How many degrees of separation

1
How many degrees of separation?

• Synthetic lethals in yeast (and humans)

2
Synthetic phenotypes
Mutant 1 X Mutant 2 Viable
Viable
Mutant 1 Mutant 2 DEAD
WHY?
3
Reasons for synthetic phenotypes(the naïve view)

• The genes encode redundant steps of the same
  pathway
• The genes are involved in the same pathway but in
  different steps
• The genes are involved in parallel pathways
• Some interesting but unexplainable biological
  connection? (an artifact?)

4
Importance of synthetic phenotypes

• Identify a component of a pathway or process not
  identifiable via single mutant genetics
• Identify processes or genes that act to buffer
  against mutation

5
Ways of finding synthetic mutants

• A mutant may actually be due to two mutations
• Tests of epistasis or enhancement between mutants
  with the same phenotype
• Enhancer or suppressor screens
• Screen directly-yeast
• Put gene of interest on plasmid with color
  selection also on plasmid
• Mutagenize and look for mutants where plasmid
  cannot be lost

6
Yeast synthetic lethals in the secretory pathway
116/237 same step 68/237 different step 53/237
not in secretion
7
Synthetic lethals-how common?

• Screens with 5 genes, between 3 and 8 interacting
  genes were found
• Yeast Proteome Database usually less than 10 per
  gene, 26 was the max
• This is an underestimate, has not been
  systematically done

Hartman, Garvik and Hartwell, 2001
8
Cassette for gene deletions
Kan R
ATG
TAA
UPTAG
DNTAG
X
X
YFG
ATG
TAA
9
Generation of haploid double mutants
Del 1-selection 1 Mat a
Del 2-selection 2 Mat a
X
a/a diploid sporulation
Selection 1, selection 2, select for haploids
Mutant 1 mutant 2
10
Screened 132 genes X 4700 deletion strains

• 132 strains either deletions or ts strains
• Each cross repeated three times
• (620,400 x 3!)
• All genes (36 million x 3)
• Multiple interactions checked by tetrad analysis
  or random spore analysis

11
How many synthetic lethals would you expect per
gene?
12
How many synthetic lethals would you expect per
gene?

• Average of 34, range from 1-146
• This compares to 8 physical interactions
• Screens with 5 genes, between 3 and 8 interacting
  genes were found
• Yeast Proteome Database usually less than 10 per
  gene, 26 was the max

13
What was interesting in the data?

• Synthetic lethals more abundant when genes had
  the same mutant phenotype P10-316
• Synthetic lethals more abundant when genes had
  the same subcellular localization P10-70
• Synthetic lethals more abundant when genes were
  known to be part of a complex P10-68

14
Network of genetically connected gene functions
15
Networks have predictive value

• CSM3-interaction pattern matched DNA replication
  checkpoint genes MRC1 and TOF1
• YMR299c-clustered with dynein/ dynactin-might be
  yeast dynein light intermediate gene

16
Do neighbors interact?
SGS1yeast homolog of human Werners syndrome
protein, a DNA helicase
17
Summary

• About 1 of double mutants generated a synthetic
  lethal phenotype
• Homeostasis/buffering?
• New hypotheses about novel genes
• Implications for variation

18
Relevance to human biology

• Challenge of understanding genetic background
• -7 modifiers of CFTR
• Digenic diseases ROM1 RDS
• Retinal outer segment membrane protein
• Retinal degeneration slow
• Together cause retinitis pigmentosum
• Many complex heterogenous diseases