Embed Size (px)
Citation preview
Real networks from 1,181bacterial species
The percolation of metabolism and bacterial
lifestyle and evolvability Germán Plata, PhD & Dennis Vitkup, PhD
Department of Systems Biology, Columbia University Medical Center, New York, NY 10032, [email protected]
Defining metabolic network connectivity
Conclusions
Impact on genome architecture
Can metabolic network diversity across bacteria be seenas a percolation process? If so, what are its consequencesfor microbial physiology, ecology and evolution?
A percolation transition in metabolism
Species interactions and evolvability
Percolation in random networks
600 800 1000 1200 1400 1600 1800
0
20
40
60
Num
ber
of gro
wth
-suport
ing
carb
on s
ourc
es
Number of reactions
400 600 800 1000 1200 1400 1600 1800
0
2
4
6
8
10
Com
ple
x m
edia
ingre
die
nts
Number of reactions
600 800 1000 1200 1400 1600 1800
0.0
0.2
0.4
0.6
0.8
1.0
Fra
ctio
n o
blig
ate
sym
bio
nts
Number of reactions
500 2000 5000 200001000 10000
1
10
100
1000
Number of
model reactions:
Moving average
Piecewise linear fit
Num
ber
of tr
anscription facto
rs
Number of genes
300
900
1800
600 800 1000 1200 1400 1600 1800
0
5
10
15
20
25
Reactions r
equired for
new
carb
on s
ourc
e
Number of reactions
600 800 1000 1200 1400 1600 1800
0.0
0.2
0.4
0.6
0.8
1.0
Cro
ss-f
eed
ing p
rob
ab
ility
Number of reactions
• Metabolic networks were defined based on the abilityof mass (carbon) to flow between metabolites (right)
• 1,181 networks were obtained from Kbase, each from adifferent bacterial genus
600 800 1000 1200 1400 1600 1800
0
50
100
150
200
250
Siz
e o
f la
rgest com
ponent
Number of reactions
Critical threshold at ~900 reactions or ~2,000 genes
Less fastidious1More free-living2
500 2000 5000 200001000 10000
200
500
100
1000
500 2000 5000 200001000 1000050
200
500
2000
100
1000
500 2000 5000 200001000 100002
5
20
50
200
500
2000
10
100
1000
500 2000 5000 200001000 10000
20
50
200
500
2000
10
100
1000
Moving average
Piecewise linear fit
Info
rma
tio
n g
en
es
Total genes
300
900
1800Number of
model reactions:
Me
tab
olic
ge
ne
s
Total genes
Moving average
Piecewise linear fit
Sig
na
ling
ge
ne
s
Total genes
Moving average
Piecewise linear fit
Oth
er
ge
ne
s
Total genes
Moving average
Piecewise linear fit
Condition-specific genes are lost more rapidly withdecreasing genome size below percolation threshold
How likely is one species to benefit from metabolic by-products of another?
How many additional reactions are needed to grow on anew carbon source?
More evolvable
Random networks, same
probability per reaction.
Optimal networks, least
number of reactions per carbonsource supporting growth.
Bacterial evolution avoids metabolic networks that arelarge and disconnected or small and connected.
• Differences in the connectivity of bacterial metabolicnetworks can be described as a percolation process
• Species with less than ~2,000 genes tend to be morefastidious, less easily evolvable, and less likely to cross-feed with a randomly chosen bacterium
• Differences are explained by metabolite interconversionssupported by the giant component present in largermetabolic networks
1. Known growth media: komodo.modelseed.org 2. Lifestyle classification: Burstein et al. 2015. Nat Commun. 6, 8493 Funding: National Institute of General Medical Sciences GM079759 to DV
Species 1 Species 2
A percolation process describes the changes in theconnectivity of a network as nodes or edges are removed
Each dot represents the metabolicnetwork of a different bacterium
Percolation threshold How many carbon sourcessupport growth?
Arrows indicate percolationthresholds
Obligate parasitic or symbiotic lifestyles are commonbelow the percolation threshold
250 500 750 1000 1250 1500 1750 2000
0
50
100
150
200
250
Siz
e o
f la
rgest com
ponent
Number of reactions
