Darwinian Genomics

1
Darwinian Genomics Csaba Pal Biological
Research Center Szeged, Hungary
2
Genomics Major revolution in the past 10 15
years with the rise of high-throughput molecular
technologies New methods for rapid and
relatively cheap measurements of biological
molecules on a global scale
3
Systematic mapping components, interactions and
functional states of the cell

• Genomics genome sequencing and annotation
• Transcriptomics mRNA levels, mRNA half-lives
• Proteomics protein levels, protein protein
  interactions, protein modifications
• Metabolomics metabolite concentrations
• Phenomics creating collections of mutant strains
  and measuring phenotypes (e.g. cell growth) under
  various conditions

4

• Darwinian genomics Testing key issues in
  evolutionary biology
• Examples
• Role of chance and necessity
• Gradual changes or jumps
• Extent and evolution of robustness against
  mutations

5

• Darwinian genomics Testing key issues in
  evolutionary biology
• Examples
• Role of chance and necessity
• Gradual changes or jumps
• Extent and evolution of robustness against
  mutations

6
Yeast (S. cerevisiae) is an ideal model organism

• Complete genome sequence/detailed biochemical
  studies
• -gt network reconstruction
• 2) Genome-scale computational models
• -gt systems level properties of cellular networks
• 3) Large-scale mutant libraries
• -gt test predictions of the models
• 4) Complete genome sequences for 30 closely
  related species
• -gt study evolution across species

7
The knock-out paradox
High-throughput single gene knock-out studies no
phenotype for most genes in the lab
8

• Why keep them during evolution?
• Keep optimal cellular performance in face of
  harmful mutations and non-heritable errors
• Allow cellular growth under wide range of
  external conditions

9
(Seemingly) dispensable genes....

• compensated by a gene duplicate (genetic
  redundancy)
• compensated by alternative genetic pathways
  (distributed robustness)
• have important functions only under specific
  environmental conditions

Gene A
Gene B
Gene A
Gene B
10
Redundancy is only apparent, most genes should
have important contribution to survival under
special environmental conditions
11
Hillenmeyer et al. Science 2008
12
Compared growth rates of 5000 single gene
knock-out strains under gt1000 environments
97 of the mutants show slow growth under at
least one condition
Hillenmeyer et al. Science 2008
13
Are these explanations mutually exclusive?

• compensated by a gene duplicate (genetic
  redundancy)
• compensated by alternative genetic pathways
  (distributed robustness)
• have important functions only under specific
  environmental conditions

Gene A
Gene B
Gene A
Gene B
14
Does the capacity to compensate the impact of
gene deletions depend on the environment?
15
The extent of compensation may depend on nutrient
availability
Environment I.
Environment II.
Environment III.
A
A
A
B
B
B
A B
A B
A B
a B
a B
a B
A b
A b
A b
a b
a b
a b
synthetic lethality
no interaction
no interaction
16
Computational tool Flux Balance Analysis (FBA)
Amino acids Carbohydrates Ribonucleotides Deoxyrib
onucleotides Lipids Phospholipids Steroles Fatty
acids
fitness

• Network reconstruction In S. cerevisiae 1400
  biochemical reactions, including transport
  processes.
• Application of constraints Specify the nutrients
  available in the environment (B,E), the key
  metabolites or biomass constituents (X, Y, Z)
  essential for survival, presence/absence of genes
• Find a particular enzymatic flux distribution -gt
  rate of biomass production (fitness)

17

• What are the advantages of flux balance analysis?
  
• Study large number of genes and environments
  simultaneously
• Predictions
• a) Changes in enzyme activity as a response to
  nutrient conditions and genetic deletions
• b) Impact of gene deletions and gene addition on
  growth rates
• 3) Good agreement between experimental studies
  and model predictions (90)

Forster et al. 2003 OMICS, Papp et al. Nature
2004
18
Interactions between mutations in metabolic
networks
A special case Synthetic lethal genetic
interactions
Redundant gene duplicates
Gene A
Gene B
A B
normal growth
a B
A b
Gene A
lethal (or sick)
a b
Gene B
Alternative cellular pathways
19
Model predictions and verification of genetic
interactions

• Using Flux Balance analysis, we simulated all
  possible single and double gene deletions (125
  000) in the metabolic network under 53 different
  nutrient conditions
• ? 98 gene pairs are synthetic lethal under at
  least one condition
• We performed lab experiments to validate them

20
Results 1) 50 of the predictions were correct
(only 0.6 expected by chance!) 2) 85 of the
interacting gene pairs show condition-dependent
synthetic lethality
unconditional synthetic lethality
21
An example
Harrison et al. (2007) PNAS 1042307-2312
22
An example
Harrison et al. (2007) PNAS 1042307-2312
23
Conclusions

• The metabolic network model can reliably predict
  (synthetic lethal) genetic interactions.
• The presence of genetic interactions (and hence
  the extent of compensation) vary extensively
  across nutrient conditions.

24
Speculations and potential implications

• Experimental design. Different environments
  should be screened to identify the majority of
  genetic interactions
• Functional genomics. Redundancy is more apparent
  than real. Many seemingly dispensable genes have
  important physiological role under specific
  conditions
• Evolution. Robustness against mutations may not
  be a directly selected trait, but rather a
  by-product of evolution of novel metabolic
  pathways towards new environmental conditions

25

• Shortcomings
• The computational model is far from perfect, and
  ignores many biological details
• Only specific genetic interactions have been
  studied
• No systematic experimental screen

Harrison et al. (2007) PNAS 1042307-2312
26

• Collaboration with Charles Boone lab
• Using robotic protocols, they map genetic
  interactions across the whole yeast genome (107
  combinations )
• They developed high-throughput protocols to
  measure fitness at high precision

27
Why study evolution?
28
Evolution of antibiotics resistance 33 Billion
annual costs in US
29
Ignoring evolution has serious health consequences
30
Evolutionary Systems Biology Group

• Projects
• Analyses of genetic interactions
• Evolution of antibiotics resistance

http//www.brc.hu/sysbiol/
31
Interactions between genes are masked by distant
gene duplicates
Confirmed by creating corresponding triple
knock-outs Overlapping enzymatic activities
between duplicates conserved across more than 100
million years of evolution