linked disease
natural
the in situ production of currently available or novel therape
community
alizable technologies to engineer commensal microbes. Her
Trends in Immunology
‘Microbiome of microbiome engineering needed for understanding host–
Engineering’
of the most pressing challenges in the field.
Review
Emerging Frontiers in Microbiome Engineering
Marı́a Eugenia Inda,1,2,5 Esther Broset,3,4,5 Timothy K. Lu,1,2,* and Cesar de la Fuente-Nunez3,*

The gut microbiome has a significant impact on health and disease and can actively contribute to
952 Trends in Immunology, October 2019, Vol. Highlights
40, No. 10
obesity, diabetes, inflammatory bowel disease, cardiovascular disease, and neurological disor- The microbiome plays a funda- https:/
ª 2019 Published by Elsevier Ltd.
ders. We do not yet have the necessary tools to fine-tune the microbial communities that consti-
tute the microbiome, though such tools could unlock extensive benefits to human health. Here,
mental role in our health and
disease.
we provide an overview of the current state of technological tools that may be used for micro-
Engineering the microbiome might
biome engineering. These tools can enable investigators to define the parameters of a healthy
enable studying the contribution of
microbiome and to determine how gut bacteria may contribute to the etiology of a variety of
individual microbes and generating
diseases. These tools may also allow us to explore the exciting prospect of developing targeted
potential therapies against meta-
therapies and personalized treatments for microbiome-linked diseases. bolic (e.g., phenylketonuria and
chronic kidney disease), inflamma-
tory, and immunological diseases,
Modulating the Microbiome: An Emerging Paradigm for Understanding and Treating among others.
Human Diseases
Current methods for probing the
The human body is home to at least as many microbial cells as human cells [1]. However, the most

How to shift a community?

microbiome include fecal micro-
salient characteristic of the interaction between microbes and the human body is not the number biota transplantation and the use of
of cells involved, but their inextricable link with each other. The microbiome (i.e., the microbial com- antibiotics, probiotics, and pre-
munities living within our bodies) can affect metabolism, development, immunity, and other aspects biotics; these can produce broad
of human health. Imbalances in the microbiome may alter an individual’s health status. The ‘hygiene alterations in the microbiome and
hypothesis’ (see Glossary) suggests, for example, that loss of microbial diversity in the human micro- its reciprocal interactions with the
biome may underlie the recent increase in the incidence of asthma [2]. immune system.

Novel tools are needed to precisely

Bacteria exert their effects not only locally where they reside, but also at other sites of the human

Hit it with a hammer! ‘Probiotic microbes’, phages, antibiotics,

reprogram microbial communities
body; they release molecules that travel to other tissues, such as the liver or brain, traditionally consid- in order to improve health and
ered sterile (Box 1). It is estimated that gut microbial metabolites represent 10% of the metabolites disease outcomes (among other
found in mammalian blood [3] and their effects on hosts are only starting to be identified, as reviewed applications).

metabolites, pH?
elsewhere [4]. As a result of these novel insights, animals are no longer considered autonomous units
but, rather, ‘holobionts’, the ensemble of the host and its microorganisms [5,6].

Existing methods to modulate the microbiome can cause sweeping alterations (e.g., fecal microbiota
transplantation, antimicrobial drugs, and prebiotics). These methods are inadequate for understand-
ing how the microbial community is structured and which particular species, strains, or secreted pro- 1Synthetic Biology Group, MIT Synthetic
teins produce the observed effects. Existing methods are also not refined enough to fine-tune dys- Biology Center, Department of Biological
biosis. For example, fecal transplant to treat Clostridium difficile diarrhea is clinically successful in Engineering and Department of Electrical
Engineering and Computer Science,

Clear understanding lacking to do this predictably so far. One

humans [7]; nevertheless, it is difficult to standardize and such therapies can also perturb the micro- Massachusetts Institute of Technology,
biome quite broadly [8], as the FDA has also warned [9]. Antibiotics frequently used to treat infections Cambridge, MA 02139, USA
cause broad-spectrum changes to microbiome communities and their overuse can lead to resistance 2Research Laboratory of Electronics,

[10], reducing the proportion of beneficial microbes within the community [11]. Prebiotics and dietary Massachusetts Institute of Technology,
Cambridge, MA 02139, USA

insight is that high nutrient levels reduce community diversity.

interventions that induce the growth of beneficial bacteria may be useful for some indications, but do
3Machine Biology Group, Departments of
not allow specific kinds of bacteria to be targeted [12]. However, current microbiome engineering is Psychiatry and Microbiology, Perelman
constrained by a lack of appropriate tools (Box 2). Thus, new strategies are being sought to determine School of Medicine, and Department of
the key mechanisms underlying microbiome-associated phenotypes and to pinpoint the physiolog- Bioengineering, University of
Pennsylvania, Philadelphia, PA, USA
ical roles of specific bacteria. These strategies may allow long-lasting, effective therapies to be imple-
4

Fig. 2. (A) In our mathematical model, species S1 and S2 share a sub-
strate containing nutrients and toxins at concentrations CN and CT . The
Predictions. In the model, positive interactions dominate
toxicity, given that sufficient nutrients are present. I
ing nutrient concentrations further or reducing toxicity
increases competition. We assumed that our bacteria

Enriching Simpli ed ‘Synthetic’ Communities species take up the same nutrients, and invest a fraction of these into toxin
degradation and the rest into population growth. Toxins cause cell death
and population decline. (B) Example results of the model (parameters in
SI Appendix, Table S3), shown as the abundance of species S1 (solid line)
and concentrations of nutrients and toxins (dashed and dotted lines, respec-
tively). In monoculture, S1 goes extinct due to toxins (Left), but survives in
coculture with S2 (Right). (C) The response of one species to the presence
of another is measured as the difference in AUC between the coculture and
monoculture (color and parameters in SI Appendix, Table S3) and shown
as a function of nutrient and toxin concentrations. At high toxin concen-
MWF lay at the point in the state space where positive i
tions are favored, and modified the environment in 3 add
experiments to test the model’s predictions.
We first increased the concentration of nutrients in the
medium by adding 1% Casamino acids (AA) (see Materi
Methods), which is a nutrient source for 3 out of the 4
(SI Appendix, Fig. S8). In this supplemented MWF m
(MWF + AA), monocultures of A. tumefaciens and C.
teroni immediately grew well, while M. saperdae and O. a

~~
trations and intermediate nutrients, interactions are positive (+ve) due to
the joint degradation of toxins (as in B). As nutrients are increased or tox-
ins decreased, competition for limited resources dominates (-ve, short for
“negative”).
Enrichment first. The lower the toxin concentration, the faster this competi-
tive effect arises (Fig. 2C). In sum, high toxicity and intermediate
nutrients, where species cannot survive alone, is where species
in our model benefit from the presence of others. We hypoth-
esized that this regime best describes the 4 species’ growth
in MWF.
Cellulose Degradation: When the 2 species have the same model parameters, positive
interactions rely on the coculture being inoculated with twice as
still suffered from its toxicity (SI Appendix, Fig. S9). Acc
to the model, we expect competition between the 2 spec
could grow. Indeed, the 2-way positive interaction betw
testosteroni and A. tumefaciens switched to negative in 1
tion (Fig. 3B), indicating that a change in nutrient comp
can radically modify bacterial interactions. The 2 speci
still experienced the environment as toxic (M. saperdae
anthropi) became the only 2 species benefiting from being
wise cocultures. They also started to benefit from A. tume
and benefited more from C. testosteroni that could grow
(and presumably detoxify faster) in this medium than in M
Second, we reduced toxicity by growing the bacteria in 1
Ideally, we would have removed toxic compounds from

7104 KATO ET AL.

Kato, S et al. App. & Env Microbiol 71:11 (2005)
FIG. 3. Network model of substrate flow and the putative roles of
each member in SF356. The solid lines indicate the flow of the sub-
strates. The thickness of each line represents the relative contribution
of a particular pathway in SF356. The dotted lines indicate promoting
or inhibiting factors with respect to the efficiency of cellulose degra-
dation. The names of the substrates are presented in boxes. The des-
ignations of the bacterial strains are presented in ovals.
many cells as the monoculture, and hence twice the degradation
effort.
APPL. ENVIRON. MICROBIOL . According to our experiments, however, positive interac-
Industrial Waste Degradation: but MWF is chemically complex and only sold as a fi
product. By removing MWF entirely, the growth mediu
tions still dominate even if the total cell number at the beginning no longer toxic, but nutrients were also reduced and ma
Five species cellulose-
aerobic static conditions, and the degradation efficiencyiswas
constant, suggesting that facilitation occurs because different become differently accessible. Caveats aside, according
significantly low even under anaerobic conditions. Thesespecies
ob-
Four species community degrading toxic
degrade different toxins (SI Appendix, Fig. S5). To better model, we expected negative interactions to increase. I
degrading community represent this effect, we extended our model in SI Appendix, sec-
servations suggested that the contributions of aerobic noncel- we found all interactions to be negative, except for M. sa
lulolytic bacteria (i.e., the consumption of oxygen, pH neutral-
ization, and stimulation of growth) (20) (Fig. 3) are essential
for effective cellulose degradation by the community. In addi-
industrial waster
tion, the growth of noncellulolytic isolates was promoted in the A B C
cell-free culture filtrate of C. straminisolvens CSK1 (Table 1).
Thus, the results are suggestive of a synergistic relationship
between C. straminisolvens CSK1 and the aerobic noncellulo-
lytic bacteria.
(ii) Clostridium sp. strain FG4. Clostridium sp. strain FG4 is
an anaerobic noncellulolytic isolate. It appears that Clostridium
sp. strain FG4 consumes saccharides derived from cellulose
degradation and produces a large amount of acetate and a
small amount of ethanol (Fig. 3), as based on the knockout
D E F
experiments (Fig. 2C to E). This prediction was also supported
by the results of the pure-culture experiments (Table 1).
Water-soluble oligosaccharides, especially cellobiose, are
Piccardi, P et al. PNAS 116:32 (2019)
known to repress cellulase expression (18, 33), to inhibit cel-
lulase activities (19, 23), and to inhibit cellulose degradation
(24) by cellulolytic clostridia. Hence, the consumption of sac-
charides by Clostridium sp. strain FG4 is assumed to have a
positive effect on achieving effective cellulose degradation.
However, although the accumulation of saccharides in !FG4

Cyanobacterial Granules:
istence of two bacterial strains in mutualistic or commensalistic
association (e.g., see references 10 and 31). In addition, some
groups have reported the construction of a defined mixed cul-
ture of more than two isolates in order to reveal metabolic
17 species cyanobacterial
sharing and synergistic interactions (2, 5, 6). For example,
Dejonghe et al. (6) isolated bacteria corresponding to all of the
was higher than that in the case of SF356 (Fig. 2C), !FG4
exhibited more effective degradation of filter paper than SF356
Fig. 3. Pairwise interaction networks under different environmental conditions. Positive/negative interactions indicate that the species at the en
(Fig. 2A). This observation indicated that Clostridium sp. strain
arrow grew significantly better/worse in the presence of the species at the beginning of the arrow in (A) MWF, (B) MWF + AA, and (C) AA medium
FG4 has a negative effect on cellulose degradation, andthickness
this
represents interaction strength as the 10-fold change in the coculture AUCs compared with monoculture AUCs, i.e., by how many o
effect exceeded the positive effects exerted by the saccharide
magnitude a species changed the AUC of another. Statistical significance and interaction strengths were calculated based on 2 experiments in A
scavenge. Although the pH value in !FG4 dropped once below Fig. 1 and SI Appendix, Fig. S1), and 1 experiment in B (SI Appendix, Fig. S9) and C (SI Appendix, Fig. S8). P values and interaction strengths are

community
major bands detected by DGGE analysis from a stable micro-
flora that degrades the herbicide linuron, and they constructed
6 before day 3, it returned to an approximately neutral value,
and the acidic pH in SF356 (the mixed culture containing
Dataset S1. (D) Monoculture and coculture growth curves of ancestral At and (E) Ct versus the same strains evolved in monoculture for 10 wk (AtT10
Coculture partners are indicated in brackets. (F) Interactions between ancestral and evolved At and Ct strains based on growth curves in D and E
by guest on February 3, 2020

a defined mixed culture that degraded linuron in order to
reveal the relevant synergistic interactions. However, they did
https://www.biorxiv.org/content/10.1101/2022.12.13.520286v1
not make note of the stable coexistence of the members of the
community. To the best of our knowledge, there appears to
have been no reports to date regarding the construction of a
Clostridium sp. strain FG4) was generally maintained during widths and asterisks are as defined for A–C. The interactions between At and Ct in A and F have different strengths and P values because they co
the cultivation period (Fig. 2B). The production of aceticdifferent
acid experimental repeats.
by Clostridium sp. strain FG4 would be expected to maintain an
acidic pH. This could account for the negative effects on cel-
lulose degradation exerted by C. straminisolvens CSK1, which Piccardi et al. PNAS | August 6, 2019 | vol. 116 | no. 32
fi
Learning from Nature to ‘engineer’ metabolically
coupled synthetic communities?
Module based
‘design’

design principles
modules
Cellular Trade Offs
Thermodynamic Inhibition Cross-Feeding
Metabolic Cycles

systems
Syntrophy
Complementary
Auxotrophy
 Modules for synthetic community engineering
Desulfovibrio
vulgaris Methanococcus
maripaludis
Syntrophy motif - anaerobic. Could
g interactions
Lactate
CH4
be used for methane production
S Pande et al Sulfate H2
ACS Synthetic Biology Letter
e cases,
udied in
as the
Auxotrophy motif - aerobic. Could
Inerney
chment be used for evolutionary stability?
., 2008)
specific Pande, S et al. ISME J 8:5 (2014)
unately,
s makes
ractions
.,iology
2002; pubs.acs.org/synthbio Research Article
Cross-feeding motif - aerobic.
Figure 1 Design strategy of synthetic cross-feeding interactions.
Moran, Overview over the genes deleted in E. coli wild type (WT) to yield
uestions one of four single-gene deletion mutants that are either auxo-
utionary
obligate
trophic for one amino acid (AA, that is, arginine, tryptophan,
leucine and histidine) or overproduce AAs, as well as double
deletion mutants (that is, ‘cross-feeders’), in which the two
Could be used for bioproduction?
mutations causing AA auxotrophy and overproduction were
g inter- combined in all possible combinations. Coculturing two of those Lalwani MA et al ACS Syn Bio 10:8 (2021)
m: why cross-feeders should result in reciprocal exchange of essential
bolite to amino acids (inset). Bernstein HC et al J Biotech 157 (2012)
r itself?
cause of Weiss TL et al Metabolic Eng. 44 (2017) Santala S et al. PLoS One 9:12 (2014)
on the among two bacterial genotypes? (2) Are cooperative
utionary cross-feeding interactions vulnerable to the exploi-
e to a tation of
non-cooperating genotypes? Our results provide
sm. For
compar- evidence for a significant fitness advantage of
obligate cross-feeding relative to metabolic auto-
Reciprocal cross-feeding motif -
related
oopera-
e rarely
nomy and suggest that the metabolic cross-feeding
interactions can stably coexist with other, non-
phototrophic. Could be used for
Second,
e cross-
cooperating genotypes—even in the absence of
spatial structure. bioproduction? Smith, MJ et al. ACS Synthetic Biology 5 (2016)
 Towards a closed microbial ecosystem?
ter marine
Joe Hanson
ome, of the
83, NASA Self-sustaining system
driven by sunlight??
neering &
cson, AZ .
duced from
and filtered
d Gorgonia
iving, for
ng to the
//www.eco-
shrimp
ecosystems
many homes,
replicated
EcoSphere by Hanson J and Folsome C
which will be reviewed in some detail below. His

on of NASA’s Jet Propulsion Laboratory resulted in the

osed ecosystems as EcoSpheres® (Figure 2.1). Lactate

Acetate
products?
system? Mn2+
stems can persist under closure. For example, what are the

pecies? What interaction patterns are stable? What, if any,

MnO2
ersity required? In general, closed ecosystems can be
Roseobacter Shewanella
AzwK-3b oneidensis MR-1
in top-down construction, a sample from a natural system Control% Freeze%both%samples%
Phototrophic granule (by Soyer OS et al)
er closure until a remaining set of species persists; in Acetate Lactate 3%days’%supernatant%
Freeze% %%
https://www.biorxiv.org/content/
ual species from stock cultures are combined in a suitable
10hr%supernatant% sample% 6%days’%supernatant%
10.1101/2022.12.13.520286v1 %% %%
Mix%with%same%volume% 9%days’%supernatant%
fresh%medium% %%
Divide%into%two%flasks%
14
Challenges in microbial ecosystem design

The importance of spatial organisation?

Roseobacter
Aerobic, liquid phase for acetate
AzwK-3b
MnO2
and Mn oxidation

Mn2+
Anaerobic phase devoid of
Shewanella
oneidensis MR-1 terminal electron acceptors,
favouring MnOx reduction
 Questions in a closed microbial ecosystem

Can system behaviour be

explained by principles of
entropy maximisation?
Vallino JJ Phil. T. Roy. Soc. B 365:1545 (2010)

Evolution of new
Lactate
Acetate
products? interactions in light of
existing interactions??
Mn2+

MnO2
Roseobacter Shewanella
AzwK-3b oneidensis MR-1
Control% Freeze%both%samples%
Acetate Lactate 3%days’%supernatant%
Freeze%
sample%
%%
6%days’%supernatant%
Evolution of system
Mix%with%same%volume%
%%
9%days’%supernatant% robustness against variations
fresh%medium%
Divide%into%two%flasks%
%%
in environmental conditions?
Closed Ecosystems….Applications?

Mass Balance

Inputs: Outputs:

Urine
O2
Feaces
Water
CO2
Food
Transpiration
Contaminants
(Microbial
Chemical)
Engineering Principles Natural Principles

Optimise Evolution & Thermo

Measure & dynamics
Model Ecology
Modular
Trade-Offs
Design

Microbial Community Engineering….

Bespoke waste Plant supporting Life-support systems

conversion and minimal for enclosed spaces
bioproduction communities

 Engineering Metabolically Coupled Communities

A re-mineralization cycle

Polymetallic (Mn-rich) nodules of biogenic origin.

tion. 10.0
Lactate ●

NO3− (in mM)

● ● ● ●
●
Blöthe et al., Environ. Sci. Technol. 2015, 49, 7692. 7.5 ● ●

5.0 ●
●
Tebo et al., Annu.Rev.EarthPlanet.Sci. 2004, 32 (1), 287. 2.5
Roseobacter
Remucal CK, Ginder-V. M, Environ.Sci.: Processes Impacts 2014, 16, 1247. 0.0
AzwK-3b
Mn
Li et al., Current2+
Biology 2016, 26, 950.
0 1 2 3
Rhoads et al., Environ. Sci. Technol. 2005, 39 (12), 4666.
Time (days)
Sw.

Acetate (in mM)

10.0 ●
Shewanella
MnO2 oneidensis MR-1
7.5 ●

5.0
Rb. 2.5
●
0.0 ● ● ● ●

0 1 2 3

Acetate Time (days)

Lactate (in mM)

10.0 ●
●
7.5 ●

5.0
Rb.
2.5
0.0 ● ● ● ●

0 1 2 3
Sw.
Time (days)

Individual growth in bi-culture further

con rmed by selective plating
fi
Exploiting Metabolically Coupled Communities
A re-mineralization cycle

We need a spatially engineered system?

Roseobacter
Aerobic, liquid phase for acetate
AzwK-3b
MnO2
and Mn oxidation

Mn2+
Anaerobic phase devoid of
Shewanella
oneidensis MR-1 terminal electron acceptors,
favouring MnOx reduction
 Workpackage A. Establishing the design principles of and experimental tools fo
controlling microbial metabolic interactions.
How does metabolic cross-feeding emerge?
The overarching vision of this workpackage is to enable rational engineering at
species. Achieving such rational engineering will require modelling and predicting
points based on their metabolism, and characterising and manipulating such interactio
will develop these tools and enable easier and systematic engineering of microbial met
Sugars Sugars
Project A1. Establishing a t
modelling framework for meta
microbial organisms display ox
they utilise under the presence o
acceptors for complete oxidati
fermentative pathways that they
Ethanol, Acetate, Lactate,
that result
Butyrate, … in incomplete oxidatio
5). It is also possible that under
conditions, e.g. at high levels of
Respiration both pathways are utilised to har
Figure 5. A cartoon representation of metabolic energy from oxidative pathways
flux among oxidative (black) and fermentative substrate into fermentation40,41,45,
(red) pathways. In presence (A) of strong electron these two modalities of growth,
acceptors (blue), the oxidative pathways would can be understood in the conte
carry most of the flux, while in their absence (B), and resource allocation.
fermentative pathways would dominate. We will develop a mo
microbial interactions that will
point. The reason for this focus is that the interplay and choice between oxidative and
in each species would exert a direct effect on metabolic interactions among species tha
the end-products of fermentative pathways.