Electronic Supplementary Material (ESI) for Molecular BioSystems
This journal is © The Royal Society of Chemistry 2012
Supplementary Information:
Network motifs provide signatures that characterize metabolism
Erin R. Shellmana,b , Charles F. Burantb and Santiago Schnellc,b
a Department
of Internal Medicine, University of Michigan Medical School, Ann Arbor, USA.
b Department
of Computational Medicine & Bioinformatics, University of Michigan Medical School,
Ann Arbor, USA. c Department of Molecular & Integrative Physiology, University of Michigan Medical School, Ann Arbor, USA. E-mail: [email protected]
Supplementary material for DOI: 10.1039/C2MB25346A
Organisms Used for Analysis
In this work, the kingdom of Animalia is represented by Homo sapiens and Mus musculus. The
metabolic reactions of H. sapiens were taken from RECON1 and the metabolic network reconstruction of M. musculus is largely based on RECON1 as well. Additional Eukaryotes include two
species of Fungi, Saccharomyces cerevisiae and Pichia pastoris, both of which are well-known model
organisms with similar life cycles and environmental requirements.
Archaea are represented by two closely related species, Methanosarcina barkeri and Methanosarcina
acetivorans and a third Halobacterium salinarum. Both methanogenic Archaea are known for
employing all three pathways of methane production, which makes them popular for use in the
development of biofuels.
There are two plants under investigation, Arabidopsis thaliana and Zea mays. Similarly, there
is one algae, Chlamydomonas reinhardtii, which belongs to the kingdom of Protista. C. reinhardtii
is motile, and uses two flagella to propel itself. It has a light-sensitive eyespot and is a popular
model organism for use in the development of alternative fuel sources.
In addition to the aforementioned organisms which comprise five of the six kingdoms of life, there
are eleven species of Bacteria that span a wide range of physiological features and environments.
S-1
Electronic Supplementary Material (ESI) for Molecular BioSystems
This journal is © The Royal Society of Chemistry 2012
Clostridium thermocellum is a Gram+ anaerobe and an extremophile that lives in thermophilic
environments. Many of the Bacteria in this set are interesting because of the role they play in human
diseases. For instance, Vibrio vulnificus is a close relative of Vibrio cholerae, the Bacterium that
causes cholera, and Escherichia coli, Staphylococcus aureus, Helicobacter pylori and Mycobacterium
tuberculosis are all well-known human flora and potential pathogens, implicated in diseases of the
respiratory system to the digestive tract.
Species
Kingdom
Nodes
Edges
Compartment
A. thaliana
Plantae
C. reinhardtii
Protista
C. thermocellum
D. ethenogenes
E. coli
H. pylori
H. salinarum
H. sapiens
Bacteria
Bacteria
Bacteria
Bacteria
Archaea
Animalia
G. sulfurreducens
M. acetivorans
M. barkeri
M. musculus
Bacteria
Archaea
Archaea
Animalia
M. tuberculosis
P. pastoris
Bacteria
Fungi
1501
50
57
660
25
260
48
516
501
908
400
526
779
184
234
189
352
85
135
466
697
542
842
182
262
205
385
85
140
486
571
19
16
225
3411
122
112
2165
58
652
56
1604
1498
2863
1194
1269
2181
402
591
351
905
173
335
908
1832
1602
2399
400
643
383
1019
176
342
1417
1774
22
20
576
Cytosol
[5]
Mitochondrion
Peroxisome
Cytosol
[3]
Golgi
Mitochondrion
Nucleus
Cytosol
[16]
Cytosol
[1]
Cytosol
[9]
Cytosol
[20]
Cytosol
[11]
Cytosol
[6]
ER
Golgi
Lysosome
Mitochondrion
Nucleus
Peroxisome
Cytosol
[14]
Cytosol
[13]
Cytosol
[10]
Cytosol
[18]
ER
Golgi
Lysosome
Mitochondrion
Nucleus
Peroxisome
Cytosol
[8]
Cytosol
[4]
ER
Golgi
Mitochondrion
S-2
Electronic Supplementary Material (ESI) for Molecular BioSystems
This journal is © The Royal Society of Chemistry 2012
R. etli
S. aureus
S. cerevisiae
Bacteria
Bacteria
Fungi
S. typhimurium
T. maritima
V. vulnificus
Z. mays
Bacteria
Bacteria
Bacteria
Plantae
36
74
350
549
528
15
11
214
30
73
852
727
831
1418
60
50
62
161
748
1657
1657
18
17
531
45
186
3102
2478
2494
2463
78
51
Nucleus
Peroxisome
Cytosol
[15]
Cytosol
[2]
Cytosol
[7]
ER
Golgi
Mitochondrion
Nucleus
Peroxisome
Cytosol
[19]
Cytosol
[21]
Cytosol
[12]
Cytosol
[17]
Mitochondrion
Peroxisome
There are more published metabolic network reconstructions available than were included in
the analyses. Criteria for inclusion were simply that the reconstructions were curated in SBML and
were readable into the COBRA toolbox in Matlab. These criteria insured that the reconstructions
were curated using similar protocols and adequately formatted and vetted for typographical errors.
Although compatability with COBRA was a requirement, neither COBRA nor Matlab were used
for analysis. Once each reconstruction was read into Matlab, we exported relevant data as plain
text files to use for motif mining. Specifically, we extracted the stoichiometric matrix, the reaction
and metabolite names, a dummy variable indicating the reversibility of each reaction and the
subsystem to which the reaction belonged (e.g. “Folate Biosynthesis,” “TCA Cycle”, “Salvage
Pathway of ATP”).
Motifs of Size Three
This work focused only on motifs of node-size three. All motifs were represented as substrate
graphs. Substrate graphs represent associativity of nodes, rather than mechanistic relationships
like those of a bipartite graph. Each graph type has its advantages and disadvanges. For instance,
when using bipartite graphs of size 3 it is possible to generate motifs that contain no biological
meaning. For example a bipartite motif might contain two nodes that represent reactions and one
S-3
Electronic Supplementary Material (ESI) for Molecular BioSystems
This journal is © The Royal Society of Chemistry 2012
that represents a metabolite, which is not a valid chemical mechanism. Similarly, because substrate
graphs are associations, we cannot know the chemical mechanism from the motif structure.
Motif Number
Motif Structure
Name
A
1
B
C
Concurrent
C
Trapping Reactions
C
Consecutive Reactions
C
Consecutive Reactions with Reversible Step
C
Trapping Reactions with Reversible Step
C
Reversible Consecutive Reactions
C
Feed-forward Reaction
C
Closed Cycle
C
Concurrent Reaction with Exchange
C
Trapping Reaction with Exchange
A
2
B
A
3
B
A
4
B
A
5
B
A
6
B
A
7
B
A
8
B
A
9
B
A
10
B
S-4
Electronic Supplementary Material (ESI) for Molecular BioSystems
This journal is © The Royal Society of Chemistry 2012
A
11
B
C
One Way Cycle with One Reversible Step
C
One Way Cycle with Two Reversible Steps
C
Reversible Cycle
A
12
B
A
13
B
Extracting Reaction Mechanisms from Substrate Graphs
As previously mentioned, one cannot infer reaction mechanisms from substrate graphs. In order
to do this as in Section 2.2, we enumerated every possible mechanism capable of yielding each of
A
the 13 motifs. For example the first motif,
B
C
, has two possible mechanisms: It could be either
C → A and C → B or C → A + B. In addition to enumerating both of these mechanisms, it is
also crucial to enumerate all the combinations of reversibility. It could be the case that the correct
mechanism for the first motif is C → A and C → B, but the C → B reaction is actually the reverse
direction.
In order to characterize each motif, we used the stoichiometric matrices from the E. coli, H.
sapiens, M. barkeri and S. cerevisiae metabolic network reconstructions. Stoichiometric matrices
contain integers that denote whether a metabolite is produced, consumed or not a participant in
a particular reaction. Negative integers denote consumption, positive integers denote production,
and zeros denote absence. First, we generated a second stoichiometric matrix that contained the
reverse mechanisms for all reversible reactions. We searched for motif mechanisms using a series of
conditional tests in R. For example to find the reaction C → A, it is we used:
which(Stoich[paste(motif1$nodeA[i]), ] > 0 &
Stoich[paste(motif1$nodeB[i]), ] == 0 &
Stoich[paste(motif1$nodeC[i]), ] < 0)
S-5
Electronic Supplementary Material (ESI) for Molecular BioSystems
This journal is © The Royal Society of Chemistry 2012
The “which” function will return the stoichiometric matrix column indices for reactions where
node A is being produced (A > 0), node B does not participate (B == 0) and node C is being
consumed (C < 0). Similar conditionals were used for all other combinations of reversibility.
Kingdom
0.5
0.0
1:
Archaea
Fungi
Plantae
Protista
-0.5
r
cu
e
tR
tio
ac
ns
Re
tio
ac
ns
Re
tio
ac
ns
ib
rs
le
ep
St
sib
er
ev
le
ep
St
Re
ac
tio
ns
Re
tio
ac
n
e
os
d
Cy
cle
ch
Ex
an
ge
ch
Ex
an
ge
ib
rs
le
ep
St
sib
le
ep
St
s
cl e
si b
d
r
r
e
e
th
ar
ve
ve
ith
Cl
ve
tiv
ve
tiv
wi
rw
8:
Re
R
Re
sw
cu
cu
n
Re
Re
fo
ith
ith
se
o
se
on
tio
dne
3:
i
t
c
n
n
e
w
w
w
1
O
a
T
ac
Fe
Co
Co
th
ns
ns
Re
th
Re
wi
7:
le
3:
tio
tio
wi
nt
g
ac
ac
sib
le
cle
in
rre
c
y
p
er
u
y
Re
Re
v
p
C
C
e
nc
ra
Re
ng
ay
tiv
ay
Co
pi
:T
W
6:
cu
W
9:
ap
10
e
ne
se
Tr
n
n
O
:
:
5
:O
Co
11
12
4:
re
n
Co
Animalia
Cytosol
Normalized z-Score
Results
n
a
Tr
2:
pp
in
g
le
Cy
Motif
Species
C. reinhardtii
H. salinarum
0.5
Cytosol
Normalized z-Score
A. thaliana
0.0
H. sapiens
M. acetivorans
M. barkeri
M. musculus
P. pastoris
-0.5
o
cti
ns
tio
ac
ns
tio
ac
ns
ep
St
ep
St
tio
ac
ns
ti
ac
on
le
c
Cy
n
ha
ge
n
ha
ge
ep
St
ep
St
s
le
le
le
c
c
le
le
ed
Re
ib
sib
sib
si b
ib
Re
Ex
Ex
os
rd
rs
rs
er
er
e
er
ith
ith
Cl
ve
tiv
ve
ev
ev
ev
wa
:
r
w
w
p
r
e
r
u
e
u
8
R
R
R
s
u
n
R
fo
ap
ec
e
ec
:R
ith
ith
o
nc
on
tio
dTr
ns
ns
13
On
cti
ac
ee
Co
Tw
2:
sw
sw
Co
Co
ea
ith
:F
Re
1:
on
on
ith
e
i
i
R
w
7
l
t
3:
t
t
w
b
n
ac
ac
si
ng
le
cle
rre
pi
er
yc
Re
Re
Cy
cu
ev
ap
e
g
on
yC
ay
in
Tr
tiv
:R
a
C
p
:
u
W
6
:
W
9
ap
10
ec
ne
Tr
ne
ns
:O
5:
:O
Co
11
12
4:
t
en
Re
a
g
in
Re
e
tiv
Re
le
c
Cy
S. cerevisiae
Z. mays
Motif
Figure S.1: Average significance profile of 3-node motifs in cytosol. The top panel shows motif
distributions by kingdom and the bottom by species.
Perhaps the most striking result is the consistency of the significance profile of the cytosol
across the species and kingdoms of life, which are nearly identical. This is a remarkable finding
given the vast range of environments and functions of the organisms and, more practically, the
S-6
Electronic Supplementary Material (ESI) for Molecular BioSystems
This journal is © The Royal Society of Chemistry 2012
Proportion
0.10
0.05
0.00
0.10
0.05
0.00
0.10
0.05
0.00
0.10
0.05
0.00
0.10
0.05
0.00
0.10
0.05
0.00
0.10
0.05
0.00
0.10
0.05
0.00
0.10
0.05
0.00
0.10
0.05
0.00
0.10
0.05
0.00
0.10
0.05
0.00
0.10
0.05
0.00
Figure S.2: Barplot showing proportion of motif participation in metabolic pathways
Alanine and Aspartate Metabolism
Alkaloid biosynthesis II
Alternate Carbon Metabolism
Amino Acid Metabolism
Aminosugar Metabolism
Anaplerotic Reactions
Arginine and Proline Metabolism
Ascorbate and Aldarate Metabolism
Asparagine metabolism
beta-Alanine metabolism
Bile Acid Biosynthesis
Biotin Metabolism
Blood Group Biosynthesis
Butanoate Metabolism
C5-Branched dibasic acid metabolism
Carnitine shuttle
Cell Envelope Biosynthesis
Central Metabolism
Cholesterol Metabolism
Chondroitin / heparan sulfate biosynthesis
Chondroitin sulfate degradation
Citric Acid Cycle
CoA Biosynthesis
CoA Catabolism
Coenzyme A Biosynthesis
Coenzyme B Biosynthesis
Coenzyme M Biosynthesis
Cofactor and Prosthetic Group Biosynthesis
CYP Metabolism
Cysteine Metabolism
D-alanine metabolism
D-arg and D-orn metabolism
Eicosanoid Metabolism
Fatty Acid Biosynthesis
Fatty acid activation
Fatty Acid Degradation
Fatty acid elongation
Fatty Acid Metabolism
Fatty acid oxidation
Folate Metabolism
Fructose and Mannose Metabolism
Galactose metabolism
Glucosamine Metabolism
Glutamate Metabolism
Glutamine Metabolism
Glutathione Metabolism
Glycerolipid Metabolism
Glycerophospholipid Metabolism
Glycine, Serine, and Threonine Metabolism
Glycolysis/Gluconeogenesis
Glycoprotein Metabolism
Glycosylphosphatidylinositol (GPI)-anchor biosynthesis
Glyoxylate and Dicarboxylate Metabolism
Heme Biosynthesis
Heparan sulfate degradation
Histidine Metabolism
Hyaluronan Metabolism
IMP Biosynthesis
Inorganic Ion Transport and Metabolism
Inositol Phosphate Metabolism
Keratan sulfate biosynthesis
Keratan sulfate degradation
Limonene and pinene degradation
Lipid & Cell Wall Metabolism
Lipopolysaccharide Biosynthesis / Recycling
Lysine Metabolism
Mannitol Metabolism
Membrane Lipid Metabolism
Methane Metabolism
Methanogenesis
Methionine Metabolism
Methylglyoxal Metabolism
Miscellaneous
Murein Recycling
N-Glycan Biosynthesis
N-Glycan Degradation
NAD Biosynthesis
NAD Metabolism
Nitrogen Metabolism
Nucleic acid degradation
Nucleotide Metabolism
Nucleotide Salvage Pathway
Nucleotide Sugar Metabolism
Nucleotides
O-Glycan Biosynthesis
Other Amino Acid Metabolism
Oxidative Phosphorylation
Pantothenate and CoA Biosynthesis
Pentose and Glucuronate Interconversions
Pentose Phosphate Pathway
Phospholipid Biosynthesis
Porphyrin and Chlorophyll Metabolism
Proline Biosynthesis
Propanoate Metabolism
Purine and Pyrimidine Biosynthesis
Purine Catabolism
Pyridoxine Metabolism
Pyrimidine Biosynthesis
Pyrimidine Catabolism
Pyruvate Metabolism
Quinone Biosynthesis
Riboflavin Metabolism
ROS Detoxification
Salvage Pathway
Selenoamino acid metabolism
Sphingolipid Metabolism
Starch and Sucrose Metabolism
Steroid Metabolism
Sterol Biosynthesis
Taurine and hypotaurine metabolism
Tetrahydramethanopterin Biosynthesis
Tetrahydrobiopterin
Thiamine Metabolism
Threonine and Lysine Metabolism
Triacylglycerol Synthesis
tRNA Charging
Tyr, Phe, Trp Biosynthesis
Tyrosine, Tryptophan, and Phenylalanine Metabolism
Ubiquinone Biosynthesis
Unassigned
Urea cycle/amino group metabolism
Valine, Leucine, and Isoleucine Metabolism
Vitamin A Metabolism
Vitamin B12 Metabolism
Vitamin B6 Metabolism
Vitamin D
Vitamins & Cofactor Biosynthesis
Xylose Metabolism
1
2
3
4
5
6
7
8
9
10
11
12
13
S-7
Electronic Supplementary Material (ESI) for Molecular BioSystems
This journal is © The Royal Society of Chemistry 2012
range of laboratories and research groups responsible for curating the reconstructions. Most of the
reconstructions have differing curation protocols, naming conventions and definitions, however the
emergent metabolic patterns are strikingly similar.
In addition to the overall consistency of the species, the within kingdom consistency is notable.
Both species of yeast show identical enrichment/suppression patterns. There are no contrary motifs
in the kingdom Animalia nor Archaea, although variation in the extent of enrichment is relatively
high between the two species of Archaea.
Fig. S.2 contains the barplots for the proportions of motif pathway participation. In the text,
this plot was reduced to only those pathways that represented at least 5% of the total for each
motif.
References
[1] M Ahsanul Islam, Elizabeth A Edwards, and Radhakrishnan Mahadevan. Characterizing
the metabolism of Dehalococcoides with a constraint-based model. PLoS Comput. Biol.,
6:e1000887, 2010.
[2] Scott A Becker and Bernhard Ø Palsson. Genome-scale reconstruction of the metabolic network
in Staphylococcus aureus N315: an initial draft to the two-dimensional annotation. BMC
Microbiol., 5:8, 2005.
[3] Roger L Chang, Lila Ghamsari, Ani Manichaikul, Erik F Y Hom, Santhanam Balaji, Weiqi
Fu, Yun Shen, Tong Hao, Bernhard Ø Palsson, Kourosh Salehi-Ashtiani, and Jason A Papin. Metabolic network reconstruction of Chlamydomonas offers insight into light-driven algal
metabolism. Mol. Syst. Biol., 7:518, 2011.
[4] Bevan KS Chung, Suresh Selvarasu, Camattari Andrea, Jimyoung Ryu, Hyeokweon Lee, Jungoh Ahn, Hongweon Lee, and Dong-Yup Lee. Genome-scale metabolic reconstruction and in
silico analysis of methylotrophic yeast Pichia pastoris for strain improvement. Microbial Cell
Factories, 9(50):1–15, 2010.
[5] Cristiana Gomes de Oliveira Dal’Molin, Lake-Ee Quek, Robin William Palfreyman,
Stevens Michael Brumbley, and Lars Keld Nielsen. AraGEM, a genome-scale reconstruction
of the primary metabolic network in Arabidopsis. Plant Physiol., 152(2):579–89, Feb 2010.
[6] Natalie C. Duarte, Scott A. Becker, Neema Jamshidi, Ines Thiele, Monica L. Mo, Thuy D.
Vo, Rohith Srivas, and Bernhard Ø. Palsson. Global reconstruction of the human metabolic
S-8
Electronic Supplementary Material (ESI) for Molecular BioSystems
This journal is © The Royal Society of Chemistry 2012
network based on genomic and bibliomic data. Proc. Natl. Acad. Sci. USA., 104:1777–1782,
2007.
[7] Natalie C Duarte, Markus J Herrgård, and Bernhard Ø Palsson. Reconstruction and validation
of Saccharomyces cerevisiae iND750, a fully compartmentalized genome-scale metabolic model.
Genome Res., 14:1298–309, 2004.
[8] Xin Fang, Anders Wallqvist, and Jaques Reifman. Development and analysis of an in vivocompatible metabolic network of Mycobacterium tuberculosis. BMC Syst. Biol., 4:160, 2010.
[9] Adam M Feist, Christopher S Henry, Jennifer L Reed, Markus Krummenacker, Andrew R
Joyce, Peter D Karp, Linda J Broadbelt, Vassily Hatzimanikatis, and Bernhard Ø Palsson.
A genome-scale metabolic reconstruction for Escherichia coli K-12 MG1655 that accounts for
1260 ORFs and thermodynamic information. Mol. Syst. Biol., 3:121, 2007.
[10] Adam M Feist, Johannes C M Scholten, Bernhard Ø Palsson, Fred J Brockman, and
Trey Ideker. Modeling methanogenesis with a genome-scale metabolic reconstruction of
Methanosarcina barkeri. Mol. Syst. Biol., 2:2006.0004, 2006.
[11] Orland Gonzalez, Susanne Gronau, Michaela Falb, Friedhelm Pfeiffer, Eduardo Mendoza,
Ralf Zimmer, and Dieter Oesterhelt. Reconstruction, modeling & analysis of Halobacterium
salinarum R-1 metabolism. Mol. BioSyst., 4:148–159, 2008.
[12] Hyun Uk Kim, Soo Young Kim, Haeyoung Jeong, Tae Yong Kim, Jae Jong Kim, Hyon E
Choy, Kyu Yang Yi, Joon Haeng Rhee, and Sang Yup Lee. Integrative genome-scale metabolic
analysis of Vibrio vulnificus for drug targeting and discovery. Mol. Syst. Biol., 18:7, 2011.
[13] Vinay Satish Kumar, James G Ferry, and Costas D Maranas. Metabolic reconstruction of the
archaeon methanogen Methanosarcina acetivorans. BMC Syst. Biol., 5:28, 2011.
[14] R. Mahadevan, D. R. Bond, J. E. Butler, A. Esteve-Nuñez, M. V. Coppi, B. O. Palsson, C. H.
Schilling, and D. R. Lovley. Characterization of metabolism in the Fe(iii)-reducing organism
Geobacter sulfurreducens by constraint-based modeling. Appl. Environ. Microbiol., 72:1558–
1568, 2006.
[15] Osbaldo Resendis-Antonio, Jennifer L Reed, Sergio Encarnación, Julio Collado-Vides, and
Bernhard Ø Palsson. Metabolic reconstruction and modeling of nitrogen fixation in Rhizobium
etli. PLoS Comput. Biol., 3:e192, 2007.
[16] Seth B. Roberts, Christopher M. Gowen, J Paul Brooks, and Stephen S. Fong. Genome-scale
metabolic analysis of Clostridium thermocellum for bioethanol production. BMC Syst. Biol.,
4:31, 2010.
[17] Rajib Saha, Patrick F. Suthers, and Costas D. Maranas. Zea mays iRS1563: A comprehensive
genome-scale metabolic reconstruction of maize metabolism. PLoS ONE, 6(7):e21784, 07 2011.
[18] Martin I Sigurdsson, Neema Jamshidi, Eirikur Steingrimsson, Ines Thiele, and Bernhard Ø
Palsson. A detailed genome-wide reconstruction of mouse metabolism based on human Recon
1. BMC Syst. Biol., 4: 140, 2010.
S-9
Electronic Supplementary Material (ESI) for Molecular BioSystems
This journal is © The Royal Society of Chemistry 2012
[19] Ines Thiele, Daniel Hyduke, Benjamin Steeb, Guy Fankam, Douglas Allen, Susanna Bazzani,
Pep Charusanti, Feng-Chi Chen, Ronan Fleming, Chao Hsiung, Sigrid De Keersmaecker, YuChieh Liao, Kathleen Marchal, Monica Mo, Emre Ozdemir, Anu Raghunathan, Jennifer Reed,
Sook-Il Shin, Sara Sigurbjornsdottir, Jonas Steinmann, Suresh Sudarsan, Neil Swainston, Inge
Thijs, Karsten Zengler, Bernhard Palsson, Joshua Adkins, and Dirk Bumann. A community
effort towards a knowledge-base and mathematical model of the human pathogen Salmonella
typhimurium LT2. BMC Syst. Biol., 5:8, 2011.
[20] Ines Thiele, Thuy D Vo, Nathan D Price, and Bernhard Ø Palsson. Expanded metabolic reconstruction of helicobacter pylori (iit341 gsm/gpr): an in silico genome-scale characterization
of single- and double-deletion mutants. J. Bacteriol., 187:5818–30, 2005.
[21] Ying Zhang, Ines Thiele, Dana Weekes, Zhanwen Li, Lukasz Jaroszewski, Krzysztof Ginalski,
Ashley M. Deacon, John Wooley, Scott A. Lesley, Ian A. Wilson, Bernhard Palsson, Andrei
Osterman, and Adam Godzik. Three-dimensional structural view of the central metabolic
network of Thermotoga maritima. Science, 325:1544–1549, 2009.
S-10